diff options
Diffstat (limited to 'fs')
479 files changed, 23102 insertions, 21629 deletions
diff --git a/fs/9p/fid.c b/fs/9p/fid.c index 3031e3233dd..14d94420457 100644 --- a/fs/9p/fid.c +++ b/fs/9p/fid.c @@ -45,7 +45,7 @@ int v9fs_fid_add(struct dentry *dentry, struct p9_fid *fid) struct v9fs_dentry *dent; P9_DPRINTK(P9_DEBUG_VFS, "fid %d dentry %s\n", - fid->fid, dentry->d_iname); + fid->fid, dentry->d_name.name); dent = dentry->d_fsdata; if (!dent) { @@ -79,7 +79,7 @@ static struct p9_fid *v9fs_fid_find(struct dentry *dentry, u32 uid, int any) struct p9_fid *fid, *ret; P9_DPRINTK(P9_DEBUG_VFS, " dentry: %s (%p) uid %d any %d\n", - dentry->d_iname, dentry, uid, any); + dentry->d_name.name, dentry, uid, any); dent = (struct v9fs_dentry *) dentry->d_fsdata; ret = NULL; if (dent) { @@ -120,7 +120,7 @@ struct p9_fid *v9fs_fid_lookup(struct dentry *dentry) switch (access) { case V9FS_ACCESS_SINGLE: case V9FS_ACCESS_USER: - uid = current->fsuid; + uid = current_fsuid(); any = 0; break; diff --git a/fs/9p/v9fs.c b/fs/9p/v9fs.c index c061c3f18e7..332b5ff02fe 100644 --- a/fs/9p/v9fs.c +++ b/fs/9p/v9fs.c @@ -30,8 +30,8 @@ #include <linux/parser.h> #include <linux/idr.h> #include <net/9p/9p.h> -#include <net/9p/transport.h> #include <net/9p/client.h> +#include <net/9p/transport.h> #include "v9fs.h" #include "v9fs_vfs.h" @@ -160,7 +160,7 @@ static int v9fs_parse_options(struct v9fs_session_info *v9ses) v9ses->flags |= V9FS_ACCESS_ANY; else { v9ses->flags |= V9FS_ACCESS_SINGLE; - v9ses->uid = simple_strtol(s, &e, 10); + v9ses->uid = simple_strtoul(s, &e, 10); if (*e != '\0') v9ses->uid = ~0; } @@ -234,7 +234,7 @@ struct p9_fid *v9fs_session_init(struct v9fs_session_info *v9ses, if (!v9ses->clnt->dotu) v9ses->flags &= ~V9FS_EXTENDED; - v9ses->maxdata = v9ses->clnt->msize; + v9ses->maxdata = v9ses->clnt->msize - P9_IOHDRSZ; /* for legacy mode, fall back to V9FS_ACCESS_ANY */ if (!v9fs_extended(v9ses) && diff --git a/fs/9p/v9fs_vfs.h b/fs/9p/v9fs_vfs.h index 57997fa14e6..c295ba786ed 100644 --- a/fs/9p/v9fs_vfs.h +++ b/fs/9p/v9fs_vfs.h @@ -46,9 +46,11 @@ extern struct dentry_operations v9fs_cached_dentry_operations; struct inode *v9fs_get_inode(struct super_block *sb, int mode); ino_t v9fs_qid2ino(struct p9_qid *qid); -void v9fs_stat2inode(struct p9_stat *, struct inode *, struct super_block *); +void v9fs_stat2inode(struct p9_wstat *, struct inode *, struct super_block *); int v9fs_dir_release(struct inode *inode, struct file *filp); int v9fs_file_open(struct inode *inode, struct file *file); -void v9fs_inode2stat(struct inode *inode, struct p9_stat *stat); +void v9fs_inode2stat(struct inode *inode, struct p9_wstat *stat); void v9fs_dentry_release(struct dentry *); int v9fs_uflags2omode(int uflags, int extended); + +ssize_t v9fs_file_readn(struct file *, char *, char __user *, u32, u64); diff --git a/fs/9p/vfs_addr.c b/fs/9p/vfs_addr.c index 97d3aed5798..6fcb1e7095c 100644 --- a/fs/9p/vfs_addr.c +++ b/fs/9p/vfs_addr.c @@ -38,7 +38,6 @@ #include "v9fs.h" #include "v9fs_vfs.h" -#include "fid.h" /** * v9fs_vfs_readpage - read an entire page in from 9P @@ -53,14 +52,12 @@ static int v9fs_vfs_readpage(struct file *filp, struct page *page) int retval; loff_t offset; char *buffer; - struct p9_fid *fid; P9_DPRINTK(P9_DEBUG_VFS, "\n"); - fid = filp->private_data; buffer = kmap(page); offset = page_offset(page); - retval = p9_client_readn(fid, buffer, offset, PAGE_CACHE_SIZE); + retval = v9fs_file_readn(filp, buffer, NULL, offset, PAGE_CACHE_SIZE); if (retval < 0) goto done; diff --git a/fs/9p/vfs_dentry.c b/fs/9p/vfs_dentry.c index f9534f18df0..06dcc7c4f23 100644 --- a/fs/9p/vfs_dentry.c +++ b/fs/9p/vfs_dentry.c @@ -52,7 +52,8 @@ static int v9fs_dentry_delete(struct dentry *dentry) { - P9_DPRINTK(P9_DEBUG_VFS, " dentry: %s (%p)\n", dentry->d_iname, dentry); + P9_DPRINTK(P9_DEBUG_VFS, " dentry: %s (%p)\n", dentry->d_name.name, + dentry); return 1; } @@ -69,7 +70,8 @@ static int v9fs_dentry_delete(struct dentry *dentry) static int v9fs_cached_dentry_delete(struct dentry *dentry) { struct inode *inode = dentry->d_inode; - P9_DPRINTK(P9_DEBUG_VFS, " dentry: %s (%p)\n", dentry->d_iname, dentry); + P9_DPRINTK(P9_DEBUG_VFS, " dentry: %s (%p)\n", dentry->d_name.name, + dentry); if(!inode) return 1; @@ -88,7 +90,8 @@ void v9fs_dentry_release(struct dentry *dentry) struct v9fs_dentry *dent; struct p9_fid *temp, *current_fid; - P9_DPRINTK(P9_DEBUG_VFS, " dentry: %s (%p)\n", dentry->d_iname, dentry); + P9_DPRINTK(P9_DEBUG_VFS, " dentry: %s (%p)\n", dentry->d_name.name, + dentry); dent = dentry->d_fsdata; if (dent) { list_for_each_entry_safe(current_fid, temp, &dent->fidlist, diff --git a/fs/9p/vfs_dir.c b/fs/9p/vfs_dir.c index e298fe19409..873cd31baa4 100644 --- a/fs/9p/vfs_dir.c +++ b/fs/9p/vfs_dir.c @@ -45,7 +45,7 @@ * */ -static inline int dt_type(struct p9_stat *mistat) +static inline int dt_type(struct p9_wstat *mistat) { unsigned long perm = mistat->mode; int rettype = DT_REG; @@ -69,32 +69,58 @@ static inline int dt_type(struct p9_stat *mistat) static int v9fs_dir_readdir(struct file *filp, void *dirent, filldir_t filldir) { int over; + struct p9_wstat st; + int err; struct p9_fid *fid; - struct v9fs_session_info *v9ses; - struct inode *inode; - struct p9_stat *st; + int buflen; + char *statbuf; + int n, i = 0; P9_DPRINTK(P9_DEBUG_VFS, "name %s\n", filp->f_path.dentry->d_name.name); - inode = filp->f_path.dentry->d_inode; - v9ses = v9fs_inode2v9ses(inode); fid = filp->private_data; - while ((st = p9_client_dirread(fid, filp->f_pos)) != NULL) { - if (IS_ERR(st)) - return PTR_ERR(st); - over = filldir(dirent, st->name.str, st->name.len, filp->f_pos, - v9fs_qid2ino(&st->qid), dt_type(st)); + buflen = fid->clnt->msize - P9_IOHDRSZ; + statbuf = kmalloc(buflen, GFP_KERNEL); + if (!statbuf) + return -ENOMEM; - if (over) + while (1) { + err = v9fs_file_readn(filp, statbuf, NULL, buflen, + fid->rdir_fpos); + if (err <= 0) break; - filp->f_pos += st->size; - kfree(st); - st = NULL; + n = err; + while (i < n) { + err = p9stat_read(statbuf + i, buflen-i, &st, + fid->clnt->dotu); + if (err) { + P9_DPRINTK(P9_DEBUG_VFS, "returned %d\n", err); + err = -EIO; + p9stat_free(&st); + goto free_and_exit; + } + + i += st.size+2; + fid->rdir_fpos += st.size+2; + + over = filldir(dirent, st.name, strlen(st.name), + filp->f_pos, v9fs_qid2ino(&st.qid), dt_type(&st)); + + filp->f_pos += st.size+2; + + p9stat_free(&st); + + if (over) { + err = 0; + goto free_and_exit; + } + } } - kfree(st); - return 0; +free_and_exit: + kfree(statbuf); + return err; } diff --git a/fs/9p/vfs_file.c b/fs/9p/vfs_file.c index 52944d2249a..68bf2af6c38 100644 --- a/fs/9p/vfs_file.c +++ b/fs/9p/vfs_file.c @@ -120,23 +120,72 @@ static int v9fs_file_lock(struct file *filp, int cmd, struct file_lock *fl) } /** - * v9fs_file_read - read from a file + * v9fs_file_readn - read from a file * @filp: file pointer to read * @data: data buffer to read data into + * @udata: user data buffer to read data into * @count: size of buffer * @offset: offset at which to read data * */ + +ssize_t +v9fs_file_readn(struct file *filp, char *data, char __user *udata, u32 count, + u64 offset) +{ + int n, total; + struct p9_fid *fid = filp->private_data; + + P9_DPRINTK(P9_DEBUG_VFS, "fid %d offset %llu count %d\n", fid->fid, + (long long unsigned) offset, count); + + n = 0; + total = 0; + do { + n = p9_client_read(fid, data, udata, offset, count); + if (n <= 0) + break; + + if (data) + data += n; + if (udata) + udata += n; + + offset += n; + count -= n; + total += n; + } while (count > 0 && n == (fid->clnt->msize - P9_IOHDRSZ)); + + if (n < 0) + total = n; + + return total; +} + +/** + * v9fs_file_read - read from a file + * @filp: file pointer to read + * @udata: user data buffer to read data into + * @count: size of buffer + * @offset: offset at which to read data + * + */ + static ssize_t -v9fs_file_read(struct file *filp, char __user * data, size_t count, +v9fs_file_read(struct file *filp, char __user *udata, size_t count, loff_t * offset) { int ret; struct p9_fid *fid; - P9_DPRINTK(P9_DEBUG_VFS, "\n"); + P9_DPRINTK(P9_DEBUG_VFS, "count %zu offset %lld\n", count, *offset); fid = filp->private_data; - ret = p9_client_uread(fid, data, *offset, count); + + if (count > (fid->clnt->msize - P9_IOHDRSZ)) + ret = v9fs_file_readn(filp, NULL, udata, count, *offset); + else + ret = p9_client_read(fid, NULL, udata, *offset, count); + if (ret > 0) *offset += ret; @@ -156,19 +205,38 @@ static ssize_t v9fs_file_write(struct file *filp, const char __user * data, size_t count, loff_t * offset) { - int ret; + int n, rsize, total = 0; struct p9_fid *fid; + struct p9_client *clnt; struct inode *inode = filp->f_path.dentry->d_inode; + int origin = *offset; P9_DPRINTK(P9_DEBUG_VFS, "data %p count %d offset %x\n", data, (int)count, (int)*offset); fid = filp->private_data; - ret = p9_client_uwrite(fid, data, *offset, count); - if (ret > 0) { - invalidate_inode_pages2_range(inode->i_mapping, *offset, - *offset+ret); - *offset += ret; + clnt = fid->clnt; + + rsize = fid->iounit; + if (!rsize || rsize > clnt->msize-P9_IOHDRSZ) + rsize = clnt->msize - P9_IOHDRSZ; + + do { + if (count < rsize) + rsize = count; + + n = p9_client_write(fid, NULL, data+total, *offset+total, + rsize); + if (n <= 0) + break; + count -= n; + total += n; + } while (count > 0); + + if (total > 0) { + invalidate_inode_pages2_range(inode->i_mapping, origin, + origin+total); + *offset += total; } if (*offset > inode->i_size) { @@ -176,7 +244,10 @@ v9fs_file_write(struct file *filp, const char __user * data, inode->i_blocks = (inode->i_size + 512 - 1) >> 9; } - return ret; + if (n < 0) + return n; + + return total; } static const struct file_operations v9fs_cached_file_operations = { diff --git a/fs/9p/vfs_inode.c b/fs/9p/vfs_inode.c index e83aa5ebe86..81f8bbf12f9 100644 --- a/fs/9p/vfs_inode.c +++ b/fs/9p/vfs_inode.c @@ -215,8 +215,8 @@ struct inode *v9fs_get_inode(struct super_block *sb, int mode) inode = new_inode(sb); if (inode) { inode->i_mode = mode; - inode->i_uid = current->fsuid; - inode->i_gid = current->fsgid; + inode->i_uid = current_fsuid(); + inode->i_gid = current_fsgid(); inode->i_blocks = 0; inode->i_rdev = 0; inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME; @@ -334,7 +334,7 @@ v9fs_inode_from_fid(struct v9fs_session_info *v9ses, struct p9_fid *fid, { int err, umode; struct inode *ret; - struct p9_stat *st; + struct p9_wstat *st; ret = NULL; st = p9_client_stat(fid); @@ -417,6 +417,8 @@ v9fs_create(struct v9fs_session_info *v9ses, struct inode *dir, struct p9_fid *dfid, *ofid, *fid; struct inode *inode; + P9_DPRINTK(P9_DEBUG_VFS, "name %s\n", dentry->d_name.name); + err = 0; ofid = NULL; fid = NULL; @@ -424,6 +426,7 @@ v9fs_create(struct v9fs_session_info *v9ses, struct inode *dir, dfid = v9fs_fid_clone(dentry->d_parent); if (IS_ERR(dfid)) { err = PTR_ERR(dfid); + P9_DPRINTK(P9_DEBUG_VFS, "fid clone failed %d\n", err); dfid = NULL; goto error; } @@ -432,18 +435,22 @@ v9fs_create(struct v9fs_session_info *v9ses, struct inode *dir, ofid = p9_client_walk(dfid, 0, NULL, 1); if (IS_ERR(ofid)) { err = PTR_ERR(ofid); + P9_DPRINTK(P9_DEBUG_VFS, "p9_client_walk failed %d\n", err); ofid = NULL; goto error; } err = p9_client_fcreate(ofid, name, perm, mode, extension); - if (err < 0) + if (err < 0) { + P9_DPRINTK(P9_DEBUG_VFS, "p9_client_fcreate failed %d\n", err); goto error; + } /* now walk from the parent so we can get unopened fid */ fid = p9_client_walk(dfid, 1, &name, 0); if (IS_ERR(fid)) { err = PTR_ERR(fid); + P9_DPRINTK(P9_DEBUG_VFS, "p9_client_walk failed %d\n", err); fid = NULL; goto error; } else @@ -453,6 +460,7 @@ v9fs_create(struct v9fs_session_info *v9ses, struct inode *dir, inode = v9fs_inode_from_fid(v9ses, fid, dir->i_sb); if (IS_ERR(inode)) { err = PTR_ERR(inode); + P9_DPRINTK(P9_DEBUG_VFS, "inode creation failed %d\n", err); goto error; } @@ -734,7 +742,7 @@ v9fs_vfs_getattr(struct vfsmount *mnt, struct dentry *dentry, int err; struct v9fs_session_info *v9ses; struct p9_fid *fid; - struct p9_stat *st; + struct p9_wstat *st; P9_DPRINTK(P9_DEBUG_VFS, "dentry: %p\n", dentry); err = -EPERM; @@ -815,10 +823,9 @@ static int v9fs_vfs_setattr(struct dentry *dentry, struct iattr *iattr) */ void -v9fs_stat2inode(struct p9_stat *stat, struct inode *inode, +v9fs_stat2inode(struct p9_wstat *stat, struct inode *inode, struct super_block *sb) { - int n; char ext[32]; struct v9fs_session_info *v9ses = sb->s_fs_info; @@ -842,11 +849,7 @@ v9fs_stat2inode(struct p9_stat *stat, struct inode *inode, int major = -1; int minor = -1; - n = stat->extension.len; - if (n > sizeof(ext)-1) - n = sizeof(ext)-1; - memmove(ext, stat->extension.str, n); - ext[n] = 0; + strncpy(ext, stat->extension, sizeof(ext)); sscanf(ext, "%c %u %u", &type, &major, &minor); switch (type) { case 'c': @@ -857,10 +860,11 @@ v9fs_stat2inode(struct p9_stat *stat, struct inode *inode, break; default: P9_DPRINTK(P9_DEBUG_ERROR, - "Unknown special type %c (%.*s)\n", type, - stat->extension.len, stat->extension.str); + "Unknown special type %c %s\n", type, + stat->extension); }; inode->i_rdev = MKDEV(major, minor); + init_special_inode(inode, inode->i_mode, inode->i_rdev); } else inode->i_rdev = 0; @@ -904,7 +908,7 @@ static int v9fs_readlink(struct dentry *dentry, char *buffer, int buflen) struct v9fs_session_info *v9ses; struct p9_fid *fid; - struct p9_stat *st; + struct p9_wstat *st; P9_DPRINTK(P9_DEBUG_VFS, " %s\n", dentry->d_name.name); retval = -EPERM; @@ -926,15 +930,10 @@ static int v9fs_readlink(struct dentry *dentry, char *buffer, int buflen) } /* copy extension buffer into buffer */ - if (st->extension.len < buflen) - buflen = st->extension.len + 1; - - memmove(buffer, st->extension.str, buflen - 1); - buffer[buflen-1] = 0; + strncpy(buffer, st->extension, buflen); P9_DPRINTK(P9_DEBUG_VFS, - "%s -> %.*s (%s)\n", dentry->d_name.name, st->extension.len, - st->extension.str, buffer); + "%s -> %s (%s)\n", dentry->d_name.name, st->extension, buffer); retval = buflen; @@ -964,7 +963,8 @@ static int v9fs_vfs_readlink(struct dentry *dentry, char __user * buffer, if (buflen > PATH_MAX) buflen = PATH_MAX; - P9_DPRINTK(P9_DEBUG_VFS, " dentry: %s (%p)\n", dentry->d_iname, dentry); + P9_DPRINTK(P9_DEBUG_VFS, " dentry: %s (%p)\n", dentry->d_name.name, + dentry); retval = v9fs_readlink(dentry, link, buflen); @@ -1023,7 +1023,8 @@ v9fs_vfs_put_link(struct dentry *dentry, struct nameidata *nd, void *p) { char *s = nd_get_link(nd); - P9_DPRINTK(P9_DEBUG_VFS, " %s %s\n", dentry->d_name.name, s); + P9_DPRINTK(P9_DEBUG_VFS, " %s %s\n", dentry->d_name.name, + IS_ERR(s) ? "<error>" : s); if (!IS_ERR(s)) __putname(s); } diff --git a/fs/9p/vfs_super.c b/fs/9p/vfs_super.c index bf59c396049..93212e40221 100644 --- a/fs/9p/vfs_super.c +++ b/fs/9p/vfs_super.c @@ -111,10 +111,10 @@ static int v9fs_get_sb(struct file_system_type *fs_type, int flags, struct inode *inode = NULL; struct dentry *root = NULL; struct v9fs_session_info *v9ses = NULL; - struct p9_stat *st = NULL; + struct p9_wstat *st = NULL; int mode = S_IRWXUGO | S_ISVTX; - uid_t uid = current->fsuid; - gid_t gid = current->fsgid; + uid_t uid = current_fsuid(); + gid_t gid = current_fsgid(); struct p9_fid *fid; int retval = 0; @@ -161,10 +161,14 @@ static int v9fs_get_sb(struct file_system_type *fs_type, int flags, sb->s_root = root; root->d_inode->i_ino = v9fs_qid2ino(&st->qid); + v9fs_stat2inode(st, root->d_inode, sb); + v9fs_fid_add(root, fid); + p9stat_free(st); kfree(st); +P9_DPRINTK(P9_DEBUG_VFS, " return simple set mount\n"); return simple_set_mnt(mnt, sb); release_sb: diff --git a/fs/Kconfig b/fs/Kconfig index 9e9d70c02a0..522469a7eca 100644 --- a/fs/Kconfig +++ b/fs/Kconfig @@ -6,61 +6,9 @@ menu "File systems" if BLOCK -config EXT2_FS - tristate "Second extended fs support" - help - Ext2 is a standard Linux file system for hard disks. - - To compile this file system support as a module, choose M here: the - module will be called ext2. - - If unsure, say Y. - -config EXT2_FS_XATTR - bool "Ext2 extended attributes" - depends on EXT2_FS - help - Extended attributes are name:value pairs associated with inodes by - the kernel or by users (see the attr(5) manual page, or visit - <http://acl.bestbits.at/> for details). - - If unsure, say N. - -config EXT2_FS_POSIX_ACL - bool "Ext2 POSIX Access Control Lists" - depends on EXT2_FS_XATTR - select FS_POSIX_ACL - help - Posix Access Control Lists (ACLs) support permissions for users and - groups beyond the owner/group/world scheme. - - To learn more about Access Control Lists, visit the Posix ACLs for - Linux website <http://acl.bestbits.at/>. - - If you don't know what Access Control Lists are, say N - -config EXT2_FS_SECURITY - bool "Ext2 Security Labels" - depends on EXT2_FS_XATTR - help - Security labels support alternative access control models - implemented by security modules like SELinux. This option - enables an extended attribute handler for file security - labels in the ext2 filesystem. - - If you are not using a security module that requires using - extended attributes for file security labels, say N. - -config EXT2_FS_XIP - bool "Ext2 execute in place support" - depends on EXT2_FS && MMU - help - Execute in place can be used on memory-backed block devices. If you - enable this option, you can select to mount block devices which are - capable of this feature without using the page cache. - - If you do not use a block device that is capable of using this, - or if unsure, say N. +source "fs/ext2/Kconfig" +source "fs/ext3/Kconfig" +source "fs/ext4/Kconfig" config FS_XIP # execute in place @@ -68,225 +16,16 @@ config FS_XIP depends on EXT2_FS_XIP default y -config EXT3_FS - tristate "Ext3 journalling file system support" - select JBD - help - This is the journalling version of the Second extended file system - (often called ext3), the de facto standard Linux file system - (method to organize files on a storage device) for hard disks. - - The journalling code included in this driver means you do not have - to run e2fsck (file system checker) on your file systems after a - crash. The journal keeps track of any changes that were being made - at the time the system crashed, and can ensure that your file system - is consistent without the need for a lengthy check. - - Other than adding the journal to the file system, the on-disk format - of ext3 is identical to ext2. It is possible to freely switch - between using the ext3 driver and the ext2 driver, as long as the - file system has been cleanly unmounted, or e2fsck is run on the file - system. - - To add a journal on an existing ext2 file system or change the - behavior of ext3 file systems, you can use the tune2fs utility ("man - tune2fs"). To modify attributes of files and directories on ext3 - file systems, use chattr ("man chattr"). You need to be using - e2fsprogs version 1.20 or later in order to create ext3 journals - (available at <http://sourceforge.net/projects/e2fsprogs/>). - - To compile this file system support as a module, choose M here: the - module will be called ext3. - -config EXT3_FS_XATTR - bool "Ext3 extended attributes" - depends on EXT3_FS - default y - help - Extended attributes are name:value pairs associated with inodes by - the kernel or by users (see the attr(5) manual page, or visit - <http://acl.bestbits.at/> for details). - - If unsure, say N. - - You need this for POSIX ACL support on ext3. - -config EXT3_FS_POSIX_ACL - bool "Ext3 POSIX Access Control Lists" - depends on EXT3_FS_XATTR - select FS_POSIX_ACL - help - Posix Access Control Lists (ACLs) support permissions for users and - groups beyond the owner/group/world scheme. - - To learn more about Access Control Lists, visit the Posix ACLs for - Linux website <http://acl.bestbits.at/>. - - If you don't know what Access Control Lists are, say N - -config EXT3_FS_SECURITY - bool "Ext3 Security Labels" - depends on EXT3_FS_XATTR - help - Security labels support alternative access control models - implemented by security modules like SELinux. This option - enables an extended attribute handler for file security - labels in the ext3 filesystem. - - If you are not using a security module that requires using - extended attributes for file security labels, say N. - -config EXT4_FS - tristate "The Extended 4 (ext4) filesystem" - select JBD2 - select CRC16 - help - This is the next generation of the ext3 filesystem. - - Unlike the change from ext2 filesystem to ext3 filesystem, - the on-disk format of ext4 is not forwards compatible with - ext3; it is based on extent maps and it supports 48-bit - physical block numbers. The ext4 filesystem also supports delayed - allocation, persistent preallocation, high resolution time stamps, - and a number of other features to improve performance and speed - up fsck time. For more information, please see the web pages at - http://ext4.wiki.kernel.org. - - The ext4 filesystem will support mounting an ext3 - filesystem; while there will be some performance gains from - the delayed allocation and inode table readahead, the best - performance gains will require enabling ext4 features in the - filesystem, or formating a new filesystem as an ext4 - filesystem initially. - - To compile this file system support as a module, choose M here. The - module will be called ext4dev. - - If unsure, say N. - -config EXT4DEV_COMPAT - bool "Enable ext4dev compatibility" - depends on EXT4_FS - help - Starting with 2.6.28, the name of the ext4 filesystem was - renamed from ext4dev to ext4. Unfortunately there are some - legacy userspace programs (such as klibc's fstype) have - "ext4dev" hardcoded. - - To enable backwards compatibility so that systems that are - still expecting to mount ext4 filesystems using ext4dev, - chose Y here. This feature will go away by 2.6.31, so - please arrange to get your userspace programs fixed! - -config EXT4_FS_XATTR - bool "Ext4 extended attributes" - depends on EXT4_FS - default y - help - Extended attributes are name:value pairs associated with inodes by - the kernel or by users (see the attr(5) manual page, or visit - <http://acl.bestbits.at/> for details). - - If unsure, say N. - - You need this for POSIX ACL support on ext4. - -config EXT4_FS_POSIX_ACL - bool "Ext4 POSIX Access Control Lists" - depends on EXT4_FS_XATTR - select FS_POSIX_ACL - help - POSIX Access Control Lists (ACLs) support permissions for users and - groups beyond the owner/group/world scheme. - - To learn more about Access Control Lists, visit the POSIX ACLs for - Linux website <http://acl.bestbits.at/>. - - If you don't know what Access Control Lists are, say N - -config EXT4_FS_SECURITY - bool "Ext4 Security Labels" - depends on EXT4_FS_XATTR - help - Security labels support alternative access control models - implemented by security modules like SELinux. This option - enables an extended attribute handler for file security - labels in the ext4 filesystem. - - If you are not using a security module that requires using - extended attributes for file security labels, say N. - -config JBD - tristate - help - This is a generic journalling layer for block devices. It is - currently used by the ext3 file system, but it could also be - used to add journal support to other file systems or block - devices such as RAID or LVM. - - If you are using the ext3 file system, you need to say Y here. - If you are not using ext3 then you will probably want to say N. - - To compile this device as a module, choose M here: the module will be - called jbd. If you are compiling ext3 into the kernel, you - cannot compile this code as a module. - -config JBD_DEBUG - bool "JBD (ext3) debugging support" - depends on JBD && DEBUG_FS - help - If you are using the ext3 journaled file system (or potentially any - other file system/device using JBD), this option allows you to - enable debugging output while the system is running, in order to - help track down any problems you are having. By default the - debugging output will be turned off. - - If you select Y here, then you will be able to turn on debugging - with "echo N > /sys/kernel/debug/jbd/jbd-debug", where N is a - number between 1 and 5, the higher the number, the more debugging - output is generated. To turn debugging off again, do - "echo 0 > /sys/kernel/debug/jbd/jbd-debug". - -config JBD2 - tristate - select CRC32 - help - This is a generic journaling layer for block devices that support - both 32-bit and 64-bit block numbers. It is currently used by - the ext4 and OCFS2 filesystems, but it could also be used to add - journal support to other file systems or block devices such - as RAID or LVM. - - If you are using ext4 or OCFS2, you need to say Y here. - If you are not using ext4 or OCFS2 then you will - probably want to say N. - - To compile this device as a module, choose M here. The module will be - called jbd2. If you are compiling ext4 or OCFS2 into the kernel, - you cannot compile this code as a module. - -config JBD2_DEBUG - bool "JBD2 (ext4) debugging support" - depends on JBD2 && DEBUG_FS - help - If you are using the ext4 journaled file system (or - potentially any other filesystem/device using JBD2), this option - allows you to enable debugging output while the system is running, - in order to help track down any problems you are having. - By default, the debugging output will be turned off. - - If you select Y here, then you will be able to turn on debugging - with "echo N > /sys/kernel/debug/jbd2/jbd2-debug", where N is a - number between 1 and 5. The higher the number, the more debugging - output is generated. To turn debugging off again, do - "echo 0 > /sys/kernel/debug/jbd2/jbd2-debug". +source "fs/jbd/Kconfig" +source "fs/jbd2/Kconfig" config FS_MBCACHE # Meta block cache for Extended Attributes (ext2/ext3/ext4) tristate - depends on EXT2_FS_XATTR || EXT3_FS_XATTR || EXT4_FS_XATTR - default y if EXT2_FS=y || EXT3_FS=y || EXT4_FS=y - default m if EXT2_FS=m || EXT3_FS=m || EXT4_FS=m + default y if EXT2_FS=y && EXT2_FS_XATTR + default y if EXT3_FS=y && EXT3_FS_XATTR + default y if EXT4_FS=y && EXT4_FS_XATTR + default m if EXT2_FS_XATTR || EXT3_FS_XATTR || EXT4_FS_XATTR config REISERFS_FS tristate "Reiserfs support" @@ -665,7 +404,7 @@ config AUTOFS4_FS N here. config FUSE_FS - tristate "Filesystem in Userspace support" + tristate "FUSE (Filesystem in Userspace) support" help With FUSE it is possible to implement a fully functional filesystem in a userspace program. @@ -1168,195 +907,7 @@ config EFS_FS To compile the EFS file system support as a module, choose M here: the module will be called efs. -config JFFS2_FS - tristate "Journalling Flash File System v2 (JFFS2) support" - select CRC32 - depends on MTD - help - JFFS2 is the second generation of the Journalling Flash File System - for use on diskless embedded devices. It provides improved wear - levelling, compression and support for hard links. You cannot use - this on normal block devices, only on 'MTD' devices. - - Further information on the design and implementation of JFFS2 is - available at <http://sources.redhat.com/jffs2/>. - -config JFFS2_FS_DEBUG - int "JFFS2 debugging verbosity (0 = quiet, 2 = noisy)" - depends on JFFS2_FS - default "0" - help - This controls the amount of debugging messages produced by the JFFS2 - code. Set it to zero for use in production systems. For evaluation, - testing and debugging, it's advisable to set it to one. This will - enable a few assertions and will print debugging messages at the - KERN_DEBUG loglevel, where they won't normally be visible. Level 2 - is unlikely to be useful - it enables extra debugging in certain - areas which at one point needed debugging, but when the bugs were - located and fixed, the detailed messages were relegated to level 2. - - If reporting bugs, please try to have available a full dump of the - messages at debug level 1 while the misbehaviour was occurring. - -config JFFS2_FS_WRITEBUFFER - bool "JFFS2 write-buffering support" - depends on JFFS2_FS - default y - help - This enables the write-buffering support in JFFS2. - - This functionality is required to support JFFS2 on the following - types of flash devices: - - NAND flash - - NOR flash with transparent ECC - - DataFlash - -config JFFS2_FS_WBUF_VERIFY - bool "Verify JFFS2 write-buffer reads" - depends on JFFS2_FS_WRITEBUFFER - default n - help - This causes JFFS2 to read back every page written through the - write-buffer, and check for errors. - -config JFFS2_SUMMARY - bool "JFFS2 summary support (EXPERIMENTAL)" - depends on JFFS2_FS && EXPERIMENTAL - default n - help - This feature makes it possible to use summary information - for faster filesystem mount. - - The summary information can be inserted into a filesystem image - by the utility 'sumtool'. - - If unsure, say 'N'. - -config JFFS2_FS_XATTR - bool "JFFS2 XATTR support (EXPERIMENTAL)" - depends on JFFS2_FS && EXPERIMENTAL - default n - help - Extended attributes are name:value pairs associated with inodes by - the kernel or by users (see the attr(5) manual page, or visit - <http://acl.bestbits.at/> for details). - - If unsure, say N. - -config JFFS2_FS_POSIX_ACL - bool "JFFS2 POSIX Access Control Lists" - depends on JFFS2_FS_XATTR - default y - select FS_POSIX_ACL - help - Posix Access Control Lists (ACLs) support permissions for users and - groups beyond the owner/group/world scheme. - - To learn more about Access Control Lists, visit the Posix ACLs for - Linux website <http://acl.bestbits.at/>. - - If you don't know what Access Control Lists are, say N - -config JFFS2_FS_SECURITY - bool "JFFS2 Security Labels" - depends on JFFS2_FS_XATTR - default y - help - Security labels support alternative access control models - implemented by security modules like SELinux. This option - enables an extended attribute handler for file security - labels in the jffs2 filesystem. - - If you are not using a security module that requires using - extended attributes for file security labels, say N. - -config JFFS2_COMPRESSION_OPTIONS - bool "Advanced compression options for JFFS2" - depends on JFFS2_FS - default n - help - Enabling this option allows you to explicitly choose which - compression modules, if any, are enabled in JFFS2. Removing - compressors can mean you cannot read existing file systems, - and enabling experimental compressors can mean that you - write a file system which cannot be read by a standard kernel. - - If unsure, you should _definitely_ say 'N'. - -config JFFS2_ZLIB - bool "JFFS2 ZLIB compression support" if JFFS2_COMPRESSION_OPTIONS - select ZLIB_INFLATE - select ZLIB_DEFLATE - depends on JFFS2_FS - default y - help - Zlib is designed to be a free, general-purpose, legally unencumbered, - lossless data-compression library for use on virtually any computer - hardware and operating system. See <http://www.gzip.org/zlib/> for - further information. - - Say 'Y' if unsure. - -config JFFS2_LZO - bool "JFFS2 LZO compression support" if JFFS2_COMPRESSION_OPTIONS - select LZO_COMPRESS - select LZO_DECOMPRESS - depends on JFFS2_FS - default n - help - minilzo-based compression. Generally works better than Zlib. - - This feature was added in July, 2007. Say 'N' if you need - compatibility with older bootloaders or kernels. - -config JFFS2_RTIME - bool "JFFS2 RTIME compression support" if JFFS2_COMPRESSION_OPTIONS - depends on JFFS2_FS - default y - help - Rtime does manage to recompress already-compressed data. Say 'Y' if unsure. - -config JFFS2_RUBIN - bool "JFFS2 RUBIN compression support" if JFFS2_COMPRESSION_OPTIONS - depends on JFFS2_FS - default n - help - RUBINMIPS and DYNRUBIN compressors. Say 'N' if unsure. - -choice - prompt "JFFS2 default compression mode" if JFFS2_COMPRESSION_OPTIONS - default JFFS2_CMODE_PRIORITY - depends on JFFS2_FS - help - You can set here the default compression mode of JFFS2 from - the available compression modes. Don't touch if unsure. - -config JFFS2_CMODE_NONE - bool "no compression" - help - Uses no compression. - -config JFFS2_CMODE_PRIORITY - bool "priority" - help - Tries the compressors in a predefined order and chooses the first - successful one. - -config JFFS2_CMODE_SIZE - bool "size (EXPERIMENTAL)" - help - Tries all compressors and chooses the one which has the smallest - result. - -config JFFS2_CMODE_FAVOURLZO - bool "Favour LZO" - help - Tries all compressors and chooses the one which has the smallest - result but gives some preference to LZO (which has faster - decompression) at the expense of size. - -endchoice - +source "fs/jffs2/Kconfig" # UBIFS File system configuration source "fs/ubifs/Kconfig" @@ -1913,148 +1464,7 @@ config SMB_NLS_REMOTE smbmount from samba 2.2.0 or later supports this. -config CIFS - tristate "CIFS support (advanced network filesystem, SMBFS successor)" - depends on INET - select NLS - help - This is the client VFS module for the Common Internet File System - (CIFS) protocol which is the successor to the Server Message Block - (SMB) protocol, the native file sharing mechanism for most early - PC operating systems. The CIFS protocol is fully supported by - file servers such as Windows 2000 (including Windows 2003, NT 4 - and Windows XP) as well by Samba (which provides excellent CIFS - server support for Linux and many other operating systems). Limited - support for OS/2 and Windows ME and similar servers is provided as - well. - - The cifs module provides an advanced network file system - client for mounting to CIFS compliant servers. It includes - support for DFS (hierarchical name space), secure per-user - session establishment via Kerberos or NTLM or NTLMv2, - safe distributed caching (oplock), optional packet - signing, Unicode and other internationalization improvements. - If you need to mount to Samba or Windows from this machine, say Y. - -config CIFS_STATS - bool "CIFS statistics" - depends on CIFS - help - Enabling this option will cause statistics for each server share - mounted by the cifs client to be displayed in /proc/fs/cifs/Stats - -config CIFS_STATS2 - bool "Extended statistics" - depends on CIFS_STATS - help - Enabling this option will allow more detailed statistics on SMB - request timing to be displayed in /proc/fs/cifs/DebugData and also - allow optional logging of slow responses to dmesg (depending on the - value of /proc/fs/cifs/cifsFYI, see fs/cifs/README for more details). - These additional statistics may have a minor effect on performance - and memory utilization. - - Unless you are a developer or are doing network performance analysis - or tuning, say N. - -config CIFS_WEAK_PW_HASH - bool "Support legacy servers which use weaker LANMAN security" - depends on CIFS - help - Modern CIFS servers including Samba and most Windows versions - (since 1997) support stronger NTLM (and even NTLMv2 and Kerberos) - security mechanisms. These hash the password more securely - than the mechanisms used in the older LANMAN version of the - SMB protocol but LANMAN based authentication is needed to - establish sessions with some old SMB servers. - - Enabling this option allows the cifs module to mount to older - LANMAN based servers such as OS/2 and Windows 95, but such - mounts may be less secure than mounts using NTLM or more recent - security mechanisms if you are on a public network. Unless you - have a need to access old SMB servers (and are on a private - network) you probably want to say N. Even if this support - is enabled in the kernel build, LANMAN authentication will not be - used automatically. At runtime LANMAN mounts are disabled but - can be set to required (or optional) either in - /proc/fs/cifs (see fs/cifs/README for more detail) or via an - option on the mount command. This support is disabled by - default in order to reduce the possibility of a downgrade - attack. - - If unsure, say N. - -config CIFS_UPCALL - bool "Kerberos/SPNEGO advanced session setup" - depends on CIFS && KEYS - help - Enables an upcall mechanism for CIFS which accesses - userspace helper utilities to provide SPNEGO packaged (RFC 4178) - Kerberos tickets which are needed to mount to certain secure servers - (for which more secure Kerberos authentication is required). If - unsure, say N. - -config CIFS_XATTR - bool "CIFS extended attributes" - depends on CIFS - help - Extended attributes are name:value pairs associated with inodes by - the kernel or by users (see the attr(5) manual page, or visit - <http://acl.bestbits.at/> for details). CIFS maps the name of - extended attributes beginning with the user namespace prefix - to SMB/CIFS EAs. EAs are stored on Windows servers without the - user namespace prefix, but their names are seen by Linux cifs clients - prefaced by the user namespace prefix. The system namespace - (used by some filesystems to store ACLs) is not supported at - this time. - - If unsure, say N. - -config CIFS_POSIX - bool "CIFS POSIX Extensions" - depends on CIFS_XATTR - help - Enabling this option will cause the cifs client to attempt to - negotiate a newer dialect with servers, such as Samba 3.0.5 - or later, that optionally can handle more POSIX like (rather - than Windows like) file behavior. It also enables - support for POSIX ACLs (getfacl and setfacl) to servers - (such as Samba 3.10 and later) which can negotiate - CIFS POSIX ACL support. If unsure, say N. - -config CIFS_DEBUG2 - bool "Enable additional CIFS debugging routines" - depends on CIFS - help - Enabling this option adds a few more debugging routines - to the cifs code which slightly increases the size of - the cifs module and can cause additional logging of debug - messages in some error paths, slowing performance. This - option can be turned off unless you are debugging - cifs problems. If unsure, say N. - -config CIFS_EXPERIMENTAL - bool "CIFS Experimental Features (EXPERIMENTAL)" - depends on CIFS && EXPERIMENTAL - help - Enables cifs features under testing. These features are - experimental and currently include DFS support and directory - change notification ie fcntl(F_DNOTIFY), as well as the upcall - mechanism which will be used for Kerberos session negotiation - and uid remapping. Some of these features also may depend on - setting a value of 1 to the pseudo-file /proc/fs/cifs/Experimental - (which is disabled by default). See the file fs/cifs/README - for more details. If unsure, say N. - -config CIFS_DFS_UPCALL - bool "DFS feature support (EXPERIMENTAL)" - depends on CIFS_EXPERIMENTAL - depends on KEYS - help - Enables an upcall mechanism for CIFS which contacts userspace - helper utilities to provide server name resolution (host names to - IP addresses) which is needed for implicit mounts of DFS junction - points. If unsure, say N. +source "fs/cifs/Kconfig" config NCP_FS tristate "NCP file system support (to mount NetWare volumes)" diff --git a/fs/Kconfig.binfmt b/fs/Kconfig.binfmt index 17c9c5ec14c..ce9fb3fbfae 100644 --- a/fs/Kconfig.binfmt +++ b/fs/Kconfig.binfmt @@ -25,7 +25,7 @@ config BINFMT_ELF config COMPAT_BINFMT_ELF bool - depends on COMPAT && MMU + depends on COMPAT && BINFMT_ELF config BINFMT_ELF_FDPIC bool "Kernel support for FDPIC ELF binaries" @@ -40,6 +40,28 @@ config BINFMT_ELF_FDPIC It is also possible to run FDPIC ELF binaries on MMU linux also. +config CORE_DUMP_DEFAULT_ELF_HEADERS + bool "Write ELF core dumps with partial segments" + default n + depends on BINFMT_ELF + help + ELF core dump files describe each memory mapping of the crashed + process, and can contain or omit the memory contents of each one. + The contents of an unmodified text mapping are omitted by default. + + For an unmodified text mapping of an ELF object, including just + the first page of the file in a core dump makes it possible to + identify the build ID bits in the file, without paying the i/o + cost and disk space to dump all the text. However, versions of + GDB before 6.7 are confused by ELF core dump files in this format. + + The core dump behavior can be controlled per process using + the /proc/PID/coredump_filter pseudo-file; this setting is + inherited. See Documentation/filesystems/proc.txt for details. + + This config option changes the default setting of coredump_filter + seen at boot time. If unsure, say N. + config BINFMT_FLAT bool "Kernel support for flat binaries" depends on !MMU && (!FRV || BROKEN) diff --git a/fs/Makefile b/fs/Makefile index b6f27dc26b7..d9f8afe6f0c 100644 --- a/fs/Makefile +++ b/fs/Makefile @@ -8,7 +8,7 @@ obj-y := open.o read_write.o file_table.o super.o \ char_dev.o stat.o exec.o pipe.o namei.o fcntl.o \ ioctl.o readdir.o select.o fifo.o dcache.o inode.o \ - attr.o bad_inode.o file.o filesystems.o namespace.o aio.o \ + attr.o bad_inode.o file.o filesystems.o namespace.o \ seq_file.o xattr.o libfs.o fs-writeback.o \ pnode.o drop_caches.o splice.o sync.o utimes.o \ stack.o @@ -27,6 +27,7 @@ obj-$(CONFIG_ANON_INODES) += anon_inodes.o obj-$(CONFIG_SIGNALFD) += signalfd.o obj-$(CONFIG_TIMERFD) += timerfd.o obj-$(CONFIG_EVENTFD) += eventfd.o +obj-$(CONFIG_AIO) += aio.o obj-$(CONFIG_FILE_LOCKING) += locks.o obj-$(CONFIG_COMPAT) += compat.o compat_ioctl.o @@ -70,7 +71,7 @@ obj-$(CONFIG_DLM) += dlm/ # Do not add any filesystems before this line obj-$(CONFIG_REISERFS_FS) += reiserfs/ obj-$(CONFIG_EXT3_FS) += ext3/ # Before ext2 so root fs can be ext3 -obj-$(CONFIG_EXT4_FS) += ext4/ # Before ext2 so root fs can be ext4dev +obj-$(CONFIG_EXT4_FS) += ext4/ # Before ext2 so root fs can be ext4 obj-$(CONFIG_JBD) += jbd/ obj-$(CONFIG_JBD2) += jbd2/ obj-$(CONFIG_EXT2_FS) += ext2/ @@ -80,8 +81,6 @@ obj-$(CONFIG_HUGETLBFS) += hugetlbfs/ obj-$(CONFIG_CODA_FS) += coda/ obj-$(CONFIG_MINIX_FS) += minix/ obj-$(CONFIG_FAT_FS) += fat/ -obj-$(CONFIG_MSDOS_FS) += msdos/ -obj-$(CONFIG_VFAT_FS) += vfat/ obj-$(CONFIG_BFS_FS) += bfs/ obj-$(CONFIG_ISO9660_FS) += isofs/ obj-$(CONFIG_HFSPLUS_FS) += hfsplus/ # Before hfs to find wrapped HFS+ diff --git a/fs/affs/inode.c b/fs/affs/inode.c index a13b334a391..415d9c67ac1 100644 --- a/fs/affs/inode.c +++ b/fs/affs/inode.c @@ -293,8 +293,8 @@ affs_new_inode(struct inode *dir) mark_buffer_dirty_inode(bh, inode); affs_brelse(bh); - inode->i_uid = current->fsuid; - inode->i_gid = current->fsgid; + inode->i_uid = current_fsuid(); + inode->i_gid = current_fsgid(); inode->i_ino = block; inode->i_nlink = 1; inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME_SEC; diff --git a/fs/affs/super.c b/fs/affs/super.c index 8989c93193e..a19d64b582a 100644 --- a/fs/affs/super.c +++ b/fs/affs/super.c @@ -163,8 +163,8 @@ parse_options(char *options, uid_t *uid, gid_t *gid, int *mode, int *reserved, s /* Fill in defaults */ - *uid = current->uid; - *gid = current->gid; + *uid = current_uid(); + *gid = current_gid(); *reserved = 2; *root = -1; *blocksize = -1; diff --git a/fs/afs/dir.c b/fs/afs/dir.c index dfda03d4397..99cf390641f 100644 --- a/fs/afs/dir.c +++ b/fs/afs/dir.c @@ -45,6 +45,7 @@ const struct file_operations afs_dir_file_operations = { .release = afs_release, .readdir = afs_readdir, .lock = afs_lock, + .llseek = generic_file_llseek, }; const struct inode_operations afs_dir_inode_operations = { diff --git a/fs/afs/file.c b/fs/afs/file.c index 525f7c56e06..a3901769a96 100644 --- a/fs/afs/file.c +++ b/fs/afs/file.c @@ -50,8 +50,8 @@ const struct address_space_operations afs_fs_aops = { .launder_page = afs_launder_page, .releasepage = afs_releasepage, .invalidatepage = afs_invalidatepage, - .prepare_write = afs_prepare_write, - .commit_write = afs_commit_write, + .write_begin = afs_write_begin, + .write_end = afs_write_end, .writepage = afs_writepage, .writepages = afs_writepages, }; diff --git a/fs/afs/internal.h b/fs/afs/internal.h index 3cb6920ff30..67f259d99cd 100644 --- a/fs/afs/internal.h +++ b/fs/afs/internal.h @@ -728,8 +728,12 @@ extern int afs_volume_release_fileserver(struct afs_vnode *, */ extern int afs_set_page_dirty(struct page *); extern void afs_put_writeback(struct afs_writeback *); -extern int afs_prepare_write(struct file *, struct page *, unsigned, unsigned); -extern int afs_commit_write(struct file *, struct page *, unsigned, unsigned); +extern int afs_write_begin(struct file *file, struct address_space *mapping, + loff_t pos, unsigned len, unsigned flags, + struct page **pagep, void **fsdata); +extern int afs_write_end(struct file *file, struct address_space *mapping, + loff_t pos, unsigned len, unsigned copied, + struct page *page, void *fsdata); extern int afs_writepage(struct page *, struct writeback_control *); extern int afs_writepages(struct address_space *, struct writeback_control *); extern int afs_write_inode(struct inode *, int); diff --git a/fs/afs/proc.c b/fs/afs/proc.c index 9f7d1ae7026..7578c1ab9e0 100644 --- a/fs/afs/proc.c +++ b/fs/afs/proc.c @@ -646,7 +646,7 @@ static int afs_proc_cell_vlservers_show(struct seq_file *m, void *v) } /* display one cell per line on subsequent lines */ - seq_printf(m, "%u.%u.%u.%u\n", NIPQUAD(addr->s_addr)); + seq_printf(m, "%pI4\n", &addr->s_addr); return 0; } @@ -737,7 +737,7 @@ static int afs_proc_cell_servers_show(struct seq_file *m, void *v) } /* display one cell per line on subsequent lines */ - sprintf(ipaddr, "%u.%u.%u.%u", NIPQUAD(server->addr)); + sprintf(ipaddr, "%pI4", &server->addr); seq_printf(m, "%3d %-15.15s %5d\n", atomic_read(&server->usage), ipaddr, server->fs_state); diff --git a/fs/afs/server.c b/fs/afs/server.c index 28f2451419e..f4909951667 100644 --- a/fs/afs/server.c +++ b/fs/afs/server.c @@ -105,7 +105,7 @@ struct afs_server *afs_lookup_server(struct afs_cell *cell, { struct afs_server *server, *candidate; - _enter("%p,"NIPQUAD_FMT, cell, NIPQUAD(addr->s_addr)); + _enter("%p,%pI4", cell, &addr->s_addr); /* quick scan of the list to see if we already have the server */ read_lock(&cell->servers_lock); @@ -168,9 +168,8 @@ found_server: server_in_two_cells: write_unlock(&cell->servers_lock); kfree(candidate); - printk(KERN_NOTICE "kAFS:" - " Server "NIPQUAD_FMT" appears to be in two cells\n", - NIPQUAD(*addr)); + printk(KERN_NOTICE "kAFS: Server %pI4 appears to be in two cells\n", + addr); _leave(" = -EEXIST"); return ERR_PTR(-EEXIST); } @@ -184,7 +183,7 @@ struct afs_server *afs_find_server(const struct in_addr *_addr) struct rb_node *p; struct in_addr addr = *_addr; - _enter(NIPQUAD_FMT, NIPQUAD(addr.s_addr)); + _enter("%pI4", &addr.s_addr); read_lock(&afs_servers_lock); diff --git a/fs/afs/write.c b/fs/afs/write.c index 065b4e10681..d6b85dab35f 100644 --- a/fs/afs/write.c +++ b/fs/afs/write.c @@ -84,15 +84,23 @@ void afs_put_writeback(struct afs_writeback *wb) * partly or wholly fill a page that's under preparation for writing */ static int afs_fill_page(struct afs_vnode *vnode, struct key *key, - unsigned start, unsigned len, struct page *page) + loff_t pos, unsigned len, struct page *page) { + loff_t i_size; + unsigned eof; int ret; - _enter(",,%u,%u", start, len); + _enter(",,%llu,%u", (unsigned long long)pos, len); - ASSERTCMP(start + len, <=, PAGE_SIZE); + ASSERTCMP(len, <=, PAGE_CACHE_SIZE); - ret = afs_vnode_fetch_data(vnode, key, start, len, page); + i_size = i_size_read(&vnode->vfs_inode); + if (pos + len > i_size) + eof = i_size; + else + eof = PAGE_CACHE_SIZE; + + ret = afs_vnode_fetch_data(vnode, key, 0, eof, page); if (ret < 0) { if (ret == -ENOENT) { _debug("got NOENT from server" @@ -107,109 +115,55 @@ static int afs_fill_page(struct afs_vnode *vnode, struct key *key, } /* - * prepare a page for being written to - */ -static int afs_prepare_page(struct afs_vnode *vnode, struct page *page, - struct key *key, unsigned offset, unsigned to) -{ - unsigned eof, tail, start, stop, len; - loff_t i_size, pos; - void *p; - int ret; - - _enter(""); - - if (offset == 0 && to == PAGE_SIZE) - return 0; - - p = kmap_atomic(page, KM_USER0); - - i_size = i_size_read(&vnode->vfs_inode); - pos = (loff_t) page->index << PAGE_SHIFT; - if (pos >= i_size) { - /* partial write, page beyond EOF */ - _debug("beyond"); - if (offset > 0) - memset(p, 0, offset); - if (to < PAGE_SIZE) - memset(p + to, 0, PAGE_SIZE - to); - kunmap_atomic(p, KM_USER0); - return 0; - } - - if (i_size - pos >= PAGE_SIZE) { - /* partial write, page entirely before EOF */ - _debug("before"); - tail = eof = PAGE_SIZE; - } else { - /* partial write, page overlaps EOF */ - eof = i_size - pos; - _debug("overlap %u", eof); - tail = max(eof, to); - if (tail < PAGE_SIZE) - memset(p + tail, 0, PAGE_SIZE - tail); - if (offset > eof) - memset(p + eof, 0, PAGE_SIZE - eof); - } - - kunmap_atomic(p, KM_USER0); - - ret = 0; - if (offset > 0 || eof > to) { - /* need to fill one or two bits that aren't going to be written - * (cover both fillers in one read if there are two) */ - start = (offset > 0) ? 0 : to; - stop = (eof > to) ? eof : offset; - len = stop - start; - _debug("wr=%u-%u av=0-%u rd=%u@%u", - offset, to, eof, start, len); - ret = afs_fill_page(vnode, key, start, len, page); - } - - _leave(" = %d", ret); - return ret; -} - -/* * prepare to perform part of a write to a page - * - the caller holds the page locked, preventing it from being written out or - * modified by anyone else */ -int afs_prepare_write(struct file *file, struct page *page, - unsigned offset, unsigned to) +int afs_write_begin(struct file *file, struct address_space *mapping, + loff_t pos, unsigned len, unsigned flags, + struct page **pagep, void **fsdata) { struct afs_writeback *candidate, *wb; struct afs_vnode *vnode = AFS_FS_I(file->f_dentry->d_inode); + struct page *page; struct key *key = file->private_data; - pgoff_t index; + unsigned from = pos & (PAGE_CACHE_SIZE - 1); + unsigned to = from + len; + pgoff_t index = pos >> PAGE_CACHE_SHIFT; int ret; _enter("{%x:%u},{%lx},%u,%u", - vnode->fid.vid, vnode->fid.vnode, page->index, offset, to); + vnode->fid.vid, vnode->fid.vnode, index, from, to); candidate = kzalloc(sizeof(*candidate), GFP_KERNEL); if (!candidate) return -ENOMEM; candidate->vnode = vnode; - candidate->first = candidate->last = page->index; - candidate->offset_first = offset; + candidate->first = candidate->last = index; + candidate->offset_first = from; candidate->to_last = to; candidate->usage = 1; candidate->state = AFS_WBACK_PENDING; init_waitqueue_head(&candidate->waitq); + page = __grab_cache_page(mapping, index); + if (!page) { + kfree(candidate); + return -ENOMEM; + } + *pagep = page; + /* page won't leak in error case: it eventually gets cleaned off LRU */ + if (!PageUptodate(page)) { _debug("not up to date"); - ret = afs_prepare_page(vnode, page, key, offset, to); + ret = afs_fill_page(vnode, key, pos, len, page); if (ret < 0) { kfree(candidate); _leave(" = %d [prep]", ret); return ret; } + SetPageUptodate(page); } try_again: - index = page->index; spin_lock(&vnode->writeback_lock); /* see if this page is already pending a writeback under a suitable key @@ -242,8 +196,8 @@ try_again: subsume_in_current_wb: _debug("subsume"); ASSERTRANGE(wb->first, <=, index, <=, wb->last); - if (index == wb->first && offset < wb->offset_first) - wb->offset_first = offset; + if (index == wb->first && from < wb->offset_first) + wb->offset_first = from; if (index == wb->last && to > wb->to_last) wb->to_last = to; spin_unlock(&vnode->writeback_lock); @@ -289,17 +243,17 @@ flush_conflicting_wb: /* * finalise part of a write to a page */ -int afs_commit_write(struct file *file, struct page *page, - unsigned offset, unsigned to) +int afs_write_end(struct file *file, struct address_space *mapping, + loff_t pos, unsigned len, unsigned copied, + struct page *page, void *fsdata) { struct afs_vnode *vnode = AFS_FS_I(file->f_dentry->d_inode); loff_t i_size, maybe_i_size; - _enter("{%x:%u},{%lx},%u,%u", - vnode->fid.vid, vnode->fid.vnode, page->index, offset, to); + _enter("{%x:%u},{%lx}", + vnode->fid.vid, vnode->fid.vnode, page->index); - maybe_i_size = (loff_t) page->index << PAGE_SHIFT; - maybe_i_size += to; + maybe_i_size = pos + copied; i_size = i_size_read(&vnode->vfs_inode); if (maybe_i_size > i_size) { @@ -310,12 +264,13 @@ int afs_commit_write(struct file *file, struct page *page, spin_unlock(&vnode->writeback_lock); } - SetPageUptodate(page); set_page_dirty(page); if (PageDirty(page)) _debug("dirtied"); + unlock_page(page); + page_cache_release(page); - return 0; + return copied; } /* @@ -191,6 +191,20 @@ static int aio_setup_ring(struct kioctx *ctx) kunmap_atomic((void *)((unsigned long)__event & PAGE_MASK), km); \ } while(0) +static void ctx_rcu_free(struct rcu_head *head) +{ + struct kioctx *ctx = container_of(head, struct kioctx, rcu_head); + unsigned nr_events = ctx->max_reqs; + + kmem_cache_free(kioctx_cachep, ctx); + + if (nr_events) { + spin_lock(&aio_nr_lock); + BUG_ON(aio_nr - nr_events > aio_nr); + aio_nr -= nr_events; + spin_unlock(&aio_nr_lock); + } +} /* __put_ioctx * Called when the last user of an aio context has gone away, @@ -198,8 +212,6 @@ static int aio_setup_ring(struct kioctx *ctx) */ static void __put_ioctx(struct kioctx *ctx) { - unsigned nr_events = ctx->max_reqs; - BUG_ON(ctx->reqs_active); cancel_delayed_work(&ctx->wq); @@ -208,14 +220,7 @@ static void __put_ioctx(struct kioctx *ctx) mmdrop(ctx->mm); ctx->mm = NULL; pr_debug("__put_ioctx: freeing %p\n", ctx); - kmem_cache_free(kioctx_cachep, ctx); - - if (nr_events) { - spin_lock(&aio_nr_lock); - BUG_ON(aio_nr - nr_events > aio_nr); - aio_nr -= nr_events; - spin_unlock(&aio_nr_lock); - } + call_rcu(&ctx->rcu_head, ctx_rcu_free); } #define get_ioctx(kioctx) do { \ @@ -235,6 +240,7 @@ static struct kioctx *ioctx_alloc(unsigned nr_events) { struct mm_struct *mm; struct kioctx *ctx; + int did_sync = 0; /* Prevent overflows */ if ((nr_events > (0x10000000U / sizeof(struct io_event))) || @@ -267,21 +273,30 @@ static struct kioctx *ioctx_alloc(unsigned nr_events) goto out_freectx; /* limit the number of system wide aios */ - spin_lock(&aio_nr_lock); - if (aio_nr + ctx->max_reqs > aio_max_nr || - aio_nr + ctx->max_reqs < aio_nr) - ctx->max_reqs = 0; - else - aio_nr += ctx->max_reqs; - spin_unlock(&aio_nr_lock); + do { + spin_lock_bh(&aio_nr_lock); + if (aio_nr + nr_events > aio_max_nr || + aio_nr + nr_events < aio_nr) + ctx->max_reqs = 0; + else + aio_nr += ctx->max_reqs; + spin_unlock_bh(&aio_nr_lock); + if (ctx->max_reqs || did_sync) + break; + + /* wait for rcu callbacks to have completed before giving up */ + synchronize_rcu(); + did_sync = 1; + ctx->max_reqs = nr_events; + } while (1); + if (ctx->max_reqs == 0) goto out_cleanup; /* now link into global list. */ - write_lock(&mm->ioctx_list_lock); - ctx->next = mm->ioctx_list; - mm->ioctx_list = ctx; - write_unlock(&mm->ioctx_list_lock); + spin_lock(&mm->ioctx_lock); + hlist_add_head_rcu(&ctx->list, &mm->ioctx_list); + spin_unlock(&mm->ioctx_lock); dprintk("aio: allocated ioctx %p[%ld]: mm=%p mask=0x%x\n", ctx, ctx->user_id, current->mm, ctx->ring_info.nr); @@ -375,11 +390,12 @@ ssize_t wait_on_sync_kiocb(struct kiocb *iocb) */ void exit_aio(struct mm_struct *mm) { - struct kioctx *ctx = mm->ioctx_list; - mm->ioctx_list = NULL; - while (ctx) { - struct kioctx *next = ctx->next; - ctx->next = NULL; + struct kioctx *ctx; + + while (!hlist_empty(&mm->ioctx_list)) { + ctx = hlist_entry(mm->ioctx_list.first, struct kioctx, list); + hlist_del_rcu(&ctx->list); + aio_cancel_all(ctx); wait_for_all_aios(ctx); @@ -394,7 +410,6 @@ void exit_aio(struct mm_struct *mm) atomic_read(&ctx->users), ctx->dead, ctx->reqs_active); put_ioctx(ctx); - ctx = next; } } @@ -555,19 +570,21 @@ int aio_put_req(struct kiocb *req) static struct kioctx *lookup_ioctx(unsigned long ctx_id) { - struct kioctx *ioctx; - struct mm_struct *mm; + struct mm_struct *mm = current->mm; + struct kioctx *ctx = NULL; + struct hlist_node *n; - mm = current->mm; - read_lock(&mm->ioctx_list_lock); - for (ioctx = mm->ioctx_list; ioctx; ioctx = ioctx->next) - if (likely(ioctx->user_id == ctx_id && !ioctx->dead)) { - get_ioctx(ioctx); + rcu_read_lock(); + + hlist_for_each_entry_rcu(ctx, n, &mm->ioctx_list, list) { + if (ctx->user_id == ctx_id && !ctx->dead) { + get_ioctx(ctx); break; } - read_unlock(&mm->ioctx_list_lock); + } - return ioctx; + rcu_read_unlock(); + return ctx; } /* @@ -1215,19 +1232,14 @@ out: static void io_destroy(struct kioctx *ioctx) { struct mm_struct *mm = current->mm; - struct kioctx **tmp; int was_dead; /* delete the entry from the list is someone else hasn't already */ - write_lock(&mm->ioctx_list_lock); + spin_lock(&mm->ioctx_lock); was_dead = ioctx->dead; ioctx->dead = 1; - for (tmp = &mm->ioctx_list; *tmp && *tmp != ioctx; - tmp = &(*tmp)->next) - ; - if (*tmp) - *tmp = ioctx->next; - write_unlock(&mm->ioctx_list_lock); + hlist_del_rcu(&ioctx->list); + spin_unlock(&mm->ioctx_lock); dprintk("aio_release(%p)\n", ioctx); if (likely(!was_dead)) diff --git a/fs/anon_inodes.c b/fs/anon_inodes.c index 3662dd44896..c16d9be1b01 100644 --- a/fs/anon_inodes.c +++ b/fs/anon_inodes.c @@ -154,8 +154,8 @@ static struct inode *anon_inode_mkinode(void) */ inode->i_state = I_DIRTY; inode->i_mode = S_IRUSR | S_IWUSR; - inode->i_uid = current->fsuid; - inode->i_gid = current->fsgid; + inode->i_uid = current_fsuid(); + inode->i_gid = current_fsgid(); inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME; return inode; } diff --git a/fs/attr.c b/fs/attr.c index 26c71ba1eed..f4360192a93 100644 --- a/fs/attr.c +++ b/fs/attr.c @@ -29,13 +29,13 @@ int inode_change_ok(struct inode *inode, struct iattr *attr) /* Make sure a caller can chown. */ if ((ia_valid & ATTR_UID) && - (current->fsuid != inode->i_uid || + (current_fsuid() != inode->i_uid || attr->ia_uid != inode->i_uid) && !capable(CAP_CHOWN)) goto error; /* Make sure caller can chgrp. */ if ((ia_valid & ATTR_GID) && - (current->fsuid != inode->i_uid || + (current_fsuid() != inode->i_uid || (!in_group_p(attr->ia_gid) && attr->ia_gid != inode->i_gid)) && !capable(CAP_CHOWN)) goto error; @@ -159,17 +159,17 @@ int notify_change(struct dentry * dentry, struct iattr * attr) if (!(attr->ia_valid & ~(ATTR_KILL_SUID | ATTR_KILL_SGID))) return 0; + error = security_inode_setattr(dentry, attr); + if (error) + return error; + if (ia_valid & ATTR_SIZE) down_write(&dentry->d_inode->i_alloc_sem); if (inode->i_op && inode->i_op->setattr) { - error = security_inode_setattr(dentry, attr); - if (!error) - error = inode->i_op->setattr(dentry, attr); + error = inode->i_op->setattr(dentry, attr); } else { error = inode_change_ok(inode, attr); - if (!error) - error = security_inode_setattr(dentry, attr); if (!error) { if ((ia_valid & ATTR_UID && attr->ia_uid != inode->i_uid) || (ia_valid & ATTR_GID && attr->ia_gid != inode->i_gid)) diff --git a/fs/autofs/inode.c b/fs/autofs/inode.c index b70eea1e8c5..c773680d5c6 100644 --- a/fs/autofs/inode.c +++ b/fs/autofs/inode.c @@ -76,8 +76,8 @@ static int parse_options(char *options, int *pipefd, uid_t *uid, gid_t *gid, substring_t args[MAX_OPT_ARGS]; int option; - *uid = current->uid; - *gid = current->gid; + *uid = current_uid(); + *gid = current_gid(); *pgrp = task_pgrp_nr(current); *minproto = *maxproto = AUTOFS_PROTO_VERSION; diff --git a/fs/autofs4/Makefile b/fs/autofs4/Makefile index f2c3b79e94d..a811c1f7d9a 100644 --- a/fs/autofs4/Makefile +++ b/fs/autofs4/Makefile @@ -4,4 +4,4 @@ obj-$(CONFIG_AUTOFS4_FS) += autofs4.o -autofs4-objs := init.o inode.o root.o symlink.o waitq.o expire.o +autofs4-objs := init.o inode.o root.o symlink.o waitq.o expire.o dev-ioctl.o diff --git a/fs/autofs4/autofs_i.h b/fs/autofs4/autofs_i.h index 69a2f5c9231..e0f16da00e5 100644 --- a/fs/autofs4/autofs_i.h +++ b/fs/autofs4/autofs_i.h @@ -14,6 +14,7 @@ /* Internal header file for autofs */ #include <linux/auto_fs4.h> +#include <linux/auto_dev-ioctl.h> #include <linux/mutex.h> #include <linux/list.h> @@ -21,6 +22,11 @@ #define AUTOFS_IOC_FIRST AUTOFS_IOC_READY #define AUTOFS_IOC_COUNT 32 +#define AUTOFS_DEV_IOCTL_IOC_FIRST (AUTOFS_DEV_IOCTL_VERSION) +#define AUTOFS_DEV_IOCTL_IOC_COUNT (AUTOFS_IOC_COUNT - 11) + +#define AUTOFS_TYPE_TRIGGER (AUTOFS_TYPE_DIRECT|AUTOFS_TYPE_OFFSET) + #include <linux/kernel.h> #include <linux/slab.h> #include <linux/time.h> @@ -35,11 +41,27 @@ /* #define DEBUG */ #ifdef DEBUG -#define DPRINTK(fmt,args...) do { printk(KERN_DEBUG "pid %d: %s: " fmt "\n" , current->pid , __func__ , ##args); } while(0) +#define DPRINTK(fmt, args...) \ +do { \ + printk(KERN_DEBUG "pid %d: %s: " fmt "\n", \ + current->pid, __func__, ##args); \ +} while (0) #else -#define DPRINTK(fmt,args...) do {} while(0) +#define DPRINTK(fmt, args...) do {} while (0) #endif +#define AUTOFS_WARN(fmt, args...) \ +do { \ + printk(KERN_WARNING "pid %d: %s: " fmt "\n", \ + current->pid, __func__, ##args); \ +} while (0) + +#define AUTOFS_ERROR(fmt, args...) \ +do { \ + printk(KERN_ERR "pid %d: %s: " fmt "\n", \ + current->pid, __func__, ##args); \ +} while (0) + /* Unified info structure. This is pointed to by both the dentry and inode structures. Each file in the filesystem has an instance of this structure. It holds a reference to the dentry, so dentries are never @@ -61,6 +83,9 @@ struct autofs_info { unsigned long last_used; atomic_t count; + uid_t uid; + gid_t gid; + mode_t mode; size_t size; @@ -92,10 +117,6 @@ struct autofs_wait_queue { #define AUTOFS_SBI_MAGIC 0x6d4a556d -#define AUTOFS_TYPE_INDIRECT 0x0001 -#define AUTOFS_TYPE_DIRECT 0x0002 -#define AUTOFS_TYPE_OFFSET 0x0004 - struct autofs_sb_info { u32 magic; int pipefd; @@ -169,6 +190,17 @@ int autofs4_expire_run(struct super_block *, struct vfsmount *, struct autofs_packet_expire __user *); int autofs4_expire_multi(struct super_block *, struct vfsmount *, struct autofs_sb_info *, int __user *); +struct dentry *autofs4_expire_direct(struct super_block *sb, + struct vfsmount *mnt, + struct autofs_sb_info *sbi, int how); +struct dentry *autofs4_expire_indirect(struct super_block *sb, + struct vfsmount *mnt, + struct autofs_sb_info *sbi, int how); + +/* Device node initialization */ + +int autofs_dev_ioctl_init(void); +void autofs_dev_ioctl_exit(void); /* Operations structures */ diff --git a/fs/autofs4/dev-ioctl.c b/fs/autofs4/dev-ioctl.c new file mode 100644 index 00000000000..63b7c7afe8d --- /dev/null +++ b/fs/autofs4/dev-ioctl.c @@ -0,0 +1,865 @@ +/* + * Copyright 2008 Red Hat, Inc. All rights reserved. + * Copyright 2008 Ian Kent <raven@themaw.net> + * + * This file is part of the Linux kernel and is made available under + * the terms of the GNU General Public License, version 2, or at your + * option, any later version, incorporated herein by reference. + */ + +#include <linux/module.h> +#include <linux/vmalloc.h> +#include <linux/miscdevice.h> +#include <linux/init.h> +#include <linux/wait.h> +#include <linux/namei.h> +#include <linux/fcntl.h> +#include <linux/file.h> +#include <linux/fdtable.h> +#include <linux/sched.h> +#include <linux/compat.h> +#include <linux/syscalls.h> +#include <linux/smp_lock.h> +#include <linux/magic.h> +#include <linux/dcache.h> +#include <linux/uaccess.h> + +#include "autofs_i.h" + +/* + * This module implements an interface for routing autofs ioctl control + * commands via a miscellaneous device file. + * + * The alternate interface is needed because we need to be able open + * an ioctl file descriptor on an autofs mount that may be covered by + * another mount. This situation arises when starting automount(8) + * or other user space daemon which uses direct mounts or offset + * mounts (used for autofs lazy mount/umount of nested mount trees), + * which have been left busy at at service shutdown. + */ + +#define AUTOFS_DEV_IOCTL_SIZE sizeof(struct autofs_dev_ioctl) + +typedef int (*ioctl_fn)(struct file *, struct autofs_sb_info *, + struct autofs_dev_ioctl *); + +static int check_name(const char *name) +{ + if (!strchr(name, '/')) + return -EINVAL; + return 0; +} + +/* + * Check a string doesn't overrun the chunk of + * memory we copied from user land. + */ +static int invalid_str(char *str, void *end) +{ + while ((void *) str <= end) + if (!*str++) + return 0; + return -EINVAL; +} + +/* + * Check that the user compiled against correct version of autofs + * misc device code. + * + * As well as checking the version compatibility this always copies + * the kernel interface version out. + */ +static int check_dev_ioctl_version(int cmd, struct autofs_dev_ioctl *param) +{ + int err = 0; + + if ((AUTOFS_DEV_IOCTL_VERSION_MAJOR != param->ver_major) || + (AUTOFS_DEV_IOCTL_VERSION_MINOR < param->ver_minor)) { + AUTOFS_WARN("ioctl control interface version mismatch: " + "kernel(%u.%u), user(%u.%u), cmd(%d)", + AUTOFS_DEV_IOCTL_VERSION_MAJOR, + AUTOFS_DEV_IOCTL_VERSION_MINOR, + param->ver_major, param->ver_minor, cmd); + err = -EINVAL; + } + + /* Fill in the kernel version. */ + param->ver_major = AUTOFS_DEV_IOCTL_VERSION_MAJOR; + param->ver_minor = AUTOFS_DEV_IOCTL_VERSION_MINOR; + + return err; +} + +/* + * Copy parameter control struct, including a possible path allocated + * at the end of the struct. + */ +static struct autofs_dev_ioctl *copy_dev_ioctl(struct autofs_dev_ioctl __user *in) +{ + struct autofs_dev_ioctl tmp, *ads; + + if (copy_from_user(&tmp, in, sizeof(tmp))) + return ERR_PTR(-EFAULT); + + if (tmp.size < sizeof(tmp)) + return ERR_PTR(-EINVAL); + + ads = kmalloc(tmp.size, GFP_KERNEL); + if (!ads) + return ERR_PTR(-ENOMEM); + + if (copy_from_user(ads, in, tmp.size)) { + kfree(ads); + return ERR_PTR(-EFAULT); + } + + return ads; +} + +static inline void free_dev_ioctl(struct autofs_dev_ioctl *param) +{ + kfree(param); + return; +} + +/* + * Check sanity of parameter control fields and if a path is present + * check that it has a "/" and is terminated. + */ +static int validate_dev_ioctl(int cmd, struct autofs_dev_ioctl *param) +{ + int err; + + err = check_dev_ioctl_version(cmd, param); + if (err) { + AUTOFS_WARN("invalid device control module version " + "supplied for cmd(0x%08x)", cmd); + goto out; + } + + if (param->size > sizeof(*param)) { + err = check_name(param->path); + if (err) { + AUTOFS_WARN("invalid path supplied for cmd(0x%08x)", + cmd); + goto out; + } + + err = invalid_str(param->path, + (void *) ((size_t) param + param->size)); + if (err) { + AUTOFS_WARN("invalid path supplied for cmd(0x%08x)", + cmd); + goto out; + } + } + + err = 0; +out: + return err; +} + +/* + * Get the autofs super block info struct from the file opened on + * the autofs mount point. + */ +static struct autofs_sb_info *autofs_dev_ioctl_sbi(struct file *f) +{ + struct autofs_sb_info *sbi = NULL; + struct inode *inode; + + if (f) { + inode = f->f_path.dentry->d_inode; + sbi = autofs4_sbi(inode->i_sb); + } + return sbi; +} + +/* Return autofs module protocol version */ +static int autofs_dev_ioctl_protover(struct file *fp, + struct autofs_sb_info *sbi, + struct autofs_dev_ioctl *param) +{ + param->arg1 = sbi->version; + return 0; +} + +/* Return autofs module protocol sub version */ +static int autofs_dev_ioctl_protosubver(struct file *fp, + struct autofs_sb_info *sbi, + struct autofs_dev_ioctl *param) +{ + param->arg1 = sbi->sub_version; + return 0; +} + +/* + * Walk down the mount stack looking for an autofs mount that + * has the requested device number (aka. new_encode_dev(sb->s_dev). + */ +static int autofs_dev_ioctl_find_super(struct nameidata *nd, dev_t devno) +{ + struct dentry *dentry; + struct inode *inode; + struct super_block *sb; + dev_t s_dev; + unsigned int err; + + err = -ENOENT; + + /* Lookup the dentry name at the base of our mount point */ + dentry = d_lookup(nd->path.dentry, &nd->last); + if (!dentry) + goto out; + + dput(nd->path.dentry); + nd->path.dentry = dentry; + + /* And follow the mount stack looking for our autofs mount */ + while (follow_down(&nd->path.mnt, &nd->path.dentry)) { + inode = nd->path.dentry->d_inode; + if (!inode) + break; + + sb = inode->i_sb; + s_dev = new_encode_dev(sb->s_dev); + if (devno == s_dev) { + if (sb->s_magic == AUTOFS_SUPER_MAGIC) { + err = 0; + break; + } + } + } +out: + return err; +} + +/* + * Walk down the mount stack looking for an autofs mount that + * has the requested mount type (ie. indirect, direct or offset). + */ +static int autofs_dev_ioctl_find_sbi_type(struct nameidata *nd, unsigned int type) +{ + struct dentry *dentry; + struct autofs_info *ino; + unsigned int err; + + err = -ENOENT; + + /* Lookup the dentry name at the base of our mount point */ + dentry = d_lookup(nd->path.dentry, &nd->last); + if (!dentry) + goto out; + + dput(nd->path.dentry); + nd->path.dentry = dentry; + + /* And follow the mount stack looking for our autofs mount */ + while (follow_down(&nd->path.mnt, &nd->path.dentry)) { + ino = autofs4_dentry_ino(nd->path.dentry); + if (ino && ino->sbi->type & type) { + err = 0; + break; + } + } +out: + return err; +} + +static void autofs_dev_ioctl_fd_install(unsigned int fd, struct file *file) +{ + struct files_struct *files = current->files; + struct fdtable *fdt; + + spin_lock(&files->file_lock); + fdt = files_fdtable(files); + BUG_ON(fdt->fd[fd] != NULL); + rcu_assign_pointer(fdt->fd[fd], file); + FD_SET(fd, fdt->close_on_exec); + spin_unlock(&files->file_lock); +} + + +/* + * Open a file descriptor on the autofs mount point corresponding + * to the given path and device number (aka. new_encode_dev(sb->s_dev)). + */ +static int autofs_dev_ioctl_open_mountpoint(const char *path, dev_t devid) +{ + struct file *filp; + struct nameidata nd; + int err, fd; + + fd = get_unused_fd(); + if (likely(fd >= 0)) { + /* Get nameidata of the parent directory */ + err = path_lookup(path, LOOKUP_PARENT, &nd); + if (err) + goto out; + + /* + * Search down, within the parent, looking for an + * autofs super block that has the device number + * corresponding to the autofs fs we want to open. + */ + err = autofs_dev_ioctl_find_super(&nd, devid); + if (err) { + path_put(&nd.path); + goto out; + } + + filp = dentry_open(nd.path.dentry, nd.path.mnt, O_RDONLY, + current_cred()); + if (IS_ERR(filp)) { + err = PTR_ERR(filp); + goto out; + } + + autofs_dev_ioctl_fd_install(fd, filp); + } + + return fd; + +out: + put_unused_fd(fd); + return err; +} + +/* Open a file descriptor on an autofs mount point */ +static int autofs_dev_ioctl_openmount(struct file *fp, + struct autofs_sb_info *sbi, + struct autofs_dev_ioctl *param) +{ + const char *path; + dev_t devid; + int err, fd; + + /* param->path has already been checked */ + if (!param->arg1) + return -EINVAL; + + param->ioctlfd = -1; + + path = param->path; + devid = param->arg1; + + err = 0; + fd = autofs_dev_ioctl_open_mountpoint(path, devid); + if (unlikely(fd < 0)) { + err = fd; + goto out; + } + + param->ioctlfd = fd; +out: + return err; +} + +/* Close file descriptor allocated above (user can also use close(2)). */ +static int autofs_dev_ioctl_closemount(struct file *fp, + struct autofs_sb_info *sbi, + struct autofs_dev_ioctl *param) +{ + return sys_close(param->ioctlfd); +} + +/* + * Send "ready" status for an existing wait (either a mount or an expire + * request). + */ +static int autofs_dev_ioctl_ready(struct file *fp, + struct autofs_sb_info *sbi, + struct autofs_dev_ioctl *param) +{ + autofs_wqt_t token; + + token = (autofs_wqt_t) param->arg1; + return autofs4_wait_release(sbi, token, 0); +} + +/* + * Send "fail" status for an existing wait (either a mount or an expire + * request). + */ +static int autofs_dev_ioctl_fail(struct file *fp, + struct autofs_sb_info *sbi, + struct autofs_dev_ioctl *param) +{ + autofs_wqt_t token; + int status; + + token = (autofs_wqt_t) param->arg1; + status = param->arg2 ? param->arg2 : -ENOENT; + return autofs4_wait_release(sbi, token, status); +} + +/* + * Set the pipe fd for kernel communication to the daemon. + * + * Normally this is set at mount using an option but if we + * are reconnecting to a busy mount then we need to use this + * to tell the autofs mount about the new kernel pipe fd. In + * order to protect mounts against incorrectly setting the + * pipefd we also require that the autofs mount be catatonic. + * + * This also sets the process group id used to identify the + * controlling process (eg. the owning automount(8) daemon). + */ +static int autofs_dev_ioctl_setpipefd(struct file *fp, + struct autofs_sb_info *sbi, + struct autofs_dev_ioctl *param) +{ + int pipefd; + int err = 0; + + if (param->arg1 == -1) + return -EINVAL; + + pipefd = param->arg1; + + mutex_lock(&sbi->wq_mutex); + if (!sbi->catatonic) { + mutex_unlock(&sbi->wq_mutex); + return -EBUSY; + } else { + struct file *pipe = fget(pipefd); + if (!pipe->f_op || !pipe->f_op->write) { + err = -EPIPE; + fput(pipe); + goto out; + } + sbi->oz_pgrp = task_pgrp_nr(current); + sbi->pipefd = pipefd; + sbi->pipe = pipe; + sbi->catatonic = 0; + } +out: + mutex_unlock(&sbi->wq_mutex); + return err; +} + +/* + * Make the autofs mount point catatonic, no longer responsive to + * mount requests. Also closes the kernel pipe file descriptor. + */ +static int autofs_dev_ioctl_catatonic(struct file *fp, + struct autofs_sb_info *sbi, + struct autofs_dev_ioctl *param) +{ + autofs4_catatonic_mode(sbi); + return 0; +} + +/* Set the autofs mount timeout */ +static int autofs_dev_ioctl_timeout(struct file *fp, + struct autofs_sb_info *sbi, + struct autofs_dev_ioctl *param) +{ + unsigned long timeout; + + timeout = param->arg1; + param->arg1 = sbi->exp_timeout / HZ; + sbi->exp_timeout = timeout * HZ; + return 0; +} + +/* + * Return the uid and gid of the last request for the mount + * + * When reconstructing an autofs mount tree with active mounts + * we need to re-connect to mounts that may have used the original + * process uid and gid (or string variations of them) for mount + * lookups within the map entry. + */ +static int autofs_dev_ioctl_requester(struct file *fp, + struct autofs_sb_info *sbi, + struct autofs_dev_ioctl *param) +{ + struct autofs_info *ino; + struct nameidata nd; + const char *path; + dev_t devid; + int err = -ENOENT; + + if (param->size <= sizeof(*param)) { + err = -EINVAL; + goto out; + } + + path = param->path; + devid = sbi->sb->s_dev; + + param->arg1 = param->arg2 = -1; + + /* Get nameidata of the parent directory */ + err = path_lookup(path, LOOKUP_PARENT, &nd); + if (err) + goto out; + + err = autofs_dev_ioctl_find_super(&nd, devid); + if (err) + goto out_release; + + ino = autofs4_dentry_ino(nd.path.dentry); + if (ino) { + err = 0; + autofs4_expire_wait(nd.path.dentry); + spin_lock(&sbi->fs_lock); + param->arg1 = ino->uid; + param->arg2 = ino->gid; + spin_unlock(&sbi->fs_lock); + } + +out_release: + path_put(&nd.path); +out: + return err; +} + +/* + * Call repeatedly until it returns -EAGAIN, meaning there's nothing + * more that can be done. + */ +static int autofs_dev_ioctl_expire(struct file *fp, + struct autofs_sb_info *sbi, + struct autofs_dev_ioctl *param) +{ + struct dentry *dentry; + struct vfsmount *mnt; + int err = -EAGAIN; + int how; + + how = param->arg1; + mnt = fp->f_path.mnt; + + if (sbi->type & AUTOFS_TYPE_TRIGGER) + dentry = autofs4_expire_direct(sbi->sb, mnt, sbi, how); + else + dentry = autofs4_expire_indirect(sbi->sb, mnt, sbi, how); + + if (dentry) { + struct autofs_info *ino = autofs4_dentry_ino(dentry); + + /* + * This is synchronous because it makes the daemon a + * little easier + */ + err = autofs4_wait(sbi, dentry, NFY_EXPIRE); + + spin_lock(&sbi->fs_lock); + if (ino->flags & AUTOFS_INF_MOUNTPOINT) { + ino->flags &= ~AUTOFS_INF_MOUNTPOINT; + sbi->sb->s_root->d_mounted++; + } + ino->flags &= ~AUTOFS_INF_EXPIRING; + complete_all(&ino->expire_complete); + spin_unlock(&sbi->fs_lock); + dput(dentry); + } + + return err; +} + +/* Check if autofs mount point is in use */ +static int autofs_dev_ioctl_askumount(struct file *fp, + struct autofs_sb_info *sbi, + struct autofs_dev_ioctl *param) +{ + param->arg1 = 0; + if (may_umount(fp->f_path.mnt)) + param->arg1 = 1; + return 0; +} + +/* + * Check if the given path is a mountpoint. + * + * If we are supplied with the file descriptor of an autofs + * mount we're looking for a specific mount. In this case + * the path is considered a mountpoint if it is itself a + * mountpoint or contains a mount, such as a multi-mount + * without a root mount. In this case we return 1 if the + * path is a mount point and the super magic of the covering + * mount if there is one or 0 if it isn't a mountpoint. + * + * If we aren't supplied with a file descriptor then we + * lookup the nameidata of the path and check if it is the + * root of a mount. If a type is given we are looking for + * a particular autofs mount and if we don't find a match + * we return fail. If the located nameidata path is the + * root of a mount we return 1 along with the super magic + * of the mount or 0 otherwise. + * + * In both cases the the device number (as returned by + * new_encode_dev()) is also returned. + */ +static int autofs_dev_ioctl_ismountpoint(struct file *fp, + struct autofs_sb_info *sbi, + struct autofs_dev_ioctl *param) +{ + struct nameidata nd; + const char *path; + unsigned int type; + int err = -ENOENT; + + if (param->size <= sizeof(*param)) { + err = -EINVAL; + goto out; + } + + path = param->path; + type = param->arg1; + + param->arg1 = 0; + param->arg2 = 0; + + if (!fp || param->ioctlfd == -1) { + if (type == AUTOFS_TYPE_ANY) { + struct super_block *sb; + + err = path_lookup(path, LOOKUP_FOLLOW, &nd); + if (err) + goto out; + + sb = nd.path.dentry->d_sb; + param->arg1 = new_encode_dev(sb->s_dev); + } else { + struct autofs_info *ino; + + err = path_lookup(path, LOOKUP_PARENT, &nd); + if (err) + goto out; + + err = autofs_dev_ioctl_find_sbi_type(&nd, type); + if (err) + goto out_release; + + ino = autofs4_dentry_ino(nd.path.dentry); + param->arg1 = autofs4_get_dev(ino->sbi); + } + + err = 0; + if (nd.path.dentry->d_inode && + nd.path.mnt->mnt_root == nd.path.dentry) { + err = 1; + param->arg2 = nd.path.dentry->d_inode->i_sb->s_magic; + } + } else { + dev_t devid = new_encode_dev(sbi->sb->s_dev); + + err = path_lookup(path, LOOKUP_PARENT, &nd); + if (err) + goto out; + + err = autofs_dev_ioctl_find_super(&nd, devid); + if (err) + goto out_release; + + param->arg1 = autofs4_get_dev(sbi); + + err = have_submounts(nd.path.dentry); + + if (nd.path.mnt->mnt_mountpoint != nd.path.mnt->mnt_root) { + if (follow_down(&nd.path.mnt, &nd.path.dentry)) { + struct inode *inode = nd.path.dentry->d_inode; + param->arg2 = inode->i_sb->s_magic; + } + } + } + +out_release: + path_put(&nd.path); +out: + return err; +} + +/* + * Our range of ioctl numbers isn't 0 based so we need to shift + * the array index by _IOC_NR(AUTOFS_CTL_IOC_FIRST) for the table + * lookup. + */ +#define cmd_idx(cmd) (cmd - _IOC_NR(AUTOFS_DEV_IOCTL_IOC_FIRST)) + +static ioctl_fn lookup_dev_ioctl(unsigned int cmd) +{ + static struct { + int cmd; + ioctl_fn fn; + } _ioctls[] = { + {cmd_idx(AUTOFS_DEV_IOCTL_VERSION_CMD), NULL}, + {cmd_idx(AUTOFS_DEV_IOCTL_PROTOVER_CMD), + autofs_dev_ioctl_protover}, + {cmd_idx(AUTOFS_DEV_IOCTL_PROTOSUBVER_CMD), + autofs_dev_ioctl_protosubver}, + {cmd_idx(AUTOFS_DEV_IOCTL_OPENMOUNT_CMD), + autofs_dev_ioctl_openmount}, + {cmd_idx(AUTOFS_DEV_IOCTL_CLOSEMOUNT_CMD), + autofs_dev_ioctl_closemount}, + {cmd_idx(AUTOFS_DEV_IOCTL_READY_CMD), + autofs_dev_ioctl_ready}, + {cmd_idx(AUTOFS_DEV_IOCTL_FAIL_CMD), + autofs_dev_ioctl_fail}, + {cmd_idx(AUTOFS_DEV_IOCTL_SETPIPEFD_CMD), + autofs_dev_ioctl_setpipefd}, + {cmd_idx(AUTOFS_DEV_IOCTL_CATATONIC_CMD), + autofs_dev_ioctl_catatonic}, + {cmd_idx(AUTOFS_DEV_IOCTL_TIMEOUT_CMD), + autofs_dev_ioctl_timeout}, + {cmd_idx(AUTOFS_DEV_IOCTL_REQUESTER_CMD), + autofs_dev_ioctl_requester}, + {cmd_idx(AUTOFS_DEV_IOCTL_EXPIRE_CMD), + autofs_dev_ioctl_expire}, + {cmd_idx(AUTOFS_DEV_IOCTL_ASKUMOUNT_CMD), + autofs_dev_ioctl_askumount}, + {cmd_idx(AUTOFS_DEV_IOCTL_ISMOUNTPOINT_CMD), + autofs_dev_ioctl_ismountpoint} + }; + unsigned int idx = cmd_idx(cmd); + + return (idx >= ARRAY_SIZE(_ioctls)) ? NULL : _ioctls[idx].fn; +} + +/* ioctl dispatcher */ +static int _autofs_dev_ioctl(unsigned int command, struct autofs_dev_ioctl __user *user) +{ + struct autofs_dev_ioctl *param; + struct file *fp; + struct autofs_sb_info *sbi; + unsigned int cmd_first, cmd; + ioctl_fn fn = NULL; + int err = 0; + + /* only root can play with this */ + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; + + cmd_first = _IOC_NR(AUTOFS_DEV_IOCTL_IOC_FIRST); + cmd = _IOC_NR(command); + + if (_IOC_TYPE(command) != _IOC_TYPE(AUTOFS_DEV_IOCTL_IOC_FIRST) || + cmd - cmd_first >= AUTOFS_DEV_IOCTL_IOC_COUNT) { + return -ENOTTY; + } + + /* Copy the parameters into kernel space. */ + param = copy_dev_ioctl(user); + if (IS_ERR(param)) + return PTR_ERR(param); + + err = validate_dev_ioctl(command, param); + if (err) + goto out; + + /* The validate routine above always sets the version */ + if (cmd == AUTOFS_DEV_IOCTL_VERSION_CMD) + goto done; + + fn = lookup_dev_ioctl(cmd); + if (!fn) { + AUTOFS_WARN("unknown command 0x%08x", command); + return -ENOTTY; + } + + fp = NULL; + sbi = NULL; + + /* + * For obvious reasons the openmount can't have a file + * descriptor yet. We don't take a reference to the + * file during close to allow for immediate release. + */ + if (cmd != AUTOFS_DEV_IOCTL_OPENMOUNT_CMD && + cmd != AUTOFS_DEV_IOCTL_CLOSEMOUNT_CMD) { + fp = fget(param->ioctlfd); + if (!fp) { + if (cmd == AUTOFS_DEV_IOCTL_ISMOUNTPOINT_CMD) + goto cont; + err = -EBADF; + goto out; + } + + if (!fp->f_op) { + err = -ENOTTY; + fput(fp); + goto out; + } + + sbi = autofs_dev_ioctl_sbi(fp); + if (!sbi || sbi->magic != AUTOFS_SBI_MAGIC) { + err = -EINVAL; + fput(fp); + goto out; + } + + /* + * Admin needs to be able to set the mount catatonic in + * order to be able to perform the re-open. + */ + if (!autofs4_oz_mode(sbi) && + cmd != AUTOFS_DEV_IOCTL_CATATONIC_CMD) { + err = -EACCES; + fput(fp); + goto out; + } + } +cont: + err = fn(fp, sbi, param); + + if (fp) + fput(fp); +done: + if (err >= 0 && copy_to_user(user, param, AUTOFS_DEV_IOCTL_SIZE)) + err = -EFAULT; +out: + free_dev_ioctl(param); + return err; +} + +static long autofs_dev_ioctl(struct file *file, uint command, ulong u) +{ + int err; + err = _autofs_dev_ioctl(command, (struct autofs_dev_ioctl __user *) u); + return (long) err; +} + +#ifdef CONFIG_COMPAT +static long autofs_dev_ioctl_compat(struct file *file, uint command, ulong u) +{ + return (long) autofs_dev_ioctl(file, command, (ulong) compat_ptr(u)); +} +#else +#define autofs_dev_ioctl_compat NULL +#endif + +static const struct file_operations _dev_ioctl_fops = { + .unlocked_ioctl = autofs_dev_ioctl, + .compat_ioctl = autofs_dev_ioctl_compat, + .owner = THIS_MODULE, +}; + +static struct miscdevice _autofs_dev_ioctl_misc = { + .minor = MISC_DYNAMIC_MINOR, + .name = AUTOFS_DEVICE_NAME, + .fops = &_dev_ioctl_fops +}; + +/* Register/deregister misc character device */ +int autofs_dev_ioctl_init(void) +{ + int r; + + r = misc_register(&_autofs_dev_ioctl_misc); + if (r) { + AUTOFS_ERROR("misc_register failed for control device"); + return r; + } + + return 0; +} + +void autofs_dev_ioctl_exit(void) +{ + misc_deregister(&_autofs_dev_ioctl_misc); + return; +} + diff --git a/fs/autofs4/expire.c b/fs/autofs4/expire.c index cdabb796ff0..4b6fb3f628c 100644 --- a/fs/autofs4/expire.c +++ b/fs/autofs4/expire.c @@ -56,12 +56,23 @@ static int autofs4_mount_busy(struct vfsmount *mnt, struct dentry *dentry) mntget(mnt); dget(dentry); - if (!autofs4_follow_mount(&mnt, &dentry)) + if (!follow_down(&mnt, &dentry)) goto done; - /* This is an autofs submount, we can't expire it */ - if (is_autofs4_dentry(dentry)) - goto done; + if (is_autofs4_dentry(dentry)) { + struct autofs_sb_info *sbi = autofs4_sbi(dentry->d_sb); + + /* This is an autofs submount, we can't expire it */ + if (sbi->type == AUTOFS_TYPE_INDIRECT) + goto done; + + /* + * Otherwise it's an offset mount and we need to check + * if we can umount its mount, if there is one. + */ + if (!d_mountpoint(dentry)) + goto done; + } /* Update the expiry counter if fs is busy */ if (!may_umount_tree(mnt)) { @@ -244,10 +255,10 @@ cont: } /* Check if we can expire a direct mount (possibly a tree) */ -static struct dentry *autofs4_expire_direct(struct super_block *sb, - struct vfsmount *mnt, - struct autofs_sb_info *sbi, - int how) +struct dentry *autofs4_expire_direct(struct super_block *sb, + struct vfsmount *mnt, + struct autofs_sb_info *sbi, + int how) { unsigned long timeout; struct dentry *root = dget(sb->s_root); @@ -283,10 +294,10 @@ static struct dentry *autofs4_expire_direct(struct super_block *sb, * - it is unused by any user process * - it has been unused for exp_timeout time */ -static struct dentry *autofs4_expire_indirect(struct super_block *sb, - struct vfsmount *mnt, - struct autofs_sb_info *sbi, - int how) +struct dentry *autofs4_expire_indirect(struct super_block *sb, + struct vfsmount *mnt, + struct autofs_sb_info *sbi, + int how) { unsigned long timeout; struct dentry *root = sb->s_root; @@ -479,7 +490,7 @@ int autofs4_expire_multi(struct super_block *sb, struct vfsmount *mnt, if (arg && get_user(do_now, arg)) return -EFAULT; - if (sbi->type & AUTOFS_TYPE_DIRECT) + if (sbi->type & AUTOFS_TYPE_TRIGGER) dentry = autofs4_expire_direct(sb, mnt, sbi, do_now); else dentry = autofs4_expire_indirect(sb, mnt, sbi, do_now); diff --git a/fs/autofs4/init.c b/fs/autofs4/init.c index 723a1c5e361..9722e4bd895 100644 --- a/fs/autofs4/init.c +++ b/fs/autofs4/init.c @@ -29,11 +29,20 @@ static struct file_system_type autofs_fs_type = { static int __init init_autofs4_fs(void) { - return register_filesystem(&autofs_fs_type); + int err; + + err = register_filesystem(&autofs_fs_type); + if (err) + return err; + + autofs_dev_ioctl_init(); + + return err; } static void __exit exit_autofs4_fs(void) { + autofs_dev_ioctl_exit(); unregister_filesystem(&autofs_fs_type); } diff --git a/fs/autofs4/inode.c b/fs/autofs4/inode.c index 45d55819203..7b19802cfef 100644 --- a/fs/autofs4/inode.c +++ b/fs/autofs4/inode.c @@ -53,6 +53,8 @@ struct autofs_info *autofs4_init_ino(struct autofs_info *ino, atomic_set(&ino->count, 0); } + ino->uid = 0; + ino->gid = 0; ino->mode = mode; ino->last_used = jiffies; @@ -233,8 +235,8 @@ static int parse_options(char *options, int *pipefd, uid_t *uid, gid_t *gid, substring_t args[MAX_OPT_ARGS]; int option; - *uid = current->uid; - *gid = current->gid; + *uid = current_uid(); + *gid = current_gid(); *pgrp = task_pgrp_nr(current); *minproto = AUTOFS_MIN_PROTO_VERSION; @@ -288,7 +290,7 @@ static int parse_options(char *options, int *pipefd, uid_t *uid, gid_t *gid, *type = AUTOFS_TYPE_DIRECT; break; case Opt_offset: - *type = AUTOFS_TYPE_DIRECT | AUTOFS_TYPE_OFFSET; + *type = AUTOFS_TYPE_OFFSET; break; default: return 1; @@ -336,7 +338,7 @@ int autofs4_fill_super(struct super_block *s, void *data, int silent) sbi->sb = s; sbi->version = 0; sbi->sub_version = 0; - sbi->type = 0; + sbi->type = AUTOFS_TYPE_INDIRECT; sbi->min_proto = 0; sbi->max_proto = 0; mutex_init(&sbi->wq_mutex); @@ -378,7 +380,7 @@ int autofs4_fill_super(struct super_block *s, void *data, int silent) } root_inode->i_fop = &autofs4_root_operations; - root_inode->i_op = sbi->type & AUTOFS_TYPE_DIRECT ? + root_inode->i_op = sbi->type & AUTOFS_TYPE_TRIGGER ? &autofs4_direct_root_inode_operations : &autofs4_indirect_root_inode_operations; diff --git a/fs/autofs4/waitq.c b/fs/autofs4/waitq.c index 35216d18d8b..e02cc8ae5eb 100644 --- a/fs/autofs4/waitq.c +++ b/fs/autofs4/waitq.c @@ -337,7 +337,7 @@ int autofs4_wait(struct autofs_sb_info *sbi, struct dentry *dentry, * is very similar for indirect mounts except only dentrys * in the root of the autofs file system may be negative. */ - if (sbi->type & (AUTOFS_TYPE_DIRECT|AUTOFS_TYPE_OFFSET)) + if (sbi->type & AUTOFS_TYPE_TRIGGER) return -ENOENT; else if (!IS_ROOT(dentry->d_parent)) return -ENOENT; @@ -348,7 +348,7 @@ int autofs4_wait(struct autofs_sb_info *sbi, struct dentry *dentry, return -ENOMEM; /* If this is a direct mount request create a dummy name */ - if (IS_ROOT(dentry) && (sbi->type & AUTOFS_TYPE_DIRECT)) + if (IS_ROOT(dentry) && sbi->type & AUTOFS_TYPE_TRIGGER) qstr.len = sprintf(name, "%p", dentry); else { qstr.len = autofs4_getpath(sbi, dentry, &name); @@ -391,8 +391,8 @@ int autofs4_wait(struct autofs_sb_info *sbi, struct dentry *dentry, memcpy(&wq->name, &qstr, sizeof(struct qstr)); wq->dev = autofs4_get_dev(sbi); wq->ino = autofs4_get_ino(sbi); - wq->uid = current->uid; - wq->gid = current->gid; + wq->uid = current_uid(); + wq->gid = current_gid(); wq->pid = current->pid; wq->tgid = current->tgid; wq->status = -EINTR; /* Status return if interrupted */ @@ -406,11 +406,11 @@ int autofs4_wait(struct autofs_sb_info *sbi, struct dentry *dentry, type = autofs_ptype_expire_multi; } else { if (notify == NFY_MOUNT) - type = (sbi->type & AUTOFS_TYPE_DIRECT) ? + type = (sbi->type & AUTOFS_TYPE_TRIGGER) ? autofs_ptype_missing_direct : autofs_ptype_missing_indirect; else - type = (sbi->type & AUTOFS_TYPE_DIRECT) ? + type = (sbi->type & AUTOFS_TYPE_TRIGGER) ? autofs_ptype_expire_direct : autofs_ptype_expire_indirect; } @@ -457,6 +457,40 @@ int autofs4_wait(struct autofs_sb_info *sbi, struct dentry *dentry, status = wq->status; + /* + * For direct and offset mounts we need to track the requester's + * uid and gid in the dentry info struct. This is so it can be + * supplied, on request, by the misc device ioctl interface. + * This is needed during daemon resatart when reconnecting + * to existing, active, autofs mounts. The uid and gid (and + * related string values) may be used for macro substitution + * in autofs mount maps. + */ + if (!status) { + struct autofs_info *ino; + struct dentry *de = NULL; + + /* direct mount or browsable map */ + ino = autofs4_dentry_ino(dentry); + if (!ino) { + /* If not lookup actual dentry used */ + de = d_lookup(dentry->d_parent, &dentry->d_name); + if (de) + ino = autofs4_dentry_ino(de); + } + + /* Set mount requester */ + if (ino) { + spin_lock(&sbi->fs_lock); + ino->uid = wq->uid; + ino->gid = wq->gid; + spin_unlock(&sbi->fs_lock); + } + + if (de) + dput(de); + } + /* Are we the last process to need status? */ mutex_lock(&sbi->wq_mutex); if (!--wq->wait_ctr) diff --git a/fs/befs/befs_fs_types.h b/fs/befs/befs_fs_types.h index e2595c2c403..7893eaa1e58 100644 --- a/fs/befs/befs_fs_types.h +++ b/fs/befs/befs_fs_types.h @@ -55,8 +55,12 @@ enum super_flags { }; #define BEFS_BYTEORDER_NATIVE 0x42494745 +#define BEFS_BYTEORDER_NATIVE_LE (__force fs32)cpu_to_le32(BEFS_BYTEORDER_NATIVE) +#define BEFS_BYTEORDER_NATIVE_BE (__force fs32)cpu_to_be32(BEFS_BYTEORDER_NATIVE) #define BEFS_SUPER_MAGIC BEFS_SUPER_MAGIC1 +#define BEFS_SUPER_MAGIC1_LE (__force fs32)cpu_to_le32(BEFS_SUPER_MAGIC1) +#define BEFS_SUPER_MAGIC1_BE (__force fs32)cpu_to_be32(BEFS_SUPER_MAGIC1) /* * Flags of inode diff --git a/fs/befs/linuxvfs.c b/fs/befs/linuxvfs.c index 9286b2af893..b6dfee37c7b 100644 --- a/fs/befs/linuxvfs.c +++ b/fs/befs/linuxvfs.c @@ -809,8 +809,8 @@ befs_fill_super(struct super_block *sb, void *data, int silent) /* account for offset of super block on x86 */ disk_sb = (befs_super_block *) bh->b_data; - if ((le32_to_cpu(disk_sb->magic1) == BEFS_SUPER_MAGIC1) || - (be32_to_cpu(disk_sb->magic1) == BEFS_SUPER_MAGIC1)) { + if ((disk_sb->magic1 == BEFS_SUPER_MAGIC1_LE) || + (disk_sb->magic1 == BEFS_SUPER_MAGIC1_BE)) { befs_debug(sb, "Using PPC superblock location"); } else { befs_debug(sb, "Using x86 superblock location"); diff --git a/fs/befs/super.c b/fs/befs/super.c index 8c3401ff6d6..41f2b4d0093 100644 --- a/fs/befs/super.c +++ b/fs/befs/super.c @@ -26,10 +26,10 @@ befs_load_sb(struct super_block *sb, befs_super_block * disk_sb) befs_sb_info *befs_sb = BEFS_SB(sb); /* Check the byte order of the filesystem */ - if (le32_to_cpu(disk_sb->fs_byte_order) == BEFS_BYTEORDER_NATIVE) + if (disk_sb->fs_byte_order == BEFS_BYTEORDER_NATIVE_LE) befs_sb->byte_order = BEFS_BYTESEX_LE; - else if (be32_to_cpu(disk_sb->fs_byte_order) == BEFS_BYTEORDER_NATIVE) - befs_sb->byte_order = BEFS_BYTESEX_BE; + else if (disk_sb->fs_byte_order == BEFS_BYTEORDER_NATIVE_BE) + befs_sb->byte_order = BEFS_BYTESEX_BE; befs_sb->magic1 = fs32_to_cpu(sb, disk_sb->magic1); befs_sb->magic2 = fs32_to_cpu(sb, disk_sb->magic2); diff --git a/fs/bfs/dir.c b/fs/bfs/dir.c index ed8feb052df..4dd1b623f93 100644 --- a/fs/bfs/dir.c +++ b/fs/bfs/dir.c @@ -80,6 +80,7 @@ const struct file_operations bfs_dir_operations = { .read = generic_read_dir, .readdir = bfs_readdir, .fsync = file_fsync, + .llseek = generic_file_llseek, }; extern void dump_imap(const char *, struct super_block *); @@ -105,8 +106,8 @@ static int bfs_create(struct inode *dir, struct dentry *dentry, int mode, } set_bit(ino, info->si_imap); info->si_freei--; - inode->i_uid = current->fsuid; - inode->i_gid = (dir->i_mode & S_ISGID) ? dir->i_gid : current->fsgid; + inode->i_uid = current_fsuid(); + inode->i_gid = (dir->i_mode & S_ISGID) ? dir->i_gid : current_fsgid(); inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME_SEC; inode->i_blocks = 0; inode->i_op = &bfs_file_inops; diff --git a/fs/binfmt_aout.c b/fs/binfmt_aout.c index 204cfd1d767..f1f3f4192a6 100644 --- a/fs/binfmt_aout.c +++ b/fs/binfmt_aout.c @@ -320,7 +320,7 @@ static int load_aout_binary(struct linux_binprm * bprm, struct pt_regs * regs) current->mm->free_area_cache = current->mm->mmap_base; current->mm->cached_hole_size = 0; - compute_creds(bprm); + install_exec_creds(bprm); current->flags &= ~PF_FORKNOEXEC; #ifdef __sparc__ if (N_MAGIC(ex) == NMAGIC) { diff --git a/fs/binfmt_elf.c b/fs/binfmt_elf.c index 655ed8d30a8..c41fa2af767 100644 --- a/fs/binfmt_elf.c +++ b/fs/binfmt_elf.c @@ -157,7 +157,7 @@ create_elf_tables(struct linux_binprm *bprm, struct elfhdr *exec, int items; elf_addr_t *elf_info; int ei_index = 0; - struct task_struct *tsk = current; + const struct cred *cred = current_cred(); struct vm_area_struct *vma; /* @@ -223,10 +223,10 @@ create_elf_tables(struct linux_binprm *bprm, struct elfhdr *exec, NEW_AUX_ENT(AT_BASE, interp_load_addr); NEW_AUX_ENT(AT_FLAGS, 0); NEW_AUX_ENT(AT_ENTRY, exec->e_entry); - NEW_AUX_ENT(AT_UID, tsk->uid); - NEW_AUX_ENT(AT_EUID, tsk->euid); - NEW_AUX_ENT(AT_GID, tsk->gid); - NEW_AUX_ENT(AT_EGID, tsk->egid); + NEW_AUX_ENT(AT_UID, cred->uid); + NEW_AUX_ENT(AT_EUID, cred->euid); + NEW_AUX_ENT(AT_GID, cred->gid); + NEW_AUX_ENT(AT_EGID, cred->egid); NEW_AUX_ENT(AT_SECURE, security_bprm_secureexec(bprm)); NEW_AUX_ENT(AT_EXECFN, bprm->exec); if (k_platform) { @@ -683,7 +683,7 @@ static int load_elf_binary(struct linux_binprm *bprm, struct pt_regs *regs) * switch really is going to happen - do this in * flush_thread(). - akpm */ - SET_PERSONALITY(loc->elf_ex, 0); + SET_PERSONALITY(loc->elf_ex); interpreter = open_exec(elf_interpreter); retval = PTR_ERR(interpreter); @@ -734,7 +734,7 @@ static int load_elf_binary(struct linux_binprm *bprm, struct pt_regs *regs) goto out_free_dentry; } else { /* Executables without an interpreter also need a personality */ - SET_PERSONALITY(loc->elf_ex, 0); + SET_PERSONALITY(loc->elf_ex); } /* Flush all traces of the currently running executable */ @@ -748,7 +748,7 @@ static int load_elf_binary(struct linux_binprm *bprm, struct pt_regs *regs) /* Do this immediately, since STACK_TOP as used in setup_arg_pages may depend on the personality. */ - SET_PERSONALITY(loc->elf_ex, 0); + SET_PERSONALITY(loc->elf_ex); if (elf_read_implies_exec(loc->elf_ex, executable_stack)) current->personality |= READ_IMPLIES_EXEC; @@ -949,14 +949,14 @@ static int load_elf_binary(struct linux_binprm *bprm, struct pt_regs *regs) set_binfmt(&elf_format); #ifdef ARCH_HAS_SETUP_ADDITIONAL_PAGES - retval = arch_setup_additional_pages(bprm, executable_stack); + retval = arch_setup_additional_pages(bprm, !!elf_interpreter); if (retval < 0) { send_sig(SIGKILL, current, 0); goto out; } #endif /* ARCH_HAS_SETUP_ADDITIONAL_PAGES */ - compute_creds(bprm); + install_exec_creds(bprm); current->flags &= ~PF_FORKNOEXEC; retval = create_elf_tables(bprm, &loc->elf_ex, load_addr, interp_load_addr); @@ -1156,16 +1156,24 @@ static int dump_seek(struct file *file, loff_t off) static unsigned long vma_dump_size(struct vm_area_struct *vma, unsigned long mm_flags) { +#define FILTER(type) (mm_flags & (1UL << MMF_DUMP_##type)) + /* The vma can be set up to tell us the answer directly. */ if (vma->vm_flags & VM_ALWAYSDUMP) goto whole; + /* Hugetlb memory check */ + if (vma->vm_flags & VM_HUGETLB) { + if ((vma->vm_flags & VM_SHARED) && FILTER(HUGETLB_SHARED)) + goto whole; + if (!(vma->vm_flags & VM_SHARED) && FILTER(HUGETLB_PRIVATE)) + goto whole; + } + /* Do not dump I/O mapped devices or special mappings */ if (vma->vm_flags & (VM_IO | VM_RESERVED)) return 0; -#define FILTER(type) (mm_flags & (1UL << MMF_DUMP_##type)) - /* By default, dump shared memory if mapped from an anonymous file. */ if (vma->vm_flags & VM_SHARED) { if (vma->vm_file->f_path.dentry->d_inode->i_nlink == 0 ? @@ -1333,20 +1341,15 @@ static void fill_prstatus(struct elf_prstatus *prstatus, prstatus->pr_pgrp = task_pgrp_vnr(p); prstatus->pr_sid = task_session_vnr(p); if (thread_group_leader(p)) { + struct task_cputime cputime; + /* - * This is the record for the group leader. Add in the - * cumulative times of previous dead threads. This total - * won't include the time of each live thread whose state - * is included in the core dump. The final total reported - * to our parent process when it calls wait4 will include - * those sums as well as the little bit more time it takes - * this and each other thread to finish dying after the - * core dump synchronization phase. + * This is the record for the group leader. It shows the + * group-wide total, not its individual thread total. */ - cputime_to_timeval(cputime_add(p->utime, p->signal->utime), - &prstatus->pr_utime); - cputime_to_timeval(cputime_add(p->stime, p->signal->stime), - &prstatus->pr_stime); + thread_group_cputime(p, &cputime); + cputime_to_timeval(cputime.utime, &prstatus->pr_utime); + cputime_to_timeval(cputime.stime, &prstatus->pr_stime); } else { cputime_to_timeval(p->utime, &prstatus->pr_utime); cputime_to_timeval(p->stime, &prstatus->pr_stime); @@ -1358,6 +1361,7 @@ static void fill_prstatus(struct elf_prstatus *prstatus, static int fill_psinfo(struct elf_prpsinfo *psinfo, struct task_struct *p, struct mm_struct *mm) { + const struct cred *cred; unsigned int i, len; /* first copy the parameters from user space */ @@ -1385,8 +1389,11 @@ static int fill_psinfo(struct elf_prpsinfo *psinfo, struct task_struct *p, psinfo->pr_zomb = psinfo->pr_sname == 'Z'; psinfo->pr_nice = task_nice(p); psinfo->pr_flag = p->flags; - SET_UID(psinfo->pr_uid, p->uid); - SET_GID(psinfo->pr_gid, p->gid); + rcu_read_lock(); + cred = __task_cred(p); + SET_UID(psinfo->pr_uid, cred->uid); + SET_GID(psinfo->pr_gid, cred->gid); + rcu_read_unlock(); strncpy(psinfo->pr_fname, p->comm, sizeof(psinfo->pr_fname)); return 0; diff --git a/fs/binfmt_elf_fdpic.c b/fs/binfmt_elf_fdpic.c index 80c1f952ef7..aa5b43205e3 100644 --- a/fs/binfmt_elf_fdpic.c +++ b/fs/binfmt_elf_fdpic.c @@ -25,6 +25,7 @@ #include <linux/fcntl.h> #include <linux/slab.h> #include <linux/pagemap.h> +#include <linux/security.h> #include <linux/highmem.h> #include <linux/highuid.h> #include <linux/personality.h> @@ -403,7 +404,7 @@ static int load_elf_fdpic_binary(struct linux_binprm *bprm, current->mm->start_stack = current->mm->start_brk + stack_size; #endif - compute_creds(bprm); + install_exec_creds(bprm); current->flags &= ~PF_FORKNOEXEC; if (create_elf_fdpic_tables(bprm, current->mm, &exec_params, &interp_params) < 0) @@ -455,26 +456,42 @@ error_kill: } /*****************************************************************************/ + +#ifndef ELF_BASE_PLATFORM +/* + * AT_BASE_PLATFORM indicates the "real" hardware/microarchitecture. + * If the arch defines ELF_BASE_PLATFORM (in asm/elf.h), the value + * will be copied to the user stack in the same manner as AT_PLATFORM. + */ +#define ELF_BASE_PLATFORM NULL +#endif + /* - * present useful information to the program + * present useful information to the program by shovelling it onto the new + * process's stack */ static int create_elf_fdpic_tables(struct linux_binprm *bprm, struct mm_struct *mm, struct elf_fdpic_params *exec_params, struct elf_fdpic_params *interp_params) { + const struct cred *cred = current_cred(); unsigned long sp, csp, nitems; elf_caddr_t __user *argv, *envp; size_t platform_len = 0, len; - char *k_platform; - char __user *u_platform, *p; + char *k_platform, *k_base_platform; + char __user *u_platform, *u_base_platform, *p; long hwcap; int loop; int nr; /* reset for each csp adjustment */ - /* we're going to shovel a whole load of stuff onto the stack */ #ifdef CONFIG_MMU - sp = bprm->p; + /* In some cases (e.g. Hyper-Threading), we want to avoid L1 evictions + * by the processes running on the same package. One thing we can do is + * to shuffle the initial stack for them, so we give the architecture + * an opportunity to do so here. + */ + sp = arch_align_stack(bprm->p); #else sp = mm->start_stack; @@ -483,11 +500,14 @@ static int create_elf_fdpic_tables(struct linux_binprm *bprm, return -EFAULT; #endif - /* get hold of platform and hardware capabilities masks for the machine - * we are running on. In some cases (Sparc), this info is impossible - * to get, in others (i386) it is merely difficult. - */ hwcap = ELF_HWCAP; + + /* + * If this architecture has a platform capability string, copy it + * to userspace. In some cases (Sparc), this info is impossible + * for userspace to get any other way, in others (i386) it is + * merely difficult. + */ k_platform = ELF_PLATFORM; u_platform = NULL; @@ -499,19 +519,20 @@ static int create_elf_fdpic_tables(struct linux_binprm *bprm, return -EFAULT; } -#if defined(__i386__) && defined(CONFIG_SMP) - /* in some cases (e.g. Hyper-Threading), we want to avoid L1 evictions - * by the processes running on the same package. One thing we can do is - * to shuffle the initial stack for them. - * - * the conditionals here are unneeded, but kept in to make the code - * behaviour the same as pre change unless we have hyperthreaded - * processors. This keeps Mr Marcelo Person happier but should be - * removed for 2.5 + /* + * If this architecture has a "base" platform capability + * string, copy it to userspace. */ - if (smp_num_siblings > 1) - sp = sp - ((current->pid % 64) << 7); -#endif + k_base_platform = ELF_BASE_PLATFORM; + u_base_platform = NULL; + + if (k_base_platform) { + platform_len = strlen(k_base_platform) + 1; + sp -= platform_len; + u_base_platform = (char __user *) sp; + if (__copy_to_user(u_base_platform, k_base_platform, platform_len) != 0) + return -EFAULT; + } sp &= ~7UL; @@ -541,9 +562,13 @@ static int create_elf_fdpic_tables(struct linux_binprm *bprm, } /* force 16 byte _final_ alignment here for generality */ -#define DLINFO_ITEMS 13 +#define DLINFO_ITEMS 15 - nitems = 1 + DLINFO_ITEMS + (k_platform ? 1 : 0) + AT_VECTOR_SIZE_ARCH; + nitems = 1 + DLINFO_ITEMS + (k_platform ? 1 : 0) + + (k_base_platform ? 1 : 0) + AT_VECTOR_SIZE_ARCH; + + if (bprm->interp_flags & BINPRM_FLAGS_EXECFD) + nitems++; csp = sp; sp -= nitems * 2 * sizeof(unsigned long); @@ -575,6 +600,19 @@ static int create_elf_fdpic_tables(struct linux_binprm *bprm, (elf_addr_t) (unsigned long) u_platform); } + if (k_base_platform) { + nr = 0; + csp -= 2 * sizeof(unsigned long); + NEW_AUX_ENT(AT_BASE_PLATFORM, + (elf_addr_t) (unsigned long) u_base_platform); + } + + if (bprm->interp_flags & BINPRM_FLAGS_EXECFD) { + nr = 0; + csp -= 2 * sizeof(unsigned long); + NEW_AUX_ENT(AT_EXECFD, bprm->interp_data); + } + nr = 0; csp -= DLINFO_ITEMS * 2 * sizeof(unsigned long); NEW_AUX_ENT(AT_HWCAP, hwcap); @@ -586,10 +624,12 @@ static int create_elf_fdpic_tables(struct linux_binprm *bprm, NEW_AUX_ENT(AT_BASE, interp_params->elfhdr_addr); NEW_AUX_ENT(AT_FLAGS, 0); NEW_AUX_ENT(AT_ENTRY, exec_params->entry_addr); - NEW_AUX_ENT(AT_UID, (elf_addr_t) current->uid); - NEW_AUX_ENT(AT_EUID, (elf_addr_t) current->euid); - NEW_AUX_ENT(AT_GID, (elf_addr_t) current->gid); - NEW_AUX_ENT(AT_EGID, (elf_addr_t) current->egid); + NEW_AUX_ENT(AT_UID, (elf_addr_t) cred->uid); + NEW_AUX_ENT(AT_EUID, (elf_addr_t) cred->euid); + NEW_AUX_ENT(AT_GID, (elf_addr_t) cred->gid); + NEW_AUX_ENT(AT_EGID, (elf_addr_t) cred->egid); + NEW_AUX_ENT(AT_SECURE, security_bprm_secureexec(bprm)); + NEW_AUX_ENT(AT_EXECFN, bprm->exec); #ifdef ARCH_DLINFO nr = 0; @@ -1351,20 +1391,15 @@ static void fill_prstatus(struct elf_prstatus *prstatus, prstatus->pr_pgrp = task_pgrp_vnr(p); prstatus->pr_sid = task_session_vnr(p); if (thread_group_leader(p)) { + struct task_cputime cputime; + /* - * This is the record for the group leader. Add in the - * cumulative times of previous dead threads. This total - * won't include the time of each live thread whose state - * is included in the core dump. The final total reported - * to our parent process when it calls wait4 will include - * those sums as well as the little bit more time it takes - * this and each other thread to finish dying after the - * core dump synchronization phase. + * This is the record for the group leader. It shows the + * group-wide total, not its individual thread total. */ - cputime_to_timeval(cputime_add(p->utime, p->signal->utime), - &prstatus->pr_utime); - cputime_to_timeval(cputime_add(p->stime, p->signal->stime), - &prstatus->pr_stime); + thread_group_cputime(p, &cputime); + cputime_to_timeval(cputime.utime, &prstatus->pr_utime); + cputime_to_timeval(cputime.stime, &prstatus->pr_stime); } else { cputime_to_timeval(p->utime, &prstatus->pr_utime); cputime_to_timeval(p->stime, &prstatus->pr_stime); @@ -1379,6 +1414,7 @@ static void fill_prstatus(struct elf_prstatus *prstatus, static int fill_psinfo(struct elf_prpsinfo *psinfo, struct task_struct *p, struct mm_struct *mm) { + const struct cred *cred; unsigned int i, len; /* first copy the parameters from user space */ @@ -1406,8 +1442,11 @@ static int fill_psinfo(struct elf_prpsinfo *psinfo, struct task_struct *p, psinfo->pr_zomb = psinfo->pr_sname == 'Z'; psinfo->pr_nice = task_nice(p); psinfo->pr_flag = p->flags; - SET_UID(psinfo->pr_uid, p->uid); - SET_GID(psinfo->pr_gid, p->gid); + rcu_read_lock(); + cred = __task_cred(p); + SET_UID(psinfo->pr_uid, cred->uid); + SET_GID(psinfo->pr_gid, cred->gid); + rcu_read_unlock(); strncpy(psinfo->pr_fname, p->comm, sizeof(psinfo->pr_fname)); return 0; diff --git a/fs/binfmt_em86.c b/fs/binfmt_em86.c index f9c88d0c8ce..32fb00b52cd 100644 --- a/fs/binfmt_em86.c +++ b/fs/binfmt_em86.c @@ -43,7 +43,7 @@ static int load_em86(struct linux_binprm *bprm,struct pt_regs *regs) return -ENOEXEC; } - bprm->sh_bang = 1; /* Well, the bang-shell is implicit... */ + bprm->recursion_depth++; /* Well, the bang-shell is implicit... */ allow_write_access(bprm->file); fput(bprm->file); bprm->file = NULL; diff --git a/fs/binfmt_flat.c b/fs/binfmt_flat.c index dfc0197905c..7bbd5c6b372 100644 --- a/fs/binfmt_flat.c +++ b/fs/binfmt_flat.c @@ -229,13 +229,13 @@ static int decompress_exec( ret = 10; if (buf[3] & EXTRA_FIELD) { ret += 2 + buf[10] + (buf[11] << 8); - if (unlikely(LBUFSIZE == ret)) { + if (unlikely(LBUFSIZE <= ret)) { DBG_FLT("binfmt_flat: buffer overflow (EXTRA)?\n"); goto out_free_buf; } } if (buf[3] & ORIG_NAME) { - for (; ret < LBUFSIZE && (buf[ret] != 0); ret++) + while (ret < LBUFSIZE && buf[ret++] != 0) ; if (unlikely(LBUFSIZE == ret)) { DBG_FLT("binfmt_flat: buffer overflow (ORIG_NAME)?\n"); @@ -243,7 +243,7 @@ static int decompress_exec( } } if (buf[3] & COMMENT) { - for (; ret < LBUFSIZE && (buf[ret] != 0); ret++) + while (ret < LBUFSIZE && buf[ret++] != 0) ; if (unlikely(LBUFSIZE == ret)) { DBG_FLT("binfmt_flat: buffer overflow (COMMENT)?\n"); @@ -880,7 +880,7 @@ static int load_flat_binary(struct linux_binprm * bprm, struct pt_regs * regs) (libinfo.lib_list[j].loaded)? libinfo.lib_list[j].start_data:UNLOADED_LIB; - compute_creds(bprm); + install_exec_creds(bprm); current->flags &= ~PF_FORKNOEXEC; set_binfmt(&flat_format); diff --git a/fs/binfmt_misc.c b/fs/binfmt_misc.c index 8d7e88e02e0..f2744ab4e5b 100644 --- a/fs/binfmt_misc.c +++ b/fs/binfmt_misc.c @@ -117,7 +117,7 @@ static int load_misc_binary(struct linux_binprm *bprm, struct pt_regs *regs) goto _ret; retval = -ENOEXEC; - if (bprm->misc_bang) + if (bprm->recursion_depth > BINPRM_MAX_RECURSION) goto _ret; /* to keep locking time low, we copy the interpreter string */ @@ -197,7 +197,7 @@ static int load_misc_binary(struct linux_binprm *bprm, struct pt_regs *regs) if (retval < 0) goto _error; - bprm->misc_bang = 1; + bprm->recursion_depth++; retval = search_binary_handler (bprm, regs); if (retval < 0) diff --git a/fs/binfmt_script.c b/fs/binfmt_script.c index 9e3963f7ebf..08343505e18 100644 --- a/fs/binfmt_script.c +++ b/fs/binfmt_script.c @@ -22,14 +22,15 @@ static int load_script(struct linux_binprm *bprm,struct pt_regs *regs) char interp[BINPRM_BUF_SIZE]; int retval; - if ((bprm->buf[0] != '#') || (bprm->buf[1] != '!') || (bprm->sh_bang)) + if ((bprm->buf[0] != '#') || (bprm->buf[1] != '!') || + (bprm->recursion_depth > BINPRM_MAX_RECURSION)) return -ENOEXEC; /* * This section does the #! interpretation. * Sorta complicated, but hopefully it will work. -TYT */ - bprm->sh_bang = 1; + bprm->recursion_depth++; allow_write_access(bprm->file); fput(bprm->file); bprm->file = NULL; diff --git a/fs/binfmt_som.c b/fs/binfmt_som.c index 68be580ba28..08644a61616 100644 --- a/fs/binfmt_som.c +++ b/fs/binfmt_som.c @@ -255,7 +255,7 @@ load_som_binary(struct linux_binprm * bprm, struct pt_regs * regs) kfree(hpuxhdr); set_binfmt(&som_format); - compute_creds(bprm); + install_exec_creds(bprm); setup_arg_pages(bprm, STACK_TOP, EXSTACK_DEFAULT); create_som_tables(bprm); @@ -306,3 +306,5 @@ static void __exit exit_som_binfmt(void) core_initcall(init_som_binfmt); module_exit(exit_som_binfmt); + +MODULE_LICENSE("GPL"); diff --git a/fs/bio-integrity.c b/fs/bio-integrity.c index 19caf7c962a..77ebc3c263d 100644 --- a/fs/bio-integrity.c +++ b/fs/bio-integrity.c @@ -111,7 +111,7 @@ void bio_integrity_free(struct bio *bio, struct bio_set *bs) && bip->bip_buf != NULL) kfree(bip->bip_buf); - mempool_free(bip->bip_vec, bs->bvec_pools[bip->bip_pool]); + bvec_free_bs(bs, bip->bip_vec, bip->bip_pool); mempool_free(bip, bs->bio_integrity_pool); bio->bi_integrity = NULL; @@ -26,9 +26,16 @@ #include <linux/mempool.h> #include <linux/workqueue.h> #include <linux/blktrace_api.h> +#include <trace/block.h> #include <scsi/sg.h> /* for struct sg_iovec */ -static struct kmem_cache *bio_slab __read_mostly; +DEFINE_TRACE(block_split); + +/* + * Test patch to inline a certain number of bi_io_vec's inside the bio + * itself, to shrink a bio data allocation from two mempool calls to one + */ +#define BIO_INLINE_VECS 4 static mempool_t *bio_split_pool __read_mostly; @@ -37,9 +44,8 @@ static mempool_t *bio_split_pool __read_mostly; * break badly! cannot be bigger than what you can fit into an * unsigned short */ - #define BV(x) { .nr_vecs = x, .name = "biovec-"__stringify(x) } -static struct biovec_slab bvec_slabs[BIOVEC_NR_POOLS] __read_mostly = { +struct biovec_slab bvec_slabs[BIOVEC_NR_POOLS] __read_mostly = { BV(1), BV(4), BV(16), BV(64), BV(128), BV(BIO_MAX_PAGES), }; #undef BV @@ -50,12 +56,121 @@ static struct biovec_slab bvec_slabs[BIOVEC_NR_POOLS] __read_mostly = { */ struct bio_set *fs_bio_set; +/* + * Our slab pool management + */ +struct bio_slab { + struct kmem_cache *slab; + unsigned int slab_ref; + unsigned int slab_size; + char name[8]; +}; +static DEFINE_MUTEX(bio_slab_lock); +static struct bio_slab *bio_slabs; +static unsigned int bio_slab_nr, bio_slab_max; + +static struct kmem_cache *bio_find_or_create_slab(unsigned int extra_size) +{ + unsigned int sz = sizeof(struct bio) + extra_size; + struct kmem_cache *slab = NULL; + struct bio_slab *bslab; + unsigned int i, entry = -1; + + mutex_lock(&bio_slab_lock); + + i = 0; + while (i < bio_slab_nr) { + struct bio_slab *bslab = &bio_slabs[i]; + + if (!bslab->slab && entry == -1) + entry = i; + else if (bslab->slab_size == sz) { + slab = bslab->slab; + bslab->slab_ref++; + break; + } + i++; + } + + if (slab) + goto out_unlock; + + if (bio_slab_nr == bio_slab_max && entry == -1) { + bio_slab_max <<= 1; + bio_slabs = krealloc(bio_slabs, + bio_slab_max * sizeof(struct bio_slab), + GFP_KERNEL); + if (!bio_slabs) + goto out_unlock; + } + if (entry == -1) + entry = bio_slab_nr++; + + bslab = &bio_slabs[entry]; + + snprintf(bslab->name, sizeof(bslab->name), "bio-%d", entry); + slab = kmem_cache_create(bslab->name, sz, 0, SLAB_HWCACHE_ALIGN, NULL); + if (!slab) + goto out_unlock; + + printk("bio: create slab <%s> at %d\n", bslab->name, entry); + bslab->slab = slab; + bslab->slab_ref = 1; + bslab->slab_size = sz; +out_unlock: + mutex_unlock(&bio_slab_lock); + return slab; +} + +static void bio_put_slab(struct bio_set *bs) +{ + struct bio_slab *bslab = NULL; + unsigned int i; + + mutex_lock(&bio_slab_lock); + + for (i = 0; i < bio_slab_nr; i++) { + if (bs->bio_slab == bio_slabs[i].slab) { + bslab = &bio_slabs[i]; + break; + } + } + + if (WARN(!bslab, KERN_ERR "bio: unable to find slab!\n")) + goto out; + + WARN_ON(!bslab->slab_ref); + + if (--bslab->slab_ref) + goto out; + + kmem_cache_destroy(bslab->slab); + bslab->slab = NULL; + +out: + mutex_unlock(&bio_slab_lock); +} + unsigned int bvec_nr_vecs(unsigned short idx) { return bvec_slabs[idx].nr_vecs; } -struct bio_vec *bvec_alloc_bs(gfp_t gfp_mask, int nr, unsigned long *idx, struct bio_set *bs) +void bvec_free_bs(struct bio_set *bs, struct bio_vec *bv, unsigned int idx) +{ + BIO_BUG_ON(idx >= BIOVEC_NR_POOLS); + + if (idx == BIOVEC_MAX_IDX) + mempool_free(bv, bs->bvec_pool); + else { + struct biovec_slab *bvs = bvec_slabs + idx; + + kmem_cache_free(bvs->slab, bv); + } +} + +struct bio_vec *bvec_alloc_bs(gfp_t gfp_mask, int nr, unsigned long *idx, + struct bio_set *bs) { struct bio_vec *bvl; @@ -64,60 +179,85 @@ struct bio_vec *bvec_alloc_bs(gfp_t gfp_mask, int nr, unsigned long *idx, struct * If not, this is a bio_kmalloc() allocation and just do a * kzalloc() for the exact number of vecs right away. */ - if (bs) { + if (!bs) + bvl = kmalloc(nr * sizeof(struct bio_vec), gfp_mask); + + /* + * see comment near bvec_array define! + */ + switch (nr) { + case 1: + *idx = 0; + break; + case 2 ... 4: + *idx = 1; + break; + case 5 ... 16: + *idx = 2; + break; + case 17 ... 64: + *idx = 3; + break; + case 65 ... 128: + *idx = 4; + break; + case 129 ... BIO_MAX_PAGES: + *idx = 5; + break; + default: + return NULL; + } + + /* + * idx now points to the pool we want to allocate from. only the + * 1-vec entry pool is mempool backed. + */ + if (*idx == BIOVEC_MAX_IDX) { +fallback: + bvl = mempool_alloc(bs->bvec_pool, gfp_mask); + } else { + struct biovec_slab *bvs = bvec_slabs + *idx; + gfp_t __gfp_mask = gfp_mask & ~(__GFP_WAIT | __GFP_IO); + /* - * see comment near bvec_array define! + * Make this allocation restricted and don't dump info on + * allocation failures, since we'll fallback to the mempool + * in case of failure. */ - switch (nr) { - case 1: - *idx = 0; - break; - case 2 ... 4: - *idx = 1; - break; - case 5 ... 16: - *idx = 2; - break; - case 17 ... 64: - *idx = 3; - break; - case 65 ... 128: - *idx = 4; - break; - case 129 ... BIO_MAX_PAGES: - *idx = 5; - break; - default: - return NULL; - } + __gfp_mask |= __GFP_NOMEMALLOC | __GFP_NORETRY | __GFP_NOWARN; /* - * idx now points to the pool we want to allocate from + * Try a slab allocation. If this fails and __GFP_WAIT + * is set, retry with the 1-entry mempool */ - bvl = mempool_alloc(bs->bvec_pools[*idx], gfp_mask); - if (bvl) - memset(bvl, 0, - bvec_nr_vecs(*idx) * sizeof(struct bio_vec)); - } else - bvl = kzalloc(nr * sizeof(struct bio_vec), gfp_mask); + bvl = kmem_cache_alloc(bvs->slab, __gfp_mask); + if (unlikely(!bvl && (gfp_mask & __GFP_WAIT))) { + *idx = BIOVEC_MAX_IDX; + goto fallback; + } + } return bvl; } -void bio_free(struct bio *bio, struct bio_set *bio_set) +void bio_free(struct bio *bio, struct bio_set *bs) { - if (bio->bi_io_vec) { - const int pool_idx = BIO_POOL_IDX(bio); + void *p; - BIO_BUG_ON(pool_idx >= BIOVEC_NR_POOLS); - - mempool_free(bio->bi_io_vec, bio_set->bvec_pools[pool_idx]); - } + if (bio_has_allocated_vec(bio)) + bvec_free_bs(bs, bio->bi_io_vec, BIO_POOL_IDX(bio)); if (bio_integrity(bio)) - bio_integrity_free(bio, bio_set); + bio_integrity_free(bio, bs); + + /* + * If we have front padding, adjust the bio pointer before freeing + */ + p = bio; + if (bs->front_pad) + p -= bs->front_pad; - mempool_free(bio, bio_set->bio_pool); + mempool_free(p, bs->bio_pool); } /* @@ -130,7 +270,8 @@ static void bio_fs_destructor(struct bio *bio) static void bio_kmalloc_destructor(struct bio *bio) { - kfree(bio->bi_io_vec); + if (bio_has_allocated_vec(bio)) + kfree(bio->bi_io_vec); kfree(bio); } @@ -154,16 +295,20 @@ void bio_init(struct bio *bio) * for a &struct bio to become free. If a %NULL @bs is passed in, we will * fall back to just using @kmalloc to allocate the required memory. * - * allocate bio and iovecs from the memory pools specified by the - * bio_set structure, or @kmalloc if none given. + * Note that the caller must set ->bi_destructor on succesful return + * of a bio, to do the appropriate freeing of the bio once the reference + * count drops to zero. **/ struct bio *bio_alloc_bioset(gfp_t gfp_mask, int nr_iovecs, struct bio_set *bs) { - struct bio *bio; + struct bio *bio = NULL; + + if (bs) { + void *p = mempool_alloc(bs->bio_pool, gfp_mask); - if (bs) - bio = mempool_alloc(bs->bio_pool, gfp_mask); - else + if (p) + bio = p + bs->front_pad; + } else bio = kmalloc(sizeof(*bio), gfp_mask); if (likely(bio)) { @@ -173,7 +318,15 @@ struct bio *bio_alloc_bioset(gfp_t gfp_mask, int nr_iovecs, struct bio_set *bs) if (likely(nr_iovecs)) { unsigned long uninitialized_var(idx); - bvl = bvec_alloc_bs(gfp_mask, nr_iovecs, &idx, bs); + if (nr_iovecs <= BIO_INLINE_VECS) { + idx = 0; + bvl = bio->bi_inline_vecs; + nr_iovecs = BIO_INLINE_VECS; + } else { + bvl = bvec_alloc_bs(gfp_mask, nr_iovecs, &idx, + bs); + nr_iovecs = bvec_nr_vecs(idx); + } if (unlikely(!bvl)) { if (bs) mempool_free(bio, bs->bio_pool); @@ -183,7 +336,7 @@ struct bio *bio_alloc_bioset(gfp_t gfp_mask, int nr_iovecs, struct bio_set *bs) goto out; } bio->bi_flags |= idx << BIO_POOL_OFFSET; - bio->bi_max_vecs = bvec_nr_vecs(idx); + bio->bi_max_vecs = nr_iovecs; } bio->bi_io_vec = bvl; } @@ -1263,7 +1416,7 @@ struct bio_pair *bio_split(struct bio *bi, int first_sectors) if (!bp) return bp; - blk_add_trace_pdu_int(bdev_get_queue(bi->bi_bdev), BLK_TA_SPLIT, bi, + trace_block_split(bdev_get_queue(bi->bi_bdev), bi, bi->bi_sector + first_sectors); BUG_ON(bi->bi_vcnt != 1); @@ -1343,30 +1496,18 @@ EXPORT_SYMBOL(bio_sector_offset); */ static int biovec_create_pools(struct bio_set *bs, int pool_entries) { - int i; + struct biovec_slab *bp = bvec_slabs + BIOVEC_MAX_IDX; - for (i = 0; i < BIOVEC_NR_POOLS; i++) { - struct biovec_slab *bp = bvec_slabs + i; - mempool_t **bvp = bs->bvec_pools + i; + bs->bvec_pool = mempool_create_slab_pool(pool_entries, bp->slab); + if (!bs->bvec_pool) + return -ENOMEM; - *bvp = mempool_create_slab_pool(pool_entries, bp->slab); - if (!*bvp) - return -ENOMEM; - } return 0; } static void biovec_free_pools(struct bio_set *bs) { - int i; - - for (i = 0; i < BIOVEC_NR_POOLS; i++) { - mempool_t *bvp = bs->bvec_pools[i]; - - if (bvp) - mempool_destroy(bvp); - } - + mempool_destroy(bs->bvec_pool); } void bioset_free(struct bio_set *bs) @@ -1376,25 +1517,49 @@ void bioset_free(struct bio_set *bs) bioset_integrity_free(bs); biovec_free_pools(bs); + bio_put_slab(bs); kfree(bs); } -struct bio_set *bioset_create(int bio_pool_size, int bvec_pool_size) +/** + * bioset_create - Create a bio_set + * @pool_size: Number of bio and bio_vecs to cache in the mempool + * @front_pad: Number of bytes to allocate in front of the returned bio + * + * Description: + * Set up a bio_set to be used with @bio_alloc_bioset. Allows the caller + * to ask for a number of bytes to be allocated in front of the bio. + * Front pad allocation is useful for embedding the bio inside + * another structure, to avoid allocating extra data to go with the bio. + * Note that the bio must be embedded at the END of that structure always, + * or things will break badly. + */ +struct bio_set *bioset_create(unsigned int pool_size, unsigned int front_pad) { - struct bio_set *bs = kzalloc(sizeof(*bs), GFP_KERNEL); + unsigned int back_pad = BIO_INLINE_VECS * sizeof(struct bio_vec); + struct bio_set *bs; + bs = kzalloc(sizeof(*bs), GFP_KERNEL); if (!bs) return NULL; - bs->bio_pool = mempool_create_slab_pool(bio_pool_size, bio_slab); + bs->front_pad = front_pad; + + bs->bio_slab = bio_find_or_create_slab(front_pad + back_pad); + if (!bs->bio_slab) { + kfree(bs); + return NULL; + } + + bs->bio_pool = mempool_create_slab_pool(pool_size, bs->bio_slab); if (!bs->bio_pool) goto bad; - if (bioset_integrity_create(bs, bio_pool_size)) + if (bioset_integrity_create(bs, pool_size)) goto bad; - if (!biovec_create_pools(bs, bvec_pool_size)) + if (!biovec_create_pools(bs, pool_size)) return bs; bad: @@ -1418,12 +1583,16 @@ static void __init biovec_init_slabs(void) static int __init init_bio(void) { - bio_slab = KMEM_CACHE(bio, SLAB_HWCACHE_ALIGN|SLAB_PANIC); + bio_slab_max = 2; + bio_slab_nr = 0; + bio_slabs = kzalloc(bio_slab_max * sizeof(struct bio_slab), GFP_KERNEL); + if (!bio_slabs) + panic("bio: can't allocate bios\n"); bio_integrity_init_slab(); biovec_init_slabs(); - fs_bio_set = bioset_create(BIO_POOL_SIZE, 2); + fs_bio_set = bioset_create(BIO_POOL_SIZE, 0); if (!fs_bio_set) panic("bio: can't allocate bios\n"); diff --git a/fs/block_dev.c b/fs/block_dev.c index d84f0469a01..99e0ae1a4c7 100644 --- a/fs/block_dev.c +++ b/fs/block_dev.c @@ -840,13 +840,12 @@ EXPORT_SYMBOL_GPL(bd_release_from_disk); * to be used for internal purposes. If you ever need it - reconsider * your API. */ -struct block_device *open_by_devnum(dev_t dev, unsigned mode) +struct block_device *open_by_devnum(dev_t dev, fmode_t mode) { struct block_device *bdev = bdget(dev); int err = -ENOMEM; - int flags = mode & FMODE_WRITE ? O_RDWR : O_RDONLY; if (bdev) - err = blkdev_get(bdev, mode, flags); + err = blkdev_get(bdev, mode); return err ? ERR_PTR(err) : bdev; } @@ -975,9 +974,7 @@ void bd_set_size(struct block_device *bdev, loff_t size) } EXPORT_SYMBOL(bd_set_size); -static int __blkdev_get(struct block_device *bdev, mode_t mode, unsigned flags, - int for_part); -static int __blkdev_put(struct block_device *bdev, int for_part); +static int __blkdev_put(struct block_device *bdev, fmode_t mode, int for_part); /* * bd_mutex locking: @@ -986,17 +983,16 @@ static int __blkdev_put(struct block_device *bdev, int for_part); * mutex_lock_nested(whole->bd_mutex, 1) */ -static int do_open(struct block_device *bdev, struct file *file, int for_part) +static int __blkdev_get(struct block_device *bdev, fmode_t mode, int for_part) { struct gendisk *disk; - struct hd_struct *part = NULL; int ret; int partno; int perm = 0; - if (file->f_mode & FMODE_READ) + if (mode & FMODE_READ) perm |= MAY_READ; - if (file->f_mode & FMODE_WRITE) + if (mode & FMODE_WRITE) perm |= MAY_WRITE; /* * hooks: /n/, see "layering violations". @@ -1007,27 +1003,27 @@ static int do_open(struct block_device *bdev, struct file *file, int for_part) return ret; } - ret = -ENXIO; - file->f_mapping = bdev->bd_inode->i_mapping; - lock_kernel(); + ret = -ENXIO; disk = get_gendisk(bdev->bd_dev, &partno); if (!disk) goto out_unlock_kernel; - part = disk_get_part(disk, partno); - if (!part) - goto out_unlock_kernel; mutex_lock_nested(&bdev->bd_mutex, for_part); if (!bdev->bd_openers) { bdev->bd_disk = disk; - bdev->bd_part = part; bdev->bd_contains = bdev; if (!partno) { struct backing_dev_info *bdi; + + ret = -ENXIO; + bdev->bd_part = disk_get_part(disk, partno); + if (!bdev->bd_part) + goto out_clear; + if (disk->fops->open) { - ret = disk->fops->open(bdev->bd_inode, file); + ret = disk->fops->open(bdev, mode); if (ret) goto out_clear; } @@ -1047,28 +1043,27 @@ static int do_open(struct block_device *bdev, struct file *file, int for_part) if (!whole) goto out_clear; BUG_ON(for_part); - ret = __blkdev_get(whole, file->f_mode, file->f_flags, 1); + ret = __blkdev_get(whole, mode, 1); if (ret) goto out_clear; bdev->bd_contains = whole; bdev->bd_inode->i_data.backing_dev_info = whole->bd_inode->i_data.backing_dev_info; + bdev->bd_part = disk_get_part(disk, partno); if (!(disk->flags & GENHD_FL_UP) || - !part || !part->nr_sects) { + !bdev->bd_part || !bdev->bd_part->nr_sects) { ret = -ENXIO; goto out_clear; } - bd_set_size(bdev, (loff_t)part->nr_sects << 9); + bd_set_size(bdev, (loff_t)bdev->bd_part->nr_sects << 9); } } else { - disk_put_part(part); put_disk(disk); module_put(disk->fops->owner); - part = NULL; disk = NULL; if (bdev->bd_contains == bdev) { if (bdev->bd_disk->fops->open) { - ret = bdev->bd_disk->fops->open(bdev->bd_inode, file); + ret = bdev->bd_disk->fops->open(bdev, mode); if (ret) goto out_unlock_bdev; } @@ -1084,18 +1079,18 @@ static int do_open(struct block_device *bdev, struct file *file, int for_part) return 0; out_clear: + disk_put_part(bdev->bd_part); bdev->bd_disk = NULL; bdev->bd_part = NULL; bdev->bd_inode->i_data.backing_dev_info = &default_backing_dev_info; if (bdev != bdev->bd_contains) - __blkdev_put(bdev->bd_contains, 1); + __blkdev_put(bdev->bd_contains, mode, 1); bdev->bd_contains = NULL; out_unlock_bdev: mutex_unlock(&bdev->bd_mutex); out_unlock_kernel: unlock_kernel(); - disk_put_part(part); if (disk) module_put(disk->fops->owner); put_disk(disk); @@ -1104,28 +1099,9 @@ static int do_open(struct block_device *bdev, struct file *file, int for_part) return ret; } -static int __blkdev_get(struct block_device *bdev, mode_t mode, unsigned flags, - int for_part) -{ - /* - * This crockload is due to bad choice of ->open() type. - * It will go away. - * For now, block device ->open() routine must _not_ - * examine anything in 'inode' argument except ->i_rdev. - */ - struct file fake_file = {}; - struct dentry fake_dentry = {}; - fake_file.f_mode = mode; - fake_file.f_flags = flags; - fake_file.f_path.dentry = &fake_dentry; - fake_dentry.d_inode = bdev->bd_inode; - - return do_open(bdev, &fake_file, for_part); -} - -int blkdev_get(struct block_device *bdev, mode_t mode, unsigned flags) +int blkdev_get(struct block_device *bdev, fmode_t mode) { - return __blkdev_get(bdev, mode, flags, 0); + return __blkdev_get(bdev, mode, 0); } EXPORT_SYMBOL(blkdev_get); @@ -1142,28 +1118,39 @@ static int blkdev_open(struct inode * inode, struct file * filp) */ filp->f_flags |= O_LARGEFILE; + if (filp->f_flags & O_NDELAY) + filp->f_mode |= FMODE_NDELAY; + if (filp->f_flags & O_EXCL) + filp->f_mode |= FMODE_EXCL; + if ((filp->f_flags & O_ACCMODE) == 3) + filp->f_mode |= FMODE_WRITE_IOCTL; + bdev = bd_acquire(inode); if (bdev == NULL) return -ENOMEM; - res = do_open(bdev, filp, 0); + filp->f_mapping = bdev->bd_inode->i_mapping; + + res = blkdev_get(bdev, filp->f_mode); if (res) return res; - if (!(filp->f_flags & O_EXCL) ) - return 0; + if (filp->f_mode & FMODE_EXCL) { + res = bd_claim(bdev, filp); + if (res) + goto out_blkdev_put; + } - if (!(res = bd_claim(bdev, filp))) - return 0; + return 0; - blkdev_put(bdev); + out_blkdev_put: + blkdev_put(bdev, filp->f_mode); return res; } -static int __blkdev_put(struct block_device *bdev, int for_part) +static int __blkdev_put(struct block_device *bdev, fmode_t mode, int for_part) { int ret = 0; - struct inode *bd_inode = bdev->bd_inode; struct gendisk *disk = bdev->bd_disk; struct block_device *victim = NULL; @@ -1178,7 +1165,7 @@ static int __blkdev_put(struct block_device *bdev, int for_part) } if (bdev->bd_contains == bdev) { if (disk->fops->release) - ret = disk->fops->release(bd_inode, NULL); + ret = disk->fops->release(disk, mode); } if (!bdev->bd_openers) { struct module *owner = disk->fops->owner; @@ -1197,13 +1184,13 @@ static int __blkdev_put(struct block_device *bdev, int for_part) mutex_unlock(&bdev->bd_mutex); bdput(bdev); if (victim) - __blkdev_put(victim, 1); + __blkdev_put(victim, mode, 1); return ret; } -int blkdev_put(struct block_device *bdev) +int blkdev_put(struct block_device *bdev, fmode_t mode) { - return __blkdev_put(bdev, 0); + return __blkdev_put(bdev, mode, 0); } EXPORT_SYMBOL(blkdev_put); @@ -1212,12 +1199,24 @@ static int blkdev_close(struct inode * inode, struct file * filp) struct block_device *bdev = I_BDEV(filp->f_mapping->host); if (bdev->bd_holder == filp) bd_release(bdev); - return blkdev_put(bdev); + return blkdev_put(bdev, filp->f_mode); } static long block_ioctl(struct file *file, unsigned cmd, unsigned long arg) { - return blkdev_ioctl(file->f_mapping->host, file, cmd, arg); + struct block_device *bdev = I_BDEV(file->f_mapping->host); + fmode_t mode = file->f_mode; + + /* + * O_NDELAY can be altered using fcntl(.., F_SETFL, ..), so we have + * to updated it before every ioctl. + */ + if (file->f_flags & O_NDELAY) + mode |= FMODE_NDELAY; + else + mode &= ~FMODE_NDELAY; + + return blkdev_ioctl(bdev, mode, cmd, arg); } static const struct address_space_operations def_blk_aops = { @@ -1253,7 +1252,7 @@ int ioctl_by_bdev(struct block_device *bdev, unsigned cmd, unsigned long arg) int res; mm_segment_t old_fs = get_fs(); set_fs(KERNEL_DS); - res = blkdev_ioctl(bdev->bd_inode, NULL, cmd, arg); + res = blkdev_ioctl(bdev, 0, cmd, arg); set_fs(old_fs); return res; } @@ -1262,39 +1261,39 @@ EXPORT_SYMBOL(ioctl_by_bdev); /** * lookup_bdev - lookup a struct block_device by name - * @pathname: special file representing the block device + * @path: special file representing the block device * * Get a reference to the blockdevice at @pathname in the current * namespace if possible and return it. Return ERR_PTR(error) * otherwise. */ -struct block_device *lookup_bdev(const char *path) +struct block_device *lookup_bdev(const char *pathname) { struct block_device *bdev; struct inode *inode; - struct nameidata nd; + struct path path; int error; - if (!path || !*path) + if (!pathname || !*pathname) return ERR_PTR(-EINVAL); - error = path_lookup(path, LOOKUP_FOLLOW, &nd); + error = kern_path(pathname, LOOKUP_FOLLOW, &path); if (error) return ERR_PTR(error); - inode = nd.path.dentry->d_inode; + inode = path.dentry->d_inode; error = -ENOTBLK; if (!S_ISBLK(inode->i_mode)) goto fail; error = -EACCES; - if (nd.path.mnt->mnt_flags & MNT_NODEV) + if (path.mnt->mnt_flags & MNT_NODEV) goto fail; error = -ENOMEM; bdev = bd_acquire(inode); if (!bdev) goto fail; out: - path_put(&nd.path); + path_put(&path); return bdev; fail: bdev = ERR_PTR(error); @@ -1303,32 +1302,29 @@ fail: EXPORT_SYMBOL(lookup_bdev); /** - * open_bdev_excl - open a block device by name and set it up for use + * open_bdev_exclusive - open a block device by name and set it up for use * * @path: special file representing the block device - * @flags: %MS_RDONLY for opening read-only + * @mode: FMODE_... combination to pass be used * @holder: owner for exclusion * * Open the blockdevice described by the special file at @path, claim it * for the @holder. */ -struct block_device *open_bdev_excl(const char *path, int flags, void *holder) +struct block_device *open_bdev_exclusive(const char *path, fmode_t mode, void *holder) { struct block_device *bdev; - mode_t mode = FMODE_READ; int error = 0; bdev = lookup_bdev(path); if (IS_ERR(bdev)) return bdev; - if (!(flags & MS_RDONLY)) - mode |= FMODE_WRITE; - error = blkdev_get(bdev, mode, 0); + error = blkdev_get(bdev, mode); if (error) return ERR_PTR(error); error = -EACCES; - if (!(flags & MS_RDONLY) && bdev_read_only(bdev)) + if ((mode & FMODE_WRITE) && bdev_read_only(bdev)) goto blkdev_put; error = bd_claim(bdev, holder); if (error) @@ -1337,26 +1333,27 @@ struct block_device *open_bdev_excl(const char *path, int flags, void *holder) return bdev; blkdev_put: - blkdev_put(bdev); + blkdev_put(bdev, mode); return ERR_PTR(error); } -EXPORT_SYMBOL(open_bdev_excl); +EXPORT_SYMBOL(open_bdev_exclusive); /** - * close_bdev_excl - release a blockdevice openen by open_bdev_excl() + * close_bdev_exclusive - close a blockdevice opened by open_bdev_exclusive() * * @bdev: blockdevice to close + * @mode: mode, must match that used to open. * - * This is the counterpart to open_bdev_excl(). + * This is the counterpart to open_bdev_exclusive(). */ -void close_bdev_excl(struct block_device *bdev) +void close_bdev_exclusive(struct block_device *bdev, fmode_t mode) { bd_release(bdev); - blkdev_put(bdev); + blkdev_put(bdev, mode); } -EXPORT_SYMBOL(close_bdev_excl); +EXPORT_SYMBOL(close_bdev_exclusive); int __invalidate_device(struct block_device *bdev) { diff --git a/fs/buffer.c b/fs/buffer.c index ac78d4c19b3..776ae091d3b 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -76,8 +76,7 @@ EXPORT_SYMBOL(__lock_buffer); void unlock_buffer(struct buffer_head *bh) { - smp_mb__before_clear_bit(); - clear_buffer_locked(bh); + clear_bit_unlock(BH_Lock, &bh->b_state); smp_mb__after_clear_bit(); wake_up_bit(&bh->b_state, BH_Lock); } @@ -100,10 +99,18 @@ __clear_page_buffers(struct page *page) page_cache_release(page); } + +static int quiet_error(struct buffer_head *bh) +{ + if (!test_bit(BH_Quiet, &bh->b_state) && printk_ratelimit()) + return 0; + return 1; +} + + static void buffer_io_error(struct buffer_head *bh) { char b[BDEVNAME_SIZE]; - printk(KERN_ERR "Buffer I/O error on device %s, logical block %Lu\n", bdevname(bh->b_bdev, b), (unsigned long long)bh->b_blocknr); @@ -145,7 +152,7 @@ void end_buffer_write_sync(struct buffer_head *bh, int uptodate) if (uptodate) { set_buffer_uptodate(bh); } else { - if (!buffer_eopnotsupp(bh) && printk_ratelimit()) { + if (!buffer_eopnotsupp(bh) && !quiet_error(bh)) { buffer_io_error(bh); printk(KERN_WARNING "lost page write due to " "I/O error on %s\n", @@ -395,7 +402,7 @@ static void end_buffer_async_read(struct buffer_head *bh, int uptodate) set_buffer_uptodate(bh); } else { clear_buffer_uptodate(bh); - if (printk_ratelimit()) + if (!quiet_error(bh)) buffer_io_error(bh); SetPageError(page); } @@ -456,7 +463,7 @@ static void end_buffer_async_write(struct buffer_head *bh, int uptodate) if (uptodate) { set_buffer_uptodate(bh); } else { - if (printk_ratelimit()) { + if (!quiet_error(bh)) { buffer_io_error(bh); printk(KERN_WARNING "lost page write due to " "I/O error on %s\n", @@ -879,6 +886,7 @@ void invalidate_inode_buffers(struct inode *inode) spin_unlock(&buffer_mapping->private_lock); } } +EXPORT_SYMBOL(invalidate_inode_buffers); /* * Remove any clean buffers from the inode's buffer list. This is called @@ -2913,6 +2921,9 @@ static void end_bio_bh_io_sync(struct bio *bio, int err) set_bit(BH_Eopnotsupp, &bh->b_state); } + if (unlikely (test_bit(BIO_QUIET,&bio->bi_flags))) + set_bit(BH_Quiet, &bh->b_state); + bh->b_end_io(bh, test_bit(BIO_UPTODATE, &bio->bi_flags)); bio_put(bio); } diff --git a/fs/char_dev.c b/fs/char_dev.c index 3cb7cda3d78..700697a7261 100644 --- a/fs/char_dev.c +++ b/fs/char_dev.c @@ -22,9 +22,6 @@ #include <linux/mutex.h> #include <linux/backing-dev.h> -#ifdef CONFIG_KMOD -#include <linux/kmod.h> -#endif #include "internal.h" /* @@ -389,15 +386,22 @@ static int chrdev_open(struct inode *inode, struct file *filp) cdev_put(new); if (ret) return ret; + + ret = -ENXIO; filp->f_op = fops_get(p->ops); - if (!filp->f_op) { - cdev_put(p); - return -ENXIO; - } - if (filp->f_op->open) + if (!filp->f_op) + goto out_cdev_put; + + if (filp->f_op->open) { ret = filp->f_op->open(inode,filp); - if (ret) - cdev_put(p); + if (ret) + goto out_cdev_put; + } + + return 0; + + out_cdev_put: + cdev_put(p); return ret; } diff --git a/fs/cifs/AUTHORS b/fs/cifs/AUTHORS index 9c136d7803d..7f7fa3c302a 100644 --- a/fs/cifs/AUTHORS +++ b/fs/cifs/AUTHORS @@ -36,7 +36,9 @@ Miklos Szeredi Kazeon team for various fixes especially for 2.4 version. Asser Ferno (Change Notify support) Shaggy (Dave Kleikamp) for inumerable small fs suggestions and some good cleanup +Gunter Kukkukk (testing and suggestions for support of old servers) Igor Mammedov (DFS support) +Jeff Layton (many, many fixes, as well as great work on the cifs Kerberos code) Test case and Bug Report contributors ------------------------------------- diff --git a/fs/cifs/CHANGES b/fs/cifs/CHANGES index 06e521a945c..080703a15f4 100644 --- a/fs/cifs/CHANGES +++ b/fs/cifs/CHANGES @@ -1,3 +1,28 @@ +Version 1.56 +------------ +Add "forcemandatorylock" mount option to allow user to use mandatory +rather than posix (advisory) byte range locks, even though server would +support posix byte range locks. Fix query of root inode when prefixpath +specified and user does not have access to query information about the +top of the share. Fix problem in 2.6.28 resolving DFS paths to +Samba servers (worked to Windows). + +Version 1.55 +------------ +Various fixes to make delete of open files behavior more predictable +(when delete of an open file fails we mark the file as "delete-on-close" +in a way that more servers accept, but only if we can first rename the +file to a temporary name). Add experimental support for more safely +handling fcntl(F_SETLEASE). Convert cifs to using blocking tcp +sends, and also let tcp autotune the socket send and receive buffers. +This reduces the number of EAGAIN errors returned by TCP/IP in +high stress workloads (and the number of retries on socket writes +when sending large SMBWriteX requests). Fix case in which a portion of +data can in some cases not get written to the file on the server before the +file is closed. Fix DFS parsing to properly handle path consumed field, +and to handle certain codepage conversions better. Fix mount and +umount race that can cause oops in mount or umount or reconnect. + Version 1.54 ------------ Fix premature write failure on congested networks (we would give up @@ -13,6 +38,7 @@ on dns_upcall (resolving DFS referralls). Fix plain text password authentication (requires setting SecurityFlags to 0x30030 to enable lanman and plain text though). Fix writes to be at correct offset when file is open with O_APPEND and file is on a directio (forcediretio) mount. +Fix bug in rewinding readdir directory searches. Add nodfs mount option. Version 1.53 ------------ diff --git a/fs/cifs/Kconfig b/fs/cifs/Kconfig new file mode 100644 index 00000000000..341a98965bd --- /dev/null +++ b/fs/cifs/Kconfig @@ -0,0 +1,142 @@ +config CIFS + tristate "CIFS support (advanced network filesystem, SMBFS successor)" + depends on INET + select NLS + help + This is the client VFS module for the Common Internet File System + (CIFS) protocol which is the successor to the Server Message Block + (SMB) protocol, the native file sharing mechanism for most early + PC operating systems. The CIFS protocol is fully supported by + file servers such as Windows 2000 (including Windows 2003, NT 4 + and Windows XP) as well by Samba (which provides excellent CIFS + server support for Linux and many other operating systems). Limited + support for OS/2 and Windows ME and similar servers is provided as + well. + + The cifs module provides an advanced network file system + client for mounting to CIFS compliant servers. It includes + support for DFS (hierarchical name space), secure per-user + session establishment via Kerberos or NTLM or NTLMv2, + safe distributed caching (oplock), optional packet + signing, Unicode and other internationalization improvements. + If you need to mount to Samba or Windows from this machine, say Y. + +config CIFS_STATS + bool "CIFS statistics" + depends on CIFS + help + Enabling this option will cause statistics for each server share + mounted by the cifs client to be displayed in /proc/fs/cifs/Stats + +config CIFS_STATS2 + bool "Extended statistics" + depends on CIFS_STATS + help + Enabling this option will allow more detailed statistics on SMB + request timing to be displayed in /proc/fs/cifs/DebugData and also + allow optional logging of slow responses to dmesg (depending on the + value of /proc/fs/cifs/cifsFYI, see fs/cifs/README for more details). + These additional statistics may have a minor effect on performance + and memory utilization. + + Unless you are a developer or are doing network performance analysis + or tuning, say N. + +config CIFS_WEAK_PW_HASH + bool "Support legacy servers which use weaker LANMAN security" + depends on CIFS + help + Modern CIFS servers including Samba and most Windows versions + (since 1997) support stronger NTLM (and even NTLMv2 and Kerberos) + security mechanisms. These hash the password more securely + than the mechanisms used in the older LANMAN version of the + SMB protocol but LANMAN based authentication is needed to + establish sessions with some old SMB servers. + + Enabling this option allows the cifs module to mount to older + LANMAN based servers such as OS/2 and Windows 95, but such + mounts may be less secure than mounts using NTLM or more recent + security mechanisms if you are on a public network. Unless you + have a need to access old SMB servers (and are on a private + network) you probably want to say N. Even if this support + is enabled in the kernel build, LANMAN authentication will not be + used automatically. At runtime LANMAN mounts are disabled but + can be set to required (or optional) either in + /proc/fs/cifs (see fs/cifs/README for more detail) or via an + option on the mount command. This support is disabled by + default in order to reduce the possibility of a downgrade + attack. + + If unsure, say N. + +config CIFS_UPCALL + bool "Kerberos/SPNEGO advanced session setup" + depends on CIFS && KEYS + help + Enables an upcall mechanism for CIFS which accesses + userspace helper utilities to provide SPNEGO packaged (RFC 4178) + Kerberos tickets which are needed to mount to certain secure servers + (for which more secure Kerberos authentication is required). If + unsure, say N. + +config CIFS_XATTR + bool "CIFS extended attributes" + depends on CIFS + help + Extended attributes are name:value pairs associated with inodes by + the kernel or by users (see the attr(5) manual page, or visit + <http://acl.bestbits.at/> for details). CIFS maps the name of + extended attributes beginning with the user namespace prefix + to SMB/CIFS EAs. EAs are stored on Windows servers without the + user namespace prefix, but their names are seen by Linux cifs clients + prefaced by the user namespace prefix. The system namespace + (used by some filesystems to store ACLs) is not supported at + this time. + + If unsure, say N. + +config CIFS_POSIX + bool "CIFS POSIX Extensions" + depends on CIFS_XATTR + help + Enabling this option will cause the cifs client to attempt to + negotiate a newer dialect with servers, such as Samba 3.0.5 + or later, that optionally can handle more POSIX like (rather + than Windows like) file behavior. It also enables + support for POSIX ACLs (getfacl and setfacl) to servers + (such as Samba 3.10 and later) which can negotiate + CIFS POSIX ACL support. If unsure, say N. + +config CIFS_DEBUG2 + bool "Enable additional CIFS debugging routines" + depends on CIFS + help + Enabling this option adds a few more debugging routines + to the cifs code which slightly increases the size of + the cifs module and can cause additional logging of debug + messages in some error paths, slowing performance. This + option can be turned off unless you are debugging + cifs problems. If unsure, say N. + +config CIFS_EXPERIMENTAL + bool "CIFS Experimental Features (EXPERIMENTAL)" + depends on CIFS && EXPERIMENTAL + help + Enables cifs features under testing. These features are + experimental and currently include DFS support and directory + change notification ie fcntl(F_DNOTIFY), as well as the upcall + mechanism which will be used for Kerberos session negotiation + and uid remapping. Some of these features also may depend on + setting a value of 1 to the pseudo-file /proc/fs/cifs/Experimental + (which is disabled by default). See the file fs/cifs/README + for more details. If unsure, say N. + +config CIFS_DFS_UPCALL + bool "DFS feature support (EXPERIMENTAL)" + depends on CIFS_EXPERIMENTAL + depends on KEYS + help + Enables an upcall mechanism for CIFS which contacts userspace + helper utilities to provide server name resolution (host names to + IP addresses) which is needed for implicit mounts of DFS junction + points. If unsure, say N. diff --git a/fs/cifs/README b/fs/cifs/README index bd2343d4c6a..da4515e3be2 100644 --- a/fs/cifs/README +++ b/fs/cifs/README @@ -463,6 +463,19 @@ A partial list of the supported mount options follows: with cifs style mandatory byte range locks (and most cifs servers do not yet support requesting advisory byte range locks). + forcemandatorylock Even if the server supports posix (advisory) byte range + locking, send only mandatory lock requests. For some + (presumably rare) applications, originally coded for + DOS/Windows, which require Windows style mandatory byte range + locking, they may be able to take advantage of this option, + forcing the cifs client to only send mandatory locks + even if the cifs server would support posix advisory locks. + "forcemand" is accepted as a shorter form of this mount + option. + nodfs Disable DFS (global name space support) even if the + server claims to support it. This can help work around + a problem with parsing of DFS paths with Samba server + versions 3.0.24 and 3.0.25. remount remount the share (often used to change from ro to rw mounts or vice versa) cifsacl Report mode bits (e.g. on stat) based on the Windows ACL for @@ -488,6 +501,19 @@ A partial list of the supported mount options follows: Note that this differs from the sign mount option in that it causes encryption of data sent over this mounted share but other shares mounted to the same server are unaffected. + locallease This option is rarely needed. Fcntl F_SETLEASE is + used by some applications such as Samba and NFSv4 server to + check to see whether a file is cacheable. CIFS has no way + to explicitly request a lease, but can check whether a file + is cacheable (oplocked). Unfortunately, even if a file + is not oplocked, it could still be cacheable (ie cifs client + could grant fcntl leases if no other local processes are using + the file) for cases for example such as when the server does not + support oplocks and the user is sure that the only updates to + the file will be from this client. Specifying this mount option + will allow the cifs client to check for leases (only) locally + for files which are not oplocked instead of denying leases + in that case. (EXPERIMENTAL) sec Security mode. Allowed values are: none attempt to connection as a null user (no name) krb5 Use Kerberos version 5 authentication @@ -638,6 +664,9 @@ requires enabling CONFIG_CIFS_EXPERIMENTAL cifsacl support needed to retrieve approximated mode bits based on the contents on the CIFS ACL. + lease support: cifs will check the oplock state before calling into + the vfs to see if we can grant a lease on a file. + DNOTIFY fcntl: needed for support of directory change notification and perhaps later for file leases) diff --git a/fs/cifs/cifs_debug.c b/fs/cifs/cifs_debug.c index 69a12aae91d..490e34bbf27 100644 --- a/fs/cifs/cifs_debug.c +++ b/fs/cifs/cifs_debug.c @@ -107,12 +107,13 @@ void cifs_dump_mids(struct TCP_Server_Info *server) #ifdef CONFIG_PROC_FS static int cifs_debug_data_proc_show(struct seq_file *m, void *v) { - struct list_head *tmp; - struct list_head *tmp1; + struct list_head *tmp1, *tmp2, *tmp3; struct mid_q_entry *mid_entry; + struct TCP_Server_Info *server; struct cifsSesInfo *ses; struct cifsTconInfo *tcon; - int i; + int i, j; + __u32 dev_type; seq_puts(m, "Display Internal CIFS Data Structures for Debugging\n" @@ -122,46 +123,78 @@ static int cifs_debug_data_proc_show(struct seq_file *m, void *v) seq_printf(m, "Servers:"); i = 0; - read_lock(&GlobalSMBSeslock); - list_for_each(tmp, &GlobalSMBSessionList) { + read_lock(&cifs_tcp_ses_lock); + list_for_each(tmp1, &cifs_tcp_ses_list) { + server = list_entry(tmp1, struct TCP_Server_Info, + tcp_ses_list); i++; - ses = list_entry(tmp, struct cifsSesInfo, cifsSessionList); - if ((ses->serverDomain == NULL) || (ses->serverOS == NULL) || - (ses->serverNOS == NULL)) { - seq_printf(m, "\nentry for %s not fully " - "displayed\n\t", ses->serverName); - } else { - seq_printf(m, - "\n%d) Name: %s Domain: %s Mounts: %d OS:" - " %s \n\tNOS: %s\tCapability: 0x%x\n\tSMB" + list_for_each(tmp2, &server->smb_ses_list) { + ses = list_entry(tmp2, struct cifsSesInfo, + smb_ses_list); + if ((ses->serverDomain == NULL) || + (ses->serverOS == NULL) || + (ses->serverNOS == NULL)) { + seq_printf(m, "\n%d) entry for %s not fully " + "displayed\n\t", i, ses->serverName); + } else { + seq_printf(m, + "\n%d) Name: %s Domain: %s Uses: %d OS:" + " %s\n\tNOS: %s\tCapability: 0x%x\n\tSMB" " session status: %d\t", i, ses->serverName, ses->serverDomain, - atomic_read(&ses->inUse), - ses->serverOS, ses->serverNOS, + ses->ses_count, ses->serverOS, ses->serverNOS, ses->capabilities, ses->status); - } - if (ses->server) { + } seq_printf(m, "TCP status: %d\n\tLocal Users To " - "Server: %d SecMode: 0x%x Req On Wire: %d", - ses->server->tcpStatus, - atomic_read(&ses->server->socketUseCount), - ses->server->secMode, - atomic_read(&ses->server->inFlight)); + "Server: %d SecMode: 0x%x Req On Wire: %d", + server->tcpStatus, server->srv_count, + server->secMode, + atomic_read(&server->inFlight)); #ifdef CONFIG_CIFS_STATS2 seq_printf(m, " In Send: %d In MaxReq Wait: %d", - atomic_read(&ses->server->inSend), - atomic_read(&ses->server->num_waiters)); + atomic_read(&server->inSend), + atomic_read(&server->num_waiters)); #endif - seq_puts(m, "\nMIDs:\n"); + seq_puts(m, "\n\tShares:"); + j = 0; + list_for_each(tmp3, &ses->tcon_list) { + tcon = list_entry(tmp3, struct cifsTconInfo, + tcon_list); + ++j; + dev_type = le32_to_cpu(tcon->fsDevInfo.DeviceType); + seq_printf(m, "\n\t%d) %s Mounts: %d ", j, + tcon->treeName, tcon->tc_count); + if (tcon->nativeFileSystem) { + seq_printf(m, "Type: %s ", + tcon->nativeFileSystem); + } + seq_printf(m, "DevInfo: 0x%x Attributes: 0x%x" + "\nPathComponentMax: %d Status: 0x%d", + le32_to_cpu(tcon->fsDevInfo.DeviceCharacteristics), + le32_to_cpu(tcon->fsAttrInfo.Attributes), + le32_to_cpu(tcon->fsAttrInfo.MaxPathNameComponentLength), + tcon->tidStatus); + if (dev_type == FILE_DEVICE_DISK) + seq_puts(m, " type: DISK "); + else if (dev_type == FILE_DEVICE_CD_ROM) + seq_puts(m, " type: CDROM "); + else + seq_printf(m, " type: %d ", dev_type); + + if (tcon->need_reconnect) + seq_puts(m, "\tDISCONNECTED "); + seq_putc(m, '\n'); + } + + seq_puts(m, "\n\tMIDs:\n"); spin_lock(&GlobalMid_Lock); - list_for_each(tmp1, &ses->server->pending_mid_q) { - mid_entry = list_entry(tmp1, struct - mid_q_entry, + list_for_each(tmp3, &server->pending_mid_q) { + mid_entry = list_entry(tmp3, struct mid_q_entry, qhead); - seq_printf(m, "State: %d com: %d pid:" + seq_printf(m, "\tState: %d com: %d pid:" " %d tsk: %p mid %d\n", mid_entry->midState, (int)mid_entry->command, @@ -171,44 +204,8 @@ static int cifs_debug_data_proc_show(struct seq_file *m, void *v) } spin_unlock(&GlobalMid_Lock); } - - } - read_unlock(&GlobalSMBSeslock); - seq_putc(m, '\n'); - - seq_puts(m, "Shares:"); - - i = 0; - read_lock(&GlobalSMBSeslock); - list_for_each(tmp, &GlobalTreeConnectionList) { - __u32 dev_type; - i++; - tcon = list_entry(tmp, struct cifsTconInfo, cifsConnectionList); - dev_type = le32_to_cpu(tcon->fsDevInfo.DeviceType); - seq_printf(m, "\n%d) %s Uses: %d ", i, - tcon->treeName, atomic_read(&tcon->useCount)); - if (tcon->nativeFileSystem) { - seq_printf(m, "Type: %s ", - tcon->nativeFileSystem); - } - seq_printf(m, "DevInfo: 0x%x Attributes: 0x%x" - "\nPathComponentMax: %d Status: %d", - le32_to_cpu(tcon->fsDevInfo.DeviceCharacteristics), - le32_to_cpu(tcon->fsAttrInfo.Attributes), - le32_to_cpu(tcon->fsAttrInfo.MaxPathNameComponentLength), - tcon->tidStatus); - if (dev_type == FILE_DEVICE_DISK) - seq_puts(m, " type: DISK "); - else if (dev_type == FILE_DEVICE_CD_ROM) - seq_puts(m, " type: CDROM "); - else - seq_printf(m, " type: %d ", dev_type); - - if (tcon->tidStatus == CifsNeedReconnect) - seq_puts(m, "\tDISCONNECTED "); } - read_unlock(&GlobalSMBSeslock); - + read_unlock(&cifs_tcp_ses_lock); seq_putc(m, '\n'); /* BB add code to dump additional info such as TCP session info now */ @@ -234,7 +231,9 @@ static ssize_t cifs_stats_proc_write(struct file *file, { char c; int rc; - struct list_head *tmp; + struct list_head *tmp1, *tmp2, *tmp3; + struct TCP_Server_Info *server; + struct cifsSesInfo *ses; struct cifsTconInfo *tcon; rc = get_user(c, buffer); @@ -242,33 +241,42 @@ static ssize_t cifs_stats_proc_write(struct file *file, return rc; if (c == '1' || c == 'y' || c == 'Y' || c == '0') { - read_lock(&GlobalSMBSeslock); #ifdef CONFIG_CIFS_STATS2 atomic_set(&totBufAllocCount, 0); atomic_set(&totSmBufAllocCount, 0); #endif /* CONFIG_CIFS_STATS2 */ - list_for_each(tmp, &GlobalTreeConnectionList) { - tcon = list_entry(tmp, struct cifsTconInfo, - cifsConnectionList); - atomic_set(&tcon->num_smbs_sent, 0); - atomic_set(&tcon->num_writes, 0); - atomic_set(&tcon->num_reads, 0); - atomic_set(&tcon->num_oplock_brks, 0); - atomic_set(&tcon->num_opens, 0); - atomic_set(&tcon->num_closes, 0); - atomic_set(&tcon->num_deletes, 0); - atomic_set(&tcon->num_mkdirs, 0); - atomic_set(&tcon->num_rmdirs, 0); - atomic_set(&tcon->num_renames, 0); - atomic_set(&tcon->num_t2renames, 0); - atomic_set(&tcon->num_ffirst, 0); - atomic_set(&tcon->num_fnext, 0); - atomic_set(&tcon->num_fclose, 0); - atomic_set(&tcon->num_hardlinks, 0); - atomic_set(&tcon->num_symlinks, 0); - atomic_set(&tcon->num_locks, 0); + read_lock(&cifs_tcp_ses_lock); + list_for_each(tmp1, &cifs_tcp_ses_list) { + server = list_entry(tmp1, struct TCP_Server_Info, + tcp_ses_list); + list_for_each(tmp2, &server->smb_ses_list) { + ses = list_entry(tmp2, struct cifsSesInfo, + smb_ses_list); + list_for_each(tmp3, &ses->tcon_list) { + tcon = list_entry(tmp3, + struct cifsTconInfo, + tcon_list); + atomic_set(&tcon->num_smbs_sent, 0); + atomic_set(&tcon->num_writes, 0); + atomic_set(&tcon->num_reads, 0); + atomic_set(&tcon->num_oplock_brks, 0); + atomic_set(&tcon->num_opens, 0); + atomic_set(&tcon->num_closes, 0); + atomic_set(&tcon->num_deletes, 0); + atomic_set(&tcon->num_mkdirs, 0); + atomic_set(&tcon->num_rmdirs, 0); + atomic_set(&tcon->num_renames, 0); + atomic_set(&tcon->num_t2renames, 0); + atomic_set(&tcon->num_ffirst, 0); + atomic_set(&tcon->num_fnext, 0); + atomic_set(&tcon->num_fclose, 0); + atomic_set(&tcon->num_hardlinks, 0); + atomic_set(&tcon->num_symlinks, 0); + atomic_set(&tcon->num_locks, 0); + } + } } - read_unlock(&GlobalSMBSeslock); + read_unlock(&cifs_tcp_ses_lock); } return count; @@ -277,7 +285,9 @@ static ssize_t cifs_stats_proc_write(struct file *file, static int cifs_stats_proc_show(struct seq_file *m, void *v) { int i; - struct list_head *tmp; + struct list_head *tmp1, *tmp2, *tmp3; + struct TCP_Server_Info *server; + struct cifsSesInfo *ses; struct cifsTconInfo *tcon; seq_printf(m, @@ -306,44 +316,55 @@ static int cifs_stats_proc_show(struct seq_file *m, void *v) GlobalCurrentXid, GlobalMaxActiveXid); i = 0; - read_lock(&GlobalSMBSeslock); - list_for_each(tmp, &GlobalTreeConnectionList) { - i++; - tcon = list_entry(tmp, struct cifsTconInfo, cifsConnectionList); - seq_printf(m, "\n%d) %s", i, tcon->treeName); - if (tcon->tidStatus == CifsNeedReconnect) - seq_puts(m, "\tDISCONNECTED "); - seq_printf(m, "\nSMBs: %d Oplock Breaks: %d", - atomic_read(&tcon->num_smbs_sent), - atomic_read(&tcon->num_oplock_brks)); - seq_printf(m, "\nReads: %d Bytes: %lld", - atomic_read(&tcon->num_reads), - (long long)(tcon->bytes_read)); - seq_printf(m, "\nWrites: %d Bytes: %lld", - atomic_read(&tcon->num_writes), - (long long)(tcon->bytes_written)); - seq_printf(m, - "\nLocks: %d HardLinks: %d Symlinks: %d", - atomic_read(&tcon->num_locks), - atomic_read(&tcon->num_hardlinks), - atomic_read(&tcon->num_symlinks)); - - seq_printf(m, "\nOpens: %d Closes: %d Deletes: %d", - atomic_read(&tcon->num_opens), - atomic_read(&tcon->num_closes), - atomic_read(&tcon->num_deletes)); - seq_printf(m, "\nMkdirs: %d Rmdirs: %d", - atomic_read(&tcon->num_mkdirs), - atomic_read(&tcon->num_rmdirs)); - seq_printf(m, "\nRenames: %d T2 Renames %d", - atomic_read(&tcon->num_renames), - atomic_read(&tcon->num_t2renames)); - seq_printf(m, "\nFindFirst: %d FNext %d FClose %d", - atomic_read(&tcon->num_ffirst), - atomic_read(&tcon->num_fnext), - atomic_read(&tcon->num_fclose)); + read_lock(&cifs_tcp_ses_lock); + list_for_each(tmp1, &cifs_tcp_ses_list) { + server = list_entry(tmp1, struct TCP_Server_Info, + tcp_ses_list); + list_for_each(tmp2, &server->smb_ses_list) { + ses = list_entry(tmp2, struct cifsSesInfo, + smb_ses_list); + list_for_each(tmp3, &ses->tcon_list) { + tcon = list_entry(tmp3, + struct cifsTconInfo, + tcon_list); + i++; + seq_printf(m, "\n%d) %s", i, tcon->treeName); + if (tcon->need_reconnect) + seq_puts(m, "\tDISCONNECTED "); + seq_printf(m, "\nSMBs: %d Oplock Breaks: %d", + atomic_read(&tcon->num_smbs_sent), + atomic_read(&tcon->num_oplock_brks)); + seq_printf(m, "\nReads: %d Bytes: %lld", + atomic_read(&tcon->num_reads), + (long long)(tcon->bytes_read)); + seq_printf(m, "\nWrites: %d Bytes: %lld", + atomic_read(&tcon->num_writes), + (long long)(tcon->bytes_written)); + seq_printf(m, "\nLocks: %d HardLinks: %d " + "Symlinks: %d", + atomic_read(&tcon->num_locks), + atomic_read(&tcon->num_hardlinks), + atomic_read(&tcon->num_symlinks)); + seq_printf(m, "\nOpens: %d Closes: %d" + "Deletes: %d", + atomic_read(&tcon->num_opens), + atomic_read(&tcon->num_closes), + atomic_read(&tcon->num_deletes)); + seq_printf(m, "\nMkdirs: %d Rmdirs: %d", + atomic_read(&tcon->num_mkdirs), + atomic_read(&tcon->num_rmdirs)); + seq_printf(m, "\nRenames: %d T2 Renames %d", + atomic_read(&tcon->num_renames), + atomic_read(&tcon->num_t2renames)); + seq_printf(m, "\nFindFirst: %d FNext %d " + "FClose %d", + atomic_read(&tcon->num_ffirst), + atomic_read(&tcon->num_fnext), + atomic_read(&tcon->num_fclose)); + } + } } - read_unlock(&GlobalSMBSeslock); + read_unlock(&cifs_tcp_ses_lock); seq_putc(m, '\n'); return 0; diff --git a/fs/cifs/cifs_dfs_ref.c b/fs/cifs/cifs_dfs_ref.c index d2c8eef84f3..85c0a74d034 100644 --- a/fs/cifs/cifs_dfs_ref.c +++ b/fs/cifs/cifs_dfs_ref.c @@ -106,7 +106,8 @@ static char *cifs_get_share_name(const char *node_name) /** * compose_mount_options - creates mount options for refferral * @sb_mountdata: parent/root DFS mount options (template) - * @ref_unc: refferral server UNC + * @dentry: point where we are going to mount + * @ref: server's referral * @devname: pointer for saving device name * * creates mount options for submount based on template options sb_mountdata @@ -116,33 +117,39 @@ static char *cifs_get_share_name(const char *node_name) * Caller is responcible for freeing retunrned value if it is not error. */ static char *compose_mount_options(const char *sb_mountdata, - const char *ref_unc, + struct dentry *dentry, + const struct dfs_info3_param *ref, char **devname) { int rc; - char *mountdata; + char *mountdata = NULL; int md_len; char *tkn_e; char *srvIP = NULL; char sep = ','; int off, noff; + char *fullpath; if (sb_mountdata == NULL) return ERR_PTR(-EINVAL); - *devname = cifs_get_share_name(ref_unc); + *devname = cifs_get_share_name(ref->node_name); rc = dns_resolve_server_name_to_ip(*devname, &srvIP); if (rc != 0) { - cERROR(1, ("%s: Failed to resolve server part of %s to IP", - __func__, *devname)); - mountdata = ERR_PTR(rc); - goto compose_mount_options_out; + cERROR(1, ("%s: Failed to resolve server part of %s to IP: %d", + __func__, *devname, rc));; + goto compose_mount_options_err; } - md_len = strlen(sb_mountdata) + strlen(srvIP) + strlen(ref_unc) + 3; + /* md_len = strlen(...) + 12 for 'sep+prefixpath=' + * assuming that we have 'unc=' and 'ip=' in + * the original sb_mountdata + */ + md_len = strlen(sb_mountdata) + strlen(srvIP) + + strlen(ref->node_name) + 12; mountdata = kzalloc(md_len+1, GFP_KERNEL); if (mountdata == NULL) { - mountdata = ERR_PTR(-ENOMEM); - goto compose_mount_options_out; + rc = -ENOMEM; + goto compose_mount_options_err; } /* copy all options except of unc,ip,prefixpath */ @@ -152,53 +159,87 @@ static char *compose_mount_options(const char *sb_mountdata, strncpy(mountdata, sb_mountdata, 5); off += 5; } - while ((tkn_e = strchr(sb_mountdata+off, sep))) { - noff = (tkn_e - (sb_mountdata+off)) + 1; - if (strnicmp(sb_mountdata+off, "unc=", 4) == 0) { + + do { + tkn_e = strchr(sb_mountdata + off, sep); + if (tkn_e == NULL) + noff = strlen(sb_mountdata + off); + else + noff = tkn_e - (sb_mountdata + off) + 1; + + if (strnicmp(sb_mountdata + off, "unc=", 4) == 0) { off += noff; continue; } - if (strnicmp(sb_mountdata+off, "ip=", 3) == 0) { + if (strnicmp(sb_mountdata + off, "ip=", 3) == 0) { off += noff; continue; } - if (strnicmp(sb_mountdata+off, "prefixpath=", 3) == 0) { + if (strnicmp(sb_mountdata + off, "prefixpath=", 11) == 0) { off += noff; continue; } - strncat(mountdata, sb_mountdata+off, noff); + strncat(mountdata, sb_mountdata + off, noff); off += noff; - } - strcat(mountdata, sb_mountdata+off); + } while (tkn_e); + strcat(mountdata, sb_mountdata + off); mountdata[md_len] = '\0'; /* copy new IP and ref share name */ - strcat(mountdata, ",ip="); + if (mountdata[strlen(mountdata) - 1] != sep) + strncat(mountdata, &sep, 1); + strcat(mountdata, "ip="); strcat(mountdata, srvIP); - strcat(mountdata, ",unc="); + strncat(mountdata, &sep, 1); + strcat(mountdata, "unc="); strcat(mountdata, *devname); /* find & copy prefixpath */ - tkn_e = strchr(ref_unc+2, '\\'); - if (tkn_e) { - tkn_e = strchr(tkn_e+1, '\\'); - if (tkn_e) { - strcat(mountdata, ",prefixpath="); - strcat(mountdata, tkn_e+1); - } + tkn_e = strchr(ref->node_name + 2, '\\'); + if (tkn_e == NULL) { + /* invalid unc, missing share name*/ + rc = -EINVAL; + goto compose_mount_options_err; } + /* + * this function gives us a path with a double backslash prefix. We + * require a single backslash for DFS. Temporarily increment fullpath + * to put it in the proper form and decrement before freeing it. + */ + fullpath = build_path_from_dentry(dentry); + if (!fullpath) { + rc = -ENOMEM; + goto compose_mount_options_err; + } + ++fullpath; + tkn_e = strchr(tkn_e + 1, '\\'); + if (tkn_e || (strlen(fullpath) - ref->path_consumed)) { + strncat(mountdata, &sep, 1); + strcat(mountdata, "prefixpath="); + if (tkn_e) + strcat(mountdata, tkn_e + 1); + strcat(mountdata, fullpath + ref->path_consumed); + } + --fullpath; + kfree(fullpath); + /*cFYI(1,("%s: parent mountdata: %s", __func__,sb_mountdata));*/ /*cFYI(1, ("%s: submount mountdata: %s", __func__, mountdata ));*/ compose_mount_options_out: kfree(srvIP); return mountdata; + +compose_mount_options_err: + kfree(mountdata); + mountdata = ERR_PTR(rc); + goto compose_mount_options_out; } static struct vfsmount *cifs_dfs_do_refmount(const struct vfsmount *mnt_parent, - struct dentry *dentry, char *ref_unc) + struct dentry *dentry, const struct dfs_info3_param *ref) { struct cifs_sb_info *cifs_sb; struct vfsmount *mnt; @@ -207,7 +248,7 @@ static struct vfsmount *cifs_dfs_do_refmount(const struct vfsmount *mnt_parent, cifs_sb = CIFS_SB(dentry->d_inode->i_sb); mountdata = compose_mount_options(cifs_sb->mountdata, - ref_unc, &devname); + dentry, ref, &devname); if (IS_ERR(mountdata)) return (struct vfsmount *)mountdata; @@ -286,13 +327,19 @@ cifs_dfs_follow_mountpoint(struct dentry *dentry, struct nameidata *nd) goto out_err; } + /* + * The MSDFS spec states that paths in DFS referral requests and + * responses must be prefixed by a single '\' character instead of + * the double backslashes usually used in the UNC. This function + * gives us the latter, so we must adjust the result. + */ full_path = build_path_from_dentry(dentry); if (full_path == NULL) { rc = -ENOMEM; goto out_err; } - rc = get_dfs_path(xid, ses , full_path, cifs_sb->local_nls, + rc = get_dfs_path(xid, ses , full_path + 1, cifs_sb->local_nls, &num_referrals, &referrals, cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR); @@ -310,7 +357,7 @@ cifs_dfs_follow_mountpoint(struct dentry *dentry, struct nameidata *nd) } mnt = cifs_dfs_do_refmount(nd->path.mnt, nd->path.dentry, - referrals[i].node_name); + referrals + i); cFYI(1, ("%s: cifs_dfs_do_refmount:%s , mnt:%p", __func__, referrals[i].node_name, mnt)); diff --git a/fs/cifs/cifs_fs_sb.h b/fs/cifs/cifs_fs_sb.h index 877c85409f1..c4c306f7b06 100644 --- a/fs/cifs/cifs_fs_sb.h +++ b/fs/cifs/cifs_fs_sb.h @@ -19,8 +19,8 @@ #define _CIFS_FS_SB_H #define CIFS_MOUNT_NO_PERM 1 /* do not do client vfs_perm check */ -#define CIFS_MOUNT_SET_UID 2 /* set current->euid in create etc. */ -#define CIFS_MOUNT_SERVER_INUM 4 /* inode numbers from uniqueid from server */ +#define CIFS_MOUNT_SET_UID 2 /* set current's euid in create etc. */ +#define CIFS_MOUNT_SERVER_INUM 4 /* inode numbers from uniqueid from server */ #define CIFS_MOUNT_DIRECT_IO 8 /* do not write nor read through page cache */ #define CIFS_MOUNT_NO_XATTR 0x10 /* if set - disable xattr support */ #define CIFS_MOUNT_MAP_SPECIAL_CHR 0x20 /* remap illegal chars in filenames */ @@ -30,7 +30,8 @@ #define CIFS_MOUNT_CIFS_ACL 0x200 /* send ACL requests to non-POSIX srv */ #define CIFS_MOUNT_OVERR_UID 0x400 /* override uid returned from server */ #define CIFS_MOUNT_OVERR_GID 0x800 /* override gid returned from server */ -#define CIFS_MOUNT_DYNPERM 0x1000 /* allow in-memory only mode setting */ +#define CIFS_MOUNT_DYNPERM 0x1000 /* allow in-memory only mode setting */ +#define CIFS_MOUNT_NOPOSIXBRL 0x2000 /* mandatory not posix byte range lock */ struct cifs_sb_info { struct cifsTconInfo *tcon; /* primary mount */ diff --git a/fs/cifs/cifs_spnego.c b/fs/cifs/cifs_spnego.c index fcee9298b62..3fd3a9df043 100644 --- a/fs/cifs/cifs_spnego.c +++ b/fs/cifs/cifs_spnego.c @@ -73,8 +73,8 @@ struct key_type cifs_spnego_key_type = { * strlen(";sec=ntlmsspi") */ #define MAX_MECH_STR_LEN 13 -/* max possible addr len eg FEDC:BA98:7654:3210:FEDC:BA98:7654:3210/60 */ -#define MAX_IPV6_ADDR_LEN 42 +/* max possible addr len eg FEDC:BA98:7654:3210:FEDC:BA98:7654:3210/128 */ +#define MAX_IPV6_ADDR_LEN 43 /* strlen of "host=" */ #define HOST_KEY_LEN 5 @@ -121,11 +121,9 @@ cifs_get_spnego_key(struct cifsSesInfo *sesInfo) /* add the server address */ if (server->addr.sockAddr.sin_family == AF_INET) - sprintf(dp, "ip4=" NIPQUAD_FMT, - NIPQUAD(server->addr.sockAddr.sin_addr)); + sprintf(dp, "ip4=%pI4", &server->addr.sockAddr.sin_addr); else if (server->addr.sockAddr.sin_family == AF_INET6) - sprintf(dp, "ip6=" NIP6_SEQFMT, - NIP6(server->addr.sockAddr6.sin6_addr)); + sprintf(dp, "ip6=%pi6", &server->addr.sockAddr6.sin6_addr); else goto out; diff --git a/fs/cifs/cifsencrypt.c b/fs/cifs/cifsencrypt.c index bd5f13d3845..d4839cf0cb2 100644 --- a/fs/cifs/cifsencrypt.c +++ b/fs/cifs/cifsencrypt.c @@ -37,7 +37,7 @@ extern void mdfour(unsigned char *out, unsigned char *in, int n); extern void E_md4hash(const unsigned char *passwd, unsigned char *p16); -extern void SMBencrypt(unsigned char *passwd, unsigned char *c8, +extern void SMBencrypt(unsigned char *passwd, const unsigned char *c8, unsigned char *p24); static int cifs_calculate_signature(const struct smb_hdr *cifs_pdu, @@ -280,25 +280,22 @@ int CalcNTLMv2_partial_mac_key(struct cifsSesInfo *ses, } #ifdef CONFIG_CIFS_WEAK_PW_HASH -void calc_lanman_hash(struct cifsSesInfo *ses, char *lnm_session_key) +void calc_lanman_hash(const char *password, const char *cryptkey, bool encrypt, + char *lnm_session_key) { int i; char password_with_pad[CIFS_ENCPWD_SIZE]; - if (ses->server == NULL) - return; - memset(password_with_pad, 0, CIFS_ENCPWD_SIZE); - if (ses->password) - strncpy(password_with_pad, ses->password, CIFS_ENCPWD_SIZE); - - if ((ses->server->secMode & SECMODE_PW_ENCRYPT) == 0) - if (extended_security & CIFSSEC_MAY_PLNTXT) { - memset(lnm_session_key, 0, CIFS_SESS_KEY_SIZE); - memcpy(lnm_session_key, password_with_pad, - CIFS_ENCPWD_SIZE); - return; - } + if (password) + strncpy(password_with_pad, password, CIFS_ENCPWD_SIZE); + + if (!encrypt && extended_security & CIFSSEC_MAY_PLNTXT) { + memset(lnm_session_key, 0, CIFS_SESS_KEY_SIZE); + memcpy(lnm_session_key, password_with_pad, + CIFS_ENCPWD_SIZE); + return; + } /* calculate old style session key */ /* calling toupper is less broken than repeatedly @@ -314,7 +311,8 @@ void calc_lanman_hash(struct cifsSesInfo *ses, char *lnm_session_key) for (i = 0; i < CIFS_ENCPWD_SIZE; i++) password_with_pad[i] = toupper(password_with_pad[i]); - SMBencrypt(password_with_pad, ses->server->cryptKey, lnm_session_key); + SMBencrypt(password_with_pad, cryptkey, lnm_session_key); + /* clear password before we return/free memory */ memset(password_with_pad, 0, CIFS_ENCPWD_SIZE); } diff --git a/fs/cifs/cifsencrypt.h b/fs/cifs/cifsencrypt.h index 152fa2dcfc6..15d2ec00647 100644 --- a/fs/cifs/cifsencrypt.h +++ b/fs/cifs/cifsencrypt.h @@ -26,7 +26,8 @@ extern void mdfour(unsigned char *out, unsigned char *in, int n); /* smbdes.c */ extern void E_P16(unsigned char *p14, unsigned char *p16); -extern void E_P24(unsigned char *p21, unsigned char *c8, unsigned char *p24); +extern void E_P24(unsigned char *p21, const unsigned char *c8, + unsigned char *p24); diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c index 25ecbd5b040..0005a194a75 100644 --- a/fs/cifs/cifsfs.c +++ b/fs/cifs/cifsfs.c @@ -66,7 +66,9 @@ unsigned int sign_CIFS_PDUs = 1; extern struct task_struct *oplockThread; /* remove sparse warning */ struct task_struct *oplockThread = NULL; /* extern struct task_struct * dnotifyThread; remove sparse warning */ +#ifdef CONFIG_CIFS_EXPERIMENTAL static struct task_struct *dnotifyThread = NULL; +#endif static const struct super_operations cifs_super_ops; unsigned int CIFSMaxBufSize = CIFS_MAX_MSGSIZE; module_param(CIFSMaxBufSize, int, 0); @@ -275,9 +277,12 @@ static int cifs_permission(struct inode *inode, int mask) cifs_sb = CIFS_SB(inode->i_sb); - if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NO_PERM) - return 0; - else /* file mode might have been restricted at mount time + if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NO_PERM) { + if ((mask & MAY_EXEC) && !execute_ok(inode)) + return -EACCES; + else + return 0; + } else /* file mode might have been restricted at mount time on the client (above and beyond ACL on servers) for servers which do not support setting and viewing mode bits, so allowing client to check permissions is useful */ @@ -309,6 +314,7 @@ cifs_alloc_inode(struct super_block *sb) file data or metadata */ cifs_inode->clientCanCacheRead = false; cifs_inode->clientCanCacheAll = false; + cifs_inode->delete_pending = false; cifs_inode->vfs_inode.i_blkbits = 14; /* 2**14 = CIFS_MAX_MSGSIZE */ /* Can not set i_flags here - they get immediately overwritten @@ -333,39 +339,58 @@ static int cifs_show_options(struct seq_file *s, struct vfsmount *m) { struct cifs_sb_info *cifs_sb; + struct cifsTconInfo *tcon; + struct TCP_Server_Info *server; cifs_sb = CIFS_SB(m->mnt_sb); if (cifs_sb) { - if (cifs_sb->tcon) { -/* BB add prepath to mount options displayed */ + tcon = cifs_sb->tcon; + if (tcon) { seq_printf(s, ",unc=%s", cifs_sb->tcon->treeName); - if (cifs_sb->tcon->ses) { - if (cifs_sb->tcon->ses->userName) + if (tcon->ses) { + if (tcon->ses->userName) seq_printf(s, ",username=%s", - cifs_sb->tcon->ses->userName); - if (cifs_sb->tcon->ses->domainName) + tcon->ses->userName); + if (tcon->ses->domainName) seq_printf(s, ",domain=%s", - cifs_sb->tcon->ses->domainName); + tcon->ses->domainName); + server = tcon->ses->server; + if (server) { + seq_printf(s, ",addr="); + switch (server->addr.sockAddr6. + sin6_family) { + case AF_INET6: + seq_printf(s, "%pI6", + &server->addr.sockAddr6.sin6_addr); + break; + case AF_INET: + seq_printf(s, "%pI4", + &server->addr.sockAddr.sin_addr.s_addr); + break; + } + } } if ((cifs_sb->mnt_cifs_flags & CIFS_MOUNT_OVERR_UID) || - !(cifs_sb->tcon->unix_ext)) + !(tcon->unix_ext)) seq_printf(s, ",uid=%d", cifs_sb->mnt_uid); if ((cifs_sb->mnt_cifs_flags & CIFS_MOUNT_OVERR_GID) || - !(cifs_sb->tcon->unix_ext)) + !(tcon->unix_ext)) seq_printf(s, ",gid=%d", cifs_sb->mnt_gid); - if (!cifs_sb->tcon->unix_ext) { + if (!tcon->unix_ext) { seq_printf(s, ",file_mode=0%o,dir_mode=0%o", cifs_sb->mnt_file_mode, cifs_sb->mnt_dir_mode); } - if (cifs_sb->tcon->seal) + if (tcon->seal) seq_printf(s, ",seal"); - if (cifs_sb->tcon->nocase) + if (tcon->nocase) seq_printf(s, ",nocase"); - if (cifs_sb->tcon->retry) + if (tcon->retry) seq_printf(s, ",hard"); } + if (cifs_sb->prepath) + seq_printf(s, ",prepath=%s", cifs_sb->prepath); if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_POSIX_PATHS) seq_printf(s, ",posixpaths"); if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_SET_UID) @@ -413,9 +438,8 @@ int cifs_xquota_set(struct super_block *sb, int quota_type, qid_t qid, xid = GetXid(); if (pTcon) { cFYI(1, ("set type: 0x%x id: %d", quota_type, qid)); - } else { + } else rc = -EIO; - } FreeXid(xid); return rc; @@ -437,9 +461,8 @@ int cifs_xquota_get(struct super_block *sb, int quota_type, qid_t qid, xid = GetXid(); if (pTcon) { cFYI(1, ("set type: 0x%x id: %d", quota_type, qid)); - } else { + } else rc = -EIO; - } FreeXid(xid); return rc; @@ -460,9 +483,8 @@ int cifs_xstate_set(struct super_block *sb, unsigned int flags, int operation) xid = GetXid(); if (pTcon) { cFYI(1, ("flags: 0x%x operation: 0x%x", flags, operation)); - } else { + } else rc = -EIO; - } FreeXid(xid); return rc; @@ -475,17 +497,16 @@ int cifs_xstate_get(struct super_block *sb, struct fs_quota_stat *qstats) struct cifs_sb_info *cifs_sb = CIFS_SB(sb); struct cifsTconInfo *pTcon; - if (cifs_sb) { + if (cifs_sb) pTcon = cifs_sb->tcon; - } else { + else return -EIO; - } + xid = GetXid(); if (pTcon) { cFYI(1, ("pqstats %p", qstats)); - } else { + } else rc = -EIO; - } FreeXid(xid); return rc; @@ -510,10 +531,11 @@ static void cifs_umount_begin(struct super_block *sb) tcon = cifs_sb->tcon; if (tcon == NULL) return; - down(&tcon->tconSem); - if (atomic_read(&tcon->useCount) == 1) + + read_lock(&cifs_tcp_ses_lock); + if (tcon->tc_count == 1) tcon->tidStatus = CifsExiting; - up(&tcon->tconSem); + read_unlock(&cifs_tcp_ses_lock); /* cancel_brl_requests(tcon); */ /* BB mark all brl mids as exiting */ /* cancel_notify_requests(tcon); */ @@ -617,6 +639,37 @@ static loff_t cifs_llseek(struct file *file, loff_t offset, int origin) return generic_file_llseek_unlocked(file, offset, origin); } +#ifdef CONFIG_CIFS_EXPERIMENTAL +static int cifs_setlease(struct file *file, long arg, struct file_lock **lease) +{ + /* note that this is called by vfs setlease with the BKL held + although I doubt that BKL is needed here in cifs */ + struct inode *inode = file->f_path.dentry->d_inode; + + if (!(S_ISREG(inode->i_mode))) + return -EINVAL; + + /* check if file is oplocked */ + if (((arg == F_RDLCK) && + (CIFS_I(inode)->clientCanCacheRead)) || + ((arg == F_WRLCK) && + (CIFS_I(inode)->clientCanCacheAll))) + return generic_setlease(file, arg, lease); + else if (CIFS_SB(inode->i_sb)->tcon->local_lease && + !CIFS_I(inode)->clientCanCacheRead) + /* If the server claims to support oplock on this + file, then we still need to check oplock even + if the local_lease mount option is set, but there + are servers which do not support oplock for which + this mount option may be useful if the user + knows that the file won't be changed on the server + by anyone else */ + return generic_setlease(file, arg, lease); + else + return -EAGAIN; +} +#endif + struct file_system_type cifs_fs_type = { .owner = THIS_MODULE, .name = "cifs", @@ -695,6 +748,7 @@ const struct file_operations cifs_file_ops = { #ifdef CONFIG_CIFS_EXPERIMENTAL .dir_notify = cifs_dir_notify, + .setlease = cifs_setlease, #endif /* CONFIG_CIFS_EXPERIMENTAL */ }; @@ -715,6 +769,7 @@ const struct file_operations cifs_file_direct_ops = { .llseek = cifs_llseek, #ifdef CONFIG_CIFS_EXPERIMENTAL .dir_notify = cifs_dir_notify, + .setlease = cifs_setlease, #endif /* CONFIG_CIFS_EXPERIMENTAL */ }; const struct file_operations cifs_file_nobrl_ops = { @@ -735,6 +790,7 @@ const struct file_operations cifs_file_nobrl_ops = { #ifdef CONFIG_CIFS_EXPERIMENTAL .dir_notify = cifs_dir_notify, + .setlease = cifs_setlease, #endif /* CONFIG_CIFS_EXPERIMENTAL */ }; @@ -754,6 +810,7 @@ const struct file_operations cifs_file_direct_nobrl_ops = { .llseek = cifs_llseek, #ifdef CONFIG_CIFS_EXPERIMENTAL .dir_notify = cifs_dir_notify, + .setlease = cifs_setlease, #endif /* CONFIG_CIFS_EXPERIMENTAL */ }; @@ -765,6 +822,7 @@ const struct file_operations cifs_dir_ops = { .dir_notify = cifs_dir_notify, #endif /* CONFIG_CIFS_EXPERIMENTAL */ .unlocked_ioctl = cifs_ioctl, + .llseek = generic_file_llseek, }; static void @@ -945,6 +1003,12 @@ static int cifs_oplock_thread(void *dummyarg) the call */ /* mutex_lock(&inode->i_mutex);*/ if (S_ISREG(inode->i_mode)) { +#ifdef CONFIG_CIFS_EXPERIMENTAL + if (CIFS_I(inode)->clientCanCacheAll == 0) + break_lease(inode, FMODE_READ); + else if (CIFS_I(inode)->clientCanCacheRead == 0) + break_lease(inode, FMODE_WRITE); +#endif rc = filemap_fdatawrite(inode->i_mapping); if (CIFS_I(inode)->clientCanCacheRead == 0) { waitrc = filemap_fdatawait( @@ -967,7 +1031,7 @@ static int cifs_oplock_thread(void *dummyarg) not bother sending an oplock release if session to server still is disconnected since oplock already released by the server in that case */ - if (pTcon->tidStatus != CifsNeedReconnect) { + if (!pTcon->need_reconnect) { rc = CIFSSMBLock(0, pTcon, netfid, 0 /* len */ , 0 /* offset */, 0, 0, LOCKING_ANDX_OPLOCK_RELEASE, @@ -982,40 +1046,40 @@ static int cifs_oplock_thread(void *dummyarg) return 0; } +#ifdef CONFIG_CIFS_EXPERIMENTAL static int cifs_dnotify_thread(void *dummyarg) { struct list_head *tmp; - struct cifsSesInfo *ses; + struct TCP_Server_Info *server; do { if (try_to_freeze()) continue; set_current_state(TASK_INTERRUPTIBLE); schedule_timeout(15*HZ); - read_lock(&GlobalSMBSeslock); /* check if any stuck requests that need to be woken up and wakeq so the thread can wake up and error out */ - list_for_each(tmp, &GlobalSMBSessionList) { - ses = list_entry(tmp, struct cifsSesInfo, - cifsSessionList); - if (ses->server && atomic_read(&ses->server->inFlight)) - wake_up_all(&ses->server->response_q); + read_lock(&cifs_tcp_ses_lock); + list_for_each(tmp, &cifs_tcp_ses_list) { + server = list_entry(tmp, struct TCP_Server_Info, + tcp_ses_list); + if (atomic_read(&server->inFlight)) + wake_up_all(&server->response_q); } - read_unlock(&GlobalSMBSeslock); + read_unlock(&cifs_tcp_ses_lock); } while (!kthread_should_stop()); return 0; } +#endif static int __init init_cifs(void) { int rc = 0; cifs_proc_init(); -/* INIT_LIST_HEAD(&GlobalServerList);*/ /* BB not implemented yet */ - INIT_LIST_HEAD(&GlobalSMBSessionList); - INIT_LIST_HEAD(&GlobalTreeConnectionList); + INIT_LIST_HEAD(&cifs_tcp_ses_list); INIT_LIST_HEAD(&GlobalOplock_Q); #ifdef CONFIG_CIFS_EXPERIMENTAL INIT_LIST_HEAD(&GlobalDnotifyReqList); @@ -1043,6 +1107,7 @@ init_cifs(void) GlobalMaxActiveXid = 0; memset(Local_System_Name, 0, 15); rwlock_init(&GlobalSMBSeslock); + rwlock_init(&cifs_tcp_ses_lock); spin_lock_init(&GlobalMid_Lock); if (cifs_max_pending < 2) { @@ -1085,16 +1150,20 @@ init_cifs(void) goto out_unregister_dfs_key_type; } +#ifdef CONFIG_CIFS_EXPERIMENTAL dnotifyThread = kthread_run(cifs_dnotify_thread, NULL, "cifsdnotifyd"); if (IS_ERR(dnotifyThread)) { rc = PTR_ERR(dnotifyThread); cERROR(1, ("error %d create dnotify thread", rc)); goto out_stop_oplock_thread; } +#endif return 0; +#ifdef CONFIG_CIFS_EXPERIMENTAL out_stop_oplock_thread: +#endif kthread_stop(oplockThread); out_unregister_dfs_key_type: #ifdef CONFIG_CIFS_DFS_UPCALL @@ -1133,8 +1202,10 @@ exit_cifs(void) cifs_destroy_inodecache(); cifs_destroy_mids(); cifs_destroy_request_bufs(); - kthread_stop(oplockThread); +#ifdef CONFIG_CIFS_EXPERIMENTAL kthread_stop(dnotifyThread); +#endif + kthread_stop(oplockThread); } MODULE_AUTHOR("Steve French <sfrench@us.ibm.com>"); diff --git a/fs/cifs/cifsfs.h b/fs/cifs/cifsfs.h index f7b4a5cd837..2ce04c73d74 100644 --- a/fs/cifs/cifsfs.h +++ b/fs/cifs/cifsfs.h @@ -101,5 +101,5 @@ extern long cifs_ioctl(struct file *filep, unsigned int cmd, unsigned long arg); extern const struct export_operations cifs_export_ops; #endif /* EXPERIMENTAL */ -#define CIFS_VERSION "1.54" +#define CIFS_VERSION "1.56" #endif /* _CIFSFS_H */ diff --git a/fs/cifs/cifsglob.h b/fs/cifs/cifsglob.h index 0d22479d99b..94c1ca0ec95 100644 --- a/fs/cifs/cifsglob.h +++ b/fs/cifs/cifsglob.h @@ -47,7 +47,11 @@ */ #define CIFS_MAX_REQ 50 -#define SERVER_NAME_LENGTH 15 +#define RFC1001_NAME_LEN 15 +#define RFC1001_NAME_LEN_WITH_NULL (RFC1001_NAME_LEN + 1) + +/* currently length of NIP6_FMT */ +#define SERVER_NAME_LENGTH 40 #define SERVER_NAME_LEN_WITH_NULL (SERVER_NAME_LENGTH + 1) /* used to define string lengths for reversing unicode strings */ @@ -85,8 +89,7 @@ enum securityEnum { }; enum protocolEnum { - IPV4 = 0, - IPV6, + TCP = 0, SCTP /* Netbios frames protocol not supported at this time */ }; @@ -122,9 +125,11 @@ struct cifs_cred { */ struct TCP_Server_Info { + struct list_head tcp_ses_list; + struct list_head smb_ses_list; + int srv_count; /* reference counter */ /* 15 character server name + 0x20 16th byte indicating type = srv */ - char server_RFC1001_name[SERVER_NAME_LEN_WITH_NULL]; - char unicode_server_Name[SERVER_NAME_LEN_WITH_NULL * 2]; + char server_RFC1001_name[RFC1001_NAME_LEN_WITH_NULL]; char *hostname; /* hostname portion of UNC string */ struct socket *ssocket; union { @@ -141,14 +146,15 @@ struct TCP_Server_Info { char versionMajor; char versionMinor; bool svlocal:1; /* local server or remote */ - atomic_t socketUseCount; /* number of open cifs sessions on socket */ + bool noblocksnd; /* use blocking sendmsg */ + bool noautotune; /* do not autotune send buf sizes */ atomic_t inFlight; /* number of requests on the wire to server */ #ifdef CONFIG_CIFS_STATS2 atomic_t inSend; /* requests trying to send */ atomic_t num_waiters; /* blocked waiting to get in sendrecv */ #endif enum statusEnum tcpStatus; /* what we think the status is */ - struct semaphore tcpSem; + struct mutex srv_mutex; struct task_struct *tsk; char server_GUID[16]; char secMode; @@ -168,7 +174,7 @@ struct TCP_Server_Info { __u16 CurrentMid; /* multiplex id - rotating counter */ char cryptKey[CIFS_CRYPTO_KEY_SIZE]; /* 16th byte of RFC1001 workstation name is always null */ - char workstation_RFC1001_name[SERVER_NAME_LEN_WITH_NULL]; + char workstation_RFC1001_name[RFC1001_NAME_LEN_WITH_NULL]; __u32 sequence_number; /* needed for CIFS PDU signature */ struct mac_key mac_signing_key; char ntlmv2_hash[16]; @@ -192,13 +198,14 @@ struct cifsUidInfo { * Session structure. One of these for each uid session with a particular host */ struct cifsSesInfo { - struct list_head cifsSessionList; + struct list_head smb_ses_list; + struct list_head tcon_list; struct semaphore sesSem; #if 0 struct cifsUidInfo *uidInfo; /* pointer to user info */ #endif struct TCP_Server_Info *server; /* pointer to server info */ - atomic_t inUse; /* # of mounts (tree connections) on this ses */ + int ses_count; /* reference counter */ enum statusEnum status; unsigned overrideSecFlg; /* if non-zero override global sec flags */ __u16 ipc_tid; /* special tid for connection to IPC share */ @@ -214,6 +221,7 @@ struct cifsSesInfo { char userName[MAX_USERNAME_SIZE + 1]; char *domainName; char *password; + bool need_reconnect:1; /* connection reset, uid now invalid */ }; /* no more than one of the following three session flags may be set */ #define CIFS_SES_NT4 1 @@ -228,16 +236,16 @@ struct cifsSesInfo { * session */ struct cifsTconInfo { - struct list_head cifsConnectionList; + struct list_head tcon_list; + int tc_count; struct list_head openFileList; - struct semaphore tconSem; struct cifsSesInfo *ses; /* pointer to session associated with */ char treeName[MAX_TREE_SIZE + 1]; /* UNC name of resource in ASCII */ char *nativeFileSystem; + char *password; /* for share-level security */ __u16 tid; /* The 2 byte tree id */ __u16 Flags; /* optional support bits */ enum statusEnum tidStatus; - atomic_t useCount; /* how many explicit/implicit mounts to share */ #ifdef CONFIG_CIFS_STATS atomic_t num_smbs_sent; atomic_t num_writes; @@ -285,6 +293,8 @@ struct cifsTconInfo { bool seal:1; /* transport encryption for this mounted share */ bool unix_ext:1; /* if false disable Linux extensions to CIFS protocol for this mount even if server would support */ + bool local_lease:1; /* check leases (only) on local system not remote */ + bool need_reconnect:1; /* connection reset, tid now invalid */ /* BB add field for back pointer to sb struct(s)? */ }; @@ -353,6 +363,7 @@ struct cifsInodeInfo { bool clientCanCacheRead:1; /* read oplock */ bool clientCanCacheAll:1; /* read and writebehind oplock */ bool oplockPending:1; + bool delete_pending:1; /* DELETE_ON_CLOSE is set */ struct inode vfs_inode; }; @@ -415,7 +426,6 @@ struct mid_q_entry { unsigned long when_sent; /* time when smb send finished */ unsigned long when_received; /* when demux complete (taken off wire) */ #endif - struct cifsSesInfo *ses; /* smb was sent to this server */ struct task_struct *tsk; /* task waiting for response */ struct smb_hdr *resp_buf; /* response buffer */ int midState; /* wish this were enum but can not pass to wait_event */ @@ -584,22 +594,30 @@ require use of the stronger protocol */ #endif /* - * The list of servers that did not respond with NT LM 0.12. - * This list helps improve performance and eliminate the messages indicating - * that we had a communications error talking to the server in this list. + * the list of TCP_Server_Info structures, ie each of the sockets + * connecting our client to a distinct server (ip address), is + * chained together by cifs_tcp_ses_list. The list of all our SMB + * sessions (and from that the tree connections) can be found + * by iterating over cifs_tcp_ses_list */ -/* Feature not supported */ -/* GLOBAL_EXTERN struct servers_not_supported *NotSuppList; */ +GLOBAL_EXTERN struct list_head cifs_tcp_ses_list; /* - * The following is a hash table of all the users we know about. + * This lock protects the cifs_tcp_ses_list, the list of smb sessions per + * tcp session, and the list of tcon's per smb session. It also protects + * the reference counters for the server, smb session, and tcon. Finally, + * changes to the tcon->tidStatus should be done while holding this lock. */ -GLOBAL_EXTERN struct smbUidInfo *GlobalUidList[UID_HASH]; +GLOBAL_EXTERN rwlock_t cifs_tcp_ses_lock; -/* GLOBAL_EXTERN struct list_head GlobalServerList; BB not implemented yet */ -GLOBAL_EXTERN struct list_head GlobalSMBSessionList; -GLOBAL_EXTERN struct list_head GlobalTreeConnectionList; -GLOBAL_EXTERN rwlock_t GlobalSMBSeslock; /* protects list inserts on 3 above */ +/* + * This lock protects the cifs_file->llist and cifs_file->flist + * list operations, and updates to some flags (cifs_file->invalidHandle) + * It will be moved to either use the tcon->stat_lock or equivalent later. + * If cifs_tcp_ses_lock and the lock below are both needed to be held, then + * the cifs_tcp_ses_lock must be grabbed first and released last. + */ +GLOBAL_EXTERN rwlock_t GlobalSMBSeslock; GLOBAL_EXTERN struct list_head GlobalOplock_Q; diff --git a/fs/cifs/cifspdu.h b/fs/cifs/cifspdu.h index d2a073edd1b..b4e2e9f0ee3 100644 --- a/fs/cifs/cifspdu.h +++ b/fs/cifs/cifspdu.h @@ -1922,7 +1922,7 @@ typedef struct smb_com_transaction2_get_dfs_refer_req { /* DFS server target type */ #define DFS_TYPE_LINK 0x0000 /* also for sysvol targets */ #define DFS_TYPE_ROOT 0x0001 - + /* Referral Entry Flags */ #define DFS_NAME_LIST_REF 0x0200 diff --git a/fs/cifs/cifsproto.h b/fs/cifs/cifsproto.h index 0cff7fe986e..06f6779988b 100644 --- a/fs/cifs/cifsproto.h +++ b/fs/cifs/cifsproto.h @@ -36,10 +36,10 @@ extern void cifs_buf_release(void *); extern struct smb_hdr *cifs_small_buf_get(void); extern void cifs_small_buf_release(void *); extern int smb_send(struct socket *, struct smb_hdr *, - unsigned int /* length */ , struct sockaddr *); + unsigned int /* length */ , struct sockaddr *, bool); extern unsigned int _GetXid(void); extern void _FreeXid(unsigned int); -#define GetXid() (int)_GetXid(); cFYI(1,("CIFS VFS: in %s as Xid: %d with uid: %d",__func__, xid,current->fsuid)); +#define GetXid() (int)_GetXid(); cFYI(1,("CIFS VFS: in %s as Xid: %d with uid: %d",__func__, xid,current_fsuid())); #define FreeXid(curr_xid) {_FreeXid(curr_xid); cFYI(1,("CIFS VFS: leaving %s (xid = %d) rc = %d",__func__,curr_xid,(int)rc));} extern char *build_path_from_dentry(struct dentry *); extern char *build_wildcard_path_from_dentry(struct dentry *direntry); @@ -330,7 +330,8 @@ extern void CalcNTLMv2_response(const struct cifsSesInfo *, char *); extern void setup_ntlmv2_rsp(struct cifsSesInfo *, char *, const struct nls_table *); #ifdef CONFIG_CIFS_WEAK_PW_HASH -extern void calc_lanman_hash(struct cifsSesInfo *ses, char *lnm_session_key); +extern void calc_lanman_hash(const char *password, const char *cryptkey, + bool encrypt, char *lnm_session_key); #endif /* CIFS_WEAK_PW_HASH */ extern int CIFSSMBCopy(int xid, struct cifsTconInfo *source_tcon, diff --git a/fs/cifs/cifssmb.c b/fs/cifs/cifssmb.c index 6f4ffe15d68..552642a507c 100644 --- a/fs/cifs/cifssmb.c +++ b/fs/cifs/cifssmb.c @@ -190,10 +190,10 @@ small_smb_init(int smb_command, int wct, struct cifsTconInfo *tcon, /* need to prevent multiple threads trying to simultaneously reconnect the same SMB session */ down(&tcon->ses->sesSem); - if (tcon->ses->status == CifsNeedReconnect) + if (tcon->ses->need_reconnect) rc = cifs_setup_session(0, tcon->ses, nls_codepage); - if (!rc && (tcon->tidStatus == CifsNeedReconnect)) { + if (!rc && (tcon->need_reconnect)) { mark_open_files_invalid(tcon); rc = CIFSTCon(0, tcon->ses, tcon->treeName, tcon, nls_codepage); @@ -337,10 +337,10 @@ smb_init(int smb_command, int wct, struct cifsTconInfo *tcon, /* need to prevent multiple threads trying to simultaneously reconnect the same SMB session */ down(&tcon->ses->sesSem); - if (tcon->ses->status == CifsNeedReconnect) + if (tcon->ses->need_reconnect) rc = cifs_setup_session(0, tcon->ses, nls_codepage); - if (!rc && (tcon->tidStatus == CifsNeedReconnect)) { + if (!rc && (tcon->need_reconnect)) { mark_open_files_invalid(tcon); rc = CIFSTCon(0, tcon->ses, tcon->treeName, tcon, nls_codepage); @@ -664,8 +664,9 @@ CIFSSMBNegotiate(unsigned int xid, struct cifsSesInfo *ses) rc = -EIO; goto neg_err_exit; } - - if (server->socketUseCount.counter > 1) { + read_lock(&cifs_tcp_ses_lock); + if (server->srv_count > 1) { + read_unlock(&cifs_tcp_ses_lock); if (memcmp(server->server_GUID, pSMBr->u.extended_response. GUID, 16) != 0) { @@ -674,9 +675,11 @@ CIFSSMBNegotiate(unsigned int xid, struct cifsSesInfo *ses) pSMBr->u.extended_response.GUID, 16); } - } else + } else { + read_unlock(&cifs_tcp_ses_lock); memcpy(server->server_GUID, pSMBr->u.extended_response.GUID, 16); + } if (count == 16) { server->secType = RawNTLMSSP; @@ -739,50 +742,31 @@ CIFSSMBTDis(const int xid, struct cifsTconInfo *tcon) int rc = 0; cFYI(1, ("In tree disconnect")); - /* - * If last user of the connection and - * connection alive - disconnect it - * If this is the last connection on the server session disconnect it - * (and inside session disconnect we should check if tcp socket needs - * to be freed and kernel thread woken up). - */ - if (tcon) - down(&tcon->tconSem); - else - return -EIO; - atomic_dec(&tcon->useCount); - if (atomic_read(&tcon->useCount) > 0) { - up(&tcon->tconSem); - return -EBUSY; - } + /* BB: do we need to check this? These should never be NULL. */ + if ((tcon->ses == NULL) || (tcon->ses->server == NULL)) + return -EIO; - /* No need to return error on this operation if tid invalidated and - closed on server already e.g. due to tcp session crashing */ - if (tcon->tidStatus == CifsNeedReconnect) { - up(&tcon->tconSem); + /* + * No need to return error on this operation if tid invalidated and + * closed on server already e.g. due to tcp session crashing. Also, + * the tcon is no longer on the list, so no need to take lock before + * checking this. + */ + if (tcon->need_reconnect) return 0; - } - if ((tcon->ses == NULL) || (tcon->ses->server == NULL)) { - up(&tcon->tconSem); - return -EIO; - } rc = small_smb_init(SMB_COM_TREE_DISCONNECT, 0, tcon, (void **)&smb_buffer); - if (rc) { - up(&tcon->tconSem); + if (rc) return rc; - } rc = SendReceiveNoRsp(xid, tcon->ses, smb_buffer, 0); if (rc) cFYI(1, ("Tree disconnect failed %d", rc)); - up(&tcon->tconSem); - /* No need to return error on this operation if tid invalidated and - closed on server already e.g. due to tcp session crashing */ + closed on server already e.g. due to tcp session crashing */ if (rc == -EAGAIN) rc = 0; @@ -796,43 +780,36 @@ CIFSSMBLogoff(const int xid, struct cifsSesInfo *ses) int rc = 0; cFYI(1, ("In SMBLogoff for session disconnect")); - if (ses) - down(&ses->sesSem); - else + + /* + * BB: do we need to check validity of ses and server? They should + * always be valid since we have an active reference. If not, that + * should probably be a BUG() + */ + if (!ses || !ses->server) return -EIO; - atomic_dec(&ses->inUse); - if (atomic_read(&ses->inUse) > 0) { - up(&ses->sesSem); - return -EBUSY; - } + down(&ses->sesSem); + if (ses->need_reconnect) + goto session_already_dead; /* no need to send SMBlogoff if uid + already closed due to reconnect */ rc = small_smb_init(SMB_COM_LOGOFF_ANDX, 2, NULL, (void **)&pSMB); if (rc) { up(&ses->sesSem); return rc; } - if (ses->server) { - pSMB->hdr.Mid = GetNextMid(ses->server); + pSMB->hdr.Mid = GetNextMid(ses->server); - if (ses->server->secMode & + if (ses->server->secMode & (SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED)) pSMB->hdr.Flags2 |= SMBFLG2_SECURITY_SIGNATURE; - } pSMB->hdr.Uid = ses->Suid; pSMB->AndXCommand = 0xFF; rc = SendReceiveNoRsp(xid, ses, (struct smb_hdr *) pSMB, 0); - if (ses->server) { - atomic_dec(&ses->server->socketUseCount); - if (atomic_read(&ses->server->socketUseCount) == 0) { - spin_lock(&GlobalMid_Lock); - ses->server->tcpStatus = CifsExiting; - spin_unlock(&GlobalMid_Lock); - rc = -ESHUTDOWN; - } - } +session_already_dead: up(&ses->sesSem); /* if session dead then we do not need to do ulogoff, @@ -1309,6 +1286,7 @@ OldOpenRetry: cpu_to_le64(le32_to_cpu(pSMBr->EndOfFile)); pfile_info->EndOfFile = pfile_info->AllocationSize; pfile_info->NumberOfLinks = cpu_to_le32(1); + pfile_info->DeletePending = 0; } } @@ -1404,12 +1382,13 @@ openRetry: if (cpu_to_le32(FILE_CREATE) == pSMBr->CreateAction) *pOplock |= CIFS_CREATE_ACTION; if (pfile_info) { - memcpy((char *)pfile_info, (char *)&pSMBr->CreationTime, - 36 /* CreationTime to Attributes */); - /* the file_info buf is endian converted by caller */ - pfile_info->AllocationSize = pSMBr->AllocationSize; - pfile_info->EndOfFile = pSMBr->EndOfFile; - pfile_info->NumberOfLinks = cpu_to_le32(1); + memcpy((char *)pfile_info, (char *)&pSMBr->CreationTime, + 36 /* CreationTime to Attributes */); + /* the file_info buf is endian converted by caller */ + pfile_info->AllocationSize = pSMBr->AllocationSize; + pfile_info->EndOfFile = pSMBr->EndOfFile; + pfile_info->NumberOfLinks = cpu_to_le32(1); + pfile_info->DeletePending = 0; } } @@ -1435,8 +1414,13 @@ CIFSSMBRead(const int xid, struct cifsTconInfo *tcon, const int netfid, cFYI(1, ("Reading %d bytes on fid %d", count, netfid)); if (tcon->ses->capabilities & CAP_LARGE_FILES) wct = 12; - else + else { wct = 10; /* old style read */ + if ((lseek >> 32) > 0) { + /* can not handle this big offset for old */ + return -EIO; + } + } *nbytes = 0; rc = small_smb_init(SMB_COM_READ_ANDX, wct, tcon, (void **) &pSMB); @@ -1452,8 +1436,6 @@ CIFSSMBRead(const int xid, struct cifsTconInfo *tcon, const int netfid, pSMB->OffsetLow = cpu_to_le32(lseek & 0xFFFFFFFF); if (wct == 12) pSMB->OffsetHigh = cpu_to_le32(lseek >> 32); - else if ((lseek >> 32) > 0) /* can not handle this big offset for old */ - return -EIO; pSMB->Remaining = 0; pSMB->MaxCount = cpu_to_le16(count & 0xFFFF); @@ -1534,14 +1516,19 @@ CIFSSMBWrite(const int xid, struct cifsTconInfo *tcon, __u32 bytes_sent; __u16 byte_count; - /* cFYI(1,("write at %lld %d bytes",offset,count));*/ + /* cFYI(1, ("write at %lld %d bytes", offset, count));*/ if (tcon->ses == NULL) return -ECONNABORTED; if (tcon->ses->capabilities & CAP_LARGE_FILES) wct = 14; - else + else { wct = 12; + if ((offset >> 32) > 0) { + /* can not handle big offset for old srv */ + return -EIO; + } + } rc = smb_init(SMB_COM_WRITE_ANDX, wct, tcon, (void **) &pSMB, (void **) &pSMBr); @@ -1556,8 +1543,6 @@ CIFSSMBWrite(const int xid, struct cifsTconInfo *tcon, pSMB->OffsetLow = cpu_to_le32(offset & 0xFFFFFFFF); if (wct == 14) pSMB->OffsetHigh = cpu_to_le32(offset >> 32); - else if ((offset >> 32) > 0) /* can not handle big offset for old srv */ - return -EIO; pSMB->Reserved = 0xFFFFFFFF; pSMB->WriteMode = 0; @@ -1579,7 +1564,7 @@ CIFSSMBWrite(const int xid, struct cifsTconInfo *tcon, pSMB->DataOffset = cpu_to_le16(offsetof(struct smb_com_write_req, Data) - 4); if (buf) - memcpy(pSMB->Data, buf, bytes_sent); + memcpy(pSMB->Data, buf, bytes_sent); else if (ubuf) { if (copy_from_user(pSMB->Data, ubuf, bytes_sent)) { cifs_buf_release(pSMB); @@ -1642,10 +1627,15 @@ CIFSSMBWrite2(const int xid, struct cifsTconInfo *tcon, cFYI(1, ("write2 at %lld %d bytes", (long long)offset, count)); - if (tcon->ses->capabilities & CAP_LARGE_FILES) + if (tcon->ses->capabilities & CAP_LARGE_FILES) { wct = 14; - else + } else { wct = 12; + if ((offset >> 32) > 0) { + /* can not handle big offset for old srv */ + return -EIO; + } + } rc = small_smb_init(SMB_COM_WRITE_ANDX, wct, tcon, (void **) &pSMB); if (rc) return rc; @@ -1658,8 +1648,6 @@ CIFSSMBWrite2(const int xid, struct cifsTconInfo *tcon, pSMB->OffsetLow = cpu_to_le32(offset & 0xFFFFFFFF); if (wct == 14) pSMB->OffsetHigh = cpu_to_le32(offset >> 32); - else if ((offset >> 32) > 0) /* can not handle big offset for old srv */ - return -EIO; pSMB->Reserved = 0xFFFFFFFF; pSMB->WriteMode = 0; pSMB->Remaining = 0; @@ -1883,10 +1871,6 @@ CIFSSMBPosixLock(const int xid, struct cifsTconInfo *tcon, rc = -EIO; /* bad smb */ goto plk_err_exit; } - if (pLockData == NULL) { - rc = -EINVAL; - goto plk_err_exit; - } data_offset = le16_to_cpu(pSMBr->t2.DataOffset); data_count = le16_to_cpu(pSMBr->t2.DataCount); if (data_count < sizeof(struct cifs_posix_lock)) { @@ -3920,6 +3904,27 @@ GetInodeNumOut: return rc; } +/* computes length of UCS string converted to host codepage + * @src: UCS string + * @maxlen: length of the input string in UCS characters + * (not in bytes) + * + * return: size of input string in host codepage + */ +static int hostlen_fromUCS(const __le16 *src, const int maxlen, + const struct nls_table *nls_codepage) { + int i; + int hostlen = 0; + char to[4]; + int charlen; + for (i = 0; (i < maxlen) && src[i]; ++i) { + charlen = nls_codepage->uni2char(le16_to_cpu(src[i]), + to, NLS_MAX_CHARSET_SIZE); + hostlen += charlen > 0 ? charlen : 1; + } + return hostlen; +} + /* parses DFS refferal V3 structure * caller is responsible for freeing target_nodes * returns: @@ -3930,7 +3935,8 @@ static int parse_DFS_referrals(TRANSACTION2_GET_DFS_REFER_RSP *pSMBr, unsigned int *num_of_nodes, struct dfs_info3_param **target_nodes, - const struct nls_table *nls_codepage) + const struct nls_table *nls_codepage, int remap, + const char *searchName) { int i, rc = 0; char *data_end; @@ -3981,7 +3987,18 @@ parse_DFS_referrals(TRANSACTION2_GET_DFS_REFER_RSP *pSMBr, struct dfs_info3_param *node = (*target_nodes)+i; node->flags = le16_to_cpu(pSMBr->DFSFlags); - node->path_consumed = le16_to_cpu(pSMBr->PathConsumed); + if (is_unicode) { + __le16 *tmp = kmalloc(strlen(searchName)*2 + 2, + GFP_KERNEL); + cifsConvertToUCS((__le16 *) tmp, searchName, + PATH_MAX, nls_codepage, remap); + node->path_consumed = hostlen_fromUCS(tmp, + le16_to_cpu(pSMBr->PathConsumed)/2, + nls_codepage); + kfree(tmp); + } else + node->path_consumed = le16_to_cpu(pSMBr->PathConsumed); + node->server_type = le16_to_cpu(ref->ServerType); node->ref_flag = le16_to_cpu(ref->ReferralEntryFlags); @@ -4114,7 +4131,8 @@ getDFSRetry: /* parse returned result into more usable form */ rc = parse_DFS_referrals(pSMBr, num_of_nodes, - target_nodes, nls_codepage); + target_nodes, nls_codepage, remap, + searchName); GetDFSRefExit: cifs_buf_release(pSMB); diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index 4c13bcdb92a..e9ea394ee07 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -89,7 +89,12 @@ struct smb_vol { bool nullauth:1; /* attempt to authenticate with null user */ bool nocase:1; /* request case insensitive filenames */ bool nobrl:1; /* disable sending byte range locks to srv */ + bool mand_lock:1; /* send mandatory not posix byte range lock reqs */ bool seal:1; /* request transport encryption on share */ + bool nodfs:1; /* Do not request DFS, even if available */ + bool local_lease:1; /* check leases only on local system, not remote */ + bool noblocksnd:1; + bool noautotune:1; unsigned int rsize; unsigned int wsize; unsigned int sockopt; @@ -97,34 +102,28 @@ struct smb_vol { char *prepath; }; -static int ipv4_connect(struct sockaddr_in *psin_server, - struct socket **csocket, - char *netb_name, - char *server_netb_name); -static int ipv6_connect(struct sockaddr_in6 *psin_server, - struct socket **csocket); - - - /* - * cifs tcp session reconnection - * - * mark tcp session as reconnecting so temporarily locked - * mark all smb sessions as reconnecting for tcp session - * reconnect tcp session - * wake up waiters on reconnection? - (not needed currently) - */ +static int ipv4_connect(struct TCP_Server_Info *server); +static int ipv6_connect(struct TCP_Server_Info *server); +/* + * cifs tcp session reconnection + * + * mark tcp session as reconnecting so temporarily locked + * mark all smb sessions as reconnecting for tcp session + * reconnect tcp session + * wake up waiters on reconnection? - (not needed currently) + */ static int cifs_reconnect(struct TCP_Server_Info *server) { int rc = 0; - struct list_head *tmp; + struct list_head *tmp, *tmp2; struct cifsSesInfo *ses; struct cifsTconInfo *tcon; struct mid_q_entry *mid_entry; spin_lock(&GlobalMid_Lock); - if (kthread_should_stop()) { + if (server->tcpStatus == CifsExiting) { /* the demux thread will exit normally next time through the loop */ spin_unlock(&GlobalMid_Lock); @@ -138,25 +137,19 @@ cifs_reconnect(struct TCP_Server_Info *server) /* before reconnecting the tcp session, mark the smb session (uid) and the tid bad so they are not used until reconnected */ - read_lock(&GlobalSMBSeslock); - list_for_each(tmp, &GlobalSMBSessionList) { - ses = list_entry(tmp, struct cifsSesInfo, cifsSessionList); - if (ses->server) { - if (ses->server == server) { - ses->status = CifsNeedReconnect; - ses->ipc_tid = 0; - } + read_lock(&cifs_tcp_ses_lock); + list_for_each(tmp, &server->smb_ses_list) { + ses = list_entry(tmp, struct cifsSesInfo, smb_ses_list); + ses->need_reconnect = true; + ses->ipc_tid = 0; + list_for_each(tmp2, &ses->tcon_list) { + tcon = list_entry(tmp2, struct cifsTconInfo, tcon_list); + tcon->need_reconnect = true; } - /* else tcp and smb sessions need reconnection */ } - list_for_each(tmp, &GlobalTreeConnectionList) { - tcon = list_entry(tmp, struct cifsTconInfo, cifsConnectionList); - if ((tcon->ses) && (tcon->ses->server == server)) - tcon->tidStatus = CifsNeedReconnect; - } - read_unlock(&GlobalSMBSeslock); + read_unlock(&cifs_tcp_ses_lock); /* do not want to be sending data on a socket we are freeing */ - down(&server->tcpSem); + mutex_lock(&server->srv_mutex); if (server->ssocket) { cFYI(1, ("State: 0x%x Flags: 0x%lx", server->ssocket->state, server->ssocket->flags)); @@ -182,26 +175,22 @@ cifs_reconnect(struct TCP_Server_Info *server) } } spin_unlock(&GlobalMid_Lock); - up(&server->tcpSem); + mutex_unlock(&server->srv_mutex); - while ((!kthread_should_stop()) && (server->tcpStatus != CifsGood)) { + while ((server->tcpStatus != CifsExiting) && + (server->tcpStatus != CifsGood)) { try_to_freeze(); - if (server->protocolType == IPV6) { - rc = ipv6_connect(&server->addr.sockAddr6, - &server->ssocket); - } else { - rc = ipv4_connect(&server->addr.sockAddr, - &server->ssocket, - server->workstation_RFC1001_name, - server->server_RFC1001_name); - } + if (server->addr.sockAddr6.sin6_family == AF_INET6) + rc = ipv6_connect(server); + else + rc = ipv4_connect(server); if (rc) { cFYI(1, ("reconnect error %d", rc)); msleep(3000); } else { atomic_inc(&tcpSesReconnectCount); spin_lock(&GlobalMid_Lock); - if (!kthread_should_stop()) + if (server->tcpStatus != CifsExiting) server->tcpStatus = CifsGood; server->sequence_number = 0; spin_unlock(&GlobalMid_Lock); @@ -356,7 +345,7 @@ cifs_demultiplex_thread(struct TCP_Server_Info *server) GFP_KERNEL); set_freezable(); - while (!kthread_should_stop()) { + while (server->tcpStatus != CifsExiting) { if (try_to_freeze()) continue; if (bigbuf == NULL) { @@ -397,7 +386,7 @@ incomplete_rcv: kernel_recvmsg(csocket, &smb_msg, &iov, 1, pdu_length, 0 /* BB other flags? */); - if (kthread_should_stop()) { + if (server->tcpStatus == CifsExiting) { break; } else if (server->tcpStatus == CifsNeedReconnect) { cFYI(1, ("Reconnect after server stopped responding")); @@ -409,9 +398,14 @@ incomplete_rcv: msleep(1); /* minimum sleep to prevent looping allowing socket to clear and app threads to set tcpStatus CifsNeedReconnect if server hung */ - if (pdu_length < 4) + if (pdu_length < 4) { + iov.iov_base = (4 - pdu_length) + + (char *)smb_buffer; + iov.iov_len = pdu_length; + smb_msg.msg_control = NULL; + smb_msg.msg_controllen = 0; goto incomplete_rcv; - else + } else continue; } else if (length <= 0) { if (server->tcpStatus == CifsNew) { @@ -522,7 +516,7 @@ incomplete_rcv: total_read += length) { length = kernel_recvmsg(csocket, &smb_msg, &iov, 1, pdu_length - total_read, 0); - if (kthread_should_stop() || + if ((server->tcpStatus == CifsExiting) || (length == -EINTR)) { /* then will exit */ reconnect = 2; @@ -646,19 +640,16 @@ multi_t2_fnd: } } /* end while !EXITING */ + /* take it off the list, if it's not already */ + write_lock(&cifs_tcp_ses_lock); + list_del_init(&server->tcp_ses_list); + write_unlock(&cifs_tcp_ses_lock); + spin_lock(&GlobalMid_Lock); server->tcpStatus = CifsExiting; spin_unlock(&GlobalMid_Lock); wake_up_all(&server->response_q); - /* don't exit until kthread_stop is called */ - set_current_state(TASK_UNINTERRUPTIBLE); - while (!kthread_should_stop()) { - schedule(); - set_current_state(TASK_UNINTERRUPTIBLE); - } - set_current_state(TASK_RUNNING); - /* check if we have blocked requests that need to free */ /* Note that cifs_max_pending is normally 50, but can be set at module install time to as little as two */ @@ -686,29 +677,29 @@ multi_t2_fnd: if (smallbuf) /* no sense logging a debug message if NULL */ cifs_small_buf_release(smallbuf); - read_lock(&GlobalSMBSeslock); + /* + * BB: we shouldn't have to do any of this. It shouldn't be + * possible to exit from the thread with active SMB sessions + */ + read_lock(&cifs_tcp_ses_lock); if (list_empty(&server->pending_mid_q)) { /* loop through server session structures attached to this and mark them dead */ - list_for_each(tmp, &GlobalSMBSessionList) { - ses = - list_entry(tmp, struct cifsSesInfo, - cifsSessionList); - if (ses->server == server) { - ses->status = CifsExiting; - ses->server = NULL; - } + list_for_each(tmp, &server->smb_ses_list) { + ses = list_entry(tmp, struct cifsSesInfo, + smb_ses_list); + ses->status = CifsExiting; + ses->server = NULL; } - read_unlock(&GlobalSMBSeslock); + read_unlock(&cifs_tcp_ses_lock); } else { /* although we can not zero the server struct pointer yet, since there are active requests which may depnd on them, mark the corresponding SMB sessions as exiting too */ - list_for_each(tmp, &GlobalSMBSessionList) { + list_for_each(tmp, &server->smb_ses_list) { ses = list_entry(tmp, struct cifsSesInfo, - cifsSessionList); - if (ses->server == server) - ses->status = CifsExiting; + smb_ses_list); + ses->status = CifsExiting; } spin_lock(&GlobalMid_Lock); @@ -723,7 +714,7 @@ multi_t2_fnd: } } spin_unlock(&GlobalMid_Lock); - read_unlock(&GlobalSMBSeslock); + read_unlock(&cifs_tcp_ses_lock); /* 1/8th of sec is more than enough time for them to exit */ msleep(125); } @@ -745,16 +736,16 @@ multi_t2_fnd: if there are any pointing to this (e.g if a crazy root user tried to kill cifsd kernel thread explicitly this might happen) */ - write_lock(&GlobalSMBSeslock); - list_for_each(tmp, &GlobalSMBSessionList) { - ses = list_entry(tmp, struct cifsSesInfo, - cifsSessionList); - if (ses->server == server) - ses->server = NULL; + /* BB: This shouldn't be necessary, see above */ + read_lock(&cifs_tcp_ses_lock); + list_for_each(tmp, &server->smb_ses_list) { + ses = list_entry(tmp, struct cifsSesInfo, smb_ses_list); + ses->server = NULL; } - write_unlock(&GlobalSMBSeslock); + read_unlock(&cifs_tcp_ses_lock); kfree(server->hostname); + task_to_wake = xchg(&server->tsk, NULL); kfree(server); length = atomic_dec_return(&tcpSesAllocCount); @@ -762,7 +753,17 @@ multi_t2_fnd: mempool_resize(cifs_req_poolp, length + cifs_min_rcv, GFP_KERNEL); - return 0; + /* if server->tsk was NULL then wait for a signal before exiting */ + if (!task_to_wake) { + set_current_state(TASK_INTERRUPTIBLE); + while (!signal_pending(current)) { + schedule(); + set_current_state(TASK_INTERRUPTIBLE); + } + set_current_state(TASK_RUNNING); + } + + module_put_and_exit(0); } /* extract the host portion of the UNC string */ @@ -822,8 +823,8 @@ cifs_parse_mount_options(char *options, const char *devname, /* null target name indicates to use *SMBSERVR default called name if we end up sending RFC1001 session initialize */ vol->target_rfc1001_name[0] = 0; - vol->linux_uid = current->uid; /* current->euid instead? */ - vol->linux_gid = current->gid; + vol->linux_uid = current_uid(); /* use current_euid() instead? */ + vol->linux_gid = current_gid(); vol->dir_mode = S_IRWXUGO; /* 2767 perms indicate mandatory locking support */ vol->file_mode = (S_IRWXUGO | S_ISGID) & (~S_IXGRP); @@ -1186,6 +1187,10 @@ cifs_parse_mount_options(char *options, const char *devname, /* ignore */ } else if (strnicmp(data, "rw", 2) == 0) { vol->rw = true; + } else if (strnicmp(data, "noblocksend", 11) == 0) { + vol->noblocksnd = 1; + } else if (strnicmp(data, "noautotune", 10) == 0) { + vol->noautotune = 1; } else if ((strnicmp(data, "suid", 4) == 0) || (strnicmp(data, "nosuid", 6) == 0) || (strnicmp(data, "exec", 4) == 0) || @@ -1218,6 +1223,8 @@ cifs_parse_mount_options(char *options, const char *devname, vol->sfu_emul = 1; } else if (strnicmp(data, "nosfu", 5) == 0) { vol->sfu_emul = 0; + } else if (strnicmp(data, "nodfs", 5) == 0) { + vol->nodfs = 1; } else if (strnicmp(data, "posixpaths", 10) == 0) { vol->posix_paths = 1; } else if (strnicmp(data, "noposixpaths", 12) == 0) { @@ -1240,6 +1247,17 @@ cifs_parse_mount_options(char *options, const char *devname, if (vol->file_mode == (S_IALLUGO & ~(S_ISUID | S_IXGRP))) vol->file_mode = S_IALLUGO; + } else if (strnicmp(data, "forcemandatorylock", 9) == 0) { + /* will take the shorter form "forcemand" as well */ + /* This mount option will force use of mandatory + (DOS/Windows style) byte range locks, instead of + using posix advisory byte range locks, even if the + Unix extensions are available and posix locks would + be supported otherwise. If Unix extensions are not + negotiated this has no effect since mandatory locks + would be used (mandatory locks is all that those + those servers support) */ + vol->mand_lock = 1; } else if (strnicmp(data, "setuids", 7) == 0) { vol->setuids = 1; } else if (strnicmp(data, "nosetuids", 9) == 0) { @@ -1268,6 +1286,10 @@ cifs_parse_mount_options(char *options, const char *devname, vol->no_psx_acl = 0; } else if (strnicmp(data, "noacl", 5) == 0) { vol->no_psx_acl = 1; +#ifdef CONFIG_CIFS_EXPERIMENTAL + } else if (strnicmp(data, "locallease", 6) == 0) { + vol->local_lease = 1; +#endif } else if (strnicmp(data, "sign", 4) == 0) { vol->secFlg |= CIFSSEC_MUST_SIGN; } else if (strnicmp(data, "seal", 4) == 0) { @@ -1331,94 +1353,295 @@ cifs_parse_mount_options(char *options, const char *devname, return 0; } -static struct cifsSesInfo * -cifs_find_tcp_session(struct in_addr *target_ip_addr, - struct in6_addr *target_ip6_addr, - char *userName, struct TCP_Server_Info **psrvTcp) +static struct TCP_Server_Info * +cifs_find_tcp_session(struct sockaddr *addr) { struct list_head *tmp; - struct cifsSesInfo *ses; - - *psrvTcp = NULL; + struct TCP_Server_Info *server; + struct sockaddr_in *addr4 = (struct sockaddr_in *) addr; + struct sockaddr_in6 *addr6 = (struct sockaddr_in6 *) addr; + + write_lock(&cifs_tcp_ses_lock); + list_for_each(tmp, &cifs_tcp_ses_list) { + server = list_entry(tmp, struct TCP_Server_Info, + tcp_ses_list); + /* + * the demux thread can exit on its own while still in CifsNew + * so don't accept any sockets in that state. Since the + * tcpStatus never changes back to CifsNew it's safe to check + * for this without a lock. + */ + if (server->tcpStatus == CifsNew) + continue; - read_lock(&GlobalSMBSeslock); - list_for_each(tmp, &GlobalSMBSessionList) { - ses = list_entry(tmp, struct cifsSesInfo, cifsSessionList); - if (!ses->server) + if (addr->sa_family == AF_INET && + (addr4->sin_addr.s_addr != + server->addr.sockAddr.sin_addr.s_addr)) + continue; + else if (addr->sa_family == AF_INET6 && + memcmp(&server->addr.sockAddr6.sin6_addr, + &addr6->sin6_addr, sizeof(addr6->sin6_addr))) continue; - if (target_ip_addr && - ses->server->addr.sockAddr.sin_addr.s_addr != target_ip_addr->s_addr) - continue; - else if (target_ip6_addr && - memcmp(&ses->server->addr.sockAddr6.sin6_addr, - target_ip6_addr, sizeof(*target_ip6_addr))) - continue; - /* BB lock server and tcp session; increment use count here?? */ + ++server->srv_count; + write_unlock(&cifs_tcp_ses_lock); + cFYI(1, ("Existing tcp session with server found")); + return server; + } + write_unlock(&cifs_tcp_ses_lock); + return NULL; +} - /* found a match on the TCP session */ - *psrvTcp = ses->server; +static void +cifs_put_tcp_session(struct TCP_Server_Info *server) +{ + struct task_struct *task; + + write_lock(&cifs_tcp_ses_lock); + if (--server->srv_count > 0) { + write_unlock(&cifs_tcp_ses_lock); + return; + } + + list_del_init(&server->tcp_ses_list); + write_unlock(&cifs_tcp_ses_lock); + + spin_lock(&GlobalMid_Lock); + server->tcpStatus = CifsExiting; + spin_unlock(&GlobalMid_Lock); - /* BB check if reconnection needed */ - if (strncmp(ses->userName, userName, MAX_USERNAME_SIZE) == 0) { - read_unlock(&GlobalSMBSeslock); - /* Found exact match on both TCP and - SMB sessions */ - return ses; + task = xchg(&server->tsk, NULL); + if (task) + force_sig(SIGKILL, task); +} + +static struct TCP_Server_Info * +cifs_get_tcp_session(struct smb_vol *volume_info) +{ + struct TCP_Server_Info *tcp_ses = NULL; + struct sockaddr addr; + struct sockaddr_in *sin_server = (struct sockaddr_in *) &addr; + struct sockaddr_in6 *sin_server6 = (struct sockaddr_in6 *) &addr; + int rc; + + memset(&addr, 0, sizeof(struct sockaddr)); + + if (volume_info->UNCip && volume_info->UNC) { + rc = cifs_inet_pton(AF_INET, volume_info->UNCip, + &sin_server->sin_addr.s_addr); + + if (rc <= 0) { + /* not ipv4 address, try ipv6 */ + rc = cifs_inet_pton(AF_INET6, volume_info->UNCip, + &sin_server6->sin6_addr.in6_u); + if (rc > 0) + addr.sa_family = AF_INET6; + } else { + addr.sa_family = AF_INET; + } + + if (rc <= 0) { + /* we failed translating address */ + rc = -EINVAL; + goto out_err; } - /* else tcp and smb sessions need reconnection */ + + cFYI(1, ("UNC: %s ip: %s", volume_info->UNC, + volume_info->UNCip)); + } else if (volume_info->UNCip) { + /* BB using ip addr as tcp_ses name to connect to the + DFS root below */ + cERROR(1, ("Connecting to DFS root not implemented yet")); + rc = -EINVAL; + goto out_err; + } else /* which tcp_sess DFS root would we conect to */ { + cERROR(1, + ("CIFS mount error: No UNC path (e.g. -o " + "unc=//192.168.1.100/public) specified")); + rc = -EINVAL; + goto out_err; } - read_unlock(&GlobalSMBSeslock); - return NULL; + /* see if we already have a matching tcp_ses */ + tcp_ses = cifs_find_tcp_session(&addr); + if (tcp_ses) + return tcp_ses; + + tcp_ses = kzalloc(sizeof(struct TCP_Server_Info), GFP_KERNEL); + if (!tcp_ses) { + rc = -ENOMEM; + goto out_err; + } + + tcp_ses->hostname = extract_hostname(volume_info->UNC); + if (IS_ERR(tcp_ses->hostname)) { + rc = PTR_ERR(tcp_ses->hostname); + goto out_err; + } + + tcp_ses->noblocksnd = volume_info->noblocksnd; + tcp_ses->noautotune = volume_info->noautotune; + atomic_set(&tcp_ses->inFlight, 0); + init_waitqueue_head(&tcp_ses->response_q); + init_waitqueue_head(&tcp_ses->request_q); + INIT_LIST_HEAD(&tcp_ses->pending_mid_q); + mutex_init(&tcp_ses->srv_mutex); + memcpy(tcp_ses->workstation_RFC1001_name, + volume_info->source_rfc1001_name, RFC1001_NAME_LEN_WITH_NULL); + memcpy(tcp_ses->server_RFC1001_name, + volume_info->target_rfc1001_name, RFC1001_NAME_LEN_WITH_NULL); + tcp_ses->sequence_number = 0; + INIT_LIST_HEAD(&tcp_ses->tcp_ses_list); + INIT_LIST_HEAD(&tcp_ses->smb_ses_list); + + /* + * at this point we are the only ones with the pointer + * to the struct since the kernel thread not created yet + * no need to spinlock this init of tcpStatus or srv_count + */ + tcp_ses->tcpStatus = CifsNew; + ++tcp_ses->srv_count; + + if (addr.sa_family == AF_INET6) { + cFYI(1, ("attempting ipv6 connect")); + /* BB should we allow ipv6 on port 139? */ + /* other OS never observed in Wild doing 139 with v6 */ + memcpy(&tcp_ses->addr.sockAddr6, sin_server6, + sizeof(struct sockaddr_in6)); + sin_server6->sin6_port = htons(volume_info->port); + rc = ipv6_connect(tcp_ses); + } else { + memcpy(&tcp_ses->addr.sockAddr, sin_server, + sizeof(struct sockaddr_in)); + sin_server->sin_port = htons(volume_info->port); + rc = ipv4_connect(tcp_ses); + } + if (rc < 0) { + cERROR(1, ("Error connecting to socket. Aborting operation")); + goto out_err; + } + + /* + * since we're in a cifs function already, we know that + * this will succeed. No need for try_module_get(). + */ + __module_get(THIS_MODULE); + tcp_ses->tsk = kthread_run((void *)(void *)cifs_demultiplex_thread, + tcp_ses, "cifsd"); + if (IS_ERR(tcp_ses->tsk)) { + rc = PTR_ERR(tcp_ses->tsk); + cERROR(1, ("error %d create cifsd thread", rc)); + module_put(THIS_MODULE); + goto out_err; + } + + /* thread spawned, put it on the list */ + write_lock(&cifs_tcp_ses_lock); + list_add(&tcp_ses->tcp_ses_list, &cifs_tcp_ses_list); + write_unlock(&cifs_tcp_ses_lock); + + return tcp_ses; + +out_err: + if (tcp_ses) { + kfree(tcp_ses->hostname); + if (tcp_ses->ssocket) + sock_release(tcp_ses->ssocket); + kfree(tcp_ses); + } + return ERR_PTR(rc); } -static struct cifsTconInfo * -find_unc(__be32 new_target_ip_addr, char *uncName, char *userName) +static struct cifsSesInfo * +cifs_find_smb_ses(struct TCP_Server_Info *server, char *username) { struct list_head *tmp; - struct cifsTconInfo *tcon; - __be32 old_ip; - - read_lock(&GlobalSMBSeslock); + struct cifsSesInfo *ses; - list_for_each(tmp, &GlobalTreeConnectionList) { - cFYI(1, ("Next tcon")); - tcon = list_entry(tmp, struct cifsTconInfo, cifsConnectionList); - if (!tcon->ses || !tcon->ses->server) + write_lock(&cifs_tcp_ses_lock); + list_for_each(tmp, &server->smb_ses_list) { + ses = list_entry(tmp, struct cifsSesInfo, smb_ses_list); + if (strncmp(ses->userName, username, MAX_USERNAME_SIZE)) continue; - old_ip = tcon->ses->server->addr.sockAddr.sin_addr.s_addr; - cFYI(1, ("old ip addr: %x == new ip %x ?", - old_ip, new_target_ip_addr)); + ++ses->ses_count; + write_unlock(&cifs_tcp_ses_lock); + return ses; + } + write_unlock(&cifs_tcp_ses_lock); + return NULL; +} - if (old_ip != new_target_ip_addr) - continue; +static void +cifs_put_smb_ses(struct cifsSesInfo *ses) +{ + int xid; + struct TCP_Server_Info *server = ses->server; - /* BB lock tcon, server, tcp session and increment use count? */ - /* found a match on the TCP session */ - /* BB check if reconnection needed */ - cFYI(1, ("IP match, old UNC: %s new: %s", - tcon->treeName, uncName)); + write_lock(&cifs_tcp_ses_lock); + if (--ses->ses_count > 0) { + write_unlock(&cifs_tcp_ses_lock); + return; + } - if (strncmp(tcon->treeName, uncName, MAX_TREE_SIZE)) - continue; + list_del_init(&ses->smb_ses_list); + write_unlock(&cifs_tcp_ses_lock); - cFYI(1, ("and old usr: %s new: %s", - tcon->treeName, uncName)); + if (ses->status == CifsGood) { + xid = GetXid(); + CIFSSMBLogoff(xid, ses); + _FreeXid(xid); + } + sesInfoFree(ses); + cifs_put_tcp_session(server); +} - if (strncmp(tcon->ses->userName, userName, MAX_USERNAME_SIZE)) +static struct cifsTconInfo * +cifs_find_tcon(struct cifsSesInfo *ses, const char *unc) +{ + struct list_head *tmp; + struct cifsTconInfo *tcon; + + write_lock(&cifs_tcp_ses_lock); + list_for_each(tmp, &ses->tcon_list) { + tcon = list_entry(tmp, struct cifsTconInfo, tcon_list); + if (tcon->tidStatus == CifsExiting) + continue; + if (strncmp(tcon->treeName, unc, MAX_TREE_SIZE)) continue; - /* matched smb session (user name) */ - read_unlock(&GlobalSMBSeslock); + ++tcon->tc_count; + write_unlock(&cifs_tcp_ses_lock); return tcon; } - - read_unlock(&GlobalSMBSeslock); + write_unlock(&cifs_tcp_ses_lock); return NULL; } +static void +cifs_put_tcon(struct cifsTconInfo *tcon) +{ + int xid; + struct cifsSesInfo *ses = tcon->ses; + + write_lock(&cifs_tcp_ses_lock); + if (--tcon->tc_count > 0) { + write_unlock(&cifs_tcp_ses_lock); + return; + } + + list_del_init(&tcon->tcon_list); + write_unlock(&cifs_tcp_ses_lock); + + xid = GetXid(); + CIFSSMBTDis(xid, tcon); + _FreeXid(xid); + + DeleteTconOplockQEntries(tcon); + tconInfoFree(tcon); + cifs_put_smb_ses(ses); +} + int get_dfs_path(int xid, struct cifsSesInfo *pSesInfo, const char *old_path, const struct nls_table *nls_codepage, unsigned int *pnum_referrals, @@ -1505,87 +1728,96 @@ static void rfc1002mangle(char *target, char *source, unsigned int length) static int -ipv4_connect(struct sockaddr_in *psin_server, struct socket **csocket, - char *netbios_name, char *target_name) +ipv4_connect(struct TCP_Server_Info *server) { int rc = 0; - int connected = 0; + bool connected = false; __be16 orig_port = 0; + struct socket *socket = server->ssocket; - if (*csocket == NULL) { + if (socket == NULL) { rc = sock_create_kern(PF_INET, SOCK_STREAM, - IPPROTO_TCP, csocket); + IPPROTO_TCP, &socket); if (rc < 0) { cERROR(1, ("Error %d creating socket", rc)); - *csocket = NULL; return rc; - } else { - /* BB other socket options to set KEEPALIVE, NODELAY? */ - cFYI(1, ("Socket created")); - (*csocket)->sk->sk_allocation = GFP_NOFS; - cifs_reclassify_socket4(*csocket); } + + /* BB other socket options to set KEEPALIVE, NODELAY? */ + cFYI(1, ("Socket created")); + server->ssocket = socket; + socket->sk->sk_allocation = GFP_NOFS; + cifs_reclassify_socket4(socket); } - psin_server->sin_family = AF_INET; - if (psin_server->sin_port) { /* user overrode default port */ - rc = (*csocket)->ops->connect(*csocket, - (struct sockaddr *) psin_server, - sizeof(struct sockaddr_in), 0); + /* user overrode default port */ + if (server->addr.sockAddr.sin_port) { + rc = socket->ops->connect(socket, (struct sockaddr *) + &server->addr.sockAddr, + sizeof(struct sockaddr_in), 0); if (rc >= 0) - connected = 1; + connected = true; } if (!connected) { /* save original port so we can retry user specified port later if fall back ports fail this time */ - orig_port = psin_server->sin_port; + orig_port = server->addr.sockAddr.sin_port; /* do not retry on the same port we just failed on */ - if (psin_server->sin_port != htons(CIFS_PORT)) { - psin_server->sin_port = htons(CIFS_PORT); - - rc = (*csocket)->ops->connect(*csocket, - (struct sockaddr *) psin_server, - sizeof(struct sockaddr_in), 0); + if (server->addr.sockAddr.sin_port != htons(CIFS_PORT)) { + server->addr.sockAddr.sin_port = htons(CIFS_PORT); + rc = socket->ops->connect(socket, + (struct sockaddr *) + &server->addr.sockAddr, + sizeof(struct sockaddr_in), 0); if (rc >= 0) - connected = 1; + connected = true; } } if (!connected) { - psin_server->sin_port = htons(RFC1001_PORT); - rc = (*csocket)->ops->connect(*csocket, (struct sockaddr *) - psin_server, + server->addr.sockAddr.sin_port = htons(RFC1001_PORT); + rc = socket->ops->connect(socket, (struct sockaddr *) + &server->addr.sockAddr, sizeof(struct sockaddr_in), 0); if (rc >= 0) - connected = 1; + connected = true; } /* give up here - unless we want to retry on different protocol families some day */ if (!connected) { if (orig_port) - psin_server->sin_port = orig_port; + server->addr.sockAddr.sin_port = orig_port; cFYI(1, ("Error %d connecting to server via ipv4", rc)); - sock_release(*csocket); - *csocket = NULL; + sock_release(socket); + server->ssocket = NULL; return rc; } - /* Eventually check for other socket options to change from - the default. sock_setsockopt not used because it expects - user space buffer */ - cFYI(1, ("sndbuf %d rcvbuf %d rcvtimeo 0x%lx", - (*csocket)->sk->sk_sndbuf, - (*csocket)->sk->sk_rcvbuf, (*csocket)->sk->sk_rcvtimeo)); - (*csocket)->sk->sk_rcvtimeo = 7 * HZ; + + + /* + * Eventually check for other socket options to change from + * the default. sock_setsockopt not used because it expects + * user space buffer + */ + socket->sk->sk_rcvtimeo = 7 * HZ; + socket->sk->sk_sndtimeo = 3 * HZ; + /* make the bufsizes depend on wsize/rsize and max requests */ - if ((*csocket)->sk->sk_sndbuf < (200 * 1024)) - (*csocket)->sk->sk_sndbuf = 200 * 1024; - if ((*csocket)->sk->sk_rcvbuf < (140 * 1024)) - (*csocket)->sk->sk_rcvbuf = 140 * 1024; + if (server->noautotune) { + if (socket->sk->sk_sndbuf < (200 * 1024)) + socket->sk->sk_sndbuf = 200 * 1024; + if (socket->sk->sk_rcvbuf < (140 * 1024)) + socket->sk->sk_rcvbuf = 140 * 1024; + } + + cFYI(1, ("sndbuf %d rcvbuf %d rcvtimeo 0x%lx", + socket->sk->sk_sndbuf, + socket->sk->sk_rcvbuf, socket->sk->sk_rcvtimeo)); /* send RFC1001 sessinit */ - if (psin_server->sin_port == htons(RFC1001_PORT)) { + if (server->addr.sockAddr.sin_port == htons(RFC1001_PORT)) { /* some servers require RFC1001 sessinit before sending negprot - BB check reconnection in case where second sessinit is sent but no second negprot */ @@ -1595,31 +1827,42 @@ ipv4_connect(struct sockaddr_in *psin_server, struct socket **csocket, GFP_KERNEL); if (ses_init_buf) { ses_init_buf->trailer.session_req.called_len = 32; - if (target_name && (target_name[0] != 0)) { - rfc1002mangle(ses_init_buf->trailer.session_req.called_name, - target_name, 16); - } else { - rfc1002mangle(ses_init_buf->trailer.session_req.called_name, - DEFAULT_CIFS_CALLED_NAME, 16); - } + if (server->server_RFC1001_name && + server->server_RFC1001_name[0] != 0) + rfc1002mangle(ses_init_buf->trailer. + session_req.called_name, + server->server_RFC1001_name, + RFC1001_NAME_LEN_WITH_NULL); + else + rfc1002mangle(ses_init_buf->trailer. + session_req.called_name, + DEFAULT_CIFS_CALLED_NAME, + RFC1001_NAME_LEN_WITH_NULL); ses_init_buf->trailer.session_req.calling_len = 32; + /* calling name ends in null (byte 16) from old smb convention. */ - if (netbios_name && (netbios_name[0] != 0)) { - rfc1002mangle(ses_init_buf->trailer.session_req.calling_name, - netbios_name, 16); - } else { - rfc1002mangle(ses_init_buf->trailer.session_req.calling_name, - "LINUX_CIFS_CLNT", 16); - } + if (server->workstation_RFC1001_name && + server->workstation_RFC1001_name[0] != 0) + rfc1002mangle(ses_init_buf->trailer. + session_req.calling_name, + server->workstation_RFC1001_name, + RFC1001_NAME_LEN_WITH_NULL); + else + rfc1002mangle(ses_init_buf->trailer. + session_req.calling_name, + "LINUX_CIFS_CLNT", + RFC1001_NAME_LEN_WITH_NULL); + ses_init_buf->trailer.session_req.scope1 = 0; ses_init_buf->trailer.session_req.scope2 = 0; smb_buf = (struct smb_hdr *)ses_init_buf; /* sizeof RFC1002_SESSION_REQUEST with no scope */ smb_buf->smb_buf_length = 0x81000044; - rc = smb_send(*csocket, smb_buf, 0x44, - (struct sockaddr *)psin_server); + rc = smb_send(socket, smb_buf, 0x44, + (struct sockaddr *) &server->addr.sockAddr, + server->noblocksnd); kfree(ses_init_buf); msleep(1); /* RFC1001 layer in at least one server requires very short break before negprot @@ -1639,75 +1882,81 @@ ipv4_connect(struct sockaddr_in *psin_server, struct socket **csocket, } static int -ipv6_connect(struct sockaddr_in6 *psin_server, struct socket **csocket) +ipv6_connect(struct TCP_Server_Info *server) { int rc = 0; - int connected = 0; + bool connected = false; __be16 orig_port = 0; + struct socket *socket = server->ssocket; - if (*csocket == NULL) { + if (socket == NULL) { rc = sock_create_kern(PF_INET6, SOCK_STREAM, - IPPROTO_TCP, csocket); + IPPROTO_TCP, &socket); if (rc < 0) { cERROR(1, ("Error %d creating ipv6 socket", rc)); - *csocket = NULL; + socket = NULL; return rc; - } else { - /* BB other socket options to set KEEPALIVE, NODELAY? */ - cFYI(1, ("ipv6 Socket created")); - (*csocket)->sk->sk_allocation = GFP_NOFS; - cifs_reclassify_socket6(*csocket); } - } - psin_server->sin6_family = AF_INET6; + /* BB other socket options to set KEEPALIVE, NODELAY? */ + cFYI(1, ("ipv6 Socket created")); + server->ssocket = socket; + socket->sk->sk_allocation = GFP_NOFS; + cifs_reclassify_socket6(socket); + } - if (psin_server->sin6_port) { /* user overrode default port */ - rc = (*csocket)->ops->connect(*csocket, - (struct sockaddr *) psin_server, + /* user overrode default port */ + if (server->addr.sockAddr6.sin6_port) { + rc = socket->ops->connect(socket, + (struct sockaddr *) &server->addr.sockAddr6, sizeof(struct sockaddr_in6), 0); if (rc >= 0) - connected = 1; + connected = true; } if (!connected) { /* save original port so we can retry user specified port later if fall back ports fail this time */ - orig_port = psin_server->sin6_port; + orig_port = server->addr.sockAddr6.sin6_port; /* do not retry on the same port we just failed on */ - if (psin_server->sin6_port != htons(CIFS_PORT)) { - psin_server->sin6_port = htons(CIFS_PORT); - - rc = (*csocket)->ops->connect(*csocket, - (struct sockaddr *) psin_server, + if (server->addr.sockAddr6.sin6_port != htons(CIFS_PORT)) { + server->addr.sockAddr6.sin6_port = htons(CIFS_PORT); + rc = socket->ops->connect(socket, (struct sockaddr *) + &server->addr.sockAddr6, sizeof(struct sockaddr_in6), 0); if (rc >= 0) - connected = 1; + connected = true; } } if (!connected) { - psin_server->sin6_port = htons(RFC1001_PORT); - rc = (*csocket)->ops->connect(*csocket, (struct sockaddr *) - psin_server, sizeof(struct sockaddr_in6), 0); + server->addr.sockAddr6.sin6_port = htons(RFC1001_PORT); + rc = socket->ops->connect(socket, (struct sockaddr *) + &server->addr.sockAddr6, + sizeof(struct sockaddr_in6), 0); if (rc >= 0) - connected = 1; + connected = true; } /* give up here - unless we want to retry on different protocol families some day */ if (!connected) { if (orig_port) - psin_server->sin6_port = orig_port; + server->addr.sockAddr6.sin6_port = orig_port; cFYI(1, ("Error %d connecting to server via ipv6", rc)); - sock_release(*csocket); - *csocket = NULL; + sock_release(socket); + server->ssocket = NULL; return rc; } - /* Eventually check for other socket options to change from - the default. sock_setsockopt not used because it expects - user space buffer */ - (*csocket)->sk->sk_rcvtimeo = 7 * HZ; + + /* + * Eventually check for other socket options to change from + * the default. sock_setsockopt not used because it expects + * user space buffer + */ + socket->sk->sk_rcvtimeo = 7 * HZ; + socket->sk->sk_sndtimeo = 3 * HZ; + server->ssocket = socket; return rc; } @@ -1845,38 +2094,124 @@ convert_delimiter(char *path, char delim) } } +static void setup_cifs_sb(struct smb_vol *pvolume_info, + struct cifs_sb_info *cifs_sb) +{ + if (pvolume_info->rsize > CIFSMaxBufSize) { + cERROR(1, ("rsize %d too large, using MaxBufSize", + pvolume_info->rsize)); + cifs_sb->rsize = CIFSMaxBufSize; + } else if ((pvolume_info->rsize) && + (pvolume_info->rsize <= CIFSMaxBufSize)) + cifs_sb->rsize = pvolume_info->rsize; + else /* default */ + cifs_sb->rsize = CIFSMaxBufSize; + + if (pvolume_info->wsize > PAGEVEC_SIZE * PAGE_CACHE_SIZE) { + cERROR(1, ("wsize %d too large, using 4096 instead", + pvolume_info->wsize)); + cifs_sb->wsize = 4096; + } else if (pvolume_info->wsize) + cifs_sb->wsize = pvolume_info->wsize; + else + cifs_sb->wsize = min_t(const int, + PAGEVEC_SIZE * PAGE_CACHE_SIZE, + 127*1024); + /* old default of CIFSMaxBufSize was too small now + that SMB Write2 can send multiple pages in kvec. + RFC1001 does not describe what happens when frame + bigger than 128K is sent so use that as max in + conjunction with 52K kvec constraint on arch with 4K + page size */ + + if (cifs_sb->rsize < 2048) { + cifs_sb->rsize = 2048; + /* Windows ME may prefer this */ + cFYI(1, ("readsize set to minimum: 2048")); + } + /* calculate prepath */ + cifs_sb->prepath = pvolume_info->prepath; + if (cifs_sb->prepath) { + cifs_sb->prepathlen = strlen(cifs_sb->prepath); + /* we can not convert the / to \ in the path + separators in the prefixpath yet because we do not + know (until reset_cifs_unix_caps is called later) + whether POSIX PATH CAP is available. We normalize + the / to \ after reset_cifs_unix_caps is called */ + pvolume_info->prepath = NULL; + } else + cifs_sb->prepathlen = 0; + cifs_sb->mnt_uid = pvolume_info->linux_uid; + cifs_sb->mnt_gid = pvolume_info->linux_gid; + cifs_sb->mnt_file_mode = pvolume_info->file_mode; + cifs_sb->mnt_dir_mode = pvolume_info->dir_mode; + cFYI(1, ("file mode: 0x%x dir mode: 0x%x", + cifs_sb->mnt_file_mode, cifs_sb->mnt_dir_mode)); + + if (pvolume_info->noperm) + cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_NO_PERM; + if (pvolume_info->setuids) + cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_SET_UID; + if (pvolume_info->server_ino) + cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_SERVER_INUM; + if (pvolume_info->remap) + cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_MAP_SPECIAL_CHR; + if (pvolume_info->no_xattr) + cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_NO_XATTR; + if (pvolume_info->sfu_emul) + cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_UNX_EMUL; + if (pvolume_info->nobrl) + cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_NO_BRL; + if (pvolume_info->mand_lock) + cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_NOPOSIXBRL; + if (pvolume_info->cifs_acl) + cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_CIFS_ACL; + if (pvolume_info->override_uid) + cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_OVERR_UID; + if (pvolume_info->override_gid) + cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_OVERR_GID; + if (pvolume_info->dynperm) + cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_DYNPERM; + if (pvolume_info->direct_io) { + cFYI(1, ("mounting share using direct i/o")); + cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_DIRECT_IO; + } + + if ((pvolume_info->cifs_acl) && (pvolume_info->dynperm)) + cERROR(1, ("mount option dynperm ignored if cifsacl " + "mount option supported")); +} + int cifs_mount(struct super_block *sb, struct cifs_sb_info *cifs_sb, char *mount_data, const char *devname) { int rc = 0; int xid; - int address_type = AF_INET; - struct socket *csocket = NULL; - struct sockaddr_in sin_server; - struct sockaddr_in6 sin_server6; - struct smb_vol volume_info; + struct smb_vol *volume_info; struct cifsSesInfo *pSesInfo = NULL; - struct cifsSesInfo *existingCifsSes = NULL; struct cifsTconInfo *tcon = NULL; struct TCP_Server_Info *srvTcp = NULL; xid = GetXid(); -/* cFYI(1, ("Entering cifs_mount. Xid: %d with: %s", xid, mount_data)); */ + volume_info = kzalloc(sizeof(struct smb_vol), GFP_KERNEL); + if (!volume_info) { + rc = -ENOMEM; + goto out; + } - memset(&volume_info, 0, sizeof(struct smb_vol)); - if (cifs_parse_mount_options(mount_data, devname, &volume_info)) { + if (cifs_parse_mount_options(mount_data, devname, volume_info)) { rc = -EINVAL; goto out; } - if (volume_info.nullauth) { + if (volume_info->nullauth) { cFYI(1, ("null user")); - volume_info.username = ""; - } else if (volume_info.username) { + volume_info->username = ""; + } else if (volume_info->username) { /* BB fixme parse for domain name here */ - cFYI(1, ("Username: %s", volume_info.username)); + cFYI(1, ("Username: %s", volume_info->username)); } else { cifserror("No username specified"); /* In userspace mount helper we can get user name from alternate @@ -1885,145 +2220,40 @@ cifs_mount(struct super_block *sb, struct cifs_sb_info *cifs_sb, goto out; } - if (volume_info.UNCip && volume_info.UNC) { - rc = cifs_inet_pton(AF_INET, volume_info.UNCip, - &sin_server.sin_addr.s_addr); - - if (rc <= 0) { - /* not ipv4 address, try ipv6 */ - rc = cifs_inet_pton(AF_INET6, volume_info.UNCip, - &sin_server6.sin6_addr.in6_u); - if (rc > 0) - address_type = AF_INET6; - } else { - address_type = AF_INET; - } - - if (rc <= 0) { - /* we failed translating address */ - rc = -EINVAL; - goto out; - } - - cFYI(1, ("UNC: %s ip: %s", volume_info.UNC, volume_info.UNCip)); - /* success */ - rc = 0; - } else if (volume_info.UNCip) { - /* BB using ip addr as server name to connect to the - DFS root below */ - cERROR(1, ("Connecting to DFS root not implemented yet")); - rc = -EINVAL; - goto out; - } else /* which servers DFS root would we conect to */ { - cERROR(1, - ("CIFS mount error: No UNC path (e.g. -o " - "unc=//192.168.1.100/public) specified")); - rc = -EINVAL; - goto out; - } /* this is needed for ASCII cp to Unicode converts */ - if (volume_info.iocharset == NULL) { + if (volume_info->iocharset == NULL) { cifs_sb->local_nls = load_nls_default(); /* load_nls_default can not return null */ } else { - cifs_sb->local_nls = load_nls(volume_info.iocharset); + cifs_sb->local_nls = load_nls(volume_info->iocharset); if (cifs_sb->local_nls == NULL) { cERROR(1, ("CIFS mount error: iocharset %s not found", - volume_info.iocharset)); + volume_info->iocharset)); rc = -ELIBACC; goto out; } } - if (address_type == AF_INET) - existingCifsSes = cifs_find_tcp_session(&sin_server.sin_addr, - NULL /* no ipv6 addr */, - volume_info.username, &srvTcp); - else if (address_type == AF_INET6) { - cFYI(1, ("looking for ipv6 address")); - existingCifsSes = cifs_find_tcp_session(NULL /* no ipv4 addr */, - &sin_server6.sin6_addr, - volume_info.username, &srvTcp); - } else { - rc = -EINVAL; + /* get a reference to a tcp session */ + srvTcp = cifs_get_tcp_session(volume_info); + if (IS_ERR(srvTcp)) { + rc = PTR_ERR(srvTcp); goto out; } - if (srvTcp) { - cFYI(1, ("Existing tcp session with server found")); - } else { /* create socket */ - if (volume_info.port) - sin_server.sin_port = htons(volume_info.port); - else - sin_server.sin_port = 0; - if (address_type == AF_INET6) { - cFYI(1, ("attempting ipv6 connect")); - /* BB should we allow ipv6 on port 139? */ - /* other OS never observed in Wild doing 139 with v6 */ - rc = ipv6_connect(&sin_server6, &csocket); - } else - rc = ipv4_connect(&sin_server, &csocket, - volume_info.source_rfc1001_name, - volume_info.target_rfc1001_name); - if (rc < 0) { - cERROR(1, ("Error connecting to IPv4 socket. " - "Aborting operation")); - if (csocket != NULL) - sock_release(csocket); - goto out; - } - - srvTcp = kzalloc(sizeof(struct TCP_Server_Info), GFP_KERNEL); - if (!srvTcp) { - rc = -ENOMEM; - sock_release(csocket); - goto out; - } else { - memcpy(&srvTcp->addr.sockAddr, &sin_server, - sizeof(struct sockaddr_in)); - atomic_set(&srvTcp->inFlight, 0); - /* BB Add code for ipv6 case too */ - srvTcp->ssocket = csocket; - srvTcp->protocolType = IPV4; - srvTcp->hostname = extract_hostname(volume_info.UNC); - if (IS_ERR(srvTcp->hostname)) { - rc = PTR_ERR(srvTcp->hostname); - sock_release(csocket); - goto out; - } - init_waitqueue_head(&srvTcp->response_q); - init_waitqueue_head(&srvTcp->request_q); - INIT_LIST_HEAD(&srvTcp->pending_mid_q); - /* at this point we are the only ones with the pointer - to the struct since the kernel thread not created yet - so no need to spinlock this init of tcpStatus */ - srvTcp->tcpStatus = CifsNew; - init_MUTEX(&srvTcp->tcpSem); - srvTcp->tsk = kthread_run((void *)(void *)cifs_demultiplex_thread, srvTcp, "cifsd"); - if (IS_ERR(srvTcp->tsk)) { - rc = PTR_ERR(srvTcp->tsk); - cERROR(1, ("error %d create cifsd thread", rc)); - srvTcp->tsk = NULL; - sock_release(csocket); - kfree(srvTcp->hostname); - goto out; - } - rc = 0; - memcpy(srvTcp->workstation_RFC1001_name, - volume_info.source_rfc1001_name, 16); - memcpy(srvTcp->server_RFC1001_name, - volume_info.target_rfc1001_name, 16); - srvTcp->sequence_number = 0; - } - } - - if (existingCifsSes) { - pSesInfo = existingCifsSes; + pSesInfo = cifs_find_smb_ses(srvTcp, volume_info->username); + if (pSesInfo) { cFYI(1, ("Existing smb sess found (status=%d)", pSesInfo->status)); + /* + * The existing SMB session already has a reference to srvTcp, + * so we can put back the extra one we got before + */ + cifs_put_tcp_session(srvTcp); + down(&pSesInfo->sesSem); - if (pSesInfo->status == CifsNeedReconnect) { + if (pSesInfo->need_reconnect) { cFYI(1, ("Session needs reconnect")); rc = cifs_setup_session(xid, pSesInfo, cifs_sb->local_nls); @@ -2032,180 +2262,117 @@ cifs_mount(struct super_block *sb, struct cifs_sb_info *cifs_sb, } else if (!rc) { cFYI(1, ("Existing smb sess not found")); pSesInfo = sesInfoAlloc(); - if (pSesInfo == NULL) + if (pSesInfo == NULL) { rc = -ENOMEM; - else { - pSesInfo->server = srvTcp; - sprintf(pSesInfo->serverName, "%u.%u.%u.%u", - NIPQUAD(sin_server.sin_addr.s_addr)); + goto mount_fail_check; } - if (!rc) { - /* volume_info.password freed at unmount */ - if (volume_info.password) { - pSesInfo->password = volume_info.password; - /* set to NULL to prevent freeing on exit */ - volume_info.password = NULL; - } - if (volume_info.username) - strncpy(pSesInfo->userName, - volume_info.username, - MAX_USERNAME_SIZE); - if (volume_info.domainname) { - int len = strlen(volume_info.domainname); - pSesInfo->domainName = - kmalloc(len + 1, GFP_KERNEL); - if (pSesInfo->domainName) - strcpy(pSesInfo->domainName, - volume_info.domainname); + /* new SMB session uses our srvTcp ref */ + pSesInfo->server = srvTcp; + if (srvTcp->addr.sockAddr6.sin6_family == AF_INET6) + sprintf(pSesInfo->serverName, "%pI6", + &srvTcp->addr.sockAddr6.sin6_addr); + else + sprintf(pSesInfo->serverName, "%pI4", + &srvTcp->addr.sockAddr.sin_addr.s_addr); + + write_lock(&cifs_tcp_ses_lock); + list_add(&pSesInfo->smb_ses_list, &srvTcp->smb_ses_list); + write_unlock(&cifs_tcp_ses_lock); + + /* volume_info->password freed at unmount */ + if (volume_info->password) { + pSesInfo->password = kstrdup(volume_info->password, + GFP_KERNEL); + if (!pSesInfo->password) { + rc = -ENOMEM; + goto mount_fail_check; } - pSesInfo->linux_uid = volume_info.linux_uid; - pSesInfo->overrideSecFlg = volume_info.secFlg; - down(&pSesInfo->sesSem); - /* BB FIXME need to pass vol->secFlgs BB */ - rc = cifs_setup_session(xid, pSesInfo, - cifs_sb->local_nls); - up(&pSesInfo->sesSem); - if (!rc) - atomic_inc(&srvTcp->socketUseCount); } + if (volume_info->username) + strncpy(pSesInfo->userName, volume_info->username, + MAX_USERNAME_SIZE); + if (volume_info->domainname) { + int len = strlen(volume_info->domainname); + pSesInfo->domainName = kmalloc(len + 1, GFP_KERNEL); + if (pSesInfo->domainName) + strcpy(pSesInfo->domainName, + volume_info->domainname); + } + pSesInfo->linux_uid = volume_info->linux_uid; + pSesInfo->overrideSecFlg = volume_info->secFlg; + down(&pSesInfo->sesSem); + + /* BB FIXME need to pass vol->secFlgs BB */ + rc = cifs_setup_session(xid, pSesInfo, + cifs_sb->local_nls); + up(&pSesInfo->sesSem); } /* search for existing tcon to this server share */ if (!rc) { - if (volume_info.rsize > CIFSMaxBufSize) { - cERROR(1, ("rsize %d too large, using MaxBufSize", - volume_info.rsize)); - cifs_sb->rsize = CIFSMaxBufSize; - } else if ((volume_info.rsize) && - (volume_info.rsize <= CIFSMaxBufSize)) - cifs_sb->rsize = volume_info.rsize; - else /* default */ - cifs_sb->rsize = CIFSMaxBufSize; - - if (volume_info.wsize > PAGEVEC_SIZE * PAGE_CACHE_SIZE) { - cERROR(1, ("wsize %d too large, using 4096 instead", - volume_info.wsize)); - cifs_sb->wsize = 4096; - } else if (volume_info.wsize) - cifs_sb->wsize = volume_info.wsize; - else - cifs_sb->wsize = - min_t(const int, PAGEVEC_SIZE * PAGE_CACHE_SIZE, - 127*1024); - /* old default of CIFSMaxBufSize was too small now - that SMB Write2 can send multiple pages in kvec. - RFC1001 does not describe what happens when frame - bigger than 128K is sent so use that as max in - conjunction with 52K kvec constraint on arch with 4K - page size */ - - if (cifs_sb->rsize < 2048) { - cifs_sb->rsize = 2048; - /* Windows ME may prefer this */ - cFYI(1, ("readsize set to minimum: 2048")); - } - /* calculate prepath */ - cifs_sb->prepath = volume_info.prepath; - if (cifs_sb->prepath) { - cifs_sb->prepathlen = strlen(cifs_sb->prepath); - /* we can not convert the / to \ in the path - separators in the prefixpath yet because we do not - know (until reset_cifs_unix_caps is called later) - whether POSIX PATH CAP is available. We normalize - the / to \ after reset_cifs_unix_caps is called */ - volume_info.prepath = NULL; - } else - cifs_sb->prepathlen = 0; - cifs_sb->mnt_uid = volume_info.linux_uid; - cifs_sb->mnt_gid = volume_info.linux_gid; - cifs_sb->mnt_file_mode = volume_info.file_mode; - cifs_sb->mnt_dir_mode = volume_info.dir_mode; - cFYI(1, ("file mode: 0x%x dir mode: 0x%x", - cifs_sb->mnt_file_mode, cifs_sb->mnt_dir_mode)); - - if (volume_info.noperm) - cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_NO_PERM; - if (volume_info.setuids) - cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_SET_UID; - if (volume_info.server_ino) - cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_SERVER_INUM; - if (volume_info.remap) - cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_MAP_SPECIAL_CHR; - if (volume_info.no_xattr) - cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_NO_XATTR; - if (volume_info.sfu_emul) - cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_UNX_EMUL; - if (volume_info.nobrl) - cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_NO_BRL; - if (volume_info.cifs_acl) - cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_CIFS_ACL; - if (volume_info.override_uid) - cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_OVERR_UID; - if (volume_info.override_gid) - cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_OVERR_GID; - if (volume_info.dynperm) - cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_DYNPERM; - if (volume_info.direct_io) { - cFYI(1, ("mounting share using direct i/o")); - cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_DIRECT_IO; - } - - if ((volume_info.cifs_acl) && (volume_info.dynperm)) - cERROR(1, ("mount option dynperm ignored if cifsacl " - "mount option supported")); + setup_cifs_sb(volume_info, cifs_sb); - tcon = - find_unc(sin_server.sin_addr.s_addr, volume_info.UNC, - volume_info.username); + tcon = cifs_find_tcon(pSesInfo, volume_info->UNC); if (tcon) { cFYI(1, ("Found match on UNC path")); - /* we can have only one retry value for a connection - to a share so for resources mounted more than once - to the same server share the last value passed in - for the retry flag is used */ - tcon->retry = volume_info.retry; - tcon->nocase = volume_info.nocase; - if (tcon->seal != volume_info.seal) + /* existing tcon already has a reference */ + cifs_put_smb_ses(pSesInfo); + if (tcon->seal != volume_info->seal) cERROR(1, ("transport encryption setting " "conflicts with existing tid")); } else { tcon = tconInfoAlloc(); - if (tcon == NULL) + if (tcon == NULL) { rc = -ENOMEM; - else { - /* check for null share name ie connecting to - * dfs root */ - - /* BB check if this works for exactly length - * three strings */ - if ((strchr(volume_info.UNC + 3, '\\') == NULL) - && (strchr(volume_info.UNC + 3, '/') == - NULL)) { -/* rc = connect_to_dfs_path(xid, pSesInfo, - "", cifs_sb->local_nls, - cifs_sb->mnt_cifs_flags & - CIFS_MOUNT_MAP_SPECIAL_CHR);*/ - cFYI(1, ("DFS root not supported")); - rc = -ENODEV; - goto out; - } else { - /* BB Do we need to wrap sesSem around - * this TCon call and Unix SetFS as - * we do on SessSetup and reconnect? */ - rc = CIFSTCon(xid, pSesInfo, - volume_info.UNC, - tcon, cifs_sb->local_nls); - cFYI(1, ("CIFS Tcon rc = %d", rc)); + goto mount_fail_check; + } + + tcon->ses = pSesInfo; + if (volume_info->password) { + tcon->password = kstrdup(volume_info->password, + GFP_KERNEL); + if (!tcon->password) { + rc = -ENOMEM; + goto mount_fail_check; } - if (!rc) { - atomic_inc(&pSesInfo->inUse); - tcon->retry = volume_info.retry; - tcon->nocase = volume_info.nocase; - tcon->seal = volume_info.seal; + } + + /* check for null share name ie connect to dfs root */ + if ((strchr(volume_info->UNC + 3, '\\') == NULL) + && (strchr(volume_info->UNC + 3, '/') == NULL)) { + /* rc = connect_to_dfs_path(...) */ + cFYI(1, ("DFS root not supported")); + rc = -ENODEV; + goto mount_fail_check; + } else { + /* BB Do we need to wrap sesSem around + * this TCon call and Unix SetFS as + * we do on SessSetup and reconnect? */ + rc = CIFSTCon(xid, pSesInfo, volume_info->UNC, + tcon, cifs_sb->local_nls); + cFYI(1, ("CIFS Tcon rc = %d", rc)); + if (volume_info->nodfs) { + tcon->Flags &= ~SMB_SHARE_IS_IN_DFS; + cFYI(1, ("DFS disabled (%d)", + tcon->Flags)); } } - } + if (rc) + goto mount_fail_check; + tcon->seal = volume_info->seal; + write_lock(&cifs_tcp_ses_lock); + list_add(&tcon->tcon_list, &pSesInfo->tcon_list); + write_unlock(&cifs_tcp_ses_lock); + } + + /* we can have only one retry value for a connection + to a share so for resources mounted more than once + to the same server share the last value passed in + for the retry flag is used */ + tcon->retry = volume_info->retry; + tcon->nocase = volume_info->nocase; + tcon->local_lease = volume_info->local_lease; } if (pSesInfo) { if (pSesInfo->capabilities & CAP_LARGE_FILES) { @@ -2217,104 +2384,66 @@ cifs_mount(struct super_block *sb, struct cifs_sb_info *cifs_sb, /* BB FIXME fix time_gran to be larger for LANMAN sessions */ sb->s_time_gran = 100; -/* on error free sesinfo and tcon struct if needed */ +mount_fail_check: + /* on error free sesinfo and tcon struct if needed */ if (rc) { - /* if session setup failed, use count is zero but - we still need to free cifsd thread */ - if (atomic_read(&srvTcp->socketUseCount) == 0) { - spin_lock(&GlobalMid_Lock); - srvTcp->tcpStatus = CifsExiting; - spin_unlock(&GlobalMid_Lock); - if (srvTcp->tsk) { - /* If we could verify that kthread_stop would - always wake up processes blocked in - tcp in recv_mesg then we could remove the - send_sig call */ - force_sig(SIGKILL, srvTcp->tsk); - kthread_stop(srvTcp->tsk); - } - } - /* If find_unc succeeded then rc == 0 so we can not end */ - if (tcon) /* up accidently freeing someone elses tcon struct */ - tconInfoFree(tcon); - if (existingCifsSes == NULL) { - if (pSesInfo) { - if ((pSesInfo->server) && - (pSesInfo->status == CifsGood)) { - int temp_rc; - temp_rc = CIFSSMBLogoff(xid, pSesInfo); - /* if the socketUseCount is now zero */ - if ((temp_rc == -ESHUTDOWN) && - (pSesInfo->server) && - (pSesInfo->server->tsk)) { - force_sig(SIGKILL, - pSesInfo->server->tsk); - kthread_stop(pSesInfo->server->tsk); - } - } else { - cFYI(1, ("No session or bad tcon")); - if ((pSesInfo->server) && - (pSesInfo->server->tsk)) { - force_sig(SIGKILL, - pSesInfo->server->tsk); - kthread_stop(pSesInfo->server->tsk); - } - } - sesInfoFree(pSesInfo); - /* pSesInfo = NULL; */ - } - } - } else { - atomic_inc(&tcon->useCount); - cifs_sb->tcon = tcon; - tcon->ses = pSesInfo; - - /* do not care if following two calls succeed - informational */ - if (!tcon->ipc) { - CIFSSMBQFSDeviceInfo(xid, tcon); - CIFSSMBQFSAttributeInfo(xid, tcon); - } - - /* tell server which Unix caps we support */ - if (tcon->ses->capabilities & CAP_UNIX) - /* reset of caps checks mount to see if unix extensions - disabled for just this mount */ - reset_cifs_unix_caps(xid, tcon, sb, &volume_info); + /* If find_unc succeeded then rc == 0 so we can not end */ + /* up accidently freeing someone elses tcon struct */ + if (tcon) + cifs_put_tcon(tcon); + else if (pSesInfo) + cifs_put_smb_ses(pSesInfo); else - tcon->unix_ext = 0; /* server does not support them */ + cifs_put_tcp_session(srvTcp); + goto out; + } + cifs_sb->tcon = tcon; - /* convert forward to back slashes in prepath here if needed */ - if ((cifs_sb->mnt_cifs_flags & CIFS_MOUNT_POSIX_PATHS) == 0) - convert_delimiter(cifs_sb->prepath, - CIFS_DIR_SEP(cifs_sb)); + /* do not care if following two calls succeed - informational */ + if (!tcon->ipc) { + CIFSSMBQFSDeviceInfo(xid, tcon); + CIFSSMBQFSAttributeInfo(xid, tcon); + } - if ((tcon->unix_ext == 0) && (cifs_sb->rsize > (1024 * 127))) { - cifs_sb->rsize = 1024 * 127; - cFYI(DBG2, - ("no very large read support, rsize now 127K")); - } - if (!(tcon->ses->capabilities & CAP_LARGE_WRITE_X)) - cifs_sb->wsize = min(cifs_sb->wsize, - (tcon->ses->server->maxBuf - - MAX_CIFS_HDR_SIZE)); - if (!(tcon->ses->capabilities & CAP_LARGE_READ_X)) - cifs_sb->rsize = min(cifs_sb->rsize, - (tcon->ses->server->maxBuf - - MAX_CIFS_HDR_SIZE)); + /* tell server which Unix caps we support */ + if (tcon->ses->capabilities & CAP_UNIX) + /* reset of caps checks mount to see if unix extensions + disabled for just this mount */ + reset_cifs_unix_caps(xid, tcon, sb, volume_info); + else + tcon->unix_ext = 0; /* server does not support them */ + + /* convert forward to back slashes in prepath here if needed */ + if ((cifs_sb->mnt_cifs_flags & CIFS_MOUNT_POSIX_PATHS) == 0) + convert_delimiter(cifs_sb->prepath, CIFS_DIR_SEP(cifs_sb)); + + if ((tcon->unix_ext == 0) && (cifs_sb->rsize > (1024 * 127))) { + cifs_sb->rsize = 1024 * 127; + cFYI(DBG2, ("no very large read support, rsize now 127K")); } + if (!(tcon->ses->capabilities & CAP_LARGE_WRITE_X)) + cifs_sb->wsize = min(cifs_sb->wsize, + (tcon->ses->server->maxBuf - MAX_CIFS_HDR_SIZE)); + if (!(tcon->ses->capabilities & CAP_LARGE_READ_X)) + cifs_sb->rsize = min(cifs_sb->rsize, + (tcon->ses->server->maxBuf - MAX_CIFS_HDR_SIZE)); - /* volume_info.password is freed above when existing session found + /* volume_info->password is freed above when existing session found (in which case it is not needed anymore) but when new sesion is created the password ptr is put in the new session structure (in which case the password will be freed at unmount time) */ out: /* zero out password before freeing */ - if (volume_info.password != NULL) { - memset(volume_info.password, 0, strlen(volume_info.password)); - kfree(volume_info.password); + if (volume_info) { + if (volume_info->password != NULL) { + memset(volume_info->password, 0, + strlen(volume_info->password)); + kfree(volume_info->password); + } + kfree(volume_info->UNC); + kfree(volume_info->prepath); + kfree(volume_info); } - kfree(volume_info.UNC); - kfree(volume_info.prepath); FreeXid(xid); return rc; } @@ -2465,7 +2594,7 @@ CIFSSessSetup(unsigned int xid, struct cifsSesInfo *ses, __u16 action = le16_to_cpu(pSMBr->resp.Action); __u16 blob_len = le16_to_cpu(pSMBr->resp.SecurityBlobLength); if (action & GUEST_LOGIN) - cFYI(1, (" Guest login")); /* BB mark SesInfo struct? */ + cFYI(1, ("Guest login")); /* BB mark SesInfo struct? */ ses->Suid = smb_buffer_response->Uid; /* UID left in wire format (little endian) */ cFYI(1, ("UID = %d ", ses->Suid)); @@ -2611,13 +2740,11 @@ CIFSSessSetup(unsigned int xid, struct cifsSesInfo *ses, len)); } } else { - cERROR(1, - (" Security Blob Length extends beyond " + cERROR(1, ("Security Blob Length extends beyond " "end of SMB")); } } else { - cERROR(1, - (" Invalid Word count %d: ", + cERROR(1, ("Invalid Word count %d: ", smb_buffer_response->WordCount)); rc = -EIO; } @@ -2775,7 +2902,7 @@ CIFSNTLMSSPNegotiateSessSetup(unsigned int xid, __u16 blob_len = le16_to_cpu(pSMBr->resp.SecurityBlobLength); if (action & GUEST_LOGIN) - cFYI(1, (" Guest login")); + cFYI(1, ("Guest login")); /* Do we want to set anything in SesInfo struct when guest login? */ bcc_ptr = pByteArea(smb_buffer_response); @@ -2783,8 +2910,7 @@ CIFSNTLMSSPNegotiateSessSetup(unsigned int xid, SecurityBlob2 = (PCHALLENGE_MESSAGE) bcc_ptr; if (SecurityBlob2->MessageType != NtLmChallenge) { - cFYI(1, - ("Unexpected NTLMSSP message type received %d", + cFYI(1, ("Unexpected NTLMSSP message type received %d", SecurityBlob2->MessageType)); } else if (ses) { ses->Suid = smb_buffer_response->Uid; /* UID left in le format */ @@ -2956,8 +3082,7 @@ CIFSNTLMSSPNegotiateSessSetup(unsigned int xid, cERROR(1, ("No session structure passed in.")); } } else { - cERROR(1, - (" Invalid Word count %d:", + cERROR(1, ("Invalid Word count %d:", smb_buffer_response->WordCount)); rc = -EIO; } @@ -3196,7 +3321,7 @@ CIFSNTLMSSPAuthSessSetup(unsigned int xid, struct cifsSesInfo *ses, __u16 action = le16_to_cpu(pSMBr->resp.Action); __u16 blob_len = le16_to_cpu(pSMBr->resp.SecurityBlobLength); if (action & GUEST_LOGIN) - cFYI(1, (" Guest login")); /* BB Should we set anything + cFYI(1, ("Guest login")); /* BB Should we set anything in SesInfo struct ? */ /* if (SecurityBlob2->MessageType != NtLm??) { cFYI("Unexpected message type on auth response is %d")); @@ -3419,12 +3544,14 @@ CIFSTCon(unsigned int xid, struct cifsSesInfo *ses, NTLMv2 password here) */ #ifdef CONFIG_CIFS_WEAK_PW_HASH if ((extended_security & CIFSSEC_MAY_LANMAN) && - (ses->server->secType == LANMAN)) - calc_lanman_hash(ses, bcc_ptr); + (ses->server->secType == LANMAN)) + calc_lanman_hash(tcon->password, ses->server->cryptKey, + ses->server->secMode & + SECMODE_PW_ENCRYPT ? true : false, + bcc_ptr); else #endif /* CIFS_WEAK_PW_HASH */ - SMBNTencrypt(ses->password, - ses->server->cryptKey, + SMBNTencrypt(tcon->password, ses->server->cryptKey, bcc_ptr); bcc_ptr += CIFS_SESS_KEY_SIZE; @@ -3471,6 +3598,7 @@ CIFSTCon(unsigned int xid, struct cifsSesInfo *ses, /* above now done in SendReceive */ if ((rc == 0) && (tcon != NULL)) { tcon->tidStatus = CifsGood; + tcon->need_reconnect = false; tcon->tid = smb_buffer_response->Tid; bcc_ptr = pByteArea(smb_buffer_response); length = strnlen(bcc_ptr, BCC(smb_buffer_response) - 2); @@ -3542,52 +3670,17 @@ int cifs_umount(struct super_block *sb, struct cifs_sb_info *cifs_sb) { int rc = 0; - int xid; - struct cifsSesInfo *ses = NULL; - struct task_struct *cifsd_task; char *tmp; - xid = GetXid(); - - if (cifs_sb->tcon) { - ses = cifs_sb->tcon->ses; /* save ptr to ses before delete tcon!*/ - rc = CIFSSMBTDis(xid, cifs_sb->tcon); - if (rc == -EBUSY) { - FreeXid(xid); - return 0; - } - DeleteTconOplockQEntries(cifs_sb->tcon); - tconInfoFree(cifs_sb->tcon); - if ((ses) && (ses->server)) { - /* save off task so we do not refer to ses later */ - cifsd_task = ses->server->tsk; - cFYI(1, ("About to do SMBLogoff ")); - rc = CIFSSMBLogoff(xid, ses); - if (rc == -EBUSY) { - FreeXid(xid); - return 0; - } else if (rc == -ESHUTDOWN) { - cFYI(1, ("Waking up socket by sending signal")); - if (cifsd_task) { - force_sig(SIGKILL, cifsd_task); - kthread_stop(cifsd_task); - } - rc = 0; - } /* else - we have an smb session - left on this socket do not kill cifsd */ - } else - cFYI(1, ("No session or bad tcon")); - } + if (cifs_sb->tcon) + cifs_put_tcon(cifs_sb->tcon); cifs_sb->tcon = NULL; tmp = cifs_sb->prepath; cifs_sb->prepathlen = 0; cifs_sb->prepath = NULL; kfree(tmp); - if (ses) - sesInfoFree(ses); - FreeXid(xid); return rc; } @@ -3701,7 +3794,10 @@ int cifs_setup_session(unsigned int xid, struct cifsSesInfo *pSesInfo, cERROR(1, ("Send error in SessSetup = %d", rc)); } else { cFYI(1, ("CIFS Session Established successfully")); + spin_lock(&GlobalMid_Lock); pSesInfo->status = CifsGood; + pSesInfo->need_reconnect = false; + spin_unlock(&GlobalMid_Lock); } ss_err_exit: diff --git a/fs/cifs/dir.c b/fs/cifs/dir.c index e962e75e6f7..838d9c720a5 100644 --- a/fs/cifs/dir.c +++ b/fs/cifs/dir.c @@ -235,11 +235,11 @@ cifs_create(struct inode *inode, struct dentry *direntry, int mode, }; if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_SET_UID) { - args.uid = (__u64) current->fsuid; + args.uid = (__u64) current_fsuid(); if (inode->i_mode & S_ISGID) args.gid = (__u64) inode->i_gid; else - args.gid = (__u64) current->fsgid; + args.gid = (__u64) current_fsgid(); } else { args.uid = NO_CHANGE_64; args.gid = NO_CHANGE_64; @@ -271,13 +271,13 @@ cifs_create(struct inode *inode, struct dentry *direntry, int mode, if ((oplock & CIFS_CREATE_ACTION) && (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_SET_UID)) { - newinode->i_uid = current->fsuid; + newinode->i_uid = current_fsuid(); if (inode->i_mode & S_ISGID) newinode->i_gid = inode->i_gid; else newinode->i_gid = - current->fsgid; + current_fsgid(); } } } @@ -375,8 +375,8 @@ int cifs_mknod(struct inode *inode, struct dentry *direntry, int mode, .device = device_number, }; if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_SET_UID) { - args.uid = (__u64) current->fsuid; - args.gid = (__u64) current->fsgid; + args.uid = (__u64) current_fsuid(); + args.gid = (__u64) current_fsgid(); } else { args.uid = NO_CHANGE_64; args.gid = NO_CHANGE_64; @@ -483,7 +483,7 @@ cifs_lookup(struct inode *parent_dir_inode, struct dentry *direntry, xid = GetXid(); - cFYI(1, (" parent inode = 0x%p name is: %s and dentry = 0x%p", + cFYI(1, ("parent inode = 0x%p name is: %s and dentry = 0x%p", parent_dir_inode, direntry->d_name.name, direntry)); /* check whether path exists */ @@ -515,12 +515,11 @@ cifs_lookup(struct inode *parent_dir_inode, struct dentry *direntry, } if (direntry->d_inode != NULL) { - cFYI(1, (" non-NULL inode in lookup")); + cFYI(1, ("non-NULL inode in lookup")); } else { - cFYI(1, (" NULL inode in lookup")); + cFYI(1, ("NULL inode in lookup")); } - cFYI(1, - (" Full path: %s inode = 0x%p", full_path, direntry->d_inode)); + cFYI(1, ("Full path: %s inode = 0x%p", full_path, direntry->d_inode)); if (pTcon->unix_ext) rc = cifs_get_inode_info_unix(&newInode, full_path, diff --git a/fs/cifs/file.c b/fs/cifs/file.c index c4a8a060512..b1e1fc6a6e6 100644 --- a/fs/cifs/file.c +++ b/fs/cifs/file.c @@ -488,12 +488,13 @@ int cifs_close(struct inode *inode, struct file *file) pTcon = cifs_sb->tcon; if (pSMBFile) { struct cifsLockInfo *li, *tmp; - + write_lock(&GlobalSMBSeslock); pSMBFile->closePend = true; if (pTcon) { /* no sense reconnecting to close a file that is already closed */ - if (pTcon->tidStatus != CifsNeedReconnect) { + if (!pTcon->need_reconnect) { + write_unlock(&GlobalSMBSeslock); timeout = 2; while ((atomic_read(&pSMBFile->wrtPending) != 0) && (timeout <= 2048)) { @@ -510,12 +511,15 @@ int cifs_close(struct inode *inode, struct file *file) timeout *= 4; } if (atomic_read(&pSMBFile->wrtPending)) - cERROR(1, - ("close with pending writes")); - rc = CIFSSMBClose(xid, pTcon, + cERROR(1, ("close with pending write")); + if (!pTcon->need_reconnect && + !pSMBFile->invalidHandle) + rc = CIFSSMBClose(xid, pTcon, pSMBFile->netfid); - } - } + } else + write_unlock(&GlobalSMBSeslock); + } else + write_unlock(&GlobalSMBSeslock); /* Delete any outstanding lock records. We'll lose them when the file is closed anyway. */ @@ -587,15 +591,18 @@ int cifs_closedir(struct inode *inode, struct file *file) pTcon = cifs_sb->tcon; cFYI(1, ("Freeing private data in close dir")); + write_lock(&GlobalSMBSeslock); if (!pCFileStruct->srch_inf.endOfSearch && !pCFileStruct->invalidHandle) { pCFileStruct->invalidHandle = true; + write_unlock(&GlobalSMBSeslock); rc = CIFSFindClose(xid, pTcon, pCFileStruct->netfid); cFYI(1, ("Closing uncompleted readdir with rc %d", rc)); /* not much we can do if it fails anyway, ignore rc */ rc = 0; - } + } else + write_unlock(&GlobalSMBSeslock); ptmp = pCFileStruct->srch_inf.ntwrk_buf_start; if (ptmp) { cFYI(1, ("closedir free smb buf in srch struct")); @@ -637,10 +644,10 @@ int cifs_lock(struct file *file, int cmd, struct file_lock *pfLock) __u64 length; bool wait_flag = false; struct cifs_sb_info *cifs_sb; - struct cifsTconInfo *pTcon; + struct cifsTconInfo *tcon; __u16 netfid; __u8 lockType = LOCKING_ANDX_LARGE_FILES; - bool posix_locking; + bool posix_locking = 0; length = 1 + pfLock->fl_end - pfLock->fl_start; rc = -EACCES; @@ -691,7 +698,7 @@ int cifs_lock(struct file *file, int cmd, struct file_lock *pfLock) cFYI(1, ("Unknown type of lock")); cifs_sb = CIFS_SB(file->f_path.dentry->d_sb); - pTcon = cifs_sb->tcon; + tcon = cifs_sb->tcon; if (file->private_data == NULL) { FreeXid(xid); @@ -699,9 +706,10 @@ int cifs_lock(struct file *file, int cmd, struct file_lock *pfLock) } netfid = ((struct cifsFileInfo *)file->private_data)->netfid; - posix_locking = (cifs_sb->tcon->ses->capabilities & CAP_UNIX) && - (CIFS_UNIX_FCNTL_CAP & le64_to_cpu(cifs_sb->tcon->fsUnixInfo.Capability)); - + if ((tcon->ses->capabilities & CAP_UNIX) && + (CIFS_UNIX_FCNTL_CAP & le64_to_cpu(tcon->fsUnixInfo.Capability)) && + ((cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NOPOSIXBRL) == 0)) + posix_locking = 1; /* BB add code here to normalize offset and length to account for negative length which we can not accept over the wire */ @@ -712,7 +720,7 @@ int cifs_lock(struct file *file, int cmd, struct file_lock *pfLock) posix_lock_type = CIFS_RDLCK; else posix_lock_type = CIFS_WRLCK; - rc = CIFSSMBPosixLock(xid, pTcon, netfid, 1 /* get */, + rc = CIFSSMBPosixLock(xid, tcon, netfid, 1 /* get */, length, pfLock, posix_lock_type, wait_flag); FreeXid(xid); @@ -720,10 +728,10 @@ int cifs_lock(struct file *file, int cmd, struct file_lock *pfLock) } /* BB we could chain these into one lock request BB */ - rc = CIFSSMBLock(xid, pTcon, netfid, length, pfLock->fl_start, + rc = CIFSSMBLock(xid, tcon, netfid, length, pfLock->fl_start, 0, 1, lockType, 0 /* wait flag */ ); if (rc == 0) { - rc = CIFSSMBLock(xid, pTcon, netfid, length, + rc = CIFSSMBLock(xid, tcon, netfid, length, pfLock->fl_start, 1 /* numUnlock */ , 0 /* numLock */ , lockType, 0 /* wait flag */ ); @@ -760,7 +768,7 @@ int cifs_lock(struct file *file, int cmd, struct file_lock *pfLock) if (numUnlock == 1) posix_lock_type = CIFS_UNLCK; - rc = CIFSSMBPosixLock(xid, pTcon, netfid, 0 /* set */, + rc = CIFSSMBPosixLock(xid, tcon, netfid, 0 /* set */, length, pfLock, posix_lock_type, wait_flag); } else { @@ -768,7 +776,7 @@ int cifs_lock(struct file *file, int cmd, struct file_lock *pfLock) (struct cifsFileInfo *)file->private_data; if (numLock) { - rc = CIFSSMBLock(xid, pTcon, netfid, length, + rc = CIFSSMBLock(xid, tcon, netfid, length, pfLock->fl_start, 0, numLock, lockType, wait_flag); @@ -789,7 +797,7 @@ int cifs_lock(struct file *file, int cmd, struct file_lock *pfLock) if (pfLock->fl_start <= li->offset && (pfLock->fl_start + length) >= (li->offset + li->length)) { - stored_rc = CIFSSMBLock(xid, pTcon, + stored_rc = CIFSSMBLock(xid, tcon, netfid, li->length, li->offset, 1, 0, li->type, false); @@ -1404,7 +1412,10 @@ retry: if ((wbc->nr_to_write -= n_iov) <= 0) done = 1; index = next; - } + } else + /* Need to re-find the pages we skipped */ + index = pvec.pages[0]->index + 1; + pagevec_release(&pvec); } if (!scanned && !done) { @@ -1465,7 +1476,11 @@ static int cifs_write_end(struct file *file, struct address_space *mapping, cFYI(1, ("write_end for page %p from pos %lld with %d bytes", page, pos, copied)); - if (!PageUptodate(page) && copied == PAGE_CACHE_SIZE) + if (PageChecked(page)) { + if (copied == len) + SetPageUptodate(page); + ClearPageChecked(page); + } else if (!PageUptodate(page) && copied == PAGE_CACHE_SIZE) SetPageUptodate(page); if (!PageUptodate(page)) { @@ -1791,7 +1806,7 @@ static void cifs_copy_cache_pages(struct address_space *mapping, SetPageUptodate(page); unlock_page(page); if (!pagevec_add(plru_pvec, page)) - __pagevec_lru_add(plru_pvec); + __pagevec_lru_add_file(plru_pvec); data += PAGE_CACHE_SIZE; } return; @@ -1824,7 +1839,7 @@ static int cifs_readpages(struct file *file, struct address_space *mapping, pTcon = cifs_sb->tcon; pagevec_init(&lru_pvec, 0); - cFYI(DBG2, ("rpages: num pages %d", num_pages)); + cFYI(DBG2, ("rpages: num pages %d", num_pages)); for (i = 0; i < num_pages; ) { unsigned contig_pages; struct page *tmp_page; @@ -1925,7 +1940,7 @@ static int cifs_readpages(struct file *file, struct address_space *mapping, bytes_read = 0; } - pagevec_lru_add(&lru_pvec); + pagevec_lru_add_file(&lru_pvec); /* need to free smb_read_data buf before exit */ if (smb_read_data) { @@ -2052,39 +2067,70 @@ static int cifs_write_begin(struct file *file, struct address_space *mapping, { pgoff_t index = pos >> PAGE_CACHE_SHIFT; loff_t offset = pos & (PAGE_CACHE_SIZE - 1); + loff_t page_start = pos & PAGE_MASK; + loff_t i_size; + struct page *page; + int rc = 0; cFYI(1, ("write_begin from %lld len %d", (long long)pos, len)); - *pagep = __grab_cache_page(mapping, index); - if (!*pagep) - return -ENOMEM; - - if (PageUptodate(*pagep)) - return 0; + page = __grab_cache_page(mapping, index); + if (!page) { + rc = -ENOMEM; + goto out; + } - /* If we are writing a full page it will be up to date, - no need to read from the server */ - if (len == PAGE_CACHE_SIZE && flags & AOP_FLAG_UNINTERRUPTIBLE) - return 0; + if (PageUptodate(page)) + goto out; - if ((file->f_flags & O_ACCMODE) != O_WRONLY) { - int rc; + /* + * If we write a full page it will be up to date, no need to read from + * the server. If the write is short, we'll end up doing a sync write + * instead. + */ + if (len == PAGE_CACHE_SIZE) + goto out; - /* might as well read a page, it is fast enough */ - rc = cifs_readpage_worker(file, *pagep, &offset); + /* + * optimize away the read when we have an oplock, and we're not + * expecting to use any of the data we'd be reading in. That + * is, when the page lies beyond the EOF, or straddles the EOF + * and the write will cover all of the existing data. + */ + if (CIFS_I(mapping->host)->clientCanCacheRead) { + i_size = i_size_read(mapping->host); + if (page_start >= i_size || + (offset == 0 && (pos + len) >= i_size)) { + zero_user_segments(page, 0, offset, + offset + len, + PAGE_CACHE_SIZE); + /* + * PageChecked means that the parts of the page + * to which we're not writing are considered up + * to date. Once the data is copied to the + * page, it can be set uptodate. + */ + SetPageChecked(page); + goto out; + } + } - /* we do not need to pass errors back - e.g. if we do not have read access to the file - because cifs_write_end will attempt synchronous writes - -- shaggy */ + if ((file->f_flags & O_ACCMODE) != O_WRONLY) { + /* + * might as well read a page, it is fast enough. If we get + * an error, we don't need to return it. cifs_write_end will + * do a sync write instead since PG_uptodate isn't set. + */ + cifs_readpage_worker(file, page, &page_start); } else { /* we could try using another file handle if there is one - but how would we lock it to prevent close of that handle racing with this read? In any case this will be written out by write_end so is fine */ } - - return 0; +out: + *pagep = page; + return rc; } const struct address_space_operations cifs_addr_ops = { diff --git a/fs/cifs/inode.c b/fs/cifs/inode.c index a8c833345fc..f247da9f4ed 100644 --- a/fs/cifs/inode.c +++ b/fs/cifs/inode.c @@ -1,7 +1,7 @@ /* * fs/cifs/inode.c * - * Copyright (C) International Business Machines Corp., 2002,2007 + * Copyright (C) International Business Machines Corp., 2002,2008 * Author(s): Steve French (sfrench@us.ibm.com) * * This library is free software; you can redistribute it and/or modify @@ -506,6 +506,7 @@ int cifs_get_inode_info(struct inode **pinode, inode = *pinode; cifsInfo = CIFS_I(inode); cifsInfo->cifsAttrs = attr; + cifsInfo->delete_pending = pfindData->DeletePending ? true : false; cFYI(1, ("Old time %ld", cifsInfo->time)); cifsInfo->time = jiffies; cFYI(1, ("New time %ld", cifsInfo->time)); @@ -620,6 +621,47 @@ static const struct inode_operations cifs_ipc_inode_ops = { .lookup = cifs_lookup, }; +static char *build_path_to_root(struct cifs_sb_info *cifs_sb) +{ + int pplen = cifs_sb->prepathlen; + int dfsplen; + char *full_path = NULL; + + /* if no prefix path, simply set path to the root of share to "" */ + if (pplen == 0) { + full_path = kmalloc(1, GFP_KERNEL); + if (full_path) + full_path[0] = 0; + return full_path; + } + + if (cifs_sb->tcon && (cifs_sb->tcon->Flags & SMB_SHARE_IS_IN_DFS)) + dfsplen = strnlen(cifs_sb->tcon->treeName, MAX_TREE_SIZE + 1); + else + dfsplen = 0; + + full_path = kmalloc(dfsplen + pplen + 1, GFP_KERNEL); + if (full_path == NULL) + return full_path; + + if (dfsplen) { + strncpy(full_path, cifs_sb->tcon->treeName, dfsplen); + /* switch slash direction in prepath depending on whether + * windows or posix style path names + */ + if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_POSIX_PATHS) { + int i; + for (i = 0; i < dfsplen; i++) { + if (full_path[i] == '\\') + full_path[i] = '/'; + } + } + } + strncpy(full_path + dfsplen, cifs_sb->prepath, pplen); + full_path[dfsplen + pplen] = 0; /* add trailing null */ + return full_path; +} + /* gets root inode */ struct inode *cifs_iget(struct super_block *sb, unsigned long ino) { @@ -627,6 +669,7 @@ struct inode *cifs_iget(struct super_block *sb, unsigned long ino) struct cifs_sb_info *cifs_sb; struct inode *inode; long rc; + char *full_path; inode = iget_locked(sb, ino); if (!inode) @@ -635,13 +678,17 @@ struct inode *cifs_iget(struct super_block *sb, unsigned long ino) return inode; cifs_sb = CIFS_SB(inode->i_sb); - xid = GetXid(); + full_path = build_path_to_root(cifs_sb); + if (full_path == NULL) + return ERR_PTR(-ENOMEM); + xid = GetXid(); if (cifs_sb->tcon->unix_ext) - rc = cifs_get_inode_info_unix(&inode, "", inode->i_sb, xid); + rc = cifs_get_inode_info_unix(&inode, full_path, inode->i_sb, + xid); else - rc = cifs_get_inode_info(&inode, "", NULL, inode->i_sb, xid, - NULL); + rc = cifs_get_inode_info(&inode, full_path, NULL, inode->i_sb, + xid, NULL); if (rc && cifs_sb->tcon->ipc) { cFYI(1, ("ipc connection - fake read inode")); inode->i_mode |= S_IFDIR; @@ -651,6 +698,7 @@ struct inode *cifs_iget(struct super_block *sb, unsigned long ino) inode->i_uid = cifs_sb->mnt_uid; inode->i_gid = cifs_sb->mnt_gid; } else if (rc) { + kfree(full_path); _FreeXid(xid); iget_failed(inode); return ERR_PTR(rc); @@ -658,6 +706,7 @@ struct inode *cifs_iget(struct super_block *sb, unsigned long ino) unlock_new_inode(inode); + kfree(full_path); /* can not call macro FreeXid here since in a void func * TODO: This is no longer true */ @@ -772,63 +821,106 @@ out: * anything else. */ static int -cifs_rename_pending_delete(char *full_path, struct inode *inode, int xid) +cifs_rename_pending_delete(char *full_path, struct dentry *dentry, int xid) { int oplock = 0; int rc; __u16 netfid; + struct inode *inode = dentry->d_inode; struct cifsInodeInfo *cifsInode = CIFS_I(inode); struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb); struct cifsTconInfo *tcon = cifs_sb->tcon; - __u32 dosattr; - FILE_BASIC_INFO *info_buf; + __u32 dosattr, origattr; + FILE_BASIC_INFO *info_buf = NULL; rc = CIFSSMBOpen(xid, tcon, full_path, FILE_OPEN, - DELETE|FILE_WRITE_ATTRIBUTES, - CREATE_NOT_DIR|CREATE_DELETE_ON_CLOSE, + DELETE|FILE_WRITE_ATTRIBUTES, CREATE_NOT_DIR, &netfid, &oplock, NULL, cifs_sb->local_nls, cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR); if (rc != 0) goto out; - /* set ATTR_HIDDEN and clear ATTR_READONLY */ - cifsInode = CIFS_I(inode); - dosattr = cifsInode->cifsAttrs & ~ATTR_READONLY; + origattr = cifsInode->cifsAttrs; + if (origattr == 0) + origattr |= ATTR_NORMAL; + + dosattr = origattr & ~ATTR_READONLY; if (dosattr == 0) dosattr |= ATTR_NORMAL; dosattr |= ATTR_HIDDEN; - info_buf = kzalloc(sizeof(*info_buf), GFP_KERNEL); - if (info_buf == NULL) { - rc = -ENOMEM; - goto out_close; + /* set ATTR_HIDDEN and clear ATTR_READONLY, but only if needed */ + if (dosattr != origattr) { + info_buf = kzalloc(sizeof(*info_buf), GFP_KERNEL); + if (info_buf == NULL) { + rc = -ENOMEM; + goto out_close; + } + info_buf->Attributes = cpu_to_le32(dosattr); + rc = CIFSSMBSetFileInfo(xid, tcon, info_buf, netfid, + current->tgid); + /* although we would like to mark the file hidden + if that fails we will still try to rename it */ + if (rc != 0) + cifsInode->cifsAttrs = dosattr; + else + dosattr = origattr; /* since not able to change them */ } - info_buf->Attributes = cpu_to_le32(dosattr); - rc = CIFSSMBSetFileInfo(xid, tcon, info_buf, netfid, current->tgid); - kfree(info_buf); - if (rc != 0) - goto out_close; - cifsInode->cifsAttrs = dosattr; - /* silly-rename the file */ - CIFSSMBRenameOpenFile(xid, tcon, netfid, NULL, cifs_sb->local_nls, + /* rename the file */ + rc = CIFSSMBRenameOpenFile(xid, tcon, netfid, NULL, cifs_sb->local_nls, cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR); + if (rc != 0) { + rc = -ETXTBSY; + goto undo_setattr; + } - /* set DELETE_ON_CLOSE */ - rc = CIFSSMBSetFileDisposition(xid, tcon, true, netfid, current->tgid); - - /* - * some samba versions return -ENOENT when we try to set the file - * disposition here. Likely a samba bug, but work around it for now - */ - if (rc == -ENOENT) - rc = 0; + /* try to set DELETE_ON_CLOSE */ + if (!cifsInode->delete_pending) { + rc = CIFSSMBSetFileDisposition(xid, tcon, true, netfid, + current->tgid); + /* + * some samba versions return -ENOENT when we try to set the + * file disposition here. Likely a samba bug, but work around + * it for now. This means that some cifsXXX files may hang + * around after they shouldn't. + * + * BB: remove this hack after more servers have the fix + */ + if (rc == -ENOENT) + rc = 0; + else if (rc != 0) { + rc = -ETXTBSY; + goto undo_rename; + } + cifsInode->delete_pending = true; + } out_close: CIFSSMBClose(xid, tcon, netfid); out: + kfree(info_buf); return rc; + + /* + * reset everything back to the original state. Don't bother + * dealing with errors here since we can't do anything about + * them anyway. + */ +undo_rename: + CIFSSMBRenameOpenFile(xid, tcon, netfid, dentry->d_name.name, + cifs_sb->local_nls, cifs_sb->mnt_cifs_flags & + CIFS_MOUNT_MAP_SPECIAL_CHR); +undo_setattr: + if (dosattr != origattr) { + info_buf->Attributes = cpu_to_le32(origattr); + if (!CIFSSMBSetFileInfo(xid, tcon, info_buf, netfid, + current->tgid)) + cifsInode->cifsAttrs = origattr; + } + + goto out_close; } int cifs_unlink(struct inode *dir, struct dentry *dentry) @@ -878,7 +970,7 @@ psx_del_no_retry: } else if (rc == -ENOENT) { d_drop(dentry); } else if (rc == -ETXTBSY) { - rc = cifs_rename_pending_delete(full_path, inode, xid); + rc = cifs_rename_pending_delete(full_path, dentry, xid); if (rc == 0) drop_nlink(inode); } else if (rc == -EACCES && dosattr == 0) { @@ -1099,11 +1191,11 @@ mkdir_get_info: .device = 0, }; if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_SET_UID) { - args.uid = (__u64)current->fsuid; + args.uid = (__u64)current_fsuid(); if (inode->i_mode & S_ISGID) args.gid = (__u64)inode->i_gid; else - args.gid = (__u64)current->fsgid; + args.gid = (__u64)current_fsgid(); } else { args.uid = NO_CHANGE_64; args.gid = NO_CHANGE_64; @@ -1140,13 +1232,13 @@ mkdir_get_info: if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_SET_UID) { direntry->d_inode->i_uid = - current->fsuid; + current_fsuid(); if (inode->i_mode & S_ISGID) direntry->d_inode->i_gid = inode->i_gid; else direntry->d_inode->i_gid = - current->fsgid; + current_fsgid(); } } } @@ -1241,22 +1333,21 @@ cifs_do_rename(int xid, struct dentry *from_dentry, const char *fromPath, return rc; } -int cifs_rename(struct inode *source_inode, struct dentry *source_direntry, - struct inode *target_inode, struct dentry *target_direntry) +int cifs_rename(struct inode *source_dir, struct dentry *source_dentry, + struct inode *target_dir, struct dentry *target_dentry) { char *fromName = NULL; char *toName = NULL; struct cifs_sb_info *cifs_sb_source; struct cifs_sb_info *cifs_sb_target; - struct cifsTconInfo *pTcon; + struct cifsTconInfo *tcon; FILE_UNIX_BASIC_INFO *info_buf_source = NULL; FILE_UNIX_BASIC_INFO *info_buf_target; - int xid; - int rc; + int xid, rc, tmprc; - cifs_sb_target = CIFS_SB(target_inode->i_sb); - cifs_sb_source = CIFS_SB(source_inode->i_sb); - pTcon = cifs_sb_source->tcon; + cifs_sb_target = CIFS_SB(target_dir->i_sb); + cifs_sb_source = CIFS_SB(source_dir->i_sb); + tcon = cifs_sb_source->tcon; xid = GetXid(); @@ -1264,7 +1355,7 @@ int cifs_rename(struct inode *source_inode, struct dentry *source_direntry, * BB: this might be allowed if same server, but different share. * Consider adding support for this */ - if (pTcon != cifs_sb_target->tcon) { + if (tcon != cifs_sb_target->tcon) { rc = -EXDEV; goto cifs_rename_exit; } @@ -1273,65 +1364,67 @@ int cifs_rename(struct inode *source_inode, struct dentry *source_direntry, * we already have the rename sem so we do not need to * grab it again here to protect the path integrity */ - fromName = build_path_from_dentry(source_direntry); + fromName = build_path_from_dentry(source_dentry); if (fromName == NULL) { rc = -ENOMEM; goto cifs_rename_exit; } - toName = build_path_from_dentry(target_direntry); + toName = build_path_from_dentry(target_dentry); if (toName == NULL) { rc = -ENOMEM; goto cifs_rename_exit; } - rc = cifs_do_rename(xid, source_direntry, fromName, - target_direntry, toName); + rc = cifs_do_rename(xid, source_dentry, fromName, + target_dentry, toName); - if (rc == -EEXIST) { - if (pTcon->unix_ext) { - /* - * Are src and dst hardlinks of same inode? We can - * only tell with unix extensions enabled - */ - info_buf_source = - kmalloc(2 * sizeof(FILE_UNIX_BASIC_INFO), - GFP_KERNEL); - if (info_buf_source == NULL) - goto unlink_target; - - info_buf_target = info_buf_source + 1; - rc = CIFSSMBUnixQPathInfo(xid, pTcon, fromName, - info_buf_source, - cifs_sb_source->local_nls, - cifs_sb_source->mnt_cifs_flags & - CIFS_MOUNT_MAP_SPECIAL_CHR); - if (rc != 0) - goto unlink_target; - - rc = CIFSSMBUnixQPathInfo(xid, pTcon, - toName, info_buf_target, - cifs_sb_target->local_nls, - /* remap based on source sb */ - cifs_sb_source->mnt_cifs_flags & - CIFS_MOUNT_MAP_SPECIAL_CHR); + if (rc == -EEXIST && tcon->unix_ext) { + /* + * Are src and dst hardlinks of same inode? We can + * only tell with unix extensions enabled + */ + info_buf_source = + kmalloc(2 * sizeof(FILE_UNIX_BASIC_INFO), + GFP_KERNEL); + if (info_buf_source == NULL) { + rc = -ENOMEM; + goto cifs_rename_exit; + } + + info_buf_target = info_buf_source + 1; + tmprc = CIFSSMBUnixQPathInfo(xid, tcon, fromName, + info_buf_source, + cifs_sb_source->local_nls, + cifs_sb_source->mnt_cifs_flags & + CIFS_MOUNT_MAP_SPECIAL_CHR); + if (tmprc != 0) + goto unlink_target; + + tmprc = CIFSSMBUnixQPathInfo(xid, tcon, + toName, info_buf_target, + cifs_sb_target->local_nls, + /* remap based on source sb */ + cifs_sb_source->mnt_cifs_flags & + CIFS_MOUNT_MAP_SPECIAL_CHR); - if (rc == 0 && (info_buf_source->UniqueId == - info_buf_target->UniqueId)) - /* same file, POSIX says that this is a noop */ - goto cifs_rename_exit; - } /* else ... BB we could add the same check for Windows by + if (tmprc == 0 && (info_buf_source->UniqueId == + info_buf_target->UniqueId)) { + /* same file, POSIX says that this is a noop */ + rc = 0; + goto cifs_rename_exit; + } + } /* else ... BB we could add the same check for Windows by checking the UniqueId via FILE_INTERNAL_INFO */ + unlink_target: - /* - * we either can not tell the files are hardlinked (as with - * Windows servers) or files are not hardlinked. Delete the - * target manually before renaming to follow POSIX rather than - * Windows semantics - */ - cifs_unlink(target_inode, target_direntry); - rc = cifs_do_rename(xid, source_direntry, fromName, - target_direntry, toName); + if ((rc == -EACCES) || (rc == -EEXIST)) { + tmprc = cifs_unlink(target_dir, target_dentry); + if (tmprc) + goto cifs_rename_exit; + + rc = cifs_do_rename(xid, source_dentry, fromName, + target_dentry, toName); } cifs_rename_exit: diff --git a/fs/cifs/ioctl.c b/fs/cifs/ioctl.c index 0088a5b5256..f94650683a0 100644 --- a/fs/cifs/ioctl.c +++ b/fs/cifs/ioctl.c @@ -65,7 +65,7 @@ long cifs_ioctl(struct file *filep, unsigned int command, unsigned long arg) switch (command) { case CIFS_IOC_CHECKUMOUNT: cFYI(1, ("User unmount attempted")); - if (cifs_sb->mnt_uid == current->uid) + if (cifs_sb->mnt_uid == current_uid()) rc = 0; else { rc = -EACCES; diff --git a/fs/cifs/misc.c b/fs/cifs/misc.c index 88786ba02d2..4c89c572891 100644 --- a/fs/cifs/misc.c +++ b/fs/cifs/misc.c @@ -75,12 +75,12 @@ sesInfoAlloc(void) ret_buf = kzalloc(sizeof(struct cifsSesInfo), GFP_KERNEL); if (ret_buf) { - write_lock(&GlobalSMBSeslock); atomic_inc(&sesInfoAllocCount); ret_buf->status = CifsNew; - list_add(&ret_buf->cifsSessionList, &GlobalSMBSessionList); + ++ret_buf->ses_count; + INIT_LIST_HEAD(&ret_buf->smb_ses_list); + INIT_LIST_HEAD(&ret_buf->tcon_list); init_MUTEX(&ret_buf->sesSem); - write_unlock(&GlobalSMBSeslock); } return ret_buf; } @@ -93,14 +93,14 @@ sesInfoFree(struct cifsSesInfo *buf_to_free) return; } - write_lock(&GlobalSMBSeslock); atomic_dec(&sesInfoAllocCount); - list_del(&buf_to_free->cifsSessionList); - write_unlock(&GlobalSMBSeslock); kfree(buf_to_free->serverOS); kfree(buf_to_free->serverDomain); kfree(buf_to_free->serverNOS); - kfree(buf_to_free->password); + if (buf_to_free->password) { + memset(buf_to_free->password, 0, strlen(buf_to_free->password)); + kfree(buf_to_free->password); + } kfree(buf_to_free->domainName); kfree(buf_to_free); } @@ -111,17 +111,14 @@ tconInfoAlloc(void) struct cifsTconInfo *ret_buf; ret_buf = kzalloc(sizeof(struct cifsTconInfo), GFP_KERNEL); if (ret_buf) { - write_lock(&GlobalSMBSeslock); atomic_inc(&tconInfoAllocCount); - list_add(&ret_buf->cifsConnectionList, - &GlobalTreeConnectionList); ret_buf->tidStatus = CifsNew; + ++ret_buf->tc_count; INIT_LIST_HEAD(&ret_buf->openFileList); - init_MUTEX(&ret_buf->tconSem); + INIT_LIST_HEAD(&ret_buf->tcon_list); #ifdef CONFIG_CIFS_STATS spin_lock_init(&ret_buf->stat_lock); #endif - write_unlock(&GlobalSMBSeslock); } return ret_buf; } @@ -133,11 +130,12 @@ tconInfoFree(struct cifsTconInfo *buf_to_free) cFYI(1, ("Null buffer passed to tconInfoFree")); return; } - write_lock(&GlobalSMBSeslock); atomic_dec(&tconInfoAllocCount); - list_del(&buf_to_free->cifsConnectionList); - write_unlock(&GlobalSMBSeslock); kfree(buf_to_free->nativeFileSystem); + if (buf_to_free->password) { + memset(buf_to_free->password, 0, strlen(buf_to_free->password)); + kfree(buf_to_free->password); + } kfree(buf_to_free); } @@ -347,13 +345,13 @@ header_assemble(struct smb_hdr *buffer, char smb_command /* command */ , /* BB Add support for establishing new tCon and SMB Session */ /* with userid/password pairs found on the smb session */ /* for other target tcp/ip addresses BB */ - if (current->fsuid != treeCon->ses->linux_uid) { + if (current_fsuid() != treeCon->ses->linux_uid) { cFYI(1, ("Multiuser mode and UID " "did not match tcon uid")); - read_lock(&GlobalSMBSeslock); - list_for_each(temp_item, &GlobalSMBSessionList) { - ses = list_entry(temp_item, struct cifsSesInfo, cifsSessionList); - if (ses->linux_uid == current->fsuid) { + read_lock(&cifs_tcp_ses_lock); + list_for_each(temp_item, &treeCon->ses->server->smb_ses_list) { + ses = list_entry(temp_item, struct cifsSesInfo, smb_ses_list); + if (ses->linux_uid == current_fsuid()) { if (ses->server == treeCon->ses->server) { cFYI(1, ("found matching uid substitute right smb_uid")); buffer->Uid = ses->Suid; @@ -364,7 +362,7 @@ header_assemble(struct smb_hdr *buffer, char smb_command /* command */ , } } } - read_unlock(&GlobalSMBSeslock); + read_unlock(&cifs_tcp_ses_lock); } } } @@ -497,9 +495,10 @@ bool is_valid_oplock_break(struct smb_hdr *buf, struct TCP_Server_Info *srv) { struct smb_com_lock_req *pSMB = (struct smb_com_lock_req *)buf; - struct list_head *tmp; - struct list_head *tmp1; + struct list_head *tmp, *tmp1, *tmp2; + struct cifsSesInfo *ses; struct cifsTconInfo *tcon; + struct cifsInodeInfo *pCifsInode; struct cifsFileInfo *netfile; cFYI(1, ("Checking for oplock break or dnotify response")); @@ -554,42 +553,45 @@ is_valid_oplock_break(struct smb_hdr *buf, struct TCP_Server_Info *srv) return false; /* look up tcon based on tid & uid */ - read_lock(&GlobalSMBSeslock); - list_for_each(tmp, &GlobalTreeConnectionList) { - tcon = list_entry(tmp, struct cifsTconInfo, cifsConnectionList); - if ((tcon->tid == buf->Tid) && (srv == tcon->ses->server)) { + read_lock(&cifs_tcp_ses_lock); + list_for_each(tmp, &srv->smb_ses_list) { + ses = list_entry(tmp, struct cifsSesInfo, smb_ses_list); + list_for_each(tmp1, &ses->tcon_list) { + tcon = list_entry(tmp1, struct cifsTconInfo, tcon_list); + if (tcon->tid != buf->Tid) + continue; + cifs_stats_inc(&tcon->num_oplock_brks); - list_for_each(tmp1, &tcon->openFileList) { - netfile = list_entry(tmp1, struct cifsFileInfo, + write_lock(&GlobalSMBSeslock); + list_for_each(tmp2, &tcon->openFileList) { + netfile = list_entry(tmp2, struct cifsFileInfo, tlist); - if (pSMB->Fid == netfile->netfid) { - struct cifsInodeInfo *pCifsInode; - read_unlock(&GlobalSMBSeslock); - cFYI(1, - ("file id match, oplock break")); - pCifsInode = - CIFS_I(netfile->pInode); - pCifsInode->clientCanCacheAll = false; - if (pSMB->OplockLevel == 0) - pCifsInode->clientCanCacheRead - = false; - pCifsInode->oplockPending = true; - AllocOplockQEntry(netfile->pInode, - netfile->netfid, - tcon); - cFYI(1, - ("about to wake up oplock thread")); - if (oplockThread) - wake_up_process(oplockThread); - return true; - } + if (pSMB->Fid != netfile->netfid) + continue; + + write_unlock(&GlobalSMBSeslock); + read_unlock(&cifs_tcp_ses_lock); + cFYI(1, ("file id match, oplock break")); + pCifsInode = CIFS_I(netfile->pInode); + pCifsInode->clientCanCacheAll = false; + if (pSMB->OplockLevel == 0) + pCifsInode->clientCanCacheRead = false; + pCifsInode->oplockPending = true; + AllocOplockQEntry(netfile->pInode, + netfile->netfid, tcon); + cFYI(1, ("about to wake up oplock thread")); + if (oplockThread) + wake_up_process(oplockThread); + + return true; } - read_unlock(&GlobalSMBSeslock); + write_unlock(&GlobalSMBSeslock); + read_unlock(&cifs_tcp_ses_lock); cFYI(1, ("No matching file for oplock break")); return true; } } - read_unlock(&GlobalSMBSeslock); + read_unlock(&cifs_tcp_ses_lock); cFYI(1, ("Can not process oplock break for non-existent connection")); return true; } diff --git a/fs/cifs/readdir.c b/fs/cifs/readdir.c index 765adf12d54..9f51f9bf029 100644 --- a/fs/cifs/readdir.c +++ b/fs/cifs/readdir.c @@ -741,11 +741,14 @@ static int find_cifs_entry(const int xid, struct cifsTconInfo *pTcon, (index_to_find < first_entry_in_buffer)) { /* close and restart search */ cFYI(1, ("search backing up - close and restart search")); + write_lock(&GlobalSMBSeslock); if (!cifsFile->srch_inf.endOfSearch && !cifsFile->invalidHandle) { cifsFile->invalidHandle = true; + write_unlock(&GlobalSMBSeslock); CIFSFindClose(xid, pTcon, cifsFile->netfid); - } + } else + write_unlock(&GlobalSMBSeslock); if (cifsFile->srch_inf.ntwrk_buf_start) { cFYI(1, ("freeing SMB ff cache buf on search rewind")); if (cifsFile->srch_inf.smallBuf) @@ -762,14 +765,15 @@ static int find_cifs_entry(const int xid, struct cifsTconInfo *pTcon, rc)); return rc; } + cifs_save_resume_key(cifsFile->srch_inf.last_entry, cifsFile); } while ((index_to_find >= cifsFile->srch_inf.index_of_last_entry) && (rc == 0) && !cifsFile->srch_inf.endOfSearch) { cFYI(1, ("calling findnext2")); - cifs_save_resume_key(cifsFile->srch_inf.last_entry, cifsFile); rc = CIFSFindNext(xid, pTcon, cifsFile->netfid, &cifsFile->srch_inf); + cifs_save_resume_key(cifsFile->srch_inf.last_entry, cifsFile); if (rc) return -ENOENT; } diff --git a/fs/cifs/sess.c b/fs/cifs/sess.c index 2851d5da0c8..5f22de7b79a 100644 --- a/fs/cifs/sess.c +++ b/fs/cifs/sess.c @@ -417,7 +417,10 @@ CIFS_SessSetup(unsigned int xid, struct cifsSesInfo *ses, int first_time, /* BB calculate hash with password */ /* and copy into bcc */ - calc_lanman_hash(ses, lnm_session_key); + calc_lanman_hash(ses->password, ses->server->cryptKey, + ses->server->secMode & SECMODE_PW_ENCRYPT ? + true : false, lnm_session_key); + ses->flags |= CIFS_SES_LANMAN; memcpy(bcc_ptr, (char *)lnm_session_key, CIFS_SESS_KEY_SIZE); bcc_ptr += CIFS_SESS_KEY_SIZE; diff --git a/fs/cifs/smbdes.c b/fs/cifs/smbdes.c index 04943c976f9..224a1f47896 100644 --- a/fs/cifs/smbdes.c +++ b/fs/cifs/smbdes.c @@ -318,7 +318,8 @@ str_to_key(unsigned char *str, unsigned char *key) } static void -smbhash(unsigned char *out, unsigned char *in, unsigned char *key, int forw) +smbhash(unsigned char *out, const unsigned char *in, unsigned char *key, + int forw) { int i; char *outb; /* outb[64] */ @@ -363,7 +364,7 @@ E_P16(unsigned char *p14, unsigned char *p16) } void -E_P24(unsigned char *p21, unsigned char *c8, unsigned char *p24) +E_P24(unsigned char *p21, const unsigned char *c8, unsigned char *p24) { smbhash(p24, c8, p21, 1); smbhash(p24 + 8, c8, p21 + 7, 1); diff --git a/fs/cifs/smbencrypt.c b/fs/cifs/smbencrypt.c index ff3232fa101..93fb09a99c6 100644 --- a/fs/cifs/smbencrypt.c +++ b/fs/cifs/smbencrypt.c @@ -49,9 +49,10 @@ /*The following definitions come from libsmb/smbencrypt.c */ -void SMBencrypt(unsigned char *passwd, unsigned char *c8, unsigned char *p24); +void SMBencrypt(unsigned char *passwd, const unsigned char *c8, + unsigned char *p24); void E_md4hash(const unsigned char *passwd, unsigned char *p16); -static void SMBOWFencrypt(unsigned char passwd[16], unsigned char *c8, +static void SMBOWFencrypt(unsigned char passwd[16], const unsigned char *c8, unsigned char p24[24]); void SMBNTencrypt(unsigned char *passwd, unsigned char *c8, unsigned char *p24); @@ -61,7 +62,7 @@ void SMBNTencrypt(unsigned char *passwd, unsigned char *c8, unsigned char *p24); encrypted password into p24 */ /* Note that password must be uppercased and null terminated */ void -SMBencrypt(unsigned char *passwd, unsigned char *c8, unsigned char *p24) +SMBencrypt(unsigned char *passwd, const unsigned char *c8, unsigned char *p24) { unsigned char p14[15], p21[21]; @@ -212,7 +213,7 @@ ntv2_owf_gen(const unsigned char owf[16], const char *user_n, /* Does the des encryption from the NT or LM MD4 hash. */ static void -SMBOWFencrypt(unsigned char passwd[16], unsigned char *c8, +SMBOWFencrypt(unsigned char passwd[16], const unsigned char *c8, unsigned char p24[24]) { unsigned char p21[21]; diff --git a/fs/cifs/transport.c b/fs/cifs/transport.c index bf0e6d8e382..7ebe6599ed3 100644 --- a/fs/cifs/transport.c +++ b/fs/cifs/transport.c @@ -37,15 +37,11 @@ extern mempool_t *cifs_mid_poolp; extern struct kmem_cache *cifs_oplock_cachep; static struct mid_q_entry * -AllocMidQEntry(const struct smb_hdr *smb_buffer, struct cifsSesInfo *ses) +AllocMidQEntry(const struct smb_hdr *smb_buffer, struct TCP_Server_Info *server) { struct mid_q_entry *temp; - if (ses == NULL) { - cERROR(1, ("Null session passed in to AllocMidQEntry")); - return NULL; - } - if (ses->server == NULL) { + if (server == NULL) { cERROR(1, ("Null TCP session in AllocMidQEntry")); return NULL; } @@ -62,12 +58,11 @@ AllocMidQEntry(const struct smb_hdr *smb_buffer, struct cifsSesInfo *ses) /* do_gettimeofday(&temp->when_sent);*/ /* easier to use jiffies */ /* when mid allocated can be before when sent */ temp->when_alloc = jiffies; - temp->ses = ses; temp->tsk = current; } spin_lock(&GlobalMid_Lock); - list_add_tail(&temp->qhead, &ses->server->pending_mid_q); + list_add_tail(&temp->qhead, &server->pending_mid_q); atomic_inc(&midCount); temp->midState = MID_REQUEST_ALLOCATED; spin_unlock(&GlobalMid_Lock); @@ -161,7 +156,7 @@ void DeleteTconOplockQEntries(struct cifsTconInfo *tcon) int smb_send(struct socket *ssocket, struct smb_hdr *smb_buffer, - unsigned int smb_buf_length, struct sockaddr *sin) + unsigned int smb_buf_length, struct sockaddr *sin, bool noblocksnd) { int rc = 0; int i = 0; @@ -178,7 +173,10 @@ smb_send(struct socket *ssocket, struct smb_hdr *smb_buffer, smb_msg.msg_namelen = sizeof(struct sockaddr); smb_msg.msg_control = NULL; smb_msg.msg_controllen = 0; - smb_msg.msg_flags = MSG_DONTWAIT + MSG_NOSIGNAL; /* BB add more flags?*/ + if (noblocksnd) + smb_msg.msg_flags = MSG_DONTWAIT + MSG_NOSIGNAL; + else + smb_msg.msg_flags = MSG_NOSIGNAL; /* smb header is converted in header_assemble. bcc and rest of SMB word area, and byte area if necessary, is converted to littleendian in @@ -229,8 +227,8 @@ smb_send(struct socket *ssocket, struct smb_hdr *smb_buffer, } static int -smb_send2(struct socket *ssocket, struct kvec *iov, int n_vec, - struct sockaddr *sin) +smb_send2(struct TCP_Server_Info *server, struct kvec *iov, int n_vec, + struct sockaddr *sin, bool noblocksnd) { int rc = 0; int i = 0; @@ -240,6 +238,7 @@ smb_send2(struct socket *ssocket, struct kvec *iov, int n_vec, unsigned int total_len; int first_vec = 0; unsigned int smb_buf_length = smb_buffer->smb_buf_length; + struct socket *ssocket = server->ssocket; if (ssocket == NULL) return -ENOTSOCK; /* BB eventually add reconnect code here */ @@ -248,7 +247,10 @@ smb_send2(struct socket *ssocket, struct kvec *iov, int n_vec, smb_msg.msg_namelen = sizeof(struct sockaddr); smb_msg.msg_control = NULL; smb_msg.msg_controllen = 0; - smb_msg.msg_flags = MSG_DONTWAIT + MSG_NOSIGNAL; /* BB add more flags?*/ + if (noblocksnd) + smb_msg.msg_flags = MSG_DONTWAIT + MSG_NOSIGNAL; + else + smb_msg.msg_flags = MSG_NOSIGNAL; /* smb header is converted in header_assemble. bcc and rest of SMB word area, and byte area if necessary, is converted to littleendian in @@ -283,8 +285,11 @@ smb_send2(struct socket *ssocket, struct kvec *iov, int n_vec, if (rc < 0) break; - if (rc >= total_len) { - WARN_ON(rc > total_len); + if (rc == total_len) { + total_len = 0; + break; + } else if (rc > total_len) { + cERROR(1, ("sent %d requested %d", rc, total_len)); break; } if (rc == 0) { @@ -312,6 +317,16 @@ smb_send2(struct socket *ssocket, struct kvec *iov, int n_vec, i = 0; /* in case we get ENOSPC on the next send */ } + if ((total_len > 0) && (total_len != smb_buf_length + 4)) { + cFYI(1, ("partial send (%d remaining), terminating session", + total_len)); + /* If we have only sent part of an SMB then the next SMB + could be taken as the remainder of this one. We need + to kill the socket so the server throws away the partial + SMB */ + server->tcpStatus = CifsNeedReconnect; + } + if (rc < 0) { cERROR(1, ("Error %d sending data on socket to server", rc)); } else @@ -329,37 +344,38 @@ static int wait_for_free_request(struct cifsSesInfo *ses, const int long_op) if (long_op == CIFS_ASYNC_OP) { /* oplock breaks must not be held up */ atomic_inc(&ses->server->inFlight); - } else { - spin_lock(&GlobalMid_Lock); - while (1) { - if (atomic_read(&ses->server->inFlight) >= - cifs_max_pending){ - spin_unlock(&GlobalMid_Lock); + return 0; + } + + spin_lock(&GlobalMid_Lock); + while (1) { + if (atomic_read(&ses->server->inFlight) >= + cifs_max_pending){ + spin_unlock(&GlobalMid_Lock); #ifdef CONFIG_CIFS_STATS2 - atomic_inc(&ses->server->num_waiters); + atomic_inc(&ses->server->num_waiters); #endif - wait_event(ses->server->request_q, - atomic_read(&ses->server->inFlight) - < cifs_max_pending); + wait_event(ses->server->request_q, + atomic_read(&ses->server->inFlight) + < cifs_max_pending); #ifdef CONFIG_CIFS_STATS2 - atomic_dec(&ses->server->num_waiters); + atomic_dec(&ses->server->num_waiters); #endif - spin_lock(&GlobalMid_Lock); - } else { - if (ses->server->tcpStatus == CifsExiting) { - spin_unlock(&GlobalMid_Lock); - return -ENOENT; - } - - /* can not count locking commands against total - as they are allowed to block on server */ - - /* update # of requests on the wire to server */ - if (long_op != CIFS_BLOCKING_OP) - atomic_inc(&ses->server->inFlight); + spin_lock(&GlobalMid_Lock); + } else { + if (ses->server->tcpStatus == CifsExiting) { spin_unlock(&GlobalMid_Lock); - break; + return -ENOENT; } + + /* can not count locking commands against total + as they are allowed to block on server */ + + /* update # of requests on the wire to server */ + if (long_op != CIFS_BLOCKING_OP) + atomic_inc(&ses->server->inFlight); + spin_unlock(&GlobalMid_Lock); + break; } } return 0; @@ -370,17 +386,21 @@ static int allocate_mid(struct cifsSesInfo *ses, struct smb_hdr *in_buf, { if (ses->server->tcpStatus == CifsExiting) { return -ENOENT; - } else if (ses->server->tcpStatus == CifsNeedReconnect) { + } + + if (ses->server->tcpStatus == CifsNeedReconnect) { cFYI(1, ("tcp session dead - return to caller to retry")); return -EAGAIN; - } else if (ses->status != CifsGood) { + } + + if (ses->status != CifsGood) { /* check if SMB session is bad because we are setting it up */ if ((in_buf->Command != SMB_COM_SESSION_SETUP_ANDX) && (in_buf->Command != SMB_COM_NEGOTIATE)) return -EAGAIN; /* else ok - we are setting up session */ } - *ppmidQ = AllocMidQEntry(in_buf, ses); + *ppmidQ = AllocMidQEntry(in_buf, ses->server); if (*ppmidQ == NULL) return -ENOMEM; return 0; @@ -395,11 +415,8 @@ static int wait_for_response(struct cifsSesInfo *ses, for (;;) { curr_timeout = timeout + jiffies; - wait_event(ses->server->response_q, - (!(midQ->midState == MID_REQUEST_SUBMITTED)) || - time_after(jiffies, curr_timeout) || - ((ses->server->tcpStatus != CifsGood) && - (ses->server->tcpStatus != CifsNew))); + wait_event_timeout(ses->server->response_q, + midQ->midState != MID_REQUEST_SUBMITTED, timeout); if (time_after(jiffies, curr_timeout) && (midQ->midState == MID_REQUEST_SUBMITTED) && @@ -501,11 +518,11 @@ SendReceive2(const unsigned int xid, struct cifsSesInfo *ses, and avoid races inside tcp sendmsg code that could cause corruption of smb data */ - down(&ses->server->tcpSem); + mutex_lock(&ses->server->srv_mutex); rc = allocate_mid(ses, in_buf, &midQ); if (rc) { - up(&ses->server->tcpSem); + mutex_unlock(&ses->server->srv_mutex); cifs_small_buf_release(in_buf); /* Update # of requests on wire to server */ atomic_dec(&ses->server->inFlight); @@ -513,19 +530,25 @@ SendReceive2(const unsigned int xid, struct cifsSesInfo *ses, return rc; } rc = cifs_sign_smb2(iov, n_vec, ses->server, &midQ->sequence_number); + if (rc) { + mutex_unlock(&ses->server->srv_mutex); + cifs_small_buf_release(in_buf); + goto out; + } midQ->midState = MID_REQUEST_SUBMITTED; #ifdef CONFIG_CIFS_STATS2 atomic_inc(&ses->server->inSend); #endif - rc = smb_send2(ses->server->ssocket, iov, n_vec, - (struct sockaddr *) &(ses->server->addr.sockAddr)); + rc = smb_send2(ses->server, iov, n_vec, + (struct sockaddr *) &(ses->server->addr.sockAddr), + ses->server->noblocksnd); #ifdef CONFIG_CIFS_STATS2 atomic_dec(&ses->server->inSend); midQ->when_sent = jiffies; #endif - up(&ses->server->tcpSem); + mutex_unlock(&ses->server->srv_mutex); cifs_small_buf_release(in_buf); if (rc < 0) @@ -560,10 +583,8 @@ SendReceive2(const unsigned int xid, struct cifsSesInfo *ses, wait_for_response(ses, midQ, timeout, 10 * HZ); spin_lock(&GlobalMid_Lock); - if (midQ->resp_buf) { - spin_unlock(&GlobalMid_Lock); - receive_len = midQ->resp_buf->smb_buf_length; - } else { + + if (midQ->resp_buf == NULL) { cERROR(1, ("No response to cmd %d mid %d", midQ->command, midQ->mid)); if (midQ->midState == MID_REQUEST_SUBMITTED) { @@ -591,53 +612,59 @@ SendReceive2(const unsigned int xid, struct cifsSesInfo *ses, return rc; } + spin_unlock(&GlobalMid_Lock); + receive_len = midQ->resp_buf->smb_buf_length; + if (receive_len > CIFSMaxBufSize + MAX_CIFS_HDR_SIZE) { cERROR(1, ("Frame too large received. Length: %d Xid: %d", receive_len, xid)); rc = -EIO; - } else { /* rcvd frame is ok */ - if (midQ->resp_buf && - (midQ->midState == MID_RESPONSE_RECEIVED)) { - - iov[0].iov_base = (char *)midQ->resp_buf; - if (midQ->largeBuf) - *pRespBufType = CIFS_LARGE_BUFFER; - else - *pRespBufType = CIFS_SMALL_BUFFER; - iov[0].iov_len = receive_len + 4; - - dump_smb(midQ->resp_buf, 80); - /* convert the length into a more usable form */ - if ((receive_len > 24) && - (ses->server->secMode & (SECMODE_SIGN_REQUIRED | - SECMODE_SIGN_ENABLED))) { - rc = cifs_verify_signature(midQ->resp_buf, + goto out; + } + + /* rcvd frame is ok */ + + if (midQ->resp_buf && + (midQ->midState == MID_RESPONSE_RECEIVED)) { + + iov[0].iov_base = (char *)midQ->resp_buf; + if (midQ->largeBuf) + *pRespBufType = CIFS_LARGE_BUFFER; + else + *pRespBufType = CIFS_SMALL_BUFFER; + iov[0].iov_len = receive_len + 4; + + dump_smb(midQ->resp_buf, 80); + /* convert the length into a more usable form */ + if ((receive_len > 24) && + (ses->server->secMode & (SECMODE_SIGN_REQUIRED | + SECMODE_SIGN_ENABLED))) { + rc = cifs_verify_signature(midQ->resp_buf, &ses->server->mac_signing_key, midQ->sequence_number+1); - if (rc) { - cERROR(1, ("Unexpected SMB signature")); - /* BB FIXME add code to kill session */ - } + if (rc) { + cERROR(1, ("Unexpected SMB signature")); + /* BB FIXME add code to kill session */ } - - /* BB special case reconnect tid and uid here? */ - rc = map_smb_to_linux_error(midQ->resp_buf, - flags & CIFS_LOG_ERROR); - - /* convert ByteCount if necessary */ - if (receive_len >= sizeof(struct smb_hdr) - 4 - /* do not count RFC1001 header */ + - (2 * midQ->resp_buf->WordCount) + 2 /* bcc */ ) - BCC(midQ->resp_buf) = - le16_to_cpu(BCC_LE(midQ->resp_buf)); - if ((flags & CIFS_NO_RESP) == 0) - midQ->resp_buf = NULL; /* mark it so buf will - not be freed by - DeleteMidQEntry */ - } else { - rc = -EIO; - cFYI(1, ("Bad MID state?")); } + + /* BB special case reconnect tid and uid here? */ + rc = map_smb_to_linux_error(midQ->resp_buf, + flags & CIFS_LOG_ERROR); + + /* convert ByteCount if necessary */ + if (receive_len >= sizeof(struct smb_hdr) - 4 + /* do not count RFC1001 header */ + + (2 * midQ->resp_buf->WordCount) + 2 /* bcc */ ) + BCC(midQ->resp_buf) = + le16_to_cpu(BCC_LE(midQ->resp_buf)); + if ((flags & CIFS_NO_RESP) == 0) + midQ->resp_buf = NULL; /* mark it so buf will + not be freed by + DeleteMidQEntry */ + } else { + rc = -EIO; + cFYI(1, ("Bad MID state?")); } out: @@ -674,6 +701,12 @@ SendReceive(const unsigned int xid, struct cifsSesInfo *ses, to the same server. We may make this configurable later or use ses->maxReq */ + if (in_buf->smb_buf_length > CIFSMaxBufSize + MAX_CIFS_HDR_SIZE - 4) { + cERROR(1, ("Illegal length, greater than maximum frame, %d", + in_buf->smb_buf_length)); + return -EIO; + } + rc = wait_for_free_request(ses, long_op); if (rc) return rc; @@ -682,41 +715,35 @@ SendReceive(const unsigned int xid, struct cifsSesInfo *ses, and avoid races inside tcp sendmsg code that could cause corruption of smb data */ - down(&ses->server->tcpSem); + mutex_lock(&ses->server->srv_mutex); rc = allocate_mid(ses, in_buf, &midQ); if (rc) { - up(&ses->server->tcpSem); + mutex_unlock(&ses->server->srv_mutex); /* Update # of requests on wire to server */ atomic_dec(&ses->server->inFlight); wake_up(&ses->server->request_q); return rc; } - if (in_buf->smb_buf_length > CIFSMaxBufSize + MAX_CIFS_HDR_SIZE - 4) { - cERROR(1, ("Illegal length, greater than maximum frame, %d", - in_buf->smb_buf_length)); - DeleteMidQEntry(midQ); - up(&ses->server->tcpSem); - /* Update # of requests on wire to server */ - atomic_dec(&ses->server->inFlight); - wake_up(&ses->server->request_q); - return -EIO; - } - rc = cifs_sign_smb(in_buf, ses->server, &midQ->sequence_number); + if (rc) { + mutex_unlock(&ses->server->srv_mutex); + goto out; + } midQ->midState = MID_REQUEST_SUBMITTED; #ifdef CONFIG_CIFS_STATS2 atomic_inc(&ses->server->inSend); #endif rc = smb_send(ses->server->ssocket, in_buf, in_buf->smb_buf_length, - (struct sockaddr *) &(ses->server->addr.sockAddr)); + (struct sockaddr *) &(ses->server->addr.sockAddr), + ses->server->noblocksnd); #ifdef CONFIG_CIFS_STATS2 atomic_dec(&ses->server->inSend); midQ->when_sent = jiffies; #endif - up(&ses->server->tcpSem); + mutex_unlock(&ses->server->srv_mutex); if (rc < 0) goto out; @@ -750,10 +777,7 @@ SendReceive(const unsigned int xid, struct cifsSesInfo *ses, wait_for_response(ses, midQ, timeout, 10 * HZ); spin_lock(&GlobalMid_Lock); - if (midQ->resp_buf) { - spin_unlock(&GlobalMid_Lock); - receive_len = midQ->resp_buf->smb_buf_length; - } else { + if (midQ->resp_buf == NULL) { cERROR(1, ("No response for cmd %d mid %d", midQ->command, midQ->mid)); if (midQ->midState == MID_REQUEST_SUBMITTED) { @@ -781,47 +805,52 @@ SendReceive(const unsigned int xid, struct cifsSesInfo *ses, return rc; } + spin_unlock(&GlobalMid_Lock); + receive_len = midQ->resp_buf->smb_buf_length; + if (receive_len > CIFSMaxBufSize + MAX_CIFS_HDR_SIZE) { cERROR(1, ("Frame too large received. Length: %d Xid: %d", receive_len, xid)); rc = -EIO; - } else { /* rcvd frame is ok */ - - if (midQ->resp_buf && out_buf - && (midQ->midState == MID_RESPONSE_RECEIVED)) { - out_buf->smb_buf_length = receive_len; - memcpy((char *)out_buf + 4, - (char *)midQ->resp_buf + 4, - receive_len); - - dump_smb(out_buf, 92); - /* convert the length into a more usable form */ - if ((receive_len > 24) && - (ses->server->secMode & (SECMODE_SIGN_REQUIRED | - SECMODE_SIGN_ENABLED))) { - rc = cifs_verify_signature(out_buf, + goto out; + } + + /* rcvd frame is ok */ + + if (midQ->resp_buf && out_buf + && (midQ->midState == MID_RESPONSE_RECEIVED)) { + out_buf->smb_buf_length = receive_len; + memcpy((char *)out_buf + 4, + (char *)midQ->resp_buf + 4, + receive_len); + + dump_smb(out_buf, 92); + /* convert the length into a more usable form */ + if ((receive_len > 24) && + (ses->server->secMode & (SECMODE_SIGN_REQUIRED | + SECMODE_SIGN_ENABLED))) { + rc = cifs_verify_signature(out_buf, &ses->server->mac_signing_key, midQ->sequence_number+1); - if (rc) { - cERROR(1, ("Unexpected SMB signature")); - /* BB FIXME add code to kill session */ - } + if (rc) { + cERROR(1, ("Unexpected SMB signature")); + /* BB FIXME add code to kill session */ } + } - *pbytes_returned = out_buf->smb_buf_length; + *pbytes_returned = out_buf->smb_buf_length; - /* BB special case reconnect tid and uid here? */ - rc = map_smb_to_linux_error(out_buf, 0 /* no log */ ); + /* BB special case reconnect tid and uid here? */ + rc = map_smb_to_linux_error(out_buf, 0 /* no log */ ); - /* convert ByteCount if necessary */ - if (receive_len >= sizeof(struct smb_hdr) - 4 - /* do not count RFC1001 header */ + - (2 * out_buf->WordCount) + 2 /* bcc */ ) - BCC(out_buf) = le16_to_cpu(BCC_LE(out_buf)); - } else { - rc = -EIO; - cERROR(1, ("Bad MID state?")); - } + /* convert ByteCount if necessary */ + if (receive_len >= sizeof(struct smb_hdr) - 4 + /* do not count RFC1001 header */ + + (2 * out_buf->WordCount) + 2 /* bcc */ ) + BCC(out_buf) = le16_to_cpu(BCC_LE(out_buf)); + } else { + rc = -EIO; + cERROR(1, ("Bad MID state?")); } out: @@ -844,15 +873,16 @@ send_nt_cancel(struct cifsTconInfo *tcon, struct smb_hdr *in_buf, header_assemble(in_buf, SMB_COM_NT_CANCEL, tcon, 0); in_buf->Mid = mid; - down(&ses->server->tcpSem); + mutex_lock(&ses->server->srv_mutex); rc = cifs_sign_smb(in_buf, ses->server, &midQ->sequence_number); if (rc) { - up(&ses->server->tcpSem); + mutex_unlock(&ses->server->srv_mutex); return rc; } rc = smb_send(ses->server->ssocket, in_buf, in_buf->smb_buf_length, - (struct sockaddr *) &(ses->server->addr.sockAddr)); - up(&ses->server->tcpSem); + (struct sockaddr *) &(ses->server->addr.sockAddr), + ses->server->noblocksnd); + mutex_unlock(&ses->server->srv_mutex); return rc; } @@ -910,6 +940,12 @@ SendReceiveBlockingLock(const unsigned int xid, struct cifsTconInfo *tcon, to the same server. We may make this configurable later or use ses->maxReq */ + if (in_buf->smb_buf_length > CIFSMaxBufSize + MAX_CIFS_HDR_SIZE - 4) { + cERROR(1, ("Illegal length, greater than maximum frame, %d", + in_buf->smb_buf_length)); + return -EIO; + } + rc = wait_for_free_request(ses, CIFS_BLOCKING_OP); if (rc) return rc; @@ -918,35 +954,33 @@ SendReceiveBlockingLock(const unsigned int xid, struct cifsTconInfo *tcon, and avoid races inside tcp sendmsg code that could cause corruption of smb data */ - down(&ses->server->tcpSem); + mutex_lock(&ses->server->srv_mutex); rc = allocate_mid(ses, in_buf, &midQ); if (rc) { - up(&ses->server->tcpSem); + mutex_unlock(&ses->server->srv_mutex); return rc; } - if (in_buf->smb_buf_length > CIFSMaxBufSize + MAX_CIFS_HDR_SIZE - 4) { - up(&ses->server->tcpSem); - cERROR(1, ("Illegal length, greater than maximum frame, %d", - in_buf->smb_buf_length)); + rc = cifs_sign_smb(in_buf, ses->server, &midQ->sequence_number); + if (rc) { DeleteMidQEntry(midQ); - return -EIO; + mutex_unlock(&ses->server->srv_mutex); + return rc; } - rc = cifs_sign_smb(in_buf, ses->server, &midQ->sequence_number); - midQ->midState = MID_REQUEST_SUBMITTED; #ifdef CONFIG_CIFS_STATS2 atomic_inc(&ses->server->inSend); #endif rc = smb_send(ses->server->ssocket, in_buf, in_buf->smb_buf_length, - (struct sockaddr *) &(ses->server->addr.sockAddr)); + (struct sockaddr *) &(ses->server->addr.sockAddr), + ses->server->noblocksnd); #ifdef CONFIG_CIFS_STATS2 atomic_dec(&ses->server->inSend); midQ->when_sent = jiffies; #endif - up(&ses->server->tcpSem); + mutex_unlock(&ses->server->srv_mutex); if (rc < 0) { DeleteMidQEntry(midQ); @@ -1028,44 +1062,48 @@ SendReceiveBlockingLock(const unsigned int xid, struct cifsTconInfo *tcon, cERROR(1, ("Frame too large received. Length: %d Xid: %d", receive_len, xid)); rc = -EIO; - } else { /* rcvd frame is ok */ - - if (midQ->resp_buf && out_buf - && (midQ->midState == MID_RESPONSE_RECEIVED)) { - out_buf->smb_buf_length = receive_len; - memcpy((char *)out_buf + 4, - (char *)midQ->resp_buf + 4, - receive_len); - - dump_smb(out_buf, 92); - /* convert the length into a more usable form */ - if ((receive_len > 24) && - (ses->server->secMode & (SECMODE_SIGN_REQUIRED | - SECMODE_SIGN_ENABLED))) { - rc = cifs_verify_signature(out_buf, - &ses->server->mac_signing_key, - midQ->sequence_number+1); - if (rc) { - cERROR(1, ("Unexpected SMB signature")); - /* BB FIXME add code to kill session */ - } - } + goto out; + } - *pbytes_returned = out_buf->smb_buf_length; + /* rcvd frame is ok */ - /* BB special case reconnect tid and uid here? */ - rc = map_smb_to_linux_error(out_buf, 0 /* no log */ ); + if ((out_buf == NULL) || (midQ->midState != MID_RESPONSE_RECEIVED)) { + rc = -EIO; + cERROR(1, ("Bad MID state?")); + goto out; + } - /* convert ByteCount if necessary */ - if (receive_len >= sizeof(struct smb_hdr) - 4 - /* do not count RFC1001 header */ + - (2 * out_buf->WordCount) + 2 /* bcc */ ) - BCC(out_buf) = le16_to_cpu(BCC_LE(out_buf)); - } else { - rc = -EIO; - cERROR(1, ("Bad MID state?")); + out_buf->smb_buf_length = receive_len; + memcpy((char *)out_buf + 4, + (char *)midQ->resp_buf + 4, + receive_len); + + dump_smb(out_buf, 92); + /* convert the length into a more usable form */ + if ((receive_len > 24) && + (ses->server->secMode & (SECMODE_SIGN_REQUIRED | + SECMODE_SIGN_ENABLED))) { + rc = cifs_verify_signature(out_buf, + &ses->server->mac_signing_key, + midQ->sequence_number+1); + if (rc) { + cERROR(1, ("Unexpected SMB signature")); + /* BB FIXME add code to kill session */ } } + + *pbytes_returned = out_buf->smb_buf_length; + + /* BB special case reconnect tid and uid here? */ + rc = map_smb_to_linux_error(out_buf, 0 /* no log */ ); + + /* convert ByteCount if necessary */ + if (receive_len >= sizeof(struct smb_hdr) - 4 + /* do not count RFC1001 header */ + + (2 * out_buf->WordCount) + 2 /* bcc */ ) + BCC(out_buf) = le16_to_cpu(BCC_LE(out_buf)); + +out: DeleteMidQEntry(midQ); if (rstart && rc == -EACCES) return -ERESTARTSYS; diff --git a/fs/coda/cache.c b/fs/coda/cache.c index 8a2370341c7..a5bf5771a22 100644 --- a/fs/coda/cache.c +++ b/fs/coda/cache.c @@ -32,8 +32,8 @@ void coda_cache_enter(struct inode *inode, int mask) struct coda_inode_info *cii = ITOC(inode); cii->c_cached_epoch = atomic_read(&permission_epoch); - if (cii->c_uid != current->fsuid) { - cii->c_uid = current->fsuid; + if (cii->c_uid != current_fsuid()) { + cii->c_uid = current_fsuid(); cii->c_cached_perm = mask; } else cii->c_cached_perm |= mask; @@ -60,7 +60,7 @@ int coda_cache_check(struct inode *inode, int mask) int hit; hit = (mask & cii->c_cached_perm) == mask && - cii->c_uid == current->fsuid && + cii->c_uid == current_fsuid() && cii->c_cached_epoch == atomic_read(&permission_epoch); return hit; diff --git a/fs/coda/dir.c b/fs/coda/dir.c index c5916228243..75b1fa90b2c 100644 --- a/fs/coda/dir.c +++ b/fs/coda/dir.c @@ -146,6 +146,9 @@ int coda_permission(struct inode *inode, int mask) if (!mask) return 0; + if ((mask & MAY_EXEC) && !execute_ok(inode)) + return -EACCES; + lock_kernel(); if (coda_cache_check(inode, mask)) diff --git a/fs/coda/file.c b/fs/coda/file.c index 29137ff3ca6..466303db2df 100644 --- a/fs/coda/file.c +++ b/fs/coda/file.c @@ -13,6 +13,7 @@ #include <linux/file.h> #include <linux/fs.h> #include <linux/stat.h> +#include <linux/cred.h> #include <linux/errno.h> #include <linux/smp_lock.h> #include <linux/string.h> @@ -174,7 +175,7 @@ int coda_release(struct inode *coda_inode, struct file *coda_file) BUG_ON(!cfi || cfi->cfi_magic != CODA_MAGIC); err = venus_close(coda_inode->i_sb, coda_i2f(coda_inode), - coda_flags, coda_file->f_uid); + coda_flags, coda_file->f_cred->fsuid); host_inode = cfi->cfi_container->f_path.dentry->d_inode; cii = ITOC(coda_inode); diff --git a/fs/coda/pioctl.c b/fs/coda/pioctl.c index c51365422aa..773f2ce9aa0 100644 --- a/fs/coda/pioctl.c +++ b/fs/coda/pioctl.c @@ -43,7 +43,7 @@ const struct file_operations coda_ioctl_operations = { /* the coda pioctl inode ops */ static int coda_ioctl_permission(struct inode *inode, int mask) { - return 0; + return (mask & MAY_EXEC) ? -EACCES : 0; } static int coda_pioctl(struct inode * inode, struct file * filp, diff --git a/fs/coda/psdev.c b/fs/coda/psdev.c index 0d9b80ec689..0376ac66c44 100644 --- a/fs/coda/psdev.c +++ b/fs/coda/psdev.c @@ -2,7 +2,7 @@ * An implementation of a loadable kernel mode driver providing * multiple kernel/user space bidirectional communications links. * - * Author: Alan Cox <alan@redhat.com> + * Author: Alan Cox <alan@lxorguk.ukuu.org.uk> * * This program is free software; you can redistribute it and/or * modify it under the terms of the GNU General Public License @@ -362,9 +362,8 @@ static int init_coda_psdev(void) goto out_chrdev; } for (i = 0; i < MAX_CODADEVS; i++) - device_create_drvdata(coda_psdev_class, NULL, - MKDEV(CODA_PSDEV_MAJOR, i), - NULL, "cfs%d", i); + device_create(coda_psdev_class, NULL, + MKDEV(CODA_PSDEV_MAJOR, i), NULL, "cfs%d", i); coda_sysctl_init(); goto out; diff --git a/fs/coda/upcall.c b/fs/coda/upcall.c index ce432bca95d..c274d949179 100644 --- a/fs/coda/upcall.c +++ b/fs/coda/upcall.c @@ -52,7 +52,7 @@ static void *alloc_upcall(int opcode, int size) inp->ih.opcode = opcode; inp->ih.pid = current->pid; inp->ih.pgid = task_pgrp_nr(current); - inp->ih.uid = current->fsuid; + inp->ih.uid = current_fsuid(); return (void*)inp; } diff --git a/fs/compat.c b/fs/compat.c index 075d0509970..d1ece79b641 100644 --- a/fs/compat.c +++ b/fs/compat.c @@ -137,6 +137,45 @@ asmlinkage long compat_sys_utimes(char __user *filename, struct compat_timeval _ return compat_sys_futimesat(AT_FDCWD, filename, t); } +static int cp_compat_stat(struct kstat *stat, struct compat_stat __user *ubuf) +{ + compat_ino_t ino = stat->ino; + typeof(ubuf->st_uid) uid = 0; + typeof(ubuf->st_gid) gid = 0; + int err; + + SET_UID(uid, stat->uid); + SET_GID(gid, stat->gid); + + if ((u64) stat->size > MAX_NON_LFS || + !old_valid_dev(stat->dev) || + !old_valid_dev(stat->rdev)) + return -EOVERFLOW; + if (sizeof(ino) < sizeof(stat->ino) && ino != stat->ino) + return -EOVERFLOW; + + if (clear_user(ubuf, sizeof(*ubuf))) + return -EFAULT; + + err = __put_user(old_encode_dev(stat->dev), &ubuf->st_dev); + err |= __put_user(ino, &ubuf->st_ino); + err |= __put_user(stat->mode, &ubuf->st_mode); + err |= __put_user(stat->nlink, &ubuf->st_nlink); + err |= __put_user(uid, &ubuf->st_uid); + err |= __put_user(gid, &ubuf->st_gid); + err |= __put_user(old_encode_dev(stat->rdev), &ubuf->st_rdev); + err |= __put_user(stat->size, &ubuf->st_size); + err |= __put_user(stat->atime.tv_sec, &ubuf->st_atime); + err |= __put_user(stat->atime.tv_nsec, &ubuf->st_atime_nsec); + err |= __put_user(stat->mtime.tv_sec, &ubuf->st_mtime); + err |= __put_user(stat->mtime.tv_nsec, &ubuf->st_mtime_nsec); + err |= __put_user(stat->ctime.tv_sec, &ubuf->st_ctime); + err |= __put_user(stat->ctime.tv_nsec, &ubuf->st_ctime_nsec); + err |= __put_user(stat->blksize, &ubuf->st_blksize); + err |= __put_user(stat->blocks, &ubuf->st_blocks); + return err; +} + asmlinkage long compat_sys_newstat(char __user * filename, struct compat_stat __user *statbuf) { @@ -830,7 +869,7 @@ asmlinkage long compat_sys_old_readdir(unsigned int fd, buf.dirent = dirent; error = vfs_readdir(file, compat_fillonedir, &buf); - if (error >= 0) + if (buf.result) error = buf.result; fput(file); @@ -917,9 +956,8 @@ asmlinkage long compat_sys_getdents(unsigned int fd, buf.error = 0; error = vfs_readdir(file, compat_filldir, &buf); - if (error < 0) - goto out_putf; - error = buf.error; + if (error >= 0) + error = buf.error; lastdirent = buf.previous; if (lastdirent) { if (put_user(file->f_pos, &lastdirent->d_off)) @@ -927,8 +965,6 @@ asmlinkage long compat_sys_getdents(unsigned int fd, else error = count - buf.count; } - -out_putf: fput(file); out: return error; @@ -1008,19 +1044,16 @@ asmlinkage long compat_sys_getdents64(unsigned int fd, buf.error = 0; error = vfs_readdir(file, compat_filldir64, &buf); - if (error < 0) - goto out_putf; - error = buf.error; + if (error >= 0) + error = buf.error; lastdirent = buf.previous; if (lastdirent) { typeof(lastdirent->d_off) d_off = file->f_pos; - error = -EFAULT; if (__put_user_unaligned(d_off, &lastdirent->d_off)) - goto out_putf; - error = count - buf.count; + error = -EFAULT; + else + error = count - buf.count; } - -out_putf: fput(file); out: return error; @@ -1239,7 +1272,7 @@ static int compat_count(compat_uptr_t __user *argv, int max) if (!p) break; argv++; - if(++i > max) + if (i++ >= max) return -E2BIG; } } @@ -1360,10 +1393,20 @@ int compat_do_execve(char * filename, if (!bprm) goto out_ret; + retval = mutex_lock_interruptible(¤t->cred_exec_mutex); + if (retval < 0) + goto out_free; + + retval = -ENOMEM; + bprm->cred = prepare_exec_creds(); + if (!bprm->cred) + goto out_unlock; + check_unsafe_exec(bprm); + file = open_exec(filename); retval = PTR_ERR(file); if (IS_ERR(file)) - goto out_kfree; + goto out_unlock; sched_exec(); @@ -1377,14 +1420,10 @@ int compat_do_execve(char * filename, bprm->argc = compat_count(argv, MAX_ARG_STRINGS); if ((retval = bprm->argc) < 0) - goto out_mm; + goto out; bprm->envc = compat_count(envp, MAX_ARG_STRINGS); if ((retval = bprm->envc) < 0) - goto out_mm; - - retval = security_bprm_alloc(bprm); - if (retval) goto out; retval = prepare_binprm(bprm); @@ -1405,19 +1444,16 @@ int compat_do_execve(char * filename, goto out; retval = search_binary_handler(bprm, regs); - if (retval >= 0) { - /* execve success */ - security_bprm_free(bprm); - acct_update_integrals(current); - free_bprm(bprm); - return retval; - } + if (retval < 0) + goto out; -out: - if (bprm->security) - security_bprm_free(bprm); + /* execve succeeded */ + mutex_unlock(¤t->cred_exec_mutex); + acct_update_integrals(current); + free_bprm(bprm); + return retval; -out_mm: +out: if (bprm->mm) mmput(bprm->mm); @@ -1427,7 +1463,10 @@ out_file: fput(bprm->file); } -out_kfree: +out_unlock: + mutex_unlock(¤t->cred_exec_mutex); + +out_free: free_bprm(bprm); out_ret: @@ -1436,6 +1475,57 @@ out_ret: #define __COMPAT_NFDBITS (8 * sizeof(compat_ulong_t)) +static int poll_select_copy_remaining(struct timespec *end_time, void __user *p, + int timeval, int ret) +{ + struct timespec ts; + + if (!p) + return ret; + + if (current->personality & STICKY_TIMEOUTS) + goto sticky; + + /* No update for zero timeout */ + if (!end_time->tv_sec && !end_time->tv_nsec) + return ret; + + ktime_get_ts(&ts); + ts = timespec_sub(*end_time, ts); + if (ts.tv_sec < 0) + ts.tv_sec = ts.tv_nsec = 0; + + if (timeval) { + struct compat_timeval rtv; + + rtv.tv_sec = ts.tv_sec; + rtv.tv_usec = ts.tv_nsec / NSEC_PER_USEC; + + if (!copy_to_user(p, &rtv, sizeof(rtv))) + return ret; + } else { + struct compat_timespec rts; + + rts.tv_sec = ts.tv_sec; + rts.tv_nsec = ts.tv_nsec; + + if (!copy_to_user(p, &rts, sizeof(rts))) + return ret; + } + /* + * If an application puts its timeval in read-only memory, we + * don't want the Linux-specific update to the timeval to + * cause a fault after the select has completed + * successfully. However, because we're not updating the + * timeval, we can't restart the system call. + */ + +sticky: + if (ret == -ERESTARTNOHAND) + ret = -EINTR; + return ret; +} + /* * Ooo, nasty. We need here to frob 32-bit unsigned longs to * 64-bit unsigned longs. @@ -1517,7 +1607,8 @@ int compat_set_fd_set(unsigned long nr, compat_ulong_t __user *ufdset, ((unsigned long) (MAX_SCHEDULE_TIMEOUT / HZ)-1) int compat_core_sys_select(int n, compat_ulong_t __user *inp, - compat_ulong_t __user *outp, compat_ulong_t __user *exp, s64 *timeout) + compat_ulong_t __user *outp, compat_ulong_t __user *exp, + struct timespec *end_time) { fd_set_bits fds; void *bits; @@ -1564,7 +1655,7 @@ int compat_core_sys_select(int n, compat_ulong_t __user *inp, zero_fd_set(n, fds.res_out); zero_fd_set(n, fds.res_ex); - ret = do_select(n, &fds, timeout); + ret = do_select(n, &fds, end_time); if (ret < 0) goto out; @@ -1590,7 +1681,7 @@ asmlinkage long compat_sys_select(int n, compat_ulong_t __user *inp, compat_ulong_t __user *outp, compat_ulong_t __user *exp, struct compat_timeval __user *tvp) { - s64 timeout = -1; + struct timespec end_time, *to = NULL; struct compat_timeval tv; int ret; @@ -1598,43 +1689,15 @@ asmlinkage long compat_sys_select(int n, compat_ulong_t __user *inp, if (copy_from_user(&tv, tvp, sizeof(tv))) return -EFAULT; - if (tv.tv_sec < 0 || tv.tv_usec < 0) + to = &end_time; + if (poll_select_set_timeout(to, + tv.tv_sec + (tv.tv_usec / USEC_PER_SEC), + (tv.tv_usec % USEC_PER_SEC) * NSEC_PER_USEC)) return -EINVAL; - - /* Cast to u64 to make GCC stop complaining */ - if ((u64)tv.tv_sec >= (u64)MAX_INT64_SECONDS) - timeout = -1; /* infinite */ - else { - timeout = DIV_ROUND_UP(tv.tv_usec, 1000000/HZ); - timeout += tv.tv_sec * HZ; - } } - ret = compat_core_sys_select(n, inp, outp, exp, &timeout); - - if (tvp) { - struct compat_timeval rtv; - - if (current->personality & STICKY_TIMEOUTS) - goto sticky; - rtv.tv_usec = jiffies_to_usecs(do_div((*(u64*)&timeout), HZ)); - rtv.tv_sec = timeout; - if (compat_timeval_compare(&rtv, &tv) >= 0) - rtv = tv; - if (copy_to_user(tvp, &rtv, sizeof(rtv))) { -sticky: - /* - * If an application puts its timeval in read-only - * memory, we don't want the Linux-specific update to - * the timeval to cause a fault after the select has - * completed successfully. However, because we're not - * updating the timeval, we can't restart the system - * call. - */ - if (ret == -ERESTARTNOHAND) - ret = -EINTR; - } - } + ret = compat_core_sys_select(n, inp, outp, exp, to); + ret = poll_select_copy_remaining(&end_time, tvp, 1, ret); return ret; } @@ -1647,15 +1710,16 @@ asmlinkage long compat_sys_pselect7(int n, compat_ulong_t __user *inp, { compat_sigset_t ss32; sigset_t ksigmask, sigsaved; - s64 timeout = MAX_SCHEDULE_TIMEOUT; struct compat_timespec ts; + struct timespec end_time, *to = NULL; int ret; if (tsp) { if (copy_from_user(&ts, tsp, sizeof(ts))) return -EFAULT; - if (ts.tv_sec < 0 || ts.tv_nsec < 0) + to = &end_time; + if (poll_select_set_timeout(to, ts.tv_sec, ts.tv_nsec)) return -EINVAL; } @@ -1670,51 +1734,8 @@ asmlinkage long compat_sys_pselect7(int n, compat_ulong_t __user *inp, sigprocmask(SIG_SETMASK, &ksigmask, &sigsaved); } - do { - if (tsp) { - if ((unsigned long)ts.tv_sec < MAX_SELECT_SECONDS) { - timeout = DIV_ROUND_UP(ts.tv_nsec, 1000000000/HZ); - timeout += ts.tv_sec * (unsigned long)HZ; - ts.tv_sec = 0; - ts.tv_nsec = 0; - } else { - ts.tv_sec -= MAX_SELECT_SECONDS; - timeout = MAX_SELECT_SECONDS * HZ; - } - } - - ret = compat_core_sys_select(n, inp, outp, exp, &timeout); - - } while (!ret && !timeout && tsp && (ts.tv_sec || ts.tv_nsec)); - - if (tsp) { - struct compat_timespec rts; - - if (current->personality & STICKY_TIMEOUTS) - goto sticky; - - rts.tv_sec = timeout / HZ; - rts.tv_nsec = (timeout % HZ) * (NSEC_PER_SEC/HZ); - if (rts.tv_nsec >= NSEC_PER_SEC) { - rts.tv_sec++; - rts.tv_nsec -= NSEC_PER_SEC; - } - if (compat_timespec_compare(&rts, &ts) >= 0) - rts = ts; - if (copy_to_user(tsp, &rts, sizeof(rts))) { -sticky: - /* - * If an application puts its timeval in read-only - * memory, we don't want the Linux-specific update to - * the timeval to cause a fault after the select has - * completed successfully. However, because we're not - * updating the timeval, we can't restart the system - * call. - */ - if (ret == -ERESTARTNOHAND) - ret = -EINTR; - } - } + ret = compat_core_sys_select(n, inp, outp, exp, to); + ret = poll_select_copy_remaining(&end_time, tsp, 0, ret); if (ret == -ERESTARTNOHAND) { /* @@ -1759,18 +1780,16 @@ asmlinkage long compat_sys_ppoll(struct pollfd __user *ufds, compat_sigset_t ss32; sigset_t ksigmask, sigsaved; struct compat_timespec ts; - s64 timeout = -1; + struct timespec end_time, *to = NULL; int ret; if (tsp) { if (copy_from_user(&ts, tsp, sizeof(ts))) return -EFAULT; - /* We assume that ts.tv_sec is always lower than - the number of seconds that can be expressed in - an s64. Otherwise the compiler bitches at us */ - timeout = DIV_ROUND_UP(ts.tv_nsec, 1000000000/HZ); - timeout += ts.tv_sec * HZ; + to = &end_time; + if (poll_select_set_timeout(to, ts.tv_sec, ts.tv_nsec)) + return -EINVAL; } if (sigmask) { @@ -1784,7 +1803,7 @@ asmlinkage long compat_sys_ppoll(struct pollfd __user *ufds, sigprocmask(SIG_SETMASK, &ksigmask, &sigsaved); } - ret = do_sys_poll(ufds, nfds, &timeout); + ret = do_sys_poll(ufds, nfds, to); /* We can restart this syscall, usually */ if (ret == -EINTR) { @@ -1802,31 +1821,7 @@ asmlinkage long compat_sys_ppoll(struct pollfd __user *ufds, } else if (sigmask) sigprocmask(SIG_SETMASK, &sigsaved, NULL); - if (tsp && timeout >= 0) { - struct compat_timespec rts; - - if (current->personality & STICKY_TIMEOUTS) - goto sticky; - /* Yes, we know it's actually an s64, but it's also positive. */ - rts.tv_nsec = jiffies_to_usecs(do_div((*(u64*)&timeout), HZ)) * - 1000; - rts.tv_sec = timeout; - if (compat_timespec_compare(&rts, &ts) >= 0) - rts = ts; - if (copy_to_user(tsp, &rts, sizeof(rts))) { -sticky: - /* - * If an application puts its timeval in read-only - * memory, we don't want the Linux-specific update to - * the timeval to cause a fault after the select has - * completed successfully. However, because we're not - * updating the timeval, we can't restart the system - * call. - */ - if (ret == -ERESTARTNOHAND && timeout >= 0) - ret = -EINTR; - } - } + ret = poll_select_copy_remaining(&end_time, tsp, 0, ret); return ret; } diff --git a/fs/configfs/symlink.c b/fs/configfs/symlink.c index bf74973b049..932a92b3148 100644 --- a/fs/configfs/symlink.c +++ b/fs/configfs/symlink.c @@ -108,18 +108,18 @@ out: } -static int get_target(const char *symname, struct nameidata *nd, +static int get_target(const char *symname, struct path *path, struct config_item **target) { int ret; - ret = path_lookup(symname, LOOKUP_FOLLOW|LOOKUP_DIRECTORY, nd); + ret = kern_path(symname, LOOKUP_FOLLOW|LOOKUP_DIRECTORY, path); if (!ret) { - if (nd->path.dentry->d_sb == configfs_sb) { - *target = configfs_get_config_item(nd->path.dentry); + if (path->dentry->d_sb == configfs_sb) { + *target = configfs_get_config_item(path->dentry); if (!*target) { ret = -ENOENT; - path_put(&nd->path); + path_put(path); } } else ret = -EPERM; @@ -132,7 +132,7 @@ static int get_target(const char *symname, struct nameidata *nd, int configfs_symlink(struct inode *dir, struct dentry *dentry, const char *symname) { int ret; - struct nameidata nd; + struct path path; struct configfs_dirent *sd; struct config_item *parent_item; struct config_item *target_item; @@ -159,7 +159,7 @@ int configfs_symlink(struct inode *dir, struct dentry *dentry, const char *symna !type->ct_item_ops->allow_link) goto out_put; - ret = get_target(symname, &nd, &target_item); + ret = get_target(symname, &path, &target_item); if (ret) goto out_put; @@ -174,7 +174,7 @@ int configfs_symlink(struct inode *dir, struct dentry *dentry, const char *symna } config_item_put(target_item); - path_put(&nd.path); + path_put(&path); out_put: config_item_put(parent_item); diff --git a/fs/dcache.c b/fs/dcache.c index e7a1a99b746..a1d86c7f3e6 100644 --- a/fs/dcache.c +++ b/fs/dcache.c @@ -69,6 +69,7 @@ struct dentry_stat_t dentry_stat = { static void __d_free(struct dentry *dentry) { + WARN_ON(!list_empty(&dentry->d_alias)); if (dname_external(dentry)) kfree(dentry->d_name.name); kmem_cache_free(dentry_cache, dentry); @@ -174,9 +175,12 @@ static struct dentry *d_kill(struct dentry *dentry) dentry_stat.nr_dentry--; /* For d_free, below */ /*drops the locks, at that point nobody can reach this dentry */ dentry_iput(dentry); - parent = dentry->d_parent; + if (IS_ROOT(dentry)) + parent = NULL; + else + parent = dentry->d_parent; d_free(dentry); - return dentry == parent ? NULL : parent; + return parent; } /* @@ -666,11 +670,12 @@ static void shrink_dcache_for_umount_subtree(struct dentry *dentry) BUG(); } - parent = dentry->d_parent; - if (parent == dentry) + if (IS_ROOT(dentry)) parent = NULL; - else + else { + parent = dentry->d_parent; atomic_dec(&parent->d_count); + } list_del(&dentry->d_u.d_child); detached++; @@ -977,6 +982,15 @@ struct dentry *d_alloc_name(struct dentry *parent, const char *name) return d_alloc(parent, &q); } +/* the caller must hold dcache_lock */ +static void __d_instantiate(struct dentry *dentry, struct inode *inode) +{ + if (inode) + list_add(&dentry->d_alias, &inode->i_dentry); + dentry->d_inode = inode; + fsnotify_d_instantiate(dentry, inode); +} + /** * d_instantiate - fill in inode information for a dentry * @entry: dentry to complete @@ -996,10 +1010,7 @@ void d_instantiate(struct dentry *entry, struct inode * inode) { BUG_ON(!list_empty(&entry->d_alias)); spin_lock(&dcache_lock); - if (inode) - list_add(&entry->d_alias, &inode->i_dentry); - entry->d_inode = inode; - fsnotify_d_instantiate(entry, inode); + __d_instantiate(entry, inode); spin_unlock(&dcache_lock); security_d_instantiate(entry, inode); } @@ -1029,7 +1040,7 @@ static struct dentry *__d_instantiate_unique(struct dentry *entry, unsigned int hash = entry->d_name.hash; if (!inode) { - entry->d_inode = NULL; + __d_instantiate(entry, NULL); return NULL; } @@ -1048,9 +1059,7 @@ static struct dentry *__d_instantiate_unique(struct dentry *entry, return alias; } - list_add(&entry->d_alias, &inode->i_dentry); - entry->d_inode = inode; - fsnotify_d_instantiate(entry, inode); + __d_instantiate(entry, inode); return NULL; } @@ -1111,69 +1120,71 @@ static inline struct hlist_head *d_hash(struct dentry *parent, } /** - * d_alloc_anon - allocate an anonymous dentry + * d_obtain_alias - find or allocate a dentry for a given inode * @inode: inode to allocate the dentry for * - * This is similar to d_alloc_root. It is used by filesystems when - * creating a dentry for a given inode, often in the process of - * mapping a filehandle to a dentry. The returned dentry may be - * anonymous, or may have a full name (if the inode was already - * in the cache). The file system may need to make further - * efforts to connect this dentry into the dcache properly. + * Obtain a dentry for an inode resulting from NFS filehandle conversion or + * similar open by handle operations. The returned dentry may be anonymous, + * or may have a full name (if the inode was already in the cache). * - * When called on a directory inode, we must ensure that - * the inode only ever has one dentry. If a dentry is - * found, that is returned instead of allocating a new one. + * When called on a directory inode, we must ensure that the inode only ever + * has one dentry. If a dentry is found, that is returned instead of + * allocating a new one. * * On successful return, the reference to the inode has been transferred - * to the dentry. If %NULL is returned (indicating kmalloc failure), - * the reference on the inode has not been released. + * to the dentry. In case of an error the reference on the inode is released. + * To make it easier to use in export operations a %NULL or IS_ERR inode may + * be passed in and will be the error will be propagate to the return value, + * with a %NULL @inode replaced by ERR_PTR(-ESTALE). */ - -struct dentry * d_alloc_anon(struct inode *inode) +struct dentry *d_obtain_alias(struct inode *inode) { static const struct qstr anonstring = { .name = "" }; struct dentry *tmp; struct dentry *res; - if ((res = d_find_alias(inode))) { - iput(inode); - return res; - } + if (!inode) + return ERR_PTR(-ESTALE); + if (IS_ERR(inode)) + return ERR_CAST(inode); - tmp = d_alloc(NULL, &anonstring); - if (!tmp) - return NULL; + res = d_find_alias(inode); + if (res) + goto out_iput; + tmp = d_alloc(NULL, &anonstring); + if (!tmp) { + res = ERR_PTR(-ENOMEM); + goto out_iput; + } tmp->d_parent = tmp; /* make sure dput doesn't croak */ - + spin_lock(&dcache_lock); res = __d_find_alias(inode, 0); - if (!res) { - /* attach a disconnected dentry */ - res = tmp; - tmp = NULL; - spin_lock(&res->d_lock); - res->d_sb = inode->i_sb; - res->d_parent = res; - res->d_inode = inode; - res->d_flags |= DCACHE_DISCONNECTED; - res->d_flags &= ~DCACHE_UNHASHED; - list_add(&res->d_alias, &inode->i_dentry); - hlist_add_head(&res->d_hash, &inode->i_sb->s_anon); - spin_unlock(&res->d_lock); - - inode = NULL; /* don't drop reference */ + if (res) { + spin_unlock(&dcache_lock); + dput(tmp); + goto out_iput; } + + /* attach a disconnected dentry */ + spin_lock(&tmp->d_lock); + tmp->d_sb = inode->i_sb; + tmp->d_inode = inode; + tmp->d_flags |= DCACHE_DISCONNECTED; + tmp->d_flags &= ~DCACHE_UNHASHED; + list_add(&tmp->d_alias, &inode->i_dentry); + hlist_add_head(&tmp->d_hash, &inode->i_sb->s_anon); + spin_unlock(&tmp->d_lock); + spin_unlock(&dcache_lock); + return tmp; - if (inode) - iput(inode); - if (tmp) - dput(tmp); + out_iput: + iput(inode); return res; } - +EXPORT_SYMBOL_GPL(d_obtain_alias); /** * d_splice_alias - splice a disconnected dentry into the tree if one exists @@ -1200,17 +1211,14 @@ struct dentry *d_splice_alias(struct inode *inode, struct dentry *dentry) new = __d_find_alias(inode, 1); if (new) { BUG_ON(!(new->d_flags & DCACHE_DISCONNECTED)); - fsnotify_d_instantiate(new, inode); spin_unlock(&dcache_lock); security_d_instantiate(new, inode); d_rehash(dentry); d_move(new, dentry); iput(inode); } else { - /* d_instantiate takes dcache_lock, so we do it by hand */ - list_add(&dentry->d_alias, &inode->i_dentry); - dentry->d_inode = inode; - fsnotify_d_instantiate(dentry, inode); + /* already taking dcache_lock, so d_add() by hand */ + __d_instantiate(dentry, inode); spin_unlock(&dcache_lock); security_d_instantiate(dentry, inode); d_rehash(dentry); @@ -1293,8 +1301,7 @@ struct dentry *d_add_ci(struct dentry *dentry, struct inode *inode, * d_instantiate() by hand because it takes dcache_lock which * we already hold. */ - list_add(&found->d_alias, &inode->i_dentry); - found->d_inode = inode; + __d_instantiate(found, inode); spin_unlock(&dcache_lock); security_d_instantiate(found, inode); return found; @@ -1456,8 +1463,6 @@ out: * d_validate - verify dentry provided from insecure source * @dentry: The dentry alleged to be valid child of @dparent * @dparent: The parent dentry (known to be valid) - * @hash: Hash of the dentry - * @len: Length of the name * * An insecure source has sent us a dentry, here we verify it and dget() it. * This is used by ncpfs in its readdir implementation. @@ -1714,18 +1719,23 @@ void d_move(struct dentry * dentry, struct dentry * target) spin_unlock(&dcache_lock); } -/* - * Helper that returns 1 if p1 is a parent of p2, else 0 +/** + * d_ancestor - search for an ancestor + * @p1: ancestor dentry + * @p2: child dentry + * + * Returns the ancestor dentry of p2 which is a child of p1, if p1 is + * an ancestor of p2, else NULL. */ -static int d_isparent(struct dentry *p1, struct dentry *p2) +struct dentry *d_ancestor(struct dentry *p1, struct dentry *p2) { struct dentry *p; - for (p = p2; p->d_parent != p; p = p->d_parent) { + for (p = p2; !IS_ROOT(p); p = p->d_parent) { if (p->d_parent == p1) - return 1; + return p; } - return 0; + return NULL; } /* @@ -1749,7 +1759,7 @@ static struct dentry *__d_unalias(struct dentry *dentry, struct dentry *alias) /* Check for loops */ ret = ERR_PTR(-ELOOP); - if (d_isparent(alias, dentry)) + if (d_ancestor(alias, dentry)) goto out_err; /* See lock_rename() */ @@ -1822,7 +1832,7 @@ struct dentry *d_materialise_unique(struct dentry *dentry, struct inode *inode) if (!inode) { actual = dentry; - dentry->d_inode = NULL; + __d_instantiate(dentry, NULL); goto found_lock; } @@ -2149,32 +2159,27 @@ out: * Caller must ensure that "new_dentry" is pinned before calling is_subdir() */ -int is_subdir(struct dentry * new_dentry, struct dentry * old_dentry) +int is_subdir(struct dentry *new_dentry, struct dentry *old_dentry) { int result; - struct dentry * saved = new_dentry; unsigned long seq; - /* need rcu_readlock to protect against the d_parent trashing due to - * d_move + /* FIXME: This is old behavior, needed? Please check callers. */ + if (new_dentry == old_dentry) + return 1; + + /* + * Need rcu_readlock to protect against the d_parent trashing + * due to d_move */ rcu_read_lock(); - do { + do { /* for restarting inner loop in case of seq retry */ - new_dentry = saved; - result = 0; seq = read_seqbegin(&rename_lock); - for (;;) { - if (new_dentry != old_dentry) { - struct dentry * parent = new_dentry->d_parent; - if (parent == new_dentry) - break; - new_dentry = parent; - continue; - } + if (d_ancestor(old_dentry, new_dentry)) result = 1; - break; - } + else + result = 0; } while (read_seqretry(&rename_lock, seq)); rcu_read_unlock(); @@ -2344,7 +2349,6 @@ void __init vfs_caches_init(unsigned long mempages) } EXPORT_SYMBOL(d_alloc); -EXPORT_SYMBOL(d_alloc_anon); EXPORT_SYMBOL(d_alloc_root); EXPORT_SYMBOL(d_delete); EXPORT_SYMBOL(d_find_alias); diff --git a/fs/devpts/inode.c b/fs/devpts/inode.c index 4a714f6c1be..5d61b7c06e1 100644 --- a/fs/devpts/inode.c +++ b/fs/devpts/inode.c @@ -222,8 +222,8 @@ int devpts_pty_new(struct inode *ptmx_inode, struct tty_struct *tty) return -ENOMEM; inode->i_ino = number+2; - inode->i_uid = config.setuid ? config.uid : current->fsuid; - inode->i_gid = config.setgid ? config.gid : current->fsgid; + inode->i_uid = config.setuid ? config.uid : current_fsuid(); + inode->i_gid = config.setgid ? config.gid : current_fsgid(); inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME; init_special_inode(inode, S_IFCHR|config.mode, device); inode->i_private = tty; diff --git a/fs/direct-io.c b/fs/direct-io.c index 9606ee848fd..af0558dbe8b 100644 --- a/fs/direct-io.c +++ b/fs/direct-io.c @@ -5,11 +5,11 @@ * * O_DIRECT * - * 04Jul2002 akpm@zip.com.au + * 04Jul2002 Andrew Morton * Initial version * 11Sep2002 janetinc@us.ibm.com * added readv/writev support. - * 29Oct2002 akpm@zip.com.au + * 29Oct2002 Andrew Morton * rewrote bio_add_page() support. * 30Oct2002 pbadari@us.ibm.com * added support for non-aligned IO. diff --git a/fs/dlm/lockspace.c b/fs/dlm/lockspace.c index d910501de6d..8d86b7960f0 100644 --- a/fs/dlm/lockspace.c +++ b/fs/dlm/lockspace.c @@ -812,7 +812,7 @@ int dlm_release_lockspace(void *lockspace, int force) error = release_lockspace(ls, force); if (!error) ls_count--; - else if (!ls_count) + if (!ls_count) threads_stop(); mutex_unlock(&ls_lock); diff --git a/fs/dlm/netlink.c b/fs/dlm/netlink.c index 18bda83cc89..aa2a5775a02 100644 --- a/fs/dlm/netlink.c +++ b/fs/dlm/netlink.c @@ -127,8 +127,8 @@ static void fill_data(struct dlm_lock_data *data, struct dlm_lkb *lkb) void dlm_timeout_warn(struct dlm_lkb *lkb) { + struct sk_buff *uninitialized_var(send_skb); struct dlm_lock_data *data; - struct sk_buff *send_skb; size_t size; int rv; diff --git a/fs/dquot.c b/fs/dquot.c index ad7e59003e0..c237ccc8581 100644 --- a/fs/dquot.c +++ b/fs/dquot.c @@ -9,8 +9,6 @@ * implementation is based on one of the several variants of the LINUX * inode-subsystem with added complexity of the diskquota system. * - * Version: $Id: dquot.c,v 6.3 1996/11/17 18:35:34 mvw Exp mvw $ - * * Author: Marco van Wieringen <mvw@planets.elm.net> * * Fixes: Dmitry Gorodchanin <pgmdsg@ibi.com>, 11 Feb 96 @@ -876,7 +874,7 @@ static inline int need_print_warning(struct dquot *dquot) switch (dquot->dq_type) { case USRQUOTA: - return current->fsuid == dquot->dq_id; + return current_fsuid() == dquot->dq_id; case GRPQUOTA: return in_group_p(dquot->dq_id); } @@ -983,7 +981,7 @@ static void send_warning(const struct dquot *dquot, const char warntype) MINOR(dquot->dq_sb->s_dev)); if (ret) goto attr_err_out; - ret = nla_put_u64(skb, QUOTA_NL_A_CAUSED_ID, current->user->uid); + ret = nla_put_u64(skb, QUOTA_NL_A_CAUSED_ID, current_uid()); if (ret) goto attr_err_out; genlmsg_end(skb, msg_head); @@ -1807,19 +1805,19 @@ int vfs_quota_on_path(struct super_block *sb, int type, int format_id, } /* Actual function called from quotactl() */ -int vfs_quota_on(struct super_block *sb, int type, int format_id, char *path, +int vfs_quota_on(struct super_block *sb, int type, int format_id, char *name, int remount) { - struct nameidata nd; + struct path path; int error; if (remount) return vfs_quota_on_remount(sb, type); - error = path_lookup(path, LOOKUP_FOLLOW, &nd); + error = kern_path(name, LOOKUP_FOLLOW, &path); if (!error) { - error = vfs_quota_on_path(sb, type, format_id, &nd.path); - path_put(&nd.path); + error = vfs_quota_on_path(sb, type, format_id, &path); + path_put(&path); } return error; } diff --git a/fs/ecryptfs/Makefile b/fs/ecryptfs/Makefile index b4755a85996..2cc9ee4ad2e 100644 --- a/fs/ecryptfs/Makefile +++ b/fs/ecryptfs/Makefile @@ -4,4 +4,4 @@ obj-$(CONFIG_ECRYPT_FS) += ecryptfs.o -ecryptfs-objs := dentry.o file.o inode.o main.o super.o mmap.o read_write.o crypto.o keystore.o messaging.o netlink.o miscdev.o kthread.o debug.o +ecryptfs-objs := dentry.o file.o inode.o main.o super.o mmap.o read_write.o crypto.o keystore.o messaging.o miscdev.o kthread.o debug.o diff --git a/fs/ecryptfs/crypto.c b/fs/ecryptfs/crypto.c index 06db79d05c1..6046239465a 100644 --- a/fs/ecryptfs/crypto.c +++ b/fs/ecryptfs/crypto.c @@ -1251,6 +1251,7 @@ struct kmem_cache *ecryptfs_header_cache_2; /** * ecryptfs_write_headers_virt * @page_virt: The virtual address to write the headers to + * @max: The size of memory allocated at page_virt * @size: Set to the number of bytes written by this function * @crypt_stat: The cryptographic context * @ecryptfs_dentry: The eCryptfs dentry @@ -1278,7 +1279,8 @@ struct kmem_cache *ecryptfs_header_cache_2; * * Returns zero on success */ -static int ecryptfs_write_headers_virt(char *page_virt, size_t *size, +static int ecryptfs_write_headers_virt(char *page_virt, size_t max, + size_t *size, struct ecryptfs_crypt_stat *crypt_stat, struct dentry *ecryptfs_dentry) { @@ -1296,7 +1298,7 @@ static int ecryptfs_write_headers_virt(char *page_virt, size_t *size, offset += written; rc = ecryptfs_generate_key_packet_set((page_virt + offset), crypt_stat, ecryptfs_dentry, &written, - PAGE_CACHE_SIZE - offset); + max - offset); if (rc) ecryptfs_printk(KERN_WARNING, "Error generating key packet " "set; rc = [%d]\n", rc); @@ -1368,14 +1370,14 @@ int ecryptfs_write_metadata(struct dentry *ecryptfs_dentry) goto out; } /* Released in this function */ - virt = kzalloc(crypt_stat->num_header_bytes_at_front, GFP_KERNEL); + virt = (char *)get_zeroed_page(GFP_KERNEL); if (!virt) { printk(KERN_ERR "%s: Out of memory\n", __func__); rc = -ENOMEM; goto out; } - rc = ecryptfs_write_headers_virt(virt, &size, crypt_stat, - ecryptfs_dentry); + rc = ecryptfs_write_headers_virt(virt, PAGE_CACHE_SIZE, &size, + crypt_stat, ecryptfs_dentry); if (unlikely(rc)) { printk(KERN_ERR "%s: Error whilst writing headers; rc = [%d]\n", __func__, rc); @@ -1393,8 +1395,7 @@ int ecryptfs_write_metadata(struct dentry *ecryptfs_dentry) goto out_free; } out_free: - memset(virt, 0, crypt_stat->num_header_bytes_at_front); - kfree(virt); + free_page((unsigned long)virt); out: return rc; } diff --git a/fs/ecryptfs/ecryptfs_kernel.h b/fs/ecryptfs/ecryptfs_kernel.h index b73fb752c5f..a75026d35d1 100644 --- a/fs/ecryptfs/ecryptfs_kernel.h +++ b/fs/ecryptfs/ecryptfs_kernel.h @@ -79,11 +79,6 @@ #define ECRYPTFS_MAX_PKI_NAME_BYTES 16 #define ECRYPTFS_DEFAULT_NUM_USERS 4 #define ECRYPTFS_MAX_NUM_USERS 32768 -#define ECRYPTFS_TRANSPORT_NETLINK 0 -#define ECRYPTFS_TRANSPORT_CONNECTOR 1 -#define ECRYPTFS_TRANSPORT_RELAYFS 2 -#define ECRYPTFS_TRANSPORT_MISCDEV 3 -#define ECRYPTFS_DEFAULT_TRANSPORT ECRYPTFS_TRANSPORT_MISCDEV #define ECRYPTFS_XATTR_NAME "user.ecryptfs" #define RFC2440_CIPHER_DES3_EDE 0x02 @@ -400,8 +395,6 @@ struct ecryptfs_msg_ctx { struct mutex mux; }; -extern unsigned int ecryptfs_transport; - struct ecryptfs_daemon; struct ecryptfs_daemon { @@ -627,31 +620,20 @@ int ecryptfs_setxattr(struct dentry *dentry, const char *name, const void *value, size_t size, int flags); int ecryptfs_read_xattr_region(char *page_virt, struct inode *ecryptfs_inode); -int ecryptfs_process_helo(unsigned int transport, uid_t euid, - struct user_namespace *user_ns, struct pid *pid); +int ecryptfs_process_helo(uid_t euid, struct user_namespace *user_ns, + struct pid *pid); int ecryptfs_process_quit(uid_t euid, struct user_namespace *user_ns, struct pid *pid); int ecryptfs_process_response(struct ecryptfs_message *msg, uid_t euid, struct user_namespace *user_ns, struct pid *pid, u32 seq); -int ecryptfs_send_message(unsigned int transport, char *data, int data_len, +int ecryptfs_send_message(char *data, int data_len, struct ecryptfs_msg_ctx **msg_ctx); int ecryptfs_wait_for_response(struct ecryptfs_msg_ctx *msg_ctx, struct ecryptfs_message **emsg); -int ecryptfs_init_messaging(unsigned int transport); -void ecryptfs_release_messaging(unsigned int transport); +int ecryptfs_init_messaging(void); +void ecryptfs_release_messaging(void); -int ecryptfs_send_netlink(char *data, int data_len, - struct ecryptfs_msg_ctx *msg_ctx, u8 msg_type, - u16 msg_flags, struct pid *daemon_pid); -int ecryptfs_init_netlink(void); -void ecryptfs_release_netlink(void); - -int ecryptfs_send_connector(char *data, int data_len, - struct ecryptfs_msg_ctx *msg_ctx, u8 msg_type, - u16 msg_flags, struct pid *daemon_pid); -int ecryptfs_init_connector(void); -void ecryptfs_release_connector(void); void ecryptfs_write_header_metadata(char *virt, struct ecryptfs_crypt_stat *crypt_stat, @@ -709,7 +691,8 @@ int ecryptfs_init_kthread(void); void ecryptfs_destroy_kthread(void); int ecryptfs_privileged_open(struct file **lower_file, struct dentry *lower_dentry, - struct vfsmount *lower_mnt); + struct vfsmount *lower_mnt, + const struct cred *cred); int ecryptfs_init_persistent_file(struct dentry *ecryptfs_dentry); #endif /* #ifndef ECRYPTFS_KERNEL_H */ diff --git a/fs/ecryptfs/file.c b/fs/ecryptfs/file.c index 9244d653743..eb3dc4c7ac0 100644 --- a/fs/ecryptfs/file.c +++ b/fs/ecryptfs/file.c @@ -71,12 +71,11 @@ struct ecryptfs_getdents_callback { void *dirent; struct dentry *dentry; filldir_t filldir; - int err; int filldir_called; int entries_written; }; -/* Inspired by generic filldir in fs/readir.c */ +/* Inspired by generic filldir in fs/readdir.c */ static int ecryptfs_filldir(void *dirent, const char *name, int namelen, loff_t offset, u64 ino, unsigned int d_type) @@ -125,18 +124,18 @@ static int ecryptfs_readdir(struct file *file, void *dirent, filldir_t filldir) buf.dirent = dirent; buf.dentry = file->f_path.dentry; buf.filldir = filldir; -retry: buf.filldir_called = 0; buf.entries_written = 0; - buf.err = 0; rc = vfs_readdir(lower_file, ecryptfs_filldir, (void *)&buf); - if (buf.err) - rc = buf.err; - if (buf.filldir_called && !buf.entries_written) - goto retry; file->f_pos = lower_file->f_pos; + if (rc < 0) + goto out; + if (buf.filldir_called && !buf.entries_written) + goto out; if (rc >= 0) - fsstack_copy_attr_atime(inode, lower_file->f_path.dentry->d_inode); + fsstack_copy_attr_atime(inode, + lower_file->f_path.dentry->d_inode); +out: return rc; } diff --git a/fs/ecryptfs/keystore.c b/fs/ecryptfs/keystore.c index f5b76a331b9..0d713b69194 100644 --- a/fs/ecryptfs/keystore.c +++ b/fs/ecryptfs/keystore.c @@ -234,8 +234,8 @@ parse_tag_65_packet(struct ecryptfs_session_key *session_key, u8 *cipher_code, } i += data_len; if (message_len < (i + m_size)) { - ecryptfs_printk(KERN_ERR, "The received netlink message is " - "shorter than expected\n"); + ecryptfs_printk(KERN_ERR, "The message received from ecryptfsd " + "is shorter than expected\n"); rc = -EIO; goto out; } @@ -438,8 +438,8 @@ decrypt_pki_encrypted_session_key(struct ecryptfs_auth_tok *auth_tok, struct ecryptfs_msg_ctx *msg_ctx; struct ecryptfs_message *msg = NULL; char *auth_tok_sig; - char *netlink_message; - size_t netlink_message_length; + char *payload; + size_t payload_len; int rc; rc = ecryptfs_get_auth_tok_sig(&auth_tok_sig, auth_tok); @@ -449,15 +449,15 @@ decrypt_pki_encrypted_session_key(struct ecryptfs_auth_tok *auth_tok, goto out; } rc = write_tag_64_packet(auth_tok_sig, &(auth_tok->session_key), - &netlink_message, &netlink_message_length); + &payload, &payload_len); if (rc) { ecryptfs_printk(KERN_ERR, "Failed to write tag 64 packet\n"); goto out; } - rc = ecryptfs_send_message(ecryptfs_transport, netlink_message, - netlink_message_length, &msg_ctx); + rc = ecryptfs_send_message(payload, payload_len, &msg_ctx); if (rc) { - ecryptfs_printk(KERN_ERR, "Error sending netlink message\n"); + ecryptfs_printk(KERN_ERR, "Error sending message to " + "ecryptfsd\n"); goto out; } rc = ecryptfs_wait_for_response(msg_ctx, &msg); @@ -1037,17 +1037,14 @@ static int decrypt_passphrase_encrypted_session_key(struct ecryptfs_auth_tok *auth_tok, struct ecryptfs_crypt_stat *crypt_stat) { - struct scatterlist dst_sg; - struct scatterlist src_sg; + struct scatterlist dst_sg[2]; + struct scatterlist src_sg[2]; struct mutex *tfm_mutex; struct blkcipher_desc desc = { .flags = CRYPTO_TFM_REQ_MAY_SLEEP }; int rc = 0; - sg_init_table(&dst_sg, 1); - sg_init_table(&src_sg, 1); - if (unlikely(ecryptfs_verbosity > 0)) { ecryptfs_printk( KERN_DEBUG, "Session key encryption key (size [%d]):\n", @@ -1066,8 +1063,8 @@ decrypt_passphrase_encrypted_session_key(struct ecryptfs_auth_tok *auth_tok, } rc = virt_to_scatterlist(auth_tok->session_key.encrypted_key, auth_tok->session_key.encrypted_key_size, - &src_sg, 1); - if (rc != 1) { + src_sg, 2); + if (rc < 1 || rc > 2) { printk(KERN_ERR "Internal error whilst attempting to convert " "auth_tok->session_key.encrypted_key to scatterlist; " "expected rc = 1; got rc = [%d]. " @@ -1079,8 +1076,8 @@ decrypt_passphrase_encrypted_session_key(struct ecryptfs_auth_tok *auth_tok, auth_tok->session_key.encrypted_key_size; rc = virt_to_scatterlist(auth_tok->session_key.decrypted_key, auth_tok->session_key.decrypted_key_size, - &dst_sg, 1); - if (rc != 1) { + dst_sg, 2); + if (rc < 1 || rc > 2) { printk(KERN_ERR "Internal error whilst attempting to convert " "auth_tok->session_key.decrypted_key to scatterlist; " "expected rc = 1; got rc = [%d]\n", rc); @@ -1096,7 +1093,7 @@ decrypt_passphrase_encrypted_session_key(struct ecryptfs_auth_tok *auth_tok, rc = -EINVAL; goto out; } - rc = crypto_blkcipher_decrypt(&desc, &dst_sg, &src_sg, + rc = crypto_blkcipher_decrypt(&desc, dst_sg, src_sg, auth_tok->session_key.encrypted_key_size); mutex_unlock(tfm_mutex); if (unlikely(rc)) { @@ -1333,23 +1330,22 @@ pki_encrypt_session_key(struct ecryptfs_auth_tok *auth_tok, struct ecryptfs_key_record *key_rec) { struct ecryptfs_msg_ctx *msg_ctx = NULL; - char *netlink_payload; - size_t netlink_payload_length; + char *payload = NULL; + size_t payload_len; struct ecryptfs_message *msg; int rc; rc = write_tag_66_packet(auth_tok->token.private_key.signature, ecryptfs_code_for_cipher_string(crypt_stat), - crypt_stat, &netlink_payload, - &netlink_payload_length); + crypt_stat, &payload, &payload_len); if (rc) { ecryptfs_printk(KERN_ERR, "Error generating tag 66 packet\n"); goto out; } - rc = ecryptfs_send_message(ecryptfs_transport, netlink_payload, - netlink_payload_length, &msg_ctx); + rc = ecryptfs_send_message(payload, payload_len, &msg_ctx); if (rc) { - ecryptfs_printk(KERN_ERR, "Error sending netlink message\n"); + ecryptfs_printk(KERN_ERR, "Error sending message to " + "ecryptfsd\n"); goto out; } rc = ecryptfs_wait_for_response(msg_ctx, &msg); @@ -1364,8 +1360,7 @@ pki_encrypt_session_key(struct ecryptfs_auth_tok *auth_tok, ecryptfs_printk(KERN_ERR, "Error parsing tag 67 packet\n"); kfree(msg); out: - if (netlink_payload) - kfree(netlink_payload); + kfree(payload); return rc; } /** @@ -1541,8 +1536,8 @@ write_tag_3_packet(char *dest, size_t *remaining_bytes, size_t i; size_t encrypted_session_key_valid = 0; char session_key_encryption_key[ECRYPTFS_MAX_KEY_BYTES]; - struct scatterlist dst_sg; - struct scatterlist src_sg; + struct scatterlist dst_sg[2]; + struct scatterlist src_sg[2]; struct mutex *tfm_mutex = NULL; u8 cipher_code; size_t packet_size_length; @@ -1621,8 +1616,8 @@ write_tag_3_packet(char *dest, size_t *remaining_bytes, ecryptfs_dump_hex(session_key_encryption_key, 16); } rc = virt_to_scatterlist(crypt_stat->key, key_rec->enc_key_size, - &src_sg, 1); - if (rc != 1) { + src_sg, 2); + if (rc < 1 || rc > 2) { ecryptfs_printk(KERN_ERR, "Error generating scatterlist " "for crypt_stat session key; expected rc = 1; " "got rc = [%d]. key_rec->enc_key_size = [%d]\n", @@ -1631,8 +1626,8 @@ write_tag_3_packet(char *dest, size_t *remaining_bytes, goto out; } rc = virt_to_scatterlist(key_rec->enc_key, key_rec->enc_key_size, - &dst_sg, 1); - if (rc != 1) { + dst_sg, 2); + if (rc < 1 || rc > 2) { ecryptfs_printk(KERN_ERR, "Error generating scatterlist " "for crypt_stat encrypted session key; " "expected rc = 1; got rc = [%d]. " @@ -1653,7 +1648,7 @@ write_tag_3_packet(char *dest, size_t *remaining_bytes, rc = 0; ecryptfs_printk(KERN_DEBUG, "Encrypting [%d] bytes of the key\n", crypt_stat->key_size); - rc = crypto_blkcipher_encrypt(&desc, &dst_sg, &src_sg, + rc = crypto_blkcipher_encrypt(&desc, dst_sg, src_sg, (*key_rec).enc_key_size); mutex_unlock(tfm_mutex); if (rc) { diff --git a/fs/ecryptfs/kthread.c b/fs/ecryptfs/kthread.c index c440c6b58b2..c6d7a4d748a 100644 --- a/fs/ecryptfs/kthread.c +++ b/fs/ecryptfs/kthread.c @@ -73,7 +73,7 @@ static int ecryptfs_threadfn(void *ignored) mntget(req->lower_mnt); (*req->lower_file) = dentry_open( req->lower_dentry, req->lower_mnt, - (O_RDWR | O_LARGEFILE)); + (O_RDWR | O_LARGEFILE), current_cred()); req->flags |= ECRYPTFS_REQ_PROCESSED; } wake_up(&req->wait); @@ -132,7 +132,8 @@ void ecryptfs_destroy_kthread(void) */ int ecryptfs_privileged_open(struct file **lower_file, struct dentry *lower_dentry, - struct vfsmount *lower_mnt) + struct vfsmount *lower_mnt, + const struct cred *cred) { struct ecryptfs_open_req *req; int rc = 0; @@ -143,7 +144,7 @@ int ecryptfs_privileged_open(struct file **lower_file, dget(lower_dentry); mntget(lower_mnt); (*lower_file) = dentry_open(lower_dentry, lower_mnt, - (O_RDWR | O_LARGEFILE)); + (O_RDWR | O_LARGEFILE), cred); if (!IS_ERR(*lower_file)) goto out; req = kmem_cache_alloc(ecryptfs_open_req_cache, GFP_KERNEL); @@ -184,7 +185,7 @@ int ecryptfs_privileged_open(struct file **lower_file, dget(lower_dentry); mntget(lower_mnt); (*lower_file) = dentry_open(lower_dentry, lower_mnt, - (O_RDONLY | O_LARGEFILE)); + (O_RDONLY | O_LARGEFILE), cred); if (IS_ERR(*lower_file)) { rc = PTR_ERR(*req->lower_file); (*lower_file) = NULL; diff --git a/fs/ecryptfs/main.c b/fs/ecryptfs/main.c index 8ebe9a5d1d9..fd630713c5c 100644 --- a/fs/ecryptfs/main.c +++ b/fs/ecryptfs/main.c @@ -30,7 +30,6 @@ #include <linux/namei.h> #include <linux/skbuff.h> #include <linux/crypto.h> -#include <linux/netlink.h> #include <linux/mount.h> #include <linux/pagemap.h> #include <linux/key.h> @@ -49,8 +48,7 @@ MODULE_PARM_DESC(ecryptfs_verbosity, "0, which is Quiet)"); /** - * Module parameter that defines the number of netlink message buffer - * elements + * Module parameter that defines the number of message buffer elements */ unsigned int ecryptfs_message_buf_len = ECRYPTFS_DEFAULT_MSG_CTX_ELEMS; @@ -60,9 +58,9 @@ MODULE_PARM_DESC(ecryptfs_message_buf_len, /** * Module parameter that defines the maximum guaranteed amount of time to wait - * for a response through netlink. The actual sleep time will be, more than + * for a response from ecryptfsd. The actual sleep time will be, more than * likely, a small amount greater than this specified value, but only less if - * the netlink message successfully arrives. + * the message successfully arrives. */ signed long ecryptfs_message_wait_timeout = ECRYPTFS_MAX_MSG_CTX_TTL / HZ; @@ -83,8 +81,6 @@ module_param(ecryptfs_number_of_users, uint, 0); MODULE_PARM_DESC(ecryptfs_number_of_users, "An estimate of the number of " "concurrent users of eCryptfs"); -unsigned int ecryptfs_transport = ECRYPTFS_DEFAULT_TRANSPORT; - void __ecryptfs_printk(const char *fmt, ...) { va_list args; @@ -119,6 +115,7 @@ void __ecryptfs_printk(const char *fmt, ...) */ int ecryptfs_init_persistent_file(struct dentry *ecryptfs_dentry) { + const struct cred *cred = current_cred(); struct ecryptfs_inode_info *inode_info = ecryptfs_inode_to_private(ecryptfs_dentry->d_inode); int rc = 0; @@ -131,7 +128,7 @@ int ecryptfs_init_persistent_file(struct dentry *ecryptfs_dentry) lower_dentry = ecryptfs_dentry_to_lower(ecryptfs_dentry); rc = ecryptfs_privileged_open(&inode_info->lower_file, - lower_dentry, lower_mnt); + lower_dentry, lower_mnt, cred); if (rc || IS_ERR(inode_info->lower_file)) { printk(KERN_ERR "Error opening lower persistent file " "for lower_dentry [0x%p] and lower_mnt [0x%p]; " @@ -475,31 +472,26 @@ out: */ static int ecryptfs_read_super(struct super_block *sb, const char *dev_name) { + struct path path; int rc; - struct nameidata nd; - struct dentry *lower_root; - struct vfsmount *lower_mnt; - memset(&nd, 0, sizeof(struct nameidata)); - rc = path_lookup(dev_name, LOOKUP_FOLLOW | LOOKUP_DIRECTORY, &nd); + rc = kern_path(dev_name, LOOKUP_FOLLOW | LOOKUP_DIRECTORY, &path); if (rc) { ecryptfs_printk(KERN_WARNING, "path_lookup() failed\n"); goto out; } - lower_root = nd.path.dentry; - lower_mnt = nd.path.mnt; - ecryptfs_set_superblock_lower(sb, lower_root->d_sb); - sb->s_maxbytes = lower_root->d_sb->s_maxbytes; - sb->s_blocksize = lower_root->d_sb->s_blocksize; - ecryptfs_set_dentry_lower(sb->s_root, lower_root); - ecryptfs_set_dentry_lower_mnt(sb->s_root, lower_mnt); - rc = ecryptfs_interpose(lower_root, sb->s_root, sb, 0); + ecryptfs_set_superblock_lower(sb, path.dentry->d_sb); + sb->s_maxbytes = path.dentry->d_sb->s_maxbytes; + sb->s_blocksize = path.dentry->d_sb->s_blocksize; + ecryptfs_set_dentry_lower(sb->s_root, path.dentry); + ecryptfs_set_dentry_lower_mnt(sb->s_root, path.mnt); + rc = ecryptfs_interpose(path.dentry, sb->s_root, sb, 0); if (rc) goto out_free; rc = 0; goto out; out_free: - path_put(&nd.path); + path_put(&path); out: return rc; } @@ -779,10 +771,11 @@ static int __init ecryptfs_init(void) "rc = [%d]\n", __func__, rc); goto out_do_sysfs_unregistration; } - rc = ecryptfs_init_messaging(ecryptfs_transport); + rc = ecryptfs_init_messaging(); if (rc) { printk(KERN_ERR "Failure occured while attempting to " - "initialize the eCryptfs netlink socket\n"); + "initialize the communications channel to " + "ecryptfsd\n"); goto out_destroy_kthread; } rc = ecryptfs_init_crypto(); @@ -797,7 +790,7 @@ static int __init ecryptfs_init(void) goto out; out_release_messaging: - ecryptfs_release_messaging(ecryptfs_transport); + ecryptfs_release_messaging(); out_destroy_kthread: ecryptfs_destroy_kthread(); out_do_sysfs_unregistration: @@ -818,7 +811,7 @@ static void __exit ecryptfs_exit(void) if (rc) printk(KERN_ERR "Failure whilst attempting to destroy crypto; " "rc = [%d]\n", rc); - ecryptfs_release_messaging(ecryptfs_transport); + ecryptfs_release_messaging(); ecryptfs_destroy_kthread(); do_sysfs_unregistration(); unregister_filesystem(&ecryptfs_fs_type); diff --git a/fs/ecryptfs/messaging.c b/fs/ecryptfs/messaging.c index 1b5c20058ac..6913f727624 100644 --- a/fs/ecryptfs/messaging.c +++ b/fs/ecryptfs/messaging.c @@ -134,12 +134,11 @@ out: } static int -ecryptfs_send_message_locked(unsigned int transport, char *data, int data_len, - u8 msg_type, struct ecryptfs_msg_ctx **msg_ctx); +ecryptfs_send_message_locked(char *data, int data_len, u8 msg_type, + struct ecryptfs_msg_ctx **msg_ctx); /** * ecryptfs_send_raw_message - * @transport: Transport type * @msg_type: Message type * @daemon: Daemon struct for recipient of message * @@ -150,38 +149,25 @@ ecryptfs_send_message_locked(unsigned int transport, char *data, int data_len, * * Returns zero on success; non-zero otherwise */ -static int ecryptfs_send_raw_message(unsigned int transport, u8 msg_type, +static int ecryptfs_send_raw_message(u8 msg_type, struct ecryptfs_daemon *daemon) { struct ecryptfs_msg_ctx *msg_ctx; int rc; - switch(transport) { - case ECRYPTFS_TRANSPORT_NETLINK: - rc = ecryptfs_send_netlink(NULL, 0, NULL, msg_type, 0, - daemon->pid); - break; - case ECRYPTFS_TRANSPORT_MISCDEV: - rc = ecryptfs_send_message_locked(transport, NULL, 0, msg_type, - &msg_ctx); - if (rc) { - printk(KERN_ERR "%s: Error whilst attempting to send " - "message via procfs; rc = [%d]\n", __func__, rc); - goto out; - } - /* Raw messages are logically context-free (e.g., no - * reply is expected), so we set the state of the - * ecryptfs_msg_ctx object to indicate that it should - * be freed as soon as the transport sends out the message. */ - mutex_lock(&msg_ctx->mux); - msg_ctx->state = ECRYPTFS_MSG_CTX_STATE_NO_REPLY; - mutex_unlock(&msg_ctx->mux); - break; - case ECRYPTFS_TRANSPORT_CONNECTOR: - case ECRYPTFS_TRANSPORT_RELAYFS: - default: - rc = -ENOSYS; + rc = ecryptfs_send_message_locked(NULL, 0, msg_type, &msg_ctx); + if (rc) { + printk(KERN_ERR "%s: Error whilst attempting to send " + "message to ecryptfsd; rc = [%d]\n", __func__, rc); + goto out; } + /* Raw messages are logically context-free (e.g., no + * reply is expected), so we set the state of the + * ecryptfs_msg_ctx object to indicate that it should + * be freed as soon as the message is sent. */ + mutex_lock(&msg_ctx->mux); + msg_ctx->state = ECRYPTFS_MSG_CTX_STATE_NO_REPLY; + mutex_unlock(&msg_ctx->mux); out: return rc; } @@ -227,7 +213,6 @@ out: /** * ecryptfs_process_helo - * @transport: The underlying transport (netlink, etc.) * @euid: The user ID owner of the message * @user_ns: The namespace in which @euid applies * @pid: The process ID for the userspace program that sent the @@ -239,8 +224,8 @@ out: * Returns zero after adding a new daemon to the hash list; * non-zero otherwise. */ -int ecryptfs_process_helo(unsigned int transport, uid_t euid, - struct user_namespace *user_ns, struct pid *pid) +int ecryptfs_process_helo(uid_t euid, struct user_namespace *user_ns, + struct pid *pid) { struct ecryptfs_daemon *new_daemon; struct ecryptfs_daemon *old_daemon; @@ -252,8 +237,7 @@ int ecryptfs_process_helo(unsigned int transport, uid_t euid, printk(KERN_WARNING "Received request from user [%d] " "to register daemon [0x%p]; unregistering daemon " "[0x%p]\n", euid, pid, old_daemon->pid); - rc = ecryptfs_send_raw_message(transport, ECRYPTFS_MSG_QUIT, - old_daemon); + rc = ecryptfs_send_raw_message(ECRYPTFS_MSG_QUIT, old_daemon); if (rc) printk(KERN_WARNING "Failed to send QUIT " "message to daemon [0x%p]; rc = [%d]\n", @@ -376,7 +360,8 @@ int ecryptfs_process_response(struct ecryptfs_message *msg, uid_t euid, struct ecryptfs_msg_ctx *msg_ctx; size_t msg_size; struct nsproxy *nsproxy; - struct user_namespace *current_user_ns; + struct user_namespace *tsk_user_ns; + uid_t ctx_euid; int rc; if (msg->index >= ecryptfs_message_buf_len) { @@ -400,9 +385,9 @@ int ecryptfs_process_response(struct ecryptfs_message *msg, uid_t euid, mutex_unlock(&ecryptfs_daemon_hash_mux); goto wake_up; } - current_user_ns = nsproxy->user_ns; - rc = ecryptfs_find_daemon_by_euid(&daemon, msg_ctx->task->euid, - current_user_ns); + tsk_user_ns = __task_cred(msg_ctx->task)->user->user_ns; + ctx_euid = task_euid(msg_ctx->task); + rc = ecryptfs_find_daemon_by_euid(&daemon, ctx_euid, tsk_user_ns); rcu_read_unlock(); mutex_unlock(&ecryptfs_daemon_hash_mux); if (rc) { @@ -410,28 +395,28 @@ int ecryptfs_process_response(struct ecryptfs_message *msg, uid_t euid, printk(KERN_WARNING "%s: User [%d] received a " "message response from process [0x%p] but does " "not have a registered daemon\n", __func__, - msg_ctx->task->euid, pid); + ctx_euid, pid); goto wake_up; } - if (msg_ctx->task->euid != euid) { + if (ctx_euid != euid) { rc = -EBADMSG; printk(KERN_WARNING "%s: Received message from user " "[%d]; expected message from user [%d]\n", __func__, - euid, msg_ctx->task->euid); + euid, ctx_euid); goto unlock; } - if (current_user_ns != user_ns) { + if (tsk_user_ns != user_ns) { rc = -EBADMSG; printk(KERN_WARNING "%s: Received message from user_ns " "[0x%p]; expected message from user_ns [0x%p]\n", - __func__, user_ns, nsproxy->user_ns); + __func__, user_ns, tsk_user_ns); goto unlock; } if (daemon->pid != pid) { rc = -EBADMSG; printk(KERN_ERR "%s: User [%d] sent a message response " "from an unrecognized process [0x%p]\n", - __func__, msg_ctx->task->euid, pid); + __func__, ctx_euid, pid); goto unlock; } if (msg_ctx->state != ECRYPTFS_MSG_CTX_STATE_PENDING) { @@ -467,8 +452,6 @@ out: /** * ecryptfs_send_message_locked - * @transport: The transport over which to send the message (i.e., - * netlink) * @data: The data to send * @data_len: The length of data * @msg_ctx: The message context allocated for the send @@ -478,18 +461,18 @@ out: * Returns zero on success; non-zero otherwise */ static int -ecryptfs_send_message_locked(unsigned int transport, char *data, int data_len, - u8 msg_type, struct ecryptfs_msg_ctx **msg_ctx) +ecryptfs_send_message_locked(char *data, int data_len, u8 msg_type, + struct ecryptfs_msg_ctx **msg_ctx) { struct ecryptfs_daemon *daemon; + uid_t euid = current_euid(); int rc; - rc = ecryptfs_find_daemon_by_euid(&daemon, current->euid, - current->nsproxy->user_ns); + rc = ecryptfs_find_daemon_by_euid(&daemon, euid, current_user_ns()); if (rc || !daemon) { rc = -ENOTCONN; printk(KERN_ERR "%s: User [%d] does not have a daemon " - "registered\n", __func__, current->euid); + "registered\n", __func__, euid); goto out; } mutex_lock(&ecryptfs_msg_ctx_lists_mux); @@ -503,20 +486,8 @@ ecryptfs_send_message_locked(unsigned int transport, char *data, int data_len, ecryptfs_msg_ctx_free_to_alloc(*msg_ctx); mutex_unlock(&(*msg_ctx)->mux); mutex_unlock(&ecryptfs_msg_ctx_lists_mux); - switch (transport) { - case ECRYPTFS_TRANSPORT_NETLINK: - rc = ecryptfs_send_netlink(data, data_len, *msg_ctx, msg_type, - 0, daemon->pid); - break; - case ECRYPTFS_TRANSPORT_MISCDEV: - rc = ecryptfs_send_miscdev(data, data_len, *msg_ctx, msg_type, - 0, daemon); - break; - case ECRYPTFS_TRANSPORT_CONNECTOR: - case ECRYPTFS_TRANSPORT_RELAYFS: - default: - rc = -ENOSYS; - } + rc = ecryptfs_send_miscdev(data, data_len, *msg_ctx, msg_type, 0, + daemon); if (rc) printk(KERN_ERR "%s: Error attempting to send message to " "userspace daemon; rc = [%d]\n", __func__, rc); @@ -526,8 +497,6 @@ out: /** * ecryptfs_send_message - * @transport: The transport over which to send the message (i.e., - * netlink) * @data: The data to send * @data_len: The length of data * @msg_ctx: The message context allocated for the send @@ -536,14 +505,14 @@ out: * * Returns zero on success; non-zero otherwise */ -int ecryptfs_send_message(unsigned int transport, char *data, int data_len, +int ecryptfs_send_message(char *data, int data_len, struct ecryptfs_msg_ctx **msg_ctx) { int rc; mutex_lock(&ecryptfs_daemon_hash_mux); - rc = ecryptfs_send_message_locked(transport, data, data_len, - ECRYPTFS_MSG_REQUEST, msg_ctx); + rc = ecryptfs_send_message_locked(data, data_len, ECRYPTFS_MSG_REQUEST, + msg_ctx); mutex_unlock(&ecryptfs_daemon_hash_mux); return rc; } @@ -586,7 +555,7 @@ sleep: return rc; } -int ecryptfs_init_messaging(unsigned int transport) +int ecryptfs_init_messaging(void) { int i; int rc = 0; @@ -639,27 +608,14 @@ int ecryptfs_init_messaging(unsigned int transport) mutex_unlock(&ecryptfs_msg_ctx_arr[i].mux); } mutex_unlock(&ecryptfs_msg_ctx_lists_mux); - switch(transport) { - case ECRYPTFS_TRANSPORT_NETLINK: - rc = ecryptfs_init_netlink(); - if (rc) - ecryptfs_release_messaging(transport); - break; - case ECRYPTFS_TRANSPORT_MISCDEV: - rc = ecryptfs_init_ecryptfs_miscdev(); - if (rc) - ecryptfs_release_messaging(transport); - break; - case ECRYPTFS_TRANSPORT_CONNECTOR: - case ECRYPTFS_TRANSPORT_RELAYFS: - default: - rc = -ENOSYS; - } + rc = ecryptfs_init_ecryptfs_miscdev(); + if (rc) + ecryptfs_release_messaging(); out: return rc; } -void ecryptfs_release_messaging(unsigned int transport) +void ecryptfs_release_messaging(void) { if (ecryptfs_msg_ctx_arr) { int i; @@ -698,17 +654,6 @@ void ecryptfs_release_messaging(unsigned int transport) kfree(ecryptfs_daemon_hash); mutex_unlock(&ecryptfs_daemon_hash_mux); } - switch(transport) { - case ECRYPTFS_TRANSPORT_NETLINK: - ecryptfs_release_netlink(); - break; - case ECRYPTFS_TRANSPORT_MISCDEV: - ecryptfs_destroy_ecryptfs_miscdev(); - break; - case ECRYPTFS_TRANSPORT_CONNECTOR: - case ECRYPTFS_TRANSPORT_RELAYFS: - default: - break; - } + ecryptfs_destroy_ecryptfs_miscdev(); return; } diff --git a/fs/ecryptfs/miscdev.c b/fs/ecryptfs/miscdev.c index b484792a099..efd95a0ed1e 100644 --- a/fs/ecryptfs/miscdev.c +++ b/fs/ecryptfs/miscdev.c @@ -42,12 +42,12 @@ ecryptfs_miscdev_poll(struct file *file, poll_table *pt) { struct ecryptfs_daemon *daemon; unsigned int mask = 0; + uid_t euid = current_euid(); int rc; mutex_lock(&ecryptfs_daemon_hash_mux); /* TODO: Just use file->private_data? */ - rc = ecryptfs_find_daemon_by_euid(&daemon, current->euid, - current->nsproxy->user_ns); + rc = ecryptfs_find_daemon_by_euid(&daemon, euid, current_user_ns()); BUG_ON(rc || !daemon); mutex_lock(&daemon->mux); mutex_unlock(&ecryptfs_daemon_hash_mux); @@ -83,6 +83,7 @@ static int ecryptfs_miscdev_open(struct inode *inode, struct file *file) { struct ecryptfs_daemon *daemon = NULL; + uid_t euid = current_euid(); int rc; mutex_lock(&ecryptfs_daemon_hash_mux); @@ -93,11 +94,9 @@ ecryptfs_miscdev_open(struct inode *inode, struct file *file) "count; rc = [%d]\n", __func__, rc); goto out_unlock_daemon_list; } - rc = ecryptfs_find_daemon_by_euid(&daemon, current->euid, - current->nsproxy->user_ns); + rc = ecryptfs_find_daemon_by_euid(&daemon, euid, current_user_ns()); if (rc || !daemon) { - rc = ecryptfs_spawn_daemon(&daemon, current->euid, - current->nsproxy->user_ns, + rc = ecryptfs_spawn_daemon(&daemon, euid, current_user_ns(), task_pid(current)); if (rc) { printk(KERN_ERR "%s: Error attempting to spawn daemon; " @@ -147,11 +146,11 @@ static int ecryptfs_miscdev_release(struct inode *inode, struct file *file) { struct ecryptfs_daemon *daemon = NULL; + uid_t euid = current_euid(); int rc; mutex_lock(&ecryptfs_daemon_hash_mux); - rc = ecryptfs_find_daemon_by_euid(&daemon, current->euid, - current->nsproxy->user_ns); + rc = ecryptfs_find_daemon_by_euid(&daemon, euid, current_user_ns()); BUG_ON(rc || !daemon); mutex_lock(&daemon->mux); BUG_ON(daemon->pid != task_pid(current)); @@ -246,12 +245,12 @@ ecryptfs_miscdev_read(struct file *file, char __user *buf, size_t count, char packet_length[3]; size_t i; size_t total_length; + uid_t euid = current_euid(); int rc; mutex_lock(&ecryptfs_daemon_hash_mux); /* TODO: Just use file->private_data? */ - rc = ecryptfs_find_daemon_by_euid(&daemon, current->euid, - current->nsproxy->user_ns); + rc = ecryptfs_find_daemon_by_euid(&daemon, euid, current_user_ns()); BUG_ON(rc || !daemon); mutex_lock(&daemon->mux); if (daemon->flags & ECRYPTFS_DAEMON_ZOMBIE) { @@ -290,8 +289,8 @@ check_list: * message from the queue; try again */ goto check_list; } - BUG_ON(current->euid != daemon->euid); - BUG_ON(current->nsproxy->user_ns != daemon->user_ns); + BUG_ON(euid != daemon->euid); + BUG_ON(current_user_ns() != daemon->user_ns); BUG_ON(task_pid(current) != daemon->pid); msg_ctx = list_first_entry(&daemon->msg_ctx_out_queue, struct ecryptfs_msg_ctx, daemon_out_list); @@ -414,6 +413,7 @@ ecryptfs_miscdev_write(struct file *file, const char __user *buf, size_t packet_size, packet_size_length, i; ssize_t sz = 0; char *data; + uid_t euid = current_euid(); int rc; if (count == 0) @@ -463,8 +463,7 @@ ecryptfs_miscdev_write(struct file *file, const char __user *buf, goto out_free; } rc = ecryptfs_miscdev_response(&data[i], packet_size, - current->euid, - current->nsproxy->user_ns, + euid, current_user_ns(), task_pid(current), seq); if (rc) printk(KERN_WARNING "%s: Failed to deliver miscdev " diff --git a/fs/ecryptfs/mmap.c b/fs/ecryptfs/mmap.c index 245c2dc02d5..04d7b3fa1ac 100644 --- a/fs/ecryptfs/mmap.c +++ b/fs/ecryptfs/mmap.c @@ -265,22 +265,34 @@ out: } /** - * ecryptfs_prepare_write + * ecryptfs_write_begin * @file: The eCryptfs file - * @page: The eCryptfs page - * @from: The start byte from which we will write - * @to: The end byte to which we will write + * @mapping: The eCryptfs object + * @pos: The file offset at which to start writing + * @len: Length of the write + * @flags: Various flags + * @pagep: Pointer to return the page + * @fsdata: Pointer to return fs data (unused) * * This function must zero any hole we create * * Returns zero on success; non-zero otherwise */ -static int ecryptfs_prepare_write(struct file *file, struct page *page, - unsigned from, unsigned to) +static int ecryptfs_write_begin(struct file *file, + struct address_space *mapping, + loff_t pos, unsigned len, unsigned flags, + struct page **pagep, void **fsdata) { + pgoff_t index = pos >> PAGE_CACHE_SHIFT; + struct page *page; loff_t prev_page_end_size; int rc = 0; + page = __grab_cache_page(mapping, index); + if (!page) + return -ENOMEM; + *pagep = page; + if (!PageUptodate(page)) { struct ecryptfs_crypt_stat *crypt_stat = &ecryptfs_inode_to_private( @@ -289,8 +301,7 @@ static int ecryptfs_prepare_write(struct file *file, struct page *page, if (!(crypt_stat->flags & ECRYPTFS_ENCRYPTED) || (crypt_stat->flags & ECRYPTFS_NEW_FILE)) { rc = ecryptfs_read_lower_page_segment( - page, page->index, 0, PAGE_CACHE_SIZE, - page->mapping->host); + page, index, 0, PAGE_CACHE_SIZE, mapping->host); if (rc) { printk(KERN_ERR "%s: Error attemping to read " "lower page segment; rc = [%d]\n", @@ -316,8 +327,8 @@ static int ecryptfs_prepare_write(struct file *file, struct page *page, SetPageUptodate(page); } else { rc = ecryptfs_read_lower_page_segment( - page, page->index, 0, PAGE_CACHE_SIZE, - page->mapping->host); + page, index, 0, PAGE_CACHE_SIZE, + mapping->host); if (rc) { printk(KERN_ERR "%s: Error reading " "page; rc = [%d]\n", @@ -339,10 +350,10 @@ static int ecryptfs_prepare_write(struct file *file, struct page *page, SetPageUptodate(page); } } - prev_page_end_size = ((loff_t)page->index << PAGE_CACHE_SHIFT); + prev_page_end_size = ((loff_t)index << PAGE_CACHE_SHIFT); /* If creating a page or more of holes, zero them out via truncate. * Note, this will increase i_size. */ - if (page->index != 0) { + if (index != 0) { if (prev_page_end_size > i_size_read(page->mapping->host)) { rc = ecryptfs_truncate(file->f_path.dentry, prev_page_end_size); @@ -357,8 +368,8 @@ static int ecryptfs_prepare_write(struct file *file, struct page *page, } /* Writing to a new page, and creating a small hole from start * of page? Zero it out. */ - if ((i_size_read(page->mapping->host) == prev_page_end_size) - && (from != 0)) + if ((i_size_read(mapping->host) == prev_page_end_size) + && (pos != 0)) zero_user(page, 0, PAGE_CACHE_SIZE); out: return rc; @@ -445,21 +456,28 @@ int ecryptfs_write_inode_size_to_metadata(struct inode *ecryptfs_inode) } /** - * ecryptfs_commit_write + * ecryptfs_write_end * @file: The eCryptfs file object + * @mapping: The eCryptfs object + * @pos: The file position + * @len: The length of the data (unused) + * @copied: The amount of data copied * @page: The eCryptfs page - * @from: Ignored (we rotate the page IV on each write) - * @to: Ignored + * @fsdata: The fsdata (unused) * * This is where we encrypt the data and pass the encrypted data to * the lower filesystem. In OpenPGP-compatible mode, we operate on * entire underlying packets. */ -static int ecryptfs_commit_write(struct file *file, struct page *page, - unsigned from, unsigned to) +static int ecryptfs_write_end(struct file *file, + struct address_space *mapping, + loff_t pos, unsigned len, unsigned copied, + struct page *page, void *fsdata) { - loff_t pos; - struct inode *ecryptfs_inode = page->mapping->host; + pgoff_t index = pos >> PAGE_CACHE_SHIFT; + unsigned from = pos & (PAGE_CACHE_SIZE - 1); + unsigned to = from + copied; + struct inode *ecryptfs_inode = mapping->host; struct ecryptfs_crypt_stat *crypt_stat = &ecryptfs_inode_to_private(file->f_path.dentry->d_inode)->crypt_stat; int rc; @@ -471,25 +489,22 @@ static int ecryptfs_commit_write(struct file *file, struct page *page, } else ecryptfs_printk(KERN_DEBUG, "Not a new file\n"); ecryptfs_printk(KERN_DEBUG, "Calling fill_zeros_to_end_of_page" - "(page w/ index = [0x%.16x], to = [%d])\n", page->index, - to); + "(page w/ index = [0x%.16x], to = [%d])\n", index, to); /* Fills in zeros if 'to' goes beyond inode size */ rc = fill_zeros_to_end_of_page(page, to); if (rc) { ecryptfs_printk(KERN_WARNING, "Error attempting to fill " - "zeros in page with index = [0x%.16x]\n", - page->index); + "zeros in page with index = [0x%.16x]\n", index); goto out; } rc = ecryptfs_encrypt_page(page); if (rc) { ecryptfs_printk(KERN_WARNING, "Error encrypting page (upper " - "index [0x%.16x])\n", page->index); + "index [0x%.16x])\n", index); goto out; } - pos = (((loff_t)page->index) << PAGE_CACHE_SHIFT) + to; - if (pos > i_size_read(ecryptfs_inode)) { - i_size_write(ecryptfs_inode, pos); + if (pos + copied > i_size_read(ecryptfs_inode)) { + i_size_write(ecryptfs_inode, pos + copied); ecryptfs_printk(KERN_DEBUG, "Expanded file size to " "[0x%.16x]\n", i_size_read(ecryptfs_inode)); } @@ -497,7 +512,11 @@ static int ecryptfs_commit_write(struct file *file, struct page *page, if (rc) printk(KERN_ERR "Error writing inode size to metadata; " "rc = [%d]\n", rc); + else + rc = copied; out: + unlock_page(page); + page_cache_release(page); return rc; } @@ -518,7 +537,7 @@ static sector_t ecryptfs_bmap(struct address_space *mapping, sector_t block) struct address_space_operations ecryptfs_aops = { .writepage = ecryptfs_writepage, .readpage = ecryptfs_readpage, - .prepare_write = ecryptfs_prepare_write, - .commit_write = ecryptfs_commit_write, + .write_begin = ecryptfs_write_begin, + .write_end = ecryptfs_write_end, .bmap = ecryptfs_bmap, }; diff --git a/fs/ecryptfs/netlink.c b/fs/ecryptfs/netlink.c deleted file mode 100644 index e0abad62b39..00000000000 --- a/fs/ecryptfs/netlink.c +++ /dev/null @@ -1,249 +0,0 @@ -/** - * eCryptfs: Linux filesystem encryption layer - * - * Copyright (C) 2004-2006 International Business Machines Corp. - * Author(s): Michael A. Halcrow <mhalcrow@us.ibm.com> - * Tyler Hicks <tyhicks@ou.edu> - * - * This program is free software; you can redistribute it and/or - * modify it under the terms of the GNU General Public License version - * 2 as published by the Free Software Foundation. - * - * This program is distributed in the hope that it will be useful, but - * WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU - * General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA - * 02111-1307, USA. - */ - -#include <net/sock.h> -#include <linux/hash.h> -#include <linux/random.h> -#include "ecryptfs_kernel.h" - -static struct sock *ecryptfs_nl_sock; - -/** - * ecryptfs_send_netlink - * @data: The data to include as the payload - * @data_len: The byte count of the data - * @msg_ctx: The netlink context that will be used to handle the - * response message - * @msg_type: The type of netlink message to send - * @msg_flags: The flags to include in the netlink header - * @daemon_pid: The process id of the daemon to send the message to - * - * Sends the data to the specified daemon pid and uses the netlink - * context element to store the data needed for validation upon - * receiving the response. The data and the netlink context can be - * null if just sending a netlink header is sufficient. Returns zero - * upon sending the message; non-zero upon error. - */ -int ecryptfs_send_netlink(char *data, int data_len, - struct ecryptfs_msg_ctx *msg_ctx, u8 msg_type, - u16 msg_flags, struct pid *daemon_pid) -{ - struct sk_buff *skb; - struct nlmsghdr *nlh; - struct ecryptfs_message *msg; - size_t payload_len; - int rc; - - payload_len = ((data && data_len) ? (sizeof(*msg) + data_len) : 0); - skb = alloc_skb(NLMSG_SPACE(payload_len), GFP_KERNEL); - if (!skb) { - rc = -ENOMEM; - ecryptfs_printk(KERN_ERR, "Failed to allocate socket buffer\n"); - goto out; - } - nlh = NLMSG_PUT(skb, pid_nr(daemon_pid), msg_ctx ? msg_ctx->counter : 0, - msg_type, payload_len); - nlh->nlmsg_flags = msg_flags; - if (msg_ctx && payload_len) { - msg = (struct ecryptfs_message *)NLMSG_DATA(nlh); - msg->index = msg_ctx->index; - msg->data_len = data_len; - memcpy(msg->data, data, data_len); - } - rc = netlink_unicast(ecryptfs_nl_sock, skb, pid_nr(daemon_pid), 0); - if (rc < 0) { - ecryptfs_printk(KERN_ERR, "Failed to send eCryptfs netlink " - "message; rc = [%d]\n", rc); - goto out; - } - rc = 0; - goto out; -nlmsg_failure: - rc = -EMSGSIZE; - kfree_skb(skb); -out: - return rc; -} - -/** - * ecryptfs_process_nl_reponse - * @skb: The socket buffer containing the netlink message of state - * RESPONSE - * - * Processes a response message after sending a operation request to - * userspace. Attempts to assign the msg to a netlink context element - * at the index specified in the msg. The sk_buff and nlmsghdr must - * be validated before this function. Returns zero upon delivery to - * desired context element; non-zero upon delivery failure or error. - */ -static int ecryptfs_process_nl_response(struct sk_buff *skb) -{ - struct nlmsghdr *nlh = nlmsg_hdr(skb); - struct ecryptfs_message *msg = NLMSG_DATA(nlh); - struct pid *pid; - int rc; - - if (skb->len - NLMSG_HDRLEN - sizeof(*msg) != msg->data_len) { - rc = -EINVAL; - ecryptfs_printk(KERN_ERR, "Received netlink message with " - "incorrectly specified data length\n"); - goto out; - } - pid = find_get_pid(NETLINK_CREDS(skb)->pid); - rc = ecryptfs_process_response(msg, NETLINK_CREDS(skb)->uid, NULL, - pid, nlh->nlmsg_seq); - put_pid(pid); - if (rc) - printk(KERN_ERR - "Error processing response message; rc = [%d]\n", rc); -out: - return rc; -} - -/** - * ecryptfs_process_nl_helo - * @skb: The socket buffer containing the nlmsghdr in HELO state - * - * Gets uid and pid of the skb and adds the values to the daemon id - * hash. Returns zero after adding a new daemon id to the hash list; - * non-zero otherwise. - */ -static int ecryptfs_process_nl_helo(struct sk_buff *skb) -{ - struct pid *pid; - int rc; - - pid = find_get_pid(NETLINK_CREDS(skb)->pid); - rc = ecryptfs_process_helo(ECRYPTFS_TRANSPORT_NETLINK, - NETLINK_CREDS(skb)->uid, NULL, pid); - put_pid(pid); - if (rc) - printk(KERN_WARNING "Error processing HELO; rc = [%d]\n", rc); - return rc; -} - -/** - * ecryptfs_process_nl_quit - * @skb: The socket buffer containing the nlmsghdr in QUIT state - * - * Gets uid and pid of the skb and deletes the corresponding daemon - * id, if it is the registered that is requesting the - * deletion. Returns zero after deleting the desired daemon id; - * non-zero otherwise. - */ -static int ecryptfs_process_nl_quit(struct sk_buff *skb) -{ - struct pid *pid; - int rc; - - pid = find_get_pid(NETLINK_CREDS(skb)->pid); - rc = ecryptfs_process_quit(NETLINK_CREDS(skb)->uid, NULL, pid); - put_pid(pid); - if (rc) - printk(KERN_WARNING - "Error processing QUIT message; rc = [%d]\n", rc); - return rc; -} - -/** - * ecryptfs_receive_nl_message - * - * Callback function called by netlink system when a message arrives. - * If the message looks to be valid, then an attempt is made to assign - * it to its desired netlink context element and wake up the process - * that is waiting for a response. - */ -static void ecryptfs_receive_nl_message(struct sk_buff *skb) -{ - struct nlmsghdr *nlh; - - nlh = nlmsg_hdr(skb); - if (!NLMSG_OK(nlh, skb->len)) { - ecryptfs_printk(KERN_ERR, "Received corrupt netlink " - "message\n"); - goto free; - } - switch (nlh->nlmsg_type) { - case ECRYPTFS_MSG_RESPONSE: - if (ecryptfs_process_nl_response(skb)) { - ecryptfs_printk(KERN_WARNING, "Failed to " - "deliver netlink response to " - "requesting operation\n"); - } - break; - case ECRYPTFS_MSG_HELO: - if (ecryptfs_process_nl_helo(skb)) { - ecryptfs_printk(KERN_WARNING, "Failed to " - "fulfill HELO request\n"); - } - break; - case ECRYPTFS_MSG_QUIT: - if (ecryptfs_process_nl_quit(skb)) { - ecryptfs_printk(KERN_WARNING, "Failed to " - "fulfill QUIT request\n"); - } - break; - default: - ecryptfs_printk(KERN_WARNING, "Dropping netlink " - "message of unrecognized type [%d]\n", - nlh->nlmsg_type); - break; - } -free: - kfree_skb(skb); -} - -/** - * ecryptfs_init_netlink - * - * Initializes the daemon id hash list, netlink context array, and - * necessary locks. Returns zero upon success; non-zero upon error. - */ -int ecryptfs_init_netlink(void) -{ - int rc; - - ecryptfs_nl_sock = netlink_kernel_create(&init_net, NETLINK_ECRYPTFS, 0, - ecryptfs_receive_nl_message, - NULL, THIS_MODULE); - if (!ecryptfs_nl_sock) { - rc = -EIO; - ecryptfs_printk(KERN_ERR, "Failed to create netlink socket\n"); - goto out; - } - ecryptfs_nl_sock->sk_sndtimeo = ECRYPTFS_DEFAULT_SEND_TIMEOUT; - rc = 0; -out: - return rc; -} - -/** - * ecryptfs_release_netlink - * - * Frees all memory used by the netlink context array and releases the - * netlink socket. - */ -void ecryptfs_release_netlink(void) -{ - netlink_kernel_release(ecryptfs_nl_sock); - ecryptfs_nl_sock = NULL; -} diff --git a/fs/efs/namei.c b/fs/efs/namei.c index 291abb11e20..c3fb5f9c4a4 100644 --- a/fs/efs/namei.c +++ b/fs/efs/namei.c @@ -112,35 +112,14 @@ struct dentry *efs_fh_to_parent(struct super_block *sb, struct fid *fid, struct dentry *efs_get_parent(struct dentry *child) { - struct dentry *parent; - struct inode *inode; + struct dentry *parent = ERR_PTR(-ENOENT); efs_ino_t ino; - long error; lock_kernel(); - - error = -ENOENT; ino = efs_find_entry(child->d_inode, "..", 2); - if (!ino) - goto fail; - - inode = efs_iget(child->d_inode->i_sb, ino); - if (IS_ERR(inode)) { - error = PTR_ERR(inode); - goto fail; - } - - error = -ENOMEM; - parent = d_alloc_anon(inode); - if (!parent) - goto fail_iput; - + if (ino) + parent = d_obtain_alias(efs_iget(child->d_inode->i_sb, ino)); unlock_kernel(); - return parent; - fail_iput: - iput(inode); - fail: - unlock_kernel(); - return ERR_PTR(error); + return parent; } diff --git a/fs/eventpoll.c b/fs/eventpoll.c index 7cc0eb756b5..96355d50534 100644 --- a/fs/eventpoll.c +++ b/fs/eventpoll.c @@ -102,6 +102,8 @@ #define EP_UNACTIVE_PTR ((void *) -1L) +#define EP_ITEM_COST (sizeof(struct epitem) + sizeof(struct eppoll_entry)) + struct epoll_filefd { struct file *file; int fd; @@ -200,6 +202,9 @@ struct eventpoll { * holding ->lock. */ struct epitem *ovflist; + + /* The user that created the eventpoll descriptor */ + struct user_struct *user; }; /* Wait structure used by the poll hooks */ @@ -227,9 +232,17 @@ struct ep_pqueue { }; /* + * Configuration options available inside /proc/sys/fs/epoll/ + */ +/* Maximum number of epoll devices, per user */ +static int max_user_instances __read_mostly; +/* Maximum number of epoll watched descriptors, per user */ +static int max_user_watches __read_mostly; + +/* * This mutex is used to serialize ep_free() and eventpoll_release_file(). */ -static struct mutex epmutex; +static DEFINE_MUTEX(epmutex); /* Safe wake up implementation */ static struct poll_safewake psw; @@ -240,6 +253,33 @@ static struct kmem_cache *epi_cache __read_mostly; /* Slab cache used to allocate "struct eppoll_entry" */ static struct kmem_cache *pwq_cache __read_mostly; +#ifdef CONFIG_SYSCTL + +#include <linux/sysctl.h> + +static int zero; + +ctl_table epoll_table[] = { + { + .procname = "max_user_instances", + .data = &max_user_instances, + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = &proc_dointvec_minmax, + .extra1 = &zero, + }, + { + .procname = "max_user_watches", + .data = &max_user_watches, + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = &proc_dointvec_minmax, + .extra1 = &zero, + }, + { .ctl_name = 0 } +}; +#endif /* CONFIG_SYSCTL */ + /* Setup the structure that is used as key for the RB tree */ static inline void ep_set_ffd(struct epoll_filefd *ffd, @@ -402,6 +442,8 @@ static int ep_remove(struct eventpoll *ep, struct epitem *epi) /* At this point it is safe to free the eventpoll item */ kmem_cache_free(epi_cache, epi); + atomic_dec(&ep->user->epoll_watches); + DNPRINTK(3, (KERN_INFO "[%p] eventpoll: ep_remove(%p, %p)\n", current, ep, file)); @@ -449,6 +491,8 @@ static void ep_free(struct eventpoll *ep) mutex_unlock(&epmutex); mutex_destroy(&ep->mtx); + atomic_dec(&ep->user->epoll_devs); + free_uid(ep->user); kfree(ep); } @@ -532,10 +576,19 @@ void eventpoll_release_file(struct file *file) static int ep_alloc(struct eventpoll **pep) { - struct eventpoll *ep = kzalloc(sizeof(*ep), GFP_KERNEL); + int error; + struct user_struct *user; + struct eventpoll *ep; - if (!ep) - return -ENOMEM; + user = get_current_user(); + error = -EMFILE; + if (unlikely(atomic_read(&user->epoll_devs) >= + max_user_instances)) + goto free_uid; + error = -ENOMEM; + ep = kzalloc(sizeof(*ep), GFP_KERNEL); + if (unlikely(!ep)) + goto free_uid; spin_lock_init(&ep->lock); mutex_init(&ep->mtx); @@ -544,12 +597,17 @@ static int ep_alloc(struct eventpoll **pep) INIT_LIST_HEAD(&ep->rdllist); ep->rbr = RB_ROOT; ep->ovflist = EP_UNACTIVE_PTR; + ep->user = user; *pep = ep; DNPRINTK(3, (KERN_INFO "[%p] eventpoll: ep_alloc() ep=%p\n", current, ep)); return 0; + +free_uid: + free_uid(user); + return error; } /* @@ -703,9 +761,11 @@ static int ep_insert(struct eventpoll *ep, struct epoll_event *event, struct epitem *epi; struct ep_pqueue epq; - error = -ENOMEM; + if (unlikely(atomic_read(&ep->user->epoll_watches) >= + max_user_watches)) + return -ENOSPC; if (!(epi = kmem_cache_alloc(epi_cache, GFP_KERNEL))) - goto error_return; + return -ENOMEM; /* Item initialization follow here ... */ INIT_LIST_HEAD(&epi->rdllink); @@ -735,6 +795,7 @@ static int ep_insert(struct eventpoll *ep, struct epoll_event *event, * install process. Namely an allocation for a wait queue failed due * high memory pressure. */ + error = -ENOMEM; if (epi->nwait < 0) goto error_unregister; @@ -765,6 +826,8 @@ static int ep_insert(struct eventpoll *ep, struct epoll_event *event, spin_unlock_irqrestore(&ep->lock, flags); + atomic_inc(&ep->user->epoll_watches); + /* We have to call this outside the lock */ if (pwake) ep_poll_safewake(&psw, &ep->poll_wait); @@ -789,7 +852,7 @@ error_unregister: spin_unlock_irqrestore(&ep->lock, flags); kmem_cache_free(epi_cache, epi); -error_return: + return error; } @@ -927,12 +990,16 @@ errxit: /* * During the time we spent in the loop above, some other events * might have been queued by the poll callback. We re-insert them - * here (in case they are not already queued, or they're one-shot). + * inside the main ready-list here. */ for (nepi = ep->ovflist; (epi = nepi) != NULL; nepi = epi->next, epi->next = EP_UNACTIVE_PTR) { - if (!ep_is_linked(&epi->rdllink) && - (epi->event.events & ~EP_PRIVATE_BITS)) + /* + * If the above loop quit with errors, the epoll item might still + * be linked to "txlist", and the list_splice() done below will + * take care of those cases. + */ + if (!ep_is_linked(&epi->rdllink)) list_add_tail(&epi->rdllink, &ep->rdllist); } /* @@ -1074,6 +1141,7 @@ asmlinkage long sys_epoll_create1(int flags) flags & O_CLOEXEC); if (fd < 0) ep_free(ep); + atomic_inc(&ep->user->epoll_devs); error_return: DNPRINTK(3, (KERN_INFO "[%p] eventpoll: sys_epoll_create(%d) = %d\n", @@ -1295,7 +1363,12 @@ asmlinkage long sys_epoll_pwait(int epfd, struct epoll_event __user *events, static int __init eventpoll_init(void) { - mutex_init(&epmutex); + struct sysinfo si; + + si_meminfo(&si); + max_user_instances = 128; + max_user_watches = (((si.totalram - si.totalhigh) / 32) << PAGE_SHIFT) / + EP_ITEM_COST; /* Initialize the structure used to perform safe poll wait head wake ups */ ep_poll_safewake_init(&psw); diff --git a/fs/exec.c b/fs/exec.c index cecee501ce7..02d2e120542 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -50,14 +50,12 @@ #include <linux/cn_proc.h> #include <linux/audit.h> #include <linux/tracehook.h> +#include <linux/kmod.h> #include <asm/uaccess.h> #include <asm/mmu_context.h> #include <asm/tlb.h> - -#ifdef CONFIG_KMOD -#include <linux/kmod.h> -#endif +#include "internal.h" #ifdef __alpha__ /* for /sbin/loader handling in search_binary_handler() */ @@ -391,7 +389,7 @@ static int count(char __user * __user * argv, int max) if (!p) break; argv++; - if(++i > max) + if (i++ >= max) return -E2BIG; cond_resched(); } @@ -775,7 +773,6 @@ static int de_thread(struct task_struct *tsk) struct signal_struct *sig = tsk->signal; struct sighand_struct *oldsighand = tsk->sighand; spinlock_t *lock = &oldsighand->siglock; - struct task_struct *leader = NULL; int count; if (thread_group_empty(tsk)) @@ -813,7 +810,7 @@ static int de_thread(struct task_struct *tsk) * and to assume its PID: */ if (!thread_group_leader(tsk)) { - leader = tsk->group_leader; + struct task_struct *leader = tsk->group_leader; sig->notify_count = -1; /* for exit_notify() */ for (;;) { @@ -825,8 +822,6 @@ static int de_thread(struct task_struct *tsk) schedule(); } - if (unlikely(task_child_reaper(tsk) == leader)) - task_active_pid_ns(tsk)->child_reaper = tsk; /* * The only record we have of the real-time age of a * process, regardless of execs it's done, is start_time. @@ -867,8 +862,9 @@ static int de_thread(struct task_struct *tsk) BUG_ON(leader->exit_state != EXIT_ZOMBIE); leader->exit_state = EXIT_DEAD; - write_unlock_irq(&tasklist_lock); + + release_task(leader); } sig->group_exit_task = NULL; @@ -877,8 +873,6 @@ static int de_thread(struct task_struct *tsk) no_thread_group: exit_itimers(sig); flush_itimer_signals(); - if (leader) - release_task(leader); if (atomic_read(&oldsighand->count) != 1) { struct sighand_struct *newsighand; @@ -985,7 +979,7 @@ int flush_old_exec(struct linux_binprm * bprm) /* This is the point of no return */ current->sas_ss_sp = current->sas_ss_size = 0; - if (current->euid == current->uid && current->egid == current->gid) + if (current_euid() == current_uid() && current_egid() == current_gid()) set_dumpable(current->mm, 1); else set_dumpable(current->mm, suid_dumpable); @@ -1012,16 +1006,17 @@ int flush_old_exec(struct linux_binprm * bprm) */ current->mm->task_size = TASK_SIZE; - if (bprm->e_uid != current->euid || bprm->e_gid != current->egid) { - suid_keys(current); - set_dumpable(current->mm, suid_dumpable); + /* install the new credentials */ + if (bprm->cred->uid != current_euid() || + bprm->cred->gid != current_egid()) { current->pdeath_signal = 0; } else if (file_permission(bprm->file, MAY_READ) || - (bprm->interp_flags & BINPRM_FLAGS_ENFORCE_NONDUMP)) { - suid_keys(current); + bprm->interp_flags & BINPRM_FLAGS_ENFORCE_NONDUMP) { set_dumpable(current->mm, suid_dumpable); } + current->personality &= ~bprm->per_clear; + /* An exec changes our domain. We are no longer part of the thread group */ @@ -1038,13 +1033,50 @@ out: EXPORT_SYMBOL(flush_old_exec); +/* + * install the new credentials for this executable + */ +void install_exec_creds(struct linux_binprm *bprm) +{ + security_bprm_committing_creds(bprm); + + commit_creds(bprm->cred); + bprm->cred = NULL; + + /* cred_exec_mutex must be held at least to this point to prevent + * ptrace_attach() from altering our determination of the task's + * credentials; any time after this it may be unlocked */ + + security_bprm_committed_creds(bprm); +} +EXPORT_SYMBOL(install_exec_creds); + +/* + * determine how safe it is to execute the proposed program + * - the caller must hold current->cred_exec_mutex to protect against + * PTRACE_ATTACH + */ +void check_unsafe_exec(struct linux_binprm *bprm) +{ + struct task_struct *p = current; + + bprm->unsafe = tracehook_unsafe_exec(p); + + if (atomic_read(&p->fs->count) > 1 || + atomic_read(&p->files->count) > 1 || + atomic_read(&p->sighand->count) > 1) + bprm->unsafe |= LSM_UNSAFE_SHARE; +} + /* * Fill the binprm structure from the inode. * Check permissions, then read the first 128 (BINPRM_BUF_SIZE) bytes + * + * This may be called multiple times for binary chains (scripts for example). */ int prepare_binprm(struct linux_binprm *bprm) { - int mode; + umode_t mode; struct inode * inode = bprm->file->f_path.dentry->d_inode; int retval; @@ -1052,14 +1084,15 @@ int prepare_binprm(struct linux_binprm *bprm) if (bprm->file->f_op == NULL) return -EACCES; - bprm->e_uid = current->euid; - bprm->e_gid = current->egid; + /* clear any previous set[ug]id data from a previous binary */ + bprm->cred->euid = current_euid(); + bprm->cred->egid = current_egid(); - if(!(bprm->file->f_path.mnt->mnt_flags & MNT_NOSUID)) { + if (!(bprm->file->f_path.mnt->mnt_flags & MNT_NOSUID)) { /* Set-uid? */ if (mode & S_ISUID) { - current->personality &= ~PER_CLEAR_ON_SETID; - bprm->e_uid = inode->i_uid; + bprm->per_clear |= PER_CLEAR_ON_SETID; + bprm->cred->euid = inode->i_uid; } /* Set-gid? */ @@ -1069,52 +1102,23 @@ int prepare_binprm(struct linux_binprm *bprm) * executable. */ if ((mode & (S_ISGID | S_IXGRP)) == (S_ISGID | S_IXGRP)) { - current->personality &= ~PER_CLEAR_ON_SETID; - bprm->e_gid = inode->i_gid; + bprm->per_clear |= PER_CLEAR_ON_SETID; + bprm->cred->egid = inode->i_gid; } } /* fill in binprm security blob */ - retval = security_bprm_set(bprm); + retval = security_bprm_set_creds(bprm); if (retval) return retval; + bprm->cred_prepared = 1; - memset(bprm->buf,0,BINPRM_BUF_SIZE); - return kernel_read(bprm->file,0,bprm->buf,BINPRM_BUF_SIZE); + memset(bprm->buf, 0, BINPRM_BUF_SIZE); + return kernel_read(bprm->file, 0, bprm->buf, BINPRM_BUF_SIZE); } EXPORT_SYMBOL(prepare_binprm); -static int unsafe_exec(struct task_struct *p) -{ - int unsafe = tracehook_unsafe_exec(p); - - if (atomic_read(&p->fs->count) > 1 || - atomic_read(&p->files->count) > 1 || - atomic_read(&p->sighand->count) > 1) - unsafe |= LSM_UNSAFE_SHARE; - - return unsafe; -} - -void compute_creds(struct linux_binprm *bprm) -{ - int unsafe; - - if (bprm->e_uid != current->uid) { - suid_keys(current); - current->pdeath_signal = 0; - } - exec_keys(current); - - task_lock(current); - unsafe = unsafe_exec(current); - security_bprm_apply_creds(bprm, unsafe); - task_unlock(current); - security_bprm_post_apply_creds(bprm); -} -EXPORT_SYMBOL(compute_creds); - /* * Arguments are '\0' separated strings found at the location bprm->p * points to; chop off the first by relocating brpm->p to right after @@ -1164,6 +1168,7 @@ EXPORT_SYMBOL(remove_arg_zero); */ int search_binary_handler(struct linux_binprm *bprm,struct pt_regs *regs) { + unsigned int depth = bprm->recursion_depth; int try,retval; struct linux_binfmt *fmt; #ifdef __alpha__ @@ -1189,7 +1194,7 @@ int search_binary_handler(struct linux_binprm *bprm,struct pt_regs *regs) return retval; /* Remember if the application is TASO. */ - bprm->sh_bang = eh->ah.entry < 0x100000000UL; + bprm->taso = eh->ah.entry < 0x100000000UL; bprm->file = file; bprm->loader = loader; @@ -1224,8 +1229,15 @@ int search_binary_handler(struct linux_binprm *bprm,struct pt_regs *regs) continue; read_unlock(&binfmt_lock); retval = fn(bprm, regs); + /* + * Restore the depth counter to its starting value + * in this call, so we don't have to rely on every + * load_binary function to restore it on return. + */ + bprm->recursion_depth = depth; if (retval >= 0) { - tracehook_report_exec(fmt, bprm, regs); + if (depth == 0) + tracehook_report_exec(fmt, bprm, regs); put_binfmt(fmt); allow_write_access(bprm->file); if (bprm->file) @@ -1247,8 +1259,8 @@ int search_binary_handler(struct linux_binprm *bprm,struct pt_regs *regs) read_unlock(&binfmt_lock); if (retval != -ENOEXEC || bprm->mm == NULL) { break; -#ifdef CONFIG_KMOD - }else{ +#ifdef CONFIG_MODULES + } else { #define printable(c) (((c)=='\t') || ((c)=='\n') || (0x20<=(c) && (c)<=0x7e)) if (printable(bprm->buf[0]) && printable(bprm->buf[1]) && @@ -1267,6 +1279,8 @@ EXPORT_SYMBOL(search_binary_handler); void free_bprm(struct linux_binprm *bprm) { free_arg_pages(bprm); + if (bprm->cred) + abort_creds(bprm->cred); kfree(bprm); } @@ -1292,10 +1306,20 @@ int do_execve(char * filename, if (!bprm) goto out_files; + retval = mutex_lock_interruptible(¤t->cred_exec_mutex); + if (retval < 0) + goto out_free; + + retval = -ENOMEM; + bprm->cred = prepare_exec_creds(); + if (!bprm->cred) + goto out_unlock; + check_unsafe_exec(bprm); + file = open_exec(filename); retval = PTR_ERR(file); if (IS_ERR(file)) - goto out_kfree; + goto out_unlock; sched_exec(); @@ -1309,14 +1333,10 @@ int do_execve(char * filename, bprm->argc = count(argv, MAX_ARG_STRINGS); if ((retval = bprm->argc) < 0) - goto out_mm; + goto out; bprm->envc = count(envp, MAX_ARG_STRINGS); if ((retval = bprm->envc) < 0) - goto out_mm; - - retval = security_bprm_alloc(bprm); - if (retval) goto out; retval = prepare_binprm(bprm); @@ -1338,21 +1358,18 @@ int do_execve(char * filename, current->flags &= ~PF_KTHREAD; retval = search_binary_handler(bprm,regs); - if (retval >= 0) { - /* execve success */ - security_bprm_free(bprm); - acct_update_integrals(current); - free_bprm(bprm); - if (displaced) - put_files_struct(displaced); - return retval; - } + if (retval < 0) + goto out; -out: - if (bprm->security) - security_bprm_free(bprm); + /* execve succeeded */ + mutex_unlock(¤t->cred_exec_mutex); + acct_update_integrals(current); + free_bprm(bprm); + if (displaced) + put_files_struct(displaced); + return retval; -out_mm: +out: if (bprm->mm) mmput (bprm->mm); @@ -1361,7 +1378,11 @@ out_file: allow_write_access(bprm->file); fput(bprm->file); } -out_kfree: + +out_unlock: + mutex_unlock(¤t->cred_exec_mutex); + +out_free: free_bprm(bprm); out_files: @@ -1391,8 +1412,9 @@ EXPORT_SYMBOL(set_binfmt); * name into corename, which must have space for at least * CORENAME_MAX_SIZE bytes plus one byte for the zero terminator. */ -static int format_corename(char *corename, int nr_threads, long signr) +static int format_corename(char *corename, long signr) { + const struct cred *cred = current_cred(); const char *pat_ptr = core_pattern; int ispipe = (*pat_ptr == '|'); char *out_ptr = corename; @@ -1429,7 +1451,7 @@ static int format_corename(char *corename, int nr_threads, long signr) /* uid */ case 'u': rc = snprintf(out_ptr, out_end - out_ptr, - "%d", current->uid); + "%d", cred->uid); if (rc > out_end - out_ptr) goto out; out_ptr += rc; @@ -1437,7 +1459,7 @@ static int format_corename(char *corename, int nr_threads, long signr) /* gid */ case 'g': rc = snprintf(out_ptr, out_end - out_ptr, - "%d", current->gid); + "%d", cred->gid); if (rc > out_end - out_ptr) goto out; out_ptr += rc; @@ -1498,8 +1520,7 @@ static int format_corename(char *corename, int nr_threads, long signr) * If core_pattern does not include a %p (as is the default) * and core_uses_pid is set, then .%pid will be appended to * the filename. Do not do this for piped commands. */ - if (!ispipe && !pid_in_pattern - && (core_uses_pid || nr_threads)) { + if (!ispipe && !pid_in_pattern && core_uses_pid) { rc = snprintf(out_ptr, out_end - out_ptr, ".%d", task_tgid_vnr(current)); if (rc > out_end - out_ptr) @@ -1714,8 +1735,9 @@ int do_coredump(long signr, int exit_code, struct pt_regs * regs) struct linux_binfmt * binfmt; struct inode * inode; struct file * file; + const struct cred *old_cred; + struct cred *cred; int retval = 0; - int fsuid = current->fsuid; int flag = 0; int ispipe = 0; unsigned long core_limit = current->signal->rlim[RLIMIT_CORE].rlim_cur; @@ -1728,12 +1750,20 @@ int do_coredump(long signr, int exit_code, struct pt_regs * regs) binfmt = current->binfmt; if (!binfmt || !binfmt->core_dump) goto fail; + + cred = prepare_creds(); + if (!cred) { + retval = -ENOMEM; + goto fail; + } + down_write(&mm->mmap_sem); /* * If another thread got here first, or we are not dumpable, bail out. */ if (mm->core_state || !get_dumpable(mm)) { up_write(&mm->mmap_sem); + put_cred(cred); goto fail; } @@ -1744,12 +1774,16 @@ int do_coredump(long signr, int exit_code, struct pt_regs * regs) */ if (get_dumpable(mm) == 2) { /* Setuid core dump mode */ flag = O_EXCL; /* Stop rewrite attacks */ - current->fsuid = 0; /* Dump root private */ + cred->fsuid = 0; /* Dump root private */ } retval = coredump_wait(exit_code, &core_state); - if (retval < 0) + if (retval < 0) { + put_cred(cred); goto fail; + } + + old_cred = override_creds(cred); /* * Clear any false indication of pending signals that might @@ -1762,7 +1796,7 @@ int do_coredump(long signr, int exit_code, struct pt_regs * regs) * uses lock_kernel() */ lock_kernel(); - ispipe = format_corename(corename, retval, signr); + ispipe = format_corename(corename, signr); unlock_kernel(); /* * Don't bother to check the RLIMIT_CORE value if core_pattern points @@ -1821,7 +1855,7 @@ int do_coredump(long signr, int exit_code, struct pt_regs * regs) * Dont allow local users get cute and trick others to coredump * into their pre-created files: */ - if (inode->i_uid != current->fsuid) + if (inode->i_uid != current_fsuid()) goto close_fail; if (!file->f_op) goto close_fail; @@ -1840,7 +1874,8 @@ fail_unlock: if (helper_argv) argv_free(helper_argv); - current->fsuid = fsuid; + revert_creds(old_cred); + put_cred(cred); coredump_finish(mm); fail: return retval; diff --git a/fs/exportfs/expfs.c b/fs/exportfs/expfs.c index cc91227d3bb..197c7db583c 100644 --- a/fs/exportfs/expfs.c +++ b/fs/exportfs/expfs.c @@ -14,6 +14,7 @@ #include <linux/module.h> #include <linux/mount.h> #include <linux/namei.h> +#include <linux/sched.h> #define dprintk(fmt, args...) do{}while(0) @@ -94,9 +95,8 @@ find_disconnected_root(struct dentry *dentry) * It may already be, as the flag isn't always updated when connection happens. */ static int -reconnect_path(struct vfsmount *mnt, struct dentry *target_dir) +reconnect_path(struct vfsmount *mnt, struct dentry *target_dir, char *nbuf) { - char nbuf[NAME_MAX+1]; int noprogress = 0; int err = -ESTALE; @@ -250,6 +250,7 @@ static int filldir_one(void * __buf, const char * name, int len, static int get_name(struct vfsmount *mnt, struct dentry *dentry, char *name, struct dentry *child) { + const struct cred *cred = current_cred(); struct inode *dir = dentry->d_inode; int error; struct file *file; @@ -264,7 +265,7 @@ static int get_name(struct vfsmount *mnt, struct dentry *dentry, /* * Open the directory ... */ - file = dentry_open(dget(dentry), mntget(mnt), O_RDONLY); + file = dentry_open(dget(dentry), mntget(mnt), O_RDONLY, cred); error = PTR_ERR(file); if (IS_ERR(file)) goto out; @@ -281,13 +282,14 @@ static int get_name(struct vfsmount *mnt, struct dentry *dentry, int old_seq = buffer.sequence; error = vfs_readdir(file, filldir_one, &buffer); + if (buffer.found) { + error = 0; + break; + } if (error < 0) break; - error = 0; - if (buffer.found) - break; error = -ENOENT; if (old_seq == buffer.sequence) break; @@ -360,6 +362,7 @@ struct dentry *exportfs_decode_fh(struct vfsmount *mnt, struct fid *fid, { const struct export_operations *nop = mnt->mnt_sb->s_export_op; struct dentry *result, *alias; + char nbuf[NAME_MAX+1]; int err; /* @@ -381,7 +384,7 @@ struct dentry *exportfs_decode_fh(struct vfsmount *mnt, struct fid *fid, * filesystem root. */ if (result->d_flags & DCACHE_DISCONNECTED) { - err = reconnect_path(mnt, result); + err = reconnect_path(mnt, result, nbuf); if (err) goto err_result; } @@ -397,7 +400,6 @@ struct dentry *exportfs_decode_fh(struct vfsmount *mnt, struct fid *fid, * It's not a directory. Life is a little more complicated. */ struct dentry *target_dir, *nresult; - char nbuf[NAME_MAX+1]; /* * See if either the dentry we just got from the filesystem @@ -433,7 +435,7 @@ struct dentry *exportfs_decode_fh(struct vfsmount *mnt, struct fid *fid, * connected to the filesystem root. The VFS really doesn't * like disconnected directories.. */ - err = reconnect_path(mnt, target_dir); + err = reconnect_path(mnt, target_dir, nbuf); if (err) { dput(target_dir); goto err_result; diff --git a/fs/ext2/Kconfig b/fs/ext2/Kconfig new file mode 100644 index 00000000000..14a6780fd03 --- /dev/null +++ b/fs/ext2/Kconfig @@ -0,0 +1,55 @@ +config EXT2_FS + tristate "Second extended fs support" + help + Ext2 is a standard Linux file system for hard disks. + + To compile this file system support as a module, choose M here: the + module will be called ext2. + + If unsure, say Y. + +config EXT2_FS_XATTR + bool "Ext2 extended attributes" + depends on EXT2_FS + help + Extended attributes are name:value pairs associated with inodes by + the kernel or by users (see the attr(5) manual page, or visit + <http://acl.bestbits.at/> for details). + + If unsure, say N. + +config EXT2_FS_POSIX_ACL + bool "Ext2 POSIX Access Control Lists" + depends on EXT2_FS_XATTR + select FS_POSIX_ACL + help + Posix Access Control Lists (ACLs) support permissions for users and + groups beyond the owner/group/world scheme. + + To learn more about Access Control Lists, visit the Posix ACLs for + Linux website <http://acl.bestbits.at/>. + + If you don't know what Access Control Lists are, say N + +config EXT2_FS_SECURITY + bool "Ext2 Security Labels" + depends on EXT2_FS_XATTR + help + Security labels support alternative access control models + implemented by security modules like SELinux. This option + enables an extended attribute handler for file security + labels in the ext2 filesystem. + + If you are not using a security module that requires using + extended attributes for file security labels, say N. + +config EXT2_FS_XIP + bool "Ext2 execute in place support" + depends on EXT2_FS && MMU + help + Execute in place can be used on memory-backed block devices. If you + enable this option, you can select to mount block devices which are + capable of this feature without using the page cache. + + If you do not use a block device that is capable of using this, + or if unsure, say N. diff --git a/fs/ext2/balloc.c b/fs/ext2/balloc.c index 10bb02c3f25..4a29d637608 100644 --- a/fs/ext2/balloc.c +++ b/fs/ext2/balloc.c @@ -1193,7 +1193,7 @@ static int ext2_has_free_blocks(struct ext2_sb_info *sbi) free_blocks = percpu_counter_read_positive(&sbi->s_freeblocks_counter); root_blocks = le32_to_cpu(sbi->s_es->s_r_blocks_count); if (free_blocks < root_blocks + 1 && !capable(CAP_SYS_RESOURCE) && - sbi->s_resuid != current->fsuid && + sbi->s_resuid != current_fsuid() && (sbi->s_resgid == 0 || !in_group_p (sbi->s_resgid))) { return 0; } @@ -1295,6 +1295,7 @@ retry_alloc: * turn off reservation for this allocation */ if (my_rsv && (free_blocks < windowsz) + && (free_blocks > 0) && (rsv_is_empty(&my_rsv->rsv_window))) my_rsv = NULL; @@ -1332,7 +1333,7 @@ retry_alloc: * free blocks is less than half of the reservation * window size. */ - if (free_blocks <= (windowsz/2)) + if (my_rsv && (free_blocks <= (windowsz/2))) continue; brelse(bitmap_bh); diff --git a/fs/ext2/dir.c b/fs/ext2/dir.c index a78c6b4af06..9a0fc400f91 100644 --- a/fs/ext2/dir.c +++ b/fs/ext2/dir.c @@ -103,7 +103,7 @@ static int ext2_commit_chunk(struct page *page, loff_t pos, unsigned len) return err; } -static void ext2_check_page(struct page *page) +static void ext2_check_page(struct page *page, int quiet) { struct inode *dir = page->mapping->host; struct super_block *sb = dir->i_sb; @@ -146,10 +146,10 @@ out: /* Too bad, we had an error */ Ebadsize: - ext2_error(sb, "ext2_check_page", - "size of directory #%lu is not a multiple of chunk size", - dir->i_ino - ); + if (!quiet) + ext2_error(sb, __func__, + "size of directory #%lu is not a multiple " + "of chunk size", dir->i_ino); goto fail; Eshort: error = "rec_len is smaller than minimal"; @@ -166,32 +166,36 @@ Espan: Einumber: error = "inode out of bounds"; bad_entry: - ext2_error (sb, "ext2_check_page", "bad entry in directory #%lu: %s - " - "offset=%lu, inode=%lu, rec_len=%d, name_len=%d", - dir->i_ino, error, (page->index<<PAGE_CACHE_SHIFT)+offs, - (unsigned long) le32_to_cpu(p->inode), - rec_len, p->name_len); + if (!quiet) + ext2_error(sb, __func__, "bad entry in directory #%lu: : %s - " + "offset=%lu, inode=%lu, rec_len=%d, name_len=%d", + dir->i_ino, error, (page->index<<PAGE_CACHE_SHIFT)+offs, + (unsigned long) le32_to_cpu(p->inode), + rec_len, p->name_len); goto fail; Eend: - p = (ext2_dirent *)(kaddr + offs); - ext2_error (sb, "ext2_check_page", - "entry in directory #%lu spans the page boundary" - "offset=%lu, inode=%lu", - dir->i_ino, (page->index<<PAGE_CACHE_SHIFT)+offs, - (unsigned long) le32_to_cpu(p->inode)); + if (!quiet) { + p = (ext2_dirent *)(kaddr + offs); + ext2_error(sb, "ext2_check_page", + "entry in directory #%lu spans the page boundary" + "offset=%lu, inode=%lu", + dir->i_ino, (page->index<<PAGE_CACHE_SHIFT)+offs, + (unsigned long) le32_to_cpu(p->inode)); + } fail: SetPageChecked(page); SetPageError(page); } -static struct page * ext2_get_page(struct inode *dir, unsigned long n) +static struct page * ext2_get_page(struct inode *dir, unsigned long n, + int quiet) { struct address_space *mapping = dir->i_mapping; struct page *page = read_mapping_page(mapping, n, NULL); if (!IS_ERR(page)) { kmap(page); if (!PageChecked(page)) - ext2_check_page(page); + ext2_check_page(page, quiet); if (PageError(page)) goto fail; } @@ -292,7 +296,7 @@ ext2_readdir (struct file * filp, void * dirent, filldir_t filldir) for ( ; n < npages; n++, offset = 0) { char *kaddr, *limit; ext2_dirent *de; - struct page *page = ext2_get_page(inode, n); + struct page *page = ext2_get_page(inode, n, 0); if (IS_ERR(page)) { ext2_error(sb, __func__, @@ -350,17 +354,18 @@ ext2_readdir (struct file * filp, void * dirent, filldir_t filldir) * (as a parameter - res_dir). Page is returned mapped and unlocked. * Entry is guaranteed to be valid. */ -struct ext2_dir_entry_2 * ext2_find_entry (struct inode * dir, - struct dentry *dentry, struct page ** res_page) +struct ext2_dir_entry_2 *ext2_find_entry (struct inode * dir, + struct qstr *child, struct page ** res_page) { - const char *name = dentry->d_name.name; - int namelen = dentry->d_name.len; + const char *name = child->name; + int namelen = child->len; unsigned reclen = EXT2_DIR_REC_LEN(namelen); unsigned long start, n; unsigned long npages = dir_pages(dir); struct page *page = NULL; struct ext2_inode_info *ei = EXT2_I(dir); ext2_dirent * de; + int dir_has_error = 0; if (npages == 0) goto out; @@ -374,7 +379,7 @@ struct ext2_dir_entry_2 * ext2_find_entry (struct inode * dir, n = start; do { char *kaddr; - page = ext2_get_page(dir, n); + page = ext2_get_page(dir, n, dir_has_error); if (!IS_ERR(page)) { kaddr = page_address(page); de = (ext2_dirent *) kaddr; @@ -391,7 +396,9 @@ struct ext2_dir_entry_2 * ext2_find_entry (struct inode * dir, de = ext2_next_entry(de); } ext2_put_page(page); - } + } else + dir_has_error = 1; + if (++n >= npages) n = 0; /* next page is past the blocks we've got */ @@ -414,7 +421,7 @@ found: struct ext2_dir_entry_2 * ext2_dotdot (struct inode *dir, struct page **p) { - struct page *page = ext2_get_page(dir, 0); + struct page *page = ext2_get_page(dir, 0, 0); ext2_dirent *de = NULL; if (!IS_ERR(page)) { @@ -424,13 +431,13 @@ struct ext2_dir_entry_2 * ext2_dotdot (struct inode *dir, struct page **p) return de; } -ino_t ext2_inode_by_name(struct inode * dir, struct dentry *dentry) +ino_t ext2_inode_by_name(struct inode *dir, struct qstr *child) { ino_t res = 0; - struct ext2_dir_entry_2 * de; + struct ext2_dir_entry_2 *de; struct page *page; - de = ext2_find_entry (dir, dentry, &page); + de = ext2_find_entry (dir, child, &page); if (de) { res = le32_to_cpu(de->inode); ext2_put_page(page); @@ -487,7 +494,7 @@ int ext2_add_link (struct dentry *dentry, struct inode *inode) for (n = 0; n <= npages; n++) { char *dir_end; - page = ext2_get_page(dir, n); + page = ext2_get_page(dir, n, 0); err = PTR_ERR(page); if (IS_ERR(page)) goto out; @@ -655,14 +662,17 @@ int ext2_empty_dir (struct inode * inode) { struct page *page = NULL; unsigned long i, npages = dir_pages(inode); + int dir_has_error = 0; for (i = 0; i < npages; i++) { char *kaddr; ext2_dirent * de; - page = ext2_get_page(inode, i); + page = ext2_get_page(inode, i, dir_has_error); - if (IS_ERR(page)) + if (IS_ERR(page)) { + dir_has_error = 1; continue; + } kaddr = page_address(page); de = (ext2_dirent *)kaddr; diff --git a/fs/ext2/ext2.h b/fs/ext2/ext2.h index bae998c1e44..3203042b36e 100644 --- a/fs/ext2/ext2.h +++ b/fs/ext2/ext2.h @@ -105,9 +105,9 @@ extern void ext2_rsv_window_add(struct super_block *sb, struct ext2_reserve_wind /* dir.c */ extern int ext2_add_link (struct dentry *, struct inode *); -extern ino_t ext2_inode_by_name(struct inode *, struct dentry *); +extern ino_t ext2_inode_by_name(struct inode *, struct qstr *); extern int ext2_make_empty(struct inode *, struct inode *); -extern struct ext2_dir_entry_2 * ext2_find_entry (struct inode *,struct dentry *, struct page **); +extern struct ext2_dir_entry_2 * ext2_find_entry (struct inode *,struct qstr *, struct page **); extern int ext2_delete_entry (struct ext2_dir_entry_2 *, struct page *); extern int ext2_empty_dir (struct inode *); extern struct ext2_dir_entry_2 * ext2_dotdot (struct inode *, struct page **); diff --git a/fs/ext2/ialloc.c b/fs/ext2/ialloc.c index f5974134676..8d0add62587 100644 --- a/fs/ext2/ialloc.c +++ b/fs/ext2/ialloc.c @@ -550,7 +550,7 @@ got: sb->s_dirt = 1; mark_buffer_dirty(bh2); - inode->i_uid = current->fsuid; + inode->i_uid = current_fsuid(); if (test_opt (sb, GRPID)) inode->i_gid = dir->i_gid; else if (dir->i_mode & S_ISGID) { @@ -558,7 +558,7 @@ got: if (S_ISDIR(mode)) mode |= S_ISGID; } else - inode->i_gid = current->fsgid; + inode->i_gid = current_fsgid(); inode->i_mode = mode; inode->i_ino = ino; diff --git a/fs/ext2/namei.c b/fs/ext2/namei.c index 80c97fd8c57..2a747252ec1 100644 --- a/fs/ext2/namei.c +++ b/fs/ext2/namei.c @@ -60,7 +60,7 @@ static struct dentry *ext2_lookup(struct inode * dir, struct dentry *dentry, str if (dentry->d_name.len > EXT2_NAME_LEN) return ERR_PTR(-ENAMETOOLONG); - ino = ext2_inode_by_name(dir, dentry); + ino = ext2_inode_by_name(dir, &dentry->d_name); inode = NULL; if (ino) { inode = ext2_iget(dir->i_sb, ino); @@ -72,27 +72,11 @@ static struct dentry *ext2_lookup(struct inode * dir, struct dentry *dentry, str struct dentry *ext2_get_parent(struct dentry *child) { - unsigned long ino; - struct dentry *parent; - struct inode *inode; - struct dentry dotdot; - - dotdot.d_name.name = ".."; - dotdot.d_name.len = 2; - - ino = ext2_inode_by_name(child->d_inode, &dotdot); + struct qstr dotdot = {.name = "..", .len = 2}; + unsigned long ino = ext2_inode_by_name(child->d_inode, &dotdot); if (!ino) return ERR_PTR(-ENOENT); - inode = ext2_iget(child->d_inode->i_sb, ino); - - if (IS_ERR(inode)) - return ERR_CAST(inode); - parent = d_alloc_anon(inode); - if (!parent) { - iput(inode); - parent = ERR_PTR(-ENOMEM); - } - return parent; + return d_obtain_alias(ext2_iget(child->d_inode->i_sb, ino)); } /* @@ -257,7 +241,7 @@ static int ext2_unlink(struct inode * dir, struct dentry *dentry) struct page * page; int err = -ENOENT; - de = ext2_find_entry (dir, dentry, &page); + de = ext2_find_entry (dir, &dentry->d_name, &page); if (!de) goto out; @@ -299,7 +283,7 @@ static int ext2_rename (struct inode * old_dir, struct dentry * old_dentry, struct ext2_dir_entry_2 * old_de; int err = -ENOENT; - old_de = ext2_find_entry (old_dir, old_dentry, &old_page); + old_de = ext2_find_entry (old_dir, &old_dentry->d_name, &old_page); if (!old_de) goto out; @@ -319,7 +303,7 @@ static int ext2_rename (struct inode * old_dir, struct dentry * old_dentry, goto out_dir; err = -ENOENT; - new_de = ext2_find_entry (new_dir, new_dentry, &new_page); + new_de = ext2_find_entry (new_dir, &new_dentry->d_name, &new_page); if (!new_de) goto out_dir; inode_inc_link_count(old_inode); diff --git a/fs/ext2/xip.c b/fs/ext2/xip.c index 4fb94c20041..b72b8588422 100644 --- a/fs/ext2/xip.c +++ b/fs/ext2/xip.c @@ -11,6 +11,7 @@ #include <linux/buffer_head.h> #include <linux/ext2_fs_sb.h> #include <linux/ext2_fs.h> +#include <linux/blkdev.h> #include "ext2.h" #include "xip.h" diff --git a/fs/ext3/Kconfig b/fs/ext3/Kconfig new file mode 100644 index 00000000000..8e0cfe44b0f --- /dev/null +++ b/fs/ext3/Kconfig @@ -0,0 +1,67 @@ +config EXT3_FS + tristate "Ext3 journalling file system support" + select JBD + help + This is the journalling version of the Second extended file system + (often called ext3), the de facto standard Linux file system + (method to organize files on a storage device) for hard disks. + + The journalling code included in this driver means you do not have + to run e2fsck (file system checker) on your file systems after a + crash. The journal keeps track of any changes that were being made + at the time the system crashed, and can ensure that your file system + is consistent without the need for a lengthy check. + + Other than adding the journal to the file system, the on-disk format + of ext3 is identical to ext2. It is possible to freely switch + between using the ext3 driver and the ext2 driver, as long as the + file system has been cleanly unmounted, or e2fsck is run on the file + system. + + To add a journal on an existing ext2 file system or change the + behavior of ext3 file systems, you can use the tune2fs utility ("man + tune2fs"). To modify attributes of files and directories on ext3 + file systems, use chattr ("man chattr"). You need to be using + e2fsprogs version 1.20 or later in order to create ext3 journals + (available at <http://sourceforge.net/projects/e2fsprogs/>). + + To compile this file system support as a module, choose M here: the + module will be called ext3. + +config EXT3_FS_XATTR + bool "Ext3 extended attributes" + depends on EXT3_FS + default y + help + Extended attributes are name:value pairs associated with inodes by + the kernel or by users (see the attr(5) manual page, or visit + <http://acl.bestbits.at/> for details). + + If unsure, say N. + + You need this for POSIX ACL support on ext3. + +config EXT3_FS_POSIX_ACL + bool "Ext3 POSIX Access Control Lists" + depends on EXT3_FS_XATTR + select FS_POSIX_ACL + help + Posix Access Control Lists (ACLs) support permissions for users and + groups beyond the owner/group/world scheme. + + To learn more about Access Control Lists, visit the Posix ACLs for + Linux website <http://acl.bestbits.at/>. + + If you don't know what Access Control Lists are, say N + +config EXT3_FS_SECURITY + bool "Ext3 Security Labels" + depends on EXT3_FS_XATTR + help + Security labels support alternative access control models + implemented by security modules like SELinux. This option + enables an extended attribute handler for file security + labels in the ext3 filesystem. + + If you are not using a security module that requires using + extended attributes for file security labels, say N. diff --git a/fs/ext3/balloc.c b/fs/ext3/balloc.c index 92fd0338a6e..0dbf1c04847 100644 --- a/fs/ext3/balloc.c +++ b/fs/ext3/balloc.c @@ -1422,7 +1422,7 @@ static int ext3_has_free_blocks(struct ext3_sb_info *sbi) free_blocks = percpu_counter_read_positive(&sbi->s_freeblocks_counter); root_blocks = le32_to_cpu(sbi->s_es->s_r_blocks_count); if (free_blocks < root_blocks + 1 && !capable(CAP_SYS_RESOURCE) && - sbi->s_resuid != current->fsuid && + sbi->s_resuid != current_fsuid() && (sbi->s_resgid == 0 || !in_group_p (sbi->s_resgid))) { return 0; } @@ -1547,6 +1547,7 @@ retry_alloc: * turn off reservation for this allocation */ if (my_rsv && (free_blocks < windowsz) + && (free_blocks > 0) && (rsv_is_empty(&my_rsv->rsv_window))) my_rsv = NULL; @@ -1585,7 +1586,7 @@ retry_alloc: * free blocks is less than half of the reservation * window size. */ - if (free_blocks <= (windowsz/2)) + if (my_rsv && (free_blocks <= (windowsz/2))) continue; brelse(bitmap_bh); diff --git a/fs/ext3/dir.c b/fs/ext3/dir.c index 2eea96ec78e..5853f4440af 100644 --- a/fs/ext3/dir.c +++ b/fs/ext3/dir.c @@ -102,6 +102,7 @@ static int ext3_readdir(struct file * filp, int err; struct inode *inode = filp->f_path.dentry->d_inode; int ret = 0; + int dir_has_error = 0; sb = inode->i_sb; @@ -148,9 +149,12 @@ static int ext3_readdir(struct file * filp, * of recovering data when there's a bad sector */ if (!bh) { - ext3_error (sb, "ext3_readdir", - "directory #%lu contains a hole at offset %lu", - inode->i_ino, (unsigned long)filp->f_pos); + if (!dir_has_error) { + ext3_error(sb, __func__, "directory #%lu " + "contains a hole at offset %lld", + inode->i_ino, filp->f_pos); + dir_has_error = 1; + } /* corrupt size? Maybe no more blocks to read */ if (filp->f_pos > inode->i_blocks << 9) break; @@ -410,7 +414,7 @@ static int call_filldir(struct file * filp, void * dirent, get_dtype(sb, fname->file_type)); if (error) { filp->f_pos = curr_pos; - info->extra_fname = fname->next; + info->extra_fname = fname; return error; } fname = fname->next; @@ -449,11 +453,12 @@ static int ext3_dx_readdir(struct file * filp, * If there are any leftover names on the hash collision * chain, return them first. */ - if (info->extra_fname && - call_filldir(filp, dirent, filldir, info->extra_fname)) - goto finished; - - if (!info->curr_node) + if (info->extra_fname) { + if (call_filldir(filp, dirent, filldir, info->extra_fname)) + goto finished; + info->extra_fname = NULL; + goto next_node; + } else if (!info->curr_node) info->curr_node = rb_first(&info->root); while (1) { @@ -484,9 +489,14 @@ static int ext3_dx_readdir(struct file * filp, info->curr_minor_hash = fname->minor_hash; if (call_filldir(filp, dirent, filldir, fname)) break; - + next_node: info->curr_node = rb_next(info->curr_node); - if (!info->curr_node) { + if (info->curr_node) { + fname = rb_entry(info->curr_node, struct fname, + rb_hash); + info->curr_hash = fname->hash; + info->curr_minor_hash = fname->minor_hash; + } else { if (info->next_hash == ~0) { filp->f_pos = EXT3_HTREE_EOF; break; diff --git a/fs/ext3/ialloc.c b/fs/ext3/ialloc.c index 47b678d73e7..490bd0ed789 100644 --- a/fs/ext3/ialloc.c +++ b/fs/ext3/ialloc.c @@ -539,7 +539,7 @@ got: percpu_counter_inc(&sbi->s_dirs_counter); sb->s_dirt = 1; - inode->i_uid = current->fsuid; + inode->i_uid = current_fsuid(); if (test_opt (sb, GRPID)) inode->i_gid = dir->i_gid; else if (dir->i_mode & S_ISGID) { @@ -547,7 +547,7 @@ got: if (S_ISDIR(mode)) mode |= S_ISGID; } else - inode->i_gid = current->fsgid; + inode->i_gid = current_fsgid(); inode->i_mode = mode; inode->i_ino = ino; diff --git a/fs/ext3/inode.c b/fs/ext3/inode.c index ebfec4d0148..f8424ad8997 100644 --- a/fs/ext3/inode.c +++ b/fs/ext3/inode.c @@ -1186,6 +1186,13 @@ write_begin_failed: ext3_journal_stop(handle); unlock_page(page); page_cache_release(page); + /* + * block_write_begin may have instantiated a few blocks + * outside i_size. Trim these off again. Don't need + * i_size_read because we hold i_mutex. + */ + if (pos + len > inode->i_size) + vmtruncate(inode, inode->i_size); } if (ret == -ENOSPC && ext3_should_retry_alloc(inode->i_sb, &retries)) goto retry; diff --git a/fs/ext3/ioctl.c b/fs/ext3/ioctl.c index 0d0c7015164..b7394d05ee8 100644 --- a/fs/ext3/ioctl.c +++ b/fs/ext3/ioctl.c @@ -239,7 +239,7 @@ setrsvsz_out: case EXT3_IOC_GROUP_EXTEND: { ext3_fsblk_t n_blocks_count; struct super_block *sb = inode->i_sb; - int err; + int err, err2; if (!capable(CAP_SYS_RESOURCE)) return -EPERM; @@ -254,8 +254,10 @@ setrsvsz_out: } err = ext3_group_extend(sb, EXT3_SB(sb)->s_es, n_blocks_count); journal_lock_updates(EXT3_SB(sb)->s_journal); - journal_flush(EXT3_SB(sb)->s_journal); + err2 = journal_flush(EXT3_SB(sb)->s_journal); journal_unlock_updates(EXT3_SB(sb)->s_journal); + if (err == 0) + err = err2; group_extend_out: mnt_drop_write(filp->f_path.mnt); return err; @@ -263,7 +265,7 @@ group_extend_out: case EXT3_IOC_GROUP_ADD: { struct ext3_new_group_data input; struct super_block *sb = inode->i_sb; - int err; + int err, err2; if (!capable(CAP_SYS_RESOURCE)) return -EPERM; @@ -280,8 +282,10 @@ group_extend_out: err = ext3_group_add(sb, &input); journal_lock_updates(EXT3_SB(sb)->s_journal); - journal_flush(EXT3_SB(sb)->s_journal); + err2 = journal_flush(EXT3_SB(sb)->s_journal); journal_unlock_updates(EXT3_SB(sb)->s_journal); + if (err == 0) + err = err2; group_add_out: mnt_drop_write(filp->f_path.mnt); return err; diff --git a/fs/ext3/namei.c b/fs/ext3/namei.c index de13e919cd8..3e5edc92aa0 100644 --- a/fs/ext3/namei.c +++ b/fs/ext3/namei.c @@ -159,7 +159,7 @@ static void dx_set_count (struct dx_entry *entries, unsigned value); static void dx_set_limit (struct dx_entry *entries, unsigned value); static unsigned dx_root_limit (struct inode *dir, unsigned infosize); static unsigned dx_node_limit (struct inode *dir); -static struct dx_frame *dx_probe(struct dentry *dentry, +static struct dx_frame *dx_probe(struct qstr *entry, struct inode *dir, struct dx_hash_info *hinfo, struct dx_frame *frame, @@ -176,8 +176,9 @@ static int ext3_htree_next_block(struct inode *dir, __u32 hash, struct dx_frame *frame, struct dx_frame *frames, __u32 *start_hash); -static struct buffer_head * ext3_dx_find_entry(struct dentry *dentry, - struct ext3_dir_entry_2 **res_dir, int *err); +static struct buffer_head * ext3_dx_find_entry(struct inode *dir, + struct qstr *entry, struct ext3_dir_entry_2 **res_dir, + int *err); static int ext3_dx_add_entry(handle_t *handle, struct dentry *dentry, struct inode *inode); @@ -342,7 +343,7 @@ struct stats dx_show_entries(struct dx_hash_info *hinfo, struct inode *dir, * back to userspace. */ static struct dx_frame * -dx_probe(struct dentry *dentry, struct inode *dir, +dx_probe(struct qstr *entry, struct inode *dir, struct dx_hash_info *hinfo, struct dx_frame *frame_in, int *err) { unsigned count, indirect; @@ -353,8 +354,6 @@ dx_probe(struct dentry *dentry, struct inode *dir, u32 hash; frame->bh = NULL; - if (dentry) - dir = dentry->d_parent->d_inode; if (!(bh = ext3_bread (NULL,dir, 0, 0, err))) goto fail; root = (struct dx_root *) bh->b_data; @@ -370,8 +369,8 @@ dx_probe(struct dentry *dentry, struct inode *dir, } hinfo->hash_version = root->info.hash_version; hinfo->seed = EXT3_SB(dir->i_sb)->s_hash_seed; - if (dentry) - ext3fs_dirhash(dentry->d_name.name, dentry->d_name.len, hinfo); + if (entry) + ext3fs_dirhash(entry->name, entry->len, hinfo); hash = hinfo->hash; if (root->info.unused_flags & 1) { @@ -803,15 +802,15 @@ static inline int ext3_match (int len, const char * const name, */ static inline int search_dirblock(struct buffer_head * bh, struct inode *dir, - struct dentry *dentry, + struct qstr *child, unsigned long offset, struct ext3_dir_entry_2 ** res_dir) { struct ext3_dir_entry_2 * de; char * dlimit; int de_len; - const char *name = dentry->d_name.name; - int namelen = dentry->d_name.len; + const char *name = child->name; + int namelen = child->len; de = (struct ext3_dir_entry_2 *) bh->b_data; dlimit = bh->b_data + dir->i_sb->s_blocksize; @@ -850,8 +849,9 @@ static inline int search_dirblock(struct buffer_head * bh, * The returned buffer_head has ->b_count elevated. The caller is expected * to brelse() it when appropriate. */ -static struct buffer_head * ext3_find_entry (struct dentry *dentry, - struct ext3_dir_entry_2 ** res_dir) +static struct buffer_head *ext3_find_entry(struct inode *dir, + struct qstr *entry, + struct ext3_dir_entry_2 **res_dir) { struct super_block * sb; struct buffer_head * bh_use[NAMEI_RA_SIZE]; @@ -863,16 +863,15 @@ static struct buffer_head * ext3_find_entry (struct dentry *dentry, buffer */ int num = 0; int nblocks, i, err; - struct inode *dir = dentry->d_parent->d_inode; int namelen; *res_dir = NULL; sb = dir->i_sb; - namelen = dentry->d_name.len; + namelen = entry->len; if (namelen > EXT3_NAME_LEN) return NULL; if (is_dx(dir)) { - bh = ext3_dx_find_entry(dentry, res_dir, &err); + bh = ext3_dx_find_entry(dir, entry, res_dir, &err); /* * On success, or if the error was file not found, * return. Otherwise, fall back to doing a search the @@ -923,7 +922,7 @@ restart: brelse(bh); goto next; } - i = search_dirblock(bh, dir, dentry, + i = search_dirblock(bh, dir, entry, block << EXT3_BLOCK_SIZE_BITS(sb), res_dir); if (i == 1) { EXT3_I(dir)->i_dir_start_lookup = block; @@ -957,8 +956,9 @@ cleanup_and_exit: return ret; } -static struct buffer_head * ext3_dx_find_entry(struct dentry *dentry, - struct ext3_dir_entry_2 **res_dir, int *err) +static struct buffer_head * ext3_dx_find_entry(struct inode *dir, + struct qstr *entry, struct ext3_dir_entry_2 **res_dir, + int *err) { struct super_block * sb; struct dx_hash_info hinfo; @@ -968,14 +968,13 @@ static struct buffer_head * ext3_dx_find_entry(struct dentry *dentry, struct buffer_head *bh; unsigned long block; int retval; - int namelen = dentry->d_name.len; - const u8 *name = dentry->d_name.name; - struct inode *dir = dentry->d_parent->d_inode; + int namelen = entry->len; + const u8 *name = entry->name; sb = dir->i_sb; /* NFS may look up ".." - look at dx_root directory block */ - if (namelen > 2 || name[0] != '.'||(name[1] != '.' && name[1] != '\0')){ - if (!(frame = dx_probe(dentry, NULL, &hinfo, frames, err))) + if (namelen > 2 || name[0] != '.'|| (namelen == 2 && name[1] != '.')) { + if (!(frame = dx_probe(entry, dir, &hinfo, frames, err))) return NULL; } else { frame = frames; @@ -1036,7 +1035,7 @@ static struct dentry *ext3_lookup(struct inode * dir, struct dentry *dentry, str if (dentry->d_name.len > EXT3_NAME_LEN) return ERR_PTR(-ENAMETOOLONG); - bh = ext3_find_entry(dentry, &de); + bh = ext3_find_entry(dir, &dentry->d_name, &de); inode = NULL; if (bh) { unsigned long ino = le32_to_cpu(de->inode); @@ -1057,18 +1056,11 @@ static struct dentry *ext3_lookup(struct inode * dir, struct dentry *dentry, str struct dentry *ext3_get_parent(struct dentry *child) { unsigned long ino; - struct dentry *parent; - struct inode *inode; - struct dentry dotdot; + struct qstr dotdot = {.name = "..", .len = 2}; struct ext3_dir_entry_2 * de; struct buffer_head *bh; - dotdot.d_name.name = ".."; - dotdot.d_name.len = 2; - dotdot.d_parent = child; /* confusing, isn't it! */ - - bh = ext3_find_entry(&dotdot, &de); - inode = NULL; + bh = ext3_find_entry(child->d_inode, &dotdot, &de); if (!bh) return ERR_PTR(-ENOENT); ino = le32_to_cpu(de->inode); @@ -1080,16 +1072,7 @@ struct dentry *ext3_get_parent(struct dentry *child) return ERR_PTR(-EIO); } - inode = ext3_iget(child->d_inode->i_sb, ino); - if (IS_ERR(inode)) - return ERR_CAST(inode); - - parent = d_alloc_anon(inode); - if (!parent) { - iput(inode); - parent = ERR_PTR(-ENOMEM); - } - return parent; + return d_obtain_alias(ext3_iget(child->d_inode->i_sb, ino)); } #define S_SHIFT 12 @@ -1503,7 +1486,7 @@ static int ext3_dx_add_entry(handle_t *handle, struct dentry *dentry, struct ext3_dir_entry_2 *de; int err; - frame = dx_probe(dentry, NULL, &hinfo, frames, &err); + frame = dx_probe(&dentry->d_name, dir, &hinfo, frames, &err); if (!frame) return err; entries = frame->entries; @@ -2056,7 +2039,7 @@ static int ext3_rmdir (struct inode * dir, struct dentry *dentry) return PTR_ERR(handle); retval = -ENOENT; - bh = ext3_find_entry (dentry, &de); + bh = ext3_find_entry(dir, &dentry->d_name, &de); if (!bh) goto end_rmdir; @@ -2118,7 +2101,7 @@ static int ext3_unlink(struct inode * dir, struct dentry *dentry) handle->h_sync = 1; retval = -ENOENT; - bh = ext3_find_entry (dentry, &de); + bh = ext3_find_entry(dir, &dentry->d_name, &de); if (!bh) goto end_unlink; @@ -2276,7 +2259,7 @@ static int ext3_rename (struct inode * old_dir, struct dentry *old_dentry, if (IS_DIRSYNC(old_dir) || IS_DIRSYNC(new_dir)) handle->h_sync = 1; - old_bh = ext3_find_entry (old_dentry, &old_de); + old_bh = ext3_find_entry(old_dir, &old_dentry->d_name, &old_de); /* * Check for inode number is _not_ due to possible IO errors. * We might rmdir the source, keep it as pwd of some process @@ -2289,7 +2272,7 @@ static int ext3_rename (struct inode * old_dir, struct dentry *old_dentry, goto end_rename; new_inode = new_dentry->d_inode; - new_bh = ext3_find_entry (new_dentry, &new_de); + new_bh = ext3_find_entry(new_dir, &new_dentry->d_name, &new_de); if (new_bh) { if (!new_inode) { brelse (new_bh); @@ -2355,7 +2338,8 @@ static int ext3_rename (struct inode * old_dir, struct dentry *old_dentry, struct buffer_head *old_bh2; struct ext3_dir_entry_2 *old_de2; - old_bh2 = ext3_find_entry(old_dentry, &old_de2); + old_bh2 = ext3_find_entry(old_dir, &old_dentry->d_name, + &old_de2); if (old_bh2) { retval = ext3_delete_entry(handle, old_dir, old_de2, old_bh2); diff --git a/fs/ext3/resize.c b/fs/ext3/resize.c index 77278e947e9..78fdf383637 100644 --- a/fs/ext3/resize.c +++ b/fs/ext3/resize.c @@ -790,7 +790,8 @@ int ext3_group_add(struct super_block *sb, struct ext3_new_group_data *input) if (reserved_gdb || gdb_off == 0) { if (!EXT3_HAS_COMPAT_FEATURE(sb, - EXT3_FEATURE_COMPAT_RESIZE_INODE)){ + EXT3_FEATURE_COMPAT_RESIZE_INODE) + || !le16_to_cpu(es->s_reserved_gdt_blocks)) { ext3_warning(sb, __func__, "No reserved GDT blocks, can't resize"); return -EPERM; diff --git a/fs/ext3/super.c b/fs/ext3/super.c index 399a96a6c55..f6c94f232ec 100644 --- a/fs/ext3/super.c +++ b/fs/ext3/super.c @@ -281,7 +281,8 @@ void ext3_abort (struct super_block * sb, const char * function, EXT3_SB(sb)->s_mount_state |= EXT3_ERROR_FS; sb->s_flags |= MS_RDONLY; EXT3_SB(sb)->s_mount_opt |= EXT3_MOUNT_ABORT; - journal_abort(EXT3_SB(sb)->s_journal, -EIO); + if (EXT3_SB(sb)->s_journal) + journal_abort(EXT3_SB(sb)->s_journal, -EIO); } void ext3_warning (struct super_block * sb, const char * function, @@ -347,7 +348,7 @@ fail: static int ext3_blkdev_put(struct block_device *bdev) { bd_release(bdev); - return blkdev_put(bdev); + return blkdev_put(bdev, FMODE_READ|FMODE_WRITE); } static int ext3_blkdev_remove(struct ext3_sb_info *sbi) @@ -390,10 +391,14 @@ static void ext3_put_super (struct super_block * sb) { struct ext3_sb_info *sbi = EXT3_SB(sb); struct ext3_super_block *es = sbi->s_es; - int i; + int i, err; ext3_xattr_put_super(sb); - journal_destroy(sbi->s_journal); + err = journal_destroy(sbi->s_journal); + sbi->s_journal = NULL; + if (err < 0) + ext3_abort(sb, __func__, "Couldn't clean up the journal"); + if (!(sb->s_flags & MS_RDONLY)) { EXT3_CLEAR_INCOMPAT_FEATURE(sb, EXT3_FEATURE_INCOMPAT_RECOVER); es->s_state = cpu_to_le16(sbi->s_mount_state); @@ -625,6 +630,9 @@ static int ext3_show_options(struct seq_file *seq, struct vfsmount *vfs) else if (test_opt(sb, DATA_FLAGS) == EXT3_MOUNT_WRITEBACK_DATA) seq_puts(seq, ",data=writeback"); + if (test_opt(sb, DATA_ERR_ABORT)) + seq_puts(seq, ",data_err=abort"); + ext3_show_quota_options(seq, sb); return 0; @@ -754,6 +762,7 @@ enum { Opt_reservation, Opt_noreservation, Opt_noload, Opt_nobh, Opt_bh, Opt_commit, Opt_journal_update, Opt_journal_inum, Opt_journal_dev, Opt_abort, Opt_data_journal, Opt_data_ordered, Opt_data_writeback, + Opt_data_err_abort, Opt_data_err_ignore, Opt_usrjquota, Opt_grpjquota, Opt_offusrjquota, Opt_offgrpjquota, Opt_jqfmt_vfsold, Opt_jqfmt_vfsv0, Opt_quota, Opt_noquota, Opt_ignore, Opt_barrier, Opt_err, Opt_resize, Opt_usrquota, @@ -796,6 +805,8 @@ static const match_table_t tokens = { {Opt_data_journal, "data=journal"}, {Opt_data_ordered, "data=ordered"}, {Opt_data_writeback, "data=writeback"}, + {Opt_data_err_abort, "data_err=abort"}, + {Opt_data_err_ignore, "data_err=ignore"}, {Opt_offusrjquota, "usrjquota="}, {Opt_usrjquota, "usrjquota=%s"}, {Opt_offgrpjquota, "grpjquota="}, @@ -1011,6 +1022,12 @@ static int parse_options (char *options, struct super_block *sb, sbi->s_mount_opt |= data_opt; } break; + case Opt_data_err_abort: + set_opt(sbi->s_mount_opt, DATA_ERR_ABORT); + break; + case Opt_data_err_ignore: + clear_opt(sbi->s_mount_opt, DATA_ERR_ABORT); + break; #ifdef CONFIG_QUOTA case Opt_usrjquota: qtype = USRQUOTA; @@ -1986,6 +2003,10 @@ static void ext3_init_journal_params(struct super_block *sb, journal_t *journal) journal->j_flags |= JFS_BARRIER; else journal->j_flags &= ~JFS_BARRIER; + if (test_opt(sb, DATA_ERR_ABORT)) + journal->j_flags |= JFS_ABORT_ON_SYNCDATA_ERR; + else + journal->j_flags &= ~JFS_ABORT_ON_SYNCDATA_ERR; spin_unlock(&journal->j_state_lock); } @@ -2050,7 +2071,7 @@ static journal_t *ext3_get_dev_journal(struct super_block *sb, if (bd_claim(bdev, sb)) { printk(KERN_ERR "EXT3: failed to claim external journal device.\n"); - blkdev_put(bdev); + blkdev_put(bdev, FMODE_READ|FMODE_WRITE); return NULL; } @@ -2280,7 +2301,9 @@ static void ext3_mark_recovery_complete(struct super_block * sb, journal_t *journal = EXT3_SB(sb)->s_journal; journal_lock_updates(journal); - journal_flush(journal); + if (journal_flush(journal) < 0) + goto out; + lock_super(sb); if (EXT3_HAS_INCOMPAT_FEATURE(sb, EXT3_FEATURE_INCOMPAT_RECOVER) && sb->s_flags & MS_RDONLY) { @@ -2289,6 +2312,8 @@ static void ext3_mark_recovery_complete(struct super_block * sb, ext3_commit_super(sb, es, 1); } unlock_super(sb); + +out: journal_unlock_updates(journal); } @@ -2350,12 +2375,9 @@ int ext3_force_commit(struct super_block *sb) /* * Ext3 always journals updates to the superblock itself, so we don't * have to propagate any other updates to the superblock on disk at this - * point. Just start an async writeback to get the buffers on their way - * to the disk. - * - * This implicitly triggers the writebehind on sync(). + * point. (We can probably nuke this function altogether, and remove + * any mention to sb->s_dirt in all of fs/ext3; eventual cleanup...) */ - static void ext3_write_super (struct super_block * sb) { if (mutex_trylock(&sb->s_lock) != 0) @@ -2365,13 +2387,12 @@ static void ext3_write_super (struct super_block * sb) static int ext3_sync_fs(struct super_block *sb, int wait) { - tid_t target; - sb->s_dirt = 0; - if (journal_start_commit(EXT3_SB(sb)->s_journal, &target)) { - if (wait) - log_wait_commit(EXT3_SB(sb)->s_journal, target); - } + if (wait) + ext3_force_commit(sb); + else + journal_start_commit(EXT3_SB(sb)->s_journal, NULL); + return 0; } @@ -2388,7 +2409,13 @@ static void ext3_write_super_lockfs(struct super_block *sb) /* Now we set up the journal barrier. */ journal_lock_updates(journal); - journal_flush(journal); + + /* + * We don't want to clear needs_recovery flag when we failed + * to flush the journal. + */ + if (journal_flush(journal) < 0) + return; /* Journal blocked and flushed, clear needs_recovery flag. */ EXT3_CLEAR_INCOMPAT_FEATURE(sb, EXT3_FEATURE_INCOMPAT_RECOVER); @@ -2767,30 +2794,30 @@ static int ext3_quota_on_mount(struct super_block *sb, int type) * Standard function to be called on quota_on */ static int ext3_quota_on(struct super_block *sb, int type, int format_id, - char *path, int remount) + char *name, int remount) { int err; - struct nameidata nd; + struct path path; if (!test_opt(sb, QUOTA)) return -EINVAL; - /* When remounting, no checks are needed and in fact, path is NULL */ + /* When remounting, no checks are needed and in fact, name is NULL */ if (remount) - return vfs_quota_on(sb, type, format_id, path, remount); + return vfs_quota_on(sb, type, format_id, name, remount); - err = path_lookup(path, LOOKUP_FOLLOW, &nd); + err = kern_path(name, LOOKUP_FOLLOW, &path); if (err) return err; /* Quotafile not on the same filesystem? */ - if (nd.path.mnt->mnt_sb != sb) { - path_put(&nd.path); + if (path.mnt->mnt_sb != sb) { + path_put(&path); return -EXDEV; } /* Journaling quota? */ if (EXT3_SB(sb)->s_qf_names[type]) { /* Quotafile not of fs root? */ - if (nd.path.dentry->d_parent->d_inode != sb->s_root->d_inode) + if (path.dentry->d_parent != sb->s_root) printk(KERN_WARNING "EXT3-fs: Quota file not on filesystem root. " "Journaled quota will not work.\n"); @@ -2800,18 +2827,22 @@ static int ext3_quota_on(struct super_block *sb, int type, int format_id, * When we journal data on quota file, we have to flush journal to see * all updates to the file when we bypass pagecache... */ - if (ext3_should_journal_data(nd.path.dentry->d_inode)) { + if (ext3_should_journal_data(path.dentry->d_inode)) { /* * We don't need to lock updates but journal_flush() could * otherwise be livelocked... */ journal_lock_updates(EXT3_SB(sb)->s_journal); - journal_flush(EXT3_SB(sb)->s_journal); + err = journal_flush(EXT3_SB(sb)->s_journal); journal_unlock_updates(EXT3_SB(sb)->s_journal); + if (err) { + path_put(&path); + return err; + } } - err = vfs_quota_on_path(sb, type, format_id, &nd.path); - path_put(&nd.path); + err = vfs_quota_on_path(sb, type, format_id, &path); + path_put(&path); return err; } diff --git a/fs/ext4/Kconfig b/fs/ext4/Kconfig new file mode 100644 index 00000000000..7505482a08f --- /dev/null +++ b/fs/ext4/Kconfig @@ -0,0 +1,79 @@ +config EXT4_FS + tristate "The Extended 4 (ext4) filesystem" + select JBD2 + select CRC16 + help + This is the next generation of the ext3 filesystem. + + Unlike the change from ext2 filesystem to ext3 filesystem, + the on-disk format of ext4 is not forwards compatible with + ext3; it is based on extent maps and it supports 48-bit + physical block numbers. The ext4 filesystem also supports delayed + allocation, persistent preallocation, high resolution time stamps, + and a number of other features to improve performance and speed + up fsck time. For more information, please see the web pages at + http://ext4.wiki.kernel.org. + + The ext4 filesystem will support mounting an ext3 + filesystem; while there will be some performance gains from + the delayed allocation and inode table readahead, the best + performance gains will require enabling ext4 features in the + filesystem, or formating a new filesystem as an ext4 + filesystem initially. + + To compile this file system support as a module, choose M here. The + module will be called ext4. + + If unsure, say N. + +config EXT4DEV_COMPAT + bool "Enable ext4dev compatibility" + depends on EXT4_FS + help + Starting with 2.6.28, the name of the ext4 filesystem was + renamed from ext4dev to ext4. Unfortunately there are some + legacy userspace programs (such as klibc's fstype) have + "ext4dev" hardcoded. + + To enable backwards compatibility so that systems that are + still expecting to mount ext4 filesystems using ext4dev, + chose Y here. This feature will go away by 2.6.31, so + please arrange to get your userspace programs fixed! + +config EXT4_FS_XATTR + bool "Ext4 extended attributes" + depends on EXT4_FS + default y + help + Extended attributes are name:value pairs associated with inodes by + the kernel or by users (see the attr(5) manual page, or visit + <http://acl.bestbits.at/> for details). + + If unsure, say N. + + You need this for POSIX ACL support on ext4. + +config EXT4_FS_POSIX_ACL + bool "Ext4 POSIX Access Control Lists" + depends on EXT4_FS_XATTR + select FS_POSIX_ACL + help + POSIX Access Control Lists (ACLs) support permissions for users and + groups beyond the owner/group/world scheme. + + To learn more about Access Control Lists, visit the POSIX ACLs for + Linux website <http://acl.bestbits.at/>. + + If you don't know what Access Control Lists are, say N + +config EXT4_FS_SECURITY + bool "Ext4 Security Labels" + depends on EXT4_FS_XATTR + help + Security labels support alternative access control models + implemented by security modules like SELinux. This option + enables an extended attribute handler for file security + labels in the ext4 filesystem. + + If you are not using a security module that requires using + extended attributes for file security labels, say N. diff --git a/fs/ext4/balloc.c b/fs/ext4/balloc.c index bd2ece22882..38b3acf5683 100644 --- a/fs/ext4/balloc.c +++ b/fs/ext4/balloc.c @@ -568,8 +568,16 @@ void ext4_free_blocks(handle_t *handle, struct inode *inode, /* this isn't the right place to decide whether block is metadata * inode.c/extents.c knows better, but for safety ... */ - if (S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode) || - ext4_should_journal_data(inode)) + if (S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode)) + metadata = 1; + + /* We need to make sure we don't reuse + * block released untill the transaction commit. + * writeback mode have weak data consistency so + * don't force data as metadata when freeing block + * for writeback mode. + */ + if (metadata == 0 && !ext4_should_writeback_data(inode)) metadata = 1; sb = inode->i_sb; @@ -581,26 +589,28 @@ void ext4_free_blocks(handle_t *handle, struct inode *inode, return; } -int ext4_claim_free_blocks(struct ext4_sb_info *sbi, - s64 nblocks) +/** + * ext4_has_free_blocks() + * @sbi: in-core super block structure. + * @nblocks: number of needed blocks + * + * Check if filesystem has nblocks free & available for allocation. + * On success return 1, return 0 on failure. + */ +int ext4_has_free_blocks(struct ext4_sb_info *sbi, s64 nblocks) { - s64 free_blocks, dirty_blocks; - s64 root_blocks = 0; + s64 free_blocks, dirty_blocks, root_blocks; struct percpu_counter *fbc = &sbi->s_freeblocks_counter; struct percpu_counter *dbc = &sbi->s_dirtyblocks_counter; free_blocks = percpu_counter_read_positive(fbc); dirty_blocks = percpu_counter_read_positive(dbc); - - if (!capable(CAP_SYS_RESOURCE) && - sbi->s_resuid != current->fsuid && - (sbi->s_resgid == 0 || !in_group_p(sbi->s_resgid))) - root_blocks = ext4_r_blocks_count(sbi->s_es); + root_blocks = ext4_r_blocks_count(sbi->s_es); if (free_blocks - (nblocks + root_blocks + dirty_blocks) < EXT4_FREEBLOCKS_WATERMARK) { - free_blocks = percpu_counter_sum(fbc); - dirty_blocks = percpu_counter_sum(dbc); + free_blocks = percpu_counter_sum_positive(fbc); + dirty_blocks = percpu_counter_sum_positive(dbc); if (dirty_blocks < 0) { printk(KERN_CRIT "Dirty block accounting " "went wrong %lld\n", @@ -608,57 +618,32 @@ int ext4_claim_free_blocks(struct ext4_sb_info *sbi, } } /* Check whether we have space after - * accounting for current dirty blocks + * accounting for current dirty blocks & root reserved blocks. */ - if (free_blocks < ((root_blocks + nblocks) + dirty_blocks)) - /* we don't have free space */ - return -ENOSPC; + if (free_blocks >= ((root_blocks + nblocks) + dirty_blocks)) + return 1; + + /* Hm, nope. Are (enough) root reserved blocks available? */ + if (sbi->s_resuid == current_fsuid() || + ((sbi->s_resgid != 0) && in_group_p(sbi->s_resgid)) || + capable(CAP_SYS_RESOURCE)) { + if (free_blocks >= (nblocks + dirty_blocks)) + return 1; + } - /* Add the blocks to nblocks */ - percpu_counter_add(dbc, nblocks); return 0; } -/** - * ext4_has_free_blocks() - * @sbi: in-core super block structure. - * @nblocks: number of neeed blocks - * - * Check if filesystem has free blocks available for allocation. - * Return the number of blocks avaible for allocation for this request - * On success, return nblocks - */ -ext4_fsblk_t ext4_has_free_blocks(struct ext4_sb_info *sbi, +int ext4_claim_free_blocks(struct ext4_sb_info *sbi, s64 nblocks) { - s64 free_blocks, dirty_blocks; - s64 root_blocks = 0; - struct percpu_counter *fbc = &sbi->s_freeblocks_counter; - struct percpu_counter *dbc = &sbi->s_dirtyblocks_counter; - - free_blocks = percpu_counter_read_positive(fbc); - dirty_blocks = percpu_counter_read_positive(dbc); - - if (!capable(CAP_SYS_RESOURCE) && - sbi->s_resuid != current->fsuid && - (sbi->s_resgid == 0 || !in_group_p(sbi->s_resgid))) - root_blocks = ext4_r_blocks_count(sbi->s_es); - - if (free_blocks - (nblocks + root_blocks + dirty_blocks) < - EXT4_FREEBLOCKS_WATERMARK) { - free_blocks = percpu_counter_sum(fbc); - dirty_blocks = percpu_counter_sum(dbc); - } - if (free_blocks <= (root_blocks + dirty_blocks)) - /* we don't have free space */ + if (ext4_has_free_blocks(sbi, nblocks)) { + percpu_counter_add(&sbi->s_dirtyblocks_counter, nblocks); return 0; - - if (free_blocks - (root_blocks + dirty_blocks) < nblocks) - return free_blocks - (root_blocks + dirty_blocks); - return nblocks; + } else + return -ENOSPC; } - /** * ext4_should_retry_alloc() * @sb: super block diff --git a/fs/ext4/dir.c b/fs/ext4/dir.c index 3ca6a2b7632..fed5b610df5 100644 --- a/fs/ext4/dir.c +++ b/fs/ext4/dir.c @@ -459,17 +459,8 @@ static int ext4_dx_readdir(struct file *filp, if (info->extra_fname) { if (call_filldir(filp, dirent, filldir, info->extra_fname)) goto finished; - info->extra_fname = NULL; - info->curr_node = rb_next(info->curr_node); - if (!info->curr_node) { - if (info->next_hash == ~0) { - filp->f_pos = EXT4_HTREE_EOF; - goto finished; - } - info->curr_hash = info->next_hash; - info->curr_minor_hash = 0; - } + goto next_node; } else if (!info->curr_node) info->curr_node = rb_first(&info->root); @@ -501,9 +492,14 @@ static int ext4_dx_readdir(struct file *filp, info->curr_minor_hash = fname->minor_hash; if (call_filldir(filp, dirent, filldir, fname)) break; - + next_node: info->curr_node = rb_next(info->curr_node); - if (!info->curr_node) { + if (info->curr_node) { + fname = rb_entry(info->curr_node, struct fname, + rb_hash); + info->curr_hash = fname->hash; + info->curr_minor_hash = fname->minor_hash; + } else { if (info->next_hash == ~0) { filp->f_pos = EXT4_HTREE_EOF; break; diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 6690a41cdd9..b0537c82702 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -511,7 +511,6 @@ do { \ /* * Mount flags */ -#define EXT4_MOUNT_CHECK 0x00001 /* Do mount-time checks */ #define EXT4_MOUNT_OLDALLOC 0x00002 /* Don't use the new Orlov allocator */ #define EXT4_MOUNT_GRPID 0x00004 /* Create files with directory's group */ #define EXT4_MOUNT_DEBUG 0x00008 /* Some debugging messages */ @@ -1004,8 +1003,7 @@ extern ext4_fsblk_t ext4_new_blocks(handle_t *handle, struct inode *inode, ext4_lblk_t iblock, ext4_fsblk_t goal, unsigned long *count, int *errp); extern int ext4_claim_free_blocks(struct ext4_sb_info *sbi, s64 nblocks); -extern ext4_fsblk_t ext4_has_free_blocks(struct ext4_sb_info *sbi, - s64 nblocks); +extern int ext4_has_free_blocks(struct ext4_sb_info *sbi, s64 nblocks); extern void ext4_free_blocks(handle_t *handle, struct inode *inode, ext4_fsblk_t block, unsigned long count, int metadata); extern void ext4_free_blocks_sb(handle_t *handle, struct super_block *sb, diff --git a/fs/ext4/ext4_sb.h b/fs/ext4/ext4_sb.h index 6a0b40d4326..445fde603df 100644 --- a/fs/ext4/ext4_sb.h +++ b/fs/ext4/ext4_sb.h @@ -99,9 +99,6 @@ struct ext4_sb_info { struct inode *s_buddy_cache; long s_blocks_reserved; spinlock_t s_reserve_lock; - struct list_head s_active_transaction; - struct list_head s_closed_transaction; - struct list_head s_committed_transaction; spinlock_t s_md_lock; tid_t s_last_transaction; unsigned short *s_mb_offsets, *s_mb_maxs; diff --git a/fs/ext4/ialloc.c b/fs/ext4/ialloc.c index fe34d74cfb1..08cac9fcace 100644 --- a/fs/ext4/ialloc.c +++ b/fs/ext4/ialloc.c @@ -718,6 +718,8 @@ got: gdp->bg_flags &= cpu_to_le16(~EXT4_BG_BLOCK_UNINIT); free = ext4_free_blocks_after_init(sb, group, gdp); gdp->bg_free_blocks_count = cpu_to_le16(free); + gdp->bg_checksum = ext4_group_desc_csum(sbi, group, + gdp); } spin_unlock(sb_bgl_lock(sbi, group)); @@ -785,7 +787,7 @@ got: spin_unlock(sb_bgl_lock(sbi, flex_group)); } - inode->i_uid = current->fsuid; + inode->i_uid = current_fsuid(); if (test_opt(sb, GRPID)) inode->i_gid = dir->i_gid; else if (dir->i_mode & S_ISGID) { @@ -793,7 +795,7 @@ got: if (S_ISDIR(mode)) mode |= S_ISGID; } else - inode->i_gid = current->fsgid; + inode->i_gid = current_fsgid(); inode->i_mode = mode; inode->i_ino = ino + group * EXT4_INODES_PER_GROUP(sb); diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 9b4ec9decfd..be21a5ae33c 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -1648,6 +1648,7 @@ static int mpage_da_submit_io(struct mpage_da_data *mpd) int ret = 0, err, nr_pages, i; unsigned long index, end; struct pagevec pvec; + long pages_skipped; BUG_ON(mpd->next_page <= mpd->first_page); pagevec_init(&pvec, 0); @@ -1655,20 +1656,30 @@ static int mpage_da_submit_io(struct mpage_da_data *mpd) end = mpd->next_page - 1; while (index <= end) { - /* XXX: optimize tail */ - nr_pages = pagevec_lookup(&pvec, mapping, index, PAGEVEC_SIZE); + /* + * We can use PAGECACHE_TAG_DIRTY lookup here because + * even though we have cleared the dirty flag on the page + * We still keep the page in the radix tree with tag + * PAGECACHE_TAG_DIRTY. See clear_page_dirty_for_io. + * The PAGECACHE_TAG_DIRTY is cleared in set_page_writeback + * which is called via the below writepage callback. + */ + nr_pages = pagevec_lookup_tag(&pvec, mapping, &index, + PAGECACHE_TAG_DIRTY, + min(end - index, + (pgoff_t)PAGEVEC_SIZE-1) + 1); if (nr_pages == 0) break; for (i = 0; i < nr_pages; i++) { struct page *page = pvec.pages[i]; - index = page->index; - if (index > end) - break; - index++; - + pages_skipped = mpd->wbc->pages_skipped; err = mapping->a_ops->writepage(page, mpd->wbc); - if (!err) + if (!err && (pages_skipped == mpd->wbc->pages_skipped)) + /* + * have successfully written the page + * without skipping the same + */ mpd->pages_written++; /* * In error case, we have to continue because @@ -2104,7 +2115,6 @@ static int mpage_da_writepages(struct address_space *mapping, struct writeback_control *wbc, struct mpage_da_data *mpd) { - long to_write; int ret; if (!mpd->get_block) @@ -2119,19 +2129,18 @@ static int mpage_da_writepages(struct address_space *mapping, mpd->pages_written = 0; mpd->retval = 0; - to_write = wbc->nr_to_write; - ret = write_cache_pages(mapping, wbc, __mpage_da_writepage, mpd); - /* * Handle last extent of pages */ if (!mpd->io_done && mpd->next_page != mpd->first_page) { if (mpage_da_map_blocks(mpd) == 0) mpage_da_submit_io(mpd); - } - wbc->nr_to_write = to_write - mpd->pages_written; + mpd->io_done = 1; + ret = MPAGE_DA_EXTENT_TAIL; + } + wbc->nr_to_write -= mpd->pages_written; return ret; } @@ -2320,6 +2329,8 @@ static int ext4_da_writepage(struct page *page, unlock_page(page); return 0; } + /* now mark the buffer_heads as dirty and uptodate */ + block_commit_write(page, 0, PAGE_CACHE_SIZE); } if (test_opt(inode->i_sb, NOBH) && ext4_should_writeback_data(inode)) @@ -2360,12 +2371,14 @@ static int ext4_da_writepages_trans_blocks(struct inode *inode) static int ext4_da_writepages(struct address_space *mapping, struct writeback_control *wbc) { + pgoff_t index; + int range_whole = 0; handle_t *handle = NULL; - loff_t range_start = 0; struct mpage_da_data mpd; struct inode *inode = mapping->host; + int no_nrwrite_index_update; + long pages_written = 0, pages_skipped; int needed_blocks, ret = 0, nr_to_writebump = 0; - long to_write, pages_skipped = 0; struct ext4_sb_info *sbi = EXT4_SB(mapping->host->i_sb); /* @@ -2385,23 +2398,26 @@ static int ext4_da_writepages(struct address_space *mapping, nr_to_writebump = sbi->s_mb_stream_request - wbc->nr_to_write; wbc->nr_to_write = sbi->s_mb_stream_request; } + if (wbc->range_start == 0 && wbc->range_end == LLONG_MAX) + range_whole = 1; - if (!wbc->range_cyclic) - /* - * If range_cyclic is not set force range_cont - * and save the old writeback_index - */ - wbc->range_cont = 1; - - range_start = wbc->range_start; - pages_skipped = wbc->pages_skipped; + if (wbc->range_cyclic) + index = mapping->writeback_index; + else + index = wbc->range_start >> PAGE_CACHE_SHIFT; mpd.wbc = wbc; mpd.inode = mapping->host; -restart_loop: - to_write = wbc->nr_to_write; - while (!ret && to_write > 0) { + /* + * we don't want write_cache_pages to update + * nr_to_write and writeback_index + */ + no_nrwrite_index_update = wbc->no_nrwrite_index_update; + wbc->no_nrwrite_index_update = 1; + pages_skipped = wbc->pages_skipped; + + while (!ret && wbc->nr_to_write > 0) { /* * we insert one extent at a time. So we need @@ -2422,48 +2438,53 @@ restart_loop: dump_stack(); goto out_writepages; } - to_write -= wbc->nr_to_write; - mpd.get_block = ext4_da_get_block_write; ret = mpage_da_writepages(mapping, wbc, &mpd); ext4_journal_stop(handle); - if (mpd.retval == -ENOSPC) + if (mpd.retval == -ENOSPC) { + /* commit the transaction which would + * free blocks released in the transaction + * and try again + */ jbd2_journal_force_commit_nested(sbi->s_journal); - - /* reset the retry count */ - if (ret == MPAGE_DA_EXTENT_TAIL) { + wbc->pages_skipped = pages_skipped; + ret = 0; + } else if (ret == MPAGE_DA_EXTENT_TAIL) { /* * got one extent now try with * rest of the pages */ - to_write += wbc->nr_to_write; + pages_written += mpd.pages_written; + wbc->pages_skipped = pages_skipped; ret = 0; - } else if (wbc->nr_to_write) { + } else if (wbc->nr_to_write) /* * There is no more writeout needed * or we requested for a noblocking writeout * and we found the device congested */ - to_write += wbc->nr_to_write; break; - } - wbc->nr_to_write = to_write; - } - - if (wbc->range_cont && (pages_skipped != wbc->pages_skipped)) { - /* We skipped pages in this loop */ - wbc->range_start = range_start; - wbc->nr_to_write = to_write + - wbc->pages_skipped - pages_skipped; - wbc->pages_skipped = pages_skipped; - goto restart_loop; } + if (pages_skipped != wbc->pages_skipped) + printk(KERN_EMERG "This should not happen leaving %s " + "with nr_to_write = %ld ret = %d\n", + __func__, wbc->nr_to_write, ret); + + /* Update index */ + index += pages_written; + if (wbc->range_cyclic || (range_whole && wbc->nr_to_write > 0)) + /* + * set the writeback_index so that range_cyclic + * mode will write it back later + */ + mapping->writeback_index = index; out_writepages: - wbc->nr_to_write = to_write - nr_to_writebump; - wbc->range_start = range_start; + if (!no_nrwrite_index_update) + wbc->no_nrwrite_index_update = 0; + wbc->nr_to_write -= nr_to_writebump; return ret; } @@ -4175,7 +4196,6 @@ static int ext4_inode_blocks_set(handle_t *handle, struct inode *inode = &(ei->vfs_inode); u64 i_blocks = inode->i_blocks; struct super_block *sb = inode->i_sb; - int err = 0; if (i_blocks <= ~0U) { /* @@ -4185,36 +4205,27 @@ static int ext4_inode_blocks_set(handle_t *handle, raw_inode->i_blocks_lo = cpu_to_le32(i_blocks); raw_inode->i_blocks_high = 0; ei->i_flags &= ~EXT4_HUGE_FILE_FL; - } else if (i_blocks <= 0xffffffffffffULL) { + return 0; + } + if (!EXT4_HAS_RO_COMPAT_FEATURE(sb, EXT4_FEATURE_RO_COMPAT_HUGE_FILE)) + return -EFBIG; + + if (i_blocks <= 0xffffffffffffULL) { /* * i_blocks can be represented in a 48 bit variable * as multiple of 512 bytes */ - err = ext4_update_rocompat_feature(handle, sb, - EXT4_FEATURE_RO_COMPAT_HUGE_FILE); - if (err) - goto err_out; - /* i_block is stored in the split 48 bit fields */ raw_inode->i_blocks_lo = cpu_to_le32(i_blocks); raw_inode->i_blocks_high = cpu_to_le16(i_blocks >> 32); ei->i_flags &= ~EXT4_HUGE_FILE_FL; } else { - /* - * i_blocks should be represented in a 48 bit variable - * as multiple of file system block size - */ - err = ext4_update_rocompat_feature(handle, sb, - EXT4_FEATURE_RO_COMPAT_HUGE_FILE); - if (err) - goto err_out; ei->i_flags |= EXT4_HUGE_FILE_FL; /* i_block is stored in file system block size */ i_blocks = i_blocks >> (inode->i_blkbits - 9); raw_inode->i_blocks_lo = cpu_to_le32(i_blocks); raw_inode->i_blocks_high = cpu_to_le16(i_blocks >> 32); } -err_out: - return err; + return 0; } /* @@ -4571,9 +4582,10 @@ static int ext4_indirect_trans_blocks(struct inode *inode, int nrblocks, static int ext4_index_trans_blocks(struct inode *inode, int nrblocks, int chunk) { if (!(EXT4_I(inode)->i_flags & EXT4_EXTENTS_FL)) - return ext4_indirect_trans_blocks(inode, nrblocks, 0); - return ext4_ext_index_trans_blocks(inode, nrblocks, 0); + return ext4_indirect_trans_blocks(inode, nrblocks, chunk); + return ext4_ext_index_trans_blocks(inode, nrblocks, chunk); } + /* * Account for index blocks, block groups bitmaps and block group * descriptor blocks if modify datablocks and index blocks diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index b580714f0d8..444ad998f72 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -2300,6 +2300,7 @@ int ext4_mb_add_groupinfo(struct super_block *sb, ext4_group_t group, } INIT_LIST_HEAD(&meta_group_info[i]->bb_prealloc_list); + meta_group_info[i]->bb_free_root.rb_node = NULL;; #ifdef DOUBLE_CHECK { @@ -2522,9 +2523,6 @@ int ext4_mb_init(struct super_block *sb, int needs_recovery) } spin_lock_init(&sbi->s_md_lock); - INIT_LIST_HEAD(&sbi->s_active_transaction); - INIT_LIST_HEAD(&sbi->s_closed_transaction); - INIT_LIST_HEAD(&sbi->s_committed_transaction); spin_lock_init(&sbi->s_bal_lock); sbi->s_mb_max_to_scan = MB_DEFAULT_MAX_TO_SCAN; @@ -2553,6 +2551,8 @@ int ext4_mb_init(struct super_block *sb, int needs_recovery) ext4_mb_init_per_dev_proc(sb); ext4_mb_history_init(sb); + sbi->s_journal->j_commit_callback = release_blocks_on_commit; + printk(KERN_INFO "EXT4-fs: mballoc enabled\n"); return 0; } @@ -2568,7 +2568,7 @@ static void ext4_mb_cleanup_pa(struct ext4_group_info *grp) pa = list_entry(cur, struct ext4_prealloc_space, pa_group_list); list_del(&pa->pa_group_list); count++; - kfree(pa); + kmem_cache_free(ext4_pspace_cachep, pa); } if (count) mb_debug("mballoc: %u PAs left\n", count); @@ -2582,15 +2582,6 @@ int ext4_mb_release(struct super_block *sb) struct ext4_group_info *grinfo; struct ext4_sb_info *sbi = EXT4_SB(sb); - /* release freed, non-committed blocks */ - spin_lock(&sbi->s_md_lock); - list_splice_init(&sbi->s_closed_transaction, - &sbi->s_committed_transaction); - list_splice_init(&sbi->s_active_transaction, - &sbi->s_committed_transaction); - spin_unlock(&sbi->s_md_lock); - ext4_mb_free_committed_blocks(sb); - if (sbi->s_group_info) { for (i = 0; i < sbi->s_groups_count; i++) { grinfo = ext4_get_group_info(sb, i); @@ -2644,61 +2635,57 @@ int ext4_mb_release(struct super_block *sb) return 0; } -static noinline_for_stack void -ext4_mb_free_committed_blocks(struct super_block *sb) +/* + * This function is called by the jbd2 layer once the commit has finished, + * so we know we can free the blocks that were released with that commit. + */ +static void release_blocks_on_commit(journal_t *journal, transaction_t *txn) { - struct ext4_sb_info *sbi = EXT4_SB(sb); - int err; - int i; - int count = 0; - int count2 = 0; - struct ext4_free_metadata *md; + struct super_block *sb = journal->j_private; struct ext4_buddy e4b; + struct ext4_group_info *db; + int err, count = 0, count2 = 0; + struct ext4_free_data *entry; + ext4_fsblk_t discard_block; + struct list_head *l, *ltmp; - if (list_empty(&sbi->s_committed_transaction)) - return; - - /* there is committed blocks to be freed yet */ - do { - /* get next array of blocks */ - md = NULL; - spin_lock(&sbi->s_md_lock); - if (!list_empty(&sbi->s_committed_transaction)) { - md = list_entry(sbi->s_committed_transaction.next, - struct ext4_free_metadata, list); - list_del(&md->list); - } - spin_unlock(&sbi->s_md_lock); - - if (md == NULL) - break; + list_for_each_safe(l, ltmp, &txn->t_private_list) { + entry = list_entry(l, struct ext4_free_data, list); mb_debug("gonna free %u blocks in group %lu (0x%p):", - md->num, md->group, md); + entry->count, entry->group, entry); - err = ext4_mb_load_buddy(sb, md->group, &e4b); + err = ext4_mb_load_buddy(sb, entry->group, &e4b); /* we expect to find existing buddy because it's pinned */ BUG_ON(err != 0); + db = e4b.bd_info; /* there are blocks to put in buddy to make them really free */ - count += md->num; + count += entry->count; count2++; - ext4_lock_group(sb, md->group); - for (i = 0; i < md->num; i++) { - mb_debug(" %u", md->blocks[i]); - mb_free_blocks(NULL, &e4b, md->blocks[i], 1); + ext4_lock_group(sb, entry->group); + /* Take it out of per group rb tree */ + rb_erase(&entry->node, &(db->bb_free_root)); + mb_free_blocks(NULL, &e4b, entry->start_blk, entry->count); + + if (!db->bb_free_root.rb_node) { + /* No more items in the per group rb tree + * balance refcounts from ext4_mb_free_metadata() + */ + page_cache_release(e4b.bd_buddy_page); + page_cache_release(e4b.bd_bitmap_page); } - mb_debug("\n"); - ext4_unlock_group(sb, md->group); - - /* balance refcounts from ext4_mb_free_metadata() */ - page_cache_release(e4b.bd_buddy_page); - page_cache_release(e4b.bd_bitmap_page); - - kfree(md); + ext4_unlock_group(sb, entry->group); + discard_block = (ext4_fsblk_t) entry->group * EXT4_BLOCKS_PER_GROUP(sb) + + entry->start_blk + + le32_to_cpu(EXT4_SB(sb)->s_es->s_first_data_block); + trace_mark(ext4_discard_blocks, "dev %s blk %llu count %u", sb->s_id, + (unsigned long long) discard_block, entry->count); + sb_issue_discard(sb, discard_block, entry->count); + + kmem_cache_free(ext4_free_ext_cachep, entry); ext4_mb_release_desc(&e4b); - - } while (md); + } mb_debug("freed %u blocks in %u structures\n", count, count2); } @@ -2712,6 +2699,7 @@ ext4_mb_free_committed_blocks(struct super_block *sb) static int ext4_mb_init_per_dev_proc(struct super_block *sb) { +#ifdef CONFIG_PROC_FS mode_t mode = S_IFREG | S_IRUGO | S_IWUSR; struct ext4_sb_info *sbi = EXT4_SB(sb); struct proc_dir_entry *proc; @@ -2735,10 +2723,14 @@ err_out: remove_proc_entry(EXT4_MB_MAX_TO_SCAN_NAME, sbi->s_proc); remove_proc_entry(EXT4_MB_STATS_NAME, sbi->s_proc); return -ENOMEM; +#else + return 0; +#endif } static int ext4_mb_destroy_per_dev_proc(struct super_block *sb) { +#ifdef CONFIG_PROC_FS struct ext4_sb_info *sbi = EXT4_SB(sb); if (sbi->s_proc == NULL) @@ -2750,7 +2742,7 @@ static int ext4_mb_destroy_per_dev_proc(struct super_block *sb) remove_proc_entry(EXT4_MB_MIN_TO_SCAN_NAME, sbi->s_proc); remove_proc_entry(EXT4_MB_MAX_TO_SCAN_NAME, sbi->s_proc); remove_proc_entry(EXT4_MB_STATS_NAME, sbi->s_proc); - +#endif return 0; } @@ -2771,6 +2763,16 @@ int __init init_ext4_mballoc(void) kmem_cache_destroy(ext4_pspace_cachep); return -ENOMEM; } + + ext4_free_ext_cachep = + kmem_cache_create("ext4_free_block_extents", + sizeof(struct ext4_free_data), + 0, SLAB_RECLAIM_ACCOUNT, NULL); + if (ext4_free_ext_cachep == NULL) { + kmem_cache_destroy(ext4_pspace_cachep); + kmem_cache_destroy(ext4_ac_cachep); + return -ENOMEM; + } return 0; } @@ -2779,6 +2781,7 @@ void exit_ext4_mballoc(void) /* XXX: synchronize_rcu(); */ kmem_cache_destroy(ext4_pspace_cachep); kmem_cache_destroy(ext4_ac_cachep); + kmem_cache_destroy(ext4_free_ext_cachep); } @@ -4324,8 +4327,6 @@ ext4_fsblk_t ext4_mb_new_blocks(handle_t *handle, goto out1; } - ext4_mb_poll_new_transaction(sb, handle); - *errp = ext4_mb_initialize_context(ac, ar); if (*errp) { ar->len = 0; @@ -4384,35 +4385,20 @@ out1: return block; } -static void ext4_mb_poll_new_transaction(struct super_block *sb, - handle_t *handle) -{ - struct ext4_sb_info *sbi = EXT4_SB(sb); - - if (sbi->s_last_transaction == handle->h_transaction->t_tid) - return; - - /* new transaction! time to close last one and free blocks for - * committed transaction. we know that only transaction can be - * active, so previos transaction can be being logged and we - * know that transaction before previous is known to be already - * logged. this means that now we may free blocks freed in all - * transactions before previous one. hope I'm clear enough ... */ - spin_lock(&sbi->s_md_lock); - if (sbi->s_last_transaction != handle->h_transaction->t_tid) { - mb_debug("new transaction %lu, old %lu\n", - (unsigned long) handle->h_transaction->t_tid, - (unsigned long) sbi->s_last_transaction); - list_splice_init(&sbi->s_closed_transaction, - &sbi->s_committed_transaction); - list_splice_init(&sbi->s_active_transaction, - &sbi->s_closed_transaction); - sbi->s_last_transaction = handle->h_transaction->t_tid; - } - spin_unlock(&sbi->s_md_lock); - - ext4_mb_free_committed_blocks(sb); +/* + * We can merge two free data extents only if the physical blocks + * are contiguous, AND the extents were freed by the same transaction, + * AND the blocks are associated with the same group. + */ +static int can_merge(struct ext4_free_data *entry1, + struct ext4_free_data *entry2) +{ + if ((entry1->t_tid == entry2->t_tid) && + (entry1->group == entry2->group) && + ((entry1->start_blk + entry1->count) == entry2->start_blk)) + return 1; + return 0; } static noinline_for_stack int @@ -4422,57 +4408,81 @@ ext4_mb_free_metadata(handle_t *handle, struct ext4_buddy *e4b, struct ext4_group_info *db = e4b->bd_info; struct super_block *sb = e4b->bd_sb; struct ext4_sb_info *sbi = EXT4_SB(sb); - struct ext4_free_metadata *md; - int i; + struct ext4_free_data *entry, *new_entry; + struct rb_node **n = &db->bb_free_root.rb_node, *node; + struct rb_node *parent = NULL, *new_node; + BUG_ON(e4b->bd_bitmap_page == NULL); BUG_ON(e4b->bd_buddy_page == NULL); - ext4_lock_group(sb, group); - for (i = 0; i < count; i++) { - md = db->bb_md_cur; - if (md && db->bb_tid != handle->h_transaction->t_tid) { - db->bb_md_cur = NULL; - md = NULL; - } + new_entry = kmem_cache_alloc(ext4_free_ext_cachep, GFP_NOFS); + new_entry->start_blk = block; + new_entry->group = group; + new_entry->count = count; + new_entry->t_tid = handle->h_transaction->t_tid; + new_node = &new_entry->node; - if (md == NULL) { + ext4_lock_group(sb, group); + if (!*n) { + /* first free block exent. We need to + protect buddy cache from being freed, + * otherwise we'll refresh it from + * on-disk bitmap and lose not-yet-available + * blocks */ + page_cache_get(e4b->bd_buddy_page); + page_cache_get(e4b->bd_bitmap_page); + } + while (*n) { + parent = *n; + entry = rb_entry(parent, struct ext4_free_data, node); + if (block < entry->start_blk) + n = &(*n)->rb_left; + else if (block >= (entry->start_blk + entry->count)) + n = &(*n)->rb_right; + else { ext4_unlock_group(sb, group); - md = kmalloc(sizeof(*md), GFP_NOFS); - if (md == NULL) - return -ENOMEM; - md->num = 0; - md->group = group; + ext4_error(sb, __func__, + "Double free of blocks %d (%d %d)\n", + block, entry->start_blk, entry->count); + return 0; + } + } - ext4_lock_group(sb, group); - if (db->bb_md_cur == NULL) { - spin_lock(&sbi->s_md_lock); - list_add(&md->list, &sbi->s_active_transaction); - spin_unlock(&sbi->s_md_lock); - /* protect buddy cache from being freed, - * otherwise we'll refresh it from - * on-disk bitmap and lose not-yet-available - * blocks */ - page_cache_get(e4b->bd_buddy_page); - page_cache_get(e4b->bd_bitmap_page); - db->bb_md_cur = md; - db->bb_tid = handle->h_transaction->t_tid; - mb_debug("new md 0x%p for group %lu\n", - md, md->group); - } else { - kfree(md); - md = db->bb_md_cur; - } + rb_link_node(new_node, parent, n); + rb_insert_color(new_node, &db->bb_free_root); + + /* Now try to see the extent can be merged to left and right */ + node = rb_prev(new_node); + if (node) { + entry = rb_entry(node, struct ext4_free_data, node); + if (can_merge(entry, new_entry)) { + new_entry->start_blk = entry->start_blk; + new_entry->count += entry->count; + rb_erase(node, &(db->bb_free_root)); + spin_lock(&sbi->s_md_lock); + list_del(&entry->list); + spin_unlock(&sbi->s_md_lock); + kmem_cache_free(ext4_free_ext_cachep, entry); } + } - BUG_ON(md->num >= EXT4_BB_MAX_BLOCKS); - md->blocks[md->num] = block + i; - md->num++; - if (md->num == EXT4_BB_MAX_BLOCKS) { - /* no more space, put full container on a sb's list */ - db->bb_md_cur = NULL; + node = rb_next(new_node); + if (node) { + entry = rb_entry(node, struct ext4_free_data, node); + if (can_merge(new_entry, entry)) { + new_entry->count += entry->count; + rb_erase(node, &(db->bb_free_root)); + spin_lock(&sbi->s_md_lock); + list_del(&entry->list); + spin_unlock(&sbi->s_md_lock); + kmem_cache_free(ext4_free_ext_cachep, entry); } } + /* Add the extent to transaction's private list */ + spin_lock(&sbi->s_md_lock); + list_add(&new_entry->list, &handle->h_transaction->t_private_list); + spin_unlock(&sbi->s_md_lock); ext4_unlock_group(sb, group); return 0; } @@ -4500,8 +4510,6 @@ void ext4_mb_free_blocks(handle_t *handle, struct inode *inode, *freed = 0; - ext4_mb_poll_new_transaction(sb, handle); - sbi = EXT4_SB(sb); es = EXT4_SB(sb)->s_es; if (block < le32_to_cpu(es->s_first_data_block) || diff --git a/fs/ext4/mballoc.h b/fs/ext4/mballoc.h index b3b4828f8b8..b5dff1fff1e 100644 --- a/fs/ext4/mballoc.h +++ b/fs/ext4/mballoc.h @@ -18,6 +18,8 @@ #include <linux/pagemap.h> #include <linux/seq_file.h> #include <linux/version.h> +#include <linux/blkdev.h> +#include <linux/marker.h> #include "ext4_jbd2.h" #include "ext4.h" #include "group.h" @@ -98,23 +100,29 @@ static struct kmem_cache *ext4_pspace_cachep; static struct kmem_cache *ext4_ac_cachep; +static struct kmem_cache *ext4_free_ext_cachep; -#ifdef EXT4_BB_MAX_BLOCKS -#undef EXT4_BB_MAX_BLOCKS -#endif -#define EXT4_BB_MAX_BLOCKS 30 +struct ext4_free_data { + /* this links the free block information from group_info */ + struct rb_node node; -struct ext4_free_metadata { - ext4_group_t group; - unsigned short num; - ext4_grpblk_t blocks[EXT4_BB_MAX_BLOCKS]; + /* this links the free block information from ext4_sb_info */ struct list_head list; + + /* group which free block extent belongs */ + ext4_group_t group; + + /* free block extent */ + ext4_grpblk_t start_blk; + ext4_grpblk_t count; + + /* transaction which freed this extent */ + tid_t t_tid; }; struct ext4_group_info { unsigned long bb_state; - unsigned long bb_tid; - struct ext4_free_metadata *bb_md_cur; + struct rb_root bb_free_root; unsigned short bb_first_free; unsigned short bb_free; unsigned short bb_fragments; @@ -261,8 +269,6 @@ struct buffer_head *read_block_bitmap(struct super_block *, ext4_group_t); static void ext4_mb_generate_from_pa(struct super_block *sb, void *bitmap, ext4_group_t group); -static void ext4_mb_poll_new_transaction(struct super_block *, handle_t *); -static void ext4_mb_free_committed_blocks(struct super_block *); static void ext4_mb_return_to_preallocation(struct inode *inode, struct ext4_buddy *e4b, sector_t block, int count); @@ -270,6 +276,7 @@ static void ext4_mb_put_pa(struct ext4_allocation_context *, struct super_block *, struct ext4_prealloc_space *pa); static int ext4_mb_init_per_dev_proc(struct super_block *sb); static int ext4_mb_destroy_per_dev_proc(struct super_block *sb); +static void release_blocks_on_commit(journal_t *journal, transaction_t *txn); static inline void ext4_lock_group(struct super_block *sb, ext4_group_t group) diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c index 92db9e94514..63adcb79298 100644 --- a/fs/ext4/namei.c +++ b/fs/ext4/namei.c @@ -1061,7 +1061,6 @@ static struct dentry *ext4_lookup(struct inode *dir, struct dentry *dentry, stru struct dentry *ext4_get_parent(struct dentry *child) { unsigned long ino; - struct dentry *parent; struct inode *inode; static const struct qstr dotdot = { .name = "..", @@ -1083,16 +1082,7 @@ struct dentry *ext4_get_parent(struct dentry *child) return ERR_PTR(-EIO); } - inode = ext4_iget(child->d_inode->i_sb, ino); - if (IS_ERR(inode)) - return ERR_CAST(inode); - - parent = d_alloc_anon(inode); - if (!parent) { - iput(inode); - parent = ERR_PTR(-ENOMEM); - } - return parent; + return d_obtain_alias(ext4_iget(child->d_inode->i_sb, ino)); } #define S_SHIFT 12 diff --git a/fs/ext4/super.c b/fs/ext4/super.c index dea8f13c2fd..04158ad74db 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -333,7 +333,8 @@ void ext4_abort(struct super_block *sb, const char *function, EXT4_SB(sb)->s_mount_state |= EXT4_ERROR_FS; sb->s_flags |= MS_RDONLY; EXT4_SB(sb)->s_mount_opt |= EXT4_MOUNT_ABORT; - jbd2_journal_abort(EXT4_SB(sb)->s_journal, -EIO); + if (EXT4_SB(sb)->s_journal) + jbd2_journal_abort(EXT4_SB(sb)->s_journal, -EIO); } void ext4_warning(struct super_block *sb, const char *function, @@ -374,66 +375,6 @@ void ext4_update_dynamic_rev(struct super_block *sb) */ } -int ext4_update_compat_feature(handle_t *handle, - struct super_block *sb, __u32 compat) -{ - int err = 0; - if (!EXT4_HAS_COMPAT_FEATURE(sb, compat)) { - err = ext4_journal_get_write_access(handle, - EXT4_SB(sb)->s_sbh); - if (err) - return err; - EXT4_SET_COMPAT_FEATURE(sb, compat); - sb->s_dirt = 1; - handle->h_sync = 1; - BUFFER_TRACE(EXT4_SB(sb)->s_sbh, - "call ext4_journal_dirty_met adata"); - err = ext4_journal_dirty_metadata(handle, - EXT4_SB(sb)->s_sbh); - } - return err; -} - -int ext4_update_rocompat_feature(handle_t *handle, - struct super_block *sb, __u32 rocompat) -{ - int err = 0; - if (!EXT4_HAS_RO_COMPAT_FEATURE(sb, rocompat)) { - err = ext4_journal_get_write_access(handle, - EXT4_SB(sb)->s_sbh); - if (err) - return err; - EXT4_SET_RO_COMPAT_FEATURE(sb, rocompat); - sb->s_dirt = 1; - handle->h_sync = 1; - BUFFER_TRACE(EXT4_SB(sb)->s_sbh, - "call ext4_journal_dirty_met adata"); - err = ext4_journal_dirty_metadata(handle, - EXT4_SB(sb)->s_sbh); - } - return err; -} - -int ext4_update_incompat_feature(handle_t *handle, - struct super_block *sb, __u32 incompat) -{ - int err = 0; - if (!EXT4_HAS_INCOMPAT_FEATURE(sb, incompat)) { - err = ext4_journal_get_write_access(handle, - EXT4_SB(sb)->s_sbh); - if (err) - return err; - EXT4_SET_INCOMPAT_FEATURE(sb, incompat); - sb->s_dirt = 1; - handle->h_sync = 1; - BUFFER_TRACE(EXT4_SB(sb)->s_sbh, - "call ext4_journal_dirty_met adata"); - err = ext4_journal_dirty_metadata(handle, - EXT4_SB(sb)->s_sbh); - } - return err; -} - /* * Open the external journal device */ @@ -459,7 +400,7 @@ fail: static int ext4_blkdev_put(struct block_device *bdev) { bd_release(bdev); - return blkdev_put(bdev); + return blkdev_put(bdev, FMODE_READ|FMODE_WRITE); } static int ext4_blkdev_remove(struct ext4_sb_info *sbi) @@ -502,14 +443,16 @@ static void ext4_put_super(struct super_block *sb) { struct ext4_sb_info *sbi = EXT4_SB(sb); struct ext4_super_block *es = sbi->s_es; - int i; + int i, err; ext4_mb_release(sb); ext4_ext_release(sb); ext4_xattr_put_super(sb); - if (jbd2_journal_destroy(sbi->s_journal) < 0) - ext4_abort(sb, __func__, "Couldn't clean up the journal"); + err = jbd2_journal_destroy(sbi->s_journal); sbi->s_journal = NULL; + if (err < 0) + ext4_abort(sb, __func__, "Couldn't clean up the journal"); + if (!(sb->s_flags & MS_RDONLY)) { EXT4_CLEAR_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_RECOVER); es->s_state = cpu_to_le16(sbi->s_mount_state); @@ -904,7 +847,7 @@ static const struct export_operations ext4_export_ops = { enum { Opt_bsd_df, Opt_minix_df, Opt_grpid, Opt_nogrpid, Opt_resgid, Opt_resuid, Opt_sb, Opt_err_cont, Opt_err_panic, Opt_err_ro, - Opt_nouid32, Opt_nocheck, Opt_debug, Opt_oldalloc, Opt_orlov, + Opt_nouid32, Opt_debug, Opt_oldalloc, Opt_orlov, Opt_user_xattr, Opt_nouser_xattr, Opt_acl, Opt_noacl, Opt_reservation, Opt_noreservation, Opt_noload, Opt_nobh, Opt_bh, Opt_commit, Opt_journal_update, Opt_journal_inum, Opt_journal_dev, @@ -915,7 +858,7 @@ enum { Opt_jqfmt_vfsold, Opt_jqfmt_vfsv0, Opt_quota, Opt_noquota, Opt_ignore, Opt_barrier, Opt_err, Opt_resize, Opt_usrquota, Opt_grpquota, Opt_extents, Opt_noextents, Opt_i_version, - Opt_mballoc, Opt_nomballoc, Opt_stripe, Opt_delalloc, Opt_nodelalloc, + Opt_stripe, Opt_delalloc, Opt_nodelalloc, Opt_inode_readahead_blks }; @@ -933,8 +876,6 @@ static const match_table_t tokens = { {Opt_err_panic, "errors=panic"}, {Opt_err_ro, "errors=remount-ro"}, {Opt_nouid32, "nouid32"}, - {Opt_nocheck, "nocheck"}, - {Opt_nocheck, "check=none"}, {Opt_debug, "debug"}, {Opt_oldalloc, "oldalloc"}, {Opt_orlov, "orlov"}, @@ -973,8 +914,6 @@ static const match_table_t tokens = { {Opt_extents, "extents"}, {Opt_noextents, "noextents"}, {Opt_i_version, "i_version"}, - {Opt_mballoc, "mballoc"}, - {Opt_nomballoc, "nomballoc"}, {Opt_stripe, "stripe=%u"}, {Opt_resize, "resize"}, {Opt_delalloc, "delalloc"}, @@ -1073,9 +1012,6 @@ static int parse_options(char *options, struct super_block *sb, case Opt_nouid32: set_opt(sbi->s_mount_opt, NO_UID32); break; - case Opt_nocheck: - clear_opt(sbi->s_mount_opt, CHECK); - break; case Opt_debug: set_opt(sbi->s_mount_opt, DEBUG); break; @@ -1522,9 +1458,8 @@ static int ext4_fill_flex_info(struct super_block *sb) /* We allocate both existing and potentially added groups */ flex_group_count = ((sbi->s_groups_count + groups_per_flex - 1) + - ((sbi->s_es->s_reserved_gdt_blocks +1 ) << - EXT4_DESC_PER_BLOCK_BITS(sb))) / - groups_per_flex; + ((le16_to_cpu(sbi->s_es->s_reserved_gdt_blocks) + 1) << + EXT4_DESC_PER_BLOCK_BITS(sb))) / groups_per_flex; sbi->s_flex_groups = kzalloc(flex_group_count * sizeof(struct flex_groups), GFP_KERNEL); if (sbi->s_flex_groups == NULL) { @@ -1618,14 +1553,14 @@ static int ext4_check_descriptors(struct super_block *sb) if (block_bitmap < first_block || block_bitmap > last_block) { printk(KERN_ERR "EXT4-fs: ext4_check_descriptors: " "Block bitmap for group %lu not in group " - "(block %llu)!", i, block_bitmap); + "(block %llu)!\n", i, block_bitmap); return 0; } inode_bitmap = ext4_inode_bitmap(sb, gdp); if (inode_bitmap < first_block || inode_bitmap > last_block) { printk(KERN_ERR "EXT4-fs: ext4_check_descriptors: " "Inode bitmap for group %lu not in group " - "(block %llu)!", i, inode_bitmap); + "(block %llu)!\n", i, inode_bitmap); return 0; } inode_table = ext4_inode_table(sb, gdp); @@ -1633,7 +1568,7 @@ static int ext4_check_descriptors(struct super_block *sb) inode_table + sbi->s_itb_per_group - 1 > last_block) { printk(KERN_ERR "EXT4-fs: ext4_check_descriptors: " "Inode table for group %lu not in group " - "(block %llu)!", i, inode_table); + "(block %llu)!\n", i, inode_table); return 0; } spin_lock(sb_bgl_lock(sbi, i)); @@ -1778,15 +1713,15 @@ static void ext4_orphan_cleanup(struct super_block *sb, * * Note, this does *not* consider any metadata overhead for vfs i_blocks. */ -static loff_t ext4_max_size(int blkbits) +static loff_t ext4_max_size(int blkbits, int has_huge_files) { loff_t res; loff_t upper_limit = MAX_LFS_FILESIZE; /* small i_blocks in vfs inode? */ - if (sizeof(blkcnt_t) < sizeof(u64)) { + if (!has_huge_files || sizeof(blkcnt_t) < sizeof(u64)) { /* - * CONFIG_LSF is not enabled implies the inode + * CONFIG_LBD is not enabled implies the inode * i_block represent total blocks in 512 bytes * 32 == size of vfs inode i_blocks * 8 */ @@ -1814,7 +1749,7 @@ static loff_t ext4_max_size(int blkbits) * block limit, and also a limit of (2^48 - 1) 512-byte sectors in i_blocks. * We need to be 1 filesystem block less than the 2^48 sector limit. */ -static loff_t ext4_max_bitmap_size(int bits) +static loff_t ext4_max_bitmap_size(int bits, int has_huge_files) { loff_t res = EXT4_NDIR_BLOCKS; int meta_blocks; @@ -1827,11 +1762,11 @@ static loff_t ext4_max_bitmap_size(int bits) * total number of 512 bytes blocks of the file */ - if (sizeof(blkcnt_t) < sizeof(u64)) { + if (!has_huge_files || sizeof(blkcnt_t) < sizeof(u64)) { /* - * CONFIG_LSF is not enabled implies the inode - * i_block represent total blocks in 512 bytes - * 32 == size of vfs inode i_blocks * 8 + * !has_huge_files or CONFIG_LBD is not enabled + * implies the inode i_block represent total blocks in + * 512 bytes 32 == size of vfs inode i_blocks * 8 */ upper_limit = (1LL << 32) - 1; @@ -1940,7 +1875,7 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) int blocksize; int db_count; int i; - int needs_recovery; + int needs_recovery, has_huge_files; __le32 features; __u64 blocks_count; int err; @@ -2081,16 +2016,18 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) sb->s_id, le32_to_cpu(features)); goto failed_mount; } - if (EXT4_HAS_RO_COMPAT_FEATURE(sb, EXT4_FEATURE_RO_COMPAT_HUGE_FILE)) { + has_huge_files = EXT4_HAS_RO_COMPAT_FEATURE(sb, + EXT4_FEATURE_RO_COMPAT_HUGE_FILE); + if (has_huge_files) { /* * Large file size enabled file system can only be - * mount if kernel is build with CONFIG_LSF + * mount if kernel is build with CONFIG_LBD */ if (sizeof(root->i_blocks) < sizeof(u64) && !(sb->s_flags & MS_RDONLY)) { printk(KERN_ERR "EXT4-fs: %s: Filesystem with huge " "files cannot be mounted read-write " - "without CONFIG_LSF.\n", sb->s_id); + "without CONFIG_LBD.\n", sb->s_id); goto failed_mount; } } @@ -2131,8 +2068,9 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) } } - sbi->s_bitmap_maxbytes = ext4_max_bitmap_size(sb->s_blocksize_bits); - sb->s_maxbytes = ext4_max_size(sb->s_blocksize_bits); + sbi->s_bitmap_maxbytes = ext4_max_bitmap_size(sb->s_blocksize_bits, + has_huge_files); + sb->s_maxbytes = ext4_max_size(sb->s_blocksize_bits, has_huge_files); if (le32_to_cpu(es->s_rev_level) == EXT4_GOOD_OLD_REV) { sbi->s_inode_size = EXT4_GOOD_OLD_INODE_SIZE; @@ -2456,6 +2394,21 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) "available.\n"); } + if (test_opt(sb, DATA_FLAGS) == EXT4_MOUNT_JOURNAL_DATA) { + printk(KERN_WARNING "EXT4-fs: Ignoring delalloc option - " + "requested data journaling mode\n"); + clear_opt(sbi->s_mount_opt, DELALLOC); + } else if (test_opt(sb, DELALLOC)) + printk(KERN_INFO "EXT4-fs: delayed allocation enabled\n"); + + ext4_ext_init(sb); + err = ext4_mb_init(sb, needs_recovery); + if (err) { + printk(KERN_ERR "EXT4-fs: failed to initalize mballoc (%d)\n", + err); + goto failed_mount4; + } + /* * akpm: core read_super() calls in here with the superblock locked. * That deadlocks, because orphan cleanup needs to lock the superblock @@ -2475,21 +2428,6 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) test_opt(sb, DATA_FLAGS) == EXT4_MOUNT_ORDERED_DATA ? "ordered": "writeback"); - if (test_opt(sb, DATA_FLAGS) == EXT4_MOUNT_JOURNAL_DATA) { - printk(KERN_WARNING "EXT4-fs: Ignoring delalloc option - " - "requested data journaling mode\n"); - clear_opt(sbi->s_mount_opt, DELALLOC); - } else if (test_opt(sb, DELALLOC)) - printk(KERN_INFO "EXT4-fs: delayed allocation enabled\n"); - - ext4_ext_init(sb); - err = ext4_mb_init(sb, needs_recovery); - if (err) { - printk(KERN_ERR "EXT4-fs: failed to initalize mballoc (%d)\n", - err); - goto failed_mount4; - } - lock_kernel(); return 0; @@ -2617,7 +2555,7 @@ static journal_t *ext4_get_dev_journal(struct super_block *sb, if (bd_claim(bdev, sb)) { printk(KERN_ERR "EXT4: failed to claim external journal device.\n"); - blkdev_put(bdev); + blkdev_put(bdev, FMODE_READ|FMODE_WRITE); return NULL; } @@ -2946,12 +2884,9 @@ int ext4_force_commit(struct super_block *sb) /* * Ext4 always journals updates to the superblock itself, so we don't * have to propagate any other updates to the superblock on disk at this - * point. Just start an async writeback to get the buffers on their way - * to the disk. - * - * This implicitly triggers the writebehind on sync(). + * point. (We can probably nuke this function altogether, and remove + * any mention to sb->s_dirt in all of fs/ext4; eventual cleanup...) */ - static void ext4_write_super(struct super_block *sb) { if (mutex_trylock(&sb->s_lock) != 0) @@ -2961,15 +2896,15 @@ static void ext4_write_super(struct super_block *sb) static int ext4_sync_fs(struct super_block *sb, int wait) { - tid_t target; + int ret = 0; trace_mark(ext4_sync_fs, "dev %s wait %d", sb->s_id, wait); sb->s_dirt = 0; - if (jbd2_journal_start_commit(EXT4_SB(sb)->s_journal, &target)) { - if (wait) - jbd2_log_wait_commit(EXT4_SB(sb)->s_journal, target); - } - return 0; + if (wait) + ret = ext4_force_commit(sb); + else + jbd2_journal_start_commit(EXT4_SB(sb)->s_journal, NULL); + return ret; } /* @@ -3392,30 +3327,30 @@ static int ext4_quota_on_mount(struct super_block *sb, int type) * Standard function to be called on quota_on */ static int ext4_quota_on(struct super_block *sb, int type, int format_id, - char *path, int remount) + char *name, int remount) { int err; - struct nameidata nd; + struct path path; if (!test_opt(sb, QUOTA)) return -EINVAL; - /* When remounting, no checks are needed and in fact, path is NULL */ + /* When remounting, no checks are needed and in fact, name is NULL */ if (remount) - return vfs_quota_on(sb, type, format_id, path, remount); + return vfs_quota_on(sb, type, format_id, name, remount); - err = path_lookup(path, LOOKUP_FOLLOW, &nd); + err = kern_path(name, LOOKUP_FOLLOW, &path); if (err) return err; /* Quotafile not on the same filesystem? */ - if (nd.path.mnt->mnt_sb != sb) { - path_put(&nd.path); + if (path.mnt->mnt_sb != sb) { + path_put(&path); return -EXDEV; } /* Journaling quota? */ if (EXT4_SB(sb)->s_qf_names[type]) { /* Quotafile not in fs root? */ - if (nd.path.dentry->d_parent->d_inode != sb->s_root->d_inode) + if (path.dentry->d_parent != sb->s_root) printk(KERN_WARNING "EXT4-fs: Quota file not on filesystem root. " "Journaled quota will not work.\n"); @@ -3425,7 +3360,7 @@ static int ext4_quota_on(struct super_block *sb, int type, int format_id, * When we journal data on quota file, we have to flush journal to see * all updates to the file when we bypass pagecache... */ - if (ext4_should_journal_data(nd.path.dentry->d_inode)) { + if (ext4_should_journal_data(path.dentry->d_inode)) { /* * We don't need to lock updates but journal_flush() could * otherwise be livelocked... @@ -3434,13 +3369,13 @@ static int ext4_quota_on(struct super_block *sb, int type, int format_id, err = jbd2_journal_flush(EXT4_SB(sb)->s_journal); jbd2_journal_unlock_updates(EXT4_SB(sb)->s_journal); if (err) { - path_put(&nd.path); + path_put(&path); return err; } } - err = vfs_quota_on_path(sb, type, format_id, &nd.path); - path_put(&nd.path); + err = vfs_quota_on_path(sb, type, format_id, &path); + path_put(&path); return err; } diff --git a/fs/fat/Makefile b/fs/fat/Makefile index bfb5f06cf2c..e06190322c1 100644 --- a/fs/fat/Makefile +++ b/fs/fat/Makefile @@ -3,5 +3,9 @@ # obj-$(CONFIG_FAT_FS) += fat.o +obj-$(CONFIG_VFAT_FS) += vfat.o +obj-$(CONFIG_MSDOS_FS) += msdos.o -fat-objs := cache.o dir.o fatent.o file.o inode.o misc.o +fat-y := cache.o dir.o fatent.o file.o inode.o misc.o +vfat-y := namei_vfat.o +msdos-y := namei_msdos.o diff --git a/fs/fat/cache.c b/fs/fat/cache.c index 3222f51c41c..b4260229808 100644 --- a/fs/fat/cache.c +++ b/fs/fat/cache.c @@ -9,8 +9,8 @@ */ #include <linux/fs.h> -#include <linux/msdos_fs.h> #include <linux/buffer_head.h> +#include "fat.h" /* this must be > 0. */ #define FAT_MAX_CACHE 8 @@ -293,10 +293,12 @@ static int fat_bmap_cluster(struct inode *inode, int cluster) } int fat_bmap(struct inode *inode, sector_t sector, sector_t *phys, - unsigned long *mapped_blocks) + unsigned long *mapped_blocks, int create) { struct super_block *sb = inode->i_sb; struct msdos_sb_info *sbi = MSDOS_SB(sb); + const unsigned long blocksize = sb->s_blocksize; + const unsigned char blocksize_bits = sb->s_blocksize_bits; sector_t last_block; int cluster, offset; @@ -309,10 +311,21 @@ int fat_bmap(struct inode *inode, sector_t sector, sector_t *phys, } return 0; } - last_block = (MSDOS_I(inode)->mmu_private + (sb->s_blocksize - 1)) - >> sb->s_blocksize_bits; - if (sector >= last_block) - return 0; + + last_block = (i_size_read(inode) + (blocksize - 1)) >> blocksize_bits; + if (sector >= last_block) { + if (!create) + return 0; + + /* + * ->mmu_private can access on only allocation path. + * (caller must hold ->i_mutex) + */ + last_block = (MSDOS_I(inode)->mmu_private + (blocksize - 1)) + >> blocksize_bits; + if (sector >= last_block) + return 0; + } cluster = sector >> (sbi->cluster_bits - sb->s_blocksize_bits); offset = sector & (sbi->sec_per_clus - 1); diff --git a/fs/fat/dir.c b/fs/fat/dir.c index cd4a0162e10..67e05835709 100644 --- a/fs/fat/dir.c +++ b/fs/fat/dir.c @@ -16,11 +16,11 @@ #include <linux/module.h> #include <linux/slab.h> #include <linux/time.h> -#include <linux/msdos_fs.h> #include <linux/smp_lock.h> #include <linux/buffer_head.h> #include <linux/compat.h> #include <asm/uaccess.h> +#include "fat.h" static inline loff_t fat_make_i_pos(struct super_block *sb, struct buffer_head *bh, @@ -77,7 +77,7 @@ next: *bh = NULL; iblock = *pos >> sb->s_blocksize_bits; - err = fat_bmap(dir, iblock, &phys, &mapped_blocks); + err = fat_bmap(dir, iblock, &phys, &mapped_blocks, 0); if (err || !phys) return -1; /* beyond EOF or error */ @@ -86,7 +86,7 @@ next: *bh = sb_bread(sb, phys); if (*bh == NULL) { printk(KERN_ERR "FAT: Directory bread(block %llu) failed\n", - (unsigned long long)phys); + (llu)phys); /* skip this block */ *pos = (iblock + 1) << sb->s_blocksize_bits; goto next; @@ -373,9 +373,10 @@ parse_record: if (de->attr == ATTR_EXT) { int status = fat_parse_long(inode, &cpos, &bh, &de, &unicode, &nr_slots); - if (status < 0) - return status; - else if (status == PARSE_INVALID) + if (status < 0) { + err = status; + goto end_of_dir; + } else if (status == PARSE_INVALID) continue; else if (status == PARSE_NOT_LONGNAME) goto parse_record; @@ -832,6 +833,7 @@ static long fat_compat_dir_ioctl(struct file *filp, unsigned cmd, #endif /* CONFIG_COMPAT */ const struct file_operations fat_dir_operations = { + .llseek = generic_file_llseek, .read = generic_read_dir, .readdir = fat_readdir, .ioctl = fat_dir_ioctl, @@ -839,6 +841,7 @@ const struct file_operations fat_dir_operations = { .compat_ioctl = fat_compat_dir_ioctl, #endif .fsync = file_fsync, + .llseek = generic_file_llseek, }; static int fat_get_short_entry(struct inode *dir, loff_t *pos, @@ -1088,6 +1091,7 @@ int fat_alloc_new_dir(struct inode *dir, struct timespec *ts) struct msdos_dir_entry *de; sector_t blknr; __le16 date, time; + u8 time_cs; int err, cluster; err = fat_alloc_clusters(dir, &cluster, 1); @@ -1101,7 +1105,7 @@ int fat_alloc_new_dir(struct inode *dir, struct timespec *ts) goto error_free; } - fat_date_unix2dos(ts->tv_sec, &time, &date, sbi->options.tz_utc); + fat_time_unix2fat(sbi, ts, &time, &date, &time_cs); de = (struct msdos_dir_entry *)bhs[0]->b_data; /* filling the new directory slots ("." and ".." entries) */ @@ -1111,13 +1115,14 @@ int fat_alloc_new_dir(struct inode *dir, struct timespec *ts) de[0].lcase = de[1].lcase = 0; de[0].time = de[1].time = time; de[0].date = de[1].date = date; - de[0].ctime_cs = de[1].ctime_cs = 0; if (sbi->options.isvfat) { /* extra timestamps */ de[0].ctime = de[1].ctime = time; + de[0].ctime_cs = de[1].ctime_cs = time_cs; de[0].adate = de[0].cdate = de[1].adate = de[1].cdate = date; } else { de[0].ctime = de[1].ctime = 0; + de[0].ctime_cs = de[1].ctime_cs = 0; de[0].adate = de[0].cdate = de[1].adate = de[1].cdate = 0; } de[0].start = cpu_to_le16(cluster); diff --git a/fs/fat/fat.h b/fs/fat/fat.h new file mode 100644 index 00000000000..ea440d65819 --- /dev/null +++ b/fs/fat/fat.h @@ -0,0 +1,329 @@ +#ifndef _FAT_H +#define _FAT_H + +#include <linux/buffer_head.h> +#include <linux/string.h> +#include <linux/nls.h> +#include <linux/fs.h> +#include <linux/mutex.h> +#include <linux/msdos_fs.h> + +/* + * vfat shortname flags + */ +#define VFAT_SFN_DISPLAY_LOWER 0x0001 /* convert to lowercase for display */ +#define VFAT_SFN_DISPLAY_WIN95 0x0002 /* emulate win95 rule for display */ +#define VFAT_SFN_DISPLAY_WINNT 0x0004 /* emulate winnt rule for display */ +#define VFAT_SFN_CREATE_WIN95 0x0100 /* emulate win95 rule for create */ +#define VFAT_SFN_CREATE_WINNT 0x0200 /* emulate winnt rule for create */ + +struct fat_mount_options { + uid_t fs_uid; + gid_t fs_gid; + unsigned short fs_fmask; + unsigned short fs_dmask; + unsigned short codepage; /* Codepage for shortname conversions */ + char *iocharset; /* Charset used for filename input/display */ + unsigned short shortname; /* flags for shortname display/create rule */ + unsigned char name_check; /* r = relaxed, n = normal, s = strict */ + unsigned short allow_utime;/* permission for setting the [am]time */ + unsigned quiet:1, /* set = fake successful chmods and chowns */ + showexec:1, /* set = only set x bit for com/exe/bat */ + sys_immutable:1, /* set = system files are immutable */ + dotsOK:1, /* set = hidden and system files are named '.filename' */ + isvfat:1, /* 0=no vfat long filename support, 1=vfat support */ + utf8:1, /* Use of UTF-8 character set (Default) */ + unicode_xlate:1, /* create escape sequences for unhandled Unicode */ + numtail:1, /* Does first alias have a numeric '~1' type tail? */ + flush:1, /* write things quickly */ + nocase:1, /* Does this need case conversion? 0=need case conversion*/ + usefree:1, /* Use free_clusters for FAT32 */ + tz_utc:1, /* Filesystem timestamps are in UTC */ + rodir:1; /* allow ATTR_RO for directory */ +}; + +#define FAT_HASH_BITS 8 +#define FAT_HASH_SIZE (1UL << FAT_HASH_BITS) + +/* + * MS-DOS file system in-core superblock data + */ +struct msdos_sb_info { + unsigned short sec_per_clus; /* sectors/cluster */ + unsigned short cluster_bits; /* log2(cluster_size) */ + unsigned int cluster_size; /* cluster size */ + unsigned char fats,fat_bits; /* number of FATs, FAT bits (12 or 16) */ + unsigned short fat_start; + unsigned long fat_length; /* FAT start & length (sec.) */ + unsigned long dir_start; + unsigned short dir_entries; /* root dir start & entries */ + unsigned long data_start; /* first data sector */ + unsigned long max_cluster; /* maximum cluster number */ + unsigned long root_cluster; /* first cluster of the root directory */ + unsigned long fsinfo_sector; /* sector number of FAT32 fsinfo */ + struct mutex fat_lock; + unsigned int prev_free; /* previously allocated cluster number */ + unsigned int free_clusters; /* -1 if undefined */ + unsigned int free_clus_valid; /* is free_clusters valid? */ + struct fat_mount_options options; + struct nls_table *nls_disk; /* Codepage used on disk */ + struct nls_table *nls_io; /* Charset used for input and display */ + const void *dir_ops; /* Opaque; default directory operations */ + int dir_per_block; /* dir entries per block */ + int dir_per_block_bits; /* log2(dir_per_block) */ + + int fatent_shift; + struct fatent_operations *fatent_ops; + + spinlock_t inode_hash_lock; + struct hlist_head inode_hashtable[FAT_HASH_SIZE]; +}; + +#define FAT_CACHE_VALID 0 /* special case for valid cache */ + +/* + * MS-DOS file system inode data in memory + */ +struct msdos_inode_info { + spinlock_t cache_lru_lock; + struct list_head cache_lru; + int nr_caches; + /* for avoiding the race between fat_free() and fat_get_cluster() */ + unsigned int cache_valid_id; + + /* NOTE: mmu_private is 64bits, so must hold ->i_mutex to access */ + loff_t mmu_private; /* physically allocated size */ + + int i_start; /* first cluster or 0 */ + int i_logstart; /* logical first cluster */ + int i_attrs; /* unused attribute bits */ + loff_t i_pos; /* on-disk position of directory entry or 0 */ + struct hlist_node i_fat_hash; /* hash by i_location */ + struct inode vfs_inode; +}; + +struct fat_slot_info { + loff_t i_pos; /* on-disk position of directory entry */ + loff_t slot_off; /* offset for slot or de start */ + int nr_slots; /* number of slots + 1(de) in filename */ + struct msdos_dir_entry *de; + struct buffer_head *bh; +}; + +static inline struct msdos_sb_info *MSDOS_SB(struct super_block *sb) +{ + return sb->s_fs_info; +} + +static inline struct msdos_inode_info *MSDOS_I(struct inode *inode) +{ + return container_of(inode, struct msdos_inode_info, vfs_inode); +} + +/* + * If ->i_mode can't hold S_IWUGO (i.e. ATTR_RO), we use ->i_attrs to + * save ATTR_RO instead of ->i_mode. + * + * If it's directory and !sbi->options.rodir, ATTR_RO isn't read-only + * bit, it's just used as flag for app. + */ +static inline int fat_mode_can_hold_ro(struct inode *inode) +{ + struct msdos_sb_info *sbi = MSDOS_SB(inode->i_sb); + mode_t mask; + + if (S_ISDIR(inode->i_mode)) { + if (!sbi->options.rodir) + return 0; + mask = ~sbi->options.fs_dmask; + } else + mask = ~sbi->options.fs_fmask; + + if (!(mask & S_IWUGO)) + return 0; + return 1; +} + +/* Convert attribute bits and a mask to the UNIX mode. */ +static inline mode_t fat_make_mode(struct msdos_sb_info *sbi, + u8 attrs, mode_t mode) +{ + if (attrs & ATTR_RO && !((attrs & ATTR_DIR) && !sbi->options.rodir)) + mode &= ~S_IWUGO; + + if (attrs & ATTR_DIR) + return (mode & ~sbi->options.fs_dmask) | S_IFDIR; + else + return (mode & ~sbi->options.fs_fmask) | S_IFREG; +} + +/* Return the FAT attribute byte for this inode */ +static inline u8 fat_make_attrs(struct inode *inode) +{ + u8 attrs = MSDOS_I(inode)->i_attrs; + if (S_ISDIR(inode->i_mode)) + attrs |= ATTR_DIR; + if (fat_mode_can_hold_ro(inode) && !(inode->i_mode & S_IWUGO)) + attrs |= ATTR_RO; + return attrs; +} + +static inline void fat_save_attrs(struct inode *inode, u8 attrs) +{ + if (fat_mode_can_hold_ro(inode)) + MSDOS_I(inode)->i_attrs = attrs & ATTR_UNUSED; + else + MSDOS_I(inode)->i_attrs = attrs & (ATTR_UNUSED | ATTR_RO); +} + +static inline unsigned char fat_checksum(const __u8 *name) +{ + unsigned char s = name[0]; + s = (s<<7) + (s>>1) + name[1]; s = (s<<7) + (s>>1) + name[2]; + s = (s<<7) + (s>>1) + name[3]; s = (s<<7) + (s>>1) + name[4]; + s = (s<<7) + (s>>1) + name[5]; s = (s<<7) + (s>>1) + name[6]; + s = (s<<7) + (s>>1) + name[7]; s = (s<<7) + (s>>1) + name[8]; + s = (s<<7) + (s>>1) + name[9]; s = (s<<7) + (s>>1) + name[10]; + return s; +} + +static inline sector_t fat_clus_to_blknr(struct msdos_sb_info *sbi, int clus) +{ + return ((sector_t)clus - FAT_START_ENT) * sbi->sec_per_clus + + sbi->data_start; +} + +static inline void fat16_towchar(wchar_t *dst, const __u8 *src, size_t len) +{ +#ifdef __BIG_ENDIAN + while (len--) { + *dst++ = src[0] | (src[1] << 8); + src += 2; + } +#else + memcpy(dst, src, len * 2); +#endif +} + +static inline void fatwchar_to16(__u8 *dst, const wchar_t *src, size_t len) +{ +#ifdef __BIG_ENDIAN + while (len--) { + dst[0] = *src & 0x00FF; + dst[1] = (*src & 0xFF00) >> 8; + dst += 2; + src++; + } +#else + memcpy(dst, src, len * 2); +#endif +} + +/* fat/cache.c */ +extern void fat_cache_inval_inode(struct inode *inode); +extern int fat_get_cluster(struct inode *inode, int cluster, + int *fclus, int *dclus); +extern int fat_bmap(struct inode *inode, sector_t sector, sector_t *phys, + unsigned long *mapped_blocks, int create); + +/* fat/dir.c */ +extern const struct file_operations fat_dir_operations; +extern int fat_search_long(struct inode *inode, const unsigned char *name, + int name_len, struct fat_slot_info *sinfo); +extern int fat_dir_empty(struct inode *dir); +extern int fat_subdirs(struct inode *dir); +extern int fat_scan(struct inode *dir, const unsigned char *name, + struct fat_slot_info *sinfo); +extern int fat_get_dotdot_entry(struct inode *dir, struct buffer_head **bh, + struct msdos_dir_entry **de, loff_t *i_pos); +extern int fat_alloc_new_dir(struct inode *dir, struct timespec *ts); +extern int fat_add_entries(struct inode *dir, void *slots, int nr_slots, + struct fat_slot_info *sinfo); +extern int fat_remove_entries(struct inode *dir, struct fat_slot_info *sinfo); + +/* fat/fatent.c */ +struct fat_entry { + int entry; + union { + u8 *ent12_p[2]; + __le16 *ent16_p; + __le32 *ent32_p; + } u; + int nr_bhs; + struct buffer_head *bhs[2]; +}; + +static inline void fatent_init(struct fat_entry *fatent) +{ + fatent->nr_bhs = 0; + fatent->entry = 0; + fatent->u.ent32_p = NULL; + fatent->bhs[0] = fatent->bhs[1] = NULL; +} + +static inline void fatent_set_entry(struct fat_entry *fatent, int entry) +{ + fatent->entry = entry; + fatent->u.ent32_p = NULL; +} + +static inline void fatent_brelse(struct fat_entry *fatent) +{ + int i; + fatent->u.ent32_p = NULL; + for (i = 0; i < fatent->nr_bhs; i++) + brelse(fatent->bhs[i]); + fatent->nr_bhs = 0; + fatent->bhs[0] = fatent->bhs[1] = NULL; +} + +extern void fat_ent_access_init(struct super_block *sb); +extern int fat_ent_read(struct inode *inode, struct fat_entry *fatent, + int entry); +extern int fat_ent_write(struct inode *inode, struct fat_entry *fatent, + int new, int wait); +extern int fat_alloc_clusters(struct inode *inode, int *cluster, + int nr_cluster); +extern int fat_free_clusters(struct inode *inode, int cluster); +extern int fat_count_free_clusters(struct super_block *sb); + +/* fat/file.c */ +extern int fat_generic_ioctl(struct inode *inode, struct file *filp, + unsigned int cmd, unsigned long arg); +extern const struct file_operations fat_file_operations; +extern const struct inode_operations fat_file_inode_operations; +extern int fat_setattr(struct dentry * dentry, struct iattr * attr); +extern void fat_truncate(struct inode *inode); +extern int fat_getattr(struct vfsmount *mnt, struct dentry *dentry, + struct kstat *stat); + +/* fat/inode.c */ +extern void fat_attach(struct inode *inode, loff_t i_pos); +extern void fat_detach(struct inode *inode); +extern struct inode *fat_iget(struct super_block *sb, loff_t i_pos); +extern struct inode *fat_build_inode(struct super_block *sb, + struct msdos_dir_entry *de, loff_t i_pos); +extern int fat_sync_inode(struct inode *inode); +extern int fat_fill_super(struct super_block *sb, void *data, int silent, + const struct inode_operations *fs_dir_inode_ops, int isvfat); + +extern int fat_flush_inodes(struct super_block *sb, struct inode *i1, + struct inode *i2); +/* fat/misc.c */ +extern void fat_fs_panic(struct super_block *s, const char *fmt, ...) + __attribute__ ((format (printf, 2, 3))) __cold; +extern void fat_clusters_flush(struct super_block *sb); +extern int fat_chain_add(struct inode *inode, int new_dclus, int nr_cluster); +extern void fat_time_fat2unix(struct msdos_sb_info *sbi, struct timespec *ts, + __le16 __time, __le16 __date, u8 time_cs); +extern void fat_time_unix2fat(struct msdos_sb_info *sbi, struct timespec *ts, + __le16 *time, __le16 *date, u8 *time_cs); +extern int fat_sync_bhs(struct buffer_head **bhs, int nr_bhs); + +int fat_cache_init(void); +void fat_cache_destroy(void); + +/* helper for printk */ +typedef unsigned long long llu; + +#endif /* !_FAT_H */ diff --git a/fs/fat/fatent.c b/fs/fat/fatent.c index fb98b3d847e..da6eea47872 100644 --- a/fs/fat/fatent.c +++ b/fs/fat/fatent.c @@ -7,6 +7,7 @@ #include <linux/fs.h> #include <linux/msdos_fs.h> #include <linux/blkdev.h> +#include "fat.h" struct fatent_operations { void (*ent_blocknr)(struct super_block *, int, int *, sector_t *); @@ -92,8 +93,7 @@ static int fat12_ent_bread(struct super_block *sb, struct fat_entry *fatent, err_brelse: brelse(bhs[0]); err: - printk(KERN_ERR "FAT: FAT read failed (blocknr %llu)\n", - (unsigned long long)blocknr); + printk(KERN_ERR "FAT: FAT read failed (blocknr %llu)\n", (llu)blocknr); return -EIO; } @@ -106,7 +106,7 @@ static int fat_ent_bread(struct super_block *sb, struct fat_entry *fatent, fatent->bhs[0] = sb_bread(sb, blocknr); if (!fatent->bhs[0]) { printk(KERN_ERR "FAT: FAT read failed (blocknr %llu)\n", - (unsigned long long)blocknr); + (llu)blocknr); return -EIO; } fatent->nr_bhs = 1; @@ -316,10 +316,20 @@ static inline int fat_ent_update_ptr(struct super_block *sb, /* Is this fatent's blocks including this entry? */ if (!fatent->nr_bhs || bhs[0]->b_blocknr != blocknr) return 0; - /* Does this entry need the next block? */ - if (sbi->fat_bits == 12 && (offset + 1) >= sb->s_blocksize) { - if (fatent->nr_bhs != 2 || bhs[1]->b_blocknr != (blocknr + 1)) - return 0; + if (sbi->fat_bits == 12) { + if ((offset + 1) < sb->s_blocksize) { + /* This entry is on bhs[0]. */ + if (fatent->nr_bhs == 2) { + brelse(bhs[1]); + fatent->nr_bhs = 1; + } + } else { + /* This entry needs the next block. */ + if (fatent->nr_bhs != 2) + return 0; + if (bhs[1]->b_blocknr != (blocknr + 1)) + return 0; + } } ops->ent_set_ptr(fatent, offset); return 1; diff --git a/fs/fat/file.c b/fs/fat/file.c index ddde37025ca..0a7f4a9918b 100644 --- a/fs/fat/file.c +++ b/fs/fat/file.c @@ -10,13 +10,13 @@ #include <linux/module.h> #include <linux/mount.h> #include <linux/time.h> -#include <linux/msdos_fs.h> #include <linux/buffer_head.h> #include <linux/writeback.h> #include <linux/backing-dev.h> #include <linux/blkdev.h> #include <linux/fsnotify.h> #include <linux/security.h> +#include "fat.h" int fat_generic_ioctl(struct inode *inode, struct file *filp, unsigned int cmd, unsigned long arg) @@ -29,10 +29,9 @@ int fat_generic_ioctl(struct inode *inode, struct file *filp, { u32 attr; - if (inode->i_ino == MSDOS_ROOT_INO) - attr = ATTR_DIR; - else - attr = fat_attr(inode); + mutex_lock(&inode->i_mutex); + attr = fat_make_attrs(inode); + mutex_unlock(&inode->i_mutex); return put_user(attr, user_attr); } @@ -62,20 +61,16 @@ int fat_generic_ioctl(struct inode *inode, struct file *filp, /* Merge in ATTR_VOLUME and ATTR_DIR */ attr |= (MSDOS_I(inode)->i_attrs & ATTR_VOLUME) | (is_dir ? ATTR_DIR : 0); - oldattr = fat_attr(inode); + oldattr = fat_make_attrs(inode); /* Equivalent to a chmod() */ ia.ia_valid = ATTR_MODE | ATTR_CTIME; ia.ia_ctime = current_fs_time(inode->i_sb); - if (is_dir) { - ia.ia_mode = MSDOS_MKMODE(attr, - S_IRWXUGO & ~sbi->options.fs_dmask) - | S_IFDIR; - } else { - ia.ia_mode = MSDOS_MKMODE(attr, - (S_IRUGO | S_IWUGO | (inode->i_mode & S_IXUGO)) - & ~sbi->options.fs_fmask) - | S_IFREG; + if (is_dir) + ia.ia_mode = fat_make_mode(sbi, attr, S_IRWXUGO); + else { + ia.ia_mode = fat_make_mode(sbi, attr, + S_IRUGO | S_IWUGO | (inode->i_mode & S_IXUGO)); } /* The root directory has no attributes */ @@ -115,7 +110,7 @@ int fat_generic_ioctl(struct inode *inode, struct file *filp, inode->i_flags &= S_IMMUTABLE; } - MSDOS_I(inode)->i_attrs = attr & ATTR_UNUSED; + fat_save_attrs(inode, attr); mark_inode_dirty(inode); up: mnt_drop_write(filp->f_path.mnt); @@ -274,7 +269,7 @@ static int fat_sanitize_mode(const struct msdos_sb_info *sbi, /* * Note, the basic check is already done by a caller of - * (attr->ia_mode & ~MSDOS_VALID_MODE) + * (attr->ia_mode & ~FAT_VALID_MODE) */ if (S_ISREG(inode->i_mode)) @@ -287,11 +282,18 @@ static int fat_sanitize_mode(const struct msdos_sb_info *sbi, /* * Of the r and x bits, all (subject to umask) must be present. Of the * w bits, either all (subject to umask) or none must be present. + * + * If fat_mode_can_hold_ro(inode) is false, can't change w bits. */ if ((perm & (S_IRUGO | S_IXUGO)) != (inode->i_mode & (S_IRUGO|S_IXUGO))) return -EPERM; - if ((perm & S_IWUGO) && ((perm & S_IWUGO) != (S_IWUGO & ~mask))) - return -EPERM; + if (fat_mode_can_hold_ro(inode)) { + if ((perm & S_IWUGO) && ((perm & S_IWUGO) != (S_IWUGO & ~mask))) + return -EPERM; + } else { + if ((perm & S_IWUGO) != (S_IWUGO & ~mask)) + return -EPERM; + } *mode_ptr &= S_IFMT | perm; @@ -302,7 +304,7 @@ static int fat_allow_set_time(struct msdos_sb_info *sbi, struct inode *inode) { mode_t allow_utime = sbi->options.allow_utime; - if (current->fsuid != inode->i_uid) { + if (current_fsuid() != inode->i_uid) { if (in_group_p(inode->i_gid)) allow_utime >>= 3; if (allow_utime & MAY_WRITE) @@ -314,13 +316,15 @@ static int fat_allow_set_time(struct msdos_sb_info *sbi, struct inode *inode) } #define TIMES_SET_FLAGS (ATTR_MTIME_SET | ATTR_ATIME_SET | ATTR_TIMES_SET) +/* valid file mode bits */ +#define FAT_VALID_MODE (S_IFREG | S_IFDIR | S_IRWXUGO) int fat_setattr(struct dentry *dentry, struct iattr *attr) { struct msdos_sb_info *sbi = MSDOS_SB(dentry->d_sb); struct inode *inode = dentry->d_inode; - int error = 0; unsigned int ia_valid; + int error; /* * Expand the file. Since inode_setattr() updates ->i_size @@ -356,7 +360,7 @@ int fat_setattr(struct dentry *dentry, struct iattr *attr) ((attr->ia_valid & ATTR_GID) && (attr->ia_gid != sbi->options.fs_gid)) || ((attr->ia_valid & ATTR_MODE) && - (attr->ia_mode & ~MSDOS_VALID_MODE))) + (attr->ia_mode & ~FAT_VALID_MODE))) error = -EPERM; if (error) { @@ -374,7 +378,8 @@ int fat_setattr(struct dentry *dentry, struct iattr *attr) attr->ia_valid &= ~ATTR_MODE; } - error = inode_setattr(inode, attr); + if (attr->ia_valid) + error = inode_setattr(inode, attr); out: return error; } diff --git a/fs/fat/inode.c b/fs/fat/inode.c index d12cdf2a040..d937aaf7737 100644 --- a/fs/fat/inode.c +++ b/fs/fat/inode.c @@ -16,7 +16,6 @@ #include <linux/slab.h> #include <linux/smp_lock.h> #include <linux/seq_file.h> -#include <linux/msdos_fs.h> #include <linux/pagemap.h> #include <linux/mpage.h> #include <linux/buffer_head.h> @@ -27,7 +26,9 @@ #include <linux/uio.h> #include <linux/writeback.h> #include <linux/log2.h> +#include <linux/hash.h> #include <asm/unaligned.h> +#include "fat.h" #ifndef CONFIG_FAT_DEFAULT_IOCHARSET /* if user don't select VFAT, this is undefined. */ @@ -63,7 +64,7 @@ static inline int __fat_get_block(struct inode *inode, sector_t iblock, sector_t phys; int err, offset; - err = fat_bmap(inode, iblock, &phys, &mapped_blocks); + err = fat_bmap(inode, iblock, &phys, &mapped_blocks, create); if (err) return err; if (phys) { @@ -93,7 +94,7 @@ static inline int __fat_get_block(struct inode *inode, sector_t iblock, *max_blocks = min(mapped_blocks, *max_blocks); MSDOS_I(inode)->mmu_private += *max_blocks << sb->s_blocksize_bits; - err = fat_bmap(inode, iblock, &phys, &mapped_blocks); + err = fat_bmap(inode, iblock, &phys, &mapped_blocks, create); if (err) return err; @@ -175,7 +176,7 @@ static ssize_t fat_direct_IO(int rw, struct kiocb *iocb, if (rw == WRITE) { /* - * FIXME: blockdev_direct_IO() doesn't use ->prepare_write(), + * FIXME: blockdev_direct_IO() doesn't use ->write_begin(), * so we need to update the ->mmu_private to block boundary. * * But we must fill the remaining area or hole by nul for @@ -198,7 +199,14 @@ static ssize_t fat_direct_IO(int rw, struct kiocb *iocb, static sector_t _fat_bmap(struct address_space *mapping, sector_t block) { - return generic_block_bmap(mapping, block, fat_get_block); + sector_t blocknr; + + /* fat_get_cluster() assumes the requested blocknr isn't truncated. */ + mutex_lock(&mapping->host->i_mutex); + blocknr = generic_block_bmap(mapping, block, fat_get_block); + mutex_unlock(&mapping->host->i_mutex); + + return blocknr; } static const struct address_space_operations fat_aops = { @@ -247,25 +255,21 @@ static void fat_hash_init(struct super_block *sb) INIT_HLIST_HEAD(&sbi->inode_hashtable[i]); } -static inline unsigned long fat_hash(struct super_block *sb, loff_t i_pos) +static inline unsigned long fat_hash(loff_t i_pos) { - unsigned long tmp = (unsigned long)i_pos | (unsigned long) sb; - tmp = tmp + (tmp >> FAT_HASH_BITS) + (tmp >> FAT_HASH_BITS * 2); - return tmp & FAT_HASH_MASK; + return hash_32(i_pos, FAT_HASH_BITS); } void fat_attach(struct inode *inode, loff_t i_pos) { - struct super_block *sb = inode->i_sb; - struct msdos_sb_info *sbi = MSDOS_SB(sb); + struct msdos_sb_info *sbi = MSDOS_SB(inode->i_sb); + struct hlist_head *head = sbi->inode_hashtable + fat_hash(i_pos); spin_lock(&sbi->inode_hash_lock); MSDOS_I(inode)->i_pos = i_pos; - hlist_add_head(&MSDOS_I(inode)->i_fat_hash, - sbi->inode_hashtable + fat_hash(sb, i_pos)); + hlist_add_head(&MSDOS_I(inode)->i_fat_hash, head); spin_unlock(&sbi->inode_hash_lock); } - EXPORT_SYMBOL_GPL(fat_attach); void fat_detach(struct inode *inode) @@ -276,13 +280,12 @@ void fat_detach(struct inode *inode) hlist_del_init(&MSDOS_I(inode)->i_fat_hash); spin_unlock(&sbi->inode_hash_lock); } - EXPORT_SYMBOL_GPL(fat_detach); struct inode *fat_iget(struct super_block *sb, loff_t i_pos) { struct msdos_sb_info *sbi = MSDOS_SB(sb); - struct hlist_head *head = sbi->inode_hashtable + fat_hash(sb, i_pos); + struct hlist_head *head = sbi->inode_hashtable + fat_hash(i_pos); struct hlist_node *_p; struct msdos_inode_info *i; struct inode *inode = NULL; @@ -341,8 +344,7 @@ static int fat_fill_inode(struct inode *inode, struct msdos_dir_entry *de) if ((de->attr & ATTR_DIR) && !IS_FREE(de->name)) { inode->i_generation &= ~1; - inode->i_mode = MSDOS_MKMODE(de->attr, - S_IRWXUGO & ~sbi->options.fs_dmask) | S_IFDIR; + inode->i_mode = fat_make_mode(sbi, de->attr, S_IRWXUGO); inode->i_op = sbi->dir_ops; inode->i_fop = &fat_dir_operations; @@ -359,10 +361,9 @@ static int fat_fill_inode(struct inode *inode, struct msdos_dir_entry *de) inode->i_nlink = fat_subdirs(inode); } else { /* not a directory */ inode->i_generation |= 1; - inode->i_mode = MSDOS_MKMODE(de->attr, - ((sbi->options.showexec && !is_exec(de->name + 8)) - ? S_IRUGO|S_IWUGO : S_IRWXUGO) - & ~sbi->options.fs_fmask) | S_IFREG; + inode->i_mode = fat_make_mode(sbi, de->attr, + ((sbi->options.showexec && !is_exec(de->name + 8)) + ? S_IRUGO|S_IWUGO : S_IRWXUGO)); MSDOS_I(inode)->i_start = le16_to_cpu(de->start); if (sbi->fat_bits == 32) MSDOS_I(inode)->i_start |= (le16_to_cpu(de->starthi) << 16); @@ -378,25 +379,16 @@ static int fat_fill_inode(struct inode *inode, struct msdos_dir_entry *de) if (sbi->options.sys_immutable) inode->i_flags |= S_IMMUTABLE; } - MSDOS_I(inode)->i_attrs = de->attr & ATTR_UNUSED; + fat_save_attrs(inode, de->attr); + inode->i_blocks = ((inode->i_size + (sbi->cluster_size - 1)) & ~((loff_t)sbi->cluster_size - 1)) >> 9; - inode->i_mtime.tv_sec = - date_dos2unix(le16_to_cpu(de->time), le16_to_cpu(de->date), - sbi->options.tz_utc); - inode->i_mtime.tv_nsec = 0; + + fat_time_fat2unix(sbi, &inode->i_mtime, de->time, de->date, 0); if (sbi->options.isvfat) { - int secs = de->ctime_cs / 100; - int csecs = de->ctime_cs % 100; - inode->i_ctime.tv_sec = - date_dos2unix(le16_to_cpu(de->ctime), - le16_to_cpu(de->cdate), - sbi->options.tz_utc) + secs; - inode->i_ctime.tv_nsec = csecs * 10000000; - inode->i_atime.tv_sec = - date_dos2unix(0, le16_to_cpu(de->adate), - sbi->options.tz_utc); - inode->i_atime.tv_nsec = 0; + fat_time_fat2unix(sbi, &inode->i_ctime, de->ctime, + de->cdate, de->ctime_cs); + fat_time_fat2unix(sbi, &inode->i_atime, 0, de->adate, 0); } else inode->i_ctime = inode->i_atime = inode->i_mtime; @@ -443,13 +435,8 @@ static void fat_delete_inode(struct inode *inode) static void fat_clear_inode(struct inode *inode) { - struct super_block *sb = inode->i_sb; - struct msdos_sb_info *sbi = MSDOS_SB(sb); - - spin_lock(&sbi->inode_hash_lock); fat_cache_inval_inode(inode); - hlist_del_init(&MSDOS_I(inode)->i_fat_hash); - spin_unlock(&sbi->inode_hash_lock); + fat_detach(inode); } static void fat_write_super(struct super_block *sb) @@ -555,6 +542,20 @@ static int fat_statfs(struct dentry *dentry, struct kstatfs *buf) return 0; } +static inline loff_t fat_i_pos_read(struct msdos_sb_info *sbi, + struct inode *inode) +{ + loff_t i_pos; +#if BITS_PER_LONG == 32 + spin_lock(&sbi->inode_hash_lock); +#endif + i_pos = MSDOS_I(inode)->i_pos; +#if BITS_PER_LONG == 32 + spin_unlock(&sbi->inode_hash_lock); +#endif + return i_pos; +} + static int fat_write_inode(struct inode *inode, int wait) { struct super_block *sb = inode->i_sb; @@ -564,9 +565,12 @@ static int fat_write_inode(struct inode *inode, int wait) loff_t i_pos; int err; + if (inode->i_ino == MSDOS_ROOT_INO) + return 0; + retry: - i_pos = MSDOS_I(inode)->i_pos; - if (inode->i_ino == MSDOS_ROOT_INO || !i_pos) + i_pos = fat_i_pos_read(sbi, inode); + if (!i_pos) return 0; bh = sb_bread(sb, i_pos >> sbi->dir_per_block_bits); @@ -588,19 +592,17 @@ retry: raw_entry->size = 0; else raw_entry->size = cpu_to_le32(inode->i_size); - raw_entry->attr = fat_attr(inode); + raw_entry->attr = fat_make_attrs(inode); raw_entry->start = cpu_to_le16(MSDOS_I(inode)->i_logstart); raw_entry->starthi = cpu_to_le16(MSDOS_I(inode)->i_logstart >> 16); - fat_date_unix2dos(inode->i_mtime.tv_sec, &raw_entry->time, - &raw_entry->date, sbi->options.tz_utc); + fat_time_unix2fat(sbi, &inode->i_mtime, &raw_entry->time, + &raw_entry->date, NULL); if (sbi->options.isvfat) { __le16 atime; - fat_date_unix2dos(inode->i_ctime.tv_sec, &raw_entry->ctime, - &raw_entry->cdate, sbi->options.tz_utc); - fat_date_unix2dos(inode->i_atime.tv_sec, &atime, - &raw_entry->adate, sbi->options.tz_utc); - raw_entry->ctime_cs = (inode->i_ctime.tv_sec & 1) * 100 + - inode->i_ctime.tv_nsec / 10000000; + fat_time_unix2fat(sbi, &inode->i_ctime, &raw_entry->ctime, + &raw_entry->cdate, &raw_entry->ctime_cs); + fat_time_unix2fat(sbi, &inode->i_atime, &atime, + &raw_entry->adate, NULL); } spin_unlock(&sbi->inode_hash_lock); mark_buffer_dirty(bh); @@ -681,33 +683,24 @@ static struct dentry *fat_fh_to_dentry(struct super_block *sb, inode = NULL; } } - if (!inode) { - /* For now, do nothing - * What we could do is: - * follow the file starting at fh[4], and record - * the ".." entry, and the name of the fh[2] entry. - * The follow the ".." file finding the next step up. - * This way we build a path to the root of - * the tree. If this works, we lookup the path and so - * get this inode into the cache. - * Finally try the fat_iget lookup again - * If that fails, then weare totally out of luck - * But all that is for another day - */ - } - if (!inode) - return ERR_PTR(-ESTALE); - - /* now to find a dentry. - * If possible, get a well-connected one + /* + * For now, do nothing if the inode is not found. + * + * What we could do is: + * + * - follow the file starting at fh[4], and record the ".." entry, + * and the name of the fh[2] entry. + * - then follow the ".." file finding the next step up. + * + * This way we build a path to the root of the tree. If this works, we + * lookup the path and so get this inode into the cache. Finally try + * the fat_iget lookup again. If that fails, then we are totally out + * of luck. But all that is for another day */ - result = d_alloc_anon(inode); - if (result == NULL) { - iput(inode); - return ERR_PTR(-ENOMEM); - } - result->d_op = sb->s_root->d_op; + result = d_obtain_alias(inode); + if (!IS_ERR(result)) + result->d_op = sb->s_root->d_op; return result; } @@ -754,15 +747,8 @@ static struct dentry *fat_get_parent(struct dentry *child) } inode = fat_build_inode(sb, de, i_pos); brelse(bh); - if (IS_ERR(inode)) { - parent = ERR_CAST(inode); - goto out; - } - parent = d_alloc_anon(inode); - if (!parent) { - iput(inode); - parent = ERR_PTR(-ENOMEM); - } + + parent = d_obtain_alias(inode); out: unlock_super(sb); @@ -835,8 +821,10 @@ static int fat_show_options(struct seq_file *m, struct vfsmount *mnt) seq_puts(m, ",uni_xlate"); if (!opts->numtail) seq_puts(m, ",nonumtail"); + if (opts->rodir) + seq_puts(m, ",rodir"); } - if (sbi->options.flush) + if (opts->flush) seq_puts(m, ",flush"); if (opts->tz_utc) seq_puts(m, ",tz=UTC"); @@ -852,7 +840,7 @@ enum { Opt_charset, Opt_shortname_lower, Opt_shortname_win95, Opt_shortname_winnt, Opt_shortname_mixed, Opt_utf8_no, Opt_utf8_yes, Opt_uni_xl_no, Opt_uni_xl_yes, Opt_nonumtail_no, Opt_nonumtail_yes, - Opt_obsolate, Opt_flush, Opt_tz_utc, Opt_err, + Opt_obsolate, Opt_flush, Opt_tz_utc, Opt_rodir, Opt_err, }; static const match_table_t fat_tokens = { @@ -924,6 +912,7 @@ static const match_table_t vfat_tokens = { {Opt_nonumtail_yes, "nonumtail=yes"}, {Opt_nonumtail_yes, "nonumtail=true"}, {Opt_nonumtail_yes, "nonumtail"}, + {Opt_rodir, "rodir"}, {Opt_err, NULL} }; @@ -937,16 +926,19 @@ static int parse_options(char *options, int is_vfat, int silent, int *debug, opts->isvfat = is_vfat; - opts->fs_uid = current->uid; - opts->fs_gid = current->gid; + opts->fs_uid = current_uid(); + opts->fs_gid = current_gid(); opts->fs_fmask = opts->fs_dmask = current->fs->umask; opts->allow_utime = -1; opts->codepage = fat_default_codepage; opts->iocharset = fat_default_iocharset; - if (is_vfat) + if (is_vfat) { opts->shortname = VFAT_SFN_DISPLAY_LOWER|VFAT_SFN_CREATE_WIN95; - else + opts->rodir = 0; + } else { opts->shortname = 0; + opts->rodir = 1; + } opts->name_check = 'n'; opts->quiet = opts->showexec = opts->sys_immutable = opts->dotsOK = 0; opts->utf8 = opts->unicode_xlate = 0; @@ -1097,6 +1089,9 @@ static int parse_options(char *options, int is_vfat, int silent, int *debug, case Opt_nonumtail_yes: /* empty or 1 or yes or true */ opts->numtail = 0; /* negated option */ break; + case Opt_rodir: + opts->rodir = 1; + break; /* obsolete mount options */ case Opt_obsolate: @@ -1142,7 +1137,7 @@ static int fat_read_root(struct inode *inode) inode->i_gid = sbi->options.fs_gid; inode->i_version++; inode->i_generation = 0; - inode->i_mode = (S_IRWXUGO & ~sbi->options.fs_dmask) | S_IFDIR; + inode->i_mode = fat_make_mode(sbi, ATTR_DIR, S_IRWXUGO); inode->i_op = sbi->dir_ops; inode->i_fop = &fat_dir_operations; if (sbi->fat_bits == 32) { @@ -1159,7 +1154,7 @@ static int fat_read_root(struct inode *inode) MSDOS_I(inode)->i_logstart = 0; MSDOS_I(inode)->mmu_private = inode->i_size; - MSDOS_I(inode)->i_attrs = ATTR_NONE; + fat_save_attrs(inode, ATTR_DIR); inode->i_mtime.tv_sec = inode->i_atime.tv_sec = inode->i_ctime.tv_sec = 0; inode->i_mtime.tv_nsec = inode->i_atime.tv_nsec = inode->i_ctime.tv_nsec = 0; inode->i_nlink = fat_subdirs(inode)+2; diff --git a/fs/fat/misc.c b/fs/fat/misc.c index 79fb98ad36d..ac39ebcc149 100644 --- a/fs/fat/misc.c +++ b/fs/fat/misc.c @@ -8,8 +8,8 @@ #include <linux/module.h> #include <linux/fs.h> -#include <linux/msdos_fs.h> #include <linux/buffer_head.h> +#include "fat.h" /* * fat_fs_panic reports a severe file system problem and sets the file system @@ -124,8 +124,9 @@ int fat_chain_add(struct inode *inode, int new_dclus, int nr_cluster) mark_inode_dirty(inode); } if (new_fclus != (inode->i_blocks >> (sbi->cluster_bits - 9))) { - fat_fs_panic(sb, "clusters badly computed (%d != %lu)", - new_fclus, inode->i_blocks >> (sbi->cluster_bits - 9)); + fat_fs_panic(sb, "clusters badly computed (%d != %llu)", + new_fclus, + (llu)(inode->i_blocks >> (sbi->cluster_bits - 9))); fat_cache_inval_inode(inode); } inode->i_blocks += nr_cluster << (sbi->cluster_bits - 9); @@ -135,65 +136,131 @@ int fat_chain_add(struct inode *inode, int new_dclus, int nr_cluster) extern struct timezone sys_tz; +/* + * The epoch of FAT timestamp is 1980. + * : bits : value + * date: 0 - 4: day (1 - 31) + * date: 5 - 8: month (1 - 12) + * date: 9 - 15: year (0 - 127) from 1980 + * time: 0 - 4: sec (0 - 29) 2sec counts + * time: 5 - 10: min (0 - 59) + * time: 11 - 15: hour (0 - 23) + */ +#define SECS_PER_MIN 60 +#define SECS_PER_HOUR (60 * 60) +#define SECS_PER_DAY (SECS_PER_HOUR * 24) +#define UNIX_SECS_1980 315532800L +#if BITS_PER_LONG == 64 +#define UNIX_SECS_2108 4354819200L +#endif +/* days between 1.1.70 and 1.1.80 (2 leap days) */ +#define DAYS_DELTA (365 * 10 + 2) +/* 120 (2100 - 1980) isn't leap year */ +#define YEAR_2100 120 +#define IS_LEAP_YEAR(y) (!((y) & 3) && (y) != YEAR_2100) + /* Linear day numbers of the respective 1sts in non-leap years. */ -static int day_n[] = { - /* Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec */ - 0, 31, 59, 90, 120, 151, 181, 212, 243, 273, 304, 334, 0, 0, 0, 0 +static time_t days_in_year[] = { + /* Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec */ + 0, 0, 31, 59, 90, 120, 151, 181, 212, 243, 273, 304, 334, 0, 0, 0, }; -/* Convert a MS-DOS time/date pair to a UNIX date (seconds since 1 1 70). */ -int date_dos2unix(unsigned short time, unsigned short date, int tz_utc) +/* Convert a FAT time/date pair to a UNIX date (seconds since 1 1 70). */ +void fat_time_fat2unix(struct msdos_sb_info *sbi, struct timespec *ts, + __le16 __time, __le16 __date, u8 time_cs) { - int month, year, secs; + u16 time = le16_to_cpu(__time), date = le16_to_cpu(__date); + time_t second, day, leap_day, month, year; - /* - * first subtract and mask after that... Otherwise, if - * date == 0, bad things happen - */ - month = ((date >> 5) - 1) & 15; - year = date >> 9; - secs = (time & 31)*2+60*((time >> 5) & 63)+(time >> 11)*3600+86400* - ((date & 31)-1+day_n[month]+(year/4)+year*365-((year & 3) == 0 && - month < 2 ? 1 : 0)+3653); - /* days since 1.1.70 plus 80's leap day */ - if (!tz_utc) - secs += sys_tz.tz_minuteswest*60; - return secs; + year = date >> 9; + month = max(1, (date >> 5) & 0xf); + day = max(1, date & 0x1f) - 1; + + leap_day = (year + 3) / 4; + if (year > YEAR_2100) /* 2100 isn't leap year */ + leap_day--; + if (IS_LEAP_YEAR(year) && month > 2) + leap_day++; + + second = (time & 0x1f) << 1; + second += ((time >> 5) & 0x3f) * SECS_PER_MIN; + second += (time >> 11) * SECS_PER_HOUR; + second += (year * 365 + leap_day + + days_in_year[month] + day + + DAYS_DELTA) * SECS_PER_DAY; + + if (!sbi->options.tz_utc) + second += sys_tz.tz_minuteswest * SECS_PER_MIN; + + if (time_cs) { + ts->tv_sec = second + (time_cs / 100); + ts->tv_nsec = (time_cs % 100) * 10000000; + } else { + ts->tv_sec = second; + ts->tv_nsec = 0; + } } -/* Convert linear UNIX date to a MS-DOS time/date pair. */ -void fat_date_unix2dos(int unix_date, __le16 *time, __le16 *date, int tz_utc) +/* Convert linear UNIX date to a FAT time/date pair. */ +void fat_time_unix2fat(struct msdos_sb_info *sbi, struct timespec *ts, + __le16 *time, __le16 *date, u8 *time_cs) { - int day, year, nl_day, month; + time_t second = ts->tv_sec; + time_t day, leap_day, month, year; - if (!tz_utc) - unix_date -= sys_tz.tz_minuteswest*60; + if (!sbi->options.tz_utc) + second -= sys_tz.tz_minuteswest * SECS_PER_MIN; /* Jan 1 GMT 00:00:00 1980. But what about another time zone? */ - if (unix_date < 315532800) - unix_date = 315532800; - - *time = cpu_to_le16((unix_date % 60)/2+(((unix_date/60) % 60) << 5)+ - (((unix_date/3600) % 24) << 11)); - day = unix_date/86400-3652; - year = day/365; - if ((year+3)/4+365*year > day) + if (second < UNIX_SECS_1980) { + *time = 0; + *date = cpu_to_le16((0 << 9) | (1 << 5) | 1); + if (time_cs) + *time_cs = 0; + return; + } +#if BITS_PER_LONG == 64 + if (second >= UNIX_SECS_2108) { + *time = cpu_to_le16((23 << 11) | (59 << 5) | 29); + *date = cpu_to_le16((127 << 9) | (12 << 5) | 31); + if (time_cs) + *time_cs = 199; + return; + } +#endif + + day = second / SECS_PER_DAY - DAYS_DELTA; + year = day / 365; + leap_day = (year + 3) / 4; + if (year > YEAR_2100) /* 2100 isn't leap year */ + leap_day--; + if (year * 365 + leap_day > day) year--; - day -= (year+3)/4+365*year; - if (day == 59 && !(year & 3)) { - nl_day = day; + leap_day = (year + 3) / 4; + if (year > YEAR_2100) /* 2100 isn't leap year */ + leap_day--; + day -= year * 365 + leap_day; + + if (IS_LEAP_YEAR(year) && day == days_in_year[3]) { month = 2; } else { - nl_day = (year & 3) || day <= 59 ? day : day-1; - for (month = 0; month < 12; month++) { - if (day_n[month] > nl_day) + if (IS_LEAP_YEAR(year) && day > days_in_year[3]) + day--; + for (month = 1; month < 12; month++) { + if (days_in_year[month + 1] > day) break; } } - *date = cpu_to_le16(nl_day-day_n[month-1]+1+(month << 5)+(year << 9)); -} + day -= days_in_year[month]; -EXPORT_SYMBOL_GPL(fat_date_unix2dos); + *time = cpu_to_le16(((second / SECS_PER_HOUR) % 24) << 11 + | ((second / SECS_PER_MIN) % 60) << 5 + | (second % SECS_PER_MIN) >> 1); + *date = cpu_to_le16((year << 9) | (month << 5) | (day + 1)); + if (time_cs) + *time_cs = (ts->tv_sec & 1) * 100 + ts->tv_nsec / 10000000; +} +EXPORT_SYMBOL_GPL(fat_time_unix2fat); int fat_sync_bhs(struct buffer_head **bhs, int nr_bhs) { diff --git a/fs/msdos/namei.c b/fs/fat/namei_msdos.c index e844b9809d2..7ba03a4acbe 100644 --- a/fs/msdos/namei.c +++ b/fs/fat/namei_msdos.c @@ -9,8 +9,8 @@ #include <linux/module.h> #include <linux/time.h> #include <linux/buffer_head.h> -#include <linux/msdos_fs.h> #include <linux/smp_lock.h> +#include "fat.h" /* Characters that are undesirable in an MS-DOS file name */ static unsigned char bad_chars[] = "*?<>|\""; @@ -203,33 +203,37 @@ static struct dentry *msdos_lookup(struct inode *dir, struct dentry *dentry, { struct super_block *sb = dir->i_sb; struct fat_slot_info sinfo; - struct inode *inode = NULL; - int res; - - dentry->d_op = &msdos_dentry_operations; + struct inode *inode; + int err; lock_super(sb); - res = msdos_find(dir, dentry->d_name.name, dentry->d_name.len, &sinfo); - if (res == -ENOENT) - goto add; - if (res < 0) - goto out; + + err = msdos_find(dir, dentry->d_name.name, dentry->d_name.len, &sinfo); + if (err) { + if (err == -ENOENT) { + inode = NULL; + goto out; + } + goto error; + } + inode = fat_build_inode(sb, sinfo.de, sinfo.i_pos); brelse(sinfo.bh); if (IS_ERR(inode)) { - res = PTR_ERR(inode); - goto out; + err = PTR_ERR(inode); + goto error; } -add: - res = 0; +out: + unlock_super(sb); + dentry->d_op = &msdos_dentry_operations; dentry = d_splice_alias(inode, dentry); if (dentry) dentry->d_op = &msdos_dentry_operations; -out: + return dentry; + +error: unlock_super(sb); - if (!res) - return dentry; - return ERR_PTR(res); + return ERR_PTR(err); } /***** Creates a directory entry (name is already formatted). */ @@ -247,7 +251,7 @@ static int msdos_add_entry(struct inode *dir, const unsigned char *name, if (is_hid) de.attr |= ATTR_HIDDEN; de.lcase = 0; - fat_date_unix2dos(ts->tv_sec, &time, &date, sbi->options.tz_utc); + fat_time_unix2fat(sbi, ts, &time, &date, NULL); de.cdate = de.adate = 0; de.ctime = 0; de.ctime_cs = 0; diff --git a/fs/vfat/namei.c b/fs/fat/namei_vfat.c index 155c10b4adb..bf326d4356a 100644 --- a/fs/vfat/namei.c +++ b/fs/fat/namei_vfat.c @@ -16,36 +16,75 @@ */ #include <linux/module.h> - #include <linux/jiffies.h> -#include <linux/msdos_fs.h> #include <linux/ctype.h> #include <linux/slab.h> #include <linux/smp_lock.h> #include <linux/buffer_head.h> #include <linux/namei.h> +#include "fat.h" -static int vfat_revalidate(struct dentry *dentry, struct nameidata *nd) +/* + * If new entry was created in the parent, it could create the 8.3 + * alias (the shortname of logname). So, the parent may have the + * negative-dentry which matches the created 8.3 alias. + * + * If it happened, the negative dentry isn't actually negative + * anymore. So, drop it. + */ +static int vfat_revalidate_shortname(struct dentry *dentry) { int ret = 1; - - if (!dentry->d_inode && - nd && !(nd->flags & LOOKUP_CONTINUE) && (nd->flags & LOOKUP_CREATE)) - /* - * negative dentry is dropped, in order to make sure - * to use the name which a user desires if this is - * create path. - */ + spin_lock(&dentry->d_lock); + if (dentry->d_time != dentry->d_parent->d_inode->i_version) ret = 0; - else { - spin_lock(&dentry->d_lock); - if (dentry->d_time != dentry->d_parent->d_inode->i_version) - ret = 0; - spin_unlock(&dentry->d_lock); - } + spin_unlock(&dentry->d_lock); return ret; } +static int vfat_revalidate(struct dentry *dentry, struct nameidata *nd) +{ + /* This is not negative dentry. Always valid. */ + if (dentry->d_inode) + return 1; + return vfat_revalidate_shortname(dentry); +} + +static int vfat_revalidate_ci(struct dentry *dentry, struct nameidata *nd) +{ + /* + * This is not negative dentry. Always valid. + * + * Note, rename() to existing directory entry will have ->d_inode, + * and will use existing name which isn't specified name by user. + * + * We may be able to drop this positive dentry here. But dropping + * positive dentry isn't good idea. So it's unsupported like + * rename("filename", "FILENAME") for now. + */ + if (dentry->d_inode) + return 1; + + /* + * This may be nfsd (or something), anyway, we can't see the + * intent of this. So, since this can be for creation, drop it. + */ + if (!nd) + return 0; + + /* + * Drop the negative dentry, in order to make sure to use the + * case sensitive name which is specified by user if this is + * for creation. + */ + if (!(nd->flags & (LOOKUP_CONTINUE | LOOKUP_PARENT))) { + if (nd->flags & LOOKUP_CREATE) + return 0; + } + + return vfat_revalidate_shortname(dentry); +} + /* returns the length of a struct qstr, ignoring trailing dots */ static unsigned int vfat_striptail_len(struct qstr *qstr) { @@ -127,25 +166,16 @@ static int vfat_cmp(struct dentry *dentry, struct qstr *a, struct qstr *b) return 1; } -static struct dentry_operations vfat_dentry_ops[4] = { - { - .d_hash = vfat_hashi, - .d_compare = vfat_cmpi, - }, - { - .d_revalidate = vfat_revalidate, - .d_hash = vfat_hashi, - .d_compare = vfat_cmpi, - }, - { - .d_hash = vfat_hash, - .d_compare = vfat_cmp, - }, - { - .d_revalidate = vfat_revalidate, - .d_hash = vfat_hash, - .d_compare = vfat_cmp, - } +static struct dentry_operations vfat_ci_dentry_ops = { + .d_revalidate = vfat_revalidate_ci, + .d_hash = vfat_hashi, + .d_compare = vfat_cmpi, +}; + +static struct dentry_operations vfat_dentry_ops = { + .d_revalidate = vfat_revalidate, + .d_hash = vfat_hash, + .d_compare = vfat_cmp, }; /* Characters that are undesirable in an MS-DOS file name */ @@ -569,6 +599,7 @@ static int vfat_build_slots(struct inode *dir, const unsigned char *name, unsigned char msdos_name[MSDOS_NAME]; wchar_t *uname; __le16 time, date; + u8 time_cs; int err, ulen, usize, i; loff_t offset; @@ -621,10 +652,10 @@ shortname: memcpy(de->name, msdos_name, MSDOS_NAME); de->attr = is_dir ? ATTR_DIR : ATTR_ARCH; de->lcase = lcase; - fat_date_unix2dos(ts->tv_sec, &time, &date, sbi->options.tz_utc); + fat_time_unix2fat(sbi, ts, &time, &date, &time_cs); de->time = de->ctime = time; de->date = de->cdate = de->adate = date; - de->ctime_cs = 0; + de->ctime_cs = time_cs; de->start = cpu_to_le16(cluster); de->starthi = cpu_to_le16(cluster >> 16); de->size = 0; @@ -683,46 +714,58 @@ static struct dentry *vfat_lookup(struct inode *dir, struct dentry *dentry, { struct super_block *sb = dir->i_sb; struct fat_slot_info sinfo; - struct inode *inode = NULL; + struct inode *inode; struct dentry *alias; - int err, table; + int err; lock_super(sb); - table = (MSDOS_SB(sb)->options.name_check == 's') ? 2 : 0; - dentry->d_op = &vfat_dentry_ops[table]; err = vfat_find(dir, &dentry->d_name, &sinfo); if (err) { - table++; + if (err == -ENOENT) { + inode = NULL; + goto out; + } goto error; } + inode = fat_build_inode(sb, sinfo.de, sinfo.i_pos); brelse(sinfo.bh); if (IS_ERR(inode)) { - unlock_super(sb); - return ERR_CAST(inode); + err = PTR_ERR(inode); + goto error; } - alias = d_find_alias(inode); - if (alias) { - if (d_invalidate(alias) == 0) - dput(alias); - else { - iput(inode); - unlock_super(sb); - return alias; - } + alias = d_find_alias(inode); + if (alias && !(alias->d_flags & DCACHE_DISCONNECTED)) { + /* + * This inode has non DCACHE_DISCONNECTED dentry. This + * means, the user did ->lookup() by an another name + * (longname vs 8.3 alias of it) in past. + * + * Switch to new one for reason of locality if possible. + */ + BUG_ON(d_unhashed(alias)); + if (!S_ISDIR(inode->i_mode)) + d_move(alias, dentry); + iput(inode); + unlock_super(sb); + return alias; } -error: +out: unlock_super(sb); - dentry->d_op = &vfat_dentry_ops[table]; + dentry->d_op = sb->s_root->d_op; dentry->d_time = dentry->d_parent->d_inode->i_version; dentry = d_splice_alias(inode, dentry); if (dentry) { - dentry->d_op = &vfat_dentry_ops[table]; + dentry->d_op = sb->s_root->d_op; dentry->d_time = dentry->d_parent->d_inode->i_version; } return dentry; + +error: + unlock_super(sb); + return ERR_PTR(err); } static int vfat_create(struct inode *dir, struct dentry *dentry, int mode, @@ -1014,9 +1057,9 @@ static int vfat_fill_super(struct super_block *sb, void *data, int silent) return res; if (MSDOS_SB(sb)->options.name_check != 's') - sb->s_root->d_op = &vfat_dentry_ops[0]; + sb->s_root->d_op = &vfat_ci_dentry_ops; else - sb->s_root->d_op = &vfat_dentry_ops[2]; + sb->s_root->d_op = &vfat_dentry_ops; return 0; } diff --git a/fs/fcntl.c b/fs/fcntl.c index ac4f7db9f13..cdc14194672 100644 --- a/fs/fcntl.c +++ b/fs/fcntl.c @@ -19,6 +19,7 @@ #include <linux/signal.h> #include <linux/rcupdate.h> #include <linux/pid_namespace.h> +#include <linux/smp_lock.h> #include <asm/poll.h> #include <asm/siginfo.h> @@ -175,6 +176,11 @@ static int setfl(int fd, struct file * filp, unsigned long arg) if (error) return error; + /* + * We still need a lock here for now to keep multiple FASYNC calls + * from racing with each other. + */ + lock_kernel(); if ((arg ^ filp->f_flags) & FASYNC) { if (filp->f_op && filp->f_op->fasync) { error = filp->f_op->fasync(fd, filp, (arg & FASYNC) != 0); @@ -185,6 +191,7 @@ static int setfl(int fd, struct file * filp, unsigned long arg) filp->f_flags = (arg & SETFL_MASK) | (filp->f_flags & ~SETFL_MASK); out: + unlock_kernel(); return error; } @@ -205,13 +212,14 @@ static void f_modown(struct file *filp, struct pid *pid, enum pid_type type, int __f_setown(struct file *filp, struct pid *pid, enum pid_type type, int force) { + const struct cred *cred = current_cred(); int err; err = security_file_set_fowner(filp); if (err) return err; - f_modown(filp, pid, type, current->uid, current->euid, force); + f_modown(filp, pid, type, cred->uid, cred->euid, force); return 0; } EXPORT_SYMBOL(__f_setown); @@ -400,10 +408,17 @@ static const long band_table[NSIGPOLL] = { static inline int sigio_perm(struct task_struct *p, struct fown_struct *fown, int sig) { - return (((fown->euid == 0) || - (fown->euid == p->suid) || (fown->euid == p->uid) || - (fown->uid == p->suid) || (fown->uid == p->uid)) && - !security_file_send_sigiotask(p, fown, sig)); + const struct cred *cred; + int ret; + + rcu_read_lock(); + cred = __task_cred(p); + ret = ((fown->euid == 0 || + fown->euid == cred->suid || fown->euid == cred->uid || + fown->uid == cred->suid || fown->uid == cred->uid) && + !security_file_send_sigiotask(p, fown, sig)); + rcu_read_unlock(); + return ret; } static void send_sigio_to_task(struct task_struct *p, diff --git a/fs/fifo.c b/fs/fifo.c index 987bf941149..f8f97b8b6d4 100644 --- a/fs/fifo.c +++ b/fs/fifo.c @@ -51,7 +51,7 @@ static int fifo_open(struct inode *inode, struct file *filp) filp->f_mode &= (FMODE_READ | FMODE_WRITE); switch (filp->f_mode) { - case 1: + case FMODE_READ: /* * O_RDONLY * POSIX.1 says that O_NONBLOCK means return with the FIFO @@ -76,7 +76,7 @@ static int fifo_open(struct inode *inode, struct file *filp) } break; - case 2: + case FMODE_WRITE: /* * O_WRONLY * POSIX.1 says that O_NONBLOCK means return -1 with @@ -98,7 +98,7 @@ static int fifo_open(struct inode *inode, struct file *filp) } break; - case 3: + case FMODE_READ | FMODE_WRITE: /* * O_RDWR * POSIX.1 leaves this case "undefined" when O_NONBLOCK is set. diff --git a/fs/file_table.c b/fs/file_table.c index f45a4493f9e..0fbcacc3ea7 100644 --- a/fs/file_table.c +++ b/fs/file_table.c @@ -36,7 +36,9 @@ static struct percpu_counter nr_files __cacheline_aligned_in_smp; static inline void file_free_rcu(struct rcu_head *head) { - struct file *f = container_of(head, struct file, f_u.fu_rcuhead); + struct file *f = container_of(head, struct file, f_u.fu_rcuhead); + + put_cred(f->f_cred); kmem_cache_free(filp_cachep, f); } @@ -94,7 +96,7 @@ int proc_nr_files(ctl_table *table, int write, struct file *filp, */ struct file *get_empty_filp(void) { - struct task_struct *tsk; + const struct cred *cred = current_cred(); static int old_max; struct file * f; @@ -118,12 +120,10 @@ struct file *get_empty_filp(void) if (security_file_alloc(f)) goto fail_sec; - tsk = current; INIT_LIST_HEAD(&f->f_u.fu_list); atomic_long_set(&f->f_count, 1); rwlock_init(&f->f_owner.lock); - f->f_uid = tsk->fsuid; - f->f_gid = tsk->fsgid; + f->f_cred = get_cred(cred); eventpoll_init_file(f); /* f->f_version: 0 */ return f; @@ -161,7 +161,7 @@ EXPORT_SYMBOL(get_empty_filp); * code should be moved into this function. */ struct file *alloc_file(struct vfsmount *mnt, struct dentry *dentry, - mode_t mode, const struct file_operations *fop) + fmode_t mode, const struct file_operations *fop) { struct file *file; struct path; @@ -193,7 +193,7 @@ EXPORT_SYMBOL(alloc_file); * of this should be moving to alloc_file(). */ int init_file(struct file *file, struct vfsmount *mnt, struct dentry *dentry, - mode_t mode, const struct file_operations *fop) + fmode_t mode, const struct file_operations *fop) { int error = 0; file->f_path.dentry = dentry; @@ -269,6 +269,10 @@ void __fput(struct file *file) eventpoll_release(file); locks_remove_flock(file); + if (unlikely(file->f_flags & FASYNC)) { + if (file->f_op && file->f_op->fasync) + file->f_op->fasync(-1, file, 0); + } if (file->f_op && file->f_op->release) file->f_op->release(inode, file); security_file_free(file); diff --git a/fs/filesystems.c b/fs/filesystems.c index f37f8726283..d0e20ced62d 100644 --- a/fs/filesystems.c +++ b/fs/filesystems.c @@ -8,6 +8,8 @@ #include <linux/syscalls.h> #include <linux/fs.h> +#include <linux/proc_fs.h> +#include <linux/seq_file.h> #include <linux/slab.h> #include <linux/kmod.h> #include <linux/init.h> @@ -214,6 +216,43 @@ int get_filesystem_list(char * buf) return len; } +#ifdef CONFIG_PROC_FS +static int filesystems_proc_show(struct seq_file *m, void *v) +{ + struct file_system_type * tmp; + + read_lock(&file_systems_lock); + tmp = file_systems; + while (tmp) { + seq_printf(m, "%s\t%s\n", + (tmp->fs_flags & FS_REQUIRES_DEV) ? "" : "nodev", + tmp->name); + tmp = tmp->next; + } + read_unlock(&file_systems_lock); + return 0; +} + +static int filesystems_proc_open(struct inode *inode, struct file *file) +{ + return single_open(file, filesystems_proc_show, NULL); +} + +static const struct file_operations filesystems_proc_fops = { + .open = filesystems_proc_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, +}; + +static int __init proc_filesystems_init(void) +{ + proc_create("filesystems", 0, NULL, &filesystems_proc_fops); + return 0; +} +module_init(proc_filesystems_init); +#endif + struct file_system_type *get_fs_type(const char *name) { struct file_system_type *fs; diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c index 25adfc3c693..d0ff0b8cf30 100644 --- a/fs/fs-writeback.c +++ b/fs/fs-writeback.c @@ -8,7 +8,7 @@ * pages against inodes. ie: data writeback. Writeout of the * inode itself is not handled here. * - * 10Apr2002 akpm@zip.com.au + * 10Apr2002 Andrew Morton * Split out of fs/inode.c * Additions for address_space-based writeback */ diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index 87250b6a868..fba571648a8 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -87,8 +87,8 @@ static void __fuse_put_request(struct fuse_req *req) static void fuse_req_init_context(struct fuse_req *req) { - req->in.h.uid = current->fsuid; - req->in.h.gid = current->fsgid; + req->in.h.uid = current_fsuid(); + req->in.h.gid = current_fsgid(); req->in.h.pid = current->pid; } @@ -1056,7 +1056,6 @@ static int fuse_dev_release(struct inode *inode, struct file *file) end_requests(fc, &fc->pending); end_requests(fc, &fc->processing); spin_unlock(&fc->lock); - fasync_helper(-1, file, 0, &fc->fasync); fuse_conn_put(fc); } diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c index fd03330cade..95bc22bdd06 100644 --- a/fs/fuse/dir.c +++ b/fs/fuse/dir.c @@ -869,18 +869,25 @@ int fuse_update_attributes(struct inode *inode, struct kstat *stat, */ int fuse_allow_task(struct fuse_conn *fc, struct task_struct *task) { + const struct cred *cred; + int ret; + if (fc->flags & FUSE_ALLOW_OTHER) return 1; - if (task->euid == fc->user_id && - task->suid == fc->user_id && - task->uid == fc->user_id && - task->egid == fc->group_id && - task->sgid == fc->group_id && - task->gid == fc->group_id) - return 1; + rcu_read_lock(); + ret = 0; + cred = __task_cred(task); + if (cred->euid == fc->user_id && + cred->suid == fc->user_id && + cred->uid == fc->user_id && + cred->egid == fc->group_id && + cred->sgid == fc->group_id && + cred->gid == fc->group_id) + ret = 1; + rcu_read_unlock(); - return 0; + return ret; } static int fuse_access(struct inode *inode, int mask) diff --git a/fs/fuse/file.c b/fs/fuse/file.c index 2bada6bbc31..34930a964b8 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -101,6 +101,8 @@ void fuse_finish_open(struct inode *inode, struct file *file, file->f_op = &fuse_direct_io_file_operations; if (!(outarg->open_flags & FOPEN_KEEP_CACHE)) invalidate_inode_pages2(inode->i_mapping); + if (outarg->open_flags & FOPEN_NONSEEKABLE) + nonseekable_open(inode, file); ff->fh = outarg->fh; file->private_data = fuse_file_get(ff); } @@ -1448,6 +1450,9 @@ static loff_t fuse_file_llseek(struct file *file, loff_t offset, int origin) mutex_lock(&inode->i_mutex); switch (origin) { case SEEK_END: + retval = fuse_update_attributes(inode, NULL, file, NULL); + if (retval) + return retval; offset += i_size_read(inode); break; case SEEK_CUR: diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h index 3a876076bdd..35accfdd747 100644 --- a/fs/fuse/fuse_i.h +++ b/fs/fuse/fuse_i.h @@ -6,6 +6,9 @@ See the file COPYING. */ +#ifndef _FS_FUSE_I_H +#define _FS_FUSE_I_H + #include <linux/fuse.h> #include <linux/fs.h> #include <linux/mount.h> @@ -655,3 +658,5 @@ void fuse_set_nowrite(struct inode *inode); void fuse_release_nowrite(struct inode *inode); u64 fuse_get_attr_version(struct fuse_conn *fc); + +#endif /* _FS_FUSE_I_H */ diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c index 6a84388cacf..2e99f34b443 100644 --- a/fs/fuse/inode.c +++ b/fs/fuse/inode.c @@ -596,12 +596,8 @@ static struct dentry *fuse_get_dentry(struct super_block *sb, if (inode->i_generation != handle->generation) goto out_iput; - entry = d_alloc_anon(inode); - err = -ENOMEM; - if (!entry) - goto out_iput; - - if (get_node_id(inode) != FUSE_ROOT_ID) { + entry = d_obtain_alias(inode); + if (!IS_ERR(entry) && get_node_id(inode) != FUSE_ROOT_ID) { entry->d_op = &fuse_dentry_operations; fuse_invalidate_entry_cache(entry); } @@ -696,17 +692,14 @@ static struct dentry *fuse_get_parent(struct dentry *child) name.name = ".."; err = fuse_lookup_name(child_inode->i_sb, get_node_id(child_inode), &name, &outarg, &inode); - if (err && err != -ENOENT) + if (err) { + if (err == -ENOENT) + return ERR_PTR(-ESTALE); return ERR_PTR(err); - if (err || !inode) - return ERR_PTR(-ESTALE); - - parent = d_alloc_anon(inode); - if (!parent) { - iput(inode); - return ERR_PTR(-ENOMEM); } - if (get_node_id(inode) != FUSE_ROOT_ID) { + + parent = d_obtain_alias(inode); + if (!IS_ERR(parent) && get_node_id(inode) != FUSE_ROOT_ID) { parent->d_op = &fuse_dentry_operations; fuse_invalidate_entry_cache(parent); } @@ -865,7 +858,7 @@ static int fuse_fill_super(struct super_block *sb, void *data, int silent) if (is_bdev) { fc->destroy_req = fuse_request_alloc(); if (!fc->destroy_req) - goto err_put_root; + goto err_free_init_req; } mutex_lock(&fuse_mutex); @@ -895,6 +888,7 @@ static int fuse_fill_super(struct super_block *sb, void *data, int silent) err_unlock: mutex_unlock(&fuse_mutex); + err_free_init_req: fuse_request_free(init_req); err_put_root: dput(root_dentry); diff --git a/fs/gfs2/inode.c b/fs/gfs2/inode.c index 7cee695fa44..d57616840e8 100644 --- a/fs/gfs2/inode.c +++ b/fs/gfs2/inode.c @@ -705,18 +705,18 @@ static void munge_mode_uid_gid(struct gfs2_inode *dip, unsigned int *mode, (dip->i_inode.i_mode & S_ISUID) && dip->i_inode.i_uid) { if (S_ISDIR(*mode)) *mode |= S_ISUID; - else if (dip->i_inode.i_uid != current->fsuid) + else if (dip->i_inode.i_uid != current_fsuid()) *mode &= ~07111; *uid = dip->i_inode.i_uid; } else - *uid = current->fsuid; + *uid = current_fsuid(); if (dip->i_inode.i_mode & S_ISGID) { if (S_ISDIR(*mode)) *mode |= S_ISGID; *gid = dip->i_inode.i_gid; } else - *gid = current->fsgid; + *gid = current_fsgid(); } static int alloc_dinode(struct gfs2_inode *dip, u64 *no_addr, u64 *generation) @@ -1124,8 +1124,8 @@ int gfs2_unlink_ok(struct gfs2_inode *dip, const struct qstr *name, return -EPERM; if ((dip->i_inode.i_mode & S_ISVTX) && - dip->i_inode.i_uid != current->fsuid && - ip->i_inode.i_uid != current->fsuid && !capable(CAP_FOWNER)) + dip->i_inode.i_uid != current_fsuid() && + ip->i_inode.i_uid != current_fsuid() && !capable(CAP_FOWNER)) return -EPERM; if (IS_APPEND(&dip->i_inode)) diff --git a/fs/gfs2/ops_export.c b/fs/gfs2/ops_export.c index 9cda8536530..bbb8c36403a 100644 --- a/fs/gfs2/ops_export.c +++ b/fs/gfs2/ops_export.c @@ -130,28 +130,17 @@ static int gfs2_get_name(struct dentry *parent, char *name, static struct dentry *gfs2_get_parent(struct dentry *child) { struct qstr dotdot; - struct inode *inode; struct dentry *dentry; - gfs2_str2qstr(&dotdot, ".."); - inode = gfs2_lookupi(child->d_inode, &dotdot, 1); - - if (!inode) - return ERR_PTR(-ENOENT); /* - * In case of an error, @inode carries the error value, and we - * have to return that as a(n invalid) pointer to dentry. + * XXX(hch): it would be a good idea to keep this around as a + * static variable. */ - if (IS_ERR(inode)) - return ERR_CAST(inode); - - dentry = d_alloc_anon(inode); - if (!dentry) { - iput(inode); - return ERR_PTR(-ENOMEM); - } + gfs2_str2qstr(&dotdot, ".."); - dentry->d_op = &gfs2_dops; + dentry = d_obtain_alias(gfs2_lookupi(child->d_inode, &dotdot, 1)); + if (!IS_ERR(dentry)) + dentry->d_op = &gfs2_dops; return dentry; } @@ -233,13 +222,9 @@ static struct dentry *gfs2_get_dentry(struct super_block *sb, gfs2_glock_dq_uninit(&i_gh); out_inode: - dentry = d_alloc_anon(inode); - if (!dentry) { - iput(inode); - return ERR_PTR(-ENOMEM); - } - - dentry->d_op = &gfs2_dops; + dentry = d_obtain_alias(inode); + if (!IS_ERR(dentry)) + dentry->d_op = &gfs2_dops; return dentry; fail_rgd: diff --git a/fs/gfs2/ops_inode.c b/fs/gfs2/ops_inode.c index 534e1e2c65c..d232991b904 100644 --- a/fs/gfs2/ops_inode.c +++ b/fs/gfs2/ops_inode.c @@ -69,7 +69,7 @@ static int gfs2_create(struct inode *dir, struct dentry *dentry, mark_inode_dirty(inode); break; } else if (PTR_ERR(inode) != -EEXIST || - (nd && (nd->intent.open.flags & O_EXCL))) { + (nd && nd->flags & LOOKUP_EXCL)) { gfs2_holder_uninit(ghs); return PTR_ERR(inode); } diff --git a/fs/hfs/catalog.c b/fs/hfs/catalog.c index ba851576ebb..6d98f116ca0 100644 --- a/fs/hfs/catalog.c +++ b/fs/hfs/catalog.c @@ -190,6 +190,10 @@ int hfs_cat_find_brec(struct super_block *sb, u32 cnid, fd->search_key->cat.ParID = rec.thread.ParID; len = fd->search_key->cat.CName.len = rec.thread.CName.len; + if (len > HFS_NAMELEN) { + printk(KERN_ERR "hfs: bad catalog namelength\n"); + return -EIO; + } memcpy(fd->search_key->cat.CName.name, rec.thread.CName.name, len); return hfs_brec_find(fd); } diff --git a/fs/hfs/inode.c b/fs/hfs/inode.c index 7e19835efa2..9435dda8f1e 100644 --- a/fs/hfs/inode.c +++ b/fs/hfs/inode.c @@ -155,8 +155,8 @@ struct inode *hfs_new_inode(struct inode *dir, struct qstr *name, int mode) hfs_cat_build_key(sb, (btree_key *)&HFS_I(inode)->cat_key, dir->i_ino, name); inode->i_ino = HFS_SB(sb)->next_id++; inode->i_mode = mode; - inode->i_uid = current->fsuid; - inode->i_gid = current->fsgid; + inode->i_uid = current_fsuid(); + inode->i_gid = current_fsgid(); inode->i_nlink = 1; inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME_SEC; HFS_I(inode)->flags = 0; @@ -511,13 +511,6 @@ void hfs_clear_inode(struct inode *inode) } } -static int hfs_permission(struct inode *inode, int mask) -{ - if (S_ISREG(inode->i_mode) && mask & MAY_EXEC) - return 0; - return generic_permission(inode, mask, NULL); -} - static int hfs_file_open(struct inode *inode, struct file *file) { if (HFS_IS_RSRC(inode)) @@ -616,7 +609,6 @@ static const struct inode_operations hfs_file_inode_operations = { .lookup = hfs_file_lookup, .truncate = hfs_file_truncate, .setattr = hfs_inode_setattr, - .permission = hfs_permission, .setxattr = hfs_setxattr, .getxattr = hfs_getxattr, .listxattr = hfs_listxattr, diff --git a/fs/hfs/super.c b/fs/hfs/super.c index 3c7c7637719..c8b5acf4b0b 100644 --- a/fs/hfs/super.c +++ b/fs/hfs/super.c @@ -210,8 +210,8 @@ static int parse_options(char *options, struct hfs_sb_info *hsb) int tmp, token; /* initialize the sb with defaults */ - hsb->s_uid = current->uid; - hsb->s_gid = current->gid; + hsb->s_uid = current_uid(); + hsb->s_gid = current_gid(); hsb->s_file_umask = 0133; hsb->s_dir_umask = 0022; hsb->s_type = hsb->s_creator = cpu_to_be32(0x3f3f3f3f); /* == '????' */ diff --git a/fs/hfsplus/bitmap.c b/fs/hfsplus/bitmap.c index d128a25b74d..ea30afc2a03 100644 --- a/fs/hfsplus/bitmap.c +++ b/fs/hfsplus/bitmap.c @@ -32,6 +32,10 @@ int hfsplus_block_allocate(struct super_block *sb, u32 size, u32 offset, u32 *ma mutex_lock(&HFSPLUS_SB(sb).alloc_file->i_mutex); mapping = HFSPLUS_SB(sb).alloc_file->i_mapping; page = read_mapping_page(mapping, offset / PAGE_CACHE_BITS, NULL); + if (IS_ERR(page)) { + start = size; + goto out; + } pptr = kmap(page); curr = pptr + (offset & (PAGE_CACHE_BITS - 1)) / 32; i = offset % 32; @@ -73,6 +77,10 @@ int hfsplus_block_allocate(struct super_block *sb, u32 size, u32 offset, u32 *ma break; page = read_mapping_page(mapping, offset / PAGE_CACHE_BITS, NULL); + if (IS_ERR(page)) { + start = size; + goto out; + } curr = pptr = kmap(page); if ((size ^ offset) / PAGE_CACHE_BITS) end = pptr + PAGE_CACHE_BITS / 32; @@ -120,6 +128,10 @@ found: offset += PAGE_CACHE_BITS; page = read_mapping_page(mapping, offset / PAGE_CACHE_BITS, NULL); + if (IS_ERR(page)) { + start = size; + goto out; + } pptr = kmap(page); curr = pptr; end = pptr + PAGE_CACHE_BITS / 32; diff --git a/fs/hfsplus/catalog.c b/fs/hfsplus/catalog.c index ba117c445e7..f6874acb2cf 100644 --- a/fs/hfsplus/catalog.c +++ b/fs/hfsplus/catalog.c @@ -168,6 +168,11 @@ int hfsplus_find_cat(struct super_block *sb, u32 cnid, return -EIO; } + if (be16_to_cpu(tmp.thread.nodeName.length) > 255) { + printk(KERN_ERR "hfs: catalog name length corrupted\n"); + return -EIO; + } + hfsplus_cat_build_key_uni(fd->search_key, be32_to_cpu(tmp.thread.parentID), &tmp.thread.nodeName); return hfs_brec_find(fd); diff --git a/fs/hfsplus/extents.c b/fs/hfsplus/extents.c index fec8f61227f..0022eec63cd 100644 --- a/fs/hfsplus/extents.c +++ b/fs/hfsplus/extents.c @@ -199,6 +199,9 @@ int hfsplus_get_block(struct inode *inode, sector_t iblock, goto done; } + if (inode->i_ino == HFSPLUS_EXT_CNID) + return -EIO; + mutex_lock(&HFSPLUS_I(inode).extents_lock); res = hfsplus_ext_read_extent(inode, ablock); if (!res) { diff --git a/fs/hfsplus/inode.c b/fs/hfsplus/inode.c index b085d64a2b6..f105ee9e1cc 100644 --- a/fs/hfsplus/inode.c +++ b/fs/hfsplus/inode.c @@ -238,22 +238,12 @@ static void hfsplus_set_perms(struct inode *inode, struct hfsplus_perm *perms) perms->dev = cpu_to_be32(HFSPLUS_I(inode).dev); } -static int hfsplus_permission(struct inode *inode, int mask) -{ - /* MAY_EXEC is also used for lookup, if no x bit is set allow lookup, - * open_exec has the same test, so it's still not executable, if a x bit - * is set fall back to standard permission check. - */ - if (S_ISREG(inode->i_mode) && mask & MAY_EXEC && !(inode->i_mode & 0111)) - return 0; - return generic_permission(inode, mask, NULL); -} - - static int hfsplus_file_open(struct inode *inode, struct file *file) { if (HFSPLUS_IS_RSRC(inode)) inode = HFSPLUS_I(inode).rsrc_inode; + if (!(file->f_flags & O_LARGEFILE) && i_size_read(inode) > MAX_NON_LFS) + return -EOVERFLOW; atomic_inc(&HFSPLUS_I(inode).opencnt); return 0; } @@ -279,7 +269,6 @@ static int hfsplus_file_release(struct inode *inode, struct file *file) static const struct inode_operations hfsplus_file_inode_operations = { .lookup = hfsplus_file_lookup, .truncate = hfsplus_file_truncate, - .permission = hfsplus_permission, .setxattr = hfsplus_setxattr, .getxattr = hfsplus_getxattr, .listxattr = hfsplus_listxattr, @@ -307,8 +296,8 @@ struct inode *hfsplus_new_inode(struct super_block *sb, int mode) inode->i_ino = HFSPLUS_SB(sb).next_cnid++; inode->i_mode = mode; - inode->i_uid = current->fsuid; - inode->i_gid = current->fsgid; + inode->i_uid = current_fsuid(); + inode->i_gid = current_fsgid(); inode->i_nlink = 1; inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME_SEC; INIT_LIST_HEAD(&HFSPLUS_I(inode).open_dir_list); diff --git a/fs/hfsplus/options.c b/fs/hfsplus/options.c index 9699c56d323..bab7f8d1bdf 100644 --- a/fs/hfsplus/options.c +++ b/fs/hfsplus/options.c @@ -49,8 +49,8 @@ void hfsplus_fill_defaults(struct hfsplus_sb_info *opts) opts->creator = HFSPLUS_DEF_CR_TYPE; opts->type = HFSPLUS_DEF_CR_TYPE; opts->umask = current->fs->umask; - opts->uid = current->uid; - opts->gid = current->gid; + opts->uid = current_uid(); + opts->gid = current_gid(); opts->part = -1; opts->session = -1; } diff --git a/fs/hfsplus/super.c b/fs/hfsplus/super.c index e834e578c93..eb74531a0a8 100644 --- a/fs/hfsplus/super.c +++ b/fs/hfsplus/super.c @@ -356,7 +356,7 @@ static int hfsplus_fill_super(struct super_block *sb, void *data, int silent) } else if (vhdr->attributes & cpu_to_be32(HFSPLUS_VOL_SOFTLOCK)) { printk(KERN_WARNING "hfs: Filesystem is marked locked, mounting read-only.\n"); sb->s_flags |= MS_RDONLY; - } else if (vhdr->attributes & cpu_to_be32(HFSPLUS_VOL_JOURNALED)) { + } else if ((vhdr->attributes & cpu_to_be32(HFSPLUS_VOL_JOURNALED)) && !(sb->s_flags & MS_RDONLY)) { printk(KERN_WARNING "hfs: write access to a journaled filesystem is not supported, " "use the force option at your own risk, mounting read-only.\n"); sb->s_flags |= MS_RDONLY; diff --git a/fs/hostfs/hostfs.h b/fs/hostfs/hostfs.h index 6ae9011b95e..2f34f8f2134 100644 --- a/fs/hostfs/hostfs.h +++ b/fs/hostfs/hostfs.h @@ -81,7 +81,7 @@ extern int do_rmdir(const char *file); extern int do_mknod(const char *file, int mode, unsigned int major, unsigned int minor); extern int link_file(const char *from, const char *to); -extern int do_readlink(char *file, char *buf, int size); +extern int hostfs_do_readlink(char *file, char *buf, int size); extern int rename_file(char *from, char *to); extern int do_statfs(char *root, long *bsize_out, long long *blocks_out, long long *bfree_out, long long *bavail_out, diff --git a/fs/hostfs/hostfs_kern.c b/fs/hostfs/hostfs_kern.c index d6ecabf4d23..3a31451ac17 100644 --- a/fs/hostfs/hostfs_kern.c +++ b/fs/hostfs/hostfs_kern.c @@ -20,7 +20,7 @@ struct hostfs_inode_info { char *host_filename; int fd; - int mode; + fmode_t mode; struct inode vfs_inode; }; @@ -168,7 +168,7 @@ static char *follow_link(char *link) if (name == NULL) goto out; - n = do_readlink(link, name, len); + n = hostfs_do_readlink(link, name, len); if (n < len) break; len *= 2; @@ -373,7 +373,8 @@ int hostfs_readdir(struct file *file, void *ent, filldir_t filldir) int hostfs_file_open(struct inode *ino, struct file *file) { char *name; - int mode = 0, r = 0, w = 0, fd; + fmode_t mode = 0; + int r = 0, w = 0, fd; mode = file->f_mode & (FMODE_READ | FMODE_WRITE); if ((mode & HOSTFS_I(ino)->mode) == mode) @@ -942,7 +943,7 @@ int hostfs_link_readpage(struct file *file, struct page *page) name = inode_name(page->mapping->host, 0); if (name == NULL) return -ENOMEM; - err = do_readlink(name, buffer, PAGE_CACHE_SIZE); + err = hostfs_do_readlink(name, buffer, PAGE_CACHE_SIZE); kfree(name); if (err == PAGE_CACHE_SIZE) err = -E2BIG; diff --git a/fs/hostfs/hostfs_user.c b/fs/hostfs/hostfs_user.c index 53fd0a67c11..b79424f9328 100644 --- a/fs/hostfs/hostfs_user.c +++ b/fs/hostfs/hostfs_user.c @@ -377,7 +377,7 @@ int link_file(const char *to, const char *from) return 0; } -int do_readlink(char *file, char *buf, int size) +int hostfs_do_readlink(char *file, char *buf, int size) { int n; diff --git a/fs/hpfs/file.c b/fs/hpfs/file.c index be8be5040e0..64ab5225920 100644 --- a/fs/hpfs/file.c +++ b/fs/hpfs/file.c @@ -143,5 +143,5 @@ const struct file_operations hpfs_file_ops = const struct inode_operations hpfs_file_iops = { .truncate = hpfs_truncate, - .setattr = hpfs_notify_change, + .setattr = hpfs_setattr, }; diff --git a/fs/hpfs/hpfs_fn.h b/fs/hpfs/hpfs_fn.h index 42ff60ccf2a..c2ea31bae31 100644 --- a/fs/hpfs/hpfs_fn.h +++ b/fs/hpfs/hpfs_fn.h @@ -275,7 +275,7 @@ void hpfs_init_inode(struct inode *); void hpfs_read_inode(struct inode *); void hpfs_write_inode(struct inode *); void hpfs_write_inode_nolock(struct inode *); -int hpfs_notify_change(struct dentry *, struct iattr *); +int hpfs_setattr(struct dentry *, struct iattr *); void hpfs_write_if_changed(struct inode *); void hpfs_delete_inode(struct inode *); diff --git a/fs/hpfs/inode.c b/fs/hpfs/inode.c index 85d3e1d9ac0..39a1bfbea31 100644 --- a/fs/hpfs/inode.c +++ b/fs/hpfs/inode.c @@ -260,19 +260,28 @@ void hpfs_write_inode_nolock(struct inode *i) brelse(bh); } -int hpfs_notify_change(struct dentry *dentry, struct iattr *attr) +int hpfs_setattr(struct dentry *dentry, struct iattr *attr) { struct inode *inode = dentry->d_inode; - int error=0; + int error = -EINVAL; + lock_kernel(); - if ( ((attr->ia_valid & ATTR_SIZE) && attr->ia_size > inode->i_size) || - (hpfs_sb(inode->i_sb)->sb_root == inode->i_ino) ) { - error = -EINVAL; - } else if ((error = inode_change_ok(inode, attr))) { - } else if ((error = inode_setattr(inode, attr))) { - } else { - hpfs_write_inode(inode); - } + if (inode->i_ino == hpfs_sb(inode->i_sb)->sb_root) + goto out_unlock; + if ((attr->ia_valid & ATTR_SIZE) && attr->ia_size > inode->i_size) + goto out_unlock; + + error = inode_change_ok(inode, attr); + if (error) + goto out_unlock; + + error = inode_setattr(inode, attr); + if (error) + goto out_unlock; + + hpfs_write_inode(inode); + + out_unlock: unlock_kernel(); return error; } diff --git a/fs/hpfs/namei.c b/fs/hpfs/namei.c index d9c59a77544..b649232dde9 100644 --- a/fs/hpfs/namei.c +++ b/fs/hpfs/namei.c @@ -92,11 +92,11 @@ static int hpfs_mkdir(struct inode *dir, struct dentry *dentry, int mode) inc_nlink(dir); insert_inode_hash(result); - if (result->i_uid != current->fsuid || - result->i_gid != current->fsgid || + if (result->i_uid != current_fsuid() || + result->i_gid != current_fsgid() || result->i_mode != (mode | S_IFDIR)) { - result->i_uid = current->fsuid; - result->i_gid = current->fsgid; + result->i_uid = current_fsuid(); + result->i_gid = current_fsgid(); result->i_mode = mode | S_IFDIR; hpfs_write_inode_nolock(result); } @@ -184,11 +184,11 @@ static int hpfs_create(struct inode *dir, struct dentry *dentry, int mode, struc insert_inode_hash(result); - if (result->i_uid != current->fsuid || - result->i_gid != current->fsgid || + if (result->i_uid != current_fsuid() || + result->i_gid != current_fsgid() || result->i_mode != (mode | S_IFREG)) { - result->i_uid = current->fsuid; - result->i_gid = current->fsgid; + result->i_uid = current_fsuid(); + result->i_gid = current_fsgid(); result->i_mode = mode | S_IFREG; hpfs_write_inode_nolock(result); } @@ -247,8 +247,8 @@ static int hpfs_mknod(struct inode *dir, struct dentry *dentry, int mode, dev_t result->i_mtime.tv_nsec = 0; result->i_atime.tv_nsec = 0; hpfs_i(result)->i_ea_size = 0; - result->i_uid = current->fsuid; - result->i_gid = current->fsgid; + result->i_uid = current_fsuid(); + result->i_gid = current_fsgid(); result->i_nlink = 1; result->i_size = 0; result->i_blocks = 1; @@ -325,8 +325,8 @@ static int hpfs_symlink(struct inode *dir, struct dentry *dentry, const char *sy result->i_atime.tv_nsec = 0; hpfs_i(result)->i_ea_size = 0; result->i_mode = S_IFLNK | 0777; - result->i_uid = current->fsuid; - result->i_gid = current->fsgid; + result->i_uid = current_fsuid(); + result->i_gid = current_fsgid(); result->i_blocks = 1; result->i_nlink = 1; result->i_size = strlen(symlink); @@ -669,5 +669,5 @@ const struct inode_operations hpfs_dir_iops = .rmdir = hpfs_rmdir, .mknod = hpfs_mknod, .rename = hpfs_rename, - .setattr = hpfs_notify_change, + .setattr = hpfs_setattr, }; diff --git a/fs/hpfs/super.c b/fs/hpfs/super.c index 29ad461d568..0d049b8919c 100644 --- a/fs/hpfs/super.c +++ b/fs/hpfs/super.c @@ -475,8 +475,8 @@ static int hpfs_fill_super(struct super_block *s, void *options, int silent) init_MUTEX(&sbi->hpfs_creation_de); - uid = current->uid; - gid = current->gid; + uid = current_uid(); + gid = current_gid(); umask = current->fs->umask; lowercase = 0; conv = CONV_BINARY; diff --git a/fs/hppfs/hppfs.c b/fs/hppfs/hppfs.c index 2b3d1828db9..b278f7f5202 100644 --- a/fs/hppfs/hppfs.c +++ b/fs/hppfs/hppfs.c @@ -426,6 +426,7 @@ static int file_mode(int fmode) static int hppfs_open(struct inode *inode, struct file *file) { + const struct cred *cred = file->f_cred; struct hppfs_private *data; struct vfsmount *proc_mnt; struct dentry *proc_dentry; @@ -446,7 +447,7 @@ static int hppfs_open(struct inode *inode, struct file *file) /* XXX This isn't closed anywhere */ data->proc_file = dentry_open(dget(proc_dentry), mntget(proc_mnt), - file_mode(file->f_mode)); + file_mode(file->f_mode), cred); err = PTR_ERR(data->proc_file); if (IS_ERR(data->proc_file)) goto out_free1; @@ -489,6 +490,7 @@ static int hppfs_open(struct inode *inode, struct file *file) static int hppfs_dir_open(struct inode *inode, struct file *file) { + const struct cred *cred = file->f_cred; struct hppfs_private *data; struct vfsmount *proc_mnt; struct dentry *proc_dentry; @@ -502,7 +504,7 @@ static int hppfs_dir_open(struct inode *inode, struct file *file) proc_dentry = HPPFS_I(inode)->proc_dentry; proc_mnt = inode->i_sb->s_fs_info; data->proc_file = dentry_open(dget(proc_dentry), mntget(proc_mnt), - file_mode(file->f_mode)); + file_mode(file->f_mode), cred); err = PTR_ERR(data->proc_file); if (IS_ERR(data->proc_file)) goto out_free; diff --git a/fs/hugetlbfs/inode.c b/fs/hugetlbfs/inode.c index 61edc701b0e..7d479ce3ace 100644 --- a/fs/hugetlbfs/inode.c +++ b/fs/hugetlbfs/inode.c @@ -551,9 +551,9 @@ static int hugetlbfs_mknod(struct inode *dir, if (S_ISDIR(mode)) mode |= S_ISGID; } else { - gid = current->fsgid; + gid = current_fsgid(); } - inode = hugetlbfs_get_inode(dir->i_sb, current->fsuid, gid, mode, dev); + inode = hugetlbfs_get_inode(dir->i_sb, current_fsuid(), gid, mode, dev); if (inode) { dir->i_ctime = dir->i_mtime = CURRENT_TIME; d_instantiate(dentry, inode); @@ -586,9 +586,9 @@ static int hugetlbfs_symlink(struct inode *dir, if (dir->i_mode & S_ISGID) gid = dir->i_gid; else - gid = current->fsgid; + gid = current_fsgid(); - inode = hugetlbfs_get_inode(dir->i_sb, current->fsuid, + inode = hugetlbfs_get_inode(dir->i_sb, current_fsuid(), gid, S_IFLNK|S_IRWXUGO, 0); if (inode) { int l = strlen(symname)+1; @@ -854,8 +854,8 @@ hugetlbfs_fill_super(struct super_block *sb, void *data, int silent) config.nr_blocks = -1; /* No limit on size by default */ config.nr_inodes = -1; /* No limit on number of inodes by default */ - config.uid = current->fsuid; - config.gid = current->fsgid; + config.uid = current_fsuid(); + config.gid = current_fsgid(); config.mode = 0755; config.hstate = &default_hstate; ret = hugetlbfs_parse_options(data, &config); @@ -951,6 +951,7 @@ struct file *hugetlb_file_setup(const char *name, size_t size) struct inode *inode; struct dentry *dentry, *root; struct qstr quick_string; + struct user_struct *user = current_user(); if (!hugetlbfs_vfsmount) return ERR_PTR(-ENOENT); @@ -958,7 +959,7 @@ struct file *hugetlb_file_setup(const char *name, size_t size) if (!can_do_hugetlb_shm()) return ERR_PTR(-EPERM); - if (!user_shm_lock(size, current->user)) + if (!user_shm_lock(size, user)) return ERR_PTR(-ENOMEM); root = hugetlbfs_vfsmount->mnt_root; @@ -970,8 +971,8 @@ struct file *hugetlb_file_setup(const char *name, size_t size) goto out_shm_unlock; error = -ENOSPC; - inode = hugetlbfs_get_inode(root->d_sb, current->fsuid, - current->fsgid, S_IFREG | S_IRWXUGO, 0); + inode = hugetlbfs_get_inode(root->d_sb, current_fsuid(), + current_fsgid(), S_IFREG | S_IRWXUGO, 0); if (!inode) goto out_dentry; @@ -998,7 +999,7 @@ out_inode: out_dentry: dput(dentry); out_shm_unlock: - user_shm_unlock(size, current->user); + user_shm_unlock(size, user); return ERR_PTR(error); } diff --git a/fs/inode.c b/fs/inode.c index 0487ddba139..098a2443196 100644 --- a/fs/inode.c +++ b/fs/inode.c @@ -108,84 +108,100 @@ static void wake_up_inode(struct inode *inode) wake_up_bit(&inode->i_state, __I_LOCK); } -static struct inode *alloc_inode(struct super_block *sb) +/** + * inode_init_always - perform inode structure intialisation + * @sb - superblock inode belongs to. + * @inode - inode to initialise + * + * These are initializations that need to be done on every inode + * allocation as the fields are not initialised by slab allocation. + */ +struct inode *inode_init_always(struct super_block *sb, struct inode *inode) { static const struct address_space_operations empty_aops; static struct inode_operations empty_iops; static const struct file_operations empty_fops; - struct inode *inode; - - if (sb->s_op->alloc_inode) - inode = sb->s_op->alloc_inode(sb); - else - inode = (struct inode *) kmem_cache_alloc(inode_cachep, GFP_KERNEL); - if (inode) { - struct address_space * const mapping = &inode->i_data; - - inode->i_sb = sb; - inode->i_blkbits = sb->s_blocksize_bits; - inode->i_flags = 0; - atomic_set(&inode->i_count, 1); - inode->i_op = &empty_iops; - inode->i_fop = &empty_fops; - inode->i_nlink = 1; - atomic_set(&inode->i_writecount, 0); - inode->i_size = 0; - inode->i_blocks = 0; - inode->i_bytes = 0; - inode->i_generation = 0; + struct address_space * const mapping = &inode->i_data; + + inode->i_sb = sb; + inode->i_blkbits = sb->s_blocksize_bits; + inode->i_flags = 0; + atomic_set(&inode->i_count, 1); + inode->i_op = &empty_iops; + inode->i_fop = &empty_fops; + inode->i_nlink = 1; + atomic_set(&inode->i_writecount, 0); + inode->i_size = 0; + inode->i_blocks = 0; + inode->i_bytes = 0; + inode->i_generation = 0; #ifdef CONFIG_QUOTA - memset(&inode->i_dquot, 0, sizeof(inode->i_dquot)); + memset(&inode->i_dquot, 0, sizeof(inode->i_dquot)); #endif - inode->i_pipe = NULL; - inode->i_bdev = NULL; - inode->i_cdev = NULL; - inode->i_rdev = 0; - inode->dirtied_when = 0; - if (security_inode_alloc(inode)) { - if (inode->i_sb->s_op->destroy_inode) - inode->i_sb->s_op->destroy_inode(inode); - else - kmem_cache_free(inode_cachep, (inode)); - return NULL; - } + inode->i_pipe = NULL; + inode->i_bdev = NULL; + inode->i_cdev = NULL; + inode->i_rdev = 0; + inode->dirtied_when = 0; + if (security_inode_alloc(inode)) { + if (inode->i_sb->s_op->destroy_inode) + inode->i_sb->s_op->destroy_inode(inode); + else + kmem_cache_free(inode_cachep, (inode)); + return NULL; + } - spin_lock_init(&inode->i_lock); - lockdep_set_class(&inode->i_lock, &sb->s_type->i_lock_key); + spin_lock_init(&inode->i_lock); + lockdep_set_class(&inode->i_lock, &sb->s_type->i_lock_key); - mutex_init(&inode->i_mutex); - lockdep_set_class(&inode->i_mutex, &sb->s_type->i_mutex_key); + mutex_init(&inode->i_mutex); + lockdep_set_class(&inode->i_mutex, &sb->s_type->i_mutex_key); - init_rwsem(&inode->i_alloc_sem); - lockdep_set_class(&inode->i_alloc_sem, &sb->s_type->i_alloc_sem_key); + init_rwsem(&inode->i_alloc_sem); + lockdep_set_class(&inode->i_alloc_sem, &sb->s_type->i_alloc_sem_key); - mapping->a_ops = &empty_aops; - mapping->host = inode; - mapping->flags = 0; - mapping_set_gfp_mask(mapping, GFP_HIGHUSER_PAGECACHE); - mapping->assoc_mapping = NULL; - mapping->backing_dev_info = &default_backing_dev_info; - mapping->writeback_index = 0; + mapping->a_ops = &empty_aops; + mapping->host = inode; + mapping->flags = 0; + mapping_set_gfp_mask(mapping, GFP_HIGHUSER_PAGECACHE); + mapping->assoc_mapping = NULL; + mapping->backing_dev_info = &default_backing_dev_info; + mapping->writeback_index = 0; - /* - * If the block_device provides a backing_dev_info for client - * inodes then use that. Otherwise the inode share the bdev's - * backing_dev_info. - */ - if (sb->s_bdev) { - struct backing_dev_info *bdi; + /* + * If the block_device provides a backing_dev_info for client + * inodes then use that. Otherwise the inode share the bdev's + * backing_dev_info. + */ + if (sb->s_bdev) { + struct backing_dev_info *bdi; - bdi = sb->s_bdev->bd_inode_backing_dev_info; - if (!bdi) - bdi = sb->s_bdev->bd_inode->i_mapping->backing_dev_info; - mapping->backing_dev_info = bdi; - } - inode->i_private = NULL; - inode->i_mapping = mapping; + bdi = sb->s_bdev->bd_inode_backing_dev_info; + if (!bdi) + bdi = sb->s_bdev->bd_inode->i_mapping->backing_dev_info; + mapping->backing_dev_info = bdi; } + inode->i_private = NULL; + inode->i_mapping = mapping; + return inode; } +EXPORT_SYMBOL(inode_init_always); + +static struct inode *alloc_inode(struct super_block *sb) +{ + struct inode *inode; + + if (sb->s_op->alloc_inode) + inode = sb->s_op->alloc_inode(sb); + else + inode = kmem_cache_alloc(inode_cachep, GFP_KERNEL); + + if (inode) + return inode_init_always(sb, inode); + return NULL; +} void destroy_inode(struct inode *inode) { @@ -196,6 +212,7 @@ void destroy_inode(struct inode *inode) else kmem_cache_free(inode_cachep, (inode)); } +EXPORT_SYMBOL(destroy_inode); /* @@ -534,6 +551,49 @@ repeat: return node ? inode : NULL; } +static unsigned long hash(struct super_block *sb, unsigned long hashval) +{ + unsigned long tmp; + + tmp = (hashval * (unsigned long)sb) ^ (GOLDEN_RATIO_PRIME + hashval) / + L1_CACHE_BYTES; + tmp = tmp ^ ((tmp ^ GOLDEN_RATIO_PRIME) >> I_HASHBITS); + return tmp & I_HASHMASK; +} + +static inline void +__inode_add_to_lists(struct super_block *sb, struct hlist_head *head, + struct inode *inode) +{ + inodes_stat.nr_inodes++; + list_add(&inode->i_list, &inode_in_use); + list_add(&inode->i_sb_list, &sb->s_inodes); + if (head) + hlist_add_head(&inode->i_hash, head); +} + +/** + * inode_add_to_lists - add a new inode to relevant lists + * @sb - superblock inode belongs to. + * @inode - inode to mark in use + * + * When an inode is allocated it needs to be accounted for, added to the in use + * list, the owning superblock and the inode hash. This needs to be done under + * the inode_lock, so export a function to do this rather than the inode lock + * itself. We calculate the hash list to add to here so it is all internal + * which requires the caller to have already set up the inode number in the + * inode to add. + */ +void inode_add_to_lists(struct super_block *sb, struct inode *inode) +{ + struct hlist_head *head = inode_hashtable + hash(sb, inode->i_ino); + + spin_lock(&inode_lock); + __inode_add_to_lists(sb, head, inode); + spin_unlock(&inode_lock); +} +EXPORT_SYMBOL_GPL(inode_add_to_lists); + /** * new_inode - obtain an inode * @sb: superblock @@ -561,9 +621,7 @@ struct inode *new_inode(struct super_block *sb) inode = alloc_inode(sb); if (inode) { spin_lock(&inode_lock); - inodes_stat.nr_inodes++; - list_add(&inode->i_list, &inode_in_use); - list_add(&inode->i_sb_list, &sb->s_inodes); + __inode_add_to_lists(sb, NULL, inode); inode->i_ino = ++last_ino; inode->i_state = 0; spin_unlock(&inode_lock); @@ -622,10 +680,7 @@ static struct inode * get_new_inode(struct super_block *sb, struct hlist_head *h if (set(inode, data)) goto set_failed; - inodes_stat.nr_inodes++; - list_add(&inode->i_list, &inode_in_use); - list_add(&inode->i_sb_list, &sb->s_inodes); - hlist_add_head(&inode->i_hash, head); + __inode_add_to_lists(sb, head, inode); inode->i_state = I_LOCK|I_NEW; spin_unlock(&inode_lock); @@ -671,10 +726,7 @@ static struct inode * get_new_inode_fast(struct super_block *sb, struct hlist_he old = find_inode_fast(sb, head, ino); if (!old) { inode->i_ino = ino; - inodes_stat.nr_inodes++; - list_add(&inode->i_list, &inode_in_use); - list_add(&inode->i_sb_list, &sb->s_inodes); - hlist_add_head(&inode->i_hash, head); + __inode_add_to_lists(sb, head, inode); inode->i_state = I_LOCK|I_NEW; spin_unlock(&inode_lock); @@ -698,16 +750,6 @@ static struct inode * get_new_inode_fast(struct super_block *sb, struct hlist_he return inode; } -static unsigned long hash(struct super_block *sb, unsigned long hashval) -{ - unsigned long tmp; - - tmp = (hashval * (unsigned long)sb) ^ (GOLDEN_RATIO_PRIME + hashval) / - L1_CACHE_BYTES; - tmp = tmp ^ ((tmp ^ GOLDEN_RATIO_PRIME) >> I_HASHBITS); - return tmp & I_HASHMASK; -} - /** * iunique - get a unique inode number * @sb: superblock @@ -1292,6 +1334,7 @@ int inode_wait(void *word) schedule(); return 0; } +EXPORT_SYMBOL(inode_wait); /* * If we try to find an inode in the inode hash while it is being diff --git a/fs/inotify.c b/fs/inotify.c index 690e72595e6..dae3f28f30d 100644 --- a/fs/inotify.c +++ b/fs/inotify.c @@ -106,6 +106,20 @@ void get_inotify_watch(struct inotify_watch *watch) } EXPORT_SYMBOL_GPL(get_inotify_watch); +int pin_inotify_watch(struct inotify_watch *watch) +{ + struct super_block *sb = watch->inode->i_sb; + spin_lock(&sb_lock); + if (sb->s_count >= S_BIAS) { + atomic_inc(&sb->s_active); + spin_unlock(&sb_lock); + atomic_inc(&watch->count); + return 1; + } + spin_unlock(&sb_lock); + return 0; +} + /** * put_inotify_watch - decrements the ref count on a given watch. cleans up * watch references if the count reaches zero. inotify_watch is freed by @@ -124,6 +138,13 @@ void put_inotify_watch(struct inotify_watch *watch) } EXPORT_SYMBOL_GPL(put_inotify_watch); +void unpin_inotify_watch(struct inotify_watch *watch) +{ + struct super_block *sb = watch->inode->i_sb; + put_inotify_watch(watch); + deactivate_super(sb); +} + /* * inotify_handle_get_wd - returns the next WD for use by the given handle * @@ -407,11 +428,13 @@ void inotify_unmount_inodes(struct list_head *list) watches = &inode->inotify_watches; list_for_each_entry_safe(watch, next_w, watches, i_list) { struct inotify_handle *ih= watch->ih; + get_inotify_watch(watch); mutex_lock(&ih->mutex); ih->in_ops->handle_event(watch, watch->wd, IN_UNMOUNT, 0, NULL, NULL); inotify_remove_watch_locked(ih, watch); mutex_unlock(&ih->mutex); + put_inotify_watch(watch); } mutex_unlock(&inode->inotify_mutex); iput(inode); @@ -479,6 +502,112 @@ void inotify_init_watch(struct inotify_watch *watch) } EXPORT_SYMBOL_GPL(inotify_init_watch); +/* + * Watch removals suck violently. To kick the watch out we need (in this + * order) inode->inotify_mutex and ih->mutex. That's fine if we have + * a hold on inode; however, for all other cases we need to make damn sure + * we don't race with umount. We can *NOT* just grab a reference to a + * watch - inotify_unmount_inodes() will happily sail past it and we'll end + * with reference to inode potentially outliving its superblock. Ideally + * we just want to grab an active reference to superblock if we can; that + * will make sure we won't go into inotify_umount_inodes() until we are + * done. Cleanup is just deactivate_super(). However, that leaves a messy + * case - what if we *are* racing with umount() and active references to + * superblock can't be acquired anymore? We can bump ->s_count, grab + * ->s_umount, which will almost certainly wait until the superblock is shut + * down and the watch in question is pining for fjords. That's fine, but + * there is a problem - we might have hit the window between ->s_active + * getting to 0 / ->s_count - below S_BIAS (i.e. the moment when superblock + * is past the point of no return and is heading for shutdown) and the + * moment when deactivate_super() acquires ->s_umount. We could just do + * drop_super() yield() and retry, but that's rather antisocial and this + * stuff is luser-triggerable. OTOH, having grabbed ->s_umount and having + * found that we'd got there first (i.e. that ->s_root is non-NULL) we know + * that we won't race with inotify_umount_inodes(). So we could grab a + * reference to watch and do the rest as above, just with drop_super() instead + * of deactivate_super(), right? Wrong. We had to drop ih->mutex before we + * could grab ->s_umount. So the watch could've been gone already. + * + * That still can be dealt with - we need to save watch->wd, do idr_find() + * and compare its result with our pointer. If they match, we either have + * the damn thing still alive or we'd lost not one but two races at once, + * the watch had been killed and a new one got created with the same ->wd + * at the same address. That couldn't have happened in inotify_destroy(), + * but inotify_rm_wd() could run into that. Still, "new one got created" + * is not a problem - we have every right to kill it or leave it alone, + * whatever's more convenient. + * + * So we can use idr_find(...) == watch && watch->inode->i_sb == sb as + * "grab it and kill it" check. If it's been our original watch, we are + * fine, if it's a newcomer - nevermind, just pretend that we'd won the + * race and kill the fscker anyway; we are safe since we know that its + * superblock won't be going away. + * + * And yes, this is far beyond mere "not very pretty"; so's the entire + * concept of inotify to start with. + */ + +/** + * pin_to_kill - pin the watch down for removal + * @ih: inotify handle + * @watch: watch to kill + * + * Called with ih->mutex held, drops it. Possible return values: + * 0 - nothing to do, it has died + * 1 - remove it, drop the reference and deactivate_super() + * 2 - remove it, drop the reference and drop_super(); we tried hard to avoid + * that variant, since it involved a lot of PITA, but that's the best that + * could've been done. + */ +static int pin_to_kill(struct inotify_handle *ih, struct inotify_watch *watch) +{ + struct super_block *sb = watch->inode->i_sb; + s32 wd = watch->wd; + + spin_lock(&sb_lock); + if (sb->s_count >= S_BIAS) { + atomic_inc(&sb->s_active); + spin_unlock(&sb_lock); + get_inotify_watch(watch); + mutex_unlock(&ih->mutex); + return 1; /* the best outcome */ + } + sb->s_count++; + spin_unlock(&sb_lock); + mutex_unlock(&ih->mutex); /* can't grab ->s_umount under it */ + down_read(&sb->s_umount); + if (likely(!sb->s_root)) { + /* fs is already shut down; the watch is dead */ + drop_super(sb); + return 0; + } + /* raced with the final deactivate_super() */ + mutex_lock(&ih->mutex); + if (idr_find(&ih->idr, wd) != watch || watch->inode->i_sb != sb) { + /* the watch is dead */ + mutex_unlock(&ih->mutex); + drop_super(sb); + return 0; + } + /* still alive or freed and reused with the same sb and wd; kill */ + get_inotify_watch(watch); + mutex_unlock(&ih->mutex); + return 2; +} + +static void unpin_and_kill(struct inotify_watch *watch, int how) +{ + struct super_block *sb = watch->inode->i_sb; + put_inotify_watch(watch); + switch (how) { + case 1: + deactivate_super(sb); + break; + case 2: + drop_super(sb); + } +} + /** * inotify_destroy - clean up and destroy an inotify instance * @ih: inotify handle @@ -490,11 +619,15 @@ void inotify_destroy(struct inotify_handle *ih) * pretty. We cannot do a simple iteration over the list, because we * do not know the inode until we iterate to the watch. But we need to * hold inode->inotify_mutex before ih->mutex. The following works. + * + * AV: it had to become even uglier to start working ;-/ */ while (1) { struct inotify_watch *watch; struct list_head *watches; + struct super_block *sb; struct inode *inode; + int how; mutex_lock(&ih->mutex); watches = &ih->watches; @@ -503,8 +636,10 @@ void inotify_destroy(struct inotify_handle *ih) break; } watch = list_first_entry(watches, struct inotify_watch, h_list); - get_inotify_watch(watch); - mutex_unlock(&ih->mutex); + sb = watch->inode->i_sb; + how = pin_to_kill(ih, watch); + if (!how) + continue; inode = watch->inode; mutex_lock(&inode->inotify_mutex); @@ -518,7 +653,7 @@ void inotify_destroy(struct inotify_handle *ih) mutex_unlock(&ih->mutex); mutex_unlock(&inode->inotify_mutex); - put_inotify_watch(watch); + unpin_and_kill(watch, how); } /* free this handle: the put matching the get in inotify_init() */ @@ -719,7 +854,9 @@ void inotify_evict_watch(struct inotify_watch *watch) int inotify_rm_wd(struct inotify_handle *ih, u32 wd) { struct inotify_watch *watch; + struct super_block *sb; struct inode *inode; + int how; mutex_lock(&ih->mutex); watch = idr_find(&ih->idr, wd); @@ -727,9 +864,12 @@ int inotify_rm_wd(struct inotify_handle *ih, u32 wd) mutex_unlock(&ih->mutex); return -EINVAL; } - get_inotify_watch(watch); + sb = watch->inode->i_sb; + how = pin_to_kill(ih, watch); + if (!how) + return 0; + inode = watch->inode; - mutex_unlock(&ih->mutex); mutex_lock(&inode->inotify_mutex); mutex_lock(&ih->mutex); @@ -740,7 +880,7 @@ int inotify_rm_wd(struct inotify_handle *ih, u32 wd) mutex_unlock(&ih->mutex); mutex_unlock(&inode->inotify_mutex); - put_inotify_watch(watch); + unpin_and_kill(watch, how); return 0; } diff --git a/fs/inotify_user.c b/fs/inotify_user.c index d85c7d931cd..e2425bbd871 100644 --- a/fs/inotify_user.c +++ b/fs/inotify_user.c @@ -537,9 +537,6 @@ static int inotify_release(struct inode *ignored, struct file *file) inotify_dev_event_dequeue(dev); mutex_unlock(&dev->ev_mutex); - if (file->f_flags & FASYNC) - inotify_fasync(-1, file, 0); - /* free this device: the put matching the get in inotify_init() */ put_inotify_dev(dev); @@ -604,7 +601,7 @@ asmlinkage long sys_inotify_init1(int flags) goto out_put_fd; } - user = get_uid(current->user); + user = get_current_user(); if (unlikely(atomic_read(&user->inotify_devs) >= inotify_max_user_instances)) { ret = -EMFILE; diff --git a/fs/internal.h b/fs/internal.h index 80aa9a02337..53af885f173 100644 --- a/fs/internal.h +++ b/fs/internal.h @@ -10,6 +10,7 @@ */ struct super_block; +struct linux_binprm; /* * block_dev.c @@ -40,6 +41,11 @@ static inline int sb_is_blkdev_sb(struct super_block *sb) extern void __init chrdev_init(void); /* + * exec.c + */ +extern void check_unsafe_exec(struct linux_binprm *); + +/* * namespace.c */ extern int copy_mount_options(const void __user *, unsigned long *); diff --git a/fs/ioctl.c b/fs/ioctl.c index d152856c371..43e8b2c0664 100644 --- a/fs/ioctl.c +++ b/fs/ioctl.c @@ -400,11 +400,9 @@ static int ioctl_fioasync(unsigned int fd, struct file *filp, /* Did FASYNC state change ? */ if ((flag ^ filp->f_flags) & FASYNC) { - if (filp->f_op && filp->f_op->fasync) { - lock_kernel(); + if (filp->f_op && filp->f_op->fasync) error = filp->f_op->fasync(fd, filp, on); - unlock_kernel(); - } else + else error = -ENOTTY; } if (error) @@ -440,11 +438,17 @@ int do_vfs_ioctl(struct file *filp, unsigned int fd, unsigned int cmd, break; case FIONBIO: + /* BKL needed to avoid races tweaking f_flags */ + lock_kernel(); error = ioctl_fionbio(filp, argp); + unlock_kernel(); break; case FIOASYNC: + /* BKL needed to avoid races tweaking f_flags */ + lock_kernel(); error = ioctl_fioasync(fd, filp, argp); + unlock_kernel(); break; case FIOQSIZE: diff --git a/fs/ioprio.c b/fs/ioprio.c index da3cc460d4d..3569e0ad86a 100644 --- a/fs/ioprio.c +++ b/fs/ioprio.c @@ -31,10 +31,16 @@ static int set_task_ioprio(struct task_struct *task, int ioprio) { int err; struct io_context *ioc; + const struct cred *cred = current_cred(), *tcred; - if (task->uid != current->euid && - task->uid != current->uid && !capable(CAP_SYS_NICE)) + rcu_read_lock(); + tcred = __task_cred(task); + if (tcred->uid != cred->euid && + tcred->uid != cred->uid && !capable(CAP_SYS_NICE)) { + rcu_read_unlock(); return -EPERM; + } + rcu_read_unlock(); err = security_task_setioprio(task, ioprio); if (err) @@ -123,7 +129,7 @@ asmlinkage long sys_ioprio_set(int which, int who, int ioprio) break; case IOPRIO_WHO_USER: if (!who) - user = current->user; + user = current_user(); else user = find_user(who); @@ -131,7 +137,7 @@ asmlinkage long sys_ioprio_set(int which, int who, int ioprio) break; do_each_thread(g, p) { - if (p->uid != who) + if (__task_cred(p)->uid != who) continue; ret = set_task_ioprio(p, ioprio); if (ret) @@ -216,7 +222,7 @@ asmlinkage long sys_ioprio_get(int which, int who) break; case IOPRIO_WHO_USER: if (!who) - user = current->user; + user = current_user(); else user = find_user(who); @@ -224,7 +230,7 @@ asmlinkage long sys_ioprio_get(int which, int who) break; do_each_thread(g, p) { - if (p->uid != user->uid) + if (__task_cred(p)->uid != user->uid) continue; tmpio = get_task_ioprio(p); if (tmpio < 0) diff --git a/fs/isofs/export.c b/fs/isofs/export.c index bb219138331..e81a30593ba 100644 --- a/fs/isofs/export.c +++ b/fs/isofs/export.c @@ -22,7 +22,7 @@ isofs_export_iget(struct super_block *sb, __u32 generation) { struct inode *inode; - struct dentry *result; + if (block == 0) return ERR_PTR(-ESTALE); inode = isofs_iget(sb, block, offset); @@ -32,12 +32,7 @@ isofs_export_iget(struct super_block *sb, iput(inode); return ERR_PTR(-ESTALE); } - result = d_alloc_anon(inode); - if (!result) { - iput(inode); - return ERR_PTR(-ENOMEM); - } - return result; + return d_obtain_alias(inode); } /* This function is surprisingly simple. The trick is understanding @@ -51,7 +46,6 @@ static struct dentry *isofs_export_get_parent(struct dentry *child) unsigned long parent_offset = 0; struct inode *child_inode = child->d_inode; struct iso_inode_info *e_child_inode = ISOFS_I(child_inode); - struct inode *parent_inode = NULL; struct iso_directory_record *de = NULL; struct buffer_head * bh = NULL; struct dentry *rv = NULL; @@ -104,28 +98,11 @@ static struct dentry *isofs_export_get_parent(struct dentry *child) /* Normalize */ isofs_normalize_block_and_offset(de, &parent_block, &parent_offset); - /* Get the inode. */ - parent_inode = isofs_iget(child_inode->i_sb, - parent_block, - parent_offset); - if (IS_ERR(parent_inode)) { - rv = ERR_CAST(parent_inode); - if (rv != ERR_PTR(-ENOMEM)) - rv = ERR_PTR(-EACCES); - goto out; - } - - /* Allocate the dentry. */ - rv = d_alloc_anon(parent_inode); - if (rv == NULL) { - rv = ERR_PTR(-ENOMEM); - goto out; - } - + rv = d_obtain_alias(isofs_iget(child_inode->i_sb, parent_block, + parent_offset)); out: - if (bh) { + if (bh) brelse(bh); - } return rv; } diff --git a/fs/jbd/Kconfig b/fs/jbd/Kconfig new file mode 100644 index 00000000000..4e28beeed15 --- /dev/null +++ b/fs/jbd/Kconfig @@ -0,0 +1,30 @@ +config JBD + tristate + help + This is a generic journalling layer for block devices. It is + currently used by the ext3 file system, but it could also be + used to add journal support to other file systems or block + devices such as RAID or LVM. + + If you are using the ext3 file system, you need to say Y here. + If you are not using ext3 then you will probably want to say N. + + To compile this device as a module, choose M here: the module will be + called jbd. If you are compiling ext3 into the kernel, you + cannot compile this code as a module. + +config JBD_DEBUG + bool "JBD (ext3) debugging support" + depends on JBD && DEBUG_FS + help + If you are using the ext3 journaled file system (or potentially any + other file system/device using JBD), this option allows you to + enable debugging output while the system is running, in order to + help track down any problems you are having. By default the + debugging output will be turned off. + + If you select Y here, then you will be able to turn on debugging + with "echo N > /sys/kernel/debug/jbd/jbd-debug", where N is a + number between 1 and 5, the higher the number, the more debugging + output is generated. To turn debugging off again, do + "echo 0 > /sys/kernel/debug/jbd/jbd-debug". diff --git a/fs/jbd/checkpoint.c b/fs/jbd/checkpoint.c index a5432bbbfb8..61f32f3868c 100644 --- a/fs/jbd/checkpoint.c +++ b/fs/jbd/checkpoint.c @@ -93,7 +93,8 @@ static int __try_to_free_cp_buf(struct journal_head *jh) int ret = 0; struct buffer_head *bh = jh2bh(jh); - if (jh->b_jlist == BJ_None && !buffer_locked(bh) && !buffer_dirty(bh)) { + if (jh->b_jlist == BJ_None && !buffer_locked(bh) && + !buffer_dirty(bh) && !buffer_write_io_error(bh)) { JBUFFER_TRACE(jh, "remove from checkpoint list"); ret = __journal_remove_checkpoint(jh) + 1; jbd_unlock_bh_state(bh); @@ -114,7 +115,7 @@ static int __try_to_free_cp_buf(struct journal_head *jh) */ void __log_wait_for_space(journal_t *journal) { - int nblocks; + int nblocks, space_left; assert_spin_locked(&journal->j_state_lock); nblocks = jbd_space_needed(journal); @@ -126,14 +127,46 @@ void __log_wait_for_space(journal_t *journal) /* * Test again, another process may have checkpointed while we - * were waiting for the checkpoint lock + * were waiting for the checkpoint lock. If there are no + * transactions ready to be checkpointed, try to recover + * journal space by calling cleanup_journal_tail(), and if + * that doesn't work, by waiting for the currently committing + * transaction to complete. If there is absolutely no way + * to make progress, this is either a BUG or corrupted + * filesystem, so abort the journal and leave a stack + * trace for forensic evidence. */ spin_lock(&journal->j_state_lock); + spin_lock(&journal->j_list_lock); nblocks = jbd_space_needed(journal); - if (__log_space_left(journal) < nblocks) { + space_left = __log_space_left(journal); + if (space_left < nblocks) { + int chkpt = journal->j_checkpoint_transactions != NULL; + tid_t tid = 0; + + if (journal->j_committing_transaction) + tid = journal->j_committing_transaction->t_tid; + spin_unlock(&journal->j_list_lock); spin_unlock(&journal->j_state_lock); - log_do_checkpoint(journal); + if (chkpt) { + log_do_checkpoint(journal); + } else if (cleanup_journal_tail(journal) == 0) { + /* We were able to recover space; yay! */ + ; + } else if (tid) { + log_wait_commit(journal, tid); + } else { + printk(KERN_ERR "%s: needed %d blocks and " + "only had %d space available\n", + __func__, nblocks, space_left); + printk(KERN_ERR "%s: no way to get more " + "journal space\n", __func__); + WARN_ON(1); + journal_abort(journal, 0); + } spin_lock(&journal->j_state_lock); + } else { + spin_unlock(&journal->j_list_lock); } mutex_unlock(&journal->j_checkpoint_mutex); } @@ -160,21 +193,25 @@ static void jbd_sync_bh(journal_t *journal, struct buffer_head *bh) * buffers. Note that we take the buffers in the opposite ordering * from the one in which they were submitted for IO. * + * Return 0 on success, and return <0 if some buffers have failed + * to be written out. + * * Called with j_list_lock held. */ -static void __wait_cp_io(journal_t *journal, transaction_t *transaction) +static int __wait_cp_io(journal_t *journal, transaction_t *transaction) { struct journal_head *jh; struct buffer_head *bh; tid_t this_tid; int released = 0; + int ret = 0; this_tid = transaction->t_tid; restart: /* Did somebody clean up the transaction in the meanwhile? */ if (journal->j_checkpoint_transactions != transaction || transaction->t_tid != this_tid) - return; + return ret; while (!released && transaction->t_checkpoint_io_list) { jh = transaction->t_checkpoint_io_list; bh = jh2bh(jh); @@ -194,6 +231,9 @@ restart: spin_lock(&journal->j_list_lock); goto restart; } + if (unlikely(buffer_write_io_error(bh))) + ret = -EIO; + /* * Now in whatever state the buffer currently is, we know that * it has been written out and so we can drop it from the list @@ -203,6 +243,8 @@ restart: journal_remove_journal_head(bh); __brelse(bh); } + + return ret; } #define NR_BATCH 64 @@ -226,7 +268,8 @@ __flush_batch(journal_t *journal, struct buffer_head **bhs, int *batch_count) * Try to flush one buffer from the checkpoint list to disk. * * Return 1 if something happened which requires us to abort the current - * scan of the checkpoint list. + * scan of the checkpoint list. Return <0 if the buffer has failed to + * be written out. * * Called with j_list_lock held and drops it if 1 is returned * Called under jbd_lock_bh_state(jh2bh(jh)), and drops it @@ -256,6 +299,9 @@ static int __process_buffer(journal_t *journal, struct journal_head *jh, log_wait_commit(journal, tid); ret = 1; } else if (!buffer_dirty(bh)) { + ret = 1; + if (unlikely(buffer_write_io_error(bh))) + ret = -EIO; J_ASSERT_JH(jh, !buffer_jbddirty(bh)); BUFFER_TRACE(bh, "remove from checkpoint"); __journal_remove_checkpoint(jh); @@ -263,7 +309,6 @@ static int __process_buffer(journal_t *journal, struct journal_head *jh, jbd_unlock_bh_state(bh); journal_remove_journal_head(bh); __brelse(bh); - ret = 1; } else { /* * Important: we are about to write the buffer, and @@ -295,6 +340,7 @@ static int __process_buffer(journal_t *journal, struct journal_head *jh, * to disk. We submit larger chunks of data at once. * * The journal should be locked before calling this function. + * Called with j_checkpoint_mutex held. */ int log_do_checkpoint(journal_t *journal) { @@ -318,6 +364,7 @@ int log_do_checkpoint(journal_t *journal) * OK, we need to start writing disk blocks. Take one transaction * and write it. */ + result = 0; spin_lock(&journal->j_list_lock); if (!journal->j_checkpoint_transactions) goto out; @@ -334,7 +381,7 @@ restart: int batch_count = 0; struct buffer_head *bhs[NR_BATCH]; struct journal_head *jh; - int retry = 0; + int retry = 0, err; while (!retry && transaction->t_checkpoint_list) { struct buffer_head *bh; @@ -347,6 +394,8 @@ restart: break; } retry = __process_buffer(journal, jh, bhs,&batch_count); + if (retry < 0 && !result) + result = retry; if (!retry && (need_resched() || spin_needbreak(&journal->j_list_lock))) { spin_unlock(&journal->j_list_lock); @@ -371,14 +420,18 @@ restart: * Now we have cleaned up the first transaction's checkpoint * list. Let's clean up the second one */ - __wait_cp_io(journal, transaction); + err = __wait_cp_io(journal, transaction); + if (!result) + result = err; } out: spin_unlock(&journal->j_list_lock); - result = cleanup_journal_tail(journal); if (result < 0) - return result; - return 0; + journal_abort(journal, result); + else + result = cleanup_journal_tail(journal); + + return (result < 0) ? result : 0; } /* @@ -394,8 +447,9 @@ out: * This is the only part of the journaling code which really needs to be * aware of transaction aborts. Checkpointing involves writing to the * main filesystem area rather than to the journal, so it can proceed - * even in abort state, but we must not update the journal superblock if - * we have an abort error outstanding. + * even in abort state, but we must not update the super block if + * checkpointing may have failed. Otherwise, we would lose some metadata + * buffers which should be written-back to the filesystem. */ int cleanup_journal_tail(journal_t *journal) @@ -404,6 +458,9 @@ int cleanup_journal_tail(journal_t *journal) tid_t first_tid; unsigned long blocknr, freed; + if (is_journal_aborted(journal)) + return 1; + /* OK, work out the oldest transaction remaining in the log, and * the log block it starts at. * diff --git a/fs/jbd/commit.c b/fs/jbd/commit.c index ae08c057e75..25719d902c5 100644 --- a/fs/jbd/commit.c +++ b/fs/jbd/commit.c @@ -482,6 +482,8 @@ void journal_commit_transaction(journal_t *journal) printk(KERN_WARNING "JBD: Detected IO errors while flushing file data " "on %s\n", bdevname(journal->j_fs_dev, b)); + if (journal->j_flags & JFS_ABORT_ON_SYNCDATA_ERR) + journal_abort(journal, err); err = 0; } @@ -518,9 +520,10 @@ void journal_commit_transaction(journal_t *journal) jh = commit_transaction->t_buffers; /* If we're in abort mode, we just un-journal the buffer and - release it for background writing. */ + release it. */ if (is_journal_aborted(journal)) { + clear_buffer_jbddirty(jh2bh(jh)); JBUFFER_TRACE(jh, "journal is aborting: refile"); journal_refile_buffer(journal, jh); /* If that was the last one, we need to clean up @@ -762,6 +765,9 @@ wait_for_iobuf: /* AKPM: bforget here */ } + if (err) + journal_abort(journal, err); + jbd_debug(3, "JBD: commit phase 6\n"); if (journal_write_commit_record(journal, commit_transaction)) @@ -852,6 +858,8 @@ restart_loop: if (buffer_jbddirty(bh)) { JBUFFER_TRACE(jh, "add to new checkpointing trans"); __journal_insert_checkpoint(jh, commit_transaction); + if (is_journal_aborted(journal)) + clear_buffer_jbddirty(bh); JBUFFER_TRACE(jh, "refile for checkpoint writeback"); __journal_refile_buffer(jh); jbd_unlock_bh_state(bh); diff --git a/fs/jbd/journal.c b/fs/jbd/journal.c index aa7143a8349..9e4fa52d7dc 100644 --- a/fs/jbd/journal.c +++ b/fs/jbd/journal.c @@ -1121,9 +1121,12 @@ recovery_error: * * Release a journal_t structure once it is no longer in use by the * journaled object. + * Return <0 if we couldn't clean up the journal. */ -void journal_destroy(journal_t *journal) +int journal_destroy(journal_t *journal) { + int err = 0; + /* Wait for the commit thread to wake up and die. */ journal_kill_thread(journal); @@ -1146,11 +1149,16 @@ void journal_destroy(journal_t *journal) J_ASSERT(journal->j_checkpoint_transactions == NULL); spin_unlock(&journal->j_list_lock); - /* We can now mark the journal as empty. */ - journal->j_tail = 0; - journal->j_tail_sequence = ++journal->j_transaction_sequence; if (journal->j_sb_buffer) { - journal_update_superblock(journal, 1); + if (!is_journal_aborted(journal)) { + /* We can now mark the journal as empty. */ + journal->j_tail = 0; + journal->j_tail_sequence = + ++journal->j_transaction_sequence; + journal_update_superblock(journal, 1); + } else { + err = -EIO; + } brelse(journal->j_sb_buffer); } @@ -1160,6 +1168,8 @@ void journal_destroy(journal_t *journal) journal_destroy_revoke(journal); kfree(journal->j_wbuf); kfree(journal); + + return err; } @@ -1359,10 +1369,16 @@ int journal_flush(journal_t *journal) spin_lock(&journal->j_list_lock); while (!err && journal->j_checkpoint_transactions != NULL) { spin_unlock(&journal->j_list_lock); + mutex_lock(&journal->j_checkpoint_mutex); err = log_do_checkpoint(journal); + mutex_unlock(&journal->j_checkpoint_mutex); spin_lock(&journal->j_list_lock); } spin_unlock(&journal->j_list_lock); + + if (is_journal_aborted(journal)) + return -EIO; + cleanup_journal_tail(journal); /* Finally, mark the journal as really needing no recovery. @@ -1384,7 +1400,7 @@ int journal_flush(journal_t *journal) J_ASSERT(journal->j_head == journal->j_tail); J_ASSERT(journal->j_tail_sequence == journal->j_transaction_sequence); spin_unlock(&journal->j_state_lock); - return err; + return 0; } /** diff --git a/fs/jbd/recovery.c b/fs/jbd/recovery.c index 43bc5e5ed06..db5e982c5dd 100644 --- a/fs/jbd/recovery.c +++ b/fs/jbd/recovery.c @@ -223,7 +223,7 @@ do { \ */ int journal_recover(journal_t *journal) { - int err; + int err, err2; journal_superblock_t * sb; struct recovery_info info; @@ -261,7 +261,10 @@ int journal_recover(journal_t *journal) journal->j_transaction_sequence = ++info.end_transaction; journal_clear_revoke(journal); - sync_blockdev(journal->j_fs_dev); + err2 = sync_blockdev(journal->j_fs_dev); + if (!err) + err = err2; + return err; } diff --git a/fs/jbd/transaction.c b/fs/jbd/transaction.c index 0540ca27a44..60d4c32c880 100644 --- a/fs/jbd/transaction.c +++ b/fs/jbd/transaction.c @@ -860,7 +860,6 @@ out: * int journal_get_undo_access() - Notify intent to modify metadata with non-rewindable consequences * @handle: transaction * @bh: buffer to undo - * @credits: store the number of taken credits here (if not NULL) * * Sometimes there is a need to distinguish between metadata which has * been committed to disk and that which has not. The ext3fs code uses @@ -954,9 +953,10 @@ int journal_dirty_data(handle_t *handle, struct buffer_head *bh) journal_t *journal = handle->h_transaction->t_journal; int need_brelse = 0; struct journal_head *jh; + int ret = 0; if (is_handle_aborted(handle)) - return 0; + return ret; jh = journal_add_journal_head(bh); JBUFFER_TRACE(jh, "entry"); @@ -1067,7 +1067,16 @@ int journal_dirty_data(handle_t *handle, struct buffer_head *bh) time if it is redirtied */ } - /* journal_clean_data_list() may have got there first */ + /* + * We cannot remove the buffer with io error from the + * committing transaction, because otherwise it would + * miss the error and the commit would not abort. + */ + if (unlikely(!buffer_uptodate(bh))) { + ret = -EIO; + goto no_journal; + } + if (jh->b_transaction != NULL) { JBUFFER_TRACE(jh, "unfile from commit"); __journal_temp_unlink_buffer(jh); @@ -1108,7 +1117,7 @@ no_journal: } JBUFFER_TRACE(jh, "exit"); journal_put_journal_head(jh); - return 0; + return ret; } /** diff --git a/fs/jbd2/Kconfig b/fs/jbd2/Kconfig new file mode 100644 index 00000000000..f32f346f4b0 --- /dev/null +++ b/fs/jbd2/Kconfig @@ -0,0 +1,33 @@ +config JBD2 + tristate + select CRC32 + help + This is a generic journaling layer for block devices that support + both 32-bit and 64-bit block numbers. It is currently used by + the ext4 and OCFS2 filesystems, but it could also be used to add + journal support to other file systems or block devices such + as RAID or LVM. + + If you are using ext4 or OCFS2, you need to say Y here. + If you are not using ext4 or OCFS2 then you will + probably want to say N. + + To compile this device as a module, choose M here. The module will be + called jbd2. If you are compiling ext4 or OCFS2 into the kernel, + you cannot compile this code as a module. + +config JBD2_DEBUG + bool "JBD2 (ext4) debugging support" + depends on JBD2 && DEBUG_FS + help + If you are using the ext4 journaled file system (or + potentially any other filesystem/device using JBD2), this option + allows you to enable debugging output while the system is running, + in order to help track down any problems you are having. + By default, the debugging output will be turned off. + + If you select Y here, then you will be able to turn on debugging + with "echo N > /sys/kernel/debug/jbd2/jbd2-debug", where N is a + number between 1 and 5. The higher the number, the more debugging + output is generated. To turn debugging off again, do + "echo 0 > /sys/kernel/debug/jbd2/jbd2-debug". diff --git a/fs/jbd2/checkpoint.c b/fs/jbd2/checkpoint.c index 9203c3332f1..9497718fe92 100644 --- a/fs/jbd2/checkpoint.c +++ b/fs/jbd2/checkpoint.c @@ -116,7 +116,7 @@ static int __try_to_free_cp_buf(struct journal_head *jh) */ void __jbd2_log_wait_for_space(journal_t *journal) { - int nblocks; + int nblocks, space_left; assert_spin_locked(&journal->j_state_lock); nblocks = jbd_space_needed(journal); @@ -129,25 +129,43 @@ void __jbd2_log_wait_for_space(journal_t *journal) /* * Test again, another process may have checkpointed while we * were waiting for the checkpoint lock. If there are no - * outstanding transactions there is nothing to checkpoint and - * we can't make progress. Abort the journal in this case. + * transactions ready to be checkpointed, try to recover + * journal space by calling cleanup_journal_tail(), and if + * that doesn't work, by waiting for the currently committing + * transaction to complete. If there is absolutely no way + * to make progress, this is either a BUG or corrupted + * filesystem, so abort the journal and leave a stack + * trace for forensic evidence. */ spin_lock(&journal->j_state_lock); spin_lock(&journal->j_list_lock); nblocks = jbd_space_needed(journal); - if (__jbd2_log_space_left(journal) < nblocks) { + space_left = __jbd2_log_space_left(journal); + if (space_left < nblocks) { int chkpt = journal->j_checkpoint_transactions != NULL; + tid_t tid = 0; + if (journal->j_committing_transaction) + tid = journal->j_committing_transaction->t_tid; spin_unlock(&journal->j_list_lock); spin_unlock(&journal->j_state_lock); if (chkpt) { jbd2_log_do_checkpoint(journal); + } else if (jbd2_cleanup_journal_tail(journal) == 0) { + /* We were able to recover space; yay! */ + ; + } else if (tid) { + jbd2_log_wait_commit(journal, tid); } else { - printk(KERN_ERR "%s: no transactions\n", - __func__); + printk(KERN_ERR "%s: needed %d blocks and " + "only had %d space available\n", + __func__, nblocks, space_left); + printk(KERN_ERR "%s: no way to get more " + "journal space in %s\n", __func__, + journal->j_devname); + WARN_ON(1); jbd2_journal_abort(journal, 0); } - spin_lock(&journal->j_state_lock); } else { spin_unlock(&journal->j_list_lock); diff --git a/fs/jbd2/commit.c b/fs/jbd2/commit.c index 0abe02c4242..ebc667bc54a 100644 --- a/fs/jbd2/commit.c +++ b/fs/jbd2/commit.c @@ -974,6 +974,9 @@ restart_loop: journal->j_committing_transaction = NULL; spin_unlock(&journal->j_state_lock); + if (journal->j_commit_callback) + journal->j_commit_callback(journal, commit_transaction); + if (commit_transaction->t_checkpoint_list == NULL && commit_transaction->t_checkpoint_io_list == NULL) { __jbd2_journal_drop_transaction(journal, commit_transaction); @@ -996,7 +999,7 @@ restart_loop: spin_unlock(&journal->j_list_lock); trace_mark(jbd2_end_commit, "dev %s transaction %d head %d", - journal->j_devname, commit_transaction->t_tid, + journal->j_devname, journal->j_commit_sequence, journal->j_tail_sequence); jbd_debug(1, "JBD: commit %d complete, head %d\n", journal->j_commit_sequence, journal->j_tail_sequence); diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c index 783de118de9..e70d657a19f 100644 --- a/fs/jbd2/journal.c +++ b/fs/jbd2/journal.c @@ -1089,6 +1089,7 @@ journal_t * jbd2_journal_init_inode (struct inode *inode) if (!journal->j_wbuf) { printk(KERN_ERR "%s: Cant allocate bhs for commit thread\n", __func__); + jbd2_stats_proc_exit(journal); kfree(journal); return NULL; } @@ -1098,6 +1099,7 @@ journal_t * jbd2_journal_init_inode (struct inode *inode) if (err) { printk(KERN_ERR "%s: Cannnot locate journal superblock\n", __func__); + jbd2_stats_proc_exit(journal); kfree(journal); return NULL; } diff --git a/fs/jbd2/transaction.c b/fs/jbd2/transaction.c index e5d540588fa..39b7805a599 100644 --- a/fs/jbd2/transaction.c +++ b/fs/jbd2/transaction.c @@ -52,6 +52,7 @@ jbd2_get_transaction(journal_t *journal, transaction_t *transaction) transaction->t_expires = jiffies + journal->j_commit_interval; spin_lock_init(&transaction->t_handle_lock); INIT_LIST_HEAD(&transaction->t_inode_list); + INIT_LIST_HEAD(&transaction->t_private_list); /* Set up the commit timer for the new transaction. */ journal->j_commit_timer.expires = round_jiffies(transaction->t_expires); diff --git a/fs/jffs2/Kconfig b/fs/jffs2/Kconfig new file mode 100644 index 00000000000..6ae169cd8fa --- /dev/null +++ b/fs/jffs2/Kconfig @@ -0,0 +1,188 @@ +config JFFS2_FS + tristate "Journalling Flash File System v2 (JFFS2) support" + select CRC32 + depends on MTD + help + JFFS2 is the second generation of the Journalling Flash File System + for use on diskless embedded devices. It provides improved wear + levelling, compression and support for hard links. You cannot use + this on normal block devices, only on 'MTD' devices. + + Further information on the design and implementation of JFFS2 is + available at <http://sources.redhat.com/jffs2/>. + +config JFFS2_FS_DEBUG + int "JFFS2 debugging verbosity (0 = quiet, 2 = noisy)" + depends on JFFS2_FS + default "0" + help + This controls the amount of debugging messages produced by the JFFS2 + code. Set it to zero for use in production systems. For evaluation, + testing and debugging, it's advisable to set it to one. This will + enable a few assertions and will print debugging messages at the + KERN_DEBUG loglevel, where they won't normally be visible. Level 2 + is unlikely to be useful - it enables extra debugging in certain + areas which at one point needed debugging, but when the bugs were + located and fixed, the detailed messages were relegated to level 2. + + If reporting bugs, please try to have available a full dump of the + messages at debug level 1 while the misbehaviour was occurring. + +config JFFS2_FS_WRITEBUFFER + bool "JFFS2 write-buffering support" + depends on JFFS2_FS + default y + help + This enables the write-buffering support in JFFS2. + + This functionality is required to support JFFS2 on the following + types of flash devices: + - NAND flash + - NOR flash with transparent ECC + - DataFlash + +config JFFS2_FS_WBUF_VERIFY + bool "Verify JFFS2 write-buffer reads" + depends on JFFS2_FS_WRITEBUFFER + default n + help + This causes JFFS2 to read back every page written through the + write-buffer, and check for errors. + +config JFFS2_SUMMARY + bool "JFFS2 summary support (EXPERIMENTAL)" + depends on JFFS2_FS && EXPERIMENTAL + default n + help + This feature makes it possible to use summary information + for faster filesystem mount. + + The summary information can be inserted into a filesystem image + by the utility 'sumtool'. + + If unsure, say 'N'. + +config JFFS2_FS_XATTR + bool "JFFS2 XATTR support (EXPERIMENTAL)" + depends on JFFS2_FS && EXPERIMENTAL + default n + help + Extended attributes are name:value pairs associated with inodes by + the kernel or by users (see the attr(5) manual page, or visit + <http://acl.bestbits.at/> for details). + + If unsure, say N. + +config JFFS2_FS_POSIX_ACL + bool "JFFS2 POSIX Access Control Lists" + depends on JFFS2_FS_XATTR + default y + select FS_POSIX_ACL + help + Posix Access Control Lists (ACLs) support permissions for users and + groups beyond the owner/group/world scheme. + + To learn more about Access Control Lists, visit the Posix ACLs for + Linux website <http://acl.bestbits.at/>. + + If you don't know what Access Control Lists are, say N + +config JFFS2_FS_SECURITY + bool "JFFS2 Security Labels" + depends on JFFS2_FS_XATTR + default y + help + Security labels support alternative access control models + implemented by security modules like SELinux. This option + enables an extended attribute handler for file security + labels in the jffs2 filesystem. + + If you are not using a security module that requires using + extended attributes for file security labels, say N. + +config JFFS2_COMPRESSION_OPTIONS + bool "Advanced compression options for JFFS2" + depends on JFFS2_FS + default n + help + Enabling this option allows you to explicitly choose which + compression modules, if any, are enabled in JFFS2. Removing + compressors can mean you cannot read existing file systems, + and enabling experimental compressors can mean that you + write a file system which cannot be read by a standard kernel. + + If unsure, you should _definitely_ say 'N'. + +config JFFS2_ZLIB + bool "JFFS2 ZLIB compression support" if JFFS2_COMPRESSION_OPTIONS + select ZLIB_INFLATE + select ZLIB_DEFLATE + depends on JFFS2_FS + default y + help + Zlib is designed to be a free, general-purpose, legally unencumbered, + lossless data-compression library for use on virtually any computer + hardware and operating system. See <http://www.gzip.org/zlib/> for + further information. + + Say 'Y' if unsure. + +config JFFS2_LZO + bool "JFFS2 LZO compression support" if JFFS2_COMPRESSION_OPTIONS + select LZO_COMPRESS + select LZO_DECOMPRESS + depends on JFFS2_FS + default n + help + minilzo-based compression. Generally works better than Zlib. + + This feature was added in July, 2007. Say 'N' if you need + compatibility with older bootloaders or kernels. + +config JFFS2_RTIME + bool "JFFS2 RTIME compression support" if JFFS2_COMPRESSION_OPTIONS + depends on JFFS2_FS + default y + help + Rtime does manage to recompress already-compressed data. Say 'Y' if unsure. + +config JFFS2_RUBIN + bool "JFFS2 RUBIN compression support" if JFFS2_COMPRESSION_OPTIONS + depends on JFFS2_FS + default n + help + RUBINMIPS and DYNRUBIN compressors. Say 'N' if unsure. + +choice + prompt "JFFS2 default compression mode" if JFFS2_COMPRESSION_OPTIONS + default JFFS2_CMODE_PRIORITY + depends on JFFS2_FS + help + You can set here the default compression mode of JFFS2 from + the available compression modes. Don't touch if unsure. + +config JFFS2_CMODE_NONE + bool "no compression" + help + Uses no compression. + +config JFFS2_CMODE_PRIORITY + bool "priority" + help + Tries the compressors in a predefined order and chooses the first + successful one. + +config JFFS2_CMODE_SIZE + bool "size (EXPERIMENTAL)" + help + Tries all compressors and chooses the one which has the smallest + result. + +config JFFS2_CMODE_FAVOURLZO + bool "Favour LZO" + help + Tries all compressors and chooses the one which has the smallest + result but gives some preference to LZO (which has faster + decompression) at the expense of size. + +endchoice diff --git a/fs/jffs2/background.c b/fs/jffs2/background.c index 8adebd3e43c..3cceef4ad2b 100644 --- a/fs/jffs2/background.c +++ b/fs/jffs2/background.c @@ -85,15 +85,15 @@ static int jffs2_garbage_collect_thread(void *_c) for (;;) { allow_signal(SIGHUP); again: + spin_lock(&c->erase_completion_lock); if (!jffs2_thread_should_wake(c)) { set_current_state (TASK_INTERRUPTIBLE); + spin_unlock(&c->erase_completion_lock); D1(printk(KERN_DEBUG "jffs2_garbage_collect_thread sleeping...\n")); - /* Yes, there's a race here; we checked jffs2_thread_should_wake() - before setting current->state to TASK_INTERRUPTIBLE. But it doesn't - matter - We don't care if we miss a wakeup, because the GC thread - is only an optimisation anyway. */ schedule(); - } + } else + spin_unlock(&c->erase_completion_lock); + /* This thread is purely an optimisation. But if it runs when other things could be running, it actually makes things a diff --git a/fs/jffs2/compr.c b/fs/jffs2/compr.c index 86739ee53b3..f25e70c1b51 100644 --- a/fs/jffs2/compr.c +++ b/fs/jffs2/compr.c @@ -53,8 +53,8 @@ static int jffs2_is_best_compression(struct jffs2_compressor *this, } /* jffs2_compress: - * @data: Pointer to uncompressed data - * @cdata: Pointer to returned pointer to buffer for compressed data + * @data_in: Pointer to uncompressed data + * @cpage_out: Pointer to returned pointer to buffer for compressed data * @datalen: On entry, holds the amount of data available for compression. * On exit, expected to hold the amount of data actually compressed. * @cdatalen: On entry, holds the amount of space available for compressed diff --git a/fs/jffs2/compr_lzo.c b/fs/jffs2/compr_lzo.c index 47b045797e4..90cb60d0978 100644 --- a/fs/jffs2/compr_lzo.c +++ b/fs/jffs2/compr_lzo.c @@ -19,7 +19,7 @@ static void *lzo_mem; static void *lzo_compress_buf; -static DEFINE_MUTEX(deflate_mutex); +static DEFINE_MUTEX(deflate_mutex); /* for lzo_mem and lzo_compress_buf */ static void free_workspace(void) { @@ -49,18 +49,21 @@ static int jffs2_lzo_compress(unsigned char *data_in, unsigned char *cpage_out, mutex_lock(&deflate_mutex); ret = lzo1x_1_compress(data_in, *sourcelen, lzo_compress_buf, &compress_size, lzo_mem); - mutex_unlock(&deflate_mutex); - if (ret != LZO_E_OK) - return -1; + goto fail; if (compress_size > *dstlen) - return -1; + goto fail; memcpy(cpage_out, lzo_compress_buf, compress_size); - *dstlen = compress_size; + mutex_unlock(&deflate_mutex); + *dstlen = compress_size; return 0; + + fail: + mutex_unlock(&deflate_mutex); + return -1; } static int jffs2_lzo_decompress(unsigned char *data_in, unsigned char *cpage_out, diff --git a/fs/jffs2/dir.c b/fs/jffs2/dir.c index cd219ef5525..6f60cc910f4 100644 --- a/fs/jffs2/dir.c +++ b/fs/jffs2/dir.c @@ -39,7 +39,8 @@ const struct file_operations jffs2_dir_operations = .read = generic_read_dir, .readdir = jffs2_readdir, .unlocked_ioctl=jffs2_ioctl, - .fsync = jffs2_fsync + .fsync = jffs2_fsync, + .llseek = generic_file_llseek, }; @@ -108,9 +109,7 @@ static struct dentry *jffs2_lookup(struct inode *dir_i, struct dentry *target, } } - d_add(target, inode); - - return NULL; + return d_splice_alias(inode, target); } /***********************************************************************/ @@ -311,7 +310,7 @@ static int jffs2_symlink (struct inode *dir_i, struct dentry *dentry, const char /* FIXME: If you care. We'd need to use frags for the target if it grows much more than this */ if (targetlen > 254) - return -EINVAL; + return -ENAMETOOLONG; ri = jffs2_alloc_raw_inode(); diff --git a/fs/jffs2/erase.c b/fs/jffs2/erase.c index dddb2a6c9e2..259461b910a 100644 --- a/fs/jffs2/erase.c +++ b/fs/jffs2/erase.c @@ -68,7 +68,7 @@ static void jffs2_erase_block(struct jffs2_sb_info *c, instr->len = c->sector_size; instr->callback = jffs2_erase_callback; instr->priv = (unsigned long)(&instr[1]); - instr->fail_addr = 0xffffffff; + instr->fail_addr = MTD_FAIL_ADDR_UNKNOWN; ((struct erase_priv_struct *)instr->priv)->jeb = jeb; ((struct erase_priv_struct *)instr->priv)->c = c; @@ -175,7 +175,7 @@ static void jffs2_erase_failed(struct jffs2_sb_info *c, struct jffs2_eraseblock { /* For NAND, if the failure did not occur at the device level for a specific physical page, don't bother updating the bad block table. */ - if (jffs2_cleanmarker_oob(c) && (bad_offset != 0xffffffff)) { + if (jffs2_cleanmarker_oob(c) && (bad_offset != MTD_FAIL_ADDR_UNKNOWN)) { /* We had a device-level failure to erase. Let's see if we've failed too many times. */ if (!jffs2_write_nand_badblock(c, jeb, bad_offset)) { diff --git a/fs/jffs2/fs.c b/fs/jffs2/fs.c index 086c4383022..249305d65d5 100644 --- a/fs/jffs2/fs.c +++ b/fs/jffs2/fs.c @@ -207,6 +207,8 @@ int jffs2_statfs(struct dentry *dentry, struct kstatfs *buf) buf->f_files = 0; buf->f_ffree = 0; buf->f_namelen = JFFS2_MAX_NAME_LEN; + buf->f_fsid.val[0] = JFFS2_SUPER_MAGIC; + buf->f_fsid.val[1] = c->mtd->index; spin_lock(&c->erase_completion_lock); avail = c->dirty_size + c->free_size; @@ -440,14 +442,14 @@ struct inode *jffs2_new_inode (struct inode *dir_i, int mode, struct jffs2_raw_i memset(ri, 0, sizeof(*ri)); /* Set OS-specific defaults for new inodes */ - ri->uid = cpu_to_je16(current->fsuid); + ri->uid = cpu_to_je16(current_fsuid()); if (dir_i->i_mode & S_ISGID) { ri->gid = cpu_to_je16(dir_i->i_gid); if (S_ISDIR(mode)) mode |= S_ISGID; } else { - ri->gid = cpu_to_je16(current->fsgid); + ri->gid = cpu_to_je16(current_fsgid()); } /* POSIX ACLs have to be processed now, at least partly. diff --git a/fs/jffs2/nodemgmt.c b/fs/jffs2/nodemgmt.c index a9bf9603c1b..21a052915aa 100644 --- a/fs/jffs2/nodemgmt.c +++ b/fs/jffs2/nodemgmt.c @@ -261,6 +261,12 @@ static int jffs2_find_nextblock(struct jffs2_sb_info *c) jffs2_sum_reset_collected(c->summary); /* reset collected summary */ +#ifdef CONFIG_JFFS2_FS_WRITEBUFFER + /* adjust write buffer offset, else we get a non contiguous write bug */ + if (!(c->wbuf_ofs % c->sector_size) && !c->wbuf_len) + c->wbuf_ofs = 0xffffffff; +#endif + D1(printk(KERN_DEBUG "jffs2_find_nextblock(): new nextblock = 0x%08x\n", c->nextblock->offset)); return 0; diff --git a/fs/jffs2/super.c b/fs/jffs2/super.c index efd401257ed..4c4e18c54a5 100644 --- a/fs/jffs2/super.c +++ b/fs/jffs2/super.c @@ -22,6 +22,7 @@ #include <linux/mtd/super.h> #include <linux/ctype.h> #include <linux/namei.h> +#include <linux/exportfs.h> #include "compr.h" #include "nodelist.h" @@ -62,6 +63,52 @@ static int jffs2_sync_fs(struct super_block *sb, int wait) return 0; } +static struct inode *jffs2_nfs_get_inode(struct super_block *sb, uint64_t ino, + uint32_t generation) +{ + /* We don't care about i_generation. We'll destroy the flash + before we start re-using inode numbers anyway. And even + if that wasn't true, we'd have other problems...*/ + return jffs2_iget(sb, ino); +} + +static struct dentry *jffs2_fh_to_dentry(struct super_block *sb, struct fid *fid, + int fh_len, int fh_type) +{ + return generic_fh_to_dentry(sb, fid, fh_len, fh_type, + jffs2_nfs_get_inode); +} + +static struct dentry *jffs2_fh_to_parent(struct super_block *sb, struct fid *fid, + int fh_len, int fh_type) +{ + return generic_fh_to_parent(sb, fid, fh_len, fh_type, + jffs2_nfs_get_inode); +} + +static struct dentry *jffs2_get_parent(struct dentry *child) +{ + struct jffs2_inode_info *f; + uint32_t pino; + + BUG_ON(!S_ISDIR(child->d_inode->i_mode)); + + f = JFFS2_INODE_INFO(child->d_inode); + + pino = f->inocache->pino_nlink; + + JFFS2_DEBUG("Parent of directory ino #%u is #%u\n", + f->inocache->ino, pino); + + return d_obtain_alias(jffs2_iget(child->d_inode->i_sb, pino)); +} + +static struct export_operations jffs2_export_ops = { + .get_parent = jffs2_get_parent, + .fh_to_dentry = jffs2_fh_to_dentry, + .fh_to_parent = jffs2_fh_to_parent, +}; + static const struct super_operations jffs2_super_operations = { .alloc_inode = jffs2_alloc_inode, @@ -104,6 +151,7 @@ static int jffs2_fill_super(struct super_block *sb, void *data, int silent) spin_lock_init(&c->inocache_lock); sb->s_op = &jffs2_super_operations; + sb->s_export_op = &jffs2_export_ops; sb->s_flags = sb->s_flags | MS_NOATIME; sb->s_xattr = jffs2_xattr_handlers; #ifdef CONFIG_JFFS2_FS_POSIX_ACL diff --git a/fs/jffs2/wbuf.c b/fs/jffs2/wbuf.c index 0e78b00035e..d9a721e6db7 100644 --- a/fs/jffs2/wbuf.c +++ b/fs/jffs2/wbuf.c @@ -679,10 +679,7 @@ static int __jffs2_flush_wbuf(struct jffs2_sb_info *c, int pad) memset(c->wbuf,0xff,c->wbuf_pagesize); /* adjust write buffer offset, else we get a non contiguous write bug */ - if (SECTOR_ADDR(c->wbuf_ofs) == SECTOR_ADDR(c->wbuf_ofs+c->wbuf_pagesize)) - c->wbuf_ofs += c->wbuf_pagesize; - else - c->wbuf_ofs = 0xffffffff; + c->wbuf_ofs += c->wbuf_pagesize; c->wbuf_len = 0; return 0; } diff --git a/fs/jfs/inode.c b/fs/jfs/inode.c index 210339784b5..b00ee9f05a0 100644 --- a/fs/jfs/inode.c +++ b/fs/jfs/inode.c @@ -59,8 +59,14 @@ struct inode *jfs_iget(struct super_block *sb, unsigned long ino) if (inode->i_size >= IDATASIZE) { inode->i_op = &page_symlink_inode_operations; inode->i_mapping->a_ops = &jfs_aops; - } else + } else { inode->i_op = &jfs_symlink_inode_operations; + /* + * The inline data should be null-terminated, but + * don't let on-disk corruption crash the kernel + */ + JFS_IP(inode)->i_inline[inode->i_size] = '\0'; + } } else { inode->i_op = &jfs_file_inode_operations; init_special_inode(inode, inode->i_mode, inode->i_rdev); diff --git a/fs/jfs/jfs_inode.c b/fs/jfs/jfs_inode.c index ed6574bee51..70022fd1c53 100644 --- a/fs/jfs/jfs_inode.c +++ b/fs/jfs/jfs_inode.c @@ -93,13 +93,13 @@ struct inode *ialloc(struct inode *parent, umode_t mode) return ERR_PTR(rc); } - inode->i_uid = current->fsuid; + inode->i_uid = current_fsuid(); if (parent->i_mode & S_ISGID) { inode->i_gid = parent->i_gid; if (S_ISDIR(mode)) mode |= S_ISGID; } else - inode->i_gid = current->fsgid; + inode->i_gid = current_fsgid(); /* * New inodes need to save sane values on disk when diff --git a/fs/jfs/jfs_logmgr.c b/fs/jfs/jfs_logmgr.c index cd2ec2988b5..335c4de6552 100644 --- a/fs/jfs/jfs_logmgr.c +++ b/fs/jfs/jfs_logmgr.c @@ -1168,7 +1168,7 @@ journal_found: bd_release(bdev); close: /* close external log device */ - blkdev_put(bdev); + blkdev_put(bdev, FMODE_READ|FMODE_WRITE); free: /* free log descriptor */ mutex_unlock(&jfs_log_mutex); @@ -1514,7 +1514,7 @@ int lmLogClose(struct super_block *sb) rc = lmLogShutdown(log); bd_release(bdev); - blkdev_put(bdev); + blkdev_put(bdev, FMODE_READ|FMODE_WRITE); kfree(log); diff --git a/fs/jfs/namei.c b/fs/jfs/namei.c index 2aba8238681..cc3cedffbfa 100644 --- a/fs/jfs/namei.c +++ b/fs/jfs/namei.c @@ -1511,25 +1511,12 @@ struct dentry *jfs_fh_to_parent(struct super_block *sb, struct fid *fid, struct dentry *jfs_get_parent(struct dentry *dentry) { - struct super_block *sb = dentry->d_inode->i_sb; - struct dentry *parent = ERR_PTR(-ENOENT); - struct inode *inode; unsigned long parent_ino; parent_ino = le32_to_cpu(JFS_IP(dentry->d_inode)->i_dtroot.header.idotdot); - inode = jfs_iget(sb, parent_ino); - if (IS_ERR(inode)) { - parent = ERR_CAST(inode); - } else { - parent = d_alloc_anon(inode); - if (!parent) { - parent = ERR_PTR(-ENOMEM); - iput(inode); - } - } - return parent; + return d_obtain_alias(jfs_iget(dentry->d_inode->i_sb, parent_ino)); } const struct inode_operations jfs_dir_inode_operations = { @@ -1560,6 +1547,7 @@ const struct file_operations jfs_dir_operations = { #ifdef CONFIG_COMPAT .compat_ioctl = jfs_compat_ioctl, #endif + .llseek = generic_file_llseek, }; static int jfs_ci_hash(struct dentry *dir, struct qstr *this) diff --git a/fs/libfs.c b/fs/libfs.c index 1add676a19d..e960a832190 100644 --- a/fs/libfs.c +++ b/fs/libfs.c @@ -732,28 +732,6 @@ out: return ret; } -/* - * This is what d_alloc_anon should have been. Once the exportfs - * argument transition has been finished I will update d_alloc_anon - * to this prototype and this wrapper will go away. --hch - */ -static struct dentry *exportfs_d_alloc(struct inode *inode) -{ - struct dentry *dentry; - - if (!inode) - return NULL; - if (IS_ERR(inode)) - return ERR_PTR(PTR_ERR(inode)); - - dentry = d_alloc_anon(inode); - if (!dentry) { - iput(inode); - dentry = ERR_PTR(-ENOMEM); - } - return dentry; -} - /** * generic_fh_to_dentry - generic helper for the fh_to_dentry export operation * @sb: filesystem to do the file handle conversion on @@ -782,7 +760,7 @@ struct dentry *generic_fh_to_dentry(struct super_block *sb, struct fid *fid, break; } - return exportfs_d_alloc(inode); + return d_obtain_alias(inode); } EXPORT_SYMBOL_GPL(generic_fh_to_dentry); @@ -815,7 +793,7 @@ struct dentry *generic_fh_to_parent(struct super_block *sb, struct fid *fid, break; } - return exportfs_d_alloc(inode); + return d_obtain_alias(inode); } EXPORT_SYMBOL_GPL(generic_fh_to_parent); @@ -836,7 +814,7 @@ EXPORT_SYMBOL(simple_getattr); EXPORT_SYMBOL(simple_link); EXPORT_SYMBOL(simple_lookup); EXPORT_SYMBOL(simple_pin_fs); -EXPORT_SYMBOL(simple_prepare_write); +EXPORT_UNUSED_SYMBOL(simple_prepare_write); EXPORT_SYMBOL(simple_readpage); EXPORT_SYMBOL(simple_release_fs); EXPORT_SYMBOL(simple_rename); diff --git a/fs/lockd/clntlock.c b/fs/lockd/clntlock.c index 8307dd64bf4..1f3b0fc0d35 100644 --- a/fs/lockd/clntlock.c +++ b/fs/lockd/clntlock.c @@ -14,6 +14,7 @@ #include <linux/sunrpc/svc.h> #include <linux/lockd/lockd.h> #include <linux/smp_lock.h> +#include <linux/kthread.h> #define NLMDBG_FACILITY NLMDBG_CLIENT @@ -60,7 +61,7 @@ struct nlm_host *nlmclnt_init(const struct nlmclnt_initdata *nlm_init) host = nlmclnt_lookup_host(nlm_init->address, nlm_init->addrlen, nlm_init->protocol, nlm_version, - nlm_init->hostname); + nlm_init->hostname, nlm_init->noresvport); if (host == NULL) { lockd_down(); return ERR_PTR(-ENOLCK); @@ -191,11 +192,15 @@ __be32 nlmclnt_grant(const struct sockaddr *addr, const struct nlm_lock *lock) void nlmclnt_recovery(struct nlm_host *host) { + struct task_struct *task; + if (!host->h_reclaiming++) { nlm_get_host(host); - __module_get(THIS_MODULE); - if (kernel_thread(reclaimer, host, CLONE_FS | CLONE_FILES) < 0) - module_put(THIS_MODULE); + task = kthread_run(reclaimer, host, "%s-reclaim", host->h_name); + if (IS_ERR(task)) + printk(KERN_ERR "lockd: unable to spawn reclaimer " + "thread. Locks for %s won't be reclaimed! " + "(%ld)\n", host->h_name, PTR_ERR(task)); } } @@ -207,7 +212,6 @@ reclaimer(void *ptr) struct file_lock *fl, *next; u32 nsmstate; - daemonize("%s-reclaim", host->h_name); allow_signal(SIGKILL); down_write(&host->h_rwsem); @@ -233,7 +237,12 @@ restart: list_for_each_entry_safe(fl, next, &host->h_reclaim, fl_u.nfs_fl.list) { list_del_init(&fl->fl_u.nfs_fl.list); - /* Why are we leaking memory here? --okir */ + /* + * sending this thread a SIGKILL will result in any unreclaimed + * locks being removed from the h_granted list. This means that + * the kernel will not attempt to reclaim them again if a new + * reclaimer thread is spawned for this host. + */ if (signalled()) continue; if (nlmclnt_reclaim(host, fl) != 0) @@ -261,5 +270,5 @@ restart: nlm_release_host(host); lockd_down(); unlock_kernel(); - module_put_and_exit(0); + return 0; } diff --git a/fs/lockd/host.c b/fs/lockd/host.c index 9fd8889097b..abdebf76b82 100644 --- a/fs/lockd/host.c +++ b/fs/lockd/host.c @@ -48,6 +48,7 @@ struct nlm_lookup_host_info { const size_t hostname_len; /* it's length */ const struct sockaddr *src_sap; /* our address (optional) */ const size_t src_len; /* it's length */ + const int noresvport; /* use non-priv port */ }; /* @@ -115,14 +116,14 @@ static void nlm_display_address(const struct sockaddr *sap, snprintf(buf, len, "unspecified"); break; case AF_INET: - snprintf(buf, len, NIPQUAD_FMT, NIPQUAD(sin->sin_addr.s_addr)); + snprintf(buf, len, "%pI4", &sin->sin_addr.s_addr); break; case AF_INET6: if (ipv6_addr_v4mapped(&sin6->sin6_addr)) - snprintf(buf, len, NIPQUAD_FMT, - NIPQUAD(sin6->sin6_addr.s6_addr32[3])); + snprintf(buf, len, "%pI4", + &sin6->sin6_addr.s6_addr32[3]); else - snprintf(buf, len, NIP6_FMT, NIP6(sin6->sin6_addr)); + snprintf(buf, len, "%pI6", &sin6->sin6_addr); break; default: snprintf(buf, len, "unsupported address family"); @@ -167,7 +168,8 @@ static struct nlm_host *nlm_lookup_host(struct nlm_lookup_host_info *ni) continue; if (host->h_server != ni->server) continue; - if (!nlm_cmp_addr(nlm_srcaddr(host), ni->src_sap)) + if (ni->server && + !nlm_cmp_addr(nlm_srcaddr(host), ni->src_sap)) continue; /* Move to head of hash chain. */ @@ -221,6 +223,7 @@ static struct nlm_host *nlm_lookup_host(struct nlm_lookup_host_info *ni) host->h_nsmstate = 0; /* real NSM state */ host->h_nsmhandle = nsm; host->h_server = ni->server; + host->h_noresvport = ni->noresvport; hlist_add_head(&host->h_hash, chain); INIT_LIST_HEAD(&host->h_lockowners); spin_lock_init(&host->h_lock); @@ -271,6 +274,7 @@ nlm_destroy_host(struct nlm_host *host) * @protocol: transport protocol to use * @version: NLM protocol version * @hostname: '\0'-terminated hostname of server + * @noresvport: 1 if non-privileged port should be used * * Returns an nlm_host structure that matches the passed-in * [server address, transport protocol, NLM version, server hostname]. @@ -280,7 +284,9 @@ nlm_destroy_host(struct nlm_host *host) struct nlm_host *nlmclnt_lookup_host(const struct sockaddr *sap, const size_t salen, const unsigned short protocol, - const u32 version, const char *hostname) + const u32 version, + const char *hostname, + int noresvport) { const struct sockaddr source = { .sa_family = AF_UNSPEC, @@ -295,6 +301,7 @@ struct nlm_host *nlmclnt_lookup_host(const struct sockaddr *sap, .hostname_len = strlen(hostname), .src_sap = &source, .src_len = sizeof(source), + .noresvport = noresvport, }; dprintk("lockd: %s(host='%s', vers=%u, proto=%s)\n", __func__, @@ -416,6 +423,8 @@ nlm_bind_host(struct nlm_host *host) */ if (!host->h_server) args.flags |= RPC_CLNT_CREATE_HARDRTRY; + if (host->h_noresvport) + args.flags |= RPC_CLNT_CREATE_NONPRIVPORT; clnt = rpc_create(&args); if (!IS_ERR(clnt)) diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c index 4e7e958e8f6..ffd3461f75e 100644 --- a/fs/lockd/mon.c +++ b/fs/lockd/mon.c @@ -179,7 +179,7 @@ static __be32 *xdr_encode_mon_name(__be32 *p, struct nsm_args *argp) if (!nsm_use_hostnames) { snprintf(buffer, XDR_ADDRBUF_LEN, - NIPQUAD_FMT, NIPQUAD(argp->addr)); + "%pI4", &argp->addr); name = buffer; } diff --git a/fs/lockd/svc.c b/fs/lockd/svc.c index c631a83931c..252d80163d0 100644 --- a/fs/lockd/svc.c +++ b/fs/lockd/svc.c @@ -45,7 +45,7 @@ static struct svc_program nlmsvc_program; struct nlmsvc_binding * nlmsvc_ops; -EXPORT_SYMBOL(nlmsvc_ops); +EXPORT_SYMBOL_GPL(nlmsvc_ops); static DEFINE_MUTEX(nlmsvc_mutex); static unsigned int nlmsvc_users; @@ -181,6 +181,7 @@ lockd(void *vrqstp) } flush_signals(current); cancel_delayed_work_sync(&grace_period_end); + locks_end_grace(&lockd_manager); if (nlmsvc_ops) nlmsvc_invalidate_all(); nlm_shutdown_hosts(); @@ -299,7 +300,7 @@ out: mutex_unlock(&nlmsvc_mutex); return error; } -EXPORT_SYMBOL(lockd_up); +EXPORT_SYMBOL_GPL(lockd_up); /* * Decrement the user count and bring down lockd if we're the last. @@ -328,7 +329,7 @@ lockd_down(void) out: mutex_unlock(&nlmsvc_mutex); } -EXPORT_SYMBOL(lockd_down); +EXPORT_SYMBOL_GPL(lockd_down); #ifdef CONFIG_SYSCTL diff --git a/fs/lockd/svc4proc.c b/fs/lockd/svc4proc.c index 014f6ce4817..4dfdcbc6bf6 100644 --- a/fs/lockd/svc4proc.c +++ b/fs/lockd/svc4proc.c @@ -434,6 +434,7 @@ nlm4svc_proc_sm_notify(struct svc_rqst *rqstp, struct nlm_reboot *argp, * reclaim all locks we hold on this server. */ memset(&saddr, 0, sizeof(saddr)); + saddr.sin_family = AF_INET; saddr.sin_addr.s_addr = argp->addr; nlm_host_rebooted(&saddr, argp->mon, argp->len, argp->state); diff --git a/fs/lockd/svcproc.c b/fs/lockd/svcproc.c index 548b0bb2b84..3ca89e2a938 100644 --- a/fs/lockd/svcproc.c +++ b/fs/lockd/svcproc.c @@ -466,6 +466,7 @@ nlmsvc_proc_sm_notify(struct svc_rqst *rqstp, struct nlm_reboot *argp, * reclaim all locks we hold on this server. */ memset(&saddr, 0, sizeof(saddr)); + saddr.sin_family = AF_INET; saddr.sin_addr.s_addr = argp->addr; nlm_host_rebooted(&saddr, argp->mon, argp->len, argp->state); diff --git a/fs/locks.c b/fs/locks.c index 5eb259e3cd3..46a2e12f7d4 100644 --- a/fs/locks.c +++ b/fs/locks.c @@ -1349,7 +1349,7 @@ int generic_setlease(struct file *filp, long arg, struct file_lock **flp) struct inode *inode = dentry->d_inode; int error, rdlease_count = 0, wrlease_count = 0; - if ((current->fsuid != inode->i_uid) && !capable(CAP_LEASE)) + if ((current_fsuid() != inode->i_uid) && !capable(CAP_LEASE)) return -EACCES; if (!S_ISREG(inode->i_mode)) return -EINVAL; @@ -1580,7 +1580,8 @@ asmlinkage long sys_flock(unsigned int fd, unsigned int cmd) cmd &= ~LOCK_NB; unlock = (cmd == LOCK_UN); - if (!unlock && !(cmd & LOCK_MAND) && !(filp->f_mode & 3)) + if (!unlock && !(cmd & LOCK_MAND) && + !(filp->f_mode & (FMODE_READ|FMODE_WRITE))) goto out_putf; error = flock_make_lock(filp, &lock, cmd); @@ -2078,6 +2079,7 @@ int vfs_cancel_lock(struct file *filp, struct file_lock *fl) EXPORT_SYMBOL_GPL(vfs_cancel_lock); #ifdef CONFIG_PROC_FS +#include <linux/proc_fs.h> #include <linux/seq_file.h> static void lock_get_status(struct seq_file *f, struct file_lock *fl, @@ -2183,12 +2185,31 @@ static void locks_stop(struct seq_file *f, void *v) unlock_kernel(); } -struct seq_operations locks_seq_operations = { +static const struct seq_operations locks_seq_operations = { .start = locks_start, .next = locks_next, .stop = locks_stop, .show = locks_show, }; + +static int locks_open(struct inode *inode, struct file *filp) +{ + return seq_open(filp, &locks_seq_operations); +} + +static const struct file_operations proc_locks_operations = { + .open = locks_open, + .read = seq_read, + .llseek = seq_lseek, + .release = seq_release, +}; + +static int __init proc_locks_init(void) +{ + proc_create("locks", 0, NULL, &proc_locks_operations); + return 0; +} +module_init(proc_locks_init); #endif /** diff --git a/fs/minix/bitmap.c b/fs/minix/bitmap.c index 703cc35e04b..3aebe322271 100644 --- a/fs/minix/bitmap.c +++ b/fs/minix/bitmap.c @@ -262,8 +262,8 @@ struct inode * minix_new_inode(const struct inode * dir, int * error) iput(inode); return NULL; } - inode->i_uid = current->fsuid; - inode->i_gid = (dir->i_mode & S_ISGID) ? dir->i_gid : current->fsgid; + inode->i_uid = current_fsuid(); + inode->i_gid = (dir->i_mode & S_ISGID) ? dir->i_gid : current_fsgid(); inode->i_ino = j; inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME_SEC; inode->i_blocks = 0; diff --git a/fs/mpage.c b/fs/mpage.c index dbcc7af76a1..552b80b3fac 100644 --- a/fs/mpage.c +++ b/fs/mpage.c @@ -6,7 +6,7 @@ * Contains functions related to preparing and submitting BIOs which contain * multiple pagecache pages. * - * 15May2002 akpm@zip.com.au + * 15May2002 Andrew Morton * Initial version * 27Jun2002 axboe@suse.de * use bio_add_page() to build bio's just the right size diff --git a/fs/msdos/Makefile b/fs/msdos/Makefile deleted file mode 100644 index ea67646fcb9..00000000000 --- a/fs/msdos/Makefile +++ /dev/null @@ -1,7 +0,0 @@ -# -# Makefile for the Linux msdos filesystem routines. -# - -obj-$(CONFIG_MSDOS_FS) += msdos.o - -msdos-y := namei.o diff --git a/fs/namei.c b/fs/namei.c index 4ea63ed5e79..af3783fff1d 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -186,7 +186,7 @@ int generic_permission(struct inode *inode, int mask, mask &= MAY_READ | MAY_WRITE | MAY_EXEC; - if (current->fsuid == inode->i_uid) + if (current_fsuid() == inode->i_uid) mode >>= 6; else { if (IS_POSIXACL(inode) && (mode & S_IRWXG) && check_acl) { @@ -212,8 +212,7 @@ int generic_permission(struct inode *inode, int mask, * Read/write DACs are always overridable. * Executable DACs are overridable if at least one exec bit is set. */ - if (!(mask & MAY_EXEC) || - (inode->i_mode & S_IXUGO) || S_ISDIR(inode->i_mode)) + if (!(mask & MAY_EXEC) || execute_ok(inode)) if (capable(CAP_DAC_OVERRIDE)) return 0; @@ -249,23 +248,11 @@ int inode_permission(struct inode *inode, int mask) } /* Ordinary permission routines do not understand MAY_APPEND. */ - if (inode->i_op && inode->i_op->permission) { + if (inode->i_op && inode->i_op->permission) retval = inode->i_op->permission(inode, mask); - if (!retval) { - /* - * Exec permission on a regular file is denied if none - * of the execute bits are set. - * - * This check should be done by the ->permission() - * method. - */ - if ((mask & MAY_EXEC) && S_ISREG(inode->i_mode) && - !(inode->i_mode & S_IXUGO)) - return -EACCES; - } - } else { + else retval = generic_permission(inode, mask, NULL); - } + if (retval) return retval; @@ -454,7 +441,7 @@ static int exec_permission_lite(struct inode *inode) if (inode->i_op && inode->i_op->permission) return -EAGAIN; - if (current->fsuid == inode->i_uid) + if (current_fsuid() == inode->i_uid) mode >>= 6; else if (in_group_p(inode->i_gid)) mode >>= 3; @@ -1106,6 +1093,15 @@ int path_lookup(const char *name, unsigned int flags, return do_path_lookup(AT_FDCWD, name, flags, nd); } +int kern_path(const char *name, unsigned int flags, struct path *path) +{ + struct nameidata nd; + int res = do_path_lookup(AT_FDCWD, name, flags, &nd); + if (!res) + *path = nd.path; + return res; +} + /** * vfs_path_lookup - lookup a file path relative to a dentry-vfsmount pair * @dentry: pointer to dentry of the base directory @@ -1138,9 +1134,16 @@ int vfs_path_lookup(struct dentry *dentry, struct vfsmount *mnt, } -static int __path_lookup_intent_open(int dfd, const char *name, - unsigned int lookup_flags, struct nameidata *nd, - int open_flags, int create_mode) +/** + * path_lookup_open - lookup a file path with open intent + * @dfd: the directory to use as base, or AT_FDCWD + * @name: pointer to file name + * @lookup_flags: lookup intent flags + * @nd: pointer to nameidata + * @open_flags: open intent flags + */ +int path_lookup_open(int dfd, const char *name, unsigned int lookup_flags, + struct nameidata *nd, int open_flags) { struct file *filp = get_empty_filp(); int err; @@ -1149,7 +1152,7 @@ static int __path_lookup_intent_open(int dfd, const char *name, return -ENFILE; nd->intent.open.file = filp; nd->intent.open.flags = open_flags; - nd->intent.open.create_mode = create_mode; + nd->intent.open.create_mode = 0; err = do_path_lookup(dfd, name, lookup_flags|LOOKUP_OPEN, nd); if (IS_ERR(nd->intent.open.file)) { if (err == 0) { @@ -1161,38 +1164,6 @@ static int __path_lookup_intent_open(int dfd, const char *name, return err; } -/** - * path_lookup_open - lookup a file path with open intent - * @dfd: the directory to use as base, or AT_FDCWD - * @name: pointer to file name - * @lookup_flags: lookup intent flags - * @nd: pointer to nameidata - * @open_flags: open intent flags - */ -int path_lookup_open(int dfd, const char *name, unsigned int lookup_flags, - struct nameidata *nd, int open_flags) -{ - return __path_lookup_intent_open(dfd, name, lookup_flags, nd, - open_flags, 0); -} - -/** - * path_lookup_create - lookup a file path with open + create intent - * @dfd: the directory to use as base, or AT_FDCWD - * @name: pointer to file name - * @lookup_flags: lookup intent flags - * @nd: pointer to nameidata - * @open_flags: open intent flags - * @create_mode: create intent flags - */ -static int path_lookup_create(int dfd, const char *name, - unsigned int lookup_flags, struct nameidata *nd, - int open_flags, int create_mode) -{ - return __path_lookup_intent_open(dfd, name, lookup_flags|LOOKUP_CREATE, - nd, open_flags, create_mode); -} - static struct dentry *__lookup_hash(struct qstr *name, struct dentry *base, struct nameidata *nd) { @@ -1363,11 +1334,13 @@ static int user_path_parent(int dfd, const char __user *path, */ static inline int check_sticky(struct inode *dir, struct inode *inode) { + uid_t fsuid = current_fsuid(); + if (!(dir->i_mode & S_ISVTX)) return 0; - if (inode->i_uid == current->fsuid) + if (inode->i_uid == fsuid) return 0; - if (dir->i_uid == current->fsuid) + if (dir->i_uid == fsuid) return 0; return !capable(CAP_FOWNER); } @@ -1407,7 +1380,7 @@ static int may_delete(struct inode *dir,struct dentry *victim,int isdir) if (IS_APPEND(dir)) return -EPERM; if (check_sticky(dir, victim->d_inode)||IS_APPEND(victim->d_inode)|| - IS_IMMUTABLE(victim->d_inode)) + IS_IMMUTABLE(victim->d_inode) || IS_SWAPFILE(victim->d_inode)) return -EPERM; if (isdir) { if (!S_ISDIR(victim->d_inode->i_mode)) @@ -1470,20 +1443,18 @@ struct dentry *lock_rename(struct dentry *p1, struct dentry *p2) mutex_lock(&p1->d_inode->i_sb->s_vfs_rename_mutex); - for (p = p1; p->d_parent != p; p = p->d_parent) { - if (p->d_parent == p2) { - mutex_lock_nested(&p2->d_inode->i_mutex, I_MUTEX_PARENT); - mutex_lock_nested(&p1->d_inode->i_mutex, I_MUTEX_CHILD); - return p; - } + p = d_ancestor(p2, p1); + if (p) { + mutex_lock_nested(&p2->d_inode->i_mutex, I_MUTEX_PARENT); + mutex_lock_nested(&p1->d_inode->i_mutex, I_MUTEX_CHILD); + return p; } - for (p = p2; p->d_parent != p; p = p->d_parent) { - if (p->d_parent == p1) { - mutex_lock_nested(&p1->d_inode->i_mutex, I_MUTEX_PARENT); - mutex_lock_nested(&p2->d_inode->i_mutex, I_MUTEX_CHILD); - return p; - } + p = d_ancestor(p1, p2); + if (p) { + mutex_lock_nested(&p1->d_inode->i_mutex, I_MUTEX_PARENT); + mutex_lock_nested(&p2->d_inode->i_mutex, I_MUTEX_CHILD); + return p; } mutex_lock_nested(&p1->d_inode->i_mutex, I_MUTEX_PARENT); @@ -1702,8 +1673,7 @@ struct file *do_filp_open(int dfd, const char *pathname, /* * Create - we need to know the parent. */ - error = path_lookup_create(dfd, pathname, LOOKUP_PARENT, - &nd, flag, mode); + error = do_path_lookup(dfd, pathname, LOOKUP_PARENT, &nd); if (error) return ERR_PTR(error); @@ -1714,10 +1684,20 @@ struct file *do_filp_open(int dfd, const char *pathname, */ error = -EISDIR; if (nd.last_type != LAST_NORM || nd.last.name[nd.last.len]) - goto exit; + goto exit_parent; + error = -ENFILE; + filp = get_empty_filp(); + if (filp == NULL) + goto exit_parent; + nd.intent.open.file = filp; + nd.intent.open.flags = flag; + nd.intent.open.create_mode = mode; dir = nd.path.dentry; nd.flags &= ~LOOKUP_PARENT; + nd.flags |= LOOKUP_CREATE | LOOKUP_OPEN; + if (flag & O_EXCL) + nd.flags |= LOOKUP_EXCL; mutex_lock(&dir->d_inode->i_mutex); path.dentry = lookup_hash(&nd); path.mnt = nd.path.mnt; @@ -1822,6 +1802,7 @@ exit_dput: exit: if (!IS_ERR(nd.intent.open.file)) release_open_intent(&nd); +exit_parent: path_put(&nd.path); return ERR_PTR(error); @@ -1914,7 +1895,7 @@ struct dentry *lookup_create(struct nameidata *nd, int is_dir) if (nd->last_type != LAST_NORM) goto fail; nd->flags &= ~LOOKUP_PARENT; - nd->flags |= LOOKUP_CREATE; + nd->flags |= LOOKUP_CREATE | LOOKUP_EXCL; nd->intent.open.flags = O_EXCL; /* @@ -2178,16 +2159,19 @@ static long do_rmdir(int dfd, const char __user *pathname) return error; switch(nd.last_type) { - case LAST_DOTDOT: - error = -ENOTEMPTY; - goto exit1; - case LAST_DOT: - error = -EINVAL; - goto exit1; - case LAST_ROOT: - error = -EBUSY; - goto exit1; + case LAST_DOTDOT: + error = -ENOTEMPTY; + goto exit1; + case LAST_DOT: + error = -EINVAL; + goto exit1; + case LAST_ROOT: + error = -EBUSY; + goto exit1; } + + nd.flags &= ~LOOKUP_PARENT; + mutex_lock_nested(&nd.path.dentry->d_inode->i_mutex, I_MUTEX_PARENT); dentry = lookup_hash(&nd); error = PTR_ERR(dentry); @@ -2265,6 +2249,9 @@ static long do_unlinkat(int dfd, const char __user *pathname) error = -EISDIR; if (nd.last_type != LAST_NORM) goto exit1; + + nd.flags &= ~LOOKUP_PARENT; + mutex_lock_nested(&nd.path.dentry->d_inode->i_mutex, I_MUTEX_PARENT); dentry = lookup_hash(&nd); error = PTR_ERR(dentry); @@ -2654,6 +2641,10 @@ asmlinkage long sys_renameat(int olddfd, const char __user *oldname, if (newnd.last_type != LAST_NORM) goto exit2; + oldnd.flags &= ~LOOKUP_PARENT; + newnd.flags &= ~LOOKUP_PARENT; + newnd.flags |= LOOKUP_RENAME_TARGET; + trap = lock_rename(new_dir, old_dir); old_dentry = lookup_hash(&oldnd); @@ -2855,6 +2846,7 @@ EXPORT_SYMBOL(__page_symlink); EXPORT_SYMBOL(page_symlink); EXPORT_SYMBOL(page_symlink_inode_operations); EXPORT_SYMBOL(path_lookup); +EXPORT_SYMBOL(kern_path); EXPORT_SYMBOL(vfs_path_lookup); EXPORT_SYMBOL(inode_permission); EXPORT_SYMBOL(vfs_permission); diff --git a/fs/namespace.c b/fs/namespace.c index 6e283c93b50..1c09cab8f7c 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -1167,19 +1167,19 @@ asmlinkage long sys_oldumount(char __user * name) #endif -static int mount_is_safe(struct nameidata *nd) +static int mount_is_safe(struct path *path) { if (capable(CAP_SYS_ADMIN)) return 0; return -EPERM; #ifdef notyet - if (S_ISLNK(nd->path.dentry->d_inode->i_mode)) + if (S_ISLNK(path->dentry->d_inode->i_mode)) return -EPERM; - if (nd->path.dentry->d_inode->i_mode & S_ISVTX) { - if (current->uid != nd->path.dentry->d_inode->i_uid) + if (path->dentry->d_inode->i_mode & S_ISVTX) { + if (current_uid() != path->dentry->d_inode->i_uid) return -EPERM; } - if (vfs_permission(nd, MAY_WRITE)) + if (inode_permission(path->dentry->d_inode, MAY_WRITE)) return -EPERM; return 0; #endif @@ -1425,11 +1425,10 @@ out_unlock: /* * recursively change the type of the mountpoint. - * noinline this do_mount helper to save do_mount stack space. */ -static noinline int do_change_type(struct nameidata *nd, int flag) +static int do_change_type(struct path *path, int flag) { - struct vfsmount *m, *mnt = nd->path.mnt; + struct vfsmount *m, *mnt = path->mnt; int recurse = flag & MS_REC; int type = flag & ~MS_REC; int err = 0; @@ -1437,7 +1436,7 @@ static noinline int do_change_type(struct nameidata *nd, int flag) if (!capable(CAP_SYS_ADMIN)) return -EPERM; - if (nd->path.dentry != nd->path.mnt->mnt_root) + if (path->dentry != path->mnt->mnt_root) return -EINVAL; down_write(&namespace_sem); @@ -1459,40 +1458,39 @@ static noinline int do_change_type(struct nameidata *nd, int flag) /* * do loopback mount. - * noinline this do_mount helper to save do_mount stack space. */ -static noinline int do_loopback(struct nameidata *nd, char *old_name, +static int do_loopback(struct path *path, char *old_name, int recurse) { - struct nameidata old_nd; + struct path old_path; struct vfsmount *mnt = NULL; - int err = mount_is_safe(nd); + int err = mount_is_safe(path); if (err) return err; if (!old_name || !*old_name) return -EINVAL; - err = path_lookup(old_name, LOOKUP_FOLLOW, &old_nd); + err = kern_path(old_name, LOOKUP_FOLLOW, &old_path); if (err) return err; down_write(&namespace_sem); err = -EINVAL; - if (IS_MNT_UNBINDABLE(old_nd.path.mnt)) + if (IS_MNT_UNBINDABLE(old_path.mnt)) goto out; - if (!check_mnt(nd->path.mnt) || !check_mnt(old_nd.path.mnt)) + if (!check_mnt(path->mnt) || !check_mnt(old_path.mnt)) goto out; err = -ENOMEM; if (recurse) - mnt = copy_tree(old_nd.path.mnt, old_nd.path.dentry, 0); + mnt = copy_tree(old_path.mnt, old_path.dentry, 0); else - mnt = clone_mnt(old_nd.path.mnt, old_nd.path.dentry, 0); + mnt = clone_mnt(old_path.mnt, old_path.dentry, 0); if (!mnt) goto out; - err = graft_tree(mnt, &nd->path); + err = graft_tree(mnt, path); if (err) { LIST_HEAD(umount_list); spin_lock(&vfsmount_lock); @@ -1503,7 +1501,7 @@ static noinline int do_loopback(struct nameidata *nd, char *old_name, out: up_write(&namespace_sem); - path_put(&old_nd.path); + path_put(&old_path); return err; } @@ -1528,33 +1526,37 @@ static int change_mount_flags(struct vfsmount *mnt, int ms_flags) * change filesystem flags. dir should be a physical root of filesystem. * If you've mounted a non-root directory somewhere and want to do remount * on it - tough luck. - * noinline this do_mount helper to save do_mount stack space. */ -static noinline int do_remount(struct nameidata *nd, int flags, int mnt_flags, +static int do_remount(struct path *path, int flags, int mnt_flags, void *data) { int err; - struct super_block *sb = nd->path.mnt->mnt_sb; + struct super_block *sb = path->mnt->mnt_sb; if (!capable(CAP_SYS_ADMIN)) return -EPERM; - if (!check_mnt(nd->path.mnt)) + if (!check_mnt(path->mnt)) return -EINVAL; - if (nd->path.dentry != nd->path.mnt->mnt_root) + if (path->dentry != path->mnt->mnt_root) return -EINVAL; down_write(&sb->s_umount); if (flags & MS_BIND) - err = change_mount_flags(nd->path.mnt, flags); + err = change_mount_flags(path->mnt, flags); else err = do_remount_sb(sb, flags, data, 0); if (!err) - nd->path.mnt->mnt_flags = mnt_flags; + path->mnt->mnt_flags = mnt_flags; up_write(&sb->s_umount); - if (!err) - security_sb_post_remount(nd->path.mnt, flags, data); + if (!err) { + security_sb_post_remount(path->mnt, flags, data); + + spin_lock(&vfsmount_lock); + touch_mnt_namespace(path->mnt->mnt_ns); + spin_unlock(&vfsmount_lock); + } return err; } @@ -1568,90 +1570,85 @@ static inline int tree_contains_unbindable(struct vfsmount *mnt) return 0; } -/* - * noinline this do_mount helper to save do_mount stack space. - */ -static noinline int do_move_mount(struct nameidata *nd, char *old_name) +static int do_move_mount(struct path *path, char *old_name) { - struct nameidata old_nd; - struct path parent_path; + struct path old_path, parent_path; struct vfsmount *p; int err = 0; if (!capable(CAP_SYS_ADMIN)) return -EPERM; if (!old_name || !*old_name) return -EINVAL; - err = path_lookup(old_name, LOOKUP_FOLLOW, &old_nd); + err = kern_path(old_name, LOOKUP_FOLLOW, &old_path); if (err) return err; down_write(&namespace_sem); - while (d_mountpoint(nd->path.dentry) && - follow_down(&nd->path.mnt, &nd->path.dentry)) + while (d_mountpoint(path->dentry) && + follow_down(&path->mnt, &path->dentry)) ; err = -EINVAL; - if (!check_mnt(nd->path.mnt) || !check_mnt(old_nd.path.mnt)) + if (!check_mnt(path->mnt) || !check_mnt(old_path.mnt)) goto out; err = -ENOENT; - mutex_lock(&nd->path.dentry->d_inode->i_mutex); - if (IS_DEADDIR(nd->path.dentry->d_inode)) + mutex_lock(&path->dentry->d_inode->i_mutex); + if (IS_DEADDIR(path->dentry->d_inode)) goto out1; - if (!IS_ROOT(nd->path.dentry) && d_unhashed(nd->path.dentry)) + if (!IS_ROOT(path->dentry) && d_unhashed(path->dentry)) goto out1; err = -EINVAL; - if (old_nd.path.dentry != old_nd.path.mnt->mnt_root) + if (old_path.dentry != old_path.mnt->mnt_root) goto out1; - if (old_nd.path.mnt == old_nd.path.mnt->mnt_parent) + if (old_path.mnt == old_path.mnt->mnt_parent) goto out1; - if (S_ISDIR(nd->path.dentry->d_inode->i_mode) != - S_ISDIR(old_nd.path.dentry->d_inode->i_mode)) + if (S_ISDIR(path->dentry->d_inode->i_mode) != + S_ISDIR(old_path.dentry->d_inode->i_mode)) goto out1; /* * Don't move a mount residing in a shared parent. */ - if (old_nd.path.mnt->mnt_parent && - IS_MNT_SHARED(old_nd.path.mnt->mnt_parent)) + if (old_path.mnt->mnt_parent && + IS_MNT_SHARED(old_path.mnt->mnt_parent)) goto out1; /* * Don't move a mount tree containing unbindable mounts to a destination * mount which is shared. */ - if (IS_MNT_SHARED(nd->path.mnt) && - tree_contains_unbindable(old_nd.path.mnt)) + if (IS_MNT_SHARED(path->mnt) && + tree_contains_unbindable(old_path.mnt)) goto out1; err = -ELOOP; - for (p = nd->path.mnt; p->mnt_parent != p; p = p->mnt_parent) - if (p == old_nd.path.mnt) + for (p = path->mnt; p->mnt_parent != p; p = p->mnt_parent) + if (p == old_path.mnt) goto out1; - err = attach_recursive_mnt(old_nd.path.mnt, &nd->path, &parent_path); + err = attach_recursive_mnt(old_path.mnt, path, &parent_path); if (err) goto out1; /* if the mount is moved, it should no longer be expire * automatically */ - list_del_init(&old_nd.path.mnt->mnt_expire); + list_del_init(&old_path.mnt->mnt_expire); out1: - mutex_unlock(&nd->path.dentry->d_inode->i_mutex); + mutex_unlock(&path->dentry->d_inode->i_mutex); out: up_write(&namespace_sem); if (!err) path_put(&parent_path); - path_put(&old_nd.path); + path_put(&old_path); return err; } /* * create a new mount for userspace and request it to be added into the * namespace's tree - * noinline this do_mount helper to save do_mount stack space. */ -static noinline int do_new_mount(struct nameidata *nd, char *type, int flags, +static int do_new_mount(struct path *path, char *type, int flags, int mnt_flags, char *name, void *data) { struct vfsmount *mnt; @@ -1667,7 +1664,7 @@ static noinline int do_new_mount(struct nameidata *nd, char *type, int flags, if (IS_ERR(mnt)) return PTR_ERR(mnt); - return do_add_mount(mnt, &nd->path, mnt_flags, NULL); + return do_add_mount(mnt, path, mnt_flags, NULL); } /* @@ -1818,8 +1815,8 @@ static void shrink_submounts(struct vfsmount *mnt, struct list_head *umounts) while (!list_empty(&graveyard)) { m = list_first_entry(&graveyard, struct vfsmount, mnt_expire); - touch_mnt_namespace(mnt->mnt_ns); - umount_tree(mnt, 1, umounts); + touch_mnt_namespace(m->mnt_ns); + umount_tree(m, 1, umounts); } } } @@ -1902,7 +1899,7 @@ int copy_mount_options(const void __user * data, unsigned long *where) long do_mount(char *dev_name, char *dir_name, char *type_page, unsigned long flags, void *data_page) { - struct nameidata nd; + struct path path; int retval = 0; int mnt_flags = 0; @@ -1940,29 +1937,29 @@ long do_mount(char *dev_name, char *dir_name, char *type_page, MS_NOATIME | MS_NODIRATIME | MS_RELATIME| MS_KERNMOUNT); /* ... and get the mountpoint */ - retval = path_lookup(dir_name, LOOKUP_FOLLOW, &nd); + retval = kern_path(dir_name, LOOKUP_FOLLOW, &path); if (retval) return retval; - retval = security_sb_mount(dev_name, &nd.path, + retval = security_sb_mount(dev_name, &path, type_page, flags, data_page); if (retval) goto dput_out; if (flags & MS_REMOUNT) - retval = do_remount(&nd, flags & ~MS_REMOUNT, mnt_flags, + retval = do_remount(&path, flags & ~MS_REMOUNT, mnt_flags, data_page); else if (flags & MS_BIND) - retval = do_loopback(&nd, dev_name, flags & MS_REC); + retval = do_loopback(&path, dev_name, flags & MS_REC); else if (flags & (MS_SHARED | MS_PRIVATE | MS_SLAVE | MS_UNBINDABLE)) - retval = do_change_type(&nd, flags); + retval = do_change_type(&path, flags); else if (flags & MS_MOVE) - retval = do_move_mount(&nd, dev_name); + retval = do_move_mount(&path, dev_name); else - retval = do_new_mount(&nd, type_page, flags, mnt_flags, + retval = do_new_mount(&path, type_page, flags, mnt_flags, dev_name, data_page); dput_out: - path_put(&nd.path); + path_put(&path); return retval; } diff --git a/fs/ncpfs/ioctl.c b/fs/ncpfs/ioctl.c index 3a97c95e1ca..6d04e050c74 100644 --- a/fs/ncpfs/ioctl.c +++ b/fs/ncpfs/ioctl.c @@ -40,10 +40,10 @@ ncp_get_fs_info(struct ncp_server * server, struct file *file, struct inode *inode = file->f_path.dentry->d_inode; struct ncp_fs_info info; - if ((file_permission(file, MAY_WRITE) != 0) - && (current->uid != server->m.mounted_uid)) { + if (file_permission(file, MAY_WRITE) != 0 + && current_uid() != server->m.mounted_uid) return -EACCES; - } + if (copy_from_user(&info, arg, sizeof(info))) return -EFAULT; @@ -70,10 +70,10 @@ ncp_get_fs_info_v2(struct ncp_server * server, struct file *file, struct inode *inode = file->f_path.dentry->d_inode; struct ncp_fs_info_v2 info2; - if ((file_permission(file, MAY_WRITE) != 0) - && (current->uid != server->m.mounted_uid)) { + if (file_permission(file, MAY_WRITE) != 0 + && current_uid() != server->m.mounted_uid) return -EACCES; - } + if (copy_from_user(&info2, arg, sizeof(info2))) return -EFAULT; @@ -141,10 +141,10 @@ ncp_get_compat_fs_info_v2(struct ncp_server * server, struct file *file, struct inode *inode = file->f_path.dentry->d_inode; struct compat_ncp_fs_info_v2 info2; - if ((file_permission(file, MAY_WRITE) != 0) - && (current->uid != server->m.mounted_uid)) { + if (file_permission(file, MAY_WRITE) != 0 + && current_uid() != server->m.mounted_uid) return -EACCES; - } + if (copy_from_user(&info2, arg, sizeof(info2))) return -EFAULT; @@ -270,16 +270,17 @@ static int __ncp_ioctl(struct inode *inode, struct file *filp, struct ncp_ioctl_request request; char* bouncebuffer; void __user *argp = (void __user *)arg; + uid_t uid = current_uid(); switch (cmd) { #ifdef CONFIG_COMPAT case NCP_IOC_NCPREQUEST_32: #endif case NCP_IOC_NCPREQUEST: - if ((file_permission(filp, MAY_WRITE) != 0) - && (current->uid != server->m.mounted_uid)) { + if (file_permission(filp, MAY_WRITE) != 0 + && uid != server->m.mounted_uid) return -EACCES; - } + #ifdef CONFIG_COMPAT if (cmd == NCP_IOC_NCPREQUEST_32) { struct compat_ncp_ioctl_request request32; @@ -356,10 +357,10 @@ static int __ncp_ioctl(struct inode *inode, struct file *filp, case NCP_IOC_GETMOUNTUID16: case NCP_IOC_GETMOUNTUID32: case NCP_IOC_GETMOUNTUID64: - if ((file_permission(filp, MAY_READ) != 0) - && (current->uid != server->m.mounted_uid)) { + if (file_permission(filp, MAY_READ) != 0 + && uid != server->m.mounted_uid) return -EACCES; - } + if (cmd == NCP_IOC_GETMOUNTUID16) { u16 uid; SET_UID(uid, server->m.mounted_uid); @@ -380,11 +381,10 @@ static int __ncp_ioctl(struct inode *inode, struct file *filp, { struct ncp_setroot_ioctl sr; - if ((file_permission(filp, MAY_READ) != 0) - && (current->uid != server->m.mounted_uid)) - { + if (file_permission(filp, MAY_READ) != 0 + && uid != server->m.mounted_uid) return -EACCES; - } + if (server->m.mounted_vol[0]) { struct dentry* dentry = inode->i_sb->s_root; @@ -408,6 +408,7 @@ static int __ncp_ioctl(struct inode *inode, struct file *filp, return -EFAULT; return 0; } + case NCP_IOC_SETROOT: { struct ncp_setroot_ioctl sr; @@ -455,11 +456,10 @@ static int __ncp_ioctl(struct inode *inode, struct file *filp, #ifdef CONFIG_NCPFS_PACKET_SIGNING case NCP_IOC_SIGN_INIT: - if ((file_permission(filp, MAY_WRITE) != 0) - && (current->uid != server->m.mounted_uid)) - { + if (file_permission(filp, MAY_WRITE) != 0 + && uid != server->m.mounted_uid) return -EACCES; - } + if (argp) { if (server->sign_wanted) { @@ -478,24 +478,22 @@ static int __ncp_ioctl(struct inode *inode, struct file *filp, return 0; case NCP_IOC_SIGN_WANTED: - if ((file_permission(filp, MAY_READ) != 0) - && (current->uid != server->m.mounted_uid)) - { + if (file_permission(filp, MAY_READ) != 0 + && uid != server->m.mounted_uid) return -EACCES; - } if (put_user(server->sign_wanted, (int __user *)argp)) return -EFAULT; return 0; + case NCP_IOC_SET_SIGN_WANTED: { int newstate; - if ((file_permission(filp, MAY_WRITE) != 0) - && (current->uid != server->m.mounted_uid)) - { + if (file_permission(filp, MAY_WRITE) != 0 + && uid != server->m.mounted_uid) return -EACCES; - } + /* get only low 8 bits... */ if (get_user(newstate, (unsigned char __user *)argp)) return -EFAULT; @@ -512,11 +510,10 @@ static int __ncp_ioctl(struct inode *inode, struct file *filp, #ifdef CONFIG_NCPFS_IOCTL_LOCKING case NCP_IOC_LOCKUNLOCK: - if ((file_permission(filp, MAY_WRITE) != 0) - && (current->uid != server->m.mounted_uid)) - { + if (file_permission(filp, MAY_WRITE) != 0 + && uid != server->m.mounted_uid) return -EACCES; - } + { struct ncp_lock_ioctl rqdata; @@ -585,9 +582,8 @@ outrel: #ifdef CONFIG_COMPAT case NCP_IOC_GETOBJECTNAME_32: - if (current->uid != server->m.mounted_uid) { + if (uid != server->m.mounted_uid) return -EACCES; - } { struct compat_ncp_objectname_ioctl user; size_t outl; @@ -609,10 +605,10 @@ outrel: return 0; } #endif + case NCP_IOC_GETOBJECTNAME: - if (current->uid != server->m.mounted_uid) { + if (uid != server->m.mounted_uid) return -EACCES; - } { struct ncp_objectname_ioctl user; size_t outl; @@ -633,13 +629,13 @@ outrel: return -EFAULT; return 0; } + #ifdef CONFIG_COMPAT case NCP_IOC_SETOBJECTNAME_32: #endif case NCP_IOC_SETOBJECTNAME: - if (current->uid != server->m.mounted_uid) { + if (uid != server->m.mounted_uid) return -EACCES; - } { struct ncp_objectname_ioctl user; void* newname; @@ -691,13 +687,13 @@ outrel: kfree(oldname); return 0; } + #ifdef CONFIG_COMPAT case NCP_IOC_GETPRIVATEDATA_32: #endif case NCP_IOC_GETPRIVATEDATA: - if (current->uid != server->m.mounted_uid) { + if (uid != server->m.mounted_uid) return -EACCES; - } { struct ncp_privatedata_ioctl user; size_t outl; @@ -736,13 +732,13 @@ outrel: return 0; } + #ifdef CONFIG_COMPAT case NCP_IOC_SETPRIVATEDATA_32: #endif case NCP_IOC_SETPRIVATEDATA: - if (current->uid != server->m.mounted_uid) { + if (uid != server->m.mounted_uid) return -EACCES; - } { struct ncp_privatedata_ioctl user; void* new; @@ -794,9 +790,10 @@ outrel: #endif /* CONFIG_NCPFS_NLS */ case NCP_IOC_SETDENTRYTTL: - if ((file_permission(filp, MAY_WRITE) != 0) && - (current->uid != server->m.mounted_uid)) + if (file_permission(filp, MAY_WRITE) != 0 && + uid != server->m.mounted_uid) return -EACCES; + { u_int32_t user; diff --git a/fs/nfs/callback.c b/fs/nfs/callback.c index 6a09760c596..3e634f2a108 100644 --- a/fs/nfs/callback.c +++ b/fs/nfs/callback.c @@ -16,6 +16,7 @@ #include <linux/mutex.h> #include <linux/freezer.h> #include <linux/kthread.h> +#include <linux/sunrpc/svcauth_gss.h> #include <net/inet_sock.h> @@ -40,6 +41,16 @@ unsigned short nfs_callback_tcpport; static const int nfs_set_port_min = 0; static const int nfs_set_port_max = 65535; +/* + * If the kernel has IPv6 support available, always listen for + * both AF_INET and AF_INET6 requests. + */ +#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE) +static const sa_family_t nfs_callback_family = AF_INET6; +#else +static const sa_family_t nfs_callback_family = AF_INET; +#endif + static int param_set_port(const char *val, struct kernel_param *kp) { char *endp; @@ -106,7 +117,7 @@ int nfs_callback_up(void) if (nfs_callback_info.users++ || nfs_callback_info.task != NULL) goto out; serv = svc_create(&nfs4_callback_program, NFS4_CALLBACK_BUFSIZE, - AF_INET, NULL); + nfs_callback_family, NULL); ret = -ENOMEM; if (!serv) goto out_err; @@ -116,7 +127,8 @@ int nfs_callback_up(void) if (ret <= 0) goto out_err; nfs_callback_tcpport = ret; - dprintk("Callback port = 0x%x\n", nfs_callback_tcpport); + dprintk("NFS: Callback listener port = %u (af %u)\n", + nfs_callback_tcpport, nfs_callback_family); nfs_callback_info.rqst = svc_prepare_thread(serv, &serv->sv_pools[0]); if (IS_ERR(nfs_callback_info.rqst)) { @@ -149,8 +161,8 @@ out: mutex_unlock(&nfs_callback_mutex); return ret; out_err: - dprintk("Couldn't create callback socket or server thread; err = %d\n", - ret); + dprintk("NFS: Couldn't create callback socket or server thread; " + "err = %d\n", ret); nfs_callback_info.users--; goto out; } @@ -171,10 +183,34 @@ void nfs_callback_down(void) mutex_unlock(&nfs_callback_mutex); } +static int check_gss_callback_principal(struct nfs_client *clp, + struct svc_rqst *rqstp) +{ + struct rpc_clnt *r = clp->cl_rpcclient; + char *p = svc_gss_principal(rqstp); + + /* + * It might just be a normal user principal, in which case + * userspace won't bother to tell us the name at all. + */ + if (p == NULL) + return SVC_DENIED; + + /* Expect a GSS_C_NT_HOSTBASED_NAME like "nfs@serverhostname" */ + + if (memcmp(p, "nfs@", 4) != 0) + return SVC_DENIED; + p += 4; + if (strcmp(p, r->cl_server) != 0) + return SVC_DENIED; + return SVC_OK; +} + static int nfs_callback_authenticate(struct svc_rqst *rqstp) { struct nfs_client *clp; RPC_IFDEBUG(char buf[RPC_MAX_ADDRBUFLEN]); + int ret = SVC_OK; /* Don't talk to strangers */ clp = nfs_find_client(svc_addr(rqstp), 4); @@ -183,21 +219,22 @@ static int nfs_callback_authenticate(struct svc_rqst *rqstp) dprintk("%s: %s NFSv4 callback!\n", __func__, svc_print_addr(rqstp, buf, sizeof(buf))); - nfs_put_client(clp); switch (rqstp->rq_authop->flavour) { case RPC_AUTH_NULL: if (rqstp->rq_proc != CB_NULL) - return SVC_DENIED; + ret = SVC_DENIED; break; case RPC_AUTH_UNIX: break; case RPC_AUTH_GSS: - /* FIXME: RPCSEC_GSS handling? */ + ret = check_gss_callback_principal(clp, rqstp); + break; default: - return SVC_DENIED; + ret = SVC_DENIED; } - return SVC_OK; + nfs_put_client(clp); + return ret; } /* diff --git a/fs/nfs/client.c b/fs/nfs/client.c index 5ee23e7058b..9b728f3565a 100644 --- a/fs/nfs/client.c +++ b/fs/nfs/client.c @@ -143,7 +143,6 @@ static struct nfs_client *nfs_alloc_client(const struct nfs_client_initdata *cl_ clp->cl_proto = cl_init->proto; #ifdef CONFIG_NFS_V4 - init_rwsem(&clp->cl_sem); INIT_LIST_HEAD(&clp->cl_delegations); spin_lock_init(&clp->cl_lock); INIT_DELAYED_WORK(&clp->cl_renewd, nfs4_renew_state); @@ -224,31 +223,54 @@ void nfs_put_client(struct nfs_client *clp) } } -static int nfs_sockaddr_match_ipaddr4(const struct sockaddr_in *sa1, - const struct sockaddr_in *sa2) +#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE) +static const struct in6_addr *nfs_map_ipv4_addr(const struct sockaddr *sa, struct in6_addr *addr_mapped) { - return sa1->sin_addr.s_addr == sa2->sin_addr.s_addr; + switch (sa->sa_family) { + default: + return NULL; + case AF_INET6: + return &((const struct sockaddr_in6 *)sa)->sin6_addr; + break; + case AF_INET: + ipv6_addr_set_v4mapped(((const struct sockaddr_in *)sa)->sin_addr.s_addr, + addr_mapped); + return addr_mapped; + } } -static int nfs_sockaddr_match_ipaddr6(const struct sockaddr_in6 *sa1, - const struct sockaddr_in6 *sa2) +static int nfs_sockaddr_match_ipaddr(const struct sockaddr *sa1, + const struct sockaddr *sa2) { - return ipv6_addr_equal(&sa1->sin6_addr, &sa2->sin6_addr); + const struct in6_addr *addr1; + const struct in6_addr *addr2; + struct in6_addr addr1_mapped; + struct in6_addr addr2_mapped; + + addr1 = nfs_map_ipv4_addr(sa1, &addr1_mapped); + if (likely(addr1 != NULL)) { + addr2 = nfs_map_ipv4_addr(sa2, &addr2_mapped); + if (likely(addr2 != NULL)) + return ipv6_addr_equal(addr1, addr2); + } + return 0; +} +#else +static int nfs_sockaddr_match_ipaddr4(const struct sockaddr_in *sa1, + const struct sockaddr_in *sa2) +{ + return sa1->sin_addr.s_addr == sa2->sin_addr.s_addr; } static int nfs_sockaddr_match_ipaddr(const struct sockaddr *sa1, const struct sockaddr *sa2) { - switch (sa1->sa_family) { - case AF_INET: - return nfs_sockaddr_match_ipaddr4((const struct sockaddr_in *)sa1, - (const struct sockaddr_in *)sa2); - case AF_INET6: - return nfs_sockaddr_match_ipaddr6((const struct sockaddr_in6 *)sa1, - (const struct sockaddr_in6 *)sa2); - } - BUG(); + if (unlikely(sa1->sa_family != AF_INET || sa2->sa_family != AF_INET)) + return 0; + return nfs_sockaddr_match_ipaddr4((const struct sockaddr_in *)sa1, + (const struct sockaddr_in *)sa2); } +#endif /* * Find a client by IP address and protocol version @@ -270,8 +292,6 @@ struct nfs_client *nfs_find_client(const struct sockaddr *addr, u32 nfsversion) if (clp->rpc_ops->version != nfsversion) continue; - if (addr->sa_family != clap->sa_family) - continue; /* Match only the IP address, not the port number */ if (!nfs_sockaddr_match_ipaddr(addr, clap)) continue; @@ -305,8 +325,6 @@ struct nfs_client *nfs_find_client_next(struct nfs_client *clp) if (clp->rpc_ops->version != nfsvers) continue; - if (sap->sa_family != clap->sa_family) - continue; /* Match only the IP address, not the port number */ if (!nfs_sockaddr_match_ipaddr(sap, clap)) continue; @@ -470,7 +488,7 @@ static void nfs_init_timeout_values(struct rpc_timeout *to, int proto, static int nfs_create_rpc_client(struct nfs_client *clp, const struct rpc_timeout *timeparms, rpc_authflavor_t flavor, - int flags) + int discrtry, int noresvport) { struct rpc_clnt *clnt = NULL; struct rpc_create_args args = { @@ -482,9 +500,13 @@ static int nfs_create_rpc_client(struct nfs_client *clp, .program = &nfs_program, .version = clp->rpc_ops->version, .authflavor = flavor, - .flags = flags, }; + if (discrtry) + args.flags |= RPC_CLNT_CREATE_DISCRTRY; + if (noresvport) + args.flags |= RPC_CLNT_CREATE_NONPRIVPORT; + if (!IS_ERR(clp->cl_rpcclient)) return 0; @@ -522,6 +544,8 @@ static int nfs_start_lockd(struct nfs_server *server) .protocol = server->flags & NFS_MOUNT_TCP ? IPPROTO_TCP : IPPROTO_UDP, .nfs_version = clp->rpc_ops->version, + .noresvport = server->flags & NFS_MOUNT_NORESVPORT ? + 1 : 0, }; if (nlm_init.nfs_version > 3) @@ -623,7 +647,8 @@ static int nfs_init_client(struct nfs_client *clp, * Create a client RPC handle for doing FSSTAT with UNIX auth only * - RFC 2623, sec 2.3.2 */ - error = nfs_create_rpc_client(clp, timeparms, RPC_AUTH_UNIX, 0); + error = nfs_create_rpc_client(clp, timeparms, RPC_AUTH_UNIX, + 0, data->flags & NFS_MOUNT_NORESVPORT); if (error < 0) goto error; nfs_mark_client_ready(clp, NFS_CS_READY); @@ -675,7 +700,7 @@ static int nfs_init_server(struct nfs_server *server, server->nfs_client = clp; /* Initialise the client representation from the mount data */ - server->flags = data->flags & NFS_MOUNT_FLAGMASK; + server->flags = data->flags; if (data->rsize) server->rsize = nfs_block_size(data->rsize, NULL); @@ -850,7 +875,6 @@ static struct nfs_server *nfs_alloc_server(void) INIT_LIST_HEAD(&server->client_link); INIT_LIST_HEAD(&server->master_link); - init_waitqueue_head(&server->active_wq); atomic_set(&server->active, 0); server->io_stats = nfs_alloc_iostats(); @@ -966,7 +990,8 @@ error: static int nfs4_init_client(struct nfs_client *clp, const struct rpc_timeout *timeparms, const char *ip_addr, - rpc_authflavor_t authflavour) + rpc_authflavor_t authflavour, + int flags) { int error; @@ -980,7 +1005,7 @@ static int nfs4_init_client(struct nfs_client *clp, clp->rpc_ops = &nfs_v4_clientops; error = nfs_create_rpc_client(clp, timeparms, authflavour, - RPC_CLNT_CREATE_DISCRTRY); + 1, flags & NFS_MOUNT_NORESVPORT); if (error < 0) goto error; memcpy(clp->cl_ipaddr, ip_addr, sizeof(clp->cl_ipaddr)); @@ -1031,7 +1056,8 @@ static int nfs4_set_client(struct nfs_server *server, error = PTR_ERR(clp); goto error; } - error = nfs4_init_client(clp, timeparms, ip_addr, authflavour); + error = nfs4_init_client(clp, timeparms, ip_addr, authflavour, + server->flags); if (error < 0) goto error_put; @@ -1060,6 +1086,10 @@ static int nfs4_init_server(struct nfs_server *server, nfs_init_timeout_values(&timeparms, data->nfs_server.protocol, data->timeo, data->retrans); + /* Initialise the client representation from the mount data */ + server->flags = data->flags; + server->caps |= NFS_CAP_ATOMIC_OPEN; + /* Get a client record */ error = nfs4_set_client(server, data->nfs_server.hostname, @@ -1072,10 +1102,6 @@ static int nfs4_init_server(struct nfs_server *server, if (error < 0) goto error; - /* Initialise the client representation from the mount data */ - server->flags = data->flags & NFS_MOUNT_FLAGMASK; - server->caps |= NFS_CAP_ATOMIC_OPEN; - if (data->rsize) server->rsize = nfs_block_size(data->rsize, NULL); if (data->wsize) @@ -1178,6 +1204,10 @@ struct nfs_server *nfs4_create_referral_server(struct nfs_clone_mount *data, parent_server = NFS_SB(data->sb); parent_client = parent_server->nfs_client; + /* Initialise the client representation from the parent server */ + nfs_server_copy_userdata(server, parent_server); + server->caps |= NFS_CAP_ATOMIC_OPEN; + /* Get a client representation. * Note: NFSv4 always uses TCP, */ error = nfs4_set_client(server, data->hostname, @@ -1190,10 +1220,6 @@ struct nfs_server *nfs4_create_referral_server(struct nfs_clone_mount *data, if (error < 0) goto error; - /* Initialise the client representation from the parent server */ - nfs_server_copy_userdata(server, parent_server); - server->caps |= NFS_CAP_ATOMIC_OPEN; - error = nfs_init_server_rpcclient(server, parent_server->client->cl_timeout, data->authflavor); if (error < 0) goto error; diff --git a/fs/nfs/delegation.c b/fs/nfs/delegation.c index cc563cfa694..968225a8801 100644 --- a/fs/nfs/delegation.c +++ b/fs/nfs/delegation.c @@ -43,6 +43,27 @@ static void nfs_free_delegation(struct nfs_delegation *delegation) put_rpccred(cred); } +void nfs_mark_delegation_referenced(struct nfs_delegation *delegation) +{ + set_bit(NFS_DELEGATION_REFERENCED, &delegation->flags); +} + +int nfs_have_delegation(struct inode *inode, fmode_t flags) +{ + struct nfs_delegation *delegation; + int ret = 0; + + flags &= FMODE_READ|FMODE_WRITE; + rcu_read_lock(); + delegation = rcu_dereference(NFS_I(inode)->delegation); + if (delegation != NULL && (delegation->type & flags) == flags) { + nfs_mark_delegation_referenced(delegation); + ret = 1; + } + rcu_read_unlock(); + return ret; +} + static int nfs_delegation_claim_locks(struct nfs_open_context *ctx, struct nfs4_state *state) { struct inode *inode = state->inode; @@ -119,7 +140,7 @@ void nfs_inode_reclaim_delegation(struct inode *inode, struct rpc_cred *cred, st delegation->maxsize = res->maxsize; oldcred = delegation->cred; delegation->cred = get_rpccred(cred); - delegation->flags &= ~NFS_DELEGATION_NEED_RECLAIM; + clear_bit(NFS_DELEGATION_NEED_RECLAIM, &delegation->flags); NFS_I(inode)->delegation_state = delegation->type; smp_wmb(); put_rpccred(oldcred); @@ -134,19 +155,35 @@ static int nfs_do_return_delegation(struct inode *inode, struct nfs_delegation * return res; } +static struct inode *nfs_delegation_grab_inode(struct nfs_delegation *delegation) +{ + struct inode *inode = NULL; + + spin_lock(&delegation->lock); + if (delegation->inode != NULL) + inode = igrab(delegation->inode); + spin_unlock(&delegation->lock); + return inode; +} + static struct nfs_delegation *nfs_detach_delegation_locked(struct nfs_inode *nfsi, const nfs4_stateid *stateid) { struct nfs_delegation *delegation = rcu_dereference(nfsi->delegation); if (delegation == NULL) goto nomatch; + spin_lock(&delegation->lock); if (stateid != NULL && memcmp(delegation->stateid.data, stateid->data, sizeof(delegation->stateid.data)) != 0) - goto nomatch; + goto nomatch_unlock; list_del_rcu(&delegation->super_list); + delegation->inode = NULL; nfsi->delegation_state = 0; rcu_assign_pointer(nfsi->delegation, NULL); + spin_unlock(&delegation->lock); return delegation; +nomatch_unlock: + spin_unlock(&delegation->lock); nomatch: return NULL; } @@ -172,6 +209,8 @@ int nfs_inode_set_delegation(struct inode *inode, struct rpc_cred *cred, struct delegation->change_attr = nfsi->change_attr; delegation->cred = get_rpccred(cred); delegation->inode = inode; + delegation->flags = 1<<NFS_DELEGATION_REFERENCED; + spin_lock_init(&delegation->lock); spin_lock(&clp->cl_lock); if (rcu_dereference(nfsi->delegation) != NULL) { @@ -226,22 +265,47 @@ static void nfs_msync_inode(struct inode *inode) */ static int __nfs_inode_return_delegation(struct inode *inode, struct nfs_delegation *delegation) { - struct nfs_client *clp = NFS_SERVER(inode)->nfs_client; struct nfs_inode *nfsi = NFS_I(inode); nfs_msync_inode(inode); - down_read(&clp->cl_sem); /* Guard against new delegated open calls */ down_write(&nfsi->rwsem); nfs_delegation_claim_opens(inode, &delegation->stateid); up_write(&nfsi->rwsem); - up_read(&clp->cl_sem); nfs_msync_inode(inode); return nfs_do_return_delegation(inode, delegation, 1); } /* + * Return all delegations that have been marked for return + */ +void nfs_client_return_marked_delegations(struct nfs_client *clp) +{ + struct nfs_delegation *delegation; + struct inode *inode; + +restart: + rcu_read_lock(); + list_for_each_entry_rcu(delegation, &clp->cl_delegations, super_list) { + if (!test_and_clear_bit(NFS_DELEGATION_RETURN, &delegation->flags)) + continue; + inode = nfs_delegation_grab_inode(delegation); + if (inode == NULL) + continue; + spin_lock(&clp->cl_lock); + delegation = nfs_detach_delegation_locked(NFS_I(inode), NULL); + spin_unlock(&clp->cl_lock); + rcu_read_unlock(); + if (delegation != NULL) + __nfs_inode_return_delegation(inode, delegation); + iput(inode); + goto restart; + } + rcu_read_unlock(); +} + +/* * This function returns the delegation without reclaiming opens * or protecting against delegation reclaims. * It is therefore really only safe to be called from @@ -279,83 +343,55 @@ int nfs_inode_return_delegation(struct inode *inode) return err; } +static void nfs_mark_return_delegation(struct nfs_client *clp, struct nfs_delegation *delegation) +{ + set_bit(NFS_DELEGATION_RETURN, &delegation->flags); + set_bit(NFS4CLNT_DELEGRETURN, &clp->cl_state); +} + /* * Return all delegations associated to a super block */ -void nfs_return_all_delegations(struct super_block *sb) +void nfs_super_return_all_delegations(struct super_block *sb) { struct nfs_client *clp = NFS_SB(sb)->nfs_client; struct nfs_delegation *delegation; - struct inode *inode; if (clp == NULL) return; -restart: rcu_read_lock(); list_for_each_entry_rcu(delegation, &clp->cl_delegations, super_list) { - if (delegation->inode->i_sb != sb) - continue; - inode = igrab(delegation->inode); - if (inode == NULL) - continue; - spin_lock(&clp->cl_lock); - delegation = nfs_detach_delegation_locked(NFS_I(inode), NULL); - spin_unlock(&clp->cl_lock); - rcu_read_unlock(); - if (delegation != NULL) - __nfs_inode_return_delegation(inode, delegation); - iput(inode); - goto restart; + spin_lock(&delegation->lock); + if (delegation->inode != NULL && delegation->inode->i_sb == sb) + set_bit(NFS_DELEGATION_RETURN, &delegation->flags); + spin_unlock(&delegation->lock); } rcu_read_unlock(); + nfs_client_return_marked_delegations(clp); } -static int nfs_do_expire_all_delegations(void *ptr) +static void nfs_client_mark_return_all_delegations(struct nfs_client *clp) { - struct nfs_client *clp = ptr; struct nfs_delegation *delegation; - struct inode *inode; - allow_signal(SIGKILL); -restart: - if (test_bit(NFS4CLNT_STATE_RECOVER, &clp->cl_state) != 0) - goto out; - if (test_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state) == 0) - goto out; rcu_read_lock(); list_for_each_entry_rcu(delegation, &clp->cl_delegations, super_list) { - inode = igrab(delegation->inode); - if (inode == NULL) - continue; - spin_lock(&clp->cl_lock); - delegation = nfs_detach_delegation_locked(NFS_I(inode), NULL); - spin_unlock(&clp->cl_lock); - rcu_read_unlock(); - if (delegation) - __nfs_inode_return_delegation(inode, delegation); - iput(inode); - goto restart; + set_bit(NFS_DELEGATION_RETURN, &delegation->flags); + set_bit(NFS4CLNT_DELEGRETURN, &clp->cl_state); } rcu_read_unlock(); -out: - nfs_put_client(clp); - module_put_and_exit(0); +} + +static void nfs_delegation_run_state_manager(struct nfs_client *clp) +{ + if (test_bit(NFS4CLNT_DELEGRETURN, &clp->cl_state)) + nfs4_schedule_state_manager(clp); } void nfs_expire_all_delegations(struct nfs_client *clp) { - struct task_struct *task; - - __module_get(THIS_MODULE); - atomic_inc(&clp->cl_count); - task = kthread_run(nfs_do_expire_all_delegations, clp, - "%s-delegreturn", - rpc_peeraddr2str(clp->cl_rpcclient, - RPC_DISPLAY_ADDR)); - if (!IS_ERR(task)) - return; - nfs_put_client(clp); - module_put(THIS_MODULE); + nfs_client_mark_return_all_delegations(clp); + nfs_delegation_run_state_manager(clp); } /* @@ -363,68 +399,29 @@ void nfs_expire_all_delegations(struct nfs_client *clp) */ void nfs_handle_cb_pathdown(struct nfs_client *clp) { - struct nfs_delegation *delegation; - struct inode *inode; - if (clp == NULL) return; -restart: + nfs_client_mark_return_all_delegations(clp); +} + +static void nfs_client_mark_return_unreferenced_delegations(struct nfs_client *clp) +{ + struct nfs_delegation *delegation; + rcu_read_lock(); list_for_each_entry_rcu(delegation, &clp->cl_delegations, super_list) { - inode = igrab(delegation->inode); - if (inode == NULL) + if (test_and_clear_bit(NFS_DELEGATION_REFERENCED, &delegation->flags)) continue; - spin_lock(&clp->cl_lock); - delegation = nfs_detach_delegation_locked(NFS_I(inode), NULL); - spin_unlock(&clp->cl_lock); - rcu_read_unlock(); - if (delegation != NULL) - __nfs_inode_return_delegation(inode, delegation); - iput(inode); - goto restart; + set_bit(NFS_DELEGATION_RETURN, &delegation->flags); + set_bit(NFS4CLNT_DELEGRETURN, &clp->cl_state); } rcu_read_unlock(); } -struct recall_threadargs { - struct inode *inode; - struct nfs_client *clp; - const nfs4_stateid *stateid; - - struct completion started; - int result; -}; - -static int recall_thread(void *data) +void nfs_expire_unreferenced_delegations(struct nfs_client *clp) { - struct recall_threadargs *args = (struct recall_threadargs *)data; - struct inode *inode = igrab(args->inode); - struct nfs_client *clp = NFS_SERVER(inode)->nfs_client; - struct nfs_inode *nfsi = NFS_I(inode); - struct nfs_delegation *delegation; - - daemonize("nfsv4-delegreturn"); - - nfs_msync_inode(inode); - down_read(&clp->cl_sem); - down_write(&nfsi->rwsem); - spin_lock(&clp->cl_lock); - delegation = nfs_detach_delegation_locked(nfsi, args->stateid); - if (delegation != NULL) - args->result = 0; - else - args->result = -ENOENT; - spin_unlock(&clp->cl_lock); - complete(&args->started); - nfs_delegation_claim_opens(inode, args->stateid); - up_write(&nfsi->rwsem); - up_read(&clp->cl_sem); - nfs_msync_inode(inode); - - if (delegation != NULL) - nfs_do_return_delegation(inode, delegation, 1); - iput(inode); - module_put_and_exit(0); + nfs_client_mark_return_unreferenced_delegations(clp); + nfs_delegation_run_state_manager(clp); } /* @@ -432,22 +429,20 @@ static int recall_thread(void *data) */ int nfs_async_inode_return_delegation(struct inode *inode, const nfs4_stateid *stateid) { - struct recall_threadargs data = { - .inode = inode, - .stateid = stateid, - }; - int status; + struct nfs_client *clp = NFS_SERVER(inode)->nfs_client; + struct nfs_delegation *delegation; - init_completion(&data.started); - __module_get(THIS_MODULE); - status = kernel_thread(recall_thread, &data, CLONE_KERNEL); - if (status < 0) - goto out_module_put; - wait_for_completion(&data.started); - return data.result; -out_module_put: - module_put(THIS_MODULE); - return status; + rcu_read_lock(); + delegation = rcu_dereference(NFS_I(inode)->delegation); + if (delegation == NULL || memcmp(delegation->stateid.data, stateid->data, + sizeof(delegation->stateid.data)) != 0) { + rcu_read_unlock(); + return -ENOENT; + } + nfs_mark_return_delegation(clp, delegation); + rcu_read_unlock(); + nfs_delegation_run_state_manager(clp); + return 0; } /* @@ -459,10 +454,14 @@ struct inode *nfs_delegation_find_inode(struct nfs_client *clp, const struct nfs struct inode *res = NULL; rcu_read_lock(); list_for_each_entry_rcu(delegation, &clp->cl_delegations, super_list) { - if (nfs_compare_fh(fhandle, &NFS_I(delegation->inode)->fh) == 0) { + spin_lock(&delegation->lock); + if (delegation->inode != NULL && + nfs_compare_fh(fhandle, &NFS_I(delegation->inode)->fh) == 0) { res = igrab(delegation->inode); - break; } + spin_unlock(&delegation->lock); + if (res != NULL) + break; } rcu_read_unlock(); return res; @@ -476,7 +475,7 @@ void nfs_delegation_mark_reclaim(struct nfs_client *clp) struct nfs_delegation *delegation; rcu_read_lock(); list_for_each_entry_rcu(delegation, &clp->cl_delegations, super_list) - delegation->flags |= NFS_DELEGATION_NEED_RECLAIM; + set_bit(NFS_DELEGATION_NEED_RECLAIM, &delegation->flags); rcu_read_unlock(); } @@ -486,17 +485,22 @@ void nfs_delegation_mark_reclaim(struct nfs_client *clp) void nfs_delegation_reap_unclaimed(struct nfs_client *clp) { struct nfs_delegation *delegation; + struct inode *inode; restart: rcu_read_lock(); list_for_each_entry_rcu(delegation, &clp->cl_delegations, super_list) { - if ((delegation->flags & NFS_DELEGATION_NEED_RECLAIM) == 0) + if (test_bit(NFS_DELEGATION_NEED_RECLAIM, &delegation->flags) == 0) + continue; + inode = nfs_delegation_grab_inode(delegation); + if (inode == NULL) continue; spin_lock(&clp->cl_lock); - delegation = nfs_detach_delegation_locked(NFS_I(delegation->inode), NULL); + delegation = nfs_detach_delegation_locked(NFS_I(inode), NULL); spin_unlock(&clp->cl_lock); rcu_read_unlock(); if (delegation != NULL) nfs_free_delegation(delegation); + iput(inode); goto restart; } rcu_read_unlock(); diff --git a/fs/nfs/delegation.h b/fs/nfs/delegation.h index f1c5e2a5d88..09f38379517 100644 --- a/fs/nfs/delegation.h +++ b/fs/nfs/delegation.h @@ -17,14 +17,20 @@ struct nfs_delegation { struct rpc_cred *cred; struct inode *inode; nfs4_stateid stateid; - int type; -#define NFS_DELEGATION_NEED_RECLAIM 1 - long flags; + fmode_t type; loff_t maxsize; __u64 change_attr; + unsigned long flags; + spinlock_t lock; struct rcu_head rcu; }; +enum { + NFS_DELEGATION_NEED_RECLAIM = 0, + NFS_DELEGATION_RETURN, + NFS_DELEGATION_REFERENCED, +}; + int nfs_inode_set_delegation(struct inode *inode, struct rpc_cred *cred, struct nfs_openres *res); void nfs_inode_reclaim_delegation(struct inode *inode, struct rpc_cred *cred, struct nfs_openres *res); int nfs_inode_return_delegation(struct inode *inode); @@ -32,9 +38,11 @@ int nfs_async_inode_return_delegation(struct inode *inode, const nfs4_stateid *s void nfs_inode_return_delegation_noreclaim(struct inode *inode); struct inode *nfs_delegation_find_inode(struct nfs_client *clp, const struct nfs_fh *fhandle); -void nfs_return_all_delegations(struct super_block *sb); +void nfs_super_return_all_delegations(struct super_block *sb); void nfs_expire_all_delegations(struct nfs_client *clp); +void nfs_expire_unreferenced_delegations(struct nfs_client *clp); void nfs_handle_cb_pathdown(struct nfs_client *clp); +void nfs_client_return_marked_delegations(struct nfs_client *clp); void nfs_delegation_mark_reclaim(struct nfs_client *clp); void nfs_delegation_reap_unclaimed(struct nfs_client *clp); @@ -45,22 +53,11 @@ int nfs4_open_delegation_recall(struct nfs_open_context *ctx, struct nfs4_state int nfs4_lock_delegation_recall(struct nfs4_state *state, struct file_lock *fl); int nfs4_copy_delegation_stateid(nfs4_stateid *dst, struct inode *inode); -static inline int nfs_have_delegation(struct inode *inode, int flags) -{ - struct nfs_delegation *delegation; - int ret = 0; - - flags &= FMODE_READ|FMODE_WRITE; - rcu_read_lock(); - delegation = rcu_dereference(NFS_I(inode)->delegation); - if (delegation != NULL && (delegation->type & flags) == flags) - ret = 1; - rcu_read_unlock(); - return ret; -} +void nfs_mark_delegation_referenced(struct nfs_delegation *delegation); +int nfs_have_delegation(struct inode *inode, fmode_t flags); #else -static inline int nfs_have_delegation(struct inode *inode, int flags) +static inline int nfs_have_delegation(struct inode *inode, fmode_t flags) { return 0; } diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c index 74f92b717f7..e35c8199f82 100644 --- a/fs/nfs/dir.c +++ b/fs/nfs/dir.c @@ -156,6 +156,7 @@ typedef struct { decode_dirent_t decode; int plus; unsigned long timestamp; + unsigned long gencount; int timestamp_valid; } nfs_readdir_descriptor_t; @@ -177,7 +178,7 @@ int nfs_readdir_filler(nfs_readdir_descriptor_t *desc, struct page *page) struct file *file = desc->file; struct inode *inode = file->f_path.dentry->d_inode; struct rpc_cred *cred = nfs_file_cred(file); - unsigned long timestamp; + unsigned long timestamp, gencount; int error; dfprintk(DIRCACHE, "NFS: %s: reading cookie %Lu into page %lu\n", @@ -186,6 +187,7 @@ int nfs_readdir_filler(nfs_readdir_descriptor_t *desc, struct page *page) again: timestamp = jiffies; + gencount = nfs_inc_attr_generation_counter(); error = NFS_PROTO(inode)->readdir(file->f_path.dentry, cred, desc->entry->cookie, page, NFS_SERVER(inode)->dtsize, desc->plus); if (error < 0) { @@ -199,6 +201,7 @@ int nfs_readdir_filler(nfs_readdir_descriptor_t *desc, struct page *page) goto error; } desc->timestamp = timestamp; + desc->gencount = gencount; desc->timestamp_valid = 1; SetPageUptodate(page); /* Ensure consistent page alignment of the data. @@ -224,9 +227,10 @@ int dir_decode(nfs_readdir_descriptor_t *desc) if (IS_ERR(p)) return PTR_ERR(p); desc->ptr = p; - if (desc->timestamp_valid) + if (desc->timestamp_valid) { desc->entry->fattr->time_start = desc->timestamp; - else + desc->entry->fattr->gencount = desc->gencount; + } else desc->entry->fattr->valid &= ~NFS_ATTR_FATTR; return 0; } @@ -471,7 +475,7 @@ int uncached_readdir(nfs_readdir_descriptor_t *desc, void *dirent, struct rpc_cred *cred = nfs_file_cred(file); struct page *page = NULL; int status; - unsigned long timestamp; + unsigned long timestamp, gencount; dfprintk(DIRCACHE, "NFS: uncached_readdir() searching for cookie %Lu\n", (unsigned long long)*desc->dir_cookie); @@ -482,6 +486,7 @@ int uncached_readdir(nfs_readdir_descriptor_t *desc, void *dirent, goto out; } timestamp = jiffies; + gencount = nfs_inc_attr_generation_counter(); status = NFS_PROTO(inode)->readdir(file->f_path.dentry, cred, *desc->dir_cookie, page, NFS_SERVER(inode)->dtsize, @@ -490,6 +495,7 @@ int uncached_readdir(nfs_readdir_descriptor_t *desc, void *dirent, desc->ptr = kmap(page); /* matching kunmap in nfs_do_filldir */ if (status >= 0) { desc->timestamp = timestamp; + desc->gencount = gencount; desc->timestamp_valid = 1; if ((status = dir_decode(desc)) == 0) desc->entry->prev_cookie = *desc->dir_cookie; @@ -655,7 +661,7 @@ static int nfs_fsync_dir(struct file *filp, struct dentry *dentry, int datasync) */ void nfs_force_lookup_revalidate(struct inode *dir) { - NFS_I(dir)->cache_change_attribute = jiffies; + NFS_I(dir)->cache_change_attribute++; } /* @@ -667,6 +673,8 @@ static int nfs_check_verifier(struct inode *dir, struct dentry *dentry) { if (IS_ROOT(dentry)) return 1; + if (NFS_SERVER(dir)->flags & NFS_MOUNT_LOOKUP_CACHE_NONE) + return 0; if (!nfs_verify_change_attribute(dir, dentry->d_time)) return 0; /* Revalidate nfsi->cache_change_attribute before we declare a match */ @@ -699,9 +707,7 @@ static int nfs_is_exclusive_create(struct inode *dir, struct nameidata *nd) { if (NFS_PROTO(dir)->version == 2) return 0; - if (nd == NULL || nfs_lookup_check_intent(nd, LOOKUP_CREATE) == 0) - return 0; - return (nd->intent.open.flags & O_EXCL) != 0; + return nd && nfs_lookup_check_intent(nd, LOOKUP_EXCL); } /* @@ -750,6 +756,8 @@ int nfs_neg_need_reval(struct inode *dir, struct dentry *dentry, /* Don't revalidate a negative dentry if we're creating a new file */ if (nd != NULL && nfs_lookup_check_intent(nd, LOOKUP_CREATE) != 0) return 0; + if (NFS_SERVER(dir)->flags & NFS_MOUNT_LOOKUP_CACHE_NONEG) + return 1; return !nfs_check_verifier(dir, dentry); } @@ -791,6 +799,9 @@ static int nfs_lookup_revalidate(struct dentry * dentry, struct nameidata *nd) goto out_bad; } + if (nfs_have_delegation(inode, FMODE_READ)) + goto out_set_verifier; + /* Force a full look up iff the parent directory has changed */ if (!nfs_is_exclusive_create(dir, nd) && nfs_check_verifier(dir, dentry)) { if (nfs_lookup_verify_inode(inode, nd)) @@ -809,6 +820,7 @@ static int nfs_lookup_revalidate(struct dentry * dentry, struct nameidata *nd) if ((error = nfs_refresh_inode(inode, &fattr)) != 0) goto out_bad; +out_set_verifier: nfs_set_verifier(dentry, nfs_save_change_attribute(dir)); out_valid: dput(parent); @@ -965,7 +977,7 @@ struct dentry_operations nfs4_dentry_operations = { * Use intent information to determine whether we need to substitute * the NFSv4-style stateful OPEN for the LOOKUP call */ -static int is_atomic_open(struct inode *dir, struct nameidata *nd) +static int is_atomic_open(struct nameidata *nd) { if (nd == NULL || nfs_lookup_check_intent(nd, LOOKUP_OPEN) == 0) return 0; @@ -988,7 +1000,7 @@ static struct dentry *nfs_atomic_lookup(struct inode *dir, struct dentry *dentry dir->i_sb->s_id, dir->i_ino, dentry->d_name.name); /* Check that we are indeed trying to open this file */ - if (!is_atomic_open(dir, nd)) + if (!is_atomic_open(nd)) goto no_open; if (dentry->d_name.len > NFS_SERVER(dir)->namelen) { @@ -999,7 +1011,7 @@ static struct dentry *nfs_atomic_lookup(struct inode *dir, struct dentry *dentry /* Let vfs_create() deal with O_EXCL. Instantiate, but don't hash * the dentry. */ - if (nd->intent.open.flags & O_EXCL) { + if (nd->flags & LOOKUP_EXCL) { d_instantiate(dentry, NULL); goto out; } @@ -1039,10 +1051,10 @@ static int nfs_open_revalidate(struct dentry *dentry, struct nameidata *nd) struct inode *dir; int openflags, ret = 0; + if (!is_atomic_open(nd)) + goto no_open; parent = dget_parent(dentry); dir = parent->d_inode; - if (!is_atomic_open(dir, nd)) - goto no_open; /* We can't create new files in nfs_open_revalidate(), so we * optimize away revalidation of negative dentries. */ @@ -1054,11 +1066,11 @@ static int nfs_open_revalidate(struct dentry *dentry, struct nameidata *nd) /* NFS only supports OPEN on regular files */ if (!S_ISREG(inode->i_mode)) - goto no_open; + goto no_open_dput; openflags = nd->intent.open.flags; /* We cannot do exclusive creation on a positive dentry */ if ((openflags & (O_CREAT|O_EXCL)) == (O_CREAT|O_EXCL)) - goto no_open; + goto no_open_dput; /* We can't create new files, or truncate existing ones here */ openflags &= ~(O_CREAT|O_TRUNC); @@ -1073,10 +1085,9 @@ out: if (!ret) d_drop(dentry); return ret; -no_open: +no_open_dput: dput(parent); - if (inode != NULL && nfs_have_delegation(inode, FMODE_READ)) - return 1; +no_open: return nfs_lookup_revalidate(dentry, nd); } #endif /* CONFIG_NFSV4 */ @@ -1507,7 +1518,7 @@ static int nfs_symlink(struct inode *dir, struct dentry *dentry, const char *sym if (!add_to_page_cache(page, dentry->d_inode->i_mapping, 0, GFP_KERNEL)) { pagevec_add(&lru_pvec, page); - pagevec_lru_add(&lru_pvec); + pagevec_lru_add_file(&lru_pvec); SetPageUptodate(page); unlock_page(page); } else @@ -1786,7 +1797,8 @@ static int nfs_access_get_cached(struct inode *inode, struct rpc_cred *cred, str cache = nfs_access_search_rbtree(inode, cred); if (cache == NULL) goto out; - if (!time_in_range(jiffies, cache->jiffies, cache->jiffies + nfsi->attrtimeo)) + if (!nfs_have_delegation(inode, FMODE_READ) && + !time_in_range_open(jiffies, cache->jiffies, cache->jiffies + nfsi->attrtimeo)) goto out_stale; res->jiffies = cache->jiffies; res->cred = cache->cred; @@ -1949,6 +1961,9 @@ force_lookup: } else res = PTR_ERR(cred); out: + if (!res && (mask & MAY_EXEC) && !execute_ok(inode)) + res = -EACCES; + dfprintk(VFS, "NFS: permission(%s/%ld), mask=0x%x, res=%d\n", inode->i_sb->s_id, inode->i_ino, mask, res); return res; diff --git a/fs/nfs/file.c b/fs/nfs/file.c index 78460657f5c..d319b49f8f0 100644 --- a/fs/nfs/file.c +++ b/fs/nfs/file.c @@ -188,13 +188,16 @@ static loff_t nfs_file_llseek(struct file *filp, loff_t offset, int origin) /* origin == SEEK_END => we must revalidate the cached file length */ if (origin == SEEK_END) { struct inode *inode = filp->f_mapping->host; + int retval = nfs_revalidate_file_size(inode, filp); if (retval < 0) return (loff_t)retval; - } - lock_kernel(); /* BKL needed? */ - loff = generic_file_llseek_unlocked(filp, offset, origin); - unlock_kernel(); + + spin_lock(&inode->i_lock); + loff = generic_file_llseek_unlocked(filp, offset, origin); + spin_unlock(&inode->i_lock); + } else + loff = generic_file_llseek_unlocked(filp, offset, origin); return loff; } @@ -699,13 +702,6 @@ static int nfs_flock(struct file *filp, int cmd, struct file_lock *fl) filp->f_path.dentry->d_name.name, fl->fl_type, fl->fl_flags); - /* - * No BSD flocks over NFS allowed. - * Note: we could try to fake a POSIX lock request here by - * using ((u32) filp | 0x80000000) or some such as the pid. - * Not sure whether that would be unique, though, or whether - * that would break in other places. - */ if (!(fl->fl_flags & FL_FLOCK)) return -ENOLCK; diff --git a/fs/nfs/getroot.c b/fs/nfs/getroot.c index fae97196daa..b7c9b2df1f2 100644 --- a/fs/nfs/getroot.c +++ b/fs/nfs/getroot.c @@ -107,11 +107,10 @@ struct dentry *nfs_get_root(struct super_block *sb, struct nfs_fh *mntfh) * if the dentry tree reaches them; however if the dentry already * exists, we'll pick it up at this point and use it as the root */ - mntroot = d_alloc_anon(inode); - if (!mntroot) { - iput(inode); + mntroot = d_obtain_alias(inode); + if (IS_ERR(mntroot)) { dprintk("nfs_get_root: get root dentry failed\n"); - return ERR_PTR(-ENOMEM); + return mntroot; } security_d_instantiate(mntroot, inode); @@ -277,11 +276,10 @@ struct dentry *nfs4_get_root(struct super_block *sb, struct nfs_fh *mntfh) * if the dentry tree reaches them; however if the dentry already * exists, we'll pick it up at this point and use it as the root */ - mntroot = d_alloc_anon(inode); - if (!mntroot) { - iput(inode); + mntroot = d_obtain_alias(inode); + if (IS_ERR(mntroot)) { dprintk("nfs_get_root: get root dentry failed\n"); - return ERR_PTR(-ENOMEM); + return mntroot; } security_d_instantiate(mntroot, inode); diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c index 52daefa2f52..0c381686171 100644 --- a/fs/nfs/inode.c +++ b/fs/nfs/inode.c @@ -5,7 +5,7 @@ * * nfs inode and superblock handling functions * - * Modularised by Alan Cox <Alan.Cox@linux.org>, while hacking some + * Modularised by Alan Cox <alan@lxorguk.ukuu.org.uk>, while hacking some * experimental NFS changes. Modularisation taken straight from SYS5 fs. * * Change to nfs_read_super() to permit NFS mounts to multi-homed hosts. @@ -305,8 +305,7 @@ nfs_fhget(struct super_block *sb, struct nfs_fh *fh, struct nfs_fattr *fattr) init_special_inode(inode, inode->i_mode, fattr->rdev); nfsi->read_cache_jiffies = fattr->time_start; - nfsi->last_updated = now; - nfsi->cache_change_attribute = now; + nfsi->attr_gencount = fattr->gencount; inode->i_atime = fattr->atime; inode->i_mtime = fattr->mtime; inode->i_ctime = fattr->ctime; @@ -453,6 +452,7 @@ out_big: void nfs_setattr_update_inode(struct inode *inode, struct iattr *attr) { if ((attr->ia_valid & (ATTR_MODE|ATTR_UID|ATTR_GID)) != 0) { + spin_lock(&inode->i_lock); if ((attr->ia_valid & ATTR_MODE) != 0) { int mode = attr->ia_mode & S_IALLUGO; mode |= inode->i_mode & ~S_IALLUGO; @@ -462,7 +462,6 @@ void nfs_setattr_update_inode(struct inode *inode, struct iattr *attr) inode->i_uid = attr->ia_uid; if ((attr->ia_valid & ATTR_GID) != 0) inode->i_gid = attr->ia_gid; - spin_lock(&inode->i_lock); NFS_I(inode)->cache_validity |= NFS_INO_INVALID_ACCESS|NFS_INO_INVALID_ACL; spin_unlock(&inode->i_lock); } @@ -472,37 +471,6 @@ void nfs_setattr_update_inode(struct inode *inode, struct iattr *attr) } } -static int nfs_wait_schedule(void *word) -{ - if (signal_pending(current)) - return -ERESTARTSYS; - schedule(); - return 0; -} - -/* - * Wait for the inode to get unlocked. - */ -static int nfs_wait_on_inode(struct inode *inode) -{ - struct nfs_inode *nfsi = NFS_I(inode); - int error; - - error = wait_on_bit_lock(&nfsi->flags, NFS_INO_REVALIDATING, - nfs_wait_schedule, TASK_KILLABLE); - - return error; -} - -static void nfs_wake_up_inode(struct inode *inode) -{ - struct nfs_inode *nfsi = NFS_I(inode); - - clear_bit(NFS_INO_REVALIDATING, &nfsi->flags); - smp_mb__after_clear_bit(); - wake_up_bit(&nfsi->flags, NFS_INO_REVALIDATING); -} - int nfs_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat) { struct inode *inode = dentry->d_inode; @@ -624,7 +592,7 @@ static void nfs_file_set_open_context(struct file *filp, struct nfs_open_context /* * Given an inode, search for an open context with the desired characteristics */ -struct nfs_open_context *nfs_find_open_context(struct inode *inode, struct rpc_cred *cred, int mode) +struct nfs_open_context *nfs_find_open_context(struct inode *inode, struct rpc_cred *cred, fmode_t mode) { struct nfs_inode *nfsi = NFS_I(inode); struct nfs_open_context *pos, *ctx = NULL; @@ -697,20 +665,15 @@ __nfs_revalidate_inode(struct nfs_server *server, struct inode *inode) dfprintk(PAGECACHE, "NFS: revalidating (%s/%Ld)\n", inode->i_sb->s_id, (long long)NFS_FILEID(inode)); - nfs_inc_stats(inode, NFSIOS_INODEREVALIDATE); if (is_bad_inode(inode)) - goto out_nowait; + goto out; if (NFS_STALE(inode)) - goto out_nowait; - - status = nfs_wait_on_inode(inode); - if (status < 0) goto out; - status = -ESTALE; if (NFS_STALE(inode)) goto out; + nfs_inc_stats(inode, NFSIOS_INODEREVALIDATE); status = NFS_PROTO(inode)->getattr(server, NFS_FH(inode), &fattr); if (status != 0) { dfprintk(PAGECACHE, "nfs_revalidate_inode: (%s/%Ld) getattr failed, error=%d\n", @@ -724,16 +687,13 @@ __nfs_revalidate_inode(struct nfs_server *server, struct inode *inode) goto out; } - spin_lock(&inode->i_lock); - status = nfs_update_inode(inode, &fattr); + status = nfs_refresh_inode(inode, &fattr); if (status) { - spin_unlock(&inode->i_lock); dfprintk(PAGECACHE, "nfs_revalidate_inode: (%s/%Ld) refresh failed, error=%d\n", inode->i_sb->s_id, (long long)NFS_FILEID(inode), status); goto out; } - spin_unlock(&inode->i_lock); if (nfsi->cache_validity & NFS_INO_INVALID_ACL) nfs_zap_acl_cache(inode); @@ -743,9 +703,6 @@ __nfs_revalidate_inode(struct nfs_server *server, struct inode *inode) (long long)NFS_FILEID(inode)); out: - nfs_wake_up_inode(inode); - - out_nowait: return status; } @@ -755,14 +712,7 @@ int nfs_attribute_timeout(struct inode *inode) if (nfs_have_delegation(inode, FMODE_READ)) return 0; - /* - * Special case: if the attribute timeout is set to 0, then always - * treat the cache as having expired (unless holding - * a delegation). - */ - if (nfsi->attrtimeo == 0) - return 1; - return !time_in_range(jiffies, nfsi->read_cache_jiffies, nfsi->read_cache_jiffies + nfsi->attrtimeo); + return !time_in_range_open(jiffies, nfsi->read_cache_jiffies, nfsi->read_cache_jiffies + nfsi->attrtimeo); } /** @@ -908,9 +858,6 @@ static int nfs_check_inode_attributes(struct inode *inode, struct nfs_fattr *fat return -EIO; } - /* Do atomic weak cache consistency updates */ - nfs_wcc_update_inode(inode, fattr); - if ((fattr->valid & NFS_ATTR_FATTR_V4) != 0 && nfsi->change_attr != fattr->change_attr) invalid |= NFS_INO_INVALID_ATTR|NFS_INO_REVAL_PAGECACHE; @@ -939,15 +886,76 @@ static int nfs_check_inode_attributes(struct inode *inode, struct nfs_fattr *fat if (invalid != 0) nfsi->cache_validity |= invalid; - else - nfsi->cache_validity &= ~(NFS_INO_INVALID_ATTR - | NFS_INO_INVALID_ATIME - | NFS_INO_REVAL_PAGECACHE); nfsi->read_cache_jiffies = fattr->time_start; return 0; } +static int nfs_ctime_need_update(const struct inode *inode, const struct nfs_fattr *fattr) +{ + return timespec_compare(&fattr->ctime, &inode->i_ctime) > 0; +} + +static int nfs_size_need_update(const struct inode *inode, const struct nfs_fattr *fattr) +{ + return nfs_size_to_loff_t(fattr->size) > i_size_read(inode); +} + +static atomic_long_t nfs_attr_generation_counter; + +static unsigned long nfs_read_attr_generation_counter(void) +{ + return atomic_long_read(&nfs_attr_generation_counter); +} + +unsigned long nfs_inc_attr_generation_counter(void) +{ + return atomic_long_inc_return(&nfs_attr_generation_counter); +} + +void nfs_fattr_init(struct nfs_fattr *fattr) +{ + fattr->valid = 0; + fattr->time_start = jiffies; + fattr->gencount = nfs_inc_attr_generation_counter(); +} + +/** + * nfs_inode_attrs_need_update - check if the inode attributes need updating + * @inode - pointer to inode + * @fattr - attributes + * + * Attempt to divine whether or not an RPC call reply carrying stale + * attributes got scheduled after another call carrying updated ones. + * + * To do so, the function first assumes that a more recent ctime means + * that the attributes in fattr are newer, however it also attempt to + * catch the case where ctime either didn't change, or went backwards + * (if someone reset the clock on the server) by looking at whether + * or not this RPC call was started after the inode was last updated. + * Note also the check for wraparound of 'attr_gencount' + * + * The function returns 'true' if it thinks the attributes in 'fattr' are + * more recent than the ones cached in the inode. + * + */ +static int nfs_inode_attrs_need_update(const struct inode *inode, const struct nfs_fattr *fattr) +{ + const struct nfs_inode *nfsi = NFS_I(inode); + + return ((long)fattr->gencount - (long)nfsi->attr_gencount) > 0 || + nfs_ctime_need_update(inode, fattr) || + nfs_size_need_update(inode, fattr) || + ((long)nfsi->attr_gencount - (long)nfs_read_attr_generation_counter() > 0); +} + +static int nfs_refresh_inode_locked(struct inode *inode, struct nfs_fattr *fattr) +{ + if (nfs_inode_attrs_need_update(inode, fattr)) + return nfs_update_inode(inode, fattr); + return nfs_check_inode_attributes(inode, fattr); +} + /** * nfs_refresh_inode - try to update the inode attribute cache * @inode - pointer to inode @@ -960,21 +968,28 @@ static int nfs_check_inode_attributes(struct inode *inode, struct nfs_fattr *fat */ int nfs_refresh_inode(struct inode *inode, struct nfs_fattr *fattr) { - struct nfs_inode *nfsi = NFS_I(inode); int status; if ((fattr->valid & NFS_ATTR_FATTR) == 0) return 0; spin_lock(&inode->i_lock); - if (time_after(fattr->time_start, nfsi->last_updated)) - status = nfs_update_inode(inode, fattr); - else - status = nfs_check_inode_attributes(inode, fattr); - + status = nfs_refresh_inode_locked(inode, fattr); spin_unlock(&inode->i_lock); return status; } +static int nfs_post_op_update_inode_locked(struct inode *inode, struct nfs_fattr *fattr) +{ + struct nfs_inode *nfsi = NFS_I(inode); + + nfsi->cache_validity |= NFS_INO_INVALID_ATTR|NFS_INO_REVAL_PAGECACHE; + if (S_ISDIR(inode->i_mode)) + nfsi->cache_validity |= NFS_INO_INVALID_DATA; + if ((fattr->valid & NFS_ATTR_FATTR) == 0) + return 0; + return nfs_refresh_inode_locked(inode, fattr); +} + /** * nfs_post_op_update_inode - try to update the inode attribute cache * @inode - pointer to inode @@ -991,14 +1006,12 @@ int nfs_refresh_inode(struct inode *inode, struct nfs_fattr *fattr) */ int nfs_post_op_update_inode(struct inode *inode, struct nfs_fattr *fattr) { - struct nfs_inode *nfsi = NFS_I(inode); + int status; spin_lock(&inode->i_lock); - nfsi->cache_validity |= NFS_INO_INVALID_ATTR|NFS_INO_REVAL_PAGECACHE; - if (S_ISDIR(inode->i_mode)) - nfsi->cache_validity |= NFS_INO_INVALID_DATA; + status = nfs_post_op_update_inode_locked(inode, fattr); spin_unlock(&inode->i_lock); - return nfs_refresh_inode(inode, fattr); + return status; } /** @@ -1014,6 +1027,15 @@ int nfs_post_op_update_inode(struct inode *inode, struct nfs_fattr *fattr) */ int nfs_post_op_update_inode_force_wcc(struct inode *inode, struct nfs_fattr *fattr) { + int status; + + spin_lock(&inode->i_lock); + /* Don't do a WCC update if these attributes are already stale */ + if ((fattr->valid & NFS_ATTR_FATTR) == 0 || + !nfs_inode_attrs_need_update(inode, fattr)) { + fattr->valid &= ~(NFS_ATTR_WCC_V4|NFS_ATTR_WCC); + goto out_noforce; + } if ((fattr->valid & NFS_ATTR_FATTR_V4) != 0 && (fattr->valid & NFS_ATTR_WCC_V4) == 0) { fattr->pre_change_attr = NFS_I(inode)->change_attr; @@ -1026,7 +1048,10 @@ int nfs_post_op_update_inode_force_wcc(struct inode *inode, struct nfs_fattr *fa fattr->pre_size = i_size_read(inode); fattr->valid |= NFS_ATTR_WCC; } - return nfs_post_op_update_inode(inode, fattr); +out_noforce: + status = nfs_post_op_update_inode_locked(inode, fattr); + spin_unlock(&inode->i_lock); + return status; } /* @@ -1092,7 +1117,7 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr) } /* If ctime has changed we should definitely clear access+acl caches */ if (!timespec_equal(&inode->i_ctime, &fattr->ctime)) - invalid |= NFS_INO_INVALID_ACCESS|NFS_INO_INVALID_ACL; + invalid |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_ACCESS|NFS_INO_INVALID_ACL; } else if (nfsi->change_attr != fattr->change_attr) { dprintk("NFS: change_attr change on server for file %s/%ld\n", inode->i_sb->s_id, inode->i_ino); @@ -1126,6 +1151,9 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr) inode->i_gid != fattr->gid) invalid |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_ACCESS|NFS_INO_INVALID_ACL; + if (inode->i_nlink != fattr->nlink) + invalid |= NFS_INO_INVALID_ATTR; + inode->i_mode = fattr->mode; inode->i_nlink = fattr->nlink; inode->i_uid = fattr->uid; @@ -1145,18 +1173,13 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr) nfs_inc_stats(inode, NFSIOS_ATTRINVALIDATE); nfsi->attrtimeo = NFS_MINATTRTIMEO(inode); nfsi->attrtimeo_timestamp = now; - nfsi->last_updated = now; + nfsi->attr_gencount = nfs_inc_attr_generation_counter(); } else { - if (!time_in_range(now, nfsi->attrtimeo_timestamp, nfsi->attrtimeo_timestamp + nfsi->attrtimeo)) { + if (!time_in_range_open(now, nfsi->attrtimeo_timestamp, nfsi->attrtimeo_timestamp + nfsi->attrtimeo)) { if ((nfsi->attrtimeo <<= 1) > NFS_MAXATTRTIMEO(inode)) nfsi->attrtimeo = NFS_MAXATTRTIMEO(inode); nfsi->attrtimeo_timestamp = now; } - /* - * Avoid jiffy wraparound issues with nfsi->last_updated - */ - if (!time_in_range(nfsi->last_updated, nfsi->read_cache_jiffies, now)) - nfsi->last_updated = nfsi->read_cache_jiffies; } invalid &= ~NFS_INO_INVALID_ATTR; /* Don't invalidate the data if we were to blame */ diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h index 24241fcbb98..340ede8f608 100644 --- a/fs/nfs/internal.h +++ b/fs/nfs/internal.h @@ -63,6 +63,20 @@ struct nfs_parsed_mount_data { struct security_mnt_opts lsm_opts; }; +/* mount_clnt.c */ +struct nfs_mount_request { + struct sockaddr *sap; + size_t salen; + char *hostname; + char *dirpath; + u32 version; + unsigned short protocol; + struct nfs_fh *fh; + int noresvport; +}; + +extern int nfs_mount(struct nfs_mount_request *info); + /* client.c */ extern struct rpc_program nfs_program; @@ -153,6 +167,7 @@ extern void nfs4_clear_inode(struct inode *); void nfs_zap_acl_cache(struct inode *inode); /* super.c */ +void nfs_parse_ip_address(char *, size_t, struct sockaddr *, size_t *); extern struct file_system_type nfs_xdev_fs_type; #ifdef CONFIG_NFS_V4 extern struct file_system_type nfs4_xdev_fs_type; @@ -163,8 +178,8 @@ extern struct rpc_stat nfs_rpcstat; extern int __init register_nfs_fs(void); extern void __exit unregister_nfs_fs(void); -extern void nfs_sb_active(struct nfs_server *server); -extern void nfs_sb_deactive(struct nfs_server *server); +extern void nfs_sb_active(struct super_block *sb); +extern void nfs_sb_deactive(struct super_block *sb); /* namespace.c */ extern char *nfs_path(const char *base, @@ -276,3 +291,23 @@ unsigned int nfs_page_array_len(unsigned int base, size_t len) PAGE_SIZE - 1) >> PAGE_SHIFT; } +#define IPV6_SCOPE_DELIMITER '%' + +/* + * Set the port number in an address. Be agnostic about the address + * family. + */ +static inline void nfs_set_port(struct sockaddr *sap, unsigned short port) +{ + struct sockaddr_in *ap = (struct sockaddr_in *)sap; + struct sockaddr_in6 *ap6 = (struct sockaddr_in6 *)sap; + + switch (sap->sa_family) { + case AF_INET: + ap->sin_port = htons(port); + break; + case AF_INET6: + ap6->sin6_port = htons(port); + break; + } +} diff --git a/fs/nfs/mount_clnt.c b/fs/nfs/mount_clnt.c index 779d2eb649c..ca905a5bb1b 100644 --- a/fs/nfs/mount_clnt.c +++ b/fs/nfs/mount_clnt.c @@ -14,6 +14,7 @@ #include <linux/sunrpc/clnt.h> #include <linux/sunrpc/sched.h> #include <linux/nfs_fs.h> +#include "internal.h" #ifdef RPC_DEBUG # define NFSDBG_FACILITY NFSDBG_MOUNT @@ -28,47 +29,43 @@ struct mnt_fhstatus { /** * nfs_mount - Obtain an NFS file handle for the given host and path - * @addr: pointer to server's address - * @len: size of server's address - * @hostname: name of server host, or NULL - * @path: pointer to string containing export path to mount - * @version: mount version to use for this request - * @protocol: transport protocol to use for thie request - * @fh: pointer to location to place returned file handle + * @info: pointer to mount request arguments * * Uses default timeout parameters specified by underlying transport. */ -int nfs_mount(struct sockaddr *addr, size_t len, char *hostname, char *path, - int version, int protocol, struct nfs_fh *fh) +int nfs_mount(struct nfs_mount_request *info) { struct mnt_fhstatus result = { - .fh = fh + .fh = info->fh }; struct rpc_message msg = { - .rpc_argp = path, + .rpc_argp = info->dirpath, .rpc_resp = &result, }; struct rpc_create_args args = { - .protocol = protocol, - .address = addr, - .addrsize = len, - .servername = hostname, + .protocol = info->protocol, + .address = info->sap, + .addrsize = info->salen, + .servername = info->hostname, .program = &mnt_program, - .version = version, + .version = info->version, .authflavor = RPC_AUTH_UNIX, - .flags = 0, }; struct rpc_clnt *mnt_clnt; int status; dprintk("NFS: sending MNT request for %s:%s\n", - (hostname ? hostname : "server"), path); + (info->hostname ? info->hostname : "server"), + info->dirpath); + + if (info->noresvport) + args.flags |= RPC_CLNT_CREATE_NONPRIVPORT; mnt_clnt = rpc_create(&args); if (IS_ERR(mnt_clnt)) goto out_clnt_err; - if (version == NFS_MNT3_VERSION) + if (info->version == NFS_MNT3_VERSION) msg.rpc_proc = &mnt_clnt->cl_procinfo[MOUNTPROC3_MNT]; else msg.rpc_proc = &mnt_clnt->cl_procinfo[MNTPROC_MNT]; @@ -98,7 +95,7 @@ out_call_err: out_mnt_err: dprintk("NFS: MNT server returned result %d\n", result.status); - status = -EACCES; + status = nfs_stat_to_errno(result.status); goto out; } diff --git a/fs/nfs/namespace.c b/fs/nfs/namespace.c index 66df08dd1ca..64a288ee046 100644 --- a/fs/nfs/namespace.c +++ b/fs/nfs/namespace.c @@ -105,7 +105,10 @@ static void * nfs_follow_mountpoint(struct dentry *dentry, struct nameidata *nd) dprintk("--> nfs_follow_mountpoint()\n"); - BUG_ON(IS_ROOT(dentry)); + err = -ESTALE; + if (IS_ROOT(dentry)) + goto out_err; + dprintk("%s: enter\n", __func__); dput(nd->path.dentry); nd->path.dentry = dget(dentry); @@ -189,7 +192,7 @@ static struct vfsmount *nfs_do_clone_mount(struct nfs_server *server, struct nfs_clone_mount *mountdata) { #ifdef CONFIG_NFS_V4 - struct vfsmount *mnt = NULL; + struct vfsmount *mnt = ERR_PTR(-EINVAL); switch (server->nfs_client->rpc_ops->version) { case 2: case 3: diff --git a/fs/nfs/nfs3acl.c b/fs/nfs/nfs3acl.c index 423842f51ac..cef62557c87 100644 --- a/fs/nfs/nfs3acl.c +++ b/fs/nfs/nfs3acl.c @@ -229,6 +229,7 @@ struct posix_acl *nfs3_proc_getacl(struct inode *inode, int type) dprintk("NFS call getacl\n"); msg.rpc_proc = &server->client_acl->cl_procinfo[ACLPROC3_GETACL]; + nfs_fattr_init(&fattr); status = rpc_call_sync(server->client_acl, &msg, 0); dprintk("NFS reply getacl: %d\n", status); @@ -322,6 +323,7 @@ static int nfs3_proc_setacls(struct inode *inode, struct posix_acl *acl, dprintk("NFS call setacl\n"); msg.rpc_proc = &server->client_acl->cl_procinfo[ACLPROC3_SETACL]; + nfs_fattr_init(&fattr); status = rpc_call_sync(server->client_acl, &msg, 0); nfs_access_zap_cache(inode); nfs_zap_acl_cache(inode); diff --git a/fs/nfs/nfs3proc.c b/fs/nfs/nfs3proc.c index 1e750e4574a..c55be7a7679 100644 --- a/fs/nfs/nfs3proc.c +++ b/fs/nfs/nfs3proc.c @@ -699,7 +699,7 @@ nfs3_proc_statfs(struct nfs_server *server, struct nfs_fh *fhandle, } static int -nfs3_proc_fsinfo(struct nfs_server *server, struct nfs_fh *fhandle, +do_proc_fsinfo(struct rpc_clnt *client, struct nfs_fh *fhandle, struct nfs_fsinfo *info) { struct rpc_message msg = { @@ -711,11 +711,27 @@ nfs3_proc_fsinfo(struct nfs_server *server, struct nfs_fh *fhandle, dprintk("NFS call fsinfo\n"); nfs_fattr_init(info->fattr); - status = rpc_call_sync(server->nfs_client->cl_rpcclient, &msg, 0); + status = rpc_call_sync(client, &msg, 0); dprintk("NFS reply fsinfo: %d\n", status); return status; } +/* + * Bare-bones access to fsinfo: this is for nfs_get_root/nfs_get_sb via + * nfs_create_server + */ +static int +nfs3_proc_fsinfo(struct nfs_server *server, struct nfs_fh *fhandle, + struct nfs_fsinfo *info) +{ + int status; + + status = do_proc_fsinfo(server->client, fhandle, info); + if (status && server->nfs_client->cl_rpcclient != server->client) + status = do_proc_fsinfo(server->nfs_client->cl_rpcclient, fhandle, info); + return status; +} + static int nfs3_proc_pathconf(struct nfs_server *server, struct nfs_fh *fhandle, struct nfs_pathconf *info) diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h index ea790645fda..4e4d3320437 100644 --- a/fs/nfs/nfs4_fs.h +++ b/fs/nfs/nfs4_fs.h @@ -38,8 +38,12 @@ struct idmap; ((err) != NFSERR_NOFILEHANDLE)) enum nfs4_client_state { - NFS4CLNT_STATE_RECOVER = 0, + NFS4CLNT_MANAGER_RUNNING = 0, + NFS4CLNT_CHECK_LEASE, NFS4CLNT_LEASE_EXPIRED, + NFS4CLNT_RECLAIM_REBOOT, + NFS4CLNT_RECLAIM_NOGRACE, + NFS4CLNT_DELEGRETURN, }; /* @@ -90,12 +94,18 @@ struct nfs4_state_owner { spinlock_t so_lock; atomic_t so_count; + unsigned long so_flags; struct list_head so_states; struct list_head so_delegations; struct nfs_seqid_counter so_seqid; struct rpc_sequence so_sequence; }; +enum { + NFS_OWNER_RECLAIM_REBOOT, + NFS_OWNER_RECLAIM_NOGRACE +}; + /* * struct nfs4_state maintains the client-side state for a given * (state_owner,inode) tuple (OPEN) or state_owner (LOCK). @@ -128,6 +138,8 @@ enum { NFS_O_RDONLY_STATE, /* OPEN stateid has read-only state */ NFS_O_WRONLY_STATE, /* OPEN stateid has write-only state */ NFS_O_RDWR_STATE, /* OPEN stateid has read/write state */ + NFS_STATE_RECLAIM_REBOOT, /* OPEN stateid server rebooted */ + NFS_STATE_RECLAIM_NOGRACE, /* OPEN stateid needs to recover state */ }; struct nfs4_state { @@ -149,7 +161,7 @@ struct nfs4_state { unsigned int n_rdonly; /* Number of read-only references */ unsigned int n_wronly; /* Number of write-only references */ unsigned int n_rdwr; /* Number of read/write references */ - int state; /* State on the server (R,W, or RW) */ + fmode_t state; /* State on the server (R,W, or RW) */ atomic_t count; }; @@ -157,9 +169,12 @@ struct nfs4_state { struct nfs4_exception { long timeout; int retry; + struct nfs4_state *state; }; struct nfs4_state_recovery_ops { + int owner_flag_bit; + int state_flag_bit; int (*recover_open)(struct nfs4_state_owner *, struct nfs4_state *); int (*recover_lock)(struct nfs4_state *, struct file_lock *); }; @@ -174,7 +189,6 @@ extern ssize_t nfs4_listxattr(struct dentry *, char *, size_t); /* nfs4proc.c */ -extern int nfs4_map_errors(int err); extern int nfs4_proc_setclientid(struct nfs_client *, u32, unsigned short, struct rpc_cred *); extern int nfs4_proc_setclientid_confirm(struct nfs_client *, struct rpc_cred *); extern int nfs4_proc_async_renew(struct nfs_client *, struct rpc_cred *); @@ -187,7 +201,7 @@ extern int nfs4_proc_fs_locations(struct inode *dir, const struct qstr *name, struct nfs4_fs_locations *fs_locations, struct page *page); extern struct nfs4_state_recovery_ops nfs4_reboot_recovery_ops; -extern struct nfs4_state_recovery_ops nfs4_network_partition_recovery_ops; +extern struct nfs4_state_recovery_ops nfs4_nograce_recovery_ops; extern const u32 nfs4_fattr_bitmap[2]; extern const u32 nfs4_statfs_bitmap[2]; @@ -202,16 +216,18 @@ extern void nfs4_kill_renewd(struct nfs_client *); extern void nfs4_renew_state(struct work_struct *); /* nfs4state.c */ -struct rpc_cred *nfs4_get_renew_cred(struct nfs_client *clp); +struct rpc_cred *nfs4_get_renew_cred_locked(struct nfs_client *clp); extern struct nfs4_state_owner * nfs4_get_state_owner(struct nfs_server *, struct rpc_cred *); extern void nfs4_put_state_owner(struct nfs4_state_owner *); extern struct nfs4_state * nfs4_get_open_state(struct inode *, struct nfs4_state_owner *); extern void nfs4_put_open_state(struct nfs4_state *); -extern void nfs4_close_state(struct path *, struct nfs4_state *, mode_t); -extern void nfs4_close_sync(struct path *, struct nfs4_state *, mode_t); -extern void nfs4_state_set_mode_locked(struct nfs4_state *, mode_t); +extern void nfs4_close_state(struct path *, struct nfs4_state *, fmode_t); +extern void nfs4_close_sync(struct path *, struct nfs4_state *, fmode_t); +extern void nfs4_state_set_mode_locked(struct nfs4_state *, fmode_t); extern void nfs4_schedule_state_recovery(struct nfs_client *); +extern void nfs4_schedule_state_manager(struct nfs_client *); +extern int nfs4_state_mark_reclaim_nograce(struct nfs_client *clp, struct nfs4_state *state); extern void nfs4_put_lock_state(struct nfs4_lock_state *lsp); extern int nfs4_set_lock_state(struct nfs4_state *state, struct file_lock *fl); extern void nfs4_copy_stateid(nfs4_stateid *, struct nfs4_state *, fl_owner_t); diff --git a/fs/nfs/nfs4namespace.c b/fs/nfs/nfs4namespace.c index b112857301f..30befc39b3c 100644 --- a/fs/nfs/nfs4namespace.c +++ b/fs/nfs/nfs4namespace.c @@ -93,21 +93,52 @@ static int nfs4_validate_fspath(const struct vfsmount *mnt_parent, return 0; } -/* - * Check if the string represents a "valid" IPv4 address - */ -static inline int valid_ipaddr4(const char *buf) +static struct vfsmount *try_location(struct nfs_clone_mount *mountdata, + char *page, char *page2, + const struct nfs4_fs_location *location) { - int rc, count, in[4]; - - rc = sscanf(buf, "%d.%d.%d.%d", &in[0], &in[1], &in[2], &in[3]); - if (rc != 4) - return -EINVAL; - for (count = 0; count < 4; count++) { - if (in[count] > 255) - return -EINVAL; + struct vfsmount *mnt = ERR_PTR(-ENOENT); + char *mnt_path; + int page2len; + unsigned int s; + + mnt_path = nfs4_pathname_string(&location->rootpath, page2, PAGE_SIZE); + if (IS_ERR(mnt_path)) + return mnt; + mountdata->mnt_path = mnt_path; + page2 += strlen(mnt_path) + 1; + page2len = PAGE_SIZE - strlen(mnt_path) - 1; + + for (s = 0; s < location->nservers; s++) { + const struct nfs4_string *buf = &location->servers[s]; + struct sockaddr_storage addr; + + if (buf->len <= 0 || buf->len >= PAGE_SIZE) + continue; + + mountdata->addr = (struct sockaddr *)&addr; + + if (memchr(buf->data, IPV6_SCOPE_DELIMITER, buf->len)) + continue; + nfs_parse_ip_address(buf->data, buf->len, + mountdata->addr, &mountdata->addrlen); + if (mountdata->addr->sa_family == AF_UNSPEC) + continue; + nfs_set_port(mountdata->addr, NFS_PORT); + + strncpy(page2, buf->data, page2len); + page2[page2len] = '\0'; + mountdata->hostname = page2; + + snprintf(page, PAGE_SIZE, "%s:%s", + mountdata->hostname, + mountdata->mnt_path); + + mnt = vfs_kern_mount(&nfs4_referral_fs_type, 0, page, mountdata); + if (!IS_ERR(mnt)) + break; } - return 0; + return mnt; } /** @@ -128,7 +159,6 @@ static struct vfsmount *nfs_follow_referral(const struct vfsmount *mnt_parent, .authflavor = NFS_SB(mnt_parent->mnt_sb)->client->cl_auth->au_flavor, }; char *page = NULL, *page2 = NULL; - unsigned int s; int loc, error; if (locations == NULL || locations->nlocations <= 0) @@ -152,53 +182,16 @@ static struct vfsmount *nfs_follow_referral(const struct vfsmount *mnt_parent, goto out; } - loc = 0; - while (loc < locations->nlocations && IS_ERR(mnt)) { + for (loc = 0; loc < locations->nlocations; loc++) { const struct nfs4_fs_location *location = &locations->locations[loc]; - char *mnt_path; if (location == NULL || location->nservers <= 0 || - location->rootpath.ncomponents == 0) { - loc++; + location->rootpath.ncomponents == 0) continue; - } - mnt_path = nfs4_pathname_string(&location->rootpath, page2, PAGE_SIZE); - if (IS_ERR(mnt_path)) { - loc++; - continue; - } - mountdata.mnt_path = mnt_path; - - s = 0; - while (s < location->nservers) { - struct sockaddr_in addr = { - .sin_family = AF_INET, - .sin_port = htons(NFS_PORT), - }; - - if (location->servers[s].len <= 0 || - valid_ipaddr4(location->servers[s].data) < 0) { - s++; - continue; - } - - mountdata.hostname = location->servers[s].data; - addr.sin_addr.s_addr = in_aton(mountdata.hostname), - mountdata.addr = (struct sockaddr *)&addr; - mountdata.addrlen = sizeof(addr); - - snprintf(page, PAGE_SIZE, "%s:%s", - mountdata.hostname, - mountdata.mnt_path); - - mnt = vfs_kern_mount(&nfs4_referral_fs_type, 0, page, &mountdata); - if (!IS_ERR(mnt)) { - break; - } - s++; - } - loc++; + mnt = try_location(&mountdata, page, page2, location); + if (!IS_ERR(mnt)) + break; } out: diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index c910413eaec..8dde84b988d 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -62,14 +62,12 @@ struct nfs4_opendata; static int _nfs4_proc_open(struct nfs4_opendata *data); static int nfs4_do_fsinfo(struct nfs_server *, struct nfs_fh *, struct nfs_fsinfo *); -static int nfs4_async_handle_error(struct rpc_task *, const struct nfs_server *); -static int nfs4_handle_exception(const struct nfs_server *server, int errorcode, struct nfs4_exception *exception); -static int nfs4_wait_clnt_recover(struct rpc_clnt *clnt, struct nfs_client *clp); +static int nfs4_async_handle_error(struct rpc_task *, const struct nfs_server *, struct nfs4_state *); static int _nfs4_proc_lookup(struct inode *dir, const struct qstr *name, struct nfs_fh *fhandle, struct nfs_fattr *fattr); static int _nfs4_proc_getattr(struct nfs_server *server, struct nfs_fh *fhandle, struct nfs_fattr *fattr); /* Prevent leaks of NFSv4 errors into userland */ -int nfs4_map_errors(int err) +static int nfs4_map_errors(int err) { if (err < -1000) { dprintk("%s could not handle NFSv4 error %d\n", @@ -195,6 +193,83 @@ static void nfs4_setup_readdir(u64 cookie, __be32 *verifier, struct dentry *dent kunmap_atomic(start, KM_USER0); } +static int nfs4_wait_bit_killable(void *word) +{ + if (fatal_signal_pending(current)) + return -ERESTARTSYS; + schedule(); + return 0; +} + +static int nfs4_wait_clnt_recover(struct nfs_client *clp) +{ + int res; + + might_sleep(); + + res = wait_on_bit(&clp->cl_state, NFS4CLNT_MANAGER_RUNNING, + nfs4_wait_bit_killable, TASK_KILLABLE); + return res; +} + +static int nfs4_delay(struct rpc_clnt *clnt, long *timeout) +{ + int res = 0; + + might_sleep(); + + if (*timeout <= 0) + *timeout = NFS4_POLL_RETRY_MIN; + if (*timeout > NFS4_POLL_RETRY_MAX) + *timeout = NFS4_POLL_RETRY_MAX; + schedule_timeout_killable(*timeout); + if (fatal_signal_pending(current)) + res = -ERESTARTSYS; + *timeout <<= 1; + return res; +} + +/* This is the error handling routine for processes that are allowed + * to sleep. + */ +static int nfs4_handle_exception(const struct nfs_server *server, int errorcode, struct nfs4_exception *exception) +{ + struct nfs_client *clp = server->nfs_client; + struct nfs4_state *state = exception->state; + int ret = errorcode; + + exception->retry = 0; + switch(errorcode) { + case 0: + return 0; + case -NFS4ERR_ADMIN_REVOKED: + case -NFS4ERR_BAD_STATEID: + case -NFS4ERR_OPENMODE: + if (state == NULL) + break; + nfs4_state_mark_reclaim_nograce(clp, state); + case -NFS4ERR_STALE_CLIENTID: + case -NFS4ERR_STALE_STATEID: + case -NFS4ERR_EXPIRED: + nfs4_schedule_state_recovery(clp); + ret = nfs4_wait_clnt_recover(clp); + if (ret == 0) + exception->retry = 1; + break; + case -NFS4ERR_FILE_OPEN: + case -NFS4ERR_GRACE: + case -NFS4ERR_DELAY: + ret = nfs4_delay(server->client, &exception->timeout); + if (ret != 0) + break; + case -NFS4ERR_OLD_STATEID: + exception->retry = 1; + } + /* We failed to handle the error */ + return nfs4_map_errors(ret); +} + + static void renew_lease(const struct nfs_server *server, unsigned long timestamp) { struct nfs_client *clp = server->nfs_client; @@ -248,7 +323,7 @@ static void nfs4_init_opendata_res(struct nfs4_opendata *p) } static struct nfs4_opendata *nfs4_opendata_alloc(struct path *path, - struct nfs4_state_owner *sp, int flags, + struct nfs4_state_owner *sp, fmode_t fmode, int flags, const struct iattr *attrs) { struct dentry *parent = dget_parent(path->dentry); @@ -268,7 +343,8 @@ static struct nfs4_opendata *nfs4_opendata_alloc(struct path *path, p->owner = sp; atomic_inc(&sp->so_count); p->o_arg.fh = NFS_FH(dir); - p->o_arg.open_flags = flags, + p->o_arg.open_flags = flags; + p->o_arg.fmode = fmode & (FMODE_READ|FMODE_WRITE); p->o_arg.clientid = server->nfs_client->cl_clientid; p->o_arg.id = sp->so_owner_id.id; p->o_arg.name = &p->path.dentry->d_name; @@ -324,10 +400,13 @@ static int nfs4_wait_for_completion_rpc_task(struct rpc_task *task) return ret; } -static int can_open_cached(struct nfs4_state *state, int mode) +static int can_open_cached(struct nfs4_state *state, fmode_t mode, int open_mode) { int ret = 0; - switch (mode & (FMODE_READ|FMODE_WRITE|O_EXCL)) { + + if (open_mode & O_EXCL) + goto out; + switch (mode & (FMODE_READ|FMODE_WRITE)) { case FMODE_READ: ret |= test_bit(NFS_O_RDONLY_STATE, &state->flags) != 0; break; @@ -337,21 +416,23 @@ static int can_open_cached(struct nfs4_state *state, int mode) case FMODE_READ|FMODE_WRITE: ret |= test_bit(NFS_O_RDWR_STATE, &state->flags) != 0; } +out: return ret; } -static int can_open_delegated(struct nfs_delegation *delegation, mode_t open_flags) +static int can_open_delegated(struct nfs_delegation *delegation, fmode_t fmode) { - if ((delegation->type & open_flags) != open_flags) + if ((delegation->type & fmode) != fmode) return 0; - if (delegation->flags & NFS_DELEGATION_NEED_RECLAIM) + if (test_bit(NFS_DELEGATION_NEED_RECLAIM, &delegation->flags)) return 0; + nfs_mark_delegation_referenced(delegation); return 1; } -static void update_open_stateflags(struct nfs4_state *state, mode_t open_flags) +static void update_open_stateflags(struct nfs4_state *state, fmode_t fmode) { - switch (open_flags) { + switch (fmode) { case FMODE_WRITE: state->n_wronly++; break; @@ -361,15 +442,15 @@ static void update_open_stateflags(struct nfs4_state *state, mode_t open_flags) case FMODE_READ|FMODE_WRITE: state->n_rdwr++; } - nfs4_state_set_mode_locked(state, state->state | open_flags); + nfs4_state_set_mode_locked(state, state->state | fmode); } -static void nfs_set_open_stateid_locked(struct nfs4_state *state, nfs4_stateid *stateid, int open_flags) +static void nfs_set_open_stateid_locked(struct nfs4_state *state, nfs4_stateid *stateid, fmode_t fmode) { if (test_bit(NFS_DELEGATED_STATE, &state->flags) == 0) memcpy(state->stateid.data, stateid->data, sizeof(state->stateid.data)); memcpy(state->open_stateid.data, stateid->data, sizeof(state->open_stateid.data)); - switch (open_flags) { + switch (fmode) { case FMODE_READ: set_bit(NFS_O_RDONLY_STATE, &state->flags); break; @@ -381,16 +462,15 @@ static void nfs_set_open_stateid_locked(struct nfs4_state *state, nfs4_stateid * } } -static void nfs_set_open_stateid(struct nfs4_state *state, nfs4_stateid *stateid, int open_flags) +static void nfs_set_open_stateid(struct nfs4_state *state, nfs4_stateid *stateid, fmode_t fmode) { write_seqlock(&state->seqlock); - nfs_set_open_stateid_locked(state, stateid, open_flags); + nfs_set_open_stateid_locked(state, stateid, fmode); write_sequnlock(&state->seqlock); } -static void update_open_stateid(struct nfs4_state *state, nfs4_stateid *open_stateid, nfs4_stateid *deleg_stateid, int open_flags) +static void __update_open_stateid(struct nfs4_state *state, nfs4_stateid *open_stateid, const nfs4_stateid *deleg_stateid, fmode_t fmode) { - open_flags &= (FMODE_READ|FMODE_WRITE); /* * Protect the call to nfs4_state_set_mode_locked and * serialise the stateid update @@ -401,20 +481,60 @@ static void update_open_stateid(struct nfs4_state *state, nfs4_stateid *open_sta set_bit(NFS_DELEGATED_STATE, &state->flags); } if (open_stateid != NULL) - nfs_set_open_stateid_locked(state, open_stateid, open_flags); + nfs_set_open_stateid_locked(state, open_stateid, fmode); write_sequnlock(&state->seqlock); spin_lock(&state->owner->so_lock); - update_open_stateflags(state, open_flags); + update_open_stateflags(state, fmode); spin_unlock(&state->owner->so_lock); } -static void nfs4_return_incompatible_delegation(struct inode *inode, mode_t open_flags) +static int update_open_stateid(struct nfs4_state *state, nfs4_stateid *open_stateid, nfs4_stateid *delegation, fmode_t fmode) +{ + struct nfs_inode *nfsi = NFS_I(state->inode); + struct nfs_delegation *deleg_cur; + int ret = 0; + + fmode &= (FMODE_READ|FMODE_WRITE); + + rcu_read_lock(); + deleg_cur = rcu_dereference(nfsi->delegation); + if (deleg_cur == NULL) + goto no_delegation; + + spin_lock(&deleg_cur->lock); + if (nfsi->delegation != deleg_cur || + (deleg_cur->type & fmode) != fmode) + goto no_delegation_unlock; + + if (delegation == NULL) + delegation = &deleg_cur->stateid; + else if (memcmp(deleg_cur->stateid.data, delegation->data, NFS4_STATEID_SIZE) != 0) + goto no_delegation_unlock; + + nfs_mark_delegation_referenced(deleg_cur); + __update_open_stateid(state, open_stateid, &deleg_cur->stateid, fmode); + ret = 1; +no_delegation_unlock: + spin_unlock(&deleg_cur->lock); +no_delegation: + rcu_read_unlock(); + + if (!ret && open_stateid != NULL) { + __update_open_stateid(state, open_stateid, NULL, fmode); + ret = 1; + } + + return ret; +} + + +static void nfs4_return_incompatible_delegation(struct inode *inode, fmode_t fmode) { struct nfs_delegation *delegation; rcu_read_lock(); delegation = rcu_dereference(NFS_I(inode)->delegation); - if (delegation == NULL || (delegation->type & open_flags) == open_flags) { + if (delegation == NULL || (delegation->type & fmode) == fmode) { rcu_read_unlock(); return; } @@ -427,27 +547,28 @@ static struct nfs4_state *nfs4_try_open_cached(struct nfs4_opendata *opendata) struct nfs4_state *state = opendata->state; struct nfs_inode *nfsi = NFS_I(state->inode); struct nfs_delegation *delegation; - int open_mode = opendata->o_arg.open_flags & (FMODE_READ|FMODE_WRITE|O_EXCL); + int open_mode = opendata->o_arg.open_flags & O_EXCL; + fmode_t fmode = opendata->o_arg.fmode; nfs4_stateid stateid; int ret = -EAGAIN; - rcu_read_lock(); - delegation = rcu_dereference(nfsi->delegation); for (;;) { - if (can_open_cached(state, open_mode)) { + if (can_open_cached(state, fmode, open_mode)) { spin_lock(&state->owner->so_lock); - if (can_open_cached(state, open_mode)) { - update_open_stateflags(state, open_mode); + if (can_open_cached(state, fmode, open_mode)) { + update_open_stateflags(state, fmode); spin_unlock(&state->owner->so_lock); - rcu_read_unlock(); goto out_return_state; } spin_unlock(&state->owner->so_lock); } - if (delegation == NULL) - break; - if (!can_open_delegated(delegation, open_mode)) + rcu_read_lock(); + delegation = rcu_dereference(nfsi->delegation); + if (delegation == NULL || + !can_open_delegated(delegation, fmode)) { + rcu_read_unlock(); break; + } /* Save the delegation */ memcpy(stateid.data, delegation->stateid.data, sizeof(stateid.data)); rcu_read_unlock(); @@ -455,19 +576,11 @@ static struct nfs4_state *nfs4_try_open_cached(struct nfs4_opendata *opendata) if (ret != 0) goto out; ret = -EAGAIN; - rcu_read_lock(); - delegation = rcu_dereference(nfsi->delegation); - /* If no delegation, try a cached open */ - if (delegation == NULL) - continue; - /* Is the delegation still valid? */ - if (memcmp(stateid.data, delegation->stateid.data, sizeof(stateid.data)) != 0) - continue; - rcu_read_unlock(); - update_open_stateid(state, NULL, &stateid, open_mode); - goto out_return_state; + + /* Try to update the stateid using the delegation */ + if (update_open_stateid(state, NULL, &stateid, fmode)) + goto out_return_state; } - rcu_read_unlock(); out: return ERR_PTR(ret); out_return_state: @@ -480,7 +593,6 @@ static struct nfs4_state *nfs4_opendata_to_nfs4_state(struct nfs4_opendata *data struct inode *inode; struct nfs4_state *state = NULL; struct nfs_delegation *delegation; - nfs4_stateid *deleg_stateid = NULL; int ret; if (!data->rpc_done) { @@ -507,7 +619,7 @@ static struct nfs4_state *nfs4_opendata_to_nfs4_state(struct nfs4_opendata *data if (delegation) delegation_flags = delegation->flags; rcu_read_unlock(); - if (!(delegation_flags & NFS_DELEGATION_NEED_RECLAIM)) + if ((delegation_flags & 1UL<<NFS_DELEGATION_NEED_RECLAIM) == 0) nfs_inode_set_delegation(state->inode, data->owner->so_cred, &data->o_res); @@ -516,12 +628,9 @@ static struct nfs4_state *nfs4_opendata_to_nfs4_state(struct nfs4_opendata *data data->owner->so_cred, &data->o_res); } - rcu_read_lock(); - delegation = rcu_dereference(NFS_I(inode)->delegation); - if (delegation != NULL) - deleg_stateid = &delegation->stateid; - update_open_stateid(state, &data->o_res.stateid, deleg_stateid, data->o_arg.open_flags); - rcu_read_unlock(); + + update_open_stateid(state, &data->o_res.stateid, NULL, + data->o_arg.fmode); iput(inode); out: return state; @@ -552,7 +661,7 @@ static struct nfs4_opendata *nfs4_open_recoverdata_alloc(struct nfs_open_context { struct nfs4_opendata *opendata; - opendata = nfs4_opendata_alloc(&ctx->path, state->owner, 0, NULL); + opendata = nfs4_opendata_alloc(&ctx->path, state->owner, 0, 0, NULL); if (opendata == NULL) return ERR_PTR(-ENOMEM); opendata->state = state; @@ -560,12 +669,13 @@ static struct nfs4_opendata *nfs4_open_recoverdata_alloc(struct nfs_open_context return opendata; } -static int nfs4_open_recover_helper(struct nfs4_opendata *opendata, mode_t openflags, struct nfs4_state **res) +static int nfs4_open_recover_helper(struct nfs4_opendata *opendata, fmode_t fmode, struct nfs4_state **res) { struct nfs4_state *newstate; int ret; - opendata->o_arg.open_flags = openflags; + opendata->o_arg.open_flags = 0; + opendata->o_arg.fmode = fmode; memset(&opendata->o_res, 0, sizeof(opendata->o_res)); memset(&opendata->c_res, 0, sizeof(opendata->c_res)); nfs4_init_opendata_res(opendata); @@ -575,7 +685,7 @@ static int nfs4_open_recover_helper(struct nfs4_opendata *opendata, mode_t openf newstate = nfs4_opendata_to_nfs4_state(opendata); if (IS_ERR(newstate)) return PTR_ERR(newstate); - nfs4_close_state(&opendata->path, newstate, openflags); + nfs4_close_state(&opendata->path, newstate, fmode); *res = newstate; return 0; } @@ -631,7 +741,7 @@ static int _nfs4_do_open_reclaim(struct nfs_open_context *ctx, struct nfs4_state { struct nfs_delegation *delegation; struct nfs4_opendata *opendata; - int delegation_type = 0; + fmode_t delegation_type = 0; int status; opendata = nfs4_open_recoverdata_alloc(ctx, state); @@ -641,7 +751,7 @@ static int _nfs4_do_open_reclaim(struct nfs_open_context *ctx, struct nfs4_state opendata->o_arg.fh = NFS_FH(state->inode); rcu_read_lock(); delegation = rcu_dereference(NFS_I(state->inode)->delegation); - if (delegation != NULL && (delegation->flags & NFS_DELEGATION_NEED_RECLAIM) != 0) + if (delegation != NULL && test_bit(NFS_DELEGATION_NEED_RECLAIM, &delegation->flags) != 0) delegation_type = delegation->type; rcu_read_unlock(); opendata->o_arg.u.delegation_type = delegation_type; @@ -744,7 +854,7 @@ static void nfs4_open_confirm_release(void *calldata) goto out_free; state = nfs4_opendata_to_nfs4_state(data); if (!IS_ERR(state)) - nfs4_close_state(&data->path, state, data->o_arg.open_flags); + nfs4_close_state(&data->path, state, data->o_arg.fmode); out_free: nfs4_opendata_put(data); } @@ -808,12 +918,12 @@ static void nfs4_open_prepare(struct rpc_task *task, void *calldata) if (data->state != NULL) { struct nfs_delegation *delegation; - if (can_open_cached(data->state, data->o_arg.open_flags & (FMODE_READ|FMODE_WRITE|O_EXCL))) + if (can_open_cached(data->state, data->o_arg.fmode, data->o_arg.open_flags)) goto out_no_action; rcu_read_lock(); delegation = rcu_dereference(NFS_I(data->state->inode)->delegation); if (delegation != NULL && - (delegation->flags & NFS_DELEGATION_NEED_RECLAIM) == 0) { + test_bit(NFS_DELEGATION_NEED_RECLAIM, &delegation->flags) == 0) { rcu_read_unlock(); goto out_no_action; } @@ -877,7 +987,7 @@ static void nfs4_open_release(void *calldata) goto out_free; state = nfs4_opendata_to_nfs4_state(data); if (!IS_ERR(state)) - nfs4_close_state(&data->path, state, data->o_arg.open_flags); + nfs4_close_state(&data->path, state, data->o_arg.fmode); out_free: nfs4_opendata_put(data); } @@ -955,10 +1065,11 @@ static int nfs4_recover_expired_lease(struct nfs_server *server) int ret; for (;;) { - ret = nfs4_wait_clnt_recover(server->client, clp); + ret = nfs4_wait_clnt_recover(clp); if (ret != 0) return ret; - if (!test_and_clear_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state)) + if (!test_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state) && + !test_bit(NFS4CLNT_CHECK_LEASE,&clp->cl_state)) break; nfs4_schedule_state_recovery(clp); } @@ -993,8 +1104,9 @@ static inline int nfs4_do_open_expired(struct nfs_open_context *ctx, struct nfs4 do { err = _nfs4_open_expired(ctx, state); - if (err == -NFS4ERR_DELAY) - nfs4_handle_exception(server, err, &exception); + if (err != -NFS4ERR_DELAY) + break; + nfs4_handle_exception(server, err, &exception); } while (exception.retry); return err; } @@ -1031,12 +1143,11 @@ static inline void nfs4_exclusive_attrset(struct nfs4_opendata *opendata, struct /* * Returns a referenced nfs4_state */ -static int _nfs4_do_open(struct inode *dir, struct path *path, int flags, struct iattr *sattr, struct rpc_cred *cred, struct nfs4_state **res) +static int _nfs4_do_open(struct inode *dir, struct path *path, fmode_t fmode, int flags, struct iattr *sattr, struct rpc_cred *cred, struct nfs4_state **res) { struct nfs4_state_owner *sp; struct nfs4_state *state = NULL; struct nfs_server *server = NFS_SERVER(dir); - struct nfs_client *clp = server->nfs_client; struct nfs4_opendata *opendata; int status; @@ -1050,12 +1161,11 @@ static int _nfs4_do_open(struct inode *dir, struct path *path, int flags, struct if (status != 0) goto err_put_state_owner; if (path->dentry->d_inode != NULL) - nfs4_return_incompatible_delegation(path->dentry->d_inode, flags & (FMODE_READ|FMODE_WRITE)); - down_read(&clp->cl_sem); + nfs4_return_incompatible_delegation(path->dentry->d_inode, fmode); status = -ENOMEM; - opendata = nfs4_opendata_alloc(path, sp, flags, sattr); + opendata = nfs4_opendata_alloc(path, sp, fmode, flags, sattr); if (opendata == NULL) - goto err_release_rwsem; + goto err_put_state_owner; if (path->dentry->d_inode != NULL) opendata->state = nfs4_get_open_state(path->dentry->d_inode, sp); @@ -1073,13 +1183,10 @@ static int _nfs4_do_open(struct inode *dir, struct path *path, int flags, struct goto err_opendata_put; nfs4_opendata_put(opendata); nfs4_put_state_owner(sp); - up_read(&clp->cl_sem); *res = state; return 0; err_opendata_put: nfs4_opendata_put(opendata); -err_release_rwsem: - up_read(&clp->cl_sem); err_put_state_owner: nfs4_put_state_owner(sp); out_err: @@ -1088,14 +1195,14 @@ out_err: } -static struct nfs4_state *nfs4_do_open(struct inode *dir, struct path *path, int flags, struct iattr *sattr, struct rpc_cred *cred) +static struct nfs4_state *nfs4_do_open(struct inode *dir, struct path *path, fmode_t fmode, int flags, struct iattr *sattr, struct rpc_cred *cred) { struct nfs4_exception exception = { }; struct nfs4_state *res; int status; do { - status = _nfs4_do_open(dir, path, flags, sattr, cred, &res); + status = _nfs4_do_open(dir, path, fmode, flags, sattr, cred, &res); if (status == 0) break; /* NOTE: BAD_SEQID means the server and client disagree about the @@ -1230,10 +1337,13 @@ static void nfs4_close_done(struct rpc_task *task, void *data) renew_lease(server, calldata->timestamp); break; case -NFS4ERR_STALE_STATEID: + case -NFS4ERR_OLD_STATEID: + case -NFS4ERR_BAD_STATEID: case -NFS4ERR_EXPIRED: - break; + if (calldata->arg.fmode == 0) + break; default: - if (nfs4_async_handle_error(task, server) == -EAGAIN) { + if (nfs4_async_handle_error(task, server, state) == -EAGAIN) { rpc_restart_call(task); return; } @@ -1272,10 +1382,10 @@ static void nfs4_close_prepare(struct rpc_task *task, void *data) nfs_fattr_init(calldata->res.fattr); if (test_bit(NFS_O_RDONLY_STATE, &state->flags) != 0) { task->tk_msg.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_OPEN_DOWNGRADE]; - calldata->arg.open_flags = FMODE_READ; + calldata->arg.fmode = FMODE_READ; } else if (test_bit(NFS_O_WRONLY_STATE, &state->flags) != 0) { task->tk_msg.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_OPEN_DOWNGRADE]; - calldata->arg.open_flags = FMODE_WRITE; + calldata->arg.fmode = FMODE_WRITE; } calldata->timestamp = jiffies; rpc_call_start(task); @@ -1328,6 +1438,7 @@ int nfs4_do_close(struct path *path, struct nfs4_state *state, int wait) calldata->arg.seqid = nfs_alloc_seqid(&state->owner->so_seqid); if (calldata->arg.seqid == NULL) goto out_free_calldata; + calldata->arg.fmode = 0; calldata->arg.bitmask = server->attr_bitmask; calldata->res.fattr = &calldata->fattr; calldata->res.seqid = calldata->arg.seqid; @@ -1354,13 +1465,13 @@ out: return status; } -static int nfs4_intent_set_file(struct nameidata *nd, struct path *path, struct nfs4_state *state) +static int nfs4_intent_set_file(struct nameidata *nd, struct path *path, struct nfs4_state *state, fmode_t fmode) { struct file *filp; int ret; /* If the open_intent is for execute, we have an extra check to make */ - if (nd->intent.open.flags & FMODE_EXEC) { + if (fmode & FMODE_EXEC) { ret = nfs_may_open(state->inode, state->owner->so_cred, nd->intent.open.flags); @@ -1376,7 +1487,7 @@ static int nfs4_intent_set_file(struct nameidata *nd, struct path *path, struct } ret = PTR_ERR(filp); out_close: - nfs4_close_sync(path, state, nd->intent.open.flags); + nfs4_close_sync(path, state, fmode & (FMODE_READ|FMODE_WRITE)); return ret; } @@ -1392,6 +1503,7 @@ nfs4_atomic_open(struct inode *dir, struct dentry *dentry, struct nameidata *nd) struct rpc_cred *cred; struct nfs4_state *state; struct dentry *res; + fmode_t fmode = nd->intent.open.flags & (FMODE_READ | FMODE_WRITE | FMODE_EXEC); if (nd->flags & LOOKUP_CREATE) { attr.ia_mode = nd->intent.open.create_mode; @@ -1409,7 +1521,7 @@ nfs4_atomic_open(struct inode *dir, struct dentry *dentry, struct nameidata *nd) parent = dentry->d_parent; /* Protect against concurrent sillydeletes */ nfs_block_sillyrename(parent); - state = nfs4_do_open(dir, &path, nd->intent.open.flags, &attr, cred); + state = nfs4_do_open(dir, &path, fmode, nd->intent.open.flags, &attr, cred); put_rpccred(cred); if (IS_ERR(state)) { if (PTR_ERR(state) == -ENOENT) { @@ -1424,7 +1536,7 @@ nfs4_atomic_open(struct inode *dir, struct dentry *dentry, struct nameidata *nd) path.dentry = res; nfs_set_verifier(path.dentry, nfs_save_change_attribute(dir)); nfs_unblock_sillyrename(parent); - nfs4_intent_set_file(nd, &path, state); + nfs4_intent_set_file(nd, &path, state, fmode); return res; } @@ -1437,11 +1549,12 @@ nfs4_open_revalidate(struct inode *dir, struct dentry *dentry, int openflags, st }; struct rpc_cred *cred; struct nfs4_state *state; + fmode_t fmode = openflags & (FMODE_READ | FMODE_WRITE); cred = rpc_lookup_cred(); if (IS_ERR(cred)) return PTR_ERR(cred); - state = nfs4_do_open(dir, &path, openflags, NULL, cred); + state = nfs4_do_open(dir, &path, fmode, openflags, NULL, cred); put_rpccred(cred); if (IS_ERR(state)) { switch (PTR_ERR(state)) { @@ -1458,10 +1571,10 @@ nfs4_open_revalidate(struct inode *dir, struct dentry *dentry, int openflags, st } if (state->inode == dentry->d_inode) { nfs_set_verifier(dentry, nfs_save_change_attribute(dir)); - nfs4_intent_set_file(nd, &path, state); + nfs4_intent_set_file(nd, &path, state, fmode); return 1; } - nfs4_close_sync(&path, state, openflags); + nfs4_close_sync(&path, state, fmode); out_drop: d_drop(dentry); return 0; @@ -1659,8 +1772,10 @@ nfs4_proc_setattr(struct dentry *dentry, struct nfs_fattr *fattr, struct nfs_open_context *ctx; ctx = nfs_file_open_context(sattr->ia_file); - cred = ctx->cred; - state = ctx->state; + if (ctx) { + cred = ctx->cred; + state = ctx->state; + } } status = nfs4_do_setattr(inode, cred, fattr, sattr, state); @@ -1885,6 +2000,7 @@ nfs4_proc_create(struct inode *dir, struct dentry *dentry, struct iattr *sattr, }; struct nfs4_state *state; struct rpc_cred *cred; + fmode_t fmode = flags & (FMODE_READ | FMODE_WRITE); int status = 0; cred = rpc_lookup_cred(); @@ -1892,7 +2008,7 @@ nfs4_proc_create(struct inode *dir, struct dentry *dentry, struct iattr *sattr, status = PTR_ERR(cred); goto out; } - state = nfs4_do_open(dir, &path, flags, sattr, cred); + state = nfs4_do_open(dir, &path, fmode, flags, sattr, cred); d_drop(dentry); if (IS_ERR(state)) { status = PTR_ERR(state); @@ -1908,9 +2024,9 @@ nfs4_proc_create(struct inode *dir, struct dentry *dentry, struct iattr *sattr, nfs_post_op_update_inode(state->inode, &fattr); } if (status == 0 && (nd->flags & LOOKUP_OPEN) != 0) - status = nfs4_intent_set_file(nd, &path, state); + status = nfs4_intent_set_file(nd, &path, state, fmode); else - nfs4_close_sync(&path, state, flags); + nfs4_close_sync(&path, state, fmode); out_putcred: put_rpccred(cred); out: @@ -1972,7 +2088,7 @@ static int nfs4_proc_unlink_done(struct rpc_task *task, struct inode *dir) { struct nfs_removeres *res = task->tk_msg.rpc_resp; - if (nfs4_async_handle_error(task, res->server) == -EAGAIN) + if (nfs4_async_handle_error(task, res->server, NULL) == -EAGAIN) return 0; update_changeattr(dir, &res->cinfo); nfs_post_op_update_inode(dir, &res->dir_attr); @@ -2400,7 +2516,7 @@ static int nfs4_read_done(struct rpc_task *task, struct nfs_read_data *data) { struct nfs_server *server = NFS_SERVER(data->inode); - if (nfs4_async_handle_error(task, server) == -EAGAIN) { + if (nfs4_async_handle_error(task, server, data->args.context->state) == -EAGAIN) { rpc_restart_call(task); return -EAGAIN; } @@ -2421,7 +2537,7 @@ static int nfs4_write_done(struct rpc_task *task, struct nfs_write_data *data) { struct inode *inode = data->inode; - if (nfs4_async_handle_error(task, NFS_SERVER(inode)) == -EAGAIN) { + if (nfs4_async_handle_error(task, NFS_SERVER(inode), data->args.context->state) == -EAGAIN) { rpc_restart_call(task); return -EAGAIN; } @@ -2447,7 +2563,7 @@ static int nfs4_commit_done(struct rpc_task *task, struct nfs_write_data *data) { struct inode *inode = data->inode; - if (nfs4_async_handle_error(task, NFS_SERVER(inode)) == -EAGAIN) { + if (nfs4_async_handle_error(task, NFS_SERVER(inode), NULL) == -EAGAIN) { rpc_restart_call(task); return -EAGAIN; } @@ -2740,19 +2856,25 @@ static int nfs4_proc_set_acl(struct inode *inode, const void *buf, size_t buflen } static int -nfs4_async_handle_error(struct rpc_task *task, const struct nfs_server *server) +nfs4_async_handle_error(struct rpc_task *task, const struct nfs_server *server, struct nfs4_state *state) { struct nfs_client *clp = server->nfs_client; if (!clp || task->tk_status >= 0) return 0; switch(task->tk_status) { + case -NFS4ERR_ADMIN_REVOKED: + case -NFS4ERR_BAD_STATEID: + case -NFS4ERR_OPENMODE: + if (state == NULL) + break; + nfs4_state_mark_reclaim_nograce(clp, state); case -NFS4ERR_STALE_CLIENTID: case -NFS4ERR_STALE_STATEID: case -NFS4ERR_EXPIRED: rpc_sleep_on(&clp->cl_rpcwaitq, task, NULL); nfs4_schedule_state_recovery(clp); - if (test_bit(NFS4CLNT_STATE_RECOVER, &clp->cl_state) == 0) + if (test_bit(NFS4CLNT_MANAGER_RUNNING, &clp->cl_state) == 0) rpc_wake_up_queued_task(&clp->cl_rpcwaitq, task); task->tk_status = 0; return -EAGAIN; @@ -2770,79 +2892,6 @@ nfs4_async_handle_error(struct rpc_task *task, const struct nfs_server *server) return 0; } -static int nfs4_wait_bit_killable(void *word) -{ - if (fatal_signal_pending(current)) - return -ERESTARTSYS; - schedule(); - return 0; -} - -static int nfs4_wait_clnt_recover(struct rpc_clnt *clnt, struct nfs_client *clp) -{ - int res; - - might_sleep(); - - rwsem_acquire(&clp->cl_sem.dep_map, 0, 0, _RET_IP_); - - res = wait_on_bit(&clp->cl_state, NFS4CLNT_STATE_RECOVER, - nfs4_wait_bit_killable, TASK_KILLABLE); - - rwsem_release(&clp->cl_sem.dep_map, 1, _RET_IP_); - return res; -} - -static int nfs4_delay(struct rpc_clnt *clnt, long *timeout) -{ - int res = 0; - - might_sleep(); - - if (*timeout <= 0) - *timeout = NFS4_POLL_RETRY_MIN; - if (*timeout > NFS4_POLL_RETRY_MAX) - *timeout = NFS4_POLL_RETRY_MAX; - schedule_timeout_killable(*timeout); - if (fatal_signal_pending(current)) - res = -ERESTARTSYS; - *timeout <<= 1; - return res; -} - -/* This is the error handling routine for processes that are allowed - * to sleep. - */ -static int nfs4_handle_exception(const struct nfs_server *server, int errorcode, struct nfs4_exception *exception) -{ - struct nfs_client *clp = server->nfs_client; - int ret = errorcode; - - exception->retry = 0; - switch(errorcode) { - case 0: - return 0; - case -NFS4ERR_STALE_CLIENTID: - case -NFS4ERR_STALE_STATEID: - case -NFS4ERR_EXPIRED: - nfs4_schedule_state_recovery(clp); - ret = nfs4_wait_clnt_recover(server->client, clp); - if (ret == 0) - exception->retry = 1; - break; - case -NFS4ERR_FILE_OPEN: - case -NFS4ERR_GRACE: - case -NFS4ERR_DELAY: - ret = nfs4_delay(server->client, &exception->timeout); - if (ret != 0) - break; - case -NFS4ERR_OLD_STATEID: - exception->retry = 1; - } - /* We failed to handle the error */ - return nfs4_map_errors(ret); -} - int nfs4_proc_setclientid(struct nfs_client *clp, u32 program, unsigned short port, struct rpc_cred *cred) { nfs4_verifier sc_verifier; @@ -2914,7 +2963,6 @@ static int _nfs4_proc_setclientid_confirm(struct nfs_client *clp, struct rpc_cre spin_lock(&clp->cl_lock); clp->cl_lease_time = fsinfo.lease_time * HZ; clp->cl_last_renewal = now; - clear_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state); spin_unlock(&clp->cl_lock); } return status; @@ -3072,7 +3120,6 @@ static int _nfs4_proc_getlk(struct nfs4_state *state, int cmd, struct file_lock struct nfs4_lock_state *lsp; int status; - down_read(&clp->cl_sem); arg.lock_owner.clientid = clp->cl_clientid; status = nfs4_set_lock_state(state, request); if (status != 0) @@ -3089,7 +3136,6 @@ static int _nfs4_proc_getlk(struct nfs4_state *state, int cmd, struct file_lock } request->fl_ops->fl_release_private(request); out: - up_read(&clp->cl_sem); return status; } @@ -3179,11 +3225,13 @@ static void nfs4_locku_done(struct rpc_task *task, void *data) sizeof(calldata->lsp->ls_stateid.data)); renew_lease(calldata->server, calldata->timestamp); break; + case -NFS4ERR_BAD_STATEID: + case -NFS4ERR_OLD_STATEID: case -NFS4ERR_STALE_STATEID: case -NFS4ERR_EXPIRED: break; default: - if (nfs4_async_handle_error(task, calldata->server) == -EAGAIN) + if (nfs4_async_handle_error(task, calldata->server, NULL) == -EAGAIN) rpc_restart_call(task); } } @@ -3246,6 +3294,7 @@ static struct rpc_task *nfs4_do_unlck(struct file_lock *fl, static int nfs4_proc_unlck(struct nfs4_state *state, int cmd, struct file_lock *request) { + struct nfs_inode *nfsi = NFS_I(state->inode); struct nfs_seqid *seqid; struct nfs4_lock_state *lsp; struct rpc_task *task; @@ -3255,8 +3304,12 @@ static int nfs4_proc_unlck(struct nfs4_state *state, int cmd, struct file_lock * status = nfs4_set_lock_state(state, request); /* Unlock _before_ we do the RPC call */ request->fl_flags |= FL_EXISTS; - if (do_vfs_lock(request->fl_file, request) == -ENOENT) + down_read(&nfsi->rwsem); + if (do_vfs_lock(request->fl_file, request) == -ENOENT) { + up_read(&nfsi->rwsem); goto out; + } + up_read(&nfsi->rwsem); if (status != 0) goto out; /* Is this a delegated lock? */ @@ -3482,7 +3535,7 @@ static int nfs4_lock_expired(struct nfs4_state *state, struct file_lock *request static int _nfs4_proc_setlk(struct nfs4_state *state, int cmd, struct file_lock *request) { - struct nfs_client *clp = state->owner->so_client; + struct nfs_inode *nfsi = NFS_I(state->inode); unsigned char fl_flags = request->fl_flags; int status; @@ -3494,19 +3547,13 @@ static int _nfs4_proc_setlk(struct nfs4_state *state, int cmd, struct file_lock status = do_vfs_lock(request->fl_file, request); if (status < 0) goto out; - down_read(&clp->cl_sem); + down_read(&nfsi->rwsem); if (test_bit(NFS_DELEGATED_STATE, &state->flags)) { - struct nfs_inode *nfsi = NFS_I(state->inode); /* Yes: cache locks! */ - down_read(&nfsi->rwsem); /* ...but avoid races with delegation recall... */ - if (test_bit(NFS_DELEGATED_STATE, &state->flags)) { - request->fl_flags = fl_flags & ~FL_SLEEP; - status = do_vfs_lock(request->fl_file, request); - up_read(&nfsi->rwsem); - goto out_unlock; - } - up_read(&nfsi->rwsem); + request->fl_flags = fl_flags & ~FL_SLEEP; + status = do_vfs_lock(request->fl_file, request); + goto out_unlock; } status = _nfs4_do_setlk(state, cmd, request, 0); if (status != 0) @@ -3516,7 +3563,7 @@ static int _nfs4_proc_setlk(struct nfs4_state *state, int cmd, struct file_lock if (do_vfs_lock(request->fl_file, request) < 0) printk(KERN_WARNING "%s: VFS is out of sync with lock manager!\n", __func__); out_unlock: - up_read(&clp->cl_sem); + up_read(&nfsi->rwsem); out: request->fl_flags = fl_flags; return status; @@ -3662,11 +3709,15 @@ int nfs4_proc_fs_locations(struct inode *dir, const struct qstr *name, } struct nfs4_state_recovery_ops nfs4_reboot_recovery_ops = { + .owner_flag_bit = NFS_OWNER_RECLAIM_REBOOT, + .state_flag_bit = NFS_STATE_RECLAIM_REBOOT, .recover_open = nfs4_open_reclaim, .recover_lock = nfs4_lock_reclaim, }; -struct nfs4_state_recovery_ops nfs4_network_partition_recovery_ops = { +struct nfs4_state_recovery_ops nfs4_nograce_recovery_ops = { + .owner_flag_bit = NFS_OWNER_RECLAIM_NOGRACE, + .state_flag_bit = NFS_STATE_RECLAIM_NOGRACE, .recover_open = nfs4_open_expired, .recover_lock = nfs4_lock_expired, }; diff --git a/fs/nfs/nfs4renewd.c b/fs/nfs/nfs4renewd.c index 3305acbbe2a..f524e932ff7 100644 --- a/fs/nfs/nfs4renewd.c +++ b/fs/nfs/nfs4renewd.c @@ -65,7 +65,6 @@ nfs4_renew_state(struct work_struct *work) long lease, timeout; unsigned long last, now; - down_read(&clp->cl_sem); dprintk("%s: start\n", __func__); /* Are there any active superblocks? */ if (list_empty(&clp->cl_superblocks)) @@ -77,17 +76,19 @@ nfs4_renew_state(struct work_struct *work) timeout = (2 * lease) / 3 + (long)last - (long)now; /* Are we close to a lease timeout? */ if (time_after(now, last + lease/3)) { - cred = nfs4_get_renew_cred(clp); + cred = nfs4_get_renew_cred_locked(clp); + spin_unlock(&clp->cl_lock); if (cred == NULL) { - set_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state); - spin_unlock(&clp->cl_lock); + if (list_empty(&clp->cl_delegations)) { + set_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state); + goto out; + } nfs_expire_all_delegations(clp); - goto out; + } else { + /* Queue an asynchronous RENEW. */ + nfs4_proc_async_renew(clp, cred); + put_rpccred(cred); } - spin_unlock(&clp->cl_lock); - /* Queue an asynchronous RENEW. */ - nfs4_proc_async_renew(clp, cred); - put_rpccred(cred); timeout = (2 * lease) / 3; spin_lock(&clp->cl_lock); } else @@ -100,12 +101,11 @@ nfs4_renew_state(struct work_struct *work) cancel_delayed_work(&clp->cl_renewd); schedule_delayed_work(&clp->cl_renewd, timeout); spin_unlock(&clp->cl_lock); + nfs_expire_unreferenced_delegations(clp); out: - up_read(&clp->cl_sem); dprintk("%s: done\n", __func__); } -/* Must be called with clp->cl_sem locked for writes */ void nfs4_schedule_state_renewal(struct nfs_client *clp) { diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c index 401ef8b28f9..2022fe47966 100644 --- a/fs/nfs/nfs4state.c +++ b/fs/nfs/nfs4state.c @@ -71,14 +71,12 @@ static int nfs4_init_client(struct nfs_client *clp, struct rpc_cred *cred) return status; } -static struct rpc_cred *nfs4_get_machine_cred(struct nfs_client *clp) +static struct rpc_cred *nfs4_get_machine_cred_locked(struct nfs_client *clp) { struct rpc_cred *cred = NULL; - spin_lock(&clp->cl_lock); if (clp->cl_machine_cred != NULL) cred = get_rpccred(clp->cl_machine_cred); - spin_unlock(&clp->cl_lock); return cred; } @@ -94,7 +92,7 @@ static void nfs4_clear_machine_cred(struct nfs_client *clp) put_rpccred(cred); } -struct rpc_cred *nfs4_get_renew_cred(struct nfs_client *clp) +struct rpc_cred *nfs4_get_renew_cred_locked(struct nfs_client *clp) { struct nfs4_state_owner *sp; struct rb_node *pos; @@ -110,13 +108,24 @@ struct rpc_cred *nfs4_get_renew_cred(struct nfs_client *clp) return cred; } +static struct rpc_cred *nfs4_get_renew_cred(struct nfs_client *clp) +{ + struct rpc_cred *cred; + + spin_lock(&clp->cl_lock); + cred = nfs4_get_renew_cred_locked(clp); + spin_unlock(&clp->cl_lock); + return cred; +} + static struct rpc_cred *nfs4_get_setclientid_cred(struct nfs_client *clp) { struct nfs4_state_owner *sp; struct rb_node *pos; struct rpc_cred *cred; - cred = nfs4_get_machine_cred(clp); + spin_lock(&clp->cl_lock); + cred = nfs4_get_machine_cred_locked(clp); if (cred != NULL) goto out; pos = rb_first(&clp->cl_state_owners); @@ -125,6 +134,7 @@ static struct rpc_cred *nfs4_get_setclientid_cred(struct nfs_client *clp) cred = get_rpccred(sp->so_cred); } out: + spin_unlock(&clp->cl_lock); return cred; } @@ -295,10 +305,6 @@ nfs4_drop_state_owner(struct nfs4_state_owner *sp) } } -/* - * Note: must be called with clp->cl_sem held in order to prevent races - * with reboot recovery! - */ struct nfs4_state_owner *nfs4_get_state_owner(struct nfs_server *server, struct rpc_cred *cred) { struct nfs_client *clp = server->nfs_client; @@ -327,10 +333,6 @@ struct nfs4_state_owner *nfs4_get_state_owner(struct nfs_server *server, struct return sp; } -/* - * Must be called with clp->cl_sem held in order to avoid races - * with state recovery... - */ void nfs4_put_state_owner(struct nfs4_state_owner *sp) { struct nfs_client *clp = sp->so_client; @@ -361,18 +363,18 @@ nfs4_alloc_open_state(void) } void -nfs4_state_set_mode_locked(struct nfs4_state *state, mode_t mode) +nfs4_state_set_mode_locked(struct nfs4_state *state, fmode_t fmode) { - if (state->state == mode) + if (state->state == fmode) return; /* NB! List reordering - see the reclaim code for why. */ - if ((mode & FMODE_WRITE) != (state->state & FMODE_WRITE)) { - if (mode & FMODE_WRITE) + if ((fmode & FMODE_WRITE) != (state->state & FMODE_WRITE)) { + if (fmode & FMODE_WRITE) list_move(&state->open_states, &state->owner->so_states); else list_move_tail(&state->open_states, &state->owner->so_states); } - state->state = mode; + state->state = fmode; } static struct nfs4_state * @@ -432,10 +434,6 @@ out: return state; } -/* - * Beware! Caller must be holding exactly one - * reference to clp->cl_sem! - */ void nfs4_put_open_state(struct nfs4_state *state) { struct inode *inode = state->inode; @@ -456,16 +454,16 @@ void nfs4_put_open_state(struct nfs4_state *state) /* * Close the current file. */ -static void __nfs4_close(struct path *path, struct nfs4_state *state, mode_t mode, int wait) +static void __nfs4_close(struct path *path, struct nfs4_state *state, fmode_t fmode, int wait) { struct nfs4_state_owner *owner = state->owner; int call_close = 0; - int newstate; + fmode_t newstate; atomic_inc(&owner->so_count); /* Protect against nfs4_find_state() */ spin_lock(&owner->so_lock); - switch (mode & (FMODE_READ | FMODE_WRITE)) { + switch (fmode & (FMODE_READ | FMODE_WRITE)) { case FMODE_READ: state->n_rdonly--; break; @@ -500,14 +498,14 @@ static void __nfs4_close(struct path *path, struct nfs4_state *state, mode_t mod nfs4_do_close(path, state, wait); } -void nfs4_close_state(struct path *path, struct nfs4_state *state, mode_t mode) +void nfs4_close_state(struct path *path, struct nfs4_state *state, fmode_t fmode) { - __nfs4_close(path, state, mode, 0); + __nfs4_close(path, state, fmode, 0); } -void nfs4_close_sync(struct path *path, struct nfs4_state *state, mode_t mode) +void nfs4_close_sync(struct path *path, struct nfs4_state *state, fmode_t fmode) { - __nfs4_close(path, state, mode, 1); + __nfs4_close(path, state, fmode, 1); } /* @@ -568,7 +566,6 @@ static void nfs4_free_lock_state(struct nfs4_lock_state *lsp) * Return a compatible lock_state. If no initialized lock_state structure * exists, return an uninitialized one. * - * The caller must be holding clp->cl_sem */ static struct nfs4_lock_state *nfs4_get_lock_state(struct nfs4_state *state, fl_owner_t owner) { @@ -770,32 +767,34 @@ unlock: return status; } -static int reclaimer(void *); +static int nfs4_run_state_manager(void *); -static inline void nfs4_clear_recover_bit(struct nfs_client *clp) +static void nfs4_clear_state_manager_bit(struct nfs_client *clp) { smp_mb__before_clear_bit(); - clear_bit(NFS4CLNT_STATE_RECOVER, &clp->cl_state); + clear_bit(NFS4CLNT_MANAGER_RUNNING, &clp->cl_state); smp_mb__after_clear_bit(); - wake_up_bit(&clp->cl_state, NFS4CLNT_STATE_RECOVER); + wake_up_bit(&clp->cl_state, NFS4CLNT_MANAGER_RUNNING); rpc_wake_up(&clp->cl_rpcwaitq); } /* - * State recovery routine + * Schedule the nfs_client asynchronous state management routine */ -static void nfs4_recover_state(struct nfs_client *clp) +void nfs4_schedule_state_manager(struct nfs_client *clp) { struct task_struct *task; + if (test_and_set_bit(NFS4CLNT_MANAGER_RUNNING, &clp->cl_state) != 0) + return; __module_get(THIS_MODULE); atomic_inc(&clp->cl_count); - task = kthread_run(reclaimer, clp, "%s-reclaim", + task = kthread_run(nfs4_run_state_manager, clp, "%s-manager", rpc_peeraddr2str(clp->cl_rpcclient, RPC_DISPLAY_ADDR)); if (!IS_ERR(task)) return; - nfs4_clear_recover_bit(clp); + nfs4_clear_state_manager_bit(clp); nfs_put_client(clp); module_put(THIS_MODULE); } @@ -807,16 +806,42 @@ void nfs4_schedule_state_recovery(struct nfs_client *clp) { if (!clp) return; - if (test_and_set_bit(NFS4CLNT_STATE_RECOVER, &clp->cl_state) == 0) - nfs4_recover_state(clp); + if (!test_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state)) + set_bit(NFS4CLNT_CHECK_LEASE, &clp->cl_state); + nfs4_schedule_state_manager(clp); } -static int nfs4_reclaim_locks(struct nfs4_state_recovery_ops *ops, struct nfs4_state *state) +static int nfs4_state_mark_reclaim_reboot(struct nfs_client *clp, struct nfs4_state *state) +{ + + set_bit(NFS_STATE_RECLAIM_REBOOT, &state->flags); + /* Don't recover state that expired before the reboot */ + if (test_bit(NFS_STATE_RECLAIM_NOGRACE, &state->flags)) { + clear_bit(NFS_STATE_RECLAIM_REBOOT, &state->flags); + return 0; + } + set_bit(NFS_OWNER_RECLAIM_REBOOT, &state->owner->so_flags); + set_bit(NFS4CLNT_RECLAIM_REBOOT, &clp->cl_state); + return 1; +} + +int nfs4_state_mark_reclaim_nograce(struct nfs_client *clp, struct nfs4_state *state) +{ + set_bit(NFS_STATE_RECLAIM_NOGRACE, &state->flags); + clear_bit(NFS_STATE_RECLAIM_REBOOT, &state->flags); + set_bit(NFS_OWNER_RECLAIM_NOGRACE, &state->owner->so_flags); + set_bit(NFS4CLNT_RECLAIM_NOGRACE, &clp->cl_state); + return 1; +} + +static int nfs4_reclaim_locks(struct nfs4_state *state, const struct nfs4_state_recovery_ops *ops) { struct inode *inode = state->inode; + struct nfs_inode *nfsi = NFS_I(inode); struct file_lock *fl; int status = 0; + down_write(&nfsi->rwsem); for (fl = inode->i_flock; fl != NULL; fl = fl->fl_next) { if (!(fl->fl_flags & (FL_POSIX|FL_FLOCK))) continue; @@ -839,12 +864,14 @@ static int nfs4_reclaim_locks(struct nfs4_state_recovery_ops *ops, struct nfs4_s goto out_err; } } + up_write(&nfsi->rwsem); return 0; out_err: + up_write(&nfsi->rwsem); return status; } -static int nfs4_reclaim_open_state(struct nfs4_state_recovery_ops *ops, struct nfs4_state_owner *sp) +static int nfs4_reclaim_open_state(struct nfs4_state_owner *sp, const struct nfs4_state_recovery_ops *ops) { struct nfs4_state *state; struct nfs4_lock_state *lock; @@ -858,28 +885,34 @@ static int nfs4_reclaim_open_state(struct nfs4_state_recovery_ops *ops, struct n * recovering after a network partition or a reboot from a * server that doesn't support a grace period. */ +restart: + spin_lock(&sp->so_lock); list_for_each_entry(state, &sp->so_states, open_states) { + if (!test_and_clear_bit(ops->state_flag_bit, &state->flags)) + continue; if (state->state == 0) continue; + atomic_inc(&state->count); + spin_unlock(&sp->so_lock); status = ops->recover_open(sp, state); if (status >= 0) { - status = nfs4_reclaim_locks(ops, state); - if (status < 0) - goto out_err; - list_for_each_entry(lock, &state->lock_states, ls_locks) { - if (!(lock->ls_flags & NFS_LOCK_INITIALIZED)) - printk("%s: Lock reclaim failed!\n", + status = nfs4_reclaim_locks(state, ops); + if (status >= 0) { + list_for_each_entry(lock, &state->lock_states, ls_locks) { + if (!(lock->ls_flags & NFS_LOCK_INITIALIZED)) + printk("%s: Lock reclaim failed!\n", __func__); + } + nfs4_put_open_state(state); + goto restart; } - continue; } switch (status) { default: printk(KERN_ERR "%s: unhandled error %d. Zeroing state\n", __func__, status); case -ENOENT: - case -NFS4ERR_RECLAIM_BAD: - case -NFS4ERR_RECLAIM_CONFLICT: + case -ESTALE: /* * Open state on this file cannot be recovered * All we can do is revert to using the zero stateid. @@ -889,84 +922,176 @@ static int nfs4_reclaim_open_state(struct nfs4_state_recovery_ops *ops, struct n /* Mark the file as being 'closed' */ state->state = 0; break; + case -NFS4ERR_RECLAIM_BAD: + case -NFS4ERR_RECLAIM_CONFLICT: + nfs4_state_mark_reclaim_nograce(sp->so_client, state); + break; case -NFS4ERR_EXPIRED: case -NFS4ERR_NO_GRACE: + nfs4_state_mark_reclaim_nograce(sp->so_client, state); case -NFS4ERR_STALE_CLIENTID: goto out_err; } + nfs4_put_open_state(state); + goto restart; } + spin_unlock(&sp->so_lock); return 0; out_err: + nfs4_put_open_state(state); return status; } -static void nfs4_state_mark_reclaim(struct nfs_client *clp) +static void nfs4_clear_open_state(struct nfs4_state *state) +{ + struct nfs4_lock_state *lock; + + clear_bit(NFS_DELEGATED_STATE, &state->flags); + clear_bit(NFS_O_RDONLY_STATE, &state->flags); + clear_bit(NFS_O_WRONLY_STATE, &state->flags); + clear_bit(NFS_O_RDWR_STATE, &state->flags); + list_for_each_entry(lock, &state->lock_states, ls_locks) { + lock->ls_seqid.flags = 0; + lock->ls_flags &= ~NFS_LOCK_INITIALIZED; + } +} + +static void nfs4_state_mark_reclaim_helper(struct nfs_client *clp, int (*mark_reclaim)(struct nfs_client *clp, struct nfs4_state *state)) { struct nfs4_state_owner *sp; struct rb_node *pos; struct nfs4_state *state; - struct nfs4_lock_state *lock; /* Reset all sequence ids to zero */ for (pos = rb_first(&clp->cl_state_owners); pos != NULL; pos = rb_next(pos)) { sp = rb_entry(pos, struct nfs4_state_owner, so_client_node); - sp->so_seqid.counter = 0; sp->so_seqid.flags = 0; spin_lock(&sp->so_lock); list_for_each_entry(state, &sp->so_states, open_states) { - clear_bit(NFS_DELEGATED_STATE, &state->flags); - clear_bit(NFS_O_RDONLY_STATE, &state->flags); - clear_bit(NFS_O_WRONLY_STATE, &state->flags); - clear_bit(NFS_O_RDWR_STATE, &state->flags); - list_for_each_entry(lock, &state->lock_states, ls_locks) { - lock->ls_seqid.counter = 0; - lock->ls_seqid.flags = 0; - lock->ls_flags &= ~NFS_LOCK_INITIALIZED; - } + if (mark_reclaim(clp, state)) + nfs4_clear_open_state(state); } spin_unlock(&sp->so_lock); } } -static int reclaimer(void *ptr) +static void nfs4_state_start_reclaim_reboot(struct nfs_client *clp) +{ + /* Mark all delegations for reclaim */ + nfs_delegation_mark_reclaim(clp); + nfs4_state_mark_reclaim_helper(clp, nfs4_state_mark_reclaim_reboot); +} + +static void nfs4_state_end_reclaim_reboot(struct nfs_client *clp) { - struct nfs_client *clp = ptr; struct nfs4_state_owner *sp; struct rb_node *pos; - struct nfs4_state_recovery_ops *ops; - struct rpc_cred *cred; + struct nfs4_state *state; + + if (!test_and_clear_bit(NFS4CLNT_RECLAIM_REBOOT, &clp->cl_state)) + return; + + for (pos = rb_first(&clp->cl_state_owners); pos != NULL; pos = rb_next(pos)) { + sp = rb_entry(pos, struct nfs4_state_owner, so_client_node); + spin_lock(&sp->so_lock); + list_for_each_entry(state, &sp->so_states, open_states) { + if (!test_and_clear_bit(NFS_STATE_RECLAIM_REBOOT, &state->flags)) + continue; + nfs4_state_mark_reclaim_nograce(clp, state); + } + spin_unlock(&sp->so_lock); + } + + nfs_delegation_reap_unclaimed(clp); +} + +static void nfs_delegation_clear_all(struct nfs_client *clp) +{ + nfs_delegation_mark_reclaim(clp); + nfs_delegation_reap_unclaimed(clp); +} + +static void nfs4_state_start_reclaim_nograce(struct nfs_client *clp) +{ + nfs_delegation_clear_all(clp); + nfs4_state_mark_reclaim_helper(clp, nfs4_state_mark_reclaim_nograce); +} + +static void nfs4_state_end_reclaim_nograce(struct nfs_client *clp) +{ + clear_bit(NFS4CLNT_RECLAIM_NOGRACE, &clp->cl_state); +} + +static void nfs4_recovery_handle_error(struct nfs_client *clp, int error) +{ + switch (error) { + case -NFS4ERR_CB_PATH_DOWN: + nfs_handle_cb_pathdown(clp); + break; + case -NFS4ERR_STALE_CLIENTID: + case -NFS4ERR_LEASE_MOVED: + set_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state); + nfs4_state_start_reclaim_reboot(clp); + break; + case -NFS4ERR_EXPIRED: + set_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state); + nfs4_state_start_reclaim_nograce(clp); + } +} + +static int nfs4_do_reclaim(struct nfs_client *clp, const struct nfs4_state_recovery_ops *ops) +{ + struct rb_node *pos; int status = 0; - allow_signal(SIGKILL); +restart: + spin_lock(&clp->cl_lock); + for (pos = rb_first(&clp->cl_state_owners); pos != NULL; pos = rb_next(pos)) { + struct nfs4_state_owner *sp = rb_entry(pos, struct nfs4_state_owner, so_client_node); + if (!test_and_clear_bit(ops->owner_flag_bit, &sp->so_flags)) + continue; + atomic_inc(&sp->so_count); + spin_unlock(&clp->cl_lock); + status = nfs4_reclaim_open_state(sp, ops); + if (status < 0) { + set_bit(ops->owner_flag_bit, &sp->so_flags); + nfs4_put_state_owner(sp); + nfs4_recovery_handle_error(clp, status); + return status; + } + nfs4_put_state_owner(sp); + goto restart; + } + spin_unlock(&clp->cl_lock); + return status; +} - /* Ensure exclusive access to NFSv4 state */ - down_write(&clp->cl_sem); - /* Are there any NFS mounts out there? */ - if (list_empty(&clp->cl_superblocks)) - goto out; -restart_loop: - ops = &nfs4_network_partition_recovery_ops; - /* Are there any open files on this volume? */ +static int nfs4_check_lease(struct nfs_client *clp) +{ + struct rpc_cred *cred; + int status = -NFS4ERR_EXPIRED; + + /* Is the client already known to have an expired lease? */ + if (test_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state)) + return 0; cred = nfs4_get_renew_cred(clp); - if (cred != NULL) { - /* Yes there are: try to renew the old lease */ - status = nfs4_proc_renew(clp, cred); - put_rpccred(cred); - switch (status) { - case 0: - case -NFS4ERR_CB_PATH_DOWN: - goto out; - case -NFS4ERR_STALE_CLIENTID: - case -NFS4ERR_LEASE_MOVED: - ops = &nfs4_reboot_recovery_ops; - } - } else { - /* "reboot" to ensure we clear all state on the server */ - clp->cl_boot_time = CURRENT_TIME; + if (cred == NULL) { + cred = nfs4_get_setclientid_cred(clp); + if (cred == NULL) + goto out; } - /* We're going to have to re-establish a clientid */ - nfs4_state_mark_reclaim(clp); - status = -ENOENT; + status = nfs4_proc_renew(clp, cred); + put_rpccred(cred); +out: + nfs4_recovery_handle_error(clp, status); + return status; +} + +static int nfs4_reclaim_lease(struct nfs_client *clp) +{ + struct rpc_cred *cred; + int status = -ENOENT; + cred = nfs4_get_setclientid_cred(clp); if (cred != NULL) { status = nfs4_init_client(clp, cred); @@ -974,42 +1099,90 @@ restart_loop: /* Handle case where the user hasn't set up machine creds */ if (status == -EACCES && cred == clp->cl_machine_cred) { nfs4_clear_machine_cred(clp); - goto restart_loop; + status = -EAGAIN; } } - if (status) - goto out_error; - /* Mark all delegations for reclaim */ - nfs_delegation_mark_reclaim(clp); - /* Note: list is protected by exclusive lock on cl->cl_sem */ - for (pos = rb_first(&clp->cl_state_owners); pos != NULL; pos = rb_next(pos)) { - sp = rb_entry(pos, struct nfs4_state_owner, so_client_node); - status = nfs4_reclaim_open_state(ops, sp); - if (status < 0) { - if (status == -NFS4ERR_NO_GRACE) { - ops = &nfs4_network_partition_recovery_ops; - status = nfs4_reclaim_open_state(ops, sp); + return status; +} + +static void nfs4_state_manager(struct nfs_client *clp) +{ + int status = 0; + + /* Ensure exclusive access to NFSv4 state */ + for(;;) { + if (test_and_clear_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state)) { + /* We're going to have to re-establish a clientid */ + status = nfs4_reclaim_lease(clp); + if (status) { + set_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state); + if (status == -EAGAIN) + continue; + goto out_error; } + clear_bit(NFS4CLNT_CHECK_LEASE, &clp->cl_state); + } + + if (test_and_clear_bit(NFS4CLNT_CHECK_LEASE, &clp->cl_state)) { + status = nfs4_check_lease(clp); + if (status != 0) + continue; + } + + /* First recover reboot state... */ + if (test_and_clear_bit(NFS4CLNT_RECLAIM_REBOOT, &clp->cl_state)) { + status = nfs4_do_reclaim(clp, &nfs4_reboot_recovery_ops); if (status == -NFS4ERR_STALE_CLIENTID) - goto restart_loop; - if (status == -NFS4ERR_EXPIRED) - goto restart_loop; + continue; + nfs4_state_end_reclaim_reboot(clp); + continue; + } + + /* Now recover expired state... */ + if (test_and_clear_bit(NFS4CLNT_RECLAIM_NOGRACE, &clp->cl_state)) { + status = nfs4_do_reclaim(clp, &nfs4_nograce_recovery_ops); + if (status < 0) { + set_bit(NFS4CLNT_RECLAIM_NOGRACE, &clp->cl_state); + if (status == -NFS4ERR_STALE_CLIENTID) + continue; + if (status == -NFS4ERR_EXPIRED) + continue; + goto out_error; + } else + nfs4_state_end_reclaim_nograce(clp); + continue; } + + if (test_and_clear_bit(NFS4CLNT_DELEGRETURN, &clp->cl_state)) { + nfs_client_return_marked_delegations(clp); + continue; + } + + nfs4_clear_state_manager_bit(clp); + /* Did we race with an attempt to give us more work? */ + if (clp->cl_state == 0) + break; + if (test_and_set_bit(NFS4CLNT_MANAGER_RUNNING, &clp->cl_state) != 0) + break; } - nfs_delegation_reap_unclaimed(clp); -out: - up_write(&clp->cl_sem); - if (status == -NFS4ERR_CB_PATH_DOWN) - nfs_handle_cb_pathdown(clp); - nfs4_clear_recover_bit(clp); + return; +out_error: + printk(KERN_WARNING "Error: state manager failed on NFSv4 server %s" + " with error %d\n", clp->cl_hostname, -status); + if (test_bit(NFS4CLNT_RECLAIM_REBOOT, &clp->cl_state)) + nfs4_state_end_reclaim_reboot(clp); + nfs4_clear_state_manager_bit(clp); +} + +static int nfs4_run_state_manager(void *ptr) +{ + struct nfs_client *clp = ptr; + + allow_signal(SIGKILL); + nfs4_state_manager(clp); nfs_put_client(clp); module_put_and_exit(0); return 0; -out_error: - printk(KERN_WARNING "Error: state recovery failed on NFSv4 server %s" - " with error %d\n", clp->cl_hostname, -status); - set_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state); - goto out; } /* diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index b916297d233..d1e4c8f8a0a 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -8,7 +8,7 @@ * * Kendrick Smith <kmsmith@umich.edu> * Andy Adamson <andros@umich.edu> - * + * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: @@ -67,7 +67,7 @@ static int nfs4_stat_to_errno(int); #define NFS4_MAXTAGLEN 0 #endif -/* lock,open owner id: +/* lock,open owner id: * we currently use size 2 (u64) out of (NFS4_OPAQUE_LIMIT >> 2) */ #define open_owner_id_maxsz (1 + 4) @@ -541,6 +541,7 @@ static struct { struct compound_hdr { int32_t status; uint32_t nops; + __be32 * nops_p; uint32_t taglen; char * tag; }; @@ -578,7 +579,7 @@ static void encode_string(struct xdr_stream *xdr, unsigned int len, const char * xdr_encode_opaque(p, str, len); } -static int encode_compound_hdr(struct xdr_stream *xdr, struct compound_hdr *hdr) +static void encode_compound_hdr(struct xdr_stream *xdr, struct compound_hdr *hdr) { __be32 *p; @@ -588,8 +589,13 @@ static int encode_compound_hdr(struct xdr_stream *xdr, struct compound_hdr *hdr) WRITE32(hdr->taglen); WRITEMEM(hdr->tag, hdr->taglen); WRITE32(NFS4_MINOR_VERSION); + hdr->nops_p = p; WRITE32(hdr->nops); - return 0; +} + +static void encode_nops(struct compound_hdr *hdr) +{ + *hdr->nops_p = htonl(hdr->nops); } static void encode_nfs4_verifier(struct xdr_stream *xdr, const nfs4_verifier *verf) @@ -601,7 +607,7 @@ static void encode_nfs4_verifier(struct xdr_stream *xdr, const nfs4_verifier *ve xdr_encode_opaque_fixed(p, verf->data, NFS4_VERIFIER_SIZE); } -static int encode_attrs(struct xdr_stream *xdr, const struct iattr *iap, const struct nfs_server *server) +static void encode_attrs(struct xdr_stream *xdr, const struct iattr *iap, const struct nfs_server *server) { char owner_name[IDMAP_NAMESZ]; char owner_group[IDMAP_NAMESZ]; @@ -612,7 +618,6 @@ static int encode_attrs(struct xdr_stream *xdr, const struct iattr *iap, const s int len; uint32_t bmval0 = 0; uint32_t bmval1 = 0; - int status; /* * We reserve enough space to write the entire attribute buffer at once. @@ -709,7 +714,7 @@ static int encode_attrs(struct xdr_stream *xdr, const struct iattr *iap, const s bmval1 |= FATTR4_WORD1_TIME_MODIFY_SET; WRITE32(NFS4_SET_TO_SERVER_TIME); } - + /* * Now we backfill the bitmap and the attribute buffer length. */ @@ -723,23 +728,20 @@ static int encode_attrs(struct xdr_stream *xdr, const struct iattr *iap, const s *q++ = htonl(bmval1); *q++ = htonl(len); - status = 0; /* out: */ - return status; } -static int encode_access(struct xdr_stream *xdr, u32 access) +static void encode_access(struct xdr_stream *xdr, u32 access, struct compound_hdr *hdr) { __be32 *p; RESERVE_SPACE(8); WRITE32(OP_ACCESS); WRITE32(access); - - return 0; + hdr->nops++; } -static int encode_close(struct xdr_stream *xdr, const struct nfs_closeargs *arg) +static void encode_close(struct xdr_stream *xdr, const struct nfs_closeargs *arg, struct compound_hdr *hdr) { __be32 *p; @@ -747,26 +749,24 @@ static int encode_close(struct xdr_stream *xdr, const struct nfs_closeargs *arg) WRITE32(OP_CLOSE); WRITE32(arg->seqid->sequence->counter); WRITEMEM(arg->stateid->data, NFS4_STATEID_SIZE); - - return 0; + hdr->nops++; } -static int encode_commit(struct xdr_stream *xdr, const struct nfs_writeargs *args) +static void encode_commit(struct xdr_stream *xdr, const struct nfs_writeargs *args, struct compound_hdr *hdr) { __be32 *p; - - RESERVE_SPACE(16); - WRITE32(OP_COMMIT); - WRITE64(args->offset); - WRITE32(args->count); - return 0; + RESERVE_SPACE(16); + WRITE32(OP_COMMIT); + WRITE64(args->offset); + WRITE32(args->count); + hdr->nops++; } -static int encode_create(struct xdr_stream *xdr, const struct nfs4_create_arg *create) +static void encode_create(struct xdr_stream *xdr, const struct nfs4_create_arg *create, struct compound_hdr *hdr) { __be32 *p; - + RESERVE_SPACE(8); WRITE32(OP_CREATE); WRITE32(create->ftype); @@ -791,64 +791,62 @@ static int encode_create(struct xdr_stream *xdr, const struct nfs4_create_arg *c RESERVE_SPACE(4 + create->name->len); WRITE32(create->name->len); WRITEMEM(create->name->name, create->name->len); + hdr->nops++; - return encode_attrs(xdr, create->attrs, create->server); + encode_attrs(xdr, create->attrs, create->server); } -static int encode_getattr_one(struct xdr_stream *xdr, uint32_t bitmap) +static void encode_getattr_one(struct xdr_stream *xdr, uint32_t bitmap, struct compound_hdr *hdr) { - __be32 *p; + __be32 *p; - RESERVE_SPACE(12); - WRITE32(OP_GETATTR); - WRITE32(1); - WRITE32(bitmap); - return 0; + RESERVE_SPACE(12); + WRITE32(OP_GETATTR); + WRITE32(1); + WRITE32(bitmap); + hdr->nops++; } -static int encode_getattr_two(struct xdr_stream *xdr, uint32_t bm0, uint32_t bm1) +static void encode_getattr_two(struct xdr_stream *xdr, uint32_t bm0, uint32_t bm1, struct compound_hdr *hdr) { - __be32 *p; + __be32 *p; - RESERVE_SPACE(16); - WRITE32(OP_GETATTR); - WRITE32(2); - WRITE32(bm0); - WRITE32(bm1); - return 0; + RESERVE_SPACE(16); + WRITE32(OP_GETATTR); + WRITE32(2); + WRITE32(bm0); + WRITE32(bm1); + hdr->nops++; } -static int encode_getfattr(struct xdr_stream *xdr, const u32* bitmask) +static void encode_getfattr(struct xdr_stream *xdr, const u32* bitmask, struct compound_hdr *hdr) { - return encode_getattr_two(xdr, - bitmask[0] & nfs4_fattr_bitmap[0], - bitmask[1] & nfs4_fattr_bitmap[1]); + encode_getattr_two(xdr, bitmask[0] & nfs4_fattr_bitmap[0], + bitmask[1] & nfs4_fattr_bitmap[1], hdr); } -static int encode_fsinfo(struct xdr_stream *xdr, const u32* bitmask) +static void encode_fsinfo(struct xdr_stream *xdr, const u32* bitmask, struct compound_hdr *hdr) { - return encode_getattr_two(xdr, bitmask[0] & nfs4_fsinfo_bitmap[0], - bitmask[1] & nfs4_fsinfo_bitmap[1]); + encode_getattr_two(xdr, bitmask[0] & nfs4_fsinfo_bitmap[0], + bitmask[1] & nfs4_fsinfo_bitmap[1], hdr); } -static int encode_fs_locations(struct xdr_stream *xdr, const u32* bitmask) +static void encode_fs_locations(struct xdr_stream *xdr, const u32* bitmask, struct compound_hdr *hdr) { - return encode_getattr_two(xdr, - bitmask[0] & nfs4_fs_locations_bitmap[0], - bitmask[1] & nfs4_fs_locations_bitmap[1]); + encode_getattr_two(xdr, bitmask[0] & nfs4_fs_locations_bitmap[0], + bitmask[1] & nfs4_fs_locations_bitmap[1], hdr); } -static int encode_getfh(struct xdr_stream *xdr) +static void encode_getfh(struct xdr_stream *xdr, struct compound_hdr *hdr) { __be32 *p; RESERVE_SPACE(4); WRITE32(OP_GETFH); - - return 0; + hdr->nops++; } -static int encode_link(struct xdr_stream *xdr, const struct qstr *name) +static void encode_link(struct xdr_stream *xdr, const struct qstr *name, struct compound_hdr *hdr) { __be32 *p; @@ -856,8 +854,7 @@ static int encode_link(struct xdr_stream *xdr, const struct qstr *name) WRITE32(OP_LINK); WRITE32(name->len); WRITEMEM(name->name, name->len); - - return 0; + hdr->nops++; } static inline int nfs4_lock_type(struct file_lock *fl, int block) @@ -878,7 +875,7 @@ static inline uint64_t nfs4_lock_length(struct file_lock *fl) * opcode,type,reclaim,offset,length,new_lock_owner = 32 * open_seqid,open_stateid,lock_seqid,lock_owner.clientid, lock_owner.id = 40 */ -static int encode_lock(struct xdr_stream *xdr, const struct nfs_lock_args *args) +static void encode_lock(struct xdr_stream *xdr, const struct nfs_lock_args *args, struct compound_hdr *hdr) { __be32 *p; @@ -904,11 +901,10 @@ static int encode_lock(struct xdr_stream *xdr, const struct nfs_lock_args *args) WRITEMEM(args->lock_stateid->data, NFS4_STATEID_SIZE); WRITE32(args->lock_seqid->sequence->counter); } - - return 0; + hdr->nops++; } -static int encode_lockt(struct xdr_stream *xdr, const struct nfs_lockt_args *args) +static void encode_lockt(struct xdr_stream *xdr, const struct nfs_lockt_args *args, struct compound_hdr *hdr) { __be32 *p; @@ -921,11 +917,10 @@ static int encode_lockt(struct xdr_stream *xdr, const struct nfs_lockt_args *arg WRITE32(16); WRITEMEM("lock id:", 8); WRITE64(args->lock_owner.id); - - return 0; + hdr->nops++; } -static int encode_locku(struct xdr_stream *xdr, const struct nfs_locku_args *args) +static void encode_locku(struct xdr_stream *xdr, const struct nfs_locku_args *args, struct compound_hdr *hdr) { __be32 *p; @@ -936,11 +931,10 @@ static int encode_locku(struct xdr_stream *xdr, const struct nfs_locku_args *arg WRITEMEM(args->stateid->data, NFS4_STATEID_SIZE); WRITE64(args->fl->fl_start); WRITE64(nfs4_lock_length(args->fl)); - - return 0; + hdr->nops++; } -static int encode_lookup(struct xdr_stream *xdr, const struct qstr *name) +static void encode_lookup(struct xdr_stream *xdr, const struct qstr *name, struct compound_hdr *hdr) { int len = name->len; __be32 *p; @@ -949,27 +943,26 @@ static int encode_lookup(struct xdr_stream *xdr, const struct qstr *name) WRITE32(OP_LOOKUP); WRITE32(len); WRITEMEM(name->name, len); - - return 0; + hdr->nops++; } -static void encode_share_access(struct xdr_stream *xdr, int open_flags) +static void encode_share_access(struct xdr_stream *xdr, fmode_t fmode) { __be32 *p; RESERVE_SPACE(8); - switch (open_flags & (FMODE_READ|FMODE_WRITE)) { - case FMODE_READ: - WRITE32(NFS4_SHARE_ACCESS_READ); - break; - case FMODE_WRITE: - WRITE32(NFS4_SHARE_ACCESS_WRITE); - break; - case FMODE_READ|FMODE_WRITE: - WRITE32(NFS4_SHARE_ACCESS_BOTH); - break; - default: - BUG(); + switch (fmode & (FMODE_READ|FMODE_WRITE)) { + case FMODE_READ: + WRITE32(NFS4_SHARE_ACCESS_READ); + break; + case FMODE_WRITE: + WRITE32(NFS4_SHARE_ACCESS_WRITE); + break; + case FMODE_READ|FMODE_WRITE: + WRITE32(NFS4_SHARE_ACCESS_BOTH); + break; + default: + WRITE32(0); } WRITE32(0); /* for linux, share_deny = 0 always */ } @@ -984,7 +977,7 @@ static inline void encode_openhdr(struct xdr_stream *xdr, const struct nfs_opena RESERVE_SPACE(8); WRITE32(OP_OPEN); WRITE32(arg->seqid->sequence->counter); - encode_share_access(xdr, arg->open_flags); + encode_share_access(xdr, arg->fmode); RESERVE_SPACE(28); WRITE64(arg->clientid); WRITE32(16); @@ -998,13 +991,13 @@ static inline void encode_createmode(struct xdr_stream *xdr, const struct nfs_op RESERVE_SPACE(4); switch(arg->open_flags & O_EXCL) { - case 0: - WRITE32(NFS4_CREATE_UNCHECKED); - encode_attrs(xdr, arg->u.attrs, arg->server); - break; - default: - WRITE32(NFS4_CREATE_EXCLUSIVE); - encode_nfs4_verifier(xdr, &arg->u.verifier); + case 0: + WRITE32(NFS4_CREATE_UNCHECKED); + encode_attrs(xdr, arg->u.attrs, arg->server); + break; + default: + WRITE32(NFS4_CREATE_EXCLUSIVE); + encode_nfs4_verifier(xdr, &arg->u.verifier); } } @@ -1014,33 +1007,33 @@ static void encode_opentype(struct xdr_stream *xdr, const struct nfs_openargs *a RESERVE_SPACE(4); switch (arg->open_flags & O_CREAT) { - case 0: - WRITE32(NFS4_OPEN_NOCREATE); - break; - default: - BUG_ON(arg->claim != NFS4_OPEN_CLAIM_NULL); - WRITE32(NFS4_OPEN_CREATE); - encode_createmode(xdr, arg); + case 0: + WRITE32(NFS4_OPEN_NOCREATE); + break; + default: + BUG_ON(arg->claim != NFS4_OPEN_CLAIM_NULL); + WRITE32(NFS4_OPEN_CREATE); + encode_createmode(xdr, arg); } } -static inline void encode_delegation_type(struct xdr_stream *xdr, int delegation_type) +static inline void encode_delegation_type(struct xdr_stream *xdr, fmode_t delegation_type) { __be32 *p; RESERVE_SPACE(4); switch (delegation_type) { - case 0: - WRITE32(NFS4_OPEN_DELEGATE_NONE); - break; - case FMODE_READ: - WRITE32(NFS4_OPEN_DELEGATE_READ); - break; - case FMODE_WRITE|FMODE_READ: - WRITE32(NFS4_OPEN_DELEGATE_WRITE); - break; - default: - BUG(); + case 0: + WRITE32(NFS4_OPEN_DELEGATE_NONE); + break; + case FMODE_READ: + WRITE32(NFS4_OPEN_DELEGATE_READ); + break; + case FMODE_WRITE|FMODE_READ: + WRITE32(NFS4_OPEN_DELEGATE_WRITE); + break; + default: + BUG(); } } @@ -1053,7 +1046,7 @@ static inline void encode_claim_null(struct xdr_stream *xdr, const struct qstr * encode_string(xdr, name->len, name->name); } -static inline void encode_claim_previous(struct xdr_stream *xdr, int type) +static inline void encode_claim_previous(struct xdr_stream *xdr, fmode_t type) { __be32 *p; @@ -1072,27 +1065,27 @@ static inline void encode_claim_delegate_cur(struct xdr_stream *xdr, const struc encode_string(xdr, name->len, name->name); } -static int encode_open(struct xdr_stream *xdr, const struct nfs_openargs *arg) +static void encode_open(struct xdr_stream *xdr, const struct nfs_openargs *arg, struct compound_hdr *hdr) { encode_openhdr(xdr, arg); encode_opentype(xdr, arg); switch (arg->claim) { - case NFS4_OPEN_CLAIM_NULL: - encode_claim_null(xdr, arg->name); - break; - case NFS4_OPEN_CLAIM_PREVIOUS: - encode_claim_previous(xdr, arg->u.delegation_type); - break; - case NFS4_OPEN_CLAIM_DELEGATE_CUR: - encode_claim_delegate_cur(xdr, arg->name, &arg->u.delegation); - break; - default: - BUG(); + case NFS4_OPEN_CLAIM_NULL: + encode_claim_null(xdr, arg->name); + break; + case NFS4_OPEN_CLAIM_PREVIOUS: + encode_claim_previous(xdr, arg->u.delegation_type); + break; + case NFS4_OPEN_CLAIM_DELEGATE_CUR: + encode_claim_delegate_cur(xdr, arg->name, &arg->u.delegation); + break; + default: + BUG(); } - return 0; + hdr->nops++; } -static int encode_open_confirm(struct xdr_stream *xdr, const struct nfs_open_confirmargs *arg) +static void encode_open_confirm(struct xdr_stream *xdr, const struct nfs_open_confirmargs *arg, struct compound_hdr *hdr) { __be32 *p; @@ -1100,11 +1093,10 @@ static int encode_open_confirm(struct xdr_stream *xdr, const struct nfs_open_con WRITE32(OP_OPEN_CONFIRM); WRITEMEM(arg->stateid->data, NFS4_STATEID_SIZE); WRITE32(arg->seqid->sequence->counter); - - return 0; + hdr->nops++; } -static int encode_open_downgrade(struct xdr_stream *xdr, const struct nfs_closeargs *arg) +static void encode_open_downgrade(struct xdr_stream *xdr, const struct nfs_closeargs *arg, struct compound_hdr *hdr) { __be32 *p; @@ -1112,12 +1104,12 @@ static int encode_open_downgrade(struct xdr_stream *xdr, const struct nfs_closea WRITE32(OP_OPEN_DOWNGRADE); WRITEMEM(arg->stateid->data, NFS4_STATEID_SIZE); WRITE32(arg->seqid->sequence->counter); - encode_share_access(xdr, arg->open_flags); - return 0; + encode_share_access(xdr, arg->fmode); + hdr->nops++; } -static int -encode_putfh(struct xdr_stream *xdr, const struct nfs_fh *fh) +static void +encode_putfh(struct xdr_stream *xdr, const struct nfs_fh *fh, struct compound_hdr *hdr) { int len = fh->size; __be32 *p; @@ -1126,18 +1118,16 @@ encode_putfh(struct xdr_stream *xdr, const struct nfs_fh *fh) WRITE32(OP_PUTFH); WRITE32(len); WRITEMEM(fh->data, len); - - return 0; + hdr->nops++; } -static int encode_putrootfh(struct xdr_stream *xdr) +static void encode_putrootfh(struct xdr_stream *xdr, struct compound_hdr *hdr) { - __be32 *p; - - RESERVE_SPACE(4); - WRITE32(OP_PUTROOTFH); + __be32 *p; - return 0; + RESERVE_SPACE(4); + WRITE32(OP_PUTROOTFH); + hdr->nops++; } static void encode_stateid(struct xdr_stream *xdr, const struct nfs_open_context *ctx) @@ -1153,7 +1143,7 @@ static void encode_stateid(struct xdr_stream *xdr, const struct nfs_open_context WRITEMEM(zero_stateid.data, NFS4_STATEID_SIZE); } -static int encode_read(struct xdr_stream *xdr, const struct nfs_readargs *args) +static void encode_read(struct xdr_stream *xdr, const struct nfs_readargs *args, struct compound_hdr *hdr) { __be32 *p; @@ -1165,11 +1155,10 @@ static int encode_read(struct xdr_stream *xdr, const struct nfs_readargs *args) RESERVE_SPACE(12); WRITE64(args->offset); WRITE32(args->count); - - return 0; + hdr->nops++; } -static int encode_readdir(struct xdr_stream *xdr, const struct nfs4_readdir_arg *readdir, struct rpc_rqst *req) +static void encode_readdir(struct xdr_stream *xdr, const struct nfs4_readdir_arg *readdir, struct rpc_rqst *req, struct compound_hdr *hdr) { uint32_t attrs[2] = { FATTR4_WORD0_RDATTR_ERROR|FATTR4_WORD0_FILEID, @@ -1191,6 +1180,7 @@ static int encode_readdir(struct xdr_stream *xdr, const struct nfs4_readdir_arg attrs[1] &= ~FATTR4_WORD1_MOUNTED_ON_FILEID; WRITE32(attrs[0] & readdir->bitmask[0]); WRITE32(attrs[1] & readdir->bitmask[1]); + hdr->nops++; dprintk("%s: cookie = %Lu, verifier = %08x:%08x, bitmap = %08x:%08x\n", __func__, (unsigned long long)readdir->cookie, @@ -1198,21 +1188,18 @@ static int encode_readdir(struct xdr_stream *xdr, const struct nfs4_readdir_arg ((u32 *)readdir->verifier.data)[1], attrs[0] & readdir->bitmask[0], attrs[1] & readdir->bitmask[1]); - - return 0; } -static int encode_readlink(struct xdr_stream *xdr, const struct nfs4_readlink *readlink, struct rpc_rqst *req) +static void encode_readlink(struct xdr_stream *xdr, const struct nfs4_readlink *readlink, struct rpc_rqst *req, struct compound_hdr *hdr) { __be32 *p; RESERVE_SPACE(4); WRITE32(OP_READLINK); - - return 0; + hdr->nops++; } -static int encode_remove(struct xdr_stream *xdr, const struct qstr *name) +static void encode_remove(struct xdr_stream *xdr, const struct qstr *name, struct compound_hdr *hdr) { __be32 *p; @@ -1220,11 +1207,10 @@ static int encode_remove(struct xdr_stream *xdr, const struct qstr *name) WRITE32(OP_REMOVE); WRITE32(name->len); WRITEMEM(name->name, name->len); - - return 0; + hdr->nops++; } -static int encode_rename(struct xdr_stream *xdr, const struct qstr *oldname, const struct qstr *newname) +static void encode_rename(struct xdr_stream *xdr, const struct qstr *oldname, const struct qstr *newname, struct compound_hdr *hdr) { __be32 *p; @@ -1232,38 +1218,35 @@ static int encode_rename(struct xdr_stream *xdr, const struct qstr *oldname, con WRITE32(OP_RENAME); WRITE32(oldname->len); WRITEMEM(oldname->name, oldname->len); - + RESERVE_SPACE(4 + newname->len); WRITE32(newname->len); WRITEMEM(newname->name, newname->len); - - return 0; + hdr->nops++; } -static int encode_renew(struct xdr_stream *xdr, const struct nfs_client *client_stateid) +static void encode_renew(struct xdr_stream *xdr, const struct nfs_client *client_stateid, struct compound_hdr *hdr) { __be32 *p; RESERVE_SPACE(12); WRITE32(OP_RENEW); WRITE64(client_stateid->cl_clientid); - - return 0; + hdr->nops++; } -static int -encode_restorefh(struct xdr_stream *xdr) +static void +encode_restorefh(struct xdr_stream *xdr, struct compound_hdr *hdr) { __be32 *p; RESERVE_SPACE(4); WRITE32(OP_RESTOREFH); - - return 0; + hdr->nops++; } static int -encode_setacl(struct xdr_stream *xdr, struct nfs_setaclargs *arg) +encode_setacl(struct xdr_stream *xdr, struct nfs_setaclargs *arg, struct compound_hdr *hdr) { __be32 *p; @@ -1278,36 +1261,32 @@ encode_setacl(struct xdr_stream *xdr, struct nfs_setaclargs *arg) RESERVE_SPACE(4); WRITE32(arg->acl_len); xdr_write_pages(xdr, arg->acl_pages, arg->acl_pgbase, arg->acl_len); + hdr->nops++; return 0; } -static int -encode_savefh(struct xdr_stream *xdr) +static void +encode_savefh(struct xdr_stream *xdr, struct compound_hdr *hdr) { __be32 *p; RESERVE_SPACE(4); WRITE32(OP_SAVEFH); - - return 0; + hdr->nops++; } -static int encode_setattr(struct xdr_stream *xdr, const struct nfs_setattrargs *arg, const struct nfs_server *server) +static void encode_setattr(struct xdr_stream *xdr, const struct nfs_setattrargs *arg, const struct nfs_server *server, struct compound_hdr *hdr) { - int status; __be32 *p; - - RESERVE_SPACE(4+NFS4_STATEID_SIZE); - WRITE32(OP_SETATTR); - WRITEMEM(arg->stateid.data, NFS4_STATEID_SIZE); - if ((status = encode_attrs(xdr, arg->iap, server))) - return status; - - return 0; + RESERVE_SPACE(4+NFS4_STATEID_SIZE); + WRITE32(OP_SETATTR); + WRITEMEM(arg->stateid.data, NFS4_STATEID_SIZE); + hdr->nops++; + encode_attrs(xdr, arg->iap, server); } -static int encode_setclientid(struct xdr_stream *xdr, const struct nfs4_setclientid *setclientid) +static void encode_setclientid(struct xdr_stream *xdr, const struct nfs4_setclientid *setclientid, struct compound_hdr *hdr) { __be32 *p; @@ -1322,23 +1301,21 @@ static int encode_setclientid(struct xdr_stream *xdr, const struct nfs4_setclien encode_string(xdr, setclientid->sc_uaddr_len, setclientid->sc_uaddr); RESERVE_SPACE(4); WRITE32(setclientid->sc_cb_ident); - - return 0; + hdr->nops++; } -static int encode_setclientid_confirm(struct xdr_stream *xdr, const struct nfs_client *client_state) +static void encode_setclientid_confirm(struct xdr_stream *xdr, const struct nfs_client *client_state, struct compound_hdr *hdr) { - __be32 *p; - - RESERVE_SPACE(12 + NFS4_VERIFIER_SIZE); - WRITE32(OP_SETCLIENTID_CONFIRM); - WRITE64(client_state->cl_clientid); - WRITEMEM(client_state->cl_confirm.data, NFS4_VERIFIER_SIZE); + __be32 *p; - return 0; + RESERVE_SPACE(12 + NFS4_VERIFIER_SIZE); + WRITE32(OP_SETCLIENTID_CONFIRM); + WRITE64(client_state->cl_clientid); + WRITEMEM(client_state->cl_confirm.data, NFS4_VERIFIER_SIZE); + hdr->nops++; } -static int encode_write(struct xdr_stream *xdr, const struct nfs_writeargs *args) +static void encode_write(struct xdr_stream *xdr, const struct nfs_writeargs *args, struct compound_hdr *hdr) { __be32 *p; @@ -1353,11 +1330,10 @@ static int encode_write(struct xdr_stream *xdr, const struct nfs_writeargs *args WRITE32(args->count); xdr_write_pages(xdr, args->pages, args->pgbase, args->count); - - return 0; + hdr->nops++; } -static int encode_delegreturn(struct xdr_stream *xdr, const nfs4_stateid *stateid) +static void encode_delegreturn(struct xdr_stream *xdr, const nfs4_stateid *stateid, struct compound_hdr *hdr) { __be32 *p; @@ -1365,8 +1341,7 @@ static int encode_delegreturn(struct xdr_stream *xdr, const nfs4_stateid *statei WRITE32(OP_DELEGRETURN); WRITEMEM(stateid->data, NFS4_STATEID_SIZE); - return 0; - + hdr->nops++; } /* * END OF "GENERIC" ENCODE ROUTINES. @@ -1379,21 +1354,16 @@ static int nfs4_xdr_enc_access(struct rpc_rqst *req, __be32 *p, const struct nfs { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 3, + .nops = 0, }; - int status; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); - status = encode_putfh(&xdr, args->fh); - if (status != 0) - goto out; - status = encode_access(&xdr, args->access); - if (status != 0) - goto out; - status = encode_getfattr(&xdr, args->bitmask); -out: - return status; + encode_putfh(&xdr, args->fh, &hdr); + encode_access(&xdr, args->access, &hdr); + encode_getfattr(&xdr, args->bitmask, &hdr); + encode_nops(&hdr); + return 0; } /* @@ -1403,21 +1373,17 @@ static int nfs4_xdr_enc_lookup(struct rpc_rqst *req, __be32 *p, const struct nfs { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 4, + .nops = 0, }; - int status; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); - if ((status = encode_putfh(&xdr, args->dir_fh)) != 0) - goto out; - if ((status = encode_lookup(&xdr, args->name)) != 0) - goto out; - if ((status = encode_getfh(&xdr)) != 0) - goto out; - status = encode_getfattr(&xdr, args->bitmask); -out: - return status; + encode_putfh(&xdr, args->dir_fh, &hdr); + encode_lookup(&xdr, args->name, &hdr); + encode_getfh(&xdr, &hdr); + encode_getfattr(&xdr, args->bitmask, &hdr); + encode_nops(&hdr); + return 0; } /* @@ -1427,18 +1393,16 @@ static int nfs4_xdr_enc_lookup_root(struct rpc_rqst *req, __be32 *p, const struc { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 3, + .nops = 0, }; - int status; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); - if ((status = encode_putrootfh(&xdr)) != 0) - goto out; - if ((status = encode_getfh(&xdr)) == 0) - status = encode_getfattr(&xdr, args->bitmask); -out: - return status; + encode_putrootfh(&xdr, &hdr); + encode_getfh(&xdr, &hdr); + encode_getfattr(&xdr, args->bitmask, &hdr); + encode_nops(&hdr); + return 0; } /* @@ -1448,19 +1412,16 @@ static int nfs4_xdr_enc_remove(struct rpc_rqst *req, __be32 *p, const struct nfs { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 3, + .nops = 0, }; - int status; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); - if ((status = encode_putfh(&xdr, args->fh)) != 0) - goto out; - if ((status = encode_remove(&xdr, &args->name)) != 0) - goto out; - status = encode_getfattr(&xdr, args->bitmask); -out: - return status; + encode_putfh(&xdr, args->fh, &hdr); + encode_remove(&xdr, &args->name, &hdr); + encode_getfattr(&xdr, args->bitmask, &hdr); + encode_nops(&hdr); + return 0; } /* @@ -1470,27 +1431,20 @@ static int nfs4_xdr_enc_rename(struct rpc_rqst *req, __be32 *p, const struct nfs { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 7, + .nops = 0, }; - int status; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); - if ((status = encode_putfh(&xdr, args->old_dir)) != 0) - goto out; - if ((status = encode_savefh(&xdr)) != 0) - goto out; - if ((status = encode_putfh(&xdr, args->new_dir)) != 0) - goto out; - if ((status = encode_rename(&xdr, args->old_name, args->new_name)) != 0) - goto out; - if ((status = encode_getfattr(&xdr, args->bitmask)) != 0) - goto out; - if ((status = encode_restorefh(&xdr)) != 0) - goto out; - status = encode_getfattr(&xdr, args->bitmask); -out: - return status; + encode_putfh(&xdr, args->old_dir, &hdr); + encode_savefh(&xdr, &hdr); + encode_putfh(&xdr, args->new_dir, &hdr); + encode_rename(&xdr, args->old_name, args->new_name, &hdr); + encode_getfattr(&xdr, args->bitmask, &hdr); + encode_restorefh(&xdr, &hdr); + encode_getfattr(&xdr, args->bitmask, &hdr); + encode_nops(&hdr); + return 0; } /* @@ -1500,27 +1454,20 @@ static int nfs4_xdr_enc_link(struct rpc_rqst *req, __be32 *p, const struct nfs4_ { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 7, + .nops = 0, }; - int status; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); - if ((status = encode_putfh(&xdr, args->fh)) != 0) - goto out; - if ((status = encode_savefh(&xdr)) != 0) - goto out; - if ((status = encode_putfh(&xdr, args->dir_fh)) != 0) - goto out; - if ((status = encode_link(&xdr, args->name)) != 0) - goto out; - if ((status = encode_getfattr(&xdr, args->bitmask)) != 0) - goto out; - if ((status = encode_restorefh(&xdr)) != 0) - goto out; - status = encode_getfattr(&xdr, args->bitmask); -out: - return status; + encode_putfh(&xdr, args->fh, &hdr); + encode_savefh(&xdr, &hdr); + encode_putfh(&xdr, args->dir_fh, &hdr); + encode_link(&xdr, args->name, &hdr); + encode_getfattr(&xdr, args->bitmask, &hdr); + encode_restorefh(&xdr, &hdr); + encode_getfattr(&xdr, args->bitmask, &hdr); + encode_nops(&hdr); + return 0; } /* @@ -1530,27 +1477,20 @@ static int nfs4_xdr_enc_create(struct rpc_rqst *req, __be32 *p, const struct nfs { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 7, + .nops = 0, }; - int status; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); - if ((status = encode_putfh(&xdr, args->dir_fh)) != 0) - goto out; - if ((status = encode_savefh(&xdr)) != 0) - goto out; - if ((status = encode_create(&xdr, args)) != 0) - goto out; - if ((status = encode_getfh(&xdr)) != 0) - goto out; - if ((status = encode_getfattr(&xdr, args->bitmask)) != 0) - goto out; - if ((status = encode_restorefh(&xdr)) != 0) - goto out; - status = encode_getfattr(&xdr, args->bitmask); -out: - return status; + encode_putfh(&xdr, args->dir_fh, &hdr); + encode_savefh(&xdr, &hdr); + encode_create(&xdr, args, &hdr); + encode_getfh(&xdr, &hdr); + encode_getfattr(&xdr, args->bitmask, &hdr); + encode_restorefh(&xdr, &hdr); + encode_getfattr(&xdr, args->bitmask, &hdr); + encode_nops(&hdr); + return 0; } /* @@ -1568,15 +1508,15 @@ static int nfs4_xdr_enc_getattr(struct rpc_rqst *req, __be32 *p, const struct nf { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 2, + .nops = 0, }; - int status; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); - if ((status = encode_putfh(&xdr, args->fh)) == 0) - status = encode_getfattr(&xdr, args->bitmask); - return status; + encode_putfh(&xdr, args->fh, &hdr); + encode_getfattr(&xdr, args->bitmask, &hdr); + encode_nops(&hdr); + return 0; } /* @@ -1584,23 +1524,18 @@ static int nfs4_xdr_enc_getattr(struct rpc_rqst *req, __be32 *p, const struct nf */ static int nfs4_xdr_enc_close(struct rpc_rqst *req, __be32 *p, struct nfs_closeargs *args) { - struct xdr_stream xdr; - struct compound_hdr hdr = { - .nops = 3, - }; - int status; - - xdr_init_encode(&xdr, &req->rq_snd_buf, p); - encode_compound_hdr(&xdr, &hdr); - status = encode_putfh(&xdr, args->fh); - if(status) - goto out; - status = encode_close(&xdr, args); - if (status != 0) - goto out; - status = encode_getfattr(&xdr, args->bitmask); -out: - return status; + struct xdr_stream xdr; + struct compound_hdr hdr = { + .nops = 0, + }; + + xdr_init_encode(&xdr, &req->rq_snd_buf, p); + encode_compound_hdr(&xdr, &hdr); + encode_putfh(&xdr, args->fh, &hdr); + encode_close(&xdr, args, &hdr); + encode_getfattr(&xdr, args->bitmask, &hdr); + encode_nops(&hdr); + return 0; } /* @@ -1610,33 +1545,20 @@ static int nfs4_xdr_enc_open(struct rpc_rqst *req, __be32 *p, struct nfs_openarg { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 7, + .nops = 0, }; - int status; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); - status = encode_putfh(&xdr, args->fh); - if (status) - goto out; - status = encode_savefh(&xdr); - if (status) - goto out; - status = encode_open(&xdr, args); - if (status) - goto out; - status = encode_getfh(&xdr); - if (status) - goto out; - status = encode_getfattr(&xdr, args->bitmask); - if (status) - goto out; - status = encode_restorefh(&xdr); - if (status) - goto out; - status = encode_getfattr(&xdr, args->bitmask); -out: - return status; + encode_putfh(&xdr, args->fh, &hdr); + encode_savefh(&xdr, &hdr); + encode_open(&xdr, args, &hdr); + encode_getfh(&xdr, &hdr); + encode_getfattr(&xdr, args->bitmask, &hdr); + encode_restorefh(&xdr, &hdr); + encode_getfattr(&xdr, args->bitmask, &hdr); + encode_nops(&hdr); + return 0; } /* @@ -1646,18 +1568,15 @@ static int nfs4_xdr_enc_open_confirm(struct rpc_rqst *req, __be32 *p, struct nfs { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 2, + .nops = 0, }; - int status; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); - status = encode_putfh(&xdr, args->fh); - if(status) - goto out; - status = encode_open_confirm(&xdr, args); -out: - return status; + encode_putfh(&xdr, args->fh, &hdr); + encode_open_confirm(&xdr, args, &hdr); + encode_nops(&hdr); + return 0; } /* @@ -1667,21 +1586,16 @@ static int nfs4_xdr_enc_open_noattr(struct rpc_rqst *req, __be32 *p, struct nfs_ { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 3, + .nops = 0, }; - int status; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); - status = encode_putfh(&xdr, args->fh); - if (status) - goto out; - status = encode_open(&xdr, args); - if (status) - goto out; - status = encode_getfattr(&xdr, args->bitmask); -out: - return status; + encode_putfh(&xdr, args->fh, &hdr); + encode_open(&xdr, args, &hdr); + encode_getfattr(&xdr, args->bitmask, &hdr); + encode_nops(&hdr); + return 0; } /* @@ -1691,21 +1605,16 @@ static int nfs4_xdr_enc_open_downgrade(struct rpc_rqst *req, __be32 *p, struct n { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 3, + .nops = 0, }; - int status; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); - status = encode_putfh(&xdr, args->fh); - if (status) - goto out; - status = encode_open_downgrade(&xdr, args); - if (status != 0) - goto out; - status = encode_getfattr(&xdr, args->bitmask); -out: - return status; + encode_putfh(&xdr, args->fh, &hdr); + encode_open_downgrade(&xdr, args, &hdr); + encode_getfattr(&xdr, args->bitmask, &hdr); + encode_nops(&hdr); + return 0; } /* @@ -1715,18 +1624,15 @@ static int nfs4_xdr_enc_lock(struct rpc_rqst *req, __be32 *p, struct nfs_lock_ar { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 2, + .nops = 0, }; - int status; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); - status = encode_putfh(&xdr, args->fh); - if(status) - goto out; - status = encode_lock(&xdr, args); -out: - return status; + encode_putfh(&xdr, args->fh, &hdr); + encode_lock(&xdr, args, &hdr); + encode_nops(&hdr); + return 0; } /* @@ -1736,18 +1642,15 @@ static int nfs4_xdr_enc_lockt(struct rpc_rqst *req, __be32 *p, struct nfs_lockt_ { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 2, + .nops = 0, }; - int status; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); - status = encode_putfh(&xdr, args->fh); - if(status) - goto out; - status = encode_lockt(&xdr, args); -out: - return status; + encode_putfh(&xdr, args->fh, &hdr); + encode_lockt(&xdr, args, &hdr); + encode_nops(&hdr); + return 0; } /* @@ -1757,18 +1660,15 @@ static int nfs4_xdr_enc_locku(struct rpc_rqst *req, __be32 *p, struct nfs_locku_ { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 2, + .nops = 0, }; - int status; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); - status = encode_putfh(&xdr, args->fh); - if(status) - goto out; - status = encode_locku(&xdr, args); -out: - return status; + encode_putfh(&xdr, args->fh, &hdr); + encode_locku(&xdr, args, &hdr); + encode_nops(&hdr); + return 0; } /* @@ -1778,18 +1678,15 @@ static int nfs4_xdr_enc_readlink(struct rpc_rqst *req, __be32 *p, const struct n { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 2, + .nops = 0, }; struct rpc_auth *auth = req->rq_task->tk_msg.rpc_cred->cr_auth; unsigned int replen; - int status; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); - status = encode_putfh(&xdr, args->fh); - if(status) - goto out; - status = encode_readlink(&xdr, args, req); + encode_putfh(&xdr, args->fh, &hdr); + encode_readlink(&xdr, args, req, &hdr); /* set up reply kvec * toplevel_status + taglen + rescount + OP_PUTFH + status @@ -1798,9 +1695,8 @@ static int nfs4_xdr_enc_readlink(struct rpc_rqst *req, __be32 *p, const struct n replen = (RPC_REPHDRSIZE + auth->au_rslack + NFS4_dec_readlink_sz) << 2; xdr_inline_pages(&req->rq_rcv_buf, replen, args->pages, args->pgbase, args->pglen); - -out: - return status; + encode_nops(&hdr); + return 0; } /* @@ -1810,18 +1706,15 @@ static int nfs4_xdr_enc_readdir(struct rpc_rqst *req, __be32 *p, const struct nf { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 2, + .nops = 0, }; struct rpc_auth *auth = req->rq_task->tk_msg.rpc_cred->cr_auth; int replen; - int status; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); - status = encode_putfh(&xdr, args->fh); - if(status) - goto out; - status = encode_readdir(&xdr, args, req); + encode_putfh(&xdr, args->fh, &hdr); + encode_readdir(&xdr, args, req, &hdr); /* set up reply kvec * toplevel_status + taglen + rescount + OP_PUTFH + status @@ -1833,9 +1726,8 @@ static int nfs4_xdr_enc_readdir(struct rpc_rqst *req, __be32 *p, const struct nf dprintk("%s: inlined page args = (%u, %p, %u, %u)\n", __func__, replen, args->pages, args->pgbase, args->count); - -out: - return status; + encode_nops(&hdr); + return 0; } /* @@ -1846,18 +1738,14 @@ static int nfs4_xdr_enc_read(struct rpc_rqst *req, __be32 *p, struct nfs_readarg struct rpc_auth *auth = req->rq_task->tk_msg.rpc_cred->cr_auth; struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 2, + .nops = 0, }; - int replen, status; + int replen; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); - status = encode_putfh(&xdr, args->fh); - if (status) - goto out; - status = encode_read(&xdr, args); - if (status) - goto out; + encode_putfh(&xdr, args->fh, &hdr); + encode_read(&xdr, args, &hdr); /* set up reply kvec * toplevel status + taglen=0 + rescount + OP_PUTFH + status @@ -1867,33 +1755,27 @@ static int nfs4_xdr_enc_read(struct rpc_rqst *req, __be32 *p, struct nfs_readarg xdr_inline_pages(&req->rq_rcv_buf, replen, args->pages, args->pgbase, args->count); req->rq_rcv_buf.flags |= XDRBUF_READ; -out: - return status; + encode_nops(&hdr); + return 0; } /* * Encode an SETATTR request */ static int nfs4_xdr_enc_setattr(struct rpc_rqst *req, __be32 *p, struct nfs_setattrargs *args) - { - struct xdr_stream xdr; - struct compound_hdr hdr = { - .nops = 3, - }; - int status; - - xdr_init_encode(&xdr, &req->rq_snd_buf, p); - encode_compound_hdr(&xdr, &hdr); - status = encode_putfh(&xdr, args->fh); - if(status) - goto out; - status = encode_setattr(&xdr, args, args->server); - if(status) - goto out; - status = encode_getfattr(&xdr, args->bitmask); -out: - return status; + struct xdr_stream xdr; + struct compound_hdr hdr = { + .nops = 0, + }; + + xdr_init_encode(&xdr, &req->rq_snd_buf, p); + encode_compound_hdr(&xdr, &hdr); + encode_putfh(&xdr, args->fh, &hdr); + encode_setattr(&xdr, args, args->server, &hdr); + encode_getfattr(&xdr, args->bitmask, &hdr); + encode_nops(&hdr); + return 0; } /* @@ -1906,22 +1788,21 @@ nfs4_xdr_enc_getacl(struct rpc_rqst *req, __be32 *p, struct xdr_stream xdr; struct rpc_auth *auth = req->rq_task->tk_msg.rpc_cred->cr_auth; struct compound_hdr hdr = { - .nops = 2, + .nops = 0, }; - int replen, status; + int replen; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); - status = encode_putfh(&xdr, args->fh); - if (status) - goto out; - status = encode_getattr_two(&xdr, FATTR4_WORD0_ACL, 0); + encode_putfh(&xdr, args->fh, &hdr); + encode_getattr_two(&xdr, FATTR4_WORD0_ACL, 0, &hdr); + /* set up reply buffer: */ replen = (RPC_REPHDRSIZE + auth->au_rslack + NFS4_dec_getacl_sz) << 2; xdr_inline_pages(&req->rq_rcv_buf, replen, args->acl_pages, args->acl_pgbase, args->acl_len); -out: - return status; + encode_nops(&hdr); + return 0; } /* @@ -1931,22 +1812,17 @@ static int nfs4_xdr_enc_write(struct rpc_rqst *req, __be32 *p, struct nfs_writea { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 3, + .nops = 0, }; - int status; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); - status = encode_putfh(&xdr, args->fh); - if (status) - goto out; - status = encode_write(&xdr, args); - if (status) - goto out; + encode_putfh(&xdr, args->fh, &hdr); + encode_write(&xdr, args, &hdr); req->rq_snd_buf.flags |= XDRBUF_WRITE; - status = encode_getfattr(&xdr, args->bitmask); -out: - return status; + encode_getfattr(&xdr, args->bitmask, &hdr); + encode_nops(&hdr); + return 0; } /* @@ -1956,21 +1832,16 @@ static int nfs4_xdr_enc_commit(struct rpc_rqst *req, __be32 *p, struct nfs_write { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 3, + .nops = 0, }; - int status; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); - status = encode_putfh(&xdr, args->fh); - if (status) - goto out; - status = encode_commit(&xdr, args); - if (status) - goto out; - status = encode_getfattr(&xdr, args->bitmask); -out: - return status; + encode_putfh(&xdr, args->fh, &hdr); + encode_commit(&xdr, args, &hdr); + encode_getfattr(&xdr, args->bitmask, &hdr); + encode_nops(&hdr); + return 0; } /* @@ -1980,16 +1851,15 @@ static int nfs4_xdr_enc_fsinfo(struct rpc_rqst *req, __be32 *p, struct nfs4_fsin { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 2, + .nops = 0, }; - int status; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); - status = encode_putfh(&xdr, args->fh); - if (!status) - status = encode_fsinfo(&xdr, args->bitmask); - return status; + encode_putfh(&xdr, args->fh, &hdr); + encode_fsinfo(&xdr, args->bitmask, &hdr); + encode_nops(&hdr); + return 0; } /* @@ -1999,17 +1869,16 @@ static int nfs4_xdr_enc_pathconf(struct rpc_rqst *req, __be32 *p, const struct n { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 2, + .nops = 0, }; - int status; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); - status = encode_putfh(&xdr, args->fh); - if (!status) - status = encode_getattr_one(&xdr, - args->bitmask[0] & nfs4_pathconf_bitmap[0]); - return status; + encode_putfh(&xdr, args->fh, &hdr); + encode_getattr_one(&xdr, args->bitmask[0] & nfs4_pathconf_bitmap[0], + &hdr); + encode_nops(&hdr); + return 0; } /* @@ -2019,18 +1888,16 @@ static int nfs4_xdr_enc_statfs(struct rpc_rqst *req, __be32 *p, const struct nfs { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 2, + .nops = 0, }; - int status; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); - status = encode_putfh(&xdr, args->fh); - if (status == 0) - status = encode_getattr_two(&xdr, - args->bitmask[0] & nfs4_statfs_bitmap[0], - args->bitmask[1] & nfs4_statfs_bitmap[1]); - return status; + encode_putfh(&xdr, args->fh, &hdr); + encode_getattr_two(&xdr, args->bitmask[0] & nfs4_statfs_bitmap[0], + args->bitmask[1] & nfs4_statfs_bitmap[1], &hdr); + encode_nops(&hdr); + return 0; } /* @@ -2040,19 +1907,18 @@ static int nfs4_xdr_enc_server_caps(struct rpc_rqst *req, __be32 *p, const struc { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 2, + .nops = 0, }; - int status; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); - status = encode_putfh(&xdr, fhandle); - if (status == 0) - status = encode_getattr_one(&xdr, FATTR4_WORD0_SUPPORTED_ATTRS| - FATTR4_WORD0_LINK_SUPPORT| - FATTR4_WORD0_SYMLINK_SUPPORT| - FATTR4_WORD0_ACLSUPPORT); - return status; + encode_putfh(&xdr, fhandle, &hdr); + encode_getattr_one(&xdr, FATTR4_WORD0_SUPPORTED_ATTRS| + FATTR4_WORD0_LINK_SUPPORT| + FATTR4_WORD0_SYMLINK_SUPPORT| + FATTR4_WORD0_ACLSUPPORT, &hdr); + encode_nops(&hdr); + return 0; } /* @@ -2062,12 +1928,14 @@ static int nfs4_xdr_enc_renew(struct rpc_rqst *req, __be32 *p, struct nfs_client { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 1, + .nops = 0, }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); - return encode_renew(&xdr, clp); + encode_renew(&xdr, clp, &hdr); + encode_nops(&hdr); + return 0; } /* @@ -2077,12 +1945,14 @@ static int nfs4_xdr_enc_setclientid(struct rpc_rqst *req, __be32 *p, struct nfs4 { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 1, + .nops = 0, }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); - return encode_setclientid(&xdr, sc); + encode_setclientid(&xdr, sc, &hdr); + encode_nops(&hdr); + return 0; } /* @@ -2092,19 +1962,17 @@ static int nfs4_xdr_enc_setclientid_confirm(struct rpc_rqst *req, __be32 *p, str { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 3, + .nops = 0, }; const u32 lease_bitmap[2] = { FATTR4_WORD0_LEASE_TIME, 0 }; - int status; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); - status = encode_setclientid_confirm(&xdr, clp); - if (!status) - status = encode_putrootfh(&xdr); - if (!status) - status = encode_fsinfo(&xdr, lease_bitmap); - return status; + encode_setclientid_confirm(&xdr, clp, &hdr); + encode_putrootfh(&xdr, &hdr); + encode_fsinfo(&xdr, lease_bitmap, &hdr); + encode_nops(&hdr); + return 0; } /* @@ -2114,21 +1982,16 @@ static int nfs4_xdr_enc_delegreturn(struct rpc_rqst *req, __be32 *p, const struc { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 3, + .nops = 0, }; - int status; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); - status = encode_putfh(&xdr, args->fhandle); - if (status != 0) - goto out; - status = encode_delegreturn(&xdr, args->stateid); - if (status != 0) - goto out; - status = encode_getfattr(&xdr, args->bitmask); -out: - return status; + encode_putfh(&xdr, args->fhandle, &hdr); + encode_delegreturn(&xdr, args->stateid, &hdr); + encode_getfattr(&xdr, args->bitmask, &hdr); + encode_nops(&hdr); + return 0; } /* @@ -2138,20 +2001,17 @@ static int nfs4_xdr_enc_fs_locations(struct rpc_rqst *req, __be32 *p, struct nfs { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 3, + .nops = 0, }; struct rpc_auth *auth = req->rq_task->tk_msg.rpc_cred->cr_auth; int replen; - int status; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); - if ((status = encode_putfh(&xdr, args->dir_fh)) != 0) - goto out; - if ((status = encode_lookup(&xdr, args->name)) != 0) - goto out; - if ((status = encode_fs_locations(&xdr, args->bitmask)) != 0) - goto out; + encode_putfh(&xdr, args->dir_fh, &hdr); + encode_lookup(&xdr, args->name, &hdr); + encode_fs_locations(&xdr, args->bitmask, &hdr); + /* set up reply * toplevel_status + OP_PUTFH + status * + OP_LOOKUP + status + OP_GETATTR + status = 7 @@ -2159,8 +2019,8 @@ static int nfs4_xdr_enc_fs_locations(struct rpc_rqst *req, __be32 *p, struct nfs replen = (RPC_REPHDRSIZE + auth->au_rslack + 7) << 2; xdr_inline_pages(&req->rq_rcv_buf, replen, &args->page, 0, PAGE_SIZE); -out: - return status; + encode_nops(&hdr); + return 0; } /* @@ -2217,11 +2077,13 @@ static int decode_compound_hdr(struct xdr_stream *xdr, struct compound_hdr *hdr) READ_BUF(8); READ32(hdr->status); READ32(hdr->taglen); - + READ_BUF(hdr->taglen + 4); hdr->tag = (char *)p; p += XDR_QUADLEN(hdr->taglen); READ32(hdr->nops); + if (unlikely(hdr->nops < 1)) + return nfs4_stat_to_errno(hdr->status); return 0; } @@ -3047,8 +2909,7 @@ static int decode_create(struct xdr_stream *xdr, struct nfs4_change_info *cinfo) static int decode_server_caps(struct xdr_stream *xdr, struct nfs4_server_caps_res *res) { __be32 *savep; - uint32_t attrlen, - bitmap[2] = {0}; + uint32_t attrlen, bitmap[2] = {0}; int status; if ((status = decode_op_hdr(xdr, OP_GETATTR)) != 0) @@ -3070,14 +2931,13 @@ xdr_error: dprintk("%s: xdr returned %d!\n", __func__, -status); return status; } - + static int decode_statfs(struct xdr_stream *xdr, struct nfs_fsstat *fsstat) { __be32 *savep; - uint32_t attrlen, - bitmap[2] = {0}; + uint32_t attrlen, bitmap[2] = {0}; int status; - + if ((status = decode_op_hdr(xdr, OP_GETATTR)) != 0) goto xdr_error; if ((status = decode_attr_bitmap(xdr, bitmap)) != 0) @@ -3107,10 +2967,9 @@ xdr_error: static int decode_pathconf(struct xdr_stream *xdr, struct nfs_pathconf *pathconf) { __be32 *savep; - uint32_t attrlen, - bitmap[2] = {0}; + uint32_t attrlen, bitmap[2] = {0}; int status; - + if ((status = decode_op_hdr(xdr, OP_GETATTR)) != 0) goto xdr_error; if ((status = decode_attr_bitmap(xdr, bitmap)) != 0) @@ -3256,7 +3115,7 @@ static int decode_getfh(struct xdr_stream *xdr, struct nfs_fh *fh) static int decode_link(struct xdr_stream *xdr, struct nfs4_change_info *cinfo) { int status; - + status = decode_op_hdr(xdr, OP_LINK); if (status) return status; @@ -3344,27 +3203,27 @@ static int decode_lookup(struct xdr_stream *xdr) /* This is too sick! */ static int decode_space_limit(struct xdr_stream *xdr, u64 *maxsize) { - __be32 *p; + __be32 *p; uint32_t limit_type, nblocks, blocksize; READ_BUF(12); READ32(limit_type); switch (limit_type) { - case 1: - READ64(*maxsize); - break; - case 2: - READ32(nblocks); - READ32(blocksize); - *maxsize = (uint64_t)nblocks * (uint64_t)blocksize; + case 1: + READ64(*maxsize); + break; + case 2: + READ32(nblocks); + READ32(blocksize); + *maxsize = (uint64_t)nblocks * (uint64_t)blocksize; } return 0; } static int decode_delegation(struct xdr_stream *xdr, struct nfs_openres *res) { - __be32 *p; - uint32_t delegation_type; + __be32 *p; + uint32_t delegation_type; READ_BUF(4); READ32(delegation_type); @@ -3375,13 +3234,14 @@ static int decode_delegation(struct xdr_stream *xdr, struct nfs_openres *res) READ_BUF(NFS4_STATEID_SIZE+4); COPYMEM(res->delegation.data, NFS4_STATEID_SIZE); READ32(res->do_recall); + switch (delegation_type) { - case NFS4_OPEN_DELEGATE_READ: - res->delegation_type = FMODE_READ; - break; - case NFS4_OPEN_DELEGATE_WRITE: - res->delegation_type = FMODE_WRITE|FMODE_READ; - if (decode_space_limit(xdr, &res->maxsize) < 0) + case NFS4_OPEN_DELEGATE_READ: + res->delegation_type = FMODE_READ; + break; + case NFS4_OPEN_DELEGATE_WRITE: + res->delegation_type = FMODE_WRITE|FMODE_READ; + if (decode_space_limit(xdr, &res->maxsize) < 0) return -EIO; } return decode_ace(xdr, NULL, res->server->nfs_client); @@ -3389,27 +3249,27 @@ static int decode_delegation(struct xdr_stream *xdr, struct nfs_openres *res) static int decode_open(struct xdr_stream *xdr, struct nfs_openres *res) { - __be32 *p; + __be32 *p; uint32_t savewords, bmlen, i; - int status; + int status; - status = decode_op_hdr(xdr, OP_OPEN); + status = decode_op_hdr(xdr, OP_OPEN); if (status != -EIO) nfs_increment_open_seqid(status, res->seqid); - if (status) - return status; - READ_BUF(NFS4_STATEID_SIZE); - COPYMEM(res->stateid.data, NFS4_STATEID_SIZE); + if (status) + return status; + READ_BUF(NFS4_STATEID_SIZE); + COPYMEM(res->stateid.data, NFS4_STATEID_SIZE); - decode_change_info(xdr, &res->cinfo); + decode_change_info(xdr, &res->cinfo); - READ_BUF(8); - READ32(res->rflags); - READ32(bmlen); - if (bmlen > 10) - goto xdr_error; + READ_BUF(8); + READ32(res->rflags); + READ32(bmlen); + if (bmlen > 10) + goto xdr_error; - READ_BUF(bmlen << 2); + READ_BUF(bmlen << 2); savewords = min_t(uint32_t, bmlen, NFS4_BITMAP_SIZE); for (i = 0; i < savewords; ++i) READ32(res->attrset[i]); @@ -3424,17 +3284,17 @@ xdr_error: static int decode_open_confirm(struct xdr_stream *xdr, struct nfs_open_confirmres *res) { - __be32 *p; + __be32 *p; int status; - status = decode_op_hdr(xdr, OP_OPEN_CONFIRM); + status = decode_op_hdr(xdr, OP_OPEN_CONFIRM); if (status != -EIO) nfs_increment_open_seqid(status, res->seqid); - if (status) - return status; - READ_BUF(NFS4_STATEID_SIZE); - COPYMEM(res->stateid.data, NFS4_STATEID_SIZE); - return 0; + if (status) + return status; + READ_BUF(NFS4_STATEID_SIZE); + COPYMEM(res->stateid.data, NFS4_STATEID_SIZE); + return 0; } static int decode_open_downgrade(struct xdr_stream *xdr, struct nfs_closeres *res) @@ -3562,7 +3422,7 @@ static int decode_readdir(struct xdr_stream *xdr, struct rpc_rqst *req, struct n dprintk("NFS: readdir reply truncated!\n"); entry[1] = 1; } -out: +out: kunmap_atomic(kaddr, KM_USER0); return 0; short_pkt: @@ -3718,7 +3578,6 @@ static int decode_setattr(struct xdr_stream *xdr, struct nfs_setattrres *res) uint32_t bmlen; int status; - status = decode_op_hdr(xdr, OP_SETATTR); if (status) return status; @@ -3738,7 +3597,7 @@ static int decode_setclientid(struct xdr_stream *xdr, struct nfs_client *clp) READ32(opnum); if (opnum != OP_SETCLIENTID) { dprintk("nfs: decode_setclientid: Server returned operation" - " %d\n", opnum); + " %d\n", opnum); return -EIO; } READ32(nfserr); @@ -3792,34 +3651,34 @@ static int decode_delegreturn(struct xdr_stream *xdr) } /* + * END OF "GENERIC" DECODE ROUTINES. + */ + +/* * Decode OPEN_DOWNGRADE response */ static int nfs4_xdr_dec_open_downgrade(struct rpc_rqst *rqstp, __be32 *p, struct nfs_closeres *res) { - struct xdr_stream xdr; - struct compound_hdr hdr; - int status; - - xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); - status = decode_compound_hdr(&xdr, &hdr); - if (status) - goto out; - status = decode_putfh(&xdr); - if (status) - goto out; - status = decode_open_downgrade(&xdr, res); + struct xdr_stream xdr; + struct compound_hdr hdr; + int status; + + xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); + status = decode_compound_hdr(&xdr, &hdr); + if (status) + goto out; + status = decode_putfh(&xdr); + if (status) + goto out; + status = decode_open_downgrade(&xdr, res); if (status != 0) goto out; decode_getfattr(&xdr, res->fattr, res->server); out: - return status; + return status; } /* - * END OF "GENERIC" DECODE ROUTINES. - */ - -/* * Decode ACCESS response */ static int nfs4_xdr_dec_access(struct rpc_rqst *rqstp, __be32 *p, struct nfs4_accessres *res) @@ -3827,7 +3686,7 @@ static int nfs4_xdr_dec_access(struct rpc_rqst *rqstp, __be32 *p, struct nfs4_ac struct xdr_stream xdr; struct compound_hdr hdr; int status; - + xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); if ((status = decode_compound_hdr(&xdr, &hdr)) != 0) goto out; @@ -3850,7 +3709,7 @@ static int nfs4_xdr_dec_lookup(struct rpc_rqst *rqstp, __be32 *p, struct nfs4_lo struct xdr_stream xdr; struct compound_hdr hdr; int status; - + xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); if ((status = decode_compound_hdr(&xdr, &hdr)) != 0) goto out; @@ -3873,7 +3732,7 @@ static int nfs4_xdr_dec_lookup_root(struct rpc_rqst *rqstp, __be32 *p, struct nf struct xdr_stream xdr; struct compound_hdr hdr; int status; - + xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); if ((status = decode_compound_hdr(&xdr, &hdr)) != 0) goto out; @@ -3893,7 +3752,7 @@ static int nfs4_xdr_dec_remove(struct rpc_rqst *rqstp, __be32 *p, struct nfs_rem struct xdr_stream xdr; struct compound_hdr hdr; int status; - + xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); if ((status = decode_compound_hdr(&xdr, &hdr)) != 0) goto out; @@ -3914,7 +3773,7 @@ static int nfs4_xdr_dec_rename(struct rpc_rqst *rqstp, __be32 *p, struct nfs4_re struct xdr_stream xdr; struct compound_hdr hdr; int status; - + xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); if ((status = decode_compound_hdr(&xdr, &hdr)) != 0) goto out; @@ -3944,7 +3803,7 @@ static int nfs4_xdr_dec_link(struct rpc_rqst *rqstp, __be32 *p, struct nfs4_link struct xdr_stream xdr; struct compound_hdr hdr; int status; - + xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); if ((status = decode_compound_hdr(&xdr, &hdr)) != 0) goto out; @@ -3977,7 +3836,7 @@ static int nfs4_xdr_dec_create(struct rpc_rqst *rqstp, __be32 *p, struct nfs4_cr struct xdr_stream xdr; struct compound_hdr hdr; int status; - + xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); if ((status = decode_compound_hdr(&xdr, &hdr)) != 0) goto out; @@ -4014,7 +3873,7 @@ static int nfs4_xdr_dec_getattr(struct rpc_rqst *rqstp, __be32 *p, struct nfs4_g struct xdr_stream xdr; struct compound_hdr hdr; int status; - + xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); status = decode_compound_hdr(&xdr, &hdr); if (status) @@ -4025,7 +3884,6 @@ static int nfs4_xdr_dec_getattr(struct rpc_rqst *rqstp, __be32 *p, struct nfs4_g status = decode_getfattr(&xdr, res->fattr, res->server); out: return status; - } /* @@ -4034,21 +3892,20 @@ out: static int nfs4_xdr_enc_setacl(struct rpc_rqst *req, __be32 *p, struct nfs_setaclargs *args) { - struct xdr_stream xdr; - struct compound_hdr hdr = { - .nops = 2, - }; - int status; - - xdr_init_encode(&xdr, &req->rq_snd_buf, p); - encode_compound_hdr(&xdr, &hdr); - status = encode_putfh(&xdr, args->fh); - if (status) - goto out; - status = encode_setacl(&xdr, args); -out: - return status; + struct xdr_stream xdr; + struct compound_hdr hdr = { + .nops = 0, + }; + int status; + + xdr_init_encode(&xdr, &req->rq_snd_buf, p); + encode_compound_hdr(&xdr, &hdr); + encode_putfh(&xdr, args->fh, &hdr); + status = encode_setacl(&xdr, args, &hdr); + encode_nops(&hdr); + return status; } + /* * Decode SETACL response */ @@ -4099,18 +3956,18 @@ out: */ static int nfs4_xdr_dec_close(struct rpc_rqst *rqstp, __be32 *p, struct nfs_closeres *res) { - struct xdr_stream xdr; - struct compound_hdr hdr; - int status; - - xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); - status = decode_compound_hdr(&xdr, &hdr); - if (status) - goto out; - status = decode_putfh(&xdr); - if (status) - goto out; - status = decode_close(&xdr, res); + struct xdr_stream xdr; + struct compound_hdr hdr; + int status; + + xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); + status = decode_compound_hdr(&xdr, &hdr); + if (status) + goto out; + status = decode_putfh(&xdr); + if (status) + goto out; + status = decode_close(&xdr, res); if (status != 0) goto out; /* @@ -4121,7 +3978,7 @@ static int nfs4_xdr_dec_close(struct rpc_rqst *rqstp, __be32 *p, struct nfs_clos */ decode_getfattr(&xdr, res->fattr, res->server); out: - return status; + return status; } /* @@ -4129,23 +3986,23 @@ out: */ static int nfs4_xdr_dec_open(struct rpc_rqst *rqstp, __be32 *p, struct nfs_openres *res) { - struct xdr_stream xdr; - struct compound_hdr hdr; - int status; - - xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); - status = decode_compound_hdr(&xdr, &hdr); - if (status) - goto out; - status = decode_putfh(&xdr); - if (status) - goto out; - status = decode_savefh(&xdr); + struct xdr_stream xdr; + struct compound_hdr hdr; + int status; + + xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); + status = decode_compound_hdr(&xdr, &hdr); + if (status) + goto out; + status = decode_putfh(&xdr); + if (status) + goto out; + status = decode_savefh(&xdr); + if (status) + goto out; + status = decode_open(&xdr, res); if (status) goto out; - status = decode_open(&xdr, res); - if (status) - goto out; if (decode_getfh(&xdr, &res->fh) != 0) goto out; if (decode_getfattr(&xdr, res->f_attr, res->server) != 0) @@ -4154,7 +4011,7 @@ static int nfs4_xdr_dec_open(struct rpc_rqst *rqstp, __be32 *p, struct nfs_openr goto out; decode_getfattr(&xdr, res->dir_attr, res->server); out: - return status; + return status; } /* @@ -4162,20 +4019,20 @@ out: */ static int nfs4_xdr_dec_open_confirm(struct rpc_rqst *rqstp, __be32 *p, struct nfs_open_confirmres *res) { - struct xdr_stream xdr; - struct compound_hdr hdr; - int status; - - xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); - status = decode_compound_hdr(&xdr, &hdr); - if (status) - goto out; - status = decode_putfh(&xdr); - if (status) - goto out; - status = decode_open_confirm(&xdr, res); + struct xdr_stream xdr; + struct compound_hdr hdr; + int status; + + xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); + status = decode_compound_hdr(&xdr, &hdr); + if (status) + goto out; + status = decode_putfh(&xdr); + if (status) + goto out; + status = decode_open_confirm(&xdr, res); out: - return status; + return status; } /* @@ -4183,23 +4040,23 @@ out: */ static int nfs4_xdr_dec_open_noattr(struct rpc_rqst *rqstp, __be32 *p, struct nfs_openres *res) { - struct xdr_stream xdr; - struct compound_hdr hdr; - int status; - - xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); - status = decode_compound_hdr(&xdr, &hdr); - if (status) - goto out; - status = decode_putfh(&xdr); - if (status) - goto out; - status = decode_open(&xdr, res); - if (status) - goto out; + struct xdr_stream xdr; + struct compound_hdr hdr; + int status; + + xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); + status = decode_compound_hdr(&xdr, &hdr); + if (status) + goto out; + status = decode_putfh(&xdr); + if (status) + goto out; + status = decode_open(&xdr, res); + if (status) + goto out; decode_getfattr(&xdr, res->f_attr, res->server); out: - return status; + return status; } /* @@ -4207,25 +4064,25 @@ out: */ static int nfs4_xdr_dec_setattr(struct rpc_rqst *rqstp, __be32 *p, struct nfs_setattrres *res) { - struct xdr_stream xdr; - struct compound_hdr hdr; - int status; - - xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); - status = decode_compound_hdr(&xdr, &hdr); - if (status) - goto out; - status = decode_putfh(&xdr); - if (status) - goto out; - status = decode_setattr(&xdr, res); - if (status) - goto out; + struct xdr_stream xdr; + struct compound_hdr hdr; + int status; + + xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); + status = decode_compound_hdr(&xdr, &hdr); + if (status) + goto out; + status = decode_putfh(&xdr); + if (status) + goto out; + status = decode_setattr(&xdr, res); + if (status) + goto out; status = decode_getfattr(&xdr, res->fattr, res->server); if (status == NFS4ERR_DELAY) status = 0; out: - return status; + return status; } /* @@ -4421,8 +4278,6 @@ static int nfs4_xdr_dec_fsinfo(struct rpc_rqst *req, __be32 *p, struct nfs_fsinf status = decode_putfh(&xdr); if (!status) status = decode_fsinfo(&xdr, fsinfo); - if (!status) - status = nfs4_stat_to_errno(hdr.status); return status; } @@ -4511,8 +4366,6 @@ static int nfs4_xdr_dec_setclientid(struct rpc_rqst *req, __be32 *p, status = decode_compound_hdr(&xdr, &hdr); if (!status) status = decode_setclientid(&xdr, clp); - if (!status) - status = nfs4_stat_to_errno(hdr.status); return status; } @@ -4533,8 +4386,6 @@ static int nfs4_xdr_dec_setclientid_confirm(struct rpc_rqst *req, __be32 *p, str status = decode_putrootfh(&xdr); if (!status) status = decode_fsinfo(&xdr, fsinfo); - if (!status) - status = nfs4_stat_to_errno(hdr.status); return status; } @@ -4715,7 +4566,7 @@ nfs4_stat_to_errno(int stat) .p_replen = NFS4_##restype##_sz, \ .p_statidx = NFSPROC4_CLNT_##proc, \ .p_name = #proc, \ - } +} struct rpc_procinfo nfs4_procedures[] = { PROC(READ, enc_read, dec_read), diff --git a/fs/nfs/nfsroot.c b/fs/nfs/nfsroot.c index 8478fc25dae..d9ef602fbc5 100644 --- a/fs/nfs/nfsroot.c +++ b/fs/nfs/nfsroot.c @@ -86,6 +86,8 @@ #include <net/ipconfig.h> #include <linux/parser.h> +#include "internal.h" + /* Define this to allow debugging output */ #undef NFSROOT_DEBUG #define NFSDBG_FACILITY NFSDBG_ROOT @@ -100,7 +102,7 @@ static char nfs_root_name[256] __initdata = ""; static __be32 servaddr __initdata = 0; /* Name of directory to mount */ -static char nfs_path[NFS_MAXPATHLEN] __initdata = { 0, }; +static char nfs_export_path[NFS_MAXPATHLEN] __initdata = { 0, }; /* NFS-related data */ static struct nfs_mount_data nfs_data __initdata = { 0, };/* NFS mount info */ @@ -312,7 +314,7 @@ static int __init root_nfs_name(char *name) printk(KERN_ERR "Root-NFS: Pathname for remote directory too long.\n"); return -1; } - sprintf(nfs_path, buf, cp); + sprintf(nfs_export_path, buf, cp); return 1; } @@ -329,7 +331,7 @@ static int __init root_nfs_addr(void) } snprintf(nfs_data.hostname, sizeof(nfs_data.hostname), - "%u.%u.%u.%u", NIPQUAD(servaddr)); + "%pI4", &servaddr); return 0; } @@ -340,7 +342,7 @@ static int __init root_nfs_addr(void) static void __init root_nfs_print(void) { printk(KERN_NOTICE "Root-NFS: Mounting %s on server %s as root\n", - nfs_path, nfs_data.hostname); + nfs_export_path, nfs_data.hostname); printk(KERN_NOTICE "Root-NFS: rsize = %d, wsize = %d, timeo = %d, retrans = %d\n", nfs_data.rsize, nfs_data.wsize, nfs_data.timeo, nfs_data.retrans); printk(KERN_NOTICE "Root-NFS: acreg (min,max) = (%d,%d), acdir (min,max) = (%d,%d)\n", @@ -421,8 +423,8 @@ static int __init root_nfs_getport(int program, int version, int proto) { struct sockaddr_in sin; - printk(KERN_NOTICE "Looking up port of RPC %d/%d on %u.%u.%u.%u\n", - program, version, NIPQUAD(servaddr)); + printk(KERN_NOTICE "Looking up port of RPC %d/%d on %pI4\n", + program, version, &servaddr); set_sockaddr(&sin, servaddr, 0); return rpcb_getport_sync(&sin, program, version, proto); } @@ -485,18 +487,23 @@ static int __init root_nfs_get_handle(void) { struct nfs_fh fh; struct sockaddr_in sin; + struct nfs_mount_request request = { + .sap = (struct sockaddr *)&sin, + .salen = sizeof(sin), + .dirpath = nfs_export_path, + .version = (nfs_data.flags & NFS_MOUNT_VER3) ? + NFS_MNT3_VERSION : NFS_MNT_VERSION, + .protocol = (nfs_data.flags & NFS_MOUNT_TCP) ? + XPRT_TRANSPORT_TCP : XPRT_TRANSPORT_UDP, + .fh = &fh, + }; int status; - int protocol = (nfs_data.flags & NFS_MOUNT_TCP) ? - XPRT_TRANSPORT_TCP : XPRT_TRANSPORT_UDP; - int version = (nfs_data.flags & NFS_MOUNT_VER3) ? - NFS_MNT3_VERSION : NFS_MNT_VERSION; set_sockaddr(&sin, servaddr, htons(mount_port)); - status = nfs_mount((struct sockaddr *) &sin, sizeof(sin), NULL, - nfs_path, version, protocol, &fh); + status = nfs_mount(&request); if (status < 0) printk(KERN_ERR "Root-NFS: Server returned error %d " - "while mounting %s\n", status, nfs_path); + "while mounting %s\n", status, nfs_export_path); else { nfs_data.root.size = fh.size; memcpy(nfs_data.root.data, fh.data, fh.size); diff --git a/fs/nfs/proc.c b/fs/nfs/proc.c index 4dbb84df1b6..193465210d7 100644 --- a/fs/nfs/proc.c +++ b/fs/nfs/proc.c @@ -65,14 +65,20 @@ nfs_proc_get_root(struct nfs_server *server, struct nfs_fh *fhandle, dprintk("%s: call getattr\n", __func__); nfs_fattr_init(fattr); - status = rpc_call_sync(server->nfs_client->cl_rpcclient, &msg, 0); + status = rpc_call_sync(server->client, &msg, 0); + /* Retry with default authentication if different */ + if (status && server->nfs_client->cl_rpcclient != server->client) + status = rpc_call_sync(server->nfs_client->cl_rpcclient, &msg, 0); dprintk("%s: reply getattr: %d\n", __func__, status); if (status) return status; dprintk("%s: call statfs\n", __func__); msg.rpc_proc = &nfs_procedures[NFSPROC_STATFS]; msg.rpc_resp = &fsinfo; - status = rpc_call_sync(server->nfs_client->cl_rpcclient, &msg, 0); + status = rpc_call_sync(server->client, &msg, 0); + /* Retry with default authentication if different */ + if (status && server->nfs_client->cl_rpcclient != server->client) + status = rpc_call_sync(server->nfs_client->cl_rpcclient, &msg, 0); dprintk("%s: reply statfs: %d\n", __func__, status); if (status) return status; diff --git a/fs/nfs/read.c b/fs/nfs/read.c index 40d17987d0e..f856004bb7f 100644 --- a/fs/nfs/read.c +++ b/fs/nfs/read.c @@ -533,12 +533,6 @@ readpage_async_filler(void *data, struct page *page) unsigned int len; int error; - error = nfs_wb_page(inode, page); - if (error) - goto out_unlock; - if (PageUptodate(page)) - goto out_unlock; - len = nfs_page_length(page); if (len == 0) return nfs_return_empty_page(page); diff --git a/fs/nfs/super.c b/fs/nfs/super.c index ffb697416cb..d6686f4786d 100644 --- a/fs/nfs/super.c +++ b/fs/nfs/super.c @@ -5,7 +5,7 @@ * * nfs superblock handling functions * - * Modularised by Alan Cox <Alan.Cox@linux.org>, while hacking some + * Modularised by Alan Cox <alan@lxorguk.ukuu.org.uk>, while hacking some * experimental NFS changes. Modularisation taken straight from SYS5 fs. * * Change to nfs_read_super() to permit NFS mounts to multi-homed hosts. @@ -75,6 +75,7 @@ enum { Opt_acl, Opt_noacl, Opt_rdirplus, Opt_nordirplus, Opt_sharecache, Opt_nosharecache, + Opt_resvport, Opt_noresvport, /* Mount options that take integer arguments */ Opt_port, @@ -91,6 +92,7 @@ enum { /* Mount options that take string arguments */ Opt_sec, Opt_proto, Opt_mountproto, Opt_mounthost, Opt_addr, Opt_mountaddr, Opt_clientaddr, + Opt_lookupcache, /* Special mount options */ Opt_userspace, Opt_deprecated, Opt_sloppy, @@ -128,6 +130,8 @@ static const match_table_t nfs_mount_option_tokens = { { Opt_nordirplus, "nordirplus" }, { Opt_sharecache, "sharecache" }, { Opt_nosharecache, "nosharecache" }, + { Opt_resvport, "resvport" }, + { Opt_noresvport, "noresvport" }, { Opt_port, "port=%u" }, { Opt_rsize, "rsize=%u" }, @@ -154,6 +158,8 @@ static const match_table_t nfs_mount_option_tokens = { { Opt_mounthost, "mounthost=%s" }, { Opt_mountaddr, "mountaddr=%s" }, + { Opt_lookupcache, "lookupcache=%s" }, + { Opt_err, NULL } }; @@ -200,6 +206,22 @@ static const match_table_t nfs_secflavor_tokens = { { Opt_sec_err, NULL } }; +enum { + Opt_lookupcache_all, Opt_lookupcache_positive, + Opt_lookupcache_none, + + Opt_lookupcache_err +}; + +static match_table_t nfs_lookupcache_tokens = { + { Opt_lookupcache_all, "all" }, + { Opt_lookupcache_positive, "pos" }, + { Opt_lookupcache_positive, "positive" }, + { Opt_lookupcache_none, "none" }, + + { Opt_lookupcache_err, NULL } +}; + static void nfs_umount_begin(struct super_block *); static int nfs_statfs(struct dentry *, struct kstatfs *); @@ -209,7 +231,6 @@ static int nfs_get_sb(struct file_system_type *, int, const char *, void *, stru static int nfs_xdev_get_sb(struct file_system_type *fs_type, int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt); static void nfs_kill_super(struct super_block *); -static void nfs_put_super(struct super_block *); static int nfs_remount(struct super_block *sb, int *flags, char *raw_data); static struct file_system_type nfs_fs_type = { @@ -232,7 +253,6 @@ static const struct super_operations nfs_sops = { .alloc_inode = nfs_alloc_inode, .destroy_inode = nfs_destroy_inode, .write_inode = nfs_write_inode, - .put_super = nfs_put_super, .statfs = nfs_statfs, .clear_inode = nfs_clear_inode, .umount_begin = nfs_umount_begin, @@ -337,26 +357,20 @@ void __exit unregister_nfs_fs(void) unregister_filesystem(&nfs_fs_type); } -void nfs_sb_active(struct nfs_server *server) +void nfs_sb_active(struct super_block *sb) { - atomic_inc(&server->active); -} + struct nfs_server *server = NFS_SB(sb); -void nfs_sb_deactive(struct nfs_server *server) -{ - if (atomic_dec_and_test(&server->active)) - wake_up(&server->active_wq); + if (atomic_inc_return(&server->active) == 1) + atomic_inc(&sb->s_active); } -static void nfs_put_super(struct super_block *sb) +void nfs_sb_deactive(struct super_block *sb) { struct nfs_server *server = NFS_SB(sb); - /* - * Make sure there are no outstanding ops to this server. - * If so, wait for them to finish before allowing the - * unmount to continue. - */ - wait_event(server->active_wq, atomic_read(&server->active) == 0); + + if (atomic_dec_and_test(&server->active)) + deactivate_super(sb); } /* @@ -451,14 +465,12 @@ static void nfs_show_mountd_options(struct seq_file *m, struct nfs_server *nfss, switch (sap->sa_family) { case AF_INET: { struct sockaddr_in *sin = (struct sockaddr_in *)sap; - seq_printf(m, ",mountaddr=" NIPQUAD_FMT, - NIPQUAD(sin->sin_addr.s_addr)); + seq_printf(m, ",mountaddr=%pI4", &sin->sin_addr.s_addr); break; } case AF_INET6: { struct sockaddr_in6 *sin6 = (struct sockaddr_in6 *)sap; - seq_printf(m, ",mountaddr=" NIP6_FMT, - NIP6(sin6->sin6_addr)); + seq_printf(m, ",mountaddr=%pI6", &sin6->sin6_addr); break; } default: @@ -503,7 +515,8 @@ static void nfs_show_mount_options(struct seq_file *m, struct nfs_server *nfss, { NFS_MOUNT_NONLM, ",nolock", "" }, { NFS_MOUNT_NOACL, ",noacl", "" }, { NFS_MOUNT_NORDIRPLUS, ",nordirplus", "" }, - { NFS_MOUNT_UNSHARED, ",nosharecache", ""}, + { NFS_MOUNT_UNSHARED, ",nosharecache", "" }, + { NFS_MOUNT_NORESVPORT, ",noresvport", "" }, { 0, NULL, NULL } }; const struct proc_nfs_info *nfs_infop; @@ -664,25 +677,6 @@ static void nfs_umount_begin(struct super_block *sb) } /* - * Set the port number in an address. Be agnostic about the address family. - */ -static void nfs_set_port(struct sockaddr *sap, unsigned short port) -{ - switch (sap->sa_family) { - case AF_INET: { - struct sockaddr_in *ap = (struct sockaddr_in *)sap; - ap->sin_port = htons(port); - break; - } - case AF_INET6: { - struct sockaddr_in6 *ap = (struct sockaddr_in6 *)sap; - ap->sin6_port = htons(port); - break; - } - } -} - -/* * Sanity-check a server address provided by the mount command. * * Address family must be initialized, and address must not be @@ -724,20 +718,22 @@ static void nfs_parse_ipv4_address(char *string, size_t str_len, *addr_len = 0; } -#define IPV6_SCOPE_DELIMITER '%' - #if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE) -static void nfs_parse_ipv6_scope_id(const char *string, const size_t str_len, - const char *delim, - struct sockaddr_in6 *sin6) +static int nfs_parse_ipv6_scope_id(const char *string, const size_t str_len, + const char *delim, + struct sockaddr_in6 *sin6) { char *p; size_t len; - if (!(ipv6_addr_type(&sin6->sin6_addr) & IPV6_ADDR_LINKLOCAL)) - return ; + if ((string + str_len) == delim) + return 1; + if (*delim != IPV6_SCOPE_DELIMITER) - return; + return 0; + + if (!(ipv6_addr_type(&sin6->sin6_addr) & IPV6_ADDR_LINKLOCAL)) + return 0; len = (string + str_len) - delim - 1; p = kstrndup(delim + 1, len, GFP_KERNEL); @@ -750,14 +746,20 @@ static void nfs_parse_ipv6_scope_id(const char *string, const size_t str_len, scope_id = dev->ifindex; dev_put(dev); } else { - /* scope_id is set to zero on error */ - strict_strtoul(p, 10, &scope_id); + if (strict_strtoul(p, 10, &scope_id) == 0) { + kfree(p); + return 0; + } } kfree(p); + sin6->sin6_scope_id = scope_id; dfprintk(MOUNT, "NFS: IPv6 scope ID = %lu\n", scope_id); + return 1; } + + return 0; } static void nfs_parse_ipv6_address(char *string, size_t str_len, @@ -773,9 +775,11 @@ static void nfs_parse_ipv6_address(char *string, size_t str_len, sin6->sin6_family = AF_INET6; *addr_len = sizeof(*sin6); - if (in6_pton(string, str_len, addr, IPV6_SCOPE_DELIMITER, &delim)) { - nfs_parse_ipv6_scope_id(string, str_len, delim, sin6); - return; + if (in6_pton(string, str_len, addr, + IPV6_SCOPE_DELIMITER, &delim) != 0) { + if (nfs_parse_ipv6_scope_id(string, str_len, + delim, sin6) != 0) + return; } } @@ -798,7 +802,7 @@ static void nfs_parse_ipv6_address(char *string, size_t str_len, * If there is a problem constructing the new sockaddr, set the address * family to AF_UNSPEC. */ -static void nfs_parse_ip_address(char *string, size_t str_len, +void nfs_parse_ip_address(char *string, size_t str_len, struct sockaddr *sap, size_t *addr_len) { unsigned int i, colons; @@ -1033,6 +1037,12 @@ static int nfs_parse_mount_options(char *raw, case Opt_nosharecache: mnt->flags |= NFS_MOUNT_UNSHARED; break; + case Opt_resvport: + mnt->flags &= ~NFS_MOUNT_NORESVPORT; + break; + case Opt_noresvport: + mnt->flags |= NFS_MOUNT_NORESVPORT; + break; /* * options that take numeric values @@ -1258,6 +1268,30 @@ static int nfs_parse_mount_options(char *raw, &mnt->mount_server.addrlen); kfree(string); break; + case Opt_lookupcache: + string = match_strdup(args); + if (string == NULL) + goto out_nomem; + token = match_token(string, + nfs_lookupcache_tokens, args); + kfree(string); + switch (token) { + case Opt_lookupcache_all: + mnt->flags &= ~(NFS_MOUNT_LOOKUP_CACHE_NONEG|NFS_MOUNT_LOOKUP_CACHE_NONE); + break; + case Opt_lookupcache_positive: + mnt->flags &= ~NFS_MOUNT_LOOKUP_CACHE_NONE; + mnt->flags |= NFS_MOUNT_LOOKUP_CACHE_NONEG; + break; + case Opt_lookupcache_none: + mnt->flags |= NFS_MOUNT_LOOKUP_CACHE_NONEG|NFS_MOUNT_LOOKUP_CACHE_NONE; + break; + default: + errors++; + dfprintk(MOUNT, "NFS: invalid " + "lookupcache argument\n"); + }; + break; /* * Special options @@ -1303,8 +1337,14 @@ out_security_failure: static int nfs_try_mount(struct nfs_parsed_mount_data *args, struct nfs_fh *root_fh) { - struct sockaddr *sap = (struct sockaddr *)&args->mount_server.address; - char *hostname; + struct nfs_mount_request request = { + .sap = (struct sockaddr *) + &args->mount_server.address, + .dirpath = args->nfs_server.export_path, + .protocol = args->mount_server.protocol, + .fh = root_fh, + .noresvport = args->flags & NFS_MOUNT_NORESVPORT, + }; int status; if (args->mount_server.version == 0) { @@ -1313,42 +1353,38 @@ static int nfs_try_mount(struct nfs_parsed_mount_data *args, else args->mount_server.version = NFS_MNT_VERSION; } + request.version = args->mount_server.version; if (args->mount_server.hostname) - hostname = args->mount_server.hostname; + request.hostname = args->mount_server.hostname; else - hostname = args->nfs_server.hostname; + request.hostname = args->nfs_server.hostname; /* * Construct the mount server's address. */ if (args->mount_server.address.ss_family == AF_UNSPEC) { - memcpy(sap, &args->nfs_server.address, + memcpy(request.sap, &args->nfs_server.address, args->nfs_server.addrlen); args->mount_server.addrlen = args->nfs_server.addrlen; } + request.salen = args->mount_server.addrlen; /* * autobind will be used if mount_server.port == 0 */ - nfs_set_port(sap, args->mount_server.port); + nfs_set_port(request.sap, args->mount_server.port); /* * Now ask the mount server to map our export path * to a file handle. */ - status = nfs_mount(sap, - args->mount_server.addrlen, - hostname, - args->nfs_server.export_path, - args->mount_server.version, - args->mount_server.protocol, - root_fh); + status = nfs_mount(&request); if (status == 0) return 0; dfprintk(MOUNT, "NFS: unable to mount server %s, error %d\n", - hostname, status); + request.hostname, status); return status; } @@ -1558,7 +1594,7 @@ static int nfs_validate_mount_data(void *options, * Translate to nfs_parsed_mount_data, which nfs_fill_super * can deal with. */ - args->flags = data->flags; + args->flags = data->flags & NFS_MOUNT_FLAGMASK; args->rsize = data->rsize; args->wsize = data->wsize; args->timeo = data->timeo; @@ -2395,7 +2431,7 @@ static void nfs4_kill_super(struct super_block *sb) { struct nfs_server *server = NFS_SB(sb); - nfs_return_all_delegations(sb); + nfs_super_return_all_delegations(sb); kill_anon_super(sb); nfs4_renewd_prepare_shutdown(server); @@ -2433,7 +2469,7 @@ static int nfs4_xdev_get_sb(struct file_system_type *fs_type, int flags, compare_super = NULL; /* Get a superblock - note that we may end up sharing one that already exists */ - s = sget(&nfs_fs_type, compare_super, nfs_set_super, &sb_mntdata); + s = sget(&nfs4_fs_type, compare_super, nfs_set_super, &sb_mntdata); if (IS_ERR(s)) { error = PTR_ERR(s); goto out_err_nosb; @@ -2518,7 +2554,7 @@ static int nfs4_referral_get_sb(struct file_system_type *fs_type, int flags, compare_super = NULL; /* Get a superblock - note that we may end up sharing one that already exists */ - s = sget(&nfs_fs_type, compare_super, nfs_set_super, &sb_mntdata); + s = sget(&nfs4_fs_type, compare_super, nfs_set_super, &sb_mntdata); if (IS_ERR(s)) { error = PTR_ERR(s); goto out_err_nosb; diff --git a/fs/nfs/unlink.c b/fs/nfs/unlink.c index f089e5839d7..ecc29534777 100644 --- a/fs/nfs/unlink.c +++ b/fs/nfs/unlink.c @@ -99,7 +99,7 @@ static void nfs_async_unlink_release(void *calldata) nfs_dec_sillycount(data->dir); nfs_free_unlinkdata(data); - nfs_sb_deactive(NFS_SB(sb)); + nfs_sb_deactive(sb); } static const struct rpc_call_ops nfs_unlink_ops = { @@ -118,6 +118,7 @@ static int nfs_do_call_unlink(struct dentry *parent, struct inode *dir, struct n .rpc_message = &msg, .callback_ops = &nfs_unlink_ops, .callback_data = data, + .workqueue = nfsiod_workqueue, .flags = RPC_TASK_ASYNC, }; struct rpc_task *task; @@ -149,7 +150,7 @@ static int nfs_do_call_unlink(struct dentry *parent, struct inode *dir, struct n nfs_dec_sillycount(dir); return 0; } - nfs_sb_active(NFS_SERVER(dir)); + nfs_sb_active(dir->i_sb); data->args.fh = NFS_FH(dir); nfs_fattr_init(&data->res.dir_attr); diff --git a/fs/nfs/write.c b/fs/nfs/write.c index 3229e217c77..9f9845859fc 100644 --- a/fs/nfs/write.c +++ b/fs/nfs/write.c @@ -1427,8 +1427,9 @@ static int nfs_write_mapping(struct address_space *mapping, int how) .bdi = mapping->backing_dev_info, .sync_mode = WB_SYNC_NONE, .nr_to_write = LONG_MAX, + .range_start = 0, + .range_end = LLONG_MAX, .for_writepages = 1, - .range_cyclic = 1, }; int ret; diff --git a/fs/nfs_common/nfsacl.c b/fs/nfs_common/nfsacl.c index c11f5375d7c..04133aacb1e 100644 --- a/fs/nfs_common/nfsacl.c +++ b/fs/nfs_common/nfsacl.c @@ -29,8 +29,8 @@ MODULE_LICENSE("GPL"); -EXPORT_SYMBOL(nfsacl_encode); -EXPORT_SYMBOL(nfsacl_decode); +EXPORT_SYMBOL_GPL(nfsacl_encode); +EXPORT_SYMBOL_GPL(nfsacl_decode); struct nfsacl_encode_desc { struct xdr_array2_desc desc; diff --git a/fs/nfsctl.c b/fs/nfsctl.c index aed8145d908..b1acbd6ab6f 100644 --- a/fs/nfsctl.c +++ b/fs/nfsctl.c @@ -10,6 +10,8 @@ #include <linux/sunrpc/svc.h> #include <linux/nfsd/nfsd.h> #include <linux/nfsd/syscall.h> +#include <linux/cred.h> +#include <linux/sched.h> #include <linux/linkage.h> #include <linux/namei.h> #include <linux/mount.h> @@ -41,7 +43,8 @@ static struct file *do_open(char *name, int flags) error = may_open(&nd, MAY_WRITE, FMODE_WRITE); if (!error) - return dentry_open(nd.path.dentry, nd.path.mnt, flags); + return dentry_open(nd.path.dentry, nd.path.mnt, flags, + current_cred()); path_put(&nd.path); return ERR_PTR(error); diff --git a/fs/nfsd/auth.c b/fs/nfsd/auth.c index 294992e9bf6..0184fe9b514 100644 --- a/fs/nfsd/auth.c +++ b/fs/nfsd/auth.c @@ -27,53 +27,70 @@ int nfsexp_flags(struct svc_rqst *rqstp, struct svc_export *exp) int nfsd_setuser(struct svc_rqst *rqstp, struct svc_export *exp) { - struct svc_cred cred = rqstp->rq_cred; + struct group_info *rqgi; + struct group_info *gi; + struct cred *new; int i; int flags = nfsexp_flags(rqstp, exp); int ret; + /* discard any old override before preparing the new set */ + revert_creds(get_cred(current->real_cred)); + new = prepare_creds(); + if (!new) + return -ENOMEM; + + new->fsuid = rqstp->rq_cred.cr_uid; + new->fsgid = rqstp->rq_cred.cr_gid; + + rqgi = rqstp->rq_cred.cr_group_info; + if (flags & NFSEXP_ALLSQUASH) { - cred.cr_uid = exp->ex_anon_uid; - cred.cr_gid = exp->ex_anon_gid; - cred.cr_group_info = groups_alloc(0); + new->fsuid = exp->ex_anon_uid; + new->fsgid = exp->ex_anon_gid; + gi = groups_alloc(0); } else if (flags & NFSEXP_ROOTSQUASH) { - struct group_info *gi; - if (!cred.cr_uid) - cred.cr_uid = exp->ex_anon_uid; - if (!cred.cr_gid) - cred.cr_gid = exp->ex_anon_gid; - gi = groups_alloc(cred.cr_group_info->ngroups); - if (gi) - for (i = 0; i < cred.cr_group_info->ngroups; i++) { - if (!GROUP_AT(cred.cr_group_info, i)) - GROUP_AT(gi, i) = exp->ex_anon_gid; - else - GROUP_AT(gi, i) = GROUP_AT(cred.cr_group_info, i); - } - cred.cr_group_info = gi; - } else - get_group_info(cred.cr_group_info); - - if (cred.cr_uid != (uid_t) -1) - current->fsuid = cred.cr_uid; - else - current->fsuid = exp->ex_anon_uid; - if (cred.cr_gid != (gid_t) -1) - current->fsgid = cred.cr_gid; - else - current->fsgid = exp->ex_anon_gid; + if (!new->fsuid) + new->fsuid = exp->ex_anon_uid; + if (!new->fsgid) + new->fsgid = exp->ex_anon_gid; - if (!cred.cr_group_info) - return -ENOMEM; - ret = set_current_groups(cred.cr_group_info); - put_group_info(cred.cr_group_info); - if ((cred.cr_uid)) { - current->cap_effective = - cap_drop_nfsd_set(current->cap_effective); + gi = groups_alloc(rqgi->ngroups); + if (!gi) + goto oom; + + for (i = 0; i < rqgi->ngroups; i++) { + if (!GROUP_AT(rqgi, i)) + GROUP_AT(gi, i) = exp->ex_anon_gid; + else + GROUP_AT(gi, i) = GROUP_AT(rqgi, i); + } } else { - current->cap_effective = - cap_raise_nfsd_set(current->cap_effective, - current->cap_permitted); + gi = get_group_info(rqgi); } + + if (new->fsuid == (uid_t) -1) + new->fsuid = exp->ex_anon_uid; + if (new->fsgid == (gid_t) -1) + new->fsgid = exp->ex_anon_gid; + + ret = set_groups(new, gi); + put_group_info(gi); + if (!ret) + goto error; + + if (new->uid) + new->cap_effective = cap_drop_nfsd_set(new->cap_effective); + else + new->cap_effective = cap_raise_nfsd_set(new->cap_effective, + new->cap_permitted); + put_cred(override_creds(new)); + return 0; + +oom: + ret = -ENOMEM; +error: + abort_creds(new); return ret; } + diff --git a/fs/nfsd/export.c b/fs/nfsd/export.c index 9dc036f1835..5839b229cd0 100644 --- a/fs/nfsd/export.c +++ b/fs/nfsd/export.c @@ -99,7 +99,7 @@ static int expkey_parse(struct cache_detail *cd, char *mesg, int mlen) int fsidtype; char *ep; struct svc_expkey key; - struct svc_expkey *ek; + struct svc_expkey *ek = NULL; if (mesg[mlen-1] != '\n') return -EINVAL; @@ -107,7 +107,8 @@ static int expkey_parse(struct cache_detail *cd, char *mesg, int mlen) buf = kmalloc(PAGE_SIZE, GFP_KERNEL); err = -ENOMEM; - if (!buf) goto out; + if (!buf) + goto out; err = -EINVAL; if ((len=qword_get(&mesg, buf, PAGE_SIZE)) <= 0) @@ -151,34 +152,32 @@ static int expkey_parse(struct cache_detail *cd, char *mesg, int mlen) /* now we want a pathname, or empty meaning NEGATIVE */ err = -EINVAL; - if ((len=qword_get(&mesg, buf, PAGE_SIZE)) < 0) + len = qword_get(&mesg, buf, PAGE_SIZE); + if (len < 0) goto out; dprintk("Path seems to be <%s>\n", buf); err = 0; if (len == 0) { set_bit(CACHE_NEGATIVE, &key.h.flags); ek = svc_expkey_update(&key, ek); - if (ek) - cache_put(&ek->h, &svc_expkey_cache); - else err = -ENOMEM; + if (!ek) + err = -ENOMEM; } else { - struct nameidata nd; - err = path_lookup(buf, 0, &nd); + err = kern_path(buf, 0, &key.ek_path); if (err) goto out; dprintk("Found the path %s\n", buf); - key.ek_path = nd.path; ek = svc_expkey_update(&key, ek); - if (ek) - cache_put(&ek->h, &svc_expkey_cache); - else + if (!ek) err = -ENOMEM; - path_put(&nd.path); + path_put(&key.ek_path); } cache_flush(); out: + if (ek) + cache_put(&ek->h, &svc_expkey_cache); if (dom) auth_domain_put(dom); kfree(buf); @@ -500,35 +499,22 @@ static int svc_export_parse(struct cache_detail *cd, char *mesg, int mlen) int len; int err; struct auth_domain *dom = NULL; - struct nameidata nd; - struct svc_export exp, *expp; + struct svc_export exp = {}, *expp; int an_int; - nd.path.dentry = NULL; - exp.ex_pathname = NULL; - - /* fs locations */ - exp.ex_fslocs.locations = NULL; - exp.ex_fslocs.locations_count = 0; - exp.ex_fslocs.migrated = 0; - - exp.ex_uuid = NULL; - - /* secinfo */ - exp.ex_nflavors = 0; - if (mesg[mlen-1] != '\n') return -EINVAL; mesg[mlen-1] = 0; buf = kmalloc(PAGE_SIZE, GFP_KERNEL); - err = -ENOMEM; - if (!buf) goto out; + if (!buf) + return -ENOMEM; /* client */ - len = qword_get(&mesg, buf, PAGE_SIZE); err = -EINVAL; - if (len <= 0) goto out; + len = qword_get(&mesg, buf, PAGE_SIZE); + if (len <= 0) + goto out; err = -ENOENT; dom = auth_domain_find(buf); @@ -537,25 +523,25 @@ static int svc_export_parse(struct cache_detail *cd, char *mesg, int mlen) /* path */ err = -EINVAL; - if ((len=qword_get(&mesg, buf, PAGE_SIZE)) <= 0) - goto out; - err = path_lookup(buf, 0, &nd); - if (err) goto out_no_path; + if ((len = qword_get(&mesg, buf, PAGE_SIZE)) <= 0) + goto out1; + + err = kern_path(buf, 0, &exp.ex_path); + if (err) + goto out1; - exp.h.flags = 0; exp.ex_client = dom; - exp.ex_path.mnt = nd.path.mnt; - exp.ex_path.dentry = nd.path.dentry; - exp.ex_pathname = kstrdup(buf, GFP_KERNEL); + err = -ENOMEM; + exp.ex_pathname = kstrdup(buf, GFP_KERNEL); if (!exp.ex_pathname) - goto out; + goto out2; /* expiry */ err = -EINVAL; exp.h.expiry_time = get_expiry(&mesg); if (exp.h.expiry_time == 0) - goto out; + goto out3; /* flags */ err = get_int(&mesg, &an_int); @@ -563,22 +549,26 @@ static int svc_export_parse(struct cache_detail *cd, char *mesg, int mlen) err = 0; set_bit(CACHE_NEGATIVE, &exp.h.flags); } else { - if (err || an_int < 0) goto out; + if (err || an_int < 0) + goto out3; exp.ex_flags= an_int; /* anon uid */ err = get_int(&mesg, &an_int); - if (err) goto out; + if (err) + goto out3; exp.ex_anon_uid= an_int; /* anon gid */ err = get_int(&mesg, &an_int); - if (err) goto out; + if (err) + goto out3; exp.ex_anon_gid= an_int; /* fsid */ err = get_int(&mesg, &an_int); - if (err) goto out; + if (err) + goto out3; exp.ex_fsid = an_int; while ((len = qword_get(&mesg, buf, PAGE_SIZE)) > 0) { @@ -604,12 +594,13 @@ static int svc_export_parse(struct cache_detail *cd, char *mesg, int mlen) */ break; if (err) - goto out; + goto out4; } - err = check_export(nd.path.dentry->d_inode, exp.ex_flags, + err = check_export(exp.ex_path.dentry->d_inode, exp.ex_flags, exp.ex_uuid); - if (err) goto out; + if (err) + goto out4; } expp = svc_export_lookup(&exp); @@ -622,15 +613,16 @@ static int svc_export_parse(struct cache_detail *cd, char *mesg, int mlen) err = -ENOMEM; else exp_put(expp); - out: +out4: nfsd4_fslocs_free(&exp.ex_fslocs); kfree(exp.ex_uuid); +out3: kfree(exp.ex_pathname); - if (nd.path.dentry) - path_put(&nd.path); - out_no_path: - if (dom) - auth_domain_put(dom); +out2: + path_put(&exp.ex_path); +out1: + auth_domain_put(dom); +out: kfree(buf); return err; } @@ -998,7 +990,7 @@ exp_export(struct nfsctl_export *nxp) struct svc_export *exp = NULL; struct svc_export new; struct svc_expkey *fsid_key = NULL; - struct nameidata nd; + struct path path; int err; /* Consistency check */ @@ -1021,12 +1013,12 @@ exp_export(struct nfsctl_export *nxp) /* Look up the dentry */ - err = path_lookup(nxp->ex_path, 0, &nd); + err = kern_path(nxp->ex_path, 0, &path); if (err) goto out_put_clp; err = -EINVAL; - exp = exp_get_by_name(clp, nd.path.mnt, nd.path.dentry, NULL); + exp = exp_get_by_name(clp, path.mnt, path.dentry, NULL); memset(&new, 0, sizeof(new)); @@ -1034,8 +1026,8 @@ exp_export(struct nfsctl_export *nxp) if ((nxp->ex_flags & NFSEXP_FSID) && (!IS_ERR(fsid_key = exp_get_fsid_key(clp, nxp->ex_dev))) && fsid_key->ek_path.mnt && - (fsid_key->ek_path.mnt != nd.path.mnt || - fsid_key->ek_path.dentry != nd.path.dentry)) + (fsid_key->ek_path.mnt != path.mnt || + fsid_key->ek_path.dentry != path.dentry)) goto finish; if (!IS_ERR(exp)) { @@ -1051,7 +1043,7 @@ exp_export(struct nfsctl_export *nxp) goto finish; } - err = check_export(nd.path.dentry->d_inode, nxp->ex_flags, NULL); + err = check_export(path.dentry->d_inode, nxp->ex_flags, NULL); if (err) goto finish; err = -ENOMEM; @@ -1064,7 +1056,7 @@ exp_export(struct nfsctl_export *nxp) if (!new.ex_pathname) goto finish; new.ex_client = clp; - new.ex_path = nd.path; + new.ex_path = path; new.ex_flags = nxp->ex_flags; new.ex_anon_uid = nxp->ex_anon_uid; new.ex_anon_gid = nxp->ex_anon_gid; @@ -1090,7 +1082,7 @@ finish: exp_put(exp); if (fsid_key && !IS_ERR(fsid_key)) cache_put(&fsid_key->h, &svc_expkey_cache); - path_put(&nd.path); + path_put(&path); out_put_clp: auth_domain_put(clp); out_unlock: @@ -1121,7 +1113,7 @@ exp_unexport(struct nfsctl_export *nxp) { struct auth_domain *dom; svc_export *exp; - struct nameidata nd; + struct path path; int err; /* Consistency check */ @@ -1138,13 +1130,13 @@ exp_unexport(struct nfsctl_export *nxp) goto out_unlock; } - err = path_lookup(nxp->ex_path, 0, &nd); + err = kern_path(nxp->ex_path, 0, &path); if (err) goto out_domain; err = -EINVAL; - exp = exp_get_by_name(dom, nd.path.mnt, nd.path.dentry, NULL); - path_put(&nd.path); + exp = exp_get_by_name(dom, path.mnt, path.dentry, NULL); + path_put(&path); if (IS_ERR(exp)) goto out_domain; @@ -1166,26 +1158,26 @@ out_unlock: * since its harder to fool a kernel module than a user space program. */ int -exp_rootfh(svc_client *clp, char *path, struct knfsd_fh *f, int maxsize) +exp_rootfh(svc_client *clp, char *name, struct knfsd_fh *f, int maxsize) { struct svc_export *exp; - struct nameidata nd; + struct path path; struct inode *inode; struct svc_fh fh; int err; err = -EPERM; /* NB: we probably ought to check that it's NUL-terminated */ - if (path_lookup(path, 0, &nd)) { - printk("nfsd: exp_rootfh path not found %s", path); + if (kern_path(name, 0, &path)) { + printk("nfsd: exp_rootfh path not found %s", name); return err; } - inode = nd.path.dentry->d_inode; + inode = path.dentry->d_inode; dprintk("nfsd: exp_rootfh(%s [%p] %s:%s/%ld)\n", - path, nd.path.dentry, clp->name, + name, path.dentry, clp->name, inode->i_sb->s_id, inode->i_ino); - exp = exp_parent(clp, nd.path.mnt, nd.path.dentry, NULL); + exp = exp_parent(clp, path.mnt, path.dentry, NULL); if (IS_ERR(exp)) { err = PTR_ERR(exp); goto out; @@ -1195,7 +1187,7 @@ exp_rootfh(svc_client *clp, char *path, struct knfsd_fh *f, int maxsize) * fh must be initialized before calling fh_compose */ fh_init(&fh, maxsize); - if (fh_compose(&fh, exp, nd.path.dentry, NULL)) + if (fh_compose(&fh, exp, path.dentry, NULL)) err = -EINVAL; else err = 0; @@ -1203,7 +1195,7 @@ exp_rootfh(svc_client *clp, char *path, struct knfsd_fh *f, int maxsize) fh_put(&fh); exp_put(exp); out: - path_put(&nd.path); + path_put(&path); return err; } diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c index 094747a1227..6d7d8c02c19 100644 --- a/fs/nfsd/nfs4callback.c +++ b/fs/nfsd/nfs4callback.c @@ -358,6 +358,7 @@ static struct rpc_program cb_program = { .nrvers = ARRAY_SIZE(nfs_cb_version), .version = nfs_cb_version, .stats = &cb_stats, + .pipe_dir_name = "/nfsd4_cb", }; /* Reference counting, callback cleanup, etc., all look racy as heck. @@ -382,8 +383,9 @@ static int do_probe_callback(void *data) .program = &cb_program, .prognumber = cb->cb_prog, .version = nfs_cb_version[1]->number, - .authflavor = RPC_AUTH_UNIX, /* XXX: need AUTH_GSS... */ + .authflavor = clp->cl_flavor, .flags = (RPC_CLNT_CREATE_NOPING | RPC_CLNT_CREATE_QUIET), + .client_name = clp->cl_principal, }; struct rpc_message msg = { .rpc_proc = &nfs4_cb_procedures[NFSPROC4_CLNT_CB_NULL], @@ -392,6 +394,11 @@ static int do_probe_callback(void *data) struct rpc_clnt *client; int status; + if (!clp->cl_principal && (clp->cl_flavor >= RPC_AUTH_GSS_KRB5)) { + status = nfserr_cb_path_down; + goto out_err; + } + /* Initialize address */ memset(&addr, 0, sizeof(addr)); addr.sin_family = AF_INET; diff --git a/fs/nfsd/nfs4recover.c b/fs/nfsd/nfs4recover.c index 145b3c877a2..0f9d6efaa62 100644 --- a/fs/nfsd/nfs4recover.c +++ b/fs/nfsd/nfs4recover.c @@ -51,23 +51,29 @@ #define NFSDDBG_FACILITY NFSDDBG_PROC /* Globals */ -static struct nameidata rec_dir; +static struct path rec_dir; static int rec_dir_init = 0; -static void -nfs4_save_user(uid_t *saveuid, gid_t *savegid) +static int +nfs4_save_creds(const struct cred **original_creds) { - *saveuid = current->fsuid; - *savegid = current->fsgid; - current->fsuid = 0; - current->fsgid = 0; + struct cred *new; + + new = prepare_creds(); + if (!new) + return -ENOMEM; + + new->fsuid = 0; + new->fsgid = 0; + *original_creds = override_creds(new); + put_cred(new); + return 0; } static void -nfs4_reset_user(uid_t saveuid, gid_t savegid) +nfs4_reset_creds(const struct cred *original) { - current->fsuid = saveuid; - current->fsgid = savegid; + revert_creds(original); } static void @@ -121,18 +127,17 @@ out_no_tfm: static void nfsd4_sync_rec_dir(void) { - mutex_lock(&rec_dir.path.dentry->d_inode->i_mutex); - nfsd_sync_dir(rec_dir.path.dentry); - mutex_unlock(&rec_dir.path.dentry->d_inode->i_mutex); + mutex_lock(&rec_dir.dentry->d_inode->i_mutex); + nfsd_sync_dir(rec_dir.dentry); + mutex_unlock(&rec_dir.dentry->d_inode->i_mutex); } int nfsd4_create_clid_dir(struct nfs4_client *clp) { + const struct cred *original_cred; char *dname = clp->cl_recdir; struct dentry *dentry; - uid_t uid; - gid_t gid; int status; dprintk("NFSD: nfsd4_create_clid_dir for \"%s\"\n", dname); @@ -140,12 +145,14 @@ nfsd4_create_clid_dir(struct nfs4_client *clp) if (!rec_dir_init || clp->cl_firststate) return 0; - nfs4_save_user(&uid, &gid); + status = nfs4_save_creds(&original_cred); + if (status < 0) + return status; /* lock the parent */ - mutex_lock(&rec_dir.path.dentry->d_inode->i_mutex); + mutex_lock(&rec_dir.dentry->d_inode->i_mutex); - dentry = lookup_one_len(dname, rec_dir.path.dentry, HEXDIR_LEN-1); + dentry = lookup_one_len(dname, rec_dir.dentry, HEXDIR_LEN-1); if (IS_ERR(dentry)) { status = PTR_ERR(dentry); goto out_unlock; @@ -155,20 +162,20 @@ nfsd4_create_clid_dir(struct nfs4_client *clp) dprintk("NFSD: nfsd4_create_clid_dir: DIRECTORY EXISTS\n"); goto out_put; } - status = mnt_want_write(rec_dir.path.mnt); + status = mnt_want_write(rec_dir.mnt); if (status) goto out_put; - status = vfs_mkdir(rec_dir.path.dentry->d_inode, dentry, S_IRWXU); - mnt_drop_write(rec_dir.path.mnt); + status = vfs_mkdir(rec_dir.dentry->d_inode, dentry, S_IRWXU); + mnt_drop_write(rec_dir.mnt); out_put: dput(dentry); out_unlock: - mutex_unlock(&rec_dir.path.dentry->d_inode->i_mutex); + mutex_unlock(&rec_dir.dentry->d_inode->i_mutex); if (status == 0) { clp->cl_firststate = 1; nfsd4_sync_rec_dir(); } - nfs4_reset_user(uid, gid); + nfs4_reset_creds(original_cred); dprintk("NFSD: nfsd4_create_clid_dir returns %d\n", status); return status; } @@ -211,22 +218,25 @@ nfsd4_build_dentrylist(void *arg, const char *name, int namlen, static int nfsd4_list_rec_dir(struct dentry *dir, recdir_func *f) { + const struct cred *original_cred; struct file *filp; struct dentry_list_arg dla = { .parent = dir, }; struct list_head *dentries = &dla.dentries; struct dentry_list *child; - uid_t uid; - gid_t gid; int status; if (!rec_dir_init) return 0; - nfs4_save_user(&uid, &gid); + status = nfs4_save_creds(&original_cred); + if (status < 0) + return status; + INIT_LIST_HEAD(dentries); - filp = dentry_open(dget(dir), mntget(rec_dir.path.mnt), O_RDONLY); + filp = dentry_open(dget(dir), mntget(rec_dir.mnt), O_RDONLY, + current_cred()); status = PTR_ERR(filp); if (IS_ERR(filp)) goto out; @@ -249,7 +259,7 @@ out: dput(child->dentry); kfree(child); } - nfs4_reset_user(uid, gid); + nfs4_reset_creds(original_cred); return status; } @@ -291,9 +301,9 @@ nfsd4_unlink_clid_dir(char *name, int namlen) dprintk("NFSD: nfsd4_unlink_clid_dir. name %.*s\n", namlen, name); - mutex_lock(&rec_dir.path.dentry->d_inode->i_mutex); - dentry = lookup_one_len(name, rec_dir.path.dentry, namlen); - mutex_unlock(&rec_dir.path.dentry->d_inode->i_mutex); + mutex_lock(&rec_dir.dentry->d_inode->i_mutex); + dentry = lookup_one_len(name, rec_dir.dentry, namlen); + mutex_unlock(&rec_dir.dentry->d_inode->i_mutex); if (IS_ERR(dentry)) { status = PTR_ERR(dentry); return status; @@ -302,7 +312,7 @@ nfsd4_unlink_clid_dir(char *name, int namlen) if (!dentry->d_inode) goto out; - status = nfsd4_clear_clid_dir(rec_dir.path.dentry, dentry); + status = nfsd4_clear_clid_dir(rec_dir.dentry, dentry); out: dput(dentry); return status; @@ -311,23 +321,26 @@ out: void nfsd4_remove_clid_dir(struct nfs4_client *clp) { - uid_t uid; - gid_t gid; + const struct cred *original_cred; int status; if (!rec_dir_init || !clp->cl_firststate) return; - status = mnt_want_write(rec_dir.path.mnt); + status = mnt_want_write(rec_dir.mnt); if (status) goto out; clp->cl_firststate = 0; - nfs4_save_user(&uid, &gid); + + status = nfs4_save_creds(&original_cred); + if (status < 0) + goto out; + status = nfsd4_unlink_clid_dir(clp->cl_recdir, HEXDIR_LEN-1); - nfs4_reset_user(uid, gid); + nfs4_reset_creds(original_cred); if (status == 0) nfsd4_sync_rec_dir(); - mnt_drop_write(rec_dir.path.mnt); + mnt_drop_write(rec_dir.mnt); out: if (status) printk("NFSD: Failed to remove expired client state directory" @@ -357,17 +370,17 @@ nfsd4_recdir_purge_old(void) { if (!rec_dir_init) return; - status = mnt_want_write(rec_dir.path.mnt); + status = mnt_want_write(rec_dir.mnt); if (status) goto out; - status = nfsd4_list_rec_dir(rec_dir.path.dentry, purge_old); + status = nfsd4_list_rec_dir(rec_dir.dentry, purge_old); if (status == 0) nfsd4_sync_rec_dir(); - mnt_drop_write(rec_dir.path.mnt); + mnt_drop_write(rec_dir.mnt); out: if (status) printk("nfsd4: failed to purge old clients from recovery" - " directory %s\n", rec_dir.path.dentry->d_name.name); + " directory %s\n", rec_dir.dentry->d_name.name); } static int @@ -387,10 +400,10 @@ int nfsd4_recdir_load(void) { int status; - status = nfsd4_list_rec_dir(rec_dir.path.dentry, load_recdir); + status = nfsd4_list_rec_dir(rec_dir.dentry, load_recdir); if (status) printk("nfsd4: failed loading clients from recovery" - " directory %s\n", rec_dir.path.dentry->d_name.name); + " directory %s\n", rec_dir.dentry->d_name.name); return status; } @@ -401,18 +414,23 @@ nfsd4_recdir_load(void) { void nfsd4_init_recdir(char *rec_dirname) { - uid_t uid = 0; - gid_t gid = 0; - int status; + const struct cred *original_cred; + int status; printk("NFSD: Using %s as the NFSv4 state recovery directory\n", rec_dirname); BUG_ON(rec_dir_init); - nfs4_save_user(&uid, &gid); + status = nfs4_save_creds(&original_cred); + if (status < 0) { + printk("NFSD: Unable to change credentials to find recovery" + " directory: error %d\n", + status); + return; + } - status = path_lookup(rec_dirname, LOOKUP_FOLLOW | LOOKUP_DIRECTORY, + status = kern_path(rec_dirname, LOOKUP_FOLLOW | LOOKUP_DIRECTORY, &rec_dir); if (status) printk("NFSD: unable to find recovery directory %s\n", @@ -420,7 +438,7 @@ nfsd4_init_recdir(char *rec_dirname) if (!status) rec_dir_init = 1; - nfs4_reset_user(uid, gid); + nfs4_reset_creds(original_cred); } void @@ -429,5 +447,5 @@ nfsd4_shutdown_recdir(void) if (!rec_dir_init) return; rec_dir_init = 0; - path_put(&rec_dir.path); + path_put(&rec_dir); } diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 0cc7ff5d5ab..13e0e074dbb 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -54,6 +54,7 @@ #include <linux/mutex.h> #include <linux/lockd/bind.h> #include <linux/module.h> +#include <linux/sunrpc/svcauth_gss.h> #define NFSDDBG_FACILITY NFSDDBG_PROC @@ -377,6 +378,7 @@ free_client(struct nfs4_client *clp) shutdown_callback_client(clp); if (clp->cl_cred.cr_group_info) put_group_info(clp->cl_cred.cr_group_info); + kfree(clp->cl_principal); kfree(clp->cl_name.data); kfree(clp); } @@ -696,6 +698,7 @@ nfsd4_setclientid(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, unsigned int strhashval; struct nfs4_client *conf, *unconf, *new; __be32 status; + char *princ; char dname[HEXDIR_LEN]; if (!check_name(clname)) @@ -719,8 +722,8 @@ nfsd4_setclientid(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, status = nfserr_clid_inuse; if (!same_creds(&conf->cl_cred, &rqstp->rq_cred) || conf->cl_addr != sin->sin_addr.s_addr) { - dprintk("NFSD: setclientid: string in use by client" - "at %u.%u.%u.%u\n", NIPQUAD(conf->cl_addr)); + dprintk("NFSD: setclientid: string in use by clientat %pI4\n", + &conf->cl_addr); goto out; } } @@ -783,6 +786,15 @@ nfsd4_setclientid(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, } copy_verf(new, &clverifier); new->cl_addr = sin->sin_addr.s_addr; + new->cl_flavor = rqstp->rq_flavor; + princ = svc_gss_principal(rqstp); + if (princ) { + new->cl_principal = kstrdup(princ, GFP_KERNEL); + if (new->cl_principal == NULL) { + free_client(new); + goto out; + } + } copy_cred(&new->cl_cred, &rqstp->rq_cred); gen_confirm(new); gen_callback(new, setclid); @@ -3261,6 +3273,7 @@ nfs4_state_shutdown(void) { cancel_rearming_delayed_workqueue(laundry_wq, &laundromat_work); destroy_workqueue(laundry_wq); + locks_end_grace(&nfsd4_manager); nfs4_lock_state(); nfs4_release_reclaim(); __nfs4_state_shutdown(); @@ -3284,17 +3297,17 @@ int nfs4_reset_recoverydir(char *recdir) { int status; - struct nameidata nd; + struct path path; - status = path_lookup(recdir, LOOKUP_FOLLOW, &nd); + status = kern_path(recdir, LOOKUP_FOLLOW, &path); if (status) return status; status = -ENOTDIR; - if (S_ISDIR(nd.path.dentry->d_inode->i_mode)) { + if (S_ISDIR(path.dentry->d_inode->i_mode)) { nfs4_set_recdir(recdir); status = 0; } - path_put(&nd.path); + path_put(&path); return status; } diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index 97543df5824..77d7b8c531a 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -330,7 +330,7 @@ static ssize_t failover_unlock_ip(struct file *file, char *buf, size_t size) return -EINVAL; /* get ipv4 address */ - if (sscanf(fo_path, NIPQUAD_FMT "%c", &b1, &b2, &b3, &b4, &c) != 4) + if (sscanf(fo_path, "%u.%u.%u.%u%c", &b1, &b2, &b3, &b4, &c) != 4) return -EINVAL; if (b1 > 255 || b2 > 255 || b3 > 255 || b4 > 255) return -EINVAL; @@ -341,7 +341,7 @@ static ssize_t failover_unlock_ip(struct file *file, char *buf, size_t size) static ssize_t failover_unlock_fs(struct file *file, char *buf, size_t size) { - struct nameidata nd; + struct path path; char *fo_path; int error; @@ -356,13 +356,13 @@ static ssize_t failover_unlock_fs(struct file *file, char *buf, size_t size) if (qword_get(&buf, fo_path, size) < 0) return -EINVAL; - error = path_lookup(fo_path, 0, &nd); + error = kern_path(fo_path, 0, &path); if (error) return error; - error = nlmsvc_unlock_all_by_sb(nd.path.mnt->mnt_sb); + error = nlmsvc_unlock_all_by_sb(path.mnt->mnt_sb); - path_put(&nd.path); + path_put(&path); return error; } diff --git a/fs/nfsd/nfsfh.c b/fs/nfsd/nfsfh.c index cd25d91895a..f0da7d9c3a9 100644 --- a/fs/nfsd/nfsfh.c +++ b/fs/nfsd/nfsfh.c @@ -186,9 +186,14 @@ static __be32 nfsd_set_fh_dentry(struct svc_rqst *rqstp, struct svc_fh *fhp) * access control settings being in effect, we cannot * fix that case easily. */ - current->cap_effective = - cap_raise_nfsd_set(current->cap_effective, - current->cap_permitted); + struct cred *new = prepare_creds(); + if (!new) + return nfserrno(-ENOMEM); + new->cap_effective = + cap_raise_nfsd_set(new->cap_effective, + new->cap_permitted); + put_cred(override_creds(new)); + put_cred(new); } else { error = nfsd_setuser_and_check_port(rqstp, exp); if (error) diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c index 59eeb46f82c..07e4f5d7baa 100644 --- a/fs/nfsd/nfssvc.c +++ b/fs/nfsd/nfssvc.c @@ -249,6 +249,10 @@ static int nfsd_init_socks(int port) if (error < 0) return error; + error = lockd_up(); + if (error < 0) + return error; + error = svc_create_xprt(nfsd_serv, "tcp", port, SVC_SOCK_DEFAULTS); if (error < 0) diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c index aa1d0d6489a..d1c5f787b36 100644 --- a/fs/nfsd/vfs.c +++ b/fs/nfsd/vfs.c @@ -410,6 +410,7 @@ out_nfserr: static ssize_t nfsd_getxattr(struct dentry *dentry, char *key, void **buf) { ssize_t buflen; + ssize_t ret; buflen = vfs_getxattr(dentry, key, NULL, 0); if (buflen <= 0) @@ -419,7 +420,10 @@ static ssize_t nfsd_getxattr(struct dentry *dentry, char *key, void **buf) if (!*buf) return -ENOMEM; - return vfs_getxattr(dentry, key, *buf, buflen); + ret = vfs_getxattr(dentry, key, *buf, buflen); + if (ret < 0) + kfree(*buf); + return ret; } #endif @@ -667,6 +671,7 @@ __be32 nfsd_open(struct svc_rqst *rqstp, struct svc_fh *fhp, int type, int access, struct file **filp) { + const struct cred *cred = current_cred(); struct dentry *dentry; struct inode *inode; int flags = O_RDONLY|O_LARGEFILE; @@ -721,7 +726,7 @@ nfsd_open(struct svc_rqst *rqstp, struct svc_fh *fhp, int type, DQUOT_INIT(inode); } *filp = dentry_open(dget(dentry), mntget(fhp->fh_export->ex_path.mnt), - flags); + flags, cred); if (IS_ERR(*filp)) host_err = PTR_ERR(*filp); out_nfserr: @@ -1165,7 +1170,7 @@ nfsd_create_setattr(struct svc_rqst *rqstp, struct svc_fh *resfhp, * send along the gid on create when it tries to implement * setgid directories via NFS: */ - if (current->fsuid != 0) + if (current_fsuid() != 0) iap->ia_valid &= ~(ATTR_UID|ATTR_GID); if (iap->ia_valid) return nfsd_setattr(rqstp, resfhp, iap, 0, (time_t)0); @@ -1814,6 +1819,113 @@ out: } /* + * We do this buffering because we must not call back into the file + * system's ->lookup() method from the filldir callback. That may well + * deadlock a number of file systems. + * + * This is based heavily on the implementation of same in XFS. + */ +struct buffered_dirent { + u64 ino; + loff_t offset; + int namlen; + unsigned int d_type; + char name[]; +}; + +struct readdir_data { + char *dirent; + size_t used; + int full; +}; + +static int nfsd_buffered_filldir(void *__buf, const char *name, int namlen, + loff_t offset, u64 ino, unsigned int d_type) +{ + struct readdir_data *buf = __buf; + struct buffered_dirent *de = (void *)(buf->dirent + buf->used); + unsigned int reclen; + + reclen = ALIGN(sizeof(struct buffered_dirent) + namlen, sizeof(u64)); + if (buf->used + reclen > PAGE_SIZE) { + buf->full = 1; + return -EINVAL; + } + + de->namlen = namlen; + de->offset = offset; + de->ino = ino; + de->d_type = d_type; + memcpy(de->name, name, namlen); + buf->used += reclen; + + return 0; +} + +static int nfsd_buffered_readdir(struct file *file, filldir_t func, + struct readdir_cd *cdp, loff_t *offsetp) +{ + struct readdir_data buf; + struct buffered_dirent *de; + int host_err; + int size; + loff_t offset; + + buf.dirent = (void *)__get_free_page(GFP_KERNEL); + if (!buf.dirent) + return -ENOMEM; + + offset = *offsetp; + + while (1) { + unsigned int reclen; + + cdp->err = nfserr_eof; /* will be cleared on successful read */ + buf.used = 0; + buf.full = 0; + + host_err = vfs_readdir(file, nfsd_buffered_filldir, &buf); + if (buf.full) + host_err = 0; + + if (host_err < 0) + break; + + size = buf.used; + + if (!size) + break; + + de = (struct buffered_dirent *)buf.dirent; + while (size > 0) { + offset = de->offset; + + if (func(cdp, de->name, de->namlen, de->offset, + de->ino, de->d_type)) + goto done; + + if (cdp->err != nfs_ok) + goto done; + + reclen = ALIGN(sizeof(*de) + de->namlen, + sizeof(u64)); + size -= reclen; + de = (struct buffered_dirent *)((char *)de + reclen); + } + offset = vfs_llseek(file, 0, SEEK_CUR); + } + + done: + free_page((unsigned long)(buf.dirent)); + + if (host_err) + return nfserrno(host_err); + + *offsetp = offset; + return cdp->err; +} + +/* * Read entries from a directory. * The NFSv3/4 verifier we ignore for now. */ @@ -1822,7 +1934,6 @@ nfsd_readdir(struct svc_rqst *rqstp, struct svc_fh *fhp, loff_t *offsetp, struct readdir_cd *cdp, filldir_t func) { __be32 err; - int host_err; struct file *file; loff_t offset = *offsetp; @@ -1836,21 +1947,7 @@ nfsd_readdir(struct svc_rqst *rqstp, struct svc_fh *fhp, loff_t *offsetp, goto out_close; } - /* - * Read the directory entries. This silly loop is necessary because - * readdir() is not guaranteed to fill up the entire buffer, but - * may choose to do less. - */ - - do { - cdp->err = nfserr_eof; /* will be cleared on successful read */ - host_err = vfs_readdir(file, func, cdp); - } while (host_err >=0 && cdp->err == nfs_ok); - if (host_err) - err = nfserrno(host_err); - else - err = cdp->err; - *offsetp = vfs_llseek(file, 0, 1); + err = nfsd_buffered_readdir(file, func, cdp, offsetp); if (err == nfserr_eof || err == nfserr_toosmall) err = nfs_ok; /* can still be found in ->err */ @@ -1905,7 +2002,7 @@ nfsd_permission(struct svc_rqst *rqstp, struct svc_export *exp, IS_APPEND(inode)? " append" : "", __mnt_is_readonly(exp->ex_path.mnt)? " ro" : ""); dprintk(" owner %d/%d user %d/%d\n", - inode->i_uid, inode->i_gid, current->fsuid, current->fsgid); + inode->i_uid, inode->i_gid, current_fsuid(), current_fsgid()); #endif /* Normally we reject any write/sattr etc access on a read-only file @@ -1948,7 +2045,7 @@ nfsd_permission(struct svc_rqst *rqstp, struct svc_export *exp, * with NFSv3. */ if ((acc & NFSD_MAY_OWNER_OVERRIDE) && - inode->i_uid == current->fsuid) + inode->i_uid == current_fsuid()) return 0; /* This assumes NFSD_MAY_{READ,WRITE,EXEC} == MAY_{READ,WRITE,EXEC} */ diff --git a/fs/nls/nls_base.c b/fs/nls/nls_base.c index 64965e1c21c..9b0efdad891 100644 --- a/fs/nls/nls_base.c +++ b/fs/nls/nls_base.c @@ -13,9 +13,7 @@ #include <linux/nls.h> #include <linux/kernel.h> #include <linux/errno.h> -#ifdef CONFIG_KMOD #include <linux/kmod.h> -#endif #include <linux/spinlock.h> static struct nls_table default_table; @@ -215,24 +213,7 @@ static struct nls_table *find_nls(char *charset) struct nls_table *load_nls(char *charset) { - struct nls_table *nls; -#ifdef CONFIG_KMOD - int ret; -#endif - - nls = find_nls(charset); - if (nls) - return nls; - -#ifdef CONFIG_KMOD - ret = request_module("nls_%s", charset); - if (ret != 0) { - printk("Unable to load NLS charset %s\n", charset); - return NULL; - } - nls = find_nls(charset); -#endif - return nls; + return try_then_request_module(find_nls(charset), "nls_%s", charset); } void unload_nls(struct nls_table *nls) diff --git a/fs/ntfs/debug.h b/fs/ntfs/debug.h index 5e6724c1afd..2142b1c68b6 100644 --- a/fs/ntfs/debug.h +++ b/fs/ntfs/debug.h @@ -30,7 +30,8 @@ extern int debug_msgs; -#if 0 /* Fool kernel-doc since it doesn't do macros yet */ +extern void __ntfs_debug(const char *file, int line, const char *function, + const char *format, ...) __attribute__ ((format (printf, 4, 5))); /** * ntfs_debug - write a debug level message to syslog * @f: a printf format string containing the message @@ -39,11 +40,6 @@ extern int debug_msgs; * ntfs_debug() writes a DEBUG level message to the syslog but only if the * driver was compiled with -DDEBUG. Otherwise, the call turns into a NOP. */ -static void ntfs_debug(const char *f, ...); -#endif - -extern void __ntfs_debug (const char *file, int line, const char *function, - const char *format, ...) __attribute__ ((format (printf, 4, 5))); #define ntfs_debug(f, a...) \ __ntfs_debug(__FILE__, __LINE__, __func__, f, ##a) diff --git a/fs/ntfs/file.c b/fs/ntfs/file.c index d020866d423..3140a4429af 100644 --- a/fs/ntfs/file.c +++ b/fs/ntfs/file.c @@ -439,7 +439,7 @@ static inline int __ntfs_grab_cache_pages(struct address_space *mapping, pages[nr] = *cached_page; page_cache_get(*cached_page); if (unlikely(!pagevec_add(lru_pvec, *cached_page))) - __pagevec_lru_add(lru_pvec); + __pagevec_lru_add_file(lru_pvec); *cached_page = NULL; } index++; @@ -2084,7 +2084,7 @@ err_out: OSYNC_METADATA|OSYNC_DATA); } } - pagevec_lru_add(&lru_pvec); + pagevec_lru_add_file(&lru_pvec); ntfs_debug("Done. Returning %s (written 0x%lx, status %li).", written ? "written" : "status", (unsigned long)written, (long)status); diff --git a/fs/ntfs/namei.c b/fs/ntfs/namei.c index 9e8a95be7a1..2ca00153b6e 100644 --- a/fs/ntfs/namei.c +++ b/fs/ntfs/namei.c @@ -304,8 +304,6 @@ static struct dentry *ntfs_get_parent(struct dentry *child_dent) ntfs_attr_search_ctx *ctx; ATTR_RECORD *attr; FILE_NAME_ATTR *fn; - struct inode *parent_vi; - struct dentry *parent_dent; unsigned long parent_ino; int err; @@ -345,24 +343,8 @@ try_next: /* Release the search context and the mft record of the child. */ ntfs_attr_put_search_ctx(ctx); unmap_mft_record(ni); - /* Get the inode of the parent directory. */ - parent_vi = ntfs_iget(vi->i_sb, parent_ino); - if (IS_ERR(parent_vi) || unlikely(is_bad_inode(parent_vi))) { - if (!IS_ERR(parent_vi)) - iput(parent_vi); - ntfs_error(vi->i_sb, "Failed to get parent directory inode " - "0x%lx of child inode 0x%lx.", parent_ino, - vi->i_ino); - return ERR_PTR(-EACCES); - } - /* Finally get a dentry for the parent directory and return it. */ - parent_dent = d_alloc_anon(parent_vi); - if (unlikely(!parent_dent)) { - iput(parent_vi); - return ERR_PTR(-ENOMEM); - } - ntfs_debug("Done for inode 0x%lx.", vi->i_ino); - return parent_dent; + + return d_obtain_alias(ntfs_iget(vi->i_sb, parent_ino)); } static struct inode *ntfs_nfs_get_inode(struct super_block *sb, diff --git a/fs/ocfs2/buffer_head_io.c b/fs/ocfs2/buffer_head_io.c index 7e947c67246..3a178ec48d7 100644 --- a/fs/ocfs2/buffer_head_io.c +++ b/fs/ocfs2/buffer_head_io.c @@ -112,7 +112,7 @@ int ocfs2_read_blocks_sync(struct ocfs2_super *osb, u64 block, bh = bhs[i]; if (buffer_jbd(bh)) { - mlog(ML_ERROR, + mlog(ML_BH_IO, "trying to sync read a jbd " "managed bh (blocknr = %llu), skipping\n", (unsigned long long)bh->b_blocknr); @@ -147,15 +147,10 @@ int ocfs2_read_blocks_sync(struct ocfs2_super *osb, u64 block, for (i = nr; i > 0; i--) { bh = bhs[i - 1]; - if (buffer_jbd(bh)) { - mlog(ML_ERROR, - "the journal got the buffer while it was " - "locked for io! (blocknr = %llu)\n", - (unsigned long long)bh->b_blocknr); - BUG(); - } + /* No need to wait on the buffer if it's managed by JBD. */ + if (!buffer_jbd(bh)) + wait_on_buffer(bh); - wait_on_buffer(bh); if (!buffer_uptodate(bh)) { /* Status won't be cleared from here on out, * so we can safely record this and loop back @@ -251,8 +246,6 @@ int ocfs2_read_blocks(struct inode *inode, u64 block, int nr, ignore_cache = 1; } - /* XXX: Can we ever get this and *not* have the cached - * flag set? */ if (buffer_jbd(bh)) { if (ignore_cache) mlog(ML_BH_IO, "trying to sync read a jbd " diff --git a/fs/ocfs2/cluster/heartbeat.c b/fs/ocfs2/cluster/heartbeat.c index 7dce1612553..6ebaa58e2c0 100644 --- a/fs/ocfs2/cluster/heartbeat.c +++ b/fs/ocfs2/cluster/heartbeat.c @@ -976,7 +976,7 @@ static void o2hb_region_release(struct config_item *item) } if (reg->hr_bdev) - blkdev_put(reg->hr_bdev); + blkdev_put(reg->hr_bdev, FMODE_READ|FMODE_WRITE); if (reg->hr_slots) kfree(reg->hr_slots); @@ -1268,7 +1268,7 @@ static ssize_t o2hb_region_dev_write(struct o2hb_region *reg, goto out; reg->hr_bdev = I_BDEV(filp->f_mapping->host); - ret = blkdev_get(reg->hr_bdev, FMODE_WRITE | FMODE_READ, 0); + ret = blkdev_get(reg->hr_bdev, FMODE_WRITE | FMODE_READ); if (ret) { reg->hr_bdev = NULL; goto out; @@ -1358,7 +1358,7 @@ out: iput(inode); if (ret < 0) { if (reg->hr_bdev) { - blkdev_put(reg->hr_bdev); + blkdev_put(reg->hr_bdev, FMODE_READ|FMODE_WRITE); reg->hr_bdev = NULL; } } diff --git a/fs/ocfs2/cluster/netdebug.c b/fs/ocfs2/cluster/netdebug.c index 52276c02f71..f8424874fa0 100644 --- a/fs/ocfs2/cluster/netdebug.c +++ b/fs/ocfs2/cluster/netdebug.c @@ -304,8 +304,8 @@ static int sc_seq_show(struct seq_file *seq, void *v) * use of it here generates a warning with -Wbitwise */ seq_printf(seq, "%p:\n" " krefs: %d\n" - " sock: %u.%u.%u.%u:%u -> " - "%u.%u.%u.%u:%u\n" + " sock: %pI4:%u -> " + "%pI4:%u\n" " remote node: %s\n" " page off: %zu\n" " handshake ok: %u\n" @@ -319,8 +319,8 @@ static int sc_seq_show(struct seq_file *seq, void *v) " func type: %u\n", sc, atomic_read(&sc->sc_kref.refcount), - NIPQUAD(saddr), inet ? ntohs(sport) : 0, - NIPQUAD(daddr), inet ? ntohs(dport) : 0, + &saddr, inet ? ntohs(sport) : 0, + &daddr, inet ? ntohs(dport) : 0, sc->sc_node->nd_name, sc->sc_page_off, sc->sc_handshake_ok, diff --git a/fs/ocfs2/cluster/nodemanager.c b/fs/ocfs2/cluster/nodemanager.c index 816a3f61330..70e8fa9e253 100644 --- a/fs/ocfs2/cluster/nodemanager.c +++ b/fs/ocfs2/cluster/nodemanager.c @@ -250,7 +250,7 @@ static ssize_t o2nm_node_ipv4_port_write(struct o2nm_node *node, static ssize_t o2nm_node_ipv4_address_read(struct o2nm_node *node, char *page) { - return sprintf(page, "%u.%u.%u.%u\n", NIPQUAD(node->nd_ipv4_address)); + return sprintf(page, "%pI4\n", &node->nd_ipv4_address); } static ssize_t o2nm_node_ipv4_address_write(struct o2nm_node *node, diff --git a/fs/ocfs2/cluster/tcp.c b/fs/ocfs2/cluster/tcp.c index 2bcf706d9dd..9fbe849f634 100644 --- a/fs/ocfs2/cluster/tcp.c +++ b/fs/ocfs2/cluster/tcp.c @@ -1597,8 +1597,8 @@ static void o2net_start_connect(struct work_struct *work) ret = sock->ops->bind(sock, (struct sockaddr *)&myaddr, sizeof(myaddr)); if (ret) { - mlog(ML_ERROR, "bind failed with %d at address %u.%u.%u.%u\n", - ret, NIPQUAD(mynode->nd_ipv4_address)); + mlog(ML_ERROR, "bind failed with %d at address %pI4\n", + ret, &mynode->nd_ipv4_address); goto out; } @@ -1790,17 +1790,16 @@ static int o2net_accept_one(struct socket *sock) node = o2nm_get_node_by_ip(sin.sin_addr.s_addr); if (node == NULL) { - mlog(ML_NOTICE, "attempt to connect from unknown node at " - "%u.%u.%u.%u:%d\n", NIPQUAD(sin.sin_addr.s_addr), - ntohs(sin.sin_port)); + mlog(ML_NOTICE, "attempt to connect from unknown node at %pI4:%d\n", + &sin.sin_addr.s_addr, ntohs(sin.sin_port)); ret = -EINVAL; goto out; } if (o2nm_this_node() > node->nd_num) { mlog(ML_NOTICE, "unexpected connect attempted from a lower " - "numbered node '%s' at " "%u.%u.%u.%u:%d with num %u\n", - node->nd_name, NIPQUAD(sin.sin_addr.s_addr), + "numbered node '%s' at " "%pI4:%d with num %u\n", + node->nd_name, &sin.sin_addr.s_addr, ntohs(sin.sin_port), node->nd_num); ret = -EINVAL; goto out; @@ -1810,8 +1809,8 @@ static int o2net_accept_one(struct socket *sock) * and tries to connect before we see their heartbeat */ if (!o2hb_check_node_heartbeating_from_callback(node->nd_num)) { mlog(ML_CONN, "attempt to connect from node '%s' at " - "%u.%u.%u.%u:%d but it isn't heartbeating\n", - node->nd_name, NIPQUAD(sin.sin_addr.s_addr), + "%pI4:%d but it isn't heartbeating\n", + node->nd_name, &sin.sin_addr.s_addr, ntohs(sin.sin_port)); ret = -EINVAL; goto out; @@ -1827,8 +1826,8 @@ static int o2net_accept_one(struct socket *sock) spin_unlock(&nn->nn_lock); if (ret) { mlog(ML_NOTICE, "attempt to connect from node '%s' at " - "%u.%u.%u.%u:%d but it already has an open connection\n", - node->nd_name, NIPQUAD(sin.sin_addr.s_addr), + "%pI4:%d but it already has an open connection\n", + node->nd_name, &sin.sin_addr.s_addr, ntohs(sin.sin_port)); goto out; } @@ -1924,15 +1923,15 @@ static int o2net_open_listening_sock(__be32 addr, __be16 port) sock->sk->sk_reuse = 1; ret = sock->ops->bind(sock, (struct sockaddr *)&sin, sizeof(sin)); if (ret < 0) { - mlog(ML_ERROR, "unable to bind socket at %u.%u.%u.%u:%u, " - "ret=%d\n", NIPQUAD(addr), ntohs(port), ret); + mlog(ML_ERROR, "unable to bind socket at %pI4:%u, " + "ret=%d\n", &addr, ntohs(port), ret); goto out; } ret = sock->ops->listen(sock, 64); if (ret < 0) { - mlog(ML_ERROR, "unable to listen on %u.%u.%u.%u:%u, ret=%d\n", - NIPQUAD(addr), ntohs(port), ret); + mlog(ML_ERROR, "unable to listen on %pI4:%u, ret=%d\n", + &addr, ntohs(port), ret); } out: diff --git a/fs/ocfs2/dlm/dlmfs.c b/fs/ocfs2/dlm/dlmfs.c index 533a789c3ef..6f7a77d5402 100644 --- a/fs/ocfs2/dlm/dlmfs.c +++ b/fs/ocfs2/dlm/dlmfs.c @@ -339,8 +339,8 @@ static struct inode *dlmfs_get_root_inode(struct super_block *sb) ip = DLMFS_I(inode); inode->i_mode = mode; - inode->i_uid = current->fsuid; - inode->i_gid = current->fsgid; + inode->i_uid = current_fsuid(); + inode->i_gid = current_fsgid(); inode->i_blocks = 0; inode->i_mapping->backing_dev_info = &dlmfs_backing_dev_info; inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME; @@ -365,8 +365,8 @@ static struct inode *dlmfs_get_inode(struct inode *parent, return NULL; inode->i_mode = mode; - inode->i_uid = current->fsuid; - inode->i_gid = current->fsgid; + inode->i_uid = current_fsuid(); + inode->i_gid = current_fsgid(); inode->i_blocks = 0; inode->i_mapping->backing_dev_info = &dlmfs_backing_dev_info; inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME; @@ -608,8 +608,10 @@ static int __init init_dlmfs_fs(void) 0, (SLAB_HWCACHE_ALIGN|SLAB_RECLAIM_ACCOUNT| SLAB_MEM_SPREAD), dlmfs_init_once); - if (!dlmfs_inode_cache) + if (!dlmfs_inode_cache) { + status = -ENOMEM; goto bail; + } cleanup_inode = 1; user_dlm_worker = create_singlethread_workqueue("user_dlm"); diff --git a/fs/ocfs2/dlm/userdlm.h b/fs/ocfs2/dlm/userdlm.h index 39ec2773849..0c3cc03c61f 100644 --- a/fs/ocfs2/dlm/userdlm.h +++ b/fs/ocfs2/dlm/userdlm.h @@ -33,7 +33,7 @@ #include <linux/workqueue.h> /* user_lock_res->l_flags flags. */ -#define USER_LOCK_ATTACHED (0x00000001) /* have we initialized +#define USER_LOCK_ATTACHED (0x00000001) /* we have initialized * the lvb */ #define USER_LOCK_BUSY (0x00000002) /* we are currently in * dlm_lock */ diff --git a/fs/ocfs2/dlmglue.c b/fs/ocfs2/dlmglue.c index ec684426034..6e6cc0a2e5f 100644 --- a/fs/ocfs2/dlmglue.c +++ b/fs/ocfs2/dlmglue.c @@ -2841,9 +2841,8 @@ static void ocfs2_unlock_ast(void *opaque, int error) lockres_clear_flags(lockres, OCFS2_LOCK_BUSY); lockres->l_unlock_action = OCFS2_UNLOCK_INVALID; - spin_unlock_irqrestore(&lockres->l_lock, flags); - wake_up(&lockres->l_event); + spin_unlock_irqrestore(&lockres->l_lock, flags); mlog_exit_void(); } diff --git a/fs/ocfs2/export.c b/fs/ocfs2/export.c index 67527cebf21..2f27b332d8b 100644 --- a/fs/ocfs2/export.c +++ b/fs/ocfs2/export.c @@ -68,14 +68,9 @@ static struct dentry *ocfs2_get_dentry(struct super_block *sb, return ERR_PTR(-ESTALE); } - result = d_alloc_anon(inode); - - if (!result) { - iput(inode); - mlog_errno(-ENOMEM); - return ERR_PTR(-ENOMEM); - } - result->d_op = &ocfs2_dentry_ops; + result = d_obtain_alias(inode); + if (!IS_ERR(result)) + result->d_op = &ocfs2_dentry_ops; mlog_exit_ptr(result); return result; @@ -86,7 +81,6 @@ static struct dentry *ocfs2_get_parent(struct dentry *child) int status; u64 blkno; struct dentry *parent; - struct inode *inode; struct inode *dir = child->d_inode; mlog_entry("(0x%p, '%.*s')\n", child, @@ -109,21 +103,9 @@ static struct dentry *ocfs2_get_parent(struct dentry *child) goto bail_unlock; } - inode = ocfs2_iget(OCFS2_SB(dir->i_sb), blkno, 0, 0); - if (IS_ERR(inode)) { - mlog(ML_ERROR, "Unable to create inode %llu\n", - (unsigned long long)blkno); - parent = ERR_PTR(-EACCES); - goto bail_unlock; - } - - parent = d_alloc_anon(inode); - if (!parent) { - iput(inode); - parent = ERR_PTR(-ENOMEM); - } - - parent->d_op = &ocfs2_dentry_ops; + parent = d_obtain_alias(ocfs2_iget(OCFS2_SB(dir->i_sb), blkno, 0, 0)); + if (!IS_ERR(parent)) + parent->d_op = &ocfs2_dentry_ops; bail_unlock: ocfs2_inode_unlock(dir, 0); diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c index 8d3225a7807..e2570a3bc2b 100644 --- a/fs/ocfs2/file.c +++ b/fs/ocfs2/file.c @@ -247,8 +247,8 @@ int ocfs2_update_inode_atime(struct inode *inode, mlog_entry_void(); handle = ocfs2_start_trans(osb, OCFS2_INODE_UPDATE_CREDITS); - if (handle == NULL) { - ret = -ENOMEM; + if (IS_ERR(handle)) { + ret = PTR_ERR(handle); mlog_errno(ret); goto out; } @@ -312,8 +312,8 @@ static int ocfs2_simple_size_update(struct inode *inode, handle_t *handle = NULL; handle = ocfs2_start_trans(osb, OCFS2_INODE_UPDATE_CREDITS); - if (handle == NULL) { - ret = -ENOMEM; + if (IS_ERR(handle)) { + ret = PTR_ERR(handle); mlog_errno(ret); goto out; } @@ -679,8 +679,7 @@ leave: /* Some parts of this taken from generic_cont_expand, which turned out * to be too fragile to do exactly what we need without us having to - * worry about recursive locking in ->prepare_write() and - * ->commit_write(). */ + * worry about recursive locking in ->write_begin() and ->write_end(). */ static int ocfs2_write_zero_page(struct inode *inode, u64 size) { @@ -1056,8 +1055,8 @@ static int __ocfs2_write_remove_suid(struct inode *inode, (unsigned long long)OCFS2_I(inode)->ip_blkno, inode->i_mode); handle = ocfs2_start_trans(osb, OCFS2_INODE_UPDATE_CREDITS); - if (handle == NULL) { - ret = -ENOMEM; + if (IS_ERR(handle)) { + ret = PTR_ERR(handle); mlog_errno(ret); goto out; } @@ -1260,8 +1259,8 @@ static int __ocfs2_remove_inode_range(struct inode *inode, } handle = ocfs2_start_trans(osb, OCFS2_REMOVE_EXTENT_CREDITS); - if (handle == NULL) { - ret = -ENOMEM; + if (IS_ERR(handle)) { + ret = PTR_ERR(handle); mlog_errno(ret); goto out; } @@ -1353,8 +1352,8 @@ static int ocfs2_zero_partial_clusters(struct inode *inode, goto out; handle = ocfs2_start_trans(osb, OCFS2_INODE_UPDATE_CREDITS); - if (handle == NULL) { - ret = -ENOMEM; + if (IS_ERR(handle)) { + ret = PTR_ERR(handle); mlog_errno(ret); goto out; } @@ -1867,6 +1866,13 @@ relock: written = generic_file_direct_write(iocb, iov, &nr_segs, *ppos, ppos, count, ocount); if (written < 0) { + /* + * direct write may have instantiated a few + * blocks outside i_size. Trim these off again. + * Don't need i_size_read because we hold i_mutex. + */ + if (*ppos + count > inode->i_size) + vmtruncate(inode, inode->i_size); ret = written; goto out_dio; } diff --git a/fs/ocfs2/inode.c b/fs/ocfs2/inode.c index 4903688f72a..7aa00d51187 100644 --- a/fs/ocfs2/inode.c +++ b/fs/ocfs2/inode.c @@ -1106,6 +1106,12 @@ void ocfs2_clear_inode(struct inode *inode) oi->ip_last_trans = 0; oi->ip_dir_start_lookup = 0; oi->ip_blkno = 0ULL; + + /* + * ip_jinode is used to track txns against this inode. We ensure that + * the journal is flushed before journal shutdown. Thus it is safe to + * have inodes get cleaned up after journal shutdown. + */ jbd2_journal_release_jbd_inode(OCFS2_SB(inode->i_sb)->journal->j_journal, &oi->ip_jinode); diff --git a/fs/ocfs2/journal.c b/fs/ocfs2/journal.c index 81e40677eec..99fe9d584f3 100644 --- a/fs/ocfs2/journal.c +++ b/fs/ocfs2/journal.c @@ -690,6 +690,7 @@ void ocfs2_journal_shutdown(struct ocfs2_super *osb) /* Shutdown the kernel journal system */ jbd2_journal_destroy(journal->j_journal); + journal->j_journal = NULL; OCFS2_I(inode)->ip_open_count--; diff --git a/fs/ocfs2/mmap.c b/fs/ocfs2/mmap.c index 3dc18d67557..eea1d24713e 100644 --- a/fs/ocfs2/mmap.c +++ b/fs/ocfs2/mmap.c @@ -113,7 +113,11 @@ static int __ocfs2_page_mkwrite(struct inode *inode, struct buffer_head *di_bh, * ocfs2_write_begin_nolock(). */ if (!PageUptodate(page) || page->mapping != inode->i_mapping) { - ret = -EINVAL; + /* + * the page has been umapped in ocfs2_data_downconvert_worker. + * So return 0 here and let VFS retry. + */ + ret = 0; goto out; } diff --git a/fs/ocfs2/namei.c b/fs/ocfs2/namei.c index 485a6aa0ad3..2545e7402ef 100644 --- a/fs/ocfs2/namei.c +++ b/fs/ocfs2/namei.c @@ -378,8 +378,8 @@ static int ocfs2_mknod_locked(struct ocfs2_super *osb, } inode = new_inode(dir->i_sb); - if (IS_ERR(inode)) { - status = PTR_ERR(inode); + if (!inode) { + status = -ENOMEM; mlog(ML_ERROR, "new_inode failed!\n"); goto leave; } @@ -421,13 +421,13 @@ static int ocfs2_mknod_locked(struct ocfs2_super *osb, fe->i_blkno = cpu_to_le64(fe_blkno); fe->i_suballoc_bit = cpu_to_le16(suballoc_bit); fe->i_suballoc_slot = cpu_to_le16(inode_ac->ac_alloc_slot); - fe->i_uid = cpu_to_le32(current->fsuid); + fe->i_uid = cpu_to_le32(current_fsuid()); if (dir->i_mode & S_ISGID) { fe->i_gid = cpu_to_le32(dir->i_gid); if (S_ISDIR(mode)) mode |= S_ISGID; } else - fe->i_gid = cpu_to_le32(current->fsgid); + fe->i_gid = cpu_to_le32(current_fsgid()); fe->i_mode = cpu_to_le16(mode); if (S_ISCHR(mode) || S_ISBLK(mode)) fe->id1.dev1.i_rdev = cpu_to_le64(huge_encode_dev(dev)); @@ -491,8 +491,10 @@ leave: brelse(*new_fe_bh); *new_fe_bh = NULL; } - if (inode) + if (inode) { + clear_nlink(inode); iput(inode); + } } mlog_exit(status); diff --git a/fs/ocfs2/ocfs2.h b/fs/ocfs2/ocfs2.h index a21a465490c..3fed9e3d899 100644 --- a/fs/ocfs2/ocfs2.h +++ b/fs/ocfs2/ocfs2.h @@ -85,7 +85,7 @@ enum ocfs2_unlock_action { }; /* ocfs2_lock_res->l_flags flags. */ -#define OCFS2_LOCK_ATTACHED (0x00000001) /* have we initialized +#define OCFS2_LOCK_ATTACHED (0x00000001) /* we have initialized * the lvb */ #define OCFS2_LOCK_BUSY (0x00000002) /* we are currently in * dlm_lock */ @@ -473,6 +473,9 @@ static inline int ocfs2_uses_extended_slot_map(struct ocfs2_super *osb) (____gd)->bg_signature); \ } while (0) +#define OCFS2_IS_VALID_XATTR_BLOCK(ptr) \ + (!strcmp((ptr)->xb_signature, OCFS2_XATTR_BLOCK_SIGNATURE)) + static inline unsigned long ino_from_blkno(struct super_block *sb, u64 blkno) { diff --git a/fs/ocfs2/ocfs2_fs.h b/fs/ocfs2/ocfs2_fs.h index f24ce3d3f95..5e0c0d0aef7 100644 --- a/fs/ocfs2/ocfs2_fs.h +++ b/fs/ocfs2/ocfs2_fs.h @@ -86,7 +86,8 @@ #define OCFS2_CLEAR_INCOMPAT_FEATURE(sb,mask) \ OCFS2_SB(sb)->s_feature_incompat &= ~(mask) -#define OCFS2_FEATURE_COMPAT_SUPP OCFS2_FEATURE_COMPAT_BACKUP_SB +#define OCFS2_FEATURE_COMPAT_SUPP (OCFS2_FEATURE_COMPAT_BACKUP_SB \ + | OCFS2_FEATURE_COMPAT_JBD2_SB) #define OCFS2_FEATURE_INCOMPAT_SUPP (OCFS2_FEATURE_INCOMPAT_LOCAL_MOUNT \ | OCFS2_FEATURE_INCOMPAT_SPARSE_ALLOC \ | OCFS2_FEATURE_INCOMPAT_INLINE_DATA \ @@ -153,6 +154,11 @@ #define OCFS2_FEATURE_COMPAT_BACKUP_SB 0x0001 /* + * The filesystem will correctly handle journal feature bits. + */ +#define OCFS2_FEATURE_COMPAT_JBD2_SB 0x0002 + +/* * Unwritten extents support. */ #define OCFS2_FEATURE_RO_COMPAT_UNWRITTEN 0x0001 @@ -742,12 +748,12 @@ struct ocfs2_group_desc */ struct ocfs2_xattr_entry { __le32 xe_name_hash; /* hash value of xattr prefix+suffix. */ - __le16 xe_name_offset; /* byte offset from the 1st etnry in the local + __le16 xe_name_offset; /* byte offset from the 1st entry in the local xattr storage(inode, xattr block or xattr bucket). */ __u8 xe_name_len; /* xattr name len, does't include prefix. */ - __u8 xe_type; /* the low 7 bits indicates the name prefix's - * type and the highest 1 bits indicate whether + __u8 xe_type; /* the low 7 bits indicate the name prefix + * type and the highest bit indicates whether * the EA is stored in the local storage. */ __le64 xe_value_size; /* real xattr value length. */ }; @@ -766,9 +772,10 @@ struct ocfs2_xattr_header { xattr. */ __le16 xh_name_value_len; /* total length of name/value length in this bucket. */ - __le16 xh_num_buckets; /* bucket nums in one extent - record, only valid in the - first bucket. */ + __le16 xh_num_buckets; /* Number of xattr buckets + in this extent record, + only valid in the first + bucket. */ __le64 xh_csum; struct ocfs2_xattr_entry xh_entries[0]; /* xattr entry list. */ }; @@ -776,8 +783,8 @@ struct ocfs2_xattr_header { /* * On disk structure for xattr value root. * - * It is used when one extended attribute's size is larger, and we will save it - * in an outside cluster. It will stored in a b-tree like file content. + * When an xattr's value is large enough, it is stored in an external + * b-tree like file data. The xattr value root points to this structure. */ struct ocfs2_xattr_value_root { /*00*/ __le32 xr_clusters; /* clusters covered by xattr value. */ diff --git a/fs/ocfs2/stack_user.c b/fs/ocfs2/stack_user.c index faec2d87935..9b76d41a8ac 100644 --- a/fs/ocfs2/stack_user.c +++ b/fs/ocfs2/stack_user.c @@ -740,6 +740,9 @@ static int user_dlm_lock_status(union ocfs2_dlm_lksb *lksb) static void *user_dlm_lvb(union ocfs2_dlm_lksb *lksb) { + if (!lksb->lksb_fsdlm.sb_lvbptr) + lksb->lksb_fsdlm.sb_lvbptr = (char *)lksb + + sizeof(struct dlm_lksb); return (void *)(lksb->lksb_fsdlm.sb_lvbptr); } diff --git a/fs/ocfs2/xattr.c b/fs/ocfs2/xattr.c index c25780a70df..74d7367ade1 100644 --- a/fs/ocfs2/xattr.c +++ b/fs/ocfs2/xattr.c @@ -3,25 +3,20 @@ * * xattr.c * - * Copyright (C) 2008 Oracle. All rights reserved. + * Copyright (C) 2004, 2008 Oracle. All rights reserved. * * CREDITS: - * Lots of code in this file is taken from ext3. + * Lots of code in this file is copy from linux/fs/ext3/xattr.c. + * Copyright (C) 2001-2003 Andreas Gruenbacher, <agruen@suse.de> * * This program is free software; you can redistribute it and/or * modify it under the terms of the GNU General Public - * License as published by the Free Software Foundation; either - * version 2 of the License, or (at your option) any later version. + * License version 2 as published by the Free Software Foundation. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU * General Public License for more details. - * - * You should have received a copy of the GNU General Public - * License along with this program; if not, write to the - * Free Software Foundation, Inc., 59 Temple Place - Suite 330, - * Boston, MA 021110-1307, USA. */ #include <linux/capability.h> @@ -83,7 +78,7 @@ struct xattr_handler *ocfs2_xattr_handlers[] = { NULL }; -static struct xattr_handler *ocfs2_xattr_handler_map[] = { +static struct xattr_handler *ocfs2_xattr_handler_map[OCFS2_XATTR_MAX] = { [OCFS2_XATTR_INDEX_USER] = &ocfs2_xattr_user_handler, [OCFS2_XATTR_INDEX_TRUSTED] = &ocfs2_xattr_trusted_handler, }; @@ -116,6 +111,10 @@ static int ocfs2_xattr_bucket_get_name_value(struct inode *inode, int *block_off, int *new_offset); +static int ocfs2_xattr_block_find(struct inode *inode, + int name_index, + const char *name, + struct ocfs2_xattr_search *xs); static int ocfs2_xattr_index_block_find(struct inode *inode, struct buffer_head *root_bh, int name_index, @@ -137,6 +136,24 @@ static int ocfs2_xattr_set_entry_index_block(struct inode *inode, static int ocfs2_delete_xattr_index_block(struct inode *inode, struct buffer_head *xb_bh); +static inline u16 ocfs2_xattr_buckets_per_cluster(struct ocfs2_super *osb) +{ + return (1 << osb->s_clustersize_bits) / OCFS2_XATTR_BUCKET_SIZE; +} + +static inline u16 ocfs2_blocks_per_xattr_bucket(struct super_block *sb) +{ + return OCFS2_XATTR_BUCKET_SIZE / (1 << sb->s_blocksize_bits); +} + +static inline u16 ocfs2_xattr_max_xe_in_bucket(struct super_block *sb) +{ + u16 len = sb->s_blocksize - + offsetof(struct ocfs2_xattr_header, xh_entries); + + return len / sizeof(struct ocfs2_xattr_entry); +} + static inline const char *ocfs2_xattr_prefix(int name_index) { struct xattr_handler *handler = NULL; @@ -542,14 +559,12 @@ static int ocfs2_xattr_block_list(struct inode *inode, mlog_errno(ret); return ret; } - /*Verify the signature of xattr block*/ - if (memcmp((void *)blk_bh->b_data, OCFS2_XATTR_BLOCK_SIGNATURE, - strlen(OCFS2_XATTR_BLOCK_SIGNATURE))) { - ret = -EFAULT; - goto cleanup; - } xb = (struct ocfs2_xattr_block *)blk_bh->b_data; + if (!OCFS2_IS_VALID_XATTR_BLOCK(xb)) { + ret = -EIO; + goto cleanup; + } if (!(le16_to_cpu(xb->xb_flags) & OCFS2_XATTR_INDEXED)) { struct ocfs2_xattr_header *header = &xb->xb_attrs.xb_header; @@ -749,47 +764,25 @@ static int ocfs2_xattr_block_get(struct inode *inode, size_t buffer_size, struct ocfs2_xattr_search *xs) { - struct ocfs2_dinode *di = (struct ocfs2_dinode *)xs->inode_bh->b_data; - struct buffer_head *blk_bh = NULL; struct ocfs2_xattr_block *xb; struct ocfs2_xattr_value_root *xv; size_t size; int ret = -ENODATA, name_offset, name_len, block_off, i; - if (!di->i_xattr_loc) - return ret; - memset(&xs->bucket, 0, sizeof(xs->bucket)); - ret = ocfs2_read_block(inode, le64_to_cpu(di->i_xattr_loc), &blk_bh); - if (ret < 0) { + ret = ocfs2_xattr_block_find(inode, name_index, name, xs); + if (ret) { mlog_errno(ret); - return ret; - } - /*Verify the signature of xattr block*/ - if (memcmp((void *)blk_bh->b_data, OCFS2_XATTR_BLOCK_SIGNATURE, - strlen(OCFS2_XATTR_BLOCK_SIGNATURE))) { - ret = -EFAULT; goto cleanup; } - xs->xattr_bh = blk_bh; - xb = (struct ocfs2_xattr_block *)blk_bh->b_data; - - if (!(le16_to_cpu(xb->xb_flags) & OCFS2_XATTR_INDEXED)) { - xs->header = &xb->xb_attrs.xb_header; - xs->base = (void *)xs->header; - xs->end = (void *)(blk_bh->b_data) + blk_bh->b_size; - xs->here = xs->header->xh_entries; - - ret = ocfs2_xattr_find_entry(name_index, name, xs); - } else - ret = ocfs2_xattr_index_block_find(inode, blk_bh, - name_index, - name, xs); - - if (ret) + if (xs->not_found) { + ret = -ENODATA; goto cleanup; + } + + xb = (struct ocfs2_xattr_block *)xs->xattr_bh->b_data; size = le64_to_cpu(xs->here->xe_value_size); if (buffer) { ret = -ERANGE; @@ -828,7 +821,8 @@ cleanup: brelse(xs->bucket.bhs[i]); memset(&xs->bucket, 0, sizeof(xs->bucket)); - brelse(blk_bh); + brelse(xs->xattr_bh); + xs->xattr_bh = NULL; return ret; } @@ -837,11 +831,11 @@ cleanup: * Copy an extended attribute into the buffer provided. * Buffer is NULL to compute the size of buffer required. */ -int ocfs2_xattr_get(struct inode *inode, - int name_index, - const char *name, - void *buffer, - size_t buffer_size) +static int ocfs2_xattr_get(struct inode *inode, + int name_index, + const char *name, + void *buffer, + size_t buffer_size) { int ret; struct ocfs2_dinode *di = NULL; @@ -871,7 +865,7 @@ int ocfs2_xattr_get(struct inode *inode, down_read(&oi->ip_xattr_sem); ret = ocfs2_xattr_ibody_get(inode, name_index, name, buffer, buffer_size, &xis); - if (ret == -ENODATA) + if (ret == -ENODATA && di->i_xattr_loc) ret = ocfs2_xattr_block_get(inode, name_index, name, buffer, buffer_size, &xbs); up_read(&oi->ip_xattr_sem); @@ -1229,7 +1223,7 @@ static int ocfs2_xattr_set_entry(struct inode *inode, free = min_offs - ((void *)last - xs->base) - sizeof(__u32); if (free < 0) - return -EFAULT; + return -EIO; if (!xs->not_found) { size_t size = 0; @@ -1514,10 +1508,9 @@ static int ocfs2_xattr_free_block(struct inode *inode, goto out; } - /*Verify the signature of xattr block*/ - if (memcmp((void *)blk_bh->b_data, OCFS2_XATTR_BLOCK_SIGNATURE, - strlen(OCFS2_XATTR_BLOCK_SIGNATURE))) { - ret = -EFAULT; + xb = (struct ocfs2_xattr_block *)blk_bh->b_data; + if (!OCFS2_IS_VALID_XATTR_BLOCK(xb)) { + ret = -EIO; goto out; } @@ -1527,7 +1520,6 @@ static int ocfs2_xattr_free_block(struct inode *inode, goto out; } - xb = (struct ocfs2_xattr_block *)blk_bh->b_data; blk = le64_to_cpu(xb->xb_blkno); bit = le16_to_cpu(xb->xb_suballoc_bit); bg_blkno = ocfs2_which_suballoc_group(blk, bit); @@ -1771,15 +1763,14 @@ static int ocfs2_xattr_block_find(struct inode *inode, mlog_errno(ret); return ret; } - /*Verify the signature of xattr block*/ - if (memcmp((void *)blk_bh->b_data, OCFS2_XATTR_BLOCK_SIGNATURE, - strlen(OCFS2_XATTR_BLOCK_SIGNATURE))) { - ret = -EFAULT; - goto cleanup; + + xb = (struct ocfs2_xattr_block *)blk_bh->b_data; + if (!OCFS2_IS_VALID_XATTR_BLOCK(xb)) { + ret = -EIO; + goto cleanup; } xs->xattr_bh = blk_bh; - xb = (struct ocfs2_xattr_block *)blk_bh->b_data; if (!(le16_to_cpu(xb->xb_flags) & OCFS2_XATTR_INDEXED)) { xs->header = &xb->xb_attrs.xb_header; @@ -1806,52 +1797,6 @@ cleanup: } /* - * When all the xattrs are deleted from index btree, the ocfs2_xattr_tree - * will be erased and ocfs2_xattr_block will have its ocfs2_xattr_header - * re-initialized. - */ -static int ocfs2_restore_xattr_block(struct inode *inode, - struct ocfs2_xattr_search *xs) -{ - int ret; - handle_t *handle; - struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); - struct ocfs2_xattr_block *xb = - (struct ocfs2_xattr_block *)xs->xattr_bh->b_data; - struct ocfs2_extent_list *el = &xb->xb_attrs.xb_root.xt_list; - u16 xb_flags = le16_to_cpu(xb->xb_flags); - - BUG_ON(!(xb_flags & OCFS2_XATTR_INDEXED) || - le16_to_cpu(el->l_next_free_rec) != 0); - - handle = ocfs2_start_trans(osb, OCFS2_XATTR_BLOCK_UPDATE_CREDITS); - if (IS_ERR(handle)) { - ret = PTR_ERR(handle); - handle = NULL; - goto out; - } - - ret = ocfs2_journal_access(handle, inode, xs->xattr_bh, - OCFS2_JOURNAL_ACCESS_WRITE); - if (ret < 0) { - mlog_errno(ret); - goto out_commit; - } - - memset(&xb->xb_attrs, 0, inode->i_sb->s_blocksize - - offsetof(struct ocfs2_xattr_block, xb_attrs)); - - xb->xb_flags = cpu_to_le16(xb_flags & ~OCFS2_XATTR_INDEXED); - - ocfs2_journal_dirty(handle, xs->xattr_bh); - -out_commit: - ocfs2_commit_trans(osb, handle); -out: - return ret; -} - -/* * ocfs2_xattr_block_set() * * Set, replace or remove an extended attribute into external block. @@ -1961,8 +1906,6 @@ out: } ret = ocfs2_xattr_set_entry_index_block(inode, xi, xs); - if (!ret && xblk->xb_attrs.xb_root.xt_list.l_next_free_rec == 0) - ret = ocfs2_restore_xattr_block(inode, xs); end: @@ -2349,7 +2292,7 @@ static int ocfs2_xattr_bucket_find(struct inode *inode, */ ret = ocfs2_read_blocks(inode, xs->bucket.bhs[0]->b_blocknr + 1, blk_per_bucket - 1, &xs->bucket.bhs[1], - OCFS2_BH_CACHED); + 0); if (ret) { mlog_errno(ret); goto out; @@ -2398,7 +2341,8 @@ static int ocfs2_xattr_index_block_find(struct inode *inode, BUG_ON(p_blkno == 0 || num_clusters == 0 || first_hash > name_hash); mlog(0, "find xattr extent rec %u clusters from %llu, the first hash " - "in the rec is %u\n", num_clusters, p_blkno, first_hash); + "in the rec is %u\n", num_clusters, (unsigned long long)p_blkno, + first_hash); ret = ocfs2_xattr_bucket_find(inode, name_index, name, name_hash, p_blkno, first_hash, num_clusters, xs); @@ -2422,11 +2366,11 @@ static int ocfs2_iterate_xattr_buckets(struct inode *inode, memset(&bucket, 0, sizeof(bucket)); mlog(0, "iterating xattr buckets in %u clusters starting from %llu\n", - clusters, blkno); + clusters, (unsigned long long)blkno); for (i = 0; i < num_buckets; i++, blkno += blk_per_bucket) { ret = ocfs2_read_blocks(inode, blkno, blk_per_bucket, - bucket.bhs, OCFS2_BH_CACHED); + bucket.bhs, 0); if (ret) { mlog_errno(ret); goto out; @@ -2440,7 +2384,8 @@ static int ocfs2_iterate_xattr_buckets(struct inode *inode, if (i == 0) num_buckets = le16_to_cpu(bucket.xh->xh_num_buckets); - mlog(0, "iterating xattr bucket %llu, first hash %u\n", blkno, + mlog(0, "iterating xattr bucket %llu, first hash %u\n", + (unsigned long long)blkno, le32_to_cpu(bucket.xh->xh_entries[0].xe_name_hash)); if (func) { ret = func(inode, &bucket, para); @@ -2694,15 +2639,15 @@ static int ocfs2_xattr_update_xattr_search(struct inode *inode, ret = ocfs2_read_blocks(inode, xs->bucket.bhs[0]->b_blocknr + 1, blk_per_bucket - 1, &xs->bucket.bhs[1], - OCFS2_BH_CACHED); + 0); if (ret) { mlog_errno(ret); return ret; } - i = xs->here - old_xh->xh_entries; - xs->here = &xs->header->xh_entries[i]; } + i = xs->here - old_xh->xh_entries; + xs->here = &xs->header->xh_entries[i]; } return ret; @@ -2776,7 +2721,8 @@ static int ocfs2_xattr_create_index_block(struct inode *inode, */ blkno = ocfs2_clusters_to_blocks(inode->i_sb, bit_off); - mlog(0, "allocate 1 cluster from %llu to xattr block\n", blkno); + mlog(0, "allocate 1 cluster from %llu to xattr block\n", + (unsigned long long)blkno); xh_bh = sb_getblk(inode->i_sb, blkno); if (!xh_bh) { @@ -2818,7 +2764,11 @@ static int ocfs2_xattr_create_index_block(struct inode *inode, if (data_bh) ocfs2_journal_dirty(handle, data_bh); - ocfs2_xattr_update_xattr_search(inode, xs, xb_bh, xh_bh); + ret = ocfs2_xattr_update_xattr_search(inode, xs, xb_bh, xh_bh); + if (ret) { + mlog_errno(ret); + goto out_commit; + } /* Change from ocfs2_xattr_header to ocfs2_xattr_tree_root */ memset(&xb->xb_attrs, 0, inode->i_sb->s_blocksize - @@ -2898,8 +2848,7 @@ static int ocfs2_defrag_xattr_bucket(struct inode *inode, if (!bhs) return -ENOMEM; - ret = ocfs2_read_blocks(inode, blkno, blk_per_bucket, bhs, - OCFS2_BH_CACHED); + ret = ocfs2_read_blocks(inode, blkno, blk_per_bucket, bhs, 0); if (ret) goto out; @@ -2942,8 +2891,8 @@ static int ocfs2_defrag_xattr_bucket(struct inode *inode, mlog(0, "adjust xattr bucket in %llu, count = %u, " "xh_free_start = %u, xh_name_value_len = %u.\n", - blkno, le16_to_cpu(xh->xh_count), xh_free_start, - le16_to_cpu(xh->xh_name_value_len)); + (unsigned long long)blkno, le16_to_cpu(xh->xh_count), + xh_free_start, le16_to_cpu(xh->xh_name_value_len)); /* * sort all the entries by their offset. @@ -3059,7 +3008,7 @@ static int ocfs2_mv_xattr_bucket_cross_cluster(struct inode *inode, prev_blkno += (num_clusters - 1) * bpc + bpc / 2; mlog(0, "move half of xattrs in cluster %llu to %llu\n", - prev_blkno, new_blkno); + (unsigned long long)prev_blkno, (unsigned long long)new_blkno); /* * We need to update the 1st half of the new cluster and @@ -3153,8 +3102,7 @@ static int ocfs2_read_xattr_bucket(struct inode *inode, if (!new) return ocfs2_read_blocks(inode, blkno, - blk_per_bucket, bhs, - OCFS2_BH_CACHED); + blk_per_bucket, bhs, 0); for (i = 0; i < blk_per_bucket; i++) { bhs[i] = sb_getblk(inode->i_sb, blkno + i); @@ -3170,26 +3118,74 @@ static int ocfs2_read_xattr_bucket(struct inode *inode, } /* - * Move half num of the xattrs in old bucket(blk) to new bucket(new_blk). + * Find the suitable pos when we divide a bucket into 2. + * We have to make sure the xattrs with the same hash value exist + * in the same bucket. + * + * If this ocfs2_xattr_header covers more than one hash value, find a + * place where the hash value changes. Try to find the most even split. + * The most common case is that all entries have different hash values, + * and the first check we make will find a place to split. + */ +static int ocfs2_xattr_find_divide_pos(struct ocfs2_xattr_header *xh) +{ + struct ocfs2_xattr_entry *entries = xh->xh_entries; + int count = le16_to_cpu(xh->xh_count); + int delta, middle = count / 2; + + /* + * We start at the middle. Each step gets farther away in both + * directions. We therefore hit the change in hash value + * nearest to the middle. Note that this loop does not execute for + * count < 2. + */ + for (delta = 0; delta < middle; delta++) { + /* Let's check delta earlier than middle */ + if (cmp_xe(&entries[middle - delta - 1], + &entries[middle - delta])) + return middle - delta; + + /* For even counts, don't walk off the end */ + if ((middle + delta + 1) == count) + continue; + + /* Now try delta past middle */ + if (cmp_xe(&entries[middle + delta], + &entries[middle + delta + 1])) + return middle + delta + 1; + } + + /* Every entry had the same hash */ + return count; +} + +/* + * Move some xattrs in old bucket(blk) to new bucket(new_blk). * first_hash will record the 1st hash of the new bucket. + * + * Normally half of the xattrs will be moved. But we have to make + * sure that the xattrs with the same hash value are stored in the + * same bucket. If all the xattrs in this bucket have the same hash + * value, the new bucket will be initialized as an empty one and the + * first_hash will be initialized as (hash_value+1). */ -static int ocfs2_half_xattr_bucket(struct inode *inode, - handle_t *handle, - u64 blk, - u64 new_blk, - u32 *first_hash, - int new_bucket_head) +static int ocfs2_divide_xattr_bucket(struct inode *inode, + handle_t *handle, + u64 blk, + u64 new_blk, + u32 *first_hash, + int new_bucket_head) { int ret, i; - u16 count, start, len, name_value_len, xe_len, name_offset; + int count, start, len, name_value_len = 0, xe_len, name_offset = 0; u16 blk_per_bucket = ocfs2_blocks_per_xattr_bucket(inode->i_sb); struct buffer_head **s_bhs, **t_bhs = NULL; struct ocfs2_xattr_header *xh; struct ocfs2_xattr_entry *xe; int blocksize = inode->i_sb->s_blocksize; - mlog(0, "move half of xattrs from bucket %llu to %llu\n", - blk, new_blk); + mlog(0, "move some of xattrs from bucket %llu to %llu\n", + (unsigned long long)blk, (unsigned long long)new_blk); s_bhs = kcalloc(blk_per_bucket, sizeof(struct buffer_head *), GFP_NOFS); if (!s_bhs) @@ -3222,21 +3218,44 @@ static int ocfs2_half_xattr_bucket(struct inode *inode, for (i = 0; i < blk_per_bucket; i++) { ret = ocfs2_journal_access(handle, inode, t_bhs[i], - OCFS2_JOURNAL_ACCESS_CREATE); + new_bucket_head ? + OCFS2_JOURNAL_ACCESS_CREATE : + OCFS2_JOURNAL_ACCESS_WRITE); if (ret) { mlog_errno(ret); goto out; } } + xh = (struct ocfs2_xattr_header *)s_bhs[0]->b_data; + count = le16_to_cpu(xh->xh_count); + start = ocfs2_xattr_find_divide_pos(xh); + + if (start == count) { + xe = &xh->xh_entries[start-1]; + + /* + * initialized a new empty bucket here. + * The hash value is set as one larger than + * that of the last entry in the previous bucket. + */ + for (i = 0; i < blk_per_bucket; i++) + memset(t_bhs[i]->b_data, 0, blocksize); + + xh = (struct ocfs2_xattr_header *)t_bhs[0]->b_data; + xh->xh_free_start = cpu_to_le16(blocksize); + xh->xh_entries[0].xe_name_hash = xe->xe_name_hash; + le32_add_cpu(&xh->xh_entries[0].xe_name_hash, 1); + + goto set_num_buckets; + } + /* copy the whole bucket to the new first. */ for (i = 0; i < blk_per_bucket; i++) memcpy(t_bhs[i]->b_data, s_bhs[i]->b_data, blocksize); /* update the new bucket. */ xh = (struct ocfs2_xattr_header *)t_bhs[0]->b_data; - count = le16_to_cpu(xh->xh_count); - start = count / 2; /* * Calculate the total name/value len and xh_free_start for @@ -3293,6 +3312,7 @@ static int ocfs2_half_xattr_bucket(struct inode *inode, xh->xh_free_start = xe->xe_name_offset; } +set_num_buckets: /* set xh->xh_num_buckets for the new xh. */ if (new_bucket_head) xh->xh_num_buckets = cpu_to_le16(1); @@ -3310,9 +3330,13 @@ static int ocfs2_half_xattr_bucket(struct inode *inode, *first_hash = le32_to_cpu(xh->xh_entries[0].xe_name_hash); /* - * Now only update the 1st block of the old bucket. - * Please note that the entry has been sorted already above. + * Now only update the 1st block of the old bucket. If we + * just added a new empty bucket, there is no need to modify + * it. */ + if (start == count) + goto out; + xh = (struct ocfs2_xattr_header *)s_bhs[0]->b_data; memset(&xh->xh_entries[start], 0, sizeof(struct ocfs2_xattr_entry) * (count - start)); @@ -3360,7 +3384,8 @@ static int ocfs2_cp_xattr_bucket(struct inode *inode, BUG_ON(s_blkno == t_blkno); mlog(0, "cp bucket %llu to %llu, target is %d\n", - s_blkno, t_blkno, t_is_new); + (unsigned long long)s_blkno, (unsigned long long)t_blkno, + t_is_new); s_bhs = kzalloc(sizeof(struct buffer_head *) * blk_per_bucket, GFP_NOFS); @@ -3384,6 +3409,8 @@ static int ocfs2_cp_xattr_bucket(struct inode *inode, for (i = 0; i < blk_per_bucket; i++) { ret = ocfs2_journal_access(handle, inode, t_bhs[i], + t_is_new ? + OCFS2_JOURNAL_ACCESS_CREATE : OCFS2_JOURNAL_ACCESS_WRITE); if (ret) goto out; @@ -3430,7 +3457,8 @@ static int ocfs2_cp_xattr_cluster(struct inode *inode, struct ocfs2_xattr_header *xh; u64 to_blk_start = to_blk; - mlog(0, "cp xattrs from cluster %llu to %llu\n", src_blk, to_blk); + mlog(0, "cp xattrs from cluster %llu to %llu\n", + (unsigned long long)src_blk, (unsigned long long)to_blk); /* * We need to update the new cluster and 1 more for the update of @@ -3495,15 +3523,15 @@ out: } /* - * Move half of the xattrs in this cluster to the new cluster. + * Move some xattrs in this cluster to the new cluster. * This function should only be called when bucket size == cluster size. * Otherwise ocfs2_mv_xattr_bucket_cross_cluster should be used instead. */ -static int ocfs2_half_xattr_cluster(struct inode *inode, - handle_t *handle, - u64 prev_blk, - u64 new_blk, - u32 *first_hash) +static int ocfs2_divide_xattr_cluster(struct inode *inode, + handle_t *handle, + u64 prev_blk, + u64 new_blk, + u32 *first_hash) { u16 blk_per_bucket = ocfs2_blocks_per_xattr_bucket(inode->i_sb); int ret, credits = 2 * blk_per_bucket; @@ -3517,8 +3545,8 @@ static int ocfs2_half_xattr_cluster(struct inode *inode, } /* Move half of the xattr in start_blk to the next bucket. */ - return ocfs2_half_xattr_bucket(inode, handle, prev_blk, - new_blk, first_hash, 1); + return ocfs2_divide_xattr_bucket(inode, handle, prev_blk, + new_blk, first_hash, 1); } /* @@ -3561,7 +3589,8 @@ static int ocfs2_adjust_xattr_cross_cluster(struct inode *inode, int bpc = ocfs2_clusters_to_blocks(inode->i_sb, 1); mlog(0, "adjust xattrs from cluster %llu len %u to %llu\n", - prev_blk, prev_clusters, new_blk); + (unsigned long long)prev_blk, prev_clusters, + (unsigned long long)new_blk); if (ocfs2_xattr_buckets_per_cluster(OCFS2_SB(inode->i_sb)) > 1) ret = ocfs2_mv_xattr_bucket_cross_cluster(inode, @@ -3580,9 +3609,9 @@ static int ocfs2_adjust_xattr_cross_cluster(struct inode *inode, last_blk, new_blk, v_start); else { - ret = ocfs2_half_xattr_cluster(inode, handle, - last_blk, new_blk, - v_start); + ret = ocfs2_divide_xattr_cluster(inode, handle, + last_blk, new_blk, + v_start); if ((*header_bh)->b_blocknr == last_blk && extend) *extend = 0; @@ -3631,7 +3660,7 @@ static int ocfs2_add_new_xattr_cluster(struct inode *inode, mlog(0, "Add new xattr cluster for %llu, previous xattr hash = %u, " "previous xattr blkno = %llu\n", (unsigned long long)OCFS2_I(inode)->ip_blkno, - prev_cpos, prev_blkno); + prev_cpos, (unsigned long long)prev_blkno); ocfs2_init_xattr_tree_extent_tree(&et, inode, root_bh); @@ -3718,7 +3747,7 @@ static int ocfs2_add_new_xattr_cluster(struct inode *inode, } } mlog(0, "Insert %u clusters at block %llu for xattr at %u\n", - num_bits, block, v_start); + num_bits, (unsigned long long)block, v_start); ret = ocfs2_insert_extent(osb, handle, inode, &et, v_start, block, num_bits, 0, meta_ac); if (ret < 0) { @@ -3763,7 +3792,7 @@ static int ocfs2_extend_xattr_bucket(struct inode *inode, u16 bucket = le16_to_cpu(first_xh->xh_num_buckets); mlog(0, "extend xattr bucket in %llu, xattr extend rec starting " - "from %llu, len = %u\n", start_blk, + "from %llu, len = %u\n", (unsigned long long)start_blk, (unsigned long long)first_bh->b_blocknr, num_clusters); BUG_ON(bucket >= num_buckets); @@ -3799,8 +3828,8 @@ static int ocfs2_extend_xattr_bucket(struct inode *inode, } /* Move half of the xattr in start_blk to the next bucket. */ - ret = ocfs2_half_xattr_bucket(inode, handle, start_blk, - start_blk + blk_per_bucket, NULL, 0); + ret = ocfs2_divide_xattr_bucket(inode, handle, start_blk, + start_blk + blk_per_bucket, NULL, 0); le16_add_cpu(&first_xh->xh_num_buckets, 1); ocfs2_journal_dirty(handle, first_bh); @@ -4101,7 +4130,7 @@ static int ocfs2_xattr_set_entry_in_bucket(struct inode *inode, ret = ocfs2_read_blocks(inode, xs->bucket.bhs[0]->b_blocknr + 1, blk_per_bucket - 1, &xs->bucket.bhs[1], - OCFS2_BH_CACHED); + 0); if (ret) { mlog_errno(ret); goto out; @@ -4148,7 +4177,7 @@ static int ocfs2_xattr_value_update_size(struct inode *inode, handle_t *handle = NULL; handle = ocfs2_start_trans(osb, 1); - if (handle == NULL) { + if (IS_ERR(handle)) { ret = -ENOMEM; mlog_errno(ret); goto out; @@ -4315,7 +4344,7 @@ static int ocfs2_rm_xattr_cluster(struct inode *inode, } handle = ocfs2_start_trans(osb, OCFS2_REMOVE_EXTENT_CREDITS); - if (handle == NULL) { + if (IS_ERR(handle)) { ret = -ENOMEM; mlog_errno(ret); goto out; @@ -4491,11 +4520,21 @@ out: return ret; } -/* check whether the xattr bucket is filled up with the same hash value. */ +/* + * check whether the xattr bucket is filled up with the same hash value. + * If we want to insert the xattr with the same hash, return -ENOSPC. + * If we want to insert a xattr with different hash value, go ahead + * and ocfs2_divide_xattr_bucket will handle this. + */ static int ocfs2_check_xattr_bucket_collision(struct inode *inode, - struct ocfs2_xattr_bucket *bucket) + struct ocfs2_xattr_bucket *bucket, + const char *name) { struct ocfs2_xattr_header *xh = bucket->xh; + u32 name_hash = ocfs2_xattr_name_hash(inode, name, strlen(name)); + + if (name_hash != le32_to_cpu(xh->xh_entries[0].xe_name_hash)) + return 0; if (xh->xh_entries[le16_to_cpu(xh->xh_count) - 1].xe_name_hash == xh->xh_entries[0].xe_name_hash) { @@ -4618,7 +4657,9 @@ try_again: * one bucket's worth, so check it here whether we need to * add a new bucket for the insert. */ - ret = ocfs2_check_xattr_bucket_collision(inode, &xs->bucket); + ret = ocfs2_check_xattr_bucket_collision(inode, + &xs->bucket, + xi->name); if (ret) { mlog_errno(ret); goto out; @@ -4729,14 +4770,11 @@ out: /* * 'trusted' attributes support */ - -#define XATTR_TRUSTED_PREFIX "trusted." - static size_t ocfs2_xattr_trusted_list(struct inode *inode, char *list, size_t list_size, const char *name, size_t name_len) { - const size_t prefix_len = sizeof(XATTR_TRUSTED_PREFIX) - 1; + const size_t prefix_len = XATTR_TRUSTED_PREFIX_LEN; const size_t total_len = prefix_len + name_len + 1; if (list && total_len <= list_size) { @@ -4773,18 +4811,14 @@ struct xattr_handler ocfs2_xattr_trusted_handler = { .set = ocfs2_xattr_trusted_set, }; - /* * 'user' attributes support */ - -#define XATTR_USER_PREFIX "user." - static size_t ocfs2_xattr_user_list(struct inode *inode, char *list, size_t list_size, const char *name, size_t name_len) { - const size_t prefix_len = sizeof(XATTR_USER_PREFIX) - 1; + const size_t prefix_len = XATTR_USER_PREFIX_LEN; const size_t total_len = prefix_len + name_len + 1; struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); diff --git a/fs/ocfs2/xattr.h b/fs/ocfs2/xattr.h index c25c7c62a05..1d8314c7656 100644 --- a/fs/ocfs2/xattr.h +++ b/fs/ocfs2/xattr.h @@ -3,24 +3,16 @@ * * xattr.h * - * Function prototypes - * - * Copyright (C) 2008 Oracle. All rights reserved. + * Copyright (C) 2004, 2008 Oracle. All rights reserved. * * This program is free software; you can redistribute it and/or * modify it under the terms of the GNU General Public - * License as published by the Free Software Foundation; either - * version 2 of the License, or (at your option) any later version. + * License version 2 as published by the Free Software Foundation. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU * General Public License for more details. - * - * You should have received a copy of the GNU General Public - * License along with this program; if not, write to the - * Free Software Foundation, Inc., 59 Temple Place - Suite 330, - * Boston, MA 021110-1307, USA. */ #ifndef OCFS2_XATTR_H @@ -40,29 +32,11 @@ enum ocfs2_xattr_type { extern struct xattr_handler ocfs2_xattr_user_handler; extern struct xattr_handler ocfs2_xattr_trusted_handler; - -extern ssize_t ocfs2_listxattr(struct dentry *, char *, size_t); -extern int ocfs2_xattr_get(struct inode *, int, const char *, void *, size_t); -extern int ocfs2_xattr_set(struct inode *, int, const char *, const void *, - size_t, int); -extern int ocfs2_xattr_remove(struct inode *inode, struct buffer_head *di_bh); extern struct xattr_handler *ocfs2_xattr_handlers[]; -static inline u16 ocfs2_xattr_buckets_per_cluster(struct ocfs2_super *osb) -{ - return (1 << osb->s_clustersize_bits) / OCFS2_XATTR_BUCKET_SIZE; -} - -static inline u16 ocfs2_blocks_per_xattr_bucket(struct super_block *sb) -{ - return OCFS2_XATTR_BUCKET_SIZE / (1 << sb->s_blocksize_bits); -} - -static inline u16 ocfs2_xattr_max_xe_in_bucket(struct super_block *sb) -{ - u16 len = sb->s_blocksize - - offsetof(struct ocfs2_xattr_header, xh_entries); +ssize_t ocfs2_listxattr(struct dentry *, char *, size_t); +int ocfs2_xattr_set(struct inode *, int, const char *, const void *, + size_t, int); +int ocfs2_xattr_remove(struct inode *, struct buffer_head *); - return len / sizeof(struct ocfs2_xattr_entry); -} #endif /* OCFS2_XATTR_H */ diff --git a/fs/omfs/dir.c b/fs/omfs/dir.c index c0757e99887..c7275cfbdcf 100644 --- a/fs/omfs/dir.c +++ b/fs/omfs/dir.c @@ -501,4 +501,5 @@ struct inode_operations omfs_dir_inops = { struct file_operations omfs_dir_operations = { .read = generic_read_dir, .readdir = omfs_readdir, + .llseek = generic_file_llseek, }; diff --git a/fs/omfs/inode.c b/fs/omfs/inode.c index cbf047a847c..6afe57c84f8 100644 --- a/fs/omfs/inode.c +++ b/fs/omfs/inode.c @@ -37,8 +37,8 @@ struct inode *omfs_new_inode(struct inode *dir, int mode) inode->i_ino = new_block; inode->i_mode = mode; - inode->i_uid = current->fsuid; - inode->i_gid = current->fsgid; + inode->i_uid = current_fsuid(); + inode->i_gid = current_fsgid(); inode->i_blocks = 0; inode->i_mapping->a_ops = &omfs_aops; @@ -420,8 +420,8 @@ static int omfs_fill_super(struct super_block *sb, void *data, int silent) sb->s_fs_info = sbi; - sbi->s_uid = current->uid; - sbi->s_gid = current->gid; + sbi->s_uid = current_uid(); + sbi->s_gid = current_gid(); sbi->s_dmask = sbi->s_fmask = current->fs->umask; if (!parse_options((char *) data, sbi)) diff --git a/fs/open.c b/fs/open.c index 5596049863b..c0a426d5766 100644 --- a/fs/open.c +++ b/fs/open.c @@ -425,39 +425,33 @@ out: */ asmlinkage long sys_faccessat(int dfd, const char __user *filename, int mode) { + const struct cred *old_cred; + struct cred *override_cred; struct path path; struct inode *inode; - int old_fsuid, old_fsgid; - kernel_cap_t uninitialized_var(old_cap); /* !SECURE_NO_SETUID_FIXUP */ int res; if (mode & ~S_IRWXO) /* where's F_OK, X_OK, W_OK, R_OK? */ return -EINVAL; - old_fsuid = current->fsuid; - old_fsgid = current->fsgid; + override_cred = prepare_creds(); + if (!override_cred) + return -ENOMEM; - current->fsuid = current->uid; - current->fsgid = current->gid; + override_cred->fsuid = override_cred->uid; + override_cred->fsgid = override_cred->gid; if (!issecure(SECURE_NO_SETUID_FIXUP)) { - /* - * Clear the capabilities if we switch to a non-root user - */ -#ifndef CONFIG_SECURITY_FILE_CAPABILITIES - /* - * FIXME: There is a race here against sys_capset. The - * capabilities can change yet we will restore the old - * value below. We should hold task_capabilities_lock, - * but we cannot because user_path_at can sleep. - */ -#endif /* ndef CONFIG_SECURITY_FILE_CAPABILITIES */ - if (current->uid) - old_cap = cap_set_effective(__cap_empty_set); + /* Clear the capabilities if we switch to a non-root user */ + if (override_cred->uid) + cap_clear(override_cred->cap_effective); else - old_cap = cap_set_effective(current->cap_permitted); + override_cred->cap_effective = + override_cred->cap_permitted; } + old_cred = override_creds(override_cred); + res = user_path_at(dfd, filename, LOOKUP_FOLLOW, &path); if (res) goto out; @@ -494,12 +488,8 @@ asmlinkage long sys_faccessat(int dfd, const char __user *filename, int mode) out_path_release: path_put(&path); out: - current->fsuid = old_fsuid; - current->fsgid = old_fsgid; - - if (!issecure(SECURE_NO_SETUID_FIXUP)) - cap_set_effective(old_cap); - + revert_creds(old_cred); + put_cred(override_cred); return res; } @@ -792,13 +782,14 @@ static inline int __get_file_write_access(struct inode *inode, static struct file *__dentry_open(struct dentry *dentry, struct vfsmount *mnt, int flags, struct file *f, - int (*open)(struct inode *, struct file *)) + int (*open)(struct inode *, struct file *), + const struct cred *cred) { struct inode *inode; int error; f->f_flags = flags; - f->f_mode = ((flags+1) & O_ACCMODE) | FMODE_LSEEK | + f->f_mode = (__force fmode_t)((flags+1) & O_ACCMODE) | FMODE_LSEEK | FMODE_PREAD | FMODE_PWRITE; inode = dentry->d_inode; if (f->f_mode & FMODE_WRITE) { @@ -816,7 +807,7 @@ static struct file *__dentry_open(struct dentry *dentry, struct vfsmount *mnt, f->f_op = fops_get(inode->i_fop); file_move(f, &inode->i_sb->s_files); - error = security_dentry_open(f); + error = security_dentry_open(f, cred); if (error) goto cleanup_all; @@ -891,6 +882,8 @@ cleanup_file: struct file *lookup_instantiate_filp(struct nameidata *nd, struct dentry *dentry, int (*open)(struct inode *, struct file *)) { + const struct cred *cred = current_cred(); + if (IS_ERR(nd->intent.open.file)) goto out; if (IS_ERR(dentry)) @@ -898,7 +891,7 @@ struct file *lookup_instantiate_filp(struct nameidata *nd, struct dentry *dentry nd->intent.open.file = __dentry_open(dget(dentry), mntget(nd->path.mnt), nd->intent.open.flags - 1, nd->intent.open.file, - open); + open, cred); out: return nd->intent.open.file; out_err: @@ -917,6 +910,7 @@ EXPORT_SYMBOL_GPL(lookup_instantiate_filp); */ struct file *nameidata_to_filp(struct nameidata *nd, int flags) { + const struct cred *cred = current_cred(); struct file *filp; /* Pick up the filp from the open intent */ @@ -924,7 +918,7 @@ struct file *nameidata_to_filp(struct nameidata *nd, int flags) /* Has the filesystem initialised the file for us? */ if (filp->f_path.dentry == NULL) filp = __dentry_open(nd->path.dentry, nd->path.mnt, flags, filp, - NULL); + NULL, cred); else path_put(&nd->path); return filp; @@ -934,7 +928,8 @@ struct file *nameidata_to_filp(struct nameidata *nd, int flags) * dentry_open() will have done dput(dentry) and mntput(mnt) if it returns an * error. */ -struct file *dentry_open(struct dentry *dentry, struct vfsmount *mnt, int flags) +struct file *dentry_open(struct dentry *dentry, struct vfsmount *mnt, int flags, + const struct cred *cred) { int error; struct file *f; @@ -959,7 +954,7 @@ struct file *dentry_open(struct dentry *dentry, struct vfsmount *mnt, int flags) return ERR_PTR(error); } - return __dentry_open(dentry, mnt, flags, f, NULL); + return __dentry_open(dentry, mnt, flags, f, NULL, cred); } EXPORT_SYMBOL(dentry_open); diff --git a/fs/openpromfs/inode.c b/fs/openpromfs/inode.c index 9f5b054f06b..d41bdc784de 100644 --- a/fs/openpromfs/inode.c +++ b/fs/openpromfs/inode.c @@ -167,6 +167,7 @@ static int openpromfs_readdir(struct file *, void *, filldir_t); static const struct file_operations openprom_operations = { .read = generic_read_dir, .readdir = openpromfs_readdir, + .llseek = generic_file_llseek, }; static struct dentry *openpromfs_lookup(struct inode *, struct dentry *, struct nameidata *); diff --git a/fs/partitions/acorn.c b/fs/partitions/acorn.c index 3d3e1663147..a97b477ac0f 100644 --- a/fs/partitions/acorn.c +++ b/fs/partitions/acorn.c @@ -275,16 +275,6 @@ adfspart_check_ADFS(struct parsed_partitions *state, struct block_device *bdev) id = data[0x1fc] & 15; put_dev_sector(sect); -#ifdef CONFIG_BLK_DEV_MFM - if (MAJOR(bdev->bd_dev) == MFM_ACORN_MAJOR) { - extern void xd_set_geometry(struct block_device *, - unsigned char, unsigned char, unsigned int); - xd_set_geometry(bdev, dr->secspertrack, heads, 1); - invalidate_bh_lrus(); - truncate_inode_pages(bdev->bd_inode->i_mapping, 0); - } -#endif - /* * Work out start of non-adfs partition. */ diff --git a/fs/partitions/check.c b/fs/partitions/check.c index 7408227c49c..6d5b213b8a9 100644 --- a/fs/partitions/check.c +++ b/fs/partitions/check.c @@ -195,6 +195,14 @@ check_partition(struct gendisk *hd, struct block_device *bdev) return ERR_PTR(res); } +static ssize_t part_partition_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct hd_struct *p = dev_to_part(dev); + + return sprintf(buf, "%d\n", p->partno); +} + static ssize_t part_start_show(struct device *dev, struct device_attribute *attr, char *buf) { @@ -260,6 +268,7 @@ ssize_t part_fail_store(struct device *dev, } #endif +static DEVICE_ATTR(partition, S_IRUGO, part_partition_show, NULL); static DEVICE_ATTR(start, S_IRUGO, part_start_show, NULL); static DEVICE_ATTR(size, S_IRUGO, part_size_show, NULL); static DEVICE_ATTR(stat, S_IRUGO, part_stat_show, NULL); @@ -269,6 +278,7 @@ static struct device_attribute dev_attr_fail = #endif static struct attribute *part_attrs[] = { + &dev_attr_partition.attr, &dev_attr_start.attr, &dev_attr_size.attr, &dev_attr_stat.attr, @@ -338,8 +348,8 @@ static ssize_t whole_disk_show(struct device *dev, static DEVICE_ATTR(whole_disk, S_IRUSR | S_IRGRP | S_IROTH, whole_disk_show, NULL); -int add_partition(struct gendisk *disk, int partno, - sector_t start, sector_t len, int flags) +struct hd_struct *add_partition(struct gendisk *disk, int partno, + sector_t start, sector_t len, int flags) { struct hd_struct *p; dev_t devt = MKDEV(0, 0); @@ -351,15 +361,15 @@ int add_partition(struct gendisk *disk, int partno, err = disk_expand_part_tbl(disk, partno); if (err) - return err; + return ERR_PTR(err); ptbl = disk->part_tbl; if (ptbl->part[partno]) - return -EBUSY; + return ERR_PTR(-EBUSY); p = kzalloc(sizeof(*p), GFP_KERNEL); if (!p) - return -ENOMEM; + return ERR_PTR(-EBUSY); if (!init_part_stats(p)) { err = -ENOMEM; @@ -385,7 +395,7 @@ int add_partition(struct gendisk *disk, int partno, err = blk_alloc_devt(p, &devt); if (err) - goto out_free; + goto out_free_stats; pdev->devt = devt; /* delay uevent until 'holders' subdir is created */ @@ -414,18 +424,20 @@ int add_partition(struct gendisk *disk, int partno, if (!ddev->uevent_suppress) kobject_uevent(&pdev->kobj, KOBJ_ADD); - return 0; + return p; +out_free_stats: + free_part_stats(p); out_free: kfree(p); - return err; + return ERR_PTR(err); out_del: kobject_put(p->holder_dir); device_del(pdev); out_put: put_device(pdev); blk_free_devt(devt); - return err; + return ERR_PTR(err); } /* Not exported, helper to add_disk(). */ @@ -475,10 +487,10 @@ void register_disk(struct gendisk *disk) goto exit; bdev->bd_invalidated = 1; - err = blkdev_get(bdev, FMODE_READ, 0); + err = blkdev_get(bdev, FMODE_READ); if (err < 0) goto exit; - blkdev_put(bdev); + blkdev_put(bdev, FMODE_READ); exit: /* announce disk after possible partitions are created */ @@ -538,20 +550,34 @@ int rescan_partitions(struct gendisk *disk, struct block_device *bdev) sector_t from = state->parts[p].from; if (!size) continue; + if (from >= get_capacity(disk)) { + printk(KERN_WARNING + "%s: p%d ignored, start %llu is behind the end of the disk\n", + disk->disk_name, p, (unsigned long long) from); + continue; + } if (from + size > get_capacity(disk)) { + /* + * we can not ignore partitions of broken tables + * created by for example camera firmware, but we + * limit them to the end of the disk to avoid + * creating invalid block devices + */ printk(KERN_WARNING - "%s: p%d exceeds device capacity\n", - disk->disk_name, p); + "%s: p%d size %llu limited to end of disk\n", + disk->disk_name, p, (unsigned long long) size); + size = get_capacity(disk) - from; } - res = add_partition(disk, p, from, size, state->parts[p].flags); - if (res) { - printk(KERN_ERR " %s: p%d could not be added: %d\n", - disk->disk_name, p, -res); + part = add_partition(disk, p, from, size, + state->parts[p].flags); + if (IS_ERR(part)) { + printk(KERN_ERR " %s: p%d could not be added: %ld\n", + disk->disk_name, p, -PTR_ERR(part)); continue; } #ifdef CONFIG_BLK_DEV_MD if (state->parts[p].flags & ADDPART_FLAG_RAID) - md_autodetect_dev(bdev->bd_dev+p); + md_autodetect_dev(part_to_dev(part)->devt); #endif } kfree(state); diff --git a/fs/pipe.c b/fs/pipe.c index fcba6542b8d..aaf797bd57b 100644 --- a/fs/pipe.c +++ b/fs/pipe.c @@ -717,14 +717,12 @@ pipe_rdwr_fasync(int fd, struct file *filp, int on) static int pipe_read_release(struct inode *inode, struct file *filp) { - pipe_read_fasync(-1, filp, 0); return pipe_release(inode, 1, 0); } static int pipe_write_release(struct inode *inode, struct file *filp) { - pipe_write_fasync(-1, filp, 0); return pipe_release(inode, 0, 1); } @@ -733,7 +731,6 @@ pipe_rdwr_release(struct inode *inode, struct file *filp) { int decr, decw; - pipe_rdwr_fasync(-1, filp, 0); decr = (filp->f_mode & FMODE_READ) != 0; decw = (filp->f_mode & FMODE_WRITE) != 0; return pipe_release(inode, decr, decw); @@ -902,8 +899,8 @@ static struct inode * get_pipe_inode(void) */ inode->i_state = I_DIRTY; inode->i_mode = S_IFIFO | S_IRUSR | S_IWUSR; - inode->i_uid = current->fsuid; - inode->i_gid = current->fsgid; + inode->i_uid = current_fsuid(); + inode->i_gid = current_fsgid(); inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME; return inode; diff --git a/fs/posix_acl.c b/fs/posix_acl.c index aec931e0997..39df95a0ec2 100644 --- a/fs/posix_acl.c +++ b/fs/posix_acl.c @@ -217,11 +217,11 @@ posix_acl_permission(struct inode *inode, const struct posix_acl *acl, int want) switch(pa->e_tag) { case ACL_USER_OBJ: /* (May have been checked already) */ - if (inode->i_uid == current->fsuid) + if (inode->i_uid == current_fsuid()) goto check_perm; break; case ACL_USER: - if (pa->e_id == current->fsuid) + if (pa->e_id == current_fsuid()) goto mask; break; case ACL_GROUP_OBJ: diff --git a/fs/proc/Makefile b/fs/proc/Makefile index ebaba021354..63d965193b2 100644 --- a/fs/proc/Makefile +++ b/fs/proc/Makefile @@ -8,11 +8,20 @@ proc-y := nommu.o task_nommu.o proc-$(CONFIG_MMU) := mmu.o task_mmu.o proc-y += inode.o root.o base.o generic.o array.o \ - proc_tty.o proc_misc.o - + proc_tty.o +proc-y += cmdline.o +proc-y += cpuinfo.o +proc-y += devices.o +proc-y += interrupts.o +proc-y += loadavg.o +proc-y += meminfo.o +proc-y += stat.o +proc-y += uptime.o +proc-y += version.o proc-$(CONFIG_PROC_SYSCTL) += proc_sysctl.o proc-$(CONFIG_NET) += proc_net.o proc-$(CONFIG_PROC_KCORE) += kcore.o proc-$(CONFIG_PROC_VMCORE) += vmcore.o proc-$(CONFIG_PROC_DEVICETREE) += proc_devtree.o proc-$(CONFIG_PRINTK) += kmsg.o +proc-$(CONFIG_PROC_PAGE_MONITOR) += page.o diff --git a/fs/proc/array.c b/fs/proc/array.c index f4bc0e78953..7e4877d9dcb 100644 --- a/fs/proc/array.c +++ b/fs/proc/array.c @@ -40,7 +40,7 @@ * * * Alan Cox : security fixes. - * <Alan.Cox@linux.org> + * <alan@lxorguk.ukuu.org.uk> * * Al Viro : safe handling of mm_struct * @@ -159,6 +159,7 @@ static inline void task_state(struct seq_file *m, struct pid_namespace *ns, struct group_info *group_info; int g; struct fdtable *fdt = NULL; + const struct cred *cred; pid_t ppid, tpid; rcu_read_lock(); @@ -170,6 +171,7 @@ static inline void task_state(struct seq_file *m, struct pid_namespace *ns, if (tracer) tpid = task_pid_nr_ns(tracer, ns); } + cred = get_cred((struct cred *) __task_cred(p)); seq_printf(m, "State:\t%s\n" "Tgid:\t%d\n" @@ -182,8 +184,8 @@ static inline void task_state(struct seq_file *m, struct pid_namespace *ns, task_tgid_nr_ns(p, ns), pid_nr_ns(pid, ns), ppid, tpid, - p->uid, p->euid, p->suid, p->fsuid, - p->gid, p->egid, p->sgid, p->fsgid); + cred->uid, cred->euid, cred->suid, cred->fsuid, + cred->gid, cred->egid, cred->sgid, cred->fsgid); task_lock(p); if (p->files) @@ -194,13 +196,12 @@ static inline void task_state(struct seq_file *m, struct pid_namespace *ns, fdt ? fdt->max_fds : 0); rcu_read_unlock(); - group_info = p->group_info; - get_group_info(group_info); + group_info = cred->group_info; task_unlock(p); for (g = 0; g < min(group_info->ngroups, NGROUPS_SMALL); g++) seq_printf(m, "%d ", GROUP_AT(group_info, g)); - put_group_info(group_info); + put_cred(cred); seq_printf(m, "\n"); } @@ -262,7 +263,7 @@ static inline void task_sig(struct seq_file *m, struct task_struct *p) blocked = p->blocked; collect_sigign_sigcatch(p, &ignored, &caught); num_threads = atomic_read(&p->signal->count); - qsize = atomic_read(&p->user->sigpending); + qsize = atomic_read(&__task_cred(p)->user->sigpending); qlim = p->signal->rlim[RLIMIT_SIGPENDING].rlim_cur; unlock_task_sighand(p, &flags); } @@ -293,10 +294,21 @@ static void render_cap_t(struct seq_file *m, const char *header, static inline void task_cap(struct seq_file *m, struct task_struct *p) { - render_cap_t(m, "CapInh:\t", &p->cap_inheritable); - render_cap_t(m, "CapPrm:\t", &p->cap_permitted); - render_cap_t(m, "CapEff:\t", &p->cap_effective); - render_cap_t(m, "CapBnd:\t", &p->cap_bset); + const struct cred *cred; + kernel_cap_t cap_inheritable, cap_permitted, cap_effective, cap_bset; + + rcu_read_lock(); + cred = __task_cred(p); + cap_inheritable = cred->cap_inheritable; + cap_permitted = cred->cap_permitted; + cap_effective = cred->cap_effective; + cap_bset = cred->cap_bset; + rcu_read_unlock(); + + render_cap_t(m, "CapInh:\t", &cap_inheritable); + render_cap_t(m, "CapPrm:\t", &cap_permitted); + render_cap_t(m, "CapEff:\t", &cap_effective); + render_cap_t(m, "CapBnd:\t", &cap_bset); } static inline void task_context_switch_counts(struct seq_file *m, @@ -388,20 +400,20 @@ static int do_task_stat(struct seq_file *m, struct pid_namespace *ns, /* add up live thread stats at the group level */ if (whole) { + struct task_cputime cputime; struct task_struct *t = task; do { min_flt += t->min_flt; maj_flt += t->maj_flt; - utime = cputime_add(utime, task_utime(t)); - stime = cputime_add(stime, task_stime(t)); gtime = cputime_add(gtime, task_gtime(t)); t = next_thread(t); } while (t != task); min_flt += sig->min_flt; maj_flt += sig->maj_flt; - utime = cputime_add(utime, sig->utime); - stime = cputime_add(stime, sig->stime); + thread_group_cputime(task, &cputime); + utime = cputime.utime; + stime = cputime.stime; gtime = cputime_add(gtime, sig->gtime); } diff --git a/fs/proc/base.c b/fs/proc/base.c index b5918ae8ca7..cad92c1ac2b 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -347,8 +347,8 @@ static int proc_pid_wchan(struct task_struct *task, char *buffer) static int proc_pid_schedstat(struct task_struct *task, char *buffer) { return sprintf(buffer, "%llu %llu %lu\n", - task->sched_info.cpu_time, - task->sched_info.run_delay, + (unsigned long long)task->se.sum_exec_runtime, + (unsigned long long)task->sched_info.run_delay, task->sched_info.pcount); } #endif @@ -371,7 +371,7 @@ static int lstats_show_proc(struct seq_file *m, void *v) task->latency_record[i].time, task->latency_record[i].max); for (q = 0; q < LT_BACKTRACEDEPTH; q++) { - char sym[KSYM_NAME_LEN]; + char sym[KSYM_SYMBOL_LEN]; char *c; if (!task->latency_record[i].backtrace[q]) break; @@ -1406,6 +1406,7 @@ static struct inode *proc_pid_make_inode(struct super_block * sb, struct task_st { struct inode * inode; struct proc_inode *ei; + const struct cred *cred; /* We need a new inode */ @@ -1428,8 +1429,11 @@ static struct inode *proc_pid_make_inode(struct super_block * sb, struct task_st inode->i_uid = 0; inode->i_gid = 0; if (task_dumpable(task)) { - inode->i_uid = task->euid; - inode->i_gid = task->egid; + rcu_read_lock(); + cred = __task_cred(task); + inode->i_uid = cred->euid; + inode->i_gid = cred->egid; + rcu_read_unlock(); } security_task_to_inode(task, inode); @@ -1445,6 +1449,8 @@ static int pid_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat { struct inode *inode = dentry->d_inode; struct task_struct *task; + const struct cred *cred; + generic_fillattr(inode, stat); rcu_read_lock(); @@ -1454,8 +1460,9 @@ static int pid_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat if (task) { if ((inode->i_mode == (S_IFDIR|S_IRUGO|S_IXUGO)) || task_dumpable(task)) { - stat->uid = task->euid; - stat->gid = task->egid; + cred = __task_cred(task); + stat->uid = cred->euid; + stat->gid = cred->egid; } } rcu_read_unlock(); @@ -1483,11 +1490,16 @@ static int pid_revalidate(struct dentry *dentry, struct nameidata *nd) { struct inode *inode = dentry->d_inode; struct task_struct *task = get_proc_task(inode); + const struct cred *cred; + if (task) { if ((inode->i_mode == (S_IFDIR|S_IRUGO|S_IXUGO)) || task_dumpable(task)) { - inode->i_uid = task->euid; - inode->i_gid = task->egid; + rcu_read_lock(); + cred = __task_cred(task); + inode->i_uid = cred->euid; + inode->i_gid = cred->egid; + rcu_read_unlock(); } else { inode->i_uid = 0; inode->i_gid = 0; @@ -1649,6 +1661,7 @@ static int tid_fd_revalidate(struct dentry *dentry, struct nameidata *nd) struct task_struct *task = get_proc_task(inode); int fd = proc_fd(inode); struct files_struct *files; + const struct cred *cred; if (task) { files = get_files_struct(task); @@ -1658,8 +1671,11 @@ static int tid_fd_revalidate(struct dentry *dentry, struct nameidata *nd) rcu_read_unlock(); put_files_struct(files); if (task_dumpable(task)) { - inode->i_uid = task->euid; - inode->i_gid = task->egid; + rcu_read_lock(); + cred = __task_cred(task); + inode->i_uid = cred->euid; + inode->i_gid = cred->egid; + rcu_read_unlock(); } else { inode->i_uid = 0; inode->i_gid = 0; @@ -1712,9 +1728,9 @@ static struct dentry *proc_fd_instantiate(struct inode *dir, file = fcheck_files(files, fd); if (!file) goto out_unlock; - if (file->f_mode & 1) + if (file->f_mode & FMODE_READ) inode->i_mode |= S_IRUSR | S_IXUSR; - if (file->f_mode & 2) + if (file->f_mode & FMODE_WRITE) inode->i_mode |= S_IWUSR | S_IXUSR; spin_unlock(&files->file_lock); put_files_struct(files); diff --git a/fs/proc/cmdline.c b/fs/proc/cmdline.c new file mode 100644 index 00000000000..82676e3fcd1 --- /dev/null +++ b/fs/proc/cmdline.c @@ -0,0 +1,29 @@ +#include <linux/fs.h> +#include <linux/init.h> +#include <linux/proc_fs.h> +#include <linux/seq_file.h> + +static int cmdline_proc_show(struct seq_file *m, void *v) +{ + seq_printf(m, "%s\n", saved_command_line); + return 0; +} + +static int cmdline_proc_open(struct inode *inode, struct file *file) +{ + return single_open(file, cmdline_proc_show, NULL); +} + +static const struct file_operations cmdline_proc_fops = { + .open = cmdline_proc_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, +}; + +static int __init proc_cmdline_init(void) +{ + proc_create("cmdline", 0, NULL, &cmdline_proc_fops); + return 0; +} +module_init(proc_cmdline_init); diff --git a/fs/proc/cpuinfo.c b/fs/proc/cpuinfo.c new file mode 100644 index 00000000000..5a1e539a234 --- /dev/null +++ b/fs/proc/cpuinfo.c @@ -0,0 +1,24 @@ +#include <linux/fs.h> +#include <linux/init.h> +#include <linux/proc_fs.h> +#include <linux/seq_file.h> + +extern const struct seq_operations cpuinfo_op; +static int cpuinfo_open(struct inode *inode, struct file *file) +{ + return seq_open(file, &cpuinfo_op); +} + +static const struct file_operations proc_cpuinfo_operations = { + .open = cpuinfo_open, + .read = seq_read, + .llseek = seq_lseek, + .release = seq_release, +}; + +static int __init proc_cpuinfo_init(void) +{ + proc_create("cpuinfo", 0, NULL, &proc_cpuinfo_operations); + return 0; +} +module_init(proc_cpuinfo_init); diff --git a/fs/proc/devices.c b/fs/proc/devices.c new file mode 100644 index 00000000000..59ee7da959c --- /dev/null +++ b/fs/proc/devices.c @@ -0,0 +1,70 @@ +#include <linux/fs.h> +#include <linux/init.h> +#include <linux/proc_fs.h> +#include <linux/seq_file.h> + +static int devinfo_show(struct seq_file *f, void *v) +{ + int i = *(loff_t *) v; + + if (i < CHRDEV_MAJOR_HASH_SIZE) { + if (i == 0) + seq_printf(f, "Character devices:\n"); + chrdev_show(f, i); + } +#ifdef CONFIG_BLOCK + else { + i -= CHRDEV_MAJOR_HASH_SIZE; + if (i == 0) + seq_printf(f, "\nBlock devices:\n"); + blkdev_show(f, i); + } +#endif + return 0; +} + +static void *devinfo_start(struct seq_file *f, loff_t *pos) +{ + if (*pos < (BLKDEV_MAJOR_HASH_SIZE + CHRDEV_MAJOR_HASH_SIZE)) + return pos; + return NULL; +} + +static void *devinfo_next(struct seq_file *f, void *v, loff_t *pos) +{ + (*pos)++; + if (*pos >= (BLKDEV_MAJOR_HASH_SIZE + CHRDEV_MAJOR_HASH_SIZE)) + return NULL; + return pos; +} + +static void devinfo_stop(struct seq_file *f, void *v) +{ + /* Nothing to do */ +} + +static const struct seq_operations devinfo_ops = { + .start = devinfo_start, + .next = devinfo_next, + .stop = devinfo_stop, + .show = devinfo_show +}; + +static int devinfo_open(struct inode *inode, struct file *filp) +{ + return seq_open(filp, &devinfo_ops); +} + +static const struct file_operations proc_devinfo_operations = { + .open = devinfo_open, + .read = seq_read, + .llseek = seq_lseek, + .release = seq_release, +}; + +static int __init proc_devices_init(void) +{ + proc_create("devices", 0, NULL, &proc_devinfo_operations); + return 0; +} +module_init(proc_devices_init); diff --git a/fs/proc/generic.c b/fs/proc/generic.c index 7821589a17d..60a359b3558 100644 --- a/fs/proc/generic.c +++ b/fs/proc/generic.c @@ -547,9 +547,8 @@ static int proc_register(struct proc_dir_entry * dir, struct proc_dir_entry * dp for (tmp = dir->subdir; tmp; tmp = tmp->next) if (strcmp(tmp->name, dp->name) == 0) { - printk(KERN_WARNING "proc_dir_entry '%s/%s' already registered\n", + WARN(1, KERN_WARNING "proc_dir_entry '%s/%s' already registered\n", dir->name, dp->name); - dump_stack(); break; } diff --git a/fs/proc/inode.c b/fs/proc/inode.c index c6b4fa7e3b4..2543fd00c65 100644 --- a/fs/proc/inode.c +++ b/fs/proc/inode.c @@ -106,14 +106,13 @@ static void init_once(void *foo) inode_init_once(&ei->vfs_inode); } -int __init proc_init_inodecache(void) +void __init proc_init_inodecache(void) { proc_inode_cachep = kmem_cache_create("proc_inode_cache", sizeof(struct proc_inode), 0, (SLAB_RECLAIM_ACCOUNT| SLAB_MEM_SPREAD|SLAB_PANIC), init_once); - return 0; } static const struct super_operations proc_sops = { diff --git a/fs/proc/internal.h b/fs/proc/internal.h index 3bfb7b8747b..3e8aeb8b61c 100644 --- a/fs/proc/internal.h +++ b/fs/proc/internal.h @@ -61,12 +61,11 @@ extern const struct file_operations proc_smaps_operations; extern const struct file_operations proc_clear_refs_operations; extern const struct file_operations proc_pagemap_operations; extern const struct file_operations proc_net_operations; -extern const struct file_operations proc_kmsg_operations; extern const struct inode_operations proc_net_inode_operations; void free_proc_entry(struct proc_dir_entry *de); -int proc_init_inodecache(void); +void proc_init_inodecache(void); static inline struct pid *proc_pid(struct inode *inode) { diff --git a/fs/proc/interrupts.c b/fs/proc/interrupts.c new file mode 100644 index 00000000000..05029c0e2f2 --- /dev/null +++ b/fs/proc/interrupts.c @@ -0,0 +1,53 @@ +#include <linux/fs.h> +#include <linux/init.h> +#include <linux/interrupt.h> +#include <linux/irqnr.h> +#include <linux/proc_fs.h> +#include <linux/seq_file.h> + +/* + * /proc/interrupts + */ +static void *int_seq_start(struct seq_file *f, loff_t *pos) +{ + return (*pos <= nr_irqs) ? pos : NULL; +} + +static void *int_seq_next(struct seq_file *f, void *v, loff_t *pos) +{ + (*pos)++; + if (*pos > nr_irqs) + return NULL; + return pos; +} + +static void int_seq_stop(struct seq_file *f, void *v) +{ + /* Nothing to do */ +} + +static const struct seq_operations int_seq_ops = { + .start = int_seq_start, + .next = int_seq_next, + .stop = int_seq_stop, + .show = show_interrupts +}; + +static int interrupts_open(struct inode *inode, struct file *filp) +{ + return seq_open(filp, &int_seq_ops); +} + +static const struct file_operations proc_interrupts_operations = { + .open = interrupts_open, + .read = seq_read, + .llseek = seq_lseek, + .release = seq_release, +}; + +static int __init proc_interrupts_init(void) +{ + proc_create("interrupts", 0, NULL, &proc_interrupts_operations); + return 0; +} +module_init(proc_interrupts_init); diff --git a/fs/proc/kcore.c b/fs/proc/kcore.c index c2370c76fb7..59b43a06887 100644 --- a/fs/proc/kcore.c +++ b/fs/proc/kcore.c @@ -27,6 +27,8 @@ #define ELF_CORE_EFLAGS 0 #endif +static struct proc_dir_entry *proc_root_kcore; + static int open_kcore(struct inode * inode, struct file * filp) { return capable(CAP_SYS_RAWIO) ? 0 : -EPERM; @@ -34,7 +36,7 @@ static int open_kcore(struct inode * inode, struct file * filp) static ssize_t read_kcore(struct file *, char __user *, size_t, loff_t *); -const struct file_operations proc_kcore_operations = { +static const struct file_operations proc_kcore_operations = { .read = read_kcore, .open = open_kcore, }; @@ -399,3 +401,13 @@ read_kcore(struct file *file, char __user *buffer, size_t buflen, loff_t *fpos) return acc; } + +static int __init proc_kcore_init(void) +{ + proc_root_kcore = proc_create("kcore", S_IRUSR, NULL, &proc_kcore_operations); + if (proc_root_kcore) + proc_root_kcore->size = + (size_t)high_memory - PAGE_OFFSET + PAGE_SIZE; + return 0; +} +module_init(proc_kcore_init); diff --git a/fs/proc/kmsg.c b/fs/proc/kmsg.c index 9fd5df3f40c..7ca78346d3f 100644 --- a/fs/proc/kmsg.c +++ b/fs/proc/kmsg.c @@ -10,13 +10,12 @@ #include <linux/time.h> #include <linux/kernel.h> #include <linux/poll.h> +#include <linux/proc_fs.h> #include <linux/fs.h> #include <asm/uaccess.h> #include <asm/io.h> -#include "internal.h" - extern wait_queue_head_t log_wait; extern int do_syslog(int type, char __user *bug, int count); @@ -49,9 +48,16 @@ static unsigned int kmsg_poll(struct file *file, poll_table *wait) } -const struct file_operations proc_kmsg_operations = { +static const struct file_operations proc_kmsg_operations = { .read = kmsg_read, .poll = kmsg_poll, .open = kmsg_open, .release = kmsg_release, }; + +static int __init proc_kmsg_init(void) +{ + proc_create("kmsg", S_IRUSR, NULL, &proc_kmsg_operations); + return 0; +} +module_init(proc_kmsg_init); diff --git a/fs/proc/loadavg.c b/fs/proc/loadavg.c new file mode 100644 index 00000000000..9bca39cf99e --- /dev/null +++ b/fs/proc/loadavg.c @@ -0,0 +1,51 @@ +#include <linux/fs.h> +#include <linux/init.h> +#include <linux/pid_namespace.h> +#include <linux/proc_fs.h> +#include <linux/sched.h> +#include <linux/seq_file.h> +#include <linux/seqlock.h> +#include <linux/time.h> + +#define LOAD_INT(x) ((x) >> FSHIFT) +#define LOAD_FRAC(x) LOAD_INT(((x) & (FIXED_1-1)) * 100) + +static int loadavg_proc_show(struct seq_file *m, void *v) +{ + int a, b, c; + unsigned long seq; + + do { + seq = read_seqbegin(&xtime_lock); + a = avenrun[0] + (FIXED_1/200); + b = avenrun[1] + (FIXED_1/200); + c = avenrun[2] + (FIXED_1/200); + } while (read_seqretry(&xtime_lock, seq)); + + seq_printf(m, "%d.%02d %d.%02d %d.%02d %ld/%d %d\n", + LOAD_INT(a), LOAD_FRAC(a), + LOAD_INT(b), LOAD_FRAC(b), + LOAD_INT(c), LOAD_FRAC(c), + nr_running(), nr_threads, + task_active_pid_ns(current)->last_pid); + return 0; +} + +static int loadavg_proc_open(struct inode *inode, struct file *file) +{ + return single_open(file, loadavg_proc_show, NULL); +} + +static const struct file_operations loadavg_proc_fops = { + .open = loadavg_proc_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, +}; + +static int __init proc_loadavg_init(void) +{ + proc_create("loadavg", 0, NULL, &loadavg_proc_fops); + return 0; +} +module_init(proc_loadavg_init); diff --git a/fs/proc/meminfo.c b/fs/proc/meminfo.c new file mode 100644 index 00000000000..b1675c4e66d --- /dev/null +++ b/fs/proc/meminfo.c @@ -0,0 +1,168 @@ +#include <linux/fs.h> +#include <linux/hugetlb.h> +#include <linux/init.h> +#include <linux/kernel.h> +#include <linux/mm.h> +#include <linux/mman.h> +#include <linux/mmzone.h> +#include <linux/proc_fs.h> +#include <linux/quicklist.h> +#include <linux/seq_file.h> +#include <linux/swap.h> +#include <linux/vmstat.h> +#include <asm/atomic.h> +#include <asm/page.h> +#include <asm/pgtable.h> +#include "internal.h" + +void __attribute__((weak)) arch_report_meminfo(struct seq_file *m) +{ +} + +static int meminfo_proc_show(struct seq_file *m, void *v) +{ + struct sysinfo i; + unsigned long committed; + unsigned long allowed; + struct vmalloc_info vmi; + long cached; + unsigned long pages[NR_LRU_LISTS]; + int lru; + +/* + * display in kilobytes. + */ +#define K(x) ((x) << (PAGE_SHIFT - 10)) + si_meminfo(&i); + si_swapinfo(&i); + committed = atomic_long_read(&vm_committed_space); + allowed = ((totalram_pages - hugetlb_total_pages()) + * sysctl_overcommit_ratio / 100) + total_swap_pages; + + cached = global_page_state(NR_FILE_PAGES) - + total_swapcache_pages - i.bufferram; + if (cached < 0) + cached = 0; + + get_vmalloc_info(&vmi); + + for (lru = LRU_BASE; lru < NR_LRU_LISTS; lru++) + pages[lru] = global_page_state(NR_LRU_BASE + lru); + + /* + * Tagged format, for easy grepping and expansion. + */ + seq_printf(m, + "MemTotal: %8lu kB\n" + "MemFree: %8lu kB\n" + "Buffers: %8lu kB\n" + "Cached: %8lu kB\n" + "SwapCached: %8lu kB\n" + "Active: %8lu kB\n" + "Inactive: %8lu kB\n" + "Active(anon): %8lu kB\n" + "Inactive(anon): %8lu kB\n" + "Active(file): %8lu kB\n" + "Inactive(file): %8lu kB\n" +#ifdef CONFIG_UNEVICTABLE_LRU + "Unevictable: %8lu kB\n" + "Mlocked: %8lu kB\n" +#endif +#ifdef CONFIG_HIGHMEM + "HighTotal: %8lu kB\n" + "HighFree: %8lu kB\n" + "LowTotal: %8lu kB\n" + "LowFree: %8lu kB\n" +#endif + "SwapTotal: %8lu kB\n" + "SwapFree: %8lu kB\n" + "Dirty: %8lu kB\n" + "Writeback: %8lu kB\n" + "AnonPages: %8lu kB\n" + "Mapped: %8lu kB\n" + "Slab: %8lu kB\n" + "SReclaimable: %8lu kB\n" + "SUnreclaim: %8lu kB\n" + "PageTables: %8lu kB\n" +#ifdef CONFIG_QUICKLIST + "Quicklists: %8lu kB\n" +#endif + "NFS_Unstable: %8lu kB\n" + "Bounce: %8lu kB\n" + "WritebackTmp: %8lu kB\n" + "CommitLimit: %8lu kB\n" + "Committed_AS: %8lu kB\n" + "VmallocTotal: %8lu kB\n" + "VmallocUsed: %8lu kB\n" + "VmallocChunk: %8lu kB\n", + K(i.totalram), + K(i.freeram), + K(i.bufferram), + K(cached), + K(total_swapcache_pages), + K(pages[LRU_ACTIVE_ANON] + pages[LRU_ACTIVE_FILE]), + K(pages[LRU_INACTIVE_ANON] + pages[LRU_INACTIVE_FILE]), + K(pages[LRU_ACTIVE_ANON]), + K(pages[LRU_INACTIVE_ANON]), + K(pages[LRU_ACTIVE_FILE]), + K(pages[LRU_INACTIVE_FILE]), +#ifdef CONFIG_UNEVICTABLE_LRU + K(pages[LRU_UNEVICTABLE]), + K(global_page_state(NR_MLOCK)), +#endif +#ifdef CONFIG_HIGHMEM + K(i.totalhigh), + K(i.freehigh), + K(i.totalram-i.totalhigh), + K(i.freeram-i.freehigh), +#endif + K(i.totalswap), + K(i.freeswap), + K(global_page_state(NR_FILE_DIRTY)), + K(global_page_state(NR_WRITEBACK)), + K(global_page_state(NR_ANON_PAGES)), + K(global_page_state(NR_FILE_MAPPED)), + K(global_page_state(NR_SLAB_RECLAIMABLE) + + global_page_state(NR_SLAB_UNRECLAIMABLE)), + K(global_page_state(NR_SLAB_RECLAIMABLE)), + K(global_page_state(NR_SLAB_UNRECLAIMABLE)), + K(global_page_state(NR_PAGETABLE)), +#ifdef CONFIG_QUICKLIST + K(quicklist_total_size()), +#endif + K(global_page_state(NR_UNSTABLE_NFS)), + K(global_page_state(NR_BOUNCE)), + K(global_page_state(NR_WRITEBACK_TEMP)), + K(allowed), + K(committed), + (unsigned long)VMALLOC_TOTAL >> 10, + vmi.used >> 10, + vmi.largest_chunk >> 10 + ); + + hugetlb_report_meminfo(m); + + arch_report_meminfo(m); + + return 0; +#undef K +} + +static int meminfo_proc_open(struct inode *inode, struct file *file) +{ + return single_open(file, meminfo_proc_show, NULL); +} + +static const struct file_operations meminfo_proc_fops = { + .open = meminfo_proc_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, +}; + +static int __init proc_meminfo_init(void) +{ + proc_create("meminfo", 0, NULL, &meminfo_proc_fops); + return 0; +} +module_init(proc_meminfo_init); diff --git a/fs/proc/page.c b/fs/proc/page.c new file mode 100644 index 00000000000..767d95a6d1b --- /dev/null +++ b/fs/proc/page.c @@ -0,0 +1,147 @@ +#include <linux/bootmem.h> +#include <linux/compiler.h> +#include <linux/fs.h> +#include <linux/init.h> +#include <linux/mm.h> +#include <linux/mmzone.h> +#include <linux/proc_fs.h> +#include <linux/seq_file.h> +#include <asm/uaccess.h> +#include "internal.h" + +#define KPMSIZE sizeof(u64) +#define KPMMASK (KPMSIZE - 1) +/* /proc/kpagecount - an array exposing page counts + * + * Each entry is a u64 representing the corresponding + * physical page count. + */ +static ssize_t kpagecount_read(struct file *file, char __user *buf, + size_t count, loff_t *ppos) +{ + u64 __user *out = (u64 __user *)buf; + struct page *ppage; + unsigned long src = *ppos; + unsigned long pfn; + ssize_t ret = 0; + u64 pcount; + + pfn = src / KPMSIZE; + count = min_t(size_t, count, (max_pfn * KPMSIZE) - src); + if (src & KPMMASK || count & KPMMASK) + return -EINVAL; + + while (count > 0) { + ppage = NULL; + if (pfn_valid(pfn)) + ppage = pfn_to_page(pfn); + pfn++; + if (!ppage) + pcount = 0; + else + pcount = page_mapcount(ppage); + + if (put_user(pcount, out++)) { + ret = -EFAULT; + break; + } + + count -= KPMSIZE; + } + + *ppos += (char __user *)out - buf; + if (!ret) + ret = (char __user *)out - buf; + return ret; +} + +static const struct file_operations proc_kpagecount_operations = { + .llseek = mem_lseek, + .read = kpagecount_read, +}; + +/* /proc/kpageflags - an array exposing page flags + * + * Each entry is a u64 representing the corresponding + * physical page flags. + */ + +/* These macros are used to decouple internal flags from exported ones */ + +#define KPF_LOCKED 0 +#define KPF_ERROR 1 +#define KPF_REFERENCED 2 +#define KPF_UPTODATE 3 +#define KPF_DIRTY 4 +#define KPF_LRU 5 +#define KPF_ACTIVE 6 +#define KPF_SLAB 7 +#define KPF_WRITEBACK 8 +#define KPF_RECLAIM 9 +#define KPF_BUDDY 10 + +#define kpf_copy_bit(flags, srcpos, dstpos) (((flags >> srcpos) & 1) << dstpos) + +static ssize_t kpageflags_read(struct file *file, char __user *buf, + size_t count, loff_t *ppos) +{ + u64 __user *out = (u64 __user *)buf; + struct page *ppage; + unsigned long src = *ppos; + unsigned long pfn; + ssize_t ret = 0; + u64 kflags, uflags; + + pfn = src / KPMSIZE; + count = min_t(unsigned long, count, (max_pfn * KPMSIZE) - src); + if (src & KPMMASK || count & KPMMASK) + return -EINVAL; + + while (count > 0) { + ppage = NULL; + if (pfn_valid(pfn)) + ppage = pfn_to_page(pfn); + pfn++; + if (!ppage) + kflags = 0; + else + kflags = ppage->flags; + + uflags = kpf_copy_bit(KPF_LOCKED, PG_locked, kflags) | + kpf_copy_bit(kflags, KPF_ERROR, PG_error) | + kpf_copy_bit(kflags, KPF_REFERENCED, PG_referenced) | + kpf_copy_bit(kflags, KPF_UPTODATE, PG_uptodate) | + kpf_copy_bit(kflags, KPF_DIRTY, PG_dirty) | + kpf_copy_bit(kflags, KPF_LRU, PG_lru) | + kpf_copy_bit(kflags, KPF_ACTIVE, PG_active) | + kpf_copy_bit(kflags, KPF_SLAB, PG_slab) | + kpf_copy_bit(kflags, KPF_WRITEBACK, PG_writeback) | + kpf_copy_bit(kflags, KPF_RECLAIM, PG_reclaim) | + kpf_copy_bit(kflags, KPF_BUDDY, PG_buddy); + + if (put_user(uflags, out++)) { + ret = -EFAULT; + break; + } + + count -= KPMSIZE; + } + + *ppos += (char __user *)out - buf; + if (!ret) + ret = (char __user *)out - buf; + return ret; +} + +static const struct file_operations proc_kpageflags_operations = { + .llseek = mem_lseek, + .read = kpageflags_read, +}; + +static int __init proc_page_init(void) +{ + proc_create("kpagecount", S_IRUSR, NULL, &proc_kpagecount_operations); + proc_create("kpageflags", S_IRUSR, NULL, &proc_kpageflags_operations); + return 0; +} +module_init(proc_page_init); diff --git a/fs/proc/proc_devtree.c b/fs/proc/proc_devtree.c index eca471bc851..de2bba5a344 100644 --- a/fs/proc/proc_devtree.c +++ b/fs/proc/proc_devtree.c @@ -4,6 +4,7 @@ * Copyright 1997 Paul Mackerras */ #include <linux/errno.h> +#include <linux/init.h> #include <linux/time.h> #include <linux/proc_fs.h> #include <linux/stat.h> @@ -214,11 +215,10 @@ void proc_device_tree_add_node(struct device_node *np, /* * Called on initialization to set up the /proc/device-tree subtree */ -void proc_device_tree_init(void) +void __init proc_device_tree_init(void) { struct device_node *root; - if ( !have_of ) - return; + proc_device_tree = proc_mkdir("device-tree", NULL); if (proc_device_tree == 0) return; diff --git a/fs/proc/proc_misc.c b/fs/proc/proc_misc.c deleted file mode 100644 index b675a49c182..00000000000 --- a/fs/proc/proc_misc.c +++ /dev/null @@ -1,940 +0,0 @@ -/* - * linux/fs/proc/proc_misc.c - * - * linux/fs/proc/array.c - * Copyright (C) 1992 by Linus Torvalds - * based on ideas by Darren Senn - * - * This used to be the part of array.c. See the rest of history and credits - * there. I took this into a separate file and switched the thing to generic - * proc_file_inode_operations, leaving in array.c only per-process stuff. - * Inumbers allocation made dynamic (via create_proc_entry()). AV, May 1999. - * - * Changes: - * Fulton Green : Encapsulated position metric calculations. - * <kernel@FultonGreen.com> - */ - -#include <linux/types.h> -#include <linux/errno.h> -#include <linux/time.h> -#include <linux/kernel.h> -#include <linux/kernel_stat.h> -#include <linux/fs.h> -#include <linux/tty.h> -#include <linux/string.h> -#include <linux/mman.h> -#include <linux/quicklist.h> -#include <linux/proc_fs.h> -#include <linux/ioport.h> -#include <linux/mm.h> -#include <linux/mmzone.h> -#include <linux/pagemap.h> -#include <linux/interrupt.h> -#include <linux/swap.h> -#include <linux/slab.h> -#include <linux/genhd.h> -#include <linux/smp.h> -#include <linux/signal.h> -#include <linux/module.h> -#include <linux/init.h> -#include <linux/seq_file.h> -#include <linux/times.h> -#include <linux/profile.h> -#include <linux/utsname.h> -#include <linux/blkdev.h> -#include <linux/hugetlb.h> -#include <linux/jiffies.h> -#include <linux/sysrq.h> -#include <linux/vmalloc.h> -#include <linux/crash_dump.h> -#include <linux/pid_namespace.h> -#include <linux/bootmem.h> -#include <asm/uaccess.h> -#include <asm/pgtable.h> -#include <asm/io.h> -#include <asm/tlb.h> -#include <asm/div64.h> -#include "internal.h" - -#define LOAD_INT(x) ((x) >> FSHIFT) -#define LOAD_FRAC(x) LOAD_INT(((x) & (FIXED_1-1)) * 100) -/* - * Warning: stuff below (imported functions) assumes that its output will fit - * into one page. For some of those functions it may be wrong. Moreover, we - * have a way to deal with that gracefully. Right now I used straightforward - * wrappers, but this needs further analysis wrt potential overflows. - */ -extern int get_hardware_list(char *); -extern int get_stram_list(char *); -extern int get_exec_domain_list(char *); - -static int proc_calc_metrics(char *page, char **start, off_t off, - int count, int *eof, int len) -{ - if (len <= off+count) *eof = 1; - *start = page + off; - len -= off; - if (len>count) len = count; - if (len<0) len = 0; - return len; -} - -static int loadavg_read_proc(char *page, char **start, off_t off, - int count, int *eof, void *data) -{ - int a, b, c; - int len; - unsigned long seq; - - do { - seq = read_seqbegin(&xtime_lock); - a = avenrun[0] + (FIXED_1/200); - b = avenrun[1] + (FIXED_1/200); - c = avenrun[2] + (FIXED_1/200); - } while (read_seqretry(&xtime_lock, seq)); - - len = sprintf(page,"%d.%02d %d.%02d %d.%02d %ld/%d %d\n", - LOAD_INT(a), LOAD_FRAC(a), - LOAD_INT(b), LOAD_FRAC(b), - LOAD_INT(c), LOAD_FRAC(c), - nr_running(), nr_threads, - task_active_pid_ns(current)->last_pid); - return proc_calc_metrics(page, start, off, count, eof, len); -} - -static int uptime_read_proc(char *page, char **start, off_t off, - int count, int *eof, void *data) -{ - struct timespec uptime; - struct timespec idle; - int len; - cputime_t idletime = cputime_add(init_task.utime, init_task.stime); - - do_posix_clock_monotonic_gettime(&uptime); - monotonic_to_bootbased(&uptime); - cputime_to_timespec(idletime, &idle); - len = sprintf(page,"%lu.%02lu %lu.%02lu\n", - (unsigned long) uptime.tv_sec, - (uptime.tv_nsec / (NSEC_PER_SEC / 100)), - (unsigned long) idle.tv_sec, - (idle.tv_nsec / (NSEC_PER_SEC / 100))); - - return proc_calc_metrics(page, start, off, count, eof, len); -} - -int __attribute__((weak)) arch_report_meminfo(char *page) -{ - return 0; -} - -static int meminfo_read_proc(char *page, char **start, off_t off, - int count, int *eof, void *data) -{ - struct sysinfo i; - int len; - unsigned long committed; - unsigned long allowed; - struct vmalloc_info vmi; - long cached; - -/* - * display in kilobytes. - */ -#define K(x) ((x) << (PAGE_SHIFT - 10)) - si_meminfo(&i); - si_swapinfo(&i); - committed = atomic_long_read(&vm_committed_space); - allowed = ((totalram_pages - hugetlb_total_pages()) - * sysctl_overcommit_ratio / 100) + total_swap_pages; - - cached = global_page_state(NR_FILE_PAGES) - - total_swapcache_pages - i.bufferram; - if (cached < 0) - cached = 0; - - get_vmalloc_info(&vmi); - - /* - * Tagged format, for easy grepping and expansion. - */ - len = sprintf(page, - "MemTotal: %8lu kB\n" - "MemFree: %8lu kB\n" - "Buffers: %8lu kB\n" - "Cached: %8lu kB\n" - "SwapCached: %8lu kB\n" - "Active: %8lu kB\n" - "Inactive: %8lu kB\n" -#ifdef CONFIG_HIGHMEM - "HighTotal: %8lu kB\n" - "HighFree: %8lu kB\n" - "LowTotal: %8lu kB\n" - "LowFree: %8lu kB\n" -#endif - "SwapTotal: %8lu kB\n" - "SwapFree: %8lu kB\n" - "Dirty: %8lu kB\n" - "Writeback: %8lu kB\n" - "AnonPages: %8lu kB\n" - "Mapped: %8lu kB\n" - "Slab: %8lu kB\n" - "SReclaimable: %8lu kB\n" - "SUnreclaim: %8lu kB\n" - "PageTables: %8lu kB\n" -#ifdef CONFIG_QUICKLIST - "Quicklists: %8lu kB\n" -#endif - "NFS_Unstable: %8lu kB\n" - "Bounce: %8lu kB\n" - "WritebackTmp: %8lu kB\n" - "CommitLimit: %8lu kB\n" - "Committed_AS: %8lu kB\n" - "VmallocTotal: %8lu kB\n" - "VmallocUsed: %8lu kB\n" - "VmallocChunk: %8lu kB\n", - K(i.totalram), - K(i.freeram), - K(i.bufferram), - K(cached), - K(total_swapcache_pages), - K(global_page_state(NR_ACTIVE)), - K(global_page_state(NR_INACTIVE)), -#ifdef CONFIG_HIGHMEM - K(i.totalhigh), - K(i.freehigh), - K(i.totalram-i.totalhigh), - K(i.freeram-i.freehigh), -#endif - K(i.totalswap), - K(i.freeswap), - K(global_page_state(NR_FILE_DIRTY)), - K(global_page_state(NR_WRITEBACK)), - K(global_page_state(NR_ANON_PAGES)), - K(global_page_state(NR_FILE_MAPPED)), - K(global_page_state(NR_SLAB_RECLAIMABLE) + - global_page_state(NR_SLAB_UNRECLAIMABLE)), - K(global_page_state(NR_SLAB_RECLAIMABLE)), - K(global_page_state(NR_SLAB_UNRECLAIMABLE)), - K(global_page_state(NR_PAGETABLE)), -#ifdef CONFIG_QUICKLIST - K(quicklist_total_size()), -#endif - K(global_page_state(NR_UNSTABLE_NFS)), - K(global_page_state(NR_BOUNCE)), - K(global_page_state(NR_WRITEBACK_TEMP)), - K(allowed), - K(committed), - (unsigned long)VMALLOC_TOTAL >> 10, - vmi.used >> 10, - vmi.largest_chunk >> 10 - ); - - len += hugetlb_report_meminfo(page + len); - - len += arch_report_meminfo(page + len); - - return proc_calc_metrics(page, start, off, count, eof, len); -#undef K -} - -static int fragmentation_open(struct inode *inode, struct file *file) -{ - (void)inode; - return seq_open(file, &fragmentation_op); -} - -static const struct file_operations fragmentation_file_operations = { - .open = fragmentation_open, - .read = seq_read, - .llseek = seq_lseek, - .release = seq_release, -}; - -static int pagetypeinfo_open(struct inode *inode, struct file *file) -{ - return seq_open(file, &pagetypeinfo_op); -} - -static const struct file_operations pagetypeinfo_file_ops = { - .open = pagetypeinfo_open, - .read = seq_read, - .llseek = seq_lseek, - .release = seq_release, -}; - -static int zoneinfo_open(struct inode *inode, struct file *file) -{ - return seq_open(file, &zoneinfo_op); -} - -static const struct file_operations proc_zoneinfo_file_operations = { - .open = zoneinfo_open, - .read = seq_read, - .llseek = seq_lseek, - .release = seq_release, -}; - -static int version_read_proc(char *page, char **start, off_t off, - int count, int *eof, void *data) -{ - int len; - - len = snprintf(page, PAGE_SIZE, linux_proc_banner, - utsname()->sysname, - utsname()->release, - utsname()->version); - return proc_calc_metrics(page, start, off, count, eof, len); -} - -extern const struct seq_operations cpuinfo_op; -static int cpuinfo_open(struct inode *inode, struct file *file) -{ - return seq_open(file, &cpuinfo_op); -} - -static const struct file_operations proc_cpuinfo_operations = { - .open = cpuinfo_open, - .read = seq_read, - .llseek = seq_lseek, - .release = seq_release, -}; - -static int devinfo_show(struct seq_file *f, void *v) -{ - int i = *(loff_t *) v; - - if (i < CHRDEV_MAJOR_HASH_SIZE) { - if (i == 0) - seq_printf(f, "Character devices:\n"); - chrdev_show(f, i); - } -#ifdef CONFIG_BLOCK - else { - i -= CHRDEV_MAJOR_HASH_SIZE; - if (i == 0) - seq_printf(f, "\nBlock devices:\n"); - blkdev_show(f, i); - } -#endif - return 0; -} - -static void *devinfo_start(struct seq_file *f, loff_t *pos) -{ - if (*pos < (BLKDEV_MAJOR_HASH_SIZE + CHRDEV_MAJOR_HASH_SIZE)) - return pos; - return NULL; -} - -static void *devinfo_next(struct seq_file *f, void *v, loff_t *pos) -{ - (*pos)++; - if (*pos >= (BLKDEV_MAJOR_HASH_SIZE + CHRDEV_MAJOR_HASH_SIZE)) - return NULL; - return pos; -} - -static void devinfo_stop(struct seq_file *f, void *v) -{ - /* Nothing to do */ -} - -static const struct seq_operations devinfo_ops = { - .start = devinfo_start, - .next = devinfo_next, - .stop = devinfo_stop, - .show = devinfo_show -}; - -static int devinfo_open(struct inode *inode, struct file *filp) -{ - return seq_open(filp, &devinfo_ops); -} - -static const struct file_operations proc_devinfo_operations = { - .open = devinfo_open, - .read = seq_read, - .llseek = seq_lseek, - .release = seq_release, -}; - -static int vmstat_open(struct inode *inode, struct file *file) -{ - return seq_open(file, &vmstat_op); -} -static const struct file_operations proc_vmstat_file_operations = { - .open = vmstat_open, - .read = seq_read, - .llseek = seq_lseek, - .release = seq_release, -}; - -#ifdef CONFIG_PROC_HARDWARE -static int hardware_read_proc(char *page, char **start, off_t off, - int count, int *eof, void *data) -{ - int len = get_hardware_list(page); - return proc_calc_metrics(page, start, off, count, eof, len); -} -#endif - -#ifdef CONFIG_STRAM_PROC -static int stram_read_proc(char *page, char **start, off_t off, - int count, int *eof, void *data) -{ - int len = get_stram_list(page); - return proc_calc_metrics(page, start, off, count, eof, len); -} -#endif - -#ifdef CONFIG_BLOCK -static int partitions_open(struct inode *inode, struct file *file) -{ - return seq_open(file, &partitions_op); -} -static const struct file_operations proc_partitions_operations = { - .open = partitions_open, - .read = seq_read, - .llseek = seq_lseek, - .release = seq_release, -}; - -static int diskstats_open(struct inode *inode, struct file *file) -{ - return seq_open(file, &diskstats_op); -} -static const struct file_operations proc_diskstats_operations = { - .open = diskstats_open, - .read = seq_read, - .llseek = seq_lseek, - .release = seq_release, -}; -#endif - -#ifdef CONFIG_MODULES -extern const struct seq_operations modules_op; -static int modules_open(struct inode *inode, struct file *file) -{ - return seq_open(file, &modules_op); -} -static const struct file_operations proc_modules_operations = { - .open = modules_open, - .read = seq_read, - .llseek = seq_lseek, - .release = seq_release, -}; -#endif - -#ifdef CONFIG_SLABINFO -static int slabinfo_open(struct inode *inode, struct file *file) -{ - return seq_open(file, &slabinfo_op); -} -static const struct file_operations proc_slabinfo_operations = { - .open = slabinfo_open, - .read = seq_read, - .write = slabinfo_write, - .llseek = seq_lseek, - .release = seq_release, -}; - -#ifdef CONFIG_DEBUG_SLAB_LEAK -extern const struct seq_operations slabstats_op; -static int slabstats_open(struct inode *inode, struct file *file) -{ - unsigned long *n = kzalloc(PAGE_SIZE, GFP_KERNEL); - int ret = -ENOMEM; - if (n) { - ret = seq_open(file, &slabstats_op); - if (!ret) { - struct seq_file *m = file->private_data; - *n = PAGE_SIZE / (2 * sizeof(unsigned long)); - m->private = n; - n = NULL; - } - kfree(n); - } - return ret; -} - -static const struct file_operations proc_slabstats_operations = { - .open = slabstats_open, - .read = seq_read, - .llseek = seq_lseek, - .release = seq_release_private, -}; -#endif -#endif - -#ifdef CONFIG_MMU -static int vmalloc_open(struct inode *inode, struct file *file) -{ - unsigned int *ptr = NULL; - int ret; - - if (NUMA_BUILD) - ptr = kmalloc(nr_node_ids * sizeof(unsigned int), GFP_KERNEL); - ret = seq_open(file, &vmalloc_op); - if (!ret) { - struct seq_file *m = file->private_data; - m->private = ptr; - } else - kfree(ptr); - return ret; -} - -static const struct file_operations proc_vmalloc_operations = { - .open = vmalloc_open, - .read = seq_read, - .llseek = seq_lseek, - .release = seq_release_private, -}; -#endif - -#ifndef arch_irq_stat_cpu -#define arch_irq_stat_cpu(cpu) 0 -#endif -#ifndef arch_irq_stat -#define arch_irq_stat() 0 -#endif - -static int show_stat(struct seq_file *p, void *v) -{ - int i; - unsigned long jif; - cputime64_t user, nice, system, idle, iowait, irq, softirq, steal; - cputime64_t guest; - u64 sum = 0; - struct timespec boottime; - unsigned int *per_irq_sum; - - per_irq_sum = kzalloc(sizeof(unsigned int)*NR_IRQS, GFP_KERNEL); - if (!per_irq_sum) - return -ENOMEM; - - user = nice = system = idle = iowait = - irq = softirq = steal = cputime64_zero; - guest = cputime64_zero; - getboottime(&boottime); - jif = boottime.tv_sec; - - for_each_possible_cpu(i) { - int j; - - user = cputime64_add(user, kstat_cpu(i).cpustat.user); - nice = cputime64_add(nice, kstat_cpu(i).cpustat.nice); - system = cputime64_add(system, kstat_cpu(i).cpustat.system); - idle = cputime64_add(idle, kstat_cpu(i).cpustat.idle); - iowait = cputime64_add(iowait, kstat_cpu(i).cpustat.iowait); - irq = cputime64_add(irq, kstat_cpu(i).cpustat.irq); - softirq = cputime64_add(softirq, kstat_cpu(i).cpustat.softirq); - steal = cputime64_add(steal, kstat_cpu(i).cpustat.steal); - guest = cputime64_add(guest, kstat_cpu(i).cpustat.guest); - for (j = 0; j < NR_IRQS; j++) { - unsigned int temp = kstat_cpu(i).irqs[j]; - sum += temp; - per_irq_sum[j] += temp; - } - sum += arch_irq_stat_cpu(i); - } - sum += arch_irq_stat(); - - seq_printf(p, "cpu %llu %llu %llu %llu %llu %llu %llu %llu %llu\n", - (unsigned long long)cputime64_to_clock_t(user), - (unsigned long long)cputime64_to_clock_t(nice), - (unsigned long long)cputime64_to_clock_t(system), - (unsigned long long)cputime64_to_clock_t(idle), - (unsigned long long)cputime64_to_clock_t(iowait), - (unsigned long long)cputime64_to_clock_t(irq), - (unsigned long long)cputime64_to_clock_t(softirq), - (unsigned long long)cputime64_to_clock_t(steal), - (unsigned long long)cputime64_to_clock_t(guest)); - for_each_online_cpu(i) { - - /* Copy values here to work around gcc-2.95.3, gcc-2.96 */ - user = kstat_cpu(i).cpustat.user; - nice = kstat_cpu(i).cpustat.nice; - system = kstat_cpu(i).cpustat.system; - idle = kstat_cpu(i).cpustat.idle; - iowait = kstat_cpu(i).cpustat.iowait; - irq = kstat_cpu(i).cpustat.irq; - softirq = kstat_cpu(i).cpustat.softirq; - steal = kstat_cpu(i).cpustat.steal; - guest = kstat_cpu(i).cpustat.guest; - seq_printf(p, - "cpu%d %llu %llu %llu %llu %llu %llu %llu %llu %llu\n", - i, - (unsigned long long)cputime64_to_clock_t(user), - (unsigned long long)cputime64_to_clock_t(nice), - (unsigned long long)cputime64_to_clock_t(system), - (unsigned long long)cputime64_to_clock_t(idle), - (unsigned long long)cputime64_to_clock_t(iowait), - (unsigned long long)cputime64_to_clock_t(irq), - (unsigned long long)cputime64_to_clock_t(softirq), - (unsigned long long)cputime64_to_clock_t(steal), - (unsigned long long)cputime64_to_clock_t(guest)); - } - seq_printf(p, "intr %llu", (unsigned long long)sum); - - for (i = 0; i < NR_IRQS; i++) - seq_printf(p, " %u", per_irq_sum[i]); - - seq_printf(p, - "\nctxt %llu\n" - "btime %lu\n" - "processes %lu\n" - "procs_running %lu\n" - "procs_blocked %lu\n", - nr_context_switches(), - (unsigned long)jif, - total_forks, - nr_running(), - nr_iowait()); - - kfree(per_irq_sum); - return 0; -} - -static int stat_open(struct inode *inode, struct file *file) -{ - unsigned size = 4096 * (1 + num_possible_cpus() / 32); - char *buf; - struct seq_file *m; - int res; - - /* don't ask for more than the kmalloc() max size, currently 128 KB */ - if (size > 128 * 1024) - size = 128 * 1024; - buf = kmalloc(size, GFP_KERNEL); - if (!buf) - return -ENOMEM; - - res = single_open(file, show_stat, NULL); - if (!res) { - m = file->private_data; - m->buf = buf; - m->size = size; - } else - kfree(buf); - return res; -} -static const struct file_operations proc_stat_operations = { - .open = stat_open, - .read = seq_read, - .llseek = seq_lseek, - .release = single_release, -}; - -/* - * /proc/interrupts - */ -static void *int_seq_start(struct seq_file *f, loff_t *pos) -{ - return (*pos <= NR_IRQS) ? pos : NULL; -} - -static void *int_seq_next(struct seq_file *f, void *v, loff_t *pos) -{ - (*pos)++; - if (*pos > NR_IRQS) - return NULL; - return pos; -} - -static void int_seq_stop(struct seq_file *f, void *v) -{ - /* Nothing to do */ -} - - -static const struct seq_operations int_seq_ops = { - .start = int_seq_start, - .next = int_seq_next, - .stop = int_seq_stop, - .show = show_interrupts -}; - -static int interrupts_open(struct inode *inode, struct file *filp) -{ - return seq_open(filp, &int_seq_ops); -} - -static const struct file_operations proc_interrupts_operations = { - .open = interrupts_open, - .read = seq_read, - .llseek = seq_lseek, - .release = seq_release, -}; - -static int filesystems_read_proc(char *page, char **start, off_t off, - int count, int *eof, void *data) -{ - int len = get_filesystem_list(page); - return proc_calc_metrics(page, start, off, count, eof, len); -} - -static int cmdline_read_proc(char *page, char **start, off_t off, - int count, int *eof, void *data) -{ - int len; - - len = sprintf(page, "%s\n", saved_command_line); - return proc_calc_metrics(page, start, off, count, eof, len); -} - -#ifdef CONFIG_FILE_LOCKING -static int locks_open(struct inode *inode, struct file *filp) -{ - return seq_open(filp, &locks_seq_operations); -} - -static const struct file_operations proc_locks_operations = { - .open = locks_open, - .read = seq_read, - .llseek = seq_lseek, - .release = seq_release, -}; -#endif /* CONFIG_FILE_LOCKING */ - -static int execdomains_read_proc(char *page, char **start, off_t off, - int count, int *eof, void *data) -{ - int len = get_exec_domain_list(page); - return proc_calc_metrics(page, start, off, count, eof, len); -} - -#ifdef CONFIG_MAGIC_SYSRQ -/* - * writing 'C' to /proc/sysrq-trigger is like sysrq-C - */ -static ssize_t write_sysrq_trigger(struct file *file, const char __user *buf, - size_t count, loff_t *ppos) -{ - if (count) { - char c; - - if (get_user(c, buf)) - return -EFAULT; - __handle_sysrq(c, NULL, 0); - } - return count; -} - -static const struct file_operations proc_sysrq_trigger_operations = { - .write = write_sysrq_trigger, -}; -#endif - -#ifdef CONFIG_PROC_PAGE_MONITOR -#define KPMSIZE sizeof(u64) -#define KPMMASK (KPMSIZE - 1) -/* /proc/kpagecount - an array exposing page counts - * - * Each entry is a u64 representing the corresponding - * physical page count. - */ -static ssize_t kpagecount_read(struct file *file, char __user *buf, - size_t count, loff_t *ppos) -{ - u64 __user *out = (u64 __user *)buf; - struct page *ppage; - unsigned long src = *ppos; - unsigned long pfn; - ssize_t ret = 0; - u64 pcount; - - pfn = src / KPMSIZE; - count = min_t(size_t, count, (max_pfn * KPMSIZE) - src); - if (src & KPMMASK || count & KPMMASK) - return -EINVAL; - - while (count > 0) { - ppage = NULL; - if (pfn_valid(pfn)) - ppage = pfn_to_page(pfn); - pfn++; - if (!ppage) - pcount = 0; - else - pcount = page_mapcount(ppage); - - if (put_user(pcount, out++)) { - ret = -EFAULT; - break; - } - - count -= KPMSIZE; - } - - *ppos += (char __user *)out - buf; - if (!ret) - ret = (char __user *)out - buf; - return ret; -} - -static struct file_operations proc_kpagecount_operations = { - .llseek = mem_lseek, - .read = kpagecount_read, -}; - -/* /proc/kpageflags - an array exposing page flags - * - * Each entry is a u64 representing the corresponding - * physical page flags. - */ - -/* These macros are used to decouple internal flags from exported ones */ - -#define KPF_LOCKED 0 -#define KPF_ERROR 1 -#define KPF_REFERENCED 2 -#define KPF_UPTODATE 3 -#define KPF_DIRTY 4 -#define KPF_LRU 5 -#define KPF_ACTIVE 6 -#define KPF_SLAB 7 -#define KPF_WRITEBACK 8 -#define KPF_RECLAIM 9 -#define KPF_BUDDY 10 - -#define kpf_copy_bit(flags, srcpos, dstpos) (((flags >> srcpos) & 1) << dstpos) - -static ssize_t kpageflags_read(struct file *file, char __user *buf, - size_t count, loff_t *ppos) -{ - u64 __user *out = (u64 __user *)buf; - struct page *ppage; - unsigned long src = *ppos; - unsigned long pfn; - ssize_t ret = 0; - u64 kflags, uflags; - - pfn = src / KPMSIZE; - count = min_t(unsigned long, count, (max_pfn * KPMSIZE) - src); - if (src & KPMMASK || count & KPMMASK) - return -EINVAL; - - while (count > 0) { - ppage = NULL; - if (pfn_valid(pfn)) - ppage = pfn_to_page(pfn); - pfn++; - if (!ppage) - kflags = 0; - else - kflags = ppage->flags; - - uflags = kpf_copy_bit(KPF_LOCKED, PG_locked, kflags) | - kpf_copy_bit(kflags, KPF_ERROR, PG_error) | - kpf_copy_bit(kflags, KPF_REFERENCED, PG_referenced) | - kpf_copy_bit(kflags, KPF_UPTODATE, PG_uptodate) | - kpf_copy_bit(kflags, KPF_DIRTY, PG_dirty) | - kpf_copy_bit(kflags, KPF_LRU, PG_lru) | - kpf_copy_bit(kflags, KPF_ACTIVE, PG_active) | - kpf_copy_bit(kflags, KPF_SLAB, PG_slab) | - kpf_copy_bit(kflags, KPF_WRITEBACK, PG_writeback) | - kpf_copy_bit(kflags, KPF_RECLAIM, PG_reclaim) | - kpf_copy_bit(kflags, KPF_BUDDY, PG_buddy); - - if (put_user(uflags, out++)) { - ret = -EFAULT; - break; - } - - count -= KPMSIZE; - } - - *ppos += (char __user *)out - buf; - if (!ret) - ret = (char __user *)out - buf; - return ret; -} - -static struct file_operations proc_kpageflags_operations = { - .llseek = mem_lseek, - .read = kpageflags_read, -}; -#endif /* CONFIG_PROC_PAGE_MONITOR */ - -struct proc_dir_entry *proc_root_kcore; - -void __init proc_misc_init(void) -{ - static struct { - char *name; - int (*read_proc)(char*,char**,off_t,int,int*,void*); - } *p, simple_ones[] = { - {"loadavg", loadavg_read_proc}, - {"uptime", uptime_read_proc}, - {"meminfo", meminfo_read_proc}, - {"version", version_read_proc}, -#ifdef CONFIG_PROC_HARDWARE - {"hardware", hardware_read_proc}, -#endif -#ifdef CONFIG_STRAM_PROC - {"stram", stram_read_proc}, -#endif - {"filesystems", filesystems_read_proc}, - {"cmdline", cmdline_read_proc}, - {"execdomains", execdomains_read_proc}, - {NULL,} - }; - for (p = simple_ones; p->name; p++) - create_proc_read_entry(p->name, 0, NULL, p->read_proc, NULL); - - proc_symlink("mounts", NULL, "self/mounts"); - - /* And now for trickier ones */ -#ifdef CONFIG_PRINTK - proc_create("kmsg", S_IRUSR, NULL, &proc_kmsg_operations); -#endif -#ifdef CONFIG_FILE_LOCKING - proc_create("locks", 0, NULL, &proc_locks_operations); -#endif - proc_create("devices", 0, NULL, &proc_devinfo_operations); - proc_create("cpuinfo", 0, NULL, &proc_cpuinfo_operations); -#ifdef CONFIG_BLOCK - proc_create("partitions", 0, NULL, &proc_partitions_operations); -#endif - proc_create("stat", 0, NULL, &proc_stat_operations); - proc_create("interrupts", 0, NULL, &proc_interrupts_operations); -#ifdef CONFIG_SLABINFO - proc_create("slabinfo",S_IWUSR|S_IRUGO,NULL,&proc_slabinfo_operations); -#ifdef CONFIG_DEBUG_SLAB_LEAK - proc_create("slab_allocators", 0, NULL, &proc_slabstats_operations); -#endif -#endif -#ifdef CONFIG_MMU - proc_create("vmallocinfo", S_IRUSR, NULL, &proc_vmalloc_operations); -#endif - proc_create("buddyinfo", S_IRUGO, NULL, &fragmentation_file_operations); - proc_create("pagetypeinfo", S_IRUGO, NULL, &pagetypeinfo_file_ops); - proc_create("vmstat", S_IRUGO, NULL, &proc_vmstat_file_operations); - proc_create("zoneinfo", S_IRUGO, NULL, &proc_zoneinfo_file_operations); -#ifdef CONFIG_BLOCK - proc_create("diskstats", 0, NULL, &proc_diskstats_operations); -#endif -#ifdef CONFIG_MODULES - proc_create("modules", 0, NULL, &proc_modules_operations); -#endif -#ifdef CONFIG_SCHEDSTATS - proc_create("schedstat", 0, NULL, &proc_schedstat_operations); -#endif -#ifdef CONFIG_PROC_KCORE - proc_root_kcore = proc_create("kcore", S_IRUSR, NULL, &proc_kcore_operations); - if (proc_root_kcore) - proc_root_kcore->size = - (size_t)high_memory - PAGE_OFFSET + PAGE_SIZE; -#endif -#ifdef CONFIG_PROC_PAGE_MONITOR - proc_create("kpagecount", S_IRUSR, NULL, &proc_kpagecount_operations); - proc_create("kpageflags", S_IRUSR, NULL, &proc_kpageflags_operations); -#endif -#ifdef CONFIG_PROC_VMCORE - proc_vmcore = proc_create("vmcore", S_IRUSR, NULL, &proc_vmcore_operations); -#endif -#ifdef CONFIG_MAGIC_SYSRQ - proc_create("sysrq-trigger", S_IWUSR, NULL, &proc_sysrq_trigger_operations); -#endif -} diff --git a/fs/proc/proc_sysctl.c b/fs/proc/proc_sysctl.c index 945a81043ba..06ed10b7da9 100644 --- a/fs/proc/proc_sysctl.c +++ b/fs/proc/proc_sysctl.c @@ -1,7 +1,7 @@ /* * /proc/sys support */ - +#include <linux/init.h> #include <linux/sysctl.h> #include <linux/proc_fs.h> #include <linux/security.h> @@ -31,6 +31,7 @@ static struct inode *proc_sys_make_inode(struct super_block *sb, inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME; inode->i_flags |= S_PRIVATE; /* tell selinux to ignore this inode */ inode->i_mode = table->mode; + inode->i_uid = inode->i_gid = 0; if (!table->child) { inode->i_mode |= S_IFREG; inode->i_op = &proc_sys_inode_operations; @@ -298,13 +299,19 @@ static int proc_sys_permission(struct inode *inode, int mask) * sysctl entries that are not writeable, * are _NOT_ writeable, capabilities or not. */ - struct ctl_table_header *head = grab_header(inode); - struct ctl_table *table = PROC_I(inode)->sysctl_entry; + struct ctl_table_header *head; + struct ctl_table *table; int error; + /* Executable files are not allowed under /proc/sys/ */ + if ((mask & MAY_EXEC) && S_ISREG(inode->i_mode)) + return -EACCES; + + head = grab_header(inode); if (IS_ERR(head)) return PTR_ERR(head); + table = PROC_I(inode)->sysctl_entry; if (!table) /* global root - r-xr-xr-x */ error = mask & MAY_WRITE ? -EACCES : 0; else /* Use the permissions on the sysctl table entry */ @@ -353,6 +360,7 @@ static const struct file_operations proc_sys_file_operations = { static const struct file_operations proc_sys_dir_file_operations = { .readdir = proc_sys_readdir, + .llseek = generic_file_llseek, }; static const struct inode_operations proc_sys_inode_operations = { @@ -395,7 +403,7 @@ static struct dentry_operations proc_sys_dentry_operations = { .d_compare = proc_sys_compare, }; -int proc_sys_init(void) +int __init proc_sys_init(void) { struct proc_dir_entry *proc_sys_root; diff --git a/fs/proc/root.c b/fs/proc/root.c index 95117538a4f..7761602af9d 100644 --- a/fs/proc/root.c +++ b/fs/proc/root.c @@ -104,9 +104,9 @@ static struct file_system_type proc_fs_type = { void __init proc_root_init(void) { - int err = proc_init_inodecache(); - if (err) - return; + int err; + + proc_init_inodecache(); err = register_filesystem(&proc_fs_type); if (err) return; @@ -117,7 +117,7 @@ void __init proc_root_init(void) return; } - proc_misc_init(); + proc_symlink("mounts", NULL, "self/mounts"); proc_net_init(); diff --git a/fs/proc/stat.c b/fs/proc/stat.c new file mode 100644 index 00000000000..3bb1cf1e742 --- /dev/null +++ b/fs/proc/stat.c @@ -0,0 +1,161 @@ +#include <linux/cpumask.h> +#include <linux/fs.h> +#include <linux/gfp.h> +#include <linux/init.h> +#include <linux/interrupt.h> +#include <linux/kernel_stat.h> +#include <linux/proc_fs.h> +#include <linux/sched.h> +#include <linux/seq_file.h> +#include <linux/slab.h> +#include <linux/time.h> +#include <asm/cputime.h> + +#ifndef arch_irq_stat_cpu +#define arch_irq_stat_cpu(cpu) 0 +#endif +#ifndef arch_irq_stat +#define arch_irq_stat() 0 +#endif + +static int show_stat(struct seq_file *p, void *v) +{ + int i, j; + unsigned long jif; + cputime64_t user, nice, system, idle, iowait, irq, softirq, steal; + cputime64_t guest; + u64 sum = 0; + struct timespec boottime; + unsigned int per_irq_sum; + + user = nice = system = idle = iowait = + irq = softirq = steal = cputime64_zero; + guest = cputime64_zero; + getboottime(&boottime); + jif = boottime.tv_sec; + + for_each_possible_cpu(i) { + user = cputime64_add(user, kstat_cpu(i).cpustat.user); + nice = cputime64_add(nice, kstat_cpu(i).cpustat.nice); + system = cputime64_add(system, kstat_cpu(i).cpustat.system); + idle = cputime64_add(idle, kstat_cpu(i).cpustat.idle); + iowait = cputime64_add(iowait, kstat_cpu(i).cpustat.iowait); + irq = cputime64_add(irq, kstat_cpu(i).cpustat.irq); + softirq = cputime64_add(softirq, kstat_cpu(i).cpustat.softirq); + steal = cputime64_add(steal, kstat_cpu(i).cpustat.steal); + guest = cputime64_add(guest, kstat_cpu(i).cpustat.guest); + for_each_irq_nr(j) { +#ifdef CONFIG_SPARSE_IRQ + if (!irq_to_desc(j)) + continue; +#endif + sum += kstat_irqs_cpu(j, i); + } + sum += arch_irq_stat_cpu(i); + } + sum += arch_irq_stat(); + + seq_printf(p, "cpu %llu %llu %llu %llu %llu %llu %llu %llu %llu\n", + (unsigned long long)cputime64_to_clock_t(user), + (unsigned long long)cputime64_to_clock_t(nice), + (unsigned long long)cputime64_to_clock_t(system), + (unsigned long long)cputime64_to_clock_t(idle), + (unsigned long long)cputime64_to_clock_t(iowait), + (unsigned long long)cputime64_to_clock_t(irq), + (unsigned long long)cputime64_to_clock_t(softirq), + (unsigned long long)cputime64_to_clock_t(steal), + (unsigned long long)cputime64_to_clock_t(guest)); + for_each_online_cpu(i) { + + /* Copy values here to work around gcc-2.95.3, gcc-2.96 */ + user = kstat_cpu(i).cpustat.user; + nice = kstat_cpu(i).cpustat.nice; + system = kstat_cpu(i).cpustat.system; + idle = kstat_cpu(i).cpustat.idle; + iowait = kstat_cpu(i).cpustat.iowait; + irq = kstat_cpu(i).cpustat.irq; + softirq = kstat_cpu(i).cpustat.softirq; + steal = kstat_cpu(i).cpustat.steal; + guest = kstat_cpu(i).cpustat.guest; + seq_printf(p, + "cpu%d %llu %llu %llu %llu %llu %llu %llu %llu %llu\n", + i, + (unsigned long long)cputime64_to_clock_t(user), + (unsigned long long)cputime64_to_clock_t(nice), + (unsigned long long)cputime64_to_clock_t(system), + (unsigned long long)cputime64_to_clock_t(idle), + (unsigned long long)cputime64_to_clock_t(iowait), + (unsigned long long)cputime64_to_clock_t(irq), + (unsigned long long)cputime64_to_clock_t(softirq), + (unsigned long long)cputime64_to_clock_t(steal), + (unsigned long long)cputime64_to_clock_t(guest)); + } + seq_printf(p, "intr %llu", (unsigned long long)sum); + + /* sum again ? it could be updated? */ + for_each_irq_nr(j) { + per_irq_sum = 0; +#ifdef CONFIG_SPARSE_IRQ + if (!irq_to_desc(j)) { + seq_printf(p, " %u", per_irq_sum); + continue; + } +#endif + for_each_possible_cpu(i) + per_irq_sum += kstat_irqs_cpu(j, i); + + seq_printf(p, " %u", per_irq_sum); + } + + seq_printf(p, + "\nctxt %llu\n" + "btime %lu\n" + "processes %lu\n" + "procs_running %lu\n" + "procs_blocked %lu\n", + nr_context_switches(), + (unsigned long)jif, + total_forks, + nr_running(), + nr_iowait()); + + return 0; +} + +static int stat_open(struct inode *inode, struct file *file) +{ + unsigned size = 4096 * (1 + num_possible_cpus() / 32); + char *buf; + struct seq_file *m; + int res; + + /* don't ask for more than the kmalloc() max size, currently 128 KB */ + if (size > 128 * 1024) + size = 128 * 1024; + buf = kmalloc(size, GFP_KERNEL); + if (!buf) + return -ENOMEM; + + res = single_open(file, show_stat, NULL); + if (!res) { + m = file->private_data; + m->buf = buf; + m->size = size; + } else + kfree(buf); + return res; +} + +static const struct file_operations proc_stat_operations = { + .open = stat_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, +}; + +static int __init proc_stat_init(void) +{ + proc_create("stat", 0, NULL, &proc_stat_operations); + return 0; +} +module_init(proc_stat_init); diff --git a/fs/proc/task_mmu.c b/fs/proc/task_mmu.c index 4806830ea2a..3a8bdd7f575 100644 --- a/fs/proc/task_mmu.c +++ b/fs/proc/task_mmu.c @@ -198,11 +198,8 @@ static int do_maps_open(struct inode *inode, struct file *file, return ret; } -static int show_map(struct seq_file *m, void *v) +static void show_map_vma(struct seq_file *m, struct vm_area_struct *vma) { - struct proc_maps_private *priv = m->private; - struct task_struct *task = priv->task; - struct vm_area_struct *vma = v; struct mm_struct *mm = vma->vm_mm; struct file *file = vma->vm_file; int flags = vma->vm_flags; @@ -254,6 +251,15 @@ static int show_map(struct seq_file *m, void *v) } } seq_putc(m, '\n'); +} + +static int show_map(struct seq_file *m, void *v) +{ + struct vm_area_struct *vma = v; + struct proc_maps_private *priv = m->private; + struct task_struct *task = priv->task; + + show_map_vma(m, vma); if (m->count < m->size) /* vma is copied successfully */ m->version = (vma != get_gate_vma(task))? vma->vm_start: 0; @@ -364,9 +370,10 @@ static int smaps_pte_range(pmd_t *pmd, unsigned long addr, unsigned long end, static int show_smap(struct seq_file *m, void *v) { + struct proc_maps_private *priv = m->private; + struct task_struct *task = priv->task; struct vm_area_struct *vma = v; struct mem_size_stats mss; - int ret; struct mm_walk smaps_walk = { .pmd_entry = smaps_pte_range, .mm = vma->vm_mm, @@ -378,9 +385,7 @@ static int show_smap(struct seq_file *m, void *v) if (vma->vm_mm && !is_vm_hugetlb_page(vma)) walk_page_range(vma->vm_start, vma->vm_end, &smaps_walk); - ret = show_map(m, v); - if (ret) - return ret; + show_map_vma(m, vma); seq_printf(m, "Size: %8lu kB\n" @@ -402,7 +407,9 @@ static int show_smap(struct seq_file *m, void *v) mss.referenced >> 10, mss.swap >> 10); - return ret; + if (m->count < m->size) /* vma is copied successfully */ + m->version = (vma != get_gate_vma(task)) ? vma->vm_start : 0; + return 0; } static const struct seq_operations proc_pid_smaps_op = { @@ -550,9 +557,9 @@ static u64 swap_pte_to_pagemap_entry(pte_t pte) return swp_type(e) | (swp_offset(e) << MAX_SWAPFILES_SHIFT); } -static unsigned long pte_to_pagemap_entry(pte_t pte) +static u64 pte_to_pagemap_entry(pte_t pte) { - unsigned long pme = 0; + u64 pme = 0; if (is_swap_pte(pte)) pme = PM_PFRAME(swap_pte_to_pagemap_entry(pte)) | PM_PSHIFT(PAGE_SHIFT) | PM_SWAP; diff --git a/fs/proc/uptime.c b/fs/proc/uptime.c new file mode 100644 index 00000000000..df26aa88fa4 --- /dev/null +++ b/fs/proc/uptime.c @@ -0,0 +1,45 @@ +#include <linux/init.h> +#include <linux/proc_fs.h> +#include <linux/sched.h> +#include <linux/time.h> +#include <asm/cputime.h> + +static int proc_calc_metrics(char *page, char **start, off_t off, + int count, int *eof, int len) +{ + if (len <= off + count) + *eof = 1; + *start = page + off; + len -= off; + if (len > count) + len = count; + if (len < 0) + len = 0; + return len; +} + +static int uptime_read_proc(char *page, char **start, off_t off, int count, + int *eof, void *data) +{ + struct timespec uptime; + struct timespec idle; + int len; + cputime_t idletime = cputime_add(init_task.utime, init_task.stime); + + do_posix_clock_monotonic_gettime(&uptime); + monotonic_to_bootbased(&uptime); + cputime_to_timespec(idletime, &idle); + len = sprintf(page, "%lu.%02lu %lu.%02lu\n", + (unsigned long) uptime.tv_sec, + (uptime.tv_nsec / (NSEC_PER_SEC / 100)), + (unsigned long) idle.tv_sec, + (idle.tv_nsec / (NSEC_PER_SEC / 100))); + return proc_calc_metrics(page, start, off, count, eof, len); +} + +static int __init proc_uptime_init(void) +{ + create_proc_read_entry("uptime", 0, NULL, uptime_read_proc, NULL); + return 0; +} +module_init(proc_uptime_init); diff --git a/fs/proc/version.c b/fs/proc/version.c new file mode 100644 index 00000000000..76817a60678 --- /dev/null +++ b/fs/proc/version.c @@ -0,0 +1,34 @@ +#include <linux/fs.h> +#include <linux/init.h> +#include <linux/kernel.h> +#include <linux/proc_fs.h> +#include <linux/seq_file.h> +#include <linux/utsname.h> + +static int version_proc_show(struct seq_file *m, void *v) +{ + seq_printf(m, linux_proc_banner, + utsname()->sysname, + utsname()->release, + utsname()->version); + return 0; +} + +static int version_proc_open(struct inode *inode, struct file *file) +{ + return single_open(file, version_proc_show, NULL); +} + +static const struct file_operations version_proc_fops = { + .open = version_proc_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, +}; + +static int __init proc_version_init(void) +{ + proc_create("version", 0, NULL, &version_proc_fops); + return 0; +} +module_init(proc_version_init); diff --git a/fs/proc/vmcore.c b/fs/proc/vmcore.c index 841368b87a2..03ec5950490 100644 --- a/fs/proc/vmcore.c +++ b/fs/proc/vmcore.c @@ -32,10 +32,7 @@ static size_t elfcorebuf_sz; /* Total size of vmcore file. */ static u64 vmcore_size; -/* Stores the physical address of elf header of crash image. */ -unsigned long long elfcorehdr_addr = ELFCORE_ADDR_MAX; - -struct proc_dir_entry *proc_vmcore = NULL; +static struct proc_dir_entry *proc_vmcore = NULL; /* Reads a page from the oldmem device from given offset. */ static ssize_t read_from_oldmem(char *buf, size_t count, @@ -165,7 +162,7 @@ static ssize_t read_vmcore(struct file *file, char __user *buffer, return acc; } -const struct file_operations proc_vmcore_operations = { +static const struct file_operations proc_vmcore_operations = { .read = read_vmcore, }; @@ -647,7 +644,7 @@ static int __init vmcore_init(void) int rc = 0; /* If elfcorehdr= has been passed in cmdline, then capture the dump.*/ - if (!(elfcorehdr_addr < ELFCORE_ADDR_MAX)) + if (!(is_vmcore_usable())) return rc; rc = parse_crash_elf_headers(); if (rc) { @@ -655,7 +652,7 @@ static int __init vmcore_init(void) return rc; } - /* Initialize /proc/vmcore size if proc is already up. */ + proc_vmcore = proc_create("vmcore", S_IRUSR, NULL, &proc_vmcore_operations); if (proc_vmcore) proc_vmcore->size = vmcore_size; return 0; diff --git a/fs/quota.c b/fs/quota.c index 7f4386ebc23..b7fe44e0161 100644 --- a/fs/quota.c +++ b/fs/quota.c @@ -79,7 +79,7 @@ static int generic_quotactl_valid(struct super_block *sb, int type, int cmd, qid /* Check privileges */ if (cmd == Q_GETQUOTA) { - if (((type == USRQUOTA && current->euid != id) || + if (((type == USRQUOTA && current_euid() != id) || (type == GRPQUOTA && !in_egroup_p(id))) && !capable(CAP_SYS_ADMIN)) return -EPERM; @@ -130,7 +130,7 @@ static int xqm_quotactl_valid(struct super_block *sb, int type, int cmd, qid_t i /* Check privileges */ if (cmd == Q_XGETQUOTA) { - if (((type == XQM_USRQUOTA && current->euid != id) || + if (((type == XQM_USRQUOTA && current_euid() != id) || (type == XQM_GRPQUOTA && !in_egroup_p(id))) && !capable(CAP_SYS_ADMIN)) return -EPERM; diff --git a/fs/ramfs/file-nommu.c b/fs/ramfs/file-nommu.c index 5145cb9125a..76acdbc3461 100644 --- a/fs/ramfs/file-nommu.c +++ b/fs/ramfs/file-nommu.c @@ -112,12 +112,12 @@ int ramfs_nommu_expand_for_mapping(struct inode *inode, size_t newsize) goto add_error; if (!pagevec_add(&lru_pvec, page)) - __pagevec_lru_add(&lru_pvec); + __pagevec_lru_add_file(&lru_pvec); unlock_page(page); } - pagevec_lru_add(&lru_pvec); + pagevec_lru_add_file(&lru_pvec); return 0; fsize_exceeded: diff --git a/fs/ramfs/inode.c b/fs/ramfs/inode.c index b13123424e4..a83a3518ae3 100644 --- a/fs/ramfs/inode.c +++ b/fs/ramfs/inode.c @@ -55,12 +55,13 @@ struct inode *ramfs_get_inode(struct super_block *sb, int mode, dev_t dev) if (inode) { inode->i_mode = mode; - inode->i_uid = current->fsuid; - inode->i_gid = current->fsgid; + inode->i_uid = current_fsuid(); + inode->i_gid = current_fsgid(); inode->i_blocks = 0; inode->i_mapping->a_ops = &ramfs_aops; inode->i_mapping->backing_dev_info = &ramfs_backing_dev_info; mapping_set_gfp_mask(inode->i_mapping, GFP_HIGHUSER); + mapping_set_unevictable(inode->i_mapping); inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME; switch (mode & S_IFMT) { default: diff --git a/fs/read_write.c b/fs/read_write.c index 9ba495d5a29..969a6d9c020 100644 --- a/fs/read_write.c +++ b/fs/read_write.c @@ -31,39 +31,61 @@ const struct file_operations generic_ro_fops = { EXPORT_SYMBOL(generic_ro_fops); +/** + * generic_file_llseek_unlocked - lockless generic llseek implementation + * @file: file structure to seek on + * @offset: file offset to seek to + * @origin: type of seek + * + * Updates the file offset to the value specified by @offset and @origin. + * Locking must be provided by the caller. + */ loff_t generic_file_llseek_unlocked(struct file *file, loff_t offset, int origin) { - loff_t retval; struct inode *inode = file->f_mapping->host; switch (origin) { - case SEEK_END: - offset += inode->i_size; - break; - case SEEK_CUR: - offset += file->f_pos; + case SEEK_END: + offset += inode->i_size; + break; + case SEEK_CUR: + offset += file->f_pos; + break; } - retval = -EINVAL; - if (offset>=0 && offset<=inode->i_sb->s_maxbytes) { - /* Special lock needed here? */ - if (offset != file->f_pos) { - file->f_pos = offset; - file->f_version = 0; - } - retval = offset; + + if (offset < 0 || offset > inode->i_sb->s_maxbytes) + return -EINVAL; + + /* Special lock needed here? */ + if (offset != file->f_pos) { + file->f_pos = offset; + file->f_version = 0; } - return retval; + + return offset; } EXPORT_SYMBOL(generic_file_llseek_unlocked); +/** + * generic_file_llseek - generic llseek implementation for regular files + * @file: file structure to seek on + * @offset: file offset to seek to + * @origin: type of seek + * + * This is a generic implemenation of ->llseek useable for all normal local + * filesystems. It just updates the file offset to the value specified by + * @offset and @origin under i_mutex. + */ loff_t generic_file_llseek(struct file *file, loff_t offset, int origin) { - loff_t n; + loff_t rval; + mutex_lock(&file->f_dentry->d_inode->i_mutex); - n = generic_file_llseek_unlocked(file, offset, origin); + rval = generic_file_llseek_unlocked(file, offset, origin); mutex_unlock(&file->f_dentry->d_inode->i_mutex); - return n; + + return rval; } EXPORT_SYMBOL(generic_file_llseek); diff --git a/fs/readdir.c b/fs/readdir.c index 93a7559bbfd..b318d9b5af2 100644 --- a/fs/readdir.c +++ b/fs/readdir.c @@ -117,7 +117,7 @@ asmlinkage long old_readdir(unsigned int fd, struct old_linux_dirent __user * di buf.dirent = dirent; error = vfs_readdir(file, fillonedir, &buf); - if (error >= 0) + if (buf.result) error = buf.result; fput(file); @@ -209,9 +209,8 @@ asmlinkage long sys_getdents(unsigned int fd, struct linux_dirent __user * diren buf.error = 0; error = vfs_readdir(file, filldir, &buf); - if (error < 0) - goto out_putf; - error = buf.error; + if (error >= 0) + error = buf.error; lastdirent = buf.previous; if (lastdirent) { if (put_user(file->f_pos, &lastdirent->d_off)) @@ -219,8 +218,6 @@ asmlinkage long sys_getdents(unsigned int fd, struct linux_dirent __user * diren else error = count - buf.count; } - -out_putf: fput(file); out: return error; @@ -293,19 +290,16 @@ asmlinkage long sys_getdents64(unsigned int fd, struct linux_dirent64 __user * d buf.error = 0; error = vfs_readdir(file, filldir64, &buf); - if (error < 0) - goto out_putf; - error = buf.error; + if (error >= 0) + error = buf.error; lastdirent = buf.previous; if (lastdirent) { typeof(lastdirent->d_off) d_off = file->f_pos; - error = -EFAULT; if (__put_user(d_off, &lastdirent->d_off)) - goto out_putf; - error = count - buf.count; + error = -EFAULT; + else + error = count - buf.count; } - -out_putf: fput(file); out: return error; diff --git a/fs/reiserfs/file.c b/fs/reiserfs/file.c index a804903d31d..33408417038 100644 --- a/fs/reiserfs/file.c +++ b/fs/reiserfs/file.c @@ -296,6 +296,7 @@ const struct file_operations reiserfs_file_operations = { .aio_write = generic_file_aio_write, .splice_read = generic_file_splice_read, .splice_write = generic_file_splice_write, + .llseek = generic_file_llseek, }; const struct inode_operations reiserfs_file_inode_operations = { diff --git a/fs/reiserfs/inode.c b/fs/reiserfs/inode.c index 5699171212a..6c4c2c69449 100644 --- a/fs/reiserfs/inode.c +++ b/fs/reiserfs/inode.c @@ -1522,7 +1522,6 @@ static struct dentry *reiserfs_get_dentry(struct super_block *sb, { struct cpu_key key; - struct dentry *result; struct inode *inode; key.on_disk_key.k_objectid = objectid; @@ -1535,16 +1534,8 @@ static struct dentry *reiserfs_get_dentry(struct super_block *sb, inode = NULL; } reiserfs_write_unlock(sb); - if (!inode) - inode = ERR_PTR(-ESTALE); - if (IS_ERR(inode)) - return ERR_CAST(inode); - result = d_alloc_anon(inode); - if (!result) { - iput(inode); - return ERR_PTR(-ENOMEM); - } - return result; + + return d_obtain_alias(inode); } struct dentry *reiserfs_fh_to_dentry(struct super_block *sb, struct fid *fid, diff --git a/fs/reiserfs/journal.c b/fs/reiserfs/journal.c index c21df71943a..9643c3bbeb3 100644 --- a/fs/reiserfs/journal.c +++ b/fs/reiserfs/journal.c @@ -2575,7 +2575,7 @@ static int release_journal_dev(struct super_block *super, if (journal->j_dev_bd != NULL) { if (journal->j_dev_bd->bd_dev != super->s_dev) bd_release(journal->j_dev_bd); - result = blkdev_put(journal->j_dev_bd); + result = blkdev_put(journal->j_dev_bd, journal->j_dev_mode); journal->j_dev_bd = NULL; } @@ -2593,7 +2593,7 @@ static int journal_init_dev(struct super_block *super, { int result; dev_t jdev; - int blkdev_mode = FMODE_READ | FMODE_WRITE; + fmode_t blkdev_mode = FMODE_READ | FMODE_WRITE; char b[BDEVNAME_SIZE]; result = 0; @@ -2608,6 +2608,7 @@ static int journal_init_dev(struct super_block *super, /* there is no "jdev" option and journal is on separate device */ if ((!jdev_name || !jdev_name[0])) { journal->j_dev_bd = open_by_devnum(jdev, blkdev_mode); + journal->j_dev_mode = blkdev_mode; if (IS_ERR(journal->j_dev_bd)) { result = PTR_ERR(journal->j_dev_bd); journal->j_dev_bd = NULL; @@ -2618,7 +2619,7 @@ static int journal_init_dev(struct super_block *super, } else if (jdev != super->s_dev) { result = bd_claim(journal->j_dev_bd, journal); if (result) { - blkdev_put(journal->j_dev_bd); + blkdev_put(journal->j_dev_bd, blkdev_mode); return result; } @@ -2628,7 +2629,9 @@ static int journal_init_dev(struct super_block *super, return 0; } - journal->j_dev_bd = open_bdev_excl(jdev_name, 0, journal); + journal->j_dev_mode = blkdev_mode; + journal->j_dev_bd = open_bdev_exclusive(jdev_name, + blkdev_mode, journal); if (IS_ERR(journal->j_dev_bd)) { result = PTR_ERR(journal->j_dev_bd); journal->j_dev_bd = NULL; diff --git a/fs/reiserfs/namei.c b/fs/reiserfs/namei.c index c1add28dd45..4f322e5ed84 100644 --- a/fs/reiserfs/namei.c +++ b/fs/reiserfs/namei.c @@ -383,7 +383,6 @@ struct dentry *reiserfs_get_parent(struct dentry *child) struct inode *inode = NULL; struct reiserfs_dir_entry de; INITIALIZE_PATH(path_to_entry); - struct dentry *parent; struct inode *dir = child->d_inode; if (dir->i_nlink == 0) { @@ -401,15 +400,7 @@ struct dentry *reiserfs_get_parent(struct dentry *child) inode = reiserfs_iget(dir->i_sb, (struct cpu_key *)&(de.de_dir_id)); reiserfs_write_unlock(dir->i_sb); - if (!inode || IS_ERR(inode)) { - return ERR_PTR(-EACCES); - } - parent = d_alloc_anon(inode); - if (!parent) { - iput(inode); - parent = ERR_PTR(-ENOMEM); - } - return parent; + return d_obtain_alias(inode); } /* add entry to the directory (entry can be hidden). @@ -582,7 +573,7 @@ static int new_inode_init(struct inode *inode, struct inode *dir, int mode) /* the quota init calls have to know who to charge the quota to, so ** we have to set uid and gid here */ - inode->i_uid = current->fsuid; + inode->i_uid = current_fsuid(); inode->i_mode = mode; /* Make inode invalid - just in case we are going to drop it before * the initialization happens */ @@ -593,7 +584,7 @@ static int new_inode_init(struct inode *inode, struct inode *dir, int mode) if (S_ISDIR(mode)) inode->i_mode |= S_ISGID; } else { - inode->i_gid = current->fsgid; + inode->i_gid = current_fsgid(); } DQUOT_INIT(inode); return 0; diff --git a/fs/reiserfs/procfs.c b/fs/reiserfs/procfs.c index b9dbeeca704..37173fa07d1 100644 --- a/fs/reiserfs/procfs.c +++ b/fs/reiserfs/procfs.c @@ -8,8 +8,6 @@ /* proc info support a la one created by Sizif@Botik.RU for PGC */ -/* $Id: procfs.c,v 1.1.8.2 2001/07/15 17:08:42 god Exp $ */ - #include <linux/module.h> #include <linux/time.h> #include <linux/seq_file.h> @@ -621,7 +619,6 @@ int reiserfs_global_version_in_proc(char *buffer, char **start, #endif /* - * $Log: procfs.c,v $ * Revision 1.1.8.2 2001/07/15 17:08:42 god * . use get_super() in procfs.c * . remove remove_save_link() from reiserfs_do_truncate() diff --git a/fs/reiserfs/super.c b/fs/reiserfs/super.c index d318c7e663f..663a91f5dce 100644 --- a/fs/reiserfs/super.c +++ b/fs/reiserfs/super.c @@ -2058,10 +2058,10 @@ static int reiserfs_quota_on_mount(struct super_block *sb, int type) * Standard function to be called on quota_on */ static int reiserfs_quota_on(struct super_block *sb, int type, int format_id, - char *path, int remount) + char *name, int remount) { int err; - struct nameidata nd; + struct path path; struct inode *inode; struct reiserfs_transaction_handle th; @@ -2069,16 +2069,16 @@ static int reiserfs_quota_on(struct super_block *sb, int type, int format_id, return -EINVAL; /* No more checks needed? Path and format_id are bogus anyway... */ if (remount) - return vfs_quota_on(sb, type, format_id, path, 1); - err = path_lookup(path, LOOKUP_FOLLOW, &nd); + return vfs_quota_on(sb, type, format_id, name, 1); + err = kern_path(name, LOOKUP_FOLLOW, &path); if (err) return err; /* Quotafile not on the same filesystem? */ - if (nd.path.mnt->mnt_sb != sb) { + if (path.mnt->mnt_sb != sb) { err = -EXDEV; goto out; } - inode = nd.path.dentry->d_inode; + inode = path.dentry->d_inode; /* We must not pack tails for quota files on reiserfs for quota IO to work */ if (!(REISERFS_I(inode)->i_flags & i_nopack_mask)) { err = reiserfs_unpack(inode, NULL); @@ -2094,7 +2094,7 @@ static int reiserfs_quota_on(struct super_block *sb, int type, int format_id, /* Journaling quota? */ if (REISERFS_SB(sb)->s_qf_names[type]) { /* Quotafile not of fs root? */ - if (nd.path.dentry->d_parent->d_inode != sb->s_root->d_inode) + if (path.dentry->d_parent != sb->s_root) reiserfs_warning(sb, "reiserfs: Quota file not on filesystem root. " "Journalled quota will not work."); @@ -2113,9 +2113,9 @@ static int reiserfs_quota_on(struct super_block *sb, int type, int format_id, if (err) goto out; } - err = vfs_quota_on_path(sb, type, format_id, &nd.path); + err = vfs_quota_on_path(sb, type, format_id, &path); out: - path_put(&nd.path); + path_put(&path); return err; } diff --git a/fs/reiserfs/xattr.c b/fs/reiserfs/xattr.c index bb3cb5b7cdb..ad92461cbfc 100644 --- a/fs/reiserfs/xattr.c +++ b/fs/reiserfs/xattr.c @@ -155,7 +155,7 @@ static struct dentry *get_xa_file_dentry(const struct inode *inode, xadir = open_xa_dir(inode, flags); if (IS_ERR(xadir)) { return ERR_CAST(xadir); - } else if (xadir && !xadir->d_inode) { + } else if (!xadir->d_inode) { dput(xadir); return ERR_PTR(-ENODATA); } diff --git a/fs/select.c b/fs/select.c index da0e88201c3..87df51eadcf 100644 --- a/fs/select.c +++ b/fs/select.c @@ -24,9 +24,64 @@ #include <linux/fdtable.h> #include <linux/fs.h> #include <linux/rcupdate.h> +#include <linux/hrtimer.h> #include <asm/uaccess.h> + +/* + * Estimate expected accuracy in ns from a timeval. + * + * After quite a bit of churning around, we've settled on + * a simple thing of taking 0.1% of the timeout as the + * slack, with a cap of 100 msec. + * "nice" tasks get a 0.5% slack instead. + * + * Consider this comment an open invitation to come up with even + * better solutions.. + */ + +static long __estimate_accuracy(struct timespec *tv) +{ + long slack; + int divfactor = 1000; + + if (task_nice(current) > 0) + divfactor = divfactor / 5; + + slack = tv->tv_nsec / divfactor; + slack += tv->tv_sec * (NSEC_PER_SEC/divfactor); + + if (slack > 100 * NSEC_PER_MSEC) + slack = 100 * NSEC_PER_MSEC; + + if (slack < 0) + slack = 0; + return slack; +} + +static long estimate_accuracy(struct timespec *tv) +{ + unsigned long ret; + struct timespec now; + + /* + * Realtime tasks get a slack of 0 for obvious reasons. + */ + + if (rt_task(current)) + return 0; + + ktime_get_ts(&now); + now = timespec_sub(*tv, now); + ret = __estimate_accuracy(&now); + if (ret < current->timer_slack_ns) + return current->timer_slack_ns; + return ret; +} + + + struct poll_table_page { struct poll_table_page * next; struct poll_table_entry * entry; @@ -130,6 +185,79 @@ static void __pollwait(struct file *filp, wait_queue_head_t *wait_address, add_wait_queue(wait_address, &entry->wait); } +/** + * poll_select_set_timeout - helper function to setup the timeout value + * @to: pointer to timespec variable for the final timeout + * @sec: seconds (from user space) + * @nsec: nanoseconds (from user space) + * + * Note, we do not use a timespec for the user space value here, That + * way we can use the function for timeval and compat interfaces as well. + * + * Returns -EINVAL if sec/nsec are not normalized. Otherwise 0. + */ +int poll_select_set_timeout(struct timespec *to, long sec, long nsec) +{ + struct timespec ts = {.tv_sec = sec, .tv_nsec = nsec}; + + if (!timespec_valid(&ts)) + return -EINVAL; + + /* Optimize for the zero timeout value here */ + if (!sec && !nsec) { + to->tv_sec = to->tv_nsec = 0; + } else { + ktime_get_ts(to); + *to = timespec_add_safe(*to, ts); + } + return 0; +} + +static int poll_select_copy_remaining(struct timespec *end_time, void __user *p, + int timeval, int ret) +{ + struct timespec rts; + struct timeval rtv; + + if (!p) + return ret; + + if (current->personality & STICKY_TIMEOUTS) + goto sticky; + + /* No update for zero timeout */ + if (!end_time->tv_sec && !end_time->tv_nsec) + return ret; + + ktime_get_ts(&rts); + rts = timespec_sub(*end_time, rts); + if (rts.tv_sec < 0) + rts.tv_sec = rts.tv_nsec = 0; + + if (timeval) { + rtv.tv_sec = rts.tv_sec; + rtv.tv_usec = rts.tv_nsec / NSEC_PER_USEC; + + if (!copy_to_user(p, &rtv, sizeof(rtv))) + return ret; + + } else if (!copy_to_user(p, &rts, sizeof(rts))) + return ret; + + /* + * If an application puts its timeval in read-only memory, we + * don't want the Linux-specific update to the timeval to + * cause a fault after the select has completed + * successfully. However, because we're not updating the + * timeval, we can't restart the system call. + */ + +sticky: + if (ret == -ERESTARTNOHAND) + ret = -EINTR; + return ret; +} + #define FDS_IN(fds, n) (fds->in + n) #define FDS_OUT(fds, n) (fds->out + n) #define FDS_EX(fds, n) (fds->ex + n) @@ -182,11 +310,13 @@ get_max: #define POLLOUT_SET (POLLWRBAND | POLLWRNORM | POLLOUT | POLLERR) #define POLLEX_SET (POLLPRI) -int do_select(int n, fd_set_bits *fds, s64 *timeout) +int do_select(int n, fd_set_bits *fds, struct timespec *end_time) { + ktime_t expire, *to = NULL; struct poll_wqueues table; poll_table *wait; - int retval, i; + int retval, i, timed_out = 0; + unsigned long slack = 0; rcu_read_lock(); retval = max_select_fd(n, fds); @@ -198,12 +328,17 @@ int do_select(int n, fd_set_bits *fds, s64 *timeout) poll_initwait(&table); wait = &table.pt; - if (!*timeout) + if (end_time && !end_time->tv_sec && !end_time->tv_nsec) { wait = NULL; + timed_out = 1; + } + + if (end_time && !timed_out) + slack = estimate_accuracy(end_time); + retval = 0; for (;;) { unsigned long *rinp, *routp, *rexp, *inp, *outp, *exp; - long __timeout; set_current_state(TASK_INTERRUPTIBLE); @@ -259,27 +394,25 @@ int do_select(int n, fd_set_bits *fds, s64 *timeout) cond_resched(); } wait = NULL; - if (retval || !*timeout || signal_pending(current)) + if (retval || timed_out || signal_pending(current)) break; if (table.error) { retval = table.error; break; } - if (*timeout < 0) { - /* Wait indefinitely */ - __timeout = MAX_SCHEDULE_TIMEOUT; - } else if (unlikely(*timeout >= (s64)MAX_SCHEDULE_TIMEOUT - 1)) { - /* Wait for longer than MAX_SCHEDULE_TIMEOUT. Do it in a loop */ - __timeout = MAX_SCHEDULE_TIMEOUT - 1; - *timeout -= __timeout; - } else { - __timeout = *timeout; - *timeout = 0; + /* + * If this is the first loop and we have a timeout + * given, then we convert to ktime_t and set the to + * pointer to the expiry value. + */ + if (end_time && !to) { + expire = timespec_to_ktime(*end_time); + to = &expire; } - __timeout = schedule_timeout(__timeout); - if (*timeout >= 0) - *timeout += __timeout; + + if (!schedule_hrtimeout_range(to, slack, HRTIMER_MODE_ABS)) + timed_out = 1; } __set_current_state(TASK_RUNNING); @@ -300,7 +433,7 @@ int do_select(int n, fd_set_bits *fds, s64 *timeout) ((unsigned long) (MAX_SCHEDULE_TIMEOUT / HZ)-1) int core_sys_select(int n, fd_set __user *inp, fd_set __user *outp, - fd_set __user *exp, s64 *timeout) + fd_set __user *exp, struct timespec *end_time) { fd_set_bits fds; void *bits; @@ -351,7 +484,7 @@ int core_sys_select(int n, fd_set __user *inp, fd_set __user *outp, zero_fd_set(n, fds.res_out); zero_fd_set(n, fds.res_ex); - ret = do_select(n, &fds, timeout); + ret = do_select(n, &fds, end_time); if (ret < 0) goto out; @@ -377,7 +510,7 @@ out_nofds: asmlinkage long sys_select(int n, fd_set __user *inp, fd_set __user *outp, fd_set __user *exp, struct timeval __user *tvp) { - s64 timeout = -1; + struct timespec end_time, *to = NULL; struct timeval tv; int ret; @@ -385,43 +518,15 @@ asmlinkage long sys_select(int n, fd_set __user *inp, fd_set __user *outp, if (copy_from_user(&tv, tvp, sizeof(tv))) return -EFAULT; - if (tv.tv_sec < 0 || tv.tv_usec < 0) + to = &end_time; + if (poll_select_set_timeout(to, + tv.tv_sec + (tv.tv_usec / USEC_PER_SEC), + (tv.tv_usec % USEC_PER_SEC) * NSEC_PER_USEC)) return -EINVAL; - - /* Cast to u64 to make GCC stop complaining */ - if ((u64)tv.tv_sec >= (u64)MAX_INT64_SECONDS) - timeout = -1; /* infinite */ - else { - timeout = DIV_ROUND_UP(tv.tv_usec, USEC_PER_SEC/HZ); - timeout += tv.tv_sec * HZ; - } } - ret = core_sys_select(n, inp, outp, exp, &timeout); - - if (tvp) { - struct timeval rtv; - - if (current->personality & STICKY_TIMEOUTS) - goto sticky; - rtv.tv_usec = jiffies_to_usecs(do_div((*(u64*)&timeout), HZ)); - rtv.tv_sec = timeout; - if (timeval_compare(&rtv, &tv) >= 0) - rtv = tv; - if (copy_to_user(tvp, &rtv, sizeof(rtv))) { -sticky: - /* - * If an application puts its timeval in read-only - * memory, we don't want the Linux-specific update to - * the timeval to cause a fault after the select has - * completed successfully. However, because we're not - * updating the timeval, we can't restart the system - * call. - */ - if (ret == -ERESTARTNOHAND) - ret = -EINTR; - } - } + ret = core_sys_select(n, inp, outp, exp, to); + ret = poll_select_copy_remaining(&end_time, tvp, 1, ret); return ret; } @@ -431,25 +536,17 @@ asmlinkage long sys_pselect7(int n, fd_set __user *inp, fd_set __user *outp, fd_set __user *exp, struct timespec __user *tsp, const sigset_t __user *sigmask, size_t sigsetsize) { - s64 timeout = MAX_SCHEDULE_TIMEOUT; sigset_t ksigmask, sigsaved; - struct timespec ts; + struct timespec ts, end_time, *to = NULL; int ret; if (tsp) { if (copy_from_user(&ts, tsp, sizeof(ts))) return -EFAULT; - if (ts.tv_sec < 0 || ts.tv_nsec < 0) + to = &end_time; + if (poll_select_set_timeout(to, ts.tv_sec, ts.tv_nsec)) return -EINVAL; - - /* Cast to u64 to make GCC stop complaining */ - if ((u64)ts.tv_sec >= (u64)MAX_INT64_SECONDS) - timeout = -1; /* infinite */ - else { - timeout = DIV_ROUND_UP(ts.tv_nsec, NSEC_PER_SEC/HZ); - timeout += ts.tv_sec * HZ; - } } if (sigmask) { @@ -463,32 +560,8 @@ asmlinkage long sys_pselect7(int n, fd_set __user *inp, fd_set __user *outp, sigprocmask(SIG_SETMASK, &ksigmask, &sigsaved); } - ret = core_sys_select(n, inp, outp, exp, &timeout); - - if (tsp) { - struct timespec rts; - - if (current->personality & STICKY_TIMEOUTS) - goto sticky; - rts.tv_nsec = jiffies_to_usecs(do_div((*(u64*)&timeout), HZ)) * - 1000; - rts.tv_sec = timeout; - if (timespec_compare(&rts, &ts) >= 0) - rts = ts; - if (copy_to_user(tsp, &rts, sizeof(rts))) { -sticky: - /* - * If an application puts its timeval in read-only - * memory, we don't want the Linux-specific update to - * the timeval to cause a fault after the select has - * completed successfully. However, because we're not - * updating the timeval, we can't restart the system - * call. - */ - if (ret == -ERESTARTNOHAND) - ret = -EINTR; - } - } + ret = core_sys_select(n, inp, outp, exp, &end_time); + ret = poll_select_copy_remaining(&end_time, tsp, 0, ret); if (ret == -ERESTARTNOHAND) { /* @@ -574,18 +647,24 @@ static inline unsigned int do_pollfd(struct pollfd *pollfd, poll_table *pwait) } static int do_poll(unsigned int nfds, struct poll_list *list, - struct poll_wqueues *wait, s64 *timeout) + struct poll_wqueues *wait, struct timespec *end_time) { - int count = 0; poll_table* pt = &wait->pt; + ktime_t expire, *to = NULL; + int timed_out = 0, count = 0; + unsigned long slack = 0; /* Optimise the no-wait case */ - if (!(*timeout)) + if (end_time && !end_time->tv_sec && !end_time->tv_nsec) { pt = NULL; + timed_out = 1; + } + + if (end_time && !timed_out) + slack = estimate_accuracy(end_time); for (;;) { struct poll_list *walk; - long __timeout; set_current_state(TASK_INTERRUPTIBLE); for (walk = list; walk != NULL; walk = walk->next) { @@ -617,27 +696,21 @@ static int do_poll(unsigned int nfds, struct poll_list *list, if (signal_pending(current)) count = -EINTR; } - if (count || !*timeout) + if (count || timed_out) break; - if (*timeout < 0) { - /* Wait indefinitely */ - __timeout = MAX_SCHEDULE_TIMEOUT; - } else if (unlikely(*timeout >= (s64)MAX_SCHEDULE_TIMEOUT-1)) { - /* - * Wait for longer than MAX_SCHEDULE_TIMEOUT. Do it in - * a loop - */ - __timeout = MAX_SCHEDULE_TIMEOUT - 1; - *timeout -= __timeout; - } else { - __timeout = *timeout; - *timeout = 0; + /* + * If this is the first loop and we have a timeout + * given, then we convert to ktime_t and set the to + * pointer to the expiry value. + */ + if (end_time && !to) { + expire = timespec_to_ktime(*end_time); + to = &expire; } - __timeout = schedule_timeout(__timeout); - if (*timeout >= 0) - *timeout += __timeout; + if (!schedule_hrtimeout_range(to, slack, HRTIMER_MODE_ABS)) + timed_out = 1; } __set_current_state(TASK_RUNNING); return count; @@ -646,7 +719,8 @@ static int do_poll(unsigned int nfds, struct poll_list *list, #define N_STACK_PPS ((sizeof(stack_pps) - sizeof(struct poll_list)) / \ sizeof(struct pollfd)) -int do_sys_poll(struct pollfd __user *ufds, unsigned int nfds, s64 *timeout) +int do_sys_poll(struct pollfd __user *ufds, unsigned int nfds, + struct timespec *end_time) { struct poll_wqueues table; int err = -EFAULT, fdcount, len, size; @@ -686,7 +760,7 @@ int do_sys_poll(struct pollfd __user *ufds, unsigned int nfds, s64 *timeout) } poll_initwait(&table); - fdcount = do_poll(nfds, head, &table, timeout); + fdcount = do_poll(nfds, head, &table, end_time); poll_freewait(&table); for (walk = head; walk; walk = walk->next) { @@ -712,16 +786,21 @@ out_fds: static long do_restart_poll(struct restart_block *restart_block) { - struct pollfd __user *ufds = (struct pollfd __user*)restart_block->arg0; - int nfds = restart_block->arg1; - s64 timeout = ((s64)restart_block->arg3<<32) | (s64)restart_block->arg2; + struct pollfd __user *ufds = restart_block->poll.ufds; + int nfds = restart_block->poll.nfds; + struct timespec *to = NULL, end_time; int ret; - ret = do_sys_poll(ufds, nfds, &timeout); + if (restart_block->poll.has_timeout) { + end_time.tv_sec = restart_block->poll.tv_sec; + end_time.tv_nsec = restart_block->poll.tv_nsec; + to = &end_time; + } + + ret = do_sys_poll(ufds, nfds, to); + if (ret == -EINTR) { restart_block->fn = do_restart_poll; - restart_block->arg2 = timeout & 0xFFFFFFFF; - restart_block->arg3 = (u64)timeout >> 32; ret = -ERESTART_RESTARTBLOCK; } return ret; @@ -730,31 +809,32 @@ static long do_restart_poll(struct restart_block *restart_block) asmlinkage long sys_poll(struct pollfd __user *ufds, unsigned int nfds, long timeout_msecs) { - s64 timeout_jiffies; + struct timespec end_time, *to = NULL; int ret; - if (timeout_msecs > 0) { -#if HZ > 1000 - /* We can only overflow if HZ > 1000 */ - if (timeout_msecs / 1000 > (s64)0x7fffffffffffffffULL / (s64)HZ) - timeout_jiffies = -1; - else -#endif - timeout_jiffies = msecs_to_jiffies(timeout_msecs) + 1; - } else { - /* Infinite (< 0) or no (0) timeout */ - timeout_jiffies = timeout_msecs; + if (timeout_msecs >= 0) { + to = &end_time; + poll_select_set_timeout(to, timeout_msecs / MSEC_PER_SEC, + NSEC_PER_MSEC * (timeout_msecs % MSEC_PER_SEC)); } - ret = do_sys_poll(ufds, nfds, &timeout_jiffies); + ret = do_sys_poll(ufds, nfds, to); + if (ret == -EINTR) { struct restart_block *restart_block; + restart_block = ¤t_thread_info()->restart_block; restart_block->fn = do_restart_poll; - restart_block->arg0 = (unsigned long)ufds; - restart_block->arg1 = nfds; - restart_block->arg2 = timeout_jiffies & 0xFFFFFFFF; - restart_block->arg3 = (u64)timeout_jiffies >> 32; + restart_block->poll.ufds = ufds; + restart_block->poll.nfds = nfds; + + if (timeout_msecs >= 0) { + restart_block->poll.tv_sec = end_time.tv_sec; + restart_block->poll.tv_nsec = end_time.tv_nsec; + restart_block->poll.has_timeout = 1; + } else + restart_block->poll.has_timeout = 0; + ret = -ERESTART_RESTARTBLOCK; } return ret; @@ -766,21 +846,16 @@ asmlinkage long sys_ppoll(struct pollfd __user *ufds, unsigned int nfds, size_t sigsetsize) { sigset_t ksigmask, sigsaved; - struct timespec ts; - s64 timeout = -1; + struct timespec ts, end_time, *to = NULL; int ret; if (tsp) { if (copy_from_user(&ts, tsp, sizeof(ts))) return -EFAULT; - /* Cast to u64 to make GCC stop complaining */ - if ((u64)ts.tv_sec >= (u64)MAX_INT64_SECONDS) - timeout = -1; /* infinite */ - else { - timeout = DIV_ROUND_UP(ts.tv_nsec, NSEC_PER_SEC/HZ); - timeout += ts.tv_sec * HZ; - } + to = &end_time; + if (poll_select_set_timeout(to, ts.tv_sec, ts.tv_nsec)) + return -EINVAL; } if (sigmask) { @@ -794,7 +869,7 @@ asmlinkage long sys_ppoll(struct pollfd __user *ufds, unsigned int nfds, sigprocmask(SIG_SETMASK, &ksigmask, &sigsaved); } - ret = do_sys_poll(ufds, nfds, &timeout); + ret = do_sys_poll(ufds, nfds, to); /* We can restart this syscall, usually */ if (ret == -EINTR) { @@ -812,31 +887,7 @@ asmlinkage long sys_ppoll(struct pollfd __user *ufds, unsigned int nfds, } else if (sigmask) sigprocmask(SIG_SETMASK, &sigsaved, NULL); - if (tsp && timeout >= 0) { - struct timespec rts; - - if (current->personality & STICKY_TIMEOUTS) - goto sticky; - /* Yes, we know it's actually an s64, but it's also positive. */ - rts.tv_nsec = jiffies_to_usecs(do_div((*(u64*)&timeout), HZ)) * - 1000; - rts.tv_sec = timeout; - if (timespec_compare(&rts, &ts) >= 0) - rts = ts; - if (copy_to_user(tsp, &rts, sizeof(rts))) { - sticky: - /* - * If an application puts its timeval in read-only - * memory, we don't want the Linux-specific update to - * the timeval to cause a fault after the select has - * completed successfully. However, because we're not - * updating the timeval, we can't restart the system - * call. - */ - if (ret == -ERESTARTNOHAND && timeout >= 0) - ret = -EINTR; - } - } + ret = poll_select_copy_remaining(&end_time, tsp, 0, ret); return ret; } diff --git a/fs/seq_file.c b/fs/seq_file.c index bd20f7f5a93..16c211558c2 100644 --- a/fs/seq_file.c +++ b/fs/seq_file.c @@ -357,7 +357,18 @@ int seq_printf(struct seq_file *m, const char *f, ...) } EXPORT_SYMBOL(seq_printf); -static char *mangle_path(char *s, char *p, char *esc) +/** + * mangle_path - mangle and copy path to buffer beginning + * @s: buffer start + * @p: beginning of path in above buffer + * @esc: set of characters that need escaping + * + * Copy the path from @p to @s, replacing each occurrence of character from + * @esc with usual octal escape. + * Returns pointer past last written character in @s, or NULL in case of + * failure. + */ +char *mangle_path(char *s, char *p, char *esc) { while (s <= p) { char c = *p++; @@ -376,6 +387,7 @@ static char *mangle_path(char *s, char *p, char *esc) } return NULL; } +EXPORT_SYMBOL(mangle_path); /* * return the absolute path of 'dentry' residing in mount 'mnt'. @@ -452,17 +464,34 @@ int seq_dentry(struct seq_file *m, struct dentry *dentry, char *esc) int seq_bitmap(struct seq_file *m, unsigned long *bits, unsigned int nr_bits) { - size_t len = bitmap_scnprintf_len(nr_bits); + if (m->count < m->size) { + int len = bitmap_scnprintf(m->buf + m->count, + m->size - m->count, bits, nr_bits); + if (m->count + len < m->size) { + m->count += len; + return 0; + } + } + m->count = m->size; + return -1; +} +EXPORT_SYMBOL(seq_bitmap); - if (m->count + len < m->size) { - bitmap_scnprintf(m->buf + m->count, m->size - m->count, - bits, nr_bits); - m->count += len; - return 0; +int seq_bitmap_list(struct seq_file *m, unsigned long *bits, + unsigned int nr_bits) +{ + if (m->count < m->size) { + int len = bitmap_scnlistprintf(m->buf + m->count, + m->size - m->count, bits, nr_bits); + if (m->count + len < m->size) { + m->count += len; + return 0; + } } m->count = m->size; return -1; } +EXPORT_SYMBOL(seq_bitmap_list); static void *single_start(struct seq_file *p, loff_t *pos) { diff --git a/fs/smbfs/dir.c b/fs/smbfs/dir.c index 48da4fa6b7d..e7ddd0328dd 100644 --- a/fs/smbfs/dir.c +++ b/fs/smbfs/dir.c @@ -667,8 +667,7 @@ smb_make_node(struct inode *dir, struct dentry *dentry, int mode, dev_t dev) attr.ia_valid = ATTR_MODE | ATTR_UID | ATTR_GID; attr.ia_mode = mode; - attr.ia_uid = current->euid; - attr.ia_gid = current->egid; + current_euid_egid(&attr.ia_uid, &attr.ia_gid); if (!new_valid_dev(dev)) return -EINVAL; diff --git a/fs/smbfs/inode.c b/fs/smbfs/inode.c index 3528f40ffb0..fc27fbfc539 100644 --- a/fs/smbfs/inode.c +++ b/fs/smbfs/inode.c @@ -586,7 +586,7 @@ static int smb_fill_super(struct super_block *sb, void *raw_data, int silent) if (parse_options(mnt, raw_data)) goto out_bad_option; } - mnt->mounted_uid = current->uid; + mnt->mounted_uid = current_uid(); smb_setcodepage(server, &mnt->codepage); /* diff --git a/fs/smbfs/proc.c b/fs/smbfs/proc.c index ee536e8a649..9468168b9af 100644 --- a/fs/smbfs/proc.c +++ b/fs/smbfs/proc.c @@ -864,7 +864,7 @@ smb_newconn(struct smb_sb_info *server, struct smb_conn_opt *opt) goto out; error = -EACCES; - if (current->uid != server->mnt->mounted_uid && + if (current_uid() != server->mnt->mounted_uid && !capable(CAP_SYS_ADMIN)) goto out; diff --git a/fs/splice.c b/fs/splice.c index a1e701c2715..1abab5cee4b 100644 --- a/fs/splice.c +++ b/fs/splice.c @@ -731,8 +731,8 @@ ssize_t splice_from_pipe(struct pipe_inode_info *pipe, struct file *out, }; /* - * The actor worker might be calling ->prepare_write and - * ->commit_write. Most of the time, these expect i_mutex to + * The actor worker might be calling ->write_begin and + * ->write_end. Most of the time, these expect i_mutex to * be held. Since this may result in an ABBA deadlock with * pipe->inode, we have to order lock acquiry here. */ diff --git a/fs/super.c b/fs/super.c index e931ae9511f..ddba069d7a9 100644 --- a/fs/super.c +++ b/fs/super.c @@ -682,7 +682,7 @@ void emergency_remount(void) * filesystems which don't use real block-devices. -- jrs */ -static struct idr unnamed_dev_idr; +static DEFINE_IDA(unnamed_dev_ida); static DEFINE_SPINLOCK(unnamed_dev_lock);/* protects the above */ int set_anon_super(struct super_block *s, void *data) @@ -691,10 +691,10 @@ int set_anon_super(struct super_block *s, void *data) int error; retry: - if (idr_pre_get(&unnamed_dev_idr, GFP_ATOMIC) == 0) + if (ida_pre_get(&unnamed_dev_ida, GFP_ATOMIC) == 0) return -ENOMEM; spin_lock(&unnamed_dev_lock); - error = idr_get_new(&unnamed_dev_idr, NULL, &dev); + error = ida_get_new(&unnamed_dev_ida, &dev); spin_unlock(&unnamed_dev_lock); if (error == -EAGAIN) /* We raced and lost with another CPU. */ @@ -704,7 +704,7 @@ int set_anon_super(struct super_block *s, void *data) if ((dev & MAX_ID_MASK) == (1 << MINORBITS)) { spin_lock(&unnamed_dev_lock); - idr_remove(&unnamed_dev_idr, dev); + ida_remove(&unnamed_dev_ida, dev); spin_unlock(&unnamed_dev_lock); return -EMFILE; } @@ -720,17 +720,12 @@ void kill_anon_super(struct super_block *sb) generic_shutdown_super(sb); spin_lock(&unnamed_dev_lock); - idr_remove(&unnamed_dev_idr, slot); + ida_remove(&unnamed_dev_ida, slot); spin_unlock(&unnamed_dev_lock); } EXPORT_SYMBOL(kill_anon_super); -void __init unnamed_dev_init(void) -{ - idr_init(&unnamed_dev_idr); -} - void kill_litter_super(struct super_block *sb) { if (sb->s_root) @@ -760,9 +755,13 @@ int get_sb_bdev(struct file_system_type *fs_type, { struct block_device *bdev; struct super_block *s; + fmode_t mode = FMODE_READ; int error = 0; - bdev = open_bdev_excl(dev_name, flags, fs_type); + if (!(flags & MS_RDONLY)) + mode |= FMODE_WRITE; + + bdev = open_bdev_exclusive(dev_name, mode, fs_type); if (IS_ERR(bdev)) return PTR_ERR(bdev); @@ -785,11 +784,12 @@ int get_sb_bdev(struct file_system_type *fs_type, goto error_bdev; } - close_bdev_excl(bdev); + close_bdev_exclusive(bdev, mode); } else { char b[BDEVNAME_SIZE]; s->s_flags = flags; + s->s_mode = mode; strlcpy(s->s_id, bdevname(bdev, b), sizeof(s->s_id)); sb_set_blocksize(s, block_size(bdev)); error = fill_super(s, data, flags & MS_SILENT ? 1 : 0); @@ -807,7 +807,7 @@ int get_sb_bdev(struct file_system_type *fs_type, error_s: error = PTR_ERR(s); error_bdev: - close_bdev_excl(bdev); + close_bdev_exclusive(bdev, mode); error: return error; } @@ -817,10 +817,11 @@ EXPORT_SYMBOL(get_sb_bdev); void kill_block_super(struct super_block *sb) { struct block_device *bdev = sb->s_bdev; + fmode_t mode = sb->s_mode; generic_shutdown_super(sb); sync_blockdev(bdev); - close_bdev_excl(bdev); + close_bdev_exclusive(bdev, mode); } EXPORT_SYMBOL(kill_block_super); @@ -913,7 +914,7 @@ vfs_kern_mount(struct file_system_type *type, int flags, const char *name, void goto out_free_secdata; BUG_ON(!mnt->mnt_sb); - error = security_sb_kern_mount(mnt->mnt_sb, secdata); + error = security_sb_kern_mount(mnt->mnt_sb, flags, secdata); if (error) goto out_sb; diff --git a/fs/sysfs/bin.c b/fs/sysfs/bin.c index 006fc64227d..66f6e58a7e4 100644 --- a/fs/sysfs/bin.c +++ b/fs/sysfs/bin.c @@ -61,6 +61,7 @@ read(struct file *file, char __user *userbuf, size_t bytes, loff_t *off) int size = dentry->d_inode->i_size; loff_t offs = *off; int count = min_t(size_t, bytes, PAGE_SIZE); + char *temp; if (size) { if (offs > size) @@ -69,23 +70,33 @@ read(struct file *file, char __user *userbuf, size_t bytes, loff_t *off) count = size - offs; } + temp = kmalloc(count, GFP_KERNEL); + if (!temp) + return -ENOMEM; + mutex_lock(&bb->mutex); count = fill_read(dentry, bb->buffer, offs, count); - if (count < 0) - goto out_unlock; + if (count < 0) { + mutex_unlock(&bb->mutex); + goto out_free; + } - if (copy_to_user(userbuf, bb->buffer, count)) { + memcpy(temp, bb->buffer, count); + + mutex_unlock(&bb->mutex); + + if (copy_to_user(userbuf, temp, count)) { count = -EFAULT; - goto out_unlock; + goto out_free; } pr_debug("offs = %lld, *off = %lld, count = %d\n", offs, *off, count); *off = offs + count; - out_unlock: - mutex_unlock(&bb->mutex); + out_free: + kfree(temp); return count; } @@ -118,6 +129,7 @@ static ssize_t write(struct file *file, const char __user *userbuf, int size = dentry->d_inode->i_size; loff_t offs = *off; int count = min_t(size_t, bytes, PAGE_SIZE); + char *temp; if (size) { if (offs > size) @@ -126,19 +138,27 @@ static ssize_t write(struct file *file, const char __user *userbuf, count = size - offs; } - mutex_lock(&bb->mutex); + temp = kmalloc(count, GFP_KERNEL); + if (!temp) + return -ENOMEM; - if (copy_from_user(bb->buffer, userbuf, count)) { + if (copy_from_user(temp, userbuf, count)) { count = -EFAULT; - goto out_unlock; + goto out_free; } + mutex_lock(&bb->mutex); + + memcpy(bb->buffer, temp, count); + count = flush_write(dentry, bb->buffer, offs, count); + mutex_unlock(&bb->mutex); + if (count > 0) *off = offs + count; - out_unlock: - mutex_unlock(&bb->mutex); +out_free: + kfree(temp); return count; } diff --git a/fs/sysfs/dir.c b/fs/sysfs/dir.c index aedaeba82ae..82d3b79d0e0 100644 --- a/fs/sysfs/dir.c +++ b/fs/sysfs/dir.c @@ -370,17 +370,17 @@ void sysfs_addrm_start(struct sysfs_addrm_cxt *acxt, memset(acxt, 0, sizeof(*acxt)); acxt->parent_sd = parent_sd; - /* Lookup parent inode. inode initialization and I_NEW - * clearing are protected by sysfs_mutex. By grabbing it and - * looking up with _nowait variant, inode state can be - * determined reliably. + /* Lookup parent inode. inode initialization is protected by + * sysfs_mutex, so inode existence can be determined by + * looking up inode while holding sysfs_mutex. */ mutex_lock(&sysfs_mutex); - inode = ilookup5_nowait(sysfs_sb, parent_sd->s_ino, sysfs_ilookup_test, - parent_sd); + inode = ilookup5(sysfs_sb, parent_sd->s_ino, sysfs_ilookup_test, + parent_sd); + if (inode) { + WARN_ON(inode->i_state & I_NEW); - if (inode && !(inode->i_state & I_NEW)) { /* parent inode available */ acxt->parent_inode = inode; @@ -393,8 +393,7 @@ void sysfs_addrm_start(struct sysfs_addrm_cxt *acxt, mutex_lock(&inode->i_mutex); mutex_lock(&sysfs_mutex); } - } else - iput(inode); + } } /** @@ -636,6 +635,7 @@ struct sysfs_dirent *sysfs_get_dirent(struct sysfs_dirent *parent_sd, return sd; } +EXPORT_SYMBOL_GPL(sysfs_get_dirent); static int create_dir(struct kobject *kobj, struct sysfs_dirent *parent_sd, const char *name, struct sysfs_dirent **p_sd) @@ -829,16 +829,12 @@ int sysfs_rename_dir(struct kobject * kobj, const char *new_name) if (!new_dentry) goto out_unlock; - /* rename kobject and sysfs_dirent */ + /* rename sysfs_dirent */ error = -ENOMEM; new_name = dup_name = kstrdup(new_name, GFP_KERNEL); if (!new_name) goto out_unlock; - error = kobject_set_name(kobj, "%s", new_name); - if (error) - goto out_unlock; - dup_name = sd->s_name; sd->s_name = new_name; @@ -987,4 +983,5 @@ static int sysfs_readdir(struct file * filp, void * dirent, filldir_t filldir) const struct file_operations sysfs_dir_operations = { .read = generic_read_dir, .readdir = sysfs_readdir, + .llseek = generic_file_llseek, }; diff --git a/fs/sysfs/file.c b/fs/sysfs/file.c index c9e4e5091da..1f4a3f87726 100644 --- a/fs/sysfs/file.c +++ b/fs/sysfs/file.c @@ -19,10 +19,18 @@ #include <linux/poll.h> #include <linux/list.h> #include <linux/mutex.h> +#include <linux/limits.h> #include <asm/uaccess.h> #include "sysfs.h" +/* used in crash dumps to help with debugging */ +static char last_sysfs_file[PATH_MAX]; +void sysfs_printk_last_file(void) +{ + printk(KERN_EMERG "last sysfs file: %s\n", last_sysfs_file); +} + /* * There's one sysfs_buffer for each open file and one * sysfs_open_dirent for each sysfs_dirent with one or more open @@ -328,6 +336,11 @@ static int sysfs_open_file(struct inode *inode, struct file *file) struct sysfs_buffer *buffer; struct sysfs_ops *ops; int error = -EACCES; + char *p; + + p = d_path(&file->f_path, last_sysfs_file, sizeof(last_sysfs_file)); + if (p) + memmove(last_sysfs_file, p, strlen(p) + 1); /* need attr_sd for attr and ops, its parent for kobj */ if (!sysfs_get_active_two(attr_sd)) @@ -440,7 +453,23 @@ static unsigned int sysfs_poll(struct file *filp, poll_table *wait) return POLLERR|POLLPRI; } -void sysfs_notify(struct kobject *k, char *dir, char *attr) +void sysfs_notify_dirent(struct sysfs_dirent *sd) +{ + struct sysfs_open_dirent *od; + + spin_lock(&sysfs_open_dirent_lock); + + od = sd->s_attr.open; + if (od) { + atomic_inc(&od->event); + wake_up_interruptible(&od->poll); + } + + spin_unlock(&sysfs_open_dirent_lock); +} +EXPORT_SYMBOL_GPL(sysfs_notify_dirent); + +void sysfs_notify(struct kobject *k, const char *dir, const char *attr) { struct sysfs_dirent *sd = k->sd; @@ -450,19 +479,8 @@ void sysfs_notify(struct kobject *k, char *dir, char *attr) sd = sysfs_find_dirent(sd, dir); if (sd && attr) sd = sysfs_find_dirent(sd, attr); - if (sd) { - struct sysfs_open_dirent *od; - - spin_lock(&sysfs_open_dirent_lock); - - od = sd->s_attr.open; - if (od) { - atomic_inc(&od->event); - wake_up_interruptible(&od->poll); - } - - spin_unlock(&sysfs_open_dirent_lock); - } + if (sd) + sysfs_notify_dirent(sd); mutex_unlock(&sysfs_mutex); } diff --git a/fs/sysfs/mount.c b/fs/sysfs/mount.c index 14f0023984d..ab343e371d6 100644 --- a/fs/sysfs/mount.c +++ b/fs/sysfs/mount.c @@ -16,6 +16,7 @@ #include <linux/mount.h> #include <linux/pagemap.h> #include <linux/init.h> +#include <linux/module.h> #include "sysfs.h" @@ -115,3 +116,17 @@ out_err: sysfs_dir_cachep = NULL; goto out; } + +#undef sysfs_get +struct sysfs_dirent *sysfs_get(struct sysfs_dirent *sd) +{ + return __sysfs_get(sd); +} +EXPORT_SYMBOL_GPL(sysfs_get); + +#undef sysfs_put +void sysfs_put(struct sysfs_dirent *sd) +{ + __sysfs_put(sd); +} +EXPORT_SYMBOL_GPL(sysfs_put); diff --git a/fs/sysfs/sysfs.h b/fs/sysfs/sysfs.h index a5db496f71c..93c6d6b27c4 100644 --- a/fs/sysfs/sysfs.h +++ b/fs/sysfs/sysfs.h @@ -124,7 +124,7 @@ int sysfs_create_subdir(struct kobject *kobj, const char *name, struct sysfs_dirent **p_sd); void sysfs_remove_subdir(struct sysfs_dirent *sd); -static inline struct sysfs_dirent *sysfs_get(struct sysfs_dirent *sd) +static inline struct sysfs_dirent *__sysfs_get(struct sysfs_dirent *sd) { if (sd) { WARN_ON(!atomic_read(&sd->s_count)); @@ -132,12 +132,14 @@ static inline struct sysfs_dirent *sysfs_get(struct sysfs_dirent *sd) } return sd; } +#define sysfs_get(sd) __sysfs_get(sd) -static inline void sysfs_put(struct sysfs_dirent *sd) +static inline void __sysfs_put(struct sysfs_dirent *sd) { if (sd && atomic_dec_and_test(&sd->s_count)) release_sysfs_dirent(sd); } +#define sysfs_put(sd) __sysfs_put(sd) /* * inode.c diff --git a/fs/sysv/ialloc.c b/fs/sysv/ialloc.c index 115ab0d6f4b..241e9765cfa 100644 --- a/fs/sysv/ialloc.c +++ b/fs/sysv/ialloc.c @@ -165,9 +165,9 @@ struct inode * sysv_new_inode(const struct inode * dir, mode_t mode) if (S_ISDIR(mode)) mode |= S_ISGID; } else - inode->i_gid = current->fsgid; + inode->i_gid = current_fsgid(); - inode->i_uid = current->fsuid; + inode->i_uid = current_fsuid(); inode->i_ino = fs16_to_cpu(sbi, ino); inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME_SEC; inode->i_blocks = 0; diff --git a/fs/timerfd.c b/fs/timerfd.c index c502c60e4f5..0862f0e49d0 100644 --- a/fs/timerfd.c +++ b/fs/timerfd.c @@ -52,11 +52,9 @@ static enum hrtimer_restart timerfd_tmrproc(struct hrtimer *htmr) static ktime_t timerfd_get_remaining(struct timerfd_ctx *ctx) { - ktime_t now, remaining; - - now = ctx->tmr.base->get_time(); - remaining = ktime_sub(ctx->tmr.expires, now); + ktime_t remaining; + remaining = hrtimer_expires_remaining(&ctx->tmr); return remaining.tv64 < 0 ? ktime_set(0, 0): remaining; } @@ -74,7 +72,7 @@ static void timerfd_setup(struct timerfd_ctx *ctx, int flags, ctx->ticks = 0; ctx->tintv = timespec_to_ktime(ktmr->it_interval); hrtimer_init(&ctx->tmr, ctx->clockid, htmode); - ctx->tmr.expires = texp; + hrtimer_set_expires(&ctx->tmr, texp); ctx->tmr.function = timerfd_tmrproc; if (texp.tv64 != 0) hrtimer_start(&ctx->tmr, texp, htmode); diff --git a/fs/ubifs/budget.c b/fs/ubifs/budget.c index 73db464cd08..4a18f084cc4 100644 --- a/fs/ubifs/budget.c +++ b/fs/ubifs/budget.c @@ -363,7 +363,7 @@ long long ubifs_calc_available(const struct ubifs_info *c, int min_idx_lebs) */ static int can_use_rp(struct ubifs_info *c) { - if (current->fsuid == c->rp_uid || capable(CAP_SYS_RESOURCE) || + if (current_fsuid() == c->rp_uid || capable(CAP_SYS_RESOURCE) || (c->rp_gid != 0 && in_group_p(c->rp_gid))) return 1; return 0; @@ -414,19 +414,21 @@ static int do_budget_space(struct ubifs_info *c) * @c->lst.empty_lebs + @c->freeable_cnt + @c->idx_gc_cnt - * @c->lst.taken_empty_lebs * - * @empty_lebs are available because they are empty. @freeable_cnt are - * available because they contain only free and dirty space and the - * index allocation always occurs after wbufs are synch'ed. - * @idx_gc_cnt are available because they are index LEBs that have been - * garbage collected (including trivial GC) and are awaiting the commit - * before they can be unmapped - note that the in-the-gaps method will - * grab these if it needs them. @taken_empty_lebs are empty_lebs that - * have already been allocated for some purpose (also includes those - * LEBs on the @idx_gc list). + * @c->lst.empty_lebs are available because they are empty. + * @c->freeable_cnt are available because they contain only free and + * dirty space, @c->idx_gc_cnt are available because they are index + * LEBs that have been garbage collected and are awaiting the commit + * before they can be used. And the in-the-gaps method will grab these + * if it needs them. @c->lst.taken_empty_lebs are empty LEBs that have + * already been allocated for some purpose. * - * Note, @taken_empty_lebs may temporarily be higher by one because of - * the way we serialize LEB allocations and budgeting. See a comment in - * 'ubifs_find_free_space()'. + * Note, @c->idx_gc_cnt is included to both @c->lst.empty_lebs (because + * these LEBs are empty) and to @c->lst.taken_empty_lebs (because they + * are taken until after the commit). + * + * Note, @c->lst.taken_empty_lebs may temporarily be higher by one + * because of the way we serialize LEB allocations and budgeting. See a + * comment in 'ubifs_find_free_space()'. */ lebs = c->lst.empty_lebs + c->freeable_cnt + c->idx_gc_cnt - c->lst.taken_empty_lebs; diff --git a/fs/ubifs/commit.c b/fs/ubifs/commit.c index 0a6aa2cc78f..b49884c8c10 100644 --- a/fs/ubifs/commit.c +++ b/fs/ubifs/commit.c @@ -234,8 +234,8 @@ int ubifs_bg_thread(void *info) int err; struct ubifs_info *c = info; - ubifs_msg("background thread \"%s\" started, PID %d", - c->bgt_name, current->pid); + dbg_msg("background thread \"%s\" started, PID %d", + c->bgt_name, current->pid); set_freezable(); while (1) { diff --git a/fs/ubifs/compress.c b/fs/ubifs/compress.c index 5bb51dac3c1..a0ada596b17 100644 --- a/fs/ubifs/compress.c +++ b/fs/ubifs/compress.c @@ -91,8 +91,6 @@ struct ubifs_compressor *ubifs_compressors[UBIFS_COMPR_TYPES_CNT]; * * Note, if the input buffer was not compressed, it is copied to the output * buffer and %UBIFS_COMPR_NONE is returned in @compr_type. - * - * This functions returns %0 on success or a negative error code on failure. */ void ubifs_compress(const void *in_buf, int in_len, void *out_buf, int *out_len, int *compr_type) diff --git a/fs/ubifs/debug.c b/fs/ubifs/debug.c index d7f7645779f..510ffa0bbda 100644 --- a/fs/ubifs/debug.c +++ b/fs/ubifs/debug.c @@ -101,21 +101,24 @@ static void sprintf_key(const struct ubifs_info *c, const union ubifs_key *key, if (c->key_fmt == UBIFS_SIMPLE_KEY_FMT) { switch (type) { case UBIFS_INO_KEY: - sprintf(p, "(%lu, %s)", key_inum(c, key), + sprintf(p, "(%lu, %s)", (unsigned long)key_inum(c, key), get_key_type(type)); break; case UBIFS_DENT_KEY: case UBIFS_XENT_KEY: - sprintf(p, "(%lu, %s, %#08x)", key_inum(c, key), + sprintf(p, "(%lu, %s, %#08x)", + (unsigned long)key_inum(c, key), get_key_type(type), key_hash(c, key)); break; case UBIFS_DATA_KEY: - sprintf(p, "(%lu, %s, %u)", key_inum(c, key), + sprintf(p, "(%lu, %s, %u)", + (unsigned long)key_inum(c, key), get_key_type(type), key_block(c, key)); break; case UBIFS_TRUN_KEY: sprintf(p, "(%lu, %s)", - key_inum(c, key), get_key_type(type)); + (unsigned long)key_inum(c, key), + get_key_type(type)); break; default: sprintf(p, "(bad key type: %#08x, %#08x)", @@ -222,30 +225,38 @@ void dbg_dump_inode(const struct ubifs_info *c, const struct inode *inode) { const struct ubifs_inode *ui = ubifs_inode(inode); - printk(KERN_DEBUG "inode %lu\n", inode->i_ino); - printk(KERN_DEBUG "size %llu\n", + printk(KERN_DEBUG "Dump in-memory inode:"); + printk(KERN_DEBUG "\tinode %lu\n", inode->i_ino); + printk(KERN_DEBUG "\tsize %llu\n", (unsigned long long)i_size_read(inode)); - printk(KERN_DEBUG "nlink %u\n", inode->i_nlink); - printk(KERN_DEBUG "uid %u\n", (unsigned int)inode->i_uid); - printk(KERN_DEBUG "gid %u\n", (unsigned int)inode->i_gid); - printk(KERN_DEBUG "atime %u.%u\n", + printk(KERN_DEBUG "\tnlink %u\n", inode->i_nlink); + printk(KERN_DEBUG "\tuid %u\n", (unsigned int)inode->i_uid); + printk(KERN_DEBUG "\tgid %u\n", (unsigned int)inode->i_gid); + printk(KERN_DEBUG "\tatime %u.%u\n", (unsigned int)inode->i_atime.tv_sec, (unsigned int)inode->i_atime.tv_nsec); - printk(KERN_DEBUG "mtime %u.%u\n", + printk(KERN_DEBUG "\tmtime %u.%u\n", (unsigned int)inode->i_mtime.tv_sec, (unsigned int)inode->i_mtime.tv_nsec); - printk(KERN_DEBUG "ctime %u.%u\n", + printk(KERN_DEBUG "\tctime %u.%u\n", (unsigned int)inode->i_ctime.tv_sec, (unsigned int)inode->i_ctime.tv_nsec); - printk(KERN_DEBUG "creat_sqnum %llu\n", ui->creat_sqnum); - printk(KERN_DEBUG "xattr_size %u\n", ui->xattr_size); - printk(KERN_DEBUG "xattr_cnt %u\n", ui->xattr_cnt); - printk(KERN_DEBUG "xattr_names %u\n", ui->xattr_names); - printk(KERN_DEBUG "dirty %u\n", ui->dirty); - printk(KERN_DEBUG "xattr %u\n", ui->xattr); - printk(KERN_DEBUG "flags %d\n", ui->flags); - printk(KERN_DEBUG "compr_type %d\n", ui->compr_type); - printk(KERN_DEBUG "data_len %d\n", ui->data_len); + printk(KERN_DEBUG "\tcreat_sqnum %llu\n", ui->creat_sqnum); + printk(KERN_DEBUG "\txattr_size %u\n", ui->xattr_size); + printk(KERN_DEBUG "\txattr_cnt %u\n", ui->xattr_cnt); + printk(KERN_DEBUG "\txattr_names %u\n", ui->xattr_names); + printk(KERN_DEBUG "\tdirty %u\n", ui->dirty); + printk(KERN_DEBUG "\txattr %u\n", ui->xattr); + printk(KERN_DEBUG "\tbulk_read %u\n", ui->xattr); + printk(KERN_DEBUG "\tsynced_i_size %llu\n", + (unsigned long long)ui->synced_i_size); + printk(KERN_DEBUG "\tui_size %llu\n", + (unsigned long long)ui->ui_size); + printk(KERN_DEBUG "\tflags %d\n", ui->flags); + printk(KERN_DEBUG "\tcompr_type %d\n", ui->compr_type); + printk(KERN_DEBUG "\tlast_page_read %lu\n", ui->last_page_read); + printk(KERN_DEBUG "\tread_in_a_row %lu\n", ui->read_in_a_row); + printk(KERN_DEBUG "\tdata_len %d\n", ui->data_len); } void dbg_dump_node(const struct ubifs_info *c, const void *node) @@ -356,8 +367,8 @@ void dbg_dump_node(const struct ubifs_info *c, const void *node) le32_to_cpu(mst->ihead_lnum)); printk(KERN_DEBUG "\tihead_offs %u\n", le32_to_cpu(mst->ihead_offs)); - printk(KERN_DEBUG "\tindex_size %u\n", - le32_to_cpu(mst->index_size)); + printk(KERN_DEBUG "\tindex_size %llu\n", + (unsigned long long)le64_to_cpu(mst->index_size)); printk(KERN_DEBUG "\tlpt_lnum %u\n", le32_to_cpu(mst->lpt_lnum)); printk(KERN_DEBUG "\tlpt_offs %u\n", @@ -647,6 +658,43 @@ void dbg_dump_lprops(struct ubifs_info *c) } } +void dbg_dump_lpt_info(struct ubifs_info *c) +{ + int i; + + spin_lock(&dbg_lock); + printk(KERN_DEBUG "\tlpt_sz: %lld\n", c->lpt_sz); + printk(KERN_DEBUG "\tpnode_sz: %d\n", c->pnode_sz); + printk(KERN_DEBUG "\tnnode_sz: %d\n", c->nnode_sz); + printk(KERN_DEBUG "\tltab_sz: %d\n", c->ltab_sz); + printk(KERN_DEBUG "\tlsave_sz: %d\n", c->lsave_sz); + printk(KERN_DEBUG "\tbig_lpt: %d\n", c->big_lpt); + printk(KERN_DEBUG "\tlpt_hght: %d\n", c->lpt_hght); + printk(KERN_DEBUG "\tpnode_cnt: %d\n", c->pnode_cnt); + printk(KERN_DEBUG "\tnnode_cnt: %d\n", c->nnode_cnt); + printk(KERN_DEBUG "\tdirty_pn_cnt: %d\n", c->dirty_pn_cnt); + printk(KERN_DEBUG "\tdirty_nn_cnt: %d\n", c->dirty_nn_cnt); + printk(KERN_DEBUG "\tlsave_cnt: %d\n", c->lsave_cnt); + printk(KERN_DEBUG "\tspace_bits: %d\n", c->space_bits); + printk(KERN_DEBUG "\tlpt_lnum_bits: %d\n", c->lpt_lnum_bits); + printk(KERN_DEBUG "\tlpt_offs_bits: %d\n", c->lpt_offs_bits); + printk(KERN_DEBUG "\tlpt_spc_bits: %d\n", c->lpt_spc_bits); + printk(KERN_DEBUG "\tpcnt_bits: %d\n", c->pcnt_bits); + printk(KERN_DEBUG "\tlnum_bits: %d\n", c->lnum_bits); + printk(KERN_DEBUG "\tLPT root is at %d:%d\n", c->lpt_lnum, c->lpt_offs); + printk(KERN_DEBUG "\tLPT head is at %d:%d\n", + c->nhead_lnum, c->nhead_offs); + printk(KERN_DEBUG "\tLPT ltab is at %d:%d\n", c->ltab_lnum, c->ltab_offs); + if (c->big_lpt) + printk(KERN_DEBUG "\tLPT lsave is at %d:%d\n", + c->lsave_lnum, c->lsave_offs); + for (i = 0; i < c->lpt_lebs; i++) + printk(KERN_DEBUG "\tLPT LEB %d free %d dirty %d tgc %d " + "cmt %d\n", i + c->lpt_first, c->ltab[i].free, + c->ltab[i].dirty, c->ltab[i].tgc, c->ltab[i].cmt); + spin_unlock(&dbg_lock); +} + void dbg_dump_leb(const struct ubifs_info *c, int lnum) { struct ubifs_scan_leb *sleb; @@ -1544,7 +1592,7 @@ static struct fsck_inode *add_inode(struct ubifs_info *c, if (inum > c->highest_inum) { ubifs_err("too high inode number, max. is %lu", - c->highest_inum); + (unsigned long)c->highest_inum); return ERR_PTR(-EINVAL); } @@ -1623,16 +1671,18 @@ static struct fsck_inode *read_add_inode(struct ubifs_info *c, ino_key_init(c, &key, inum); err = ubifs_lookup_level0(c, &key, &znode, &n); if (!err) { - ubifs_err("inode %lu not found in index", inum); + ubifs_err("inode %lu not found in index", (unsigned long)inum); return ERR_PTR(-ENOENT); } else if (err < 0) { - ubifs_err("error %d while looking up inode %lu", err, inum); + ubifs_err("error %d while looking up inode %lu", + err, (unsigned long)inum); return ERR_PTR(err); } zbr = &znode->zbranch[n]; if (zbr->len < UBIFS_INO_NODE_SZ) { - ubifs_err("bad node %lu node length %d", inum, zbr->len); + ubifs_err("bad node %lu node length %d", + (unsigned long)inum, zbr->len); return ERR_PTR(-EINVAL); } @@ -1652,7 +1702,7 @@ static struct fsck_inode *read_add_inode(struct ubifs_info *c, kfree(ino); if (IS_ERR(fscki)) { ubifs_err("error %ld while adding inode %lu node", - PTR_ERR(fscki), inum); + PTR_ERR(fscki), (unsigned long)inum); return fscki; } @@ -1741,7 +1791,8 @@ static int check_leaf(struct ubifs_info *c, struct ubifs_zbranch *zbr, if (IS_ERR(fscki)) { err = PTR_ERR(fscki); ubifs_err("error %d while processing data node and " - "trying to find inode node %lu", err, inum); + "trying to find inode node %lu", + err, (unsigned long)inum); goto out_dump; } @@ -1774,7 +1825,8 @@ static int check_leaf(struct ubifs_info *c, struct ubifs_zbranch *zbr, if (IS_ERR(fscki)) { err = PTR_ERR(fscki); ubifs_err("error %d while processing entry node and " - "trying to find inode node %lu", err, inum); + "trying to find inode node %lu", + err, (unsigned long)inum); goto out_dump; } @@ -1787,7 +1839,7 @@ static int check_leaf(struct ubifs_info *c, struct ubifs_zbranch *zbr, err = PTR_ERR(fscki); ubifs_err("error %d while processing entry node and " "trying to find parent inode node %lu", - err, inum); + err, (unsigned long)inum); goto out_dump; } @@ -1878,7 +1930,8 @@ static int check_inodes(struct ubifs_info *c, struct fsck_data *fsckd) fscki->references != 1) { ubifs_err("directory inode %lu has %d " "direntries which refer it, but " - "should be 1", fscki->inum, + "should be 1", + (unsigned long)fscki->inum, fscki->references); goto out_dump; } @@ -1886,27 +1939,29 @@ static int check_inodes(struct ubifs_info *c, struct fsck_data *fsckd) fscki->references != 0) { ubifs_err("root inode %lu has non-zero (%d) " "direntries which refer it", - fscki->inum, fscki->references); + (unsigned long)fscki->inum, + fscki->references); goto out_dump; } if (fscki->calc_sz != fscki->size) { ubifs_err("directory inode %lu size is %lld, " "but calculated size is %lld", - fscki->inum, fscki->size, - fscki->calc_sz); + (unsigned long)fscki->inum, + fscki->size, fscki->calc_sz); goto out_dump; } if (fscki->calc_cnt != fscki->nlink) { ubifs_err("directory inode %lu nlink is %d, " "but calculated nlink is %d", - fscki->inum, fscki->nlink, - fscki->calc_cnt); + (unsigned long)fscki->inum, + fscki->nlink, fscki->calc_cnt); goto out_dump; } } else { if (fscki->references != fscki->nlink) { ubifs_err("inode %lu nlink is %d, but " - "calculated nlink is %d", fscki->inum, + "calculated nlink is %d", + (unsigned long)fscki->inum, fscki->nlink, fscki->references); goto out_dump; } @@ -1914,20 +1969,21 @@ static int check_inodes(struct ubifs_info *c, struct fsck_data *fsckd) if (fscki->xattr_sz != fscki->calc_xsz) { ubifs_err("inode %lu has xattr size %u, but " "calculated size is %lld", - fscki->inum, fscki->xattr_sz, + (unsigned long)fscki->inum, fscki->xattr_sz, fscki->calc_xsz); goto out_dump; } if (fscki->xattr_cnt != fscki->calc_xcnt) { ubifs_err("inode %lu has %u xattrs, but " - "calculated count is %lld", fscki->inum, + "calculated count is %lld", + (unsigned long)fscki->inum, fscki->xattr_cnt, fscki->calc_xcnt); goto out_dump; } if (fscki->xattr_nms != fscki->calc_xnms) { ubifs_err("inode %lu has xattr names' size %u, but " "calculated names' size is %lld", - fscki->inum, fscki->xattr_nms, + (unsigned long)fscki->inum, fscki->xattr_nms, fscki->calc_xnms); goto out_dump; } @@ -1940,11 +1996,12 @@ out_dump: ino_key_init(c, &key, fscki->inum); err = ubifs_lookup_level0(c, &key, &znode, &n); if (!err) { - ubifs_err("inode %lu not found in index", fscki->inum); + ubifs_err("inode %lu not found in index", + (unsigned long)fscki->inum); return -ENOENT; } else if (err < 0) { ubifs_err("error %d while looking up inode %lu", - err, fscki->inum); + err, (unsigned long)fscki->inum); return err; } @@ -1962,7 +2019,7 @@ out_dump: } ubifs_msg("dump of the inode %lu sitting in LEB %d:%d", - fscki->inum, zbr->lnum, zbr->offs); + (unsigned long)fscki->inum, zbr->lnum, zbr->offs); dbg_dump_node(c, ino); kfree(ino); return -EINVAL; diff --git a/fs/ubifs/debug.h b/fs/ubifs/debug.h index 50315fc5718..33d6b95071e 100644 --- a/fs/ubifs/debug.h +++ b/fs/ubifs/debug.h @@ -224,6 +224,7 @@ void dbg_dump_lstats(const struct ubifs_lp_stats *lst); void dbg_dump_budg(struct ubifs_info *c); void dbg_dump_lprop(const struct ubifs_info *c, const struct ubifs_lprops *lp); void dbg_dump_lprops(struct ubifs_info *c); +void dbg_dump_lpt_info(struct ubifs_info *c); void dbg_dump_leb(const struct ubifs_info *c, int lnum); void dbg_dump_znode(const struct ubifs_info *c, const struct ubifs_znode *znode); @@ -249,6 +250,8 @@ int dbg_old_index_check_init(struct ubifs_info *c, struct ubifs_zbranch *zroot); int dbg_check_old_index(struct ubifs_info *c, struct ubifs_zbranch *zroot); int dbg_check_cats(struct ubifs_info *c); int dbg_check_ltab(struct ubifs_info *c); +int dbg_chk_lpt_free_spc(struct ubifs_info *c); +int dbg_chk_lpt_sz(struct ubifs_info *c, int action, int len); int dbg_check_synced_i_size(struct inode *inode); int dbg_check_dir_size(struct ubifs_info *c, const struct inode *dir); int dbg_check_tnc(struct ubifs_info *c, int extra); @@ -367,6 +370,7 @@ static inline int dbg_change(struct ubi_volume_desc *desc, int lnum, #define dbg_dump_budg(c) ({}) #define dbg_dump_lprop(c, lp) ({}) #define dbg_dump_lprops(c) ({}) +#define dbg_dump_lpt_info(c) ({}) #define dbg_dump_leb(c, lnum) ({}) #define dbg_dump_znode(c, znode) ({}) #define dbg_dump_heap(c, heap, cat) ({}) @@ -379,6 +383,8 @@ static inline int dbg_change(struct ubi_volume_desc *desc, int lnum, #define dbg_check_old_index(c, zroot) 0 #define dbg_check_cats(c) 0 #define dbg_check_ltab(c) 0 +#define dbg_chk_lpt_free_spc(c) 0 +#define dbg_chk_lpt_sz(c, action, len) 0 #define dbg_check_synced_i_size(inode) 0 #define dbg_check_dir_size(c, dir) 0 #define dbg_check_tnc(c, x) 0 diff --git a/fs/ubifs/dir.c b/fs/ubifs/dir.c index 526c01ec800..f448ab1f9c3 100644 --- a/fs/ubifs/dir.c +++ b/fs/ubifs/dir.c @@ -104,13 +104,13 @@ struct inode *ubifs_new_inode(struct ubifs_info *c, const struct inode *dir, */ inode->i_flags |= (S_NOCMTIME); - inode->i_uid = current->fsuid; + inode->i_uid = current_fsuid(); if (dir->i_mode & S_ISGID) { inode->i_gid = dir->i_gid; if (S_ISDIR(mode)) mode |= S_ISGID; } else - inode->i_gid = current->fsgid; + inode->i_gid = current_fsgid(); inode->i_mode = mode; inode->i_mtime = inode->i_atime = inode->i_ctime = ubifs_current_time(inode); @@ -161,7 +161,7 @@ struct inode *ubifs_new_inode(struct ubifs_info *c, const struct inode *dir, return ERR_PTR(-EINVAL); } ubifs_warn("running out of inode numbers (current %lu, max %d)", - c->highest_inum, INUM_WATERMARK); + (unsigned long)c->highest_inum, INUM_WATERMARK); } inode->i_ino = ++c->highest_inum; @@ -428,7 +428,8 @@ static int ubifs_readdir(struct file *file, void *dirent, filldir_t filldir) dbg_gen("feed '%s', ino %llu, new f_pos %#x", dent->name, (unsigned long long)le64_to_cpu(dent->inum), key_hash_flash(c, &dent->key)); - ubifs_assert(dent->ch.sqnum > ubifs_inode(dir)->creat_sqnum); + ubifs_assert(le64_to_cpu(dent->ch.sqnum) > + ubifs_inode(dir)->creat_sqnum); nm.len = le16_to_cpu(dent->nlen); over = filldir(dirent, dent->name, nm.len, file->f_pos, diff --git a/fs/ubifs/file.c b/fs/ubifs/file.c index 3d698e2022b..2624411d975 100644 --- a/fs/ubifs/file.c +++ b/fs/ubifs/file.c @@ -72,7 +72,7 @@ static int read_block(struct inode *inode, void *addr, unsigned int block, return err; } - ubifs_assert(dn->ch.sqnum > ubifs_inode(inode)->creat_sqnum); + ubifs_assert(le64_to_cpu(dn->ch.sqnum) > ubifs_inode(inode)->creat_sqnum); len = le32_to_cpu(dn->size); if (len <= 0 || len > UBIFS_BLOCK_SIZE) @@ -147,6 +147,12 @@ static int do_readpage(struct page *page) err = ret; if (err != -ENOENT) break; + } else if (block + 1 == beyond) { + int dlen = le32_to_cpu(dn->size); + int ilen = i_size & (UBIFS_BLOCK_SIZE - 1); + + if (ilen && ilen < dlen) + memset(addr + ilen, 0, dlen - ilen); } } if (++i >= UBIFS_BLOCKS_PER_PAGE) @@ -577,8 +583,297 @@ out: return copied; } +/** + * populate_page - copy data nodes into a page for bulk-read. + * @c: UBIFS file-system description object + * @page: page + * @bu: bulk-read information + * @n: next zbranch slot + * + * This function returns %0 on success and a negative error code on failure. + */ +static int populate_page(struct ubifs_info *c, struct page *page, + struct bu_info *bu, int *n) +{ + int i = 0, nn = *n, offs = bu->zbranch[0].offs, hole = 0, read = 0; + struct inode *inode = page->mapping->host; + loff_t i_size = i_size_read(inode); + unsigned int page_block; + void *addr, *zaddr; + pgoff_t end_index; + + dbg_gen("ino %lu, pg %lu, i_size %lld, flags %#lx", + inode->i_ino, page->index, i_size, page->flags); + + addr = zaddr = kmap(page); + + end_index = (i_size - 1) >> PAGE_CACHE_SHIFT; + if (!i_size || page->index > end_index) { + hole = 1; + memset(addr, 0, PAGE_CACHE_SIZE); + goto out_hole; + } + + page_block = page->index << UBIFS_BLOCKS_PER_PAGE_SHIFT; + while (1) { + int err, len, out_len, dlen; + + if (nn >= bu->cnt) { + hole = 1; + memset(addr, 0, UBIFS_BLOCK_SIZE); + } else if (key_block(c, &bu->zbranch[nn].key) == page_block) { + struct ubifs_data_node *dn; + + dn = bu->buf + (bu->zbranch[nn].offs - offs); + + ubifs_assert(le64_to_cpu(dn->ch.sqnum) > + ubifs_inode(inode)->creat_sqnum); + + len = le32_to_cpu(dn->size); + if (len <= 0 || len > UBIFS_BLOCK_SIZE) + goto out_err; + + dlen = le32_to_cpu(dn->ch.len) - UBIFS_DATA_NODE_SZ; + out_len = UBIFS_BLOCK_SIZE; + err = ubifs_decompress(&dn->data, dlen, addr, &out_len, + le16_to_cpu(dn->compr_type)); + if (err || len != out_len) + goto out_err; + + if (len < UBIFS_BLOCK_SIZE) + memset(addr + len, 0, UBIFS_BLOCK_SIZE - len); + + nn += 1; + read = (i << UBIFS_BLOCK_SHIFT) + len; + } else if (key_block(c, &bu->zbranch[nn].key) < page_block) { + nn += 1; + continue; + } else { + hole = 1; + memset(addr, 0, UBIFS_BLOCK_SIZE); + } + if (++i >= UBIFS_BLOCKS_PER_PAGE) + break; + addr += UBIFS_BLOCK_SIZE; + page_block += 1; + } + + if (end_index == page->index) { + int len = i_size & (PAGE_CACHE_SIZE - 1); + + if (len && len < read) + memset(zaddr + len, 0, read - len); + } + +out_hole: + if (hole) { + SetPageChecked(page); + dbg_gen("hole"); + } + + SetPageUptodate(page); + ClearPageError(page); + flush_dcache_page(page); + kunmap(page); + *n = nn; + return 0; + +out_err: + ClearPageUptodate(page); + SetPageError(page); + flush_dcache_page(page); + kunmap(page); + ubifs_err("bad data node (block %u, inode %lu)", + page_block, inode->i_ino); + return -EINVAL; +} + +/** + * ubifs_do_bulk_read - do bulk-read. + * @c: UBIFS file-system description object + * @bu: bulk-read information + * @page1: first page to read + * + * This function returns %1 if the bulk-read is done, otherwise %0 is returned. + */ +static int ubifs_do_bulk_read(struct ubifs_info *c, struct bu_info *bu, + struct page *page1) +{ + pgoff_t offset = page1->index, end_index; + struct address_space *mapping = page1->mapping; + struct inode *inode = mapping->host; + struct ubifs_inode *ui = ubifs_inode(inode); + int err, page_idx, page_cnt, ret = 0, n = 0; + int allocate = bu->buf ? 0 : 1; + loff_t isize; + + err = ubifs_tnc_get_bu_keys(c, bu); + if (err) + goto out_warn; + + if (bu->eof) { + /* Turn off bulk-read at the end of the file */ + ui->read_in_a_row = 1; + ui->bulk_read = 0; + } + + page_cnt = bu->blk_cnt >> UBIFS_BLOCKS_PER_PAGE_SHIFT; + if (!page_cnt) { + /* + * This happens when there are multiple blocks per page and the + * blocks for the first page we are looking for, are not + * together. If all the pages were like this, bulk-read would + * reduce performance, so we turn it off for a while. + */ + goto out_bu_off; + } + + if (bu->cnt) { + if (allocate) { + /* + * Allocate bulk-read buffer depending on how many data + * nodes we are going to read. + */ + bu->buf_len = bu->zbranch[bu->cnt - 1].offs + + bu->zbranch[bu->cnt - 1].len - + bu->zbranch[0].offs; + ubifs_assert(bu->buf_len > 0); + ubifs_assert(bu->buf_len <= c->leb_size); + bu->buf = kmalloc(bu->buf_len, GFP_NOFS | __GFP_NOWARN); + if (!bu->buf) + goto out_bu_off; + } + + err = ubifs_tnc_bulk_read(c, bu); + if (err) + goto out_warn; + } + + err = populate_page(c, page1, bu, &n); + if (err) + goto out_warn; + + unlock_page(page1); + ret = 1; + + isize = i_size_read(inode); + if (isize == 0) + goto out_free; + end_index = ((isize - 1) >> PAGE_CACHE_SHIFT); + + for (page_idx = 1; page_idx < page_cnt; page_idx++) { + pgoff_t page_offset = offset + page_idx; + struct page *page; + + if (page_offset > end_index) + break; + page = find_or_create_page(mapping, page_offset, + GFP_NOFS | __GFP_COLD); + if (!page) + break; + if (!PageUptodate(page)) + err = populate_page(c, page, bu, &n); + unlock_page(page); + page_cache_release(page); + if (err) + break; + } + + ui->last_page_read = offset + page_idx - 1; + +out_free: + if (allocate) + kfree(bu->buf); + return ret; + +out_warn: + ubifs_warn("ignoring error %d and skipping bulk-read", err); + goto out_free; + +out_bu_off: + ui->read_in_a_row = ui->bulk_read = 0; + goto out_free; +} + +/** + * ubifs_bulk_read - determine whether to bulk-read and, if so, do it. + * @page: page from which to start bulk-read. + * + * Some flash media are capable of reading sequentially at faster rates. UBIFS + * bulk-read facility is designed to take advantage of that, by reading in one + * go consecutive data nodes that are also located consecutively in the same + * LEB. This function returns %1 if a bulk-read is done and %0 otherwise. + */ +static int ubifs_bulk_read(struct page *page) +{ + struct inode *inode = page->mapping->host; + struct ubifs_info *c = inode->i_sb->s_fs_info; + struct ubifs_inode *ui = ubifs_inode(inode); + pgoff_t index = page->index, last_page_read = ui->last_page_read; + struct bu_info *bu; + int err = 0, allocated = 0; + + ui->last_page_read = index; + if (!c->bulk_read) + return 0; + + /* + * Bulk-read is protected by @ui->ui_mutex, but it is an optimization, + * so don't bother if we cannot lock the mutex. + */ + if (!mutex_trylock(&ui->ui_mutex)) + return 0; + + if (index != last_page_read + 1) { + /* Turn off bulk-read if we stop reading sequentially */ + ui->read_in_a_row = 1; + if (ui->bulk_read) + ui->bulk_read = 0; + goto out_unlock; + } + + if (!ui->bulk_read) { + ui->read_in_a_row += 1; + if (ui->read_in_a_row < 3) + goto out_unlock; + /* Three reads in a row, so switch on bulk-read */ + ui->bulk_read = 1; + } + + /* + * If possible, try to use pre-allocated bulk-read information, which + * is protected by @c->bu_mutex. + */ + if (mutex_trylock(&c->bu_mutex)) + bu = &c->bu; + else { + bu = kmalloc(sizeof(struct bu_info), GFP_NOFS | __GFP_NOWARN); + if (!bu) + goto out_unlock; + + bu->buf = NULL; + allocated = 1; + } + + bu->buf_len = c->max_bu_buf_len; + data_key_init(c, &bu->key, inode->i_ino, + page->index << UBIFS_BLOCKS_PER_PAGE_SHIFT); + err = ubifs_do_bulk_read(c, bu, page); + + if (!allocated) + mutex_unlock(&c->bu_mutex); + else + kfree(bu); + +out_unlock: + mutex_unlock(&ui->ui_mutex); + return err; +} + static int ubifs_readpage(struct file *file, struct page *page) { + if (ubifs_bulk_read(page)) + return 0; do_readpage(page); unlock_page(page); return 0; diff --git a/fs/ubifs/find.c b/fs/ubifs/find.c index 47814cde240..717d79c97c5 100644 --- a/fs/ubifs/find.c +++ b/fs/ubifs/find.c @@ -901,11 +901,11 @@ static int get_idx_gc_leb(struct ubifs_info *c) * it is needed now for this commit. */ lp = ubifs_lpt_lookup_dirty(c, lnum); - if (unlikely(IS_ERR(lp))) + if (IS_ERR(lp)) return PTR_ERR(lp); lp = ubifs_change_lp(c, lp, LPROPS_NC, LPROPS_NC, lp->flags | LPROPS_INDEX, -1); - if (unlikely(IS_ERR(lp))) + if (IS_ERR(lp)) return PTR_ERR(lp); dbg_find("LEB %d, dirty %d and free %d flags %#x", lp->lnum, lp->dirty, lp->free, lp->flags); diff --git a/fs/ubifs/gc.c b/fs/ubifs/gc.c index 02aba36fe3d..0bef6501d58 100644 --- a/fs/ubifs/gc.c +++ b/fs/ubifs/gc.c @@ -96,6 +96,48 @@ static int switch_gc_head(struct ubifs_info *c) } /** + * joinup - bring data nodes for an inode together. + * @c: UBIFS file-system description object + * @sleb: describes scanned LEB + * @inum: inode number + * @blk: block number + * @data: list to which to add data nodes + * + * This function looks at the first few nodes in the scanned LEB @sleb and adds + * them to @data if they are data nodes from @inum and have a larger block + * number than @blk. This function returns %0 on success and a negative error + * code on failure. + */ +static int joinup(struct ubifs_info *c, struct ubifs_scan_leb *sleb, ino_t inum, + unsigned int blk, struct list_head *data) +{ + int err, cnt = 6, lnum = sleb->lnum, offs; + struct ubifs_scan_node *snod, *tmp; + union ubifs_key *key; + + list_for_each_entry_safe(snod, tmp, &sleb->nodes, list) { + key = &snod->key; + if (key_inum(c, key) == inum && + key_type(c, key) == UBIFS_DATA_KEY && + key_block(c, key) > blk) { + offs = snod->offs; + err = ubifs_tnc_has_node(c, key, 0, lnum, offs, 0); + if (err < 0) + return err; + list_del(&snod->list); + if (err) { + list_add_tail(&snod->list, data); + blk = key_block(c, key); + } else + kfree(snod); + cnt = 6; + } else if (--cnt == 0) + break; + } + return 0; +} + +/** * move_nodes - move nodes. * @c: UBIFS file-system description object * @sleb: describes nodes to move @@ -116,16 +158,21 @@ static int switch_gc_head(struct ubifs_info *c) static int move_nodes(struct ubifs_info *c, struct ubifs_scan_leb *sleb) { struct ubifs_scan_node *snod, *tmp; - struct list_head large, medium, small; + struct list_head data, large, medium, small; struct ubifs_wbuf *wbuf = &c->jheads[GCHD].wbuf; int avail, err, min = INT_MAX; + unsigned int blk = 0; + ino_t inum = 0; + INIT_LIST_HEAD(&data); INIT_LIST_HEAD(&large); INIT_LIST_HEAD(&medium); INIT_LIST_HEAD(&small); - list_for_each_entry_safe(snod, tmp, &sleb->nodes, list) { - struct list_head *lst; + while (!list_empty(&sleb->nodes)) { + struct list_head *lst = sleb->nodes.next; + + snod = list_entry(lst, struct ubifs_scan_node, list); ubifs_assert(snod->type != UBIFS_IDX_NODE); ubifs_assert(snod->type != UBIFS_REF_NODE); @@ -136,7 +183,6 @@ static int move_nodes(struct ubifs_info *c, struct ubifs_scan_leb *sleb) if (err < 0) goto out; - lst = &snod->list; list_del(lst); if (!err) { /* The node is obsolete, remove it from the list */ @@ -145,15 +191,30 @@ static int move_nodes(struct ubifs_info *c, struct ubifs_scan_leb *sleb) } /* - * Sort the list of nodes so that large nodes go first, and - * small nodes go last. + * Sort the list of nodes so that data nodes go first, large + * nodes go second, and small nodes go last. */ - if (snod->len > MEDIUM_NODE_WM) - list_add(lst, &large); + if (key_type(c, &snod->key) == UBIFS_DATA_KEY) { + if (inum != key_inum(c, &snod->key)) { + if (inum) { + /* + * Try to move data nodes from the same + * inode together. + */ + err = joinup(c, sleb, inum, blk, &data); + if (err) + goto out; + } + inum = key_inum(c, &snod->key); + blk = key_block(c, &snod->key); + } + list_add_tail(lst, &data); + } else if (snod->len > MEDIUM_NODE_WM) + list_add_tail(lst, &large); else if (snod->len > SMALL_NODE_WM) - list_add(lst, &medium); + list_add_tail(lst, &medium); else - list_add(lst, &small); + list_add_tail(lst, &small); /* And find the smallest node */ if (snod->len < min) @@ -164,6 +225,7 @@ static int move_nodes(struct ubifs_info *c, struct ubifs_scan_leb *sleb) * Join the tree lists so that we'd have one roughly sorted list * ('large' will be the head of the joined list). */ + list_splice(&data, &large); list_splice(&medium, large.prev); list_splice(&small, large.prev); @@ -653,7 +715,7 @@ int ubifs_gc_start_commit(struct ubifs_info *c) */ while (1) { lp = ubifs_fast_find_freeable(c); - if (unlikely(IS_ERR(lp))) { + if (IS_ERR(lp)) { err = PTR_ERR(lp); goto out; } @@ -665,7 +727,7 @@ int ubifs_gc_start_commit(struct ubifs_info *c) if (err) goto out; lp = ubifs_change_lp(c, lp, c->leb_size, 0, lp->flags, 0); - if (unlikely(IS_ERR(lp))) { + if (IS_ERR(lp)) { err = PTR_ERR(lp); goto out; } @@ -680,7 +742,7 @@ int ubifs_gc_start_commit(struct ubifs_info *c) /* Record index freeable LEBs for unmapping after commit */ while (1) { lp = ubifs_fast_find_frdi_idx(c); - if (unlikely(IS_ERR(lp))) { + if (IS_ERR(lp)) { err = PTR_ERR(lp); goto out; } @@ -696,7 +758,7 @@ int ubifs_gc_start_commit(struct ubifs_info *c) /* Don't release the LEB until after the next commit */ flags = (lp->flags | LPROPS_TAKEN) ^ LPROPS_INDEX; lp = ubifs_change_lp(c, lp, c->leb_size, 0, flags, 1); - if (unlikely(IS_ERR(lp))) { + if (IS_ERR(lp)) { err = PTR_ERR(lp); kfree(idx_gc); goto out; diff --git a/fs/ubifs/io.c b/fs/ubifs/io.c index 054363f2b20..01682713af6 100644 --- a/fs/ubifs/io.c +++ b/fs/ubifs/io.c @@ -62,6 +62,7 @@ void ubifs_ro_mode(struct ubifs_info *c, int err) { if (!c->ro_media) { c->ro_media = 1; + c->no_chk_data_crc = 0; ubifs_warn("switched to read-only mode, error %d", err); dbg_dump_stack(); } @@ -74,6 +75,7 @@ void ubifs_ro_mode(struct ubifs_info *c, int err) * @lnum: logical eraseblock number * @offs: offset within the logical eraseblock * @quiet: print no messages + * @chk_crc: indicates whether to always check the CRC * * This function checks node magic number and CRC checksum. This function also * validates node length to prevent UBIFS from becoming crazy when an attacker @@ -85,7 +87,7 @@ void ubifs_ro_mode(struct ubifs_info *c, int err) * or magic. */ int ubifs_check_node(const struct ubifs_info *c, const void *buf, int lnum, - int offs, int quiet) + int offs, int quiet, int chk_crc) { int err = -EINVAL, type, node_len; uint32_t crc, node_crc, magic; @@ -121,6 +123,10 @@ int ubifs_check_node(const struct ubifs_info *c, const void *buf, int lnum, node_len > c->ranges[type].max_len) goto out_len; + if (!chk_crc && type == UBIFS_DATA_NODE && !c->always_chk_crc) + if (c->no_chk_data_crc) + return 0; + crc = crc32(UBIFS_CRC32_INIT, buf + 8, node_len - 8); node_crc = le32_to_cpu(ch->crc); if (crc != node_crc) { @@ -722,7 +728,7 @@ int ubifs_read_node_wbuf(struct ubifs_wbuf *wbuf, void *buf, int type, int len, goto out; } - err = ubifs_check_node(c, buf, lnum, offs, 0); + err = ubifs_check_node(c, buf, lnum, offs, 0, 0); if (err) { ubifs_err("expected node type %d", type); return err; @@ -781,7 +787,7 @@ int ubifs_read_node(const struct ubifs_info *c, void *buf, int type, int len, goto out; } - err = ubifs_check_node(c, buf, lnum, offs, 0); + err = ubifs_check_node(c, buf, lnum, offs, 0, 0); if (err) { ubifs_err("expected node type %d", type); return err; diff --git a/fs/ubifs/journal.c b/fs/ubifs/journal.c index 22993f867d1..f91b745908e 100644 --- a/fs/ubifs/journal.c +++ b/fs/ubifs/journal.c @@ -690,8 +690,9 @@ int ubifs_jnl_write_data(struct ubifs_info *c, const struct inode *inode, int dlen = UBIFS_DATA_NODE_SZ + UBIFS_BLOCK_SIZE * WORST_COMPR_FACTOR; struct ubifs_inode *ui = ubifs_inode(inode); - dbg_jnl("ino %lu, blk %u, len %d, key %s", key_inum(c, key), - key_block(c, key), len, DBGKEY(key)); + dbg_jnl("ino %lu, blk %u, len %d, key %s", + (unsigned long)key_inum(c, key), key_block(c, key), len, + DBGKEY(key)); ubifs_assert(len <= UBIFS_BLOCK_SIZE); data = kmalloc(dlen, GFP_NOFS); @@ -1128,7 +1129,8 @@ int ubifs_jnl_truncate(struct ubifs_info *c, const struct inode *inode, ino_t inum = inode->i_ino; unsigned int blk; - dbg_jnl("ino %lu, size %lld -> %lld", inum, old_size, new_size); + dbg_jnl("ino %lu, size %lld -> %lld", + (unsigned long)inum, old_size, new_size); ubifs_assert(!ui->data_len); ubifs_assert(S_ISREG(inode->i_mode)); ubifs_assert(mutex_is_locked(&ui->ui_mutex)); diff --git a/fs/ubifs/key.h b/fs/ubifs/key.h index 8f747600754..3f1f16bc25c 100644 --- a/fs/ubifs/key.h +++ b/fs/ubifs/key.h @@ -345,7 +345,7 @@ static inline int key_type_flash(const struct ubifs_info *c, const void *k) { const union ubifs_key *key = k; - return le32_to_cpu(key->u32[1]) >> UBIFS_S_KEY_BLOCK_BITS; + return le32_to_cpu(key->j32[1]) >> UBIFS_S_KEY_BLOCK_BITS; } /** @@ -416,7 +416,7 @@ static inline unsigned int key_block_flash(const struct ubifs_info *c, { const union ubifs_key *key = k; - return le32_to_cpu(key->u32[1]) & UBIFS_S_KEY_BLOCK_MASK; + return le32_to_cpu(key->j32[1]) & UBIFS_S_KEY_BLOCK_MASK; } /** @@ -484,7 +484,7 @@ static inline void key_copy(const struct ubifs_info *c, * @key2: the second key to compare * * This function compares 2 keys and returns %-1 if @key1 is less than - * @key2, 0 if the keys are equivalent and %1 if @key1 is greater than @key2. + * @key2, %0 if the keys are equivalent and %1 if @key1 is greater than @key2. */ static inline int keys_cmp(const struct ubifs_info *c, const union ubifs_key *key1, @@ -503,6 +503,26 @@ static inline int keys_cmp(const struct ubifs_info *c, } /** + * keys_eq - determine if keys are equivalent. + * @c: UBIFS file-system description object + * @key1: the first key to compare + * @key2: the second key to compare + * + * This function compares 2 keys and returns %1 if @key1 is equal to @key2 and + * %0 if not. + */ +static inline int keys_eq(const struct ubifs_info *c, + const union ubifs_key *key1, + const union ubifs_key *key2) +{ + if (key1->u32[0] != key2->u32[0]) + return 0; + if (key1->u32[1] != key2->u32[1]) + return 0; + return 1; +} + +/** * is_hash_key - is a key vulnerable to hash collisions. * @c: UBIFS file-system description object * @key: key diff --git a/fs/ubifs/lprops.c b/fs/ubifs/lprops.c index 2ba93da71b6..f27176e9b70 100644 --- a/fs/ubifs/lprops.c +++ b/fs/ubifs/lprops.c @@ -125,6 +125,7 @@ static void adjust_lpt_heap(struct ubifs_info *c, struct ubifs_lpt_heap *heap, } } } + /* Not greater than parent, so compare to children */ while (1) { /* Compare to left child */ @@ -460,18 +461,6 @@ static void change_category(struct ubifs_info *c, struct ubifs_lprops *lprops) } /** - * ubifs_get_lprops - get reference to LEB properties. - * @c: the UBIFS file-system description object - * - * This function locks lprops. Lprops have to be unlocked by - * 'ubifs_release_lprops()'. - */ -void ubifs_get_lprops(struct ubifs_info *c) -{ - mutex_lock(&c->lp_mutex); -} - -/** * calc_dark - calculate LEB dark space size. * @c: the UBIFS file-system description object * @spc: amount of free and dirty space in the LEB @@ -576,7 +565,6 @@ const struct ubifs_lprops *ubifs_change_lp(struct ubifs_info *c, ubifs_assert(!(lprops->free & 7) && !(lprops->dirty & 7)); spin_lock(&c->space_lock); - if ((lprops->flags & LPROPS_TAKEN) && lprops->free == c->leb_size) c->lst.taken_empty_lebs -= 1; @@ -637,31 +625,12 @@ const struct ubifs_lprops *ubifs_change_lp(struct ubifs_info *c, c->lst.taken_empty_lebs += 1; change_category(c, lprops); - c->idx_gc_cnt += idx_gc_cnt; - spin_unlock(&c->space_lock); - return lprops; } /** - * ubifs_release_lprops - release lprops lock. - * @c: the UBIFS file-system description object - * - * This function has to be called after each 'ubifs_get_lprops()' call to - * unlock lprops. - */ -void ubifs_release_lprops(struct ubifs_info *c) -{ - ubifs_assert(mutex_is_locked(&c->lp_mutex)); - ubifs_assert(c->lst.empty_lebs >= 0 && - c->lst.empty_lebs <= c->main_lebs); - - mutex_unlock(&c->lp_mutex); -} - -/** * ubifs_get_lp_stats - get lprops statistics. * @c: UBIFS file-system description object * @st: return statistics @@ -1262,7 +1231,6 @@ static int scan_check_cb(struct ubifs_info *c, } ubifs_scan_destroy(sleb); - return LPT_SCAN_CONTINUE; out_print: diff --git a/fs/ubifs/lpt.c b/fs/ubifs/lpt.c index 9ff2463177e..db8bd0e518b 100644 --- a/fs/ubifs/lpt.c +++ b/fs/ubifs/lpt.c @@ -109,7 +109,8 @@ static void do_calc_lpt_geom(struct ubifs_info *c) c->lpt_sz = (long long)c->pnode_cnt * c->pnode_sz; c->lpt_sz += (long long)c->nnode_cnt * c->nnode_sz; c->lpt_sz += c->ltab_sz; - c->lpt_sz += c->lsave_sz; + if (c->big_lpt) + c->lpt_sz += c->lsave_sz; /* Add wastage */ sz = c->lpt_sz; @@ -287,25 +288,56 @@ uint32_t ubifs_unpack_bits(uint8_t **addr, int *pos, int nrbits) const int k = 32 - nrbits; uint8_t *p = *addr; int b = *pos; - uint32_t val; + uint32_t uninitialized_var(val); + const int bytes = (nrbits + b + 7) >> 3; ubifs_assert(nrbits > 0); ubifs_assert(nrbits <= 32); ubifs_assert(*pos >= 0); ubifs_assert(*pos < 8); if (b) { - val = p[1] | ((uint32_t)p[2] << 8) | ((uint32_t)p[3] << 16) | - ((uint32_t)p[4] << 24); + switch (bytes) { + case 2: + val = p[1]; + break; + case 3: + val = p[1] | ((uint32_t)p[2] << 8); + break; + case 4: + val = p[1] | ((uint32_t)p[2] << 8) | + ((uint32_t)p[3] << 16); + break; + case 5: + val = p[1] | ((uint32_t)p[2] << 8) | + ((uint32_t)p[3] << 16) | + ((uint32_t)p[4] << 24); + } val <<= (8 - b); val |= *p >> b; nrbits += b; - } else - val = p[0] | ((uint32_t)p[1] << 8) | ((uint32_t)p[2] << 16) | - ((uint32_t)p[3] << 24); + } else { + switch (bytes) { + case 1: + val = p[0]; + break; + case 2: + val = p[0] | ((uint32_t)p[1] << 8); + break; + case 3: + val = p[0] | ((uint32_t)p[1] << 8) | + ((uint32_t)p[2] << 16); + break; + case 4: + val = p[0] | ((uint32_t)p[1] << 8) | + ((uint32_t)p[2] << 16) | + ((uint32_t)p[3] << 24); + break; + } + } val <<= k; val >>= k; b = nrbits & 7; - p += nrbits / 8; + p += nrbits >> 3; *addr = p; *pos = b; ubifs_assert((val >> nrbits) == 0 || nrbits - b == 32); diff --git a/fs/ubifs/lpt_commit.c b/fs/ubifs/lpt_commit.c index 5f0b83e20af..a41434b4278 100644 --- a/fs/ubifs/lpt_commit.c +++ b/fs/ubifs/lpt_commit.c @@ -177,8 +177,6 @@ static int alloc_lpt_leb(struct ubifs_info *c, int *lnum) return 0; } } - dbg_err("last LEB %d", *lnum); - dump_stack(); return -ENOSPC; } @@ -193,6 +191,9 @@ static int layout_cnodes(struct ubifs_info *c) int lnum, offs, len, alen, done_lsave, done_ltab, err; struct ubifs_cnode *cnode; + err = dbg_chk_lpt_sz(c, 0, 0); + if (err) + return err; cnode = c->lpt_cnext; if (!cnode) return 0; @@ -206,6 +207,7 @@ static int layout_cnodes(struct ubifs_info *c) c->lsave_lnum = lnum; c->lsave_offs = offs; offs += c->lsave_sz; + dbg_chk_lpt_sz(c, 1, c->lsave_sz); } if (offs + c->ltab_sz <= c->leb_size) { @@ -213,6 +215,7 @@ static int layout_cnodes(struct ubifs_info *c) c->ltab_lnum = lnum; c->ltab_offs = offs; offs += c->ltab_sz; + dbg_chk_lpt_sz(c, 1, c->ltab_sz); } do { @@ -226,9 +229,10 @@ static int layout_cnodes(struct ubifs_info *c) while (offs + len > c->leb_size) { alen = ALIGN(offs, c->min_io_size); upd_ltab(c, lnum, c->leb_size - alen, alen - offs); + dbg_chk_lpt_sz(c, 2, alen - offs); err = alloc_lpt_leb(c, &lnum); if (err) - return err; + goto no_space; offs = 0; ubifs_assert(lnum >= c->lpt_first && lnum <= c->lpt_last); @@ -238,6 +242,7 @@ static int layout_cnodes(struct ubifs_info *c) c->lsave_lnum = lnum; c->lsave_offs = offs; offs += c->lsave_sz; + dbg_chk_lpt_sz(c, 1, c->lsave_sz); continue; } if (!done_ltab) { @@ -245,6 +250,7 @@ static int layout_cnodes(struct ubifs_info *c) c->ltab_lnum = lnum; c->ltab_offs = offs; offs += c->ltab_sz; + dbg_chk_lpt_sz(c, 1, c->ltab_sz); continue; } break; @@ -257,6 +263,7 @@ static int layout_cnodes(struct ubifs_info *c) c->lpt_offs = offs; } offs += len; + dbg_chk_lpt_sz(c, 1, len); cnode = cnode->cnext; } while (cnode && cnode != c->lpt_cnext); @@ -265,9 +272,10 @@ static int layout_cnodes(struct ubifs_info *c) if (offs + c->lsave_sz > c->leb_size) { alen = ALIGN(offs, c->min_io_size); upd_ltab(c, lnum, c->leb_size - alen, alen - offs); + dbg_chk_lpt_sz(c, 2, alen - offs); err = alloc_lpt_leb(c, &lnum); if (err) - return err; + goto no_space; offs = 0; ubifs_assert(lnum >= c->lpt_first && lnum <= c->lpt_last); @@ -276,6 +284,7 @@ static int layout_cnodes(struct ubifs_info *c) c->lsave_lnum = lnum; c->lsave_offs = offs; offs += c->lsave_sz; + dbg_chk_lpt_sz(c, 1, c->lsave_sz); } /* Make sure to place LPT's own lprops table */ @@ -283,9 +292,10 @@ static int layout_cnodes(struct ubifs_info *c) if (offs + c->ltab_sz > c->leb_size) { alen = ALIGN(offs, c->min_io_size); upd_ltab(c, lnum, c->leb_size - alen, alen - offs); + dbg_chk_lpt_sz(c, 2, alen - offs); err = alloc_lpt_leb(c, &lnum); if (err) - return err; + goto no_space; offs = 0; ubifs_assert(lnum >= c->lpt_first && lnum <= c->lpt_last); @@ -294,11 +304,23 @@ static int layout_cnodes(struct ubifs_info *c) c->ltab_lnum = lnum; c->ltab_offs = offs; offs += c->ltab_sz; + dbg_chk_lpt_sz(c, 1, c->ltab_sz); } alen = ALIGN(offs, c->min_io_size); upd_ltab(c, lnum, c->leb_size - alen, alen - offs); + dbg_chk_lpt_sz(c, 4, alen - offs); + err = dbg_chk_lpt_sz(c, 3, alen); + if (err) + return err; return 0; + +no_space: + ubifs_err("LPT out of space"); + dbg_err("LPT out of space at LEB %d:%d needing %d, done_ltab %d, " + "done_lsave %d", lnum, offs, len, done_ltab, done_lsave); + dbg_dump_lpt_info(c); + return err; } /** @@ -333,8 +355,6 @@ static int realloc_lpt_leb(struct ubifs_info *c, int *lnum) *lnum = i + c->lpt_first; return 0; } - dbg_err("last LEB %d", *lnum); - dump_stack(); return -ENOSPC; } @@ -369,12 +389,14 @@ static int write_cnodes(struct ubifs_info *c) done_lsave = 1; ubifs_pack_lsave(c, buf + offs, c->lsave); offs += c->lsave_sz; + dbg_chk_lpt_sz(c, 1, c->lsave_sz); } if (offs + c->ltab_sz <= c->leb_size) { done_ltab = 1; ubifs_pack_ltab(c, buf + offs, c->ltab_cmt); offs += c->ltab_sz; + dbg_chk_lpt_sz(c, 1, c->ltab_sz); } /* Loop for each cnode */ @@ -392,10 +414,12 @@ static int write_cnodes(struct ubifs_info *c) alen, UBI_SHORTTERM); if (err) return err; + dbg_chk_lpt_sz(c, 4, alen - wlen); } + dbg_chk_lpt_sz(c, 2, 0); err = realloc_lpt_leb(c, &lnum); if (err) - return err; + goto no_space; offs = 0; from = 0; ubifs_assert(lnum >= c->lpt_first && @@ -408,12 +432,14 @@ static int write_cnodes(struct ubifs_info *c) done_lsave = 1; ubifs_pack_lsave(c, buf + offs, c->lsave); offs += c->lsave_sz; + dbg_chk_lpt_sz(c, 1, c->lsave_sz); continue; } if (!done_ltab) { done_ltab = 1; ubifs_pack_ltab(c, buf + offs, c->ltab_cmt); offs += c->ltab_sz; + dbg_chk_lpt_sz(c, 1, c->ltab_sz); continue; } break; @@ -435,6 +461,7 @@ static int write_cnodes(struct ubifs_info *c) clear_bit(COW_ZNODE, &cnode->flags); smp_mb__after_clear_bit(); offs += len; + dbg_chk_lpt_sz(c, 1, len); cnode = cnode->cnext; } while (cnode && cnode != c->lpt_cnext); @@ -448,9 +475,10 @@ static int write_cnodes(struct ubifs_info *c) UBI_SHORTTERM); if (err) return err; + dbg_chk_lpt_sz(c, 2, alen - wlen); err = realloc_lpt_leb(c, &lnum); if (err) - return err; + goto no_space; offs = 0; ubifs_assert(lnum >= c->lpt_first && lnum <= c->lpt_last); @@ -461,6 +489,7 @@ static int write_cnodes(struct ubifs_info *c) done_lsave = 1; ubifs_pack_lsave(c, buf + offs, c->lsave); offs += c->lsave_sz; + dbg_chk_lpt_sz(c, 1, c->lsave_sz); } /* Make sure to place LPT's own lprops table */ @@ -473,9 +502,10 @@ static int write_cnodes(struct ubifs_info *c) UBI_SHORTTERM); if (err) return err; + dbg_chk_lpt_sz(c, 2, alen - wlen); err = realloc_lpt_leb(c, &lnum); if (err) - return err; + goto no_space; offs = 0; ubifs_assert(lnum >= c->lpt_first && lnum <= c->lpt_last); @@ -486,6 +516,7 @@ static int write_cnodes(struct ubifs_info *c) done_ltab = 1; ubifs_pack_ltab(c, buf + offs, c->ltab_cmt); offs += c->ltab_sz; + dbg_chk_lpt_sz(c, 1, c->ltab_sz); } /* Write remaining data in buffer */ @@ -495,6 +526,12 @@ static int write_cnodes(struct ubifs_info *c) err = ubifs_leb_write(c, lnum, buf + from, from, alen, UBI_SHORTTERM); if (err) return err; + + dbg_chk_lpt_sz(c, 4, alen - wlen); + err = dbg_chk_lpt_sz(c, 3, ALIGN(offs, c->min_io_size)); + if (err) + return err; + c->nhead_lnum = lnum; c->nhead_offs = ALIGN(offs, c->min_io_size); @@ -503,7 +540,15 @@ static int write_cnodes(struct ubifs_info *c) dbg_lp("LPT ltab is at %d:%d", c->ltab_lnum, c->ltab_offs); if (c->big_lpt) dbg_lp("LPT lsave is at %d:%d", c->lsave_lnum, c->lsave_offs); + return 0; + +no_space: + ubifs_err("LPT out of space mismatch"); + dbg_err("LPT out of space mismatch at LEB %d:%d needing %d, done_ltab " + "%d, done_lsave %d", lnum, offs, len, done_ltab, done_lsave); + dbg_dump_lpt_info(c); + return err; } /** @@ -526,8 +571,6 @@ static struct ubifs_pnode *next_pnode(struct ubifs_info *c, /* We assume here that LEB zero is never an LPT LEB */ if (nnode->nbranch[iip].lnum) return ubifs_get_pnode(c, nnode, iip); - else - return NULL; } /* Go up while can't go right */ @@ -1044,6 +1087,8 @@ static int is_a_node(struct ubifs_info *c, uint8_t *buf, int len) int pos = 0, node_type, node_len; uint16_t crc, calc_crc; + if (len < UBIFS_LPT_CRC_BYTES + (UBIFS_LPT_TYPE_BITS + 7) / 8) + return 0; node_type = ubifs_unpack_bits(&addr, &pos, UBIFS_LPT_TYPE_BITS); if (node_type == UBIFS_LPT_NOT_A_NODE) return 0; @@ -1156,6 +1201,9 @@ int ubifs_lpt_start_commit(struct ubifs_info *c) dbg_lp(""); mutex_lock(&c->lp_mutex); + err = dbg_chk_lpt_free_spc(c); + if (err) + goto out; err = dbg_check_ltab(c); if (err) goto out; @@ -1645,4 +1693,121 @@ int dbg_check_ltab(struct ubifs_info *c) return 0; } +/** + * dbg_chk_lpt_free_spc - check LPT free space is enough to write entire LPT. + * @c: the UBIFS file-system description object + * + * This function returns %0 on success and a negative error code on failure. + */ +int dbg_chk_lpt_free_spc(struct ubifs_info *c) +{ + long long free = 0; + int i; + + for (i = 0; i < c->lpt_lebs; i++) { + if (c->ltab[i].tgc || c->ltab[i].cmt) + continue; + if (i + c->lpt_first == c->nhead_lnum) + free += c->leb_size - c->nhead_offs; + else if (c->ltab[i].free == c->leb_size) + free += c->leb_size; + } + if (free < c->lpt_sz) { + dbg_err("LPT space error: free %lld lpt_sz %lld", + free, c->lpt_sz); + dbg_dump_lpt_info(c); + return -EINVAL; + } + return 0; +} + +/** + * dbg_chk_lpt_sz - check LPT does not write more than LPT size. + * @c: the UBIFS file-system description object + * @action: action + * @len: length written + * + * This function returns %0 on success and a negative error code on failure. + */ +int dbg_chk_lpt_sz(struct ubifs_info *c, int action, int len) +{ + long long chk_lpt_sz, lpt_sz; + int err = 0; + + switch (action) { + case 0: + c->chk_lpt_sz = 0; + c->chk_lpt_sz2 = 0; + c->chk_lpt_lebs = 0; + c->chk_lpt_wastage = 0; + if (c->dirty_pn_cnt > c->pnode_cnt) { + dbg_err("dirty pnodes %d exceed max %d", + c->dirty_pn_cnt, c->pnode_cnt); + err = -EINVAL; + } + if (c->dirty_nn_cnt > c->nnode_cnt) { + dbg_err("dirty nnodes %d exceed max %d", + c->dirty_nn_cnt, c->nnode_cnt); + err = -EINVAL; + } + return err; + case 1: + c->chk_lpt_sz += len; + return 0; + case 2: + c->chk_lpt_sz += len; + c->chk_lpt_wastage += len; + c->chk_lpt_lebs += 1; + return 0; + case 3: + chk_lpt_sz = c->leb_size; + chk_lpt_sz *= c->chk_lpt_lebs; + chk_lpt_sz += len - c->nhead_offs; + if (c->chk_lpt_sz != chk_lpt_sz) { + dbg_err("LPT wrote %lld but space used was %lld", + c->chk_lpt_sz, chk_lpt_sz); + err = -EINVAL; + } + if (c->chk_lpt_sz > c->lpt_sz) { + dbg_err("LPT wrote %lld but lpt_sz is %lld", + c->chk_lpt_sz, c->lpt_sz); + err = -EINVAL; + } + if (c->chk_lpt_sz2 && c->chk_lpt_sz != c->chk_lpt_sz2) { + dbg_err("LPT layout size %lld but wrote %lld", + c->chk_lpt_sz, c->chk_lpt_sz2); + err = -EINVAL; + } + if (c->chk_lpt_sz2 && c->new_nhead_offs != len) { + dbg_err("LPT new nhead offs: expected %d was %d", + c->new_nhead_offs, len); + err = -EINVAL; + } + lpt_sz = (long long)c->pnode_cnt * c->pnode_sz; + lpt_sz += (long long)c->nnode_cnt * c->nnode_sz; + lpt_sz += c->ltab_sz; + if (c->big_lpt) + lpt_sz += c->lsave_sz; + if (c->chk_lpt_sz - c->chk_lpt_wastage > lpt_sz) { + dbg_err("LPT chk_lpt_sz %lld + waste %lld exceeds %lld", + c->chk_lpt_sz, c->chk_lpt_wastage, lpt_sz); + err = -EINVAL; + } + if (err) + dbg_dump_lpt_info(c); + c->chk_lpt_sz2 = c->chk_lpt_sz; + c->chk_lpt_sz = 0; + c->chk_lpt_wastage = 0; + c->chk_lpt_lebs = 0; + c->new_nhead_offs = len; + return err; + case 4: + c->chk_lpt_sz += len; + c->chk_lpt_wastage += len; + return 0; + default: + return -EINVAL; + } +} + #endif /* CONFIG_UBIFS_FS_DEBUG */ diff --git a/fs/ubifs/misc.h b/fs/ubifs/misc.h index 4c12a9215d7..4fa81d867e4 100644 --- a/fs/ubifs/misc.h +++ b/fs/ubifs/misc.h @@ -310,4 +310,31 @@ static inline int ubifs_tnc_lookup(struct ubifs_info *c, return ubifs_tnc_locate(c, key, node, NULL, NULL); } +/** + * ubifs_get_lprops - get reference to LEB properties. + * @c: the UBIFS file-system description object + * + * This function locks lprops. Lprops have to be unlocked by + * 'ubifs_release_lprops()'. + */ +static inline void ubifs_get_lprops(struct ubifs_info *c) +{ + mutex_lock(&c->lp_mutex); +} + +/** + * ubifs_release_lprops - release lprops lock. + * @c: the UBIFS file-system description object + * + * This function has to be called after each 'ubifs_get_lprops()' call to + * unlock lprops. + */ +static inline void ubifs_release_lprops(struct ubifs_info *c) +{ + ubifs_assert(mutex_is_locked(&c->lp_mutex)); + ubifs_assert(c->lst.empty_lebs >= 0 && + c->lst.empty_lebs <= c->main_lebs); + mutex_unlock(&c->lp_mutex); +} + #endif /* __UBIFS_MISC_H__ */ diff --git a/fs/ubifs/orphan.c b/fs/ubifs/orphan.c index 02d3462f4d3..9bd5a43d452 100644 --- a/fs/ubifs/orphan.c +++ b/fs/ubifs/orphan.c @@ -105,7 +105,7 @@ int ubifs_add_orphan(struct ubifs_info *c, ino_t inum) list_add_tail(&orphan->list, &c->orph_list); list_add_tail(&orphan->new_list, &c->orph_new); spin_unlock(&c->orphan_lock); - dbg_gen("ino %lu", inum); + dbg_gen("ino %lu", (unsigned long)inum); return 0; } @@ -132,14 +132,16 @@ void ubifs_delete_orphan(struct ubifs_info *c, ino_t inum) else { if (o->dnext) { spin_unlock(&c->orphan_lock); - dbg_gen("deleted twice ino %lu", inum); + dbg_gen("deleted twice ino %lu", + (unsigned long)inum); return; } if (o->cnext) { o->dnext = c->orph_dnext; c->orph_dnext = o; spin_unlock(&c->orphan_lock); - dbg_gen("delete later ino %lu", inum); + dbg_gen("delete later ino %lu", + (unsigned long)inum); return; } rb_erase(p, &c->orph_tree); @@ -151,12 +153,12 @@ void ubifs_delete_orphan(struct ubifs_info *c, ino_t inum) } spin_unlock(&c->orphan_lock); kfree(o); - dbg_gen("inum %lu", inum); + dbg_gen("inum %lu", (unsigned long)inum); return; } } spin_unlock(&c->orphan_lock); - dbg_err("missing orphan ino %lu", inum); + dbg_err("missing orphan ino %lu", (unsigned long)inum); dbg_dump_stack(); } @@ -448,7 +450,7 @@ static void erase_deleted(struct ubifs_info *c) rb_erase(&orphan->rb, &c->orph_tree); list_del(&orphan->list); c->tot_orphans -= 1; - dbg_gen("deleting orphan ino %lu", orphan->inum); + dbg_gen("deleting orphan ino %lu", (unsigned long)orphan->inum); kfree(orphan); } c->orph_dnext = NULL; @@ -536,8 +538,8 @@ static int insert_dead_orphan(struct ubifs_info *c, ino_t inum) list_add_tail(&orphan->list, &c->orph_list); orphan->dnext = c->orph_dnext; c->orph_dnext = orphan; - dbg_mnt("ino %lu, new %d, tot %d", - inum, c->new_orphans, c->tot_orphans); + dbg_mnt("ino %lu, new %d, tot %d", (unsigned long)inum, + c->new_orphans, c->tot_orphans); return 0; } @@ -609,7 +611,8 @@ static int do_kill_orphans(struct ubifs_info *c, struct ubifs_scan_leb *sleb, n = (le32_to_cpu(orph->ch.len) - UBIFS_ORPH_NODE_SZ) >> 3; for (i = 0; i < n; i++) { inum = le64_to_cpu(orph->inos[i]); - dbg_rcvry("deleting orphaned inode %lu", inum); + dbg_rcvry("deleting orphaned inode %lu", + (unsigned long)inum); err = ubifs_tnc_remove_ino(c, inum); if (err) return err; @@ -840,8 +843,8 @@ static int dbg_orphan_check(struct ubifs_info *c, struct ubifs_zbranch *zbr, if (inum != ci->last_ino) { /* Lowest node type is the inode node, so it comes first */ if (key_type(c, &zbr->key) != UBIFS_INO_KEY) - ubifs_err("found orphan node ino %lu, type %d", inum, - key_type(c, &zbr->key)); + ubifs_err("found orphan node ino %lu, type %d", + (unsigned long)inum, key_type(c, &zbr->key)); ci->last_ino = inum; ci->tot_inos += 1; err = ubifs_tnc_read_node(c, zbr, ci->node); @@ -853,7 +856,8 @@ static int dbg_orphan_check(struct ubifs_info *c, struct ubifs_zbranch *zbr, /* Must be recorded as an orphan */ if (!dbg_find_check_orphan(&ci->root, inum) && !dbg_find_orphan(c, inum)) { - ubifs_err("missing orphan, ino %lu", inum); + ubifs_err("missing orphan, ino %lu", + (unsigned long)inum); ci->missing += 1; } } diff --git a/fs/ubifs/recovery.c b/fs/ubifs/recovery.c index 77d26c141cf..90acac603e6 100644 --- a/fs/ubifs/recovery.c +++ b/fs/ubifs/recovery.c @@ -168,12 +168,12 @@ static int write_rcvrd_mst_node(struct ubifs_info *c, struct ubifs_mst_node *mst) { int err = 0, lnum = UBIFS_MST_LNUM, sz = c->mst_node_alsz; - uint32_t save_flags; + __le32 save_flags; dbg_rcvry("recovery"); save_flags = mst->flags; - mst->flags = cpu_to_le32(le32_to_cpu(mst->flags) | UBIFS_MST_RCVRY); + mst->flags |= cpu_to_le32(UBIFS_MST_RCVRY); ubifs_prepare_node(c, mst, UBIFS_MST_NODE_SZ, 1); err = ubi_leb_change(c->ubi, lnum, mst, sz, UBI_SHORTTERM); @@ -1435,13 +1435,13 @@ static int fix_size_in_place(struct ubifs_info *c, struct size_entry *e) err = ubi_leb_change(c->ubi, lnum, c->sbuf, len, UBI_UNKNOWN); if (err) goto out; - dbg_rcvry("inode %lu at %d:%d size %lld -> %lld ", e->inum, lnum, offs, - i_size, e->d_size); + dbg_rcvry("inode %lu at %d:%d size %lld -> %lld ", + (unsigned long)e->inum, lnum, offs, i_size, e->d_size); return 0; out: ubifs_warn("inode %lu failed to fix size %lld -> %lld error %d", - e->inum, e->i_size, e->d_size, err); + (unsigned long)e->inum, e->i_size, e->d_size, err); return err; } @@ -1472,7 +1472,8 @@ int ubifs_recover_size(struct ubifs_info *c) return err; if (err == -ENOENT) { /* Remove data nodes that have no inode */ - dbg_rcvry("removing ino %lu", e->inum); + dbg_rcvry("removing ino %lu", + (unsigned long)e->inum); err = ubifs_tnc_remove_ino(c, e->inum); if (err) return err; @@ -1493,8 +1494,8 @@ int ubifs_recover_size(struct ubifs_info *c) return PTR_ERR(inode); if (inode->i_size < e->d_size) { dbg_rcvry("ino %lu size %lld -> %lld", - e->inum, e->d_size, - inode->i_size); + (unsigned long)e->inum, + e->d_size, inode->i_size); inode->i_size = e->d_size; ubifs_inode(inode)->ui_size = e->d_size; e->inode = inode; diff --git a/fs/ubifs/replay.c b/fs/ubifs/replay.c index 7399692af85..21f7d047c30 100644 --- a/fs/ubifs/replay.c +++ b/fs/ubifs/replay.c @@ -1065,7 +1065,7 @@ int ubifs_replay_journal(struct ubifs_info *c) ubifs_assert(c->bud_bytes <= c->max_bud_bytes || c->need_recovery); dbg_mnt("finished, log head LEB %d:%d, max_sqnum %llu, " "highest_inum %lu", c->lhead_lnum, c->lhead_offs, c->max_sqnum, - c->highest_inum); + (unsigned long)c->highest_inum); out: destroy_replay_tree(c); destroy_bud_list(c); diff --git a/fs/ubifs/sb.c b/fs/ubifs/sb.c index 2bf753b3888..0f392351dc5 100644 --- a/fs/ubifs/sb.c +++ b/fs/ubifs/sb.c @@ -81,6 +81,7 @@ static int create_default_filesystem(struct ubifs_info *c) int lpt_lebs, lpt_first, orph_lebs, big_lpt, ino_waste, sup_flags = 0; int min_leb_cnt = UBIFS_MIN_LEB_CNT; uint64_t tmp64, main_bytes; + __le64 tmp_le64; /* Some functions called from here depend on the @c->key_len filed */ c->key_len = UBIFS_SK_LEN; @@ -295,10 +296,10 @@ static int create_default_filesystem(struct ubifs_info *c) ino->ch.node_type = UBIFS_INO_NODE; ino->creat_sqnum = cpu_to_le64(++c->max_sqnum); ino->nlink = cpu_to_le32(2); - tmp = cpu_to_le64(CURRENT_TIME_SEC.tv_sec); - ino->atime_sec = tmp; - ino->ctime_sec = tmp; - ino->mtime_sec = tmp; + tmp_le64 = cpu_to_le64(CURRENT_TIME_SEC.tv_sec); + ino->atime_sec = tmp_le64; + ino->ctime_sec = tmp_le64; + ino->mtime_sec = tmp_le64; ino->atime_nsec = 0; ino->ctime_nsec = 0; ino->mtime_nsec = 0; diff --git a/fs/ubifs/scan.c b/fs/ubifs/scan.c index acf5c5fffc6..0ed82479b44 100644 --- a/fs/ubifs/scan.c +++ b/fs/ubifs/scan.c @@ -87,7 +87,7 @@ int ubifs_scan_a_node(const struct ubifs_info *c, void *buf, int len, int lnum, dbg_scan("scanning %s", dbg_ntype(ch->node_type)); - if (ubifs_check_node(c, buf, lnum, offs, quiet)) + if (ubifs_check_node(c, buf, lnum, offs, quiet, 1)) return SCANNED_A_CORRUPT_NODE; if (ch->node_type == UBIFS_PAD_NODE) { diff --git a/fs/ubifs/super.c b/fs/ubifs/super.c index 9a9220333b3..d80b2aef42b 100644 --- a/fs/ubifs/super.c +++ b/fs/ubifs/super.c @@ -36,6 +36,12 @@ #include <linux/mount.h> #include "ubifs.h" +/* + * Maximum amount of memory we may 'kmalloc()' without worrying that we are + * allocating too much. + */ +#define UBIFS_KMALLOC_OK (128*1024) + /* Slab cache for UBIFS inodes */ struct kmem_cache *ubifs_inode_slab; @@ -401,6 +407,16 @@ static int ubifs_show_options(struct seq_file *s, struct vfsmount *mnt) else if (c->mount_opts.unmount_mode == 1) seq_printf(s, ",norm_unmount"); + if (c->mount_opts.bulk_read == 2) + seq_printf(s, ",bulk_read"); + else if (c->mount_opts.bulk_read == 1) + seq_printf(s, ",no_bulk_read"); + + if (c->mount_opts.chk_data_crc == 2) + seq_printf(s, ",chk_data_crc"); + else if (c->mount_opts.chk_data_crc == 1) + seq_printf(s, ",no_chk_data_crc"); + return 0; } @@ -408,13 +424,26 @@ static int ubifs_sync_fs(struct super_block *sb, int wait) { struct ubifs_info *c = sb->s_fs_info; int i, ret = 0, err; + long long bud_bytes; - if (c->jheads) + if (c->jheads) { for (i = 0; i < c->jhead_cnt; i++) { err = ubifs_wbuf_sync(&c->jheads[i].wbuf); if (err && !ret) ret = err; } + + /* Commit the journal unless it has too little data */ + spin_lock(&c->buds_lock); + bud_bytes = c->bud_bytes; + spin_unlock(&c->buds_lock); + if (bud_bytes > c->leb_size) { + err = ubifs_run_commit(c); + if (err) + return err; + } + } + /* * We ought to call sync for c->ubi but it does not have one. If it had * it would in turn call mtd->sync, however mtd operations are @@ -538,6 +567,11 @@ static int init_constants_early(struct ubifs_info *c) * calculations when reporting free space. */ c->leb_overhead = c->leb_size % UBIFS_MAX_DATA_NODE_SZ; + + /* Buffer size for bulk-reads */ + c->max_bu_buf_len = UBIFS_MAX_BULK_READ * UBIFS_MAX_DATA_NODE_SZ; + if (c->max_bu_buf_len > c->leb_size) + c->max_bu_buf_len = c->leb_size; return 0; } @@ -840,17 +874,29 @@ static int check_volume_empty(struct ubifs_info *c) * * Opt_fast_unmount: do not run a journal commit before un-mounting * Opt_norm_unmount: run a journal commit before un-mounting + * Opt_bulk_read: enable bulk-reads + * Opt_no_bulk_read: disable bulk-reads + * Opt_chk_data_crc: check CRCs when reading data nodes + * Opt_no_chk_data_crc: do not check CRCs when reading data nodes * Opt_err: just end of array marker */ enum { Opt_fast_unmount, Opt_norm_unmount, + Opt_bulk_read, + Opt_no_bulk_read, + Opt_chk_data_crc, + Opt_no_chk_data_crc, Opt_err, }; static const match_table_t tokens = { {Opt_fast_unmount, "fast_unmount"}, {Opt_norm_unmount, "norm_unmount"}, + {Opt_bulk_read, "bulk_read"}, + {Opt_no_bulk_read, "no_bulk_read"}, + {Opt_chk_data_crc, "chk_data_crc"}, + {Opt_no_chk_data_crc, "no_chk_data_crc"}, {Opt_err, NULL}, }; @@ -888,6 +934,22 @@ static int ubifs_parse_options(struct ubifs_info *c, char *options, c->mount_opts.unmount_mode = 1; c->fast_unmount = 0; break; + case Opt_bulk_read: + c->mount_opts.bulk_read = 2; + c->bulk_read = 1; + break; + case Opt_no_bulk_read: + c->mount_opts.bulk_read = 1; + c->bulk_read = 0; + break; + case Opt_chk_data_crc: + c->mount_opts.chk_data_crc = 2; + c->no_chk_data_crc = 0; + break; + case Opt_no_chk_data_crc: + c->mount_opts.chk_data_crc = 1; + c->no_chk_data_crc = 1; + break; default: ubifs_err("unrecognized mount option \"%s\" " "or missing value", p); @@ -929,6 +991,34 @@ static void destroy_journal(struct ubifs_info *c) } /** + * bu_init - initialize bulk-read information. + * @c: UBIFS file-system description object + */ +static void bu_init(struct ubifs_info *c) +{ + ubifs_assert(c->bulk_read == 1); + + if (c->bu.buf) + return; /* Already initialized */ + +again: + c->bu.buf = kmalloc(c->max_bu_buf_len, GFP_KERNEL | __GFP_NOWARN); + if (!c->bu.buf) { + if (c->max_bu_buf_len > UBIFS_KMALLOC_OK) { + c->max_bu_buf_len = UBIFS_KMALLOC_OK; + goto again; + } + + /* Just disable bulk-read */ + ubifs_warn("Cannot allocate %d bytes of memory for bulk-read, " + "disabling it", c->max_bu_buf_len); + c->mount_opts.bulk_read = 1; + c->bulk_read = 0; + return; + } +} + +/** * mount_ubifs - mount UBIFS file-system. * @c: UBIFS file-system description object * @@ -996,6 +1086,15 @@ static int mount_ubifs(struct ubifs_info *c) goto out_free; } + if (c->bulk_read == 1) + bu_init(c); + + /* + * We have to check all CRCs, even for data nodes, when we mount the FS + * (specifically, when we are replaying). + */ + c->always_chk_crc = 1; + err = ubifs_read_superblock(c); if (err) goto out_free; @@ -1032,8 +1131,6 @@ static int mount_ubifs(struct ubifs_info *c) /* Create background thread */ c->bgt = kthread_create(ubifs_bg_thread, c, c->bgt_name); - if (!c->bgt) - c->bgt = ERR_PTR(-EINVAL); if (IS_ERR(c->bgt)) { err = PTR_ERR(c->bgt); c->bgt = NULL; @@ -1139,24 +1236,28 @@ static int mount_ubifs(struct ubifs_info *c) if (err) goto out_infos; + c->always_chk_crc = 0; + ubifs_msg("mounted UBI device %d, volume %d, name \"%s\"", c->vi.ubi_num, c->vi.vol_id, c->vi.name); if (mounted_read_only) ubifs_msg("mounted read-only"); x = (long long)c->main_lebs * c->leb_size; - ubifs_msg("file system size: %lld bytes (%lld KiB, %lld MiB, %d LEBs)", - x, x >> 10, x >> 20, c->main_lebs); + ubifs_msg("file system size: %lld bytes (%lld KiB, %lld MiB, %d " + "LEBs)", x, x >> 10, x >> 20, c->main_lebs); x = (long long)c->log_lebs * c->leb_size + c->max_bud_bytes; - ubifs_msg("journal size: %lld bytes (%lld KiB, %lld MiB, %d LEBs)", - x, x >> 10, x >> 20, c->log_lebs + c->max_bud_cnt); - ubifs_msg("default compressor: %s", ubifs_compr_name(c->default_compr)); - ubifs_msg("media format %d, latest format %d", + ubifs_msg("journal size: %lld bytes (%lld KiB, %lld MiB, %d " + "LEBs)", x, x >> 10, x >> 20, c->log_lebs + c->max_bud_cnt); + ubifs_msg("media format: %d (latest is %d)", c->fmt_version, UBIFS_FORMAT_VERSION); + ubifs_msg("default compressor: %s", ubifs_compr_name(c->default_compr)); + ubifs_msg("reserved for root: %llu bytes (%llu KiB)", + c->report_rp_size, c->report_rp_size >> 10); dbg_msg("compiled on: " __DATE__ " at " __TIME__); dbg_msg("min. I/O unit size: %d bytes", c->min_io_size); dbg_msg("LEB size: %d bytes (%d KiB)", - c->leb_size, c->leb_size / 1024); + c->leb_size, c->leb_size >> 10); dbg_msg("data journal heads: %d", c->jhead_cnt - NONDATA_JHEADS_CNT); dbg_msg("UUID: %02X%02X%02X%02X-%02X%02X" @@ -1222,6 +1323,7 @@ out_cbuf: out_dereg: dbg_failure_mode_deregistration(c); out_free: + kfree(c->bu.buf); vfree(c->ileb_buf); vfree(c->sbuf); kfree(c->bottom_up_buf); @@ -1258,10 +1360,11 @@ static void ubifs_umount(struct ubifs_info *c) kfree(c->cbuf); kfree(c->rcvrd_mst_node); kfree(c->mst_node); + kfree(c->bu.buf); + vfree(c->ileb_buf); vfree(c->sbuf); kfree(c->bottom_up_buf); UBIFS_DBG(vfree(c->dbg_buf)); - vfree(c->ileb_buf); dbg_failure_mode_deregistration(c); } @@ -1282,6 +1385,7 @@ static int ubifs_remount_rw(struct ubifs_info *c) mutex_lock(&c->umount_mutex); c->remounting_rw = 1; + c->always_chk_crc = 1; /* Check for enough free space */ if (ubifs_calc_available(c, c->min_idx_lebs) <= 0) { @@ -1345,20 +1449,20 @@ static int ubifs_remount_rw(struct ubifs_info *c) /* Create background thread */ c->bgt = kthread_create(ubifs_bg_thread, c, c->bgt_name); - if (!c->bgt) - c->bgt = ERR_PTR(-EINVAL); if (IS_ERR(c->bgt)) { err = PTR_ERR(c->bgt); c->bgt = NULL; ubifs_err("cannot spawn \"%s\", error %d", c->bgt_name, err); - return err; + goto out; } wake_up_process(c->bgt); c->orph_buf = vmalloc(c->leb_size); - if (!c->orph_buf) - return -ENOMEM; + if (!c->orph_buf) { + err = -ENOMEM; + goto out; + } /* Check for enough log space */ lnum = c->lhead_lnum + 1; @@ -1385,6 +1489,7 @@ static int ubifs_remount_rw(struct ubifs_info *c) dbg_gen("re-mounted read-write"); c->vfs_sb->s_flags &= ~MS_RDONLY; c->remounting_rw = 0; + c->always_chk_crc = 0; mutex_unlock(&c->umount_mutex); return 0; @@ -1400,6 +1505,7 @@ out: c->ileb_buf = NULL; ubifs_lpt_free(c, 1); c->remounting_rw = 0; + c->always_chk_crc = 0; mutex_unlock(&c->umount_mutex); return err; } @@ -1408,12 +1514,9 @@ out: * commit_on_unmount - commit the journal when un-mounting. * @c: UBIFS file-system description object * - * This function is called during un-mounting and it commits the journal unless - * the "fast unmount" mode is enabled. It also avoids committing the journal if - * it contains too few data. - * - * Sometimes recovery requires the journal to be committed at least once, and - * this function takes care about this. + * This function is called during un-mounting and re-mounting, and it commits + * the journal unless the "fast unmount" mode is enabled. It also avoids + * committing the journal if it contains too few data. */ static void commit_on_unmount(struct ubifs_info *c) { @@ -1559,6 +1662,7 @@ static int ubifs_remount_fs(struct super_block *sb, int *flags, char *data) ubifs_err("invalid or unknown remount parameter"); return err; } + if ((sb->s_flags & MS_RDONLY) && !(*flags & MS_RDONLY)) { err = ubifs_remount_rw(c); if (err) @@ -1566,6 +1670,14 @@ static int ubifs_remount_fs(struct super_block *sb, int *flags, char *data) } else if (!(sb->s_flags & MS_RDONLY) && (*flags & MS_RDONLY)) ubifs_remount_ro(c); + if (c->bulk_read == 1) + bu_init(c); + else { + dbg_gen("disable bulk-read"); + kfree(c->bu.buf); + c->bu.buf = NULL; + } + return 0; } @@ -1656,6 +1768,7 @@ static int ubifs_fill_super(struct super_block *sb, void *data, int silent) mutex_init(&c->log_mutex); mutex_init(&c->mst_mutex); mutex_init(&c->umount_mutex); + mutex_init(&c->bu_mutex); init_waitqueue_head(&c->cmt_wq); c->buds = RB_ROOT; c->old_idx = RB_ROOT; diff --git a/fs/ubifs/tnc.c b/fs/ubifs/tnc.c index 7634c597088..6eef5344a14 100644 --- a/fs/ubifs/tnc.c +++ b/fs/ubifs/tnc.c @@ -284,7 +284,7 @@ static struct ubifs_znode *dirty_cow_znode(struct ubifs_info *c, } zn = copy_znode(c, znode); - if (unlikely(IS_ERR(zn))) + if (IS_ERR(zn)) return zn; if (zbr->len) { @@ -470,6 +470,10 @@ static int try_read_node(const struct ubifs_info *c, void *buf, int type, if (node_len != len) return 0; + if (type == UBIFS_DATA_NODE && !c->always_chk_crc) + if (c->no_chk_data_crc) + return 0; + crc = crc32(UBIFS_CRC32_INIT, buf + 8, node_len - 8); node_crc = le32_to_cpu(ch->crc); if (crc != node_crc) @@ -1128,7 +1132,7 @@ static struct ubifs_znode *dirty_cow_bottom_up(struct ubifs_info *c, ubifs_assert(znode == c->zroot.znode); znode = dirty_cow_znode(c, &c->zroot); } - if (unlikely(IS_ERR(znode)) || !p) + if (IS_ERR(znode) || !p) break; ubifs_assert(path[p - 1] >= 0); ubifs_assert(path[p - 1] < znode->child_cnt); @@ -1492,6 +1496,294 @@ out: } /** + * ubifs_tnc_get_bu_keys - lookup keys for bulk-read. + * @c: UBIFS file-system description object + * @bu: bulk-read parameters and results + * + * Lookup consecutive data node keys for the same inode that reside + * consecutively in the same LEB. This function returns zero in case of success + * and a negative error code in case of failure. + * + * Note, if the bulk-read buffer length (@bu->buf_len) is known, this function + * makes sure bulk-read nodes fit the buffer. Otherwise, this function prepares + * maxumum possible amount of nodes for bulk-read. + */ +int ubifs_tnc_get_bu_keys(struct ubifs_info *c, struct bu_info *bu) +{ + int n, err = 0, lnum = -1, uninitialized_var(offs); + int uninitialized_var(len); + unsigned int block = key_block(c, &bu->key); + struct ubifs_znode *znode; + + bu->cnt = 0; + bu->blk_cnt = 0; + bu->eof = 0; + + mutex_lock(&c->tnc_mutex); + /* Find first key */ + err = ubifs_lookup_level0(c, &bu->key, &znode, &n); + if (err < 0) + goto out; + if (err) { + /* Key found */ + len = znode->zbranch[n].len; + /* The buffer must be big enough for at least 1 node */ + if (len > bu->buf_len) { + err = -EINVAL; + goto out; + } + /* Add this key */ + bu->zbranch[bu->cnt++] = znode->zbranch[n]; + bu->blk_cnt += 1; + lnum = znode->zbranch[n].lnum; + offs = ALIGN(znode->zbranch[n].offs + len, 8); + } + while (1) { + struct ubifs_zbranch *zbr; + union ubifs_key *key; + unsigned int next_block; + + /* Find next key */ + err = tnc_next(c, &znode, &n); + if (err) + goto out; + zbr = &znode->zbranch[n]; + key = &zbr->key; + /* See if there is another data key for this file */ + if (key_inum(c, key) != key_inum(c, &bu->key) || + key_type(c, key) != UBIFS_DATA_KEY) { + err = -ENOENT; + goto out; + } + if (lnum < 0) { + /* First key found */ + lnum = zbr->lnum; + offs = ALIGN(zbr->offs + zbr->len, 8); + len = zbr->len; + if (len > bu->buf_len) { + err = -EINVAL; + goto out; + } + } else { + /* + * The data nodes must be in consecutive positions in + * the same LEB. + */ + if (zbr->lnum != lnum || zbr->offs != offs) + goto out; + offs += ALIGN(zbr->len, 8); + len = ALIGN(len, 8) + zbr->len; + /* Must not exceed buffer length */ + if (len > bu->buf_len) + goto out; + } + /* Allow for holes */ + next_block = key_block(c, key); + bu->blk_cnt += (next_block - block - 1); + if (bu->blk_cnt >= UBIFS_MAX_BULK_READ) + goto out; + block = next_block; + /* Add this key */ + bu->zbranch[bu->cnt++] = *zbr; + bu->blk_cnt += 1; + /* See if we have room for more */ + if (bu->cnt >= UBIFS_MAX_BULK_READ) + goto out; + if (bu->blk_cnt >= UBIFS_MAX_BULK_READ) + goto out; + } +out: + if (err == -ENOENT) { + bu->eof = 1; + err = 0; + } + bu->gc_seq = c->gc_seq; + mutex_unlock(&c->tnc_mutex); + if (err) + return err; + /* + * An enormous hole could cause bulk-read to encompass too many + * page cache pages, so limit the number here. + */ + if (bu->blk_cnt > UBIFS_MAX_BULK_READ) + bu->blk_cnt = UBIFS_MAX_BULK_READ; + /* + * Ensure that bulk-read covers a whole number of page cache + * pages. + */ + if (UBIFS_BLOCKS_PER_PAGE == 1 || + !(bu->blk_cnt & (UBIFS_BLOCKS_PER_PAGE - 1))) + return 0; + if (bu->eof) { + /* At the end of file we can round up */ + bu->blk_cnt += UBIFS_BLOCKS_PER_PAGE - 1; + return 0; + } + /* Exclude data nodes that do not make up a whole page cache page */ + block = key_block(c, &bu->key) + bu->blk_cnt; + block &= ~(UBIFS_BLOCKS_PER_PAGE - 1); + while (bu->cnt) { + if (key_block(c, &bu->zbranch[bu->cnt - 1].key) < block) + break; + bu->cnt -= 1; + } + return 0; +} + +/** + * read_wbuf - bulk-read from a LEB with a wbuf. + * @wbuf: wbuf that may overlap the read + * @buf: buffer into which to read + * @len: read length + * @lnum: LEB number from which to read + * @offs: offset from which to read + * + * This functions returns %0 on success or a negative error code on failure. + */ +static int read_wbuf(struct ubifs_wbuf *wbuf, void *buf, int len, int lnum, + int offs) +{ + const struct ubifs_info *c = wbuf->c; + int rlen, overlap; + + dbg_io("LEB %d:%d, length %d", lnum, offs, len); + ubifs_assert(wbuf && lnum >= 0 && lnum < c->leb_cnt && offs >= 0); + ubifs_assert(!(offs & 7) && offs < c->leb_size); + ubifs_assert(offs + len <= c->leb_size); + + spin_lock(&wbuf->lock); + overlap = (lnum == wbuf->lnum && offs + len > wbuf->offs); + if (!overlap) { + /* We may safely unlock the write-buffer and read the data */ + spin_unlock(&wbuf->lock); + return ubi_read(c->ubi, lnum, buf, offs, len); + } + + /* Don't read under wbuf */ + rlen = wbuf->offs - offs; + if (rlen < 0) + rlen = 0; + + /* Copy the rest from the write-buffer */ + memcpy(buf + rlen, wbuf->buf + offs + rlen - wbuf->offs, len - rlen); + spin_unlock(&wbuf->lock); + + if (rlen > 0) + /* Read everything that goes before write-buffer */ + return ubi_read(c->ubi, lnum, buf, offs, rlen); + + return 0; +} + +/** + * validate_data_node - validate data nodes for bulk-read. + * @c: UBIFS file-system description object + * @buf: buffer containing data node to validate + * @zbr: zbranch of data node to validate + * + * This functions returns %0 on success or a negative error code on failure. + */ +static int validate_data_node(struct ubifs_info *c, void *buf, + struct ubifs_zbranch *zbr) +{ + union ubifs_key key1; + struct ubifs_ch *ch = buf; + int err, len; + + if (ch->node_type != UBIFS_DATA_NODE) { + ubifs_err("bad node type (%d but expected %d)", + ch->node_type, UBIFS_DATA_NODE); + goto out_err; + } + + err = ubifs_check_node(c, buf, zbr->lnum, zbr->offs, 0, 0); + if (err) { + ubifs_err("expected node type %d", UBIFS_DATA_NODE); + goto out; + } + + len = le32_to_cpu(ch->len); + if (len != zbr->len) { + ubifs_err("bad node length %d, expected %d", len, zbr->len); + goto out_err; + } + + /* Make sure the key of the read node is correct */ + key_read(c, buf + UBIFS_KEY_OFFSET, &key1); + if (!keys_eq(c, &zbr->key, &key1)) { + ubifs_err("bad key in node at LEB %d:%d", + zbr->lnum, zbr->offs); + dbg_tnc("looked for key %s found node's key %s", + DBGKEY(&zbr->key), DBGKEY1(&key1)); + goto out_err; + } + + return 0; + +out_err: + err = -EINVAL; +out: + ubifs_err("bad node at LEB %d:%d", zbr->lnum, zbr->offs); + dbg_dump_node(c, buf); + dbg_dump_stack(); + return err; +} + +/** + * ubifs_tnc_bulk_read - read a number of data nodes in one go. + * @c: UBIFS file-system description object + * @bu: bulk-read parameters and results + * + * This functions reads and validates the data nodes that were identified by the + * 'ubifs_tnc_get_bu_keys()' function. This functions returns %0 on success, + * -EAGAIN to indicate a race with GC, or another negative error code on + * failure. + */ +int ubifs_tnc_bulk_read(struct ubifs_info *c, struct bu_info *bu) +{ + int lnum = bu->zbranch[0].lnum, offs = bu->zbranch[0].offs, len, err, i; + struct ubifs_wbuf *wbuf; + void *buf; + + len = bu->zbranch[bu->cnt - 1].offs; + len += bu->zbranch[bu->cnt - 1].len - offs; + if (len > bu->buf_len) { + ubifs_err("buffer too small %d vs %d", bu->buf_len, len); + return -EINVAL; + } + + /* Do the read */ + wbuf = ubifs_get_wbuf(c, lnum); + if (wbuf) + err = read_wbuf(wbuf, bu->buf, len, lnum, offs); + else + err = ubi_read(c->ubi, lnum, bu->buf, offs, len); + + /* Check for a race with GC */ + if (maybe_leb_gced(c, lnum, bu->gc_seq)) + return -EAGAIN; + + if (err && err != -EBADMSG) { + ubifs_err("failed to read from LEB %d:%d, error %d", + lnum, offs, err); + dbg_dump_stack(); + dbg_tnc("key %s", DBGKEY(&bu->key)); + return err; + } + + /* Validate the nodes read */ + buf = bu->buf; + for (i = 0; i < bu->cnt; i++) { + err = validate_data_node(c, buf, &bu->zbranch[i]); + if (err) + return err; + buf = buf + ALIGN(bu->zbranch[i].len, 8); + } + + return 0; +} + +/** * do_lookup_nm- look up a "hashed" node. * @c: UBIFS file-system description object * @key: node key to lookup @@ -1675,7 +1967,7 @@ static int tnc_insert(struct ubifs_info *c, struct ubifs_znode *znode, { struct ubifs_znode *zn, *zi, *zp; int i, keep, move, appending = 0; - union ubifs_key *key = &zbr->key; + union ubifs_key *key = &zbr->key, *key1; ubifs_assert(n >= 0 && n <= c->fanout); @@ -1716,20 +2008,33 @@ again: zn->level = znode->level; /* Decide where to split */ - if (znode->level == 0 && n == c->fanout && - key_type(c, key) == UBIFS_DATA_KEY) { - union ubifs_key *key1; - - /* - * If this is an inode which is being appended - do not split - * it because no other zbranches can be inserted between - * zbranches of consecutive data nodes anyway. - */ - key1 = &znode->zbranch[n - 1].key; - if (key_inum(c, key1) == key_inum(c, key) && - key_type(c, key1) == UBIFS_DATA_KEY && - key_block(c, key1) == key_block(c, key) - 1) - appending = 1; + if (znode->level == 0 && key_type(c, key) == UBIFS_DATA_KEY) { + /* Try not to split consecutive data keys */ + if (n == c->fanout) { + key1 = &znode->zbranch[n - 1].key; + if (key_inum(c, key1) == key_inum(c, key) && + key_type(c, key1) == UBIFS_DATA_KEY) + appending = 1; + } else + goto check_split; + } else if (appending && n != c->fanout) { + /* Try not to split consecutive data keys */ + appending = 0; +check_split: + if (n >= (c->fanout + 1) / 2) { + key1 = &znode->zbranch[0].key; + if (key_inum(c, key1) == key_inum(c, key) && + key_type(c, key1) == UBIFS_DATA_KEY) { + key1 = &znode->zbranch[n].key; + if (key_inum(c, key1) != key_inum(c, key) || + key_type(c, key1) != UBIFS_DATA_KEY) { + keep = n; + move = c->fanout - keep; + zi = znode; + goto do_split; + } + } + } } if (appending) { @@ -1759,6 +2064,8 @@ again: zbr->znode->parent = zn; } +do_split: + __set_bit(DIRTY_ZNODE, &zn->flags); atomic_long_inc(&c->dirty_zn_cnt); @@ -1785,14 +2092,11 @@ again: /* Insert new znode (produced by spitting) into the parent */ if (zp) { - i = n; + if (n == 0 && zi == znode && znode->iip == 0) + correct_parent_keys(c, znode); + /* Locate insertion point */ n = znode->iip + 1; - if (appending && n != c->fanout) - appending = 0; - - if (i == 0 && zi == znode && znode->iip == 0) - correct_parent_keys(c, znode); /* Tail recursion */ zbr->key = zn->zbranch[0].key; @@ -2378,7 +2682,7 @@ int ubifs_tnc_remove_ino(struct ubifs_info *c, ino_t inum) struct ubifs_dent_node *xent, *pxent = NULL; struct qstr nm = { .name = NULL }; - dbg_tnc("ino %lu", inum); + dbg_tnc("ino %lu", (unsigned long)inum); /* * Walk all extended attribute entries and remove them together with @@ -2398,7 +2702,8 @@ int ubifs_tnc_remove_ino(struct ubifs_info *c, ino_t inum) } xattr_inum = le64_to_cpu(xent->inum); - dbg_tnc("xent '%s', ino %lu", xent->name, xattr_inum); + dbg_tnc("xent '%s', ino %lu", xent->name, + (unsigned long)xattr_inum); nm.name = xent->name; nm.len = le16_to_cpu(xent->nlen); diff --git a/fs/ubifs/tnc_misc.c b/fs/ubifs/tnc_misc.c index a25c1cc1f8d..b48db999903 100644 --- a/fs/ubifs/tnc_misc.c +++ b/fs/ubifs/tnc_misc.c @@ -480,8 +480,8 @@ int ubifs_tnc_read_node(struct ubifs_info *c, struct ubifs_zbranch *zbr, } /* Make sure the key of the read node is correct */ - key_read(c, key, &key1); - if (memcmp(node + UBIFS_KEY_OFFSET, &key1, c->key_len)) { + key_read(c, node + UBIFS_KEY_OFFSET, &key1); + if (!keys_eq(c, key, &key1)) { ubifs_err("bad key in node at LEB %d:%d", zbr->lnum, zbr->offs); dbg_tnc("looked for key %s found node's key %s", diff --git a/fs/ubifs/ubifs-media.h b/fs/ubifs/ubifs-media.h index a9ecbd9af20..0b378042a3a 100644 --- a/fs/ubifs/ubifs-media.h +++ b/fs/ubifs/ubifs-media.h @@ -75,7 +75,6 @@ */ #define UBIFS_BLOCK_SIZE 4096 #define UBIFS_BLOCK_SHIFT 12 -#define UBIFS_BLOCK_MASK 0x00000FFF /* UBIFS padding byte pattern (must not be first or last byte of node magic) */ #define UBIFS_PADDING_BYTE 0xCE diff --git a/fs/ubifs/ubifs.h b/fs/ubifs/ubifs.h index 17c620b93ee..46b172560a0 100644 --- a/fs/ubifs/ubifs.h +++ b/fs/ubifs/ubifs.h @@ -142,6 +142,9 @@ /* Maximum expected tree height for use by bottom_up_buf */ #define BOTTOM_UP_HEIGHT 64 +/* Maximum number of data nodes to bulk-read */ +#define UBIFS_MAX_BULK_READ 32 + /* * Lockdep classes for UBIFS inode @ui_mutex. */ @@ -328,9 +331,10 @@ struct ubifs_gced_idx_leb { * this inode * @dirty: non-zero if the inode is dirty * @xattr: non-zero if this is an extended attribute inode + * @bulk_read: non-zero if bulk-read should be used * @ui_mutex: serializes inode write-back with the rest of VFS operations, - * serializes "clean <-> dirty" state changes, protects @dirty, - * @ui_size, and @xattr_size + * serializes "clean <-> dirty" state changes, serializes bulk-read, + * protects @dirty, @bulk_read, @ui_size, and @xattr_size * @ui_lock: protects @synced_i_size * @synced_i_size: synchronized size of inode, i.e. the value of inode size * currently stored on the flash; used only for regular file @@ -338,6 +342,8 @@ struct ubifs_gced_idx_leb { * @ui_size: inode size used by UBIFS when writing to flash * @flags: inode flags (@UBIFS_COMPR_FL, etc) * @compr_type: default compression type used for this inode + * @last_page_read: page number of last page read (for bulk read) + * @read_in_a_row: number of consecutive pages read in a row (for bulk read) * @data_len: length of the data attached to the inode * @data: inode's data * @@ -379,12 +385,15 @@ struct ubifs_inode { unsigned int xattr_names; unsigned int dirty:1; unsigned int xattr:1; + unsigned int bulk_read:1; struct mutex ui_mutex; spinlock_t ui_lock; loff_t synced_i_size; loff_t ui_size; int flags; int compr_type; + pgoff_t last_page_read; + pgoff_t read_in_a_row; int data_len; void *data; }; @@ -698,8 +707,8 @@ struct ubifs_jhead { * struct ubifs_zbranch - key/coordinate/length branch stored in znodes. * @key: key * @znode: znode address in memory - * @lnum: LEB number of the indexing node - * @offs: offset of the indexing node within @lnum + * @lnum: LEB number of the target node (indexing node or data node) + * @offs: target node offset within @lnum * @len: target node length */ struct ubifs_zbranch { @@ -744,6 +753,28 @@ struct ubifs_znode { }; /** + * struct bu_info - bulk-read information. + * @key: first data node key + * @zbranch: zbranches of data nodes to bulk read + * @buf: buffer to read into + * @buf_len: buffer length + * @gc_seq: GC sequence number to detect races with GC + * @cnt: number of data nodes for bulk read + * @blk_cnt: number of data blocks including holes + * @oef: end of file reached + */ +struct bu_info { + union ubifs_key key; + struct ubifs_zbranch zbranch[UBIFS_MAX_BULK_READ]; + void *buf; + int buf_len; + int gc_seq; + int cnt; + int blk_cnt; + int eof; +}; + +/** * struct ubifs_node_range - node length range description data structure. * @len: fixed node length * @min_len: minimum possible node length @@ -862,9 +893,13 @@ struct ubifs_orphan { /** * struct ubifs_mount_opts - UBIFS-specific mount options information. * @unmount_mode: selected unmount mode (%0 default, %1 normal, %2 fast) + * @bulk_read: enable bulk-reads + * @chk_data_crc: check CRCs when reading data nodes */ struct ubifs_mount_opts { unsigned int unmount_mode:2; + unsigned int bulk_read:2; + unsigned int chk_data_crc:2; }; /** @@ -905,13 +940,12 @@ struct ubifs_mount_opts { * @cmt_state: commit state * @cs_lock: commit state lock * @cmt_wq: wait queue to sleep on if the log is full and a commit is running + * * @fast_unmount: do not run journal commit before un-mounting * @big_lpt: flag that LPT is too big to write whole during commit - * @check_lpt_free: flag that indicates LPT GC may be needed - * @nospace: non-zero if the file-system does not have flash space (used as - * optimization) - * @nospace_rp: the same as @nospace, but additionally means that even reserved - * pool is full + * @no_chk_data_crc: do not check CRCs when reading data nodes (except during + * recovery) + * @bulk_read: enable bulk-reads * * @tnc_mutex: protects the Tree Node Cache (TNC), @zroot, @cnext, @enext, and * @calc_idx_sz @@ -936,6 +970,10 @@ struct ubifs_mount_opts { * @mst_offs: offset of valid master node * @mst_mutex: protects the master node area, @mst_node, and @mst_offs * + * @max_bu_buf_len: maximum bulk-read buffer length + * @bu_mutex: protects the pre-allocated bulk-read buffer and @c->bu + * @bu: pre-allocated bulk-read information + * * @log_lebs: number of logical eraseblocks in the log * @log_bytes: log size in bytes * @log_last: last LEB of the log @@ -977,12 +1015,17 @@ struct ubifs_mount_opts { * but which still have to be taken into account because * the index has not been committed so far * @space_lock: protects @budg_idx_growth, @budg_data_growth, @budg_dd_growth, - * @budg_uncommited_idx, @min_idx_lebs, @old_idx_sz, and @lst; + * @budg_uncommited_idx, @min_idx_lebs, @old_idx_sz, @lst, + * @nospace, and @nospace_rp; * @min_idx_lebs: minimum number of LEBs required for the index * @old_idx_sz: size of index on flash * @calc_idx_sz: temporary variable which is used to calculate new index size * (contains accurate new index size at end of TNC commit start) * @lst: lprops statistics + * @nospace: non-zero if the file-system does not have flash space (used as + * optimization) + * @nospace_rp: the same as @nospace, but additionally means that even reserved + * pool is full * * @page_budget: budget for a page * @inode_budget: budget for an inode @@ -1061,6 +1104,7 @@ struct ubifs_mount_opts { * @lpt_drty_flgs: dirty flags for LPT special nodes e.g. ltab * @dirty_nn_cnt: number of dirty nnodes * @dirty_pn_cnt: number of dirty pnodes + * @check_lpt_free: flag that indicates LPT GC may be needed * @lpt_sz: LPT size * @lpt_nod_buf: buffer for an on-flash nnode or pnode * @lpt_buf: buffer of LEB size used by LPT @@ -1102,6 +1146,7 @@ struct ubifs_mount_opts { * @rcvrd_mst_node: recovered master node to write when mounting ro to rw * @size_tree: inode size information for recovery * @remounting_rw: set while remounting from ro to rw (sb flags have MS_RDONLY) + * @always_chk_crc: always check CRCs (while mounting and remounting rw) * @mount_opts: UBIFS-specific mount options * * @dbg_buf: a buffer of LEB size used for debugging purposes @@ -1146,11 +1191,11 @@ struct ubifs_info { int cmt_state; spinlock_t cs_lock; wait_queue_head_t cmt_wq; + unsigned int fast_unmount:1; unsigned int big_lpt:1; - unsigned int check_lpt_free:1; - unsigned int nospace:1; - unsigned int nospace_rp:1; + unsigned int no_chk_data_crc:1; + unsigned int bulk_read:1; struct mutex tnc_mutex; struct ubifs_zbranch zroot; @@ -1176,6 +1221,10 @@ struct ubifs_info { int mst_offs; struct mutex mst_mutex; + int max_bu_buf_len; + struct mutex bu_mutex; + struct bu_info bu; + int log_lebs; long long log_bytes; int log_last; @@ -1218,6 +1267,8 @@ struct ubifs_info { unsigned long long old_idx_sz; unsigned long long calc_idx_sz; struct ubifs_lp_stats lst; + unsigned int nospace:1; + unsigned int nospace_rp:1; int page_budget; int inode_budget; @@ -1294,6 +1345,7 @@ struct ubifs_info { int lpt_drty_flgs; int dirty_nn_cnt; int dirty_pn_cnt; + int check_lpt_free; long long lpt_sz; void *lpt_nod_buf; void *lpt_buf; @@ -1335,6 +1387,7 @@ struct ubifs_info { struct ubifs_mst_node *rcvrd_mst_node; struct rb_root size_tree; int remounting_rw; + int always_chk_crc; struct ubifs_mount_opts mount_opts; #ifdef CONFIG_UBIFS_FS_DEBUG @@ -1347,6 +1400,12 @@ struct ubifs_info { unsigned long fail_timeout; unsigned int fail_cnt; unsigned int fail_cnt_max; + long long chk_lpt_sz; + long long chk_lpt_sz2; + long long chk_lpt_wastage; + int chk_lpt_lebs; + int new_nhead_lnum; + int new_nhead_offs; #endif }; @@ -1377,7 +1436,7 @@ int ubifs_read_node_wbuf(struct ubifs_wbuf *wbuf, void *buf, int type, int len, int ubifs_write_node(struct ubifs_info *c, void *node, int len, int lnum, int offs, int dtype); int ubifs_check_node(const struct ubifs_info *c, const void *buf, int lnum, - int offs, int quiet); + int offs, int quiet, int chk_crc); void ubifs_prepare_node(struct ubifs_info *c, void *buf, int len, int pad); void ubifs_prep_grp_node(struct ubifs_info *c, void *node, int len, int last); int ubifs_io_init(struct ubifs_info *c); @@ -1490,6 +1549,8 @@ void destroy_old_idx(struct ubifs_info *c); int is_idx_node_in_tnc(struct ubifs_info *c, union ubifs_key *key, int level, int lnum, int offs); int insert_old_idx_znode(struct ubifs_info *c, struct ubifs_znode *znode); +int ubifs_tnc_get_bu_keys(struct ubifs_info *c, struct bu_info *bu); +int ubifs_tnc_bulk_read(struct ubifs_info *c, struct bu_info *bu); /* tnc_misc.c */ struct ubifs_znode *ubifs_tnc_levelorder_next(struct ubifs_znode *zr, @@ -1586,12 +1647,10 @@ int ubifs_lpt_post_commit(struct ubifs_info *c); void ubifs_lpt_free(struct ubifs_info *c, int wr_only); /* lprops.c */ -void ubifs_get_lprops(struct ubifs_info *c); const struct ubifs_lprops *ubifs_change_lp(struct ubifs_info *c, const struct ubifs_lprops *lp, int free, int dirty, int flags, int idx_gc_cnt); -void ubifs_release_lprops(struct ubifs_info *c); void ubifs_get_lp_stats(struct ubifs_info *c, struct ubifs_lp_stats *stats); void ubifs_add_to_cat(struct ubifs_info *c, struct ubifs_lprops *lprops, int cat); diff --git a/fs/ubifs/xattr.c b/fs/ubifs/xattr.c index 649bec78b64..cfd31e229c8 100644 --- a/fs/ubifs/xattr.c +++ b/fs/ubifs/xattr.c @@ -446,7 +446,7 @@ ssize_t ubifs_listxattr(struct dentry *dentry, char *buffer, size_t size) int type; xent = ubifs_tnc_next_ent(c, &key, &nm); - if (unlikely(IS_ERR(xent))) { + if (IS_ERR(xent)) { err = PTR_ERR(xent); break; } diff --git a/fs/udf/ialloc.c b/fs/udf/ialloc.c index a4f2b3ce45b..31fc84297dd 100644 --- a/fs/udf/ialloc.c +++ b/fs/udf/ialloc.c @@ -126,13 +126,13 @@ struct inode *udf_new_inode(struct inode *dir, int mode, int *err) } mutex_unlock(&sbi->s_alloc_mutex); inode->i_mode = mode; - inode->i_uid = current->fsuid; + inode->i_uid = current_fsuid(); if (dir->i_mode & S_ISGID) { inode->i_gid = dir->i_gid; if (S_ISDIR(mode)) mode |= S_ISGID; } else { - inode->i_gid = current->fsgid; + inode->i_gid = current_fsgid(); } iinfo->i_location.logicalBlockNum = block; diff --git a/fs/udf/inode.c b/fs/udf/inode.c index 6e74b117aaf..30ebde490f7 100644 --- a/fs/udf/inode.c +++ b/fs/udf/inode.c @@ -106,6 +106,7 @@ void udf_clear_inode(struct inode *inode) udf_truncate_tail_extent(inode); unlock_kernel(); write_inode_now(inode, 0); + invalidate_inode_buffers(inode); } iinfo = UDF_I(inode); kfree(iinfo->i_ext.i_data); diff --git a/fs/udf/namei.c b/fs/udf/namei.c index d3231947db1..f84bfaa8d94 100644 --- a/fs/udf/namei.c +++ b/fs/udf/namei.c @@ -142,7 +142,7 @@ int udf_write_fi(struct inode *inode, struct fileIdentDesc *cfi, } static struct fileIdentDesc *udf_find_entry(struct inode *dir, - struct dentry *dentry, + struct qstr *child, struct udf_fileident_bh *fibh, struct fileIdentDesc *cfi) { @@ -159,8 +159,8 @@ static struct fileIdentDesc *udf_find_entry(struct inode *dir, sector_t offset; struct extent_position epos = {}; struct udf_inode_info *dinfo = UDF_I(dir); - int isdotdot = dentry->d_name.len == 2 && - dentry->d_name.name[0] == '.' && dentry->d_name.name[1] == '.'; + int isdotdot = child->len == 2 && + child->name[0] == '.' && child->name[1] == '.'; size = udf_ext0_offset(dir) + dir->i_size; f_pos = udf_ext0_offset(dir); @@ -238,8 +238,7 @@ static struct fileIdentDesc *udf_find_entry(struct inode *dir, continue; flen = udf_get_filename(dir->i_sb, nameptr, fname, lfi); - if (flen && udf_match(flen, fname, dentry->d_name.len, - dentry->d_name.name)) + if (flen && udf_match(flen, fname, child->len, child->name)) goto out_ok; } @@ -283,7 +282,7 @@ static struct dentry *udf_lookup(struct inode *dir, struct dentry *dentry, } else #endif /* UDF_RECOVERY */ - if (udf_find_entry(dir, dentry, &fibh, &cfi)) { + if (udf_find_entry(dir, &dentry->d_name, &fibh, &cfi)) { if (fibh.sbh != fibh.ebh) brelse(fibh.ebh); brelse(fibh.sbh); @@ -605,7 +604,7 @@ static int udf_mknod(struct inode *dir, struct dentry *dentry, int mode, goto out; iinfo = UDF_I(inode); - inode->i_uid = current->fsuid; + inode->i_uid = current_fsuid(); init_special_inode(inode, mode, rdev); fi = udf_add_entry(dir, dentry, &fibh, &cfi, &err); if (!fi) { @@ -783,7 +782,7 @@ static int udf_rmdir(struct inode *dir, struct dentry *dentry) retval = -ENOENT; lock_kernel(); - fi = udf_find_entry(dir, dentry, &fibh, &cfi); + fi = udf_find_entry(dir, &dentry->d_name, &fibh, &cfi); if (!fi) goto out; @@ -829,7 +828,7 @@ static int udf_unlink(struct inode *dir, struct dentry *dentry) retval = -ENOENT; lock_kernel(); - fi = udf_find_entry(dir, dentry, &fibh, &cfi); + fi = udf_find_entry(dir, &dentry->d_name, &fibh, &cfi); if (!fi) goto out; @@ -1113,7 +1112,7 @@ static int udf_rename(struct inode *old_dir, struct dentry *old_dentry, struct udf_inode_info *old_iinfo = UDF_I(old_inode); lock_kernel(); - ofi = udf_find_entry(old_dir, old_dentry, &ofibh, &ocfi); + ofi = udf_find_entry(old_dir, &old_dentry->d_name, &ofibh, &ocfi); if (ofi) { if (ofibh.sbh != ofibh.ebh) brelse(ofibh.ebh); @@ -1124,7 +1123,7 @@ static int udf_rename(struct inode *old_dir, struct dentry *old_dentry, != old_inode->i_ino) goto end_rename; - nfi = udf_find_entry(new_dir, new_dentry, &nfibh, &ncfi); + nfi = udf_find_entry(new_dir, &new_dentry->d_name, &nfibh, &ncfi); if (nfi) { if (!new_inode) { if (nfibh.sbh != nfibh.ebh) @@ -1192,7 +1191,7 @@ static int udf_rename(struct inode *old_dir, struct dentry *old_dentry, udf_write_fi(new_dir, &ncfi, nfi, &nfibh, NULL, NULL); /* The old fid may have moved - find it again */ - ofi = udf_find_entry(old_dir, old_dentry, &ofibh, &ocfi); + ofi = udf_find_entry(old_dir, &old_dentry->d_name, &ofibh, &ocfi); udf_delete_entry(old_dir, ofi, &ofibh, &ocfi); if (new_inode) { @@ -1243,15 +1242,11 @@ end_rename: static struct dentry *udf_get_parent(struct dentry *child) { - struct dentry *parent; struct inode *inode = NULL; - struct dentry dotdot; + struct qstr dotdot = {.name = "..", .len = 2}; struct fileIdentDesc cfi; struct udf_fileident_bh fibh; - dotdot.d_name.name = ".."; - dotdot.d_name.len = 2; - lock_kernel(); if (!udf_find_entry(child->d_inode, &dotdot, &fibh, &cfi)) goto out_unlock; @@ -1266,13 +1261,7 @@ static struct dentry *udf_get_parent(struct dentry *child) goto out_unlock; unlock_kernel(); - parent = d_alloc_anon(inode); - if (!parent) { - iput(inode); - parent = ERR_PTR(-ENOMEM); - } - - return parent; + return d_obtain_alias(inode); out_unlock: unlock_kernel(); return ERR_PTR(-EACCES); @@ -1283,7 +1272,6 @@ static struct dentry *udf_nfs_get_inode(struct super_block *sb, u32 block, u16 partref, __u32 generation) { struct inode *inode; - struct dentry *result; kernel_lb_addr loc; if (block == 0) @@ -1300,12 +1288,7 @@ static struct dentry *udf_nfs_get_inode(struct super_block *sb, u32 block, iput(inode); return ERR_PTR(-ESTALE); } - result = d_alloc_anon(inode); - if (!result) { - iput(inode); - return ERR_PTR(-ENOMEM); - } - return result; + return d_obtain_alias(inode); } static struct dentry *udf_fh_to_dentry(struct super_block *sb, diff --git a/fs/ufs/dir.c b/fs/ufs/dir.c index df0bef18742..dbbbc466876 100644 --- a/fs/ufs/dir.c +++ b/fs/ufs/dir.c @@ -667,4 +667,5 @@ const struct file_operations ufs_dir_operations = { .read = generic_read_dir, .readdir = ufs_readdir, .fsync = file_fsync, + .llseek = generic_file_llseek, }; diff --git a/fs/ufs/ialloc.c b/fs/ufs/ialloc.c index ac181f6806a..6f5dcf00609 100644 --- a/fs/ufs/ialloc.c +++ b/fs/ufs/ialloc.c @@ -304,13 +304,13 @@ cg_found: inode->i_ino = cg * uspi->s_ipg + bit; inode->i_mode = mode; - inode->i_uid = current->fsuid; + inode->i_uid = current_fsuid(); if (dir->i_mode & S_ISGID) { inode->i_gid = dir->i_gid; if (S_ISDIR(mode)) inode->i_mode |= S_ISGID; } else - inode->i_gid = current->fsgid; + inode->i_gid = current_fsgid(); inode->i_blocks = 0; inode->i_generation = 0; diff --git a/fs/vfat/Makefile b/fs/vfat/Makefile deleted file mode 100644 index 40f2798a4f0..00000000000 --- a/fs/vfat/Makefile +++ /dev/null @@ -1,7 +0,0 @@ -# -# Makefile for the linux vfat-filesystem routines. -# - -obj-$(CONFIG_VFAT_FS) += vfat.o - -vfat-y := namei.o diff --git a/fs/xfs/Makefile b/fs/xfs/Makefile index 737c9a42536..c3dc491fff8 100644 --- a/fs/xfs/Makefile +++ b/fs/xfs/Makefile @@ -85,13 +85,13 @@ xfs-y += xfs_alloc.o \ xfs_trans_inode.o \ xfs_trans_item.o \ xfs_utils.o \ - xfs_vfsops.o \ xfs_vnodeops.o \ xfs_rw.o \ xfs_dmops.o \ xfs_qmops.o -xfs-$(CONFIG_XFS_TRACE) += xfs_dir2_trace.o +xfs-$(CONFIG_XFS_TRACE) += xfs_btree_trace.o \ + xfs_dir2_trace.o # Objects in linux/ xfs-y += $(addprefix $(XFS_LINUX)/, \ @@ -106,7 +106,7 @@ xfs-y += $(addprefix $(XFS_LINUX)/, \ xfs_iops.o \ xfs_lrw.o \ xfs_super.o \ - xfs_vnode.o \ + xfs_sync.o \ xfs_xattr.o) # Objects in support/ diff --git a/fs/xfs/linux-2.6/sv.h b/fs/xfs/linux-2.6/sv.h index 351a8f454bd..4dfc7c37081 100644 --- a/fs/xfs/linux-2.6/sv.h +++ b/fs/xfs/linux-2.6/sv.h @@ -32,23 +32,15 @@ typedef struct sv_s { wait_queue_head_t waiters; } sv_t; -#define SV_FIFO 0x0 /* sv_t is FIFO type */ -#define SV_LIFO 0x2 /* sv_t is LIFO type */ -#define SV_PRIO 0x4 /* sv_t is PRIO type */ -#define SV_KEYED 0x6 /* sv_t is KEYED type */ -#define SV_DEFAULT SV_FIFO - - -static inline void _sv_wait(sv_t *sv, spinlock_t *lock, int state, - unsigned long timeout) +static inline void _sv_wait(sv_t *sv, spinlock_t *lock) { DECLARE_WAITQUEUE(wait, current); add_wait_queue_exclusive(&sv->waiters, &wait); - __set_current_state(state); + __set_current_state(TASK_UNINTERRUPTIBLE); spin_unlock(lock); - schedule_timeout(timeout); + schedule(); remove_wait_queue(&sv->waiters, &wait); } @@ -58,13 +50,7 @@ static inline void _sv_wait(sv_t *sv, spinlock_t *lock, int state, #define sv_destroy(sv) \ /*NOTHING*/ #define sv_wait(sv, pri, lock, s) \ - _sv_wait(sv, lock, TASK_UNINTERRUPTIBLE, MAX_SCHEDULE_TIMEOUT) -#define sv_wait_sig(sv, pri, lock, s) \ - _sv_wait(sv, lock, TASK_INTERRUPTIBLE, MAX_SCHEDULE_TIMEOUT) -#define sv_timedwait(sv, pri, lock, s, svf, ts, rts) \ - _sv_wait(sv, lock, TASK_UNINTERRUPTIBLE, timespec_to_jiffies(ts)) -#define sv_timedwait_sig(sv, pri, lock, s, svf, ts, rts) \ - _sv_wait(sv, lock, TASK_INTERRUPTIBLE, timespec_to_jiffies(ts)) + _sv_wait(sv, lock) #define sv_signal(sv) \ wake_up(&(sv)->waiters) #define sv_broadcast(sv) \ diff --git a/fs/xfs/linux-2.6/xfs_aops.c b/fs/xfs/linux-2.6/xfs_aops.c index a44d68eb50b..de3a198f771 100644 --- a/fs/xfs/linux-2.6/xfs_aops.c +++ b/fs/xfs/linux-2.6/xfs_aops.c @@ -42,6 +42,40 @@ #include <linux/pagevec.h> #include <linux/writeback.h> + +/* + * Prime number of hash buckets since address is used as the key. + */ +#define NVSYNC 37 +#define to_ioend_wq(v) (&xfs_ioend_wq[((unsigned long)v) % NVSYNC]) +static wait_queue_head_t xfs_ioend_wq[NVSYNC]; + +void __init +xfs_ioend_init(void) +{ + int i; + + for (i = 0; i < NVSYNC; i++) + init_waitqueue_head(&xfs_ioend_wq[i]); +} + +void +xfs_ioend_wait( + xfs_inode_t *ip) +{ + wait_queue_head_t *wq = to_ioend_wq(ip); + + wait_event(*wq, (atomic_read(&ip->i_iocount) == 0)); +} + +STATIC void +xfs_ioend_wake( + xfs_inode_t *ip) +{ + if (atomic_dec_and_test(&ip->i_iocount)) + wake_up(to_ioend_wq(ip)); +} + STATIC void xfs_count_page_state( struct page *page, @@ -146,16 +180,25 @@ xfs_destroy_ioend( xfs_ioend_t *ioend) { struct buffer_head *bh, *next; + struct xfs_inode *ip = XFS_I(ioend->io_inode); for (bh = ioend->io_buffer_head; bh; bh = next) { next = bh->b_private; bh->b_end_io(bh, !ioend->io_error); } - if (unlikely(ioend->io_error)) { - vn_ioerror(XFS_I(ioend->io_inode), ioend->io_error, - __FILE__,__LINE__); + + /* + * Volume managers supporting multiple paths can send back ENODEV + * when the final path disappears. In this case continuing to fill + * the page cache with dirty data which cannot be written out is + * evil, so prevent that. + */ + if (unlikely(ioend->io_error == -ENODEV)) { + xfs_do_force_shutdown(ip->i_mount, SHUTDOWN_DEVICE_REQ, + __FILE__, __LINE__); } - vn_iowake(XFS_I(ioend->io_inode)); + + xfs_ioend_wake(ip); mempool_free(ioend, xfs_ioend_pool); } @@ -191,7 +234,7 @@ xfs_setfilesize( ip->i_d.di_size = isize; ip->i_update_core = 1; ip->i_update_size = 1; - mark_inode_dirty_sync(ioend->io_inode); + xfs_mark_inode_dirty_sync(ip); } xfs_iunlock(ip, XFS_ILOCK_EXCL); @@ -317,14 +360,9 @@ xfs_map_blocks( xfs_iomap_t *mapp, int flags) { - xfs_inode_t *ip = XFS_I(inode); - int error, nmaps = 1; - - error = xfs_iomap(ip, offset, count, - flags, mapp, &nmaps); - if (!error && (flags & (BMAPI_WRITE|BMAPI_ALLOCATE))) - xfs_iflags_set(ip, XFS_IMODIFIED); - return -error; + int nmaps = 1; + + return -xfs_iomap(XFS_I(inode), offset, count, flags, mapp, &nmaps); } STATIC_INLINE int @@ -512,7 +550,7 @@ xfs_cancel_ioend( unlock_buffer(bh); } while ((bh = next_bh) != NULL); - vn_iowake(XFS_I(ioend->io_inode)); + xfs_ioend_wake(XFS_I(ioend->io_inode)); mempool_free(ioend, xfs_ioend_pool); } while ((ioend = next) != NULL); } diff --git a/fs/xfs/linux-2.6/xfs_aops.h b/fs/xfs/linux-2.6/xfs_aops.h index 3ba0631a381..7b26f5ff969 100644 --- a/fs/xfs/linux-2.6/xfs_aops.h +++ b/fs/xfs/linux-2.6/xfs_aops.h @@ -43,4 +43,7 @@ typedef struct xfs_ioend { extern const struct address_space_operations xfs_address_space_operations; extern int xfs_get_blocks(struct inode *, sector_t, struct buffer_head *, int); +extern void xfs_ioend_init(void); +extern void xfs_ioend_wait(struct xfs_inode *); + #endif /* __XFS_AOPS_H__ */ diff --git a/fs/xfs/linux-2.6/xfs_buf.c b/fs/xfs/linux-2.6/xfs_buf.c index 36d5fcd3f59..cb329edc925 100644 --- a/fs/xfs/linux-2.6/xfs_buf.c +++ b/fs/xfs/linux-2.6/xfs_buf.c @@ -630,6 +630,29 @@ xfs_buf_get_flags( return NULL; } +STATIC int +_xfs_buf_read( + xfs_buf_t *bp, + xfs_buf_flags_t flags) +{ + int status; + + XB_TRACE(bp, "_xfs_buf_read", (unsigned long)flags); + + ASSERT(!(flags & (XBF_DELWRI|XBF_WRITE))); + ASSERT(bp->b_bn != XFS_BUF_DADDR_NULL); + + bp->b_flags &= ~(XBF_WRITE | XBF_ASYNC | XBF_DELWRI | \ + XBF_READ_AHEAD | _XBF_RUN_QUEUES); + bp->b_flags |= flags & (XBF_READ | XBF_ASYNC | \ + XBF_READ_AHEAD | _XBF_RUN_QUEUES); + + status = xfs_buf_iorequest(bp); + if (!status && !(flags & XBF_ASYNC)) + status = xfs_buf_iowait(bp); + return status; +} + xfs_buf_t * xfs_buf_read_flags( xfs_buftarg_t *target, @@ -646,7 +669,7 @@ xfs_buf_read_flags( if (!XFS_BUF_ISDONE(bp)) { XB_TRACE(bp, "read", (unsigned long)flags); XFS_STATS_INC(xb_get_read); - xfs_buf_iostart(bp, flags); + _xfs_buf_read(bp, flags); } else if (flags & XBF_ASYNC) { XB_TRACE(bp, "read_async", (unsigned long)flags); /* @@ -1048,50 +1071,39 @@ xfs_buf_ioerror( XB_TRACE(bp, "ioerror", (unsigned long)error); } -/* - * Initiate I/O on a buffer, based on the flags supplied. - * The b_iodone routine in the buffer supplied will only be called - * when all of the subsidiary I/O requests, if any, have been completed. - */ int -xfs_buf_iostart( - xfs_buf_t *bp, - xfs_buf_flags_t flags) +xfs_bawrite( + void *mp, + struct xfs_buf *bp) { - int status = 0; + XB_TRACE(bp, "bawrite", 0); - XB_TRACE(bp, "iostart", (unsigned long)flags); + ASSERT(bp->b_bn != XFS_BUF_DADDR_NULL); - if (flags & XBF_DELWRI) { - bp->b_flags &= ~(XBF_READ | XBF_WRITE | XBF_ASYNC); - bp->b_flags |= flags & (XBF_DELWRI | XBF_ASYNC); - xfs_buf_delwri_queue(bp, 1); - return 0; - } + xfs_buf_delwri_dequeue(bp); - bp->b_flags &= ~(XBF_READ | XBF_WRITE | XBF_ASYNC | XBF_DELWRI | \ - XBF_READ_AHEAD | _XBF_RUN_QUEUES); - bp->b_flags |= flags & (XBF_READ | XBF_WRITE | XBF_ASYNC | \ - XBF_READ_AHEAD | _XBF_RUN_QUEUES); + bp->b_flags &= ~(XBF_READ | XBF_DELWRI | XBF_READ_AHEAD); + bp->b_flags |= (XBF_WRITE | XBF_ASYNC | _XBF_RUN_QUEUES); + + bp->b_mount = mp; + bp->b_strat = xfs_bdstrat_cb; + return xfs_bdstrat_cb(bp); +} - BUG_ON(bp->b_bn == XFS_BUF_DADDR_NULL); +void +xfs_bdwrite( + void *mp, + struct xfs_buf *bp) +{ + XB_TRACE(bp, "bdwrite", 0); - /* For writes allow an alternate strategy routine to precede - * the actual I/O request (which may not be issued at all in - * a shutdown situation, for example). - */ - status = (flags & XBF_WRITE) ? - xfs_buf_iostrategy(bp) : xfs_buf_iorequest(bp); + bp->b_strat = xfs_bdstrat_cb; + bp->b_mount = mp; - /* Wait for I/O if we are not an async request. - * Note: async I/O request completion will release the buffer, - * and that can already be done by this point. So using the - * buffer pointer from here on, after async I/O, is invalid. - */ - if (!status && !(flags & XBF_ASYNC)) - status = xfs_buf_iowait(bp); + bp->b_flags &= ~XBF_READ; + bp->b_flags |= (XBF_DELWRI | XBF_ASYNC); - return status; + xfs_buf_delwri_queue(bp, 1); } STATIC_INLINE void @@ -1114,8 +1126,7 @@ xfs_buf_bio_end_io( unsigned int blocksize = bp->b_target->bt_bsize; struct bio_vec *bvec = bio->bi_io_vec + bio->bi_vcnt - 1; - if (!test_bit(BIO_UPTODATE, &bio->bi_flags)) - bp->b_error = EIO; + xfs_buf_ioerror(bp, -error); do { struct page *page = bvec->bv_page; diff --git a/fs/xfs/linux-2.6/xfs_buf.h b/fs/xfs/linux-2.6/xfs_buf.h index 456519a088c..288ae7c4c80 100644 --- a/fs/xfs/linux-2.6/xfs_buf.h +++ b/fs/xfs/linux-2.6/xfs_buf.h @@ -168,7 +168,7 @@ typedef struct xfs_buf { struct completion b_iowait; /* queue for I/O waiters */ void *b_fspriv; void *b_fspriv2; - void *b_fspriv3; + struct xfs_mount *b_mount; unsigned short b_error; /* error code on I/O */ unsigned int b_page_count; /* size of page array */ unsigned int b_offset; /* page offset in first page */ @@ -214,9 +214,10 @@ extern void xfs_buf_lock(xfs_buf_t *); extern void xfs_buf_unlock(xfs_buf_t *); /* Buffer Read and Write Routines */ +extern int xfs_bawrite(void *mp, xfs_buf_t *bp); +extern void xfs_bdwrite(void *mp, xfs_buf_t *bp); extern void xfs_buf_ioend(xfs_buf_t *, int); extern void xfs_buf_ioerror(xfs_buf_t *, int); -extern int xfs_buf_iostart(xfs_buf_t *, xfs_buf_flags_t); extern int xfs_buf_iorequest(xfs_buf_t *); extern int xfs_buf_iowait(xfs_buf_t *); extern void xfs_buf_iomove(xfs_buf_t *, size_t, size_t, xfs_caddr_t, @@ -311,10 +312,6 @@ extern void xfs_buf_trace(xfs_buf_t *, char *, void *, void *); #define XFS_BUF_UNORDERED(bp) ((bp)->b_flags &= ~XBF_ORDERED) #define XFS_BUF_ISORDERED(bp) ((bp)->b_flags & XBF_ORDERED) -#define XFS_BUF_SHUT(bp) do { } while (0) -#define XFS_BUF_UNSHUT(bp) do { } while (0) -#define XFS_BUF_ISSHUT(bp) (0) - #define XFS_BUF_HOLD(bp) xfs_buf_hold(bp) #define XFS_BUF_READ(bp) ((bp)->b_flags |= XBF_READ) #define XFS_BUF_UNREAD(bp) ((bp)->b_flags &= ~XBF_READ) @@ -334,8 +331,6 @@ extern void xfs_buf_trace(xfs_buf_t *, char *, void *, void *); #define XFS_BUF_SET_FSPRIVATE(bp, val) ((bp)->b_fspriv = (void*)(val)) #define XFS_BUF_FSPRIVATE2(bp, type) ((type)(bp)->b_fspriv2) #define XFS_BUF_SET_FSPRIVATE2(bp, val) ((bp)->b_fspriv2 = (void*)(val)) -#define XFS_BUF_FSPRIVATE3(bp, type) ((type)(bp)->b_fspriv3) -#define XFS_BUF_SET_FSPRIVATE3(bp, val) ((bp)->b_fspriv3 = (void*)(val)) #define XFS_BUF_SET_START(bp) do { } while (0) #define XFS_BUF_SET_BRELSE_FUNC(bp, func) ((bp)->b_relse = (func)) @@ -366,14 +361,6 @@ extern void xfs_buf_trace(xfs_buf_t *, char *, void *, void *); #define XFS_BUF_TARGET(bp) ((bp)->b_target) #define XFS_BUFTARG_NAME(target) xfs_buf_target_name(target) -static inline int xfs_bawrite(void *mp, xfs_buf_t *bp) -{ - bp->b_fspriv3 = mp; - bp->b_strat = xfs_bdstrat_cb; - xfs_buf_delwri_dequeue(bp); - return xfs_buf_iostart(bp, XBF_WRITE | XBF_ASYNC | _XBF_RUN_QUEUES); -} - static inline void xfs_buf_relse(xfs_buf_t *bp) { if (!bp->b_relse) @@ -414,17 +401,6 @@ static inline int XFS_bwrite(xfs_buf_t *bp) return error; } -/* - * No error can be returned from xfs_buf_iostart for delwri - * buffers as they are queued and no I/O is issued. - */ -static inline void xfs_bdwrite(void *mp, xfs_buf_t *bp) -{ - bp->b_strat = xfs_bdstrat_cb; - bp->b_fspriv3 = mp; - (void)xfs_buf_iostart(bp, XBF_DELWRI | XBF_ASYNC); -} - #define XFS_bdstrat(bp) xfs_buf_iorequest(bp) #define xfs_iowait(bp) xfs_buf_iowait(bp) diff --git a/fs/xfs/linux-2.6/xfs_cred.h b/fs/xfs/linux-2.6/xfs_cred.h index 652721ce0ea..55bddf3b609 100644 --- a/fs/xfs/linux-2.6/xfs_cred.h +++ b/fs/xfs/linux-2.6/xfs_cred.h @@ -23,16 +23,6 @@ /* * Credentials */ -typedef struct cred { - /* EMPTY */ -} cred_t; - -extern struct cred *sys_cred; - -/* this is a hack.. (assumes sys_cred is the only cred_t in the system) */ -static inline int capable_cred(cred_t *cr, int cid) -{ - return (cr == sys_cred) ? 1 : capable(cid); -} +typedef const struct cred cred_t; #endif /* __XFS_CRED_H__ */ diff --git a/fs/xfs/linux-2.6/xfs_export.c b/fs/xfs/linux-2.6/xfs_export.c index 24fd598af84..595751f7835 100644 --- a/fs/xfs/linux-2.6/xfs_export.c +++ b/fs/xfs/linux-2.6/xfs_export.c @@ -29,7 +29,6 @@ #include "xfs_vnodeops.h" #include "xfs_bmap_btree.h" #include "xfs_inode.h" -#include "xfs_vfsops.h" /* * Note that we only accept fileids which are long enough rather than allow @@ -148,7 +147,6 @@ xfs_fs_fh_to_dentry(struct super_block *sb, struct fid *fid, { struct xfs_fid64 *fid64 = (struct xfs_fid64 *)fid; struct inode *inode = NULL; - struct dentry *result; if (fh_len < xfs_fileid_length(fileid_type)) return NULL; @@ -164,16 +162,7 @@ xfs_fs_fh_to_dentry(struct super_block *sb, struct fid *fid, break; } - if (!inode) - return NULL; - if (IS_ERR(inode)) - return ERR_CAST(inode); - result = d_alloc_anon(inode); - if (!result) { - iput(inode); - return ERR_PTR(-ENOMEM); - } - return result; + return d_obtain_alias(inode); } STATIC struct dentry * @@ -182,7 +171,6 @@ xfs_fs_fh_to_parent(struct super_block *sb, struct fid *fid, { struct xfs_fid64 *fid64 = (struct xfs_fid64 *)fid; struct inode *inode = NULL; - struct dentry *result; switch (fileid_type) { case FILEID_INO32_GEN_PARENT: @@ -195,16 +183,7 @@ xfs_fs_fh_to_parent(struct super_block *sb, struct fid *fid, break; } - if (!inode) - return NULL; - if (IS_ERR(inode)) - return ERR_CAST(inode); - result = d_alloc_anon(inode); - if (!result) { - iput(inode); - return ERR_PTR(-ENOMEM); - } - return result; + return d_obtain_alias(inode); } STATIC struct dentry * @@ -213,18 +192,12 @@ xfs_fs_get_parent( { int error; struct xfs_inode *cip; - struct dentry *parent; error = xfs_lookup(XFS_I(child->d_inode), &xfs_name_dotdot, &cip, NULL); if (unlikely(error)) return ERR_PTR(-error); - parent = d_alloc_anon(VFS_I(cip)); - if (unlikely(!parent)) { - iput(VFS_I(cip)); - return ERR_PTR(-ENOMEM); - } - return parent; + return d_obtain_alias(VFS_I(cip)); } const struct export_operations xfs_export_operations = { diff --git a/fs/xfs/linux-2.6/xfs_file.c b/fs/xfs/linux-2.6/xfs_file.c index 5311c1acdd4..e14c4e3aea0 100644 --- a/fs/xfs/linux-2.6/xfs_file.c +++ b/fs/xfs/linux-2.6/xfs_file.c @@ -36,89 +36,54 @@ #include "xfs_inode.h" #include "xfs_error.h" #include "xfs_rw.h" -#include "xfs_ioctl32.h" #include "xfs_vnodeops.h" +#include "xfs_da_btree.h" +#include "xfs_ioctl.h" #include <linux/dcache.h> #include <linux/smp_lock.h> static struct vm_operations_struct xfs_file_vm_ops; -STATIC_INLINE ssize_t -__xfs_file_read( +STATIC ssize_t +xfs_file_aio_read( struct kiocb *iocb, const struct iovec *iov, unsigned long nr_segs, - int ioflags, loff_t pos) { struct file *file = iocb->ki_filp; + int ioflags = IO_ISAIO; BUG_ON(iocb->ki_pos != pos); if (unlikely(file->f_flags & O_DIRECT)) ioflags |= IO_ISDIRECT; + if (file->f_mode & FMODE_NOCMTIME) + ioflags |= IO_INVIS; return xfs_read(XFS_I(file->f_path.dentry->d_inode), iocb, iov, nr_segs, &iocb->ki_pos, ioflags); } STATIC ssize_t -xfs_file_aio_read( - struct kiocb *iocb, - const struct iovec *iov, - unsigned long nr_segs, - loff_t pos) -{ - return __xfs_file_read(iocb, iov, nr_segs, IO_ISAIO, pos); -} - -STATIC ssize_t -xfs_file_aio_read_invis( - struct kiocb *iocb, - const struct iovec *iov, - unsigned long nr_segs, - loff_t pos) -{ - return __xfs_file_read(iocb, iov, nr_segs, IO_ISAIO|IO_INVIS, pos); -} - -STATIC_INLINE ssize_t -__xfs_file_write( +xfs_file_aio_write( struct kiocb *iocb, const struct iovec *iov, unsigned long nr_segs, - int ioflags, loff_t pos) { - struct file *file = iocb->ki_filp; + struct file *file = iocb->ki_filp; + int ioflags = IO_ISAIO; BUG_ON(iocb->ki_pos != pos); if (unlikely(file->f_flags & O_DIRECT)) ioflags |= IO_ISDIRECT; + if (file->f_mode & FMODE_NOCMTIME) + ioflags |= IO_INVIS; return xfs_write(XFS_I(file->f_mapping->host), iocb, iov, nr_segs, &iocb->ki_pos, ioflags); } STATIC ssize_t -xfs_file_aio_write( - struct kiocb *iocb, - const struct iovec *iov, - unsigned long nr_segs, - loff_t pos) -{ - return __xfs_file_write(iocb, iov, nr_segs, IO_ISAIO, pos); -} - -STATIC ssize_t -xfs_file_aio_write_invis( - struct kiocb *iocb, - const struct iovec *iov, - unsigned long nr_segs, - loff_t pos) -{ - return __xfs_file_write(iocb, iov, nr_segs, IO_ISAIO|IO_INVIS, pos); -} - -STATIC ssize_t xfs_file_splice_read( struct file *infilp, loff_t *ppos, @@ -126,20 +91,13 @@ xfs_file_splice_read( size_t len, unsigned int flags) { - return xfs_splice_read(XFS_I(infilp->f_path.dentry->d_inode), - infilp, ppos, pipe, len, flags, 0); -} + int ioflags = 0; + + if (infilp->f_mode & FMODE_NOCMTIME) + ioflags |= IO_INVIS; -STATIC ssize_t -xfs_file_splice_read_invis( - struct file *infilp, - loff_t *ppos, - struct pipe_inode_info *pipe, - size_t len, - unsigned int flags) -{ return xfs_splice_read(XFS_I(infilp->f_path.dentry->d_inode), - infilp, ppos, pipe, len, flags, IO_INVIS); + infilp, ppos, pipe, len, flags, ioflags); } STATIC ssize_t @@ -150,30 +108,49 @@ xfs_file_splice_write( size_t len, unsigned int flags) { - return xfs_splice_write(XFS_I(outfilp->f_path.dentry->d_inode), - pipe, outfilp, ppos, len, flags, 0); -} + int ioflags = 0; + + if (outfilp->f_mode & FMODE_NOCMTIME) + ioflags |= IO_INVIS; -STATIC ssize_t -xfs_file_splice_write_invis( - struct pipe_inode_info *pipe, - struct file *outfilp, - loff_t *ppos, - size_t len, - unsigned int flags) -{ return xfs_splice_write(XFS_I(outfilp->f_path.dentry->d_inode), - pipe, outfilp, ppos, len, flags, IO_INVIS); + pipe, outfilp, ppos, len, flags, ioflags); } STATIC int xfs_file_open( struct inode *inode, - struct file *filp) + struct file *file) { - if (!(filp->f_flags & O_LARGEFILE) && i_size_read(inode) > MAX_NON_LFS) + if (!(file->f_flags & O_LARGEFILE) && i_size_read(inode) > MAX_NON_LFS) return -EFBIG; - return -xfs_open(XFS_I(inode)); + if (XFS_FORCED_SHUTDOWN(XFS_M(inode->i_sb))) + return -EIO; + return 0; +} + +STATIC int +xfs_dir_open( + struct inode *inode, + struct file *file) +{ + struct xfs_inode *ip = XFS_I(inode); + int mode; + int error; + + error = xfs_file_open(inode, file); + if (error) + return error; + + /* + * If there are any blocks, read-ahead block 0 as we're almost + * certain to have the next operation be a read there. + */ + mode = xfs_ilock_map_shared(ip); + if (ip->i_d.di_nextents > 0) + xfs_da_reada_buf(NULL, ip, 0, XFS_DATA_FORK); + xfs_iunlock(ip, mode); + return 0; } STATIC int @@ -204,15 +181,6 @@ xfs_file_fsync( return -xfs_fsync(XFS_I(dentry->d_inode)); } -/* - * Unfortunately we can't just use the clean and simple readdir implementation - * below, because nfs might call back into ->lookup from the filldir callback - * and that will deadlock the low-level btree code. - * - * Hopefully we'll find a better workaround that allows to use the optimal - * version at least for local readdirs for 2.6.25. - */ -#if 0 STATIC int xfs_file_readdir( struct file *filp, @@ -236,7 +204,7 @@ xfs_file_readdir( * point we can change the ->readdir prototype to include the * buffer size. */ - bufsize = (size_t)min_t(loff_t, PAGE_SIZE, inode->i_size); + bufsize = (size_t)min_t(loff_t, PAGE_SIZE, ip->i_d.di_size); error = xfs_readdir(ip, dirent, bufsize, (xfs_off_t *)&filp->f_pos, filldir); @@ -244,125 +212,6 @@ xfs_file_readdir( return -error; return 0; } -#else - -struct hack_dirent { - u64 ino; - loff_t offset; - int namlen; - unsigned int d_type; - char name[]; -}; - -struct hack_callback { - char *dirent; - size_t len; - size_t used; -}; - -STATIC int -xfs_hack_filldir( - void *__buf, - const char *name, - int namlen, - loff_t offset, - u64 ino, - unsigned int d_type) -{ - struct hack_callback *buf = __buf; - struct hack_dirent *de = (struct hack_dirent *)(buf->dirent + buf->used); - unsigned int reclen; - - reclen = ALIGN(sizeof(struct hack_dirent) + namlen, sizeof(u64)); - if (buf->used + reclen > buf->len) - return -EINVAL; - - de->namlen = namlen; - de->offset = offset; - de->ino = ino; - de->d_type = d_type; - memcpy(de->name, name, namlen); - buf->used += reclen; - return 0; -} - -STATIC int -xfs_file_readdir( - struct file *filp, - void *dirent, - filldir_t filldir) -{ - struct inode *inode = filp->f_path.dentry->d_inode; - xfs_inode_t *ip = XFS_I(inode); - struct hack_callback buf; - struct hack_dirent *de; - int error; - loff_t size; - int eof = 0; - xfs_off_t start_offset, curr_offset, offset; - - /* - * Try fairly hard to get memory - */ - buf.len = PAGE_CACHE_SIZE; - do { - buf.dirent = kmalloc(buf.len, GFP_KERNEL); - if (buf.dirent) - break; - buf.len >>= 1; - } while (buf.len >= 1024); - - if (!buf.dirent) - return -ENOMEM; - - curr_offset = filp->f_pos; - if (curr_offset == 0x7fffffff) - offset = 0xffffffff; - else - offset = filp->f_pos; - - while (!eof) { - unsigned int reclen; - - start_offset = offset; - - buf.used = 0; - error = -xfs_readdir(ip, &buf, buf.len, &offset, - xfs_hack_filldir); - if (error || offset == start_offset) { - size = 0; - break; - } - - size = buf.used; - de = (struct hack_dirent *)buf.dirent; - while (size > 0) { - curr_offset = de->offset /* & 0x7fffffff */; - if (filldir(dirent, de->name, de->namlen, - curr_offset & 0x7fffffff, - de->ino, de->d_type)) { - goto done; - } - - reclen = ALIGN(sizeof(struct hack_dirent) + de->namlen, - sizeof(u64)); - size -= reclen; - de = (struct hack_dirent *)((char *)de + reclen); - } - } - - done: - if (!error) { - if (size == 0) - filp->f_pos = offset & 0x7fffffff; - else if (de) - filp->f_pos = curr_offset; - } - - kfree(buf.dirent); - return error; -} -#endif STATIC int xfs_file_mmap( @@ -376,48 +225,6 @@ xfs_file_mmap( return 0; } -STATIC long -xfs_file_ioctl( - struct file *filp, - unsigned int cmd, - unsigned long p) -{ - int error; - struct inode *inode = filp->f_path.dentry->d_inode; - - error = xfs_ioctl(XFS_I(inode), filp, 0, cmd, (void __user *)p); - xfs_iflags_set(XFS_I(inode), XFS_IMODIFIED); - - /* NOTE: some of the ioctl's return positive #'s as a - * byte count indicating success, such as - * readlink_by_handle. So we don't "sign flip" - * like most other routines. This means true - * errors need to be returned as a negative value. - */ - return error; -} - -STATIC long -xfs_file_ioctl_invis( - struct file *filp, - unsigned int cmd, - unsigned long p) -{ - int error; - struct inode *inode = filp->f_path.dentry->d_inode; - - error = xfs_ioctl(XFS_I(inode), filp, IO_INVIS, cmd, (void __user *)p); - xfs_iflags_set(XFS_I(inode), XFS_IMODIFIED); - - /* NOTE: some of the ioctl's return positive #'s as a - * byte count indicating success, such as - * readlink_by_handle. So we don't "sign flip" - * like most other routines. This means true - * errors need to be returned as a negative value. - */ - return error; -} - /* * mmap()d file has taken write protection fault and is being made * writable. We can set the page state up correctly for a writable @@ -453,26 +260,8 @@ const struct file_operations xfs_file_operations = { #endif }; -const struct file_operations xfs_invis_file_operations = { - .llseek = generic_file_llseek, - .read = do_sync_read, - .write = do_sync_write, - .aio_read = xfs_file_aio_read_invis, - .aio_write = xfs_file_aio_write_invis, - .splice_read = xfs_file_splice_read_invis, - .splice_write = xfs_file_splice_write_invis, - .unlocked_ioctl = xfs_file_ioctl_invis, -#ifdef CONFIG_COMPAT - .compat_ioctl = xfs_file_compat_invis_ioctl, -#endif - .mmap = xfs_file_mmap, - .open = xfs_file_open, - .release = xfs_file_release, - .fsync = xfs_file_fsync, -}; - - const struct file_operations xfs_dir_file_operations = { + .open = xfs_dir_open, .read = generic_read_dir, .readdir = xfs_file_readdir, .llseek = generic_file_llseek, diff --git a/fs/xfs/linux-2.6/xfs_fs_subr.c b/fs/xfs/linux-2.6/xfs_fs_subr.c index 36caa6d957d..5aeb7777696 100644 --- a/fs/xfs/linux-2.6/xfs_fs_subr.c +++ b/fs/xfs/linux-2.6/xfs_fs_subr.c @@ -24,6 +24,10 @@ int fs_noerr(void) { return 0; } int fs_nosys(void) { return ENOSYS; } void fs_noval(void) { return; } +/* + * note: all filemap functions return negative error codes. These + * need to be inverted before returning to the xfs core functions. + */ void xfs_tosspages( xfs_inode_t *ip, @@ -53,7 +57,7 @@ xfs_flushinval_pages( if (!ret) truncate_inode_pages(mapping, first); } - return ret; + return -ret; } int @@ -72,10 +76,23 @@ xfs_flush_pages( xfs_iflags_clear(ip, XFS_ITRUNCATED); ret = filemap_fdatawrite(mapping); if (flags & XFS_B_ASYNC) - return ret; + return -ret; ret2 = filemap_fdatawait(mapping); if (!ret) ret = ret2; } - return ret; + return -ret; +} + +int +xfs_wait_on_pages( + xfs_inode_t *ip, + xfs_off_t first, + xfs_off_t last) +{ + struct address_space *mapping = VFS_I(ip)->i_mapping; + + if (mapping_tagged(mapping, PAGECACHE_TAG_WRITEBACK)) + return -filemap_fdatawait(mapping); + return 0; } diff --git a/fs/xfs/linux-2.6/xfs_globals.c b/fs/xfs/linux-2.6/xfs_globals.c index ef90e64641e..2ae8b1ccb02 100644 --- a/fs/xfs/linux-2.6/xfs_globals.c +++ b/fs/xfs/linux-2.6/xfs_globals.c @@ -26,7 +26,6 @@ */ xfs_param_t xfs_params = { /* MIN DFLT MAX */ - .restrict_chown = { 0, 1, 1 }, .sgid_inherit = { 0, 0, 1 }, .symlink_mode = { 0, 0, 1 }, .panic_mask = { 0, 0, 255 }, @@ -43,10 +42,3 @@ xfs_param_t xfs_params = { .inherit_nodfrg = { 0, 1, 1 }, .fstrm_timer = { 1, 30*100, 3600*100}, }; - -/* - * Global system credential structure. - */ -static cred_t sys_cred_val; -cred_t *sys_cred = &sys_cred_val; - diff --git a/fs/xfs/linux-2.6/xfs_globals.h b/fs/xfs/linux-2.6/xfs_globals.h index 2770b0085ee..69f71caf061 100644 --- a/fs/xfs/linux-2.6/xfs_globals.h +++ b/fs/xfs/linux-2.6/xfs_globals.h @@ -19,6 +19,5 @@ #define __XFS_GLOBALS_H__ extern uint64_t xfs_panic_mask; /* set to cause more panics */ -extern struct cred *sys_cred; #endif /* __XFS_GLOBALS_H__ */ diff --git a/fs/xfs/linux-2.6/xfs_ioctl.c b/fs/xfs/linux-2.6/xfs_ioctl.c index 48799ba7e3e..67205f6198b 100644 --- a/fs/xfs/linux-2.6/xfs_ioctl.c +++ b/fs/xfs/linux-2.6/xfs_ioctl.c @@ -68,26 +68,22 @@ * XFS_IOC_PATH_TO_HANDLE * returns full handle for a path */ -STATIC int +int xfs_find_handle( unsigned int cmd, - void __user *arg) + xfs_fsop_handlereq_t *hreq) { int hsize; xfs_handle_t handle; - xfs_fsop_handlereq_t hreq; struct inode *inode; - if (copy_from_user(&hreq, arg, sizeof(hreq))) - return -XFS_ERROR(EFAULT); - memset((char *)&handle, 0, sizeof(handle)); switch (cmd) { case XFS_IOC_PATH_TO_FSHANDLE: case XFS_IOC_PATH_TO_HANDLE: { struct path path; - int error = user_lpath((const char __user *)hreq.path, &path); + int error = user_lpath((const char __user *)hreq->path, &path); if (error) return error; @@ -101,7 +97,7 @@ xfs_find_handle( case XFS_IOC_FD_TO_HANDLE: { struct file *file; - file = fget(hreq.fd); + file = fget(hreq->fd); if (!file) return -EBADF; @@ -158,8 +154,8 @@ xfs_find_handle( } /* now copy our handle into the user buffer & write out the size */ - if (copy_to_user(hreq.ohandle, &handle, hsize) || - copy_to_user(hreq.ohandlen, &hsize, sizeof(__s32))) { + if (copy_to_user(hreq->ohandle, &handle, hsize) || + copy_to_user(hreq->ohandlen, &hsize, sizeof(__s32))) { iput(inode); return -XFS_ERROR(EFAULT); } @@ -249,27 +245,25 @@ xfs_vget_fsop_handlereq( return 0; } -STATIC int +int xfs_open_by_handle( xfs_mount_t *mp, - void __user *arg, + xfs_fsop_handlereq_t *hreq, struct file *parfilp, struct inode *parinode) { + const struct cred *cred = current_cred(); int error; int new_fd; int permflag; struct file *filp; struct inode *inode; struct dentry *dentry; - xfs_fsop_handlereq_t hreq; if (!capable(CAP_SYS_ADMIN)) return -XFS_ERROR(EPERM); - if (copy_from_user(&hreq, arg, sizeof(xfs_fsop_handlereq_t))) - return -XFS_ERROR(EFAULT); - error = xfs_vget_fsop_handlereq(mp, parinode, &hreq, &inode); + error = xfs_vget_fsop_handlereq(mp, parinode, hreq, &inode); if (error) return -error; @@ -280,10 +274,10 @@ xfs_open_by_handle( } #if BITS_PER_LONG != 32 - hreq.oflags |= O_LARGEFILE; + hreq->oflags |= O_LARGEFILE; #endif /* Put open permission in namei format. */ - permflag = hreq.oflags; + permflag = hreq->oflags; if ((permflag+1) & O_ACCMODE) permflag++; if (permflag & O_TRUNC) @@ -311,26 +305,26 @@ xfs_open_by_handle( return new_fd; } - dentry = d_alloc_anon(inode); - if (dentry == NULL) { - iput(inode); + dentry = d_obtain_alias(inode); + if (IS_ERR(dentry)) { put_unused_fd(new_fd); - return -XFS_ERROR(ENOMEM); + return PTR_ERR(dentry); } /* Ensure umount returns EBUSY on umounts while this file is open. */ mntget(parfilp->f_path.mnt); /* Create file pointer. */ - filp = dentry_open(dentry, parfilp->f_path.mnt, hreq.oflags); + filp = dentry_open(dentry, parfilp->f_path.mnt, hreq->oflags, cred); if (IS_ERR(filp)) { put_unused_fd(new_fd); return -XFS_ERROR(-PTR_ERR(filp)); } + if (inode->i_mode & S_IFREG) { /* invisible operation should not change atime */ filp->f_flags |= O_NOATIME; - filp->f_op = &xfs_invis_file_operations; + filp->f_mode |= FMODE_NOCMTIME; } fd_install(new_fd, filp); @@ -363,24 +357,21 @@ do_readlink( } -STATIC int +int xfs_readlink_by_handle( xfs_mount_t *mp, - void __user *arg, + xfs_fsop_handlereq_t *hreq, struct inode *parinode) { struct inode *inode; - xfs_fsop_handlereq_t hreq; __u32 olen; void *link; int error; if (!capable(CAP_SYS_ADMIN)) return -XFS_ERROR(EPERM); - if (copy_from_user(&hreq, arg, sizeof(xfs_fsop_handlereq_t))) - return -XFS_ERROR(EFAULT); - error = xfs_vget_fsop_handlereq(mp, parinode, &hreq, &inode); + error = xfs_vget_fsop_handlereq(mp, parinode, hreq, &inode); if (error) return -error; @@ -390,7 +381,7 @@ xfs_readlink_by_handle( goto out_iput; } - if (copy_from_user(&olen, hreq.ohandlen, sizeof(__u32))) { + if (copy_from_user(&olen, hreq->ohandlen, sizeof(__u32))) { error = -XFS_ERROR(EFAULT); goto out_iput; } @@ -402,7 +393,7 @@ xfs_readlink_by_handle( error = -xfs_readlink(XFS_I(inode), link); if (error) goto out_kfree; - error = do_readlink(hreq.ohandle, olen, link); + error = do_readlink(hreq->ohandle, olen, link); if (error) goto out_kfree; @@ -501,7 +492,7 @@ xfs_attrlist_by_handle( return -error; } -STATIC int +int xfs_attrmulti_attr_get( struct inode *inode, char *name, @@ -530,7 +521,7 @@ xfs_attrmulti_attr_get( return error; } -STATIC int +int xfs_attrmulti_attr_set( struct inode *inode, char *name, @@ -560,7 +551,7 @@ xfs_attrmulti_attr_set( return error; } -STATIC int +int xfs_attrmulti_attr_remove( struct inode *inode, char *name, @@ -662,19 +653,26 @@ xfs_attrmulti_by_handle( return -error; } -STATIC int +int xfs_ioc_space( struct xfs_inode *ip, struct inode *inode, struct file *filp, int ioflags, unsigned int cmd, - void __user *arg) + xfs_flock64_t *bf) { - xfs_flock64_t bf; int attr_flags = 0; int error; + /* + * Only allow the sys admin to reserve space unless + * unwritten extents are enabled. + */ + if (!xfs_sb_version_hasextflgbit(&ip->i_mount->m_sb) && + !capable(CAP_SYS_ADMIN)) + return -XFS_ERROR(EPERM); + if (inode->i_flags & (S_IMMUTABLE|S_APPEND)) return -XFS_ERROR(EPERM); @@ -684,16 +682,12 @@ xfs_ioc_space( if (!S_ISREG(inode->i_mode)) return -XFS_ERROR(EINVAL); - if (copy_from_user(&bf, arg, sizeof(bf))) - return -XFS_ERROR(EFAULT); - if (filp->f_flags & (O_NDELAY|O_NONBLOCK)) attr_flags |= XFS_ATTR_NONBLOCK; if (ioflags & IO_INVIS) attr_flags |= XFS_ATTR_DMI; - error = xfs_change_file_space(ip, cmd, &bf, filp->f_pos, - NULL, attr_flags); + error = xfs_change_file_space(ip, cmd, bf, filp->f_pos, attr_flags); return -error; } @@ -1008,7 +1002,7 @@ xfs_ioctl_setattr( * to the file owner ID, except in cases where the * CAP_FSETID capability is applicable. */ - if (current->fsuid != ip->i_d.di_uid && !capable(CAP_FOWNER)) { + if (current_fsuid() != ip->i_d.di_uid && !capable(CAP_FOWNER)) { code = XFS_ERROR(EPERM); goto error_return; } @@ -1105,10 +1099,6 @@ xfs_ioctl_setattr( /* * Change file ownership. Must be the owner or privileged. - * If the system was configured with the "restricted_chown" - * option, the owner is not permitted to give away the file, - * and can change the group id only to a group of which he - * or she is a member. */ if (mask & FSX_PROJID) { /* @@ -1137,7 +1127,7 @@ xfs_ioctl_setattr( * the superblock version number since projids didn't * exist before DINODE_VERSION_2 and SB_VERSION_NLINK. */ - if (ip->i_d.di_version == XFS_DINODE_VERSION_1) + if (ip->i_d.di_version == 1) xfs_bump_ino_vers2(tp, ip); } @@ -1256,43 +1246,67 @@ xfs_ioc_setxflags( } STATIC int +xfs_getbmap_format(void **ap, struct getbmapx *bmv, int *full) +{ + struct getbmap __user *base = *ap; + + /* copy only getbmap portion (not getbmapx) */ + if (copy_to_user(base, bmv, sizeof(struct getbmap))) + return XFS_ERROR(EFAULT); + + *ap += sizeof(struct getbmap); + return 0; +} + +STATIC int xfs_ioc_getbmap( struct xfs_inode *ip, int ioflags, unsigned int cmd, void __user *arg) { - struct getbmap bm; - int iflags; + struct getbmapx bmx; int error; - if (copy_from_user(&bm, arg, sizeof(bm))) + if (copy_from_user(&bmx, arg, sizeof(struct getbmapx))) return -XFS_ERROR(EFAULT); - if (bm.bmv_count < 2) + if (bmx.bmv_count < 2) return -XFS_ERROR(EINVAL); - iflags = (cmd == XFS_IOC_GETBMAPA ? BMV_IF_ATTRFORK : 0); + bmx.bmv_iflags = (cmd == XFS_IOC_GETBMAPA ? BMV_IF_ATTRFORK : 0); if (ioflags & IO_INVIS) - iflags |= BMV_IF_NO_DMAPI_READ; + bmx.bmv_iflags |= BMV_IF_NO_DMAPI_READ; - error = xfs_getbmap(ip, &bm, (struct getbmap __user *)arg+1, iflags); + error = xfs_getbmap(ip, &bmx, xfs_getbmap_format, + (struct getbmap *)arg+1); if (error) return -error; - if (copy_to_user(arg, &bm, sizeof(bm))) + /* copy back header - only size of getbmap */ + if (copy_to_user(arg, &bmx, sizeof(struct getbmap))) return -XFS_ERROR(EFAULT); return 0; } STATIC int +xfs_getbmapx_format(void **ap, struct getbmapx *bmv, int *full) +{ + struct getbmapx __user *base = *ap; + + if (copy_to_user(base, bmv, sizeof(struct getbmapx))) + return XFS_ERROR(EFAULT); + + *ap += sizeof(struct getbmapx); + return 0; +} + +STATIC int xfs_ioc_getbmapx( struct xfs_inode *ip, void __user *arg) { struct getbmapx bmx; - struct getbmap bm; - int iflags; int error; if (copy_from_user(&bmx, arg, sizeof(bmx))) @@ -1301,46 +1315,46 @@ xfs_ioc_getbmapx( if (bmx.bmv_count < 2) return -XFS_ERROR(EINVAL); - /* - * Map input getbmapx structure to a getbmap - * structure for xfs_getbmap. - */ - GETBMAP_CONVERT(bmx, bm); - - iflags = bmx.bmv_iflags; - - if (iflags & (~BMV_IF_VALID)) + if (bmx.bmv_iflags & (~BMV_IF_VALID)) return -XFS_ERROR(EINVAL); - iflags |= BMV_IF_EXTENDED; - - error = xfs_getbmap(ip, &bm, (struct getbmapx __user *)arg+1, iflags); + error = xfs_getbmap(ip, &bmx, xfs_getbmapx_format, + (struct getbmapx *)arg+1); if (error) return -error; - GETBMAP_CONVERT(bm, bmx); - - if (copy_to_user(arg, &bmx, sizeof(bmx))) + /* copy back header */ + if (copy_to_user(arg, &bmx, sizeof(struct getbmapx))) return -XFS_ERROR(EFAULT); return 0; } -int -xfs_ioctl( - xfs_inode_t *ip, +/* + * Note: some of the ioctl's return positive numbers as a + * byte count indicating success, such as readlink_by_handle. + * So we don't "sign flip" like most other routines. This means + * true errors need to be returned as a negative value. + */ +long +xfs_file_ioctl( struct file *filp, - int ioflags, unsigned int cmd, - void __user *arg) + unsigned long p) { struct inode *inode = filp->f_path.dentry->d_inode; - xfs_mount_t *mp = ip->i_mount; + struct xfs_inode *ip = XFS_I(inode); + struct xfs_mount *mp = ip->i_mount; + void __user *arg = (void __user *)p; + int ioflags = 0; int error; - xfs_itrace_entry(XFS_I(inode)); - switch (cmd) { + if (filp->f_mode & FMODE_NOCMTIME) + ioflags |= IO_INVIS; + xfs_itrace_entry(ip); + + switch (cmd) { case XFS_IOC_ALLOCSP: case XFS_IOC_FREESP: case XFS_IOC_RESVSP: @@ -1348,17 +1362,13 @@ xfs_ioctl( case XFS_IOC_ALLOCSP64: case XFS_IOC_FREESP64: case XFS_IOC_RESVSP64: - case XFS_IOC_UNRESVSP64: - /* - * Only allow the sys admin to reserve space unless - * unwritten extents are enabled. - */ - if (!xfs_sb_version_hasextflgbit(&mp->m_sb) && - !capable(CAP_SYS_ADMIN)) - return -EPERM; - - return xfs_ioc_space(ip, inode, filp, ioflags, cmd, arg); + case XFS_IOC_UNRESVSP64: { + xfs_flock64_t bf; + if (copy_from_user(&bf, arg, sizeof(bf))) + return -XFS_ERROR(EFAULT); + return xfs_ioc_space(ip, inode, filp, ioflags, cmd, &bf); + } case XFS_IOC_DIOINFO: { struct dioattr da; xfs_buftarg_t *target = @@ -1418,18 +1428,30 @@ xfs_ioctl( case XFS_IOC_FD_TO_HANDLE: case XFS_IOC_PATH_TO_HANDLE: - case XFS_IOC_PATH_TO_FSHANDLE: - return xfs_find_handle(cmd, arg); + case XFS_IOC_PATH_TO_FSHANDLE: { + xfs_fsop_handlereq_t hreq; - case XFS_IOC_OPEN_BY_HANDLE: - return xfs_open_by_handle(mp, arg, filp, inode); + if (copy_from_user(&hreq, arg, sizeof(hreq))) + return -XFS_ERROR(EFAULT); + return xfs_find_handle(cmd, &hreq); + } + case XFS_IOC_OPEN_BY_HANDLE: { + xfs_fsop_handlereq_t hreq; + if (copy_from_user(&hreq, arg, sizeof(xfs_fsop_handlereq_t))) + return -XFS_ERROR(EFAULT); + return xfs_open_by_handle(mp, &hreq, filp, inode); + } case XFS_IOC_FSSETDM_BY_HANDLE: return xfs_fssetdm_by_handle(mp, arg, inode); - case XFS_IOC_READLINK_BY_HANDLE: - return xfs_readlink_by_handle(mp, arg, inode); + case XFS_IOC_READLINK_BY_HANDLE: { + xfs_fsop_handlereq_t hreq; + if (copy_from_user(&hreq, arg, sizeof(xfs_fsop_handlereq_t))) + return -XFS_ERROR(EFAULT); + return xfs_readlink_by_handle(mp, &hreq, inode); + } case XFS_IOC_ATTRLIST_BY_HANDLE: return xfs_attrlist_by_handle(mp, arg, inode); @@ -1437,7 +1459,11 @@ xfs_ioctl( return xfs_attrmulti_by_handle(mp, arg, filp, inode); case XFS_IOC_SWAPEXT: { - error = xfs_swapext((struct xfs_swapext __user *)arg); + struct xfs_swapext sxp; + + if (copy_from_user(&sxp, arg, sizeof(xfs_swapext_t))) + return -XFS_ERROR(EFAULT); + error = xfs_swapext(&sxp); return -error; } @@ -1493,9 +1519,6 @@ xfs_ioctl( case XFS_IOC_FSGROWFSDATA: { xfs_growfs_data_t in; - if (!capable(CAP_SYS_ADMIN)) - return -EPERM; - if (copy_from_user(&in, arg, sizeof(in))) return -XFS_ERROR(EFAULT); @@ -1506,9 +1529,6 @@ xfs_ioctl( case XFS_IOC_FSGROWFSLOG: { xfs_growfs_log_t in; - if (!capable(CAP_SYS_ADMIN)) - return -EPERM; - if (copy_from_user(&in, arg, sizeof(in))) return -XFS_ERROR(EFAULT); @@ -1519,9 +1539,6 @@ xfs_ioctl( case XFS_IOC_FSGROWFSRT: { xfs_growfs_rt_t in; - if (!capable(CAP_SYS_ADMIN)) - return -EPERM; - if (copy_from_user(&in, arg, sizeof(in))) return -XFS_ERROR(EFAULT); diff --git a/fs/xfs/linux-2.6/xfs_ioctl.h b/fs/xfs/linux-2.6/xfs_ioctl.h new file mode 100644 index 00000000000..8c16bf2d7e0 --- /dev/null +++ b/fs/xfs/linux-2.6/xfs_ioctl.h @@ -0,0 +1,82 @@ +/* + * Copyright (c) 2008 Silicon Graphics, Inc. + * All Rights Reserved. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it would be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write the Free Software Foundation, + * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA + */ +#ifndef __XFS_IOCTL_H__ +#define __XFS_IOCTL_H__ + +extern int +xfs_ioc_space( + struct xfs_inode *ip, + struct inode *inode, + struct file *filp, + int ioflags, + unsigned int cmd, + xfs_flock64_t *bf); + +extern int +xfs_find_handle( + unsigned int cmd, + xfs_fsop_handlereq_t *hreq); + +extern int +xfs_open_by_handle( + xfs_mount_t *mp, + xfs_fsop_handlereq_t *hreq, + struct file *parfilp, + struct inode *parinode); + +extern int +xfs_readlink_by_handle( + xfs_mount_t *mp, + xfs_fsop_handlereq_t *hreq, + struct inode *parinode); + +extern int +xfs_attrmulti_attr_get( + struct inode *inode, + char *name, + char __user *ubuf, + __uint32_t *len, + __uint32_t flags); + +extern int + xfs_attrmulti_attr_set( + struct inode *inode, + char *name, + const char __user *ubuf, + __uint32_t len, + __uint32_t flags); + +extern int +xfs_attrmulti_attr_remove( + struct inode *inode, + char *name, + __uint32_t flags); + +extern long +xfs_file_ioctl( + struct file *filp, + unsigned int cmd, + unsigned long p); + +extern long +xfs_file_compat_ioctl( + struct file *file, + unsigned int cmd, + unsigned long arg); + +#endif diff --git a/fs/xfs/linux-2.6/xfs_ioctl32.c b/fs/xfs/linux-2.6/xfs_ioctl32.c index a4b254eb43b..0504cece9f6 100644 --- a/fs/xfs/linux-2.6/xfs_ioctl32.c +++ b/fs/xfs/linux-2.6/xfs_ioctl32.c @@ -16,11 +16,7 @@ * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA */ #include <linux/compat.h> -#include <linux/init.h> #include <linux/ioctl.h> -#include <linux/syscalls.h> -#include <linux/types.h> -#include <linux/fs.h> #include <asm/uaccess.h> #include "xfs.h" #include "xfs_fs.h" @@ -36,7 +32,6 @@ #include "xfs_bmap_btree.h" #include "xfs_attr_sf.h" #include "xfs_dir2_sf.h" -#include "xfs_vfs.h" #include "xfs_vnode.h" #include "xfs_dinode.h" #include "xfs_inode.h" @@ -44,221 +39,219 @@ #include "xfs_error.h" #include "xfs_dfrag.h" #include "xfs_vnodeops.h" +#include "xfs_fsops.h" +#include "xfs_alloc.h" +#include "xfs_rtalloc.h" +#include "xfs_attr.h" +#include "xfs_ioctl.h" #include "xfs_ioctl32.h" #define _NATIVE_IOC(cmd, type) \ _IOC(_IOC_DIR(cmd), _IOC_TYPE(cmd), _IOC_NR(cmd), sizeof(type)) -#if defined(CONFIG_IA64) || defined(CONFIG_X86_64) -#define BROKEN_X86_ALIGNMENT -#define _PACKED __attribute__((packed)) -/* on ia32 l_start is on a 32-bit boundary */ -typedef struct xfs_flock64_32 { - __s16 l_type; - __s16 l_whence; - __s64 l_start __attribute__((packed)); - /* len == 0 means until end of file */ - __s64 l_len __attribute__((packed)); - __s32 l_sysid; - __u32 l_pid; - __s32 l_pad[4]; /* reserve area */ -} xfs_flock64_32_t; - -#define XFS_IOC_ALLOCSP_32 _IOW ('X', 10, struct xfs_flock64_32) -#define XFS_IOC_FREESP_32 _IOW ('X', 11, struct xfs_flock64_32) -#define XFS_IOC_ALLOCSP64_32 _IOW ('X', 36, struct xfs_flock64_32) -#define XFS_IOC_FREESP64_32 _IOW ('X', 37, struct xfs_flock64_32) -#define XFS_IOC_RESVSP_32 _IOW ('X', 40, struct xfs_flock64_32) -#define XFS_IOC_UNRESVSP_32 _IOW ('X', 41, struct xfs_flock64_32) -#define XFS_IOC_RESVSP64_32 _IOW ('X', 42, struct xfs_flock64_32) -#define XFS_IOC_UNRESVSP64_32 _IOW ('X', 43, struct xfs_flock64_32) - -/* just account for different alignment */ -STATIC unsigned long -xfs_ioctl32_flock( - unsigned long arg) +#ifdef BROKEN_X86_ALIGNMENT +STATIC int +xfs_compat_flock64_copyin( + xfs_flock64_t *bf, + compat_xfs_flock64_t __user *arg32) { - xfs_flock64_32_t __user *p32 = (void __user *)arg; - xfs_flock64_t __user *p = compat_alloc_user_space(sizeof(*p)); - - if (copy_in_user(&p->l_type, &p32->l_type, sizeof(s16)) || - copy_in_user(&p->l_whence, &p32->l_whence, sizeof(s16)) || - copy_in_user(&p->l_start, &p32->l_start, sizeof(s64)) || - copy_in_user(&p->l_len, &p32->l_len, sizeof(s64)) || - copy_in_user(&p->l_sysid, &p32->l_sysid, sizeof(s32)) || - copy_in_user(&p->l_pid, &p32->l_pid, sizeof(u32)) || - copy_in_user(&p->l_pad, &p32->l_pad, 4*sizeof(u32))) - return -EFAULT; - - return (unsigned long)p; + if (get_user(bf->l_type, &arg32->l_type) || + get_user(bf->l_whence, &arg32->l_whence) || + get_user(bf->l_start, &arg32->l_start) || + get_user(bf->l_len, &arg32->l_len) || + get_user(bf->l_sysid, &arg32->l_sysid) || + get_user(bf->l_pid, &arg32->l_pid) || + copy_from_user(bf->l_pad, &arg32->l_pad, 4*sizeof(u32))) + return -XFS_ERROR(EFAULT); + return 0; } -typedef struct compat_xfs_fsop_geom_v1 { - __u32 blocksize; /* filesystem (data) block size */ - __u32 rtextsize; /* realtime extent size */ - __u32 agblocks; /* fsblocks in an AG */ - __u32 agcount; /* number of allocation groups */ - __u32 logblocks; /* fsblocks in the log */ - __u32 sectsize; /* (data) sector size, bytes */ - __u32 inodesize; /* inode size in bytes */ - __u32 imaxpct; /* max allowed inode space(%) */ - __u64 datablocks; /* fsblocks in data subvolume */ - __u64 rtblocks; /* fsblocks in realtime subvol */ - __u64 rtextents; /* rt extents in realtime subvol*/ - __u64 logstart; /* starting fsblock of the log */ - unsigned char uuid[16]; /* unique id of the filesystem */ - __u32 sunit; /* stripe unit, fsblocks */ - __u32 swidth; /* stripe width, fsblocks */ - __s32 version; /* structure version */ - __u32 flags; /* superblock version flags */ - __u32 logsectsize; /* log sector size, bytes */ - __u32 rtsectsize; /* realtime sector size, bytes */ - __u32 dirblocksize; /* directory block size, bytes */ -} __attribute__((packed)) compat_xfs_fsop_geom_v1_t; - -#define XFS_IOC_FSGEOMETRY_V1_32 \ - _IOR ('X', 100, struct compat_xfs_fsop_geom_v1) - -STATIC unsigned long xfs_ioctl32_geom_v1(unsigned long arg) +STATIC int +xfs_compat_ioc_fsgeometry_v1( + struct xfs_mount *mp, + compat_xfs_fsop_geom_v1_t __user *arg32) { - compat_xfs_fsop_geom_v1_t __user *p32 = (void __user *)arg; - xfs_fsop_geom_v1_t __user *p = compat_alloc_user_space(sizeof(*p)); + xfs_fsop_geom_t fsgeo; + int error; - if (copy_in_user(p, p32, sizeof(*p32))) - return -EFAULT; - return (unsigned long)p; + error = xfs_fs_geometry(mp, &fsgeo, 3); + if (error) + return -error; + /* The 32-bit variant simply has some padding at the end */ + if (copy_to_user(arg32, &fsgeo, sizeof(struct compat_xfs_fsop_geom_v1))) + return -XFS_ERROR(EFAULT); + return 0; } -typedef struct compat_xfs_inogrp { - __u64 xi_startino; /* starting inode number */ - __s32 xi_alloccount; /* # bits set in allocmask */ - __u64 xi_allocmask; /* mask of allocated inodes */ -} __attribute__((packed)) compat_xfs_inogrp_t; - -STATIC int xfs_inumbers_fmt_compat( - void __user *ubuffer, - const xfs_inogrp_t *buffer, - long count, - long *written) +STATIC int +xfs_compat_growfs_data_copyin( + struct xfs_growfs_data *in, + compat_xfs_growfs_data_t __user *arg32) { - compat_xfs_inogrp_t __user *p32 = ubuffer; - long i; + if (get_user(in->newblocks, &arg32->newblocks) || + get_user(in->imaxpct, &arg32->imaxpct)) + return -XFS_ERROR(EFAULT); + return 0; +} + +STATIC int +xfs_compat_growfs_rt_copyin( + struct xfs_growfs_rt *in, + compat_xfs_growfs_rt_t __user *arg32) +{ + if (get_user(in->newblocks, &arg32->newblocks) || + get_user(in->extsize, &arg32->extsize)) + return -XFS_ERROR(EFAULT); + return 0; +} + +STATIC int +xfs_inumbers_fmt_compat( + void __user *ubuffer, + const xfs_inogrp_t *buffer, + long count, + long *written) +{ + compat_xfs_inogrp_t __user *p32 = ubuffer; + long i; for (i = 0; i < count; i++) { if (put_user(buffer[i].xi_startino, &p32[i].xi_startino) || put_user(buffer[i].xi_alloccount, &p32[i].xi_alloccount) || put_user(buffer[i].xi_allocmask, &p32[i].xi_allocmask)) - return -EFAULT; + return -XFS_ERROR(EFAULT); } *written = count * sizeof(*p32); return 0; } #else - #define xfs_inumbers_fmt_compat xfs_inumbers_fmt -#define _PACKED +#endif /* BROKEN_X86_ALIGNMENT */ -#endif +STATIC int +xfs_ioctl32_bstime_copyin( + xfs_bstime_t *bstime, + compat_xfs_bstime_t __user *bstime32) +{ + compat_time_t sec32; /* tv_sec differs on 64 vs. 32 */ -/* XFS_IOC_FSBULKSTAT and friends */ + if (get_user(sec32, &bstime32->tv_sec) || + get_user(bstime->tv_nsec, &bstime32->tv_nsec)) + return -XFS_ERROR(EFAULT); + bstime->tv_sec = sec32; + return 0; +} + +/* xfs_bstat_t has differing alignment on intel, & bstime_t sizes everywhere */ +STATIC int +xfs_ioctl32_bstat_copyin( + xfs_bstat_t *bstat, + compat_xfs_bstat_t __user *bstat32) +{ + if (get_user(bstat->bs_ino, &bstat32->bs_ino) || + get_user(bstat->bs_mode, &bstat32->bs_mode) || + get_user(bstat->bs_nlink, &bstat32->bs_nlink) || + get_user(bstat->bs_uid, &bstat32->bs_uid) || + get_user(bstat->bs_gid, &bstat32->bs_gid) || + get_user(bstat->bs_rdev, &bstat32->bs_rdev) || + get_user(bstat->bs_blksize, &bstat32->bs_blksize) || + get_user(bstat->bs_size, &bstat32->bs_size) || + xfs_ioctl32_bstime_copyin(&bstat->bs_atime, &bstat32->bs_atime) || + xfs_ioctl32_bstime_copyin(&bstat->bs_mtime, &bstat32->bs_mtime) || + xfs_ioctl32_bstime_copyin(&bstat->bs_ctime, &bstat32->bs_ctime) || + get_user(bstat->bs_blocks, &bstat32->bs_size) || + get_user(bstat->bs_xflags, &bstat32->bs_size) || + get_user(bstat->bs_extsize, &bstat32->bs_extsize) || + get_user(bstat->bs_extents, &bstat32->bs_extents) || + get_user(bstat->bs_gen, &bstat32->bs_gen) || + get_user(bstat->bs_projid, &bstat32->bs_projid) || + get_user(bstat->bs_dmevmask, &bstat32->bs_dmevmask) || + get_user(bstat->bs_dmstate, &bstat32->bs_dmstate) || + get_user(bstat->bs_aextents, &bstat32->bs_aextents)) + return -XFS_ERROR(EFAULT); + return 0; +} -typedef struct compat_xfs_bstime { - __s32 tv_sec; /* seconds */ - __s32 tv_nsec; /* and nanoseconds */ -} compat_xfs_bstime_t; +/* XFS_IOC_FSBULKSTAT and friends */ -STATIC int xfs_bstime_store_compat( - compat_xfs_bstime_t __user *p32, - const xfs_bstime_t *p) +STATIC int +xfs_bstime_store_compat( + compat_xfs_bstime_t __user *p32, + const xfs_bstime_t *p) { - __s32 sec32; + __s32 sec32; sec32 = p->tv_sec; if (put_user(sec32, &p32->tv_sec) || put_user(p->tv_nsec, &p32->tv_nsec)) - return -EFAULT; + return -XFS_ERROR(EFAULT); return 0; } -typedef struct compat_xfs_bstat { - __u64 bs_ino; /* inode number */ - __u16 bs_mode; /* type and mode */ - __u16 bs_nlink; /* number of links */ - __u32 bs_uid; /* user id */ - __u32 bs_gid; /* group id */ - __u32 bs_rdev; /* device value */ - __s32 bs_blksize; /* block size */ - __s64 bs_size; /* file size */ - compat_xfs_bstime_t bs_atime; /* access time */ - compat_xfs_bstime_t bs_mtime; /* modify time */ - compat_xfs_bstime_t bs_ctime; /* inode change time */ - int64_t bs_blocks; /* number of blocks */ - __u32 bs_xflags; /* extended flags */ - __s32 bs_extsize; /* extent size */ - __s32 bs_extents; /* number of extents */ - __u32 bs_gen; /* generation count */ - __u16 bs_projid; /* project id */ - unsigned char bs_pad[14]; /* pad space, unused */ - __u32 bs_dmevmask; /* DMIG event mask */ - __u16 bs_dmstate; /* DMIG state info */ - __u16 bs_aextents; /* attribute number of extents */ -} _PACKED compat_xfs_bstat_t; - -STATIC int xfs_bulkstat_one_fmt_compat( +/* Return 0 on success or positive error (to xfs_bulkstat()) */ +STATIC int +xfs_bulkstat_one_fmt_compat( void __user *ubuffer, + int ubsize, + int *ubused, const xfs_bstat_t *buffer) { - compat_xfs_bstat_t __user *p32 = ubuffer; - - if (put_user(buffer->bs_ino, &p32->bs_ino) || - put_user(buffer->bs_mode, &p32->bs_mode) || - put_user(buffer->bs_nlink, &p32->bs_nlink) || - put_user(buffer->bs_uid, &p32->bs_uid) || - put_user(buffer->bs_gid, &p32->bs_gid) || - put_user(buffer->bs_rdev, &p32->bs_rdev) || - put_user(buffer->bs_blksize, &p32->bs_blksize) || - put_user(buffer->bs_size, &p32->bs_size) || + compat_xfs_bstat_t __user *p32 = ubuffer; + + if (ubsize < sizeof(*p32)) + return XFS_ERROR(ENOMEM); + + if (put_user(buffer->bs_ino, &p32->bs_ino) || + put_user(buffer->bs_mode, &p32->bs_mode) || + put_user(buffer->bs_nlink, &p32->bs_nlink) || + put_user(buffer->bs_uid, &p32->bs_uid) || + put_user(buffer->bs_gid, &p32->bs_gid) || + put_user(buffer->bs_rdev, &p32->bs_rdev) || + put_user(buffer->bs_blksize, &p32->bs_blksize) || + put_user(buffer->bs_size, &p32->bs_size) || xfs_bstime_store_compat(&p32->bs_atime, &buffer->bs_atime) || xfs_bstime_store_compat(&p32->bs_mtime, &buffer->bs_mtime) || xfs_bstime_store_compat(&p32->bs_ctime, &buffer->bs_ctime) || - put_user(buffer->bs_blocks, &p32->bs_blocks) || - put_user(buffer->bs_xflags, &p32->bs_xflags) || - put_user(buffer->bs_extsize, &p32->bs_extsize) || - put_user(buffer->bs_extents, &p32->bs_extents) || - put_user(buffer->bs_gen, &p32->bs_gen) || - put_user(buffer->bs_projid, &p32->bs_projid) || - put_user(buffer->bs_dmevmask, &p32->bs_dmevmask) || - put_user(buffer->bs_dmstate, &p32->bs_dmstate) || + put_user(buffer->bs_blocks, &p32->bs_blocks) || + put_user(buffer->bs_xflags, &p32->bs_xflags) || + put_user(buffer->bs_extsize, &p32->bs_extsize) || + put_user(buffer->bs_extents, &p32->bs_extents) || + put_user(buffer->bs_gen, &p32->bs_gen) || + put_user(buffer->bs_projid, &p32->bs_projid) || + put_user(buffer->bs_dmevmask, &p32->bs_dmevmask) || + put_user(buffer->bs_dmstate, &p32->bs_dmstate) || put_user(buffer->bs_aextents, &p32->bs_aextents)) - return -EFAULT; - return sizeof(*p32); + return XFS_ERROR(EFAULT); + if (ubused) + *ubused = sizeof(*p32); + return 0; } - - -typedef struct compat_xfs_fsop_bulkreq { - compat_uptr_t lastip; /* last inode # pointer */ - __s32 icount; /* count of entries in buffer */ - compat_uptr_t ubuffer; /* user buffer for inode desc. */ - compat_uptr_t ocount; /* output count pointer */ -} compat_xfs_fsop_bulkreq_t; - -#define XFS_IOC_FSBULKSTAT_32 \ - _IOWR('X', 101, struct compat_xfs_fsop_bulkreq) -#define XFS_IOC_FSBULKSTAT_SINGLE_32 \ - _IOWR('X', 102, struct compat_xfs_fsop_bulkreq) -#define XFS_IOC_FSINUMBERS_32 \ - _IOWR('X', 103, struct compat_xfs_fsop_bulkreq) +STATIC int +xfs_bulkstat_one_compat( + xfs_mount_t *mp, /* mount point for filesystem */ + xfs_ino_t ino, /* inode number to get data for */ + void __user *buffer, /* buffer to place output in */ + int ubsize, /* size of buffer */ + void *private_data, /* my private data */ + xfs_daddr_t bno, /* starting bno of inode cluster */ + int *ubused, /* bytes used by me */ + void *dibuff, /* on-disk inode buffer */ + int *stat) /* BULKSTAT_RV_... */ +{ + return xfs_bulkstat_one_int(mp, ino, buffer, ubsize, + xfs_bulkstat_one_fmt_compat, bno, + ubused, dibuff, stat); +} /* copied from xfs_ioctl.c */ STATIC int -xfs_ioc_bulkstat_compat( - xfs_mount_t *mp, - unsigned int cmd, - void __user *arg) +xfs_compat_ioc_bulkstat( + xfs_mount_t *mp, + unsigned int cmd, + compat_xfs_fsop_bulkreq_t __user *p32) { - compat_xfs_fsop_bulkreq_t __user *p32 = (void __user *)arg; u32 addr; xfs_fsop_bulkreq_t bulkreq; int count; /* # of records returned */ @@ -270,20 +263,20 @@ xfs_ioc_bulkstat_compat( /* should be called again (unused here, but used in dmapi) */ if (!capable(CAP_SYS_ADMIN)) - return -EPERM; + return -XFS_ERROR(EPERM); if (XFS_FORCED_SHUTDOWN(mp)) return -XFS_ERROR(EIO); if (get_user(addr, &p32->lastip)) - return -EFAULT; + return -XFS_ERROR(EFAULT); bulkreq.lastip = compat_ptr(addr); if (get_user(bulkreq.icount, &p32->icount) || get_user(addr, &p32->ubuffer)) - return -EFAULT; + return -XFS_ERROR(EFAULT); bulkreq.ubuffer = compat_ptr(addr); if (get_user(addr, &p32->ocount)) - return -EFAULT; + return -XFS_ERROR(EFAULT); bulkreq.ocount = compat_ptr(addr); if (copy_from_user(&inlast, bulkreq.lastip, sizeof(__s64))) @@ -295,17 +288,22 @@ xfs_ioc_bulkstat_compat( if (bulkreq.ubuffer == NULL) return -XFS_ERROR(EINVAL); - if (cmd == XFS_IOC_FSINUMBERS) + if (cmd == XFS_IOC_FSINUMBERS_32) { error = xfs_inumbers(mp, &inlast, &count, bulkreq.ubuffer, xfs_inumbers_fmt_compat); - else { - /* declare a var to get a warning in case the type changes */ - bulkstat_one_fmt_pf formatter = xfs_bulkstat_one_fmt_compat; + } else if (cmd == XFS_IOC_FSBULKSTAT_SINGLE_32) { + int res; + + error = xfs_bulkstat_one_compat(mp, inlast, bulkreq.ubuffer, + sizeof(compat_xfs_bstat_t), + NULL, 0, NULL, NULL, &res); + } else if (cmd == XFS_IOC_FSBULKSTAT_32) { error = xfs_bulkstat(mp, &inlast, &count, - xfs_bulkstat_one, formatter, + xfs_bulkstat_one_compat, NULL, sizeof(compat_xfs_bstat_t), bulkreq.ubuffer, BULKSTAT_FG_QUICK, &done); - } + } else + error = XFS_ERROR(EINVAL); if (error) return -error; @@ -321,63 +319,306 @@ xfs_ioc_bulkstat_compat( return 0; } +STATIC int +xfs_compat_handlereq_copyin( + xfs_fsop_handlereq_t *hreq, + compat_xfs_fsop_handlereq_t __user *arg32) +{ + compat_xfs_fsop_handlereq_t hreq32; + + if (copy_from_user(&hreq32, arg32, sizeof(compat_xfs_fsop_handlereq_t))) + return -XFS_ERROR(EFAULT); + + hreq->fd = hreq32.fd; + hreq->path = compat_ptr(hreq32.path); + hreq->oflags = hreq32.oflags; + hreq->ihandle = compat_ptr(hreq32.ihandle); + hreq->ihandlen = hreq32.ihandlen; + hreq->ohandle = compat_ptr(hreq32.ohandle); + hreq->ohandlen = compat_ptr(hreq32.ohandlen); + return 0; +} -typedef struct compat_xfs_fsop_handlereq { - __u32 fd; /* fd for FD_TO_HANDLE */ - compat_uptr_t path; /* user pathname */ - __u32 oflags; /* open flags */ - compat_uptr_t ihandle; /* user supplied handle */ - __u32 ihandlen; /* user supplied length */ - compat_uptr_t ohandle; /* user buffer for handle */ - compat_uptr_t ohandlen; /* user buffer length */ -} compat_xfs_fsop_handlereq_t; - -#define XFS_IOC_PATH_TO_FSHANDLE_32 \ - _IOWR('X', 104, struct compat_xfs_fsop_handlereq) -#define XFS_IOC_PATH_TO_HANDLE_32 \ - _IOWR('X', 105, struct compat_xfs_fsop_handlereq) -#define XFS_IOC_FD_TO_HANDLE_32 \ - _IOWR('X', 106, struct compat_xfs_fsop_handlereq) -#define XFS_IOC_OPEN_BY_HANDLE_32 \ - _IOWR('X', 107, struct compat_xfs_fsop_handlereq) -#define XFS_IOC_READLINK_BY_HANDLE_32 \ - _IOWR('X', 108, struct compat_xfs_fsop_handlereq) - -STATIC unsigned long xfs_ioctl32_fshandle(unsigned long arg) +/* + * Convert userspace handle data into inode. + * + * We use the fact that all the fsop_handlereq ioctl calls have a data + * structure argument whose first component is always a xfs_fsop_handlereq_t, + * so we can pass that sub structure into this handy, shared routine. + * + * If no error, caller must always iput the returned inode. + */ +STATIC int +xfs_vget_fsop_handlereq_compat( + xfs_mount_t *mp, + struct inode *parinode, /* parent inode pointer */ + compat_xfs_fsop_handlereq_t *hreq, + struct inode **inode) { - compat_xfs_fsop_handlereq_t __user *p32 = (void __user *)arg; - xfs_fsop_handlereq_t __user *p = compat_alloc_user_space(sizeof(*p)); - u32 addr; - - if (copy_in_user(&p->fd, &p32->fd, sizeof(__u32)) || - get_user(addr, &p32->path) || - put_user(compat_ptr(addr), &p->path) || - copy_in_user(&p->oflags, &p32->oflags, sizeof(__u32)) || - get_user(addr, &p32->ihandle) || - put_user(compat_ptr(addr), &p->ihandle) || - copy_in_user(&p->ihandlen, &p32->ihandlen, sizeof(__u32)) || - get_user(addr, &p32->ohandle) || - put_user(compat_ptr(addr), &p->ohandle) || - get_user(addr, &p32->ohandlen) || - put_user(compat_ptr(addr), &p->ohandlen)) - return -EFAULT; - - return (unsigned long)p; + void __user *hanp; + size_t hlen; + xfs_fid_t *xfid; + xfs_handle_t *handlep; + xfs_handle_t handle; + xfs_inode_t *ip; + xfs_ino_t ino; + __u32 igen; + int error; + + /* + * Only allow handle opens under a directory. + */ + if (!S_ISDIR(parinode->i_mode)) + return XFS_ERROR(ENOTDIR); + + hanp = compat_ptr(hreq->ihandle); + hlen = hreq->ihandlen; + handlep = &handle; + + if (hlen < sizeof(handlep->ha_fsid) || hlen > sizeof(*handlep)) + return XFS_ERROR(EINVAL); + if (copy_from_user(handlep, hanp, hlen)) + return XFS_ERROR(EFAULT); + if (hlen < sizeof(*handlep)) + memset(((char *)handlep) + hlen, 0, sizeof(*handlep) - hlen); + if (hlen > sizeof(handlep->ha_fsid)) { + if (handlep->ha_fid.fid_len != + (hlen - sizeof(handlep->ha_fsid) - + sizeof(handlep->ha_fid.fid_len)) || + handlep->ha_fid.fid_pad) + return XFS_ERROR(EINVAL); + } + + /* + * Crack the handle, obtain the inode # & generation # + */ + xfid = (struct xfs_fid *)&handlep->ha_fid; + if (xfid->fid_len == sizeof(*xfid) - sizeof(xfid->fid_len)) { + ino = xfid->fid_ino; + igen = xfid->fid_gen; + } else { + return XFS_ERROR(EINVAL); + } + + /* + * Get the XFS inode, building a Linux inode to go with it. + */ + error = xfs_iget(mp, NULL, ino, 0, XFS_ILOCK_SHARED, &ip, 0); + if (error) + return error; + if (ip == NULL) + return XFS_ERROR(EIO); + if (ip->i_d.di_gen != igen) { + xfs_iput_new(ip, XFS_ILOCK_SHARED); + return XFS_ERROR(ENOENT); + } + + xfs_iunlock(ip, XFS_ILOCK_SHARED); + + *inode = VFS_I(ip); + return 0; } +STATIC int +xfs_compat_attrlist_by_handle( + xfs_mount_t *mp, + void __user *arg, + struct inode *parinode) +{ + int error; + attrlist_cursor_kern_t *cursor; + compat_xfs_fsop_attrlist_handlereq_t al_hreq; + struct inode *inode; + char *kbuf; + + if (!capable(CAP_SYS_ADMIN)) + return -XFS_ERROR(EPERM); + if (copy_from_user(&al_hreq, arg, + sizeof(compat_xfs_fsop_attrlist_handlereq_t))) + return -XFS_ERROR(EFAULT); + if (al_hreq.buflen > XATTR_LIST_MAX) + return -XFS_ERROR(EINVAL); + + /* + * Reject flags, only allow namespaces. + */ + if (al_hreq.flags & ~(ATTR_ROOT | ATTR_SECURE)) + return -XFS_ERROR(EINVAL); + + error = xfs_vget_fsop_handlereq_compat(mp, parinode, &al_hreq.hreq, + &inode); + if (error) + goto out; + + kbuf = kmalloc(al_hreq.buflen, GFP_KERNEL); + if (!kbuf) + goto out_vn_rele; + + cursor = (attrlist_cursor_kern_t *)&al_hreq.pos; + error = xfs_attr_list(XFS_I(inode), kbuf, al_hreq.buflen, + al_hreq.flags, cursor); + if (error) + goto out_kfree; + + if (copy_to_user(compat_ptr(al_hreq.buffer), kbuf, al_hreq.buflen)) + error = -EFAULT; + + out_kfree: + kfree(kbuf); + out_vn_rele: + iput(inode); + out: + return -error; +} -STATIC long -xfs_compat_ioctl( - int mode, - struct file *file, - unsigned cmd, - unsigned long arg) +STATIC int +xfs_compat_attrmulti_by_handle( + xfs_mount_t *mp, + void __user *arg, + struct inode *parinode) +{ + int error; + compat_xfs_attr_multiop_t *ops; + compat_xfs_fsop_attrmulti_handlereq_t am_hreq; + struct inode *inode; + unsigned int i, size; + char *attr_name; + + if (!capable(CAP_SYS_ADMIN)) + return -XFS_ERROR(EPERM); + if (copy_from_user(&am_hreq, arg, + sizeof(compat_xfs_fsop_attrmulti_handlereq_t))) + return -XFS_ERROR(EFAULT); + + error = xfs_vget_fsop_handlereq_compat(mp, parinode, &am_hreq.hreq, + &inode); + if (error) + goto out; + + error = E2BIG; + size = am_hreq.opcount * sizeof(compat_xfs_attr_multiop_t); + if (!size || size > 16 * PAGE_SIZE) + goto out_vn_rele; + + error = ENOMEM; + ops = kmalloc(size, GFP_KERNEL); + if (!ops) + goto out_vn_rele; + + error = EFAULT; + if (copy_from_user(ops, compat_ptr(am_hreq.ops), size)) + goto out_kfree_ops; + + attr_name = kmalloc(MAXNAMELEN, GFP_KERNEL); + if (!attr_name) + goto out_kfree_ops; + + + error = 0; + for (i = 0; i < am_hreq.opcount; i++) { + ops[i].am_error = strncpy_from_user(attr_name, + compat_ptr(ops[i].am_attrname), + MAXNAMELEN); + if (ops[i].am_error == 0 || ops[i].am_error == MAXNAMELEN) + error = -ERANGE; + if (ops[i].am_error < 0) + break; + + switch (ops[i].am_opcode) { + case ATTR_OP_GET: + ops[i].am_error = xfs_attrmulti_attr_get(inode, + attr_name, + compat_ptr(ops[i].am_attrvalue), + &ops[i].am_length, ops[i].am_flags); + break; + case ATTR_OP_SET: + ops[i].am_error = xfs_attrmulti_attr_set(inode, + attr_name, + compat_ptr(ops[i].am_attrvalue), + ops[i].am_length, ops[i].am_flags); + break; + case ATTR_OP_REMOVE: + ops[i].am_error = xfs_attrmulti_attr_remove(inode, + attr_name, ops[i].am_flags); + break; + default: + ops[i].am_error = EINVAL; + } + } + + if (copy_to_user(compat_ptr(am_hreq.ops), ops, size)) + error = XFS_ERROR(EFAULT); + + kfree(attr_name); + out_kfree_ops: + kfree(ops); + out_vn_rele: + iput(inode); + out: + return -error; +} + +STATIC int +xfs_compat_fssetdm_by_handle( + xfs_mount_t *mp, + void __user *arg, + struct inode *parinode) +{ + int error; + struct fsdmidata fsd; + compat_xfs_fsop_setdm_handlereq_t dmhreq; + struct inode *inode; + + if (!capable(CAP_MKNOD)) + return -XFS_ERROR(EPERM); + if (copy_from_user(&dmhreq, arg, + sizeof(compat_xfs_fsop_setdm_handlereq_t))) + return -XFS_ERROR(EFAULT); + + error = xfs_vget_fsop_handlereq_compat(mp, parinode, &dmhreq.hreq, + &inode); + if (error) + return -error; + + if (IS_IMMUTABLE(inode) || IS_APPEND(inode)) { + error = -XFS_ERROR(EPERM); + goto out; + } + + if (copy_from_user(&fsd, compat_ptr(dmhreq.data), sizeof(fsd))) { + error = -XFS_ERROR(EFAULT); + goto out; + } + + error = -xfs_set_dmattrs(XFS_I(inode), fsd.fsd_dmevmask, + fsd.fsd_dmstate); + +out: + iput(inode); + return error; +} + +long +xfs_file_compat_ioctl( + struct file *filp, + unsigned cmd, + unsigned long p) { - struct inode *inode = file->f_path.dentry->d_inode; - int error; + struct inode *inode = filp->f_path.dentry->d_inode; + struct xfs_inode *ip = XFS_I(inode); + struct xfs_mount *mp = ip->i_mount; + void __user *arg = (void __user *)p; + int ioflags = 0; + int error; + + if (filp->f_mode & FMODE_NOCMTIME) + ioflags |= IO_INVIS; + + xfs_itrace_entry(ip); switch (cmd) { + /* No size or alignment issues on any arch */ case XFS_IOC_DIOINFO: case XFS_IOC_FSGEOMETRY: case XFS_IOC_FSGETXATTR: @@ -387,48 +628,18 @@ xfs_compat_ioctl( case XFS_IOC_GETBMAP: case XFS_IOC_GETBMAPA: case XFS_IOC_GETBMAPX: -/* not handled - case XFS_IOC_FSSETDM_BY_HANDLE: - case XFS_IOC_ATTRLIST_BY_HANDLE: - case XFS_IOC_ATTRMULTI_BY_HANDLE: -*/ case XFS_IOC_FSCOUNTS: case XFS_IOC_SET_RESBLKS: case XFS_IOC_GET_RESBLKS: - case XFS_IOC_FSGROWFSDATA: case XFS_IOC_FSGROWFSLOG: - case XFS_IOC_FSGROWFSRT: case XFS_IOC_FREEZE: case XFS_IOC_THAW: case XFS_IOC_GOINGDOWN: case XFS_IOC_ERROR_INJECTION: case XFS_IOC_ERROR_CLEARALL: - break; - - case XFS_IOC32_GETXFLAGS: - case XFS_IOC32_SETXFLAGS: - case XFS_IOC32_GETVERSION: - cmd = _NATIVE_IOC(cmd, long); - break; -#ifdef BROKEN_X86_ALIGNMENT - /* xfs_flock_t has wrong u32 vs u64 alignment */ - case XFS_IOC_ALLOCSP_32: - case XFS_IOC_FREESP_32: - case XFS_IOC_ALLOCSP64_32: - case XFS_IOC_FREESP64_32: - case XFS_IOC_RESVSP_32: - case XFS_IOC_UNRESVSP_32: - case XFS_IOC_RESVSP64_32: - case XFS_IOC_UNRESVSP64_32: - arg = xfs_ioctl32_flock(arg); - cmd = _NATIVE_IOC(cmd, struct xfs_flock64); - break; - case XFS_IOC_FSGEOMETRY_V1_32: - arg = xfs_ioctl32_geom_v1(arg); - cmd = _NATIVE_IOC(cmd, struct xfs_fsop_geom_v1); - break; - -#else /* These are handled fine if no alignment issues */ + return xfs_file_ioctl(filp, cmd, p); +#ifndef BROKEN_X86_ALIGNMENT + /* These are handled fine if no alignment issues */ case XFS_IOC_ALLOCSP: case XFS_IOC_FREESP: case XFS_IOC_RESVSP: @@ -438,51 +649,97 @@ xfs_compat_ioctl( case XFS_IOC_RESVSP64: case XFS_IOC_UNRESVSP64: case XFS_IOC_FSGEOMETRY_V1: - break; + case XFS_IOC_FSGROWFSDATA: + case XFS_IOC_FSGROWFSRT: + return xfs_file_ioctl(filp, cmd, p); +#else + case XFS_IOC_ALLOCSP_32: + case XFS_IOC_FREESP_32: + case XFS_IOC_ALLOCSP64_32: + case XFS_IOC_FREESP64_32: + case XFS_IOC_RESVSP_32: + case XFS_IOC_UNRESVSP_32: + case XFS_IOC_RESVSP64_32: + case XFS_IOC_UNRESVSP64_32: { + struct xfs_flock64 bf; - /* xfs_bstat_t still has wrong u32 vs u64 alignment */ - case XFS_IOC_SWAPEXT: - break; + if (xfs_compat_flock64_copyin(&bf, arg)) + return -XFS_ERROR(EFAULT); + cmd = _NATIVE_IOC(cmd, struct xfs_flock64); + return xfs_ioc_space(ip, inode, filp, ioflags, cmd, &bf); + } + case XFS_IOC_FSGEOMETRY_V1_32: + return xfs_compat_ioc_fsgeometry_v1(mp, arg); + case XFS_IOC_FSGROWFSDATA_32: { + struct xfs_growfs_data in; + + if (xfs_compat_growfs_data_copyin(&in, arg)) + return -XFS_ERROR(EFAULT); + error = xfs_growfs_data(mp, &in); + return -error; + } + case XFS_IOC_FSGROWFSRT_32: { + struct xfs_growfs_rt in; + if (xfs_compat_growfs_rt_copyin(&in, arg)) + return -XFS_ERROR(EFAULT); + error = xfs_growfs_rt(mp, &in); + return -error; + } #endif + /* long changes size, but xfs only copiese out 32 bits */ + case XFS_IOC_GETXFLAGS_32: + case XFS_IOC_SETXFLAGS_32: + case XFS_IOC_GETVERSION_32: + cmd = _NATIVE_IOC(cmd, long); + return xfs_file_ioctl(filp, cmd, p); + case XFS_IOC_SWAPEXT: { + struct xfs_swapext sxp; + struct compat_xfs_swapext __user *sxu = arg; + + /* Bulk copy in up to the sx_stat field, then copy bstat */ + if (copy_from_user(&sxp, sxu, + offsetof(struct xfs_swapext, sx_stat)) || + xfs_ioctl32_bstat_copyin(&sxp.sx_stat, &sxu->sx_stat)) + return -XFS_ERROR(EFAULT); + error = xfs_swapext(&sxp); + return -error; + } case XFS_IOC_FSBULKSTAT_32: case XFS_IOC_FSBULKSTAT_SINGLE_32: case XFS_IOC_FSINUMBERS_32: - cmd = _NATIVE_IOC(cmd, struct xfs_fsop_bulkreq); - return xfs_ioc_bulkstat_compat(XFS_I(inode)->i_mount, - cmd, (void __user*)arg); + return xfs_compat_ioc_bulkstat(mp, cmd, arg); case XFS_IOC_FD_TO_HANDLE_32: case XFS_IOC_PATH_TO_HANDLE_32: - case XFS_IOC_PATH_TO_FSHANDLE_32: - case XFS_IOC_OPEN_BY_HANDLE_32: - case XFS_IOC_READLINK_BY_HANDLE_32: - arg = xfs_ioctl32_fshandle(arg); + case XFS_IOC_PATH_TO_FSHANDLE_32: { + struct xfs_fsop_handlereq hreq; + + if (xfs_compat_handlereq_copyin(&hreq, arg)) + return -XFS_ERROR(EFAULT); cmd = _NATIVE_IOC(cmd, struct xfs_fsop_handlereq); - break; - default: - return -ENOIOCTLCMD; + return xfs_find_handle(cmd, &hreq); } + case XFS_IOC_OPEN_BY_HANDLE_32: { + struct xfs_fsop_handlereq hreq; - error = xfs_ioctl(XFS_I(inode), file, mode, cmd, (void __user *)arg); - xfs_iflags_set(XFS_I(inode), XFS_IMODIFIED); - - return error; -} - -long -xfs_file_compat_ioctl( - struct file *file, - unsigned cmd, - unsigned long arg) -{ - return xfs_compat_ioctl(0, file, cmd, arg); -} + if (xfs_compat_handlereq_copyin(&hreq, arg)) + return -XFS_ERROR(EFAULT); + return xfs_open_by_handle(mp, &hreq, filp, inode); + } + case XFS_IOC_READLINK_BY_HANDLE_32: { + struct xfs_fsop_handlereq hreq; -long -xfs_file_compat_invis_ioctl( - struct file *file, - unsigned cmd, - unsigned long arg) -{ - return xfs_compat_ioctl(IO_INVIS, file, cmd, arg); + if (xfs_compat_handlereq_copyin(&hreq, arg)) + return -XFS_ERROR(EFAULT); + return xfs_readlink_by_handle(mp, &hreq, inode); + } + case XFS_IOC_ATTRLIST_BY_HANDLE_32: + return xfs_compat_attrlist_by_handle(mp, arg, inode); + case XFS_IOC_ATTRMULTI_BY_HANDLE_32: + return xfs_compat_attrmulti_by_handle(mp, arg, inode); + case XFS_IOC_FSSETDM_BY_HANDLE_32: + return xfs_compat_fssetdm_by_handle(mp, arg, inode); + default: + return -XFS_ERROR(ENOIOCTLCMD); + } } diff --git a/fs/xfs/linux-2.6/xfs_ioctl32.h b/fs/xfs/linux-2.6/xfs_ioctl32.h index 02de6e62ee3..1024c4f8ba0 100644 --- a/fs/xfs/linux-2.6/xfs_ioctl32.h +++ b/fs/xfs/linux-2.6/xfs_ioctl32.h @@ -18,7 +18,217 @@ #ifndef __XFS_IOCTL32_H__ #define __XFS_IOCTL32_H__ -extern long xfs_file_compat_ioctl(struct file *, unsigned, unsigned long); -extern long xfs_file_compat_invis_ioctl(struct file *, unsigned, unsigned long); +#include <linux/compat.h> + +/* + * on 32-bit arches, ioctl argument structures may have different sizes + * and/or alignment. We define compat structures which match the + * 32-bit sizes/alignments here, and their associated ioctl numbers. + * + * xfs_ioctl32.c contains routines to copy these structures in and out. + */ + +/* stock kernel-level ioctls we support */ +#define XFS_IOC_GETXFLAGS_32 FS_IOC32_GETFLAGS +#define XFS_IOC_SETXFLAGS_32 FS_IOC32_SETFLAGS +#define XFS_IOC_GETVERSION_32 FS_IOC32_GETVERSION + +/* + * On intel, even if sizes match, alignment and/or padding may differ. + */ +#if defined(CONFIG_IA64) || defined(CONFIG_X86_64) +#define BROKEN_X86_ALIGNMENT +#define __compat_packed __attribute__((packed)) +#else +#define __compat_packed +#endif + +typedef struct compat_xfs_bstime { + compat_time_t tv_sec; /* seconds */ + __s32 tv_nsec; /* and nanoseconds */ +} compat_xfs_bstime_t; + +typedef struct compat_xfs_bstat { + __u64 bs_ino; /* inode number */ + __u16 bs_mode; /* type and mode */ + __u16 bs_nlink; /* number of links */ + __u32 bs_uid; /* user id */ + __u32 bs_gid; /* group id */ + __u32 bs_rdev; /* device value */ + __s32 bs_blksize; /* block size */ + __s64 bs_size; /* file size */ + compat_xfs_bstime_t bs_atime; /* access time */ + compat_xfs_bstime_t bs_mtime; /* modify time */ + compat_xfs_bstime_t bs_ctime; /* inode change time */ + int64_t bs_blocks; /* number of blocks */ + __u32 bs_xflags; /* extended flags */ + __s32 bs_extsize; /* extent size */ + __s32 bs_extents; /* number of extents */ + __u32 bs_gen; /* generation count */ + __u16 bs_projid; /* project id */ + unsigned char bs_pad[14]; /* pad space, unused */ + __u32 bs_dmevmask; /* DMIG event mask */ + __u16 bs_dmstate; /* DMIG state info */ + __u16 bs_aextents; /* attribute number of extents */ +} __compat_packed compat_xfs_bstat_t; + +typedef struct compat_xfs_fsop_bulkreq { + compat_uptr_t lastip; /* last inode # pointer */ + __s32 icount; /* count of entries in buffer */ + compat_uptr_t ubuffer; /* user buffer for inode desc. */ + compat_uptr_t ocount; /* output count pointer */ +} compat_xfs_fsop_bulkreq_t; + +#define XFS_IOC_FSBULKSTAT_32 \ + _IOWR('X', 101, struct compat_xfs_fsop_bulkreq) +#define XFS_IOC_FSBULKSTAT_SINGLE_32 \ + _IOWR('X', 102, struct compat_xfs_fsop_bulkreq) +#define XFS_IOC_FSINUMBERS_32 \ + _IOWR('X', 103, struct compat_xfs_fsop_bulkreq) + +typedef struct compat_xfs_fsop_handlereq { + __u32 fd; /* fd for FD_TO_HANDLE */ + compat_uptr_t path; /* user pathname */ + __u32 oflags; /* open flags */ + compat_uptr_t ihandle; /* user supplied handle */ + __u32 ihandlen; /* user supplied length */ + compat_uptr_t ohandle; /* user buffer for handle */ + compat_uptr_t ohandlen; /* user buffer length */ +} compat_xfs_fsop_handlereq_t; + +#define XFS_IOC_PATH_TO_FSHANDLE_32 \ + _IOWR('X', 104, struct compat_xfs_fsop_handlereq) +#define XFS_IOC_PATH_TO_HANDLE_32 \ + _IOWR('X', 105, struct compat_xfs_fsop_handlereq) +#define XFS_IOC_FD_TO_HANDLE_32 \ + _IOWR('X', 106, struct compat_xfs_fsop_handlereq) +#define XFS_IOC_OPEN_BY_HANDLE_32 \ + _IOWR('X', 107, struct compat_xfs_fsop_handlereq) +#define XFS_IOC_READLINK_BY_HANDLE_32 \ + _IOWR('X', 108, struct compat_xfs_fsop_handlereq) + +/* The bstat field in the swapext struct needs translation */ +typedef struct compat_xfs_swapext { + __int64_t sx_version; /* version */ + __int64_t sx_fdtarget; /* fd of target file */ + __int64_t sx_fdtmp; /* fd of tmp file */ + xfs_off_t sx_offset; /* offset into file */ + xfs_off_t sx_length; /* leng from offset */ + char sx_pad[16]; /* pad space, unused */ + compat_xfs_bstat_t sx_stat; /* stat of target b4 copy */ +} __compat_packed compat_xfs_swapext_t; + +#define XFS_IOC_SWAPEXT_32 _IOWR('X', 109, struct compat_xfs_swapext) + +typedef struct compat_xfs_fsop_attrlist_handlereq { + struct compat_xfs_fsop_handlereq hreq; /* handle interface structure */ + struct xfs_attrlist_cursor pos; /* opaque cookie, list offset */ + __u32 flags; /* which namespace to use */ + __u32 buflen; /* length of buffer supplied */ + compat_uptr_t buffer; /* returned names */ +} __compat_packed compat_xfs_fsop_attrlist_handlereq_t; + +/* Note: actually this is read/write */ +#define XFS_IOC_ATTRLIST_BY_HANDLE_32 \ + _IOW('X', 122, struct compat_xfs_fsop_attrlist_handlereq) + +/* am_opcodes defined in xfs_fs.h */ +typedef struct compat_xfs_attr_multiop { + __u32 am_opcode; + __s32 am_error; + compat_uptr_t am_attrname; + compat_uptr_t am_attrvalue; + __u32 am_length; + __u32 am_flags; +} compat_xfs_attr_multiop_t; + +typedef struct compat_xfs_fsop_attrmulti_handlereq { + struct compat_xfs_fsop_handlereq hreq; /* handle interface structure */ + __u32 opcount;/* count of following multiop */ + /* ptr to compat_xfs_attr_multiop */ + compat_uptr_t ops; /* attr_multi data */ +} compat_xfs_fsop_attrmulti_handlereq_t; + +#define XFS_IOC_ATTRMULTI_BY_HANDLE_32 \ + _IOW('X', 123, struct compat_xfs_fsop_attrmulti_handlereq) + +typedef struct compat_xfs_fsop_setdm_handlereq { + struct compat_xfs_fsop_handlereq hreq; /* handle information */ + /* ptr to struct fsdmidata */ + compat_uptr_t data; /* DMAPI data */ +} compat_xfs_fsop_setdm_handlereq_t; + +#define XFS_IOC_FSSETDM_BY_HANDLE_32 \ + _IOW('X', 121, struct compat_xfs_fsop_setdm_handlereq) + +#ifdef BROKEN_X86_ALIGNMENT +/* on ia32 l_start is on a 32-bit boundary */ +typedef struct compat_xfs_flock64 { + __s16 l_type; + __s16 l_whence; + __s64 l_start __attribute__((packed)); + /* len == 0 means until end of file */ + __s64 l_len __attribute__((packed)); + __s32 l_sysid; + __u32 l_pid; + __s32 l_pad[4]; /* reserve area */ +} compat_xfs_flock64_t; + +#define XFS_IOC_ALLOCSP_32 _IOW('X', 10, struct compat_xfs_flock64) +#define XFS_IOC_FREESP_32 _IOW('X', 11, struct compat_xfs_flock64) +#define XFS_IOC_ALLOCSP64_32 _IOW('X', 36, struct compat_xfs_flock64) +#define XFS_IOC_FREESP64_32 _IOW('X', 37, struct compat_xfs_flock64) +#define XFS_IOC_RESVSP_32 _IOW('X', 40, struct compat_xfs_flock64) +#define XFS_IOC_UNRESVSP_32 _IOW('X', 41, struct compat_xfs_flock64) +#define XFS_IOC_RESVSP64_32 _IOW('X', 42, struct compat_xfs_flock64) +#define XFS_IOC_UNRESVSP64_32 _IOW('X', 43, struct compat_xfs_flock64) + +typedef struct compat_xfs_fsop_geom_v1 { + __u32 blocksize; /* filesystem (data) block size */ + __u32 rtextsize; /* realtime extent size */ + __u32 agblocks; /* fsblocks in an AG */ + __u32 agcount; /* number of allocation groups */ + __u32 logblocks; /* fsblocks in the log */ + __u32 sectsize; /* (data) sector size, bytes */ + __u32 inodesize; /* inode size in bytes */ + __u32 imaxpct; /* max allowed inode space(%) */ + __u64 datablocks; /* fsblocks in data subvolume */ + __u64 rtblocks; /* fsblocks in realtime subvol */ + __u64 rtextents; /* rt extents in realtime subvol*/ + __u64 logstart; /* starting fsblock of the log */ + unsigned char uuid[16]; /* unique id of the filesystem */ + __u32 sunit; /* stripe unit, fsblocks */ + __u32 swidth; /* stripe width, fsblocks */ + __s32 version; /* structure version */ + __u32 flags; /* superblock version flags */ + __u32 logsectsize; /* log sector size, bytes */ + __u32 rtsectsize; /* realtime sector size, bytes */ + __u32 dirblocksize; /* directory block size, bytes */ +} __attribute__((packed)) compat_xfs_fsop_geom_v1_t; + +#define XFS_IOC_FSGEOMETRY_V1_32 \ + _IOR('X', 100, struct compat_xfs_fsop_geom_v1) + +typedef struct compat_xfs_inogrp { + __u64 xi_startino; /* starting inode number */ + __s32 xi_alloccount; /* # bits set in allocmask */ + __u64 xi_allocmask; /* mask of allocated inodes */ +} __attribute__((packed)) compat_xfs_inogrp_t; + +/* These growfs input structures have padding on the end, so must translate */ +typedef struct compat_xfs_growfs_data { + __u64 newblocks; /* new data subvol size, fsblocks */ + __u32 imaxpct; /* new inode space percentage limit */ +} __attribute__((packed)) compat_xfs_growfs_data_t; + +typedef struct compat_xfs_growfs_rt { + __u64 newblocks; /* new realtime size, fsblocks */ + __u32 extsize; /* new realtime extent size, fsblocks */ +} __attribute__((packed)) compat_xfs_growfs_rt_t; + +#define XFS_IOC_FSGROWFSDATA_32 _IOW('X', 110, struct compat_xfs_growfs_data) +#define XFS_IOC_FSGROWFSRT_32 _IOW('X', 112, struct compat_xfs_growfs_rt) + +#endif /* BROKEN_X86_ALIGNMENT */ #endif /* __XFS_IOCTL32_H__ */ diff --git a/fs/xfs/linux-2.6/xfs_iops.c b/fs/xfs/linux-2.6/xfs_iops.c index 095d271f343..7aa53fefc67 100644 --- a/fs/xfs/linux-2.6/xfs_iops.c +++ b/fs/xfs/linux-2.6/xfs_iops.c @@ -53,6 +53,7 @@ #include <linux/namei.h> #include <linux/security.h> #include <linux/falloc.h> +#include <linux/fiemap.h> /* * Bring the atime in the XFS inode uptodate. @@ -64,14 +65,14 @@ xfs_synchronize_atime( { struct inode *inode = VFS_I(ip); - if (inode) { + if (!(inode->i_state & I_CLEAR)) { ip->i_d.di_atime.t_sec = (__int32_t)inode->i_atime.tv_sec; ip->i_d.di_atime.t_nsec = (__int32_t)inode->i_atime.tv_nsec; } } /* - * If the linux inode exists, mark it dirty. + * If the linux inode is valid, mark it dirty. * Used when commiting a dirty inode into a transaction so that * the inode will get written back by the linux code */ @@ -81,7 +82,7 @@ xfs_mark_inode_dirty_sync( { struct inode *inode = VFS_I(ip); - if (inode) + if (!(inode->i_state & (I_WILL_FREE|I_FREEING|I_CLEAR))) mark_inode_dirty_sync(inode); } @@ -128,7 +129,7 @@ xfs_ichgtime( if (sync_it) { SYNCHRONIZE(); ip->i_update_core = 1; - mark_inode_dirty_sync(inode); + xfs_mark_inode_dirty_sync(ip); } } @@ -158,8 +159,6 @@ xfs_init_security( } error = xfs_attr_set(ip, name, value, length, ATTR_SECURE); - if (!error) - xfs_iflags_set(ip, XFS_IMODIFIED); kfree(name); kfree(value); @@ -260,7 +259,6 @@ xfs_vn_mknod( error = _ACL_INHERIT(inode, mode, default_acl); if (unlikely(error)) goto out_cleanup_inode; - xfs_iflags_set(ip, XFS_IMODIFIED); _ACL_FREE(default_acl); } @@ -366,21 +364,17 @@ xfs_vn_link( struct inode *dir, struct dentry *dentry) { - struct inode *inode; /* inode of guy being linked to */ + struct inode *inode = old_dentry->d_inode; struct xfs_name name; int error; - inode = old_dentry->d_inode; xfs_dentry_to_name(&name, dentry); - igrab(inode); error = xfs_link(XFS_I(dir), XFS_I(inode), &name); - if (unlikely(error)) { - iput(inode); + if (unlikely(error)) return -error; - } - xfs_iflags_set(XFS_I(dir), XFS_IMODIFIED); + atomic_inc(&inode->i_count); d_instantiate(dentry, inode); return 0; } @@ -601,7 +595,7 @@ xfs_vn_setattr( struct dentry *dentry, struct iattr *iattr) { - return -xfs_setattr(XFS_I(dentry->d_inode), iattr, 0, NULL); + return -xfs_setattr(XFS_I(dentry->d_inode), iattr, 0); } /* @@ -642,7 +636,7 @@ xfs_vn_fallocate( xfs_ilock(ip, XFS_IOLOCK_EXCL); error = xfs_change_file_space(ip, XFS_IOC_RESVSP, &bf, - 0, NULL, XFS_ATTR_NOLOCK); + 0, XFS_ATTR_NOLOCK); if (!error && !(mode & FALLOC_FL_KEEP_SIZE) && offset + len > i_size_read(inode)) new_size = offset + len; @@ -653,7 +647,7 @@ xfs_vn_fallocate( iattr.ia_valid = ATTR_SIZE; iattr.ia_size = new_size; - error = xfs_setattr(ip, &iattr, XFS_ATTR_NOLOCK, NULL); + error = xfs_setattr(ip, &iattr, XFS_ATTR_NOLOCK); } xfs_iunlock(ip, XFS_IOLOCK_EXCL); @@ -661,6 +655,88 @@ out_error: return error; } +#define XFS_FIEMAP_FLAGS (FIEMAP_FLAG_SYNC|FIEMAP_FLAG_XATTR) + +/* + * Call fiemap helper to fill in user data. + * Returns positive errors to xfs_getbmap. + */ +STATIC int +xfs_fiemap_format( + void **arg, + struct getbmapx *bmv, + int *full) +{ + int error; + struct fiemap_extent_info *fieinfo = *arg; + u32 fiemap_flags = 0; + u64 logical, physical, length; + + /* Do nothing for a hole */ + if (bmv->bmv_block == -1LL) + return 0; + + logical = BBTOB(bmv->bmv_offset); + physical = BBTOB(bmv->bmv_block); + length = BBTOB(bmv->bmv_length); + + if (bmv->bmv_oflags & BMV_OF_PREALLOC) + fiemap_flags |= FIEMAP_EXTENT_UNWRITTEN; + else if (bmv->bmv_oflags & BMV_OF_DELALLOC) { + fiemap_flags |= FIEMAP_EXTENT_DELALLOC; + physical = 0; /* no block yet */ + } + if (bmv->bmv_oflags & BMV_OF_LAST) + fiemap_flags |= FIEMAP_EXTENT_LAST; + + error = fiemap_fill_next_extent(fieinfo, logical, physical, + length, fiemap_flags); + if (error > 0) { + error = 0; + *full = 1; /* user array now full */ + } + + return -error; +} + +STATIC int +xfs_vn_fiemap( + struct inode *inode, + struct fiemap_extent_info *fieinfo, + u64 start, + u64 length) +{ + xfs_inode_t *ip = XFS_I(inode); + struct getbmapx bm; + int error; + + error = fiemap_check_flags(fieinfo, XFS_FIEMAP_FLAGS); + if (error) + return error; + + /* Set up bmap header for xfs internal routine */ + bm.bmv_offset = BTOBB(start); + /* Special case for whole file */ + if (length == FIEMAP_MAX_OFFSET) + bm.bmv_length = -1LL; + else + bm.bmv_length = BTOBB(length); + + /* our formatter will tell xfs_getbmap when to stop. */ + bm.bmv_count = MAXEXTNUM; + bm.bmv_iflags = BMV_IF_PREALLOC; + if (fieinfo->fi_flags & FIEMAP_FLAG_XATTR) + bm.bmv_iflags |= BMV_IF_ATTRFORK; + if (!(fieinfo->fi_flags & FIEMAP_FLAG_SYNC)) + bm.bmv_iflags |= BMV_IF_DELALLOC; + + error = xfs_getbmap(ip, &bm, xfs_fiemap_format, fieinfo); + if (error) + return -error; + + return 0; +} + static const struct inode_operations xfs_inode_operations = { .permission = xfs_vn_permission, .truncate = xfs_vn_truncate, @@ -671,6 +747,7 @@ static const struct inode_operations xfs_inode_operations = { .removexattr = generic_removexattr, .listxattr = xfs_vn_listxattr, .fallocate = xfs_vn_fallocate, + .fiemap = xfs_vn_fiemap, }; static const struct inode_operations xfs_dir_inode_operations = { @@ -766,12 +843,20 @@ xfs_diflags_to_iflags( * When reading existing inodes from disk this is called directly * from xfs_iget, when creating a new inode it is called from * xfs_ialloc after setting up the inode. + * + * We are always called with an uninitialised linux inode here. + * We need to initialise the necessary fields and take a reference + * on it. */ void xfs_setup_inode( struct xfs_inode *ip) { - struct inode *inode = ip->i_vnode; + struct inode *inode = &ip->i_vnode; + + inode->i_ino = ip->i_ino; + inode->i_state = I_NEW|I_LOCK; + inode_add_to_lists(ip->i_mount->m_super, inode); inode->i_mode = ip->i_d.di_mode; inode->i_nlink = ip->i_d.di_nlink; @@ -799,7 +884,6 @@ xfs_setup_inode( inode->i_ctime.tv_sec = ip->i_d.di_ctime.t_sec; inode->i_ctime.tv_nsec = ip->i_d.di_ctime.t_nsec; xfs_diflags_to_iflags(inode, ip); - xfs_iflags_clear(ip, XFS_IMODIFIED); switch (inode->i_mode & S_IFMT) { case S_IFREG: diff --git a/fs/xfs/linux-2.6/xfs_iops.h b/fs/xfs/linux-2.6/xfs_iops.h index 8b1a1e31dc2..ef41c92ce66 100644 --- a/fs/xfs/linux-2.6/xfs_iops.h +++ b/fs/xfs/linux-2.6/xfs_iops.h @@ -22,7 +22,6 @@ struct xfs_inode; extern const struct file_operations xfs_file_operations; extern const struct file_operations xfs_dir_file_operations; -extern const struct file_operations xfs_invis_file_operations; extern ssize_t xfs_vn_listxattr(struct dentry *, char *data, size_t size); diff --git a/fs/xfs/linux-2.6/xfs_linux.h b/fs/xfs/linux-2.6/xfs_linux.h index cc0f7b3a979..507492d6dcc 100644 --- a/fs/xfs/linux-2.6/xfs_linux.h +++ b/fs/xfs/linux-2.6/xfs_linux.h @@ -21,18 +21,12 @@ #include <linux/types.h> /* - * Some types are conditional depending on the target system. * XFS_BIG_BLKNOS needs block layer disk addresses to be 64 bits. - * XFS_BIG_INUMS needs the VFS inode number to be 64 bits, as well - * as requiring XFS_BIG_BLKNOS to be set. + * XFS_BIG_INUMS requires XFS_BIG_BLKNOS to be set. */ #if defined(CONFIG_LBD) || (BITS_PER_LONG == 64) # define XFS_BIG_BLKNOS 1 -# if BITS_PER_LONG == 64 -# define XFS_BIG_INUMS 1 -# else -# define XFS_BIG_INUMS 0 -# endif +# define XFS_BIG_INUMS 1 #else # define XFS_BIG_BLKNOS 0 # define XFS_BIG_INUMS 0 @@ -77,6 +71,7 @@ #include <linux/spinlock.h> #include <linux/random.h> #include <linux/ctype.h> +#include <linux/writeback.h> #include <asm/page.h> #include <asm/div64.h> @@ -85,7 +80,6 @@ #include <asm/byteorder.h> #include <asm/unaligned.h> -#include <xfs_vfs.h> #include <xfs_cred.h> #include <xfs_vnode.h> #include <xfs_stats.h> @@ -107,7 +101,6 @@ #undef HAVE_PERCPU_SB /* per cpu superblock counters are a 2.6 feature */ #endif -#define restricted_chown xfs_params.restrict_chown.val #define irix_sgid_inherit xfs_params.sgid_inherit.val #define irix_symlink_mode xfs_params.symlink_mode.val #define xfs_panic_mask xfs_params.panic_mask.val diff --git a/fs/xfs/linux-2.6/xfs_lrw.c b/fs/xfs/linux-2.6/xfs_lrw.c index 1957e5357d0..7e90daa0d1d 100644 --- a/fs/xfs/linux-2.6/xfs_lrw.c +++ b/fs/xfs/linux-2.6/xfs_lrw.c @@ -51,7 +51,6 @@ #include "xfs_vnodeops.h" #include <linux/capability.h> -#include <linux/mount.h> #include <linux/writeback.h> @@ -243,7 +242,7 @@ xfs_read( if (unlikely(ioflags & IO_ISDIRECT)) { if (inode->i_mapping->nrpages) - ret = xfs_flushinval_pages(ip, (*offset & PAGE_CACHE_MASK), + ret = -xfs_flushinval_pages(ip, (*offset & PAGE_CACHE_MASK), -1, FI_REMAPF_LOCKED); mutex_unlock(&inode->i_mutex); if (ret) { @@ -668,15 +667,8 @@ start: if (new_size > xip->i_size) xip->i_new_size = new_size; - /* - * We're not supposed to change timestamps in readonly-mounted - * filesystems. Throw it away if anyone asks us. - */ - if (likely(!(ioflags & IO_INVIS) && - !mnt_want_write(file->f_path.mnt))) { + if (likely(!(ioflags & IO_INVIS))) xfs_ichgtime(xip, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG); - mnt_drop_write(file->f_path.mnt); - } /* * If the offset is beyond the size of the file, we have a couple @@ -715,7 +707,6 @@ start: } } -retry: /* We can write back this queue in page reclaim */ current->backing_dev_info = mapping->backing_dev_info; @@ -771,6 +762,17 @@ retry: if (ret == -EIOCBQUEUED && !(ioflags & IO_ISAIO)) ret = wait_on_sync_kiocb(iocb); + isize = i_size_read(inode); + if (unlikely(ret < 0 && ret != -EFAULT && *offset > isize)) + *offset = isize; + + if (*offset > xip->i_size) { + xfs_ilock(xip, XFS_ILOCK_EXCL); + if (*offset > xip->i_size) + xip->i_size = *offset; + xfs_iunlock(xip, XFS_ILOCK_EXCL); + } + if (ret == -ENOSPC && DM_EVENT_ENABLED(xip, DM_EVENT_NOSPACE) && !(ioflags & IO_INVIS)) { xfs_iunlock(xip, iolock); @@ -784,20 +786,7 @@ retry: xfs_ilock(xip, iolock); if (error) goto out_unlock_internal; - pos = xip->i_size; - ret = 0; - goto retry; - } - - isize = i_size_read(inode); - if (unlikely(ret < 0 && ret != -EFAULT && *offset > isize)) - *offset = isize; - - if (*offset > xip->i_size) { - xfs_ilock(xip, XFS_ILOCK_EXCL); - if (*offset > xip->i_size) - xip->i_size = *offset; - xfs_iunlock(xip, XFS_ILOCK_EXCL); + goto start; } error = -ret; @@ -855,13 +844,7 @@ retry: int xfs_bdstrat_cb(struct xfs_buf *bp) { - xfs_mount_t *mp; - - mp = XFS_BUF_FSPRIVATE3(bp, xfs_mount_t *); - if (!XFS_FORCED_SHUTDOWN(mp)) { - xfs_buf_iorequest(bp); - return 0; - } else { + if (XFS_FORCED_SHUTDOWN(bp->b_mount)) { xfs_buftrace("XFS__BDSTRAT IOERROR", bp); /* * Metadata write that didn't get logged but @@ -874,6 +857,9 @@ xfs_bdstrat_cb(struct xfs_buf *bp) else return (xfs_bioerror(bp)); } + + xfs_buf_iorequest(bp); + return 0; } /* diff --git a/fs/xfs/linux-2.6/xfs_stats.c b/fs/xfs/linux-2.6/xfs_stats.c index 3d5b67c075c..c3526d445f6 100644 --- a/fs/xfs/linux-2.6/xfs_stats.c +++ b/fs/xfs/linux-2.6/xfs_stats.c @@ -53,11 +53,15 @@ xfs_read_xfsstats( { "icluster", XFSSTAT_END_INODE_CLUSTER }, { "vnodes", XFSSTAT_END_VNODE_OPS }, { "buf", XFSSTAT_END_BUF }, + { "abtb2", XFSSTAT_END_ABTB_V2 }, + { "abtc2", XFSSTAT_END_ABTC_V2 }, + { "bmbt2", XFSSTAT_END_BMBT_V2 }, + { "ibt2", XFSSTAT_END_IBT_V2 }, }; /* Loop over all stats groups */ for (i=j=len = 0; i < ARRAY_SIZE(xstats); i++) { - len += sprintf(buffer + len, xstats[i].desc); + len += sprintf(buffer + len, "%s", xstats[i].desc); /* inner loop does each group */ while (j < xstats[i].endpoint) { val = 0; diff --git a/fs/xfs/linux-2.6/xfs_stats.h b/fs/xfs/linux-2.6/xfs_stats.h index e83820febc9..736854b1ca1 100644 --- a/fs/xfs/linux-2.6/xfs_stats.h +++ b/fs/xfs/linux-2.6/xfs_stats.h @@ -118,6 +118,71 @@ struct xfsstats { __uint32_t xb_page_retries; __uint32_t xb_page_found; __uint32_t xb_get_read; +/* Version 2 btree counters */ +#define XFSSTAT_END_ABTB_V2 (XFSSTAT_END_BUF+15) + __uint32_t xs_abtb_2_lookup; + __uint32_t xs_abtb_2_compare; + __uint32_t xs_abtb_2_insrec; + __uint32_t xs_abtb_2_delrec; + __uint32_t xs_abtb_2_newroot; + __uint32_t xs_abtb_2_killroot; + __uint32_t xs_abtb_2_increment; + __uint32_t xs_abtb_2_decrement; + __uint32_t xs_abtb_2_lshift; + __uint32_t xs_abtb_2_rshift; + __uint32_t xs_abtb_2_split; + __uint32_t xs_abtb_2_join; + __uint32_t xs_abtb_2_alloc; + __uint32_t xs_abtb_2_free; + __uint32_t xs_abtb_2_moves; +#define XFSSTAT_END_ABTC_V2 (XFSSTAT_END_ABTB_V2+15) + __uint32_t xs_abtc_2_lookup; + __uint32_t xs_abtc_2_compare; + __uint32_t xs_abtc_2_insrec; + __uint32_t xs_abtc_2_delrec; + __uint32_t xs_abtc_2_newroot; + __uint32_t xs_abtc_2_killroot; + __uint32_t xs_abtc_2_increment; + __uint32_t xs_abtc_2_decrement; + __uint32_t xs_abtc_2_lshift; + __uint32_t xs_abtc_2_rshift; + __uint32_t xs_abtc_2_split; + __uint32_t xs_abtc_2_join; + __uint32_t xs_abtc_2_alloc; + __uint32_t xs_abtc_2_free; + __uint32_t xs_abtc_2_moves; +#define XFSSTAT_END_BMBT_V2 (XFSSTAT_END_ABTC_V2+15) + __uint32_t xs_bmbt_2_lookup; + __uint32_t xs_bmbt_2_compare; + __uint32_t xs_bmbt_2_insrec; + __uint32_t xs_bmbt_2_delrec; + __uint32_t xs_bmbt_2_newroot; + __uint32_t xs_bmbt_2_killroot; + __uint32_t xs_bmbt_2_increment; + __uint32_t xs_bmbt_2_decrement; + __uint32_t xs_bmbt_2_lshift; + __uint32_t xs_bmbt_2_rshift; + __uint32_t xs_bmbt_2_split; + __uint32_t xs_bmbt_2_join; + __uint32_t xs_bmbt_2_alloc; + __uint32_t xs_bmbt_2_free; + __uint32_t xs_bmbt_2_moves; +#define XFSSTAT_END_IBT_V2 (XFSSTAT_END_BMBT_V2+15) + __uint32_t xs_ibt_2_lookup; + __uint32_t xs_ibt_2_compare; + __uint32_t xs_ibt_2_insrec; + __uint32_t xs_ibt_2_delrec; + __uint32_t xs_ibt_2_newroot; + __uint32_t xs_ibt_2_killroot; + __uint32_t xs_ibt_2_increment; + __uint32_t xs_ibt_2_decrement; + __uint32_t xs_ibt_2_lshift; + __uint32_t xs_ibt_2_rshift; + __uint32_t xs_ibt_2_split; + __uint32_t xs_ibt_2_join; + __uint32_t xs_ibt_2_alloc; + __uint32_t xs_ibt_2_free; + __uint32_t xs_ibt_2_moves; /* Extra precision counters */ __uint64_t xs_xstrat_bytes; __uint64_t xs_write_bytes; diff --git a/fs/xfs/linux-2.6/xfs_super.c b/fs/xfs/linux-2.6/xfs_super.c index 7227b2efef2..36f6cc703ef 100644 --- a/fs/xfs/linux-2.6/xfs_super.c +++ b/fs/xfs/linux-2.6/xfs_super.c @@ -18,7 +18,6 @@ #include "xfs.h" #include "xfs_bit.h" #include "xfs_log.h" -#include "xfs_clnt.h" #include "xfs_inum.h" #include "xfs_trans.h" #include "xfs_sb.h" @@ -36,6 +35,7 @@ #include "xfs_dinode.h" #include "xfs_inode.h" #include "xfs_btree.h" +#include "xfs_btree_trace.h" #include "xfs_ialloc.h" #include "xfs_bmap.h" #include "xfs_rtalloc.h" @@ -48,7 +48,6 @@ #include "xfs_buf_item.h" #include "xfs_utils.h" #include "xfs_vnodeops.h" -#include "xfs_vfsops.h" #include "xfs_version.h" #include "xfs_log_priv.h" #include "xfs_trans_priv.h" @@ -58,6 +57,7 @@ #include "xfs_extfree_item.h" #include "xfs_mru_cache.h" #include "xfs_inode_item.h" +#include "xfs_sync.h" #include <linux/namei.h> #include <linux/init.h> @@ -70,36 +70,9 @@ static struct quotactl_ops xfs_quotactl_operations; static struct super_operations xfs_super_operations; -static kmem_zone_t *xfs_vnode_zone; static kmem_zone_t *xfs_ioend_zone; mempool_t *xfs_ioend_pool; -STATIC struct xfs_mount_args * -xfs_args_allocate( - struct super_block *sb, - int silent) -{ - struct xfs_mount_args *args; - - args = kzalloc(sizeof(struct xfs_mount_args), GFP_KERNEL); - if (!args) - return NULL; - - args->logbufs = args->logbufsize = -1; - strncpy(args->fsname, sb->s_id, MAXNAMELEN); - - /* Copy the already-parsed mount(2) flags we're interested in */ - if (sb->s_flags & MS_DIRSYNC) - args->flags |= XFSMNT_DIRSYNC; - if (sb->s_flags & MS_SYNCHRONOUS) - args->flags |= XFSMNT_WSYNC; - if (silent) - args->flags |= XFSMNT_QUIET; - args->flags |= XFSMNT_32BITINODES; - - return args; -} - #define MNTOPT_LOGBUFS "logbufs" /* number of XFS log buffers */ #define MNTOPT_LOGBSIZE "logbsize" /* size of XFS log buffers */ #define MNTOPT_LOGDEV "logdev" /* log device */ @@ -188,26 +161,54 @@ suffix_strtoul(char *s, char **endp, unsigned int base) return simple_strtoul((const char *)s, endp, base) << shift_left_factor; } +/* + * This function fills in xfs_mount_t fields based on mount args. + * Note: the superblock has _not_ yet been read in. + * + * Note that this function leaks the various device name allocations on + * failure. The caller takes care of them. + */ STATIC int xfs_parseargs( struct xfs_mount *mp, char *options, - struct xfs_mount_args *args, - int update) + char **mtpt) { + struct super_block *sb = mp->m_super; char *this_char, *value, *eov; - int dsunit, dswidth, vol_dsunit, vol_dswidth; - int iosize; + int dsunit = 0; + int dswidth = 0; + int iosize = 0; int dmapi_implies_ikeep = 1; + uchar_t iosizelog = 0; + + /* + * Copy binary VFS mount flags we are interested in. + */ + if (sb->s_flags & MS_RDONLY) + mp->m_flags |= XFS_MOUNT_RDONLY; + if (sb->s_flags & MS_DIRSYNC) + mp->m_flags |= XFS_MOUNT_DIRSYNC; + if (sb->s_flags & MS_SYNCHRONOUS) + mp->m_flags |= XFS_MOUNT_WSYNC; + + /* + * Set some default flags that could be cleared by the mount option + * parsing. + */ + mp->m_flags |= XFS_MOUNT_BARRIER; + mp->m_flags |= XFS_MOUNT_COMPAT_IOSIZE; + mp->m_flags |= XFS_MOUNT_SMALL_INUMS; - args->flags |= XFSMNT_BARRIER; - args->flags2 |= XFSMNT2_COMPAT_IOSIZE; + /* + * These can be overridden by the mount option parsing. + */ + mp->m_logbufs = -1; + mp->m_logbsize = -1; if (!options) goto done; - iosize = dsunit = dswidth = vol_dsunit = vol_dswidth = 0; - while ((this_char = strsep(&options, ",")) != NULL) { if (!*this_char) continue; @@ -221,7 +222,7 @@ xfs_parseargs( this_char); return EINVAL; } - args->logbufs = simple_strtoul(value, &eov, 10); + mp->m_logbufs = simple_strtoul(value, &eov, 10); } else if (!strcmp(this_char, MNTOPT_LOGBSIZE)) { if (!value || !*value) { cmn_err(CE_WARN, @@ -229,7 +230,7 @@ xfs_parseargs( this_char); return EINVAL; } - args->logbufsize = suffix_strtoul(value, &eov, 10); + mp->m_logbsize = suffix_strtoul(value, &eov, 10); } else if (!strcmp(this_char, MNTOPT_LOGDEV)) { if (!value || !*value) { cmn_err(CE_WARN, @@ -237,7 +238,9 @@ xfs_parseargs( this_char); return EINVAL; } - strncpy(args->logname, value, MAXNAMELEN); + mp->m_logname = kstrndup(value, MAXNAMELEN, GFP_KERNEL); + if (!mp->m_logname) + return ENOMEM; } else if (!strcmp(this_char, MNTOPT_MTPT)) { if (!value || !*value) { cmn_err(CE_WARN, @@ -245,7 +248,9 @@ xfs_parseargs( this_char); return EINVAL; } - strncpy(args->mtpt, value, MAXNAMELEN); + *mtpt = kstrndup(value, MAXNAMELEN, GFP_KERNEL); + if (!*mtpt) + return ENOMEM; } else if (!strcmp(this_char, MNTOPT_RTDEV)) { if (!value || !*value) { cmn_err(CE_WARN, @@ -253,7 +258,9 @@ xfs_parseargs( this_char); return EINVAL; } - strncpy(args->rtname, value, MAXNAMELEN); + mp->m_rtname = kstrndup(value, MAXNAMELEN, GFP_KERNEL); + if (!mp->m_rtname) + return ENOMEM; } else if (!strcmp(this_char, MNTOPT_BIOSIZE)) { if (!value || !*value) { cmn_err(CE_WARN, @@ -262,8 +269,7 @@ xfs_parseargs( return EINVAL; } iosize = simple_strtoul(value, &eov, 10); - args->flags |= XFSMNT_IOSIZE; - args->iosizelog = (uint8_t) iosize; + iosizelog = ffs(iosize) - 1; } else if (!strcmp(this_char, MNTOPT_ALLOCSIZE)) { if (!value || !*value) { cmn_err(CE_WARN, @@ -272,8 +278,7 @@ xfs_parseargs( return EINVAL; } iosize = suffix_strtoul(value, &eov, 10); - args->flags |= XFSMNT_IOSIZE; - args->iosizelog = ffs(iosize) - 1; + iosizelog = ffs(iosize) - 1; } else if (!strcmp(this_char, MNTOPT_GRPID) || !strcmp(this_char, MNTOPT_BSDGROUPS)) { mp->m_flags |= XFS_MOUNT_GRPID; @@ -281,23 +286,25 @@ xfs_parseargs( !strcmp(this_char, MNTOPT_SYSVGROUPS)) { mp->m_flags &= ~XFS_MOUNT_GRPID; } else if (!strcmp(this_char, MNTOPT_WSYNC)) { - args->flags |= XFSMNT_WSYNC; + mp->m_flags |= XFS_MOUNT_WSYNC; } else if (!strcmp(this_char, MNTOPT_OSYNCISOSYNC)) { - args->flags |= XFSMNT_OSYNCISOSYNC; + mp->m_flags |= XFS_MOUNT_OSYNCISOSYNC; } else if (!strcmp(this_char, MNTOPT_NORECOVERY)) { - args->flags |= XFSMNT_NORECOVERY; + mp->m_flags |= XFS_MOUNT_NORECOVERY; } else if (!strcmp(this_char, MNTOPT_INO64)) { - args->flags |= XFSMNT_INO64; -#if !XFS_BIG_INUMS +#if XFS_BIG_INUMS + mp->m_flags |= XFS_MOUNT_INO64; + mp->m_inoadd = XFS_INO64_OFFSET; +#else cmn_err(CE_WARN, "XFS: %s option not allowed on this system", this_char); return EINVAL; #endif } else if (!strcmp(this_char, MNTOPT_NOALIGN)) { - args->flags |= XFSMNT_NOALIGN; + mp->m_flags |= XFS_MOUNT_NOALIGN; } else if (!strcmp(this_char, MNTOPT_SWALLOC)) { - args->flags |= XFSMNT_SWALLOC; + mp->m_flags |= XFS_MOUNT_SWALLOC; } else if (!strcmp(this_char, MNTOPT_SUNIT)) { if (!value || !*value) { cmn_err(CE_WARN, @@ -315,7 +322,7 @@ xfs_parseargs( } dswidth = simple_strtoul(value, &eov, 10); } else if (!strcmp(this_char, MNTOPT_64BITINODE)) { - args->flags &= ~XFSMNT_32BITINODES; + mp->m_flags &= ~XFS_MOUNT_SMALL_INUMS; #if !XFS_BIG_INUMS cmn_err(CE_WARN, "XFS: %s option not allowed on this system", @@ -323,56 +330,61 @@ xfs_parseargs( return EINVAL; #endif } else if (!strcmp(this_char, MNTOPT_NOUUID)) { - args->flags |= XFSMNT_NOUUID; + mp->m_flags |= XFS_MOUNT_NOUUID; } else if (!strcmp(this_char, MNTOPT_BARRIER)) { - args->flags |= XFSMNT_BARRIER; + mp->m_flags |= XFS_MOUNT_BARRIER; } else if (!strcmp(this_char, MNTOPT_NOBARRIER)) { - args->flags &= ~XFSMNT_BARRIER; + mp->m_flags &= ~XFS_MOUNT_BARRIER; } else if (!strcmp(this_char, MNTOPT_IKEEP)) { - args->flags |= XFSMNT_IKEEP; + mp->m_flags |= XFS_MOUNT_IKEEP; } else if (!strcmp(this_char, MNTOPT_NOIKEEP)) { dmapi_implies_ikeep = 0; - args->flags &= ~XFSMNT_IKEEP; + mp->m_flags &= ~XFS_MOUNT_IKEEP; } else if (!strcmp(this_char, MNTOPT_LARGEIO)) { - args->flags2 &= ~XFSMNT2_COMPAT_IOSIZE; + mp->m_flags &= ~XFS_MOUNT_COMPAT_IOSIZE; } else if (!strcmp(this_char, MNTOPT_NOLARGEIO)) { - args->flags2 |= XFSMNT2_COMPAT_IOSIZE; + mp->m_flags |= XFS_MOUNT_COMPAT_IOSIZE; } else if (!strcmp(this_char, MNTOPT_ATTR2)) { - args->flags |= XFSMNT_ATTR2; + mp->m_flags |= XFS_MOUNT_ATTR2; } else if (!strcmp(this_char, MNTOPT_NOATTR2)) { - args->flags &= ~XFSMNT_ATTR2; - args->flags |= XFSMNT_NOATTR2; + mp->m_flags &= ~XFS_MOUNT_ATTR2; + mp->m_flags |= XFS_MOUNT_NOATTR2; } else if (!strcmp(this_char, MNTOPT_FILESTREAM)) { - args->flags2 |= XFSMNT2_FILESTREAMS; + mp->m_flags |= XFS_MOUNT_FILESTREAMS; } else if (!strcmp(this_char, MNTOPT_NOQUOTA)) { - args->flags &= ~(XFSMNT_UQUOTAENF|XFSMNT_UQUOTA); - args->flags &= ~(XFSMNT_GQUOTAENF|XFSMNT_GQUOTA); + mp->m_qflags &= ~(XFS_UQUOTA_ACCT | XFS_UQUOTA_ACTIVE | + XFS_GQUOTA_ACCT | XFS_GQUOTA_ACTIVE | + XFS_PQUOTA_ACCT | XFS_PQUOTA_ACTIVE | + XFS_UQUOTA_ENFD | XFS_OQUOTA_ENFD); } else if (!strcmp(this_char, MNTOPT_QUOTA) || !strcmp(this_char, MNTOPT_UQUOTA) || !strcmp(this_char, MNTOPT_USRQUOTA)) { - args->flags |= XFSMNT_UQUOTA | XFSMNT_UQUOTAENF; + mp->m_qflags |= (XFS_UQUOTA_ACCT | XFS_UQUOTA_ACTIVE | + XFS_UQUOTA_ENFD); } else if (!strcmp(this_char, MNTOPT_QUOTANOENF) || !strcmp(this_char, MNTOPT_UQUOTANOENF)) { - args->flags |= XFSMNT_UQUOTA; - args->flags &= ~XFSMNT_UQUOTAENF; + mp->m_qflags |= (XFS_UQUOTA_ACCT | XFS_UQUOTA_ACTIVE); + mp->m_qflags &= ~XFS_UQUOTA_ENFD; } else if (!strcmp(this_char, MNTOPT_PQUOTA) || !strcmp(this_char, MNTOPT_PRJQUOTA)) { - args->flags |= XFSMNT_PQUOTA | XFSMNT_PQUOTAENF; + mp->m_qflags |= (XFS_PQUOTA_ACCT | XFS_PQUOTA_ACTIVE | + XFS_OQUOTA_ENFD); } else if (!strcmp(this_char, MNTOPT_PQUOTANOENF)) { - args->flags |= XFSMNT_PQUOTA; - args->flags &= ~XFSMNT_PQUOTAENF; + mp->m_qflags |= (XFS_PQUOTA_ACCT | XFS_PQUOTA_ACTIVE); + mp->m_qflags &= ~XFS_OQUOTA_ENFD; } else if (!strcmp(this_char, MNTOPT_GQUOTA) || !strcmp(this_char, MNTOPT_GRPQUOTA)) { - args->flags |= XFSMNT_GQUOTA | XFSMNT_GQUOTAENF; + mp->m_qflags |= (XFS_GQUOTA_ACCT | XFS_GQUOTA_ACTIVE | + XFS_OQUOTA_ENFD); } else if (!strcmp(this_char, MNTOPT_GQUOTANOENF)) { - args->flags |= XFSMNT_GQUOTA; - args->flags &= ~XFSMNT_GQUOTAENF; + mp->m_qflags |= (XFS_GQUOTA_ACCT | XFS_GQUOTA_ACTIVE); + mp->m_qflags &= ~XFS_OQUOTA_ENFD; } else if (!strcmp(this_char, MNTOPT_DMAPI)) { - args->flags |= XFSMNT_DMAPI; + mp->m_flags |= XFS_MOUNT_DMAPI; } else if (!strcmp(this_char, MNTOPT_XDSM)) { - args->flags |= XFSMNT_DMAPI; + mp->m_flags |= XFS_MOUNT_DMAPI; } else if (!strcmp(this_char, MNTOPT_DMI)) { - args->flags |= XFSMNT_DMAPI; + mp->m_flags |= XFS_MOUNT_DMAPI; } else if (!strcmp(this_char, "ihashsize")) { cmn_err(CE_WARN, "XFS: ihashsize no longer used, option is deprecated."); @@ -390,27 +402,29 @@ xfs_parseargs( } } - if (args->flags & XFSMNT_NORECOVERY) { - if ((mp->m_flags & XFS_MOUNT_RDONLY) == 0) { - cmn_err(CE_WARN, - "XFS: no-recovery mounts must be read-only."); - return EINVAL; - } + /* + * no recovery flag requires a read-only mount + */ + if ((mp->m_flags & XFS_MOUNT_NORECOVERY) && + !(mp->m_flags & XFS_MOUNT_RDONLY)) { + cmn_err(CE_WARN, "XFS: no-recovery mounts must be read-only."); + return EINVAL; } - if ((args->flags & XFSMNT_NOALIGN) && (dsunit || dswidth)) { + if ((mp->m_flags & XFS_MOUNT_NOALIGN) && (dsunit || dswidth)) { cmn_err(CE_WARN, "XFS: sunit and swidth options incompatible with the noalign option"); return EINVAL; } - if ((args->flags & XFSMNT_GQUOTA) && (args->flags & XFSMNT_PQUOTA)) { + if ((mp->m_qflags & (XFS_GQUOTA_ACCT | XFS_GQUOTA_ACTIVE)) && + (mp->m_qflags & (XFS_PQUOTA_ACCT | XFS_PQUOTA_ACTIVE))) { cmn_err(CE_WARN, "XFS: cannot mount with both project and group quota"); return EINVAL; } - if ((args->flags & XFSMNT_DMAPI) && *args->mtpt == '\0') { + if ((mp->m_flags & XFS_MOUNT_DMAPI) && (!*mtpt || *mtpt[0] == '\0')) { printk("XFS: %s option needs the mount point option as well\n", MNTOPT_DMAPI); return EINVAL; @@ -438,27 +452,66 @@ xfs_parseargs( * Note that if "ikeep" or "noikeep" mount options are * supplied, then they are honored. */ - if ((args->flags & XFSMNT_DMAPI) && dmapi_implies_ikeep) - args->flags |= XFSMNT_IKEEP; + if ((mp->m_flags & XFS_MOUNT_DMAPI) && dmapi_implies_ikeep) + mp->m_flags |= XFS_MOUNT_IKEEP; - if ((args->flags & XFSMNT_NOALIGN) != XFSMNT_NOALIGN) { +done: + if (!(mp->m_flags & XFS_MOUNT_NOALIGN)) { + /* + * At this point the superblock has not been read + * in, therefore we do not know the block size. + * Before the mount call ends we will convert + * these to FSBs. + */ if (dsunit) { - args->sunit = dsunit; - args->flags |= XFSMNT_RETERR; - } else { - args->sunit = vol_dsunit; + mp->m_dalign = dsunit; + mp->m_flags |= XFS_MOUNT_RETERR; } - dswidth ? (args->swidth = dswidth) : - (args->swidth = vol_dswidth); - } else { - args->sunit = args->swidth = 0; + + if (dswidth) + mp->m_swidth = dswidth; + } + + if (mp->m_logbufs != -1 && + mp->m_logbufs != 0 && + (mp->m_logbufs < XLOG_MIN_ICLOGS || + mp->m_logbufs > XLOG_MAX_ICLOGS)) { + cmn_err(CE_WARN, + "XFS: invalid logbufs value: %d [not %d-%d]", + mp->m_logbufs, XLOG_MIN_ICLOGS, XLOG_MAX_ICLOGS); + return XFS_ERROR(EINVAL); + } + if (mp->m_logbsize != -1 && + mp->m_logbsize != 0 && + (mp->m_logbsize < XLOG_MIN_RECORD_BSIZE || + mp->m_logbsize > XLOG_MAX_RECORD_BSIZE || + !is_power_of_2(mp->m_logbsize))) { + cmn_err(CE_WARN, + "XFS: invalid logbufsize: %d [not 16k,32k,64k,128k or 256k]", + mp->m_logbsize); + return XFS_ERROR(EINVAL); + } + + mp->m_fsname = kstrndup(sb->s_id, MAXNAMELEN, GFP_KERNEL); + if (!mp->m_fsname) + return ENOMEM; + mp->m_fsname_len = strlen(mp->m_fsname) + 1; + + if (iosizelog) { + if (iosizelog > XFS_MAX_IO_LOG || + iosizelog < XFS_MIN_IO_LOG) { + cmn_err(CE_WARN, + "XFS: invalid log iosize: %d [not %d-%d]", + iosizelog, XFS_MIN_IO_LOG, + XFS_MAX_IO_LOG); + return XFS_ERROR(EINVAL); + } + + mp->m_flags |= XFS_MOUNT_DFLT_IOSIZE; + mp->m_readio_log = iosizelog; + mp->m_writeio_log = iosizelog; } -done: - if (args->flags & XFSMNT_32BITINODES) - mp->m_flags |= XFS_MOUNT_SMALL_INUMS; - if (args->flags2) - args->flags |= XFSMNT_FLAGS2; return 0; } @@ -589,7 +642,7 @@ xfs_blkdev_get( { int error = 0; - *bdevp = open_bdev_excl(name, 0, mp); + *bdevp = open_bdev_exclusive(name, FMODE_READ|FMODE_WRITE, mp); if (IS_ERR(*bdevp)) { error = PTR_ERR(*bdevp); printk("XFS: Invalid device [%s], error=%d\n", name, error); @@ -603,7 +656,7 @@ xfs_blkdev_put( struct block_device *bdev) { if (bdev) - close_bdev_excl(bdev); + close_bdev_exclusive(bdev, FMODE_READ|FMODE_WRITE); } /* @@ -704,8 +757,7 @@ xfs_close_devices( */ STATIC int xfs_open_devices( - struct xfs_mount *mp, - struct xfs_mount_args *args) + struct xfs_mount *mp) { struct block_device *ddev = mp->m_super->s_bdev; struct block_device *logdev = NULL, *rtdev = NULL; @@ -714,14 +766,14 @@ xfs_open_devices( /* * Open real time and log devices - order is important. */ - if (args->logname[0]) { - error = xfs_blkdev_get(mp, args->logname, &logdev); + if (mp->m_logname) { + error = xfs_blkdev_get(mp, mp->m_logname, &logdev); if (error) goto out; } - if (args->rtname[0]) { - error = xfs_blkdev_get(mp, args->rtname, &rtdev); + if (mp->m_rtname) { + error = xfs_blkdev_get(mp, mp->m_rtname, &rtdev); if (error) goto out_close_logdev; @@ -813,18 +865,18 @@ xfs_setup_devices( */ void xfsaild_wakeup( - xfs_mount_t *mp, + struct xfs_ail *ailp, xfs_lsn_t threshold_lsn) { - mp->m_ail.xa_target = threshold_lsn; - wake_up_process(mp->m_ail.xa_task); + ailp->xa_target = threshold_lsn; + wake_up_process(ailp->xa_task); } int xfsaild( void *data) { - xfs_mount_t *mp = (xfs_mount_t *)data; + struct xfs_ail *ailp = data; xfs_lsn_t last_pushed_lsn = 0; long tout = 0; @@ -836,11 +888,11 @@ xfsaild( /* swsusp */ try_to_freeze(); - ASSERT(mp->m_log); - if (XFS_FORCED_SHUTDOWN(mp)) + ASSERT(ailp->xa_mount->m_log); + if (XFS_FORCED_SHUTDOWN(ailp->xa_mount)) continue; - tout = xfsaild_push(mp, &last_pushed_lsn); + tout = xfsaild_push(ailp, &last_pushed_lsn); } return 0; @@ -848,43 +900,82 @@ xfsaild( int xfsaild_start( - xfs_mount_t *mp) + struct xfs_ail *ailp) { - mp->m_ail.xa_target = 0; - mp->m_ail.xa_task = kthread_run(xfsaild, mp, "xfsaild"); - if (IS_ERR(mp->m_ail.xa_task)) - return -PTR_ERR(mp->m_ail.xa_task); + ailp->xa_target = 0; + ailp->xa_task = kthread_run(xfsaild, ailp, "xfsaild"); + if (IS_ERR(ailp->xa_task)) + return -PTR_ERR(ailp->xa_task); return 0; } void xfsaild_stop( - xfs_mount_t *mp) + struct xfs_ail *ailp) { - kthread_stop(mp->m_ail.xa_task); + kthread_stop(ailp->xa_task); } - +/* Catch misguided souls that try to use this interface on XFS */ STATIC struct inode * xfs_fs_alloc_inode( struct super_block *sb) { - return kmem_zone_alloc(xfs_vnode_zone, KM_SLEEP); + BUG(); + return NULL; } +/* + * Now that the generic code is guaranteed not to be accessing + * the linux inode, we can reclaim the inode. + */ STATIC void xfs_fs_destroy_inode( - struct inode *inode) + struct inode *inode) { - kmem_zone_free(xfs_vnode_zone, inode); + xfs_inode_t *ip = XFS_I(inode); + + XFS_STATS_INC(vn_reclaim); + if (xfs_reclaim(ip)) + panic("%s: cannot reclaim 0x%p\n", __func__, inode); } +/* + * Slab object creation initialisation for the XFS inode. + * This covers only the idempotent fields in the XFS inode; + * all other fields need to be initialised on allocation + * from the slab. This avoids the need to repeatedly intialise + * fields in the xfs inode that left in the initialise state + * when freeing the inode. + */ STATIC void xfs_fs_inode_init_once( - void *vnode) + void *inode) { - inode_init_once((struct inode *)vnode); + struct xfs_inode *ip = inode; + + memset(ip, 0, sizeof(struct xfs_inode)); + + /* vfs inode */ + inode_init_once(VFS_I(ip)); + + /* xfs inode */ + atomic_set(&ip->i_iocount, 0); + atomic_set(&ip->i_pincount, 0); + spin_lock_init(&ip->i_flags_lock); + init_waitqueue_head(&ip->i_ipin_wait); + /* + * Because we want to use a counting completion, complete + * the flush completion once to allow a single access to + * the flush completion without blocking. + */ + init_completion(&ip->i_flush); + complete(&ip->i_flush); + + mrlock_init(&ip->i_lock, MRLOCK_ALLOW_EQUAL_PRI|MRLOCK_BARRIER, + "xfsino", ip->i_ino); + mrlock_init(&ip->i_iolock, MRLOCK_BARRIER, "xfsio", ip->i_ino); } /* @@ -898,21 +989,26 @@ xfs_fs_write_inode( struct inode *inode, int sync) { + struct xfs_inode *ip = XFS_I(inode); int error = 0; int flags = 0; - xfs_itrace_entry(XFS_I(inode)); + xfs_itrace_entry(ip); if (sync) { - filemap_fdatawait(inode->i_mapping); + error = xfs_wait_on_pages(ip, 0, -1); + if (error) + goto out_error; flags |= FLUSH_SYNC; } - error = xfs_inode_flush(XFS_I(inode), flags); + error = xfs_inode_flush(ip, flags); + +out_error: /* * if we failed to write out the inode then mark * it dirty again so we'll try again later. */ if (error) - mark_inode_dirty_sync(inode); + xfs_mark_inode_dirty_sync(ip); return -error; } @@ -923,164 +1019,12 @@ xfs_fs_clear_inode( { xfs_inode_t *ip = XFS_I(inode); - /* - * ip can be null when xfs_iget_core calls xfs_idestroy if we - * find an inode with di_mode == 0 but without IGET_CREATE set. - */ - if (ip) { - xfs_itrace_entry(ip); - XFS_STATS_INC(vn_rele); - XFS_STATS_INC(vn_remove); - XFS_STATS_INC(vn_reclaim); - XFS_STATS_DEC(vn_active); - - xfs_inactive(ip); - xfs_iflags_clear(ip, XFS_IMODIFIED); - if (xfs_reclaim(ip)) - panic("%s: cannot reclaim 0x%p\n", __func__, inode); - } - - ASSERT(XFS_I(inode) == NULL); -} - -/* - * Enqueue a work item to be picked up by the vfs xfssyncd thread. - * Doing this has two advantages: - * - It saves on stack space, which is tight in certain situations - * - It can be used (with care) as a mechanism to avoid deadlocks. - * Flushing while allocating in a full filesystem requires both. - */ -STATIC void -xfs_syncd_queue_work( - struct xfs_mount *mp, - void *data, - void (*syncer)(struct xfs_mount *, void *)) -{ - struct bhv_vfs_sync_work *work; - - work = kmem_alloc(sizeof(struct bhv_vfs_sync_work), KM_SLEEP); - INIT_LIST_HEAD(&work->w_list); - work->w_syncer = syncer; - work->w_data = data; - work->w_mount = mp; - spin_lock(&mp->m_sync_lock); - list_add_tail(&work->w_list, &mp->m_sync_list); - spin_unlock(&mp->m_sync_lock); - wake_up_process(mp->m_sync_task); -} - -/* - * Flush delayed allocate data, attempting to free up reserved space - * from existing allocations. At this point a new allocation attempt - * has failed with ENOSPC and we are in the process of scratching our - * heads, looking about for more room... - */ -STATIC void -xfs_flush_inode_work( - struct xfs_mount *mp, - void *arg) -{ - struct inode *inode = arg; - filemap_flush(inode->i_mapping); - iput(inode); -} - -void -xfs_flush_inode( - xfs_inode_t *ip) -{ - struct inode *inode = VFS_I(ip); - - igrab(inode); - xfs_syncd_queue_work(ip->i_mount, inode, xfs_flush_inode_work); - delay(msecs_to_jiffies(500)); -} - -/* - * This is the "bigger hammer" version of xfs_flush_inode_work... - * (IOW, "If at first you don't succeed, use a Bigger Hammer"). - */ -STATIC void -xfs_flush_device_work( - struct xfs_mount *mp, - void *arg) -{ - struct inode *inode = arg; - sync_blockdev(mp->m_super->s_bdev); - iput(inode); -} - -void -xfs_flush_device( - xfs_inode_t *ip) -{ - struct inode *inode = VFS_I(ip); - - igrab(inode); - xfs_syncd_queue_work(ip->i_mount, inode, xfs_flush_device_work); - delay(msecs_to_jiffies(500)); - xfs_log_force(ip->i_mount, (xfs_lsn_t)0, XFS_LOG_FORCE|XFS_LOG_SYNC); -} - -STATIC void -xfs_sync_worker( - struct xfs_mount *mp, - void *unused) -{ - int error; - - if (!(mp->m_flags & XFS_MOUNT_RDONLY)) - error = xfs_sync(mp, SYNC_FSDATA | SYNC_BDFLUSH | SYNC_ATTR); - mp->m_sync_seq++; - wake_up(&mp->m_wait_single_sync_task); -} - -STATIC int -xfssyncd( - void *arg) -{ - struct xfs_mount *mp = arg; - long timeleft; - bhv_vfs_sync_work_t *work, *n; - LIST_HEAD (tmp); - - set_freezable(); - timeleft = xfs_syncd_centisecs * msecs_to_jiffies(10); - for (;;) { - timeleft = schedule_timeout_interruptible(timeleft); - /* swsusp */ - try_to_freeze(); - if (kthread_should_stop() && list_empty(&mp->m_sync_list)) - break; - - spin_lock(&mp->m_sync_lock); - /* - * We can get woken by laptop mode, to do a sync - - * that's the (only!) case where the list would be - * empty with time remaining. - */ - if (!timeleft || list_empty(&mp->m_sync_list)) { - if (!timeleft) - timeleft = xfs_syncd_centisecs * - msecs_to_jiffies(10); - INIT_LIST_HEAD(&mp->m_sync_work.w_list); - list_add_tail(&mp->m_sync_work.w_list, - &mp->m_sync_list); - } - list_for_each_entry_safe(work, n, &mp->m_sync_list, w_list) - list_move(&work->w_list, &tmp); - spin_unlock(&mp->m_sync_lock); - - list_for_each_entry_safe(work, n, &tmp, w_list) { - (*work->w_syncer)(mp, work->w_data); - list_del(&work->w_list); - if (work == &mp->m_sync_work) - continue; - kmem_free(work); - } - } + xfs_itrace_entry(ip); + XFS_STATS_INC(vn_rele); + XFS_STATS_INC(vn_remove); + XFS_STATS_DEC(vn_active); - return 0; + xfs_inactive(ip); } STATIC void @@ -1099,11 +1043,9 @@ xfs_fs_put_super( struct xfs_mount *mp = XFS_M(sb); struct xfs_inode *rip = mp->m_rootip; int unmount_event_flags = 0; - int error; - kthread_stop(mp->m_sync_task); - - xfs_sync(mp, SYNC_ATTR | SYNC_DELWRI); + xfs_syncd_stop(mp); + xfs_sync_inodes(mp, SYNC_ATTR|SYNC_DELWRI); #ifdef HAVE_DMAPI if (mp->m_flags & XFS_MOUNT_DMAPI) { @@ -1128,18 +1070,6 @@ xfs_fs_put_super( xfs_filestream_unmount(mp); XFS_bflush(mp->m_ddev_targp); - error = xfs_unmount_flush(mp, 0); - WARN_ON(error); - - /* - * If we're forcing a shutdown, typically because of a media error, - * we want to make sure we invalidate dirty pages that belong to - * referenced vnodes as well. - */ - if (XFS_FORCED_SHUTDOWN(mp)) { - error = xfs_sync(mp, SYNC_WAIT | SYNC_CLOSE); - ASSERT(error != EFSCORRUPTED); - } if (mp->m_flags & XFS_MOUNT_DMAPI) { XFS_SEND_UNMOUNT(mp, rip, DM_RIGHT_NULL, 0, 0, @@ -1161,7 +1091,7 @@ xfs_fs_write_super( struct super_block *sb) { if (!(sb->s_flags & MS_RDONLY)) - xfs_sync(XFS_M(sb), SYNC_FSDATA); + xfs_sync_fsdata(XFS_M(sb), 0); sb->s_dirt = 0; } @@ -1172,7 +1102,6 @@ xfs_fs_sync_super( { struct xfs_mount *mp = XFS_M(sb); int error; - int flags; /* * Treat a sync operation like a freeze. This is to work @@ -1186,20 +1115,10 @@ xfs_fs_sync_super( * dirty the Linux inode until after the transaction I/O * completes. */ - if (wait || unlikely(sb->s_frozen == SB_FREEZE_WRITE)) { - /* - * First stage of freeze - no more writers will make progress - * now we are here, so we flush delwri and delalloc buffers - * here, then wait for all I/O to complete. Data is frozen at - * that point. Metadata is not frozen, transactions can still - * occur here so don't bother flushing the buftarg (i.e - * SYNC_QUIESCE) because it'll just get dirty again. - */ - flags = SYNC_DATA_QUIESCE; - } else - flags = SYNC_FSDATA; - - error = xfs_sync(mp, flags); + if (wait || unlikely(sb->s_frozen == SB_FREEZE_WRITE)) + error = xfs_quiesce_data(mp); + else + error = xfs_sync_fsdata(mp, 0); sb->s_dirt = 0; if (unlikely(laptop_mode)) { @@ -1323,7 +1242,7 @@ xfs_fs_remount( "XFS: mount option \"%s\" not supported for remount\n", p); return -EINVAL; #else - return 0; + break; #endif } } @@ -1337,9 +1256,8 @@ xfs_fs_remount( /* rw -> ro */ if (!(mp->m_flags & XFS_MOUNT_RDONLY) && (*flags & MS_RDONLY)) { - xfs_filestream_flush(mp); - xfs_sync(mp, SYNC_DATA_QUIESCE); - xfs_attr_quiesce(mp); + xfs_quiesce_data(mp); + xfs_quiesce_attr(mp); mp->m_flags |= XFS_MOUNT_RDONLY; } @@ -1348,7 +1266,7 @@ xfs_fs_remount( /* * Second stage of a freeze. The data is already frozen so we only - * need to take care of themetadata. Once that's done write a dummy + * need to take care of the metadata. Once that's done write a dummy * record to dirty the log in case of a crash while frozen. */ STATIC void @@ -1357,7 +1275,7 @@ xfs_fs_lockfs( { struct xfs_mount *mp = XFS_M(sb); - xfs_attr_quiesce(mp); + xfs_quiesce_attr(mp); xfs_fs_log_dummy(mp); } @@ -1422,175 +1340,28 @@ xfs_fs_setxquota( /* * This function fills in xfs_mount_t fields based on mount args. - * Note: the superblock has _not_ yet been read in. - */ -STATIC int -xfs_start_flags( - struct xfs_mount_args *ap, - struct xfs_mount *mp) -{ - int error; - - /* Values are in BBs */ - if ((ap->flags & XFSMNT_NOALIGN) != XFSMNT_NOALIGN) { - /* - * At this point the superblock has not been read - * in, therefore we do not know the block size. - * Before the mount call ends we will convert - * these to FSBs. - */ - mp->m_dalign = ap->sunit; - mp->m_swidth = ap->swidth; - } - - if (ap->logbufs != -1 && - ap->logbufs != 0 && - (ap->logbufs < XLOG_MIN_ICLOGS || - ap->logbufs > XLOG_MAX_ICLOGS)) { - cmn_err(CE_WARN, - "XFS: invalid logbufs value: %d [not %d-%d]", - ap->logbufs, XLOG_MIN_ICLOGS, XLOG_MAX_ICLOGS); - return XFS_ERROR(EINVAL); - } - mp->m_logbufs = ap->logbufs; - if (ap->logbufsize != -1 && - ap->logbufsize != 0 && - (ap->logbufsize < XLOG_MIN_RECORD_BSIZE || - ap->logbufsize > XLOG_MAX_RECORD_BSIZE || - !is_power_of_2(ap->logbufsize))) { - cmn_err(CE_WARN, - "XFS: invalid logbufsize: %d [not 16k,32k,64k,128k or 256k]", - ap->logbufsize); - return XFS_ERROR(EINVAL); - } - - error = ENOMEM; - - mp->m_logbsize = ap->logbufsize; - mp->m_fsname_len = strlen(ap->fsname) + 1; - - mp->m_fsname = kstrdup(ap->fsname, GFP_KERNEL); - if (!mp->m_fsname) - goto out; - - if (ap->rtname[0]) { - mp->m_rtname = kstrdup(ap->rtname, GFP_KERNEL); - if (!mp->m_rtname) - goto out_free_fsname; - - } - - if (ap->logname[0]) { - mp->m_logname = kstrdup(ap->logname, GFP_KERNEL); - if (!mp->m_logname) - goto out_free_rtname; - } - - if (ap->flags & XFSMNT_WSYNC) - mp->m_flags |= XFS_MOUNT_WSYNC; -#if XFS_BIG_INUMS - if (ap->flags & XFSMNT_INO64) { - mp->m_flags |= XFS_MOUNT_INO64; - mp->m_inoadd = XFS_INO64_OFFSET; - } -#endif - if (ap->flags & XFSMNT_RETERR) - mp->m_flags |= XFS_MOUNT_RETERR; - if (ap->flags & XFSMNT_NOALIGN) - mp->m_flags |= XFS_MOUNT_NOALIGN; - if (ap->flags & XFSMNT_SWALLOC) - mp->m_flags |= XFS_MOUNT_SWALLOC; - if (ap->flags & XFSMNT_OSYNCISOSYNC) - mp->m_flags |= XFS_MOUNT_OSYNCISOSYNC; - if (ap->flags & XFSMNT_32BITINODES) - mp->m_flags |= XFS_MOUNT_32BITINODES; - - if (ap->flags & XFSMNT_IOSIZE) { - if (ap->iosizelog > XFS_MAX_IO_LOG || - ap->iosizelog < XFS_MIN_IO_LOG) { - cmn_err(CE_WARN, - "XFS: invalid log iosize: %d [not %d-%d]", - ap->iosizelog, XFS_MIN_IO_LOG, - XFS_MAX_IO_LOG); - return XFS_ERROR(EINVAL); - } - - mp->m_flags |= XFS_MOUNT_DFLT_IOSIZE; - mp->m_readio_log = mp->m_writeio_log = ap->iosizelog; - } - - if (ap->flags & XFSMNT_IKEEP) - mp->m_flags |= XFS_MOUNT_IKEEP; - if (ap->flags & XFSMNT_DIRSYNC) - mp->m_flags |= XFS_MOUNT_DIRSYNC; - if (ap->flags & XFSMNT_ATTR2) - mp->m_flags |= XFS_MOUNT_ATTR2; - if (ap->flags & XFSMNT_NOATTR2) - mp->m_flags |= XFS_MOUNT_NOATTR2; - - if (ap->flags2 & XFSMNT2_COMPAT_IOSIZE) - mp->m_flags |= XFS_MOUNT_COMPAT_IOSIZE; - - /* - * no recovery flag requires a read-only mount - */ - if (ap->flags & XFSMNT_NORECOVERY) { - if (!(mp->m_flags & XFS_MOUNT_RDONLY)) { - cmn_err(CE_WARN, - "XFS: tried to mount a FS read-write without recovery!"); - return XFS_ERROR(EINVAL); - } - mp->m_flags |= XFS_MOUNT_NORECOVERY; - } - - if (ap->flags & XFSMNT_NOUUID) - mp->m_flags |= XFS_MOUNT_NOUUID; - if (ap->flags & XFSMNT_BARRIER) - mp->m_flags |= XFS_MOUNT_BARRIER; - else - mp->m_flags &= ~XFS_MOUNT_BARRIER; - - if (ap->flags2 & XFSMNT2_FILESTREAMS) - mp->m_flags |= XFS_MOUNT_FILESTREAMS; - - if (ap->flags & XFSMNT_DMAPI) - mp->m_flags |= XFS_MOUNT_DMAPI; - return 0; - - - out_free_rtname: - kfree(mp->m_rtname); - out_free_fsname: - kfree(mp->m_fsname); - out: - return error; -} - -/* - * This function fills in xfs_mount_t fields based on mount args. * Note: the superblock _has_ now been read in. */ STATIC int xfs_finish_flags( - struct xfs_mount_args *ap, struct xfs_mount *mp) { int ronly = (mp->m_flags & XFS_MOUNT_RDONLY); /* Fail a mount where the logbuf is smaller then the log stripe */ if (xfs_sb_version_haslogv2(&mp->m_sb)) { - if ((ap->logbufsize <= 0) && - (mp->m_sb.sb_logsunit > XLOG_BIG_RECORD_BSIZE)) { + if (mp->m_logbsize <= 0 && + mp->m_sb.sb_logsunit > XLOG_BIG_RECORD_BSIZE) { mp->m_logbsize = mp->m_sb.sb_logsunit; - } else if (ap->logbufsize > 0 && - ap->logbufsize < mp->m_sb.sb_logsunit) { + } else if (mp->m_logbsize > 0 && + mp->m_logbsize < mp->m_sb.sb_logsunit) { cmn_err(CE_WARN, "XFS: logbuf size must be greater than or equal to log stripe size"); return XFS_ERROR(EINVAL); } } else { /* Fail a mount if the logbuf is larger than 32K */ - if (ap->logbufsize > XLOG_BIG_RECORD_BSIZE) { + if (mp->m_logbsize > XLOG_BIG_RECORD_BSIZE) { cmn_err(CE_WARN, "XFS: logbuf size for version 1 logs must be 16K or 32K"); return XFS_ERROR(EINVAL); @@ -1602,7 +1373,7 @@ xfs_finish_flags( * told by noattr2 to turn it off */ if (xfs_sb_version_hasattr2(&mp->m_sb) && - !(ap->flags & XFSMNT_NOATTR2)) + !(mp->m_flags & XFS_MOUNT_NOATTR2)) mp->m_flags |= XFS_MOUNT_ATTR2; /* @@ -1614,48 +1385,6 @@ xfs_finish_flags( return XFS_ERROR(EROFS); } - /* - * check for shared mount. - */ - if (ap->flags & XFSMNT_SHARED) { - if (!xfs_sb_version_hasshared(&mp->m_sb)) - return XFS_ERROR(EINVAL); - - /* - * For IRIX 6.5, shared mounts must have the shared - * version bit set, have the persistent readonly - * field set, must be version 0 and can only be mounted - * read-only. - */ - if (!ronly || !(mp->m_sb.sb_flags & XFS_SBF_READONLY) || - (mp->m_sb.sb_shared_vn != 0)) - return XFS_ERROR(EINVAL); - - mp->m_flags |= XFS_MOUNT_SHARED; - - /* - * Shared XFS V0 can't deal with DMI. Return EINVAL. - */ - if (mp->m_sb.sb_shared_vn == 0 && (ap->flags & XFSMNT_DMAPI)) - return XFS_ERROR(EINVAL); - } - - if (ap->flags & XFSMNT_UQUOTA) { - mp->m_qflags |= (XFS_UQUOTA_ACCT | XFS_UQUOTA_ACTIVE); - if (ap->flags & XFSMNT_UQUOTAENF) - mp->m_qflags |= XFS_UQUOTA_ENFD; - } - - if (ap->flags & XFSMNT_GQUOTA) { - mp->m_qflags |= (XFS_GQUOTA_ACCT | XFS_GQUOTA_ACTIVE); - if (ap->flags & XFSMNT_GQUOTAENF) - mp->m_qflags |= XFS_OQUOTA_ENFD; - } else if (ap->flags & XFSMNT_PQUOTA) { - mp->m_qflags |= (XFS_PQUOTA_ACCT | XFS_PQUOTA_ACTIVE); - if (ap->flags & XFSMNT_PQUOTAENF) - mp->m_qflags |= XFS_OQUOTA_ENFD; - } - return 0; } @@ -1667,19 +1396,14 @@ xfs_fs_fill_super( { struct inode *root; struct xfs_mount *mp = NULL; - struct xfs_mount_args *args; int flags = 0, error = ENOMEM; - - args = xfs_args_allocate(sb, silent); - if (!args) - return -ENOMEM; + char *mtpt = NULL; mp = kzalloc(sizeof(struct xfs_mount), GFP_KERNEL); if (!mp) - goto out_free_args; + goto out; spin_lock_init(&mp->m_sb_lock); - mutex_init(&mp->m_ilock); mutex_init(&mp->m_growlock); atomic_set(&mp->m_active_trans, 0); INIT_LIST_HEAD(&mp->m_sync_list); @@ -1689,12 +1413,9 @@ xfs_fs_fill_super( mp->m_super = sb; sb->s_fs_info = mp; - if (sb->s_flags & MS_RDONLY) - mp->m_flags |= XFS_MOUNT_RDONLY; - - error = xfs_parseargs(mp, (char *)data, args, 0); + error = xfs_parseargs(mp, (char *)data, &mtpt); if (error) - goto out_free_mp; + goto out_free_fsname; sb_min_blocksize(sb, BBSIZE); sb->s_xattr = xfs_xattr_handlers; @@ -1702,33 +1423,28 @@ xfs_fs_fill_super( sb->s_qcop = &xfs_quotactl_operations; sb->s_op = &xfs_super_operations; - error = xfs_dmops_get(mp, args); + error = xfs_dmops_get(mp); if (error) - goto out_free_mp; - error = xfs_qmops_get(mp, args); + goto out_free_fsname; + error = xfs_qmops_get(mp); if (error) goto out_put_dmops; - if (args->flags & XFSMNT_QUIET) + if (silent) flags |= XFS_MFSI_QUIET; - error = xfs_open_devices(mp, args); + error = xfs_open_devices(mp); if (error) goto out_put_qmops; if (xfs_icsb_init_counters(mp)) mp->m_flags |= XFS_MOUNT_NO_PERCPU_SB; - /* - * Setup flags based on mount(2) options and then the superblock - */ - error = xfs_start_flags(args, mp); - if (error) - goto out_free_fsname; error = xfs_readsb(mp, flags); if (error) - goto out_free_fsname; - error = xfs_finish_flags(args, mp); + goto out_destroy_counters; + + error = xfs_finish_flags(mp); if (error) goto out_free_sb; @@ -1747,7 +1463,7 @@ xfs_fs_fill_super( if (error) goto out_filestream_unmount; - XFS_SEND_MOUNT(mp, DM_RIGHT_NULL, args->mtpt, args->fsname); + XFS_SEND_MOUNT(mp, DM_RIGHT_NULL, mtpt, mp->m_fsname); sb->s_dirt = 1; sb->s_magic = XFS_SB_MAGIC; @@ -1772,35 +1488,31 @@ xfs_fs_fill_super( goto fail_vnrele; } - mp->m_sync_work.w_syncer = xfs_sync_worker; - mp->m_sync_work.w_mount = mp; - mp->m_sync_task = kthread_run(xfssyncd, mp, "xfssyncd"); - if (IS_ERR(mp->m_sync_task)) { - error = -PTR_ERR(mp->m_sync_task); + error = xfs_syncd_init(mp); + if (error) goto fail_vnrele; - } - xfs_itrace_exit(XFS_I(sb->s_root->d_inode)); + kfree(mtpt); - kfree(args); + xfs_itrace_exit(XFS_I(sb->s_root->d_inode)); return 0; out_filestream_unmount: xfs_filestream_unmount(mp); out_free_sb: xfs_freesb(mp); - out_free_fsname: - xfs_free_fsname(mp); + out_destroy_counters: xfs_icsb_destroy_counters(mp); xfs_close_devices(mp); out_put_qmops: xfs_qmops_put(mp); out_put_dmops: xfs_dmops_put(mp); - out_free_mp: + out_free_fsname: + xfs_free_fsname(mp); + kfree(mtpt); kfree(mp); - out_free_args: - kfree(args); + out: return -error; fail_vnrele: @@ -1820,8 +1532,6 @@ xfs_fs_fill_super( xfs_filestream_unmount(mp); XFS_bflush(mp->m_ddev_targp); - error = xfs_unmount_flush(mp, 0); - WARN_ON(error); xfs_unmountfs(mp); goto out_free_sb; @@ -1882,10 +1592,19 @@ xfs_alloc_trace_bufs(void) if (!xfs_bmap_trace_buf) goto out_free_alloc_trace; #endif -#ifdef XFS_BMBT_TRACE +#ifdef XFS_BTREE_TRACE + xfs_allocbt_trace_buf = ktrace_alloc(XFS_ALLOCBT_TRACE_SIZE, + KM_MAYFAIL); + if (!xfs_allocbt_trace_buf) + goto out_free_bmap_trace; + + xfs_inobt_trace_buf = ktrace_alloc(XFS_INOBT_TRACE_SIZE, KM_MAYFAIL); + if (!xfs_inobt_trace_buf) + goto out_free_allocbt_trace; + xfs_bmbt_trace_buf = ktrace_alloc(XFS_BMBT_TRACE_SIZE, KM_MAYFAIL); if (!xfs_bmbt_trace_buf) - goto out_free_bmap_trace; + goto out_free_inobt_trace; #endif #ifdef XFS_ATTR_TRACE xfs_attr_trace_buf = ktrace_alloc(XFS_ATTR_TRACE_SIZE, KM_MAYFAIL); @@ -1907,8 +1626,12 @@ xfs_alloc_trace_bufs(void) ktrace_free(xfs_attr_trace_buf); out_free_bmbt_trace: #endif -#ifdef XFS_BMBT_TRACE +#ifdef XFS_BTREE_TRACE ktrace_free(xfs_bmbt_trace_buf); + out_free_inobt_trace: + ktrace_free(xfs_inobt_trace_buf); + out_free_allocbt_trace: + ktrace_free(xfs_allocbt_trace_buf); out_free_bmap_trace: #endif #ifdef XFS_BMAP_TRACE @@ -1931,8 +1654,10 @@ xfs_free_trace_bufs(void) #ifdef XFS_ATTR_TRACE ktrace_free(xfs_attr_trace_buf); #endif -#ifdef XFS_BMBT_TRACE +#ifdef XFS_BTREE_TRACE ktrace_free(xfs_bmbt_trace_buf); + ktrace_free(xfs_inobt_trace_buf); + ktrace_free(xfs_allocbt_trace_buf); #endif #ifdef XFS_BMAP_TRACE ktrace_free(xfs_bmap_trace_buf); @@ -1945,16 +1670,10 @@ xfs_free_trace_bufs(void) STATIC int __init xfs_init_zones(void) { - xfs_vnode_zone = kmem_zone_init_flags(sizeof(struct inode), "xfs_vnode", - KM_ZONE_HWALIGN | KM_ZONE_RECLAIM | - KM_ZONE_SPREAD, - xfs_fs_inode_init_once); - if (!xfs_vnode_zone) - goto out; xfs_ioend_zone = kmem_zone_init(sizeof(xfs_ioend_t), "xfs_ioend"); if (!xfs_ioend_zone) - goto out_destroy_vnode_zone; + goto out; xfs_ioend_pool = mempool_create_slab_pool(4 * MAX_BUF_PER_PAGE, xfs_ioend_zone); @@ -1970,6 +1689,7 @@ xfs_init_zones(void) "xfs_bmap_free_item"); if (!xfs_bmap_free_item_zone) goto out_destroy_log_ticket_zone; + xfs_btree_cur_zone = kmem_zone_init(sizeof(xfs_btree_cur_t), "xfs_btree_cur"); if (!xfs_btree_cur_zone) @@ -2017,8 +1737,8 @@ xfs_init_zones(void) xfs_inode_zone = kmem_zone_init_flags(sizeof(xfs_inode_t), "xfs_inode", - KM_ZONE_HWALIGN | KM_ZONE_RECLAIM | - KM_ZONE_SPREAD, NULL); + KM_ZONE_HWALIGN | KM_ZONE_RECLAIM | KM_ZONE_SPREAD, + xfs_fs_inode_init_once); if (!xfs_inode_zone) goto out_destroy_efi_zone; @@ -2066,8 +1786,6 @@ xfs_init_zones(void) mempool_destroy(xfs_ioend_pool); out_destroy_ioend_zone: kmem_zone_destroy(xfs_ioend_zone); - out_destroy_vnode_zone: - kmem_zone_destroy(xfs_vnode_zone); out: return -ENOMEM; } @@ -2092,7 +1810,6 @@ xfs_destroy_zones(void) kmem_zone_destroy(xfs_log_ticket_zone); mempool_destroy(xfs_ioend_pool); kmem_zone_destroy(xfs_ioend_zone); - kmem_zone_destroy(xfs_vnode_zone); } @@ -2100,13 +1817,12 @@ STATIC int __init init_xfs_fs(void) { int error; - static char message[] __initdata = KERN_INFO \ - XFS_VERSION_STRING " with " XFS_BUILD_OPTIONS " enabled\n"; - printk(message); + printk(KERN_INFO XFS_VERSION_STRING " with " + XFS_BUILD_OPTIONS " enabled\n"); ktrace_init(64); - vn_init(); + xfs_ioend_init(); xfs_dir_startup(); error = xfs_init_zones(); diff --git a/fs/xfs/linux-2.6/xfs_super.h b/fs/xfs/linux-2.6/xfs_super.h index fe2ef4e6a0f..d5d776d4cd6 100644 --- a/fs/xfs/linux-2.6/xfs_super.h +++ b/fs/xfs/linux-2.6/xfs_super.h @@ -20,24 +20,12 @@ #include <linux/exportfs.h> -#ifdef CONFIG_XFS_DMAPI -# define vfs_insertdmapi(vfs) vfs_insertops(vfsp, &xfs_dmops) -# define vfs_initdmapi() dmapi_init() -# define vfs_exitdmapi() dmapi_uninit() -#else -# define vfs_insertdmapi(vfs) do { } while (0) -# define vfs_initdmapi() do { } while (0) -# define vfs_exitdmapi() do { } while (0) -#endif - #ifdef CONFIG_XFS_QUOTA -# define vfs_insertquota(vfs) vfs_insertops(vfsp, &xfs_qmops) extern void xfs_qm_init(void); extern void xfs_qm_exit(void); # define vfs_initquota() xfs_qm_init() # define vfs_exitquota() xfs_qm_exit() #else -# define vfs_insertquota(vfs) do { } while (0) # define vfs_initquota() do { } while (0) # define vfs_exitquota() do { } while (0) #endif @@ -101,9 +89,6 @@ struct block_device; extern __uint64_t xfs_max_file_offset(unsigned int); -extern void xfs_flush_inode(struct xfs_inode *); -extern void xfs_flush_device(struct xfs_inode *); - extern void xfs_blkdev_issue_flush(struct xfs_buftarg *); extern const struct export_operations xfs_export_operations; diff --git a/fs/xfs/linux-2.6/xfs_sync.c b/fs/xfs/linux-2.6/xfs_sync.c new file mode 100644 index 00000000000..2ed035354c2 --- /dev/null +++ b/fs/xfs/linux-2.6/xfs_sync.c @@ -0,0 +1,762 @@ +/* + * Copyright (c) 2000-2005 Silicon Graphics, Inc. + * All Rights Reserved. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it would be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write the Free Software Foundation, + * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA + */ +#include "xfs.h" +#include "xfs_fs.h" +#include "xfs_types.h" +#include "xfs_bit.h" +#include "xfs_log.h" +#include "xfs_inum.h" +#include "xfs_trans.h" +#include "xfs_sb.h" +#include "xfs_ag.h" +#include "xfs_dir2.h" +#include "xfs_dmapi.h" +#include "xfs_mount.h" +#include "xfs_bmap_btree.h" +#include "xfs_alloc_btree.h" +#include "xfs_ialloc_btree.h" +#include "xfs_btree.h" +#include "xfs_dir2_sf.h" +#include "xfs_attr_sf.h" +#include "xfs_inode.h" +#include "xfs_dinode.h" +#include "xfs_error.h" +#include "xfs_mru_cache.h" +#include "xfs_filestream.h" +#include "xfs_vnodeops.h" +#include "xfs_utils.h" +#include "xfs_buf_item.h" +#include "xfs_inode_item.h" +#include "xfs_rw.h" + +#include <linux/kthread.h> +#include <linux/freezer.h> + +/* + * Sync all the inodes in the given AG according to the + * direction given by the flags. + */ +STATIC int +xfs_sync_inodes_ag( + xfs_mount_t *mp, + int ag, + int flags) +{ + xfs_perag_t *pag = &mp->m_perag[ag]; + int nr_found; + uint32_t first_index = 0; + int error = 0; + int last_error = 0; + int fflag = XFS_B_ASYNC; + + if (flags & SYNC_DELWRI) + fflag = XFS_B_DELWRI; + if (flags & SYNC_WAIT) + fflag = 0; /* synchronous overrides all */ + + do { + struct inode *inode; + xfs_inode_t *ip = NULL; + int lock_flags = XFS_ILOCK_SHARED; + + /* + * use a gang lookup to find the next inode in the tree + * as the tree is sparse and a gang lookup walks to find + * the number of objects requested. + */ + read_lock(&pag->pag_ici_lock); + nr_found = radix_tree_gang_lookup(&pag->pag_ici_root, + (void**)&ip, first_index, 1); + + if (!nr_found) { + read_unlock(&pag->pag_ici_lock); + break; + } + + /* + * Update the index for the next lookup. Catch overflows + * into the next AG range which can occur if we have inodes + * in the last block of the AG and we are currently + * pointing to the last inode. + */ + first_index = XFS_INO_TO_AGINO(mp, ip->i_ino + 1); + if (first_index < XFS_INO_TO_AGINO(mp, ip->i_ino)) { + read_unlock(&pag->pag_ici_lock); + break; + } + + /* nothing to sync during shutdown */ + if (XFS_FORCED_SHUTDOWN(mp)) { + read_unlock(&pag->pag_ici_lock); + return 0; + } + + /* + * If we can't get a reference on the inode, it must be + * in reclaim. Leave it for the reclaim code to flush. + */ + inode = VFS_I(ip); + if (!igrab(inode)) { + read_unlock(&pag->pag_ici_lock); + continue; + } + read_unlock(&pag->pag_ici_lock); + + /* avoid new or bad inodes */ + if (is_bad_inode(inode) || + xfs_iflags_test(ip, XFS_INEW)) { + IRELE(ip); + continue; + } + + /* + * If we have to flush data or wait for I/O completion + * we need to hold the iolock. + */ + if ((flags & SYNC_DELWRI) && VN_DIRTY(inode)) { + xfs_ilock(ip, XFS_IOLOCK_SHARED); + lock_flags |= XFS_IOLOCK_SHARED; + error = xfs_flush_pages(ip, 0, -1, fflag, FI_NONE); + if (flags & SYNC_IOWAIT) + xfs_ioend_wait(ip); + } + xfs_ilock(ip, XFS_ILOCK_SHARED); + + if ((flags & SYNC_ATTR) && !xfs_inode_clean(ip)) { + if (flags & SYNC_WAIT) { + xfs_iflock(ip); + if (!xfs_inode_clean(ip)) + error = xfs_iflush(ip, XFS_IFLUSH_SYNC); + else + xfs_ifunlock(ip); + } else if (xfs_iflock_nowait(ip)) { + if (!xfs_inode_clean(ip)) + error = xfs_iflush(ip, XFS_IFLUSH_DELWRI); + else + xfs_ifunlock(ip); + } + } + xfs_iput(ip, lock_flags); + + if (error) + last_error = error; + /* + * bail out if the filesystem is corrupted. + */ + if (error == EFSCORRUPTED) + return XFS_ERROR(error); + + } while (nr_found); + + return last_error; +} + +int +xfs_sync_inodes( + xfs_mount_t *mp, + int flags) +{ + int error; + int last_error; + int i; + int lflags = XFS_LOG_FORCE; + + if (mp->m_flags & XFS_MOUNT_RDONLY) + return 0; + error = 0; + last_error = 0; + + if (flags & SYNC_WAIT) + lflags |= XFS_LOG_SYNC; + + for (i = 0; i < mp->m_sb.sb_agcount; i++) { + if (!mp->m_perag[i].pag_ici_init) + continue; + error = xfs_sync_inodes_ag(mp, i, flags); + if (error) + last_error = error; + if (error == EFSCORRUPTED) + break; + } + if (flags & SYNC_DELWRI) + xfs_log_force(mp, 0, lflags); + + return XFS_ERROR(last_error); +} + +STATIC int +xfs_commit_dummy_trans( + struct xfs_mount *mp, + uint log_flags) +{ + struct xfs_inode *ip = mp->m_rootip; + struct xfs_trans *tp; + int error; + + /* + * Put a dummy transaction in the log to tell recovery + * that all others are OK. + */ + tp = xfs_trans_alloc(mp, XFS_TRANS_DUMMY1); + error = xfs_trans_reserve(tp, 0, XFS_ICHANGE_LOG_RES(mp), 0, 0, 0); + if (error) { + xfs_trans_cancel(tp, 0); + return error; + } + + xfs_ilock(ip, XFS_ILOCK_EXCL); + + xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL); + xfs_trans_ihold(tp, ip); + xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); + /* XXX(hch): ignoring the error here.. */ + error = xfs_trans_commit(tp, 0); + + xfs_iunlock(ip, XFS_ILOCK_EXCL); + + xfs_log_force(mp, 0, log_flags); + return 0; +} + +int +xfs_sync_fsdata( + struct xfs_mount *mp, + int flags) +{ + struct xfs_buf *bp; + struct xfs_buf_log_item *bip; + int error = 0; + + /* + * If this is xfssyncd() then only sync the superblock if we can + * lock it without sleeping and it is not pinned. + */ + if (flags & SYNC_BDFLUSH) { + ASSERT(!(flags & SYNC_WAIT)); + + bp = xfs_getsb(mp, XFS_BUF_TRYLOCK); + if (!bp) + goto out; + + bip = XFS_BUF_FSPRIVATE(bp, struct xfs_buf_log_item *); + if (!bip || !xfs_buf_item_dirty(bip) || XFS_BUF_ISPINNED(bp)) + goto out_brelse; + } else { + bp = xfs_getsb(mp, 0); + + /* + * If the buffer is pinned then push on the log so we won't + * get stuck waiting in the write for someone, maybe + * ourselves, to flush the log. + * + * Even though we just pushed the log above, we did not have + * the superblock buffer locked at that point so it can + * become pinned in between there and here. + */ + if (XFS_BUF_ISPINNED(bp)) + xfs_log_force(mp, 0, XFS_LOG_FORCE); + } + + + if (flags & SYNC_WAIT) + XFS_BUF_UNASYNC(bp); + else + XFS_BUF_ASYNC(bp); + + return xfs_bwrite(mp, bp); + + out_brelse: + xfs_buf_relse(bp); + out: + return error; +} + +/* + * When remounting a filesystem read-only or freezing the filesystem, we have + * two phases to execute. This first phase is syncing the data before we + * quiesce the filesystem, and the second is flushing all the inodes out after + * we've waited for all the transactions created by the first phase to + * complete. The second phase ensures that the inodes are written to their + * location on disk rather than just existing in transactions in the log. This + * means after a quiesce there is no log replay required to write the inodes to + * disk (this is the main difference between a sync and a quiesce). + */ +/* + * First stage of freeze - no writers will make progress now we are here, + * so we flush delwri and delalloc buffers here, then wait for all I/O to + * complete. Data is frozen at that point. Metadata is not frozen, + * transactions can still occur here so don't bother flushing the buftarg + * because it'll just get dirty again. + */ +int +xfs_quiesce_data( + struct xfs_mount *mp) +{ + int error; + + /* push non-blocking */ + xfs_sync_inodes(mp, SYNC_DELWRI|SYNC_BDFLUSH); + XFS_QM_DQSYNC(mp, SYNC_BDFLUSH); + xfs_filestream_flush(mp); + + /* push and block */ + xfs_sync_inodes(mp, SYNC_DELWRI|SYNC_WAIT|SYNC_IOWAIT); + XFS_QM_DQSYNC(mp, SYNC_WAIT); + + /* write superblock and hoover up shutdown errors */ + error = xfs_sync_fsdata(mp, 0); + + /* flush data-only devices */ + if (mp->m_rtdev_targp) + XFS_bflush(mp->m_rtdev_targp); + + return error; +} + +STATIC void +xfs_quiesce_fs( + struct xfs_mount *mp) +{ + int count = 0, pincount; + + xfs_flush_buftarg(mp->m_ddev_targp, 0); + xfs_reclaim_inodes(mp, 0, XFS_IFLUSH_DELWRI_ELSE_ASYNC); + + /* + * This loop must run at least twice. The first instance of the loop + * will flush most meta data but that will generate more meta data + * (typically directory updates). Which then must be flushed and + * logged before we can write the unmount record. + */ + do { + xfs_sync_inodes(mp, SYNC_ATTR|SYNC_WAIT); + pincount = xfs_flush_buftarg(mp->m_ddev_targp, 1); + if (!pincount) { + delay(50); + count++; + } + } while (count < 2); +} + +/* + * Second stage of a quiesce. The data is already synced, now we have to take + * care of the metadata. New transactions are already blocked, so we need to + * wait for any remaining transactions to drain out before proceding. + */ +void +xfs_quiesce_attr( + struct xfs_mount *mp) +{ + int error = 0; + + /* wait for all modifications to complete */ + while (atomic_read(&mp->m_active_trans) > 0) + delay(100); + + /* flush inodes and push all remaining buffers out to disk */ + xfs_quiesce_fs(mp); + + ASSERT_ALWAYS(atomic_read(&mp->m_active_trans) == 0); + + /* Push the superblock and write an unmount record */ + error = xfs_log_sbcount(mp, 1); + if (error) + xfs_fs_cmn_err(CE_WARN, mp, + "xfs_attr_quiesce: failed to log sb changes. " + "Frozen image may not be consistent."); + xfs_log_unmount_write(mp); + xfs_unmountfs_writesb(mp); +} + +/* + * Enqueue a work item to be picked up by the vfs xfssyncd thread. + * Doing this has two advantages: + * - It saves on stack space, which is tight in certain situations + * - It can be used (with care) as a mechanism to avoid deadlocks. + * Flushing while allocating in a full filesystem requires both. + */ +STATIC void +xfs_syncd_queue_work( + struct xfs_mount *mp, + void *data, + void (*syncer)(struct xfs_mount *, void *)) +{ + struct bhv_vfs_sync_work *work; + + work = kmem_alloc(sizeof(struct bhv_vfs_sync_work), KM_SLEEP); + INIT_LIST_HEAD(&work->w_list); + work->w_syncer = syncer; + work->w_data = data; + work->w_mount = mp; + spin_lock(&mp->m_sync_lock); + list_add_tail(&work->w_list, &mp->m_sync_list); + spin_unlock(&mp->m_sync_lock); + wake_up_process(mp->m_sync_task); +} + +/* + * Flush delayed allocate data, attempting to free up reserved space + * from existing allocations. At this point a new allocation attempt + * has failed with ENOSPC and we are in the process of scratching our + * heads, looking about for more room... + */ +STATIC void +xfs_flush_inode_work( + struct xfs_mount *mp, + void *arg) +{ + struct inode *inode = arg; + filemap_flush(inode->i_mapping); + iput(inode); +} + +void +xfs_flush_inode( + xfs_inode_t *ip) +{ + struct inode *inode = VFS_I(ip); + + igrab(inode); + xfs_syncd_queue_work(ip->i_mount, inode, xfs_flush_inode_work); + delay(msecs_to_jiffies(500)); +} + +/* + * This is the "bigger hammer" version of xfs_flush_inode_work... + * (IOW, "If at first you don't succeed, use a Bigger Hammer"). + */ +STATIC void +xfs_flush_device_work( + struct xfs_mount *mp, + void *arg) +{ + struct inode *inode = arg; + sync_blockdev(mp->m_super->s_bdev); + iput(inode); +} + +void +xfs_flush_device( + xfs_inode_t *ip) +{ + struct inode *inode = VFS_I(ip); + + igrab(inode); + xfs_syncd_queue_work(ip->i_mount, inode, xfs_flush_device_work); + delay(msecs_to_jiffies(500)); + xfs_log_force(ip->i_mount, (xfs_lsn_t)0, XFS_LOG_FORCE|XFS_LOG_SYNC); +} + +/* + * Every sync period we need to unpin all items, reclaim inodes, sync + * quota and write out the superblock. We might need to cover the log + * to indicate it is idle. + */ +STATIC void +xfs_sync_worker( + struct xfs_mount *mp, + void *unused) +{ + int error; + + if (!(mp->m_flags & XFS_MOUNT_RDONLY)) { + xfs_log_force(mp, (xfs_lsn_t)0, XFS_LOG_FORCE); + xfs_reclaim_inodes(mp, 0, XFS_IFLUSH_DELWRI_ELSE_ASYNC); + /* dgc: errors ignored here */ + error = XFS_QM_DQSYNC(mp, SYNC_BDFLUSH); + error = xfs_sync_fsdata(mp, SYNC_BDFLUSH); + if (xfs_log_need_covered(mp)) + error = xfs_commit_dummy_trans(mp, XFS_LOG_FORCE); + } + mp->m_sync_seq++; + wake_up(&mp->m_wait_single_sync_task); +} + +STATIC int +xfssyncd( + void *arg) +{ + struct xfs_mount *mp = arg; + long timeleft; + bhv_vfs_sync_work_t *work, *n; + LIST_HEAD (tmp); + + set_freezable(); + timeleft = xfs_syncd_centisecs * msecs_to_jiffies(10); + for (;;) { + timeleft = schedule_timeout_interruptible(timeleft); + /* swsusp */ + try_to_freeze(); + if (kthread_should_stop() && list_empty(&mp->m_sync_list)) + break; + + spin_lock(&mp->m_sync_lock); + /* + * We can get woken by laptop mode, to do a sync - + * that's the (only!) case where the list would be + * empty with time remaining. + */ + if (!timeleft || list_empty(&mp->m_sync_list)) { + if (!timeleft) + timeleft = xfs_syncd_centisecs * + msecs_to_jiffies(10); + INIT_LIST_HEAD(&mp->m_sync_work.w_list); + list_add_tail(&mp->m_sync_work.w_list, + &mp->m_sync_list); + } + list_for_each_entry_safe(work, n, &mp->m_sync_list, w_list) + list_move(&work->w_list, &tmp); + spin_unlock(&mp->m_sync_lock); + + list_for_each_entry_safe(work, n, &tmp, w_list) { + (*work->w_syncer)(mp, work->w_data); + list_del(&work->w_list); + if (work == &mp->m_sync_work) + continue; + kmem_free(work); + } + } + + return 0; +} + +int +xfs_syncd_init( + struct xfs_mount *mp) +{ + mp->m_sync_work.w_syncer = xfs_sync_worker; + mp->m_sync_work.w_mount = mp; + mp->m_sync_task = kthread_run(xfssyncd, mp, "xfssyncd"); + if (IS_ERR(mp->m_sync_task)) + return -PTR_ERR(mp->m_sync_task); + return 0; +} + +void +xfs_syncd_stop( + struct xfs_mount *mp) +{ + kthread_stop(mp->m_sync_task); +} + +int +xfs_reclaim_inode( + xfs_inode_t *ip, + int locked, + int sync_mode) +{ + xfs_perag_t *pag = xfs_get_perag(ip->i_mount, ip->i_ino); + + /* The hash lock here protects a thread in xfs_iget_core from + * racing with us on linking the inode back with a vnode. + * Once we have the XFS_IRECLAIM flag set it will not touch + * us. + */ + write_lock(&pag->pag_ici_lock); + spin_lock(&ip->i_flags_lock); + if (__xfs_iflags_test(ip, XFS_IRECLAIM) || + !__xfs_iflags_test(ip, XFS_IRECLAIMABLE)) { + spin_unlock(&ip->i_flags_lock); + write_unlock(&pag->pag_ici_lock); + if (locked) { + xfs_ifunlock(ip); + xfs_iunlock(ip, XFS_ILOCK_EXCL); + } + return 1; + } + __xfs_iflags_set(ip, XFS_IRECLAIM); + spin_unlock(&ip->i_flags_lock); + write_unlock(&pag->pag_ici_lock); + xfs_put_perag(ip->i_mount, pag); + + /* + * If the inode is still dirty, then flush it out. If the inode + * is not in the AIL, then it will be OK to flush it delwri as + * long as xfs_iflush() does not keep any references to the inode. + * We leave that decision up to xfs_iflush() since it has the + * knowledge of whether it's OK to simply do a delwri flush of + * the inode or whether we need to wait until the inode is + * pulled from the AIL. + * We get the flush lock regardless, though, just to make sure + * we don't free it while it is being flushed. + */ + if (!locked) { + xfs_ilock(ip, XFS_ILOCK_EXCL); + xfs_iflock(ip); + } + + /* + * In the case of a forced shutdown we rely on xfs_iflush() to + * wait for the inode to be unpinned before returning an error. + */ + if (!is_bad_inode(VFS_I(ip)) && xfs_iflush(ip, sync_mode) == 0) { + /* synchronize with xfs_iflush_done */ + xfs_iflock(ip); + xfs_ifunlock(ip); + } + + xfs_iunlock(ip, XFS_ILOCK_EXCL); + xfs_ireclaim(ip); + return 0; +} + +/* + * We set the inode flag atomically with the radix tree tag. + * Once we get tag lookups on the radix tree, this inode flag + * can go away. + */ +void +xfs_inode_set_reclaim_tag( + xfs_inode_t *ip) +{ + xfs_mount_t *mp = ip->i_mount; + xfs_perag_t *pag = xfs_get_perag(mp, ip->i_ino); + + read_lock(&pag->pag_ici_lock); + spin_lock(&ip->i_flags_lock); + radix_tree_tag_set(&pag->pag_ici_root, + XFS_INO_TO_AGINO(mp, ip->i_ino), XFS_ICI_RECLAIM_TAG); + __xfs_iflags_set(ip, XFS_IRECLAIMABLE); + spin_unlock(&ip->i_flags_lock); + read_unlock(&pag->pag_ici_lock); + xfs_put_perag(mp, pag); +} + +void +__xfs_inode_clear_reclaim_tag( + xfs_mount_t *mp, + xfs_perag_t *pag, + xfs_inode_t *ip) +{ + radix_tree_tag_clear(&pag->pag_ici_root, + XFS_INO_TO_AGINO(mp, ip->i_ino), XFS_ICI_RECLAIM_TAG); +} + +void +xfs_inode_clear_reclaim_tag( + xfs_inode_t *ip) +{ + xfs_mount_t *mp = ip->i_mount; + xfs_perag_t *pag = xfs_get_perag(mp, ip->i_ino); + + read_lock(&pag->pag_ici_lock); + spin_lock(&ip->i_flags_lock); + __xfs_inode_clear_reclaim_tag(mp, pag, ip); + spin_unlock(&ip->i_flags_lock); + read_unlock(&pag->pag_ici_lock); + xfs_put_perag(mp, pag); +} + + +STATIC void +xfs_reclaim_inodes_ag( + xfs_mount_t *mp, + int ag, + int noblock, + int mode) +{ + xfs_inode_t *ip = NULL; + xfs_perag_t *pag = &mp->m_perag[ag]; + int nr_found; + uint32_t first_index; + int skipped; + +restart: + first_index = 0; + skipped = 0; + do { + /* + * use a gang lookup to find the next inode in the tree + * as the tree is sparse and a gang lookup walks to find + * the number of objects requested. + */ + read_lock(&pag->pag_ici_lock); + nr_found = radix_tree_gang_lookup_tag(&pag->pag_ici_root, + (void**)&ip, first_index, 1, + XFS_ICI_RECLAIM_TAG); + + if (!nr_found) { + read_unlock(&pag->pag_ici_lock); + break; + } + + /* + * Update the index for the next lookup. Catch overflows + * into the next AG range which can occur if we have inodes + * in the last block of the AG and we are currently + * pointing to the last inode. + */ + first_index = XFS_INO_TO_AGINO(mp, ip->i_ino + 1); + if (first_index < XFS_INO_TO_AGINO(mp, ip->i_ino)) { + read_unlock(&pag->pag_ici_lock); + break; + } + + /* ignore if already under reclaim */ + if (xfs_iflags_test(ip, XFS_IRECLAIM)) { + read_unlock(&pag->pag_ici_lock); + continue; + } + + if (noblock) { + if (!xfs_ilock_nowait(ip, XFS_ILOCK_EXCL)) { + read_unlock(&pag->pag_ici_lock); + continue; + } + if (xfs_ipincount(ip) || + !xfs_iflock_nowait(ip)) { + xfs_iunlock(ip, XFS_ILOCK_EXCL); + read_unlock(&pag->pag_ici_lock); + continue; + } + } + read_unlock(&pag->pag_ici_lock); + + /* + * hmmm - this is an inode already in reclaim. Do + * we even bother catching it here? + */ + if (xfs_reclaim_inode(ip, noblock, mode)) + skipped++; + } while (nr_found); + + if (skipped) { + delay(1); + goto restart; + } + return; + +} + +int +xfs_reclaim_inodes( + xfs_mount_t *mp, + int noblock, + int mode) +{ + int i; + + for (i = 0; i < mp->m_sb.sb_agcount; i++) { + if (!mp->m_perag[i].pag_ici_init) + continue; + xfs_reclaim_inodes_ag(mp, i, noblock, mode); + } + return 0; +} + + diff --git a/fs/xfs/linux-2.6/xfs_sync.h b/fs/xfs/linux-2.6/xfs_sync.h new file mode 100644 index 00000000000..5f6de1efe1f --- /dev/null +++ b/fs/xfs/linux-2.6/xfs_sync.h @@ -0,0 +1,55 @@ +/* + * Copyright (c) 2000-2006 Silicon Graphics, Inc. + * All Rights Reserved. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it would be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write the Free Software Foundation, + * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA + */ +#ifndef XFS_SYNC_H +#define XFS_SYNC_H 1 + +struct xfs_mount; + +typedef struct bhv_vfs_sync_work { + struct list_head w_list; + struct xfs_mount *w_mount; + void *w_data; /* syncer routine argument */ + void (*w_syncer)(struct xfs_mount *, void *); +} bhv_vfs_sync_work_t; + +#define SYNC_ATTR 0x0001 /* sync attributes */ +#define SYNC_DELWRI 0x0002 /* look at delayed writes */ +#define SYNC_WAIT 0x0004 /* wait for i/o to complete */ +#define SYNC_BDFLUSH 0x0008 /* BDFLUSH is calling -- don't block */ +#define SYNC_IOWAIT 0x0010 /* wait for all I/O to complete */ + +int xfs_syncd_init(struct xfs_mount *mp); +void xfs_syncd_stop(struct xfs_mount *mp); + +int xfs_sync_inodes(struct xfs_mount *mp, int flags); +int xfs_sync_fsdata(struct xfs_mount *mp, int flags); + +int xfs_quiesce_data(struct xfs_mount *mp); +void xfs_quiesce_attr(struct xfs_mount *mp); + +void xfs_flush_inode(struct xfs_inode *ip); +void xfs_flush_device(struct xfs_inode *ip); + +int xfs_reclaim_inode(struct xfs_inode *ip, int locked, int sync_mode); +int xfs_reclaim_inodes(struct xfs_mount *mp, int noblock, int mode); + +void xfs_inode_set_reclaim_tag(struct xfs_inode *ip); +void xfs_inode_clear_reclaim_tag(struct xfs_inode *ip); +void __xfs_inode_clear_reclaim_tag(struct xfs_mount *mp, struct xfs_perag *pag, + struct xfs_inode *ip); +#endif diff --git a/fs/xfs/linux-2.6/xfs_sysctl.c b/fs/xfs/linux-2.6/xfs_sysctl.c index 7dacb5bbde3..916c0ffb608 100644 --- a/fs/xfs/linux-2.6/xfs_sysctl.c +++ b/fs/xfs/linux-2.6/xfs_sysctl.c @@ -56,17 +56,6 @@ xfs_stats_clear_proc_handler( static ctl_table xfs_table[] = { { - .ctl_name = XFS_RESTRICT_CHOWN, - .procname = "restrict_chown", - .data = &xfs_params.restrict_chown.val, - .maxlen = sizeof(int), - .mode = 0644, - .proc_handler = &proc_dointvec_minmax, - .strategy = &sysctl_intvec, - .extra1 = &xfs_params.restrict_chown.min, - .extra2 = &xfs_params.restrict_chown.max - }, - { .ctl_name = XFS_SGID_INHERIT, .procname = "irix_sgid_inherit", .data = &xfs_params.sgid_inherit.val, diff --git a/fs/xfs/linux-2.6/xfs_sysctl.h b/fs/xfs/linux-2.6/xfs_sysctl.h index 4aadb8056c3..b9937d450f8 100644 --- a/fs/xfs/linux-2.6/xfs_sysctl.h +++ b/fs/xfs/linux-2.6/xfs_sysctl.h @@ -31,7 +31,6 @@ typedef struct xfs_sysctl_val { } xfs_sysctl_val_t; typedef struct xfs_param { - xfs_sysctl_val_t restrict_chown;/* Root/non-root can give away files.*/ xfs_sysctl_val_t sgid_inherit; /* Inherit S_ISGID if process' GID is * not a member of parent dir GID. */ xfs_sysctl_val_t symlink_mode; /* Link creat mode affected by umask */ @@ -68,7 +67,7 @@ typedef struct xfs_param { enum { /* XFS_REFCACHE_SIZE = 1 */ /* XFS_REFCACHE_PURGE = 2 */ - XFS_RESTRICT_CHOWN = 3, + /* XFS_RESTRICT_CHOWN = 3 */ XFS_SGID_INHERIT = 4, XFS_SYMLINK_MODE = 5, XFS_PANIC_MASK = 6, diff --git a/fs/xfs/linux-2.6/xfs_vfs.h b/fs/xfs/linux-2.6/xfs_vfs.h deleted file mode 100644 index 7e60c7776b1..00000000000 --- a/fs/xfs/linux-2.6/xfs_vfs.h +++ /dev/null @@ -1,77 +0,0 @@ -/* - * Copyright (c) 2000-2006 Silicon Graphics, Inc. - * All Rights Reserved. - * - * This program is free software; you can redistribute it and/or - * modify it under the terms of the GNU General Public License as - * published by the Free Software Foundation. - * - * This program is distributed in the hope that it would be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write the Free Software Foundation, - * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA - */ -#ifndef __XFS_VFS_H__ -#define __XFS_VFS_H__ - -#include <linux/vfs.h> -#include "xfs_fs.h" - -struct inode; - -struct fid; -struct cred; -struct seq_file; -struct super_block; -struct xfs_inode; -struct xfs_mount; -struct xfs_mount_args; - -typedef struct kstatfs bhv_statvfs_t; - -typedef struct bhv_vfs_sync_work { - struct list_head w_list; - struct xfs_mount *w_mount; - void *w_data; /* syncer routine argument */ - void (*w_syncer)(struct xfs_mount *, void *); -} bhv_vfs_sync_work_t; - -#define SYNC_ATTR 0x0001 /* sync attributes */ -#define SYNC_CLOSE 0x0002 /* close file system down */ -#define SYNC_DELWRI 0x0004 /* look at delayed writes */ -#define SYNC_WAIT 0x0008 /* wait for i/o to complete */ -#define SYNC_BDFLUSH 0x0010 /* BDFLUSH is calling -- don't block */ -#define SYNC_FSDATA 0x0020 /* flush fs data (e.g. superblocks) */ -#define SYNC_REFCACHE 0x0040 /* prune some of the nfs ref cache */ -#define SYNC_REMOUNT 0x0080 /* remount readonly, no dummy LRs */ -#define SYNC_IOWAIT 0x0100 /* wait for all I/O to complete */ - -/* - * When remounting a filesystem read-only or freezing the filesystem, - * we have two phases to execute. This first phase is syncing the data - * before we quiesce the fielsystem, and the second is flushing all the - * inodes out after we've waited for all the transactions created by - * the first phase to complete. The second phase uses SYNC_INODE_QUIESCE - * to ensure that the inodes are written to their location on disk - * rather than just existing in transactions in the log. This means - * after a quiesce there is no log replay required to write the inodes - * to disk (this is the main difference between a sync and a quiesce). - */ -#define SYNC_DATA_QUIESCE (SYNC_DELWRI|SYNC_FSDATA|SYNC_WAIT|SYNC_IOWAIT) -#define SYNC_INODE_QUIESCE (SYNC_REMOUNT|SYNC_ATTR|SYNC_WAIT) - -#define SHUTDOWN_META_IO_ERROR 0x0001 /* write attempt to metadata failed */ -#define SHUTDOWN_LOG_IO_ERROR 0x0002 /* write attempt to the log failed */ -#define SHUTDOWN_FORCE_UMOUNT 0x0004 /* shutdown from a forced unmount */ -#define SHUTDOWN_CORRUPT_INCORE 0x0008 /* corrupt in-memory data structures */ -#define SHUTDOWN_REMOTE_REQ 0x0010 /* shutdown came from remote cell */ -#define SHUTDOWN_DEVICE_REQ 0x0020 /* failed all paths to the device */ - -#define xfs_test_for_freeze(mp) ((mp)->m_super->s_frozen) -#define xfs_wait_for_freeze(mp,l) vfs_check_frozen((mp)->m_super, (l)) - -#endif /* __XFS_VFS_H__ */ diff --git a/fs/xfs/linux-2.6/xfs_vnode.c b/fs/xfs/linux-2.6/xfs_vnode.c deleted file mode 100644 index b52528bbbff..00000000000 --- a/fs/xfs/linux-2.6/xfs_vnode.c +++ /dev/null @@ -1,145 +0,0 @@ -/* - * Copyright (c) 2000-2003,2005 Silicon Graphics, Inc. - * All Rights Reserved. - * - * This program is free software; you can redistribute it and/or - * modify it under the terms of the GNU General Public License as - * published by the Free Software Foundation. - * - * This program is distributed in the hope that it would be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write the Free Software Foundation, - * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA - */ -#include "xfs.h" -#include "xfs_vnodeops.h" -#include "xfs_bmap_btree.h" -#include "xfs_inode.h" - -/* - * And this gunk is needed for xfs_mount.h" - */ -#include "xfs_log.h" -#include "xfs_trans.h" -#include "xfs_sb.h" -#include "xfs_dmapi.h" -#include "xfs_inum.h" -#include "xfs_ag.h" -#include "xfs_mount.h" - - -/* - * Dedicated vnode inactive/reclaim sync wait queues. - * Prime number of hash buckets since address is used as the key. - */ -#define NVSYNC 37 -#define vptosync(v) (&vsync[((unsigned long)v) % NVSYNC]) -static wait_queue_head_t vsync[NVSYNC]; - -void __init -vn_init(void) -{ - int i; - - for (i = 0; i < NVSYNC; i++) - init_waitqueue_head(&vsync[i]); -} - -void -vn_iowait( - xfs_inode_t *ip) -{ - wait_queue_head_t *wq = vptosync(ip); - - wait_event(*wq, (atomic_read(&ip->i_iocount) == 0)); -} - -void -vn_iowake( - xfs_inode_t *ip) -{ - if (atomic_dec_and_test(&ip->i_iocount)) - wake_up(vptosync(ip)); -} - -/* - * Volume managers supporting multiple paths can send back ENODEV when the - * final path disappears. In this case continuing to fill the page cache - * with dirty data which cannot be written out is evil, so prevent that. - */ -void -vn_ioerror( - xfs_inode_t *ip, - int error, - char *f, - int l) -{ - if (unlikely(error == -ENODEV)) - xfs_do_force_shutdown(ip->i_mount, SHUTDOWN_DEVICE_REQ, f, l); -} - -#ifdef XFS_INODE_TRACE - -/* - * Reference count of Linux inode if present, -1 if the xfs_inode - * has no associated Linux inode. - */ -static inline int xfs_icount(struct xfs_inode *ip) -{ - struct inode *vp = VFS_I(ip); - - if (vp) - return vn_count(vp); - return -1; -} - -#define KTRACE_ENTER(ip, vk, s, line, ra) \ - ktrace_enter( (ip)->i_trace, \ -/* 0 */ (void *)(__psint_t)(vk), \ -/* 1 */ (void *)(s), \ -/* 2 */ (void *)(__psint_t) line, \ -/* 3 */ (void *)(__psint_t)xfs_icount(ip), \ -/* 4 */ (void *)(ra), \ -/* 5 */ NULL, \ -/* 6 */ (void *)(__psint_t)current_cpu(), \ -/* 7 */ (void *)(__psint_t)current_pid(), \ -/* 8 */ (void *)__return_address, \ -/* 9 */ NULL, NULL, NULL, NULL, NULL, NULL, NULL) - -/* - * Vnode tracing code. - */ -void -_xfs_itrace_entry(xfs_inode_t *ip, const char *func, inst_t *ra) -{ - KTRACE_ENTER(ip, INODE_KTRACE_ENTRY, func, 0, ra); -} - -void -_xfs_itrace_exit(xfs_inode_t *ip, const char *func, inst_t *ra) -{ - KTRACE_ENTER(ip, INODE_KTRACE_EXIT, func, 0, ra); -} - -void -xfs_itrace_hold(xfs_inode_t *ip, char *file, int line, inst_t *ra) -{ - KTRACE_ENTER(ip, INODE_KTRACE_HOLD, file, line, ra); -} - -void -_xfs_itrace_ref(xfs_inode_t *ip, char *file, int line, inst_t *ra) -{ - KTRACE_ENTER(ip, INODE_KTRACE_REF, file, line, ra); -} - -void -xfs_itrace_rele(xfs_inode_t *ip, char *file, int line, inst_t *ra) -{ - KTRACE_ENTER(ip, INODE_KTRACE_RELE, file, line, ra); -} -#endif /* XFS_INODE_TRACE */ diff --git a/fs/xfs/linux-2.6/xfs_vnode.h b/fs/xfs/linux-2.6/xfs_vnode.h index 683ce16210f..f65983a230d 100644 --- a/fs/xfs/linux-2.6/xfs_vnode.h +++ b/fs/xfs/linux-2.6/xfs_vnode.h @@ -18,7 +18,10 @@ #ifndef __XFS_VNODE_H__ #define __XFS_VNODE_H__ +#include "xfs_fs.h" + struct file; +struct xfs_inode; struct xfs_iomap; struct attrlist_cursor_kern; @@ -51,40 +54,6 @@ struct attrlist_cursor_kern; Prevent VM access to the pages until the operation completes. */ - -extern void vn_init(void); - -/* - * Yeah, these don't take vnode anymore at all, all this should be - * cleaned up at some point. - */ -extern void vn_iowait(struct xfs_inode *ip); -extern void vn_iowake(struct xfs_inode *ip); -extern void vn_ioerror(struct xfs_inode *ip, int error, char *f, int l); - -static inline int vn_count(struct inode *vp) -{ - return atomic_read(&vp->i_count); -} - -#define IHOLD(ip) \ -do { \ - ASSERT(atomic_read(&VFS_I(ip)->i_count) > 0) ; \ - atomic_inc(&(VFS_I(ip)->i_count)); \ - xfs_itrace_hold((ip), __FILE__, __LINE__, (inst_t *)__return_address); \ -} while (0) - -#define IRELE(ip) \ -do { \ - xfs_itrace_rele((ip), __FILE__, __LINE__, (inst_t *)__return_address); \ - iput(VFS_I(ip)); \ -} while (0) - -static inline struct inode *vn_grab(struct inode *vp) -{ - return igrab(vp); -} - /* * Dealing with bad inodes */ @@ -121,39 +90,4 @@ static inline void vn_atime_to_time_t(struct inode *vp, time_t *tt) PAGECACHE_TAG_DIRTY) -/* - * Tracking vnode activity. - */ -#if defined(XFS_INODE_TRACE) - -#define INODE_TRACE_SIZE 16 /* number of trace entries */ -#define INODE_KTRACE_ENTRY 1 -#define INODE_KTRACE_EXIT 2 -#define INODE_KTRACE_HOLD 3 -#define INODE_KTRACE_REF 4 -#define INODE_KTRACE_RELE 5 - -extern void _xfs_itrace_entry(struct xfs_inode *, const char *, inst_t *); -extern void _xfs_itrace_exit(struct xfs_inode *, const char *, inst_t *); -extern void xfs_itrace_hold(struct xfs_inode *, char *, int, inst_t *); -extern void _xfs_itrace_ref(struct xfs_inode *, char *, int, inst_t *); -extern void xfs_itrace_rele(struct xfs_inode *, char *, int, inst_t *); -#define xfs_itrace_entry(ip) \ - _xfs_itrace_entry(ip, __func__, (inst_t *)__return_address) -#define xfs_itrace_exit(ip) \ - _xfs_itrace_exit(ip, __func__, (inst_t *)__return_address) -#define xfs_itrace_exit_tag(ip, tag) \ - _xfs_itrace_exit(ip, tag, (inst_t *)__return_address) -#define xfs_itrace_ref(ip) \ - _xfs_itrace_ref(ip, __FILE__, __LINE__, (inst_t *)__return_address) - -#else -#define xfs_itrace_entry(a) -#define xfs_itrace_exit(a) -#define xfs_itrace_exit_tag(a, b) -#define xfs_itrace_hold(a, b, c, d) -#define xfs_itrace_ref(a) -#define xfs_itrace_rele(a, b, c, d) -#endif - #endif /* __XFS_VNODE_H__ */ diff --git a/fs/xfs/quota/xfs_dquot.c b/fs/xfs/quota/xfs_dquot.c index f2705f2fd43..591ca6602bf 100644 --- a/fs/xfs/quota/xfs_dquot.c +++ b/fs/xfs/quota/xfs_dquot.c @@ -101,7 +101,7 @@ xfs_qm_dqinit( if (brandnewdquot) { dqp->dq_flnext = dqp->dq_flprev = dqp; mutex_init(&dqp->q_qlock); - sv_init(&dqp->q_pinwait, SV_DEFAULT, "pdq"); + init_waitqueue_head(&dqp->q_pinwait); /* * Because we want to use a counting completion, complete @@ -131,7 +131,7 @@ xfs_qm_dqinit( dqp->q_res_bcount = 0; dqp->q_res_icount = 0; dqp->q_res_rtbcount = 0; - dqp->q_pincount = 0; + atomic_set(&dqp->q_pincount, 0); dqp->q_hash = NULL; ASSERT(dqp->dq_flnext == dqp->dq_flprev); @@ -1221,16 +1221,14 @@ xfs_qm_dqflush( xfs_dqtrace_entry(dqp, "DQFLUSH"); /* - * If not dirty, nada. + * If not dirty, or it's pinned and we are not supposed to + * block, nada. */ - if (!XFS_DQ_IS_DIRTY(dqp)) { + if (!XFS_DQ_IS_DIRTY(dqp) || + (!(flags & XFS_QMOPT_SYNC) && atomic_read(&dqp->q_pincount) > 0)) { xfs_dqfunlock(dqp); - return (0); + return 0; } - - /* - * Cant flush a pinned dquot. Wait for it. - */ xfs_qm_dqunpin_wait(dqp); /* @@ -1274,10 +1272,8 @@ xfs_qm_dqflush( dqp->dq_flags &= ~(XFS_DQ_DIRTY); mp = dqp->q_mount; - /* lsn is 64 bits */ - spin_lock(&mp->m_ail_lock); - dqp->q_logitem.qli_flush_lsn = dqp->q_logitem.qli_item.li_lsn; - spin_unlock(&mp->m_ail_lock); + xfs_trans_ail_copy_lsn(mp->m_ail, &dqp->q_logitem.qli_flush_lsn, + &dqp->q_logitem.qli_item.li_lsn); /* * Attach an iodone routine so that we can remove this dquot from the @@ -1323,8 +1319,10 @@ xfs_qm_dqflush_done( xfs_dq_logitem_t *qip) { xfs_dquot_t *dqp; + struct xfs_ail *ailp; dqp = qip->qli_dquot; + ailp = qip->qli_item.li_ailp; /* * We only want to pull the item from the AIL if its @@ -1337,15 +1335,12 @@ xfs_qm_dqflush_done( if ((qip->qli_item.li_flags & XFS_LI_IN_AIL) && qip->qli_item.li_lsn == qip->qli_flush_lsn) { - spin_lock(&dqp->q_mount->m_ail_lock); - /* - * xfs_trans_delete_ail() drops the AIL lock. - */ + /* xfs_trans_ail_delete() drops the AIL lock. */ + spin_lock(&ailp->xa_lock); if (qip->qli_item.li_lsn == qip->qli_flush_lsn) - xfs_trans_delete_ail(dqp->q_mount, - (xfs_log_item_t*)qip); + xfs_trans_ail_delete(ailp, (xfs_log_item_t*)qip); else - spin_unlock(&dqp->q_mount->m_ail_lock); + spin_unlock(&ailp->xa_lock); } /* @@ -1375,7 +1370,7 @@ xfs_dqunlock( mutex_unlock(&(dqp->q_qlock)); if (dqp->q_logitem.qli_dquot == dqp) { /* Once was dqp->q_mount, but might just have been cleared */ - xfs_trans_unlocked_item(dqp->q_logitem.qli_item.li_mountp, + xfs_trans_unlocked_item(dqp->q_logitem.qli_item.li_ailp, (xfs_log_item_t*)&(dqp->q_logitem)); } } @@ -1489,7 +1484,7 @@ xfs_qm_dqpurge( "xfs_qm_dqpurge: dquot %p flush failed", dqp); xfs_dqflock(dqp); } - ASSERT(dqp->q_pincount == 0); + ASSERT(atomic_read(&dqp->q_pincount) == 0); ASSERT(XFS_FORCED_SHUTDOWN(mp) || !(dqp->q_logitem.qli_item.li_flags & XFS_LI_IN_AIL)); diff --git a/fs/xfs/quota/xfs_dquot.h b/fs/xfs/quota/xfs_dquot.h index 8958d0faf8d..7e455337e2b 100644 --- a/fs/xfs/quota/xfs_dquot.h +++ b/fs/xfs/quota/xfs_dquot.h @@ -83,8 +83,8 @@ typedef struct xfs_dquot { xfs_qcnt_t q_res_rtbcount;/* total realtime blks used+reserved */ mutex_t q_qlock; /* quota lock */ struct completion q_flush; /* flush completion queue */ - uint q_pincount; /* pin count for this dquot */ - sv_t q_pinwait; /* sync var for pinning */ + atomic_t q_pincount; /* dquot pin count */ + wait_queue_head_t q_pinwait; /* dquot pinning wait queue */ #ifdef XFS_DQUOT_TRACE struct ktrace *q_trace; /* trace header structure */ #endif diff --git a/fs/xfs/quota/xfs_dquot_item.c b/fs/xfs/quota/xfs_dquot_item.c index f028644caa5..1728f6a7c4f 100644 --- a/fs/xfs/quota/xfs_dquot_item.c +++ b/fs/xfs/quota/xfs_dquot_item.c @@ -88,25 +88,22 @@ xfs_qm_dquot_logitem_format( /* * Increment the pin count of the given dquot. - * This value is protected by pinlock spinlock in the xQM structure. */ STATIC void xfs_qm_dquot_logitem_pin( xfs_dq_logitem_t *logitem) { - xfs_dquot_t *dqp; + xfs_dquot_t *dqp = logitem->qli_dquot; - dqp = logitem->qli_dquot; ASSERT(XFS_DQ_IS_LOCKED(dqp)); - spin_lock(&(XFS_DQ_TO_QINF(dqp)->qi_pinlock)); - dqp->q_pincount++; - spin_unlock(&(XFS_DQ_TO_QINF(dqp)->qi_pinlock)); + atomic_inc(&dqp->q_pincount); } /* * Decrement the pin count of the given dquot, and wake up * anyone in xfs_dqwait_unpin() if the count goes to 0. The - * dquot must have been previously pinned with a call to xfs_dqpin(). + * dquot must have been previously pinned with a call to + * xfs_qm_dquot_logitem_pin(). */ /* ARGSUSED */ STATIC void @@ -114,16 +111,11 @@ xfs_qm_dquot_logitem_unpin( xfs_dq_logitem_t *logitem, int stale) { - xfs_dquot_t *dqp; + xfs_dquot_t *dqp = logitem->qli_dquot; - dqp = logitem->qli_dquot; - ASSERT(dqp->q_pincount > 0); - spin_lock(&(XFS_DQ_TO_QINF(dqp)->qi_pinlock)); - dqp->q_pincount--; - if (dqp->q_pincount == 0) { - sv_broadcast(&dqp->q_pinwait); - } - spin_unlock(&(XFS_DQ_TO_QINF(dqp)->qi_pinlock)); + ASSERT(atomic_read(&dqp->q_pincount) > 0); + if (atomic_dec_and_test(&dqp->q_pincount)) + wake_up(&dqp->q_pinwait); } /* ARGSUSED */ @@ -193,21 +185,14 @@ xfs_qm_dqunpin_wait( xfs_dquot_t *dqp) { ASSERT(XFS_DQ_IS_LOCKED(dqp)); - if (dqp->q_pincount == 0) { + if (atomic_read(&dqp->q_pincount) == 0) return; - } /* * Give the log a push so we don't wait here too long. */ xfs_log_force(dqp->q_mount, (xfs_lsn_t)0, XFS_LOG_FORCE); - spin_lock(&(XFS_DQ_TO_QINF(dqp)->qi_pinlock)); - if (dqp->q_pincount == 0) { - spin_unlock(&(XFS_DQ_TO_QINF(dqp)->qi_pinlock)); - return; - } - sv_wait(&(dqp->q_pinwait), PINOD, - &(XFS_DQ_TO_QINF(dqp)->qi_pinlock), s); + wait_event(dqp->q_pinwait, (atomic_read(&dqp->q_pincount) == 0)); } /* @@ -310,7 +295,7 @@ xfs_qm_dquot_logitem_trylock( uint retval; dqp = qip->qli_dquot; - if (dqp->q_pincount > 0) + if (atomic_read(&dqp->q_pincount) > 0) return (XFS_ITEM_PINNED); if (! xfs_qm_dqlock_nowait(dqp)) @@ -568,14 +553,16 @@ xfs_qm_qoffend_logitem_committed( xfs_lsn_t lsn) { xfs_qoff_logitem_t *qfs; + struct xfs_ail *ailp; qfs = qfe->qql_start_lip; - spin_lock(&qfs->qql_item.li_mountp->m_ail_lock); + ailp = qfs->qql_item.li_ailp; + spin_lock(&ailp->xa_lock); /* * Delete the qoff-start logitem from the AIL. - * xfs_trans_delete_ail() drops the AIL lock. + * xfs_trans_ail_delete() drops the AIL lock. */ - xfs_trans_delete_ail(qfs->qql_item.li_mountp, (xfs_log_item_t *)qfs); + xfs_trans_ail_delete(ailp, (xfs_log_item_t *)qfs); kmem_free(qfs); kmem_free(qfe); return (xfs_lsn_t)-1; diff --git a/fs/xfs/quota/xfs_qm.c b/fs/xfs/quota/xfs_qm.c index df0ffef9775..6b13960cf31 100644 --- a/fs/xfs/quota/xfs_qm.c +++ b/fs/xfs/quota/xfs_qm.c @@ -20,7 +20,6 @@ #include "xfs_bit.h" #include "xfs_log.h" #include "xfs_inum.h" -#include "xfs_clnt.h" #include "xfs_trans.h" #include "xfs_sb.h" #include "xfs_ag.h" @@ -396,13 +395,10 @@ xfs_qm_mount_quotas( /* * Called from the vfsops layer. */ -int +void xfs_qm_unmount_quotas( xfs_mount_t *mp) { - xfs_inode_t *uqp, *gqp; - int error = 0; - /* * Release the dquots that root inode, et al might be holding, * before we flush quotas and blow away the quotainfo structure. @@ -415,43 +411,18 @@ xfs_qm_unmount_quotas( xfs_qm_dqdetach(mp->m_rsumip); /* - * Flush out the quota inodes. + * Release the quota inodes. */ - uqp = gqp = NULL; if (mp->m_quotainfo) { - if ((uqp = mp->m_quotainfo->qi_uquotaip) != NULL) { - xfs_ilock(uqp, XFS_ILOCK_EXCL); - xfs_iflock(uqp); - error = xfs_iflush(uqp, XFS_IFLUSH_SYNC); - xfs_iunlock(uqp, XFS_ILOCK_EXCL); - if (unlikely(error == EFSCORRUPTED)) { - XFS_ERROR_REPORT("xfs_qm_unmount_quotas(1)", - XFS_ERRLEVEL_LOW, mp); - goto out; - } + if (mp->m_quotainfo->qi_uquotaip) { + IRELE(mp->m_quotainfo->qi_uquotaip); + mp->m_quotainfo->qi_uquotaip = NULL; } - if ((gqp = mp->m_quotainfo->qi_gquotaip) != NULL) { - xfs_ilock(gqp, XFS_ILOCK_EXCL); - xfs_iflock(gqp); - error = xfs_iflush(gqp, XFS_IFLUSH_SYNC); - xfs_iunlock(gqp, XFS_ILOCK_EXCL); - if (unlikely(error == EFSCORRUPTED)) { - XFS_ERROR_REPORT("xfs_qm_unmount_quotas(2)", - XFS_ERRLEVEL_LOW, mp); - goto out; - } + if (mp->m_quotainfo->qi_gquotaip) { + IRELE(mp->m_quotainfo->qi_gquotaip); + mp->m_quotainfo->qi_gquotaip = NULL; } } - if (uqp) { - IRELE(uqp); - mp->m_quotainfo->qi_uquotaip = NULL; - } - if (gqp) { - IRELE(gqp); - mp->m_quotainfo->qi_gquotaip = NULL; - } -out: - return XFS_ERROR(error); } /* @@ -987,14 +958,10 @@ xfs_qm_dqdetach( } /* - * This is called by VFS_SYNC and flags arg determines the caller, - * and its motives, as done in xfs_sync. - * - * vfs_sync: SYNC_FSDATA|SYNC_ATTR|SYNC_BDFLUSH 0x31 - * syscall sync: SYNC_FSDATA|SYNC_ATTR|SYNC_DELWRI 0x25 - * umountroot : SYNC_WAIT | SYNC_CLOSE | SYNC_ATTR | SYNC_FSDATA + * This is called to sync quotas. We can be told to use non-blocking + * semantics by either the SYNC_BDFLUSH flag or the absence of the + * SYNC_WAIT flag. */ - int xfs_qm_sync( xfs_mount_t *mp, @@ -1137,7 +1104,6 @@ xfs_qm_init_quotainfo( return error; } - spin_lock_init(&qinf->qi_pinlock); xfs_qm_list_init(&qinf->qi_dqlist, "mpdqlist", 0); qinf->qi_dqreclaims = 0; @@ -1234,7 +1200,6 @@ xfs_qm_destroy_quotainfo( */ xfs_qm_rele_quotafs_ref(mp); - spinlock_destroy(&qi->qi_pinlock); xfs_qm_list_destroy(&qi->qi_dqlist); if (qi->qi_uquotaip) { diff --git a/fs/xfs/quota/xfs_qm.h b/fs/xfs/quota/xfs_qm.h index 44f25349e47..ddf09166387 100644 --- a/fs/xfs/quota/xfs_qm.h +++ b/fs/xfs/quota/xfs_qm.h @@ -106,7 +106,6 @@ typedef struct xfs_qm { typedef struct xfs_quotainfo { xfs_inode_t *qi_uquotaip; /* user quota inode */ xfs_inode_t *qi_gquotaip; /* group quota inode */ - spinlock_t qi_pinlock; /* dquot pinning lock */ xfs_dqlist_t qi_dqlist; /* all dquots in filesys */ int qi_dqreclaims; /* a change here indicates a removal in the dqlist */ @@ -168,7 +167,7 @@ extern void xfs_qm_destroy_quotainfo(xfs_mount_t *); extern void xfs_qm_mount_quotas(xfs_mount_t *); extern int xfs_qm_quotacheck(xfs_mount_t *); extern void xfs_qm_unmount_quotadestroy(xfs_mount_t *); -extern int xfs_qm_unmount_quotas(xfs_mount_t *); +extern void xfs_qm_unmount_quotas(xfs_mount_t *); extern int xfs_qm_write_sb_changes(xfs_mount_t *, __int64_t); extern int xfs_qm_sync(xfs_mount_t *, int); diff --git a/fs/xfs/quota/xfs_qm_bhv.c b/fs/xfs/quota/xfs_qm_bhv.c index eea2e60b456..bc6c5cca3e1 100644 --- a/fs/xfs/quota/xfs_qm_bhv.c +++ b/fs/xfs/quota/xfs_qm_bhv.c @@ -20,7 +20,6 @@ #include "xfs_bit.h" #include "xfs_log.h" #include "xfs_inum.h" -#include "xfs_clnt.h" #include "xfs_trans.h" #include "xfs_sb.h" #include "xfs_ag.h" @@ -51,7 +50,7 @@ STATIC void xfs_fill_statvfs_from_dquot( - bhv_statvfs_t *statp, + struct kstatfs *statp, xfs_disk_dquot_t *dp) { __uint64_t limit; @@ -88,7 +87,7 @@ xfs_fill_statvfs_from_dquot( STATIC void xfs_qm_statvfs( xfs_inode_t *ip, - bhv_statvfs_t *statp) + struct kstatfs *statp) { xfs_mount_t *mp = ip->i_mount; xfs_dquot_t *dqp; diff --git a/fs/xfs/quota/xfs_qm_syscalls.c b/fs/xfs/quota/xfs_qm_syscalls.c index 1a3b803dfa5..68139b38aed 100644 --- a/fs/xfs/quota/xfs_qm_syscalls.c +++ b/fs/xfs/quota/xfs_qm_syscalls.c @@ -127,7 +127,7 @@ xfs_qm_quotactl( break; case Q_XQUOTASYNC: - return (xfs_sync_inodes(mp, SYNC_DELWRI, NULL)); + return xfs_sync_inodes(mp, SYNC_DELWRI); default: break; @@ -1022,101 +1022,104 @@ xfs_qm_export_flags( /* - * Go thru all the inodes in the file system, releasing their dquots. - * Note that the mount structure gets modified to indicate that quotas are off - * AFTER this, in the case of quotaoff. This also gets called from - * xfs_rootumount. + * Release all the dquots on the inodes in an AG. */ -void -xfs_qm_dqrele_all_inodes( - struct xfs_mount *mp, - uint flags) +STATIC void +xfs_qm_dqrele_inodes_ag( + xfs_mount_t *mp, + int ag, + uint flags) { - xfs_inode_t *ip, *topino; - uint ireclaims; - struct inode *vp; - boolean_t vnode_refd; + xfs_inode_t *ip = NULL; + xfs_perag_t *pag = &mp->m_perag[ag]; + int first_index = 0; + int nr_found; - ASSERT(mp->m_quotainfo); - - XFS_MOUNT_ILOCK(mp); -again: - ip = mp->m_inodes; - if (ip == NULL) { - XFS_MOUNT_IUNLOCK(mp); - return; - } do { - /* Skip markers inserted by xfs_sync */ - if (ip->i_mount == NULL) { - ip = ip->i_mnext; - continue; + /* + * use a gang lookup to find the next inode in the tree + * as the tree is sparse and a gang lookup walks to find + * the number of objects requested. + */ + read_lock(&pag->pag_ici_lock); + nr_found = radix_tree_gang_lookup(&pag->pag_ici_root, + (void**)&ip, first_index, 1); + + if (!nr_found) { + read_unlock(&pag->pag_ici_lock); + break; } - /* Root inode, rbmip and rsumip have associated blocks */ + + /* + * Update the index for the next lookup. Catch overflows + * into the next AG range which can occur if we have inodes + * in the last block of the AG and we are currently + * pointing to the last inode. + */ + first_index = XFS_INO_TO_AGINO(mp, ip->i_ino + 1); + if (first_index < XFS_INO_TO_AGINO(mp, ip->i_ino)) { + read_unlock(&pag->pag_ici_lock); + break; + } + + /* skip quota inodes */ if (ip == XFS_QI_UQIP(mp) || ip == XFS_QI_GQIP(mp)) { ASSERT(ip->i_udquot == NULL); ASSERT(ip->i_gdquot == NULL); - ip = ip->i_mnext; + read_unlock(&pag->pag_ici_lock); continue; } - vp = VFS_I(ip); - if (!vp) { - ASSERT(ip->i_udquot == NULL); - ASSERT(ip->i_gdquot == NULL); - ip = ip->i_mnext; + + /* + * If we can't get a reference on the inode, it must be + * in reclaim. Leave it for the reclaim code to flush. + */ + if (!igrab(VFS_I(ip))) { + read_unlock(&pag->pag_ici_lock); continue; } - vnode_refd = B_FALSE; - if (xfs_ilock_nowait(ip, XFS_ILOCK_EXCL) == 0) { - ireclaims = mp->m_ireclaims; - topino = mp->m_inodes; - vp = vn_grab(vp); - if (!vp) - goto again; - - XFS_MOUNT_IUNLOCK(mp); - /* XXX restart limit ? */ - xfs_ilock(ip, XFS_ILOCK_EXCL); - vnode_refd = B_TRUE; - } else { - ireclaims = mp->m_ireclaims; - topino = mp->m_inodes; - XFS_MOUNT_IUNLOCK(mp); + read_unlock(&pag->pag_ici_lock); + + /* avoid new inodes though we shouldn't find any here */ + if (xfs_iflags_test(ip, XFS_INEW)) { + IRELE(ip); + continue; } - /* - * We don't keep the mountlock across the dqrele() call, - * since it can take a while.. - */ + xfs_ilock(ip, XFS_ILOCK_EXCL); if ((flags & XFS_UQUOTA_ACCT) && ip->i_udquot) { xfs_qm_dqrele(ip->i_udquot); ip->i_udquot = NULL; } - if (flags & (XFS_PQUOTA_ACCT|XFS_GQUOTA_ACCT) && ip->i_gdquot) { + if (flags & (XFS_PQUOTA_ACCT|XFS_GQUOTA_ACCT) && + ip->i_gdquot) { xfs_qm_dqrele(ip->i_gdquot); ip->i_gdquot = NULL; } - xfs_iunlock(ip, XFS_ILOCK_EXCL); - /* - * Wait until we've dropped the ilock and mountlock to - * do the vn_rele. Or be condemned to an eternity in the - * inactive code in hell. - */ - if (vnode_refd) - IRELE(ip); - XFS_MOUNT_ILOCK(mp); - /* - * If an inode was inserted or removed, we gotta - * start over again. - */ - if (topino != mp->m_inodes || mp->m_ireclaims != ireclaims) { - /* XXX use a sentinel */ - goto again; - } - ip = ip->i_mnext; - } while (ip != mp->m_inodes); + xfs_iput(ip, XFS_ILOCK_EXCL); + + } while (nr_found); +} + +/* + * Go thru all the inodes in the file system, releasing their dquots. + * Note that the mount structure gets modified to indicate that quotas are off + * AFTER this, in the case of quotaoff. This also gets called from + * xfs_rootumount. + */ +void +xfs_qm_dqrele_all_inodes( + struct xfs_mount *mp, + uint flags) +{ + int i; - XFS_MOUNT_IUNLOCK(mp); + ASSERT(mp->m_quotainfo); + for (i = 0; i < mp->m_sb.sb_agcount; i++) { + if (!mp->m_perag[i].pag_ici_init) + continue; + xfs_qm_dqrele_inodes_ag(mp, i, flags); + } } /*------------------------------------------------------------------------*/ diff --git a/fs/xfs/support/debug.c b/fs/xfs/support/debug.c index c27abef7b84..ae548296542 100644 --- a/fs/xfs/support/debug.c +++ b/fs/xfs/support/debug.c @@ -18,6 +18,13 @@ #include <xfs.h> #include "debug.h" +/* xfs_mount.h drags a lot of crap in, sorry.. */ +#include "xfs_sb.h" +#include "xfs_inum.h" +#include "xfs_ag.h" +#include "xfs_dmapi.h" +#include "xfs_mount.h" + static char message[1024]; /* keep it off the stack */ static DEFINE_SPINLOCK(xfs_err_lock); @@ -55,22 +62,42 @@ cmn_err(register int level, char *fmt, ...) } void -icmn_err(register int level, char *fmt, va_list ap) +xfs_fs_vcmn_err( + int level, + struct xfs_mount *mp, + char *fmt, + va_list ap) { - ulong flags; - int len; + unsigned long flags; + int len = 0; level &= XFS_ERR_MASK; - if(level > XFS_MAX_ERR_LEVEL) + if (level > XFS_MAX_ERR_LEVEL) level = XFS_MAX_ERR_LEVEL; + spin_lock_irqsave(&xfs_err_lock,flags); - len = vsnprintf(message, sizeof(message), fmt, ap); + + if (mp) { + len = sprintf(message, "Filesystem \"%s\": ", mp->m_fsname); + + /* + * Skip the printk if we can't print anything useful + * due to an over-long device name. + */ + if (len >= sizeof(message)) + goto out; + } + + len = vsnprintf(message + len, sizeof(message) - len, fmt, ap); if (len >= sizeof(message)) len = sizeof(message) - 1; if (message[len-1] == '\n') message[len-1] = 0; + printk("%s%s\n", err_level[level], message); + out: spin_unlock_irqrestore(&xfs_err_lock,flags); + BUG_ON(level == CE_PANIC); } @@ -84,5 +111,5 @@ assfail(char *expr, char *file, int line) void xfs_hex_dump(void *p, int length) { - print_hex_dump(KERN_ALERT, "", DUMP_PREFIX_OFFSET, 16, 1, p, length, 1); + print_hex_dump(KERN_ALERT, "", DUMP_PREFIX_ADDRESS, 16, 1, p, length, 1); } diff --git a/fs/xfs/support/debug.h b/fs/xfs/support/debug.h index 75845f95081..6f4fd37c67a 100644 --- a/fs/xfs/support/debug.h +++ b/fs/xfs/support/debug.h @@ -27,8 +27,6 @@ #define CE_ALERT 1 /* alert */ #define CE_PANIC 0 /* panic */ -extern void icmn_err(int, char *, va_list) - __attribute__ ((format (printf, 2, 0))); extern void cmn_err(int, char *, ...) __attribute__ ((format (printf, 2, 3))); extern void assfail(char *expr, char *f, int l); diff --git a/fs/xfs/support/ktrace.c b/fs/xfs/support/ktrace.c index a34ef05489b..2d494c26717 100644 --- a/fs/xfs/support/ktrace.c +++ b/fs/xfs/support/ktrace.c @@ -113,21 +113,16 @@ ktrace_alloc(int nentries, unsigned int __nocast sleep) void ktrace_free(ktrace_t *ktp) { - int entries_size; - if (ktp == (ktrace_t *)NULL) return; /* * Special treatment for the Vnode trace buffer. */ - if (ktp->kt_nentries == ktrace_zentries) { + if (ktp->kt_nentries == ktrace_zentries) kmem_zone_free(ktrace_ent_zone, ktp->kt_entries); - } else { - entries_size = (int)(ktp->kt_nentries * sizeof(ktrace_entry_t)); - + else kmem_free(ktp->kt_entries); - } kmem_zone_free(ktrace_hdr_zone, ktp); } diff --git a/fs/xfs/xfs.h b/fs/xfs/xfs.h index 540e4c98982..17254b529c5 100644 --- a/fs/xfs/xfs.h +++ b/fs/xfs/xfs.h @@ -30,7 +30,7 @@ #define XFS_ATTR_TRACE 1 #define XFS_BLI_TRACE 1 #define XFS_BMAP_TRACE 1 -#define XFS_BMBT_TRACE 1 +#define XFS_BTREE_TRACE 1 #define XFS_DIR2_TRACE 1 #define XFS_DQUOT_TRACE 1 #define XFS_ILOCK_TRACE 1 diff --git a/fs/xfs/xfs_acl.c b/fs/xfs/xfs_acl.c index b2f639a1416..a8cdd73999a 100644 --- a/fs/xfs/xfs_acl.c +++ b/fs/xfs/xfs_acl.c @@ -366,7 +366,7 @@ xfs_acl_allow_set( return ENOTDIR; if (vp->i_sb->s_flags & MS_RDONLY) return EROFS; - if (XFS_I(vp)->i_d.di_uid != current->fsuid && !capable(CAP_FOWNER)) + if (XFS_I(vp)->i_d.di_uid != current_fsuid() && !capable(CAP_FOWNER)) return EPERM; return 0; } @@ -413,13 +413,13 @@ xfs_acl_access( switch (fap->acl_entry[i].ae_tag) { case ACL_USER_OBJ: seen_userobj = 1; - if (fuid != current->fsuid) + if (fuid != current_fsuid()) continue; matched.ae_tag = ACL_USER_OBJ; matched.ae_perm = allows; break; case ACL_USER: - if (fap->acl_entry[i].ae_id != current->fsuid) + if (fap->acl_entry[i].ae_id != current_fsuid()) continue; matched.ae_tag = ACL_USER; matched.ae_perm = allows; @@ -758,7 +758,7 @@ xfs_acl_setmode( if (gap && nomask) iattr.ia_mode |= gap->ae_perm << 3; - return xfs_setattr(XFS_I(vp), &iattr, 0, sys_cred); + return xfs_setattr(XFS_I(vp), &iattr, 0); } /* diff --git a/fs/xfs/xfs_ag.h b/fs/xfs/xfs_ag.h index 61b292a9fb4..f2e21817a22 100644 --- a/fs/xfs/xfs_ag.h +++ b/fs/xfs/xfs_ag.h @@ -91,6 +91,8 @@ typedef struct xfs_agf { #define XFS_AGF_BLOCK(mp) XFS_HDR_BLOCK(mp, XFS_AGF_DADDR(mp)) #define XFS_BUF_TO_AGF(bp) ((xfs_agf_t *)XFS_BUF_PTR(bp)) +extern int xfs_read_agf(struct xfs_mount *mp, struct xfs_trans *tp, + xfs_agnumber_t agno, int flags, struct xfs_buf **bpp); /* * Size of the unlinked inode hash table in the agi. @@ -142,6 +144,9 @@ typedef struct xfs_agi { #define XFS_AGI_BLOCK(mp) XFS_HDR_BLOCK(mp, XFS_AGI_DADDR(mp)) #define XFS_BUF_TO_AGI(bp) ((xfs_agi_t *)XFS_BUF_PTR(bp)) +extern int xfs_read_agi(struct xfs_mount *mp, struct xfs_trans *tp, + xfs_agnumber_t agno, struct xfs_buf **bpp); + /* * The third a.g. block contains the a.g. freelist, an array * of block pointers to blocks owned by the allocation btree code. @@ -192,17 +197,23 @@ typedef struct xfs_perag xfs_agino_t pagi_freecount; /* number of free inodes */ xfs_agino_t pagi_count; /* number of allocated inodes */ int pagb_count; /* pagb slots in use */ + xfs_perag_busy_t *pagb_list; /* unstable blocks */ #ifdef __KERNEL__ spinlock_t pagb_lock; /* lock for pagb_list */ -#endif - xfs_perag_busy_t *pagb_list; /* unstable blocks */ + atomic_t pagf_fstrms; /* # of filestreams active in this AG */ int pag_ici_init; /* incore inode cache initialised */ rwlock_t pag_ici_lock; /* incore inode lock */ struct radix_tree_root pag_ici_root; /* incore inode cache root */ +#endif } xfs_perag_t; +/* + * tags for inode radix tree + */ +#define XFS_ICI_RECLAIM_TAG 0 /* inode is to be reclaimed */ + #define XFS_AG_MAXLEVELS(mp) ((mp)->m_ag_maxlevels) #define XFS_MIN_FREELIST_RAW(bl,cl,mp) \ (MIN(bl + 1, XFS_AG_MAXLEVELS(mp)) + MIN(cl + 1, XFS_AG_MAXLEVELS(mp))) diff --git a/fs/xfs/xfs_alloc.c b/fs/xfs/xfs_alloc.c index 1956f83489f..028e44e58ea 100644 --- a/fs/xfs/xfs_alloc.c +++ b/fs/xfs/xfs_alloc.c @@ -90,6 +90,92 @@ STATIC int xfs_alloc_ag_vextent_small(xfs_alloc_arg_t *, */ /* + * Lookup the record equal to [bno, len] in the btree given by cur. + */ +STATIC int /* error */ +xfs_alloc_lookup_eq( + struct xfs_btree_cur *cur, /* btree cursor */ + xfs_agblock_t bno, /* starting block of extent */ + xfs_extlen_t len, /* length of extent */ + int *stat) /* success/failure */ +{ + cur->bc_rec.a.ar_startblock = bno; + cur->bc_rec.a.ar_blockcount = len; + return xfs_btree_lookup(cur, XFS_LOOKUP_EQ, stat); +} + +/* + * Lookup the first record greater than or equal to [bno, len] + * in the btree given by cur. + */ +STATIC int /* error */ +xfs_alloc_lookup_ge( + struct xfs_btree_cur *cur, /* btree cursor */ + xfs_agblock_t bno, /* starting block of extent */ + xfs_extlen_t len, /* length of extent */ + int *stat) /* success/failure */ +{ + cur->bc_rec.a.ar_startblock = bno; + cur->bc_rec.a.ar_blockcount = len; + return xfs_btree_lookup(cur, XFS_LOOKUP_GE, stat); +} + +/* + * Lookup the first record less than or equal to [bno, len] + * in the btree given by cur. + */ +STATIC int /* error */ +xfs_alloc_lookup_le( + struct xfs_btree_cur *cur, /* btree cursor */ + xfs_agblock_t bno, /* starting block of extent */ + xfs_extlen_t len, /* length of extent */ + int *stat) /* success/failure */ +{ + cur->bc_rec.a.ar_startblock = bno; + cur->bc_rec.a.ar_blockcount = len; + return xfs_btree_lookup(cur, XFS_LOOKUP_LE, stat); +} + +/* + * Update the record referred to by cur to the value given + * by [bno, len]. + * This either works (return 0) or gets an EFSCORRUPTED error. + */ +STATIC int /* error */ +xfs_alloc_update( + struct xfs_btree_cur *cur, /* btree cursor */ + xfs_agblock_t bno, /* starting block of extent */ + xfs_extlen_t len) /* length of extent */ +{ + union xfs_btree_rec rec; + + rec.alloc.ar_startblock = cpu_to_be32(bno); + rec.alloc.ar_blockcount = cpu_to_be32(len); + return xfs_btree_update(cur, &rec); +} + +/* + * Get the data from the pointed-to record. + */ +STATIC int /* error */ +xfs_alloc_get_rec( + struct xfs_btree_cur *cur, /* btree cursor */ + xfs_agblock_t *bno, /* output: starting block of extent */ + xfs_extlen_t *len, /* output: length of extent */ + int *stat) /* output: success/failure */ +{ + union xfs_btree_rec *rec; + int error; + + error = xfs_btree_get_rec(cur, &rec, stat); + if (!error && *stat == 1) { + *bno = be32_to_cpu(rec->alloc.ar_startblock); + *len = be32_to_cpu(rec->alloc.ar_blockcount); + } + return error; +} + +/* * Compute aligned version of the found extent. * Takes alignment and min length into account. */ @@ -294,21 +380,20 @@ xfs_alloc_fixup_trees( return error; XFS_WANT_CORRUPTED_RETURN(i == 1); } + #ifdef DEBUG - { - xfs_alloc_block_t *bnoblock; - xfs_alloc_block_t *cntblock; - - if (bno_cur->bc_nlevels == 1 && - cnt_cur->bc_nlevels == 1) { - bnoblock = XFS_BUF_TO_ALLOC_BLOCK(bno_cur->bc_bufs[0]); - cntblock = XFS_BUF_TO_ALLOC_BLOCK(cnt_cur->bc_bufs[0]); - XFS_WANT_CORRUPTED_RETURN( - be16_to_cpu(bnoblock->bb_numrecs) == - be16_to_cpu(cntblock->bb_numrecs)); - } + if (bno_cur->bc_nlevels == 1 && cnt_cur->bc_nlevels == 1) { + struct xfs_btree_block *bnoblock; + struct xfs_btree_block *cntblock; + + bnoblock = XFS_BUF_TO_BLOCK(bno_cur->bc_bufs[0]); + cntblock = XFS_BUF_TO_BLOCK(cnt_cur->bc_bufs[0]); + + XFS_WANT_CORRUPTED_RETURN( + bnoblock->bb_numrecs == cntblock->bb_numrecs); } #endif + /* * Deal with all four cases: the allocated record is contained * within the freespace record, so we can have new freespace @@ -333,7 +418,7 @@ xfs_alloc_fixup_trees( /* * Delete the entry from the by-size btree. */ - if ((error = xfs_alloc_delete(cnt_cur, &i))) + if ((error = xfs_btree_delete(cnt_cur, &i))) return error; XFS_WANT_CORRUPTED_RETURN(i == 1); /* @@ -343,7 +428,7 @@ xfs_alloc_fixup_trees( if ((error = xfs_alloc_lookup_eq(cnt_cur, nfbno1, nflen1, &i))) return error; XFS_WANT_CORRUPTED_RETURN(i == 0); - if ((error = xfs_alloc_insert(cnt_cur, &i))) + if ((error = xfs_btree_insert(cnt_cur, &i))) return error; XFS_WANT_CORRUPTED_RETURN(i == 1); } @@ -351,7 +436,7 @@ xfs_alloc_fixup_trees( if ((error = xfs_alloc_lookup_eq(cnt_cur, nfbno2, nflen2, &i))) return error; XFS_WANT_CORRUPTED_RETURN(i == 0); - if ((error = xfs_alloc_insert(cnt_cur, &i))) + if ((error = xfs_btree_insert(cnt_cur, &i))) return error; XFS_WANT_CORRUPTED_RETURN(i == 1); } @@ -362,7 +447,7 @@ xfs_alloc_fixup_trees( /* * No remaining freespace, just delete the by-block tree entry. */ - if ((error = xfs_alloc_delete(bno_cur, &i))) + if ((error = xfs_btree_delete(bno_cur, &i))) return error; XFS_WANT_CORRUPTED_RETURN(i == 1); } else { @@ -379,7 +464,7 @@ xfs_alloc_fixup_trees( if ((error = xfs_alloc_lookup_eq(bno_cur, nfbno2, nflen2, &i))) return error; XFS_WANT_CORRUPTED_RETURN(i == 0); - if ((error = xfs_alloc_insert(bno_cur, &i))) + if ((error = xfs_btree_insert(bno_cur, &i))) return error; XFS_WANT_CORRUPTED_RETURN(i == 1); } @@ -640,8 +725,8 @@ xfs_alloc_ag_vextent_exact( /* * Allocate/initialize a cursor for the by-number freespace btree. */ - bno_cur = xfs_btree_init_cursor(args->mp, args->tp, args->agbp, - args->agno, XFS_BTNUM_BNO, NULL, 0); + bno_cur = xfs_allocbt_init_cursor(args->mp, args->tp, args->agbp, + args->agno, XFS_BTNUM_BNO); /* * Lookup bno and minlen in the btree (minlen is irrelevant, really). * Look for the closest free block <= bno, it must contain bno @@ -696,8 +781,8 @@ xfs_alloc_ag_vextent_exact( * We are allocating agbno for rlen [agbno .. end] * Allocate/initialize a cursor for the by-size btree. */ - cnt_cur = xfs_btree_init_cursor(args->mp, args->tp, args->agbp, - args->agno, XFS_BTNUM_CNT, NULL, 0); + cnt_cur = xfs_allocbt_init_cursor(args->mp, args->tp, args->agbp, + args->agno, XFS_BTNUM_CNT); ASSERT(args->agbno + args->len <= be32_to_cpu(XFS_BUF_TO_AGF(args->agbp)->agf_length)); if ((error = xfs_alloc_fixup_trees(cnt_cur, bno_cur, fbno, flen, @@ -759,8 +844,8 @@ xfs_alloc_ag_vextent_near( /* * Get a cursor for the by-size btree. */ - cnt_cur = xfs_btree_init_cursor(args->mp, args->tp, args->agbp, - args->agno, XFS_BTNUM_CNT, NULL, 0); + cnt_cur = xfs_allocbt_init_cursor(args->mp, args->tp, args->agbp, + args->agno, XFS_BTNUM_CNT); ltlen = 0; bno_cur_lt = bno_cur_gt = NULL; /* @@ -818,7 +903,7 @@ xfs_alloc_ag_vextent_near( XFS_WANT_CORRUPTED_GOTO(i == 1, error0); if (ltlen >= args->minlen) break; - if ((error = xfs_alloc_increment(cnt_cur, 0, &i))) + if ((error = xfs_btree_increment(cnt_cur, 0, &i))) goto error0; } while (i); ASSERT(ltlen >= args->minlen); @@ -828,7 +913,7 @@ xfs_alloc_ag_vextent_near( i = cnt_cur->bc_ptrs[0]; for (j = 1, blen = 0, bdiff = 0; !error && j && (blen < args->maxlen || bdiff > 0); - error = xfs_alloc_increment(cnt_cur, 0, &j)) { + error = xfs_btree_increment(cnt_cur, 0, &j)) { /* * For each entry, decide if it's better than * the previous best entry. @@ -886,8 +971,8 @@ xfs_alloc_ag_vextent_near( /* * Set up a cursor for the by-bno tree. */ - bno_cur_lt = xfs_btree_init_cursor(args->mp, args->tp, - args->agbp, args->agno, XFS_BTNUM_BNO, NULL, 0); + bno_cur_lt = xfs_allocbt_init_cursor(args->mp, args->tp, + args->agbp, args->agno, XFS_BTNUM_BNO); /* * Fix up the btree entries. */ @@ -914,8 +999,8 @@ xfs_alloc_ag_vextent_near( /* * Allocate and initialize the cursor for the leftward search. */ - bno_cur_lt = xfs_btree_init_cursor(args->mp, args->tp, args->agbp, - args->agno, XFS_BTNUM_BNO, NULL, 0); + bno_cur_lt = xfs_allocbt_init_cursor(args->mp, args->tp, args->agbp, + args->agno, XFS_BTNUM_BNO); /* * Lookup <= bno to find the leftward search's starting point. */ @@ -938,7 +1023,7 @@ xfs_alloc_ag_vextent_near( * Increment the cursor, so we will point at the entry just right * of the leftward entry if any, or to the leftmost entry. */ - if ((error = xfs_alloc_increment(bno_cur_gt, 0, &i))) + if ((error = xfs_btree_increment(bno_cur_gt, 0, &i))) goto error0; if (!i) { /* @@ -961,7 +1046,7 @@ xfs_alloc_ag_vextent_near( args->minlen, <bnoa, <lena); if (ltlena >= args->minlen) break; - if ((error = xfs_alloc_decrement(bno_cur_lt, 0, &i))) + if ((error = xfs_btree_decrement(bno_cur_lt, 0, &i))) goto error0; if (!i) { xfs_btree_del_cursor(bno_cur_lt, @@ -977,7 +1062,7 @@ xfs_alloc_ag_vextent_near( args->minlen, >bnoa, >lena); if (gtlena >= args->minlen) break; - if ((error = xfs_alloc_increment(bno_cur_gt, 0, &i))) + if ((error = xfs_btree_increment(bno_cur_gt, 0, &i))) goto error0; if (!i) { xfs_btree_del_cursor(bno_cur_gt, @@ -1066,7 +1151,7 @@ xfs_alloc_ag_vextent_near( /* * Fell off the right end. */ - if ((error = xfs_alloc_increment( + if ((error = xfs_btree_increment( bno_cur_gt, 0, &i))) goto error0; if (!i) { @@ -1162,7 +1247,7 @@ xfs_alloc_ag_vextent_near( /* * Fell off the left end. */ - if ((error = xfs_alloc_decrement( + if ((error = xfs_btree_decrement( bno_cur_lt, 0, &i))) goto error0; if (!i) { @@ -1267,8 +1352,8 @@ xfs_alloc_ag_vextent_size( /* * Allocate and initialize a cursor for the by-size btree. */ - cnt_cur = xfs_btree_init_cursor(args->mp, args->tp, args->agbp, - args->agno, XFS_BTNUM_CNT, NULL, 0); + cnt_cur = xfs_allocbt_init_cursor(args->mp, args->tp, args->agbp, + args->agno, XFS_BTNUM_CNT); bno_cur = NULL; /* * Look for an entry >= maxlen+alignment-1 blocks. @@ -1321,7 +1406,7 @@ xfs_alloc_ag_vextent_size( bestflen = flen; bestfbno = fbno; for (;;) { - if ((error = xfs_alloc_decrement(cnt_cur, 0, &i))) + if ((error = xfs_btree_decrement(cnt_cur, 0, &i))) goto error0; if (i == 0) break; @@ -1372,8 +1457,8 @@ xfs_alloc_ag_vextent_size( /* * Allocate and initialize a cursor for the by-block tree. */ - bno_cur = xfs_btree_init_cursor(args->mp, args->tp, args->agbp, - args->agno, XFS_BTNUM_BNO, NULL, 0); + bno_cur = xfs_allocbt_init_cursor(args->mp, args->tp, args->agbp, + args->agno, XFS_BTNUM_BNO); if ((error = xfs_alloc_fixup_trees(cnt_cur, bno_cur, fbno, flen, rbno, rlen, XFSA_FIXUP_CNT_OK))) goto error0; @@ -1416,7 +1501,7 @@ xfs_alloc_ag_vextent_small( xfs_extlen_t flen; int i; - if ((error = xfs_alloc_decrement(ccur, 0, &i))) + if ((error = xfs_btree_decrement(ccur, 0, &i))) goto error0; if (i) { if ((error = xfs_alloc_get_rec(ccur, &fbno, &flen, &i))) @@ -1515,8 +1600,7 @@ xfs_free_ag_extent( /* * Allocate and initialize a cursor for the by-block btree. */ - bno_cur = xfs_btree_init_cursor(mp, tp, agbp, agno, XFS_BTNUM_BNO, NULL, - 0); + bno_cur = xfs_allocbt_init_cursor(mp, tp, agbp, agno, XFS_BTNUM_BNO); cnt_cur = NULL; /* * Look for a neighboring block on the left (lower block numbers) @@ -1549,7 +1633,7 @@ xfs_free_ag_extent( * Look for a neighboring block on the right (higher block numbers) * that is contiguous with this space. */ - if ((error = xfs_alloc_increment(bno_cur, 0, &haveright))) + if ((error = xfs_btree_increment(bno_cur, 0, &haveright))) goto error0; if (haveright) { /* @@ -1575,8 +1659,7 @@ xfs_free_ag_extent( /* * Now allocate and initialize a cursor for the by-size tree. */ - cnt_cur = xfs_btree_init_cursor(mp, tp, agbp, agno, XFS_BTNUM_CNT, NULL, - 0); + cnt_cur = xfs_allocbt_init_cursor(mp, tp, agbp, agno, XFS_BTNUM_CNT); /* * Have both left and right contiguous neighbors. * Merge all three into a single free block. @@ -1588,7 +1671,7 @@ xfs_free_ag_extent( if ((error = xfs_alloc_lookup_eq(cnt_cur, ltbno, ltlen, &i))) goto error0; XFS_WANT_CORRUPTED_GOTO(i == 1, error0); - if ((error = xfs_alloc_delete(cnt_cur, &i))) + if ((error = xfs_btree_delete(cnt_cur, &i))) goto error0; XFS_WANT_CORRUPTED_GOTO(i == 1, error0); /* @@ -1597,19 +1680,19 @@ xfs_free_ag_extent( if ((error = xfs_alloc_lookup_eq(cnt_cur, gtbno, gtlen, &i))) goto error0; XFS_WANT_CORRUPTED_GOTO(i == 1, error0); - if ((error = xfs_alloc_delete(cnt_cur, &i))) + if ((error = xfs_btree_delete(cnt_cur, &i))) goto error0; XFS_WANT_CORRUPTED_GOTO(i == 1, error0); /* * Delete the old by-block entry for the right block. */ - if ((error = xfs_alloc_delete(bno_cur, &i))) + if ((error = xfs_btree_delete(bno_cur, &i))) goto error0; XFS_WANT_CORRUPTED_GOTO(i == 1, error0); /* * Move the by-block cursor back to the left neighbor. */ - if ((error = xfs_alloc_decrement(bno_cur, 0, &i))) + if ((error = xfs_btree_decrement(bno_cur, 0, &i))) goto error0; XFS_WANT_CORRUPTED_GOTO(i == 1, error0); #ifdef DEBUG @@ -1648,14 +1731,14 @@ xfs_free_ag_extent( if ((error = xfs_alloc_lookup_eq(cnt_cur, ltbno, ltlen, &i))) goto error0; XFS_WANT_CORRUPTED_GOTO(i == 1, error0); - if ((error = xfs_alloc_delete(cnt_cur, &i))) + if ((error = xfs_btree_delete(cnt_cur, &i))) goto error0; XFS_WANT_CORRUPTED_GOTO(i == 1, error0); /* * Back up the by-block cursor to the left neighbor, and * update its length. */ - if ((error = xfs_alloc_decrement(bno_cur, 0, &i))) + if ((error = xfs_btree_decrement(bno_cur, 0, &i))) goto error0; XFS_WANT_CORRUPTED_GOTO(i == 1, error0); nbno = ltbno; @@ -1674,7 +1757,7 @@ xfs_free_ag_extent( if ((error = xfs_alloc_lookup_eq(cnt_cur, gtbno, gtlen, &i))) goto error0; XFS_WANT_CORRUPTED_GOTO(i == 1, error0); - if ((error = xfs_alloc_delete(cnt_cur, &i))) + if ((error = xfs_btree_delete(cnt_cur, &i))) goto error0; XFS_WANT_CORRUPTED_GOTO(i == 1, error0); /* @@ -1693,7 +1776,7 @@ xfs_free_ag_extent( else { nbno = bno; nlen = len; - if ((error = xfs_alloc_insert(bno_cur, &i))) + if ((error = xfs_btree_insert(bno_cur, &i))) goto error0; XFS_WANT_CORRUPTED_GOTO(i == 1, error0); } @@ -1705,7 +1788,7 @@ xfs_free_ag_extent( if ((error = xfs_alloc_lookup_eq(cnt_cur, nbno, nlen, &i))) goto error0; XFS_WANT_CORRUPTED_GOTO(i == 0, error0); - if ((error = xfs_alloc_insert(cnt_cur, &i))) + if ((error = xfs_btree_insert(cnt_cur, &i))) goto error0; XFS_WANT_CORRUPTED_GOTO(i == 1, error0); xfs_btree_del_cursor(cnt_cur, XFS_BTREE_NOERROR); @@ -2150,51 +2233,83 @@ xfs_alloc_put_freelist( * Read in the allocation group header (free/alloc section). */ int /* error */ -xfs_alloc_read_agf( - xfs_mount_t *mp, /* mount point structure */ - xfs_trans_t *tp, /* transaction pointer */ - xfs_agnumber_t agno, /* allocation group number */ - int flags, /* XFS_ALLOC_FLAG_... */ - xfs_buf_t **bpp) /* buffer for the ag freelist header */ +xfs_read_agf( + struct xfs_mount *mp, /* mount point structure */ + struct xfs_trans *tp, /* transaction pointer */ + xfs_agnumber_t agno, /* allocation group number */ + int flags, /* XFS_BUF_ */ + struct xfs_buf **bpp) /* buffer for the ag freelist header */ { - xfs_agf_t *agf; /* ag freelist header */ + struct xfs_agf *agf; /* ag freelist header */ int agf_ok; /* set if agf is consistent */ - xfs_buf_t *bp; /* return value */ - xfs_perag_t *pag; /* per allocation group data */ int error; ASSERT(agno != NULLAGNUMBER); error = xfs_trans_read_buf( mp, tp, mp->m_ddev_targp, XFS_AG_DADDR(mp, agno, XFS_AGF_DADDR(mp)), - XFS_FSS_TO_BB(mp, 1), - (flags & XFS_ALLOC_FLAG_TRYLOCK) ? XFS_BUF_TRYLOCK : 0U, - &bp); + XFS_FSS_TO_BB(mp, 1), flags, bpp); if (error) return error; - ASSERT(!bp || !XFS_BUF_GETERROR(bp)); - if (!bp) { - *bpp = NULL; + if (!*bpp) return 0; - } + + ASSERT(!XFS_BUF_GETERROR(*bpp)); + agf = XFS_BUF_TO_AGF(*bpp); + /* * Validate the magic number of the agf block. */ - agf = XFS_BUF_TO_AGF(bp); agf_ok = be32_to_cpu(agf->agf_magicnum) == XFS_AGF_MAGIC && XFS_AGF_GOOD_VERSION(be32_to_cpu(agf->agf_versionnum)) && be32_to_cpu(agf->agf_freeblks) <= be32_to_cpu(agf->agf_length) && be32_to_cpu(agf->agf_flfirst) < XFS_AGFL_SIZE(mp) && be32_to_cpu(agf->agf_fllast) < XFS_AGFL_SIZE(mp) && - be32_to_cpu(agf->agf_flcount) <= XFS_AGFL_SIZE(mp); + be32_to_cpu(agf->agf_flcount) <= XFS_AGFL_SIZE(mp) && + be32_to_cpu(agf->agf_seqno) == agno; + if (xfs_sb_version_haslazysbcount(&mp->m_sb)) + agf_ok = agf_ok && be32_to_cpu(agf->agf_btreeblks) <= + be32_to_cpu(agf->agf_length); if (unlikely(XFS_TEST_ERROR(!agf_ok, mp, XFS_ERRTAG_ALLOC_READ_AGF, XFS_RANDOM_ALLOC_READ_AGF))) { XFS_CORRUPTION_ERROR("xfs_alloc_read_agf", XFS_ERRLEVEL_LOW, mp, agf); - xfs_trans_brelse(tp, bp); + xfs_trans_brelse(tp, *bpp); return XFS_ERROR(EFSCORRUPTED); } + + XFS_BUF_SET_VTYPE_REF(*bpp, B_FS_AGF, XFS_AGF_REF); + return 0; +} + +/* + * Read in the allocation group header (free/alloc section). + */ +int /* error */ +xfs_alloc_read_agf( + struct xfs_mount *mp, /* mount point structure */ + struct xfs_trans *tp, /* transaction pointer */ + xfs_agnumber_t agno, /* allocation group number */ + int flags, /* XFS_ALLOC_FLAG_... */ + struct xfs_buf **bpp) /* buffer for the ag freelist header */ +{ + struct xfs_agf *agf; /* ag freelist header */ + struct xfs_perag *pag; /* per allocation group data */ + int error; + + ASSERT(agno != NULLAGNUMBER); + + error = xfs_read_agf(mp, tp, agno, + (flags & XFS_ALLOC_FLAG_TRYLOCK) ? XFS_BUF_TRYLOCK : 0, + bpp); + if (error) + return error; + if (!*bpp) + return 0; + ASSERT(!XFS_BUF_GETERROR(*bpp)); + + agf = XFS_BUF_TO_AGF(*bpp); pag = &mp->m_perag[agno]; if (!pag->pagf_init) { pag->pagf_freeblks = be32_to_cpu(agf->agf_freeblks); @@ -2213,6 +2328,7 @@ xfs_alloc_read_agf( #ifdef DEBUG else if (!XFS_FORCED_SHUTDOWN(mp)) { ASSERT(pag->pagf_freeblks == be32_to_cpu(agf->agf_freeblks)); + ASSERT(pag->pagf_btreeblks == be32_to_cpu(agf->agf_btreeblks)); ASSERT(pag->pagf_flcount == be32_to_cpu(agf->agf_flcount)); ASSERT(pag->pagf_longest == be32_to_cpu(agf->agf_longest)); ASSERT(pag->pagf_levels[XFS_BTNUM_BNOi] == @@ -2221,8 +2337,6 @@ xfs_alloc_read_agf( be32_to_cpu(agf->agf_levels[XFS_BTNUM_CNTi])); } #endif - XFS_BUF_SET_VTYPE_REF(bp, B_FS_AGF, XFS_AGF_REF); - *bpp = bp; return 0; } diff --git a/fs/xfs/xfs_alloc.h b/fs/xfs/xfs_alloc.h index 5aec15d0651..588172796f7 100644 --- a/fs/xfs/xfs_alloc.h +++ b/fs/xfs/xfs_alloc.h @@ -121,6 +121,19 @@ extern ktrace_t *xfs_alloc_trace_buf; #define XFS_ALLOC_KTRACE_BUSYSEARCH 6 #endif +void +xfs_alloc_mark_busy(xfs_trans_t *tp, + xfs_agnumber_t agno, + xfs_agblock_t bno, + xfs_extlen_t len); + +void +xfs_alloc_clear_busy(xfs_trans_t *tp, + xfs_agnumber_t ag, + int idx); + +#endif /* __KERNEL__ */ + /* * Compute and fill in value of m_ag_maxlevels. */ @@ -196,18 +209,4 @@ xfs_free_extent( xfs_fsblock_t bno, /* starting block number of extent */ xfs_extlen_t len); /* length of extent */ -void -xfs_alloc_mark_busy(xfs_trans_t *tp, - xfs_agnumber_t agno, - xfs_agblock_t bno, - xfs_extlen_t len); - -void -xfs_alloc_clear_busy(xfs_trans_t *tp, - xfs_agnumber_t ag, - int idx); - - -#endif /* __KERNEL__ */ - #endif /* __XFS_ALLOC_H__ */ diff --git a/fs/xfs/xfs_alloc_btree.c b/fs/xfs/xfs_alloc_btree.c index 3ce2645508a..733cb75a8c5 100644 --- a/fs/xfs/xfs_alloc_btree.c +++ b/fs/xfs/xfs_alloc_btree.c @@ -35,2177 +35,464 @@ #include "xfs_dinode.h" #include "xfs_inode.h" #include "xfs_btree.h" +#include "xfs_btree_trace.h" #include "xfs_ialloc.h" #include "xfs_alloc.h" #include "xfs_error.h" -/* - * Prototypes for internal functions. - */ -STATIC void xfs_alloc_log_block(xfs_trans_t *, xfs_buf_t *, int); -STATIC void xfs_alloc_log_keys(xfs_btree_cur_t *, xfs_buf_t *, int, int); -STATIC void xfs_alloc_log_ptrs(xfs_btree_cur_t *, xfs_buf_t *, int, int); -STATIC void xfs_alloc_log_recs(xfs_btree_cur_t *, xfs_buf_t *, int, int); -STATIC int xfs_alloc_lshift(xfs_btree_cur_t *, int, int *); -STATIC int xfs_alloc_newroot(xfs_btree_cur_t *, int *); -STATIC int xfs_alloc_rshift(xfs_btree_cur_t *, int, int *); -STATIC int xfs_alloc_split(xfs_btree_cur_t *, int, xfs_agblock_t *, - xfs_alloc_key_t *, xfs_btree_cur_t **, int *); -STATIC int xfs_alloc_updkey(xfs_btree_cur_t *, xfs_alloc_key_t *, int); +STATIC struct xfs_btree_cur * +xfs_allocbt_dup_cursor( + struct xfs_btree_cur *cur) +{ + return xfs_allocbt_init_cursor(cur->bc_mp, cur->bc_tp, + cur->bc_private.a.agbp, cur->bc_private.a.agno, + cur->bc_btnum); +} -/* - * Internal functions. - */ +STATIC void +xfs_allocbt_set_root( + struct xfs_btree_cur *cur, + union xfs_btree_ptr *ptr, + int inc) +{ + struct xfs_buf *agbp = cur->bc_private.a.agbp; + struct xfs_agf *agf = XFS_BUF_TO_AGF(agbp); + xfs_agnumber_t seqno = be32_to_cpu(agf->agf_seqno); + int btnum = cur->bc_btnum; -/* - * Single level of the xfs_alloc_delete record deletion routine. - * Delete record pointed to by cur/level. - * Remove the record from its block then rebalance the tree. - * Return 0 for error, 1 for done, 2 to go on to the next level. - */ -STATIC int /* error */ -xfs_alloc_delrec( - xfs_btree_cur_t *cur, /* btree cursor */ - int level, /* level removing record from */ - int *stat) /* fail/done/go-on */ + ASSERT(ptr->s != 0); + + agf->agf_roots[btnum] = ptr->s; + be32_add_cpu(&agf->agf_levels[btnum], inc); + cur->bc_mp->m_perag[seqno].pagf_levels[btnum] += inc; + + xfs_alloc_log_agf(cur->bc_tp, agbp, XFS_AGF_ROOTS | XFS_AGF_LEVELS); +} + +STATIC int +xfs_allocbt_alloc_block( + struct xfs_btree_cur *cur, + union xfs_btree_ptr *start, + union xfs_btree_ptr *new, + int length, + int *stat) { - xfs_agf_t *agf; /* allocation group freelist header */ - xfs_alloc_block_t *block; /* btree block record/key lives in */ - xfs_agblock_t bno; /* btree block number */ - xfs_buf_t *bp; /* buffer for block */ - int error; /* error return value */ - int i; /* loop index */ - xfs_alloc_key_t key; /* kp points here if block is level 0 */ - xfs_agblock_t lbno; /* left block's block number */ - xfs_buf_t *lbp; /* left block's buffer pointer */ - xfs_alloc_block_t *left; /* left btree block */ - xfs_alloc_key_t *lkp=NULL; /* left block key pointer */ - xfs_alloc_ptr_t *lpp=NULL; /* left block address pointer */ - int lrecs=0; /* number of records in left block */ - xfs_alloc_rec_t *lrp; /* left block record pointer */ - xfs_mount_t *mp; /* mount structure */ - int ptr; /* index in btree block for this rec */ - xfs_agblock_t rbno; /* right block's block number */ - xfs_buf_t *rbp; /* right block's buffer pointer */ - xfs_alloc_block_t *right; /* right btree block */ - xfs_alloc_key_t *rkp; /* right block key pointer */ - xfs_alloc_ptr_t *rpp; /* right block address pointer */ - int rrecs=0; /* number of records in right block */ - int numrecs; - xfs_alloc_rec_t *rrp; /* right block record pointer */ - xfs_btree_cur_t *tcur; /* temporary btree cursor */ + int error; + xfs_agblock_t bno; - /* - * Get the index of the entry being deleted, check for nothing there. - */ - ptr = cur->bc_ptrs[level]; - if (ptr == 0) { - *stat = 0; - return 0; - } - /* - * Get the buffer & block containing the record or key/ptr. - */ - bp = cur->bc_bufs[level]; - block = XFS_BUF_TO_ALLOC_BLOCK(bp); -#ifdef DEBUG - if ((error = xfs_btree_check_sblock(cur, block, level, bp))) + XFS_BTREE_TRACE_CURSOR(cur, XBT_ENTRY); + + /* Allocate the new block from the freelist. If we can't, give up. */ + error = xfs_alloc_get_freelist(cur->bc_tp, cur->bc_private.a.agbp, + &bno, 1); + if (error) { + XFS_BTREE_TRACE_CURSOR(cur, XBT_ERROR); return error; -#endif - /* - * Fail if we're off the end of the block. - */ - numrecs = be16_to_cpu(block->bb_numrecs); - if (ptr > numrecs) { + } + + if (bno == NULLAGBLOCK) { + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); *stat = 0; return 0; } - XFS_STATS_INC(xs_abt_delrec); - /* - * It's a nonleaf. Excise the key and ptr being deleted, by - * sliding the entries past them down one. - * Log the changed areas of the block. - */ - if (level > 0) { - lkp = XFS_ALLOC_KEY_ADDR(block, 1, cur); - lpp = XFS_ALLOC_PTR_ADDR(block, 1, cur); -#ifdef DEBUG - for (i = ptr; i < numrecs; i++) { - if ((error = xfs_btree_check_sptr(cur, be32_to_cpu(lpp[i]), level))) - return error; - } -#endif - if (ptr < numrecs) { - memmove(&lkp[ptr - 1], &lkp[ptr], - (numrecs - ptr) * sizeof(*lkp)); - memmove(&lpp[ptr - 1], &lpp[ptr], - (numrecs - ptr) * sizeof(*lpp)); - xfs_alloc_log_ptrs(cur, bp, ptr, numrecs - 1); - xfs_alloc_log_keys(cur, bp, ptr, numrecs - 1); - } - } - /* - * It's a leaf. Excise the record being deleted, by sliding the - * entries past it down one. Log the changed areas of the block. - */ - else { - lrp = XFS_ALLOC_REC_ADDR(block, 1, cur); - if (ptr < numrecs) { - memmove(&lrp[ptr - 1], &lrp[ptr], - (numrecs - ptr) * sizeof(*lrp)); - xfs_alloc_log_recs(cur, bp, ptr, numrecs - 1); - } - /* - * If it's the first record in the block, we'll need a key - * structure to pass up to the next level (updkey). - */ - if (ptr == 1) { - key.ar_startblock = lrp->ar_startblock; - key.ar_blockcount = lrp->ar_blockcount; - lkp = &key; - } - } - /* - * Decrement and log the number of entries in the block. - */ - numrecs--; - block->bb_numrecs = cpu_to_be16(numrecs); - xfs_alloc_log_block(cur->bc_tp, bp, XFS_BB_NUMRECS); - /* - * See if the longest free extent in the allocation group was - * changed by this operation. True if it's the by-size btree, and - * this is the leaf level, and there is no right sibling block, - * and this was the last record. - */ - agf = XFS_BUF_TO_AGF(cur->bc_private.a.agbp); - mp = cur->bc_mp; - if (level == 0 && - cur->bc_btnum == XFS_BTNUM_CNT && - be32_to_cpu(block->bb_rightsib) == NULLAGBLOCK && - ptr > numrecs) { - ASSERT(ptr == numrecs + 1); - /* - * There are still records in the block. Grab the size - * from the last one. - */ - if (numrecs) { - rrp = XFS_ALLOC_REC_ADDR(block, numrecs, cur); - agf->agf_longest = rrp->ar_blockcount; - } - /* - * No free extents left. - */ - else - agf->agf_longest = 0; - mp->m_perag[be32_to_cpu(agf->agf_seqno)].pagf_longest = - be32_to_cpu(agf->agf_longest); - xfs_alloc_log_agf(cur->bc_tp, cur->bc_private.a.agbp, - XFS_AGF_LONGEST); - } - /* - * Is this the root level? If so, we're almost done. - */ - if (level == cur->bc_nlevels - 1) { - /* - * If this is the root level, - * and there's only one entry left, - * and it's NOT the leaf level, - * then we can get rid of this level. - */ - if (numrecs == 1 && level > 0) { - /* - * lpp is still set to the first pointer in the block. - * Make it the new root of the btree. - */ - bno = be32_to_cpu(agf->agf_roots[cur->bc_btnum]); - agf->agf_roots[cur->bc_btnum] = *lpp; - be32_add_cpu(&agf->agf_levels[cur->bc_btnum], -1); - mp->m_perag[be32_to_cpu(agf->agf_seqno)].pagf_levels[cur->bc_btnum]--; - /* - * Put this buffer/block on the ag's freelist. - */ - error = xfs_alloc_put_freelist(cur->bc_tp, - cur->bc_private.a.agbp, NULL, bno, 1); - if (error) - return error; - /* - * Since blocks move to the free list without the - * coordination used in xfs_bmap_finish, we can't allow - * block to be available for reallocation and - * non-transaction writing (user data) until we know - * that the transaction that moved it to the free list - * is permanently on disk. We track the blocks by - * declaring these blocks as "busy"; the busy list is - * maintained on a per-ag basis and each transaction - * records which entries should be removed when the - * iclog commits to disk. If a busy block is - * allocated, the iclog is pushed up to the LSN - * that freed the block. - */ - xfs_alloc_mark_busy(cur->bc_tp, - be32_to_cpu(agf->agf_seqno), bno, 1); + xfs_trans_agbtree_delta(cur->bc_tp, 1); + new->s = cpu_to_be32(bno); - xfs_trans_agbtree_delta(cur->bc_tp, -1); - xfs_alloc_log_agf(cur->bc_tp, cur->bc_private.a.agbp, - XFS_AGF_ROOTS | XFS_AGF_LEVELS); - /* - * Update the cursor so there's one fewer level. - */ - xfs_btree_setbuf(cur, level, NULL); - cur->bc_nlevels--; - } else if (level > 0 && - (error = xfs_alloc_decrement(cur, level, &i))) - return error; - *stat = 1; - return 0; - } - /* - * If we deleted the leftmost entry in the block, update the - * key values above us in the tree. - */ - if (ptr == 1 && (error = xfs_alloc_updkey(cur, lkp, level + 1))) - return error; - /* - * If the number of records remaining in the block is at least - * the minimum, we're done. - */ - if (numrecs >= XFS_ALLOC_BLOCK_MINRECS(level, cur)) { - if (level > 0 && (error = xfs_alloc_decrement(cur, level, &i))) - return error; - *stat = 1; - return 0; - } - /* - * Otherwise, we have to move some records around to keep the - * tree balanced. Look at the left and right sibling blocks to - * see if we can re-balance by moving only one record. - */ - rbno = be32_to_cpu(block->bb_rightsib); - lbno = be32_to_cpu(block->bb_leftsib); - bno = NULLAGBLOCK; - ASSERT(rbno != NULLAGBLOCK || lbno != NULLAGBLOCK); - /* - * Duplicate the cursor so our btree manipulations here won't - * disrupt the next level up. - */ - if ((error = xfs_btree_dup_cursor(cur, &tcur))) - return error; - /* - * If there's a right sibling, see if it's ok to shift an entry - * out of it. - */ - if (rbno != NULLAGBLOCK) { - /* - * Move the temp cursor to the last entry in the next block. - * Actually any entry but the first would suffice. - */ - i = xfs_btree_lastrec(tcur, level); - XFS_WANT_CORRUPTED_GOTO(i == 1, error0); - if ((error = xfs_alloc_increment(tcur, level, &i))) - goto error0; - XFS_WANT_CORRUPTED_GOTO(i == 1, error0); - i = xfs_btree_lastrec(tcur, level); - XFS_WANT_CORRUPTED_GOTO(i == 1, error0); - /* - * Grab a pointer to the block. - */ - rbp = tcur->bc_bufs[level]; - right = XFS_BUF_TO_ALLOC_BLOCK(rbp); -#ifdef DEBUG - if ((error = xfs_btree_check_sblock(cur, right, level, rbp))) - goto error0; -#endif - /* - * Grab the current block number, for future use. - */ - bno = be32_to_cpu(right->bb_leftsib); - /* - * If right block is full enough so that removing one entry - * won't make it too empty, and left-shifting an entry out - * of right to us works, we're done. - */ - if (be16_to_cpu(right->bb_numrecs) - 1 >= - XFS_ALLOC_BLOCK_MINRECS(level, cur)) { - if ((error = xfs_alloc_lshift(tcur, level, &i))) - goto error0; - if (i) { - ASSERT(be16_to_cpu(block->bb_numrecs) >= - XFS_ALLOC_BLOCK_MINRECS(level, cur)); - xfs_btree_del_cursor(tcur, - XFS_BTREE_NOERROR); - if (level > 0 && - (error = xfs_alloc_decrement(cur, level, - &i))) - return error; - *stat = 1; - return 0; - } - } - /* - * Otherwise, grab the number of records in right for - * future reference, and fix up the temp cursor to point - * to our block again (last record). - */ - rrecs = be16_to_cpu(right->bb_numrecs); - if (lbno != NULLAGBLOCK) { - i = xfs_btree_firstrec(tcur, level); - XFS_WANT_CORRUPTED_GOTO(i == 1, error0); - if ((error = xfs_alloc_decrement(tcur, level, &i))) - goto error0; - XFS_WANT_CORRUPTED_GOTO(i == 1, error0); - } - } - /* - * If there's a left sibling, see if it's ok to shift an entry - * out of it. - */ - if (lbno != NULLAGBLOCK) { - /* - * Move the temp cursor to the first entry in the - * previous block. - */ - i = xfs_btree_firstrec(tcur, level); - XFS_WANT_CORRUPTED_GOTO(i == 1, error0); - if ((error = xfs_alloc_decrement(tcur, level, &i))) - goto error0; - XFS_WANT_CORRUPTED_GOTO(i == 1, error0); - xfs_btree_firstrec(tcur, level); - /* - * Grab a pointer to the block. - */ - lbp = tcur->bc_bufs[level]; - left = XFS_BUF_TO_ALLOC_BLOCK(lbp); -#ifdef DEBUG - if ((error = xfs_btree_check_sblock(cur, left, level, lbp))) - goto error0; -#endif - /* - * Grab the current block number, for future use. - */ - bno = be32_to_cpu(left->bb_rightsib); - /* - * If left block is full enough so that removing one entry - * won't make it too empty, and right-shifting an entry out - * of left to us works, we're done. - */ - if (be16_to_cpu(left->bb_numrecs) - 1 >= - XFS_ALLOC_BLOCK_MINRECS(level, cur)) { - if ((error = xfs_alloc_rshift(tcur, level, &i))) - goto error0; - if (i) { - ASSERT(be16_to_cpu(block->bb_numrecs) >= - XFS_ALLOC_BLOCK_MINRECS(level, cur)); - xfs_btree_del_cursor(tcur, - XFS_BTREE_NOERROR); - if (level == 0) - cur->bc_ptrs[0]++; - *stat = 1; - return 0; - } - } - /* - * Otherwise, grab the number of records in right for - * future reference. - */ - lrecs = be16_to_cpu(left->bb_numrecs); - } - /* - * Delete the temp cursor, we're done with it. - */ - xfs_btree_del_cursor(tcur, XFS_BTREE_NOERROR); - /* - * If here, we need to do a join to keep the tree balanced. - */ - ASSERT(bno != NULLAGBLOCK); - /* - * See if we can join with the left neighbor block. - */ - if (lbno != NULLAGBLOCK && - lrecs + numrecs <= XFS_ALLOC_BLOCK_MAXRECS(level, cur)) { - /* - * Set "right" to be the starting block, - * "left" to be the left neighbor. - */ - rbno = bno; - right = block; - rrecs = be16_to_cpu(right->bb_numrecs); - rbp = bp; - if ((error = xfs_btree_read_bufs(mp, cur->bc_tp, - cur->bc_private.a.agno, lbno, 0, &lbp, - XFS_ALLOC_BTREE_REF))) - return error; - left = XFS_BUF_TO_ALLOC_BLOCK(lbp); - lrecs = be16_to_cpu(left->bb_numrecs); - if ((error = xfs_btree_check_sblock(cur, left, level, lbp))) - return error; - } - /* - * If that won't work, see if we can join with the right neighbor block. - */ - else if (rbno != NULLAGBLOCK && - rrecs + numrecs <= XFS_ALLOC_BLOCK_MAXRECS(level, cur)) { - /* - * Set "left" to be the starting block, - * "right" to be the right neighbor. - */ - lbno = bno; - left = block; - lrecs = be16_to_cpu(left->bb_numrecs); - lbp = bp; - if ((error = xfs_btree_read_bufs(mp, cur->bc_tp, - cur->bc_private.a.agno, rbno, 0, &rbp, - XFS_ALLOC_BTREE_REF))) - return error; - right = XFS_BUF_TO_ALLOC_BLOCK(rbp); - rrecs = be16_to_cpu(right->bb_numrecs); - if ((error = xfs_btree_check_sblock(cur, right, level, rbp))) - return error; - } - /* - * Otherwise, we can't fix the imbalance. - * Just return. This is probably a logic error, but it's not fatal. - */ - else { - if (level > 0 && (error = xfs_alloc_decrement(cur, level, &i))) - return error; - *stat = 1; - return 0; - } - /* - * We're now going to join "left" and "right" by moving all the stuff - * in "right" to "left" and deleting "right". - */ - if (level > 0) { - /* - * It's a non-leaf. Move keys and pointers. - */ - lkp = XFS_ALLOC_KEY_ADDR(left, lrecs + 1, cur); - lpp = XFS_ALLOC_PTR_ADDR(left, lrecs + 1, cur); - rkp = XFS_ALLOC_KEY_ADDR(right, 1, cur); - rpp = XFS_ALLOC_PTR_ADDR(right, 1, cur); -#ifdef DEBUG - for (i = 0; i < rrecs; i++) { - if ((error = xfs_btree_check_sptr(cur, be32_to_cpu(rpp[i]), level))) - return error; - } -#endif - memcpy(lkp, rkp, rrecs * sizeof(*lkp)); - memcpy(lpp, rpp, rrecs * sizeof(*lpp)); - xfs_alloc_log_keys(cur, lbp, lrecs + 1, lrecs + rrecs); - xfs_alloc_log_ptrs(cur, lbp, lrecs + 1, lrecs + rrecs); - } else { - /* - * It's a leaf. Move records. - */ - lrp = XFS_ALLOC_REC_ADDR(left, lrecs + 1, cur); - rrp = XFS_ALLOC_REC_ADDR(right, 1, cur); - memcpy(lrp, rrp, rrecs * sizeof(*lrp)); - xfs_alloc_log_recs(cur, lbp, lrecs + 1, lrecs + rrecs); - } - /* - * If we joined with the left neighbor, set the buffer in the - * cursor to the left block, and fix up the index. - */ - if (bp != lbp) { - xfs_btree_setbuf(cur, level, lbp); - cur->bc_ptrs[level] += lrecs; - } - /* - * If we joined with the right neighbor and there's a level above - * us, increment the cursor at that level. - */ - else if (level + 1 < cur->bc_nlevels && - (error = xfs_alloc_increment(cur, level + 1, &i))) - return error; - /* - * Fix up the number of records in the surviving block. - */ - lrecs += rrecs; - left->bb_numrecs = cpu_to_be16(lrecs); - /* - * Fix up the right block pointer in the surviving block, and log it. - */ - left->bb_rightsib = right->bb_rightsib; - xfs_alloc_log_block(cur->bc_tp, lbp, XFS_BB_NUMRECS | XFS_BB_RIGHTSIB); - /* - * If there is a right sibling now, make it point to the - * remaining block. - */ - if (be32_to_cpu(left->bb_rightsib) != NULLAGBLOCK) { - xfs_alloc_block_t *rrblock; - xfs_buf_t *rrbp; + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); + *stat = 1; + return 0; +} - if ((error = xfs_btree_read_bufs(mp, cur->bc_tp, - cur->bc_private.a.agno, be32_to_cpu(left->bb_rightsib), 0, - &rrbp, XFS_ALLOC_BTREE_REF))) - return error; - rrblock = XFS_BUF_TO_ALLOC_BLOCK(rrbp); - if ((error = xfs_btree_check_sblock(cur, rrblock, level, rrbp))) - return error; - rrblock->bb_leftsib = cpu_to_be32(lbno); - xfs_alloc_log_block(cur->bc_tp, rrbp, XFS_BB_LEFTSIB); - } - /* - * Free the deleting block by putting it on the freelist. - */ - error = xfs_alloc_put_freelist(cur->bc_tp, - cur->bc_private.a.agbp, NULL, rbno, 1); +STATIC int +xfs_allocbt_free_block( + struct xfs_btree_cur *cur, + struct xfs_buf *bp) +{ + struct xfs_buf *agbp = cur->bc_private.a.agbp; + struct xfs_agf *agf = XFS_BUF_TO_AGF(agbp); + xfs_agblock_t bno; + int error; + + bno = XFS_DADDR_TO_AGBNO(cur->bc_mp, XFS_BUF_ADDR(bp)); + error = xfs_alloc_put_freelist(cur->bc_tp, agbp, NULL, bno, 1); if (error) return error; + /* - * Since blocks move to the free list without the coordination - * used in xfs_bmap_finish, we can't allow block to be available - * for reallocation and non-transaction writing (user data) - * until we know that the transaction that moved it to the free - * list is permanently on disk. We track the blocks by declaring - * these blocks as "busy"; the busy list is maintained on a - * per-ag basis and each transaction records which entries - * should be removed when the iclog commits to disk. If a - * busy block is allocated, the iclog is pushed up to the + * Since blocks move to the free list without the coordination used in + * xfs_bmap_finish, we can't allow block to be available for + * reallocation and non-transaction writing (user data) until we know + * that the transaction that moved it to the free list is permanently + * on disk. We track the blocks by declaring these blocks as "busy"; + * the busy list is maintained on a per-ag basis and each transaction + * records which entries should be removed when the iclog commits to + * disk. If a busy block is allocated, the iclog is pushed up to the * LSN that freed the block. */ xfs_alloc_mark_busy(cur->bc_tp, be32_to_cpu(agf->agf_seqno), bno, 1); xfs_trans_agbtree_delta(cur->bc_tp, -1); - - /* - * Adjust the current level's cursor so that we're left referring - * to the right node, after we're done. - * If this leaves the ptr value 0 our caller will fix it up. - */ - if (level > 0) - cur->bc_ptrs[level]--; - /* - * Return value means the next level up has something to do. - */ - *stat = 2; return 0; - -error0: - xfs_btree_del_cursor(tcur, XFS_BTREE_ERROR); - return error; } /* - * Insert one record/level. Return information to the caller - * allowing the next level up to proceed if necessary. + * Update the longest extent in the AGF */ -STATIC int /* error */ -xfs_alloc_insrec( - xfs_btree_cur_t *cur, /* btree cursor */ - int level, /* level to insert record at */ - xfs_agblock_t *bnop, /* i/o: block number inserted */ - xfs_alloc_rec_t *recp, /* i/o: record data inserted */ - xfs_btree_cur_t **curp, /* output: new cursor replacing cur */ - int *stat) /* output: success/failure */ +STATIC void +xfs_allocbt_update_lastrec( + struct xfs_btree_cur *cur, + struct xfs_btree_block *block, + union xfs_btree_rec *rec, + int ptr, + int reason) { - xfs_agf_t *agf; /* allocation group freelist header */ - xfs_alloc_block_t *block; /* btree block record/key lives in */ - xfs_buf_t *bp; /* buffer for block */ - int error; /* error return value */ - int i; /* loop index */ - xfs_alloc_key_t key; /* key value being inserted */ - xfs_alloc_key_t *kp; /* pointer to btree keys */ - xfs_agblock_t nbno; /* block number of allocated block */ - xfs_btree_cur_t *ncur; /* new cursor to be used at next lvl */ - xfs_alloc_key_t nkey; /* new key value, from split */ - xfs_alloc_rec_t nrec; /* new record value, for caller */ + struct xfs_agf *agf = XFS_BUF_TO_AGF(cur->bc_private.a.agbp); + xfs_agnumber_t seqno = be32_to_cpu(agf->agf_seqno); + __be32 len; int numrecs; - int optr; /* old ptr value */ - xfs_alloc_ptr_t *pp; /* pointer to btree addresses */ - int ptr; /* index in btree block for this rec */ - xfs_alloc_rec_t *rp; /* pointer to btree records */ - ASSERT(be32_to_cpu(recp->ar_blockcount) > 0); + ASSERT(cur->bc_btnum == XFS_BTNUM_CNT); + + switch (reason) { + case LASTREC_UPDATE: + /* + * If this is the last leaf block and it's the last record, + * then update the size of the longest extent in the AG. + */ + if (ptr != xfs_btree_get_numrecs(block)) + return; + len = rec->alloc.ar_blockcount; + break; + case LASTREC_INSREC: + if (be32_to_cpu(rec->alloc.ar_blockcount) <= + be32_to_cpu(agf->agf_longest)) + return; + len = rec->alloc.ar_blockcount; + break; + case LASTREC_DELREC: + numrecs = xfs_btree_get_numrecs(block); + if (ptr <= numrecs) + return; + ASSERT(ptr == numrecs + 1); - /* - * GCC doesn't understand the (arguably complex) control flow in - * this function and complains about uninitialized structure fields - * without this. - */ - memset(&nrec, 0, sizeof(nrec)); + if (numrecs) { + xfs_alloc_rec_t *rrp; - /* - * If we made it to the root level, allocate a new root block - * and we're done. - */ - if (level >= cur->bc_nlevels) { - XFS_STATS_INC(xs_abt_insrec); - if ((error = xfs_alloc_newroot(cur, &i))) - return error; - *bnop = NULLAGBLOCK; - *stat = i; - return 0; - } - /* - * Make a key out of the record data to be inserted, and save it. - */ - key.ar_startblock = recp->ar_startblock; - key.ar_blockcount = recp->ar_blockcount; - optr = ptr = cur->bc_ptrs[level]; - /* - * If we're off the left edge, return failure. - */ - if (ptr == 0) { - *stat = 0; - return 0; - } - XFS_STATS_INC(xs_abt_insrec); - /* - * Get pointers to the btree buffer and block. - */ - bp = cur->bc_bufs[level]; - block = XFS_BUF_TO_ALLOC_BLOCK(bp); - numrecs = be16_to_cpu(block->bb_numrecs); -#ifdef DEBUG - if ((error = xfs_btree_check_sblock(cur, block, level, bp))) - return error; - /* - * Check that the new entry is being inserted in the right place. - */ - if (ptr <= numrecs) { - if (level == 0) { - rp = XFS_ALLOC_REC_ADDR(block, ptr, cur); - xfs_btree_check_rec(cur->bc_btnum, recp, rp); + rrp = XFS_ALLOC_REC_ADDR(cur->bc_mp, block, numrecs); + len = rrp->ar_blockcount; } else { - kp = XFS_ALLOC_KEY_ADDR(block, ptr, cur); - xfs_btree_check_key(cur->bc_btnum, &key, kp); - } - } -#endif - nbno = NULLAGBLOCK; - ncur = NULL; - /* - * If the block is full, we can't insert the new entry until we - * make the block un-full. - */ - if (numrecs == XFS_ALLOC_BLOCK_MAXRECS(level, cur)) { - /* - * First, try shifting an entry to the right neighbor. - */ - if ((error = xfs_alloc_rshift(cur, level, &i))) - return error; - if (i) { - /* nothing */ - } - /* - * Next, try shifting an entry to the left neighbor. - */ - else { - if ((error = xfs_alloc_lshift(cur, level, &i))) - return error; - if (i) - optr = ptr = cur->bc_ptrs[level]; - else { - /* - * Next, try splitting the current block in - * half. If this works we have to re-set our - * variables because we could be in a - * different block now. - */ - if ((error = xfs_alloc_split(cur, level, &nbno, - &nkey, &ncur, &i))) - return error; - if (i) { - bp = cur->bc_bufs[level]; - block = XFS_BUF_TO_ALLOC_BLOCK(bp); -#ifdef DEBUG - if ((error = - xfs_btree_check_sblock(cur, - block, level, bp))) - return error; -#endif - ptr = cur->bc_ptrs[level]; - nrec.ar_startblock = nkey.ar_startblock; - nrec.ar_blockcount = nkey.ar_blockcount; - } - /* - * Otherwise the insert fails. - */ - else { - *stat = 0; - return 0; - } - } - } - } - /* - * At this point we know there's room for our new entry in the block - * we're pointing at. - */ - numrecs = be16_to_cpu(block->bb_numrecs); - if (level > 0) { - /* - * It's a non-leaf entry. Make a hole for the new data - * in the key and ptr regions of the block. - */ - kp = XFS_ALLOC_KEY_ADDR(block, 1, cur); - pp = XFS_ALLOC_PTR_ADDR(block, 1, cur); -#ifdef DEBUG - for (i = numrecs; i >= ptr; i--) { - if ((error = xfs_btree_check_sptr(cur, be32_to_cpu(pp[i - 1]), level))) - return error; + len = 0; } -#endif - memmove(&kp[ptr], &kp[ptr - 1], - (numrecs - ptr + 1) * sizeof(*kp)); - memmove(&pp[ptr], &pp[ptr - 1], - (numrecs - ptr + 1) * sizeof(*pp)); -#ifdef DEBUG - if ((error = xfs_btree_check_sptr(cur, *bnop, level))) - return error; -#endif - /* - * Now stuff the new data in, bump numrecs and log the new data. - */ - kp[ptr - 1] = key; - pp[ptr - 1] = cpu_to_be32(*bnop); - numrecs++; - block->bb_numrecs = cpu_to_be16(numrecs); - xfs_alloc_log_keys(cur, bp, ptr, numrecs); - xfs_alloc_log_ptrs(cur, bp, ptr, numrecs); -#ifdef DEBUG - if (ptr < numrecs) - xfs_btree_check_key(cur->bc_btnum, kp + ptr - 1, - kp + ptr); -#endif - } else { - /* - * It's a leaf entry. Make a hole for the new record. - */ - rp = XFS_ALLOC_REC_ADDR(block, 1, cur); - memmove(&rp[ptr], &rp[ptr - 1], - (numrecs - ptr + 1) * sizeof(*rp)); - /* - * Now stuff the new record in, bump numrecs - * and log the new data. - */ - rp[ptr - 1] = *recp; - numrecs++; - block->bb_numrecs = cpu_to_be16(numrecs); - xfs_alloc_log_recs(cur, bp, ptr, numrecs); -#ifdef DEBUG - if (ptr < numrecs) - xfs_btree_check_rec(cur->bc_btnum, rp + ptr - 1, - rp + ptr); -#endif - } - /* - * Log the new number of records in the btree header. - */ - xfs_alloc_log_block(cur->bc_tp, bp, XFS_BB_NUMRECS); - /* - * If we inserted at the start of a block, update the parents' keys. - */ - if (optr == 1 && (error = xfs_alloc_updkey(cur, &key, level + 1))) - return error; - /* - * Look to see if the longest extent in the allocation group - * needs to be updated. - */ - agf = XFS_BUF_TO_AGF(cur->bc_private.a.agbp); - if (level == 0 && - cur->bc_btnum == XFS_BTNUM_CNT && - be32_to_cpu(block->bb_rightsib) == NULLAGBLOCK && - be32_to_cpu(recp->ar_blockcount) > be32_to_cpu(agf->agf_longest)) { - /* - * If this is a leaf in the by-size btree and there - * is no right sibling block and this block is bigger - * than the previous longest block, update it. - */ - agf->agf_longest = recp->ar_blockcount; - cur->bc_mp->m_perag[be32_to_cpu(agf->agf_seqno)].pagf_longest - = be32_to_cpu(recp->ar_blockcount); - xfs_alloc_log_agf(cur->bc_tp, cur->bc_private.a.agbp, - XFS_AGF_LONGEST); + break; + default: + ASSERT(0); + return; } - /* - * Return the new block number, if any. - * If there is one, give back a record value and a cursor too. - */ - *bnop = nbno; - if (nbno != NULLAGBLOCK) { - *recp = nrec; - *curp = ncur; - } - *stat = 1; - return 0; + + agf->agf_longest = len; + cur->bc_mp->m_perag[seqno].pagf_longest = be32_to_cpu(len); + xfs_alloc_log_agf(cur->bc_tp, cur->bc_private.a.agbp, XFS_AGF_LONGEST); } -/* - * Log header fields from a btree block. - */ -STATIC void -xfs_alloc_log_block( - xfs_trans_t *tp, /* transaction pointer */ - xfs_buf_t *bp, /* buffer containing btree block */ - int fields) /* mask of fields: XFS_BB_... */ +STATIC int +xfs_allocbt_get_minrecs( + struct xfs_btree_cur *cur, + int level) { - int first; /* first byte offset logged */ - int last; /* last byte offset logged */ - static const short offsets[] = { /* table of offsets */ - offsetof(xfs_alloc_block_t, bb_magic), - offsetof(xfs_alloc_block_t, bb_level), - offsetof(xfs_alloc_block_t, bb_numrecs), - offsetof(xfs_alloc_block_t, bb_leftsib), - offsetof(xfs_alloc_block_t, bb_rightsib), - sizeof(xfs_alloc_block_t) - }; + return cur->bc_mp->m_alloc_mnr[level != 0]; +} - xfs_btree_offsets(fields, offsets, XFS_BB_NUM_BITS, &first, &last); - xfs_trans_log_buf(tp, bp, first, last); +STATIC int +xfs_allocbt_get_maxrecs( + struct xfs_btree_cur *cur, + int level) +{ + return cur->bc_mp->m_alloc_mxr[level != 0]; } -/* - * Log keys from a btree block (nonleaf). - */ STATIC void -xfs_alloc_log_keys( - xfs_btree_cur_t *cur, /* btree cursor */ - xfs_buf_t *bp, /* buffer containing btree block */ - int kfirst, /* index of first key to log */ - int klast) /* index of last key to log */ +xfs_allocbt_init_key_from_rec( + union xfs_btree_key *key, + union xfs_btree_rec *rec) { - xfs_alloc_block_t *block; /* btree block to log from */ - int first; /* first byte offset logged */ - xfs_alloc_key_t *kp; /* key pointer in btree block */ - int last; /* last byte offset logged */ + ASSERT(rec->alloc.ar_startblock != 0); - block = XFS_BUF_TO_ALLOC_BLOCK(bp); - kp = XFS_ALLOC_KEY_ADDR(block, 1, cur); - first = (int)((xfs_caddr_t)&kp[kfirst - 1] - (xfs_caddr_t)block); - last = (int)(((xfs_caddr_t)&kp[klast] - 1) - (xfs_caddr_t)block); - xfs_trans_log_buf(cur->bc_tp, bp, first, last); + key->alloc.ar_startblock = rec->alloc.ar_startblock; + key->alloc.ar_blockcount = rec->alloc.ar_blockcount; } -/* - * Log block pointer fields from a btree block (nonleaf). - */ STATIC void -xfs_alloc_log_ptrs( - xfs_btree_cur_t *cur, /* btree cursor */ - xfs_buf_t *bp, /* buffer containing btree block */ - int pfirst, /* index of first pointer to log */ - int plast) /* index of last pointer to log */ +xfs_allocbt_init_rec_from_key( + union xfs_btree_key *key, + union xfs_btree_rec *rec) { - xfs_alloc_block_t *block; /* btree block to log from */ - int first; /* first byte offset logged */ - int last; /* last byte offset logged */ - xfs_alloc_ptr_t *pp; /* block-pointer pointer in btree blk */ + ASSERT(key->alloc.ar_startblock != 0); - block = XFS_BUF_TO_ALLOC_BLOCK(bp); - pp = XFS_ALLOC_PTR_ADDR(block, 1, cur); - first = (int)((xfs_caddr_t)&pp[pfirst - 1] - (xfs_caddr_t)block); - last = (int)(((xfs_caddr_t)&pp[plast] - 1) - (xfs_caddr_t)block); - xfs_trans_log_buf(cur->bc_tp, bp, first, last); + rec->alloc.ar_startblock = key->alloc.ar_startblock; + rec->alloc.ar_blockcount = key->alloc.ar_blockcount; } -/* - * Log records from a btree block (leaf). - */ STATIC void -xfs_alloc_log_recs( - xfs_btree_cur_t *cur, /* btree cursor */ - xfs_buf_t *bp, /* buffer containing btree block */ - int rfirst, /* index of first record to log */ - int rlast) /* index of last record to log */ +xfs_allocbt_init_rec_from_cur( + struct xfs_btree_cur *cur, + union xfs_btree_rec *rec) { - xfs_alloc_block_t *block; /* btree block to log from */ - int first; /* first byte offset logged */ - int last; /* last byte offset logged */ - xfs_alloc_rec_t *rp; /* record pointer for btree block */ - + ASSERT(cur->bc_rec.a.ar_startblock != 0); - block = XFS_BUF_TO_ALLOC_BLOCK(bp); - rp = XFS_ALLOC_REC_ADDR(block, 1, cur); -#ifdef DEBUG - { - xfs_agf_t *agf; - xfs_alloc_rec_t *p; - - agf = XFS_BUF_TO_AGF(cur->bc_private.a.agbp); - for (p = &rp[rfirst - 1]; p <= &rp[rlast - 1]; p++) - ASSERT(be32_to_cpu(p->ar_startblock) + - be32_to_cpu(p->ar_blockcount) <= - be32_to_cpu(agf->agf_length)); - } -#endif - first = (int)((xfs_caddr_t)&rp[rfirst - 1] - (xfs_caddr_t)block); - last = (int)(((xfs_caddr_t)&rp[rlast] - 1) - (xfs_caddr_t)block); - xfs_trans_log_buf(cur->bc_tp, bp, first, last); + rec->alloc.ar_startblock = cpu_to_be32(cur->bc_rec.a.ar_startblock); + rec->alloc.ar_blockcount = cpu_to_be32(cur->bc_rec.a.ar_blockcount); } -/* - * Lookup the record. The cursor is made to point to it, based on dir. - * Return 0 if can't find any such record, 1 for success. - */ -STATIC int /* error */ -xfs_alloc_lookup( - xfs_btree_cur_t *cur, /* btree cursor */ - xfs_lookup_t dir, /* <=, ==, or >= */ - int *stat) /* success/failure */ +STATIC void +xfs_allocbt_init_ptr_from_cur( + struct xfs_btree_cur *cur, + union xfs_btree_ptr *ptr) { - xfs_agblock_t agbno; /* a.g. relative btree block number */ - xfs_agnumber_t agno; /* allocation group number */ - xfs_alloc_block_t *block=NULL; /* current btree block */ - int diff; /* difference for the current key */ - int error; /* error return value */ - int keyno=0; /* current key number */ - int level; /* level in the btree */ - xfs_mount_t *mp; /* file system mount point */ - - XFS_STATS_INC(xs_abt_lookup); - /* - * Get the allocation group header, and the root block number. - */ - mp = cur->bc_mp; - - { - xfs_agf_t *agf; /* a.g. freespace header */ - - agf = XFS_BUF_TO_AGF(cur->bc_private.a.agbp); - agno = be32_to_cpu(agf->agf_seqno); - agbno = be32_to_cpu(agf->agf_roots[cur->bc_btnum]); - } - /* - * Iterate over each level in the btree, starting at the root. - * For each level above the leaves, find the key we need, based - * on the lookup record, then follow the corresponding block - * pointer down to the next level. - */ - for (level = cur->bc_nlevels - 1, diff = 1; level >= 0; level--) { - xfs_buf_t *bp; /* buffer pointer for btree block */ - xfs_daddr_t d; /* disk address of btree block */ - - /* - * Get the disk address we're looking for. - */ - d = XFS_AGB_TO_DADDR(mp, agno, agbno); - /* - * If the old buffer at this level is for a different block, - * throw it away, otherwise just use it. - */ - bp = cur->bc_bufs[level]; - if (bp && XFS_BUF_ADDR(bp) != d) - bp = NULL; - if (!bp) { - /* - * Need to get a new buffer. Read it, then - * set it in the cursor, releasing the old one. - */ - if ((error = xfs_btree_read_bufs(mp, cur->bc_tp, agno, - agbno, 0, &bp, XFS_ALLOC_BTREE_REF))) - return error; - xfs_btree_setbuf(cur, level, bp); - /* - * Point to the btree block, now that we have the buffer - */ - block = XFS_BUF_TO_ALLOC_BLOCK(bp); - if ((error = xfs_btree_check_sblock(cur, block, level, - bp))) - return error; - } else - block = XFS_BUF_TO_ALLOC_BLOCK(bp); - /* - * If we already had a key match at a higher level, we know - * we need to use the first entry in this block. - */ - if (diff == 0) - keyno = 1; - /* - * Otherwise we need to search this block. Do a binary search. - */ - else { - int high; /* high entry number */ - xfs_alloc_key_t *kkbase=NULL;/* base of keys in block */ - xfs_alloc_rec_t *krbase=NULL;/* base of records in block */ - int low; /* low entry number */ - - /* - * Get a pointer to keys or records. - */ - if (level > 0) - kkbase = XFS_ALLOC_KEY_ADDR(block, 1, cur); - else - krbase = XFS_ALLOC_REC_ADDR(block, 1, cur); - /* - * Set low and high entry numbers, 1-based. - */ - low = 1; - if (!(high = be16_to_cpu(block->bb_numrecs))) { - /* - * If the block is empty, the tree must - * be an empty leaf. - */ - ASSERT(level == 0 && cur->bc_nlevels == 1); - cur->bc_ptrs[0] = dir != XFS_LOOKUP_LE; - *stat = 0; - return 0; - } - /* - * Binary search the block. - */ - while (low <= high) { - xfs_extlen_t blockcount; /* key value */ - xfs_agblock_t startblock; /* key value */ - - XFS_STATS_INC(xs_abt_compare); - /* - * keyno is average of low and high. - */ - keyno = (low + high) >> 1; - /* - * Get startblock & blockcount. - */ - if (level > 0) { - xfs_alloc_key_t *kkp; - - kkp = kkbase + keyno - 1; - startblock = be32_to_cpu(kkp->ar_startblock); - blockcount = be32_to_cpu(kkp->ar_blockcount); - } else { - xfs_alloc_rec_t *krp; + struct xfs_agf *agf = XFS_BUF_TO_AGF(cur->bc_private.a.agbp); - krp = krbase + keyno - 1; - startblock = be32_to_cpu(krp->ar_startblock); - blockcount = be32_to_cpu(krp->ar_blockcount); - } - /* - * Compute difference to get next direction. - */ - if (cur->bc_btnum == XFS_BTNUM_BNO) - diff = (int)startblock - - (int)cur->bc_rec.a.ar_startblock; - else if (!(diff = (int)blockcount - - (int)cur->bc_rec.a.ar_blockcount)) - diff = (int)startblock - - (int)cur->bc_rec.a.ar_startblock; - /* - * Less than, move right. - */ - if (diff < 0) - low = keyno + 1; - /* - * Greater than, move left. - */ - else if (diff > 0) - high = keyno - 1; - /* - * Equal, we're done. - */ - else - break; - } - } - /* - * If there are more levels, set up for the next level - * by getting the block number and filling in the cursor. - */ - if (level > 0) { - /* - * If we moved left, need the previous key number, - * unless there isn't one. - */ - if (diff > 0 && --keyno < 1) - keyno = 1; - agbno = be32_to_cpu(*XFS_ALLOC_PTR_ADDR(block, keyno, cur)); -#ifdef DEBUG - if ((error = xfs_btree_check_sptr(cur, agbno, level))) - return error; -#endif - cur->bc_ptrs[level] = keyno; - } - } - /* - * Done with the search. - * See if we need to adjust the results. - */ - if (dir != XFS_LOOKUP_LE && diff < 0) { - keyno++; - /* - * If ge search and we went off the end of the block, but it's - * not the last block, we're in the wrong block. - */ - if (dir == XFS_LOOKUP_GE && - keyno > be16_to_cpu(block->bb_numrecs) && - be32_to_cpu(block->bb_rightsib) != NULLAGBLOCK) { - int i; + ASSERT(cur->bc_private.a.agno == be32_to_cpu(agf->agf_seqno)); + ASSERT(agf->agf_roots[cur->bc_btnum] != 0); - cur->bc_ptrs[0] = keyno; - if ((error = xfs_alloc_increment(cur, 0, &i))) - return error; - XFS_WANT_CORRUPTED_RETURN(i == 1); - *stat = 1; - return 0; - } - } - else if (dir == XFS_LOOKUP_LE && diff > 0) - keyno--; - cur->bc_ptrs[0] = keyno; - /* - * Return if we succeeded or not. - */ - if (keyno == 0 || keyno > be16_to_cpu(block->bb_numrecs)) - *stat = 0; - else - *stat = ((dir != XFS_LOOKUP_EQ) || (diff == 0)); - return 0; + ptr->s = agf->agf_roots[cur->bc_btnum]; } -/* - * Move 1 record left from cur/level if possible. - * Update cur to reflect the new path. - */ -STATIC int /* error */ -xfs_alloc_lshift( - xfs_btree_cur_t *cur, /* btree cursor */ - int level, /* level to shift record on */ - int *stat) /* success/failure */ +STATIC __int64_t +xfs_allocbt_key_diff( + struct xfs_btree_cur *cur, + union xfs_btree_key *key) { - int error; /* error return value */ -#ifdef DEBUG - int i; /* loop index */ -#endif - xfs_alloc_key_t key; /* key value for leaf level upward */ - xfs_buf_t *lbp; /* buffer for left neighbor block */ - xfs_alloc_block_t *left; /* left neighbor btree block */ - int nrec; /* new number of left block entries */ - xfs_buf_t *rbp; /* buffer for right (current) block */ - xfs_alloc_block_t *right; /* right (current) btree block */ - xfs_alloc_key_t *rkp=NULL; /* key pointer for right block */ - xfs_alloc_ptr_t *rpp=NULL; /* address pointer for right block */ - xfs_alloc_rec_t *rrp=NULL; /* record pointer for right block */ + xfs_alloc_rec_incore_t *rec = &cur->bc_rec.a; + xfs_alloc_key_t *kp = &key->alloc; + __int64_t diff; - /* - * Set up variables for this block as "right". - */ - rbp = cur->bc_bufs[level]; - right = XFS_BUF_TO_ALLOC_BLOCK(rbp); -#ifdef DEBUG - if ((error = xfs_btree_check_sblock(cur, right, level, rbp))) - return error; -#endif - /* - * If we've got no left sibling then we can't shift an entry left. - */ - if (be32_to_cpu(right->bb_leftsib) == NULLAGBLOCK) { - *stat = 0; - return 0; - } - /* - * If the cursor entry is the one that would be moved, don't - * do it... it's too complicated. - */ - if (cur->bc_ptrs[level] <= 1) { - *stat = 0; - return 0; - } - /* - * Set up the left neighbor as "left". - */ - if ((error = xfs_btree_read_bufs(cur->bc_mp, cur->bc_tp, - cur->bc_private.a.agno, be32_to_cpu(right->bb_leftsib), - 0, &lbp, XFS_ALLOC_BTREE_REF))) - return error; - left = XFS_BUF_TO_ALLOC_BLOCK(lbp); - if ((error = xfs_btree_check_sblock(cur, left, level, lbp))) - return error; - /* - * If it's full, it can't take another entry. - */ - if (be16_to_cpu(left->bb_numrecs) == XFS_ALLOC_BLOCK_MAXRECS(level, cur)) { - *stat = 0; - return 0; + if (cur->bc_btnum == XFS_BTNUM_BNO) { + return (__int64_t)be32_to_cpu(kp->ar_startblock) - + rec->ar_startblock; } - nrec = be16_to_cpu(left->bb_numrecs) + 1; - /* - * If non-leaf, copy a key and a ptr to the left block. - */ - if (level > 0) { - xfs_alloc_key_t *lkp; /* key pointer for left block */ - xfs_alloc_ptr_t *lpp; /* address pointer for left block */ - lkp = XFS_ALLOC_KEY_ADDR(left, nrec, cur); - rkp = XFS_ALLOC_KEY_ADDR(right, 1, cur); - *lkp = *rkp; - xfs_alloc_log_keys(cur, lbp, nrec, nrec); - lpp = XFS_ALLOC_PTR_ADDR(left, nrec, cur); - rpp = XFS_ALLOC_PTR_ADDR(right, 1, cur); -#ifdef DEBUG - if ((error = xfs_btree_check_sptr(cur, be32_to_cpu(*rpp), level))) - return error; -#endif - *lpp = *rpp; - xfs_alloc_log_ptrs(cur, lbp, nrec, nrec); - xfs_btree_check_key(cur->bc_btnum, lkp - 1, lkp); - } - /* - * If leaf, copy a record to the left block. - */ - else { - xfs_alloc_rec_t *lrp; /* record pointer for left block */ + diff = (__int64_t)be32_to_cpu(kp->ar_blockcount) - rec->ar_blockcount; + if (diff) + return diff; - lrp = XFS_ALLOC_REC_ADDR(left, nrec, cur); - rrp = XFS_ALLOC_REC_ADDR(right, 1, cur); - *lrp = *rrp; - xfs_alloc_log_recs(cur, lbp, nrec, nrec); - xfs_btree_check_rec(cur->bc_btnum, lrp - 1, lrp); - } - /* - * Bump and log left's numrecs, decrement and log right's numrecs. - */ - be16_add_cpu(&left->bb_numrecs, 1); - xfs_alloc_log_block(cur->bc_tp, lbp, XFS_BB_NUMRECS); - be16_add_cpu(&right->bb_numrecs, -1); - xfs_alloc_log_block(cur->bc_tp, rbp, XFS_BB_NUMRECS); - /* - * Slide the contents of right down one entry. - */ - if (level > 0) { -#ifdef DEBUG - for (i = 0; i < be16_to_cpu(right->bb_numrecs); i++) { - if ((error = xfs_btree_check_sptr(cur, be32_to_cpu(rpp[i + 1]), - level))) - return error; - } -#endif - memmove(rkp, rkp + 1, be16_to_cpu(right->bb_numrecs) * sizeof(*rkp)); - memmove(rpp, rpp + 1, be16_to_cpu(right->bb_numrecs) * sizeof(*rpp)); - xfs_alloc_log_keys(cur, rbp, 1, be16_to_cpu(right->bb_numrecs)); - xfs_alloc_log_ptrs(cur, rbp, 1, be16_to_cpu(right->bb_numrecs)); - } else { - memmove(rrp, rrp + 1, be16_to_cpu(right->bb_numrecs) * sizeof(*rrp)); - xfs_alloc_log_recs(cur, rbp, 1, be16_to_cpu(right->bb_numrecs)); - key.ar_startblock = rrp->ar_startblock; - key.ar_blockcount = rrp->ar_blockcount; - rkp = &key; - } - /* - * Update the parent key values of right. - */ - if ((error = xfs_alloc_updkey(cur, rkp, level + 1))) - return error; - /* - * Slide the cursor value left one. - */ - cur->bc_ptrs[level]--; - *stat = 1; - return 0; + return (__int64_t)be32_to_cpu(kp->ar_startblock) - rec->ar_startblock; } -/* - * Allocate a new root block, fill it in. - */ -STATIC int /* error */ -xfs_alloc_newroot( - xfs_btree_cur_t *cur, /* btree cursor */ - int *stat) /* success/failure */ +STATIC int +xfs_allocbt_kill_root( + struct xfs_btree_cur *cur, + struct xfs_buf *bp, + int level, + union xfs_btree_ptr *newroot) { - int error; /* error return value */ - xfs_agblock_t lbno; /* left block number */ - xfs_buf_t *lbp; /* left btree buffer */ - xfs_alloc_block_t *left; /* left btree block */ - xfs_mount_t *mp; /* mount structure */ - xfs_agblock_t nbno; /* new block number */ - xfs_buf_t *nbp; /* new (root) buffer */ - xfs_alloc_block_t *new; /* new (root) btree block */ - int nptr; /* new value for key index, 1 or 2 */ - xfs_agblock_t rbno; /* right block number */ - xfs_buf_t *rbp; /* right btree buffer */ - xfs_alloc_block_t *right; /* right btree block */ - - mp = cur->bc_mp; + int error; - ASSERT(cur->bc_nlevels < XFS_AG_MAXLEVELS(mp)); - /* - * Get a buffer from the freelist blocks, for the new root. - */ - error = xfs_alloc_get_freelist(cur->bc_tp, - cur->bc_private.a.agbp, &nbno, 1); - if (error) - return error; - /* - * None available, we fail. - */ - if (nbno == NULLAGBLOCK) { - *stat = 0; - return 0; - } - xfs_trans_agbtree_delta(cur->bc_tp, 1); - nbp = xfs_btree_get_bufs(mp, cur->bc_tp, cur->bc_private.a.agno, nbno, - 0); - new = XFS_BUF_TO_ALLOC_BLOCK(nbp); - /* - * Set the root data in the a.g. freespace structure. - */ - { - xfs_agf_t *agf; /* a.g. freespace header */ - xfs_agnumber_t seqno; + XFS_BTREE_TRACE_CURSOR(cur, XBT_ENTRY); + XFS_BTREE_STATS_INC(cur, killroot); - agf = XFS_BUF_TO_AGF(cur->bc_private.a.agbp); - agf->agf_roots[cur->bc_btnum] = cpu_to_be32(nbno); - be32_add_cpu(&agf->agf_levels[cur->bc_btnum], 1); - seqno = be32_to_cpu(agf->agf_seqno); - mp->m_perag[seqno].pagf_levels[cur->bc_btnum]++; - xfs_alloc_log_agf(cur->bc_tp, cur->bc_private.a.agbp, - XFS_AGF_ROOTS | XFS_AGF_LEVELS); - } /* - * At the previous root level there are now two blocks: the old - * root, and the new block generated when it was split. - * We don't know which one the cursor is pointing at, so we - * set up variables "left" and "right" for each case. + * Update the root pointer, decreasing the level by 1 and then + * free the old root. */ - lbp = cur->bc_bufs[cur->bc_nlevels - 1]; - left = XFS_BUF_TO_ALLOC_BLOCK(lbp); -#ifdef DEBUG - if ((error = xfs_btree_check_sblock(cur, left, cur->bc_nlevels - 1, lbp))) + xfs_allocbt_set_root(cur, newroot, -1); + error = xfs_allocbt_free_block(cur, bp); + if (error) { + XFS_BTREE_TRACE_CURSOR(cur, XBT_ERROR); return error; -#endif - if (be32_to_cpu(left->bb_rightsib) != NULLAGBLOCK) { - /* - * Our block is left, pick up the right block. - */ - lbno = XFS_DADDR_TO_AGBNO(mp, XFS_BUF_ADDR(lbp)); - rbno = be32_to_cpu(left->bb_rightsib); - if ((error = xfs_btree_read_bufs(mp, cur->bc_tp, - cur->bc_private.a.agno, rbno, 0, &rbp, - XFS_ALLOC_BTREE_REF))) - return error; - right = XFS_BUF_TO_ALLOC_BLOCK(rbp); - if ((error = xfs_btree_check_sblock(cur, right, - cur->bc_nlevels - 1, rbp))) - return error; - nptr = 1; - } else { - /* - * Our block is right, pick up the left block. - */ - rbp = lbp; - right = left; - rbno = XFS_DADDR_TO_AGBNO(mp, XFS_BUF_ADDR(rbp)); - lbno = be32_to_cpu(right->bb_leftsib); - if ((error = xfs_btree_read_bufs(mp, cur->bc_tp, - cur->bc_private.a.agno, lbno, 0, &lbp, - XFS_ALLOC_BTREE_REF))) - return error; - left = XFS_BUF_TO_ALLOC_BLOCK(lbp); - if ((error = xfs_btree_check_sblock(cur, left, - cur->bc_nlevels - 1, lbp))) - return error; - nptr = 2; } - /* - * Fill in the new block's btree header and log it. - */ - new->bb_magic = cpu_to_be32(xfs_magics[cur->bc_btnum]); - new->bb_level = cpu_to_be16(cur->bc_nlevels); - new->bb_numrecs = cpu_to_be16(2); - new->bb_leftsib = cpu_to_be32(NULLAGBLOCK); - new->bb_rightsib = cpu_to_be32(NULLAGBLOCK); - xfs_alloc_log_block(cur->bc_tp, nbp, XFS_BB_ALL_BITS); - ASSERT(lbno != NULLAGBLOCK && rbno != NULLAGBLOCK); - /* - * Fill in the key data in the new root. - */ - { - xfs_alloc_key_t *kp; /* btree key pointer */ - kp = XFS_ALLOC_KEY_ADDR(new, 1, cur); - if (be16_to_cpu(left->bb_level) > 0) { - kp[0] = *XFS_ALLOC_KEY_ADDR(left, 1, cur); - kp[1] = *XFS_ALLOC_KEY_ADDR(right, 1, cur); - } else { - xfs_alloc_rec_t *rp; /* btree record pointer */ + XFS_BTREE_STATS_INC(cur, free); - rp = XFS_ALLOC_REC_ADDR(left, 1, cur); - kp[0].ar_startblock = rp->ar_startblock; - kp[0].ar_blockcount = rp->ar_blockcount; - rp = XFS_ALLOC_REC_ADDR(right, 1, cur); - kp[1].ar_startblock = rp->ar_startblock; - kp[1].ar_blockcount = rp->ar_blockcount; - } - } - xfs_alloc_log_keys(cur, nbp, 1, 2); - /* - * Fill in the pointer data in the new root. - */ - { - xfs_alloc_ptr_t *pp; /* btree address pointer */ + xfs_btree_setbuf(cur, level, NULL); + cur->bc_nlevels--; - pp = XFS_ALLOC_PTR_ADDR(new, 1, cur); - pp[0] = cpu_to_be32(lbno); - pp[1] = cpu_to_be32(rbno); - } - xfs_alloc_log_ptrs(cur, nbp, 1, 2); - /* - * Fix up the cursor. - */ - xfs_btree_setbuf(cur, cur->bc_nlevels, nbp); - cur->bc_ptrs[cur->bc_nlevels] = nptr; - cur->bc_nlevels++; - *stat = 1; + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); return 0; } -/* - * Move 1 record right from cur/level if possible. - * Update cur to reflect the new path. - */ -STATIC int /* error */ -xfs_alloc_rshift( - xfs_btree_cur_t *cur, /* btree cursor */ - int level, /* level to shift record on */ - int *stat) /* success/failure */ -{ - int error; /* error return value */ - int i; /* loop index */ - xfs_alloc_key_t key; /* key value for leaf level upward */ - xfs_buf_t *lbp; /* buffer for left (current) block */ - xfs_alloc_block_t *left; /* left (current) btree block */ - xfs_buf_t *rbp; /* buffer for right neighbor block */ - xfs_alloc_block_t *right; /* right neighbor btree block */ - xfs_alloc_key_t *rkp; /* key pointer for right block */ - xfs_btree_cur_t *tcur; /* temporary cursor */ - - /* - * Set up variables for this block as "left". - */ - lbp = cur->bc_bufs[level]; - left = XFS_BUF_TO_ALLOC_BLOCK(lbp); -#ifdef DEBUG - if ((error = xfs_btree_check_sblock(cur, left, level, lbp))) - return error; -#endif - /* - * If we've got no right sibling then we can't shift an entry right. - */ - if (be32_to_cpu(left->bb_rightsib) == NULLAGBLOCK) { - *stat = 0; - return 0; - } - /* - * If the cursor entry is the one that would be moved, don't - * do it... it's too complicated. - */ - if (cur->bc_ptrs[level] >= be16_to_cpu(left->bb_numrecs)) { - *stat = 0; - return 0; - } - /* - * Set up the right neighbor as "right". - */ - if ((error = xfs_btree_read_bufs(cur->bc_mp, cur->bc_tp, - cur->bc_private.a.agno, be32_to_cpu(left->bb_rightsib), - 0, &rbp, XFS_ALLOC_BTREE_REF))) - return error; - right = XFS_BUF_TO_ALLOC_BLOCK(rbp); - if ((error = xfs_btree_check_sblock(cur, right, level, rbp))) - return error; - /* - * If it's full, it can't take another entry. - */ - if (be16_to_cpu(right->bb_numrecs) == XFS_ALLOC_BLOCK_MAXRECS(level, cur)) { - *stat = 0; - return 0; - } - /* - * Make a hole at the start of the right neighbor block, then - * copy the last left block entry to the hole. - */ - if (level > 0) { - xfs_alloc_key_t *lkp; /* key pointer for left block */ - xfs_alloc_ptr_t *lpp; /* address pointer for left block */ - xfs_alloc_ptr_t *rpp; /* address pointer for right block */ - - lkp = XFS_ALLOC_KEY_ADDR(left, be16_to_cpu(left->bb_numrecs), cur); - lpp = XFS_ALLOC_PTR_ADDR(left, be16_to_cpu(left->bb_numrecs), cur); - rkp = XFS_ALLOC_KEY_ADDR(right, 1, cur); - rpp = XFS_ALLOC_PTR_ADDR(right, 1, cur); #ifdef DEBUG - for (i = be16_to_cpu(right->bb_numrecs) - 1; i >= 0; i--) { - if ((error = xfs_btree_check_sptr(cur, be32_to_cpu(rpp[i]), level))) - return error; - } -#endif - memmove(rkp + 1, rkp, be16_to_cpu(right->bb_numrecs) * sizeof(*rkp)); - memmove(rpp + 1, rpp, be16_to_cpu(right->bb_numrecs) * sizeof(*rpp)); -#ifdef DEBUG - if ((error = xfs_btree_check_sptr(cur, be32_to_cpu(*lpp), level))) - return error; -#endif - *rkp = *lkp; - *rpp = *lpp; - xfs_alloc_log_keys(cur, rbp, 1, be16_to_cpu(right->bb_numrecs) + 1); - xfs_alloc_log_ptrs(cur, rbp, 1, be16_to_cpu(right->bb_numrecs) + 1); - xfs_btree_check_key(cur->bc_btnum, rkp, rkp + 1); +STATIC int +xfs_allocbt_keys_inorder( + struct xfs_btree_cur *cur, + union xfs_btree_key *k1, + union xfs_btree_key *k2) +{ + if (cur->bc_btnum == XFS_BTNUM_BNO) { + return be32_to_cpu(k1->alloc.ar_startblock) < + be32_to_cpu(k2->alloc.ar_startblock); } else { - xfs_alloc_rec_t *lrp; /* record pointer for left block */ - xfs_alloc_rec_t *rrp; /* record pointer for right block */ - - lrp = XFS_ALLOC_REC_ADDR(left, be16_to_cpu(left->bb_numrecs), cur); - rrp = XFS_ALLOC_REC_ADDR(right, 1, cur); - memmove(rrp + 1, rrp, be16_to_cpu(right->bb_numrecs) * sizeof(*rrp)); - *rrp = *lrp; - xfs_alloc_log_recs(cur, rbp, 1, be16_to_cpu(right->bb_numrecs) + 1); - key.ar_startblock = rrp->ar_startblock; - key.ar_blockcount = rrp->ar_blockcount; - rkp = &key; - xfs_btree_check_rec(cur->bc_btnum, rrp, rrp + 1); + return be32_to_cpu(k1->alloc.ar_blockcount) < + be32_to_cpu(k2->alloc.ar_blockcount) || + (k1->alloc.ar_blockcount == k2->alloc.ar_blockcount && + be32_to_cpu(k1->alloc.ar_startblock) < + be32_to_cpu(k2->alloc.ar_startblock)); } - /* - * Decrement and log left's numrecs, bump and log right's numrecs. - */ - be16_add_cpu(&left->bb_numrecs, -1); - xfs_alloc_log_block(cur->bc_tp, lbp, XFS_BB_NUMRECS); - be16_add_cpu(&right->bb_numrecs, 1); - xfs_alloc_log_block(cur->bc_tp, rbp, XFS_BB_NUMRECS); - /* - * Using a temporary cursor, update the parent key values of the - * block on the right. - */ - if ((error = xfs_btree_dup_cursor(cur, &tcur))) - return error; - i = xfs_btree_lastrec(tcur, level); - XFS_WANT_CORRUPTED_GOTO(i == 1, error0); - if ((error = xfs_alloc_increment(tcur, level, &i)) || - (error = xfs_alloc_updkey(tcur, rkp, level + 1))) - goto error0; - xfs_btree_del_cursor(tcur, XFS_BTREE_NOERROR); - *stat = 1; - return 0; -error0: - xfs_btree_del_cursor(tcur, XFS_BTREE_ERROR); - return error; } -/* - * Split cur/level block in half. - * Return new block number and its first record (to be inserted into parent). - */ -STATIC int /* error */ -xfs_alloc_split( - xfs_btree_cur_t *cur, /* btree cursor */ - int level, /* level to split */ - xfs_agblock_t *bnop, /* output: block number allocated */ - xfs_alloc_key_t *keyp, /* output: first key of new block */ - xfs_btree_cur_t **curp, /* output: new cursor */ - int *stat) /* success/failure */ +STATIC int +xfs_allocbt_recs_inorder( + struct xfs_btree_cur *cur, + union xfs_btree_rec *r1, + union xfs_btree_rec *r2) { - int error; /* error return value */ - int i; /* loop index/record number */ - xfs_agblock_t lbno; /* left (current) block number */ - xfs_buf_t *lbp; /* buffer for left block */ - xfs_alloc_block_t *left; /* left (current) btree block */ - xfs_agblock_t rbno; /* right (new) block number */ - xfs_buf_t *rbp; /* buffer for right block */ - xfs_alloc_block_t *right; /* right (new) btree block */ - - /* - * Allocate the new block from the freelist. - * If we can't do it, we're toast. Give up. - */ - error = xfs_alloc_get_freelist(cur->bc_tp, - cur->bc_private.a.agbp, &rbno, 1); - if (error) - return error; - if (rbno == NULLAGBLOCK) { - *stat = 0; - return 0; - } - xfs_trans_agbtree_delta(cur->bc_tp, 1); - rbp = xfs_btree_get_bufs(cur->bc_mp, cur->bc_tp, cur->bc_private.a.agno, - rbno, 0); - /* - * Set up the new block as "right". - */ - right = XFS_BUF_TO_ALLOC_BLOCK(rbp); - /* - * "Left" is the current (according to the cursor) block. - */ - lbp = cur->bc_bufs[level]; - left = XFS_BUF_TO_ALLOC_BLOCK(lbp); -#ifdef DEBUG - if ((error = xfs_btree_check_sblock(cur, left, level, lbp))) - return error; -#endif - /* - * Fill in the btree header for the new block. - */ - right->bb_magic = cpu_to_be32(xfs_magics[cur->bc_btnum]); - right->bb_level = left->bb_level; - right->bb_numrecs = cpu_to_be16(be16_to_cpu(left->bb_numrecs) / 2); - /* - * Make sure that if there's an odd number of entries now, that - * each new block will have the same number of entries. - */ - if ((be16_to_cpu(left->bb_numrecs) & 1) && - cur->bc_ptrs[level] <= be16_to_cpu(right->bb_numrecs) + 1) - be16_add_cpu(&right->bb_numrecs, 1); - i = be16_to_cpu(left->bb_numrecs) - be16_to_cpu(right->bb_numrecs) + 1; - /* - * For non-leaf blocks, copy keys and addresses over to the new block. - */ - if (level > 0) { - xfs_alloc_key_t *lkp; /* left btree key pointer */ - xfs_alloc_ptr_t *lpp; /* left btree address pointer */ - xfs_alloc_key_t *rkp; /* right btree key pointer */ - xfs_alloc_ptr_t *rpp; /* right btree address pointer */ - - lkp = XFS_ALLOC_KEY_ADDR(left, i, cur); - lpp = XFS_ALLOC_PTR_ADDR(left, i, cur); - rkp = XFS_ALLOC_KEY_ADDR(right, 1, cur); - rpp = XFS_ALLOC_PTR_ADDR(right, 1, cur); -#ifdef DEBUG - for (i = 0; i < be16_to_cpu(right->bb_numrecs); i++) { - if ((error = xfs_btree_check_sptr(cur, be32_to_cpu(lpp[i]), level))) - return error; - } -#endif - memcpy(rkp, lkp, be16_to_cpu(right->bb_numrecs) * sizeof(*rkp)); - memcpy(rpp, lpp, be16_to_cpu(right->bb_numrecs) * sizeof(*rpp)); - xfs_alloc_log_keys(cur, rbp, 1, be16_to_cpu(right->bb_numrecs)); - xfs_alloc_log_ptrs(cur, rbp, 1, be16_to_cpu(right->bb_numrecs)); - *keyp = *rkp; + if (cur->bc_btnum == XFS_BTNUM_BNO) { + return be32_to_cpu(r1->alloc.ar_startblock) + + be32_to_cpu(r1->alloc.ar_blockcount) <= + be32_to_cpu(r2->alloc.ar_startblock); + } else { + return be32_to_cpu(r1->alloc.ar_blockcount) < + be32_to_cpu(r2->alloc.ar_blockcount) || + (r1->alloc.ar_blockcount == r2->alloc.ar_blockcount && + be32_to_cpu(r1->alloc.ar_startblock) < + be32_to_cpu(r2->alloc.ar_startblock)); } - /* - * For leaf blocks, copy records over to the new block. - */ - else { - xfs_alloc_rec_t *lrp; /* left btree record pointer */ - xfs_alloc_rec_t *rrp; /* right btree record pointer */ +} +#endif /* DEBUG */ - lrp = XFS_ALLOC_REC_ADDR(left, i, cur); - rrp = XFS_ALLOC_REC_ADDR(right, 1, cur); - memcpy(rrp, lrp, be16_to_cpu(right->bb_numrecs) * sizeof(*rrp)); - xfs_alloc_log_recs(cur, rbp, 1, be16_to_cpu(right->bb_numrecs)); - keyp->ar_startblock = rrp->ar_startblock; - keyp->ar_blockcount = rrp->ar_blockcount; - } - /* - * Find the left block number by looking in the buffer. - * Adjust numrecs, sibling pointers. - */ - lbno = XFS_DADDR_TO_AGBNO(cur->bc_mp, XFS_BUF_ADDR(lbp)); - be16_add_cpu(&left->bb_numrecs, -(be16_to_cpu(right->bb_numrecs))); - right->bb_rightsib = left->bb_rightsib; - left->bb_rightsib = cpu_to_be32(rbno); - right->bb_leftsib = cpu_to_be32(lbno); - xfs_alloc_log_block(cur->bc_tp, rbp, XFS_BB_ALL_BITS); - xfs_alloc_log_block(cur->bc_tp, lbp, XFS_BB_NUMRECS | XFS_BB_RIGHTSIB); - /* - * If there's a block to the new block's right, make that block - * point back to right instead of to left. - */ - if (be32_to_cpu(right->bb_rightsib) != NULLAGBLOCK) { - xfs_alloc_block_t *rrblock; /* rr btree block */ - xfs_buf_t *rrbp; /* buffer for rrblock */ +#ifdef XFS_BTREE_TRACE +ktrace_t *xfs_allocbt_trace_buf; - if ((error = xfs_btree_read_bufs(cur->bc_mp, cur->bc_tp, - cur->bc_private.a.agno, be32_to_cpu(right->bb_rightsib), 0, - &rrbp, XFS_ALLOC_BTREE_REF))) - return error; - rrblock = XFS_BUF_TO_ALLOC_BLOCK(rrbp); - if ((error = xfs_btree_check_sblock(cur, rrblock, level, rrbp))) - return error; - rrblock->bb_leftsib = cpu_to_be32(rbno); - xfs_alloc_log_block(cur->bc_tp, rrbp, XFS_BB_LEFTSIB); - } - /* - * If the cursor is really in the right block, move it there. - * If it's just pointing past the last entry in left, then we'll - * insert there, so don't change anything in that case. - */ - if (cur->bc_ptrs[level] > be16_to_cpu(left->bb_numrecs) + 1) { - xfs_btree_setbuf(cur, level, rbp); - cur->bc_ptrs[level] -= be16_to_cpu(left->bb_numrecs); - } - /* - * If there are more levels, we'll need another cursor which refers to - * the right block, no matter where this cursor was. - */ - if (level + 1 < cur->bc_nlevels) { - if ((error = xfs_btree_dup_cursor(cur, curp))) - return error; - (*curp)->bc_ptrs[level + 1]++; - } - *bnop = rbno; - *stat = 1; - return 0; +STATIC void +xfs_allocbt_trace_enter( + struct xfs_btree_cur *cur, + const char *func, + char *s, + int type, + int line, + __psunsigned_t a0, + __psunsigned_t a1, + __psunsigned_t a2, + __psunsigned_t a3, + __psunsigned_t a4, + __psunsigned_t a5, + __psunsigned_t a6, + __psunsigned_t a7, + __psunsigned_t a8, + __psunsigned_t a9, + __psunsigned_t a10) +{ + ktrace_enter(xfs_allocbt_trace_buf, (void *)(__psint_t)type, + (void *)func, (void *)s, NULL, (void *)cur, + (void *)a0, (void *)a1, (void *)a2, (void *)a3, + (void *)a4, (void *)a5, (void *)a6, (void *)a7, + (void *)a8, (void *)a9, (void *)a10); } -/* - * Update keys at all levels from here to the root along the cursor's path. - */ -STATIC int /* error */ -xfs_alloc_updkey( - xfs_btree_cur_t *cur, /* btree cursor */ - xfs_alloc_key_t *keyp, /* new key value to update to */ - int level) /* starting level for update */ +STATIC void +xfs_allocbt_trace_cursor( + struct xfs_btree_cur *cur, + __uint32_t *s0, + __uint64_t *l0, + __uint64_t *l1) { - int ptr; /* index of key in block */ - - /* - * Go up the tree from this level toward the root. - * At each level, update the key value to the value input. - * Stop when we reach a level where the cursor isn't pointing - * at the first entry in the block. - */ - for (ptr = 1; ptr == 1 && level < cur->bc_nlevels; level++) { - xfs_alloc_block_t *block; /* btree block */ - xfs_buf_t *bp; /* buffer for block */ -#ifdef DEBUG - int error; /* error return value */ -#endif - xfs_alloc_key_t *kp; /* ptr to btree block keys */ - - bp = cur->bc_bufs[level]; - block = XFS_BUF_TO_ALLOC_BLOCK(bp); -#ifdef DEBUG - if ((error = xfs_btree_check_sblock(cur, block, level, bp))) - return error; -#endif - ptr = cur->bc_ptrs[level]; - kp = XFS_ALLOC_KEY_ADDR(block, ptr, cur); - *kp = *keyp; - xfs_alloc_log_keys(cur, bp, ptr, ptr); - } - return 0; + *s0 = cur->bc_private.a.agno; + *l0 = cur->bc_rec.a.ar_startblock; + *l1 = cur->bc_rec.a.ar_blockcount; } -/* - * Externally visible routines. - */ - -/* - * Decrement cursor by one record at the level. - * For nonzero levels the leaf-ward information is untouched. - */ -int /* error */ -xfs_alloc_decrement( - xfs_btree_cur_t *cur, /* btree cursor */ - int level, /* level in btree, 0 is leaf */ - int *stat) /* success/failure */ +STATIC void +xfs_allocbt_trace_key( + struct xfs_btree_cur *cur, + union xfs_btree_key *key, + __uint64_t *l0, + __uint64_t *l1) { - xfs_alloc_block_t *block; /* btree block */ - int error; /* error return value */ - int lev; /* btree level */ - - ASSERT(level < cur->bc_nlevels); - /* - * Read-ahead to the left at this level. - */ - xfs_btree_readahead(cur, level, XFS_BTCUR_LEFTRA); - /* - * Decrement the ptr at this level. If we're still in the block - * then we're done. - */ - if (--cur->bc_ptrs[level] > 0) { - *stat = 1; - return 0; - } - /* - * Get a pointer to the btree block. - */ - block = XFS_BUF_TO_ALLOC_BLOCK(cur->bc_bufs[level]); -#ifdef DEBUG - if ((error = xfs_btree_check_sblock(cur, block, level, - cur->bc_bufs[level]))) - return error; -#endif - /* - * If we just went off the left edge of the tree, return failure. - */ - if (be32_to_cpu(block->bb_leftsib) == NULLAGBLOCK) { - *stat = 0; - return 0; - } - /* - * March up the tree decrementing pointers. - * Stop when we don't go off the left edge of a block. - */ - for (lev = level + 1; lev < cur->bc_nlevels; lev++) { - if (--cur->bc_ptrs[lev] > 0) - break; - /* - * Read-ahead the left block, we're going to read it - * in the next loop. - */ - xfs_btree_readahead(cur, lev, XFS_BTCUR_LEFTRA); - } - /* - * If we went off the root then we are seriously confused. - */ - ASSERT(lev < cur->bc_nlevels); - /* - * Now walk back down the tree, fixing up the cursor's buffer - * pointers and key numbers. - */ - for (block = XFS_BUF_TO_ALLOC_BLOCK(cur->bc_bufs[lev]); lev > level; ) { - xfs_agblock_t agbno; /* block number of btree block */ - xfs_buf_t *bp; /* buffer pointer for block */ - - agbno = be32_to_cpu(*XFS_ALLOC_PTR_ADDR(block, cur->bc_ptrs[lev], cur)); - if ((error = xfs_btree_read_bufs(cur->bc_mp, cur->bc_tp, - cur->bc_private.a.agno, agbno, 0, &bp, - XFS_ALLOC_BTREE_REF))) - return error; - lev--; - xfs_btree_setbuf(cur, lev, bp); - block = XFS_BUF_TO_ALLOC_BLOCK(bp); - if ((error = xfs_btree_check_sblock(cur, block, lev, bp))) - return error; - cur->bc_ptrs[lev] = be16_to_cpu(block->bb_numrecs); - } - *stat = 1; - return 0; + *l0 = be32_to_cpu(key->alloc.ar_startblock); + *l1 = be32_to_cpu(key->alloc.ar_blockcount); } -/* - * Delete the record pointed to by cur. - * The cursor refers to the place where the record was (could be inserted) - * when the operation returns. - */ -int /* error */ -xfs_alloc_delete( - xfs_btree_cur_t *cur, /* btree cursor */ - int *stat) /* success/failure */ +STATIC void +xfs_allocbt_trace_record( + struct xfs_btree_cur *cur, + union xfs_btree_rec *rec, + __uint64_t *l0, + __uint64_t *l1, + __uint64_t *l2) { - int error; /* error return value */ - int i; /* result code */ - int level; /* btree level */ - - /* - * Go up the tree, starting at leaf level. - * If 2 is returned then a join was done; go to the next level. - * Otherwise we are done. - */ - for (level = 0, i = 2; i == 2; level++) { - if ((error = xfs_alloc_delrec(cur, level, &i))) - return error; - } - if (i == 0) { - for (level = 1; level < cur->bc_nlevels; level++) { - if (cur->bc_ptrs[level] == 0) { - if ((error = xfs_alloc_decrement(cur, level, &i))) - return error; - break; - } - } - } - *stat = i; - return 0; + *l0 = be32_to_cpu(rec->alloc.ar_startblock); + *l1 = be32_to_cpu(rec->alloc.ar_blockcount); + *l2 = 0; } +#endif /* XFS_BTREE_TRACE */ + +static const struct xfs_btree_ops xfs_allocbt_ops = { + .rec_len = sizeof(xfs_alloc_rec_t), + .key_len = sizeof(xfs_alloc_key_t), + + .dup_cursor = xfs_allocbt_dup_cursor, + .set_root = xfs_allocbt_set_root, + .kill_root = xfs_allocbt_kill_root, + .alloc_block = xfs_allocbt_alloc_block, + .free_block = xfs_allocbt_free_block, + .update_lastrec = xfs_allocbt_update_lastrec, + .get_minrecs = xfs_allocbt_get_minrecs, + .get_maxrecs = xfs_allocbt_get_maxrecs, + .init_key_from_rec = xfs_allocbt_init_key_from_rec, + .init_rec_from_key = xfs_allocbt_init_rec_from_key, + .init_rec_from_cur = xfs_allocbt_init_rec_from_cur, + .init_ptr_from_cur = xfs_allocbt_init_ptr_from_cur, + .key_diff = xfs_allocbt_key_diff, -/* - * Get the data from the pointed-to record. - */ -int /* error */ -xfs_alloc_get_rec( - xfs_btree_cur_t *cur, /* btree cursor */ - xfs_agblock_t *bno, /* output: starting block of extent */ - xfs_extlen_t *len, /* output: length of extent */ - int *stat) /* output: success/failure */ -{ - xfs_alloc_block_t *block; /* btree block */ #ifdef DEBUG - int error; /* error return value */ + .keys_inorder = xfs_allocbt_keys_inorder, + .recs_inorder = xfs_allocbt_recs_inorder, #endif - int ptr; /* record number */ - ptr = cur->bc_ptrs[0]; - block = XFS_BUF_TO_ALLOC_BLOCK(cur->bc_bufs[0]); -#ifdef DEBUG - if ((error = xfs_btree_check_sblock(cur, block, 0, cur->bc_bufs[0]))) - return error; +#ifdef XFS_BTREE_TRACE + .trace_enter = xfs_allocbt_trace_enter, + .trace_cursor = xfs_allocbt_trace_cursor, + .trace_key = xfs_allocbt_trace_key, + .trace_record = xfs_allocbt_trace_record, #endif - /* - * Off the right end or left end, return failure. - */ - if (ptr > be16_to_cpu(block->bb_numrecs) || ptr <= 0) { - *stat = 0; - return 0; - } - /* - * Point to the record and extract its data. - */ - { - xfs_alloc_rec_t *rec; /* record data */ - - rec = XFS_ALLOC_REC_ADDR(block, ptr, cur); - *bno = be32_to_cpu(rec->ar_startblock); - *len = be32_to_cpu(rec->ar_blockcount); - } - *stat = 1; - return 0; -} +}; /* - * Increment cursor by one record at the level. - * For nonzero levels the leaf-ward information is untouched. + * Allocate a new allocation btree cursor. */ -int /* error */ -xfs_alloc_increment( - xfs_btree_cur_t *cur, /* btree cursor */ - int level, /* level in btree, 0 is leaf */ - int *stat) /* success/failure */ +struct xfs_btree_cur * /* new alloc btree cursor */ +xfs_allocbt_init_cursor( + struct xfs_mount *mp, /* file system mount point */ + struct xfs_trans *tp, /* transaction pointer */ + struct xfs_buf *agbp, /* buffer for agf structure */ + xfs_agnumber_t agno, /* allocation group number */ + xfs_btnum_t btnum) /* btree identifier */ { - xfs_alloc_block_t *block; /* btree block */ - xfs_buf_t *bp; /* tree block buffer */ - int error; /* error return value */ - int lev; /* btree level */ - - ASSERT(level < cur->bc_nlevels); - /* - * Read-ahead to the right at this level. - */ - xfs_btree_readahead(cur, level, XFS_BTCUR_RIGHTRA); - /* - * Get a pointer to the btree block. - */ - bp = cur->bc_bufs[level]; - block = XFS_BUF_TO_ALLOC_BLOCK(bp); -#ifdef DEBUG - if ((error = xfs_btree_check_sblock(cur, block, level, bp))) - return error; -#endif - /* - * Increment the ptr at this level. If we're still in the block - * then we're done. - */ - if (++cur->bc_ptrs[level] <= be16_to_cpu(block->bb_numrecs)) { - *stat = 1; - return 0; - } - /* - * If we just went off the right edge of the tree, return failure. - */ - if (be32_to_cpu(block->bb_rightsib) == NULLAGBLOCK) { - *stat = 0; - return 0; - } - /* - * March up the tree incrementing pointers. - * Stop when we don't go off the right edge of a block. - */ - for (lev = level + 1; lev < cur->bc_nlevels; lev++) { - bp = cur->bc_bufs[lev]; - block = XFS_BUF_TO_ALLOC_BLOCK(bp); -#ifdef DEBUG - if ((error = xfs_btree_check_sblock(cur, block, lev, bp))) - return error; -#endif - if (++cur->bc_ptrs[lev] <= be16_to_cpu(block->bb_numrecs)) - break; - /* - * Read-ahead the right block, we're going to read it - * in the next loop. - */ - xfs_btree_readahead(cur, lev, XFS_BTCUR_RIGHTRA); - } - /* - * If we went off the root then we are seriously confused. - */ - ASSERT(lev < cur->bc_nlevels); - /* - * Now walk back down the tree, fixing up the cursor's buffer - * pointers and key numbers. - */ - for (bp = cur->bc_bufs[lev], block = XFS_BUF_TO_ALLOC_BLOCK(bp); - lev > level; ) { - xfs_agblock_t agbno; /* block number of btree block */ + struct xfs_agf *agf = XFS_BUF_TO_AGF(agbp); + struct xfs_btree_cur *cur; - agbno = be32_to_cpu(*XFS_ALLOC_PTR_ADDR(block, cur->bc_ptrs[lev], cur)); - if ((error = xfs_btree_read_bufs(cur->bc_mp, cur->bc_tp, - cur->bc_private.a.agno, agbno, 0, &bp, - XFS_ALLOC_BTREE_REF))) - return error; - lev--; - xfs_btree_setbuf(cur, lev, bp); - block = XFS_BUF_TO_ALLOC_BLOCK(bp); - if ((error = xfs_btree_check_sblock(cur, block, lev, bp))) - return error; - cur->bc_ptrs[lev] = 1; - } - *stat = 1; - return 0; -} + ASSERT(btnum == XFS_BTNUM_BNO || btnum == XFS_BTNUM_CNT); -/* - * Insert the current record at the point referenced by cur. - * The cursor may be inconsistent on return if splits have been done. - */ -int /* error */ -xfs_alloc_insert( - xfs_btree_cur_t *cur, /* btree cursor */ - int *stat) /* success/failure */ -{ - int error; /* error return value */ - int i; /* result value, 0 for failure */ - int level; /* current level number in btree */ - xfs_agblock_t nbno; /* new block number (split result) */ - xfs_btree_cur_t *ncur; /* new cursor (split result) */ - xfs_alloc_rec_t nrec; /* record being inserted this level */ - xfs_btree_cur_t *pcur; /* previous level's cursor */ + cur = kmem_zone_zalloc(xfs_btree_cur_zone, KM_SLEEP); - level = 0; - nbno = NULLAGBLOCK; - nrec.ar_startblock = cpu_to_be32(cur->bc_rec.a.ar_startblock); - nrec.ar_blockcount = cpu_to_be32(cur->bc_rec.a.ar_blockcount); - ncur = NULL; - pcur = cur; - /* - * Loop going up the tree, starting at the leaf level. - * Stop when we don't get a split block, that must mean that - * the insert is finished with this level. - */ - do { - /* - * Insert nrec/nbno into this level of the tree. - * Note if we fail, nbno will be null. - */ - if ((error = xfs_alloc_insrec(pcur, level++, &nbno, &nrec, &ncur, - &i))) { - if (pcur != cur) - xfs_btree_del_cursor(pcur, XFS_BTREE_ERROR); - return error; - } - /* - * See if the cursor we just used is trash. - * Can't trash the caller's cursor, but otherwise we should - * if ncur is a new cursor or we're about to be done. - */ - if (pcur != cur && (ncur || nbno == NULLAGBLOCK)) { - cur->bc_nlevels = pcur->bc_nlevels; - xfs_btree_del_cursor(pcur, XFS_BTREE_NOERROR); - } - /* - * If we got a new cursor, switch to it. - */ - if (ncur) { - pcur = ncur; - ncur = NULL; - } - } while (nbno != NULLAGBLOCK); - *stat = i; - return 0; -} + cur->bc_tp = tp; + cur->bc_mp = mp; + cur->bc_nlevels = be32_to_cpu(agf->agf_levels[btnum]); + cur->bc_btnum = btnum; + cur->bc_blocklog = mp->m_sb.sb_blocklog; -/* - * Lookup the record equal to [bno, len] in the btree given by cur. - */ -int /* error */ -xfs_alloc_lookup_eq( - xfs_btree_cur_t *cur, /* btree cursor */ - xfs_agblock_t bno, /* starting block of extent */ - xfs_extlen_t len, /* length of extent */ - int *stat) /* success/failure */ -{ - cur->bc_rec.a.ar_startblock = bno; - cur->bc_rec.a.ar_blockcount = len; - return xfs_alloc_lookup(cur, XFS_LOOKUP_EQ, stat); -} + cur->bc_ops = &xfs_allocbt_ops; + if (btnum == XFS_BTNUM_CNT) + cur->bc_flags = XFS_BTREE_LASTREC_UPDATE; -/* - * Lookup the first record greater than or equal to [bno, len] - * in the btree given by cur. - */ -int /* error */ -xfs_alloc_lookup_ge( - xfs_btree_cur_t *cur, /* btree cursor */ - xfs_agblock_t bno, /* starting block of extent */ - xfs_extlen_t len, /* length of extent */ - int *stat) /* success/failure */ -{ - cur->bc_rec.a.ar_startblock = bno; - cur->bc_rec.a.ar_blockcount = len; - return xfs_alloc_lookup(cur, XFS_LOOKUP_GE, stat); -} + cur->bc_private.a.agbp = agbp; + cur->bc_private.a.agno = agno; -/* - * Lookup the first record less than or equal to [bno, len] - * in the btree given by cur. - */ -int /* error */ -xfs_alloc_lookup_le( - xfs_btree_cur_t *cur, /* btree cursor */ - xfs_agblock_t bno, /* starting block of extent */ - xfs_extlen_t len, /* length of extent */ - int *stat) /* success/failure */ -{ - cur->bc_rec.a.ar_startblock = bno; - cur->bc_rec.a.ar_blockcount = len; - return xfs_alloc_lookup(cur, XFS_LOOKUP_LE, stat); + return cur; } /* - * Update the record referred to by cur, to the value given by [bno, len]. - * This either works (return 0) or gets an EFSCORRUPTED error. + * Calculate number of records in an alloc btree block. */ -int /* error */ -xfs_alloc_update( - xfs_btree_cur_t *cur, /* btree cursor */ - xfs_agblock_t bno, /* starting block of extent */ - xfs_extlen_t len) /* length of extent */ +int +xfs_allocbt_maxrecs( + struct xfs_mount *mp, + int blocklen, + int leaf) { - xfs_alloc_block_t *block; /* btree block to update */ - int error; /* error return value */ - int ptr; /* current record number (updating) */ + blocklen -= XFS_ALLOC_BLOCK_LEN(mp); - ASSERT(len > 0); - /* - * Pick up the a.g. freelist struct and the current block. - */ - block = XFS_BUF_TO_ALLOC_BLOCK(cur->bc_bufs[0]); -#ifdef DEBUG - if ((error = xfs_btree_check_sblock(cur, block, 0, cur->bc_bufs[0]))) - return error; -#endif - /* - * Get the address of the rec to be updated. - */ - ptr = cur->bc_ptrs[0]; - { - xfs_alloc_rec_t *rp; /* pointer to updated record */ - - rp = XFS_ALLOC_REC_ADDR(block, ptr, cur); - /* - * Fill in the new contents and log them. - */ - rp->ar_startblock = cpu_to_be32(bno); - rp->ar_blockcount = cpu_to_be32(len); - xfs_alloc_log_recs(cur, cur->bc_bufs[0], ptr, ptr); - } - /* - * If it's the by-size btree and it's the last leaf block and - * it's the last record... then update the size of the longest - * extent in the a.g., which we cache in the a.g. freelist header. - */ - if (cur->bc_btnum == XFS_BTNUM_CNT && - be32_to_cpu(block->bb_rightsib) == NULLAGBLOCK && - ptr == be16_to_cpu(block->bb_numrecs)) { - xfs_agf_t *agf; /* a.g. freespace header */ - xfs_agnumber_t seqno; - - agf = XFS_BUF_TO_AGF(cur->bc_private.a.agbp); - seqno = be32_to_cpu(agf->agf_seqno); - cur->bc_mp->m_perag[seqno].pagf_longest = len; - agf->agf_longest = cpu_to_be32(len); - xfs_alloc_log_agf(cur->bc_tp, cur->bc_private.a.agbp, - XFS_AGF_LONGEST); - } - /* - * Updating first record in leaf. Pass new key value up to our parent. - */ - if (ptr == 1) { - xfs_alloc_key_t key; /* key containing [bno, len] */ - - key.ar_startblock = cpu_to_be32(bno); - key.ar_blockcount = cpu_to_be32(len); - if ((error = xfs_alloc_updkey(cur, &key, 1))) - return error; - } - return 0; + if (leaf) + return blocklen / sizeof(xfs_alloc_rec_t); + return blocklen / (sizeof(xfs_alloc_key_t) + sizeof(xfs_alloc_ptr_t)); } diff --git a/fs/xfs/xfs_alloc_btree.h b/fs/xfs/xfs_alloc_btree.h index 5bd1a2c8bd0..a6caa0022c9 100644 --- a/fs/xfs/xfs_alloc_btree.h +++ b/fs/xfs/xfs_alloc_btree.h @@ -24,7 +24,6 @@ struct xfs_buf; struct xfs_btree_cur; -struct xfs_btree_sblock; struct xfs_mount; /* @@ -50,16 +49,6 @@ typedef struct xfs_alloc_rec_incore { /* btree pointer type */ typedef __be32 xfs_alloc_ptr_t; -/* btree block header type */ -typedef struct xfs_btree_sblock xfs_alloc_block_t; - -#define XFS_BUF_TO_ALLOC_BLOCK(bp) ((xfs_alloc_block_t *)XFS_BUF_PTR(bp)) - -/* - * Real block structures have a size equal to the disk block size. - */ -#define XFS_ALLOC_BLOCK_MAXRECS(lev,cur) ((cur)->bc_mp->m_alloc_mxr[lev != 0]) -#define XFS_ALLOC_BLOCK_MINRECS(lev,cur) ((cur)->bc_mp->m_alloc_mnr[lev != 0]) /* * Minimum and maximum blocksize and sectorsize. @@ -83,73 +72,39 @@ typedef struct xfs_btree_sblock xfs_alloc_block_t; #define XFS_CNT_BLOCK(mp) ((xfs_agblock_t)(XFS_BNO_BLOCK(mp) + 1)) /* - * Record, key, and pointer address macros for btree blocks. - */ -#define XFS_ALLOC_REC_ADDR(bb,i,cur) \ - XFS_BTREE_REC_ADDR(xfs_alloc, bb, i) - -#define XFS_ALLOC_KEY_ADDR(bb,i,cur) \ - XFS_BTREE_KEY_ADDR(xfs_alloc, bb, i) - -#define XFS_ALLOC_PTR_ADDR(bb,i,cur) \ - XFS_BTREE_PTR_ADDR(xfs_alloc, bb, i, XFS_ALLOC_BLOCK_MAXRECS(1, cur)) - -/* - * Decrement cursor by one record at the level. - * For nonzero levels the leaf-ward information is untouched. - */ -extern int xfs_alloc_decrement(struct xfs_btree_cur *cur, int level, int *stat); - -/* - * Delete the record pointed to by cur. - * The cursor refers to the place where the record was (could be inserted) - * when the operation returns. - */ -extern int xfs_alloc_delete(struct xfs_btree_cur *cur, int *stat); - -/* - * Get the data from the pointed-to record. - */ -extern int xfs_alloc_get_rec(struct xfs_btree_cur *cur, xfs_agblock_t *bno, - xfs_extlen_t *len, int *stat); - -/* - * Increment cursor by one record at the level. - * For nonzero levels the leaf-ward information is untouched. - */ -extern int xfs_alloc_increment(struct xfs_btree_cur *cur, int level, int *stat); - -/* - * Insert the current record at the point referenced by cur. - * The cursor may be inconsistent on return if splits have been done. - */ -extern int xfs_alloc_insert(struct xfs_btree_cur *cur, int *stat); - -/* - * Lookup the record equal to [bno, len] in the btree given by cur. - */ -extern int xfs_alloc_lookup_eq(struct xfs_btree_cur *cur, xfs_agblock_t bno, - xfs_extlen_t len, int *stat); - -/* - * Lookup the first record greater than or equal to [bno, len] - * in the btree given by cur. - */ -extern int xfs_alloc_lookup_ge(struct xfs_btree_cur *cur, xfs_agblock_t bno, - xfs_extlen_t len, int *stat); - -/* - * Lookup the first record less than or equal to [bno, len] - * in the btree given by cur. + * Btree block header size depends on a superblock flag. + * + * (not quite yet, but soon) */ -extern int xfs_alloc_lookup_le(struct xfs_btree_cur *cur, xfs_agblock_t bno, - xfs_extlen_t len, int *stat); +#define XFS_ALLOC_BLOCK_LEN(mp) XFS_BTREE_SBLOCK_LEN /* - * Update the record referred to by cur, to the value given by [bno, len]. - * This either works (return 0) or gets an EFSCORRUPTED error. - */ -extern int xfs_alloc_update(struct xfs_btree_cur *cur, xfs_agblock_t bno, - xfs_extlen_t len); + * Record, key, and pointer address macros for btree blocks. + * + * (note that some of these may appear unused, but they are used in userspace) + */ +#define XFS_ALLOC_REC_ADDR(mp, block, index) \ + ((xfs_alloc_rec_t *) \ + ((char *)(block) + \ + XFS_ALLOC_BLOCK_LEN(mp) + \ + (((index) - 1) * sizeof(xfs_alloc_rec_t)))) + +#define XFS_ALLOC_KEY_ADDR(mp, block, index) \ + ((xfs_alloc_key_t *) \ + ((char *)(block) + \ + XFS_ALLOC_BLOCK_LEN(mp) + \ + ((index) - 1) * sizeof(xfs_alloc_key_t))) + +#define XFS_ALLOC_PTR_ADDR(mp, block, index, maxrecs) \ + ((xfs_alloc_ptr_t *) \ + ((char *)(block) + \ + XFS_ALLOC_BLOCK_LEN(mp) + \ + (maxrecs) * sizeof(xfs_alloc_key_t) + \ + ((index) - 1) * sizeof(xfs_alloc_ptr_t))) + +extern struct xfs_btree_cur *xfs_allocbt_init_cursor(struct xfs_mount *, + struct xfs_trans *, struct xfs_buf *, + xfs_agnumber_t, xfs_btnum_t); +extern int xfs_allocbt_maxrecs(struct xfs_mount *, int, int); #endif /* __XFS_ALLOC_BTREE_H__ */ diff --git a/fs/xfs/xfs_arch.h b/fs/xfs/xfs_arch.h index 0b3b5efe848..53d5e70d136 100644 --- a/fs/xfs/xfs_arch.h +++ b/fs/xfs/xfs_arch.h @@ -41,21 +41,36 @@ #endif #ifdef XFS_NATIVE_HOST -#define cpu_to_be16(val) ((__be16)(val)) -#define cpu_to_be32(val) ((__be32)(val)) -#define cpu_to_be64(val) ((__be64)(val)) -#define be16_to_cpu(val) ((__uint16_t)(val)) -#define be32_to_cpu(val) ((__uint32_t)(val)) -#define be64_to_cpu(val) ((__uint64_t)(val)) +#define cpu_to_be16(val) ((__force __be16)(__u16)(val)) +#define cpu_to_be32(val) ((__force __be32)(__u32)(val)) +#define cpu_to_be64(val) ((__force __be64)(__u64)(val)) +#define be16_to_cpu(val) ((__force __u16)(__be16)(val)) +#define be32_to_cpu(val) ((__force __u32)(__be32)(val)) +#define be64_to_cpu(val) ((__force __u64)(__be64)(val)) #else -#define cpu_to_be16(val) (__swab16((__uint16_t)(val))) -#define cpu_to_be32(val) (__swab32((__uint32_t)(val))) -#define cpu_to_be64(val) (__swab64((__uint64_t)(val))) -#define be16_to_cpu(val) (__swab16((__be16)(val))) -#define be32_to_cpu(val) (__swab32((__be32)(val))) -#define be64_to_cpu(val) (__swab64((__be64)(val))) +#define cpu_to_be16(val) ((__force __be16)__swab16((__u16)(val))) +#define cpu_to_be32(val) ((__force __be32)__swab32((__u32)(val))) +#define cpu_to_be64(val) ((__force __be64)__swab64((__u64)(val))) +#define be16_to_cpu(val) (__swab16((__force __u16)(__be16)(val))) +#define be32_to_cpu(val) (__swab32((__force __u32)(__be32)(val))) +#define be64_to_cpu(val) (__swab64((__force __u64)(__be64)(val))) #endif +static inline void be16_add_cpu(__be16 *a, __s16 b) +{ + *a = cpu_to_be16(be16_to_cpu(*a) + b); +} + +static inline void be32_add_cpu(__be32 *a, __s32 b) +{ + *a = cpu_to_be32(be32_to_cpu(*a) + b); +} + +static inline void be64_add_cpu(__be64 *a, __s64 b) +{ + *a = cpu_to_be64(be64_to_cpu(*a) + b); +} + #endif /* __KERNEL__ */ /* do we need conversion? */ diff --git a/fs/xfs/xfs_bit.h b/fs/xfs/xfs_bit.h index 8e0e463dae2..bca7b243c31 100644 --- a/fs/xfs/xfs_bit.h +++ b/fs/xfs/xfs_bit.h @@ -61,8 +61,7 @@ static inline int xfs_highbit64(__uint64_t v) /* Get low bit set out of 32-bit argument, -1 if none set */ static inline int xfs_lowbit32(__uint32_t v) { - unsigned long t = v; - return (v) ? find_first_bit(&t, 32) : -1; + return ffs(v) - 1; } /* Get low bit set out of 64-bit argument, -1 if none set */ diff --git a/fs/xfs/xfs_bmap.c b/fs/xfs/xfs_bmap.c index a1aab9275d5..138308e70d1 100644 --- a/fs/xfs/xfs_bmap.c +++ b/fs/xfs/xfs_bmap.c @@ -393,8 +393,8 @@ xfs_bmap_count_leaves( STATIC void xfs_bmap_disk_count_leaves( - xfs_extnum_t idx, - xfs_bmbt_block_t *block, + struct xfs_mount *mp, + struct xfs_btree_block *block, int numrecs, int *count); @@ -402,6 +402,53 @@ xfs_bmap_disk_count_leaves( * Bmap internal routines. */ +STATIC int /* error */ +xfs_bmbt_lookup_eq( + struct xfs_btree_cur *cur, + xfs_fileoff_t off, + xfs_fsblock_t bno, + xfs_filblks_t len, + int *stat) /* success/failure */ +{ + cur->bc_rec.b.br_startoff = off; + cur->bc_rec.b.br_startblock = bno; + cur->bc_rec.b.br_blockcount = len; + return xfs_btree_lookup(cur, XFS_LOOKUP_EQ, stat); +} + +STATIC int /* error */ +xfs_bmbt_lookup_ge( + struct xfs_btree_cur *cur, + xfs_fileoff_t off, + xfs_fsblock_t bno, + xfs_filblks_t len, + int *stat) /* success/failure */ +{ + cur->bc_rec.b.br_startoff = off; + cur->bc_rec.b.br_startblock = bno; + cur->bc_rec.b.br_blockcount = len; + return xfs_btree_lookup(cur, XFS_LOOKUP_GE, stat); +} + +/* +* Update the record referred to by cur to the value given + * by [off, bno, len, state]. + * This either works (return 0) or gets an EFSCORRUPTED error. + */ +STATIC int +xfs_bmbt_update( + struct xfs_btree_cur *cur, + xfs_fileoff_t off, + xfs_fsblock_t bno, + xfs_filblks_t len, + xfs_exntst_t state) +{ + union xfs_btree_rec rec; + + xfs_bmbt_disk_set_allf(&rec.bmbt, off, bno, len, state); + return xfs_btree_update(cur, &rec); +} + /* * Called from xfs_bmap_add_attrfork to handle btree format files. */ @@ -422,15 +469,14 @@ xfs_bmap_add_attrfork_btree( if (ip->i_df.if_broot_bytes <= XFS_IFORK_DSIZE(ip)) *flags |= XFS_ILOG_DBROOT; else { - cur = xfs_btree_init_cursor(mp, tp, NULL, 0, XFS_BTNUM_BMAP, ip, - XFS_DATA_FORK); + cur = xfs_bmbt_init_cursor(mp, tp, ip, XFS_DATA_FORK); cur->bc_private.b.flist = flist; cur->bc_private.b.firstblock = *firstblock; if ((error = xfs_bmbt_lookup_ge(cur, 0, 0, 0, &stat))) goto error0; /* must be at least one entry */ XFS_WANT_CORRUPTED_GOTO(stat == 1, error0); - if ((error = xfs_bmbt_newroot(cur, flags, &stat))) + if ((error = xfs_btree_new_iroot(cur, flags, &stat))) goto error0; if (stat == 0) { xfs_btree_del_cursor(cur, XFS_BTREE_NOERROR); @@ -818,10 +864,10 @@ xfs_bmap_add_extent_delay_real( RIGHT.br_blockcount, &i))) goto done; XFS_WANT_CORRUPTED_GOTO(i == 1, done); - if ((error = xfs_bmbt_delete(cur, &i))) + if ((error = xfs_btree_delete(cur, &i))) goto done; XFS_WANT_CORRUPTED_GOTO(i == 1, done); - if ((error = xfs_bmbt_decrement(cur, 0, &i))) + if ((error = xfs_btree_decrement(cur, 0, &i))) goto done; XFS_WANT_CORRUPTED_GOTO(i == 1, done); if ((error = xfs_bmbt_update(cur, LEFT.br_startoff, @@ -931,7 +977,7 @@ xfs_bmap_add_extent_delay_real( goto done; XFS_WANT_CORRUPTED_GOTO(i == 0, done); cur->bc_rec.b.br_state = XFS_EXT_NORM; - if ((error = xfs_bmbt_insert(cur, &i))) + if ((error = xfs_btree_insert(cur, &i))) goto done; XFS_WANT_CORRUPTED_GOTO(i == 1, done); } @@ -1007,7 +1053,7 @@ xfs_bmap_add_extent_delay_real( goto done; XFS_WANT_CORRUPTED_GOTO(i == 0, done); cur->bc_rec.b.br_state = XFS_EXT_NORM; - if ((error = xfs_bmbt_insert(cur, &i))) + if ((error = xfs_btree_insert(cur, &i))) goto done; XFS_WANT_CORRUPTED_GOTO(i == 1, done); } @@ -1097,7 +1143,7 @@ xfs_bmap_add_extent_delay_real( goto done; XFS_WANT_CORRUPTED_GOTO(i == 0, done); cur->bc_rec.b.br_state = XFS_EXT_NORM; - if ((error = xfs_bmbt_insert(cur, &i))) + if ((error = xfs_btree_insert(cur, &i))) goto done; XFS_WANT_CORRUPTED_GOTO(i == 1, done); } @@ -1152,7 +1198,7 @@ xfs_bmap_add_extent_delay_real( goto done; XFS_WANT_CORRUPTED_GOTO(i == 0, done); cur->bc_rec.b.br_state = XFS_EXT_NORM; - if ((error = xfs_bmbt_insert(cur, &i))) + if ((error = xfs_btree_insert(cur, &i))) goto done; XFS_WANT_CORRUPTED_GOTO(i == 1, done); } @@ -1379,16 +1425,16 @@ xfs_bmap_add_extent_unwritten_real( RIGHT.br_blockcount, &i))) goto done; XFS_WANT_CORRUPTED_GOTO(i == 1, done); - if ((error = xfs_bmbt_delete(cur, &i))) + if ((error = xfs_btree_delete(cur, &i))) goto done; XFS_WANT_CORRUPTED_GOTO(i == 1, done); - if ((error = xfs_bmbt_decrement(cur, 0, &i))) + if ((error = xfs_btree_decrement(cur, 0, &i))) goto done; XFS_WANT_CORRUPTED_GOTO(i == 1, done); - if ((error = xfs_bmbt_delete(cur, &i))) + if ((error = xfs_btree_delete(cur, &i))) goto done; XFS_WANT_CORRUPTED_GOTO(i == 1, done); - if ((error = xfs_bmbt_decrement(cur, 0, &i))) + if ((error = xfs_btree_decrement(cur, 0, &i))) goto done; XFS_WANT_CORRUPTED_GOTO(i == 1, done); if ((error = xfs_bmbt_update(cur, LEFT.br_startoff, @@ -1428,10 +1474,10 @@ xfs_bmap_add_extent_unwritten_real( &i))) goto done; XFS_WANT_CORRUPTED_GOTO(i == 1, done); - if ((error = xfs_bmbt_delete(cur, &i))) + if ((error = xfs_btree_delete(cur, &i))) goto done; XFS_WANT_CORRUPTED_GOTO(i == 1, done); - if ((error = xfs_bmbt_decrement(cur, 0, &i))) + if ((error = xfs_btree_decrement(cur, 0, &i))) goto done; XFS_WANT_CORRUPTED_GOTO(i == 1, done); if ((error = xfs_bmbt_update(cur, LEFT.br_startoff, @@ -1471,10 +1517,10 @@ xfs_bmap_add_extent_unwritten_real( RIGHT.br_blockcount, &i))) goto done; XFS_WANT_CORRUPTED_GOTO(i == 1, done); - if ((error = xfs_bmbt_delete(cur, &i))) + if ((error = xfs_btree_delete(cur, &i))) goto done; XFS_WANT_CORRUPTED_GOTO(i == 1, done); - if ((error = xfs_bmbt_decrement(cur, 0, &i))) + if ((error = xfs_btree_decrement(cur, 0, &i))) goto done; XFS_WANT_CORRUPTED_GOTO(i == 1, done); if ((error = xfs_bmbt_update(cur, new->br_startoff, @@ -1557,7 +1603,7 @@ xfs_bmap_add_extent_unwritten_real( PREV.br_blockcount - new->br_blockcount, oldext))) goto done; - if ((error = xfs_bmbt_decrement(cur, 0, &i))) + if ((error = xfs_btree_decrement(cur, 0, &i))) goto done; if (xfs_bmbt_update(cur, LEFT.br_startoff, LEFT.br_startblock, @@ -1605,7 +1651,7 @@ xfs_bmap_add_extent_unwritten_real( oldext))) goto done; cur->bc_rec.b = *new; - if ((error = xfs_bmbt_insert(cur, &i))) + if ((error = xfs_btree_insert(cur, &i))) goto done; XFS_WANT_CORRUPTED_GOTO(i == 1, done); } @@ -1647,7 +1693,7 @@ xfs_bmap_add_extent_unwritten_real( PREV.br_blockcount - new->br_blockcount, oldext))) goto done; - if ((error = xfs_bmbt_increment(cur, 0, &i))) + if ((error = xfs_btree_increment(cur, 0, &i))) goto done; if ((error = xfs_bmbt_update(cur, new->br_startoff, new->br_startblock, @@ -1695,7 +1741,7 @@ xfs_bmap_add_extent_unwritten_real( goto done; XFS_WANT_CORRUPTED_GOTO(i == 0, done); cur->bc_rec.b.br_state = XFS_EXT_NORM; - if ((error = xfs_bmbt_insert(cur, &i))) + if ((error = xfs_btree_insert(cur, &i))) goto done; XFS_WANT_CORRUPTED_GOTO(i == 1, done); } @@ -1743,7 +1789,7 @@ xfs_bmap_add_extent_unwritten_real( cur->bc_rec.b = PREV; cur->bc_rec.b.br_blockcount = new->br_startoff - PREV.br_startoff; - if ((error = xfs_bmbt_insert(cur, &i))) + if ((error = xfs_btree_insert(cur, &i))) goto done; XFS_WANT_CORRUPTED_GOTO(i == 1, done); /* @@ -1758,7 +1804,7 @@ xfs_bmap_add_extent_unwritten_real( XFS_WANT_CORRUPTED_GOTO(i == 0, done); /* new middle extent - newext */ cur->bc_rec.b.br_state = new->br_state; - if ((error = xfs_bmbt_insert(cur, &i))) + if ((error = xfs_btree_insert(cur, &i))) goto done; XFS_WANT_CORRUPTED_GOTO(i == 1, done); } @@ -2106,10 +2152,10 @@ xfs_bmap_add_extent_hole_real( right.br_blockcount, &i))) goto done; XFS_WANT_CORRUPTED_GOTO(i == 1, done); - if ((error = xfs_bmbt_delete(cur, &i))) + if ((error = xfs_btree_delete(cur, &i))) goto done; XFS_WANT_CORRUPTED_GOTO(i == 1, done); - if ((error = xfs_bmbt_decrement(cur, 0, &i))) + if ((error = xfs_btree_decrement(cur, 0, &i))) goto done; XFS_WANT_CORRUPTED_GOTO(i == 1, done); if ((error = xfs_bmbt_update(cur, left.br_startoff, @@ -2218,7 +2264,7 @@ xfs_bmap_add_extent_hole_real( goto done; XFS_WANT_CORRUPTED_GOTO(i == 0, done); cur->bc_rec.b.br_state = new->br_state; - if ((error = xfs_bmbt_insert(cur, &i))) + if ((error = xfs_btree_insert(cur, &i))) goto done; XFS_WANT_CORRUPTED_GOTO(i == 1, done); } @@ -2996,24 +3042,24 @@ xfs_bmap_btree_to_extents( int whichfork) /* data or attr fork */ { /* REFERENCED */ - xfs_bmbt_block_t *cblock;/* child btree block */ + struct xfs_btree_block *cblock;/* child btree block */ xfs_fsblock_t cbno; /* child block number */ xfs_buf_t *cbp; /* child block's buffer */ int error; /* error return value */ xfs_ifork_t *ifp; /* inode fork data */ xfs_mount_t *mp; /* mount point structure */ __be64 *pp; /* ptr to block address */ - xfs_bmbt_block_t *rblock;/* root btree block */ + struct xfs_btree_block *rblock;/* root btree block */ + mp = ip->i_mount; ifp = XFS_IFORK_PTR(ip, whichfork); ASSERT(ifp->if_flags & XFS_IFEXTENTS); ASSERT(XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_BTREE); rblock = ifp->if_broot; ASSERT(be16_to_cpu(rblock->bb_level) == 1); ASSERT(be16_to_cpu(rblock->bb_numrecs) == 1); - ASSERT(XFS_BMAP_BROOT_MAXRECS(ifp->if_broot_bytes) == 1); - mp = ip->i_mount; - pp = XFS_BMAP_BROOT_PTR_ADDR(rblock, 1, ifp->if_broot_bytes); + ASSERT(xfs_bmbt_maxrecs(mp, ifp->if_broot_bytes, 0) == 1); + pp = XFS_BMAP_BROOT_PTR_ADDR(mp, rblock, 1, ifp->if_broot_bytes); cbno = be64_to_cpu(*pp); *logflagsp = 0; #ifdef DEBUG @@ -3023,8 +3069,8 @@ xfs_bmap_btree_to_extents( if ((error = xfs_btree_read_bufl(mp, tp, cbno, 0, &cbp, XFS_BMAP_BTREE_REF))) return error; - cblock = XFS_BUF_TO_BMBT_BLOCK(cbp); - if ((error = xfs_btree_check_lblock(cur, cblock, 0, cbp))) + cblock = XFS_BUF_TO_BLOCK(cbp); + if ((error = xfs_btree_check_block(cur, cblock, 0, cbp))) return error; xfs_bmap_add_free(cbno, 1, cur->bc_private.b.flist, mp); ip->i_d.di_nblocks--; @@ -3170,7 +3216,7 @@ xfs_bmap_del_extent( flags |= XFS_ILOG_FEXT(whichfork); break; } - if ((error = xfs_bmbt_delete(cur, &i))) + if ((error = xfs_btree_delete(cur, &i))) goto done; XFS_WANT_CORRUPTED_GOTO(i == 1, done); break; @@ -3254,10 +3300,10 @@ xfs_bmap_del_extent( got.br_startblock, temp, got.br_state))) goto done; - if ((error = xfs_bmbt_increment(cur, 0, &i))) + if ((error = xfs_btree_increment(cur, 0, &i))) goto done; cur->bc_rec.b = new; - error = xfs_bmbt_insert(cur, &i); + error = xfs_btree_insert(cur, &i); if (error && error != ENOSPC) goto done; /* @@ -3404,11 +3450,11 @@ xfs_bmap_extents_to_btree( int *logflagsp, /* inode logging flags */ int whichfork) /* data or attr fork */ { - xfs_bmbt_block_t *ablock; /* allocated (child) bt block */ + struct xfs_btree_block *ablock; /* allocated (child) bt block */ xfs_buf_t *abp; /* buffer for ablock */ xfs_alloc_arg_t args; /* allocation arguments */ xfs_bmbt_rec_t *arp; /* child record pointer */ - xfs_bmbt_block_t *block; /* btree root block */ + struct xfs_btree_block *block; /* btree root block */ xfs_btree_cur_t *cur; /* bmap btree cursor */ xfs_bmbt_rec_host_t *ep; /* extent record pointer */ int error; /* error return value */ @@ -3428,6 +3474,7 @@ xfs_bmap_extents_to_btree( */ xfs_iroot_realloc(ip, 1, whichfork); ifp->if_flags |= XFS_IFBROOT; + /* * Fill in the root. */ @@ -3435,14 +3482,14 @@ xfs_bmap_extents_to_btree( block->bb_magic = cpu_to_be32(XFS_BMAP_MAGIC); block->bb_level = cpu_to_be16(1); block->bb_numrecs = cpu_to_be16(1); - block->bb_leftsib = cpu_to_be64(NULLDFSBNO); - block->bb_rightsib = cpu_to_be64(NULLDFSBNO); + block->bb_u.l.bb_leftsib = cpu_to_be64(NULLDFSBNO); + block->bb_u.l.bb_rightsib = cpu_to_be64(NULLDFSBNO); + /* * Need a cursor. Can't allocate until bb_level is filled in. */ mp = ip->i_mount; - cur = xfs_btree_init_cursor(mp, tp, NULL, 0, XFS_BTNUM_BMAP, ip, - whichfork); + cur = xfs_bmbt_init_cursor(mp, tp, ip, whichfork); cur->bc_private.b.firstblock = *firstblock; cur->bc_private.b.flist = flist; cur->bc_private.b.flags = wasdel ? XFS_BTCUR_BPRV_WASDEL : 0; @@ -3489,12 +3536,12 @@ xfs_bmap_extents_to_btree( /* * Fill in the child block. */ - ablock = XFS_BUF_TO_BMBT_BLOCK(abp); + ablock = XFS_BUF_TO_BLOCK(abp); ablock->bb_magic = cpu_to_be32(XFS_BMAP_MAGIC); ablock->bb_level = 0; - ablock->bb_leftsib = cpu_to_be64(NULLDFSBNO); - ablock->bb_rightsib = cpu_to_be64(NULLDFSBNO); - arp = XFS_BMAP_REC_IADDR(ablock, 1, cur); + ablock->bb_u.l.bb_leftsib = cpu_to_be64(NULLDFSBNO); + ablock->bb_u.l.bb_rightsib = cpu_to_be64(NULLDFSBNO); + arp = XFS_BMBT_REC_ADDR(mp, ablock, 1); nextents = ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t); for (cnt = i = 0; i < nextents; i++) { ep = xfs_iext_get_ext(ifp, i); @@ -3505,21 +3552,24 @@ xfs_bmap_extents_to_btree( } } ASSERT(cnt == XFS_IFORK_NEXTENTS(ip, whichfork)); - ablock->bb_numrecs = cpu_to_be16(cnt); + xfs_btree_set_numrecs(ablock, cnt); + /* * Fill in the root key and pointer. */ - kp = XFS_BMAP_KEY_IADDR(block, 1, cur); - arp = XFS_BMAP_REC_IADDR(ablock, 1, cur); + kp = XFS_BMBT_KEY_ADDR(mp, block, 1); + arp = XFS_BMBT_REC_ADDR(mp, ablock, 1); kp->br_startoff = cpu_to_be64(xfs_bmbt_disk_get_startoff(arp)); - pp = XFS_BMAP_PTR_IADDR(block, 1, cur); + pp = XFS_BMBT_PTR_ADDR(mp, block, 1, xfs_bmbt_get_maxrecs(cur, + be16_to_cpu(block->bb_level))); *pp = cpu_to_be64(args.fsbno); + /* * Do all this logging at the end so that * the root is at the right level. */ - xfs_bmbt_log_block(cur, abp, XFS_BB_ALL_BITS); - xfs_bmbt_log_recs(cur, abp, 1, be16_to_cpu(ablock->bb_numrecs)); + xfs_btree_log_block(cur, abp, XFS_BB_ALL_BITS); + xfs_btree_log_recs(cur, abp, 1, be16_to_cpu(ablock->bb_numrecs)); ASSERT(*curp == NULL); *curp = cur; *logflagsp = XFS_ILOG_CORE | XFS_ILOG_FBROOT(whichfork); @@ -4176,7 +4226,7 @@ xfs_bmap_compute_maxlevels( maxleafents = MAXAEXTNUM; sz = XFS_BMDR_SPACE_CALC(MINABTPTRS); } - maxrootrecs = (int)XFS_BTREE_BLOCK_MAXRECS(sz, xfs_bmdr, 0); + maxrootrecs = xfs_bmdr_maxrecs(mp, sz, 0); minleafrecs = mp->m_bmap_dmnr[0]; minnoderecs = mp->m_bmap_dmnr[1]; maxblocks = (maxleafents + minleafrecs - 1) / minleafrecs; @@ -4242,9 +4292,15 @@ xfs_bmap_finish( * We have a new transaction, so we should return committed=1, * even though we're returning an error. */ - if (error) { + if (error) return error; - } + + /* + * transaction commit worked ok so we can drop the extra ticket + * reference that we gained in xfs_trans_dup() + */ + xfs_log_ticket_put(ntp->t_ticket); + if ((error = xfs_trans_reserve(ntp, 0, logres, 0, XFS_TRANS_PERM_LOG_RES, logcount))) return error; @@ -4474,6 +4530,22 @@ xfs_bmap_one_block( return rval; } +STATIC int +xfs_bmap_sanity_check( + struct xfs_mount *mp, + struct xfs_buf *bp, + int level) +{ + struct xfs_btree_block *block = XFS_BUF_TO_BLOCK(bp); + + if (be32_to_cpu(block->bb_magic) != XFS_BMAP_MAGIC || + be16_to_cpu(block->bb_level) != level || + be16_to_cpu(block->bb_numrecs) == 0 || + be16_to_cpu(block->bb_numrecs) > mp->m_bmap_dmxr[level != 0]) + return 0; + return 1; +} + /* * Read in the extents to if_extents. * All inode fields are set up by caller, we just traverse the btree @@ -4486,7 +4558,7 @@ xfs_bmap_read_extents( xfs_inode_t *ip, /* incore inode */ int whichfork) /* data or attr fork */ { - xfs_bmbt_block_t *block; /* current btree block */ + struct xfs_btree_block *block; /* current btree block */ xfs_fsblock_t bno; /* block # of "block" */ xfs_buf_t *bp; /* buffer for "block" */ int error; /* error return value */ @@ -4510,7 +4582,7 @@ xfs_bmap_read_extents( */ level = be16_to_cpu(block->bb_level); ASSERT(level > 0); - pp = XFS_BMAP_BROOT_PTR_ADDR(block, 1, ifp->if_broot_bytes); + pp = XFS_BMAP_BROOT_PTR_ADDR(mp, block, 1, ifp->if_broot_bytes); bno = be64_to_cpu(*pp); ASSERT(bno != NULLDFSBNO); ASSERT(XFS_FSB_TO_AGNO(mp, bno) < mp->m_sb.sb_agcount); @@ -4523,13 +4595,13 @@ xfs_bmap_read_extents( if ((error = xfs_btree_read_bufl(mp, tp, bno, 0, &bp, XFS_BMAP_BTREE_REF))) return error; - block = XFS_BUF_TO_BMBT_BLOCK(bp); + block = XFS_BUF_TO_BLOCK(bp); XFS_WANT_CORRUPTED_GOTO( - XFS_BMAP_SANITY_CHECK(mp, block, level), + xfs_bmap_sanity_check(mp, bp, level), error0); if (level == 0) break; - pp = XFS_BTREE_PTR_ADDR(xfs_bmbt, block, 1, mp->m_bmap_dmxr[1]); + pp = XFS_BMBT_PTR_ADDR(mp, block, 1, mp->m_bmap_dmxr[1]); bno = be64_to_cpu(*pp); XFS_WANT_CORRUPTED_GOTO(XFS_FSB_SANITY_CHECK(mp, bno), error0); xfs_trans_brelse(tp, bp); @@ -4549,7 +4621,7 @@ xfs_bmap_read_extents( xfs_extnum_t start; - num_recs = be16_to_cpu(block->bb_numrecs); + num_recs = xfs_btree_get_numrecs(block); if (unlikely(i + num_recs > room)) { ASSERT(i + num_recs <= room); xfs_fs_repair_cmn_err(CE_WARN, ip->i_mount, @@ -4561,18 +4633,18 @@ xfs_bmap_read_extents( goto error0; } XFS_WANT_CORRUPTED_GOTO( - XFS_BMAP_SANITY_CHECK(mp, block, 0), + xfs_bmap_sanity_check(mp, bp, 0), error0); /* * Read-ahead the next leaf block, if any. */ - nextbno = be64_to_cpu(block->bb_rightsib); + nextbno = be64_to_cpu(block->bb_u.l.bb_rightsib); if (nextbno != NULLFSBLOCK) xfs_btree_reada_bufl(mp, nextbno, 1); /* * Copy records into the extent records. */ - frp = XFS_BTREE_REC_ADDR(xfs_bmbt, block, 1); + frp = XFS_BMBT_REC_ADDR(mp, block, 1); start = i; for (j = 0; j < num_recs; j++, i++, frp++) { xfs_bmbt_rec_host_t *trp = xfs_iext_get_ext(ifp, i); @@ -4603,7 +4675,7 @@ xfs_bmap_read_extents( if ((error = xfs_btree_read_bufl(mp, tp, bno, 0, &bp, XFS_BMAP_BTREE_REF))) return error; - block = XFS_BUF_TO_BMBT_BLOCK(bp); + block = XFS_BUF_TO_BLOCK(bp); } ASSERT(i == (ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t))); ASSERT(i == XFS_IFORK_NEXTENTS(ip, whichfork)); @@ -5029,8 +5101,7 @@ xfs_bmapi( if (abno == NULLFSBLOCK) break; if ((ifp->if_flags & XFS_IFBROOT) && !cur) { - cur = xfs_btree_init_cursor(mp, - tp, NULL, 0, XFS_BTNUM_BMAP, + cur = xfs_bmbt_init_cursor(mp, tp, ip, whichfork); cur->bc_private.b.firstblock = *firstblock; @@ -5147,9 +5218,8 @@ xfs_bmapi( */ ASSERT(mval->br_blockcount <= len); if ((ifp->if_flags & XFS_IFBROOT) && !cur) { - cur = xfs_btree_init_cursor(mp, - tp, NULL, 0, XFS_BTNUM_BMAP, - ip, whichfork); + cur = xfs_bmbt_init_cursor(mp, + tp, ip, whichfork); cur->bc_private.b.firstblock = *firstblock; cur->bc_private.b.flist = flist; @@ -5440,8 +5510,7 @@ xfs_bunmapi( logflags = 0; if (ifp->if_flags & XFS_IFBROOT) { ASSERT(XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_BTREE); - cur = xfs_btree_init_cursor(mp, tp, NULL, 0, XFS_BTNUM_BMAP, ip, - whichfork); + cur = xfs_bmbt_init_cursor(mp, tp, ip, whichfork); cur->bc_private.b.firstblock = *firstblock; cur->bc_private.b.flist = flist; cur->bc_private.b.flags = 0; @@ -5742,14 +5811,17 @@ error0: STATIC int xfs_getbmapx_fix_eof_hole( xfs_inode_t *ip, /* xfs incore inode pointer */ - struct getbmap *out, /* output structure */ + struct getbmapx *out, /* output structure */ int prealloced, /* this is a file with - * preallocated data space */ + * preallocated data space */ __int64_t end, /* last block requested */ xfs_fsblock_t startblock) { __int64_t fixlen; xfs_mount_t *mp; /* file system mount point */ + xfs_ifork_t *ifp; /* inode fork pointer */ + xfs_extnum_t lastx; /* last extent pointer */ + xfs_fileoff_t fileblock; if (startblock == HOLESTARTBLOCK) { mp = ip->i_mount; @@ -5763,21 +5835,33 @@ xfs_getbmapx_fix_eof_hole( out->bmv_length = fixlen; } } else { - out->bmv_block = XFS_FSB_TO_DB(ip, startblock); + if (startblock == DELAYSTARTBLOCK) + out->bmv_block = -2; + else + out->bmv_block = XFS_FSB_TO_DB(ip, startblock); + fileblock = XFS_BB_TO_FSB(ip->i_mount, out->bmv_offset); + ifp = XFS_IFORK_PTR(ip, XFS_DATA_FORK); + if (xfs_iext_bno_to_ext(ifp, fileblock, &lastx) && + (lastx == (ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t))-1)) + out->bmv_oflags |= BMV_OF_LAST; } return 1; } /* - * Fcntl interface to xfs_bmapi. + * Get inode's extents as described in bmv, and format for output. + * Calls formatter to fill the user's buffer until all extents + * are mapped, until the passed-in bmv->bmv_count slots have + * been filled, or until the formatter short-circuits the loop, + * if it is tracking filled-in extents on its own. */ int /* error code */ xfs_getbmap( xfs_inode_t *ip, - struct getbmap *bmv, /* user bmap structure */ - void __user *ap, /* pointer to user's array */ - int interface) /* interface flags */ + struct getbmapx *bmv, /* user bmap structure */ + xfs_bmap_format_t formatter, /* format to user */ + void *arg) /* formatter arg */ { __int64_t bmvend; /* last block requested */ int error; /* return value */ @@ -5790,19 +5874,17 @@ xfs_getbmap( int nexleft; /* # of user extents left */ int subnex; /* # of bmapi's can do */ int nmap; /* number of map entries */ - struct getbmap out; /* output structure */ + struct getbmapx out; /* output structure */ int whichfork; /* data or attr fork */ int prealloced; /* this is a file with * preallocated data space */ - int sh_unwritten; /* true, if unwritten */ - /* extents listed separately */ + int iflags; /* interface flags */ int bmapi_flags; /* flags for xfs_bmapi */ - __int32_t oflags; /* getbmapx bmv_oflags field */ mp = ip->i_mount; + iflags = bmv->bmv_iflags; - whichfork = interface & BMV_IF_ATTRFORK ? XFS_ATTR_FORK : XFS_DATA_FORK; - sh_unwritten = (interface & BMV_IF_PREALLOC) != 0; + whichfork = iflags & BMV_IF_ATTRFORK ? XFS_ATTR_FORK : XFS_DATA_FORK; /* If the BMV_IF_NO_DMAPI_READ interface bit specified, do not * generate a DMAPI read event. Otherwise, if the DM_EVENT_READ @@ -5817,7 +5899,7 @@ xfs_getbmap( * could misinterpret holes in a DMAPI file as true holes, * when in fact they may represent offline user data. */ - if ((interface & BMV_IF_NO_DMAPI_READ) == 0 && + if ((iflags & BMV_IF_NO_DMAPI_READ) == 0 && DM_EVENT_ENABLED(ip, DM_EVENT_READ) && whichfork == XFS_DATA_FORK) { error = XFS_SEND_DATA(mp, DM_EVENT_READ, ip, 0, 0, 0, NULL); @@ -5873,8 +5955,9 @@ xfs_getbmap( xfs_ilock(ip, XFS_IOLOCK_SHARED); - if (whichfork == XFS_DATA_FORK && - (ip->i_delayed_blks || ip->i_size > ip->i_d.di_size)) { + if (((iflags & BMV_IF_DELALLOC) == 0) && + (whichfork == XFS_DATA_FORK) && + (ip->i_delayed_blks || ip->i_size > ip->i_d.di_size)) { /* xfs_fsize_t last_byte = xfs_file_last_byte(ip); */ error = xfs_flush_pages(ip, (xfs_off_t)0, -1, 0, FI_REMAPF); @@ -5884,7 +5967,8 @@ xfs_getbmap( } } - ASSERT(whichfork == XFS_ATTR_FORK || ip->i_delayed_blks == 0); + ASSERT(whichfork == XFS_ATTR_FORK || (iflags & BMV_IF_DELALLOC) || + ip->i_delayed_blks == 0); lock = xfs_ilock_map_shared(ip); @@ -5896,7 +5980,7 @@ xfs_getbmap( nex = XFS_IFORK_NEXTENTS(ip, whichfork) * 2 + 1; bmapi_flags = XFS_BMAPI_AFLAG(whichfork) | - ((sh_unwritten) ? 0 : XFS_BMAPI_IGSTATE); + ((iflags & BMV_IF_PREALLOC) ? 0 : XFS_BMAPI_IGSTATE); /* * Allocate enough space to handle "subnex" maps at a time. @@ -5906,9 +5990,12 @@ xfs_getbmap( bmv->bmv_entries = 0; - if (XFS_IFORK_NEXTENTS(ip, whichfork) == 0) { - error = 0; - goto unlock_and_return; + if ((XFS_IFORK_NEXTENTS(ip, whichfork) == 0)) { + if (((iflags & BMV_IF_DELALLOC) == 0) || + whichfork == XFS_ATTR_FORK) { + error = 0; + goto unlock_and_return; + } } nexleft = nex; @@ -5924,52 +6011,40 @@ xfs_getbmap( ASSERT(nmap <= subnex); for (i = 0; i < nmap && nexleft && bmv->bmv_length; i++) { - nexleft--; - oflags = (map[i].br_state == XFS_EXT_UNWRITTEN) ? - BMV_OF_PREALLOC : 0; + out.bmv_oflags = 0; + if (map[i].br_state == XFS_EXT_UNWRITTEN) + out.bmv_oflags |= BMV_OF_PREALLOC; + else if (map[i].br_startblock == DELAYSTARTBLOCK) + out.bmv_oflags |= BMV_OF_DELALLOC; out.bmv_offset = XFS_FSB_TO_BB(mp, map[i].br_startoff); out.bmv_length = XFS_FSB_TO_BB(mp, map[i].br_blockcount); - ASSERT(map[i].br_startblock != DELAYSTARTBLOCK); + out.bmv_unused1 = out.bmv_unused2 = 0; + ASSERT(((iflags & BMV_IF_DELALLOC) != 0) || + (map[i].br_startblock != DELAYSTARTBLOCK)); if (map[i].br_startblock == HOLESTARTBLOCK && whichfork == XFS_ATTR_FORK) { /* came to the end of attribute fork */ + out.bmv_oflags |= BMV_OF_LAST; goto unlock_and_return; } else { + int full = 0; /* user array is full */ + if (!xfs_getbmapx_fix_eof_hole(ip, &out, prealloced, bmvend, map[i].br_startblock)) { goto unlock_and_return; } - /* return either getbmap/getbmapx structure. */ - if (interface & BMV_IF_EXTENDED) { - struct getbmapx outx; - - GETBMAP_CONVERT(out,outx); - outx.bmv_oflags = oflags; - outx.bmv_unused1 = outx.bmv_unused2 = 0; - if (copy_to_user(ap, &outx, - sizeof(outx))) { - error = XFS_ERROR(EFAULT); - goto unlock_and_return; - } - } else { - if (copy_to_user(ap, &out, - sizeof(out))) { - error = XFS_ERROR(EFAULT); - goto unlock_and_return; - } - } + /* format results & advance arg */ + error = formatter(&arg, &out, &full); + if (error || full) + goto unlock_and_return; + nexleft--; bmv->bmv_offset = out.bmv_offset + out.bmv_length; bmv->bmv_length = MAX((__int64_t)0, (__int64_t)(bmvend - bmv->bmv_offset)); bmv->bmv_entries++; - ap = (interface & BMV_IF_EXTENDED) ? - (void __user *) - ((struct getbmapx __user *)ap + 1) : - (void __user *) - ((struct getbmap __user *)ap + 1); } } } while (nmap && nexleft && bmv->bmv_length); @@ -6131,7 +6206,7 @@ xfs_bmap_get_bp( void xfs_check_block( - xfs_bmbt_block_t *block, + struct xfs_btree_block *block, xfs_mount_t *mp, int root, short sz) @@ -6143,36 +6218,29 @@ xfs_check_block( ASSERT(be16_to_cpu(block->bb_level) > 0); prevp = NULL; - for( i = 1; i <= be16_to_cpu(block->bb_numrecs); i++) { + for( i = 1; i <= xfs_btree_get_numrecs(block); i++) { dmxr = mp->m_bmap_dmxr[0]; - - if (root) { - keyp = XFS_BMAP_BROOT_KEY_ADDR(block, i, sz); - } else { - keyp = XFS_BTREE_KEY_ADDR(xfs_bmbt, block, i); - } + keyp = XFS_BMBT_KEY_ADDR(mp, block, i); if (prevp) { - xfs_btree_check_key(XFS_BTNUM_BMAP, prevp, keyp); + ASSERT(be64_to_cpu(prevp->br_startoff) < + be64_to_cpu(keyp->br_startoff)); } prevp = keyp; /* * Compare the block numbers to see if there are dups. */ + if (root) + pp = XFS_BMAP_BROOT_PTR_ADDR(mp, block, i, sz); + else + pp = XFS_BMBT_PTR_ADDR(mp, block, i, dmxr); - if (root) { - pp = XFS_BMAP_BROOT_PTR_ADDR(block, i, sz); - } else { - pp = XFS_BTREE_PTR_ADDR(xfs_bmbt, block, i, dmxr); - } for (j = i+1; j <= be16_to_cpu(block->bb_numrecs); j++) { - if (root) { - thispa = XFS_BMAP_BROOT_PTR_ADDR(block, j, sz); - } else { - thispa = XFS_BTREE_PTR_ADDR(xfs_bmbt, block, j, - dmxr); - } + if (root) + thispa = XFS_BMAP_BROOT_PTR_ADDR(mp, block, j, sz); + else + thispa = XFS_BMBT_PTR_ADDR(mp, block, j, dmxr); if (*thispa == *pp) { cmn_err(CE_WARN, "%s: thispa(%d) == pp(%d) %Ld", __func__, j, i, @@ -6195,7 +6263,7 @@ xfs_bmap_check_leaf_extents( xfs_inode_t *ip, /* incore inode pointer */ int whichfork) /* data or attr fork */ { - xfs_bmbt_block_t *block; /* current btree block */ + struct xfs_btree_block *block; /* current btree block */ xfs_fsblock_t bno; /* block # of "block" */ xfs_buf_t *bp; /* buffer for "block" */ int error; /* error return value */ @@ -6223,7 +6291,7 @@ xfs_bmap_check_leaf_extents( level = be16_to_cpu(block->bb_level); ASSERT(level > 0); xfs_check_block(block, mp, 1, ifp->if_broot_bytes); - pp = XFS_BMAP_BROOT_PTR_ADDR(block, 1, ifp->if_broot_bytes); + pp = XFS_BMAP_BROOT_PTR_ADDR(mp, block, 1, ifp->if_broot_bytes); bno = be64_to_cpu(*pp); ASSERT(bno != NULLDFSBNO); @@ -6245,9 +6313,9 @@ xfs_bmap_check_leaf_extents( if (!bp && (error = xfs_btree_read_bufl(mp, NULL, bno, 0, &bp, XFS_BMAP_BTREE_REF))) goto error_norelse; - block = XFS_BUF_TO_BMBT_BLOCK(bp); + block = XFS_BUF_TO_BLOCK(bp); XFS_WANT_CORRUPTED_GOTO( - XFS_BMAP_SANITY_CHECK(mp, block, level), + xfs_bmap_sanity_check(mp, bp, level), error0); if (level == 0) break; @@ -6258,7 +6326,7 @@ xfs_bmap_check_leaf_extents( */ xfs_check_block(block, mp, 0, 0); - pp = XFS_BTREE_PTR_ADDR(xfs_bmbt, block, 1, mp->m_bmap_dmxr[1]); + pp = XFS_BMBT_PTR_ADDR(mp, block, 1, mp->m_bmap_dmxr[1]); bno = be64_to_cpu(*pp); XFS_WANT_CORRUPTED_GOTO(XFS_FSB_SANITY_CHECK(mp, bno), error0); if (bp_release) { @@ -6280,13 +6348,13 @@ xfs_bmap_check_leaf_extents( xfs_extnum_t num_recs; - num_recs = be16_to_cpu(block->bb_numrecs); + num_recs = xfs_btree_get_numrecs(block); /* * Read-ahead the next leaf block, if any. */ - nextbno = be64_to_cpu(block->bb_rightsib); + nextbno = be64_to_cpu(block->bb_u.l.bb_rightsib); /* * Check all the extents to make sure they are OK. @@ -6294,13 +6362,17 @@ xfs_bmap_check_leaf_extents( * conform with the first entry in this one. */ - ep = XFS_BTREE_REC_ADDR(xfs_bmbt, block, 1); + ep = XFS_BMBT_REC_ADDR(mp, block, 1); if (i) { - xfs_btree_check_rec(XFS_BTNUM_BMAP, &last, ep); + ASSERT(xfs_bmbt_disk_get_startoff(&last) + + xfs_bmbt_disk_get_blockcount(&last) <= + xfs_bmbt_disk_get_startoff(ep)); } for (j = 1; j < num_recs; j++) { - nextp = XFS_BTREE_REC_ADDR(xfs_bmbt, block, j + 1); - xfs_btree_check_rec(XFS_BTNUM_BMAP, ep, nextp); + nextp = XFS_BMBT_REC_ADDR(mp, block, j + 1); + ASSERT(xfs_bmbt_disk_get_startoff(ep) + + xfs_bmbt_disk_get_blockcount(ep) <= + xfs_bmbt_disk_get_startoff(nextp)); ep = nextp; } @@ -6326,7 +6398,7 @@ xfs_bmap_check_leaf_extents( if (!bp && (error = xfs_btree_read_bufl(mp, NULL, bno, 0, &bp, XFS_BMAP_BTREE_REF))) goto error_norelse; - block = XFS_BUF_TO_BMBT_BLOCK(bp); + block = XFS_BUF_TO_BLOCK(bp); } if (bp_release) { bp_release = 0; @@ -6356,7 +6428,7 @@ xfs_bmap_count_blocks( int whichfork, /* data or attr fork */ int *count) /* out: count of blocks */ { - xfs_bmbt_block_t *block; /* current btree block */ + struct xfs_btree_block *block; /* current btree block */ xfs_fsblock_t bno; /* block # of "block" */ xfs_ifork_t *ifp; /* fork structure */ int level; /* btree level, for checking */ @@ -6379,7 +6451,7 @@ xfs_bmap_count_blocks( block = ifp->if_broot; level = be16_to_cpu(block->bb_level); ASSERT(level > 0); - pp = XFS_BMAP_BROOT_PTR_ADDR(block, 1, ifp->if_broot_bytes); + pp = XFS_BMAP_BROOT_PTR_ADDR(mp, block, 1, ifp->if_broot_bytes); bno = be64_to_cpu(*pp); ASSERT(bno != NULLDFSBNO); ASSERT(XFS_FSB_TO_AGNO(mp, bno) < mp->m_sb.sb_agcount); @@ -6413,29 +6485,29 @@ xfs_bmap_count_tree( __be64 *pp; xfs_fsblock_t bno = blockno; xfs_fsblock_t nextbno; - xfs_bmbt_block_t *block, *nextblock; + struct xfs_btree_block *block, *nextblock; int numrecs; if ((error = xfs_btree_read_bufl(mp, tp, bno, 0, &bp, XFS_BMAP_BTREE_REF))) return error; *count += 1; - block = XFS_BUF_TO_BMBT_BLOCK(bp); + block = XFS_BUF_TO_BLOCK(bp); if (--level) { /* Not at node above leafs, count this level of nodes */ - nextbno = be64_to_cpu(block->bb_rightsib); + nextbno = be64_to_cpu(block->bb_u.l.bb_rightsib); while (nextbno != NULLFSBLOCK) { if ((error = xfs_btree_read_bufl(mp, tp, nextbno, 0, &nbp, XFS_BMAP_BTREE_REF))) return error; *count += 1; - nextblock = XFS_BUF_TO_BMBT_BLOCK(nbp); - nextbno = be64_to_cpu(nextblock->bb_rightsib); + nextblock = XFS_BUF_TO_BLOCK(nbp); + nextbno = be64_to_cpu(nextblock->bb_u.l.bb_rightsib); xfs_trans_brelse(tp, nbp); } /* Dive to the next level */ - pp = XFS_BTREE_PTR_ADDR(xfs_bmbt, block, 1, mp->m_bmap_dmxr[1]); + pp = XFS_BMBT_PTR_ADDR(mp, block, 1, mp->m_bmap_dmxr[1]); bno = be64_to_cpu(*pp); if (unlikely((error = xfs_bmap_count_tree(mp, tp, ifp, bno, level, count)) < 0)) { @@ -6448,9 +6520,9 @@ xfs_bmap_count_tree( } else { /* count all level 1 nodes and their leaves */ for (;;) { - nextbno = be64_to_cpu(block->bb_rightsib); + nextbno = be64_to_cpu(block->bb_u.l.bb_rightsib); numrecs = be16_to_cpu(block->bb_numrecs); - xfs_bmap_disk_count_leaves(0, block, numrecs, count); + xfs_bmap_disk_count_leaves(mp, block, numrecs, count); xfs_trans_brelse(tp, bp); if (nextbno == NULLFSBLOCK) break; @@ -6459,7 +6531,7 @@ xfs_bmap_count_tree( XFS_BMAP_BTREE_REF))) return error; *count += 1; - block = XFS_BUF_TO_BMBT_BLOCK(bp); + block = XFS_BUF_TO_BLOCK(bp); } } return 0; @@ -6489,8 +6561,8 @@ xfs_bmap_count_leaves( */ STATIC void xfs_bmap_disk_count_leaves( - xfs_extnum_t idx, - xfs_bmbt_block_t *block, + struct xfs_mount *mp, + struct xfs_btree_block *block, int numrecs, int *count) { @@ -6498,7 +6570,7 @@ xfs_bmap_disk_count_leaves( xfs_bmbt_rec_t *frp; for (b = 1; b <= numrecs; b++) { - frp = XFS_BTREE_REC_ADDR(xfs_bmbt, block, idx + b); + frp = XFS_BMBT_REC_ADDR(mp, block, b); *count += xfs_bmbt_disk_get_blockcount(frp); } } diff --git a/fs/xfs/xfs_bmap.h b/fs/xfs/xfs_bmap.h index 9f3e3a836d1..284571c05ed 100644 --- a/fs/xfs/xfs_bmap.h +++ b/fs/xfs/xfs_bmap.h @@ -137,9 +137,7 @@ typedef struct xfs_bmalloca { char conv; /* overwriting unwritten extents */ } xfs_bmalloca_t; -#ifdef __KERNEL__ - -#if defined(XFS_BMAP_TRACE) +#if defined(__KERNEL__) && defined(XFS_BMAP_TRACE) /* * Trace operations for bmap extent tracing */ @@ -163,9 +161,12 @@ xfs_bmap_trace_exlist( int whichfork); /* data or attr fork */ #define XFS_BMAP_TRACE_EXLIST(ip,c,w) \ xfs_bmap_trace_exlist(__func__,ip,c,w) -#else + +#else /* __KERNEL__ && XFS_BMAP_TRACE */ + #define XFS_BMAP_TRACE_EXLIST(ip,c,w) -#endif + +#endif /* __KERNEL__ && XFS_BMAP_TRACE */ /* * Convert inode from non-attributed to attributed. @@ -206,20 +207,6 @@ xfs_bmap_compute_maxlevels( int whichfork); /* data or attr fork */ /* - * Routine to be called at transaction's end by xfs_bmapi, xfs_bunmapi - * caller. Frees all the extents that need freeing, which must be done - * last due to locking considerations. - * - * Return 1 if the given transaction was committed and a new one allocated, - * and 0 otherwise. - */ -int /* error */ -xfs_bmap_finish( - struct xfs_trans **tp, /* transaction pointer addr */ - xfs_bmap_free_t *flist, /* i/o: list extents to free */ - int *committed); /* xact committed or not */ - -/* * Returns the file-relative block number of the first unused block in the file. * This is the lowest-address hole if the file has holes, else the first block * past the end of file. @@ -344,14 +331,43 @@ xfs_bunmapi( int *done); /* set if not done yet */ /* - * Fcntl interface to xfs_bmapi. + * Check an extent list, which has just been read, for + * any bit in the extent flag field. + */ +int +xfs_check_nostate_extents( + struct xfs_ifork *ifp, + xfs_extnum_t idx, + xfs_extnum_t num); + +#ifdef __KERNEL__ + +/* + * Routine to be called at transaction's end by xfs_bmapi, xfs_bunmapi + * caller. Frees all the extents that need freeing, which must be done + * last due to locking considerations. + * + * Return 1 if the given transaction was committed and a new one allocated, + * and 0 otherwise. + */ +int /* error */ +xfs_bmap_finish( + struct xfs_trans **tp, /* transaction pointer addr */ + xfs_bmap_free_t *flist, /* i/o: list extents to free */ + int *committed); /* xact committed or not */ + +/* bmap to userspace formatter - copy to user & advance pointer */ +typedef int (*xfs_bmap_format_t)(void **, struct getbmapx *, int *); + +/* + * Get inode's extents as described in bmv, and format for output. */ int /* error code */ xfs_getbmap( xfs_inode_t *ip, - struct getbmap *bmv, /* user bmap structure */ - void __user *ap, /* pointer to user's array */ - int iflags); /* interface flags */ + struct getbmapx *bmv, /* user bmap structure */ + xfs_bmap_format_t formatter, /* format to user */ + void *arg); /* formatter arg */ /* * Check if the endoff is outside the last extent. If so the caller will grow @@ -375,16 +391,6 @@ xfs_bmap_count_blocks( int *count); /* - * Check an extent list, which has just been read, for - * any bit in the extent flag field. - */ -int -xfs_check_nostate_extents( - struct xfs_ifork *ifp, - xfs_extnum_t idx, - xfs_extnum_t num); - -/* * Search the extent records for the entry containing block bno. * If bno lies in a hole, point to the next entry. If bno lies * past eof, *eofp will be set, and *prevp will contain the last diff --git a/fs/xfs/xfs_bmap_btree.c b/fs/xfs/xfs_bmap_btree.c index 23efad29a5c..8f1ec73725d 100644 --- a/fs/xfs/xfs_bmap_btree.c +++ b/fs/xfs/xfs_bmap_btree.c @@ -37,1406 +37,13 @@ #include "xfs_inode_item.h" #include "xfs_alloc.h" #include "xfs_btree.h" +#include "xfs_btree_trace.h" #include "xfs_ialloc.h" #include "xfs_itable.h" #include "xfs_bmap.h" #include "xfs_error.h" #include "xfs_quota.h" -#if defined(XFS_BMBT_TRACE) -ktrace_t *xfs_bmbt_trace_buf; -#endif - -/* - * Prototypes for internal btree functions. - */ - - -STATIC int xfs_bmbt_killroot(xfs_btree_cur_t *); -STATIC void xfs_bmbt_log_keys(xfs_btree_cur_t *, xfs_buf_t *, int, int); -STATIC void xfs_bmbt_log_ptrs(xfs_btree_cur_t *, xfs_buf_t *, int, int); -STATIC int xfs_bmbt_lshift(xfs_btree_cur_t *, int, int *); -STATIC int xfs_bmbt_rshift(xfs_btree_cur_t *, int, int *); -STATIC int xfs_bmbt_split(xfs_btree_cur_t *, int, xfs_fsblock_t *, - __uint64_t *, xfs_btree_cur_t **, int *); -STATIC int xfs_bmbt_updkey(xfs_btree_cur_t *, xfs_bmbt_key_t *, int); - - -#if defined(XFS_BMBT_TRACE) - -static char ARGS[] = "args"; -static char ENTRY[] = "entry"; -static char ERROR[] = "error"; -#undef EXIT -static char EXIT[] = "exit"; - -/* - * Add a trace buffer entry for the arguments given to the routine, - * generic form. - */ -STATIC void -xfs_bmbt_trace_enter( - const char *func, - xfs_btree_cur_t *cur, - char *s, - int type, - int line, - __psunsigned_t a0, - __psunsigned_t a1, - __psunsigned_t a2, - __psunsigned_t a3, - __psunsigned_t a4, - __psunsigned_t a5, - __psunsigned_t a6, - __psunsigned_t a7, - __psunsigned_t a8, - __psunsigned_t a9, - __psunsigned_t a10) -{ - xfs_inode_t *ip; - int whichfork; - - ip = cur->bc_private.b.ip; - whichfork = cur->bc_private.b.whichfork; - ktrace_enter(xfs_bmbt_trace_buf, - (void *)((__psint_t)type | (whichfork << 8) | (line << 16)), - (void *)func, (void *)s, (void *)ip, (void *)cur, - (void *)a0, (void *)a1, (void *)a2, (void *)a3, - (void *)a4, (void *)a5, (void *)a6, (void *)a7, - (void *)a8, (void *)a9, (void *)a10); - ASSERT(ip->i_btrace); - ktrace_enter(ip->i_btrace, - (void *)((__psint_t)type | (whichfork << 8) | (line << 16)), - (void *)func, (void *)s, (void *)ip, (void *)cur, - (void *)a0, (void *)a1, (void *)a2, (void *)a3, - (void *)a4, (void *)a5, (void *)a6, (void *)a7, - (void *)a8, (void *)a9, (void *)a10); -} -/* - * Add a trace buffer entry for arguments, for a buffer & 1 integer arg. - */ -STATIC void -xfs_bmbt_trace_argbi( - const char *func, - xfs_btree_cur_t *cur, - xfs_buf_t *b, - int i, - int line) -{ - xfs_bmbt_trace_enter(func, cur, ARGS, XFS_BMBT_KTRACE_ARGBI, line, - (__psunsigned_t)b, i, 0, 0, - 0, 0, 0, 0, - 0, 0, 0); -} - -/* - * Add a trace buffer entry for arguments, for a buffer & 2 integer args. - */ -STATIC void -xfs_bmbt_trace_argbii( - const char *func, - xfs_btree_cur_t *cur, - xfs_buf_t *b, - int i0, - int i1, - int line) -{ - xfs_bmbt_trace_enter(func, cur, ARGS, XFS_BMBT_KTRACE_ARGBII, line, - (__psunsigned_t)b, i0, i1, 0, - 0, 0, 0, 0, - 0, 0, 0); -} - -/* - * Add a trace buffer entry for arguments, for 3 block-length args - * and an integer arg. - */ -STATIC void -xfs_bmbt_trace_argfffi( - const char *func, - xfs_btree_cur_t *cur, - xfs_dfiloff_t o, - xfs_dfsbno_t b, - xfs_dfilblks_t i, - int j, - int line) -{ - xfs_bmbt_trace_enter(func, cur, ARGS, XFS_BMBT_KTRACE_ARGFFFI, line, - o >> 32, (int)o, b >> 32, (int)b, - i >> 32, (int)i, (int)j, 0, - 0, 0, 0); -} - -/* - * Add a trace buffer entry for arguments, for one integer arg. - */ -STATIC void -xfs_bmbt_trace_argi( - const char *func, - xfs_btree_cur_t *cur, - int i, - int line) -{ - xfs_bmbt_trace_enter(func, cur, ARGS, XFS_BMBT_KTRACE_ARGI, line, - i, 0, 0, 0, - 0, 0, 0, 0, - 0, 0, 0); -} - -/* - * Add a trace buffer entry for arguments, for int, fsblock, key. - */ -STATIC void -xfs_bmbt_trace_argifk( - const char *func, - xfs_btree_cur_t *cur, - int i, - xfs_fsblock_t f, - xfs_dfiloff_t o, - int line) -{ - xfs_bmbt_trace_enter(func, cur, ARGS, XFS_BMBT_KTRACE_ARGIFK, line, - i, (xfs_dfsbno_t)f >> 32, (int)f, o >> 32, - (int)o, 0, 0, 0, - 0, 0, 0); -} - -/* - * Add a trace buffer entry for arguments, for int, fsblock, rec. - */ -STATIC void -xfs_bmbt_trace_argifr( - const char *func, - xfs_btree_cur_t *cur, - int i, - xfs_fsblock_t f, - xfs_bmbt_rec_t *r, - int line) -{ - xfs_dfsbno_t b; - xfs_dfilblks_t c; - xfs_dfsbno_t d; - xfs_dfiloff_t o; - xfs_bmbt_irec_t s; - - d = (xfs_dfsbno_t)f; - xfs_bmbt_disk_get_all(r, &s); - o = (xfs_dfiloff_t)s.br_startoff; - b = (xfs_dfsbno_t)s.br_startblock; - c = s.br_blockcount; - xfs_bmbt_trace_enter(func, cur, ARGS, XFS_BMBT_KTRACE_ARGIFR, line, - i, d >> 32, (int)d, o >> 32, - (int)o, b >> 32, (int)b, c >> 32, - (int)c, 0, 0); -} - -/* - * Add a trace buffer entry for arguments, for int, key. - */ -STATIC void -xfs_bmbt_trace_argik( - const char *func, - xfs_btree_cur_t *cur, - int i, - xfs_bmbt_key_t *k, - int line) -{ - xfs_dfiloff_t o; - - o = be64_to_cpu(k->br_startoff); - xfs_bmbt_trace_enter(func, cur, ARGS, XFS_BMBT_KTRACE_ARGIFK, line, - i, o >> 32, (int)o, 0, - 0, 0, 0, 0, - 0, 0, 0); -} - -/* - * Add a trace buffer entry for the cursor/operation. - */ -STATIC void -xfs_bmbt_trace_cursor( - const char *func, - xfs_btree_cur_t *cur, - char *s, - int line) -{ - xfs_bmbt_rec_host_t r; - - xfs_bmbt_set_all(&r, &cur->bc_rec.b); - xfs_bmbt_trace_enter(func, cur, s, XFS_BMBT_KTRACE_CUR, line, - (cur->bc_nlevels << 24) | (cur->bc_private.b.flags << 16) | - cur->bc_private.b.allocated, - r.l0 >> 32, (int)r.l0, - r.l1 >> 32, (int)r.l1, - (unsigned long)cur->bc_bufs[0], (unsigned long)cur->bc_bufs[1], - (unsigned long)cur->bc_bufs[2], (unsigned long)cur->bc_bufs[3], - (cur->bc_ptrs[0] << 16) | cur->bc_ptrs[1], - (cur->bc_ptrs[2] << 16) | cur->bc_ptrs[3]); -} - -#define XFS_BMBT_TRACE_ARGBI(c,b,i) \ - xfs_bmbt_trace_argbi(__func__, c, b, i, __LINE__) -#define XFS_BMBT_TRACE_ARGBII(c,b,i,j) \ - xfs_bmbt_trace_argbii(__func__, c, b, i, j, __LINE__) -#define XFS_BMBT_TRACE_ARGFFFI(c,o,b,i,j) \ - xfs_bmbt_trace_argfffi(__func__, c, o, b, i, j, __LINE__) -#define XFS_BMBT_TRACE_ARGI(c,i) \ - xfs_bmbt_trace_argi(__func__, c, i, __LINE__) -#define XFS_BMBT_TRACE_ARGIFK(c,i,f,s) \ - xfs_bmbt_trace_argifk(__func__, c, i, f, s, __LINE__) -#define XFS_BMBT_TRACE_ARGIFR(c,i,f,r) \ - xfs_bmbt_trace_argifr(__func__, c, i, f, r, __LINE__) -#define XFS_BMBT_TRACE_ARGIK(c,i,k) \ - xfs_bmbt_trace_argik(__func__, c, i, k, __LINE__) -#define XFS_BMBT_TRACE_CURSOR(c,s) \ - xfs_bmbt_trace_cursor(__func__, c, s, __LINE__) -#else -#define XFS_BMBT_TRACE_ARGBI(c,b,i) -#define XFS_BMBT_TRACE_ARGBII(c,b,i,j) -#define XFS_BMBT_TRACE_ARGFFFI(c,o,b,i,j) -#define XFS_BMBT_TRACE_ARGI(c,i) -#define XFS_BMBT_TRACE_ARGIFK(c,i,f,s) -#define XFS_BMBT_TRACE_ARGIFR(c,i,f,r) -#define XFS_BMBT_TRACE_ARGIK(c,i,k) -#define XFS_BMBT_TRACE_CURSOR(c,s) -#endif /* XFS_BMBT_TRACE */ - - -/* - * Internal functions. - */ - -/* - * Delete record pointed to by cur/level. - */ -STATIC int /* error */ -xfs_bmbt_delrec( - xfs_btree_cur_t *cur, - int level, - int *stat) /* success/failure */ -{ - xfs_bmbt_block_t *block; /* bmap btree block */ - xfs_fsblock_t bno; /* fs-relative block number */ - xfs_buf_t *bp; /* buffer for block */ - int error; /* error return value */ - int i; /* loop counter */ - int j; /* temp state */ - xfs_bmbt_key_t key; /* bmap btree key */ - xfs_bmbt_key_t *kp=NULL; /* pointer to bmap btree key */ - xfs_fsblock_t lbno; /* left sibling block number */ - xfs_buf_t *lbp; /* left buffer pointer */ - xfs_bmbt_block_t *left; /* left btree block */ - xfs_bmbt_key_t *lkp; /* left btree key */ - xfs_bmbt_ptr_t *lpp; /* left address pointer */ - int lrecs=0; /* left record count */ - xfs_bmbt_rec_t *lrp; /* left record pointer */ - xfs_mount_t *mp; /* file system mount point */ - xfs_bmbt_ptr_t *pp; /* pointer to bmap block addr */ - int ptr; /* key/record index */ - xfs_fsblock_t rbno; /* right sibling block number */ - xfs_buf_t *rbp; /* right buffer pointer */ - xfs_bmbt_block_t *right; /* right btree block */ - xfs_bmbt_key_t *rkp; /* right btree key */ - xfs_bmbt_rec_t *rp; /* pointer to bmap btree rec */ - xfs_bmbt_ptr_t *rpp; /* right address pointer */ - xfs_bmbt_block_t *rrblock; /* right-right btree block */ - xfs_buf_t *rrbp; /* right-right buffer pointer */ - int rrecs=0; /* right record count */ - xfs_bmbt_rec_t *rrp; /* right record pointer */ - xfs_btree_cur_t *tcur; /* temporary btree cursor */ - int numrecs; /* temporary numrec count */ - int numlrecs, numrrecs; - - XFS_BMBT_TRACE_CURSOR(cur, ENTRY); - XFS_BMBT_TRACE_ARGI(cur, level); - ptr = cur->bc_ptrs[level]; - tcur = NULL; - if (ptr == 0) { - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *stat = 0; - return 0; - } - block = xfs_bmbt_get_block(cur, level, &bp); - numrecs = be16_to_cpu(block->bb_numrecs); -#ifdef DEBUG - if ((error = xfs_btree_check_lblock(cur, block, level, bp))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - goto error0; - } -#endif - if (ptr > numrecs) { - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *stat = 0; - return 0; - } - XFS_STATS_INC(xs_bmbt_delrec); - if (level > 0) { - kp = XFS_BMAP_KEY_IADDR(block, 1, cur); - pp = XFS_BMAP_PTR_IADDR(block, 1, cur); -#ifdef DEBUG - for (i = ptr; i < numrecs; i++) { - if ((error = xfs_btree_check_lptr_disk(cur, pp[i], level))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - goto error0; - } - } -#endif - if (ptr < numrecs) { - memmove(&kp[ptr - 1], &kp[ptr], - (numrecs - ptr) * sizeof(*kp)); - memmove(&pp[ptr - 1], &pp[ptr], - (numrecs - ptr) * sizeof(*pp)); - xfs_bmbt_log_ptrs(cur, bp, ptr, numrecs - 1); - xfs_bmbt_log_keys(cur, bp, ptr, numrecs - 1); - } - } else { - rp = XFS_BMAP_REC_IADDR(block, 1, cur); - if (ptr < numrecs) { - memmove(&rp[ptr - 1], &rp[ptr], - (numrecs - ptr) * sizeof(*rp)); - xfs_bmbt_log_recs(cur, bp, ptr, numrecs - 1); - } - if (ptr == 1) { - key.br_startoff = - cpu_to_be64(xfs_bmbt_disk_get_startoff(rp)); - kp = &key; - } - } - numrecs--; - block->bb_numrecs = cpu_to_be16(numrecs); - xfs_bmbt_log_block(cur, bp, XFS_BB_NUMRECS); - /* - * We're at the root level. - * First, shrink the root block in-memory. - * Try to get rid of the next level down. - * If we can't then there's nothing left to do. - */ - if (level == cur->bc_nlevels - 1) { - xfs_iroot_realloc(cur->bc_private.b.ip, -1, - cur->bc_private.b.whichfork); - if ((error = xfs_bmbt_killroot(cur))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - goto error0; - } - if (level > 0 && (error = xfs_bmbt_decrement(cur, level, &j))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - goto error0; - } - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *stat = 1; - return 0; - } - if (ptr == 1 && (error = xfs_bmbt_updkey(cur, kp, level + 1))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - goto error0; - } - if (numrecs >= XFS_BMAP_BLOCK_IMINRECS(level, cur)) { - if (level > 0 && (error = xfs_bmbt_decrement(cur, level, &j))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - goto error0; - } - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *stat = 1; - return 0; - } - rbno = be64_to_cpu(block->bb_rightsib); - lbno = be64_to_cpu(block->bb_leftsib); - /* - * One child of root, need to get a chance to copy its contents - * into the root and delete it. Can't go up to next level, - * there's nothing to delete there. - */ - if (lbno == NULLFSBLOCK && rbno == NULLFSBLOCK && - level == cur->bc_nlevels - 2) { - if ((error = xfs_bmbt_killroot(cur))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - goto error0; - } - if (level > 0 && (error = xfs_bmbt_decrement(cur, level, &i))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - goto error0; - } - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *stat = 1; - return 0; - } - ASSERT(rbno != NULLFSBLOCK || lbno != NULLFSBLOCK); - if ((error = xfs_btree_dup_cursor(cur, &tcur))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - goto error0; - } - bno = NULLFSBLOCK; - if (rbno != NULLFSBLOCK) { - i = xfs_btree_lastrec(tcur, level); - XFS_WANT_CORRUPTED_GOTO(i == 1, error0); - if ((error = xfs_bmbt_increment(tcur, level, &i))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - goto error0; - } - XFS_WANT_CORRUPTED_GOTO(i == 1, error0); - i = xfs_btree_lastrec(tcur, level); - XFS_WANT_CORRUPTED_GOTO(i == 1, error0); - rbp = tcur->bc_bufs[level]; - right = XFS_BUF_TO_BMBT_BLOCK(rbp); -#ifdef DEBUG - if ((error = xfs_btree_check_lblock(cur, right, level, rbp))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - goto error0; - } -#endif - bno = be64_to_cpu(right->bb_leftsib); - if (be16_to_cpu(right->bb_numrecs) - 1 >= - XFS_BMAP_BLOCK_IMINRECS(level, cur)) { - if ((error = xfs_bmbt_lshift(tcur, level, &i))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - goto error0; - } - if (i) { - ASSERT(be16_to_cpu(block->bb_numrecs) >= - XFS_BMAP_BLOCK_IMINRECS(level, tcur)); - xfs_btree_del_cursor(tcur, XFS_BTREE_NOERROR); - tcur = NULL; - if (level > 0) { - if ((error = xfs_bmbt_decrement(cur, - level, &i))) { - XFS_BMBT_TRACE_CURSOR(cur, - ERROR); - goto error0; - } - } - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *stat = 1; - return 0; - } - } - rrecs = be16_to_cpu(right->bb_numrecs); - if (lbno != NULLFSBLOCK) { - i = xfs_btree_firstrec(tcur, level); - XFS_WANT_CORRUPTED_GOTO(i == 1, error0); - if ((error = xfs_bmbt_decrement(tcur, level, &i))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - goto error0; - } - XFS_WANT_CORRUPTED_GOTO(i == 1, error0); - } - } - if (lbno != NULLFSBLOCK) { - i = xfs_btree_firstrec(tcur, level); - XFS_WANT_CORRUPTED_GOTO(i == 1, error0); - /* - * decrement to last in block - */ - if ((error = xfs_bmbt_decrement(tcur, level, &i))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - goto error0; - } - i = xfs_btree_firstrec(tcur, level); - XFS_WANT_CORRUPTED_GOTO(i == 1, error0); - lbp = tcur->bc_bufs[level]; - left = XFS_BUF_TO_BMBT_BLOCK(lbp); -#ifdef DEBUG - if ((error = xfs_btree_check_lblock(cur, left, level, lbp))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - goto error0; - } -#endif - bno = be64_to_cpu(left->bb_rightsib); - if (be16_to_cpu(left->bb_numrecs) - 1 >= - XFS_BMAP_BLOCK_IMINRECS(level, cur)) { - if ((error = xfs_bmbt_rshift(tcur, level, &i))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - goto error0; - } - if (i) { - ASSERT(be16_to_cpu(block->bb_numrecs) >= - XFS_BMAP_BLOCK_IMINRECS(level, tcur)); - xfs_btree_del_cursor(tcur, XFS_BTREE_NOERROR); - tcur = NULL; - if (level == 0) - cur->bc_ptrs[0]++; - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *stat = 1; - return 0; - } - } - lrecs = be16_to_cpu(left->bb_numrecs); - } - xfs_btree_del_cursor(tcur, XFS_BTREE_NOERROR); - tcur = NULL; - mp = cur->bc_mp; - ASSERT(bno != NULLFSBLOCK); - if (lbno != NULLFSBLOCK && - lrecs + be16_to_cpu(block->bb_numrecs) <= XFS_BMAP_BLOCK_IMAXRECS(level, cur)) { - rbno = bno; - right = block; - rbp = bp; - if ((error = xfs_btree_read_bufl(mp, cur->bc_tp, lbno, 0, &lbp, - XFS_BMAP_BTREE_REF))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - goto error0; - } - left = XFS_BUF_TO_BMBT_BLOCK(lbp); - if ((error = xfs_btree_check_lblock(cur, left, level, lbp))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - goto error0; - } - } else if (rbno != NULLFSBLOCK && - rrecs + be16_to_cpu(block->bb_numrecs) <= - XFS_BMAP_BLOCK_IMAXRECS(level, cur)) { - lbno = bno; - left = block; - lbp = bp; - if ((error = xfs_btree_read_bufl(mp, cur->bc_tp, rbno, 0, &rbp, - XFS_BMAP_BTREE_REF))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - goto error0; - } - right = XFS_BUF_TO_BMBT_BLOCK(rbp); - if ((error = xfs_btree_check_lblock(cur, right, level, rbp))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - goto error0; - } - lrecs = be16_to_cpu(left->bb_numrecs); - } else { - if (level > 0 && (error = xfs_bmbt_decrement(cur, level, &i))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - goto error0; - } - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *stat = 1; - return 0; - } - numlrecs = be16_to_cpu(left->bb_numrecs); - numrrecs = be16_to_cpu(right->bb_numrecs); - if (level > 0) { - lkp = XFS_BMAP_KEY_IADDR(left, numlrecs + 1, cur); - lpp = XFS_BMAP_PTR_IADDR(left, numlrecs + 1, cur); - rkp = XFS_BMAP_KEY_IADDR(right, 1, cur); - rpp = XFS_BMAP_PTR_IADDR(right, 1, cur); -#ifdef DEBUG - for (i = 0; i < numrrecs; i++) { - if ((error = xfs_btree_check_lptr_disk(cur, rpp[i], level))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - goto error0; - } - } -#endif - memcpy(lkp, rkp, numrrecs * sizeof(*lkp)); - memcpy(lpp, rpp, numrrecs * sizeof(*lpp)); - xfs_bmbt_log_keys(cur, lbp, numlrecs + 1, numlrecs + numrrecs); - xfs_bmbt_log_ptrs(cur, lbp, numlrecs + 1, numlrecs + numrrecs); - } else { - lrp = XFS_BMAP_REC_IADDR(left, numlrecs + 1, cur); - rrp = XFS_BMAP_REC_IADDR(right, 1, cur); - memcpy(lrp, rrp, numrrecs * sizeof(*lrp)); - xfs_bmbt_log_recs(cur, lbp, numlrecs + 1, numlrecs + numrrecs); - } - be16_add_cpu(&left->bb_numrecs, numrrecs); - left->bb_rightsib = right->bb_rightsib; - xfs_bmbt_log_block(cur, lbp, XFS_BB_RIGHTSIB | XFS_BB_NUMRECS); - if (be64_to_cpu(left->bb_rightsib) != NULLDFSBNO) { - if ((error = xfs_btree_read_bufl(mp, cur->bc_tp, - be64_to_cpu(left->bb_rightsib), - 0, &rrbp, XFS_BMAP_BTREE_REF))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - goto error0; - } - rrblock = XFS_BUF_TO_BMBT_BLOCK(rrbp); - if ((error = xfs_btree_check_lblock(cur, rrblock, level, rrbp))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - goto error0; - } - rrblock->bb_leftsib = cpu_to_be64(lbno); - xfs_bmbt_log_block(cur, rrbp, XFS_BB_LEFTSIB); - } - xfs_bmap_add_free(XFS_DADDR_TO_FSB(mp, XFS_BUF_ADDR(rbp)), 1, - cur->bc_private.b.flist, mp); - cur->bc_private.b.ip->i_d.di_nblocks--; - xfs_trans_log_inode(cur->bc_tp, cur->bc_private.b.ip, XFS_ILOG_CORE); - XFS_TRANS_MOD_DQUOT_BYINO(mp, cur->bc_tp, cur->bc_private.b.ip, - XFS_TRANS_DQ_BCOUNT, -1L); - xfs_trans_binval(cur->bc_tp, rbp); - if (bp != lbp) { - cur->bc_bufs[level] = lbp; - cur->bc_ptrs[level] += lrecs; - cur->bc_ra[level] = 0; - } else if ((error = xfs_bmbt_increment(cur, level + 1, &i))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - goto error0; - } - if (level > 0) - cur->bc_ptrs[level]--; - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *stat = 2; - return 0; - -error0: - if (tcur) - xfs_btree_del_cursor(tcur, XFS_BTREE_ERROR); - return error; -} - -/* - * Insert one record/level. Return information to the caller - * allowing the next level up to proceed if necessary. - */ -STATIC int /* error */ -xfs_bmbt_insrec( - xfs_btree_cur_t *cur, - int level, - xfs_fsblock_t *bnop, - xfs_bmbt_rec_t *recp, - xfs_btree_cur_t **curp, - int *stat) /* no-go/done/continue */ -{ - xfs_bmbt_block_t *block; /* bmap btree block */ - xfs_buf_t *bp; /* buffer for block */ - int error; /* error return value */ - int i; /* loop index */ - xfs_bmbt_key_t key; /* bmap btree key */ - xfs_bmbt_key_t *kp=NULL; /* pointer to bmap btree key */ - int logflags; /* inode logging flags */ - xfs_fsblock_t nbno; /* new block number */ - struct xfs_btree_cur *ncur; /* new btree cursor */ - __uint64_t startoff; /* new btree key value */ - xfs_bmbt_rec_t nrec; /* new record count */ - int optr; /* old key/record index */ - xfs_bmbt_ptr_t *pp; /* pointer to bmap block addr */ - int ptr; /* key/record index */ - xfs_bmbt_rec_t *rp=NULL; /* pointer to bmap btree rec */ - int numrecs; - - ASSERT(level < cur->bc_nlevels); - XFS_BMBT_TRACE_CURSOR(cur, ENTRY); - XFS_BMBT_TRACE_ARGIFR(cur, level, *bnop, recp); - ncur = NULL; - key.br_startoff = cpu_to_be64(xfs_bmbt_disk_get_startoff(recp)); - optr = ptr = cur->bc_ptrs[level]; - if (ptr == 0) { - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *stat = 0; - return 0; - } - XFS_STATS_INC(xs_bmbt_insrec); - block = xfs_bmbt_get_block(cur, level, &bp); - numrecs = be16_to_cpu(block->bb_numrecs); -#ifdef DEBUG - if ((error = xfs_btree_check_lblock(cur, block, level, bp))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } - if (ptr <= numrecs) { - if (level == 0) { - rp = XFS_BMAP_REC_IADDR(block, ptr, cur); - xfs_btree_check_rec(XFS_BTNUM_BMAP, recp, rp); - } else { - kp = XFS_BMAP_KEY_IADDR(block, ptr, cur); - xfs_btree_check_key(XFS_BTNUM_BMAP, &key, kp); - } - } -#endif - nbno = NULLFSBLOCK; - if (numrecs == XFS_BMAP_BLOCK_IMAXRECS(level, cur)) { - if (numrecs < XFS_BMAP_BLOCK_DMAXRECS(level, cur)) { - /* - * A root block, that can be made bigger. - */ - xfs_iroot_realloc(cur->bc_private.b.ip, 1, - cur->bc_private.b.whichfork); - block = xfs_bmbt_get_block(cur, level, &bp); - } else if (level == cur->bc_nlevels - 1) { - if ((error = xfs_bmbt_newroot(cur, &logflags, stat)) || - *stat == 0) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } - xfs_trans_log_inode(cur->bc_tp, cur->bc_private.b.ip, - logflags); - block = xfs_bmbt_get_block(cur, level, &bp); - } else { - if ((error = xfs_bmbt_rshift(cur, level, &i))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } - if (i) { - /* nothing */ - } else { - if ((error = xfs_bmbt_lshift(cur, level, &i))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } - if (i) { - optr = ptr = cur->bc_ptrs[level]; - } else { - if ((error = xfs_bmbt_split(cur, level, - &nbno, &startoff, &ncur, - &i))) { - XFS_BMBT_TRACE_CURSOR(cur, - ERROR); - return error; - } - if (i) { - block = xfs_bmbt_get_block( - cur, level, &bp); -#ifdef DEBUG - if ((error = - xfs_btree_check_lblock(cur, - block, level, bp))) { - XFS_BMBT_TRACE_CURSOR( - cur, ERROR); - return error; - } -#endif - ptr = cur->bc_ptrs[level]; - xfs_bmbt_disk_set_allf(&nrec, - startoff, 0, 0, - XFS_EXT_NORM); - } else { - XFS_BMBT_TRACE_CURSOR(cur, - EXIT); - *stat = 0; - return 0; - } - } - } - } - } - numrecs = be16_to_cpu(block->bb_numrecs); - if (level > 0) { - kp = XFS_BMAP_KEY_IADDR(block, 1, cur); - pp = XFS_BMAP_PTR_IADDR(block, 1, cur); -#ifdef DEBUG - for (i = numrecs; i >= ptr; i--) { - if ((error = xfs_btree_check_lptr_disk(cur, pp[i - 1], - level))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } - } -#endif - memmove(&kp[ptr], &kp[ptr - 1], - (numrecs - ptr + 1) * sizeof(*kp)); - memmove(&pp[ptr], &pp[ptr - 1], - (numrecs - ptr + 1) * sizeof(*pp)); -#ifdef DEBUG - if ((error = xfs_btree_check_lptr(cur, *bnop, level))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } -#endif - kp[ptr - 1] = key; - pp[ptr - 1] = cpu_to_be64(*bnop); - numrecs++; - block->bb_numrecs = cpu_to_be16(numrecs); - xfs_bmbt_log_keys(cur, bp, ptr, numrecs); - xfs_bmbt_log_ptrs(cur, bp, ptr, numrecs); - } else { - rp = XFS_BMAP_REC_IADDR(block, 1, cur); - memmove(&rp[ptr], &rp[ptr - 1], - (numrecs - ptr + 1) * sizeof(*rp)); - rp[ptr - 1] = *recp; - numrecs++; - block->bb_numrecs = cpu_to_be16(numrecs); - xfs_bmbt_log_recs(cur, bp, ptr, numrecs); - } - xfs_bmbt_log_block(cur, bp, XFS_BB_NUMRECS); -#ifdef DEBUG - if (ptr < numrecs) { - if (level == 0) - xfs_btree_check_rec(XFS_BTNUM_BMAP, rp + ptr - 1, - rp + ptr); - else - xfs_btree_check_key(XFS_BTNUM_BMAP, kp + ptr - 1, - kp + ptr); - } -#endif - if (optr == 1 && (error = xfs_bmbt_updkey(cur, &key, level + 1))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } - *bnop = nbno; - if (nbno != NULLFSBLOCK) { - *recp = nrec; - *curp = ncur; - } - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *stat = 1; - return 0; -} - -STATIC int -xfs_bmbt_killroot( - xfs_btree_cur_t *cur) -{ - xfs_bmbt_block_t *block; - xfs_bmbt_block_t *cblock; - xfs_buf_t *cbp; - xfs_bmbt_key_t *ckp; - xfs_bmbt_ptr_t *cpp; -#ifdef DEBUG - int error; -#endif - int i; - xfs_bmbt_key_t *kp; - xfs_inode_t *ip; - xfs_ifork_t *ifp; - int level; - xfs_bmbt_ptr_t *pp; - - XFS_BMBT_TRACE_CURSOR(cur, ENTRY); - level = cur->bc_nlevels - 1; - ASSERT(level >= 1); - /* - * Don't deal with the root block needs to be a leaf case. - * We're just going to turn the thing back into extents anyway. - */ - if (level == 1) { - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - return 0; - } - block = xfs_bmbt_get_block(cur, level, &cbp); - /* - * Give up if the root has multiple children. - */ - if (be16_to_cpu(block->bb_numrecs) != 1) { - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - return 0; - } - /* - * Only do this if the next level will fit. - * Then the data must be copied up to the inode, - * instead of freeing the root you free the next level. - */ - cbp = cur->bc_bufs[level - 1]; - cblock = XFS_BUF_TO_BMBT_BLOCK(cbp); - if (be16_to_cpu(cblock->bb_numrecs) > XFS_BMAP_BLOCK_DMAXRECS(level, cur)) { - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - return 0; - } - ASSERT(be64_to_cpu(cblock->bb_leftsib) == NULLDFSBNO); - ASSERT(be64_to_cpu(cblock->bb_rightsib) == NULLDFSBNO); - ip = cur->bc_private.b.ip; - ifp = XFS_IFORK_PTR(ip, cur->bc_private.b.whichfork); - ASSERT(XFS_BMAP_BLOCK_IMAXRECS(level, cur) == - XFS_BMAP_BROOT_MAXRECS(ifp->if_broot_bytes)); - i = (int)(be16_to_cpu(cblock->bb_numrecs) - XFS_BMAP_BLOCK_IMAXRECS(level, cur)); - if (i) { - xfs_iroot_realloc(ip, i, cur->bc_private.b.whichfork); - block = ifp->if_broot; - } - be16_add_cpu(&block->bb_numrecs, i); - ASSERT(block->bb_numrecs == cblock->bb_numrecs); - kp = XFS_BMAP_KEY_IADDR(block, 1, cur); - ckp = XFS_BMAP_KEY_IADDR(cblock, 1, cur); - memcpy(kp, ckp, be16_to_cpu(block->bb_numrecs) * sizeof(*kp)); - pp = XFS_BMAP_PTR_IADDR(block, 1, cur); - cpp = XFS_BMAP_PTR_IADDR(cblock, 1, cur); -#ifdef DEBUG - for (i = 0; i < be16_to_cpu(cblock->bb_numrecs); i++) { - if ((error = xfs_btree_check_lptr_disk(cur, cpp[i], level - 1))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } - } -#endif - memcpy(pp, cpp, be16_to_cpu(block->bb_numrecs) * sizeof(*pp)); - xfs_bmap_add_free(XFS_DADDR_TO_FSB(cur->bc_mp, XFS_BUF_ADDR(cbp)), 1, - cur->bc_private.b.flist, cur->bc_mp); - ip->i_d.di_nblocks--; - XFS_TRANS_MOD_DQUOT_BYINO(cur->bc_mp, cur->bc_tp, ip, - XFS_TRANS_DQ_BCOUNT, -1L); - xfs_trans_binval(cur->bc_tp, cbp); - cur->bc_bufs[level - 1] = NULL; - be16_add_cpu(&block->bb_level, -1); - xfs_trans_log_inode(cur->bc_tp, ip, - XFS_ILOG_CORE | XFS_ILOG_FBROOT(cur->bc_private.b.whichfork)); - cur->bc_nlevels--; - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - return 0; -} - -/* - * Log key values from the btree block. - */ -STATIC void -xfs_bmbt_log_keys( - xfs_btree_cur_t *cur, - xfs_buf_t *bp, - int kfirst, - int klast) -{ - xfs_trans_t *tp; - - XFS_BMBT_TRACE_CURSOR(cur, ENTRY); - XFS_BMBT_TRACE_ARGBII(cur, bp, kfirst, klast); - tp = cur->bc_tp; - if (bp) { - xfs_bmbt_block_t *block; - int first; - xfs_bmbt_key_t *kp; - int last; - - block = XFS_BUF_TO_BMBT_BLOCK(bp); - kp = XFS_BMAP_KEY_DADDR(block, 1, cur); - first = (int)((xfs_caddr_t)&kp[kfirst - 1] - (xfs_caddr_t)block); - last = (int)(((xfs_caddr_t)&kp[klast] - 1) - (xfs_caddr_t)block); - xfs_trans_log_buf(tp, bp, first, last); - } else { - xfs_inode_t *ip; - - ip = cur->bc_private.b.ip; - xfs_trans_log_inode(tp, ip, - XFS_ILOG_FBROOT(cur->bc_private.b.whichfork)); - } - XFS_BMBT_TRACE_CURSOR(cur, EXIT); -} - -/* - * Log pointer values from the btree block. - */ -STATIC void -xfs_bmbt_log_ptrs( - xfs_btree_cur_t *cur, - xfs_buf_t *bp, - int pfirst, - int plast) -{ - xfs_trans_t *tp; - - XFS_BMBT_TRACE_CURSOR(cur, ENTRY); - XFS_BMBT_TRACE_ARGBII(cur, bp, pfirst, plast); - tp = cur->bc_tp; - if (bp) { - xfs_bmbt_block_t *block; - int first; - int last; - xfs_bmbt_ptr_t *pp; - - block = XFS_BUF_TO_BMBT_BLOCK(bp); - pp = XFS_BMAP_PTR_DADDR(block, 1, cur); - first = (int)((xfs_caddr_t)&pp[pfirst - 1] - (xfs_caddr_t)block); - last = (int)(((xfs_caddr_t)&pp[plast] - 1) - (xfs_caddr_t)block); - xfs_trans_log_buf(tp, bp, first, last); - } else { - xfs_inode_t *ip; - - ip = cur->bc_private.b.ip; - xfs_trans_log_inode(tp, ip, - XFS_ILOG_FBROOT(cur->bc_private.b.whichfork)); - } - XFS_BMBT_TRACE_CURSOR(cur, EXIT); -} - -/* - * Lookup the record. The cursor is made to point to it, based on dir. - */ -STATIC int /* error */ -xfs_bmbt_lookup( - xfs_btree_cur_t *cur, - xfs_lookup_t dir, - int *stat) /* success/failure */ -{ - xfs_bmbt_block_t *block=NULL; - xfs_buf_t *bp; - xfs_daddr_t d; - xfs_sfiloff_t diff; - int error; /* error return value */ - xfs_fsblock_t fsbno=0; - int high; - int i; - int keyno=0; - xfs_bmbt_key_t *kkbase=NULL; - xfs_bmbt_key_t *kkp; - xfs_bmbt_rec_t *krbase=NULL; - xfs_bmbt_rec_t *krp; - int level; - int low; - xfs_mount_t *mp; - xfs_bmbt_ptr_t *pp; - xfs_bmbt_irec_t *rp; - xfs_fileoff_t startoff; - xfs_trans_t *tp; - - XFS_STATS_INC(xs_bmbt_lookup); - XFS_BMBT_TRACE_CURSOR(cur, ENTRY); - XFS_BMBT_TRACE_ARGI(cur, (int)dir); - tp = cur->bc_tp; - mp = cur->bc_mp; - rp = &cur->bc_rec.b; - for (level = cur->bc_nlevels - 1, diff = 1; level >= 0; level--) { - if (level < cur->bc_nlevels - 1) { - d = XFS_FSB_TO_DADDR(mp, fsbno); - bp = cur->bc_bufs[level]; - if (bp && XFS_BUF_ADDR(bp) != d) - bp = NULL; - if (!bp) { - if ((error = xfs_btree_read_bufl(mp, tp, fsbno, - 0, &bp, XFS_BMAP_BTREE_REF))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } - xfs_btree_setbuf(cur, level, bp); - block = XFS_BUF_TO_BMBT_BLOCK(bp); - if ((error = xfs_btree_check_lblock(cur, block, - level, bp))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } - } else - block = XFS_BUF_TO_BMBT_BLOCK(bp); - } else - block = xfs_bmbt_get_block(cur, level, &bp); - if (diff == 0) - keyno = 1; - else { - if (level > 0) - kkbase = XFS_BMAP_KEY_IADDR(block, 1, cur); - else - krbase = XFS_BMAP_REC_IADDR(block, 1, cur); - low = 1; - if (!(high = be16_to_cpu(block->bb_numrecs))) { - ASSERT(level == 0); - cur->bc_ptrs[0] = dir != XFS_LOOKUP_LE; - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *stat = 0; - return 0; - } - while (low <= high) { - XFS_STATS_INC(xs_bmbt_compare); - keyno = (low + high) >> 1; - if (level > 0) { - kkp = kkbase + keyno - 1; - startoff = be64_to_cpu(kkp->br_startoff); - } else { - krp = krbase + keyno - 1; - startoff = xfs_bmbt_disk_get_startoff(krp); - } - diff = (xfs_sfiloff_t) - (startoff - rp->br_startoff); - if (diff < 0) - low = keyno + 1; - else if (diff > 0) - high = keyno - 1; - else - break; - } - } - if (level > 0) { - if (diff > 0 && --keyno < 1) - keyno = 1; - pp = XFS_BMAP_PTR_IADDR(block, keyno, cur); - fsbno = be64_to_cpu(*pp); -#ifdef DEBUG - if ((error = xfs_btree_check_lptr(cur, fsbno, level))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } -#endif - cur->bc_ptrs[level] = keyno; - } - } - if (dir != XFS_LOOKUP_LE && diff < 0) { - keyno++; - /* - * If ge search and we went off the end of the block, but it's - * not the last block, we're in the wrong block. - */ - if (dir == XFS_LOOKUP_GE && keyno > be16_to_cpu(block->bb_numrecs) && - be64_to_cpu(block->bb_rightsib) != NULLDFSBNO) { - cur->bc_ptrs[0] = keyno; - if ((error = xfs_bmbt_increment(cur, 0, &i))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } - XFS_WANT_CORRUPTED_RETURN(i == 1); - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *stat = 1; - return 0; - } - } - else if (dir == XFS_LOOKUP_LE && diff > 0) - keyno--; - cur->bc_ptrs[0] = keyno; - if (keyno == 0 || keyno > be16_to_cpu(block->bb_numrecs)) { - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *stat = 0; - } else { - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *stat = ((dir != XFS_LOOKUP_EQ) || (diff == 0)); - } - return 0; -} - -/* - * Move 1 record left from cur/level if possible. - * Update cur to reflect the new path. - */ -STATIC int /* error */ -xfs_bmbt_lshift( - xfs_btree_cur_t *cur, - int level, - int *stat) /* success/failure */ -{ - int error; /* error return value */ -#ifdef DEBUG - int i; /* loop counter */ -#endif - xfs_bmbt_key_t key; /* bmap btree key */ - xfs_buf_t *lbp; /* left buffer pointer */ - xfs_bmbt_block_t *left; /* left btree block */ - xfs_bmbt_key_t *lkp=NULL; /* left btree key */ - xfs_bmbt_ptr_t *lpp; /* left address pointer */ - int lrecs; /* left record count */ - xfs_bmbt_rec_t *lrp=NULL; /* left record pointer */ - xfs_mount_t *mp; /* file system mount point */ - xfs_buf_t *rbp; /* right buffer pointer */ - xfs_bmbt_block_t *right; /* right btree block */ - xfs_bmbt_key_t *rkp=NULL; /* right btree key */ - xfs_bmbt_ptr_t *rpp=NULL; /* right address pointer */ - xfs_bmbt_rec_t *rrp=NULL; /* right record pointer */ - int rrecs; /* right record count */ - - XFS_BMBT_TRACE_CURSOR(cur, ENTRY); - XFS_BMBT_TRACE_ARGI(cur, level); - if (level == cur->bc_nlevels - 1) { - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *stat = 0; - return 0; - } - rbp = cur->bc_bufs[level]; - right = XFS_BUF_TO_BMBT_BLOCK(rbp); -#ifdef DEBUG - if ((error = xfs_btree_check_lblock(cur, right, level, rbp))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } -#endif - if (be64_to_cpu(right->bb_leftsib) == NULLDFSBNO) { - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *stat = 0; - return 0; - } - if (cur->bc_ptrs[level] <= 1) { - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *stat = 0; - return 0; - } - mp = cur->bc_mp; - if ((error = xfs_btree_read_bufl(mp, cur->bc_tp, be64_to_cpu(right->bb_leftsib), 0, - &lbp, XFS_BMAP_BTREE_REF))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } - left = XFS_BUF_TO_BMBT_BLOCK(lbp); - if ((error = xfs_btree_check_lblock(cur, left, level, lbp))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } - if (be16_to_cpu(left->bb_numrecs) == XFS_BMAP_BLOCK_IMAXRECS(level, cur)) { - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *stat = 0; - return 0; - } - lrecs = be16_to_cpu(left->bb_numrecs) + 1; - if (level > 0) { - lkp = XFS_BMAP_KEY_IADDR(left, lrecs, cur); - rkp = XFS_BMAP_KEY_IADDR(right, 1, cur); - *lkp = *rkp; - xfs_bmbt_log_keys(cur, lbp, lrecs, lrecs); - lpp = XFS_BMAP_PTR_IADDR(left, lrecs, cur); - rpp = XFS_BMAP_PTR_IADDR(right, 1, cur); -#ifdef DEBUG - if ((error = xfs_btree_check_lptr_disk(cur, *rpp, level))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } -#endif - *lpp = *rpp; - xfs_bmbt_log_ptrs(cur, lbp, lrecs, lrecs); - } else { - lrp = XFS_BMAP_REC_IADDR(left, lrecs, cur); - rrp = XFS_BMAP_REC_IADDR(right, 1, cur); - *lrp = *rrp; - xfs_bmbt_log_recs(cur, lbp, lrecs, lrecs); - } - left->bb_numrecs = cpu_to_be16(lrecs); - xfs_bmbt_log_block(cur, lbp, XFS_BB_NUMRECS); -#ifdef DEBUG - if (level > 0) - xfs_btree_check_key(XFS_BTNUM_BMAP, lkp - 1, lkp); - else - xfs_btree_check_rec(XFS_BTNUM_BMAP, lrp - 1, lrp); -#endif - rrecs = be16_to_cpu(right->bb_numrecs) - 1; - right->bb_numrecs = cpu_to_be16(rrecs); - xfs_bmbt_log_block(cur, rbp, XFS_BB_NUMRECS); - if (level > 0) { -#ifdef DEBUG - for (i = 0; i < rrecs; i++) { - if ((error = xfs_btree_check_lptr_disk(cur, rpp[i + 1], - level))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } - } -#endif - memmove(rkp, rkp + 1, rrecs * sizeof(*rkp)); - memmove(rpp, rpp + 1, rrecs * sizeof(*rpp)); - xfs_bmbt_log_keys(cur, rbp, 1, rrecs); - xfs_bmbt_log_ptrs(cur, rbp, 1, rrecs); - } else { - memmove(rrp, rrp + 1, rrecs * sizeof(*rrp)); - xfs_bmbt_log_recs(cur, rbp, 1, rrecs); - key.br_startoff = cpu_to_be64(xfs_bmbt_disk_get_startoff(rrp)); - rkp = &key; - } - if ((error = xfs_bmbt_updkey(cur, rkp, level + 1))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } - cur->bc_ptrs[level]--; - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *stat = 1; - return 0; -} - -/* - * Move 1 record right from cur/level if possible. - * Update cur to reflect the new path. - */ -STATIC int /* error */ -xfs_bmbt_rshift( - xfs_btree_cur_t *cur, - int level, - int *stat) /* success/failure */ -{ - int error; /* error return value */ - int i; /* loop counter */ - xfs_bmbt_key_t key; /* bmap btree key */ - xfs_buf_t *lbp; /* left buffer pointer */ - xfs_bmbt_block_t *left; /* left btree block */ - xfs_bmbt_key_t *lkp; /* left btree key */ - xfs_bmbt_ptr_t *lpp; /* left address pointer */ - xfs_bmbt_rec_t *lrp; /* left record pointer */ - xfs_mount_t *mp; /* file system mount point */ - xfs_buf_t *rbp; /* right buffer pointer */ - xfs_bmbt_block_t *right; /* right btree block */ - xfs_bmbt_key_t *rkp; /* right btree key */ - xfs_bmbt_ptr_t *rpp; /* right address pointer */ - xfs_bmbt_rec_t *rrp=NULL; /* right record pointer */ - struct xfs_btree_cur *tcur; /* temporary btree cursor */ - - XFS_BMBT_TRACE_CURSOR(cur, ENTRY); - XFS_BMBT_TRACE_ARGI(cur, level); - if (level == cur->bc_nlevels - 1) { - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *stat = 0; - return 0; - } - lbp = cur->bc_bufs[level]; - left = XFS_BUF_TO_BMBT_BLOCK(lbp); -#ifdef DEBUG - if ((error = xfs_btree_check_lblock(cur, left, level, lbp))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } -#endif - if (be64_to_cpu(left->bb_rightsib) == NULLDFSBNO) { - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *stat = 0; - return 0; - } - if (cur->bc_ptrs[level] >= be16_to_cpu(left->bb_numrecs)) { - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *stat = 0; - return 0; - } - mp = cur->bc_mp; - if ((error = xfs_btree_read_bufl(mp, cur->bc_tp, be64_to_cpu(left->bb_rightsib), 0, - &rbp, XFS_BMAP_BTREE_REF))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } - right = XFS_BUF_TO_BMBT_BLOCK(rbp); - if ((error = xfs_btree_check_lblock(cur, right, level, rbp))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } - if (be16_to_cpu(right->bb_numrecs) == XFS_BMAP_BLOCK_IMAXRECS(level, cur)) { - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *stat = 0; - return 0; - } - if (level > 0) { - lkp = XFS_BMAP_KEY_IADDR(left, be16_to_cpu(left->bb_numrecs), cur); - lpp = XFS_BMAP_PTR_IADDR(left, be16_to_cpu(left->bb_numrecs), cur); - rkp = XFS_BMAP_KEY_IADDR(right, 1, cur); - rpp = XFS_BMAP_PTR_IADDR(right, 1, cur); -#ifdef DEBUG - for (i = be16_to_cpu(right->bb_numrecs) - 1; i >= 0; i--) { - if ((error = xfs_btree_check_lptr_disk(cur, rpp[i], level))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } - } -#endif - memmove(rkp + 1, rkp, be16_to_cpu(right->bb_numrecs) * sizeof(*rkp)); - memmove(rpp + 1, rpp, be16_to_cpu(right->bb_numrecs) * sizeof(*rpp)); -#ifdef DEBUG - if ((error = xfs_btree_check_lptr_disk(cur, *lpp, level))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } -#endif - *rkp = *lkp; - *rpp = *lpp; - xfs_bmbt_log_keys(cur, rbp, 1, be16_to_cpu(right->bb_numrecs) + 1); - xfs_bmbt_log_ptrs(cur, rbp, 1, be16_to_cpu(right->bb_numrecs) + 1); - } else { - lrp = XFS_BMAP_REC_IADDR(left, be16_to_cpu(left->bb_numrecs), cur); - rrp = XFS_BMAP_REC_IADDR(right, 1, cur); - memmove(rrp + 1, rrp, be16_to_cpu(right->bb_numrecs) * sizeof(*rrp)); - *rrp = *lrp; - xfs_bmbt_log_recs(cur, rbp, 1, be16_to_cpu(right->bb_numrecs) + 1); - key.br_startoff = cpu_to_be64(xfs_bmbt_disk_get_startoff(rrp)); - rkp = &key; - } - be16_add_cpu(&left->bb_numrecs, -1); - xfs_bmbt_log_block(cur, lbp, XFS_BB_NUMRECS); - be16_add_cpu(&right->bb_numrecs, 1); -#ifdef DEBUG - if (level > 0) - xfs_btree_check_key(XFS_BTNUM_BMAP, rkp, rkp + 1); - else - xfs_btree_check_rec(XFS_BTNUM_BMAP, rrp, rrp + 1); -#endif - xfs_bmbt_log_block(cur, rbp, XFS_BB_NUMRECS); - if ((error = xfs_btree_dup_cursor(cur, &tcur))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } - i = xfs_btree_lastrec(tcur, level); - XFS_WANT_CORRUPTED_GOTO(i == 1, error0); - if ((error = xfs_bmbt_increment(tcur, level, &i))) { - XFS_BMBT_TRACE_CURSOR(tcur, ERROR); - goto error1; - } - XFS_WANT_CORRUPTED_GOTO(i == 1, error0); - if ((error = xfs_bmbt_updkey(tcur, rkp, level + 1))) { - XFS_BMBT_TRACE_CURSOR(tcur, ERROR); - goto error1; - } - xfs_btree_del_cursor(tcur, XFS_BTREE_NOERROR); - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *stat = 1; - return 0; -error0: - XFS_BMBT_TRACE_CURSOR(cur, ERROR); -error1: - xfs_btree_del_cursor(tcur, XFS_BTREE_ERROR); - return error; -} - /* * Determine the extent state. */ @@ -1453,229 +60,15 @@ xfs_extent_state( return XFS_EXT_NORM; } - -/* - * Split cur/level block in half. - * Return new block number and its first record (to be inserted into parent). - */ -STATIC int /* error */ -xfs_bmbt_split( - xfs_btree_cur_t *cur, - int level, - xfs_fsblock_t *bnop, - __uint64_t *startoff, - xfs_btree_cur_t **curp, - int *stat) /* success/failure */ -{ - xfs_alloc_arg_t args; /* block allocation args */ - int error; /* error return value */ - int i; /* loop counter */ - xfs_fsblock_t lbno; /* left sibling block number */ - xfs_buf_t *lbp; /* left buffer pointer */ - xfs_bmbt_block_t *left; /* left btree block */ - xfs_bmbt_key_t *lkp; /* left btree key */ - xfs_bmbt_ptr_t *lpp; /* left address pointer */ - xfs_bmbt_rec_t *lrp; /* left record pointer */ - xfs_buf_t *rbp; /* right buffer pointer */ - xfs_bmbt_block_t *right; /* right btree block */ - xfs_bmbt_key_t *rkp; /* right btree key */ - xfs_bmbt_ptr_t *rpp; /* right address pointer */ - xfs_bmbt_block_t *rrblock; /* right-right btree block */ - xfs_buf_t *rrbp; /* right-right buffer pointer */ - xfs_bmbt_rec_t *rrp; /* right record pointer */ - - XFS_BMBT_TRACE_CURSOR(cur, ENTRY); - XFS_BMBT_TRACE_ARGIFK(cur, level, *bnop, *startoff); - args.tp = cur->bc_tp; - args.mp = cur->bc_mp; - lbp = cur->bc_bufs[level]; - lbno = XFS_DADDR_TO_FSB(args.mp, XFS_BUF_ADDR(lbp)); - left = XFS_BUF_TO_BMBT_BLOCK(lbp); - args.fsbno = cur->bc_private.b.firstblock; - args.firstblock = args.fsbno; - args.minleft = 0; - if (args.fsbno == NULLFSBLOCK) { - args.fsbno = lbno; - args.type = XFS_ALLOCTYPE_START_BNO; - /* - * Make sure there is sufficient room left in the AG to - * complete a full tree split for an extent insert. If - * we are converting the middle part of an extent then - * we may need space for two tree splits. - * - * We are relying on the caller to make the correct block - * reservation for this operation to succeed. If the - * reservation amount is insufficient then we may fail a - * block allocation here and corrupt the filesystem. - */ - args.minleft = xfs_trans_get_block_res(args.tp); - } else if (cur->bc_private.b.flist->xbf_low) - args.type = XFS_ALLOCTYPE_START_BNO; - else - args.type = XFS_ALLOCTYPE_NEAR_BNO; - args.mod = args.alignment = args.total = args.isfl = - args.userdata = args.minalignslop = 0; - args.minlen = args.maxlen = args.prod = 1; - args.wasdel = cur->bc_private.b.flags & XFS_BTCUR_BPRV_WASDEL; - if (!args.wasdel && xfs_trans_get_block_res(args.tp) == 0) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return XFS_ERROR(ENOSPC); - } - if ((error = xfs_alloc_vextent(&args))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } - if (args.fsbno == NULLFSBLOCK && args.minleft) { - /* - * Could not find an AG with enough free space to satisfy - * a full btree split. Try again without minleft and if - * successful activate the lowspace algorithm. - */ - args.fsbno = 0; - args.type = XFS_ALLOCTYPE_FIRST_AG; - args.minleft = 0; - if ((error = xfs_alloc_vextent(&args))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } - cur->bc_private.b.flist->xbf_low = 1; - } - if (args.fsbno == NULLFSBLOCK) { - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *stat = 0; - return 0; - } - ASSERT(args.len == 1); - cur->bc_private.b.firstblock = args.fsbno; - cur->bc_private.b.allocated++; - cur->bc_private.b.ip->i_d.di_nblocks++; - xfs_trans_log_inode(args.tp, cur->bc_private.b.ip, XFS_ILOG_CORE); - XFS_TRANS_MOD_DQUOT_BYINO(args.mp, args.tp, cur->bc_private.b.ip, - XFS_TRANS_DQ_BCOUNT, 1L); - rbp = xfs_btree_get_bufl(args.mp, args.tp, args.fsbno, 0); - right = XFS_BUF_TO_BMBT_BLOCK(rbp); -#ifdef DEBUG - if ((error = xfs_btree_check_lblock(cur, left, level, rbp))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } -#endif - right->bb_magic = cpu_to_be32(XFS_BMAP_MAGIC); - right->bb_level = left->bb_level; - right->bb_numrecs = cpu_to_be16(be16_to_cpu(left->bb_numrecs) / 2); - if ((be16_to_cpu(left->bb_numrecs) & 1) && - cur->bc_ptrs[level] <= be16_to_cpu(right->bb_numrecs) + 1) - be16_add_cpu(&right->bb_numrecs, 1); - i = be16_to_cpu(left->bb_numrecs) - be16_to_cpu(right->bb_numrecs) + 1; - if (level > 0) { - lkp = XFS_BMAP_KEY_IADDR(left, i, cur); - lpp = XFS_BMAP_PTR_IADDR(left, i, cur); - rkp = XFS_BMAP_KEY_IADDR(right, 1, cur); - rpp = XFS_BMAP_PTR_IADDR(right, 1, cur); -#ifdef DEBUG - for (i = 0; i < be16_to_cpu(right->bb_numrecs); i++) { - if ((error = xfs_btree_check_lptr_disk(cur, lpp[i], level))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } - } -#endif - memcpy(rkp, lkp, be16_to_cpu(right->bb_numrecs) * sizeof(*rkp)); - memcpy(rpp, lpp, be16_to_cpu(right->bb_numrecs) * sizeof(*rpp)); - xfs_bmbt_log_keys(cur, rbp, 1, be16_to_cpu(right->bb_numrecs)); - xfs_bmbt_log_ptrs(cur, rbp, 1, be16_to_cpu(right->bb_numrecs)); - *startoff = be64_to_cpu(rkp->br_startoff); - } else { - lrp = XFS_BMAP_REC_IADDR(left, i, cur); - rrp = XFS_BMAP_REC_IADDR(right, 1, cur); - memcpy(rrp, lrp, be16_to_cpu(right->bb_numrecs) * sizeof(*rrp)); - xfs_bmbt_log_recs(cur, rbp, 1, be16_to_cpu(right->bb_numrecs)); - *startoff = xfs_bmbt_disk_get_startoff(rrp); - } - be16_add_cpu(&left->bb_numrecs, -(be16_to_cpu(right->bb_numrecs))); - right->bb_rightsib = left->bb_rightsib; - left->bb_rightsib = cpu_to_be64(args.fsbno); - right->bb_leftsib = cpu_to_be64(lbno); - xfs_bmbt_log_block(cur, rbp, XFS_BB_ALL_BITS); - xfs_bmbt_log_block(cur, lbp, XFS_BB_NUMRECS | XFS_BB_RIGHTSIB); - if (be64_to_cpu(right->bb_rightsib) != NULLDFSBNO) { - if ((error = xfs_btree_read_bufl(args.mp, args.tp, - be64_to_cpu(right->bb_rightsib), 0, &rrbp, - XFS_BMAP_BTREE_REF))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } - rrblock = XFS_BUF_TO_BMBT_BLOCK(rrbp); - if ((error = xfs_btree_check_lblock(cur, rrblock, level, rrbp))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } - rrblock->bb_leftsib = cpu_to_be64(args.fsbno); - xfs_bmbt_log_block(cur, rrbp, XFS_BB_LEFTSIB); - } - if (cur->bc_ptrs[level] > be16_to_cpu(left->bb_numrecs) + 1) { - xfs_btree_setbuf(cur, level, rbp); - cur->bc_ptrs[level] -= be16_to_cpu(left->bb_numrecs); - } - if (level + 1 < cur->bc_nlevels) { - if ((error = xfs_btree_dup_cursor(cur, curp))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } - (*curp)->bc_ptrs[level + 1]++; - } - *bnop = args.fsbno; - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *stat = 1; - return 0; -} - - -/* - * Update keys for the record. - */ -STATIC int -xfs_bmbt_updkey( - xfs_btree_cur_t *cur, - xfs_bmbt_key_t *keyp, /* on-disk format */ - int level) -{ - xfs_bmbt_block_t *block; - xfs_buf_t *bp; -#ifdef DEBUG - int error; -#endif - xfs_bmbt_key_t *kp; - int ptr; - - ASSERT(level >= 1); - XFS_BMBT_TRACE_CURSOR(cur, ENTRY); - XFS_BMBT_TRACE_ARGIK(cur, level, keyp); - for (ptr = 1; ptr == 1 && level < cur->bc_nlevels; level++) { - block = xfs_bmbt_get_block(cur, level, &bp); -#ifdef DEBUG - if ((error = xfs_btree_check_lblock(cur, block, level, bp))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } -#endif - ptr = cur->bc_ptrs[level]; - kp = XFS_BMAP_KEY_IADDR(block, ptr, cur); - *kp = *keyp; - xfs_bmbt_log_keys(cur, bp, ptr, ptr); - } - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - return 0; -} - /* * Convert on-disk form of btree root to in-memory form. */ void xfs_bmdr_to_bmbt( + struct xfs_mount *mp, xfs_bmdr_block_t *dblock, int dblocklen, - xfs_bmbt_block_t *rblock, + struct xfs_btree_block *rblock, int rblocklen) { int dmxr; @@ -1688,129 +81,19 @@ xfs_bmdr_to_bmbt( rblock->bb_level = dblock->bb_level; ASSERT(be16_to_cpu(rblock->bb_level) > 0); rblock->bb_numrecs = dblock->bb_numrecs; - rblock->bb_leftsib = cpu_to_be64(NULLDFSBNO); - rblock->bb_rightsib = cpu_to_be64(NULLDFSBNO); - dmxr = (int)XFS_BTREE_BLOCK_MAXRECS(dblocklen, xfs_bmdr, 0); - fkp = XFS_BTREE_KEY_ADDR(xfs_bmdr, dblock, 1); - tkp = XFS_BMAP_BROOT_KEY_ADDR(rblock, 1, rblocklen); - fpp = XFS_BTREE_PTR_ADDR(xfs_bmdr, dblock, 1, dmxr); - tpp = XFS_BMAP_BROOT_PTR_ADDR(rblock, 1, rblocklen); + rblock->bb_u.l.bb_leftsib = cpu_to_be64(NULLDFSBNO); + rblock->bb_u.l.bb_rightsib = cpu_to_be64(NULLDFSBNO); + dmxr = xfs_bmdr_maxrecs(mp, dblocklen, 0); + fkp = XFS_BMDR_KEY_ADDR(dblock, 1); + tkp = XFS_BMBT_KEY_ADDR(mp, rblock, 1); + fpp = XFS_BMDR_PTR_ADDR(dblock, 1, dmxr); + tpp = XFS_BMAP_BROOT_PTR_ADDR(mp, rblock, 1, rblocklen); dmxr = be16_to_cpu(dblock->bb_numrecs); memcpy(tkp, fkp, sizeof(*fkp) * dmxr); memcpy(tpp, fpp, sizeof(*fpp) * dmxr); } /* - * Decrement cursor by one record at the level. - * For nonzero levels the leaf-ward information is untouched. - */ -int /* error */ -xfs_bmbt_decrement( - xfs_btree_cur_t *cur, - int level, - int *stat) /* success/failure */ -{ - xfs_bmbt_block_t *block; - xfs_buf_t *bp; - int error; /* error return value */ - xfs_fsblock_t fsbno; - int lev; - xfs_mount_t *mp; - xfs_trans_t *tp; - - XFS_BMBT_TRACE_CURSOR(cur, ENTRY); - XFS_BMBT_TRACE_ARGI(cur, level); - ASSERT(level < cur->bc_nlevels); - if (level < cur->bc_nlevels - 1) - xfs_btree_readahead(cur, level, XFS_BTCUR_LEFTRA); - if (--cur->bc_ptrs[level] > 0) { - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *stat = 1; - return 0; - } - block = xfs_bmbt_get_block(cur, level, &bp); -#ifdef DEBUG - if ((error = xfs_btree_check_lblock(cur, block, level, bp))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } -#endif - if (be64_to_cpu(block->bb_leftsib) == NULLDFSBNO) { - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *stat = 0; - return 0; - } - for (lev = level + 1; lev < cur->bc_nlevels; lev++) { - if (--cur->bc_ptrs[lev] > 0) - break; - if (lev < cur->bc_nlevels - 1) - xfs_btree_readahead(cur, lev, XFS_BTCUR_LEFTRA); - } - if (lev == cur->bc_nlevels) { - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *stat = 0; - return 0; - } - tp = cur->bc_tp; - mp = cur->bc_mp; - for (block = xfs_bmbt_get_block(cur, lev, &bp); lev > level; ) { - fsbno = be64_to_cpu(*XFS_BMAP_PTR_IADDR(block, cur->bc_ptrs[lev], cur)); - if ((error = xfs_btree_read_bufl(mp, tp, fsbno, 0, &bp, - XFS_BMAP_BTREE_REF))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } - lev--; - xfs_btree_setbuf(cur, lev, bp); - block = XFS_BUF_TO_BMBT_BLOCK(bp); - if ((error = xfs_btree_check_lblock(cur, block, lev, bp))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } - cur->bc_ptrs[lev] = be16_to_cpu(block->bb_numrecs); - } - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *stat = 1; - return 0; -} - -/* - * Delete the record pointed to by cur. - */ -int /* error */ -xfs_bmbt_delete( - xfs_btree_cur_t *cur, - int *stat) /* success/failure */ -{ - int error; /* error return value */ - int i; - int level; - - XFS_BMBT_TRACE_CURSOR(cur, ENTRY); - for (level = 0, i = 2; i == 2; level++) { - if ((error = xfs_bmbt_delrec(cur, level, &i))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } - } - if (i == 0) { - for (level = 1; level < cur->bc_nlevels; level++) { - if (cur->bc_ptrs[level] == 0) { - if ((error = xfs_bmbt_decrement(cur, level, - &i))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } - break; - } - } - } - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *stat = i; - return 0; -} - -/* * Convert a compressed bmap extent record to an uncompressed form. * This code must be in sync with the routines xfs_bmbt_get_startoff, * xfs_bmbt_get_startblock, xfs_bmbt_get_blockcount and xfs_bmbt_get_state. @@ -1864,31 +147,6 @@ xfs_bmbt_get_all( } /* - * Get the block pointer for the given level of the cursor. - * Fill in the buffer pointer, if applicable. - */ -xfs_bmbt_block_t * -xfs_bmbt_get_block( - xfs_btree_cur_t *cur, - int level, - xfs_buf_t **bpp) -{ - xfs_ifork_t *ifp; - xfs_bmbt_block_t *rval; - - if (level < cur->bc_nlevels - 1) { - *bpp = cur->bc_bufs[level]; - rval = XFS_BUF_TO_BMBT_BLOCK(*bpp); - } else { - *bpp = NULL; - ifp = XFS_IFORK_PTR(cur->bc_private.b.ip, - cur->bc_private.b.whichfork); - rval = ifp->if_broot; - } - return rval; -} - -/* * Extract the blockcount field from an in memory bmap extent record. */ xfs_filblks_t @@ -1950,7 +208,8 @@ xfs_bmbt_disk_get_all( xfs_bmbt_rec_t *r, xfs_bmbt_irec_t *s) { - __xfs_bmbt_get_all(be64_to_cpu(r->l0), be64_to_cpu(r->l1), s); + __xfs_bmbt_get_all(get_unaligned_be64(&r->l0), + get_unaligned_be64(&r->l1), s); } /* @@ -1974,348 +233,6 @@ xfs_bmbt_disk_get_startoff( XFS_MASK64LO(64 - BMBT_EXNTFLAG_BITLEN)) >> 9; } -/* - * Increment cursor by one record at the level. - * For nonzero levels the leaf-ward information is untouched. - */ -int /* error */ -xfs_bmbt_increment( - xfs_btree_cur_t *cur, - int level, - int *stat) /* success/failure */ -{ - xfs_bmbt_block_t *block; - xfs_buf_t *bp; - int error; /* error return value */ - xfs_fsblock_t fsbno; - int lev; - xfs_mount_t *mp; - xfs_trans_t *tp; - - XFS_BMBT_TRACE_CURSOR(cur, ENTRY); - XFS_BMBT_TRACE_ARGI(cur, level); - ASSERT(level < cur->bc_nlevels); - if (level < cur->bc_nlevels - 1) - xfs_btree_readahead(cur, level, XFS_BTCUR_RIGHTRA); - block = xfs_bmbt_get_block(cur, level, &bp); -#ifdef DEBUG - if ((error = xfs_btree_check_lblock(cur, block, level, bp))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } -#endif - if (++cur->bc_ptrs[level] <= be16_to_cpu(block->bb_numrecs)) { - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *stat = 1; - return 0; - } - if (be64_to_cpu(block->bb_rightsib) == NULLDFSBNO) { - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *stat = 0; - return 0; - } - for (lev = level + 1; lev < cur->bc_nlevels; lev++) { - block = xfs_bmbt_get_block(cur, lev, &bp); -#ifdef DEBUG - if ((error = xfs_btree_check_lblock(cur, block, lev, bp))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } -#endif - if (++cur->bc_ptrs[lev] <= be16_to_cpu(block->bb_numrecs)) - break; - if (lev < cur->bc_nlevels - 1) - xfs_btree_readahead(cur, lev, XFS_BTCUR_RIGHTRA); - } - if (lev == cur->bc_nlevels) { - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *stat = 0; - return 0; - } - tp = cur->bc_tp; - mp = cur->bc_mp; - for (block = xfs_bmbt_get_block(cur, lev, &bp); lev > level; ) { - fsbno = be64_to_cpu(*XFS_BMAP_PTR_IADDR(block, cur->bc_ptrs[lev], cur)); - if ((error = xfs_btree_read_bufl(mp, tp, fsbno, 0, &bp, - XFS_BMAP_BTREE_REF))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } - lev--; - xfs_btree_setbuf(cur, lev, bp); - block = XFS_BUF_TO_BMBT_BLOCK(bp); - if ((error = xfs_btree_check_lblock(cur, block, lev, bp))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } - cur->bc_ptrs[lev] = 1; - } - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *stat = 1; - return 0; -} - -/* - * Insert the current record at the point referenced by cur. - * - * A multi-level split of the tree on insert will invalidate the original - * cursor. All callers of this function should assume that the cursor is - * no longer valid and revalidate it. - */ -int /* error */ -xfs_bmbt_insert( - xfs_btree_cur_t *cur, - int *stat) /* success/failure */ -{ - int error; /* error return value */ - int i; - int level; - xfs_fsblock_t nbno; - xfs_btree_cur_t *ncur; - xfs_bmbt_rec_t nrec; - xfs_btree_cur_t *pcur; - - XFS_BMBT_TRACE_CURSOR(cur, ENTRY); - level = 0; - nbno = NULLFSBLOCK; - xfs_bmbt_disk_set_all(&nrec, &cur->bc_rec.b); - ncur = NULL; - pcur = cur; - do { - if ((error = xfs_bmbt_insrec(pcur, level++, &nbno, &nrec, &ncur, - &i))) { - if (pcur != cur) - xfs_btree_del_cursor(pcur, XFS_BTREE_ERROR); - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } - XFS_WANT_CORRUPTED_GOTO(i == 1, error0); - if (pcur != cur && (ncur || nbno == NULLFSBLOCK)) { - cur->bc_nlevels = pcur->bc_nlevels; - cur->bc_private.b.allocated += - pcur->bc_private.b.allocated; - pcur->bc_private.b.allocated = 0; - ASSERT((cur->bc_private.b.firstblock != NULLFSBLOCK) || - XFS_IS_REALTIME_INODE(cur->bc_private.b.ip)); - cur->bc_private.b.firstblock = - pcur->bc_private.b.firstblock; - ASSERT(cur->bc_private.b.flist == - pcur->bc_private.b.flist); - xfs_btree_del_cursor(pcur, XFS_BTREE_NOERROR); - } - if (ncur) { - pcur = ncur; - ncur = NULL; - } - } while (nbno != NULLFSBLOCK); - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *stat = i; - return 0; -error0: - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; -} - -/* - * Log fields from the btree block header. - */ -void -xfs_bmbt_log_block( - xfs_btree_cur_t *cur, - xfs_buf_t *bp, - int fields) -{ - int first; - int last; - xfs_trans_t *tp; - static const short offsets[] = { - offsetof(xfs_bmbt_block_t, bb_magic), - offsetof(xfs_bmbt_block_t, bb_level), - offsetof(xfs_bmbt_block_t, bb_numrecs), - offsetof(xfs_bmbt_block_t, bb_leftsib), - offsetof(xfs_bmbt_block_t, bb_rightsib), - sizeof(xfs_bmbt_block_t) - }; - - XFS_BMBT_TRACE_CURSOR(cur, ENTRY); - XFS_BMBT_TRACE_ARGBI(cur, bp, fields); - tp = cur->bc_tp; - if (bp) { - xfs_btree_offsets(fields, offsets, XFS_BB_NUM_BITS, &first, - &last); - xfs_trans_log_buf(tp, bp, first, last); - } else - xfs_trans_log_inode(tp, cur->bc_private.b.ip, - XFS_ILOG_FBROOT(cur->bc_private.b.whichfork)); - XFS_BMBT_TRACE_CURSOR(cur, EXIT); -} - -/* - * Log record values from the btree block. - */ -void -xfs_bmbt_log_recs( - xfs_btree_cur_t *cur, - xfs_buf_t *bp, - int rfirst, - int rlast) -{ - xfs_bmbt_block_t *block; - int first; - int last; - xfs_bmbt_rec_t *rp; - xfs_trans_t *tp; - - XFS_BMBT_TRACE_CURSOR(cur, ENTRY); - XFS_BMBT_TRACE_ARGBII(cur, bp, rfirst, rlast); - ASSERT(bp); - tp = cur->bc_tp; - block = XFS_BUF_TO_BMBT_BLOCK(bp); - rp = XFS_BMAP_REC_DADDR(block, 1, cur); - first = (int)((xfs_caddr_t)&rp[rfirst - 1] - (xfs_caddr_t)block); - last = (int)(((xfs_caddr_t)&rp[rlast] - 1) - (xfs_caddr_t)block); - xfs_trans_log_buf(tp, bp, first, last); - XFS_BMBT_TRACE_CURSOR(cur, EXIT); -} - -int /* error */ -xfs_bmbt_lookup_eq( - xfs_btree_cur_t *cur, - xfs_fileoff_t off, - xfs_fsblock_t bno, - xfs_filblks_t len, - int *stat) /* success/failure */ -{ - cur->bc_rec.b.br_startoff = off; - cur->bc_rec.b.br_startblock = bno; - cur->bc_rec.b.br_blockcount = len; - return xfs_bmbt_lookup(cur, XFS_LOOKUP_EQ, stat); -} - -int /* error */ -xfs_bmbt_lookup_ge( - xfs_btree_cur_t *cur, - xfs_fileoff_t off, - xfs_fsblock_t bno, - xfs_filblks_t len, - int *stat) /* success/failure */ -{ - cur->bc_rec.b.br_startoff = off; - cur->bc_rec.b.br_startblock = bno; - cur->bc_rec.b.br_blockcount = len; - return xfs_bmbt_lookup(cur, XFS_LOOKUP_GE, stat); -} - -/* - * Give the bmap btree a new root block. Copy the old broot contents - * down into a real block and make the broot point to it. - */ -int /* error */ -xfs_bmbt_newroot( - xfs_btree_cur_t *cur, /* btree cursor */ - int *logflags, /* logging flags for inode */ - int *stat) /* return status - 0 fail */ -{ - xfs_alloc_arg_t args; /* allocation arguments */ - xfs_bmbt_block_t *block; /* bmap btree block */ - xfs_buf_t *bp; /* buffer for block */ - xfs_bmbt_block_t *cblock; /* child btree block */ - xfs_bmbt_key_t *ckp; /* child key pointer */ - xfs_bmbt_ptr_t *cpp; /* child ptr pointer */ - int error; /* error return code */ -#ifdef DEBUG - int i; /* loop counter */ -#endif - xfs_bmbt_key_t *kp; /* pointer to bmap btree key */ - int level; /* btree level */ - xfs_bmbt_ptr_t *pp; /* pointer to bmap block addr */ - - XFS_BMBT_TRACE_CURSOR(cur, ENTRY); - level = cur->bc_nlevels - 1; - block = xfs_bmbt_get_block(cur, level, &bp); - /* - * Copy the root into a real block. - */ - args.mp = cur->bc_mp; - pp = XFS_BMAP_PTR_IADDR(block, 1, cur); - args.tp = cur->bc_tp; - args.fsbno = cur->bc_private.b.firstblock; - args.mod = args.minleft = args.alignment = args.total = args.isfl = - args.userdata = args.minalignslop = 0; - args.minlen = args.maxlen = args.prod = 1; - args.wasdel = cur->bc_private.b.flags & XFS_BTCUR_BPRV_WASDEL; - args.firstblock = args.fsbno; - if (args.fsbno == NULLFSBLOCK) { -#ifdef DEBUG - if ((error = xfs_btree_check_lptr_disk(cur, *pp, level))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } -#endif - args.fsbno = be64_to_cpu(*pp); - args.type = XFS_ALLOCTYPE_START_BNO; - } else if (cur->bc_private.b.flist->xbf_low) - args.type = XFS_ALLOCTYPE_START_BNO; - else - args.type = XFS_ALLOCTYPE_NEAR_BNO; - if ((error = xfs_alloc_vextent(&args))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } - if (args.fsbno == NULLFSBLOCK) { - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *stat = 0; - return 0; - } - ASSERT(args.len == 1); - cur->bc_private.b.firstblock = args.fsbno; - cur->bc_private.b.allocated++; - cur->bc_private.b.ip->i_d.di_nblocks++; - XFS_TRANS_MOD_DQUOT_BYINO(args.mp, args.tp, cur->bc_private.b.ip, - XFS_TRANS_DQ_BCOUNT, 1L); - bp = xfs_btree_get_bufl(args.mp, cur->bc_tp, args.fsbno, 0); - cblock = XFS_BUF_TO_BMBT_BLOCK(bp); - *cblock = *block; - be16_add_cpu(&block->bb_level, 1); - block->bb_numrecs = cpu_to_be16(1); - cur->bc_nlevels++; - cur->bc_ptrs[level + 1] = 1; - kp = XFS_BMAP_KEY_IADDR(block, 1, cur); - ckp = XFS_BMAP_KEY_IADDR(cblock, 1, cur); - memcpy(ckp, kp, be16_to_cpu(cblock->bb_numrecs) * sizeof(*kp)); - cpp = XFS_BMAP_PTR_IADDR(cblock, 1, cur); -#ifdef DEBUG - for (i = 0; i < be16_to_cpu(cblock->bb_numrecs); i++) { - if ((error = xfs_btree_check_lptr_disk(cur, pp[i], level))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } - } -#endif - memcpy(cpp, pp, be16_to_cpu(cblock->bb_numrecs) * sizeof(*pp)); -#ifdef DEBUG - if ((error = xfs_btree_check_lptr(cur, args.fsbno, level))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } -#endif - *pp = cpu_to_be64(args.fsbno); - xfs_iroot_realloc(cur->bc_private.b.ip, 1 - be16_to_cpu(cblock->bb_numrecs), - cur->bc_private.b.whichfork); - xfs_btree_setbuf(cur, level, bp); - /* - * Do all this logging at the end so that - * the root is at the right level. - */ - xfs_bmbt_log_block(cur, bp, XFS_BB_ALL_BITS); - xfs_bmbt_log_keys(cur, bp, 1, be16_to_cpu(cblock->bb_numrecs)); - xfs_bmbt_log_ptrs(cur, bp, 1, be16_to_cpu(cblock->bb_numrecs)); - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - *logflags |= - XFS_ILOG_CORE | XFS_ILOG_FBROOT(cur->bc_private.b.whichfork); - *stat = 1; - return 0; -} /* * Set all the fields in a bmap extent record from the arguments. @@ -2512,7 +429,8 @@ xfs_bmbt_set_state( */ void xfs_bmbt_to_bmdr( - xfs_bmbt_block_t *rblock, + struct xfs_mount *mp, + struct xfs_btree_block *rblock, int rblocklen, xfs_bmdr_block_t *dblock, int dblocklen) @@ -2524,67 +442,22 @@ xfs_bmbt_to_bmdr( __be64 *tpp; ASSERT(be32_to_cpu(rblock->bb_magic) == XFS_BMAP_MAGIC); - ASSERT(be64_to_cpu(rblock->bb_leftsib) == NULLDFSBNO); - ASSERT(be64_to_cpu(rblock->bb_rightsib) == NULLDFSBNO); + ASSERT(be64_to_cpu(rblock->bb_u.l.bb_leftsib) == NULLDFSBNO); + ASSERT(be64_to_cpu(rblock->bb_u.l.bb_rightsib) == NULLDFSBNO); ASSERT(be16_to_cpu(rblock->bb_level) > 0); dblock->bb_level = rblock->bb_level; dblock->bb_numrecs = rblock->bb_numrecs; - dmxr = (int)XFS_BTREE_BLOCK_MAXRECS(dblocklen, xfs_bmdr, 0); - fkp = XFS_BMAP_BROOT_KEY_ADDR(rblock, 1, rblocklen); - tkp = XFS_BTREE_KEY_ADDR(xfs_bmdr, dblock, 1); - fpp = XFS_BMAP_BROOT_PTR_ADDR(rblock, 1, rblocklen); - tpp = XFS_BTREE_PTR_ADDR(xfs_bmdr, dblock, 1, dmxr); + dmxr = xfs_bmdr_maxrecs(mp, dblocklen, 0); + fkp = XFS_BMBT_KEY_ADDR(mp, rblock, 1); + tkp = XFS_BMDR_KEY_ADDR(dblock, 1); + fpp = XFS_BMAP_BROOT_PTR_ADDR(mp, rblock, 1, rblocklen); + tpp = XFS_BMDR_PTR_ADDR(dblock, 1, dmxr); dmxr = be16_to_cpu(dblock->bb_numrecs); memcpy(tkp, fkp, sizeof(*fkp) * dmxr); memcpy(tpp, fpp, sizeof(*fpp) * dmxr); } /* - * Update the record to the passed values. - */ -int -xfs_bmbt_update( - xfs_btree_cur_t *cur, - xfs_fileoff_t off, - xfs_fsblock_t bno, - xfs_filblks_t len, - xfs_exntst_t state) -{ - xfs_bmbt_block_t *block; - xfs_buf_t *bp; - int error; - xfs_bmbt_key_t key; - int ptr; - xfs_bmbt_rec_t *rp; - - XFS_BMBT_TRACE_CURSOR(cur, ENTRY); - XFS_BMBT_TRACE_ARGFFFI(cur, (xfs_dfiloff_t)off, (xfs_dfsbno_t)bno, - (xfs_dfilblks_t)len, (int)state); - block = xfs_bmbt_get_block(cur, 0, &bp); -#ifdef DEBUG - if ((error = xfs_btree_check_lblock(cur, block, 0, bp))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } -#endif - ptr = cur->bc_ptrs[0]; - rp = XFS_BMAP_REC_IADDR(block, ptr, cur); - xfs_bmbt_disk_set_allf(rp, off, bno, len, state); - xfs_bmbt_log_recs(cur, bp, ptr, ptr); - if (ptr > 1) { - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - return 0; - } - key.br_startoff = cpu_to_be64(off); - if ((error = xfs_bmbt_updkey(cur, &key, 1))) { - XFS_BMBT_TRACE_CURSOR(cur, ERROR); - return error; - } - XFS_BMBT_TRACE_CURSOR(cur, EXIT); - return 0; -} - -/* * Check extent records, which have just been read, for * any bit in the extent flag field. ASSERT on debug * kernels, as this condition should not occur. @@ -2608,3 +481,451 @@ xfs_check_nostate_extents( } return 0; } + + +STATIC struct xfs_btree_cur * +xfs_bmbt_dup_cursor( + struct xfs_btree_cur *cur) +{ + struct xfs_btree_cur *new; + + new = xfs_bmbt_init_cursor(cur->bc_mp, cur->bc_tp, + cur->bc_private.b.ip, cur->bc_private.b.whichfork); + + /* + * Copy the firstblock, flist, and flags values, + * since init cursor doesn't get them. + */ + new->bc_private.b.firstblock = cur->bc_private.b.firstblock; + new->bc_private.b.flist = cur->bc_private.b.flist; + new->bc_private.b.flags = cur->bc_private.b.flags; + + return new; +} + +STATIC void +xfs_bmbt_update_cursor( + struct xfs_btree_cur *src, + struct xfs_btree_cur *dst) +{ + ASSERT((dst->bc_private.b.firstblock != NULLFSBLOCK) || + (dst->bc_private.b.ip->i_d.di_flags & XFS_DIFLAG_REALTIME)); + ASSERT(dst->bc_private.b.flist == src->bc_private.b.flist); + + dst->bc_private.b.allocated += src->bc_private.b.allocated; + dst->bc_private.b.firstblock = src->bc_private.b.firstblock; + + src->bc_private.b.allocated = 0; +} + +STATIC int +xfs_bmbt_alloc_block( + struct xfs_btree_cur *cur, + union xfs_btree_ptr *start, + union xfs_btree_ptr *new, + int length, + int *stat) +{ + xfs_alloc_arg_t args; /* block allocation args */ + int error; /* error return value */ + + memset(&args, 0, sizeof(args)); + args.tp = cur->bc_tp; + args.mp = cur->bc_mp; + args.fsbno = cur->bc_private.b.firstblock; + args.firstblock = args.fsbno; + + if (args.fsbno == NULLFSBLOCK) { + args.fsbno = be64_to_cpu(start->l); + args.type = XFS_ALLOCTYPE_START_BNO; + /* + * Make sure there is sufficient room left in the AG to + * complete a full tree split for an extent insert. If + * we are converting the middle part of an extent then + * we may need space for two tree splits. + * + * We are relying on the caller to make the correct block + * reservation for this operation to succeed. If the + * reservation amount is insufficient then we may fail a + * block allocation here and corrupt the filesystem. + */ + args.minleft = xfs_trans_get_block_res(args.tp); + } else if (cur->bc_private.b.flist->xbf_low) { + args.type = XFS_ALLOCTYPE_START_BNO; + } else { + args.type = XFS_ALLOCTYPE_NEAR_BNO; + } + + args.minlen = args.maxlen = args.prod = 1; + args.wasdel = cur->bc_private.b.flags & XFS_BTCUR_BPRV_WASDEL; + if (!args.wasdel && xfs_trans_get_block_res(args.tp) == 0) { + error = XFS_ERROR(ENOSPC); + goto error0; + } + error = xfs_alloc_vextent(&args); + if (error) + goto error0; + + if (args.fsbno == NULLFSBLOCK && args.minleft) { + /* + * Could not find an AG with enough free space to satisfy + * a full btree split. Try again without minleft and if + * successful activate the lowspace algorithm. + */ + args.fsbno = 0; + args.type = XFS_ALLOCTYPE_FIRST_AG; + args.minleft = 0; + error = xfs_alloc_vextent(&args); + if (error) + goto error0; + cur->bc_private.b.flist->xbf_low = 1; + } + if (args.fsbno == NULLFSBLOCK) { + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); + *stat = 0; + return 0; + } + ASSERT(args.len == 1); + cur->bc_private.b.firstblock = args.fsbno; + cur->bc_private.b.allocated++; + cur->bc_private.b.ip->i_d.di_nblocks++; + xfs_trans_log_inode(args.tp, cur->bc_private.b.ip, XFS_ILOG_CORE); + XFS_TRANS_MOD_DQUOT_BYINO(args.mp, args.tp, cur->bc_private.b.ip, + XFS_TRANS_DQ_BCOUNT, 1L); + + new->l = cpu_to_be64(args.fsbno); + + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); + *stat = 1; + return 0; + + error0: + XFS_BTREE_TRACE_CURSOR(cur, XBT_ERROR); + return error; +} + +STATIC int +xfs_bmbt_free_block( + struct xfs_btree_cur *cur, + struct xfs_buf *bp) +{ + struct xfs_mount *mp = cur->bc_mp; + struct xfs_inode *ip = cur->bc_private.b.ip; + struct xfs_trans *tp = cur->bc_tp; + xfs_fsblock_t fsbno = XFS_DADDR_TO_FSB(mp, XFS_BUF_ADDR(bp)); + + xfs_bmap_add_free(fsbno, 1, cur->bc_private.b.flist, mp); + ip->i_d.di_nblocks--; + + xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); + XFS_TRANS_MOD_DQUOT_BYINO(mp, tp, ip, XFS_TRANS_DQ_BCOUNT, -1L); + xfs_trans_binval(tp, bp); + return 0; +} + +STATIC int +xfs_bmbt_get_minrecs( + struct xfs_btree_cur *cur, + int level) +{ + if (level == cur->bc_nlevels - 1) { + struct xfs_ifork *ifp; + + ifp = XFS_IFORK_PTR(cur->bc_private.b.ip, + cur->bc_private.b.whichfork); + + return xfs_bmbt_maxrecs(cur->bc_mp, + ifp->if_broot_bytes, level == 0) / 2; + } + + return cur->bc_mp->m_bmap_dmnr[level != 0]; +} + +int +xfs_bmbt_get_maxrecs( + struct xfs_btree_cur *cur, + int level) +{ + if (level == cur->bc_nlevels - 1) { + struct xfs_ifork *ifp; + + ifp = XFS_IFORK_PTR(cur->bc_private.b.ip, + cur->bc_private.b.whichfork); + + return xfs_bmbt_maxrecs(cur->bc_mp, + ifp->if_broot_bytes, level == 0); + } + + return cur->bc_mp->m_bmap_dmxr[level != 0]; + +} + +/* + * Get the maximum records we could store in the on-disk format. + * + * For non-root nodes this is equivalent to xfs_bmbt_get_maxrecs, but + * for the root node this checks the available space in the dinode fork + * so that we can resize the in-memory buffer to match it. After a + * resize to the maximum size this function returns the same value + * as xfs_bmbt_get_maxrecs for the root node, too. + */ +STATIC int +xfs_bmbt_get_dmaxrecs( + struct xfs_btree_cur *cur, + int level) +{ + if (level != cur->bc_nlevels - 1) + return cur->bc_mp->m_bmap_dmxr[level != 0]; + return xfs_bmdr_maxrecs(cur->bc_mp, cur->bc_private.b.forksize, + level == 0); +} + +STATIC void +xfs_bmbt_init_key_from_rec( + union xfs_btree_key *key, + union xfs_btree_rec *rec) +{ + key->bmbt.br_startoff = + cpu_to_be64(xfs_bmbt_disk_get_startoff(&rec->bmbt)); +} + +STATIC void +xfs_bmbt_init_rec_from_key( + union xfs_btree_key *key, + union xfs_btree_rec *rec) +{ + ASSERT(key->bmbt.br_startoff != 0); + + xfs_bmbt_disk_set_allf(&rec->bmbt, be64_to_cpu(key->bmbt.br_startoff), + 0, 0, XFS_EXT_NORM); +} + +STATIC void +xfs_bmbt_init_rec_from_cur( + struct xfs_btree_cur *cur, + union xfs_btree_rec *rec) +{ + xfs_bmbt_disk_set_all(&rec->bmbt, &cur->bc_rec.b); +} + +STATIC void +xfs_bmbt_init_ptr_from_cur( + struct xfs_btree_cur *cur, + union xfs_btree_ptr *ptr) +{ + ptr->l = 0; +} + +STATIC __int64_t +xfs_bmbt_key_diff( + struct xfs_btree_cur *cur, + union xfs_btree_key *key) +{ + return (__int64_t)be64_to_cpu(key->bmbt.br_startoff) - + cur->bc_rec.b.br_startoff; +} + +#ifdef DEBUG +STATIC int +xfs_bmbt_keys_inorder( + struct xfs_btree_cur *cur, + union xfs_btree_key *k1, + union xfs_btree_key *k2) +{ + return be64_to_cpu(k1->bmbt.br_startoff) < + be64_to_cpu(k2->bmbt.br_startoff); +} + +STATIC int +xfs_bmbt_recs_inorder( + struct xfs_btree_cur *cur, + union xfs_btree_rec *r1, + union xfs_btree_rec *r2) +{ + return xfs_bmbt_disk_get_startoff(&r1->bmbt) + + xfs_bmbt_disk_get_blockcount(&r1->bmbt) <= + xfs_bmbt_disk_get_startoff(&r2->bmbt); +} +#endif /* DEBUG */ + +#ifdef XFS_BTREE_TRACE +ktrace_t *xfs_bmbt_trace_buf; + +STATIC void +xfs_bmbt_trace_enter( + struct xfs_btree_cur *cur, + const char *func, + char *s, + int type, + int line, + __psunsigned_t a0, + __psunsigned_t a1, + __psunsigned_t a2, + __psunsigned_t a3, + __psunsigned_t a4, + __psunsigned_t a5, + __psunsigned_t a6, + __psunsigned_t a7, + __psunsigned_t a8, + __psunsigned_t a9, + __psunsigned_t a10) +{ + struct xfs_inode *ip = cur->bc_private.b.ip; + int whichfork = cur->bc_private.b.whichfork; + + ktrace_enter(xfs_bmbt_trace_buf, + (void *)((__psint_t)type | (whichfork << 8) | (line << 16)), + (void *)func, (void *)s, (void *)ip, (void *)cur, + (void *)a0, (void *)a1, (void *)a2, (void *)a3, + (void *)a4, (void *)a5, (void *)a6, (void *)a7, + (void *)a8, (void *)a9, (void *)a10); + ktrace_enter(ip->i_btrace, + (void *)((__psint_t)type | (whichfork << 8) | (line << 16)), + (void *)func, (void *)s, (void *)ip, (void *)cur, + (void *)a0, (void *)a1, (void *)a2, (void *)a3, + (void *)a4, (void *)a5, (void *)a6, (void *)a7, + (void *)a8, (void *)a9, (void *)a10); +} + +STATIC void +xfs_bmbt_trace_cursor( + struct xfs_btree_cur *cur, + __uint32_t *s0, + __uint64_t *l0, + __uint64_t *l1) +{ + struct xfs_bmbt_rec_host r; + + xfs_bmbt_set_all(&r, &cur->bc_rec.b); + + *s0 = (cur->bc_nlevels << 24) | + (cur->bc_private.b.flags << 16) | + cur->bc_private.b.allocated; + *l0 = r.l0; + *l1 = r.l1; +} + +STATIC void +xfs_bmbt_trace_key( + struct xfs_btree_cur *cur, + union xfs_btree_key *key, + __uint64_t *l0, + __uint64_t *l1) +{ + *l0 = be64_to_cpu(key->bmbt.br_startoff); + *l1 = 0; +} + +STATIC void +xfs_bmbt_trace_record( + struct xfs_btree_cur *cur, + union xfs_btree_rec *rec, + __uint64_t *l0, + __uint64_t *l1, + __uint64_t *l2) +{ + struct xfs_bmbt_irec irec; + + xfs_bmbt_disk_get_all(&rec->bmbt, &irec); + *l0 = irec.br_startoff; + *l1 = irec.br_startblock; + *l2 = irec.br_blockcount; +} +#endif /* XFS_BTREE_TRACE */ + +static const struct xfs_btree_ops xfs_bmbt_ops = { + .rec_len = sizeof(xfs_bmbt_rec_t), + .key_len = sizeof(xfs_bmbt_key_t), + + .dup_cursor = xfs_bmbt_dup_cursor, + .update_cursor = xfs_bmbt_update_cursor, + .alloc_block = xfs_bmbt_alloc_block, + .free_block = xfs_bmbt_free_block, + .get_maxrecs = xfs_bmbt_get_maxrecs, + .get_minrecs = xfs_bmbt_get_minrecs, + .get_dmaxrecs = xfs_bmbt_get_dmaxrecs, + .init_key_from_rec = xfs_bmbt_init_key_from_rec, + .init_rec_from_key = xfs_bmbt_init_rec_from_key, + .init_rec_from_cur = xfs_bmbt_init_rec_from_cur, + .init_ptr_from_cur = xfs_bmbt_init_ptr_from_cur, + .key_diff = xfs_bmbt_key_diff, + +#ifdef DEBUG + .keys_inorder = xfs_bmbt_keys_inorder, + .recs_inorder = xfs_bmbt_recs_inorder, +#endif + +#ifdef XFS_BTREE_TRACE + .trace_enter = xfs_bmbt_trace_enter, + .trace_cursor = xfs_bmbt_trace_cursor, + .trace_key = xfs_bmbt_trace_key, + .trace_record = xfs_bmbt_trace_record, +#endif +}; + +/* + * Allocate a new bmap btree cursor. + */ +struct xfs_btree_cur * /* new bmap btree cursor */ +xfs_bmbt_init_cursor( + struct xfs_mount *mp, /* file system mount point */ + struct xfs_trans *tp, /* transaction pointer */ + struct xfs_inode *ip, /* inode owning the btree */ + int whichfork) /* data or attr fork */ +{ + struct xfs_ifork *ifp = XFS_IFORK_PTR(ip, whichfork); + struct xfs_btree_cur *cur; + + cur = kmem_zone_zalloc(xfs_btree_cur_zone, KM_SLEEP); + + cur->bc_tp = tp; + cur->bc_mp = mp; + cur->bc_nlevels = be16_to_cpu(ifp->if_broot->bb_level) + 1; + cur->bc_btnum = XFS_BTNUM_BMAP; + cur->bc_blocklog = mp->m_sb.sb_blocklog; + + cur->bc_ops = &xfs_bmbt_ops; + cur->bc_flags = XFS_BTREE_LONG_PTRS | XFS_BTREE_ROOT_IN_INODE; + + cur->bc_private.b.forksize = XFS_IFORK_SIZE(ip, whichfork); + cur->bc_private.b.ip = ip; + cur->bc_private.b.firstblock = NULLFSBLOCK; + cur->bc_private.b.flist = NULL; + cur->bc_private.b.allocated = 0; + cur->bc_private.b.flags = 0; + cur->bc_private.b.whichfork = whichfork; + + return cur; +} + +/* + * Calculate number of records in a bmap btree block. + */ +int +xfs_bmbt_maxrecs( + struct xfs_mount *mp, + int blocklen, + int leaf) +{ + blocklen -= XFS_BMBT_BLOCK_LEN(mp); + + if (leaf) + return blocklen / sizeof(xfs_bmbt_rec_t); + return blocklen / (sizeof(xfs_bmbt_key_t) + sizeof(xfs_bmbt_ptr_t)); +} + +/* + * Calculate number of records in a bmap btree inode root. + */ +int +xfs_bmdr_maxrecs( + struct xfs_mount *mp, + int blocklen, + int leaf) +{ + blocklen -= sizeof(xfs_bmdr_block_t); + + if (leaf) + return blocklen / sizeof(xfs_bmdr_rec_t); + return blocklen / (sizeof(xfs_bmdr_key_t) + sizeof(xfs_bmdr_ptr_t)); +} diff --git a/fs/xfs/xfs_bmap_btree.h b/fs/xfs/xfs_bmap_btree.h index cd0d4b4bb81..a4555abb662 100644 --- a/fs/xfs/xfs_bmap_btree.h +++ b/fs/xfs/xfs_bmap_btree.h @@ -21,9 +21,10 @@ #define XFS_BMAP_MAGIC 0x424d4150 /* 'BMAP' */ struct xfs_btree_cur; -struct xfs_btree_lblock; +struct xfs_btree_block; struct xfs_mount; struct xfs_inode; +struct xfs_trans; /* * Bmap root header, on-disk form only. @@ -145,71 +146,60 @@ typedef struct xfs_bmbt_key { /* btree pointer type */ typedef __be64 xfs_bmbt_ptr_t, xfs_bmdr_ptr_t; -/* btree block header type */ -typedef struct xfs_btree_lblock xfs_bmbt_block_t; - -#define XFS_BUF_TO_BMBT_BLOCK(bp) ((xfs_bmbt_block_t *)XFS_BUF_PTR(bp)) - -#define XFS_BMAP_RBLOCK_DSIZE(lev,cur) ((cur)->bc_private.b.forksize) -#define XFS_BMAP_RBLOCK_ISIZE(lev,cur) \ - ((int)XFS_IFORK_PTR((cur)->bc_private.b.ip, \ - (cur)->bc_private.b.whichfork)->if_broot_bytes) - -#define XFS_BMAP_BLOCK_DMAXRECS(lev,cur) \ - (((lev) == (cur)->bc_nlevels - 1 ? \ - XFS_BTREE_BLOCK_MAXRECS(XFS_BMAP_RBLOCK_DSIZE(lev,cur), \ - xfs_bmdr, (lev) == 0) : \ - ((cur)->bc_mp->m_bmap_dmxr[(lev) != 0]))) -#define XFS_BMAP_BLOCK_IMAXRECS(lev,cur) \ - (((lev) == (cur)->bc_nlevels - 1 ? \ - XFS_BTREE_BLOCK_MAXRECS(XFS_BMAP_RBLOCK_ISIZE(lev,cur),\ - xfs_bmbt, (lev) == 0) : \ - ((cur)->bc_mp->m_bmap_dmxr[(lev) != 0]))) - -#define XFS_BMAP_BLOCK_DMINRECS(lev,cur) \ - (((lev) == (cur)->bc_nlevels - 1 ? \ - XFS_BTREE_BLOCK_MINRECS(XFS_BMAP_RBLOCK_DSIZE(lev,cur),\ - xfs_bmdr, (lev) == 0) : \ - ((cur)->bc_mp->m_bmap_dmnr[(lev) != 0]))) -#define XFS_BMAP_BLOCK_IMINRECS(lev,cur) \ - (((lev) == (cur)->bc_nlevels - 1 ? \ - XFS_BTREE_BLOCK_MINRECS(XFS_BMAP_RBLOCK_ISIZE(lev,cur),\ - xfs_bmbt, (lev) == 0) : \ - ((cur)->bc_mp->m_bmap_dmnr[(lev) != 0]))) - -#define XFS_BMAP_REC_DADDR(bb,i,cur) (XFS_BTREE_REC_ADDR(xfs_bmbt, bb, i)) - -#define XFS_BMAP_REC_IADDR(bb,i,cur) (XFS_BTREE_REC_ADDR(xfs_bmbt, bb, i)) - -#define XFS_BMAP_KEY_DADDR(bb,i,cur) \ - (XFS_BTREE_KEY_ADDR(xfs_bmbt, bb, i)) - -#define XFS_BMAP_KEY_IADDR(bb,i,cur) \ - (XFS_BTREE_KEY_ADDR(xfs_bmbt, bb, i)) - -#define XFS_BMAP_PTR_DADDR(bb,i,cur) \ - (XFS_BTREE_PTR_ADDR(xfs_bmbt, bb, i, XFS_BMAP_BLOCK_DMAXRECS( \ - be16_to_cpu((bb)->bb_level), cur))) -#define XFS_BMAP_PTR_IADDR(bb,i,cur) \ - (XFS_BTREE_PTR_ADDR(xfs_bmbt, bb, i, XFS_BMAP_BLOCK_IMAXRECS( \ - be16_to_cpu((bb)->bb_level), cur))) +/* + * Btree block header size depends on a superblock flag. + * + * (not quite yet, but soon) + */ +#define XFS_BMBT_BLOCK_LEN(mp) XFS_BTREE_LBLOCK_LEN + +#define XFS_BMBT_REC_ADDR(mp, block, index) \ + ((xfs_bmbt_rec_t *) \ + ((char *)(block) + \ + XFS_BMBT_BLOCK_LEN(mp) + \ + ((index) - 1) * sizeof(xfs_bmbt_rec_t))) + +#define XFS_BMBT_KEY_ADDR(mp, block, index) \ + ((xfs_bmbt_key_t *) \ + ((char *)(block) + \ + XFS_BMBT_BLOCK_LEN(mp) + \ + ((index) - 1) * sizeof(xfs_bmbt_key_t))) + +#define XFS_BMBT_PTR_ADDR(mp, block, index, maxrecs) \ + ((xfs_bmbt_ptr_t *) \ + ((char *)(block) + \ + XFS_BMBT_BLOCK_LEN(mp) + \ + (maxrecs) * sizeof(xfs_bmbt_key_t) + \ + ((index) - 1) * sizeof(xfs_bmbt_ptr_t))) + +#define XFS_BMDR_REC_ADDR(block, index) \ + ((xfs_bmdr_rec_t *) \ + ((char *)(block) + \ + sizeof(struct xfs_bmdr_block) + \ + ((index) - 1) * sizeof(xfs_bmdr_rec_t))) + +#define XFS_BMDR_KEY_ADDR(block, index) \ + ((xfs_bmdr_key_t *) \ + ((char *)(block) + \ + sizeof(struct xfs_bmdr_block) + \ + ((index) - 1) * sizeof(xfs_bmdr_key_t))) + +#define XFS_BMDR_PTR_ADDR(block, index, maxrecs) \ + ((xfs_bmdr_ptr_t *) \ + ((char *)(block) + \ + sizeof(struct xfs_bmdr_block) + \ + (maxrecs) * sizeof(xfs_bmdr_key_t) + \ + ((index) - 1) * sizeof(xfs_bmdr_ptr_t))) /* * These are to be used when we know the size of the block and * we don't have a cursor. */ -#define XFS_BMAP_BROOT_REC_ADDR(bb,i,sz) \ - (XFS_BTREE_REC_ADDR(xfs_bmbt,bb,i)) -#define XFS_BMAP_BROOT_KEY_ADDR(bb,i,sz) \ - (XFS_BTREE_KEY_ADDR(xfs_bmbt,bb,i)) -#define XFS_BMAP_BROOT_PTR_ADDR(bb,i,sz) \ - (XFS_BTREE_PTR_ADDR(xfs_bmbt,bb,i,XFS_BMAP_BROOT_MAXRECS(sz))) - -#define XFS_BMAP_BROOT_NUMRECS(bb) be16_to_cpu((bb)->bb_numrecs) -#define XFS_BMAP_BROOT_MAXRECS(sz) XFS_BTREE_BLOCK_MAXRECS(sz,xfs_bmbt,0) +#define XFS_BMAP_BROOT_PTR_ADDR(mp, bb, i, sz) \ + XFS_BMBT_PTR_ADDR(mp, bb, i, xfs_bmbt_maxrecs(mp, sz, 0)) #define XFS_BMAP_BROOT_SPACE_CALC(nrecs) \ - (int)(sizeof(xfs_bmbt_block_t) + \ + (int)(XFS_BTREE_LBLOCK_LEN + \ ((nrecs) * (sizeof(xfs_bmbt_key_t) + sizeof(xfs_bmbt_ptr_t)))) #define XFS_BMAP_BROOT_SPACE(bb) \ @@ -223,42 +213,12 @@ typedef struct xfs_btree_lblock xfs_bmbt_block_t; */ #define XFS_BM_MAXLEVELS(mp,w) ((mp)->m_bm_maxlevels[(w)]) -#define XFS_BMAP_SANITY_CHECK(mp,bb,level) \ - (be32_to_cpu((bb)->bb_magic) == XFS_BMAP_MAGIC && \ - be16_to_cpu((bb)->bb_level) == level && \ - be16_to_cpu((bb)->bb_numrecs) > 0 && \ - be16_to_cpu((bb)->bb_numrecs) <= (mp)->m_bmap_dmxr[(level) != 0]) - - -#ifdef __KERNEL__ - -#if defined(XFS_BMBT_TRACE) -/* - * Trace buffer entry types. - */ -#define XFS_BMBT_KTRACE_ARGBI 1 -#define XFS_BMBT_KTRACE_ARGBII 2 -#define XFS_BMBT_KTRACE_ARGFFFI 3 -#define XFS_BMBT_KTRACE_ARGI 4 -#define XFS_BMBT_KTRACE_ARGIFK 5 -#define XFS_BMBT_KTRACE_ARGIFR 6 -#define XFS_BMBT_KTRACE_ARGIK 7 -#define XFS_BMBT_KTRACE_CUR 8 - -#define XFS_BMBT_TRACE_SIZE 4096 /* size of global trace buffer */ -#define XFS_BMBT_KTRACE_SIZE 32 /* size of per-inode trace buffer */ -extern ktrace_t *xfs_bmbt_trace_buf; -#endif - /* * Prototypes for xfs_bmap.c to call. */ -extern void xfs_bmdr_to_bmbt(xfs_bmdr_block_t *, int, xfs_bmbt_block_t *, int); -extern int xfs_bmbt_decrement(struct xfs_btree_cur *, int, int *); -extern int xfs_bmbt_delete(struct xfs_btree_cur *, int *); +extern void xfs_bmdr_to_bmbt(struct xfs_mount *, xfs_bmdr_block_t *, int, + struct xfs_btree_block *, int); extern void xfs_bmbt_get_all(xfs_bmbt_rec_host_t *r, xfs_bmbt_irec_t *s); -extern xfs_bmbt_block_t *xfs_bmbt_get_block(struct xfs_btree_cur *cur, - int, struct xfs_buf **bpp); extern xfs_filblks_t xfs_bmbt_get_blockcount(xfs_bmbt_rec_host_t *r); extern xfs_fsblock_t xfs_bmbt_get_startblock(xfs_bmbt_rec_host_t *r); extern xfs_fileoff_t xfs_bmbt_get_startoff(xfs_bmbt_rec_host_t *r); @@ -268,22 +228,6 @@ extern void xfs_bmbt_disk_get_all(xfs_bmbt_rec_t *r, xfs_bmbt_irec_t *s); extern xfs_filblks_t xfs_bmbt_disk_get_blockcount(xfs_bmbt_rec_t *r); extern xfs_fileoff_t xfs_bmbt_disk_get_startoff(xfs_bmbt_rec_t *r); -extern int xfs_bmbt_increment(struct xfs_btree_cur *, int, int *); -extern int xfs_bmbt_insert(struct xfs_btree_cur *, int *); -extern void xfs_bmbt_log_block(struct xfs_btree_cur *, struct xfs_buf *, int); -extern void xfs_bmbt_log_recs(struct xfs_btree_cur *, struct xfs_buf *, int, - int); -extern int xfs_bmbt_lookup_eq(struct xfs_btree_cur *, xfs_fileoff_t, - xfs_fsblock_t, xfs_filblks_t, int *); -extern int xfs_bmbt_lookup_ge(struct xfs_btree_cur *, xfs_fileoff_t, - xfs_fsblock_t, xfs_filblks_t, int *); - -/* - * Give the bmap btree a new root block. Copy the old broot contents - * down into a real block and make the broot point to it. - */ -extern int xfs_bmbt_newroot(struct xfs_btree_cur *cur, int *lflags, int *stat); - extern void xfs_bmbt_set_all(xfs_bmbt_rec_host_t *r, xfs_bmbt_irec_t *s); extern void xfs_bmbt_set_allf(xfs_bmbt_rec_host_t *r, xfs_fileoff_t o, xfs_fsblock_t b, xfs_filblks_t c, xfs_exntst_t v); @@ -296,10 +240,15 @@ extern void xfs_bmbt_disk_set_all(xfs_bmbt_rec_t *r, xfs_bmbt_irec_t *s); extern void xfs_bmbt_disk_set_allf(xfs_bmbt_rec_t *r, xfs_fileoff_t o, xfs_fsblock_t b, xfs_filblks_t c, xfs_exntst_t v); -extern void xfs_bmbt_to_bmdr(xfs_bmbt_block_t *, int, xfs_bmdr_block_t *, int); -extern int xfs_bmbt_update(struct xfs_btree_cur *, xfs_fileoff_t, - xfs_fsblock_t, xfs_filblks_t, xfs_exntst_t); +extern void xfs_bmbt_to_bmdr(struct xfs_mount *, struct xfs_btree_block *, int, + xfs_bmdr_block_t *, int); + +extern int xfs_bmbt_get_maxrecs(struct xfs_btree_cur *, int level); +extern int xfs_bmdr_maxrecs(struct xfs_mount *, int blocklen, int leaf); +extern int xfs_bmbt_maxrecs(struct xfs_mount *, int blocklen, int leaf); + +extern struct xfs_btree_cur *xfs_bmbt_init_cursor(struct xfs_mount *, + struct xfs_trans *, struct xfs_inode *, int); -#endif /* __KERNEL__ */ #endif /* __XFS_BMAP_BTREE_H__ */ diff --git a/fs/xfs/xfs_btree.c b/fs/xfs/xfs_btree.c index cc593a84c34..7ed59267420 100644 --- a/fs/xfs/xfs_btree.c +++ b/fs/xfs/xfs_btree.c @@ -34,7 +34,9 @@ #include "xfs_attr_sf.h" #include "xfs_dinode.h" #include "xfs_inode.h" +#include "xfs_inode_item.h" #include "xfs_btree.h" +#include "xfs_btree_trace.h" #include "xfs_ialloc.h" #include "xfs_error.h" @@ -50,135 +52,33 @@ const __uint32_t xfs_magics[XFS_BTNUM_MAX] = { XFS_ABTB_MAGIC, XFS_ABTC_MAGIC, XFS_BMAP_MAGIC, XFS_IBT_MAGIC }; -/* - * Checking routine: return maxrecs for the block. - */ -STATIC int /* number of records fitting in block */ -xfs_btree_maxrecs( - xfs_btree_cur_t *cur, /* btree cursor */ - xfs_btree_block_t *block) /* generic btree block pointer */ -{ - switch (cur->bc_btnum) { - case XFS_BTNUM_BNO: - case XFS_BTNUM_CNT: - return (int)XFS_ALLOC_BLOCK_MAXRECS( - be16_to_cpu(block->bb_h.bb_level), cur); - case XFS_BTNUM_BMAP: - return (int)XFS_BMAP_BLOCK_IMAXRECS( - be16_to_cpu(block->bb_h.bb_level), cur); - case XFS_BTNUM_INO: - return (int)XFS_INOBT_BLOCK_MAXRECS( - be16_to_cpu(block->bb_h.bb_level), cur); - default: - ASSERT(0); - return 0; - } -} - -/* - * External routines. - */ - -#ifdef DEBUG -/* - * Debug routine: check that block header is ok. - */ -void -xfs_btree_check_block( - xfs_btree_cur_t *cur, /* btree cursor */ - xfs_btree_block_t *block, /* generic btree block pointer */ - int level, /* level of the btree block */ - xfs_buf_t *bp) /* buffer containing block, if any */ -{ - if (XFS_BTREE_LONG_PTRS(cur->bc_btnum)) - xfs_btree_check_lblock(cur, (xfs_btree_lblock_t *)block, level, - bp); - else - xfs_btree_check_sblock(cur, (xfs_btree_sblock_t *)block, level, - bp); -} - -/* - * Debug routine: check that keys are in the right order. - */ -void -xfs_btree_check_key( - xfs_btnum_t btnum, /* btree identifier */ - void *ak1, /* pointer to left (lower) key */ - void *ak2) /* pointer to right (higher) key */ -{ - switch (btnum) { - case XFS_BTNUM_BNO: { - xfs_alloc_key_t *k1; - xfs_alloc_key_t *k2; - - k1 = ak1; - k2 = ak2; - ASSERT(be32_to_cpu(k1->ar_startblock) < be32_to_cpu(k2->ar_startblock)); - break; - } - case XFS_BTNUM_CNT: { - xfs_alloc_key_t *k1; - xfs_alloc_key_t *k2; - - k1 = ak1; - k2 = ak2; - ASSERT(be32_to_cpu(k1->ar_blockcount) < be32_to_cpu(k2->ar_blockcount) || - (k1->ar_blockcount == k2->ar_blockcount && - be32_to_cpu(k1->ar_startblock) < be32_to_cpu(k2->ar_startblock))); - break; - } - case XFS_BTNUM_BMAP: { - xfs_bmbt_key_t *k1; - xfs_bmbt_key_t *k2; - - k1 = ak1; - k2 = ak2; - ASSERT(be64_to_cpu(k1->br_startoff) < be64_to_cpu(k2->br_startoff)); - break; - } - case XFS_BTNUM_INO: { - xfs_inobt_key_t *k1; - xfs_inobt_key_t *k2; - - k1 = ak1; - k2 = ak2; - ASSERT(be32_to_cpu(k1->ir_startino) < be32_to_cpu(k2->ir_startino)); - break; - } - default: - ASSERT(0); - } -} -#endif /* DEBUG */ -/* - * Checking routine: check that long form block header is ok. - */ -/* ARGSUSED */ -int /* error (0 or EFSCORRUPTED) */ +STATIC int /* error (0 or EFSCORRUPTED) */ xfs_btree_check_lblock( - xfs_btree_cur_t *cur, /* btree cursor */ - xfs_btree_lblock_t *block, /* btree long form block pointer */ + struct xfs_btree_cur *cur, /* btree cursor */ + struct xfs_btree_block *block, /* btree long form block pointer */ int level, /* level of the btree block */ - xfs_buf_t *bp) /* buffer for block, if any */ + struct xfs_buf *bp) /* buffer for block, if any */ { int lblock_ok; /* block passes checks */ - xfs_mount_t *mp; /* file system mount point */ + struct xfs_mount *mp; /* file system mount point */ mp = cur->bc_mp; lblock_ok = be32_to_cpu(block->bb_magic) == xfs_magics[cur->bc_btnum] && be16_to_cpu(block->bb_level) == level && be16_to_cpu(block->bb_numrecs) <= - xfs_btree_maxrecs(cur, (xfs_btree_block_t *)block) && - block->bb_leftsib && - (be64_to_cpu(block->bb_leftsib) == NULLDFSBNO || - XFS_FSB_SANITY_CHECK(mp, be64_to_cpu(block->bb_leftsib))) && - block->bb_rightsib && - (be64_to_cpu(block->bb_rightsib) == NULLDFSBNO || - XFS_FSB_SANITY_CHECK(mp, be64_to_cpu(block->bb_rightsib))); - if (unlikely(XFS_TEST_ERROR(!lblock_ok, mp, XFS_ERRTAG_BTREE_CHECK_LBLOCK, + cur->bc_ops->get_maxrecs(cur, level) && + block->bb_u.l.bb_leftsib && + (be64_to_cpu(block->bb_u.l.bb_leftsib) == NULLDFSBNO || + XFS_FSB_SANITY_CHECK(mp, + be64_to_cpu(block->bb_u.l.bb_leftsib))) && + block->bb_u.l.bb_rightsib && + (be64_to_cpu(block->bb_u.l.bb_rightsib) == NULLDFSBNO || + XFS_FSB_SANITY_CHECK(mp, + be64_to_cpu(block->bb_u.l.bb_rightsib))); + if (unlikely(XFS_TEST_ERROR(!lblock_ok, mp, + XFS_ERRTAG_BTREE_CHECK_LBLOCK, XFS_RANDOM_BTREE_CHECK_LBLOCK))) { if (bp) xfs_buftrace("LBTREE ERROR", bp); @@ -189,98 +89,15 @@ xfs_btree_check_lblock( return 0; } -/* - * Checking routine: check that (long) pointer is ok. - */ -int /* error (0 or EFSCORRUPTED) */ -xfs_btree_check_lptr( - xfs_btree_cur_t *cur, /* btree cursor */ - xfs_dfsbno_t ptr, /* btree block disk address */ - int level) /* btree block level */ -{ - xfs_mount_t *mp; /* file system mount point */ - - mp = cur->bc_mp; - XFS_WANT_CORRUPTED_RETURN( - level > 0 && - ptr != NULLDFSBNO && - XFS_FSB_SANITY_CHECK(mp, ptr)); - return 0; -} - -#ifdef DEBUG -/* - * Debug routine: check that records are in the right order. - */ -void -xfs_btree_check_rec( - xfs_btnum_t btnum, /* btree identifier */ - void *ar1, /* pointer to left (lower) record */ - void *ar2) /* pointer to right (higher) record */ -{ - switch (btnum) { - case XFS_BTNUM_BNO: { - xfs_alloc_rec_t *r1; - xfs_alloc_rec_t *r2; - - r1 = ar1; - r2 = ar2; - ASSERT(be32_to_cpu(r1->ar_startblock) + - be32_to_cpu(r1->ar_blockcount) <= - be32_to_cpu(r2->ar_startblock)); - break; - } - case XFS_BTNUM_CNT: { - xfs_alloc_rec_t *r1; - xfs_alloc_rec_t *r2; - - r1 = ar1; - r2 = ar2; - ASSERT(be32_to_cpu(r1->ar_blockcount) < be32_to_cpu(r2->ar_blockcount) || - (r1->ar_blockcount == r2->ar_blockcount && - be32_to_cpu(r1->ar_startblock) < be32_to_cpu(r2->ar_startblock))); - break; - } - case XFS_BTNUM_BMAP: { - xfs_bmbt_rec_t *r1; - xfs_bmbt_rec_t *r2; - - r1 = ar1; - r2 = ar2; - ASSERT(xfs_bmbt_disk_get_startoff(r1) + - xfs_bmbt_disk_get_blockcount(r1) <= - xfs_bmbt_disk_get_startoff(r2)); - break; - } - case XFS_BTNUM_INO: { - xfs_inobt_rec_t *r1; - xfs_inobt_rec_t *r2; - - r1 = ar1; - r2 = ar2; - ASSERT(be32_to_cpu(r1->ir_startino) + XFS_INODES_PER_CHUNK <= - be32_to_cpu(r2->ir_startino)); - break; - } - default: - ASSERT(0); - } -} -#endif /* DEBUG */ - -/* - * Checking routine: check that block header is ok. - */ -/* ARGSUSED */ -int /* error (0 or EFSCORRUPTED) */ +STATIC int /* error (0 or EFSCORRUPTED) */ xfs_btree_check_sblock( - xfs_btree_cur_t *cur, /* btree cursor */ - xfs_btree_sblock_t *block, /* btree short form block pointer */ + struct xfs_btree_cur *cur, /* btree cursor */ + struct xfs_btree_block *block, /* btree short form block pointer */ int level, /* level of the btree block */ - xfs_buf_t *bp) /* buffer containing block */ + struct xfs_buf *bp) /* buffer containing block */ { - xfs_buf_t *agbp; /* buffer for ag. freespace struct */ - xfs_agf_t *agf; /* ag. freespace structure */ + struct xfs_buf *agbp; /* buffer for ag. freespace struct */ + struct xfs_agf *agf; /* ag. freespace structure */ xfs_agblock_t agflen; /* native ag. freespace length */ int sblock_ok; /* block passes checks */ @@ -291,13 +108,13 @@ xfs_btree_check_sblock( be32_to_cpu(block->bb_magic) == xfs_magics[cur->bc_btnum] && be16_to_cpu(block->bb_level) == level && be16_to_cpu(block->bb_numrecs) <= - xfs_btree_maxrecs(cur, (xfs_btree_block_t *)block) && - (be32_to_cpu(block->bb_leftsib) == NULLAGBLOCK || - be32_to_cpu(block->bb_leftsib) < agflen) && - block->bb_leftsib && - (be32_to_cpu(block->bb_rightsib) == NULLAGBLOCK || - be32_to_cpu(block->bb_rightsib) < agflen) && - block->bb_rightsib; + cur->bc_ops->get_maxrecs(cur, level) && + (be32_to_cpu(block->bb_u.s.bb_leftsib) == NULLAGBLOCK || + be32_to_cpu(block->bb_u.s.bb_leftsib) < agflen) && + block->bb_u.s.bb_leftsib && + (be32_to_cpu(block->bb_u.s.bb_rightsib) == NULLAGBLOCK || + be32_to_cpu(block->bb_u.s.bb_rightsib) < agflen) && + block->bb_u.s.bb_rightsib; if (unlikely(XFS_TEST_ERROR(!sblock_ok, cur->bc_mp, XFS_ERRTAG_BTREE_CHECK_SBLOCK, XFS_RANDOM_BTREE_CHECK_SBLOCK))) { @@ -311,27 +128,78 @@ xfs_btree_check_sblock( } /* - * Checking routine: check that (short) pointer is ok. + * Debug routine: check that block header is ok. + */ +int +xfs_btree_check_block( + struct xfs_btree_cur *cur, /* btree cursor */ + struct xfs_btree_block *block, /* generic btree block pointer */ + int level, /* level of the btree block */ + struct xfs_buf *bp) /* buffer containing block, if any */ +{ + if (cur->bc_flags & XFS_BTREE_LONG_PTRS) + return xfs_btree_check_lblock(cur, block, level, bp); + else + return xfs_btree_check_sblock(cur, block, level, bp); +} + +/* + * Check that (long) pointer is ok. */ int /* error (0 or EFSCORRUPTED) */ +xfs_btree_check_lptr( + struct xfs_btree_cur *cur, /* btree cursor */ + xfs_dfsbno_t bno, /* btree block disk address */ + int level) /* btree block level */ +{ + XFS_WANT_CORRUPTED_RETURN( + level > 0 && + bno != NULLDFSBNO && + XFS_FSB_SANITY_CHECK(cur->bc_mp, bno)); + return 0; +} + +#ifdef DEBUG +/* + * Check that (short) pointer is ok. + */ +STATIC int /* error (0 or EFSCORRUPTED) */ xfs_btree_check_sptr( - xfs_btree_cur_t *cur, /* btree cursor */ - xfs_agblock_t ptr, /* btree block disk address */ - int level) /* btree block level */ + struct xfs_btree_cur *cur, /* btree cursor */ + xfs_agblock_t bno, /* btree block disk address */ + int level) /* btree block level */ { - xfs_buf_t *agbp; /* buffer for ag. freespace struct */ - xfs_agf_t *agf; /* ag. freespace structure */ + xfs_agblock_t agblocks = cur->bc_mp->m_sb.sb_agblocks; - agbp = cur->bc_private.a.agbp; - agf = XFS_BUF_TO_AGF(agbp); XFS_WANT_CORRUPTED_RETURN( level > 0 && - ptr != NULLAGBLOCK && ptr != 0 && - ptr < be32_to_cpu(agf->agf_length)); + bno != NULLAGBLOCK && + bno != 0 && + bno < agblocks); return 0; } /* + * Check that block ptr is ok. + */ +STATIC int /* error (0 or EFSCORRUPTED) */ +xfs_btree_check_ptr( + struct xfs_btree_cur *cur, /* btree cursor */ + union xfs_btree_ptr *ptr, /* btree block disk address */ + int index, /* offset from ptr to check */ + int level) /* btree block level */ +{ + if (cur->bc_flags & XFS_BTREE_LONG_PTRS) { + return xfs_btree_check_lptr(cur, + be64_to_cpu((&ptr->l)[index]), level); + } else { + return xfs_btree_check_sptr(cur, + be32_to_cpu((&ptr->s)[index]), level); + } +} +#endif + +/* * Delete the btree cursor. */ void @@ -387,16 +255,17 @@ xfs_btree_dup_cursor( tp = cur->bc_tp; mp = cur->bc_mp; + /* * Allocate a new cursor like the old one. */ - new = xfs_btree_init_cursor(mp, tp, cur->bc_private.a.agbp, - cur->bc_private.a.agno, cur->bc_btnum, cur->bc_private.b.ip, - cur->bc_private.b.whichfork); + new = cur->bc_ops->dup_cursor(cur); + /* * Copy the record currently in the cursor. */ new->bc_rec = cur->bc_rec; + /* * For each level current, re-get the buffer and copy the ptr value. */ @@ -416,46 +285,174 @@ xfs_btree_dup_cursor( } else new->bc_bufs[i] = NULL; } - /* - * For bmap btrees, copy the firstblock, flist, and flags values, - * since init cursor doesn't get them. - */ - if (new->bc_btnum == XFS_BTNUM_BMAP) { - new->bc_private.b.firstblock = cur->bc_private.b.firstblock; - new->bc_private.b.flist = cur->bc_private.b.flist; - new->bc_private.b.flags = cur->bc_private.b.flags; - } *ncur = new; return 0; } /* + * XFS btree block layout and addressing: + * + * There are two types of blocks in the btree: leaf and non-leaf blocks. + * + * The leaf record start with a header then followed by records containing + * the values. A non-leaf block also starts with the same header, and + * then first contains lookup keys followed by an equal number of pointers + * to the btree blocks at the previous level. + * + * +--------+-------+-------+-------+-------+-------+-------+ + * Leaf: | header | rec 1 | rec 2 | rec 3 | rec 4 | rec 5 | rec N | + * +--------+-------+-------+-------+-------+-------+-------+ + * + * +--------+-------+-------+-------+-------+-------+-------+ + * Non-Leaf: | header | key 1 | key 2 | key N | ptr 1 | ptr 2 | ptr N | + * +--------+-------+-------+-------+-------+-------+-------+ + * + * The header is called struct xfs_btree_block for reasons better left unknown + * and comes in different versions for short (32bit) and long (64bit) block + * pointers. The record and key structures are defined by the btree instances + * and opaque to the btree core. The block pointers are simple disk endian + * integers, available in a short (32bit) and long (64bit) variant. + * + * The helpers below calculate the offset of a given record, key or pointer + * into a btree block (xfs_btree_*_offset) or return a pointer to the given + * record, key or pointer (xfs_btree_*_addr). Note that all addressing + * inside the btree block is done using indices starting at one, not zero! + */ + +/* + * Return size of the btree block header for this btree instance. + */ +static inline size_t xfs_btree_block_len(struct xfs_btree_cur *cur) +{ + return (cur->bc_flags & XFS_BTREE_LONG_PTRS) ? + XFS_BTREE_LBLOCK_LEN : + XFS_BTREE_SBLOCK_LEN; +} + +/* + * Return size of btree block pointers for this btree instance. + */ +static inline size_t xfs_btree_ptr_len(struct xfs_btree_cur *cur) +{ + return (cur->bc_flags & XFS_BTREE_LONG_PTRS) ? + sizeof(__be64) : sizeof(__be32); +} + +/* + * Calculate offset of the n-th record in a btree block. + */ +STATIC size_t +xfs_btree_rec_offset( + struct xfs_btree_cur *cur, + int n) +{ + return xfs_btree_block_len(cur) + + (n - 1) * cur->bc_ops->rec_len; +} + +/* + * Calculate offset of the n-th key in a btree block. + */ +STATIC size_t +xfs_btree_key_offset( + struct xfs_btree_cur *cur, + int n) +{ + return xfs_btree_block_len(cur) + + (n - 1) * cur->bc_ops->key_len; +} + +/* + * Calculate offset of the n-th block pointer in a btree block. + */ +STATIC size_t +xfs_btree_ptr_offset( + struct xfs_btree_cur *cur, + int n, + int level) +{ + return xfs_btree_block_len(cur) + + cur->bc_ops->get_maxrecs(cur, level) * cur->bc_ops->key_len + + (n - 1) * xfs_btree_ptr_len(cur); +} + +/* + * Return a pointer to the n-th record in the btree block. + */ +STATIC union xfs_btree_rec * +xfs_btree_rec_addr( + struct xfs_btree_cur *cur, + int n, + struct xfs_btree_block *block) +{ + return (union xfs_btree_rec *) + ((char *)block + xfs_btree_rec_offset(cur, n)); +} + +/* + * Return a pointer to the n-th key in the btree block. + */ +STATIC union xfs_btree_key * +xfs_btree_key_addr( + struct xfs_btree_cur *cur, + int n, + struct xfs_btree_block *block) +{ + return (union xfs_btree_key *) + ((char *)block + xfs_btree_key_offset(cur, n)); +} + +/* + * Return a pointer to the n-th block pointer in the btree block. + */ +STATIC union xfs_btree_ptr * +xfs_btree_ptr_addr( + struct xfs_btree_cur *cur, + int n, + struct xfs_btree_block *block) +{ + int level = xfs_btree_get_level(block); + + ASSERT(block->bb_level != 0); + + return (union xfs_btree_ptr *) + ((char *)block + xfs_btree_ptr_offset(cur, n, level)); +} + +/* + * Get a the root block which is stored in the inode. + * + * For now this btree implementation assumes the btree root is always + * stored in the if_broot field of an inode fork. + */ +STATIC struct xfs_btree_block * +xfs_btree_get_iroot( + struct xfs_btree_cur *cur) +{ + struct xfs_ifork *ifp; + + ifp = XFS_IFORK_PTR(cur->bc_private.b.ip, cur->bc_private.b.whichfork); + return (struct xfs_btree_block *)ifp->if_broot; +} + +/* * Retrieve the block pointer from the cursor at the given level. - * This may be a bmap btree root or from a buffer. + * This may be an inode btree root or from a buffer. */ -STATIC xfs_btree_block_t * /* generic btree block pointer */ +STATIC struct xfs_btree_block * /* generic btree block pointer */ xfs_btree_get_block( - xfs_btree_cur_t *cur, /* btree cursor */ + struct xfs_btree_cur *cur, /* btree cursor */ int level, /* level in btree */ - xfs_buf_t **bpp) /* buffer containing the block */ -{ - xfs_btree_block_t *block; /* return value */ - xfs_buf_t *bp; /* return buffer */ - xfs_ifork_t *ifp; /* inode fork pointer */ - int whichfork; /* data or attr fork */ - - if (cur->bc_btnum == XFS_BTNUM_BMAP && level == cur->bc_nlevels - 1) { - whichfork = cur->bc_private.b.whichfork; - ifp = XFS_IFORK_PTR(cur->bc_private.b.ip, whichfork); - block = (xfs_btree_block_t *)ifp->if_broot; - bp = NULL; - } else { - bp = cur->bc_bufs[level]; - block = XFS_BUF_TO_BLOCK(bp); + struct xfs_buf **bpp) /* buffer containing the block */ +{ + if ((cur->bc_flags & XFS_BTREE_ROOT_IN_INODE) && + (level == cur->bc_nlevels - 1)) { + *bpp = NULL; + return xfs_btree_get_iroot(cur); } - ASSERT(block != NULL); - *bpp = bp; - return block; + + *bpp = cur->bc_bufs[level]; + return XFS_BUF_TO_BLOCK(*bpp); } /* @@ -505,97 +502,6 @@ xfs_btree_get_bufs( } /* - * Allocate a new btree cursor. - * The cursor is either for allocation (A) or bmap (B) or inodes (I). - */ -xfs_btree_cur_t * /* new btree cursor */ -xfs_btree_init_cursor( - xfs_mount_t *mp, /* file system mount point */ - xfs_trans_t *tp, /* transaction pointer */ - xfs_buf_t *agbp, /* (A only) buffer for agf structure */ - /* (I only) buffer for agi structure */ - xfs_agnumber_t agno, /* (AI only) allocation group number */ - xfs_btnum_t btnum, /* btree identifier */ - xfs_inode_t *ip, /* (B only) inode owning the btree */ - int whichfork) /* (B only) data or attr fork */ -{ - xfs_agf_t *agf; /* (A) allocation group freespace */ - xfs_agi_t *agi; /* (I) allocation group inodespace */ - xfs_btree_cur_t *cur; /* return value */ - xfs_ifork_t *ifp; /* (I) inode fork pointer */ - int nlevels=0; /* number of levels in the btree */ - - ASSERT(xfs_btree_cur_zone != NULL); - /* - * Allocate a new cursor. - */ - cur = kmem_zone_zalloc(xfs_btree_cur_zone, KM_SLEEP); - /* - * Deduce the number of btree levels from the arguments. - */ - switch (btnum) { - case XFS_BTNUM_BNO: - case XFS_BTNUM_CNT: - agf = XFS_BUF_TO_AGF(agbp); - nlevels = be32_to_cpu(agf->agf_levels[btnum]); - break; - case XFS_BTNUM_BMAP: - ifp = XFS_IFORK_PTR(ip, whichfork); - nlevels = be16_to_cpu(ifp->if_broot->bb_level) + 1; - break; - case XFS_BTNUM_INO: - agi = XFS_BUF_TO_AGI(agbp); - nlevels = be32_to_cpu(agi->agi_level); - break; - default: - ASSERT(0); - } - /* - * Fill in the common fields. - */ - cur->bc_tp = tp; - cur->bc_mp = mp; - cur->bc_nlevels = nlevels; - cur->bc_btnum = btnum; - cur->bc_blocklog = mp->m_sb.sb_blocklog; - /* - * Fill in private fields. - */ - switch (btnum) { - case XFS_BTNUM_BNO: - case XFS_BTNUM_CNT: - /* - * Allocation btree fields. - */ - cur->bc_private.a.agbp = agbp; - cur->bc_private.a.agno = agno; - break; - case XFS_BTNUM_INO: - /* - * Inode allocation btree fields. - */ - cur->bc_private.a.agbp = agbp; - cur->bc_private.a.agno = agno; - break; - case XFS_BTNUM_BMAP: - /* - * Bmap btree fields. - */ - cur->bc_private.b.forksize = XFS_IFORK_SIZE(ip, whichfork); - cur->bc_private.b.ip = ip; - cur->bc_private.b.firstblock = NULLFSBLOCK; - cur->bc_private.b.flist = NULL; - cur->bc_private.b.allocated = 0; - cur->bc_private.b.flags = 0; - cur->bc_private.b.whichfork = whichfork; - break; - default: - ASSERT(0); - } - return cur; -} - -/* * Check for the cursor referring to the last block at the given level. */ int /* 1=is last block, 0=not last block */ @@ -603,12 +509,12 @@ xfs_btree_islastblock( xfs_btree_cur_t *cur, /* btree cursor */ int level) /* level to check */ { - xfs_btree_block_t *block; /* generic btree block pointer */ + struct xfs_btree_block *block; /* generic btree block pointer */ xfs_buf_t *bp; /* buffer containing block */ block = xfs_btree_get_block(cur, level, &bp); xfs_btree_check_block(cur, block, level, bp); - if (XFS_BTREE_LONG_PTRS(cur->bc_btnum)) + if (cur->bc_flags & XFS_BTREE_LONG_PTRS) return be64_to_cpu(block->bb_u.l.bb_rightsib) == NULLDFSBNO; else return be32_to_cpu(block->bb_u.s.bb_rightsib) == NULLAGBLOCK; @@ -618,12 +524,12 @@ xfs_btree_islastblock( * Change the cursor to point to the first record at the given level. * Other levels are unaffected. */ -int /* success=1, failure=0 */ +STATIC int /* success=1, failure=0 */ xfs_btree_firstrec( xfs_btree_cur_t *cur, /* btree cursor */ int level) /* level to change */ { - xfs_btree_block_t *block; /* generic btree block pointer */ + struct xfs_btree_block *block; /* generic btree block pointer */ xfs_buf_t *bp; /* buffer containing block */ /* @@ -634,7 +540,7 @@ xfs_btree_firstrec( /* * It's empty, there is no such record. */ - if (!block->bb_h.bb_numrecs) + if (!block->bb_numrecs) return 0; /* * Set the ptr value to 1, that's the first record/key. @@ -647,12 +553,12 @@ xfs_btree_firstrec( * Change the cursor to point to the last record in the current block * at the given level. Other levels are unaffected. */ -int /* success=1, failure=0 */ +STATIC int /* success=1, failure=0 */ xfs_btree_lastrec( xfs_btree_cur_t *cur, /* btree cursor */ int level) /* level to change */ { - xfs_btree_block_t *block; /* generic btree block pointer */ + struct xfs_btree_block *block; /* generic btree block pointer */ xfs_buf_t *bp; /* buffer containing block */ /* @@ -663,12 +569,12 @@ xfs_btree_lastrec( /* * It's empty, there is no such record. */ - if (!block->bb_h.bb_numrecs) + if (!block->bb_numrecs) return 0; /* * Set the ptr value to numrecs, that's the last record/key. */ - cur->bc_ptrs[level] = be16_to_cpu(block->bb_h.bb_numrecs); + cur->bc_ptrs[level] = be16_to_cpu(block->bb_numrecs); return 1; } @@ -817,66 +723,84 @@ xfs_btree_reada_bufs( xfs_baread(mp->m_ddev_targp, d, mp->m_bsize * count); } +STATIC int +xfs_btree_readahead_lblock( + struct xfs_btree_cur *cur, + int lr, + struct xfs_btree_block *block) +{ + int rval = 0; + xfs_fsblock_t left = be64_to_cpu(block->bb_u.l.bb_leftsib); + xfs_fsblock_t right = be64_to_cpu(block->bb_u.l.bb_rightsib); + + if ((lr & XFS_BTCUR_LEFTRA) && left != NULLDFSBNO) { + xfs_btree_reada_bufl(cur->bc_mp, left, 1); + rval++; + } + + if ((lr & XFS_BTCUR_RIGHTRA) && right != NULLDFSBNO) { + xfs_btree_reada_bufl(cur->bc_mp, right, 1); + rval++; + } + + return rval; +} + +STATIC int +xfs_btree_readahead_sblock( + struct xfs_btree_cur *cur, + int lr, + struct xfs_btree_block *block) +{ + int rval = 0; + xfs_agblock_t left = be32_to_cpu(block->bb_u.s.bb_leftsib); + xfs_agblock_t right = be32_to_cpu(block->bb_u.s.bb_rightsib); + + + if ((lr & XFS_BTCUR_LEFTRA) && left != NULLAGBLOCK) { + xfs_btree_reada_bufs(cur->bc_mp, cur->bc_private.a.agno, + left, 1); + rval++; + } + + if ((lr & XFS_BTCUR_RIGHTRA) && right != NULLAGBLOCK) { + xfs_btree_reada_bufs(cur->bc_mp, cur->bc_private.a.agno, + right, 1); + rval++; + } + + return rval; +} + /* * Read-ahead btree blocks, at the given level. * Bits in lr are set from XFS_BTCUR_{LEFT,RIGHT}RA. */ -int -xfs_btree_readahead_core( - xfs_btree_cur_t *cur, /* btree cursor */ +STATIC int +xfs_btree_readahead( + struct xfs_btree_cur *cur, /* btree cursor */ int lev, /* level in btree */ int lr) /* left/right bits */ { - xfs_alloc_block_t *a; - xfs_bmbt_block_t *b; - xfs_inobt_block_t *i; - int rval = 0; + struct xfs_btree_block *block; + + /* + * No readahead needed if we are at the root level and the + * btree root is stored in the inode. + */ + if ((cur->bc_flags & XFS_BTREE_ROOT_IN_INODE) && + (lev == cur->bc_nlevels - 1)) + return 0; + + if ((cur->bc_ra[lev] | lr) == cur->bc_ra[lev]) + return 0; - ASSERT(cur->bc_bufs[lev] != NULL); cur->bc_ra[lev] |= lr; - switch (cur->bc_btnum) { - case XFS_BTNUM_BNO: - case XFS_BTNUM_CNT: - a = XFS_BUF_TO_ALLOC_BLOCK(cur->bc_bufs[lev]); - if ((lr & XFS_BTCUR_LEFTRA) && be32_to_cpu(a->bb_leftsib) != NULLAGBLOCK) { - xfs_btree_reada_bufs(cur->bc_mp, cur->bc_private.a.agno, - be32_to_cpu(a->bb_leftsib), 1); - rval++; - } - if ((lr & XFS_BTCUR_RIGHTRA) && be32_to_cpu(a->bb_rightsib) != NULLAGBLOCK) { - xfs_btree_reada_bufs(cur->bc_mp, cur->bc_private.a.agno, - be32_to_cpu(a->bb_rightsib), 1); - rval++; - } - break; - case XFS_BTNUM_BMAP: - b = XFS_BUF_TO_BMBT_BLOCK(cur->bc_bufs[lev]); - if ((lr & XFS_BTCUR_LEFTRA) && be64_to_cpu(b->bb_leftsib) != NULLDFSBNO) { - xfs_btree_reada_bufl(cur->bc_mp, be64_to_cpu(b->bb_leftsib), 1); - rval++; - } - if ((lr & XFS_BTCUR_RIGHTRA) && be64_to_cpu(b->bb_rightsib) != NULLDFSBNO) { - xfs_btree_reada_bufl(cur->bc_mp, be64_to_cpu(b->bb_rightsib), 1); - rval++; - } - break; - case XFS_BTNUM_INO: - i = XFS_BUF_TO_INOBT_BLOCK(cur->bc_bufs[lev]); - if ((lr & XFS_BTCUR_LEFTRA) && be32_to_cpu(i->bb_leftsib) != NULLAGBLOCK) { - xfs_btree_reada_bufs(cur->bc_mp, cur->bc_private.a.agno, - be32_to_cpu(i->bb_leftsib), 1); - rval++; - } - if ((lr & XFS_BTCUR_RIGHTRA) && be32_to_cpu(i->bb_rightsib) != NULLAGBLOCK) { - xfs_btree_reada_bufs(cur->bc_mp, cur->bc_private.a.agno, - be32_to_cpu(i->bb_rightsib), 1); - rval++; - } - break; - default: - ASSERT(0); - } - return rval; + block = XFS_BUF_TO_BLOCK(cur->bc_bufs[lev]); + + if (cur->bc_flags & XFS_BTREE_LONG_PTRS) + return xfs_btree_readahead_lblock(cur, lr, block); + return xfs_btree_readahead_sblock(cur, lr, block); } /* @@ -889,7 +813,7 @@ xfs_btree_setbuf( int lev, /* level in btree */ xfs_buf_t *bp) /* new buffer to set */ { - xfs_btree_block_t *b; /* btree block */ + struct xfs_btree_block *b; /* btree block */ xfs_buf_t *obp; /* old buffer pointer */ obp = cur->bc_bufs[lev]; @@ -900,7 +824,7 @@ xfs_btree_setbuf( if (!bp) return; b = XFS_BUF_TO_BLOCK(bp); - if (XFS_BTREE_LONG_PTRS(cur->bc_btnum)) { + if (cur->bc_flags & XFS_BTREE_LONG_PTRS) { if (be64_to_cpu(b->bb_u.l.bb_leftsib) == NULLDFSBNO) cur->bc_ra[lev] |= XFS_BTCUR_LEFTRA; if (be64_to_cpu(b->bb_u.l.bb_rightsib) == NULLDFSBNO) @@ -912,3 +836,2855 @@ xfs_btree_setbuf( cur->bc_ra[lev] |= XFS_BTCUR_RIGHTRA; } } + +STATIC int +xfs_btree_ptr_is_null( + struct xfs_btree_cur *cur, + union xfs_btree_ptr *ptr) +{ + if (cur->bc_flags & XFS_BTREE_LONG_PTRS) + return be64_to_cpu(ptr->l) == NULLFSBLOCK; + else + return be32_to_cpu(ptr->s) == NULLAGBLOCK; +} + +STATIC void +xfs_btree_set_ptr_null( + struct xfs_btree_cur *cur, + union xfs_btree_ptr *ptr) +{ + if (cur->bc_flags & XFS_BTREE_LONG_PTRS) + ptr->l = cpu_to_be64(NULLFSBLOCK); + else + ptr->s = cpu_to_be32(NULLAGBLOCK); +} + +/* + * Get/set/init sibling pointers + */ +STATIC void +xfs_btree_get_sibling( + struct xfs_btree_cur *cur, + struct xfs_btree_block *block, + union xfs_btree_ptr *ptr, + int lr) +{ + ASSERT(lr == XFS_BB_LEFTSIB || lr == XFS_BB_RIGHTSIB); + + if (cur->bc_flags & XFS_BTREE_LONG_PTRS) { + if (lr == XFS_BB_RIGHTSIB) + ptr->l = block->bb_u.l.bb_rightsib; + else + ptr->l = block->bb_u.l.bb_leftsib; + } else { + if (lr == XFS_BB_RIGHTSIB) + ptr->s = block->bb_u.s.bb_rightsib; + else + ptr->s = block->bb_u.s.bb_leftsib; + } +} + +STATIC void +xfs_btree_set_sibling( + struct xfs_btree_cur *cur, + struct xfs_btree_block *block, + union xfs_btree_ptr *ptr, + int lr) +{ + ASSERT(lr == XFS_BB_LEFTSIB || lr == XFS_BB_RIGHTSIB); + + if (cur->bc_flags & XFS_BTREE_LONG_PTRS) { + if (lr == XFS_BB_RIGHTSIB) + block->bb_u.l.bb_rightsib = ptr->l; + else + block->bb_u.l.bb_leftsib = ptr->l; + } else { + if (lr == XFS_BB_RIGHTSIB) + block->bb_u.s.bb_rightsib = ptr->s; + else + block->bb_u.s.bb_leftsib = ptr->s; + } +} + +STATIC void +xfs_btree_init_block( + struct xfs_btree_cur *cur, + int level, + int numrecs, + struct xfs_btree_block *new) /* new block */ +{ + new->bb_magic = cpu_to_be32(xfs_magics[cur->bc_btnum]); + new->bb_level = cpu_to_be16(level); + new->bb_numrecs = cpu_to_be16(numrecs); + + if (cur->bc_flags & XFS_BTREE_LONG_PTRS) { + new->bb_u.l.bb_leftsib = cpu_to_be64(NULLFSBLOCK); + new->bb_u.l.bb_rightsib = cpu_to_be64(NULLFSBLOCK); + } else { + new->bb_u.s.bb_leftsib = cpu_to_be32(NULLAGBLOCK); + new->bb_u.s.bb_rightsib = cpu_to_be32(NULLAGBLOCK); + } +} + +/* + * Return true if ptr is the last record in the btree and + * we need to track updateÑ• to this record. The decision + * will be further refined in the update_lastrec method. + */ +STATIC int +xfs_btree_is_lastrec( + struct xfs_btree_cur *cur, + struct xfs_btree_block *block, + int level) +{ + union xfs_btree_ptr ptr; + + if (level > 0) + return 0; + if (!(cur->bc_flags & XFS_BTREE_LASTREC_UPDATE)) + return 0; + + xfs_btree_get_sibling(cur, block, &ptr, XFS_BB_RIGHTSIB); + if (!xfs_btree_ptr_is_null(cur, &ptr)) + return 0; + return 1; +} + +STATIC void +xfs_btree_buf_to_ptr( + struct xfs_btree_cur *cur, + struct xfs_buf *bp, + union xfs_btree_ptr *ptr) +{ + if (cur->bc_flags & XFS_BTREE_LONG_PTRS) + ptr->l = cpu_to_be64(XFS_DADDR_TO_FSB(cur->bc_mp, + XFS_BUF_ADDR(bp))); + else { + ptr->s = cpu_to_be32(XFS_DADDR_TO_AGBNO(cur->bc_mp, + XFS_BUF_ADDR(bp))); + } +} + +STATIC xfs_daddr_t +xfs_btree_ptr_to_daddr( + struct xfs_btree_cur *cur, + union xfs_btree_ptr *ptr) +{ + if (cur->bc_flags & XFS_BTREE_LONG_PTRS) { + ASSERT(be64_to_cpu(ptr->l) != NULLFSBLOCK); + + return XFS_FSB_TO_DADDR(cur->bc_mp, be64_to_cpu(ptr->l)); + } else { + ASSERT(cur->bc_private.a.agno != NULLAGNUMBER); + ASSERT(be32_to_cpu(ptr->s) != NULLAGBLOCK); + + return XFS_AGB_TO_DADDR(cur->bc_mp, cur->bc_private.a.agno, + be32_to_cpu(ptr->s)); + } +} + +STATIC void +xfs_btree_set_refs( + struct xfs_btree_cur *cur, + struct xfs_buf *bp) +{ + switch (cur->bc_btnum) { + case XFS_BTNUM_BNO: + case XFS_BTNUM_CNT: + XFS_BUF_SET_VTYPE_REF(*bpp, B_FS_MAP, XFS_ALLOC_BTREE_REF); + break; + case XFS_BTNUM_INO: + XFS_BUF_SET_VTYPE_REF(*bpp, B_FS_INOMAP, XFS_INO_BTREE_REF); + break; + case XFS_BTNUM_BMAP: + XFS_BUF_SET_VTYPE_REF(*bpp, B_FS_MAP, XFS_BMAP_BTREE_REF); + break; + default: + ASSERT(0); + } +} + +STATIC int +xfs_btree_get_buf_block( + struct xfs_btree_cur *cur, + union xfs_btree_ptr *ptr, + int flags, + struct xfs_btree_block **block, + struct xfs_buf **bpp) +{ + struct xfs_mount *mp = cur->bc_mp; + xfs_daddr_t d; + + /* need to sort out how callers deal with failures first */ + ASSERT(!(flags & XFS_BUF_TRYLOCK)); + + d = xfs_btree_ptr_to_daddr(cur, ptr); + *bpp = xfs_trans_get_buf(cur->bc_tp, mp->m_ddev_targp, d, + mp->m_bsize, flags); + + ASSERT(*bpp); + ASSERT(!XFS_BUF_GETERROR(*bpp)); + + *block = XFS_BUF_TO_BLOCK(*bpp); + return 0; +} + +/* + * Read in the buffer at the given ptr and return the buffer and + * the block pointer within the buffer. + */ +STATIC int +xfs_btree_read_buf_block( + struct xfs_btree_cur *cur, + union xfs_btree_ptr *ptr, + int level, + int flags, + struct xfs_btree_block **block, + struct xfs_buf **bpp) +{ + struct xfs_mount *mp = cur->bc_mp; + xfs_daddr_t d; + int error; + + /* need to sort out how callers deal with failures first */ + ASSERT(!(flags & XFS_BUF_TRYLOCK)); + + d = xfs_btree_ptr_to_daddr(cur, ptr); + error = xfs_trans_read_buf(mp, cur->bc_tp, mp->m_ddev_targp, d, + mp->m_bsize, flags, bpp); + if (error) + return error; + + ASSERT(*bpp != NULL); + ASSERT(!XFS_BUF_GETERROR(*bpp)); + + xfs_btree_set_refs(cur, *bpp); + *block = XFS_BUF_TO_BLOCK(*bpp); + + error = xfs_btree_check_block(cur, *block, level, *bpp); + if (error) + xfs_trans_brelse(cur->bc_tp, *bpp); + return error; +} + +/* + * Copy keys from one btree block to another. + */ +STATIC void +xfs_btree_copy_keys( + struct xfs_btree_cur *cur, + union xfs_btree_key *dst_key, + union xfs_btree_key *src_key, + int numkeys) +{ + ASSERT(numkeys >= 0); + memcpy(dst_key, src_key, numkeys * cur->bc_ops->key_len); +} + +/* + * Copy records from one btree block to another. + */ +STATIC void +xfs_btree_copy_recs( + struct xfs_btree_cur *cur, + union xfs_btree_rec *dst_rec, + union xfs_btree_rec *src_rec, + int numrecs) +{ + ASSERT(numrecs >= 0); + memcpy(dst_rec, src_rec, numrecs * cur->bc_ops->rec_len); +} + +/* + * Copy block pointers from one btree block to another. + */ +STATIC void +xfs_btree_copy_ptrs( + struct xfs_btree_cur *cur, + union xfs_btree_ptr *dst_ptr, + union xfs_btree_ptr *src_ptr, + int numptrs) +{ + ASSERT(numptrs >= 0); + memcpy(dst_ptr, src_ptr, numptrs * xfs_btree_ptr_len(cur)); +} + +/* + * Shift keys one index left/right inside a single btree block. + */ +STATIC void +xfs_btree_shift_keys( + struct xfs_btree_cur *cur, + union xfs_btree_key *key, + int dir, + int numkeys) +{ + char *dst_key; + + ASSERT(numkeys >= 0); + ASSERT(dir == 1 || dir == -1); + + dst_key = (char *)key + (dir * cur->bc_ops->key_len); + memmove(dst_key, key, numkeys * cur->bc_ops->key_len); +} + +/* + * Shift records one index left/right inside a single btree block. + */ +STATIC void +xfs_btree_shift_recs( + struct xfs_btree_cur *cur, + union xfs_btree_rec *rec, + int dir, + int numrecs) +{ + char *dst_rec; + + ASSERT(numrecs >= 0); + ASSERT(dir == 1 || dir == -1); + + dst_rec = (char *)rec + (dir * cur->bc_ops->rec_len); + memmove(dst_rec, rec, numrecs * cur->bc_ops->rec_len); +} + +/* + * Shift block pointers one index left/right inside a single btree block. + */ +STATIC void +xfs_btree_shift_ptrs( + struct xfs_btree_cur *cur, + union xfs_btree_ptr *ptr, + int dir, + int numptrs) +{ + char *dst_ptr; + + ASSERT(numptrs >= 0); + ASSERT(dir == 1 || dir == -1); + + dst_ptr = (char *)ptr + (dir * xfs_btree_ptr_len(cur)); + memmove(dst_ptr, ptr, numptrs * xfs_btree_ptr_len(cur)); +} + +/* + * Log key values from the btree block. + */ +STATIC void +xfs_btree_log_keys( + struct xfs_btree_cur *cur, + struct xfs_buf *bp, + int first, + int last) +{ + XFS_BTREE_TRACE_CURSOR(cur, XBT_ENTRY); + XFS_BTREE_TRACE_ARGBII(cur, bp, first, last); + + if (bp) { + xfs_trans_log_buf(cur->bc_tp, bp, + xfs_btree_key_offset(cur, first), + xfs_btree_key_offset(cur, last + 1) - 1); + } else { + xfs_trans_log_inode(cur->bc_tp, cur->bc_private.b.ip, + xfs_ilog_fbroot(cur->bc_private.b.whichfork)); + } + + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); +} + +/* + * Log record values from the btree block. + */ +void +xfs_btree_log_recs( + struct xfs_btree_cur *cur, + struct xfs_buf *bp, + int first, + int last) +{ + XFS_BTREE_TRACE_CURSOR(cur, XBT_ENTRY); + XFS_BTREE_TRACE_ARGBII(cur, bp, first, last); + + xfs_trans_log_buf(cur->bc_tp, bp, + xfs_btree_rec_offset(cur, first), + xfs_btree_rec_offset(cur, last + 1) - 1); + + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); +} + +/* + * Log block pointer fields from a btree block (nonleaf). + */ +STATIC void +xfs_btree_log_ptrs( + struct xfs_btree_cur *cur, /* btree cursor */ + struct xfs_buf *bp, /* buffer containing btree block */ + int first, /* index of first pointer to log */ + int last) /* index of last pointer to log */ +{ + XFS_BTREE_TRACE_CURSOR(cur, XBT_ENTRY); + XFS_BTREE_TRACE_ARGBII(cur, bp, first, last); + + if (bp) { + struct xfs_btree_block *block = XFS_BUF_TO_BLOCK(bp); + int level = xfs_btree_get_level(block); + + xfs_trans_log_buf(cur->bc_tp, bp, + xfs_btree_ptr_offset(cur, first, level), + xfs_btree_ptr_offset(cur, last + 1, level) - 1); + } else { + xfs_trans_log_inode(cur->bc_tp, cur->bc_private.b.ip, + xfs_ilog_fbroot(cur->bc_private.b.whichfork)); + } + + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); +} + +/* + * Log fields from a btree block header. + */ +void +xfs_btree_log_block( + struct xfs_btree_cur *cur, /* btree cursor */ + struct xfs_buf *bp, /* buffer containing btree block */ + int fields) /* mask of fields: XFS_BB_... */ +{ + int first; /* first byte offset logged */ + int last; /* last byte offset logged */ + static const short soffsets[] = { /* table of offsets (short) */ + offsetof(struct xfs_btree_block, bb_magic), + offsetof(struct xfs_btree_block, bb_level), + offsetof(struct xfs_btree_block, bb_numrecs), + offsetof(struct xfs_btree_block, bb_u.s.bb_leftsib), + offsetof(struct xfs_btree_block, bb_u.s.bb_rightsib), + XFS_BTREE_SBLOCK_LEN + }; + static const short loffsets[] = { /* table of offsets (long) */ + offsetof(struct xfs_btree_block, bb_magic), + offsetof(struct xfs_btree_block, bb_level), + offsetof(struct xfs_btree_block, bb_numrecs), + offsetof(struct xfs_btree_block, bb_u.l.bb_leftsib), + offsetof(struct xfs_btree_block, bb_u.l.bb_rightsib), + XFS_BTREE_LBLOCK_LEN + }; + + XFS_BTREE_TRACE_CURSOR(cur, XBT_ENTRY); + XFS_BTREE_TRACE_ARGBI(cur, bp, fields); + + if (bp) { + xfs_btree_offsets(fields, + (cur->bc_flags & XFS_BTREE_LONG_PTRS) ? + loffsets : soffsets, + XFS_BB_NUM_BITS, &first, &last); + xfs_trans_log_buf(cur->bc_tp, bp, first, last); + } else { + xfs_trans_log_inode(cur->bc_tp, cur->bc_private.b.ip, + xfs_ilog_fbroot(cur->bc_private.b.whichfork)); + } + + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); +} + +/* + * Increment cursor by one record at the level. + * For nonzero levels the leaf-ward information is untouched. + */ +int /* error */ +xfs_btree_increment( + struct xfs_btree_cur *cur, + int level, + int *stat) /* success/failure */ +{ + struct xfs_btree_block *block; + union xfs_btree_ptr ptr; + struct xfs_buf *bp; + int error; /* error return value */ + int lev; + + XFS_BTREE_TRACE_CURSOR(cur, XBT_ENTRY); + XFS_BTREE_TRACE_ARGI(cur, level); + + ASSERT(level < cur->bc_nlevels); + + /* Read-ahead to the right at this level. */ + xfs_btree_readahead(cur, level, XFS_BTCUR_RIGHTRA); + + /* Get a pointer to the btree block. */ + block = xfs_btree_get_block(cur, level, &bp); + +#ifdef DEBUG + error = xfs_btree_check_block(cur, block, level, bp); + if (error) + goto error0; +#endif + + /* We're done if we remain in the block after the increment. */ + if (++cur->bc_ptrs[level] <= xfs_btree_get_numrecs(block)) + goto out1; + + /* Fail if we just went off the right edge of the tree. */ + xfs_btree_get_sibling(cur, block, &ptr, XFS_BB_RIGHTSIB); + if (xfs_btree_ptr_is_null(cur, &ptr)) + goto out0; + + XFS_BTREE_STATS_INC(cur, increment); + + /* + * March up the tree incrementing pointers. + * Stop when we don't go off the right edge of a block. + */ + for (lev = level + 1; lev < cur->bc_nlevels; lev++) { + block = xfs_btree_get_block(cur, lev, &bp); + +#ifdef DEBUG + error = xfs_btree_check_block(cur, block, lev, bp); + if (error) + goto error0; +#endif + + if (++cur->bc_ptrs[lev] <= xfs_btree_get_numrecs(block)) + break; + + /* Read-ahead the right block for the next loop. */ + xfs_btree_readahead(cur, lev, XFS_BTCUR_RIGHTRA); + } + + /* + * If we went off the root then we are either seriously + * confused or have the tree root in an inode. + */ + if (lev == cur->bc_nlevels) { + if (cur->bc_flags & XFS_BTREE_ROOT_IN_INODE) + goto out0; + ASSERT(0); + error = EFSCORRUPTED; + goto error0; + } + ASSERT(lev < cur->bc_nlevels); + + /* + * Now walk back down the tree, fixing up the cursor's buffer + * pointers and key numbers. + */ + for (block = xfs_btree_get_block(cur, lev, &bp); lev > level; ) { + union xfs_btree_ptr *ptrp; + + ptrp = xfs_btree_ptr_addr(cur, cur->bc_ptrs[lev], block); + error = xfs_btree_read_buf_block(cur, ptrp, --lev, + 0, &block, &bp); + if (error) + goto error0; + + xfs_btree_setbuf(cur, lev, bp); + cur->bc_ptrs[lev] = 1; + } +out1: + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); + *stat = 1; + return 0; + +out0: + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); + *stat = 0; + return 0; + +error0: + XFS_BTREE_TRACE_CURSOR(cur, XBT_ERROR); + return error; +} + +/* + * Decrement cursor by one record at the level. + * For nonzero levels the leaf-ward information is untouched. + */ +int /* error */ +xfs_btree_decrement( + struct xfs_btree_cur *cur, + int level, + int *stat) /* success/failure */ +{ + struct xfs_btree_block *block; + xfs_buf_t *bp; + int error; /* error return value */ + int lev; + union xfs_btree_ptr ptr; + + XFS_BTREE_TRACE_CURSOR(cur, XBT_ENTRY); + XFS_BTREE_TRACE_ARGI(cur, level); + + ASSERT(level < cur->bc_nlevels); + + /* Read-ahead to the left at this level. */ + xfs_btree_readahead(cur, level, XFS_BTCUR_LEFTRA); + + /* We're done if we remain in the block after the decrement. */ + if (--cur->bc_ptrs[level] > 0) + goto out1; + + /* Get a pointer to the btree block. */ + block = xfs_btree_get_block(cur, level, &bp); + +#ifdef DEBUG + error = xfs_btree_check_block(cur, block, level, bp); + if (error) + goto error0; +#endif + + /* Fail if we just went off the left edge of the tree. */ + xfs_btree_get_sibling(cur, block, &ptr, XFS_BB_LEFTSIB); + if (xfs_btree_ptr_is_null(cur, &ptr)) + goto out0; + + XFS_BTREE_STATS_INC(cur, decrement); + + /* + * March up the tree decrementing pointers. + * Stop when we don't go off the left edge of a block. + */ + for (lev = level + 1; lev < cur->bc_nlevels; lev++) { + if (--cur->bc_ptrs[lev] > 0) + break; + /* Read-ahead the left block for the next loop. */ + xfs_btree_readahead(cur, lev, XFS_BTCUR_LEFTRA); + } + + /* + * If we went off the root then we are seriously confused. + * or the root of the tree is in an inode. + */ + if (lev == cur->bc_nlevels) { + if (cur->bc_flags & XFS_BTREE_ROOT_IN_INODE) + goto out0; + ASSERT(0); + error = EFSCORRUPTED; + goto error0; + } + ASSERT(lev < cur->bc_nlevels); + + /* + * Now walk back down the tree, fixing up the cursor's buffer + * pointers and key numbers. + */ + for (block = xfs_btree_get_block(cur, lev, &bp); lev > level; ) { + union xfs_btree_ptr *ptrp; + + ptrp = xfs_btree_ptr_addr(cur, cur->bc_ptrs[lev], block); + error = xfs_btree_read_buf_block(cur, ptrp, --lev, + 0, &block, &bp); + if (error) + goto error0; + xfs_btree_setbuf(cur, lev, bp); + cur->bc_ptrs[lev] = xfs_btree_get_numrecs(block); + } +out1: + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); + *stat = 1; + return 0; + +out0: + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); + *stat = 0; + return 0; + +error0: + XFS_BTREE_TRACE_CURSOR(cur, XBT_ERROR); + return error; +} + +STATIC int +xfs_btree_lookup_get_block( + struct xfs_btree_cur *cur, /* btree cursor */ + int level, /* level in the btree */ + union xfs_btree_ptr *pp, /* ptr to btree block */ + struct xfs_btree_block **blkp) /* return btree block */ +{ + struct xfs_buf *bp; /* buffer pointer for btree block */ + int error = 0; + + /* special case the root block if in an inode */ + if ((cur->bc_flags & XFS_BTREE_ROOT_IN_INODE) && + (level == cur->bc_nlevels - 1)) { + *blkp = xfs_btree_get_iroot(cur); + return 0; + } + + /* + * If the old buffer at this level for the disk address we are + * looking for re-use it. + * + * Otherwise throw it away and get a new one. + */ + bp = cur->bc_bufs[level]; + if (bp && XFS_BUF_ADDR(bp) == xfs_btree_ptr_to_daddr(cur, pp)) { + *blkp = XFS_BUF_TO_BLOCK(bp); + return 0; + } + + error = xfs_btree_read_buf_block(cur, pp, level, 0, blkp, &bp); + if (error) + return error; + + xfs_btree_setbuf(cur, level, bp); + return 0; +} + +/* + * Get current search key. For level 0 we don't actually have a key + * structure so we make one up from the record. For all other levels + * we just return the right key. + */ +STATIC union xfs_btree_key * +xfs_lookup_get_search_key( + struct xfs_btree_cur *cur, + int level, + int keyno, + struct xfs_btree_block *block, + union xfs_btree_key *kp) +{ + if (level == 0) { + cur->bc_ops->init_key_from_rec(kp, + xfs_btree_rec_addr(cur, keyno, block)); + return kp; + } + + return xfs_btree_key_addr(cur, keyno, block); +} + +/* + * Lookup the record. The cursor is made to point to it, based on dir. + * Return 0 if can't find any such record, 1 for success. + */ +int /* error */ +xfs_btree_lookup( + struct xfs_btree_cur *cur, /* btree cursor */ + xfs_lookup_t dir, /* <=, ==, or >= */ + int *stat) /* success/failure */ +{ + struct xfs_btree_block *block; /* current btree block */ + __int64_t diff; /* difference for the current key */ + int error; /* error return value */ + int keyno; /* current key number */ + int level; /* level in the btree */ + union xfs_btree_ptr *pp; /* ptr to btree block */ + union xfs_btree_ptr ptr; /* ptr to btree block */ + + XFS_BTREE_TRACE_CURSOR(cur, XBT_ENTRY); + XFS_BTREE_TRACE_ARGI(cur, dir); + + XFS_BTREE_STATS_INC(cur, lookup); + + block = NULL; + keyno = 0; + + /* initialise start pointer from cursor */ + cur->bc_ops->init_ptr_from_cur(cur, &ptr); + pp = &ptr; + + /* + * Iterate over each level in the btree, starting at the root. + * For each level above the leaves, find the key we need, based + * on the lookup record, then follow the corresponding block + * pointer down to the next level. + */ + for (level = cur->bc_nlevels - 1, diff = 1; level >= 0; level--) { + /* Get the block we need to do the lookup on. */ + error = xfs_btree_lookup_get_block(cur, level, pp, &block); + if (error) + goto error0; + + if (diff == 0) { + /* + * If we already had a key match at a higher level, we + * know we need to use the first entry in this block. + */ + keyno = 1; + } else { + /* Otherwise search this block. Do a binary search. */ + + int high; /* high entry number */ + int low; /* low entry number */ + + /* Set low and high entry numbers, 1-based. */ + low = 1; + high = xfs_btree_get_numrecs(block); + if (!high) { + /* Block is empty, must be an empty leaf. */ + ASSERT(level == 0 && cur->bc_nlevels == 1); + + cur->bc_ptrs[0] = dir != XFS_LOOKUP_LE; + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); + *stat = 0; + return 0; + } + + /* Binary search the block. */ + while (low <= high) { + union xfs_btree_key key; + union xfs_btree_key *kp; + + XFS_BTREE_STATS_INC(cur, compare); + + /* keyno is average of low and high. */ + keyno = (low + high) >> 1; + + /* Get current search key */ + kp = xfs_lookup_get_search_key(cur, level, + keyno, block, &key); + + /* + * Compute difference to get next direction: + * - less than, move right + * - greater than, move left + * - equal, we're done + */ + diff = cur->bc_ops->key_diff(cur, kp); + if (diff < 0) + low = keyno + 1; + else if (diff > 0) + high = keyno - 1; + else + break; + } + } + + /* + * If there are more levels, set up for the next level + * by getting the block number and filling in the cursor. + */ + if (level > 0) { + /* + * If we moved left, need the previous key number, + * unless there isn't one. + */ + if (diff > 0 && --keyno < 1) + keyno = 1; + pp = xfs_btree_ptr_addr(cur, keyno, block); + +#ifdef DEBUG + error = xfs_btree_check_ptr(cur, pp, 0, level); + if (error) + goto error0; +#endif + cur->bc_ptrs[level] = keyno; + } + } + + /* Done with the search. See if we need to adjust the results. */ + if (dir != XFS_LOOKUP_LE && diff < 0) { + keyno++; + /* + * If ge search and we went off the end of the block, but it's + * not the last block, we're in the wrong block. + */ + xfs_btree_get_sibling(cur, block, &ptr, XFS_BB_RIGHTSIB); + if (dir == XFS_LOOKUP_GE && + keyno > xfs_btree_get_numrecs(block) && + !xfs_btree_ptr_is_null(cur, &ptr)) { + int i; + + cur->bc_ptrs[0] = keyno; + error = xfs_btree_increment(cur, 0, &i); + if (error) + goto error0; + XFS_WANT_CORRUPTED_RETURN(i == 1); + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); + *stat = 1; + return 0; + } + } else if (dir == XFS_LOOKUP_LE && diff > 0) + keyno--; + cur->bc_ptrs[0] = keyno; + + /* Return if we succeeded or not. */ + if (keyno == 0 || keyno > xfs_btree_get_numrecs(block)) + *stat = 0; + else if (dir != XFS_LOOKUP_EQ || diff == 0) + *stat = 1; + else + *stat = 0; + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); + return 0; + +error0: + XFS_BTREE_TRACE_CURSOR(cur, XBT_ERROR); + return error; +} + +/* + * Update keys at all levels from here to the root along the cursor's path. + */ +STATIC int +xfs_btree_updkey( + struct xfs_btree_cur *cur, + union xfs_btree_key *keyp, + int level) +{ + struct xfs_btree_block *block; + struct xfs_buf *bp; + union xfs_btree_key *kp; + int ptr; + + XFS_BTREE_TRACE_CURSOR(cur, XBT_ENTRY); + XFS_BTREE_TRACE_ARGIK(cur, level, keyp); + + ASSERT(!(cur->bc_flags & XFS_BTREE_ROOT_IN_INODE) || level >= 1); + + /* + * Go up the tree from this level toward the root. + * At each level, update the key value to the value input. + * Stop when we reach a level where the cursor isn't pointing + * at the first entry in the block. + */ + for (ptr = 1; ptr == 1 && level < cur->bc_nlevels; level++) { +#ifdef DEBUG + int error; +#endif + block = xfs_btree_get_block(cur, level, &bp); +#ifdef DEBUG + error = xfs_btree_check_block(cur, block, level, bp); + if (error) { + XFS_BTREE_TRACE_CURSOR(cur, XBT_ERROR); + return error; + } +#endif + ptr = cur->bc_ptrs[level]; + kp = xfs_btree_key_addr(cur, ptr, block); + xfs_btree_copy_keys(cur, kp, keyp, 1); + xfs_btree_log_keys(cur, bp, ptr, ptr); + } + + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); + return 0; +} + +/* + * Update the record referred to by cur to the value in the + * given record. This either works (return 0) or gets an + * EFSCORRUPTED error. + */ +int +xfs_btree_update( + struct xfs_btree_cur *cur, + union xfs_btree_rec *rec) +{ + struct xfs_btree_block *block; + struct xfs_buf *bp; + int error; + int ptr; + union xfs_btree_rec *rp; + + XFS_BTREE_TRACE_CURSOR(cur, XBT_ENTRY); + XFS_BTREE_TRACE_ARGR(cur, rec); + + /* Pick up the current block. */ + block = xfs_btree_get_block(cur, 0, &bp); + +#ifdef DEBUG + error = xfs_btree_check_block(cur, block, 0, bp); + if (error) + goto error0; +#endif + /* Get the address of the rec to be updated. */ + ptr = cur->bc_ptrs[0]; + rp = xfs_btree_rec_addr(cur, ptr, block); + + /* Fill in the new contents and log them. */ + xfs_btree_copy_recs(cur, rp, rec, 1); + xfs_btree_log_recs(cur, bp, ptr, ptr); + + /* + * If we are tracking the last record in the tree and + * we are at the far right edge of the tree, update it. + */ + if (xfs_btree_is_lastrec(cur, block, 0)) { + cur->bc_ops->update_lastrec(cur, block, rec, + ptr, LASTREC_UPDATE); + } + + /* Updating first rec in leaf. Pass new key value up to our parent. */ + if (ptr == 1) { + union xfs_btree_key key; + + cur->bc_ops->init_key_from_rec(&key, rec); + error = xfs_btree_updkey(cur, &key, 1); + if (error) + goto error0; + } + + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); + return 0; + +error0: + XFS_BTREE_TRACE_CURSOR(cur, XBT_ERROR); + return error; +} + +/* + * Move 1 record left from cur/level if possible. + * Update cur to reflect the new path. + */ +STATIC int /* error */ +xfs_btree_lshift( + struct xfs_btree_cur *cur, + int level, + int *stat) /* success/failure */ +{ + union xfs_btree_key key; /* btree key */ + struct xfs_buf *lbp; /* left buffer pointer */ + struct xfs_btree_block *left; /* left btree block */ + int lrecs; /* left record count */ + struct xfs_buf *rbp; /* right buffer pointer */ + struct xfs_btree_block *right; /* right btree block */ + int rrecs; /* right record count */ + union xfs_btree_ptr lptr; /* left btree pointer */ + union xfs_btree_key *rkp = NULL; /* right btree key */ + union xfs_btree_ptr *rpp = NULL; /* right address pointer */ + union xfs_btree_rec *rrp = NULL; /* right record pointer */ + int error; /* error return value */ + + XFS_BTREE_TRACE_CURSOR(cur, XBT_ENTRY); + XFS_BTREE_TRACE_ARGI(cur, level); + + if ((cur->bc_flags & XFS_BTREE_ROOT_IN_INODE) && + level == cur->bc_nlevels - 1) + goto out0; + + /* Set up variables for this block as "right". */ + right = xfs_btree_get_block(cur, level, &rbp); + +#ifdef DEBUG + error = xfs_btree_check_block(cur, right, level, rbp); + if (error) + goto error0; +#endif + + /* If we've got no left sibling then we can't shift an entry left. */ + xfs_btree_get_sibling(cur, right, &lptr, XFS_BB_LEFTSIB); + if (xfs_btree_ptr_is_null(cur, &lptr)) + goto out0; + + /* + * If the cursor entry is the one that would be moved, don't + * do it... it's too complicated. + */ + if (cur->bc_ptrs[level] <= 1) + goto out0; + + /* Set up the left neighbor as "left". */ + error = xfs_btree_read_buf_block(cur, &lptr, level, 0, &left, &lbp); + if (error) + goto error0; + + /* If it's full, it can't take another entry. */ + lrecs = xfs_btree_get_numrecs(left); + if (lrecs == cur->bc_ops->get_maxrecs(cur, level)) + goto out0; + + rrecs = xfs_btree_get_numrecs(right); + + /* + * We add one entry to the left side and remove one for the right side. + * Accout for it here, the changes will be updated on disk and logged + * later. + */ + lrecs++; + rrecs--; + + XFS_BTREE_STATS_INC(cur, lshift); + XFS_BTREE_STATS_ADD(cur, moves, 1); + + /* + * If non-leaf, copy a key and a ptr to the left block. + * Log the changes to the left block. + */ + if (level > 0) { + /* It's a non-leaf. Move keys and pointers. */ + union xfs_btree_key *lkp; /* left btree key */ + union xfs_btree_ptr *lpp; /* left address pointer */ + + lkp = xfs_btree_key_addr(cur, lrecs, left); + rkp = xfs_btree_key_addr(cur, 1, right); + + lpp = xfs_btree_ptr_addr(cur, lrecs, left); + rpp = xfs_btree_ptr_addr(cur, 1, right); +#ifdef DEBUG + error = xfs_btree_check_ptr(cur, rpp, 0, level); + if (error) + goto error0; +#endif + xfs_btree_copy_keys(cur, lkp, rkp, 1); + xfs_btree_copy_ptrs(cur, lpp, rpp, 1); + + xfs_btree_log_keys(cur, lbp, lrecs, lrecs); + xfs_btree_log_ptrs(cur, lbp, lrecs, lrecs); + + ASSERT(cur->bc_ops->keys_inorder(cur, + xfs_btree_key_addr(cur, lrecs - 1, left), lkp)); + } else { + /* It's a leaf. Move records. */ + union xfs_btree_rec *lrp; /* left record pointer */ + + lrp = xfs_btree_rec_addr(cur, lrecs, left); + rrp = xfs_btree_rec_addr(cur, 1, right); + + xfs_btree_copy_recs(cur, lrp, rrp, 1); + xfs_btree_log_recs(cur, lbp, lrecs, lrecs); + + ASSERT(cur->bc_ops->recs_inorder(cur, + xfs_btree_rec_addr(cur, lrecs - 1, left), lrp)); + } + + xfs_btree_set_numrecs(left, lrecs); + xfs_btree_log_block(cur, lbp, XFS_BB_NUMRECS); + + xfs_btree_set_numrecs(right, rrecs); + xfs_btree_log_block(cur, rbp, XFS_BB_NUMRECS); + + /* + * Slide the contents of right down one entry. + */ + XFS_BTREE_STATS_ADD(cur, moves, rrecs - 1); + if (level > 0) { + /* It's a nonleaf. operate on keys and ptrs */ +#ifdef DEBUG + int i; /* loop index */ + + for (i = 0; i < rrecs; i++) { + error = xfs_btree_check_ptr(cur, rpp, i + 1, level); + if (error) + goto error0; + } +#endif + xfs_btree_shift_keys(cur, + xfs_btree_key_addr(cur, 2, right), + -1, rrecs); + xfs_btree_shift_ptrs(cur, + xfs_btree_ptr_addr(cur, 2, right), + -1, rrecs); + + xfs_btree_log_keys(cur, rbp, 1, rrecs); + xfs_btree_log_ptrs(cur, rbp, 1, rrecs); + } else { + /* It's a leaf. operate on records */ + xfs_btree_shift_recs(cur, + xfs_btree_rec_addr(cur, 2, right), + -1, rrecs); + xfs_btree_log_recs(cur, rbp, 1, rrecs); + + /* + * If it's the first record in the block, we'll need a key + * structure to pass up to the next level (updkey). + */ + cur->bc_ops->init_key_from_rec(&key, + xfs_btree_rec_addr(cur, 1, right)); + rkp = &key; + } + + /* Update the parent key values of right. */ + error = xfs_btree_updkey(cur, rkp, level + 1); + if (error) + goto error0; + + /* Slide the cursor value left one. */ + cur->bc_ptrs[level]--; + + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); + *stat = 1; + return 0; + +out0: + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); + *stat = 0; + return 0; + +error0: + XFS_BTREE_TRACE_CURSOR(cur, XBT_ERROR); + return error; +} + +/* + * Move 1 record right from cur/level if possible. + * Update cur to reflect the new path. + */ +STATIC int /* error */ +xfs_btree_rshift( + struct xfs_btree_cur *cur, + int level, + int *stat) /* success/failure */ +{ + union xfs_btree_key key; /* btree key */ + struct xfs_buf *lbp; /* left buffer pointer */ + struct xfs_btree_block *left; /* left btree block */ + struct xfs_buf *rbp; /* right buffer pointer */ + struct xfs_btree_block *right; /* right btree block */ + struct xfs_btree_cur *tcur; /* temporary btree cursor */ + union xfs_btree_ptr rptr; /* right block pointer */ + union xfs_btree_key *rkp; /* right btree key */ + int rrecs; /* right record count */ + int lrecs; /* left record count */ + int error; /* error return value */ + int i; /* loop counter */ + + XFS_BTREE_TRACE_CURSOR(cur, XBT_ENTRY); + XFS_BTREE_TRACE_ARGI(cur, level); + + if ((cur->bc_flags & XFS_BTREE_ROOT_IN_INODE) && + (level == cur->bc_nlevels - 1)) + goto out0; + + /* Set up variables for this block as "left". */ + left = xfs_btree_get_block(cur, level, &lbp); + +#ifdef DEBUG + error = xfs_btree_check_block(cur, left, level, lbp); + if (error) + goto error0; +#endif + + /* If we've got no right sibling then we can't shift an entry right. */ + xfs_btree_get_sibling(cur, left, &rptr, XFS_BB_RIGHTSIB); + if (xfs_btree_ptr_is_null(cur, &rptr)) + goto out0; + + /* + * If the cursor entry is the one that would be moved, don't + * do it... it's too complicated. + */ + lrecs = xfs_btree_get_numrecs(left); + if (cur->bc_ptrs[level] >= lrecs) + goto out0; + + /* Set up the right neighbor as "right". */ + error = xfs_btree_read_buf_block(cur, &rptr, level, 0, &right, &rbp); + if (error) + goto error0; + + /* If it's full, it can't take another entry. */ + rrecs = xfs_btree_get_numrecs(right); + if (rrecs == cur->bc_ops->get_maxrecs(cur, level)) + goto out0; + + XFS_BTREE_STATS_INC(cur, rshift); + XFS_BTREE_STATS_ADD(cur, moves, rrecs); + + /* + * Make a hole at the start of the right neighbor block, then + * copy the last left block entry to the hole. + */ + if (level > 0) { + /* It's a nonleaf. make a hole in the keys and ptrs */ + union xfs_btree_key *lkp; + union xfs_btree_ptr *lpp; + union xfs_btree_ptr *rpp; + + lkp = xfs_btree_key_addr(cur, lrecs, left); + lpp = xfs_btree_ptr_addr(cur, lrecs, left); + rkp = xfs_btree_key_addr(cur, 1, right); + rpp = xfs_btree_ptr_addr(cur, 1, right); + +#ifdef DEBUG + for (i = rrecs - 1; i >= 0; i--) { + error = xfs_btree_check_ptr(cur, rpp, i, level); + if (error) + goto error0; + } +#endif + + xfs_btree_shift_keys(cur, rkp, 1, rrecs); + xfs_btree_shift_ptrs(cur, rpp, 1, rrecs); + +#ifdef DEBUG + error = xfs_btree_check_ptr(cur, lpp, 0, level); + if (error) + goto error0; +#endif + + /* Now put the new data in, and log it. */ + xfs_btree_copy_keys(cur, rkp, lkp, 1); + xfs_btree_copy_ptrs(cur, rpp, lpp, 1); + + xfs_btree_log_keys(cur, rbp, 1, rrecs + 1); + xfs_btree_log_ptrs(cur, rbp, 1, rrecs + 1); + + ASSERT(cur->bc_ops->keys_inorder(cur, rkp, + xfs_btree_key_addr(cur, 2, right))); + } else { + /* It's a leaf. make a hole in the records */ + union xfs_btree_rec *lrp; + union xfs_btree_rec *rrp; + + lrp = xfs_btree_rec_addr(cur, lrecs, left); + rrp = xfs_btree_rec_addr(cur, 1, right); + + xfs_btree_shift_recs(cur, rrp, 1, rrecs); + + /* Now put the new data in, and log it. */ + xfs_btree_copy_recs(cur, rrp, lrp, 1); + xfs_btree_log_recs(cur, rbp, 1, rrecs + 1); + + cur->bc_ops->init_key_from_rec(&key, rrp); + rkp = &key; + + ASSERT(cur->bc_ops->recs_inorder(cur, rrp, + xfs_btree_rec_addr(cur, 2, right))); + } + + /* + * Decrement and log left's numrecs, bump and log right's numrecs. + */ + xfs_btree_set_numrecs(left, --lrecs); + xfs_btree_log_block(cur, lbp, XFS_BB_NUMRECS); + + xfs_btree_set_numrecs(right, ++rrecs); + xfs_btree_log_block(cur, rbp, XFS_BB_NUMRECS); + + /* + * Using a temporary cursor, update the parent key values of the + * block on the right. + */ + error = xfs_btree_dup_cursor(cur, &tcur); + if (error) + goto error0; + i = xfs_btree_lastrec(tcur, level); + XFS_WANT_CORRUPTED_GOTO(i == 1, error0); + + error = xfs_btree_increment(tcur, level, &i); + if (error) + goto error1; + + error = xfs_btree_updkey(tcur, rkp, level + 1); + if (error) + goto error1; + + xfs_btree_del_cursor(tcur, XFS_BTREE_NOERROR); + + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); + *stat = 1; + return 0; + +out0: + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); + *stat = 0; + return 0; + +error0: + XFS_BTREE_TRACE_CURSOR(cur, XBT_ERROR); + return error; + +error1: + XFS_BTREE_TRACE_CURSOR(tcur, XBT_ERROR); + xfs_btree_del_cursor(tcur, XFS_BTREE_ERROR); + return error; +} + +/* + * Split cur/level block in half. + * Return new block number and the key to its first + * record (to be inserted into parent). + */ +STATIC int /* error */ +xfs_btree_split( + struct xfs_btree_cur *cur, + int level, + union xfs_btree_ptr *ptrp, + union xfs_btree_key *key, + struct xfs_btree_cur **curp, + int *stat) /* success/failure */ +{ + union xfs_btree_ptr lptr; /* left sibling block ptr */ + struct xfs_buf *lbp; /* left buffer pointer */ + struct xfs_btree_block *left; /* left btree block */ + union xfs_btree_ptr rptr; /* right sibling block ptr */ + struct xfs_buf *rbp; /* right buffer pointer */ + struct xfs_btree_block *right; /* right btree block */ + union xfs_btree_ptr rrptr; /* right-right sibling ptr */ + struct xfs_buf *rrbp; /* right-right buffer pointer */ + struct xfs_btree_block *rrblock; /* right-right btree block */ + int lrecs; + int rrecs; + int src_index; + int error; /* error return value */ +#ifdef DEBUG + int i; +#endif + + XFS_BTREE_TRACE_CURSOR(cur, XBT_ENTRY); + XFS_BTREE_TRACE_ARGIPK(cur, level, *ptrp, key); + + XFS_BTREE_STATS_INC(cur, split); + + /* Set up left block (current one). */ + left = xfs_btree_get_block(cur, level, &lbp); + +#ifdef DEBUG + error = xfs_btree_check_block(cur, left, level, lbp); + if (error) + goto error0; +#endif + + xfs_btree_buf_to_ptr(cur, lbp, &lptr); + + /* Allocate the new block. If we can't do it, we're toast. Give up. */ + error = cur->bc_ops->alloc_block(cur, &lptr, &rptr, 1, stat); + if (error) + goto error0; + if (*stat == 0) + goto out0; + XFS_BTREE_STATS_INC(cur, alloc); + + /* Set up the new block as "right". */ + error = xfs_btree_get_buf_block(cur, &rptr, 0, &right, &rbp); + if (error) + goto error0; + + /* Fill in the btree header for the new right block. */ + xfs_btree_init_block(cur, xfs_btree_get_level(left), 0, right); + + /* + * Split the entries between the old and the new block evenly. + * Make sure that if there's an odd number of entries now, that + * each new block will have the same number of entries. + */ + lrecs = xfs_btree_get_numrecs(left); + rrecs = lrecs / 2; + if ((lrecs & 1) && cur->bc_ptrs[level] <= rrecs + 1) + rrecs++; + src_index = (lrecs - rrecs + 1); + + XFS_BTREE_STATS_ADD(cur, moves, rrecs); + + /* + * Copy btree block entries from the left block over to the + * new block, the right. Update the right block and log the + * changes. + */ + if (level > 0) { + /* It's a non-leaf. Move keys and pointers. */ + union xfs_btree_key *lkp; /* left btree key */ + union xfs_btree_ptr *lpp; /* left address pointer */ + union xfs_btree_key *rkp; /* right btree key */ + union xfs_btree_ptr *rpp; /* right address pointer */ + + lkp = xfs_btree_key_addr(cur, src_index, left); + lpp = xfs_btree_ptr_addr(cur, src_index, left); + rkp = xfs_btree_key_addr(cur, 1, right); + rpp = xfs_btree_ptr_addr(cur, 1, right); + +#ifdef DEBUG + for (i = src_index; i < rrecs; i++) { + error = xfs_btree_check_ptr(cur, lpp, i, level); + if (error) + goto error0; + } +#endif + + xfs_btree_copy_keys(cur, rkp, lkp, rrecs); + xfs_btree_copy_ptrs(cur, rpp, lpp, rrecs); + + xfs_btree_log_keys(cur, rbp, 1, rrecs); + xfs_btree_log_ptrs(cur, rbp, 1, rrecs); + + /* Grab the keys to the entries moved to the right block */ + xfs_btree_copy_keys(cur, key, rkp, 1); + } else { + /* It's a leaf. Move records. */ + union xfs_btree_rec *lrp; /* left record pointer */ + union xfs_btree_rec *rrp; /* right record pointer */ + + lrp = xfs_btree_rec_addr(cur, src_index, left); + rrp = xfs_btree_rec_addr(cur, 1, right); + + xfs_btree_copy_recs(cur, rrp, lrp, rrecs); + xfs_btree_log_recs(cur, rbp, 1, rrecs); + + cur->bc_ops->init_key_from_rec(key, + xfs_btree_rec_addr(cur, 1, right)); + } + + + /* + * Find the left block number by looking in the buffer. + * Adjust numrecs, sibling pointers. + */ + xfs_btree_get_sibling(cur, left, &rrptr, XFS_BB_RIGHTSIB); + xfs_btree_set_sibling(cur, right, &rrptr, XFS_BB_RIGHTSIB); + xfs_btree_set_sibling(cur, right, &lptr, XFS_BB_LEFTSIB); + xfs_btree_set_sibling(cur, left, &rptr, XFS_BB_RIGHTSIB); + + lrecs -= rrecs; + xfs_btree_set_numrecs(left, lrecs); + xfs_btree_set_numrecs(right, xfs_btree_get_numrecs(right) + rrecs); + + xfs_btree_log_block(cur, rbp, XFS_BB_ALL_BITS); + xfs_btree_log_block(cur, lbp, XFS_BB_NUMRECS | XFS_BB_RIGHTSIB); + + /* + * If there's a block to the new block's right, make that block + * point back to right instead of to left. + */ + if (!xfs_btree_ptr_is_null(cur, &rrptr)) { + error = xfs_btree_read_buf_block(cur, &rrptr, level, + 0, &rrblock, &rrbp); + if (error) + goto error0; + xfs_btree_set_sibling(cur, rrblock, &rptr, XFS_BB_LEFTSIB); + xfs_btree_log_block(cur, rrbp, XFS_BB_LEFTSIB); + } + /* + * If the cursor is really in the right block, move it there. + * If it's just pointing past the last entry in left, then we'll + * insert there, so don't change anything in that case. + */ + if (cur->bc_ptrs[level] > lrecs + 1) { + xfs_btree_setbuf(cur, level, rbp); + cur->bc_ptrs[level] -= lrecs; + } + /* + * If there are more levels, we'll need another cursor which refers + * the right block, no matter where this cursor was. + */ + if (level + 1 < cur->bc_nlevels) { + error = xfs_btree_dup_cursor(cur, curp); + if (error) + goto error0; + (*curp)->bc_ptrs[level + 1]++; + } + *ptrp = rptr; + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); + *stat = 1; + return 0; +out0: + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); + *stat = 0; + return 0; + +error0: + XFS_BTREE_TRACE_CURSOR(cur, XBT_ERROR); + return error; +} + +/* + * Copy the old inode root contents into a real block and make the + * broot point to it. + */ +int /* error */ +xfs_btree_new_iroot( + struct xfs_btree_cur *cur, /* btree cursor */ + int *logflags, /* logging flags for inode */ + int *stat) /* return status - 0 fail */ +{ + struct xfs_buf *cbp; /* buffer for cblock */ + struct xfs_btree_block *block; /* btree block */ + struct xfs_btree_block *cblock; /* child btree block */ + union xfs_btree_key *ckp; /* child key pointer */ + union xfs_btree_ptr *cpp; /* child ptr pointer */ + union xfs_btree_key *kp; /* pointer to btree key */ + union xfs_btree_ptr *pp; /* pointer to block addr */ + union xfs_btree_ptr nptr; /* new block addr */ + int level; /* btree level */ + int error; /* error return code */ +#ifdef DEBUG + int i; /* loop counter */ +#endif + + XFS_BTREE_TRACE_CURSOR(cur, XBT_ENTRY); + XFS_BTREE_STATS_INC(cur, newroot); + + ASSERT(cur->bc_flags & XFS_BTREE_ROOT_IN_INODE); + + level = cur->bc_nlevels - 1; + + block = xfs_btree_get_iroot(cur); + pp = xfs_btree_ptr_addr(cur, 1, block); + + /* Allocate the new block. If we can't do it, we're toast. Give up. */ + error = cur->bc_ops->alloc_block(cur, pp, &nptr, 1, stat); + if (error) + goto error0; + if (*stat == 0) { + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); + return 0; + } + XFS_BTREE_STATS_INC(cur, alloc); + + /* Copy the root into a real block. */ + error = xfs_btree_get_buf_block(cur, &nptr, 0, &cblock, &cbp); + if (error) + goto error0; + + memcpy(cblock, block, xfs_btree_block_len(cur)); + + be16_add_cpu(&block->bb_level, 1); + xfs_btree_set_numrecs(block, 1); + cur->bc_nlevels++; + cur->bc_ptrs[level + 1] = 1; + + kp = xfs_btree_key_addr(cur, 1, block); + ckp = xfs_btree_key_addr(cur, 1, cblock); + xfs_btree_copy_keys(cur, ckp, kp, xfs_btree_get_numrecs(cblock)); + + cpp = xfs_btree_ptr_addr(cur, 1, cblock); +#ifdef DEBUG + for (i = 0; i < be16_to_cpu(cblock->bb_numrecs); i++) { + error = xfs_btree_check_ptr(cur, pp, i, level); + if (error) + goto error0; + } +#endif + xfs_btree_copy_ptrs(cur, cpp, pp, xfs_btree_get_numrecs(cblock)); + +#ifdef DEBUG + error = xfs_btree_check_ptr(cur, &nptr, 0, level); + if (error) + goto error0; +#endif + xfs_btree_copy_ptrs(cur, pp, &nptr, 1); + + xfs_iroot_realloc(cur->bc_private.b.ip, + 1 - xfs_btree_get_numrecs(cblock), + cur->bc_private.b.whichfork); + + xfs_btree_setbuf(cur, level, cbp); + + /* + * Do all this logging at the end so that + * the root is at the right level. + */ + xfs_btree_log_block(cur, cbp, XFS_BB_ALL_BITS); + xfs_btree_log_keys(cur, cbp, 1, be16_to_cpu(cblock->bb_numrecs)); + xfs_btree_log_ptrs(cur, cbp, 1, be16_to_cpu(cblock->bb_numrecs)); + + *logflags |= + XFS_ILOG_CORE | XFS_ILOG_FBROOT(cur->bc_private.b.whichfork); + *stat = 1; + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); + return 0; +error0: + XFS_BTREE_TRACE_CURSOR(cur, XBT_ERROR); + return error; +} + +/* + * Allocate a new root block, fill it in. + */ +STATIC int /* error */ +xfs_btree_new_root( + struct xfs_btree_cur *cur, /* btree cursor */ + int *stat) /* success/failure */ +{ + struct xfs_btree_block *block; /* one half of the old root block */ + struct xfs_buf *bp; /* buffer containing block */ + int error; /* error return value */ + struct xfs_buf *lbp; /* left buffer pointer */ + struct xfs_btree_block *left; /* left btree block */ + struct xfs_buf *nbp; /* new (root) buffer */ + struct xfs_btree_block *new; /* new (root) btree block */ + int nptr; /* new value for key index, 1 or 2 */ + struct xfs_buf *rbp; /* right buffer pointer */ + struct xfs_btree_block *right; /* right btree block */ + union xfs_btree_ptr rptr; + union xfs_btree_ptr lptr; + + XFS_BTREE_TRACE_CURSOR(cur, XBT_ENTRY); + XFS_BTREE_STATS_INC(cur, newroot); + + /* initialise our start point from the cursor */ + cur->bc_ops->init_ptr_from_cur(cur, &rptr); + + /* Allocate the new block. If we can't do it, we're toast. Give up. */ + error = cur->bc_ops->alloc_block(cur, &rptr, &lptr, 1, stat); + if (error) + goto error0; + if (*stat == 0) + goto out0; + XFS_BTREE_STATS_INC(cur, alloc); + + /* Set up the new block. */ + error = xfs_btree_get_buf_block(cur, &lptr, 0, &new, &nbp); + if (error) + goto error0; + + /* Set the root in the holding structure increasing the level by 1. */ + cur->bc_ops->set_root(cur, &lptr, 1); + + /* + * At the previous root level there are now two blocks: the old root, + * and the new block generated when it was split. We don't know which + * one the cursor is pointing at, so we set up variables "left" and + * "right" for each case. + */ + block = xfs_btree_get_block(cur, cur->bc_nlevels - 1, &bp); + +#ifdef DEBUG + error = xfs_btree_check_block(cur, block, cur->bc_nlevels - 1, bp); + if (error) + goto error0; +#endif + + xfs_btree_get_sibling(cur, block, &rptr, XFS_BB_RIGHTSIB); + if (!xfs_btree_ptr_is_null(cur, &rptr)) { + /* Our block is left, pick up the right block. */ + lbp = bp; + xfs_btree_buf_to_ptr(cur, lbp, &lptr); + left = block; + error = xfs_btree_read_buf_block(cur, &rptr, + cur->bc_nlevels - 1, 0, &right, &rbp); + if (error) + goto error0; + bp = rbp; + nptr = 1; + } else { + /* Our block is right, pick up the left block. */ + rbp = bp; + xfs_btree_buf_to_ptr(cur, rbp, &rptr); + right = block; + xfs_btree_get_sibling(cur, right, &lptr, XFS_BB_LEFTSIB); + error = xfs_btree_read_buf_block(cur, &lptr, + cur->bc_nlevels - 1, 0, &left, &lbp); + if (error) + goto error0; + bp = lbp; + nptr = 2; + } + /* Fill in the new block's btree header and log it. */ + xfs_btree_init_block(cur, cur->bc_nlevels, 2, new); + xfs_btree_log_block(cur, nbp, XFS_BB_ALL_BITS); + ASSERT(!xfs_btree_ptr_is_null(cur, &lptr) && + !xfs_btree_ptr_is_null(cur, &rptr)); + + /* Fill in the key data in the new root. */ + if (xfs_btree_get_level(left) > 0) { + xfs_btree_copy_keys(cur, + xfs_btree_key_addr(cur, 1, new), + xfs_btree_key_addr(cur, 1, left), 1); + xfs_btree_copy_keys(cur, + xfs_btree_key_addr(cur, 2, new), + xfs_btree_key_addr(cur, 1, right), 1); + } else { + cur->bc_ops->init_key_from_rec( + xfs_btree_key_addr(cur, 1, new), + xfs_btree_rec_addr(cur, 1, left)); + cur->bc_ops->init_key_from_rec( + xfs_btree_key_addr(cur, 2, new), + xfs_btree_rec_addr(cur, 1, right)); + } + xfs_btree_log_keys(cur, nbp, 1, 2); + + /* Fill in the pointer data in the new root. */ + xfs_btree_copy_ptrs(cur, + xfs_btree_ptr_addr(cur, 1, new), &lptr, 1); + xfs_btree_copy_ptrs(cur, + xfs_btree_ptr_addr(cur, 2, new), &rptr, 1); + xfs_btree_log_ptrs(cur, nbp, 1, 2); + + /* Fix up the cursor. */ + xfs_btree_setbuf(cur, cur->bc_nlevels, nbp); + cur->bc_ptrs[cur->bc_nlevels] = nptr; + cur->bc_nlevels++; + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); + *stat = 1; + return 0; +error0: + XFS_BTREE_TRACE_CURSOR(cur, XBT_ERROR); + return error; +out0: + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); + *stat = 0; + return 0; +} + +STATIC int +xfs_btree_make_block_unfull( + struct xfs_btree_cur *cur, /* btree cursor */ + int level, /* btree level */ + int numrecs,/* # of recs in block */ + int *oindex,/* old tree index */ + int *index, /* new tree index */ + union xfs_btree_ptr *nptr, /* new btree ptr */ + struct xfs_btree_cur **ncur, /* new btree cursor */ + union xfs_btree_rec *nrec, /* new record */ + int *stat) +{ + union xfs_btree_key key; /* new btree key value */ + int error = 0; + + if ((cur->bc_flags & XFS_BTREE_ROOT_IN_INODE) && + level == cur->bc_nlevels - 1) { + struct xfs_inode *ip = cur->bc_private.b.ip; + + if (numrecs < cur->bc_ops->get_dmaxrecs(cur, level)) { + /* A root block that can be made bigger. */ + + xfs_iroot_realloc(ip, 1, cur->bc_private.b.whichfork); + } else { + /* A root block that needs replacing */ + int logflags = 0; + + error = xfs_btree_new_iroot(cur, &logflags, stat); + if (error || *stat == 0) + return error; + + xfs_trans_log_inode(cur->bc_tp, ip, logflags); + } + + return 0; + } + + /* First, try shifting an entry to the right neighbor. */ + error = xfs_btree_rshift(cur, level, stat); + if (error || *stat) + return error; + + /* Next, try shifting an entry to the left neighbor. */ + error = xfs_btree_lshift(cur, level, stat); + if (error) + return error; + + if (*stat) { + *oindex = *index = cur->bc_ptrs[level]; + return 0; + } + + /* + * Next, try splitting the current block in half. + * + * If this works we have to re-set our variables because we + * could be in a different block now. + */ + error = xfs_btree_split(cur, level, nptr, &key, ncur, stat); + if (error || *stat == 0) + return error; + + + *index = cur->bc_ptrs[level]; + cur->bc_ops->init_rec_from_key(&key, nrec); + return 0; +} + +/* + * Insert one record/level. Return information to the caller + * allowing the next level up to proceed if necessary. + */ +STATIC int +xfs_btree_insrec( + struct xfs_btree_cur *cur, /* btree cursor */ + int level, /* level to insert record at */ + union xfs_btree_ptr *ptrp, /* i/o: block number inserted */ + union xfs_btree_rec *recp, /* i/o: record data inserted */ + struct xfs_btree_cur **curp, /* output: new cursor replacing cur */ + int *stat) /* success/failure */ +{ + struct xfs_btree_block *block; /* btree block */ + struct xfs_buf *bp; /* buffer for block */ + union xfs_btree_key key; /* btree key */ + union xfs_btree_ptr nptr; /* new block ptr */ + struct xfs_btree_cur *ncur; /* new btree cursor */ + union xfs_btree_rec nrec; /* new record count */ + int optr; /* old key/record index */ + int ptr; /* key/record index */ + int numrecs;/* number of records */ + int error; /* error return value */ +#ifdef DEBUG + int i; +#endif + + XFS_BTREE_TRACE_CURSOR(cur, XBT_ENTRY); + XFS_BTREE_TRACE_ARGIPR(cur, level, *ptrp, recp); + + ncur = NULL; + + /* + * If we have an external root pointer, and we've made it to the + * root level, allocate a new root block and we're done. + */ + if (!(cur->bc_flags & XFS_BTREE_ROOT_IN_INODE) && + (level >= cur->bc_nlevels)) { + error = xfs_btree_new_root(cur, stat); + xfs_btree_set_ptr_null(cur, ptrp); + + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); + return error; + } + + /* If we're off the left edge, return failure. */ + ptr = cur->bc_ptrs[level]; + if (ptr == 0) { + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); + *stat = 0; + return 0; + } + + /* Make a key out of the record data to be inserted, and save it. */ + cur->bc_ops->init_key_from_rec(&key, recp); + + optr = ptr; + + XFS_BTREE_STATS_INC(cur, insrec); + + /* Get pointers to the btree buffer and block. */ + block = xfs_btree_get_block(cur, level, &bp); + numrecs = xfs_btree_get_numrecs(block); + +#ifdef DEBUG + error = xfs_btree_check_block(cur, block, level, bp); + if (error) + goto error0; + + /* Check that the new entry is being inserted in the right place. */ + if (ptr <= numrecs) { + if (level == 0) { + ASSERT(cur->bc_ops->recs_inorder(cur, recp, + xfs_btree_rec_addr(cur, ptr, block))); + } else { + ASSERT(cur->bc_ops->keys_inorder(cur, &key, + xfs_btree_key_addr(cur, ptr, block))); + } + } +#endif + + /* + * If the block is full, we can't insert the new entry until we + * make the block un-full. + */ + xfs_btree_set_ptr_null(cur, &nptr); + if (numrecs == cur->bc_ops->get_maxrecs(cur, level)) { + error = xfs_btree_make_block_unfull(cur, level, numrecs, + &optr, &ptr, &nptr, &ncur, &nrec, stat); + if (error || *stat == 0) + goto error0; + } + + /* + * The current block may have changed if the block was + * previously full and we have just made space in it. + */ + block = xfs_btree_get_block(cur, level, &bp); + numrecs = xfs_btree_get_numrecs(block); + +#ifdef DEBUG + error = xfs_btree_check_block(cur, block, level, bp); + if (error) + return error; +#endif + + /* + * At this point we know there's room for our new entry in the block + * we're pointing at. + */ + XFS_BTREE_STATS_ADD(cur, moves, numrecs - ptr + 1); + + if (level > 0) { + /* It's a nonleaf. make a hole in the keys and ptrs */ + union xfs_btree_key *kp; + union xfs_btree_ptr *pp; + + kp = xfs_btree_key_addr(cur, ptr, block); + pp = xfs_btree_ptr_addr(cur, ptr, block); + +#ifdef DEBUG + for (i = numrecs - ptr; i >= 0; i--) { + error = xfs_btree_check_ptr(cur, pp, i, level); + if (error) + return error; + } +#endif + + xfs_btree_shift_keys(cur, kp, 1, numrecs - ptr + 1); + xfs_btree_shift_ptrs(cur, pp, 1, numrecs - ptr + 1); + +#ifdef DEBUG + error = xfs_btree_check_ptr(cur, ptrp, 0, level); + if (error) + goto error0; +#endif + + /* Now put the new data in, bump numrecs and log it. */ + xfs_btree_copy_keys(cur, kp, &key, 1); + xfs_btree_copy_ptrs(cur, pp, ptrp, 1); + numrecs++; + xfs_btree_set_numrecs(block, numrecs); + xfs_btree_log_ptrs(cur, bp, ptr, numrecs); + xfs_btree_log_keys(cur, bp, ptr, numrecs); +#ifdef DEBUG + if (ptr < numrecs) { + ASSERT(cur->bc_ops->keys_inorder(cur, kp, + xfs_btree_key_addr(cur, ptr + 1, block))); + } +#endif + } else { + /* It's a leaf. make a hole in the records */ + union xfs_btree_rec *rp; + + rp = xfs_btree_rec_addr(cur, ptr, block); + + xfs_btree_shift_recs(cur, rp, 1, numrecs - ptr + 1); + + /* Now put the new data in, bump numrecs and log it. */ + xfs_btree_copy_recs(cur, rp, recp, 1); + xfs_btree_set_numrecs(block, ++numrecs); + xfs_btree_log_recs(cur, bp, ptr, numrecs); +#ifdef DEBUG + if (ptr < numrecs) { + ASSERT(cur->bc_ops->recs_inorder(cur, rp, + xfs_btree_rec_addr(cur, ptr + 1, block))); + } +#endif + } + + /* Log the new number of records in the btree header. */ + xfs_btree_log_block(cur, bp, XFS_BB_NUMRECS); + + /* If we inserted at the start of a block, update the parents' keys. */ + if (optr == 1) { + error = xfs_btree_updkey(cur, &key, level + 1); + if (error) + goto error0; + } + + /* + * If we are tracking the last record in the tree and + * we are at the far right edge of the tree, update it. + */ + if (xfs_btree_is_lastrec(cur, block, level)) { + cur->bc_ops->update_lastrec(cur, block, recp, + ptr, LASTREC_INSREC); + } + + /* + * Return the new block number, if any. + * If there is one, give back a record value and a cursor too. + */ + *ptrp = nptr; + if (!xfs_btree_ptr_is_null(cur, &nptr)) { + *recp = nrec; + *curp = ncur; + } + + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); + *stat = 1; + return 0; + +error0: + XFS_BTREE_TRACE_CURSOR(cur, XBT_ERROR); + return error; +} + +/* + * Insert the record at the point referenced by cur. + * + * A multi-level split of the tree on insert will invalidate the original + * cursor. All callers of this function should assume that the cursor is + * no longer valid and revalidate it. + */ +int +xfs_btree_insert( + struct xfs_btree_cur *cur, + int *stat) +{ + int error; /* error return value */ + int i; /* result value, 0 for failure */ + int level; /* current level number in btree */ + union xfs_btree_ptr nptr; /* new block number (split result) */ + struct xfs_btree_cur *ncur; /* new cursor (split result) */ + struct xfs_btree_cur *pcur; /* previous level's cursor */ + union xfs_btree_rec rec; /* record to insert */ + + level = 0; + ncur = NULL; + pcur = cur; + + xfs_btree_set_ptr_null(cur, &nptr); + cur->bc_ops->init_rec_from_cur(cur, &rec); + + /* + * Loop going up the tree, starting at the leaf level. + * Stop when we don't get a split block, that must mean that + * the insert is finished with this level. + */ + do { + /* + * Insert nrec/nptr into this level of the tree. + * Note if we fail, nptr will be null. + */ + error = xfs_btree_insrec(pcur, level, &nptr, &rec, &ncur, &i); + if (error) { + if (pcur != cur) + xfs_btree_del_cursor(pcur, XFS_BTREE_ERROR); + goto error0; + } + + XFS_WANT_CORRUPTED_GOTO(i == 1, error0); + level++; + + /* + * See if the cursor we just used is trash. + * Can't trash the caller's cursor, but otherwise we should + * if ncur is a new cursor or we're about to be done. + */ + if (pcur != cur && + (ncur || xfs_btree_ptr_is_null(cur, &nptr))) { + /* Save the state from the cursor before we trash it */ + if (cur->bc_ops->update_cursor) + cur->bc_ops->update_cursor(pcur, cur); + cur->bc_nlevels = pcur->bc_nlevels; + xfs_btree_del_cursor(pcur, XFS_BTREE_NOERROR); + } + /* If we got a new cursor, switch to it. */ + if (ncur) { + pcur = ncur; + ncur = NULL; + } + } while (!xfs_btree_ptr_is_null(cur, &nptr)); + + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); + *stat = i; + return 0; +error0: + XFS_BTREE_TRACE_CURSOR(cur, XBT_ERROR); + return error; +} + +/* + * Try to merge a non-leaf block back into the inode root. + * + * Note: the killroot names comes from the fact that we're effectively + * killing the old root block. But because we can't just delete the + * inode we have to copy the single block it was pointing to into the + * inode. + */ +int +xfs_btree_kill_iroot( + struct xfs_btree_cur *cur) +{ + int whichfork = cur->bc_private.b.whichfork; + struct xfs_inode *ip = cur->bc_private.b.ip; + struct xfs_ifork *ifp = XFS_IFORK_PTR(ip, whichfork); + struct xfs_btree_block *block; + struct xfs_btree_block *cblock; + union xfs_btree_key *kp; + union xfs_btree_key *ckp; + union xfs_btree_ptr *pp; + union xfs_btree_ptr *cpp; + struct xfs_buf *cbp; + int level; + int index; + int numrecs; +#ifdef DEBUG + union xfs_btree_ptr ptr; + int i; +#endif + + XFS_BTREE_TRACE_CURSOR(cur, XBT_ENTRY); + + ASSERT(cur->bc_flags & XFS_BTREE_ROOT_IN_INODE); + ASSERT(cur->bc_nlevels > 1); + + /* + * Don't deal with the root block needs to be a leaf case. + * We're just going to turn the thing back into extents anyway. + */ + level = cur->bc_nlevels - 1; + if (level == 1) + goto out0; + + /* + * Give up if the root has multiple children. + */ + block = xfs_btree_get_iroot(cur); + if (xfs_btree_get_numrecs(block) != 1) + goto out0; + + cblock = xfs_btree_get_block(cur, level - 1, &cbp); + numrecs = xfs_btree_get_numrecs(cblock); + + /* + * Only do this if the next level will fit. + * Then the data must be copied up to the inode, + * instead of freeing the root you free the next level. + */ + if (numrecs > cur->bc_ops->get_dmaxrecs(cur, level)) + goto out0; + + XFS_BTREE_STATS_INC(cur, killroot); + +#ifdef DEBUG + xfs_btree_get_sibling(cur, block, &ptr, XFS_BB_LEFTSIB); + ASSERT(xfs_btree_ptr_is_null(cur, &ptr)); + xfs_btree_get_sibling(cur, block, &ptr, XFS_BB_RIGHTSIB); + ASSERT(xfs_btree_ptr_is_null(cur, &ptr)); +#endif + + index = numrecs - cur->bc_ops->get_maxrecs(cur, level); + if (index) { + xfs_iroot_realloc(cur->bc_private.b.ip, index, + cur->bc_private.b.whichfork); + block = ifp->if_broot; + } + + be16_add_cpu(&block->bb_numrecs, index); + ASSERT(block->bb_numrecs == cblock->bb_numrecs); + + kp = xfs_btree_key_addr(cur, 1, block); + ckp = xfs_btree_key_addr(cur, 1, cblock); + xfs_btree_copy_keys(cur, kp, ckp, numrecs); + + pp = xfs_btree_ptr_addr(cur, 1, block); + cpp = xfs_btree_ptr_addr(cur, 1, cblock); +#ifdef DEBUG + for (i = 0; i < numrecs; i++) { + int error; + + error = xfs_btree_check_ptr(cur, cpp, i, level - 1); + if (error) { + XFS_BTREE_TRACE_CURSOR(cur, XBT_ERROR); + return error; + } + } +#endif + xfs_btree_copy_ptrs(cur, pp, cpp, numrecs); + + cur->bc_ops->free_block(cur, cbp); + XFS_BTREE_STATS_INC(cur, free); + + cur->bc_bufs[level - 1] = NULL; + be16_add_cpu(&block->bb_level, -1); + xfs_trans_log_inode(cur->bc_tp, ip, + XFS_ILOG_CORE | XFS_ILOG_FBROOT(cur->bc_private.b.whichfork)); + cur->bc_nlevels--; +out0: + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); + return 0; +} + +STATIC int +xfs_btree_dec_cursor( + struct xfs_btree_cur *cur, + int level, + int *stat) +{ + int error; + int i; + + if (level > 0) { + error = xfs_btree_decrement(cur, level, &i); + if (error) + return error; + } + + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); + *stat = 1; + return 0; +} + +/* + * Single level of the btree record deletion routine. + * Delete record pointed to by cur/level. + * Remove the record from its block then rebalance the tree. + * Return 0 for error, 1 for done, 2 to go on to the next level. + */ +STATIC int /* error */ +xfs_btree_delrec( + struct xfs_btree_cur *cur, /* btree cursor */ + int level, /* level removing record from */ + int *stat) /* fail/done/go-on */ +{ + struct xfs_btree_block *block; /* btree block */ + union xfs_btree_ptr cptr; /* current block ptr */ + struct xfs_buf *bp; /* buffer for block */ + int error; /* error return value */ + int i; /* loop counter */ + union xfs_btree_key key; /* storage for keyp */ + union xfs_btree_key *keyp = &key; /* passed to the next level */ + union xfs_btree_ptr lptr; /* left sibling block ptr */ + struct xfs_buf *lbp; /* left buffer pointer */ + struct xfs_btree_block *left; /* left btree block */ + int lrecs = 0; /* left record count */ + int ptr; /* key/record index */ + union xfs_btree_ptr rptr; /* right sibling block ptr */ + struct xfs_buf *rbp; /* right buffer pointer */ + struct xfs_btree_block *right; /* right btree block */ + struct xfs_btree_block *rrblock; /* right-right btree block */ + struct xfs_buf *rrbp; /* right-right buffer pointer */ + int rrecs = 0; /* right record count */ + struct xfs_btree_cur *tcur; /* temporary btree cursor */ + int numrecs; /* temporary numrec count */ + + XFS_BTREE_TRACE_CURSOR(cur, XBT_ENTRY); + XFS_BTREE_TRACE_ARGI(cur, level); + + tcur = NULL; + + /* Get the index of the entry being deleted, check for nothing there. */ + ptr = cur->bc_ptrs[level]; + if (ptr == 0) { + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); + *stat = 0; + return 0; + } + + /* Get the buffer & block containing the record or key/ptr. */ + block = xfs_btree_get_block(cur, level, &bp); + numrecs = xfs_btree_get_numrecs(block); + +#ifdef DEBUG + error = xfs_btree_check_block(cur, block, level, bp); + if (error) + goto error0; +#endif + + /* Fail if we're off the end of the block. */ + if (ptr > numrecs) { + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); + *stat = 0; + return 0; + } + + XFS_BTREE_STATS_INC(cur, delrec); + XFS_BTREE_STATS_ADD(cur, moves, numrecs - ptr); + + /* Excise the entries being deleted. */ + if (level > 0) { + /* It's a nonleaf. operate on keys and ptrs */ + union xfs_btree_key *lkp; + union xfs_btree_ptr *lpp; + + lkp = xfs_btree_key_addr(cur, ptr + 1, block); + lpp = xfs_btree_ptr_addr(cur, ptr + 1, block); + +#ifdef DEBUG + for (i = 0; i < numrecs - ptr; i++) { + error = xfs_btree_check_ptr(cur, lpp, i, level); + if (error) + goto error0; + } +#endif + + if (ptr < numrecs) { + xfs_btree_shift_keys(cur, lkp, -1, numrecs - ptr); + xfs_btree_shift_ptrs(cur, lpp, -1, numrecs - ptr); + xfs_btree_log_keys(cur, bp, ptr, numrecs - 1); + xfs_btree_log_ptrs(cur, bp, ptr, numrecs - 1); + } + + /* + * If it's the first record in the block, we'll need to pass a + * key up to the next level (updkey). + */ + if (ptr == 1) + keyp = xfs_btree_key_addr(cur, 1, block); + } else { + /* It's a leaf. operate on records */ + if (ptr < numrecs) { + xfs_btree_shift_recs(cur, + xfs_btree_rec_addr(cur, ptr + 1, block), + -1, numrecs - ptr); + xfs_btree_log_recs(cur, bp, ptr, numrecs - 1); + } + + /* + * If it's the first record in the block, we'll need a key + * structure to pass up to the next level (updkey). + */ + if (ptr == 1) { + cur->bc_ops->init_key_from_rec(&key, + xfs_btree_rec_addr(cur, 1, block)); + keyp = &key; + } + } + + /* + * Decrement and log the number of entries in the block. + */ + xfs_btree_set_numrecs(block, --numrecs); + xfs_btree_log_block(cur, bp, XFS_BB_NUMRECS); + + /* + * If we are tracking the last record in the tree and + * we are at the far right edge of the tree, update it. + */ + if (xfs_btree_is_lastrec(cur, block, level)) { + cur->bc_ops->update_lastrec(cur, block, NULL, + ptr, LASTREC_DELREC); + } + + /* + * We're at the root level. First, shrink the root block in-memory. + * Try to get rid of the next level down. If we can't then there's + * nothing left to do. + */ + if (level == cur->bc_nlevels - 1) { + if (cur->bc_flags & XFS_BTREE_ROOT_IN_INODE) { + xfs_iroot_realloc(cur->bc_private.b.ip, -1, + cur->bc_private.b.whichfork); + + error = xfs_btree_kill_iroot(cur); + if (error) + goto error0; + + error = xfs_btree_dec_cursor(cur, level, stat); + if (error) + goto error0; + *stat = 1; + return 0; + } + + /* + * If this is the root level, and there's only one entry left, + * and it's NOT the leaf level, then we can get rid of this + * level. + */ + if (numrecs == 1 && level > 0) { + union xfs_btree_ptr *pp; + /* + * pp is still set to the first pointer in the block. + * Make it the new root of the btree. + */ + pp = xfs_btree_ptr_addr(cur, 1, block); + error = cur->bc_ops->kill_root(cur, bp, level, pp); + if (error) + goto error0; + } else if (level > 0) { + error = xfs_btree_dec_cursor(cur, level, stat); + if (error) + goto error0; + } + *stat = 1; + return 0; + } + + /* + * If we deleted the leftmost entry in the block, update the + * key values above us in the tree. + */ + if (ptr == 1) { + error = xfs_btree_updkey(cur, keyp, level + 1); + if (error) + goto error0; + } + + /* + * If the number of records remaining in the block is at least + * the minimum, we're done. + */ + if (numrecs >= cur->bc_ops->get_minrecs(cur, level)) { + error = xfs_btree_dec_cursor(cur, level, stat); + if (error) + goto error0; + return 0; + } + + /* + * Otherwise, we have to move some records around to keep the + * tree balanced. Look at the left and right sibling blocks to + * see if we can re-balance by moving only one record. + */ + xfs_btree_get_sibling(cur, block, &rptr, XFS_BB_RIGHTSIB); + xfs_btree_get_sibling(cur, block, &lptr, XFS_BB_LEFTSIB); + + if (cur->bc_flags & XFS_BTREE_ROOT_IN_INODE) { + /* + * One child of root, need to get a chance to copy its contents + * into the root and delete it. Can't go up to next level, + * there's nothing to delete there. + */ + if (xfs_btree_ptr_is_null(cur, &rptr) && + xfs_btree_ptr_is_null(cur, &lptr) && + level == cur->bc_nlevels - 2) { + error = xfs_btree_kill_iroot(cur); + if (!error) + error = xfs_btree_dec_cursor(cur, level, stat); + if (error) + goto error0; + return 0; + } + } + + ASSERT(!xfs_btree_ptr_is_null(cur, &rptr) || + !xfs_btree_ptr_is_null(cur, &lptr)); + + /* + * Duplicate the cursor so our btree manipulations here won't + * disrupt the next level up. + */ + error = xfs_btree_dup_cursor(cur, &tcur); + if (error) + goto error0; + + /* + * If there's a right sibling, see if it's ok to shift an entry + * out of it. + */ + if (!xfs_btree_ptr_is_null(cur, &rptr)) { + /* + * Move the temp cursor to the last entry in the next block. + * Actually any entry but the first would suffice. + */ + i = xfs_btree_lastrec(tcur, level); + XFS_WANT_CORRUPTED_GOTO(i == 1, error0); + + error = xfs_btree_increment(tcur, level, &i); + if (error) + goto error0; + XFS_WANT_CORRUPTED_GOTO(i == 1, error0); + + i = xfs_btree_lastrec(tcur, level); + XFS_WANT_CORRUPTED_GOTO(i == 1, error0); + + /* Grab a pointer to the block. */ + right = xfs_btree_get_block(tcur, level, &rbp); +#ifdef DEBUG + error = xfs_btree_check_block(tcur, right, level, rbp); + if (error) + goto error0; +#endif + /* Grab the current block number, for future use. */ + xfs_btree_get_sibling(tcur, right, &cptr, XFS_BB_LEFTSIB); + + /* + * If right block is full enough so that removing one entry + * won't make it too empty, and left-shifting an entry out + * of right to us works, we're done. + */ + if (xfs_btree_get_numrecs(right) - 1 >= + cur->bc_ops->get_minrecs(tcur, level)) { + error = xfs_btree_lshift(tcur, level, &i); + if (error) + goto error0; + if (i) { + ASSERT(xfs_btree_get_numrecs(block) >= + cur->bc_ops->get_minrecs(tcur, level)); + + xfs_btree_del_cursor(tcur, XFS_BTREE_NOERROR); + tcur = NULL; + + error = xfs_btree_dec_cursor(cur, level, stat); + if (error) + goto error0; + return 0; + } + } + + /* + * Otherwise, grab the number of records in right for + * future reference, and fix up the temp cursor to point + * to our block again (last record). + */ + rrecs = xfs_btree_get_numrecs(right); + if (!xfs_btree_ptr_is_null(cur, &lptr)) { + i = xfs_btree_firstrec(tcur, level); + XFS_WANT_CORRUPTED_GOTO(i == 1, error0); + + error = xfs_btree_decrement(tcur, level, &i); + if (error) + goto error0; + XFS_WANT_CORRUPTED_GOTO(i == 1, error0); + } + } + + /* + * If there's a left sibling, see if it's ok to shift an entry + * out of it. + */ + if (!xfs_btree_ptr_is_null(cur, &lptr)) { + /* + * Move the temp cursor to the first entry in the + * previous block. + */ + i = xfs_btree_firstrec(tcur, level); + XFS_WANT_CORRUPTED_GOTO(i == 1, error0); + + error = xfs_btree_decrement(tcur, level, &i); + if (error) + goto error0; + i = xfs_btree_firstrec(tcur, level); + XFS_WANT_CORRUPTED_GOTO(i == 1, error0); + + /* Grab a pointer to the block. */ + left = xfs_btree_get_block(tcur, level, &lbp); +#ifdef DEBUG + error = xfs_btree_check_block(cur, left, level, lbp); + if (error) + goto error0; +#endif + /* Grab the current block number, for future use. */ + xfs_btree_get_sibling(tcur, left, &cptr, XFS_BB_RIGHTSIB); + + /* + * If left block is full enough so that removing one entry + * won't make it too empty, and right-shifting an entry out + * of left to us works, we're done. + */ + if (xfs_btree_get_numrecs(left) - 1 >= + cur->bc_ops->get_minrecs(tcur, level)) { + error = xfs_btree_rshift(tcur, level, &i); + if (error) + goto error0; + if (i) { + ASSERT(xfs_btree_get_numrecs(block) >= + cur->bc_ops->get_minrecs(tcur, level)); + xfs_btree_del_cursor(tcur, XFS_BTREE_NOERROR); + tcur = NULL; + if (level == 0) + cur->bc_ptrs[0]++; + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); + *stat = 1; + return 0; + } + } + + /* + * Otherwise, grab the number of records in right for + * future reference. + */ + lrecs = xfs_btree_get_numrecs(left); + } + + /* Delete the temp cursor, we're done with it. */ + xfs_btree_del_cursor(tcur, XFS_BTREE_NOERROR); + tcur = NULL; + + /* If here, we need to do a join to keep the tree balanced. */ + ASSERT(!xfs_btree_ptr_is_null(cur, &cptr)); + + if (!xfs_btree_ptr_is_null(cur, &lptr) && + lrecs + xfs_btree_get_numrecs(block) <= + cur->bc_ops->get_maxrecs(cur, level)) { + /* + * Set "right" to be the starting block, + * "left" to be the left neighbor. + */ + rptr = cptr; + right = block; + rbp = bp; + error = xfs_btree_read_buf_block(cur, &lptr, level, + 0, &left, &lbp); + if (error) + goto error0; + + /* + * If that won't work, see if we can join with the right neighbor block. + */ + } else if (!xfs_btree_ptr_is_null(cur, &rptr) && + rrecs + xfs_btree_get_numrecs(block) <= + cur->bc_ops->get_maxrecs(cur, level)) { + /* + * Set "left" to be the starting block, + * "right" to be the right neighbor. + */ + lptr = cptr; + left = block; + lbp = bp; + error = xfs_btree_read_buf_block(cur, &rptr, level, + 0, &right, &rbp); + if (error) + goto error0; + + /* + * Otherwise, we can't fix the imbalance. + * Just return. This is probably a logic error, but it's not fatal. + */ + } else { + error = xfs_btree_dec_cursor(cur, level, stat); + if (error) + goto error0; + return 0; + } + + rrecs = xfs_btree_get_numrecs(right); + lrecs = xfs_btree_get_numrecs(left); + + /* + * We're now going to join "left" and "right" by moving all the stuff + * in "right" to "left" and deleting "right". + */ + XFS_BTREE_STATS_ADD(cur, moves, rrecs); + if (level > 0) { + /* It's a non-leaf. Move keys and pointers. */ + union xfs_btree_key *lkp; /* left btree key */ + union xfs_btree_ptr *lpp; /* left address pointer */ + union xfs_btree_key *rkp; /* right btree key */ + union xfs_btree_ptr *rpp; /* right address pointer */ + + lkp = xfs_btree_key_addr(cur, lrecs + 1, left); + lpp = xfs_btree_ptr_addr(cur, lrecs + 1, left); + rkp = xfs_btree_key_addr(cur, 1, right); + rpp = xfs_btree_ptr_addr(cur, 1, right); +#ifdef DEBUG + for (i = 1; i < rrecs; i++) { + error = xfs_btree_check_ptr(cur, rpp, i, level); + if (error) + goto error0; + } +#endif + xfs_btree_copy_keys(cur, lkp, rkp, rrecs); + xfs_btree_copy_ptrs(cur, lpp, rpp, rrecs); + + xfs_btree_log_keys(cur, lbp, lrecs + 1, lrecs + rrecs); + xfs_btree_log_ptrs(cur, lbp, lrecs + 1, lrecs + rrecs); + } else { + /* It's a leaf. Move records. */ + union xfs_btree_rec *lrp; /* left record pointer */ + union xfs_btree_rec *rrp; /* right record pointer */ + + lrp = xfs_btree_rec_addr(cur, lrecs + 1, left); + rrp = xfs_btree_rec_addr(cur, 1, right); + + xfs_btree_copy_recs(cur, lrp, rrp, rrecs); + xfs_btree_log_recs(cur, lbp, lrecs + 1, lrecs + rrecs); + } + + XFS_BTREE_STATS_INC(cur, join); + + /* + * Fix up the the number of records and right block pointer in the + * surviving block, and log it. + */ + xfs_btree_set_numrecs(left, lrecs + rrecs); + xfs_btree_get_sibling(cur, right, &cptr, XFS_BB_RIGHTSIB), + xfs_btree_set_sibling(cur, left, &cptr, XFS_BB_RIGHTSIB); + xfs_btree_log_block(cur, lbp, XFS_BB_NUMRECS | XFS_BB_RIGHTSIB); + + /* If there is a right sibling, point it to the remaining block. */ + xfs_btree_get_sibling(cur, left, &cptr, XFS_BB_RIGHTSIB); + if (!xfs_btree_ptr_is_null(cur, &cptr)) { + error = xfs_btree_read_buf_block(cur, &cptr, level, + 0, &rrblock, &rrbp); + if (error) + goto error0; + xfs_btree_set_sibling(cur, rrblock, &lptr, XFS_BB_LEFTSIB); + xfs_btree_log_block(cur, rrbp, XFS_BB_LEFTSIB); + } + + /* Free the deleted block. */ + error = cur->bc_ops->free_block(cur, rbp); + if (error) + goto error0; + XFS_BTREE_STATS_INC(cur, free); + + /* + * If we joined with the left neighbor, set the buffer in the + * cursor to the left block, and fix up the index. + */ + if (bp != lbp) { + cur->bc_bufs[level] = lbp; + cur->bc_ptrs[level] += lrecs; + cur->bc_ra[level] = 0; + } + /* + * If we joined with the right neighbor and there's a level above + * us, increment the cursor at that level. + */ + else if ((cur->bc_flags & XFS_BTREE_ROOT_IN_INODE) || + (level + 1 < cur->bc_nlevels)) { + error = xfs_btree_increment(cur, level + 1, &i); + if (error) + goto error0; + } + + /* + * Readjust the ptr at this level if it's not a leaf, since it's + * still pointing at the deletion point, which makes the cursor + * inconsistent. If this makes the ptr 0, the caller fixes it up. + * We can't use decrement because it would change the next level up. + */ + if (level > 0) + cur->bc_ptrs[level]--; + + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); + /* Return value means the next level up has something to do. */ + *stat = 2; + return 0; + +error0: + XFS_BTREE_TRACE_CURSOR(cur, XBT_ERROR); + if (tcur) + xfs_btree_del_cursor(tcur, XFS_BTREE_ERROR); + return error; +} + +/* + * Delete the record pointed to by cur. + * The cursor refers to the place where the record was (could be inserted) + * when the operation returns. + */ +int /* error */ +xfs_btree_delete( + struct xfs_btree_cur *cur, + int *stat) /* success/failure */ +{ + int error; /* error return value */ + int level; + int i; + + XFS_BTREE_TRACE_CURSOR(cur, XBT_ENTRY); + + /* + * Go up the tree, starting at leaf level. + * + * If 2 is returned then a join was done; go to the next level. + * Otherwise we are done. + */ + for (level = 0, i = 2; i == 2; level++) { + error = xfs_btree_delrec(cur, level, &i); + if (error) + goto error0; + } + + if (i == 0) { + for (level = 1; level < cur->bc_nlevels; level++) { + if (cur->bc_ptrs[level] == 0) { + error = xfs_btree_decrement(cur, level, &i); + if (error) + goto error0; + break; + } + } + } + + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); + *stat = i; + return 0; +error0: + XFS_BTREE_TRACE_CURSOR(cur, XBT_ERROR); + return error; +} + +/* + * Get the data from the pointed-to record. + */ +int /* error */ +xfs_btree_get_rec( + struct xfs_btree_cur *cur, /* btree cursor */ + union xfs_btree_rec **recp, /* output: btree record */ + int *stat) /* output: success/failure */ +{ + struct xfs_btree_block *block; /* btree block */ + struct xfs_buf *bp; /* buffer pointer */ + int ptr; /* record number */ +#ifdef DEBUG + int error; /* error return value */ +#endif + + ptr = cur->bc_ptrs[0]; + block = xfs_btree_get_block(cur, 0, &bp); + +#ifdef DEBUG + error = xfs_btree_check_block(cur, block, 0, bp); + if (error) + return error; +#endif + + /* + * Off the right end or left end, return failure. + */ + if (ptr > xfs_btree_get_numrecs(block) || ptr <= 0) { + *stat = 0; + return 0; + } + + /* + * Point to the record and extract its data. + */ + *recp = xfs_btree_rec_addr(cur, ptr, block); + *stat = 1; + return 0; +} diff --git a/fs/xfs/xfs_btree.h b/fs/xfs/xfs_btree.h index 1f528a2a375..789fffdf8b2 100644 --- a/fs/xfs/xfs_btree.h +++ b/fs/xfs/xfs_btree.h @@ -39,39 +39,19 @@ extern kmem_zone_t *xfs_btree_cur_zone; #define XFS_BTNUM_INO ((xfs_btnum_t)XFS_BTNUM_INOi) /* - * Short form header: space allocation btrees. - */ -typedef struct xfs_btree_sblock { - __be32 bb_magic; /* magic number for block type */ - __be16 bb_level; /* 0 is a leaf */ - __be16 bb_numrecs; /* current # of data records */ - __be32 bb_leftsib; /* left sibling block or NULLAGBLOCK */ - __be32 bb_rightsib; /* right sibling block or NULLAGBLOCK */ -} xfs_btree_sblock_t; - -/* - * Long form header: bmap btrees. - */ -typedef struct xfs_btree_lblock { - __be32 bb_magic; /* magic number for block type */ - __be16 bb_level; /* 0 is a leaf */ - __be16 bb_numrecs; /* current # of data records */ - __be64 bb_leftsib; /* left sibling block or NULLDFSBNO */ - __be64 bb_rightsib; /* right sibling block or NULLDFSBNO */ -} xfs_btree_lblock_t; - -/* - * Combined header and structure, used by common code. + * Generic btree header. + * + * This is a comination of the actual format used on disk for short and long + * format btrees. The first three fields are shared by both format, but + * the pointers are different and should be used with care. + * + * To get the size of the actual short or long form headers please use + * the size macros below. Never use sizeof(xfs_btree_block). */ -typedef struct xfs_btree_hdr -{ +struct xfs_btree_block { __be32 bb_magic; /* magic number for block type */ __be16 bb_level; /* 0 is a leaf */ __be16 bb_numrecs; /* current # of data records */ -} xfs_btree_hdr_t; - -typedef struct xfs_btree_block { - xfs_btree_hdr_t bb_h; /* header */ union { struct { __be32 bb_leftsib; @@ -82,7 +62,36 @@ typedef struct xfs_btree_block { __be64 bb_rightsib; } l; /* long form pointers */ } bb_u; /* rest */ -} xfs_btree_block_t; +}; + +#define XFS_BTREE_SBLOCK_LEN 16 /* size of a short form block */ +#define XFS_BTREE_LBLOCK_LEN 24 /* size of a long form block */ + + +/* + * Generic key, ptr and record wrapper structures. + * + * These are disk format structures, and are converted where necessary + * by the btree specific code that needs to interpret them. + */ +union xfs_btree_ptr { + __be32 s; /* short form ptr */ + __be64 l; /* long form ptr */ +}; + +union xfs_btree_key { + xfs_bmbt_key_t bmbt; + xfs_bmdr_key_t bmbr; /* bmbt root block */ + xfs_alloc_key_t alloc; + xfs_inobt_key_t inobt; +}; + +union xfs_btree_rec { + xfs_bmbt_rec_t bmbt; + xfs_bmdr_rec_t bmbr; /* bmbt root block */ + xfs_alloc_rec_t alloc; + xfs_inobt_rec_t inobt; +}; /* * For logging record fields. @@ -96,46 +105,131 @@ typedef struct xfs_btree_block { #define XFS_BB_ALL_BITS ((1 << XFS_BB_NUM_BITS) - 1) /* - * Boolean to select which form of xfs_btree_block_t.bb_u to use. - */ -#define XFS_BTREE_LONG_PTRS(btnum) ((btnum) == XFS_BTNUM_BMAP) - -/* * Magic numbers for btree blocks. */ extern const __uint32_t xfs_magics[]; /* - * Maximum and minimum records in a btree block. - * Given block size, type prefix, and leaf flag (0 or 1). - * The divisor below is equivalent to lf ? (e1) : (e2) but that produces - * compiler warnings. - */ -#define XFS_BTREE_BLOCK_MAXRECS(bsz,t,lf) \ - ((int)(((bsz) - (uint)sizeof(t ## _block_t)) / \ - (((lf) * (uint)sizeof(t ## _rec_t)) + \ - ((1 - (lf)) * \ - ((uint)sizeof(t ## _key_t) + (uint)sizeof(t ## _ptr_t)))))) -#define XFS_BTREE_BLOCK_MINRECS(bsz,t,lf) \ - (XFS_BTREE_BLOCK_MAXRECS(bsz,t,lf) / 2) - -/* - * Record, key, and pointer address calculation macros. - * Given block size, type prefix, block pointer, and index of requested entry - * (first entry numbered 1). - */ -#define XFS_BTREE_REC_ADDR(t,bb,i) \ - ((t ## _rec_t *)((char *)(bb) + sizeof(t ## _block_t) + \ - ((i) - 1) * sizeof(t ## _rec_t))) -#define XFS_BTREE_KEY_ADDR(t,bb,i) \ - ((t ## _key_t *)((char *)(bb) + sizeof(t ## _block_t) + \ - ((i) - 1) * sizeof(t ## _key_t))) -#define XFS_BTREE_PTR_ADDR(t,bb,i,mxr) \ - ((t ## _ptr_t *)((char *)(bb) + sizeof(t ## _block_t) + \ - (mxr) * sizeof(t ## _key_t) + ((i) - 1) * sizeof(t ## _ptr_t))) + * Generic stats interface + */ +#define __XFS_BTREE_STATS_INC(type, stat) \ + XFS_STATS_INC(xs_ ## type ## _2_ ## stat) +#define XFS_BTREE_STATS_INC(cur, stat) \ +do { \ + switch (cur->bc_btnum) { \ + case XFS_BTNUM_BNO: __XFS_BTREE_STATS_INC(abtb, stat); break; \ + case XFS_BTNUM_CNT: __XFS_BTREE_STATS_INC(abtc, stat); break; \ + case XFS_BTNUM_BMAP: __XFS_BTREE_STATS_INC(bmbt, stat); break; \ + case XFS_BTNUM_INO: __XFS_BTREE_STATS_INC(ibt, stat); break; \ + case XFS_BTNUM_MAX: ASSERT(0); /* fucking gcc */ ; break; \ + } \ +} while (0) + +#define __XFS_BTREE_STATS_ADD(type, stat, val) \ + XFS_STATS_ADD(xs_ ## type ## _2_ ## stat, val) +#define XFS_BTREE_STATS_ADD(cur, stat, val) \ +do { \ + switch (cur->bc_btnum) { \ + case XFS_BTNUM_BNO: __XFS_BTREE_STATS_ADD(abtb, stat, val); break; \ + case XFS_BTNUM_CNT: __XFS_BTREE_STATS_ADD(abtc, stat, val); break; \ + case XFS_BTNUM_BMAP: __XFS_BTREE_STATS_ADD(bmbt, stat, val); break; \ + case XFS_BTNUM_INO: __XFS_BTREE_STATS_ADD(ibt, stat, val); break; \ + case XFS_BTNUM_MAX: ASSERT(0); /* fucking gcc */ ; break; \ + } \ +} while (0) #define XFS_BTREE_MAXLEVELS 8 /* max of all btrees */ +struct xfs_btree_ops { + /* size of the key and record structures */ + size_t key_len; + size_t rec_len; + + /* cursor operations */ + struct xfs_btree_cur *(*dup_cursor)(struct xfs_btree_cur *); + void (*update_cursor)(struct xfs_btree_cur *src, + struct xfs_btree_cur *dst); + + /* update btree root pointer */ + void (*set_root)(struct xfs_btree_cur *cur, + union xfs_btree_ptr *nptr, int level_change); + int (*kill_root)(struct xfs_btree_cur *cur, struct xfs_buf *bp, + int level, union xfs_btree_ptr *newroot); + + /* block allocation / freeing */ + int (*alloc_block)(struct xfs_btree_cur *cur, + union xfs_btree_ptr *start_bno, + union xfs_btree_ptr *new_bno, + int length, int *stat); + int (*free_block)(struct xfs_btree_cur *cur, struct xfs_buf *bp); + + /* update last record information */ + void (*update_lastrec)(struct xfs_btree_cur *cur, + struct xfs_btree_block *block, + union xfs_btree_rec *rec, + int ptr, int reason); + + /* records in block/level */ + int (*get_minrecs)(struct xfs_btree_cur *cur, int level); + int (*get_maxrecs)(struct xfs_btree_cur *cur, int level); + + /* records on disk. Matter for the root in inode case. */ + int (*get_dmaxrecs)(struct xfs_btree_cur *cur, int level); + + /* init values of btree structures */ + void (*init_key_from_rec)(union xfs_btree_key *key, + union xfs_btree_rec *rec); + void (*init_rec_from_key)(union xfs_btree_key *key, + union xfs_btree_rec *rec); + void (*init_rec_from_cur)(struct xfs_btree_cur *cur, + union xfs_btree_rec *rec); + void (*init_ptr_from_cur)(struct xfs_btree_cur *cur, + union xfs_btree_ptr *ptr); + + /* difference between key value and cursor value */ + __int64_t (*key_diff)(struct xfs_btree_cur *cur, + union xfs_btree_key *key); + +#ifdef DEBUG + /* check that k1 is lower than k2 */ + int (*keys_inorder)(struct xfs_btree_cur *cur, + union xfs_btree_key *k1, + union xfs_btree_key *k2); + + /* check that r1 is lower than r2 */ + int (*recs_inorder)(struct xfs_btree_cur *cur, + union xfs_btree_rec *r1, + union xfs_btree_rec *r2); +#endif + + /* btree tracing */ +#ifdef XFS_BTREE_TRACE + void (*trace_enter)(struct xfs_btree_cur *, const char *, + char *, int, int, __psunsigned_t, + __psunsigned_t, __psunsigned_t, + __psunsigned_t, __psunsigned_t, + __psunsigned_t, __psunsigned_t, + __psunsigned_t, __psunsigned_t, + __psunsigned_t, __psunsigned_t); + void (*trace_cursor)(struct xfs_btree_cur *, __uint32_t *, + __uint64_t *, __uint64_t *); + void (*trace_key)(struct xfs_btree_cur *, + union xfs_btree_key *, __uint64_t *, + __uint64_t *); + void (*trace_record)(struct xfs_btree_cur *, + union xfs_btree_rec *, __uint64_t *, + __uint64_t *, __uint64_t *); +#endif +}; + +/* + * Reasons for the update_lastrec method to be called. + */ +#define LASTREC_UPDATE 0 +#define LASTREC_INSREC 1 +#define LASTREC_DELREC 2 + + /* * Btree cursor structure. * This collects all information needed by the btree code in one place. @@ -144,6 +238,8 @@ typedef struct xfs_btree_cur { struct xfs_trans *bc_tp; /* transaction we're in, if any */ struct xfs_mount *bc_mp; /* file system mount struct */ + const struct xfs_btree_ops *bc_ops; + uint bc_flags; /* btree features - below */ union { xfs_alloc_rec_incore_t a; xfs_bmbt_irec_t b; @@ -175,94 +271,40 @@ typedef struct xfs_btree_cur } bc_private; /* per-btree type data */ } xfs_btree_cur_t; +/* cursor flags */ +#define XFS_BTREE_LONG_PTRS (1<<0) /* pointers are 64bits long */ +#define XFS_BTREE_ROOT_IN_INODE (1<<1) /* root may be variable size */ +#define XFS_BTREE_LASTREC_UPDATE (1<<2) /* track last rec externally */ + + #define XFS_BTREE_NOERROR 0 #define XFS_BTREE_ERROR 1 /* * Convert from buffer to btree block header. */ -#define XFS_BUF_TO_BLOCK(bp) ((xfs_btree_block_t *)XFS_BUF_PTR(bp)) -#define XFS_BUF_TO_LBLOCK(bp) ((xfs_btree_lblock_t *)XFS_BUF_PTR(bp)) -#define XFS_BUF_TO_SBLOCK(bp) ((xfs_btree_sblock_t *)XFS_BUF_PTR(bp)) +#define XFS_BUF_TO_BLOCK(bp) ((struct xfs_btree_block *)XFS_BUF_PTR(bp)) -#ifdef __KERNEL__ - -#ifdef DEBUG /* - * Debug routine: check that block header is ok. + * Check that block header is ok. */ -void +int xfs_btree_check_block( - xfs_btree_cur_t *cur, /* btree cursor */ - xfs_btree_block_t *block, /* generic btree block pointer */ - int level, /* level of the btree block */ - struct xfs_buf *bp); /* buffer containing block, if any */ - -/* - * Debug routine: check that keys are in the right order. - */ -void -xfs_btree_check_key( - xfs_btnum_t btnum, /* btree identifier */ - void *ak1, /* pointer to left (lower) key */ - void *ak2); /* pointer to right (higher) key */ - -/* - * Debug routine: check that records are in the right order. - */ -void -xfs_btree_check_rec( - xfs_btnum_t btnum, /* btree identifier */ - void *ar1, /* pointer to left (lower) record */ - void *ar2); /* pointer to right (higher) record */ -#else -#define xfs_btree_check_block(a,b,c,d) -#define xfs_btree_check_key(a,b,c) -#define xfs_btree_check_rec(a,b,c) -#endif /* DEBUG */ - -/* - * Checking routine: check that long form block header is ok. - */ -int /* error (0 or EFSCORRUPTED) */ -xfs_btree_check_lblock( - xfs_btree_cur_t *cur, /* btree cursor */ - xfs_btree_lblock_t *block, /* btree long form block pointer */ + struct xfs_btree_cur *cur, /* btree cursor */ + struct xfs_btree_block *block, /* generic btree block pointer */ int level, /* level of the btree block */ struct xfs_buf *bp); /* buffer containing block, if any */ /* - * Checking routine: check that (long) pointer is ok. + * Check that (long) pointer is ok. */ int /* error (0 or EFSCORRUPTED) */ xfs_btree_check_lptr( - xfs_btree_cur_t *cur, /* btree cursor */ + struct xfs_btree_cur *cur, /* btree cursor */ xfs_dfsbno_t ptr, /* btree block disk address */ int level); /* btree block level */ -#define xfs_btree_check_lptr_disk(cur, ptr, level) \ - xfs_btree_check_lptr(cur, be64_to_cpu(ptr), level) - -/* - * Checking routine: check that short form block header is ok. - */ -int /* error (0 or EFSCORRUPTED) */ -xfs_btree_check_sblock( - xfs_btree_cur_t *cur, /* btree cursor */ - xfs_btree_sblock_t *block, /* btree short form block pointer */ - int level, /* level of the btree block */ - struct xfs_buf *bp); /* buffer containing block */ - -/* - * Checking routine: check that (short) pointer is ok. - */ -int /* error (0 or EFSCORRUPTED) */ -xfs_btree_check_sptr( - xfs_btree_cur_t *cur, /* btree cursor */ - xfs_agblock_t ptr, /* btree block disk address */ - int level); /* btree block level */ - /* * Delete the btree cursor. */ @@ -281,15 +323,6 @@ xfs_btree_dup_cursor( xfs_btree_cur_t **ncur);/* output cursor */ /* - * Change the cursor to point to the first record in the current block - * at the given level. Other levels are unaffected. - */ -int /* success=1, failure=0 */ -xfs_btree_firstrec( - xfs_btree_cur_t *cur, /* btree cursor */ - int level); /* level to change */ - -/* * Get a buffer for the block, return it with no data read. * Long-form addressing. */ @@ -313,20 +346,6 @@ xfs_btree_get_bufs( uint lock); /* lock flags for get_buf */ /* - * Allocate a new btree cursor. - * The cursor is either for allocation (A) or bmap (B). - */ -xfs_btree_cur_t * /* new btree cursor */ -xfs_btree_init_cursor( - struct xfs_mount *mp, /* file system mount point */ - struct xfs_trans *tp, /* transaction pointer */ - struct xfs_buf *agbp, /* (A only) buffer for agf structure */ - xfs_agnumber_t agno, /* (A only) allocation group number */ - xfs_btnum_t btnum, /* btree identifier */ - struct xfs_inode *ip, /* (B only) inode owning the btree */ - int whichfork); /* (B only) data/attr fork */ - -/* * Check for the cursor referring to the last block at the given level. */ int /* 1=is last block, 0=not last block */ @@ -335,15 +354,6 @@ xfs_btree_islastblock( int level); /* level to check */ /* - * Change the cursor to point to the last record in the current block - * at the given level. Other levels are unaffected. - */ -int /* success=1, failure=0 */ -xfs_btree_lastrec( - xfs_btree_cur_t *cur, /* btree cursor */ - int level); /* level to change */ - -/* * Compute first and last byte offsets for the fields given. * Interprets the offsets table, which contains struct field offsets. */ @@ -404,39 +414,53 @@ xfs_btree_reada_bufs( xfs_extlen_t count); /* count of filesystem blocks */ /* - * Read-ahead btree blocks, at the given level. - * Bits in lr are set from XFS_BTCUR_{LEFT,RIGHT}RA. + * Set the buffer for level "lev" in the cursor to bp, releasing + * any previous buffer. */ -int /* readahead block count */ -xfs_btree_readahead_core( +void +xfs_btree_setbuf( xfs_btree_cur_t *cur, /* btree cursor */ int lev, /* level in btree */ - int lr); /* left/right bits */ + struct xfs_buf *bp); /* new buffer to set */ -static inline int /* readahead block count */ -xfs_btree_readahead( - xfs_btree_cur_t *cur, /* btree cursor */ - int lev, /* level in btree */ - int lr) /* left/right bits */ -{ - if ((cur->bc_ra[lev] | lr) == cur->bc_ra[lev]) - return 0; - return xfs_btree_readahead_core(cur, lev, lr); -} +/* + * Common btree core entry points. + */ +int xfs_btree_increment(struct xfs_btree_cur *, int, int *); +int xfs_btree_decrement(struct xfs_btree_cur *, int, int *); +int xfs_btree_lookup(struct xfs_btree_cur *, xfs_lookup_t, int *); +int xfs_btree_update(struct xfs_btree_cur *, union xfs_btree_rec *); +int xfs_btree_new_iroot(struct xfs_btree_cur *, int *, int *); +int xfs_btree_kill_iroot(struct xfs_btree_cur *); +int xfs_btree_insert(struct xfs_btree_cur *, int *); +int xfs_btree_delete(struct xfs_btree_cur *, int *); +int xfs_btree_get_rec(struct xfs_btree_cur *, union xfs_btree_rec **, int *); +/* + * Internal btree helpers also used by xfs_bmap.c. + */ +void xfs_btree_log_block(struct xfs_btree_cur *, struct xfs_buf *, int); +void xfs_btree_log_recs(struct xfs_btree_cur *, struct xfs_buf *, int, int); /* - * Set the buffer for level "lev" in the cursor to bp, releasing - * any previous buffer. + * Helpers. */ -void -xfs_btree_setbuf( - xfs_btree_cur_t *cur, /* btree cursor */ - int lev, /* level in btree */ - struct xfs_buf *bp); /* new buffer to set */ +static inline int xfs_btree_get_numrecs(struct xfs_btree_block *block) +{ + return be16_to_cpu(block->bb_numrecs); +} + +static inline void xfs_btree_set_numrecs(struct xfs_btree_block *block, + __uint16_t numrecs) +{ + block->bb_numrecs = cpu_to_be16(numrecs); +} -#endif /* __KERNEL__ */ +static inline int xfs_btree_get_level(struct xfs_btree_block *block) +{ + return be16_to_cpu(block->bb_level); +} /* diff --git a/fs/xfs/xfs_btree_trace.c b/fs/xfs/xfs_btree_trace.c new file mode 100644 index 00000000000..44ff942a0fd --- /dev/null +++ b/fs/xfs/xfs_btree_trace.c @@ -0,0 +1,249 @@ +/* + * Copyright (c) 2008 Silicon Graphics, Inc. + * All Rights Reserved. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it would be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write the Free Software Foundation, + * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA + */ +#include "xfs.h" +#include "xfs_types.h" +#include "xfs_inum.h" +#include "xfs_bmap_btree.h" +#include "xfs_alloc_btree.h" +#include "xfs_ialloc_btree.h" +#include "xfs_inode.h" +#include "xfs_btree.h" +#include "xfs_btree_trace.h" + +STATIC void +xfs_btree_trace_ptr( + struct xfs_btree_cur *cur, + union xfs_btree_ptr ptr, + __psunsigned_t *high, + __psunsigned_t *low) +{ + if (cur->bc_flags & XFS_BTREE_LONG_PTRS) { + __u64 val = be64_to_cpu(ptr.l); + *high = val >> 32; + *low = (int)val; + } else { + *high = 0; + *low = be32_to_cpu(ptr.s); + } +} + +/* + * Add a trace buffer entry for arguments, for a buffer & 1 integer arg. + */ +void +xfs_btree_trace_argbi( + const char *func, + struct xfs_btree_cur *cur, + struct xfs_buf *b, + int i, + int line) +{ + cur->bc_ops->trace_enter(cur, func, XBT_ARGS, XFS_BTREE_KTRACE_ARGBI, + line, (__psunsigned_t)b, i, 0, 0, 0, 0, 0, + 0, 0, 0, 0); +} + +/* + * Add a trace buffer entry for arguments, for a buffer & 2 integer args. + */ +void +xfs_btree_trace_argbii( + const char *func, + struct xfs_btree_cur *cur, + struct xfs_buf *b, + int i0, + int i1, + int line) +{ + cur->bc_ops->trace_enter(cur, func, XBT_ARGS, XFS_BTREE_KTRACE_ARGBII, + line, (__psunsigned_t)b, i0, i1, 0, 0, 0, 0, + 0, 0, 0, 0); +} + +/* + * Add a trace buffer entry for arguments, for 3 block-length args + * and an integer arg. + */ +void +xfs_btree_trace_argfffi( + const char *func, + struct xfs_btree_cur *cur, + xfs_dfiloff_t o, + xfs_dfsbno_t b, + xfs_dfilblks_t i, + int j, + int line) +{ + cur->bc_ops->trace_enter(cur, func, XBT_ARGS, XFS_BTREE_KTRACE_ARGFFFI, + line, + o >> 32, (int)o, + b >> 32, (int)b, + i >> 32, (int)i, + (int)j, 0, 0, 0, 0); +} + +/* + * Add a trace buffer entry for arguments, for one integer arg. + */ +void +xfs_btree_trace_argi( + const char *func, + struct xfs_btree_cur *cur, + int i, + int line) +{ + cur->bc_ops->trace_enter(cur, func, XBT_ARGS, XFS_BTREE_KTRACE_ARGI, + line, i, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0); +} + +/* + * Add a trace buffer entry for arguments, for int, fsblock, key. + */ +void +xfs_btree_trace_argipk( + const char *func, + struct xfs_btree_cur *cur, + int i, + union xfs_btree_ptr ptr, + union xfs_btree_key *key, + int line) +{ + __psunsigned_t high, low; + __uint64_t l0, l1; + + xfs_btree_trace_ptr(cur, ptr, &high, &low); + cur->bc_ops->trace_key(cur, key, &l0, &l1); + cur->bc_ops->trace_enter(cur, func, XBT_ARGS, XFS_BTREE_KTRACE_ARGIPK, + line, i, high, low, + l0 >> 32, (int)l0, + l1 >> 32, (int)l1, + 0, 0, 0, 0); +} + +/* + * Add a trace buffer entry for arguments, for int, fsblock, rec. + */ +void +xfs_btree_trace_argipr( + const char *func, + struct xfs_btree_cur *cur, + int i, + union xfs_btree_ptr ptr, + union xfs_btree_rec *rec, + int line) +{ + __psunsigned_t high, low; + __uint64_t l0, l1, l2; + + xfs_btree_trace_ptr(cur, ptr, &high, &low); + cur->bc_ops->trace_record(cur, rec, &l0, &l1, &l2); + cur->bc_ops->trace_enter(cur, func, XBT_ARGS, XFS_BTREE_KTRACE_ARGIPR, + line, i, + high, low, + l0 >> 32, (int)l0, + l1 >> 32, (int)l1, + l2 >> 32, (int)l2, + 0, 0); +} + +/* + * Add a trace buffer entry for arguments, for int, key. + */ +void +xfs_btree_trace_argik( + const char *func, + struct xfs_btree_cur *cur, + int i, + union xfs_btree_key *key, + int line) +{ + __uint64_t l0, l1; + + cur->bc_ops->trace_key(cur, key, &l0, &l1); + cur->bc_ops->trace_enter(cur, func, XBT_ARGS, XFS_BTREE_KTRACE_ARGIK, + line, i, + l0 >> 32, (int)l0, + l1 >> 32, (int)l1, + 0, 0, 0, 0, 0, 0); +} + +/* + * Add a trace buffer entry for arguments, for record. + */ +void +xfs_btree_trace_argr( + const char *func, + struct xfs_btree_cur *cur, + union xfs_btree_rec *rec, + int line) +{ + __uint64_t l0, l1, l2; + + cur->bc_ops->trace_record(cur, rec, &l0, &l1, &l2); + cur->bc_ops->trace_enter(cur, func, XBT_ARGS, XFS_BTREE_KTRACE_ARGR, + line, + l0 >> 32, (int)l0, + l1 >> 32, (int)l1, + l2 >> 32, (int)l2, + 0, 0, 0, 0, 0); +} + +/* + * Add a trace buffer entry for the cursor/operation. + */ +void +xfs_btree_trace_cursor( + const char *func, + struct xfs_btree_cur *cur, + int type, + int line) +{ + __uint32_t s0; + __uint64_t l0, l1; + char *s; + + switch (type) { + case XBT_ARGS: + s = "args"; + break; + case XBT_ENTRY: + s = "entry"; + break; + case XBT_ERROR: + s = "error"; + break; + case XBT_EXIT: + s = "exit"; + break; + default: + s = "unknown"; + break; + } + + cur->bc_ops->trace_cursor(cur, &s0, &l0, &l1); + cur->bc_ops->trace_enter(cur, func, s, XFS_BTREE_KTRACE_CUR, line, + s0, + l0 >> 32, (int)l0, + l1 >> 32, (int)l1, + (__psunsigned_t)cur->bc_bufs[0], + (__psunsigned_t)cur->bc_bufs[1], + (__psunsigned_t)cur->bc_bufs[2], + (__psunsigned_t)cur->bc_bufs[3], + (cur->bc_ptrs[0] << 16) | cur->bc_ptrs[1], + (cur->bc_ptrs[2] << 16) | cur->bc_ptrs[3]); +} diff --git a/fs/xfs/xfs_btree_trace.h b/fs/xfs/xfs_btree_trace.h new file mode 100644 index 00000000000..b3f5eb3c3c6 --- /dev/null +++ b/fs/xfs/xfs_btree_trace.h @@ -0,0 +1,116 @@ +/* + * Copyright (c) 2008 Silicon Graphics, Inc. + * All Rights Reserved. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it would be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write the Free Software Foundation, + * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA + */ +#ifndef __XFS_BTREE_TRACE_H__ +#define __XFS_BTREE_TRACE_H__ + +struct xfs_btree_cur; +struct xfs_buf; + + +/* + * Trace hooks. + * i,j = integer (32 bit) + * b = btree block buffer (xfs_buf_t) + * p = btree ptr + * r = btree record + * k = btree key + */ + +#ifdef XFS_BTREE_TRACE + +/* + * Trace buffer entry types. + */ +#define XFS_BTREE_KTRACE_ARGBI 1 +#define XFS_BTREE_KTRACE_ARGBII 2 +#define XFS_BTREE_KTRACE_ARGFFFI 3 +#define XFS_BTREE_KTRACE_ARGI 4 +#define XFS_BTREE_KTRACE_ARGIPK 5 +#define XFS_BTREE_KTRACE_ARGIPR 6 +#define XFS_BTREE_KTRACE_ARGIK 7 +#define XFS_BTREE_KTRACE_ARGR 8 +#define XFS_BTREE_KTRACE_CUR 9 + +/* + * Sub-types for cursor traces. + */ +#define XBT_ARGS 0 +#define XBT_ENTRY 1 +#define XBT_ERROR 2 +#define XBT_EXIT 3 + +void xfs_btree_trace_argbi(const char *, struct xfs_btree_cur *, + struct xfs_buf *, int, int); +void xfs_btree_trace_argbii(const char *, struct xfs_btree_cur *, + struct xfs_buf *, int, int, int); +void xfs_btree_trace_argfffi(const char *, struct xfs_btree_cur *, + xfs_dfiloff_t, xfs_dfsbno_t, xfs_dfilblks_t, int, int); +void xfs_btree_trace_argi(const char *, struct xfs_btree_cur *, int, int); +void xfs_btree_trace_argipk(const char *, struct xfs_btree_cur *, int, + union xfs_btree_ptr, union xfs_btree_key *, int); +void xfs_btree_trace_argipr(const char *, struct xfs_btree_cur *, int, + union xfs_btree_ptr, union xfs_btree_rec *, int); +void xfs_btree_trace_argik(const char *, struct xfs_btree_cur *, int, + union xfs_btree_key *, int); +void xfs_btree_trace_argr(const char *, struct xfs_btree_cur *, + union xfs_btree_rec *, int); +void xfs_btree_trace_cursor(const char *, struct xfs_btree_cur *, int, int); + + +#define XFS_ALLOCBT_TRACE_SIZE 4096 /* size of global trace buffer */ +extern ktrace_t *xfs_allocbt_trace_buf; + +#define XFS_INOBT_TRACE_SIZE 4096 /* size of global trace buffer */ +extern ktrace_t *xfs_inobt_trace_buf; + +#define XFS_BMBT_TRACE_SIZE 4096 /* size of global trace buffer */ +#define XFS_BMBT_KTRACE_SIZE 32 /* size of per-inode trace buffer */ +extern ktrace_t *xfs_bmbt_trace_buf; + + +#define XFS_BTREE_TRACE_ARGBI(c, b, i) \ + xfs_btree_trace_argbi(__func__, c, b, i, __LINE__) +#define XFS_BTREE_TRACE_ARGBII(c, b, i, j) \ + xfs_btree_trace_argbii(__func__, c, b, i, j, __LINE__) +#define XFS_BTREE_TRACE_ARGFFFI(c, o, b, i, j) \ + xfs_btree_trace_argfffi(__func__, c, o, b, i, j, __LINE__) +#define XFS_BTREE_TRACE_ARGI(c, i) \ + xfs_btree_trace_argi(__func__, c, i, __LINE__) +#define XFS_BTREE_TRACE_ARGIPK(c, i, p, k) \ + xfs_btree_trace_argipk(__func__, c, i, p, k, __LINE__) +#define XFS_BTREE_TRACE_ARGIPR(c, i, p, r) \ + xfs_btree_trace_argipr(__func__, c, i, p, r, __LINE__) +#define XFS_BTREE_TRACE_ARGIK(c, i, k) \ + xfs_btree_trace_argik(__func__, c, i, k, __LINE__) +#define XFS_BTREE_TRACE_ARGR(c, r) \ + xfs_btree_trace_argr(__func__, c, r, __LINE__) +#define XFS_BTREE_TRACE_CURSOR(c, t) \ + xfs_btree_trace_cursor(__func__, c, t, __LINE__) +#else +#define XFS_BTREE_TRACE_ARGBI(c, b, i) +#define XFS_BTREE_TRACE_ARGBII(c, b, i, j) +#define XFS_BTREE_TRACE_ARGFFFI(c, o, b, i, j) +#define XFS_BTREE_TRACE_ARGI(c, i) +#define XFS_BTREE_TRACE_ARGIPK(c, i, p, s) +#define XFS_BTREE_TRACE_ARGIPR(c, i, p, r) +#define XFS_BTREE_TRACE_ARGIK(c, i, k) +#define XFS_BTREE_TRACE_ARGR(c, r) +#define XFS_BTREE_TRACE_CURSOR(c, t) +#endif /* XFS_BTREE_TRACE */ + +#endif /* __XFS_BTREE_TRACE_H__ */ diff --git a/fs/xfs/xfs_buf_item.c b/fs/xfs/xfs_buf_item.c index 002fc2617c8..92af4098c7e 100644 --- a/fs/xfs/xfs_buf_item.c +++ b/fs/xfs/xfs_buf_item.c @@ -375,7 +375,7 @@ xfs_buf_item_unpin( xfs_buf_log_item_t *bip, int stale) { - xfs_mount_t *mp; + struct xfs_ail *ailp; xfs_buf_t *bp; int freed; @@ -387,7 +387,7 @@ xfs_buf_item_unpin( xfs_buftrace("XFS_UNPIN", bp); freed = atomic_dec_and_test(&bip->bli_refcount); - mp = bip->bli_item.li_mountp; + ailp = bip->bli_item.li_ailp; xfs_bunpin(bp); if (freed && stale) { ASSERT(bip->bli_flags & XFS_BLI_STALE); @@ -399,17 +399,17 @@ xfs_buf_item_unpin( xfs_buftrace("XFS_UNPIN STALE", bp); /* * If we get called here because of an IO error, we may - * or may not have the item on the AIL. xfs_trans_delete_ail() + * or may not have the item on the AIL. xfs_trans_ail_delete() * will take care of that situation. - * xfs_trans_delete_ail() drops the AIL lock. + * xfs_trans_ail_delete() drops the AIL lock. */ if (bip->bli_flags & XFS_BLI_STALE_INODE) { xfs_buf_do_callbacks(bp, (xfs_log_item_t *)bip); XFS_BUF_SET_FSPRIVATE(bp, NULL); XFS_BUF_CLR_IODONE_FUNC(bp); } else { - spin_lock(&mp->m_ail_lock); - xfs_trans_delete_ail(mp, (xfs_log_item_t *)bip); + spin_lock(&ailp->xa_lock); + xfs_trans_ail_delete(ailp, (xfs_log_item_t *)bip); xfs_buf_item_relse(bp); ASSERT(XFS_BUF_FSPRIVATE(bp, void *) == NULL); } @@ -707,8 +707,8 @@ xfs_buf_item_init( * the first. If we do already have one, there is * nothing to do here so return. */ - if (XFS_BUF_FSPRIVATE3(bp, xfs_mount_t *) != mp) - XFS_BUF_SET_FSPRIVATE3(bp, mp); + if (bp->b_mount != mp) + bp->b_mount = mp; XFS_BUF_SET_BDSTRAT_FUNC(bp, xfs_bdstrat_cb); if (XFS_BUF_FSPRIVATE(bp, void *) != NULL) { lip = XFS_BUF_FSPRIVATE(bp, xfs_log_item_t *); @@ -731,6 +731,7 @@ xfs_buf_item_init( bip->bli_item.li_type = XFS_LI_BUF; bip->bli_item.li_ops = &xfs_buf_item_ops; bip->bli_item.li_mountp = mp; + bip->bli_item.li_ailp = mp->m_ail; bip->bli_buf = bp; xfs_buf_hold(bp); bip->bli_format.blf_type = XFS_LI_BUF; @@ -997,21 +998,7 @@ xfs_buf_iodone_callbacks( xfs_buf_do_callbacks(bp, lip); XFS_BUF_SET_FSPRIVATE(bp, NULL); XFS_BUF_CLR_IODONE_FUNC(bp); - - /* - * XFS_SHUT flag gets set when we go thru the - * entire buffer cache and deliberately start - * throwing away delayed write buffers. - * Since there's no biowait done on those, - * we should just brelse them. - */ - if (XFS_BUF_ISSHUT(bp)) { - XFS_BUF_UNSHUT(bp); - xfs_buf_relse(bp); - } else { - xfs_biodone(bp); - } - + xfs_biodone(bp); return; } @@ -1122,27 +1109,23 @@ xfs_buf_iodone( xfs_buf_t *bp, xfs_buf_log_item_t *bip) { - struct xfs_mount *mp; + struct xfs_ail *ailp = bip->bli_item.li_ailp; ASSERT(bip->bli_buf == bp); xfs_buf_rele(bp); - mp = bip->bli_item.li_mountp; /* * If we are forcibly shutting down, this may well be * off the AIL already. That's because we simulate the * log-committed callbacks to unpin these buffers. Or we may never * have put this item on AIL because of the transaction was - * aborted forcibly. xfs_trans_delete_ail() takes care of these. + * aborted forcibly. xfs_trans_ail_delete() takes care of these. * * Either way, AIL is useless if we're forcing a shutdown. */ - spin_lock(&mp->m_ail_lock); - /* - * xfs_trans_delete_ail() drops the AIL lock. - */ - xfs_trans_delete_ail(mp, (xfs_log_item_t *)bip); + spin_lock(&ailp->xa_lock); + xfs_trans_ail_delete(ailp, (xfs_log_item_t *)bip); xfs_buf_item_free(bip); } diff --git a/fs/xfs/xfs_clnt.h b/fs/xfs/xfs_clnt.h deleted file mode 100644 index d2ce5dd70d8..00000000000 --- a/fs/xfs/xfs_clnt.h +++ /dev/null @@ -1,105 +0,0 @@ -/* - * Copyright (c) 2000-2005 Silicon Graphics, Inc. - * All Rights Reserved. - * - * This program is free software; you can redistribute it and/or - * modify it under the terms of the GNU General Public License as - * published by the Free Software Foundation. - * - * This program is distributed in the hope that it would be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write the Free Software Foundation, - * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA - */ -#ifndef __XFS_CLNT_H__ -#define __XFS_CLNT_H__ - -/* - * XFS arguments structure, constructed from the arguments we - * are passed via the mount system call. - * - * NOTE: The mount system call is handled differently between - * Linux and IRIX. In IRIX we worked work with a binary data - * structure coming in across the syscall interface from user - * space (the mount userspace knows about each filesystem type - * and the set of valid options for it, and converts the users - * argument string into a binary structure _before_ making the - * system call), and the ABI issues that this implies. - * - * In Linux, we are passed a comma separated set of options; - * ie. a NULL terminated string of characters. Userspace mount - * code does not have any knowledge of mount options expected by - * each filesystem type and so each filesystem parses its mount - * options in kernel space. - * - * For the Linux port, we kept this structure pretty much intact - * and use it internally (because the existing code groks it). - */ -struct xfs_mount_args { - int flags; /* flags -> see XFSMNT_... macros below */ - int flags2; /* flags -> see XFSMNT2_... macros below */ - int logbufs; /* Number of log buffers, -1 to default */ - int logbufsize; /* Size of log buffers, -1 to default */ - char fsname[MAXNAMELEN+1]; /* data device name */ - char rtname[MAXNAMELEN+1]; /* realtime device filename */ - char logname[MAXNAMELEN+1]; /* journal device filename */ - char mtpt[MAXNAMELEN+1]; /* filesystem mount point */ - int sunit; /* stripe unit (BBs) */ - int swidth; /* stripe width (BBs), multiple of sunit */ - uchar_t iosizelog; /* log2 of the preferred I/O size */ - int ihashsize; /* inode hash table size (buckets) */ -}; - -/* - * XFS mount option flags -- args->flags1 - */ -#define XFSMNT_ATTR2 0x00000001 /* allow ATTR2 EA format */ -#define XFSMNT_WSYNC 0x00000002 /* safe mode nfs mount - * compatible */ -#define XFSMNT_INO64 0x00000004 /* move inode numbers up - * past 2^32 */ -#define XFSMNT_UQUOTA 0x00000008 /* user quota accounting */ -#define XFSMNT_PQUOTA 0x00000010 /* IRIX prj quota accounting */ -#define XFSMNT_UQUOTAENF 0x00000020 /* user quota limit - * enforcement */ -#define XFSMNT_PQUOTAENF 0x00000040 /* IRIX project quota limit - * enforcement */ -#define XFSMNT_QUIET 0x00000080 /* don't report mount errors */ -#define XFSMNT_NOALIGN 0x00000200 /* don't allocate at - * stripe boundaries*/ -#define XFSMNT_RETERR 0x00000400 /* return error to user */ -#define XFSMNT_NORECOVERY 0x00000800 /* no recovery, implies - * read-only mount */ -#define XFSMNT_SHARED 0x00001000 /* shared XFS mount */ -#define XFSMNT_IOSIZE 0x00002000 /* optimize for I/O size */ -#define XFSMNT_OSYNCISOSYNC 0x00004000 /* o_sync is REALLY o_sync */ - /* (osyncisdsync is default) */ -#define XFSMNT_NOATTR2 0x00008000 /* turn off ATTR2 EA format */ -#define XFSMNT_32BITINODES 0x00200000 /* restrict inodes to 32 - * bits of address space */ -#define XFSMNT_GQUOTA 0x00400000 /* group quota accounting */ -#define XFSMNT_GQUOTAENF 0x00800000 /* group quota limit - * enforcement */ -#define XFSMNT_NOUUID 0x01000000 /* Ignore fs uuid */ -#define XFSMNT_DMAPI 0x02000000 /* enable dmapi/xdsm */ -#define XFSMNT_BARRIER 0x04000000 /* use write barriers */ -#define XFSMNT_IKEEP 0x08000000 /* inode cluster delete */ -#define XFSMNT_SWALLOC 0x10000000 /* turn on stripe width - * allocation */ -#define XFSMNT_DIRSYNC 0x40000000 /* sync creat,link,unlink,rename - * symlink,mkdir,rmdir,mknod */ -#define XFSMNT_FLAGS2 0x80000000 /* more flags set in flags2 */ - -/* - * XFS mount option flags -- args->flags2 - */ -#define XFSMNT2_COMPAT_IOSIZE 0x00000001 /* don't report large preferred - * I/O size in stat(2) */ -#define XFSMNT2_FILESTREAMS 0x00000002 /* enable the filestreams - * allocator */ - -#endif /* __XFS_CLNT_H__ */ diff --git a/fs/xfs/xfs_da_btree.c b/fs/xfs/xfs_da_btree.c index 9e561a9cefc..a11a8390bf6 100644 --- a/fs/xfs/xfs_da_btree.c +++ b/fs/xfs/xfs_da_btree.c @@ -1566,11 +1566,14 @@ xfs_da_grow_inode(xfs_da_args_t *args, xfs_dablk_t *new_blkno) int nmap, error, w, count, c, got, i, mapi; xfs_trans_t *tp; xfs_mount_t *mp; + xfs_drfsbno_t nblks; dp = args->dp; mp = dp->i_mount; w = args->whichfork; tp = args->trans; + nblks = dp->i_d.di_nblocks; + /* * For new directories adjust the file offset and block count. */ @@ -1647,6 +1650,8 @@ xfs_da_grow_inode(xfs_da_args_t *args, xfs_dablk_t *new_blkno) } if (mapp != &map) kmem_free(mapp); + /* account for newly allocated blocks in reserved blocks total */ + args->total -= dp->i_d.di_nblocks - nblks; *new_blkno = (xfs_dablk_t)bno; return 0; } diff --git a/fs/xfs/xfs_da_btree.h b/fs/xfs/xfs_da_btree.h index 8be0b00ede9..70b710c1792 100644 --- a/fs/xfs/xfs_da_btree.h +++ b/fs/xfs/xfs_da_btree.h @@ -72,27 +72,7 @@ typedef struct xfs_da_intnode { typedef struct xfs_da_node_hdr xfs_da_node_hdr_t; typedef struct xfs_da_node_entry xfs_da_node_entry_t; -#define XFS_DA_MAXHASH ((xfs_dahash_t)-1) /* largest valid hash value */ - #define XFS_LBSIZE(mp) (mp)->m_sb.sb_blocksize -#define XFS_LBLOG(mp) (mp)->m_sb.sb_blocklog - -#define XFS_DA_MAKE_BNOENTRY(mp,bno,entry) \ - (((bno) << (mp)->m_dircook_elog) | (entry)) -#define XFS_DA_MAKE_COOKIE(mp,bno,entry,hash) \ - (((xfs_off_t)XFS_DA_MAKE_BNOENTRY(mp, bno, entry) << 32) | (hash)) -#define XFS_DA_COOKIE_HASH(mp,cookie) ((xfs_dahash_t)cookie) -#define XFS_DA_COOKIE_BNO(mp,cookie) \ - ((((xfs_off_t)(cookie) >> 31) == -1LL ? \ - (xfs_dablk_t)0 : \ - (xfs_dablk_t)((xfs_off_t)(cookie) >> \ - ((mp)->m_dircook_elog + 32)))) -#define XFS_DA_COOKIE_ENTRY(mp,cookie) \ - ((((xfs_off_t)(cookie) >> 31) == -1LL ? \ - (xfs_dablk_t)0 : \ - (xfs_dablk_t)(((xfs_off_t)(cookie) >> 32) & \ - ((1 << (mp)->m_dircook_elog) - 1)))) - /*======================================================================== * Btree searching and modification structure definitions. @@ -226,9 +206,8 @@ struct xfs_nameops { }; -#ifdef __KERNEL__ /*======================================================================== - * Function prototypes for the kernel. + * Function prototypes. *========================================================================*/ /* @@ -289,6 +268,5 @@ xfs_daddr_t xfs_da_blkno(xfs_dabuf_t *dabuf); extern struct kmem_zone *xfs_da_state_zone; extern struct kmem_zone *xfs_dabuf_zone; -#endif /* __KERNEL__ */ #endif /* __XFS_DA_BTREE_H__ */ diff --git a/fs/xfs/xfs_dfrag.c b/fs/xfs/xfs_dfrag.c index 75b0cd4da0e..b4c1ee71349 100644 --- a/fs/xfs/xfs_dfrag.c +++ b/fs/xfs/xfs_dfrag.c @@ -49,9 +49,8 @@ */ int xfs_swapext( - xfs_swapext_t __user *sxu) + xfs_swapext_t *sxp) { - xfs_swapext_t *sxp; xfs_inode_t *ip, *tip; struct file *file, *target_file; int error = 0; @@ -62,11 +61,6 @@ xfs_swapext( goto out; } - if (copy_from_user(sxp, sxu, sizeof(xfs_swapext_t))) { - error = XFS_ERROR(EFAULT); - goto out_free_sxp; - } - /* Pull information for the target fd */ file = fget((int)sxp->sx_fdtarget); if (!file) { diff --git a/fs/xfs/xfs_dfrag.h b/fs/xfs/xfs_dfrag.h index da178205be6..4f55a630655 100644 --- a/fs/xfs/xfs_dfrag.h +++ b/fs/xfs/xfs_dfrag.h @@ -46,7 +46,7 @@ typedef struct xfs_swapext /* * Syscall interface for xfs_swapext */ -int xfs_swapext(struct xfs_swapext __user *sx); +int xfs_swapext(struct xfs_swapext *sx); int xfs_swap_extents(struct xfs_inode *ip, struct xfs_inode *tip, struct xfs_swapext *sxp); diff --git a/fs/xfs/xfs_dinode.h b/fs/xfs/xfs_dinode.h index c9065eaf2a4..162e8726df5 100644 --- a/fs/xfs/xfs_dinode.h +++ b/fs/xfs/xfs_dinode.h @@ -18,32 +18,29 @@ #ifndef __XFS_DINODE_H__ #define __XFS_DINODE_H__ -struct xfs_buf; -struct xfs_mount; +#define XFS_DINODE_MAGIC 0x494e /* 'IN' */ +#define XFS_DINODE_GOOD_VERSION(v) (((v) == 1 || (v) == 2)) -#define XFS_DINODE_VERSION_1 1 -#define XFS_DINODE_VERSION_2 2 -#define XFS_DINODE_GOOD_VERSION(v) \ - (((v) == XFS_DINODE_VERSION_1 || (v) == XFS_DINODE_VERSION_2)) -#define XFS_DINODE_MAGIC 0x494e /* 'IN' */ - -/* - * Disk inode structure. - * This is just the header; the inode is expanded to fill a variable size - * with the last field expanding. It is split into the core and "other" - * because we only need the core part in the in-core inode. - */ typedef struct xfs_timestamp { __be32 t_sec; /* timestamp seconds */ __be32 t_nsec; /* timestamp nanoseconds */ } xfs_timestamp_t; /* - * Note: Coordinate changes to this structure with the XFS_DI_* #defines - * below, the offsets table in xfs_ialloc_log_di() and struct xfs_icdinode - * in xfs_inode.h. + * On-disk inode structure. + * + * This is just the header or "dinode core", the inode is expanded to fill a + * variable size the leftover area split into a data and an attribute fork. + * The format of the data and attribute fork depends on the format of the + * inode as indicated by di_format and di_aformat. To access the data and + * attribute use the XFS_DFORK_PTR, XFS_DFORK_DPTR, and XFS_DFORK_PTR macros + * below. + * + * There is a very similar struct icdinode in xfs_inode which matches the + * layout of the first 96 bytes of this structure, but is kept in native + * format instead of big endian. */ -typedef struct xfs_dinode_core { +typedef struct xfs_dinode { __be16 di_magic; /* inode magic # = XFS_DINODE_MAGIC */ __be16 di_mode; /* mode and type of file */ __u8 di_version; /* inode version */ @@ -69,34 +66,12 @@ typedef struct xfs_dinode_core { __be16 di_dmstate; /* DMIG state info */ __be16 di_flags; /* random flags, XFS_DIFLAG_... */ __be32 di_gen; /* generation number */ -} xfs_dinode_core_t; -#define DI_MAX_FLUSH 0xffff + /* di_next_unlinked is the only non-core field in the old dinode */ + __be32 di_next_unlinked;/* agi unlinked list ptr */ +} __attribute__((packed)) xfs_dinode_t; -typedef struct xfs_dinode -{ - xfs_dinode_core_t di_core; - /* - * In adding anything between the core and the union, be - * sure to update the macros like XFS_LITINO below and - * XFS_BMAP_RBLOCK_DSIZE in xfs_bmap_btree.h. - */ - __be32 di_next_unlinked;/* agi unlinked list ptr */ - union { - xfs_bmdr_block_t di_bmbt; /* btree root block */ - xfs_bmbt_rec_32_t di_bmx[1]; /* extent list */ - xfs_dir2_sf_t di_dir2sf; /* shortform directory v2 */ - char di_c[1]; /* local contents */ - __be32 di_dev; /* device for S_IFCHR/S_IFBLK */ - uuid_t di_muuid; /* mount point value */ - char di_symlink[1]; /* local symbolic link */ - } di_u; - union { - xfs_bmdr_block_t di_abmbt; /* btree root block */ - xfs_bmbt_rec_32_t di_abmx[1]; /* extent list */ - xfs_attr_shortform_t di_attrsf; /* shortform attribute list */ - } di_a; -} xfs_dinode_t; +#define DI_MAX_FLUSH 0xffff /* * The 32 bit link count in the inode theoretically maxes out at UINT_MAX. @@ -107,50 +82,14 @@ typedef struct xfs_dinode #define XFS_MAXLINK_1 65535U /* - * Bit names for logging disk inodes only - */ -#define XFS_DI_MAGIC 0x0000001 -#define XFS_DI_MODE 0x0000002 -#define XFS_DI_VERSION 0x0000004 -#define XFS_DI_FORMAT 0x0000008 -#define XFS_DI_ONLINK 0x0000010 -#define XFS_DI_UID 0x0000020 -#define XFS_DI_GID 0x0000040 -#define XFS_DI_NLINK 0x0000080 -#define XFS_DI_PROJID 0x0000100 -#define XFS_DI_PAD 0x0000200 -#define XFS_DI_ATIME 0x0000400 -#define XFS_DI_MTIME 0x0000800 -#define XFS_DI_CTIME 0x0001000 -#define XFS_DI_SIZE 0x0002000 -#define XFS_DI_NBLOCKS 0x0004000 -#define XFS_DI_EXTSIZE 0x0008000 -#define XFS_DI_NEXTENTS 0x0010000 -#define XFS_DI_NAEXTENTS 0x0020000 -#define XFS_DI_FORKOFF 0x0040000 -#define XFS_DI_AFORMAT 0x0080000 -#define XFS_DI_DMEVMASK 0x0100000 -#define XFS_DI_DMSTATE 0x0200000 -#define XFS_DI_FLAGS 0x0400000 -#define XFS_DI_GEN 0x0800000 -#define XFS_DI_NEXT_UNLINKED 0x1000000 -#define XFS_DI_U 0x2000000 -#define XFS_DI_A 0x4000000 -#define XFS_DI_NUM_BITS 27 -#define XFS_DI_ALL_BITS ((1 << XFS_DI_NUM_BITS) - 1) -#define XFS_DI_CORE_BITS (XFS_DI_ALL_BITS & ~(XFS_DI_U|XFS_DI_A)) - -/* * Values for di_format */ -typedef enum xfs_dinode_fmt -{ - XFS_DINODE_FMT_DEV, /* CHR, BLK: di_dev */ - XFS_DINODE_FMT_LOCAL, /* DIR, REG: di_c */ - /* LNK: di_symlink */ - XFS_DINODE_FMT_EXTENTS, /* DIR, REG, LNK: di_bmx */ - XFS_DINODE_FMT_BTREE, /* DIR, REG, LNK: di_bmbt */ - XFS_DINODE_FMT_UUID /* MNT: di_uuid */ +typedef enum xfs_dinode_fmt { + XFS_DINODE_FMT_DEV, /* xfs_dev_t */ + XFS_DINODE_FMT_LOCAL, /* bulk data */ + XFS_DINODE_FMT_EXTENTS, /* struct xfs_bmbt_rec */ + XFS_DINODE_FMT_BTREE, /* struct xfs_bmdr_block */ + XFS_DINODE_FMT_UUID /* uuid_t */ } xfs_dinode_fmt_t; /* @@ -166,13 +105,13 @@ typedef enum xfs_dinode_fmt */ #define XFS_LITINO(mp) ((mp)->m_litino) #define XFS_BROOT_SIZE_ADJ \ - (sizeof(xfs_bmbt_block_t) - sizeof(xfs_bmdr_block_t)) + (XFS_BTREE_LBLOCK_LEN - sizeof(xfs_bmdr_block_t)) /* * Inode data & attribute fork sizes, per inode. */ -#define XFS_DFORK_Q(dip) ((dip)->di_core.di_forkoff != 0) -#define XFS_DFORK_BOFF(dip) ((int)((dip)->di_core.di_forkoff << 3)) +#define XFS_DFORK_Q(dip) ((dip)->di_forkoff != 0) +#define XFS_DFORK_BOFF(dip) ((int)((dip)->di_forkoff << 3)) #define XFS_DFORK_DSIZE(dip,mp) \ (XFS_DFORK_Q(dip) ? \ @@ -187,23 +126,42 @@ typedef enum xfs_dinode_fmt XFS_DFORK_DSIZE(dip, mp) : \ XFS_DFORK_ASIZE(dip, mp)) -#define XFS_DFORK_DPTR(dip) ((dip)->di_u.di_c) +/* + * Return pointers to the data or attribute forks. + */ +#define XFS_DFORK_DPTR(dip) \ + ((char *)(dip) + sizeof(struct xfs_dinode)) #define XFS_DFORK_APTR(dip) \ - ((dip)->di_u.di_c + XFS_DFORK_BOFF(dip)) + (XFS_DFORK_DPTR(dip) + XFS_DFORK_BOFF(dip)) #define XFS_DFORK_PTR(dip,w) \ ((w) == XFS_DATA_FORK ? XFS_DFORK_DPTR(dip) : XFS_DFORK_APTR(dip)) + #define XFS_DFORK_FORMAT(dip,w) \ ((w) == XFS_DATA_FORK ? \ - (dip)->di_core.di_format : \ - (dip)->di_core.di_aformat) + (dip)->di_format : \ + (dip)->di_aformat) #define XFS_DFORK_NEXTENTS(dip,w) \ ((w) == XFS_DATA_FORK ? \ - be32_to_cpu((dip)->di_core.di_nextents) : \ - be16_to_cpu((dip)->di_core.di_anextents)) + be32_to_cpu((dip)->di_nextents) : \ + be16_to_cpu((dip)->di_anextents)) #define XFS_BUF_TO_DINODE(bp) ((xfs_dinode_t *)XFS_BUF_PTR(bp)) /* + * For block and character special files the 32bit dev_t is stored at the + * beginning of the data fork. + */ +static inline xfs_dev_t xfs_dinode_get_rdev(struct xfs_dinode *dip) +{ + return be32_to_cpu(*(__be32 *)XFS_DFORK_DPTR(dip)); +} + +static inline void xfs_dinode_put_rdev(struct xfs_dinode *dip, xfs_dev_t rdev) +{ + *(__be32 *)XFS_DFORK_DPTR(dip) = cpu_to_be32(rdev); +} + +/* * Values for di_flags * There should be a one-to-one correspondence between these flags and the * XFS_XFLAG_s. diff --git a/fs/xfs/xfs_dir2.c b/fs/xfs/xfs_dir2.c index 80e0dc51361..1afb12278b8 100644 --- a/fs/xfs/xfs_dir2.c +++ b/fs/xfs/xfs_dir2.c @@ -525,11 +525,13 @@ xfs_dir2_grow_inode( xfs_mount_t *mp; int nmap; /* number of bmap entries */ xfs_trans_t *tp; + xfs_drfsbno_t nblks; xfs_dir2_trace_args_s("grow_inode", args, space); dp = args->dp; tp = args->trans; mp = dp->i_mount; + nblks = dp->i_d.di_nblocks; /* * Set lowest possible block in the space requested. */ @@ -622,7 +624,11 @@ xfs_dir2_grow_inode( */ if (mapp != &map) kmem_free(mapp); + + /* account for newly allocated blocks in reserved blocks total */ + args->total -= dp->i_d.di_nblocks - nblks; *dbp = xfs_dir2_da_to_db(mp, (xfs_dablk_t)bno); + /* * Update file's size if this is the data space and it grew. */ diff --git a/fs/xfs/xfs_dir2_sf.h b/fs/xfs/xfs_dir2_sf.h index deecc9d238f..6ac44b550d3 100644 --- a/fs/xfs/xfs_dir2_sf.h +++ b/fs/xfs/xfs_dir2_sf.h @@ -34,13 +34,6 @@ struct xfs_mount; struct xfs_trans; /* - * Maximum size of a shortform directory. - */ -#define XFS_DIR2_SF_MAX_SIZE \ - (XFS_DINODE_MAX_SIZE - (uint)sizeof(xfs_dinode_core_t) - \ - (uint)sizeof(xfs_agino_t)) - -/* * Inode number stored as 8 8-bit values. */ typedef struct { __uint8_t i[8]; } xfs_dir2_ino8_t; diff --git a/fs/xfs/xfs_dmops.c b/fs/xfs/xfs_dmops.c index a1e55fb9d5d..e71e2581c0c 100644 --- a/fs/xfs/xfs_dmops.c +++ b/fs/xfs/xfs_dmops.c @@ -25,7 +25,6 @@ #include "xfs_inum.h" #include "xfs_ag.h" #include "xfs_mount.h" -#include "xfs_clnt.h" static struct xfs_dmops xfs_dmcore_stub = { @@ -38,9 +37,9 @@ static struct xfs_dmops xfs_dmcore_stub = { }; int -xfs_dmops_get(struct xfs_mount *mp, struct xfs_mount_args *args) +xfs_dmops_get(struct xfs_mount *mp) { - if (args->flags & XFSMNT_DMAPI) { + if (mp->m_flags & XFS_MOUNT_DMAPI) { cmn_err(CE_WARN, "XFS: dmapi support not available in this kernel."); return EINVAL; diff --git a/fs/xfs/xfs_error.c b/fs/xfs/xfs_error.c index f227ecd1a29..92d5cd5bf4f 100644 --- a/fs/xfs/xfs_error.c +++ b/fs/xfs/xfs_error.c @@ -153,21 +153,6 @@ xfs_errortag_clearall(xfs_mount_t *mp, int loud) } #endif /* DEBUG */ -static void -xfs_fs_vcmn_err(int level, xfs_mount_t *mp, char *fmt, va_list ap) -{ - if (mp != NULL) { - char *newfmt; - int len = 16 + mp->m_fsname_len + strlen(fmt); - - newfmt = kmem_alloc(len, KM_SLEEP); - sprintf(newfmt, "Filesystem \"%s\": %s", mp->m_fsname, fmt); - icmn_err(level, newfmt, ap); - kmem_free(newfmt); - } else { - icmn_err(level, fmt, ap); - } -} void xfs_fs_cmn_err(int level, xfs_mount_t *mp, char *fmt, ...) diff --git a/fs/xfs/xfs_error.h b/fs/xfs/xfs_error.h index 11543f10b0c..0c93051c465 100644 --- a/fs/xfs/xfs_error.h +++ b/fs/xfs/xfs_error.h @@ -159,11 +159,15 @@ extern int xfs_errortag_clearall(xfs_mount_t *mp, int loud); #define XFS_PTAG_FSBLOCK_ZERO 0x00000080 struct xfs_mount; -/* PRINTFLIKE4 */ + +extern void xfs_fs_vcmn_err(int level, struct xfs_mount *mp, + char *fmt, va_list ap) + __attribute__ ((format (printf, 3, 0))); extern void xfs_cmn_err(int panic_tag, int level, struct xfs_mount *mp, - char *fmt, ...); -/* PRINTFLIKE3 */ -extern void xfs_fs_cmn_err(int level, struct xfs_mount *mp, char *fmt, ...); + char *fmt, ...) + __attribute__ ((format (printf, 4, 5))); +extern void xfs_fs_cmn_err(int level, struct xfs_mount *mp, char *fmt, ...) + __attribute__ ((format (printf, 3, 4))); extern void xfs_hex_dump(void *p, int length); diff --git a/fs/xfs/xfs_extfree_item.c b/fs/xfs/xfs_extfree_item.c index 8aa28f751b2..05a4bdd4be3 100644 --- a/fs/xfs/xfs_extfree_item.c +++ b/fs/xfs/xfs_extfree_item.c @@ -108,19 +108,16 @@ xfs_efi_item_pin(xfs_efi_log_item_t *efip) STATIC void xfs_efi_item_unpin(xfs_efi_log_item_t *efip, int stale) { - xfs_mount_t *mp; + struct xfs_ail *ailp = efip->efi_item.li_ailp; - mp = efip->efi_item.li_mountp; - spin_lock(&mp->m_ail_lock); + spin_lock(&ailp->xa_lock); if (efip->efi_flags & XFS_EFI_CANCELED) { - /* - * xfs_trans_delete_ail() drops the AIL lock. - */ - xfs_trans_delete_ail(mp, (xfs_log_item_t *)efip); + /* xfs_trans_ail_delete() drops the AIL lock. */ + xfs_trans_ail_delete(ailp, (xfs_log_item_t *)efip); xfs_efi_item_free(efip); } else { efip->efi_flags |= XFS_EFI_COMMITTED; - spin_unlock(&mp->m_ail_lock); + spin_unlock(&ailp->xa_lock); } } @@ -134,26 +131,23 @@ xfs_efi_item_unpin(xfs_efi_log_item_t *efip, int stale) STATIC void xfs_efi_item_unpin_remove(xfs_efi_log_item_t *efip, xfs_trans_t *tp) { - xfs_mount_t *mp; + struct xfs_ail *ailp = efip->efi_item.li_ailp; xfs_log_item_desc_t *lidp; - mp = efip->efi_item.li_mountp; - spin_lock(&mp->m_ail_lock); + spin_lock(&ailp->xa_lock); if (efip->efi_flags & XFS_EFI_CANCELED) { /* * free the xaction descriptor pointing to this item */ lidp = xfs_trans_find_item(tp, (xfs_log_item_t *) efip); xfs_trans_free_item(tp, lidp); - /* - * pull the item off the AIL. - * xfs_trans_delete_ail() drops the AIL lock. - */ - xfs_trans_delete_ail(mp, (xfs_log_item_t *)efip); + + /* xfs_trans_ail_delete() drops the AIL lock. */ + xfs_trans_ail_delete(ailp, (xfs_log_item_t *)efip); xfs_efi_item_free(efip); } else { efip->efi_flags |= XFS_EFI_COMMITTED; - spin_unlock(&mp->m_ail_lock); + spin_unlock(&ailp->xa_lock); } } @@ -268,6 +262,7 @@ xfs_efi_init(xfs_mount_t *mp, efip->efi_item.li_type = XFS_LI_EFI; efip->efi_item.li_ops = &xfs_efi_item_ops; efip->efi_item.li_mountp = mp; + efip->efi_item.li_ailp = mp->m_ail; efip->efi_format.efi_nextents = nextents; efip->efi_format.efi_id = (__psint_t)(void*)efip; @@ -345,25 +340,22 @@ void xfs_efi_release(xfs_efi_log_item_t *efip, uint nextents) { - xfs_mount_t *mp; - int extents_left; + struct xfs_ail *ailp = efip->efi_item.li_ailp; + int extents_left; - mp = efip->efi_item.li_mountp; ASSERT(efip->efi_next_extent > 0); ASSERT(efip->efi_flags & XFS_EFI_COMMITTED); - spin_lock(&mp->m_ail_lock); + spin_lock(&ailp->xa_lock); ASSERT(efip->efi_next_extent >= nextents); efip->efi_next_extent -= nextents; extents_left = efip->efi_next_extent; if (extents_left == 0) { - /* - * xfs_trans_delete_ail() drops the AIL lock. - */ - xfs_trans_delete_ail(mp, (xfs_log_item_t *)efip); + /* xfs_trans_ail_delete() drops the AIL lock. */ + xfs_trans_ail_delete(ailp, (xfs_log_item_t *)efip); xfs_efi_item_free(efip); } else { - spin_unlock(&mp->m_ail_lock); + spin_unlock(&ailp->xa_lock); } } @@ -565,6 +557,7 @@ xfs_efd_init(xfs_mount_t *mp, efdp->efd_item.li_type = XFS_LI_EFD; efdp->efd_item.li_ops = &xfs_efd_item_ops; efdp->efd_item.li_mountp = mp; + efdp->efd_item.li_ailp = mp->m_ail; efdp->efd_efip = efip; efdp->efd_format.efd_nextents = nextents; efdp->efd_format.efd_efi_id = efip->efi_format.efi_id; diff --git a/fs/xfs/xfs_fs.h b/fs/xfs/xfs_fs.h index 01c0cc88d3f..589c41c3844 100644 --- a/fs/xfs/xfs_fs.h +++ b/fs/xfs/xfs_fs.h @@ -113,22 +113,14 @@ struct getbmapx { #define BMV_IF_ATTRFORK 0x1 /* return attr fork rather than data */ #define BMV_IF_NO_DMAPI_READ 0x2 /* Do not generate DMAPI read event */ #define BMV_IF_PREALLOC 0x4 /* rtn status BMV_OF_PREALLOC if req */ -#define BMV_IF_VALID (BMV_IF_ATTRFORK|BMV_IF_NO_DMAPI_READ|BMV_IF_PREALLOC) -#ifdef __KERNEL__ -#define BMV_IF_EXTENDED 0x40000000 /* getpmapx if set */ -#endif +#define BMV_IF_DELALLOC 0x8 /* rtn status BMV_OF_DELALLOC if req */ +#define BMV_IF_VALID \ + (BMV_IF_ATTRFORK|BMV_IF_NO_DMAPI_READ|BMV_IF_PREALLOC|BMV_IF_DELALLOC) /* bmv_oflags values - returned for for each non-header segment */ #define BMV_OF_PREALLOC 0x1 /* segment = unwritten pre-allocation */ - -/* Convert getbmap <-> getbmapx - move fields from p1 to p2. */ -#define GETBMAP_CONVERT(p1,p2) { \ - p2.bmv_offset = p1.bmv_offset; \ - p2.bmv_block = p1.bmv_block; \ - p2.bmv_length = p1.bmv_length; \ - p2.bmv_count = p1.bmv_count; \ - p2.bmv_entries = p1.bmv_entries; } - +#define BMV_OF_DELALLOC 0x2 /* segment = delayed allocation */ +#define BMV_OF_LAST 0x4 /* segment is the last in the file */ /* * Structure for XFS_IOC_FSSETDM. @@ -426,10 +418,6 @@ typedef struct xfs_handle { #define XFS_IOC_GETXFLAGS FS_IOC_GETFLAGS #define XFS_IOC_SETXFLAGS FS_IOC_SETFLAGS #define XFS_IOC_GETVERSION FS_IOC_GETVERSION -/* 32-bit compat counterparts */ -#define XFS_IOC32_GETXFLAGS FS_IOC32_GETFLAGS -#define XFS_IOC32_SETXFLAGS FS_IOC32_SETFLAGS -#define XFS_IOC32_GETVERSION FS_IOC32_GETVERSION /* * ioctl commands that replace IRIX fcntl()'s diff --git a/fs/xfs/xfs_fsops.c b/fs/xfs/xfs_fsops.c index 84583cf73db..852b6d32e8d 100644 --- a/fs/xfs/xfs_fsops.c +++ b/fs/xfs/xfs_fsops.c @@ -126,7 +126,7 @@ xfs_growfs_data_private( xfs_extlen_t agsize; xfs_extlen_t tmpsize; xfs_alloc_rec_t *arec; - xfs_btree_sblock_t *block; + struct xfs_btree_block *block; xfs_buf_t *bp; int bucket; int dpct; @@ -251,14 +251,14 @@ xfs_growfs_data_private( bp = xfs_buf_get(mp->m_ddev_targp, XFS_AGB_TO_DADDR(mp, agno, XFS_BNO_BLOCK(mp)), BTOBB(mp->m_sb.sb_blocksize), 0); - block = XFS_BUF_TO_SBLOCK(bp); + block = XFS_BUF_TO_BLOCK(bp); memset(block, 0, mp->m_sb.sb_blocksize); block->bb_magic = cpu_to_be32(XFS_ABTB_MAGIC); block->bb_level = 0; block->bb_numrecs = cpu_to_be16(1); - block->bb_leftsib = cpu_to_be32(NULLAGBLOCK); - block->bb_rightsib = cpu_to_be32(NULLAGBLOCK); - arec = XFS_BTREE_REC_ADDR(xfs_alloc, block, 1); + block->bb_u.s.bb_leftsib = cpu_to_be32(NULLAGBLOCK); + block->bb_u.s.bb_rightsib = cpu_to_be32(NULLAGBLOCK); + arec = XFS_ALLOC_REC_ADDR(mp, block, 1); arec->ar_startblock = cpu_to_be32(XFS_PREALLOC_BLOCKS(mp)); arec->ar_blockcount = cpu_to_be32( agsize - be32_to_cpu(arec->ar_startblock)); @@ -272,14 +272,14 @@ xfs_growfs_data_private( bp = xfs_buf_get(mp->m_ddev_targp, XFS_AGB_TO_DADDR(mp, agno, XFS_CNT_BLOCK(mp)), BTOBB(mp->m_sb.sb_blocksize), 0); - block = XFS_BUF_TO_SBLOCK(bp); + block = XFS_BUF_TO_BLOCK(bp); memset(block, 0, mp->m_sb.sb_blocksize); block->bb_magic = cpu_to_be32(XFS_ABTC_MAGIC); block->bb_level = 0; block->bb_numrecs = cpu_to_be16(1); - block->bb_leftsib = cpu_to_be32(NULLAGBLOCK); - block->bb_rightsib = cpu_to_be32(NULLAGBLOCK); - arec = XFS_BTREE_REC_ADDR(xfs_alloc, block, 1); + block->bb_u.s.bb_leftsib = cpu_to_be32(NULLAGBLOCK); + block->bb_u.s.bb_rightsib = cpu_to_be32(NULLAGBLOCK); + arec = XFS_ALLOC_REC_ADDR(mp, block, 1); arec->ar_startblock = cpu_to_be32(XFS_PREALLOC_BLOCKS(mp)); arec->ar_blockcount = cpu_to_be32( agsize - be32_to_cpu(arec->ar_startblock)); @@ -294,13 +294,13 @@ xfs_growfs_data_private( bp = xfs_buf_get(mp->m_ddev_targp, XFS_AGB_TO_DADDR(mp, agno, XFS_IBT_BLOCK(mp)), BTOBB(mp->m_sb.sb_blocksize), 0); - block = XFS_BUF_TO_SBLOCK(bp); + block = XFS_BUF_TO_BLOCK(bp); memset(block, 0, mp->m_sb.sb_blocksize); block->bb_magic = cpu_to_be32(XFS_IBT_MAGIC); block->bb_level = 0; block->bb_numrecs = 0; - block->bb_leftsib = cpu_to_be32(NULLAGBLOCK); - block->bb_rightsib = cpu_to_be32(NULLAGBLOCK); + block->bb_u.s.bb_leftsib = cpu_to_be32(NULLAGBLOCK); + block->bb_u.s.bb_rightsib = cpu_to_be32(NULLAGBLOCK); error = xfs_bwrite(mp, bp); if (error) { goto error0; @@ -435,6 +435,9 @@ xfs_growfs_data( xfs_growfs_data_t *in) { int error; + + if (!capable(CAP_SYS_ADMIN)) + return XFS_ERROR(EPERM); if (!mutex_trylock(&mp->m_growlock)) return XFS_ERROR(EWOULDBLOCK); error = xfs_growfs_data_private(mp, in); @@ -448,6 +451,9 @@ xfs_growfs_log( xfs_growfs_log_t *in) { int error; + + if (!capable(CAP_SYS_ADMIN)) + return XFS_ERROR(EPERM); if (!mutex_trylock(&mp->m_growlock)) return XFS_ERROR(EWOULDBLOCK); error = xfs_growfs_log_private(mp, in); diff --git a/fs/xfs/xfs_ialloc.c b/fs/xfs/xfs_ialloc.c index aad8c5da38a..e6ebbaeb4dc 100644 --- a/fs/xfs/xfs_ialloc.c +++ b/fs/xfs/xfs_ialloc.c @@ -41,68 +41,6 @@ #include "xfs_error.h" #include "xfs_bmap.h" -/* - * Log specified fields for the inode given by bp and off. - */ -STATIC void -xfs_ialloc_log_di( - xfs_trans_t *tp, /* transaction pointer */ - xfs_buf_t *bp, /* inode buffer */ - int off, /* index of inode in buffer */ - int fields) /* bitmask of fields to log */ -{ - int first; /* first byte number */ - int ioffset; /* off in bytes */ - int last; /* last byte number */ - xfs_mount_t *mp; /* mount point structure */ - static const short offsets[] = { /* field offsets */ - /* keep in sync with bits */ - offsetof(xfs_dinode_core_t, di_magic), - offsetof(xfs_dinode_core_t, di_mode), - offsetof(xfs_dinode_core_t, di_version), - offsetof(xfs_dinode_core_t, di_format), - offsetof(xfs_dinode_core_t, di_onlink), - offsetof(xfs_dinode_core_t, di_uid), - offsetof(xfs_dinode_core_t, di_gid), - offsetof(xfs_dinode_core_t, di_nlink), - offsetof(xfs_dinode_core_t, di_projid), - offsetof(xfs_dinode_core_t, di_pad), - offsetof(xfs_dinode_core_t, di_atime), - offsetof(xfs_dinode_core_t, di_mtime), - offsetof(xfs_dinode_core_t, di_ctime), - offsetof(xfs_dinode_core_t, di_size), - offsetof(xfs_dinode_core_t, di_nblocks), - offsetof(xfs_dinode_core_t, di_extsize), - offsetof(xfs_dinode_core_t, di_nextents), - offsetof(xfs_dinode_core_t, di_anextents), - offsetof(xfs_dinode_core_t, di_forkoff), - offsetof(xfs_dinode_core_t, di_aformat), - offsetof(xfs_dinode_core_t, di_dmevmask), - offsetof(xfs_dinode_core_t, di_dmstate), - offsetof(xfs_dinode_core_t, di_flags), - offsetof(xfs_dinode_core_t, di_gen), - offsetof(xfs_dinode_t, di_next_unlinked), - offsetof(xfs_dinode_t, di_u), - offsetof(xfs_dinode_t, di_a), - sizeof(xfs_dinode_t) - }; - - - ASSERT(offsetof(xfs_dinode_t, di_core) == 0); - ASSERT((fields & (XFS_DI_U|XFS_DI_A)) == 0); - mp = tp->t_mountp; - /* - * Get the inode-relative first and last bytes for these fields - */ - xfs_btree_offsets(fields, offsets, XFS_DI_NUM_BITS, &first, &last); - /* - * Convert to buffer offsets and log it. - */ - ioffset = off << mp->m_sb.sb_inodelog; - first += ioffset; - last += ioffset; - xfs_trans_log_buf(tp, bp, first, last); -} /* * Allocation group level functions. @@ -119,6 +57,102 @@ xfs_ialloc_cluster_alignment( } /* + * Lookup the record equal to ino in the btree given by cur. + */ +STATIC int /* error */ +xfs_inobt_lookup_eq( + struct xfs_btree_cur *cur, /* btree cursor */ + xfs_agino_t ino, /* starting inode of chunk */ + __int32_t fcnt, /* free inode count */ + xfs_inofree_t free, /* free inode mask */ + int *stat) /* success/failure */ +{ + cur->bc_rec.i.ir_startino = ino; + cur->bc_rec.i.ir_freecount = fcnt; + cur->bc_rec.i.ir_free = free; + return xfs_btree_lookup(cur, XFS_LOOKUP_EQ, stat); +} + +/* + * Lookup the first record greater than or equal to ino + * in the btree given by cur. + */ +int /* error */ +xfs_inobt_lookup_ge( + struct xfs_btree_cur *cur, /* btree cursor */ + xfs_agino_t ino, /* starting inode of chunk */ + __int32_t fcnt, /* free inode count */ + xfs_inofree_t free, /* free inode mask */ + int *stat) /* success/failure */ +{ + cur->bc_rec.i.ir_startino = ino; + cur->bc_rec.i.ir_freecount = fcnt; + cur->bc_rec.i.ir_free = free; + return xfs_btree_lookup(cur, XFS_LOOKUP_GE, stat); +} + +/* + * Lookup the first record less than or equal to ino + * in the btree given by cur. + */ +int /* error */ +xfs_inobt_lookup_le( + struct xfs_btree_cur *cur, /* btree cursor */ + xfs_agino_t ino, /* starting inode of chunk */ + __int32_t fcnt, /* free inode count */ + xfs_inofree_t free, /* free inode mask */ + int *stat) /* success/failure */ +{ + cur->bc_rec.i.ir_startino = ino; + cur->bc_rec.i.ir_freecount = fcnt; + cur->bc_rec.i.ir_free = free; + return xfs_btree_lookup(cur, XFS_LOOKUP_LE, stat); +} + +/* + * Update the record referred to by cur to the value given + * by [ino, fcnt, free]. + * This either works (return 0) or gets an EFSCORRUPTED error. + */ +STATIC int /* error */ +xfs_inobt_update( + struct xfs_btree_cur *cur, /* btree cursor */ + xfs_agino_t ino, /* starting inode of chunk */ + __int32_t fcnt, /* free inode count */ + xfs_inofree_t free) /* free inode mask */ +{ + union xfs_btree_rec rec; + + rec.inobt.ir_startino = cpu_to_be32(ino); + rec.inobt.ir_freecount = cpu_to_be32(fcnt); + rec.inobt.ir_free = cpu_to_be64(free); + return xfs_btree_update(cur, &rec); +} + +/* + * Get the data from the pointed-to record. + */ +int /* error */ +xfs_inobt_get_rec( + struct xfs_btree_cur *cur, /* btree cursor */ + xfs_agino_t *ino, /* output: starting inode of chunk */ + __int32_t *fcnt, /* output: number of free inodes */ + xfs_inofree_t *free, /* output: free inode mask */ + int *stat) /* output: success/failure */ +{ + union xfs_btree_rec *rec; + int error; + + error = xfs_btree_get_rec(cur, &rec, stat); + if (!error && *stat == 1) { + *ino = be32_to_cpu(rec->inobt.ir_startino); + *fcnt = be32_to_cpu(rec->inobt.ir_freecount); + *free = be64_to_cpu(rec->inobt.ir_free); + } + return error; +} + +/* * Allocate new inodes in the allocation group specified by agbp. * Return 0 for success, else error code. */ @@ -287,9 +321,9 @@ xfs_ialloc_ag_alloc( * able to use the file system. */ if (xfs_sb_version_hasnlink(&args.mp->m_sb)) - version = XFS_DINODE_VERSION_2; + version = 2; else - version = XFS_DINODE_VERSION_1; + version = 1; /* * Seed the new inode cluster with a random generation number. This @@ -310,18 +344,25 @@ xfs_ialloc_ag_alloc( XFS_BUF_LOCK); ASSERT(fbuf); ASSERT(!XFS_BUF_GETERROR(fbuf)); + /* - * Set initial values for the inodes in this buffer. + * Initialize all inodes in this buffer and then log them. + * + * XXX: It would be much better if we had just one transaction to + * log a whole cluster of inodes instead of all the indivdual + * transactions causing a lot of log traffic. */ xfs_biozero(fbuf, 0, ninodes << args.mp->m_sb.sb_inodelog); for (i = 0; i < ninodes; i++) { + int ioffset = i << args.mp->m_sb.sb_inodelog; + uint isize = sizeof(struct xfs_dinode); + free = XFS_MAKE_IPTR(args.mp, fbuf, i); - free->di_core.di_magic = cpu_to_be16(XFS_DINODE_MAGIC); - free->di_core.di_version = version; - free->di_core.di_gen = cpu_to_be32(gen); + free->di_magic = cpu_to_be16(XFS_DINODE_MAGIC); + free->di_version = version; + free->di_gen = cpu_to_be32(gen); free->di_next_unlinked = cpu_to_be32(NULLAGINO); - xfs_ialloc_log_di(tp, fbuf, i, - XFS_DI_CORE_BITS | XFS_DI_NEXT_UNLINKED); + xfs_trans_log_buf(tp, fbuf, ioffset, ioffset + isize - 1); } xfs_trans_inode_alloc_buf(tp, fbuf); } @@ -335,8 +376,7 @@ xfs_ialloc_ag_alloc( /* * Insert records describing the new inode chunk into the btree. */ - cur = xfs_btree_init_cursor(args.mp, tp, agbp, agno, - XFS_BTNUM_INO, (xfs_inode_t *)0, 0); + cur = xfs_inobt_init_cursor(args.mp, tp, agbp, agno); for (thisino = newino; thisino < newino + newlen; thisino += XFS_INODES_PER_CHUNK) { @@ -346,7 +386,7 @@ xfs_ialloc_ag_alloc( return error; } ASSERT(i == 0); - if ((error = xfs_inobt_insert(cur, &i))) { + if ((error = xfs_btree_insert(cur, &i))) { xfs_btree_del_cursor(cur, XFS_BTREE_ERROR); return error; } @@ -676,8 +716,7 @@ nextag: */ agno = tagno; *IO_agbp = NULL; - cur = xfs_btree_init_cursor(mp, tp, agbp, be32_to_cpu(agi->agi_seqno), - XFS_BTNUM_INO, (xfs_inode_t *)0, 0); + cur = xfs_inobt_init_cursor(mp, tp, agbp, be32_to_cpu(agi->agi_seqno)); /* * If pagino is 0 (this is the root inode allocation) use newino. * This must work because we've just allocated some. @@ -697,7 +736,7 @@ nextag: goto error0; XFS_WANT_CORRUPTED_GOTO(i == 1, error0); freecount += rec.ir_freecount; - if ((error = xfs_inobt_increment(cur, 0, &i))) + if ((error = xfs_btree_increment(cur, 0, &i))) goto error0; } while (i == 1); @@ -741,7 +780,7 @@ nextag: /* * Search left with tcur, back up 1 record. */ - if ((error = xfs_inobt_decrement(tcur, 0, &i))) + if ((error = xfs_btree_decrement(tcur, 0, &i))) goto error1; doneleft = !i; if (!doneleft) { @@ -755,7 +794,7 @@ nextag: /* * Search right with cur, go forward 1 record. */ - if ((error = xfs_inobt_increment(cur, 0, &i))) + if ((error = xfs_btree_increment(cur, 0, &i))) goto error1; doneright = !i; if (!doneright) { @@ -817,7 +856,7 @@ nextag: * further left. */ if (useleft) { - if ((error = xfs_inobt_decrement(tcur, 0, + if ((error = xfs_btree_decrement(tcur, 0, &i))) goto error1; doneleft = !i; @@ -837,7 +876,7 @@ nextag: * further right. */ else { - if ((error = xfs_inobt_increment(cur, 0, + if ((error = xfs_btree_increment(cur, 0, &i))) goto error1; doneright = !i; @@ -892,7 +931,7 @@ nextag: XFS_WANT_CORRUPTED_GOTO(i == 1, error0); if (rec.ir_freecount > 0) break; - if ((error = xfs_inobt_increment(cur, 0, &i))) + if ((error = xfs_btree_increment(cur, 0, &i))) goto error0; XFS_WANT_CORRUPTED_GOTO(i == 1, error0); } @@ -926,7 +965,7 @@ nextag: goto error0; XFS_WANT_CORRUPTED_GOTO(i == 1, error0); freecount += rec.ir_freecount; - if ((error = xfs_inobt_increment(cur, 0, &i))) + if ((error = xfs_btree_increment(cur, 0, &i))) goto error0; } while (i == 1); ASSERT(freecount == be32_to_cpu(agi->agi_freecount) || @@ -1022,8 +1061,7 @@ xfs_difree( /* * Initialize the cursor. */ - cur = xfs_btree_init_cursor(mp, tp, agbp, agno, XFS_BTNUM_INO, - (xfs_inode_t *)0, 0); + cur = xfs_inobt_init_cursor(mp, tp, agbp, agno); #ifdef DEBUG if (cur->bc_nlevels == 1) { int freecount = 0; @@ -1036,7 +1074,7 @@ xfs_difree( goto error0; if (i) { freecount += rec.ir_freecount; - if ((error = xfs_inobt_increment(cur, 0, &i))) + if ((error = xfs_btree_increment(cur, 0, &i))) goto error0; } } while (i == 1); @@ -1098,8 +1136,8 @@ xfs_difree( xfs_trans_mod_sb(tp, XFS_TRANS_SB_ICOUNT, -ilen); xfs_trans_mod_sb(tp, XFS_TRANS_SB_IFREE, -(ilen - 1)); - if ((error = xfs_inobt_delete(cur, &i))) { - cmn_err(CE_WARN, "xfs_difree: xfs_inobt_delete returned an error %d on %s.\n", + if ((error = xfs_btree_delete(cur, &i))) { + cmn_err(CE_WARN, "xfs_difree: xfs_btree_delete returned an error %d on %s.\n", error, mp->m_fsname); goto error0; } @@ -1141,7 +1179,7 @@ xfs_difree( goto error0; if (i) { freecount += rec.ir_freecount; - if ((error = xfs_inobt_increment(cur, 0, &i))) + if ((error = xfs_btree_increment(cur, 0, &i))) goto error0; } } while (i == 1); @@ -1158,36 +1196,28 @@ error0: } /* - * Return the location of the inode in bno/off, for mapping it into a buffer. + * Return the location of the inode in imap, for mapping it into a buffer. */ -/*ARGSUSED*/ int -xfs_dilocate( - xfs_mount_t *mp, /* file system mount structure */ - xfs_trans_t *tp, /* transaction pointer */ +xfs_imap( + xfs_mount_t *mp, /* file system mount structure */ + xfs_trans_t *tp, /* transaction pointer */ xfs_ino_t ino, /* inode to locate */ - xfs_fsblock_t *bno, /* output: block containing inode */ - int *len, /* output: num blocks in inode cluster */ - int *off, /* output: index in block of inode */ - uint flags) /* flags concerning inode lookup */ + struct xfs_imap *imap, /* location map structure */ + uint flags) /* flags for inode btree lookup */ { xfs_agblock_t agbno; /* block number of inode in the alloc group */ - xfs_buf_t *agbp; /* agi buffer */ xfs_agino_t agino; /* inode number within alloc group */ xfs_agnumber_t agno; /* allocation group number */ int blks_per_cluster; /* num blocks per inode cluster */ xfs_agblock_t chunk_agbno; /* first block in inode chunk */ - xfs_agino_t chunk_agino; /* first agino in inode chunk */ - __int32_t chunk_cnt; /* count of free inodes in chunk */ - xfs_inofree_t chunk_free; /* mask of free inodes in chunk */ xfs_agblock_t cluster_agbno; /* first block in inode cluster */ - xfs_btree_cur_t *cur; /* inode btree cursor */ int error; /* error code */ - int i; /* temp state */ int offset; /* index of inode in its buffer */ int offset_agbno; /* blks from chunk start to inode */ ASSERT(ino != NULLFSINO); + /* * Split up the inode number into its parts. */ @@ -1198,24 +1228,24 @@ xfs_dilocate( ino != XFS_AGINO_TO_INO(mp, agno, agino)) { #ifdef DEBUG /* no diagnostics for bulkstat, ino comes from userspace */ - if (flags & XFS_IMAP_BULKSTAT) + if (flags & XFS_IGET_BULKSTAT) return XFS_ERROR(EINVAL); if (agno >= mp->m_sb.sb_agcount) { xfs_fs_cmn_err(CE_ALERT, mp, - "xfs_dilocate: agno (%d) >= " + "xfs_imap: agno (%d) >= " "mp->m_sb.sb_agcount (%d)", agno, mp->m_sb.sb_agcount); } if (agbno >= mp->m_sb.sb_agblocks) { xfs_fs_cmn_err(CE_ALERT, mp, - "xfs_dilocate: agbno (0x%llx) >= " + "xfs_imap: agbno (0x%llx) >= " "mp->m_sb.sb_agblocks (0x%lx)", (unsigned long long) agbno, (unsigned long) mp->m_sb.sb_agblocks); } if (ino != XFS_AGINO_TO_INO(mp, agno, agino)) { xfs_fs_cmn_err(CE_ALERT, mp, - "xfs_dilocate: ino (0x%llx) != " + "xfs_imap: ino (0x%llx) != " "XFS_AGINO_TO_INO(mp, agno, agino) " "(0x%llx)", ino, XFS_AGINO_TO_INO(mp, agno, agino)); @@ -1224,65 +1254,89 @@ xfs_dilocate( #endif /* DEBUG */ return XFS_ERROR(EINVAL); } - if ((mp->m_sb.sb_blocksize >= XFS_INODE_CLUSTER_SIZE(mp)) || - !(flags & XFS_IMAP_LOOKUP)) { + + /* + * If the inode cluster size is the same as the blocksize or + * smaller we get to the buffer by simple arithmetics. + */ + if (XFS_INODE_CLUSTER_SIZE(mp) <= mp->m_sb.sb_blocksize) { offset = XFS_INO_TO_OFFSET(mp, ino); ASSERT(offset < mp->m_sb.sb_inopblock); - *bno = XFS_AGB_TO_FSB(mp, agno, agbno); - *off = offset; - *len = 1; + + imap->im_blkno = XFS_AGB_TO_DADDR(mp, agno, agbno); + imap->im_len = XFS_FSB_TO_BB(mp, 1); + imap->im_boffset = (ushort)(offset << mp->m_sb.sb_inodelog); return 0; } + blks_per_cluster = XFS_INODE_CLUSTER_SIZE(mp) >> mp->m_sb.sb_blocklog; - if (*bno != NULLFSBLOCK) { + + /* + * If we get a block number passed from bulkstat we can use it to + * find the buffer easily. + */ + if (imap->im_blkno) { offset = XFS_INO_TO_OFFSET(mp, ino); ASSERT(offset < mp->m_sb.sb_inopblock); - cluster_agbno = XFS_FSB_TO_AGBNO(mp, *bno); - *off = ((agbno - cluster_agbno) * mp->m_sb.sb_inopblock) + - offset; - *len = blks_per_cluster; + + cluster_agbno = XFS_DADDR_TO_AGBNO(mp, imap->im_blkno); + offset += (agbno - cluster_agbno) * mp->m_sb.sb_inopblock; + + imap->im_len = XFS_FSB_TO_BB(mp, blks_per_cluster); + imap->im_boffset = (ushort)(offset << mp->m_sb.sb_inodelog); return 0; } + + /* + * If the inode chunks are aligned then use simple maths to + * find the location. Otherwise we have to do a btree + * lookup to find the location. + */ if (mp->m_inoalign_mask) { offset_agbno = agbno & mp->m_inoalign_mask; chunk_agbno = agbno - offset_agbno; } else { + xfs_btree_cur_t *cur; /* inode btree cursor */ + xfs_agino_t chunk_agino; /* first agino in inode chunk */ + __int32_t chunk_cnt; /* count of free inodes in chunk */ + xfs_inofree_t chunk_free; /* mask of free inodes in chunk */ + xfs_buf_t *agbp; /* agi buffer */ + int i; /* temp state */ + down_read(&mp->m_peraglock); error = xfs_ialloc_read_agi(mp, tp, agno, &agbp); up_read(&mp->m_peraglock); if (error) { -#ifdef DEBUG - xfs_fs_cmn_err(CE_ALERT, mp, "xfs_dilocate: " + xfs_fs_cmn_err(CE_ALERT, mp, "xfs_imap: " "xfs_ialloc_read_agi() returned " "error %d, agno %d", error, agno); -#endif /* DEBUG */ return error; } - cur = xfs_btree_init_cursor(mp, tp, agbp, agno, XFS_BTNUM_INO, - (xfs_inode_t *)0, 0); - if ((error = xfs_inobt_lookup_le(cur, agino, 0, 0, &i))) { -#ifdef DEBUG - xfs_fs_cmn_err(CE_ALERT, mp, "xfs_dilocate: " + + cur = xfs_inobt_init_cursor(mp, tp, agbp, agno); + error = xfs_inobt_lookup_le(cur, agino, 0, 0, &i); + if (error) { + xfs_fs_cmn_err(CE_ALERT, mp, "xfs_imap: " "xfs_inobt_lookup_le() failed"); -#endif /* DEBUG */ goto error0; } - if ((error = xfs_inobt_get_rec(cur, &chunk_agino, &chunk_cnt, - &chunk_free, &i))) { -#ifdef DEBUG - xfs_fs_cmn_err(CE_ALERT, mp, "xfs_dilocate: " + + error = xfs_inobt_get_rec(cur, &chunk_agino, &chunk_cnt, + &chunk_free, &i); + if (error) { + xfs_fs_cmn_err(CE_ALERT, mp, "xfs_imap: " "xfs_inobt_get_rec() failed"); -#endif /* DEBUG */ goto error0; } if (i == 0) { #ifdef DEBUG - xfs_fs_cmn_err(CE_ALERT, mp, "xfs_dilocate: " + xfs_fs_cmn_err(CE_ALERT, mp, "xfs_imap: " "xfs_inobt_get_rec() failed"); #endif /* DEBUG */ error = XFS_ERROR(EINVAL); } + error0: xfs_trans_brelse(tp, agbp); xfs_btree_del_cursor(cur, XFS_BTREE_NOERROR); if (error) @@ -1290,19 +1344,35 @@ xfs_dilocate( chunk_agbno = XFS_AGINO_TO_AGBNO(mp, chunk_agino); offset_agbno = agbno - chunk_agbno; } + ASSERT(agbno >= chunk_agbno); cluster_agbno = chunk_agbno + ((offset_agbno / blks_per_cluster) * blks_per_cluster); offset = ((agbno - cluster_agbno) * mp->m_sb.sb_inopblock) + XFS_INO_TO_OFFSET(mp, ino); - *bno = XFS_AGB_TO_FSB(mp, agno, cluster_agbno); - *off = offset; - *len = blks_per_cluster; + + imap->im_blkno = XFS_AGB_TO_DADDR(mp, agno, cluster_agbno); + imap->im_len = XFS_FSB_TO_BB(mp, blks_per_cluster); + imap->im_boffset = (ushort)(offset << mp->m_sb.sb_inodelog); + + /* + * If the inode number maps to a block outside the bounds + * of the file system then return NULL rather than calling + * read_buf and panicing when we get an error from the + * driver. + */ + if ((imap->im_blkno + imap->im_len) > + XFS_FSB_TO_BB(mp, mp->m_sb.sb_dblocks)) { + xfs_fs_cmn_err(CE_ALERT, mp, "xfs_imap: " + "(imap->im_blkno (0x%llx) + imap->im_len (0x%llx)) > " + " XFS_FSB_TO_BB(mp, mp->m_sb.sb_dblocks) (0x%llx)", + (unsigned long long) imap->im_blkno, + (unsigned long long) imap->im_len, + XFS_FSB_TO_BB(mp, mp->m_sb.sb_dblocks)); + return XFS_ERROR(EINVAL); + } + return 0; -error0: - xfs_trans_brelse(tp, agbp); - xfs_btree_del_cursor(cur, XFS_BTREE_ERROR); - return error; } /* @@ -1370,70 +1440,95 @@ xfs_ialloc_log_agi( xfs_trans_log_buf(tp, bp, first, last); } +#ifdef DEBUG +STATIC void +xfs_check_agi_unlinked( + struct xfs_agi *agi) +{ + int i; + + for (i = 0; i < XFS_AGI_UNLINKED_BUCKETS; i++) + ASSERT(agi->agi_unlinked[i]); +} +#else +#define xfs_check_agi_unlinked(agi) +#endif + /* * Read in the allocation group header (inode allocation section) */ int -xfs_ialloc_read_agi( - xfs_mount_t *mp, /* file system mount structure */ - xfs_trans_t *tp, /* transaction pointer */ - xfs_agnumber_t agno, /* allocation group number */ - xfs_buf_t **bpp) /* allocation group hdr buf */ +xfs_read_agi( + struct xfs_mount *mp, /* file system mount structure */ + struct xfs_trans *tp, /* transaction pointer */ + xfs_agnumber_t agno, /* allocation group number */ + struct xfs_buf **bpp) /* allocation group hdr buf */ { - xfs_agi_t *agi; /* allocation group header */ - int agi_ok; /* agi is consistent */ - xfs_buf_t *bp; /* allocation group hdr buf */ - xfs_perag_t *pag; /* per allocation group data */ - int error; + struct xfs_agi *agi; /* allocation group header */ + int agi_ok; /* agi is consistent */ + int error; ASSERT(agno != NULLAGNUMBER); - error = xfs_trans_read_buf( - mp, tp, mp->m_ddev_targp, + + error = xfs_trans_read_buf(mp, tp, mp->m_ddev_targp, XFS_AG_DADDR(mp, agno, XFS_AGI_DADDR(mp)), - XFS_FSS_TO_BB(mp, 1), 0, &bp); + XFS_FSS_TO_BB(mp, 1), 0, bpp); if (error) return error; - ASSERT(bp && !XFS_BUF_GETERROR(bp)); + + ASSERT(*bpp && !XFS_BUF_GETERROR(*bpp)); + agi = XFS_BUF_TO_AGI(*bpp); /* * Validate the magic number of the agi block. */ - agi = XFS_BUF_TO_AGI(bp); - agi_ok = - be32_to_cpu(agi->agi_magicnum) == XFS_AGI_MAGIC && - XFS_AGI_GOOD_VERSION(be32_to_cpu(agi->agi_versionnum)); + agi_ok = be32_to_cpu(agi->agi_magicnum) == XFS_AGI_MAGIC && + XFS_AGI_GOOD_VERSION(be32_to_cpu(agi->agi_versionnum)) && + be32_to_cpu(agi->agi_seqno) == agno; if (unlikely(XFS_TEST_ERROR(!agi_ok, mp, XFS_ERRTAG_IALLOC_READ_AGI, XFS_RANDOM_IALLOC_READ_AGI))) { - XFS_CORRUPTION_ERROR("xfs_ialloc_read_agi", XFS_ERRLEVEL_LOW, + XFS_CORRUPTION_ERROR("xfs_read_agi", XFS_ERRLEVEL_LOW, mp, agi); - xfs_trans_brelse(tp, bp); + xfs_trans_brelse(tp, *bpp); return XFS_ERROR(EFSCORRUPTED); } + + XFS_BUF_SET_VTYPE_REF(*bpp, B_FS_AGI, XFS_AGI_REF); + + xfs_check_agi_unlinked(agi); + return 0; +} + +int +xfs_ialloc_read_agi( + struct xfs_mount *mp, /* file system mount structure */ + struct xfs_trans *tp, /* transaction pointer */ + xfs_agnumber_t agno, /* allocation group number */ + struct xfs_buf **bpp) /* allocation group hdr buf */ +{ + struct xfs_agi *agi; /* allocation group header */ + struct xfs_perag *pag; /* per allocation group data */ + int error; + + error = xfs_read_agi(mp, tp, agno, bpp); + if (error) + return error; + + agi = XFS_BUF_TO_AGI(*bpp); pag = &mp->m_perag[agno]; + if (!pag->pagi_init) { pag->pagi_freecount = be32_to_cpu(agi->agi_freecount); pag->pagi_count = be32_to_cpu(agi->agi_count); pag->pagi_init = 1; - } else { - /* - * It's possible for these to be out of sync if - * we are in the middle of a forced shutdown. - */ - ASSERT(pag->pagi_freecount == be32_to_cpu(agi->agi_freecount) || - XFS_FORCED_SHUTDOWN(mp)); } -#ifdef DEBUG - { - int i; - - for (i = 0; i < XFS_AGI_UNLINKED_BUCKETS; i++) - ASSERT(agi->agi_unlinked[i]); - } -#endif - - XFS_BUF_SET_VTYPE_REF(bp, B_FS_AGI, XFS_AGI_REF); - *bpp = bp; + /* + * It's possible for these to be out of sync if + * we are in the middle of a forced shutdown. + */ + ASSERT(pag->pagi_freecount == be32_to_cpu(agi->agi_freecount) || + XFS_FORCED_SHUTDOWN(mp)); return 0; } diff --git a/fs/xfs/xfs_ialloc.h b/fs/xfs/xfs_ialloc.h index 4e30ec1d13b..50f558a4e0a 100644 --- a/fs/xfs/xfs_ialloc.h +++ b/fs/xfs/xfs_ialloc.h @@ -20,6 +20,7 @@ struct xfs_buf; struct xfs_dinode; +struct xfs_imap; struct xfs_mount; struct xfs_trans; @@ -56,7 +57,6 @@ static inline int xfs_ialloc_find_free(xfs_inofree_t *fp) } -#ifdef __KERNEL__ /* * Allocate an inode on disk. * Mode is used to tell whether the new inode will need space, and whether @@ -105,17 +105,14 @@ xfs_difree( xfs_ino_t *first_ino); /* first inode in deleted cluster */ /* - * Return the location of the inode in bno/len/off, - * for mapping it into a buffer. + * Return the location of the inode in imap, for mapping it into a buffer. */ int -xfs_dilocate( +xfs_imap( struct xfs_mount *mp, /* file system mount structure */ struct xfs_trans *tp, /* transaction pointer */ xfs_ino_t ino, /* inode to locate */ - xfs_fsblock_t *bno, /* output: block containing inode */ - int *len, /* output: num blocks in cluster*/ - int *off, /* output: index in block of inode */ + struct xfs_imap *imap, /* location map structure */ uint flags); /* flags for inode btree lookup */ /* @@ -154,6 +151,24 @@ xfs_ialloc_pagi_init( struct xfs_trans *tp, /* transaction pointer */ xfs_agnumber_t agno); /* allocation group number */ -#endif /* __KERNEL__ */ +/* + * Lookup the first record greater than or equal to ino + * in the btree given by cur. + */ +int xfs_inobt_lookup_ge(struct xfs_btree_cur *cur, xfs_agino_t ino, + __int32_t fcnt, xfs_inofree_t free, int *stat); + +/* + * Lookup the first record less than or equal to ino + * in the btree given by cur. + */ +int xfs_inobt_lookup_le(struct xfs_btree_cur *cur, xfs_agino_t ino, + __int32_t fcnt, xfs_inofree_t free, int *stat); + +/* + * Get the data from the pointed-to record. + */ +extern int xfs_inobt_get_rec(struct xfs_btree_cur *cur, xfs_agino_t *ino, + __int32_t *fcnt, xfs_inofree_t *free, int *stat); #endif /* __XFS_IALLOC_H__ */ diff --git a/fs/xfs/xfs_ialloc_btree.c b/fs/xfs/xfs_ialloc_btree.c index 83502f3edef..99f2408e8d8 100644 --- a/fs/xfs/xfs_ialloc_btree.c +++ b/fs/xfs/xfs_ialloc_btree.c @@ -35,2044 +35,349 @@ #include "xfs_dinode.h" #include "xfs_inode.h" #include "xfs_btree.h" +#include "xfs_btree_trace.h" #include "xfs_ialloc.h" #include "xfs_alloc.h" #include "xfs_error.h" -STATIC void xfs_inobt_log_block(xfs_trans_t *, xfs_buf_t *, int); -STATIC void xfs_inobt_log_keys(xfs_btree_cur_t *, xfs_buf_t *, int, int); -STATIC void xfs_inobt_log_ptrs(xfs_btree_cur_t *, xfs_buf_t *, int, int); -STATIC void xfs_inobt_log_recs(xfs_btree_cur_t *, xfs_buf_t *, int, int); -STATIC int xfs_inobt_lshift(xfs_btree_cur_t *, int, int *); -STATIC int xfs_inobt_newroot(xfs_btree_cur_t *, int *); -STATIC int xfs_inobt_rshift(xfs_btree_cur_t *, int, int *); -STATIC int xfs_inobt_split(xfs_btree_cur_t *, int, xfs_agblock_t *, - xfs_inobt_key_t *, xfs_btree_cur_t **, int *); -STATIC int xfs_inobt_updkey(xfs_btree_cur_t *, xfs_inobt_key_t *, int); -/* - * Single level of the xfs_inobt_delete record deletion routine. - * Delete record pointed to by cur/level. - * Remove the record from its block then rebalance the tree. - * Return 0 for error, 1 for done, 2 to go on to the next level. - */ -STATIC int /* error */ -xfs_inobt_delrec( - xfs_btree_cur_t *cur, /* btree cursor */ - int level, /* level removing record from */ - int *stat) /* fail/done/go-on */ +STATIC int +xfs_inobt_get_minrecs( + struct xfs_btree_cur *cur, + int level) { - xfs_buf_t *agbp; /* buffer for a.g. inode header */ - xfs_mount_t *mp; /* mount structure */ - xfs_agi_t *agi; /* allocation group inode header */ - xfs_inobt_block_t *block; /* btree block record/key lives in */ - xfs_agblock_t bno; /* btree block number */ - xfs_buf_t *bp; /* buffer for block */ - int error; /* error return value */ - int i; /* loop index */ - xfs_inobt_key_t key; /* kp points here if block is level 0 */ - xfs_inobt_key_t *kp = NULL; /* pointer to btree keys */ - xfs_agblock_t lbno; /* left block's block number */ - xfs_buf_t *lbp; /* left block's buffer pointer */ - xfs_inobt_block_t *left; /* left btree block */ - xfs_inobt_key_t *lkp; /* left block key pointer */ - xfs_inobt_ptr_t *lpp; /* left block address pointer */ - int lrecs = 0; /* number of records in left block */ - xfs_inobt_rec_t *lrp; /* left block record pointer */ - xfs_inobt_ptr_t *pp = NULL; /* pointer to btree addresses */ - int ptr; /* index in btree block for this rec */ - xfs_agblock_t rbno; /* right block's block number */ - xfs_buf_t *rbp; /* right block's buffer pointer */ - xfs_inobt_block_t *right; /* right btree block */ - xfs_inobt_key_t *rkp; /* right block key pointer */ - xfs_inobt_rec_t *rp; /* pointer to btree records */ - xfs_inobt_ptr_t *rpp; /* right block address pointer */ - int rrecs = 0; /* number of records in right block */ - int numrecs; - xfs_inobt_rec_t *rrp; /* right block record pointer */ - xfs_btree_cur_t *tcur; /* temporary btree cursor */ - - mp = cur->bc_mp; - - /* - * Get the index of the entry being deleted, check for nothing there. - */ - ptr = cur->bc_ptrs[level]; - if (ptr == 0) { - *stat = 0; - return 0; - } - - /* - * Get the buffer & block containing the record or key/ptr. - */ - bp = cur->bc_bufs[level]; - block = XFS_BUF_TO_INOBT_BLOCK(bp); -#ifdef DEBUG - if ((error = xfs_btree_check_sblock(cur, block, level, bp))) - return error; -#endif - /* - * Fail if we're off the end of the block. - */ + return cur->bc_mp->m_inobt_mnr[level != 0]; +} - numrecs = be16_to_cpu(block->bb_numrecs); - if (ptr > numrecs) { - *stat = 0; - return 0; - } - /* - * It's a nonleaf. Excise the key and ptr being deleted, by - * sliding the entries past them down one. - * Log the changed areas of the block. - */ - if (level > 0) { - kp = XFS_INOBT_KEY_ADDR(block, 1, cur); - pp = XFS_INOBT_PTR_ADDR(block, 1, cur); -#ifdef DEBUG - for (i = ptr; i < numrecs; i++) { - if ((error = xfs_btree_check_sptr(cur, be32_to_cpu(pp[i]), level))) - return error; - } -#endif - if (ptr < numrecs) { - memmove(&kp[ptr - 1], &kp[ptr], - (numrecs - ptr) * sizeof(*kp)); - memmove(&pp[ptr - 1], &pp[ptr], - (numrecs - ptr) * sizeof(*kp)); - xfs_inobt_log_keys(cur, bp, ptr, numrecs - 1); - xfs_inobt_log_ptrs(cur, bp, ptr, numrecs - 1); - } - } - /* - * It's a leaf. Excise the record being deleted, by sliding the - * entries past it down one. Log the changed areas of the block. - */ - else { - rp = XFS_INOBT_REC_ADDR(block, 1, cur); - if (ptr < numrecs) { - memmove(&rp[ptr - 1], &rp[ptr], - (numrecs - ptr) * sizeof(*rp)); - xfs_inobt_log_recs(cur, bp, ptr, numrecs - 1); - } - /* - * If it's the first record in the block, we'll need a key - * structure to pass up to the next level (updkey). - */ - if (ptr == 1) { - key.ir_startino = rp->ir_startino; - kp = &key; - } - } - /* - * Decrement and log the number of entries in the block. - */ - numrecs--; - block->bb_numrecs = cpu_to_be16(numrecs); - xfs_inobt_log_block(cur->bc_tp, bp, XFS_BB_NUMRECS); - /* - * Is this the root level? If so, we're almost done. - */ - if (level == cur->bc_nlevels - 1) { - /* - * If this is the root level, - * and there's only one entry left, - * and it's NOT the leaf level, - * then we can get rid of this level. - */ - if (numrecs == 1 && level > 0) { - agbp = cur->bc_private.a.agbp; - agi = XFS_BUF_TO_AGI(agbp); - /* - * pp is still set to the first pointer in the block. - * Make it the new root of the btree. - */ - bno = be32_to_cpu(agi->agi_root); - agi->agi_root = *pp; - be32_add_cpu(&agi->agi_level, -1); - /* - * Free the block. - */ - if ((error = xfs_free_extent(cur->bc_tp, - XFS_AGB_TO_FSB(mp, cur->bc_private.a.agno, bno), 1))) - return error; - xfs_trans_binval(cur->bc_tp, bp); - xfs_ialloc_log_agi(cur->bc_tp, agbp, - XFS_AGI_ROOT | XFS_AGI_LEVEL); - /* - * Update the cursor so there's one fewer level. - */ - cur->bc_bufs[level] = NULL; - cur->bc_nlevels--; - } else if (level > 0 && - (error = xfs_inobt_decrement(cur, level, &i))) - return error; - *stat = 1; - return 0; - } - /* - * If we deleted the leftmost entry in the block, update the - * key values above us in the tree. - */ - if (ptr == 1 && (error = xfs_inobt_updkey(cur, kp, level + 1))) - return error; - /* - * If the number of records remaining in the block is at least - * the minimum, we're done. - */ - if (numrecs >= XFS_INOBT_BLOCK_MINRECS(level, cur)) { - if (level > 0 && - (error = xfs_inobt_decrement(cur, level, &i))) - return error; - *stat = 1; - return 0; - } - /* - * Otherwise, we have to move some records around to keep the - * tree balanced. Look at the left and right sibling blocks to - * see if we can re-balance by moving only one record. - */ - rbno = be32_to_cpu(block->bb_rightsib); - lbno = be32_to_cpu(block->bb_leftsib); - bno = NULLAGBLOCK; - ASSERT(rbno != NULLAGBLOCK || lbno != NULLAGBLOCK); - /* - * Duplicate the cursor so our btree manipulations here won't - * disrupt the next level up. - */ - if ((error = xfs_btree_dup_cursor(cur, &tcur))) - return error; - /* - * If there's a right sibling, see if it's ok to shift an entry - * out of it. - */ - if (rbno != NULLAGBLOCK) { - /* - * Move the temp cursor to the last entry in the next block. - * Actually any entry but the first would suffice. - */ - i = xfs_btree_lastrec(tcur, level); - XFS_WANT_CORRUPTED_GOTO(i == 1, error0); - if ((error = xfs_inobt_increment(tcur, level, &i))) - goto error0; - XFS_WANT_CORRUPTED_GOTO(i == 1, error0); - i = xfs_btree_lastrec(tcur, level); - XFS_WANT_CORRUPTED_GOTO(i == 1, error0); - /* - * Grab a pointer to the block. - */ - rbp = tcur->bc_bufs[level]; - right = XFS_BUF_TO_INOBT_BLOCK(rbp); -#ifdef DEBUG - if ((error = xfs_btree_check_sblock(cur, right, level, rbp))) - goto error0; -#endif - /* - * Grab the current block number, for future use. - */ - bno = be32_to_cpu(right->bb_leftsib); - /* - * If right block is full enough so that removing one entry - * won't make it too empty, and left-shifting an entry out - * of right to us works, we're done. - */ - if (be16_to_cpu(right->bb_numrecs) - 1 >= - XFS_INOBT_BLOCK_MINRECS(level, cur)) { - if ((error = xfs_inobt_lshift(tcur, level, &i))) - goto error0; - if (i) { - ASSERT(be16_to_cpu(block->bb_numrecs) >= - XFS_INOBT_BLOCK_MINRECS(level, cur)); - xfs_btree_del_cursor(tcur, - XFS_BTREE_NOERROR); - if (level > 0 && - (error = xfs_inobt_decrement(cur, level, - &i))) - return error; - *stat = 1; - return 0; - } - } - /* - * Otherwise, grab the number of records in right for - * future reference, and fix up the temp cursor to point - * to our block again (last record). - */ - rrecs = be16_to_cpu(right->bb_numrecs); - if (lbno != NULLAGBLOCK) { - xfs_btree_firstrec(tcur, level); - if ((error = xfs_inobt_decrement(tcur, level, &i))) - goto error0; - } - } - /* - * If there's a left sibling, see if it's ok to shift an entry - * out of it. - */ - if (lbno != NULLAGBLOCK) { - /* - * Move the temp cursor to the first entry in the - * previous block. - */ - xfs_btree_firstrec(tcur, level); - if ((error = xfs_inobt_decrement(tcur, level, &i))) - goto error0; - xfs_btree_firstrec(tcur, level); - /* - * Grab a pointer to the block. - */ - lbp = tcur->bc_bufs[level]; - left = XFS_BUF_TO_INOBT_BLOCK(lbp); -#ifdef DEBUG - if ((error = xfs_btree_check_sblock(cur, left, level, lbp))) - goto error0; -#endif - /* - * Grab the current block number, for future use. - */ - bno = be32_to_cpu(left->bb_rightsib); - /* - * If left block is full enough so that removing one entry - * won't make it too empty, and right-shifting an entry out - * of left to us works, we're done. - */ - if (be16_to_cpu(left->bb_numrecs) - 1 >= - XFS_INOBT_BLOCK_MINRECS(level, cur)) { - if ((error = xfs_inobt_rshift(tcur, level, &i))) - goto error0; - if (i) { - ASSERT(be16_to_cpu(block->bb_numrecs) >= - XFS_INOBT_BLOCK_MINRECS(level, cur)); - xfs_btree_del_cursor(tcur, - XFS_BTREE_NOERROR); - if (level == 0) - cur->bc_ptrs[0]++; - *stat = 1; - return 0; - } - } - /* - * Otherwise, grab the number of records in right for - * future reference. - */ - lrecs = be16_to_cpu(left->bb_numrecs); - } - /* - * Delete the temp cursor, we're done with it. - */ - xfs_btree_del_cursor(tcur, XFS_BTREE_NOERROR); - /* - * If here, we need to do a join to keep the tree balanced. - */ - ASSERT(bno != NULLAGBLOCK); - /* - * See if we can join with the left neighbor block. - */ - if (lbno != NULLAGBLOCK && - lrecs + numrecs <= XFS_INOBT_BLOCK_MAXRECS(level, cur)) { - /* - * Set "right" to be the starting block, - * "left" to be the left neighbor. - */ - rbno = bno; - right = block; - rrecs = be16_to_cpu(right->bb_numrecs); - rbp = bp; - if ((error = xfs_btree_read_bufs(mp, cur->bc_tp, - cur->bc_private.a.agno, lbno, 0, &lbp, - XFS_INO_BTREE_REF))) - return error; - left = XFS_BUF_TO_INOBT_BLOCK(lbp); - lrecs = be16_to_cpu(left->bb_numrecs); - if ((error = xfs_btree_check_sblock(cur, left, level, lbp))) - return error; - } - /* - * If that won't work, see if we can join with the right neighbor block. - */ - else if (rbno != NULLAGBLOCK && - rrecs + numrecs <= XFS_INOBT_BLOCK_MAXRECS(level, cur)) { - /* - * Set "left" to be the starting block, - * "right" to be the right neighbor. - */ - lbno = bno; - left = block; - lrecs = be16_to_cpu(left->bb_numrecs); - lbp = bp; - if ((error = xfs_btree_read_bufs(mp, cur->bc_tp, - cur->bc_private.a.agno, rbno, 0, &rbp, - XFS_INO_BTREE_REF))) - return error; - right = XFS_BUF_TO_INOBT_BLOCK(rbp); - rrecs = be16_to_cpu(right->bb_numrecs); - if ((error = xfs_btree_check_sblock(cur, right, level, rbp))) - return error; - } - /* - * Otherwise, we can't fix the imbalance. - * Just return. This is probably a logic error, but it's not fatal. - */ - else { - if (level > 0 && (error = xfs_inobt_decrement(cur, level, &i))) - return error; - *stat = 1; - return 0; - } - /* - * We're now going to join "left" and "right" by moving all the stuff - * in "right" to "left" and deleting "right". - */ - if (level > 0) { - /* - * It's a non-leaf. Move keys and pointers. - */ - lkp = XFS_INOBT_KEY_ADDR(left, lrecs + 1, cur); - lpp = XFS_INOBT_PTR_ADDR(left, lrecs + 1, cur); - rkp = XFS_INOBT_KEY_ADDR(right, 1, cur); - rpp = XFS_INOBT_PTR_ADDR(right, 1, cur); -#ifdef DEBUG - for (i = 0; i < rrecs; i++) { - if ((error = xfs_btree_check_sptr(cur, be32_to_cpu(rpp[i]), level))) - return error; - } -#endif - memcpy(lkp, rkp, rrecs * sizeof(*lkp)); - memcpy(lpp, rpp, rrecs * sizeof(*lpp)); - xfs_inobt_log_keys(cur, lbp, lrecs + 1, lrecs + rrecs); - xfs_inobt_log_ptrs(cur, lbp, lrecs + 1, lrecs + rrecs); - } else { - /* - * It's a leaf. Move records. - */ - lrp = XFS_INOBT_REC_ADDR(left, lrecs + 1, cur); - rrp = XFS_INOBT_REC_ADDR(right, 1, cur); - memcpy(lrp, rrp, rrecs * sizeof(*lrp)); - xfs_inobt_log_recs(cur, lbp, lrecs + 1, lrecs + rrecs); - } - /* - * If we joined with the left neighbor, set the buffer in the - * cursor to the left block, and fix up the index. - */ - if (bp != lbp) { - xfs_btree_setbuf(cur, level, lbp); - cur->bc_ptrs[level] += lrecs; - } - /* - * If we joined with the right neighbor and there's a level above - * us, increment the cursor at that level. - */ - else if (level + 1 < cur->bc_nlevels && - (error = xfs_alloc_increment(cur, level + 1, &i))) - return error; - /* - * Fix up the number of records in the surviving block. - */ - lrecs += rrecs; - left->bb_numrecs = cpu_to_be16(lrecs); - /* - * Fix up the right block pointer in the surviving block, and log it. - */ - left->bb_rightsib = right->bb_rightsib; - xfs_inobt_log_block(cur->bc_tp, lbp, XFS_BB_NUMRECS | XFS_BB_RIGHTSIB); - /* - * If there is a right sibling now, make it point to the - * remaining block. - */ - if (be32_to_cpu(left->bb_rightsib) != NULLAGBLOCK) { - xfs_inobt_block_t *rrblock; - xfs_buf_t *rrbp; +STATIC struct xfs_btree_cur * +xfs_inobt_dup_cursor( + struct xfs_btree_cur *cur) +{ + return xfs_inobt_init_cursor(cur->bc_mp, cur->bc_tp, + cur->bc_private.a.agbp, cur->bc_private.a.agno); +} - if ((error = xfs_btree_read_bufs(mp, cur->bc_tp, - cur->bc_private.a.agno, be32_to_cpu(left->bb_rightsib), 0, - &rrbp, XFS_INO_BTREE_REF))) - return error; - rrblock = XFS_BUF_TO_INOBT_BLOCK(rrbp); - if ((error = xfs_btree_check_sblock(cur, rrblock, level, rrbp))) - return error; - rrblock->bb_leftsib = cpu_to_be32(lbno); - xfs_inobt_log_block(cur->bc_tp, rrbp, XFS_BB_LEFTSIB); - } - /* - * Free the deleting block. - */ - if ((error = xfs_free_extent(cur->bc_tp, XFS_AGB_TO_FSB(mp, - cur->bc_private.a.agno, rbno), 1))) - return error; - xfs_trans_binval(cur->bc_tp, rbp); - /* - * Readjust the ptr at this level if it's not a leaf, since it's - * still pointing at the deletion point, which makes the cursor - * inconsistent. If this makes the ptr 0, the caller fixes it up. - * We can't use decrement because it would change the next level up. - */ - if (level > 0) - cur->bc_ptrs[level]--; - /* - * Return value means the next level up has something to do. - */ - *stat = 2; - return 0; +STATIC void +xfs_inobt_set_root( + struct xfs_btree_cur *cur, + union xfs_btree_ptr *nptr, + int inc) /* level change */ +{ + struct xfs_buf *agbp = cur->bc_private.a.agbp; + struct xfs_agi *agi = XFS_BUF_TO_AGI(agbp); -error0: - xfs_btree_del_cursor(tcur, XFS_BTREE_ERROR); - return error; + agi->agi_root = nptr->s; + be32_add_cpu(&agi->agi_level, inc); + xfs_ialloc_log_agi(cur->bc_tp, agbp, XFS_AGI_ROOT | XFS_AGI_LEVEL); } -/* - * Insert one record/level. Return information to the caller - * allowing the next level up to proceed if necessary. - */ -STATIC int /* error */ -xfs_inobt_insrec( - xfs_btree_cur_t *cur, /* btree cursor */ - int level, /* level to insert record at */ - xfs_agblock_t *bnop, /* i/o: block number inserted */ - xfs_inobt_rec_t *recp, /* i/o: record data inserted */ - xfs_btree_cur_t **curp, /* output: new cursor replacing cur */ - int *stat) /* success/failure */ +STATIC int +xfs_inobt_alloc_block( + struct xfs_btree_cur *cur, + union xfs_btree_ptr *start, + union xfs_btree_ptr *new, + int length, + int *stat) { - xfs_inobt_block_t *block; /* btree block record/key lives in */ - xfs_buf_t *bp; /* buffer for block */ - int error; /* error return value */ - int i; /* loop index */ - xfs_inobt_key_t key; /* key value being inserted */ - xfs_inobt_key_t *kp=NULL; /* pointer to btree keys */ - xfs_agblock_t nbno; /* block number of allocated block */ - xfs_btree_cur_t *ncur; /* new cursor to be used at next lvl */ - xfs_inobt_key_t nkey; /* new key value, from split */ - xfs_inobt_rec_t nrec; /* new record value, for caller */ - int numrecs; - int optr; /* old ptr value */ - xfs_inobt_ptr_t *pp; /* pointer to btree addresses */ - int ptr; /* index in btree block for this rec */ - xfs_inobt_rec_t *rp=NULL; /* pointer to btree records */ + xfs_alloc_arg_t args; /* block allocation args */ + int error; /* error return value */ + xfs_agblock_t sbno = be32_to_cpu(start->s); - /* - * GCC doesn't understand the (arguably complex) control flow in - * this function and complains about uninitialized structure fields - * without this. - */ - memset(&nrec, 0, sizeof(nrec)); + XFS_BTREE_TRACE_CURSOR(cur, XBT_ENTRY); - /* - * If we made it to the root level, allocate a new root block - * and we're done. - */ - if (level >= cur->bc_nlevels) { - error = xfs_inobt_newroot(cur, &i); - *bnop = NULLAGBLOCK; - *stat = i; + memset(&args, 0, sizeof(args)); + args.tp = cur->bc_tp; + args.mp = cur->bc_mp; + args.fsbno = XFS_AGB_TO_FSB(args.mp, cur->bc_private.a.agno, sbno); + args.minlen = 1; + args.maxlen = 1; + args.prod = 1; + args.type = XFS_ALLOCTYPE_NEAR_BNO; + + error = xfs_alloc_vextent(&args); + if (error) { + XFS_BTREE_TRACE_CURSOR(cur, XBT_ERROR); return error; } - /* - * Make a key out of the record data to be inserted, and save it. - */ - key.ir_startino = recp->ir_startino; - optr = ptr = cur->bc_ptrs[level]; - /* - * If we're off the left edge, return failure. - */ - if (ptr == 0) { + if (args.fsbno == NULLFSBLOCK) { + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); *stat = 0; return 0; } - /* - * Get pointers to the btree buffer and block. - */ - bp = cur->bc_bufs[level]; - block = XFS_BUF_TO_INOBT_BLOCK(bp); - numrecs = be16_to_cpu(block->bb_numrecs); -#ifdef DEBUG - if ((error = xfs_btree_check_sblock(cur, block, level, bp))) - return error; - /* - * Check that the new entry is being inserted in the right place. - */ - if (ptr <= numrecs) { - if (level == 0) { - rp = XFS_INOBT_REC_ADDR(block, ptr, cur); - xfs_btree_check_rec(cur->bc_btnum, recp, rp); - } else { - kp = XFS_INOBT_KEY_ADDR(block, ptr, cur); - xfs_btree_check_key(cur->bc_btnum, &key, kp); - } - } -#endif - nbno = NULLAGBLOCK; - ncur = NULL; - /* - * If the block is full, we can't insert the new entry until we - * make the block un-full. - */ - if (numrecs == XFS_INOBT_BLOCK_MAXRECS(level, cur)) { - /* - * First, try shifting an entry to the right neighbor. - */ - if ((error = xfs_inobt_rshift(cur, level, &i))) - return error; - if (i) { - /* nothing */ - } - /* - * Next, try shifting an entry to the left neighbor. - */ - else { - if ((error = xfs_inobt_lshift(cur, level, &i))) - return error; - if (i) { - optr = ptr = cur->bc_ptrs[level]; - } else { - /* - * Next, try splitting the current block - * in half. If this works we have to - * re-set our variables because - * we could be in a different block now. - */ - if ((error = xfs_inobt_split(cur, level, &nbno, - &nkey, &ncur, &i))) - return error; - if (i) { - bp = cur->bc_bufs[level]; - block = XFS_BUF_TO_INOBT_BLOCK(bp); -#ifdef DEBUG - if ((error = xfs_btree_check_sblock(cur, - block, level, bp))) - return error; -#endif - ptr = cur->bc_ptrs[level]; - nrec.ir_startino = nkey.ir_startino; - } else { - /* - * Otherwise the insert fails. - */ - *stat = 0; - return 0; - } - } - } - } - /* - * At this point we know there's room for our new entry in the block - * we're pointing at. - */ - numrecs = be16_to_cpu(block->bb_numrecs); - if (level > 0) { - /* - * It's a non-leaf entry. Make a hole for the new data - * in the key and ptr regions of the block. - */ - kp = XFS_INOBT_KEY_ADDR(block, 1, cur); - pp = XFS_INOBT_PTR_ADDR(block, 1, cur); -#ifdef DEBUG - for (i = numrecs; i >= ptr; i--) { - if ((error = xfs_btree_check_sptr(cur, be32_to_cpu(pp[i - 1]), level))) - return error; - } -#endif - memmove(&kp[ptr], &kp[ptr - 1], - (numrecs - ptr + 1) * sizeof(*kp)); - memmove(&pp[ptr], &pp[ptr - 1], - (numrecs - ptr + 1) * sizeof(*pp)); - /* - * Now stuff the new data in, bump numrecs and log the new data. - */ -#ifdef DEBUG - if ((error = xfs_btree_check_sptr(cur, *bnop, level))) - return error; -#endif - kp[ptr - 1] = key; - pp[ptr - 1] = cpu_to_be32(*bnop); - numrecs++; - block->bb_numrecs = cpu_to_be16(numrecs); - xfs_inobt_log_keys(cur, bp, ptr, numrecs); - xfs_inobt_log_ptrs(cur, bp, ptr, numrecs); - } else { - /* - * It's a leaf entry. Make a hole for the new record. - */ - rp = XFS_INOBT_REC_ADDR(block, 1, cur); - memmove(&rp[ptr], &rp[ptr - 1], - (numrecs - ptr + 1) * sizeof(*rp)); - /* - * Now stuff the new record in, bump numrecs - * and log the new data. - */ - rp[ptr - 1] = *recp; - numrecs++; - block->bb_numrecs = cpu_to_be16(numrecs); - xfs_inobt_log_recs(cur, bp, ptr, numrecs); - } - /* - * Log the new number of records in the btree header. - */ - xfs_inobt_log_block(cur->bc_tp, bp, XFS_BB_NUMRECS); -#ifdef DEBUG - /* - * Check that the key/record is in the right place, now. - */ - if (ptr < numrecs) { - if (level == 0) - xfs_btree_check_rec(cur->bc_btnum, rp + ptr - 1, - rp + ptr); - else - xfs_btree_check_key(cur->bc_btnum, kp + ptr - 1, - kp + ptr); - } -#endif - /* - * If we inserted at the start of a block, update the parents' keys. - */ - if (optr == 1 && (error = xfs_inobt_updkey(cur, &key, level + 1))) - return error; - /* - * Return the new block number, if any. - * If there is one, give back a record value and a cursor too. - */ - *bnop = nbno; - if (nbno != NULLAGBLOCK) { - *recp = nrec; - *curp = ncur; - } + ASSERT(args.len == 1); + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); + + new->s = cpu_to_be32(XFS_FSB_TO_AGBNO(args.mp, args.fsbno)); *stat = 1; return 0; } -/* - * Log header fields from a btree block. - */ -STATIC void -xfs_inobt_log_block( - xfs_trans_t *tp, /* transaction pointer */ - xfs_buf_t *bp, /* buffer containing btree block */ - int fields) /* mask of fields: XFS_BB_... */ +STATIC int +xfs_inobt_free_block( + struct xfs_btree_cur *cur, + struct xfs_buf *bp) { - int first; /* first byte offset logged */ - int last; /* last byte offset logged */ - static const short offsets[] = { /* table of offsets */ - offsetof(xfs_inobt_block_t, bb_magic), - offsetof(xfs_inobt_block_t, bb_level), - offsetof(xfs_inobt_block_t, bb_numrecs), - offsetof(xfs_inobt_block_t, bb_leftsib), - offsetof(xfs_inobt_block_t, bb_rightsib), - sizeof(xfs_inobt_block_t) - }; + xfs_fsblock_t fsbno; + int error; - xfs_btree_offsets(fields, offsets, XFS_BB_NUM_BITS, &first, &last); - xfs_trans_log_buf(tp, bp, first, last); + fsbno = XFS_DADDR_TO_FSB(cur->bc_mp, XFS_BUF_ADDR(bp)); + error = xfs_free_extent(cur->bc_tp, fsbno, 1); + if (error) + return error; + + xfs_trans_binval(cur->bc_tp, bp); + return error; } -/* - * Log keys from a btree block (nonleaf). - */ -STATIC void -xfs_inobt_log_keys( - xfs_btree_cur_t *cur, /* btree cursor */ - xfs_buf_t *bp, /* buffer containing btree block */ - int kfirst, /* index of first key to log */ - int klast) /* index of last key to log */ +STATIC int +xfs_inobt_get_maxrecs( + struct xfs_btree_cur *cur, + int level) { - xfs_inobt_block_t *block; /* btree block to log from */ - int first; /* first byte offset logged */ - xfs_inobt_key_t *kp; /* key pointer in btree block */ - int last; /* last byte offset logged */ - - block = XFS_BUF_TO_INOBT_BLOCK(bp); - kp = XFS_INOBT_KEY_ADDR(block, 1, cur); - first = (int)((xfs_caddr_t)&kp[kfirst - 1] - (xfs_caddr_t)block); - last = (int)(((xfs_caddr_t)&kp[klast] - 1) - (xfs_caddr_t)block); - xfs_trans_log_buf(cur->bc_tp, bp, first, last); + return cur->bc_mp->m_inobt_mxr[level != 0]; } -/* - * Log block pointer fields from a btree block (nonleaf). - */ STATIC void -xfs_inobt_log_ptrs( - xfs_btree_cur_t *cur, /* btree cursor */ - xfs_buf_t *bp, /* buffer containing btree block */ - int pfirst, /* index of first pointer to log */ - int plast) /* index of last pointer to log */ +xfs_inobt_init_key_from_rec( + union xfs_btree_key *key, + union xfs_btree_rec *rec) { - xfs_inobt_block_t *block; /* btree block to log from */ - int first; /* first byte offset logged */ - int last; /* last byte offset logged */ - xfs_inobt_ptr_t *pp; /* block-pointer pointer in btree blk */ - - block = XFS_BUF_TO_INOBT_BLOCK(bp); - pp = XFS_INOBT_PTR_ADDR(block, 1, cur); - first = (int)((xfs_caddr_t)&pp[pfirst - 1] - (xfs_caddr_t)block); - last = (int)(((xfs_caddr_t)&pp[plast] - 1) - (xfs_caddr_t)block); - xfs_trans_log_buf(cur->bc_tp, bp, first, last); + key->inobt.ir_startino = rec->inobt.ir_startino; } -/* - * Log records from a btree block (leaf). - */ STATIC void -xfs_inobt_log_recs( - xfs_btree_cur_t *cur, /* btree cursor */ - xfs_buf_t *bp, /* buffer containing btree block */ - int rfirst, /* index of first record to log */ - int rlast) /* index of last record to log */ +xfs_inobt_init_rec_from_key( + union xfs_btree_key *key, + union xfs_btree_rec *rec) { - xfs_inobt_block_t *block; /* btree block to log from */ - int first; /* first byte offset logged */ - int last; /* last byte offset logged */ - xfs_inobt_rec_t *rp; /* record pointer for btree block */ + rec->inobt.ir_startino = key->inobt.ir_startino; +} - block = XFS_BUF_TO_INOBT_BLOCK(bp); - rp = XFS_INOBT_REC_ADDR(block, 1, cur); - first = (int)((xfs_caddr_t)&rp[rfirst - 1] - (xfs_caddr_t)block); - last = (int)(((xfs_caddr_t)&rp[rlast] - 1) - (xfs_caddr_t)block); - xfs_trans_log_buf(cur->bc_tp, bp, first, last); +STATIC void +xfs_inobt_init_rec_from_cur( + struct xfs_btree_cur *cur, + union xfs_btree_rec *rec) +{ + rec->inobt.ir_startino = cpu_to_be32(cur->bc_rec.i.ir_startino); + rec->inobt.ir_freecount = cpu_to_be32(cur->bc_rec.i.ir_freecount); + rec->inobt.ir_free = cpu_to_be64(cur->bc_rec.i.ir_free); } /* - * Lookup the record. The cursor is made to point to it, based on dir. - * Return 0 if can't find any such record, 1 for success. + * intial value of ptr for lookup */ -STATIC int /* error */ -xfs_inobt_lookup( - xfs_btree_cur_t *cur, /* btree cursor */ - xfs_lookup_t dir, /* <=, ==, or >= */ - int *stat) /* success/failure */ +STATIC void +xfs_inobt_init_ptr_from_cur( + struct xfs_btree_cur *cur, + union xfs_btree_ptr *ptr) { - xfs_agblock_t agbno; /* a.g. relative btree block number */ - xfs_agnumber_t agno; /* allocation group number */ - xfs_inobt_block_t *block=NULL; /* current btree block */ - __int64_t diff; /* difference for the current key */ - int error; /* error return value */ - int keyno=0; /* current key number */ - int level; /* level in the btree */ - xfs_mount_t *mp; /* file system mount point */ - - /* - * Get the allocation group header, and the root block number. - */ - mp = cur->bc_mp; - { - xfs_agi_t *agi; /* a.g. inode header */ - - agi = XFS_BUF_TO_AGI(cur->bc_private.a.agbp); - agno = be32_to_cpu(agi->agi_seqno); - agbno = be32_to_cpu(agi->agi_root); - } - /* - * Iterate over each level in the btree, starting at the root. - * For each level above the leaves, find the key we need, based - * on the lookup record, then follow the corresponding block - * pointer down to the next level. - */ - for (level = cur->bc_nlevels - 1, diff = 1; level >= 0; level--) { - xfs_buf_t *bp; /* buffer pointer for btree block */ - xfs_daddr_t d; /* disk address of btree block */ - - /* - * Get the disk address we're looking for. - */ - d = XFS_AGB_TO_DADDR(mp, agno, agbno); - /* - * If the old buffer at this level is for a different block, - * throw it away, otherwise just use it. - */ - bp = cur->bc_bufs[level]; - if (bp && XFS_BUF_ADDR(bp) != d) - bp = NULL; - if (!bp) { - /* - * Need to get a new buffer. Read it, then - * set it in the cursor, releasing the old one. - */ - if ((error = xfs_btree_read_bufs(mp, cur->bc_tp, - agno, agbno, 0, &bp, XFS_INO_BTREE_REF))) - return error; - xfs_btree_setbuf(cur, level, bp); - /* - * Point to the btree block, now that we have the buffer - */ - block = XFS_BUF_TO_INOBT_BLOCK(bp); - if ((error = xfs_btree_check_sblock(cur, block, level, - bp))) - return error; - } else - block = XFS_BUF_TO_INOBT_BLOCK(bp); - /* - * If we already had a key match at a higher level, we know - * we need to use the first entry in this block. - */ - if (diff == 0) - keyno = 1; - /* - * Otherwise we need to search this block. Do a binary search. - */ - else { - int high; /* high entry number */ - xfs_inobt_key_t *kkbase=NULL;/* base of keys in block */ - xfs_inobt_rec_t *krbase=NULL;/* base of records in block */ - int low; /* low entry number */ + struct xfs_agi *agi = XFS_BUF_TO_AGI(cur->bc_private.a.agbp); - /* - * Get a pointer to keys or records. - */ - if (level > 0) - kkbase = XFS_INOBT_KEY_ADDR(block, 1, cur); - else - krbase = XFS_INOBT_REC_ADDR(block, 1, cur); - /* - * Set low and high entry numbers, 1-based. - */ - low = 1; - if (!(high = be16_to_cpu(block->bb_numrecs))) { - /* - * If the block is empty, the tree must - * be an empty leaf. - */ - ASSERT(level == 0 && cur->bc_nlevels == 1); - cur->bc_ptrs[0] = dir != XFS_LOOKUP_LE; - *stat = 0; - return 0; - } - /* - * Binary search the block. - */ - while (low <= high) { - xfs_agino_t startino; /* key value */ - - /* - * keyno is average of low and high. - */ - keyno = (low + high) >> 1; - /* - * Get startino. - */ - if (level > 0) { - xfs_inobt_key_t *kkp; - - kkp = kkbase + keyno - 1; - startino = be32_to_cpu(kkp->ir_startino); - } else { - xfs_inobt_rec_t *krp; - - krp = krbase + keyno - 1; - startino = be32_to_cpu(krp->ir_startino); - } - /* - * Compute difference to get next direction. - */ - diff = (__int64_t) - startino - cur->bc_rec.i.ir_startino; - /* - * Less than, move right. - */ - if (diff < 0) - low = keyno + 1; - /* - * Greater than, move left. - */ - else if (diff > 0) - high = keyno - 1; - /* - * Equal, we're done. - */ - else - break; - } - } - /* - * If there are more levels, set up for the next level - * by getting the block number and filling in the cursor. - */ - if (level > 0) { - /* - * If we moved left, need the previous key number, - * unless there isn't one. - */ - if (diff > 0 && --keyno < 1) - keyno = 1; - agbno = be32_to_cpu(*XFS_INOBT_PTR_ADDR(block, keyno, cur)); -#ifdef DEBUG - if ((error = xfs_btree_check_sptr(cur, agbno, level))) - return error; -#endif - cur->bc_ptrs[level] = keyno; - } - } - /* - * Done with the search. - * See if we need to adjust the results. - */ - if (dir != XFS_LOOKUP_LE && diff < 0) { - keyno++; - /* - * If ge search and we went off the end of the block, but it's - * not the last block, we're in the wrong block. - */ - if (dir == XFS_LOOKUP_GE && - keyno > be16_to_cpu(block->bb_numrecs) && - be32_to_cpu(block->bb_rightsib) != NULLAGBLOCK) { - int i; + ASSERT(cur->bc_private.a.agno == be32_to_cpu(agi->agi_seqno)); - cur->bc_ptrs[0] = keyno; - if ((error = xfs_inobt_increment(cur, 0, &i))) - return error; - ASSERT(i == 1); - *stat = 1; - return 0; - } - } - else if (dir == XFS_LOOKUP_LE && diff > 0) - keyno--; - cur->bc_ptrs[0] = keyno; - /* - * Return if we succeeded or not. - */ - if (keyno == 0 || keyno > be16_to_cpu(block->bb_numrecs)) - *stat = 0; - else - *stat = ((dir != XFS_LOOKUP_EQ) || (diff == 0)); - return 0; + ptr->s = agi->agi_root; } -/* - * Move 1 record left from cur/level if possible. - * Update cur to reflect the new path. - */ -STATIC int /* error */ -xfs_inobt_lshift( - xfs_btree_cur_t *cur, /* btree cursor */ - int level, /* level to shift record on */ - int *stat) /* success/failure */ +STATIC __int64_t +xfs_inobt_key_diff( + struct xfs_btree_cur *cur, + union xfs_btree_key *key) { - int error; /* error return value */ -#ifdef DEBUG - int i; /* loop index */ -#endif - xfs_inobt_key_t key; /* key value for leaf level upward */ - xfs_buf_t *lbp; /* buffer for left neighbor block */ - xfs_inobt_block_t *left; /* left neighbor btree block */ - xfs_inobt_key_t *lkp=NULL; /* key pointer for left block */ - xfs_inobt_ptr_t *lpp; /* address pointer for left block */ - xfs_inobt_rec_t *lrp=NULL; /* record pointer for left block */ - int nrec; /* new number of left block entries */ - xfs_buf_t *rbp; /* buffer for right (current) block */ - xfs_inobt_block_t *right; /* right (current) btree block */ - xfs_inobt_key_t *rkp=NULL; /* key pointer for right block */ - xfs_inobt_ptr_t *rpp=NULL; /* address pointer for right block */ - xfs_inobt_rec_t *rrp=NULL; /* record pointer for right block */ - - /* - * Set up variables for this block as "right". - */ - rbp = cur->bc_bufs[level]; - right = XFS_BUF_TO_INOBT_BLOCK(rbp); -#ifdef DEBUG - if ((error = xfs_btree_check_sblock(cur, right, level, rbp))) - return error; -#endif - /* - * If we've got no left sibling then we can't shift an entry left. - */ - if (be32_to_cpu(right->bb_leftsib) == NULLAGBLOCK) { - *stat = 0; - return 0; - } - /* - * If the cursor entry is the one that would be moved, don't - * do it... it's too complicated. - */ - if (cur->bc_ptrs[level] <= 1) { - *stat = 0; - return 0; - } - /* - * Set up the left neighbor as "left". - */ - if ((error = xfs_btree_read_bufs(cur->bc_mp, cur->bc_tp, - cur->bc_private.a.agno, be32_to_cpu(right->bb_leftsib), - 0, &lbp, XFS_INO_BTREE_REF))) - return error; - left = XFS_BUF_TO_INOBT_BLOCK(lbp); - if ((error = xfs_btree_check_sblock(cur, left, level, lbp))) - return error; - /* - * If it's full, it can't take another entry. - */ - if (be16_to_cpu(left->bb_numrecs) == XFS_INOBT_BLOCK_MAXRECS(level, cur)) { - *stat = 0; - return 0; - } - nrec = be16_to_cpu(left->bb_numrecs) + 1; - /* - * If non-leaf, copy a key and a ptr to the left block. - */ - if (level > 0) { - lkp = XFS_INOBT_KEY_ADDR(left, nrec, cur); - rkp = XFS_INOBT_KEY_ADDR(right, 1, cur); - *lkp = *rkp; - xfs_inobt_log_keys(cur, lbp, nrec, nrec); - lpp = XFS_INOBT_PTR_ADDR(left, nrec, cur); - rpp = XFS_INOBT_PTR_ADDR(right, 1, cur); -#ifdef DEBUG - if ((error = xfs_btree_check_sptr(cur, be32_to_cpu(*rpp), level))) - return error; -#endif - *lpp = *rpp; - xfs_inobt_log_ptrs(cur, lbp, nrec, nrec); - } - /* - * If leaf, copy a record to the left block. - */ - else { - lrp = XFS_INOBT_REC_ADDR(left, nrec, cur); - rrp = XFS_INOBT_REC_ADDR(right, 1, cur); - *lrp = *rrp; - xfs_inobt_log_recs(cur, lbp, nrec, nrec); - } - /* - * Bump and log left's numrecs, decrement and log right's numrecs. - */ - be16_add_cpu(&left->bb_numrecs, 1); - xfs_inobt_log_block(cur->bc_tp, lbp, XFS_BB_NUMRECS); -#ifdef DEBUG - if (level > 0) - xfs_btree_check_key(cur->bc_btnum, lkp - 1, lkp); - else - xfs_btree_check_rec(cur->bc_btnum, lrp - 1, lrp); -#endif - be16_add_cpu(&right->bb_numrecs, -1); - xfs_inobt_log_block(cur->bc_tp, rbp, XFS_BB_NUMRECS); - /* - * Slide the contents of right down one entry. - */ - if (level > 0) { -#ifdef DEBUG - for (i = 0; i < be16_to_cpu(right->bb_numrecs); i++) { - if ((error = xfs_btree_check_sptr(cur, be32_to_cpu(rpp[i + 1]), - level))) - return error; - } -#endif - memmove(rkp, rkp + 1, be16_to_cpu(right->bb_numrecs) * sizeof(*rkp)); - memmove(rpp, rpp + 1, be16_to_cpu(right->bb_numrecs) * sizeof(*rpp)); - xfs_inobt_log_keys(cur, rbp, 1, be16_to_cpu(right->bb_numrecs)); - xfs_inobt_log_ptrs(cur, rbp, 1, be16_to_cpu(right->bb_numrecs)); - } else { - memmove(rrp, rrp + 1, be16_to_cpu(right->bb_numrecs) * sizeof(*rrp)); - xfs_inobt_log_recs(cur, rbp, 1, be16_to_cpu(right->bb_numrecs)); - key.ir_startino = rrp->ir_startino; - rkp = &key; - } - /* - * Update the parent key values of right. - */ - if ((error = xfs_inobt_updkey(cur, rkp, level + 1))) - return error; - /* - * Slide the cursor value left one. - */ - cur->bc_ptrs[level]--; - *stat = 1; - return 0; + return (__int64_t)be32_to_cpu(key->inobt.ir_startino) - + cur->bc_rec.i.ir_startino; } -/* - * Allocate a new root block, fill it in. - */ -STATIC int /* error */ -xfs_inobt_newroot( - xfs_btree_cur_t *cur, /* btree cursor */ - int *stat) /* success/failure */ +STATIC int +xfs_inobt_kill_root( + struct xfs_btree_cur *cur, + struct xfs_buf *bp, + int level, + union xfs_btree_ptr *newroot) { - xfs_agi_t *agi; /* a.g. inode header */ - xfs_alloc_arg_t args; /* allocation argument structure */ - xfs_inobt_block_t *block; /* one half of the old root block */ - xfs_buf_t *bp; /* buffer containing block */ - int error; /* error return value */ - xfs_inobt_key_t *kp; /* btree key pointer */ - xfs_agblock_t lbno; /* left block number */ - xfs_buf_t *lbp; /* left buffer pointer */ - xfs_inobt_block_t *left; /* left btree block */ - xfs_buf_t *nbp; /* new (root) buffer */ - xfs_inobt_block_t *new; /* new (root) btree block */ - int nptr; /* new value for key index, 1 or 2 */ - xfs_inobt_ptr_t *pp; /* btree address pointer */ - xfs_agblock_t rbno; /* right block number */ - xfs_buf_t *rbp; /* right buffer pointer */ - xfs_inobt_block_t *right; /* right btree block */ - xfs_inobt_rec_t *rp; /* btree record pointer */ + int error; - ASSERT(cur->bc_nlevels < XFS_IN_MAXLEVELS(cur->bc_mp)); + XFS_BTREE_TRACE_CURSOR(cur, XBT_ENTRY); + XFS_BTREE_STATS_INC(cur, killroot); /* - * Get a block & a buffer. + * Update the root pointer, decreasing the level by 1 and then + * free the old root. */ - agi = XFS_BUF_TO_AGI(cur->bc_private.a.agbp); - args.tp = cur->bc_tp; - args.mp = cur->bc_mp; - args.fsbno = XFS_AGB_TO_FSB(args.mp, cur->bc_private.a.agno, - be32_to_cpu(agi->agi_root)); - args.mod = args.minleft = args.alignment = args.total = args.wasdel = - args.isfl = args.userdata = args.minalignslop = 0; - args.minlen = args.maxlen = args.prod = 1; - args.type = XFS_ALLOCTYPE_NEAR_BNO; - if ((error = xfs_alloc_vextent(&args))) + xfs_inobt_set_root(cur, newroot, -1); + error = xfs_inobt_free_block(cur, bp); + if (error) { + XFS_BTREE_TRACE_CURSOR(cur, XBT_ERROR); return error; - /* - * None available, we fail. - */ - if (args.fsbno == NULLFSBLOCK) { - *stat = 0; - return 0; - } - ASSERT(args.len == 1); - nbp = xfs_btree_get_bufs(args.mp, args.tp, args.agno, args.agbno, 0); - new = XFS_BUF_TO_INOBT_BLOCK(nbp); - /* - * Set the root data in the a.g. inode structure. - */ - agi->agi_root = cpu_to_be32(args.agbno); - be32_add_cpu(&agi->agi_level, 1); - xfs_ialloc_log_agi(args.tp, cur->bc_private.a.agbp, - XFS_AGI_ROOT | XFS_AGI_LEVEL); - /* - * At the previous root level there are now two blocks: the old - * root, and the new block generated when it was split. - * We don't know which one the cursor is pointing at, so we - * set up variables "left" and "right" for each case. - */ - bp = cur->bc_bufs[cur->bc_nlevels - 1]; - block = XFS_BUF_TO_INOBT_BLOCK(bp); -#ifdef DEBUG - if ((error = xfs_btree_check_sblock(cur, block, cur->bc_nlevels - 1, bp))) - return error; -#endif - if (be32_to_cpu(block->bb_rightsib) != NULLAGBLOCK) { - /* - * Our block is left, pick up the right block. - */ - lbp = bp; - lbno = XFS_DADDR_TO_AGBNO(args.mp, XFS_BUF_ADDR(lbp)); - left = block; - rbno = be32_to_cpu(left->bb_rightsib); - if ((error = xfs_btree_read_bufs(args.mp, args.tp, args.agno, - rbno, 0, &rbp, XFS_INO_BTREE_REF))) - return error; - bp = rbp; - right = XFS_BUF_TO_INOBT_BLOCK(rbp); - if ((error = xfs_btree_check_sblock(cur, right, - cur->bc_nlevels - 1, rbp))) - return error; - nptr = 1; - } else { - /* - * Our block is right, pick up the left block. - */ - rbp = bp; - rbno = XFS_DADDR_TO_AGBNO(args.mp, XFS_BUF_ADDR(rbp)); - right = block; - lbno = be32_to_cpu(right->bb_leftsib); - if ((error = xfs_btree_read_bufs(args.mp, args.tp, args.agno, - lbno, 0, &lbp, XFS_INO_BTREE_REF))) - return error; - bp = lbp; - left = XFS_BUF_TO_INOBT_BLOCK(lbp); - if ((error = xfs_btree_check_sblock(cur, left, - cur->bc_nlevels - 1, lbp))) - return error; - nptr = 2; - } - /* - * Fill in the new block's btree header and log it. - */ - new->bb_magic = cpu_to_be32(xfs_magics[cur->bc_btnum]); - new->bb_level = cpu_to_be16(cur->bc_nlevels); - new->bb_numrecs = cpu_to_be16(2); - new->bb_leftsib = cpu_to_be32(NULLAGBLOCK); - new->bb_rightsib = cpu_to_be32(NULLAGBLOCK); - xfs_inobt_log_block(args.tp, nbp, XFS_BB_ALL_BITS); - ASSERT(lbno != NULLAGBLOCK && rbno != NULLAGBLOCK); - /* - * Fill in the key data in the new root. - */ - kp = XFS_INOBT_KEY_ADDR(new, 1, cur); - if (be16_to_cpu(left->bb_level) > 0) { - kp[0] = *XFS_INOBT_KEY_ADDR(left, 1, cur); - kp[1] = *XFS_INOBT_KEY_ADDR(right, 1, cur); - } else { - rp = XFS_INOBT_REC_ADDR(left, 1, cur); - kp[0].ir_startino = rp->ir_startino; - rp = XFS_INOBT_REC_ADDR(right, 1, cur); - kp[1].ir_startino = rp->ir_startino; } - xfs_inobt_log_keys(cur, nbp, 1, 2); - /* - * Fill in the pointer data in the new root. - */ - pp = XFS_INOBT_PTR_ADDR(new, 1, cur); - pp[0] = cpu_to_be32(lbno); - pp[1] = cpu_to_be32(rbno); - xfs_inobt_log_ptrs(cur, nbp, 1, 2); - /* - * Fix up the cursor. - */ - xfs_btree_setbuf(cur, cur->bc_nlevels, nbp); - cur->bc_ptrs[cur->bc_nlevels] = nptr; - cur->bc_nlevels++; - *stat = 1; - return 0; -} -/* - * Move 1 record right from cur/level if possible. - * Update cur to reflect the new path. - */ -STATIC int /* error */ -xfs_inobt_rshift( - xfs_btree_cur_t *cur, /* btree cursor */ - int level, /* level to shift record on */ - int *stat) /* success/failure */ -{ - int error; /* error return value */ - int i; /* loop index */ - xfs_inobt_key_t key; /* key value for leaf level upward */ - xfs_buf_t *lbp; /* buffer for left (current) block */ - xfs_inobt_block_t *left; /* left (current) btree block */ - xfs_inobt_key_t *lkp; /* key pointer for left block */ - xfs_inobt_ptr_t *lpp; /* address pointer for left block */ - xfs_inobt_rec_t *lrp; /* record pointer for left block */ - xfs_buf_t *rbp; /* buffer for right neighbor block */ - xfs_inobt_block_t *right; /* right neighbor btree block */ - xfs_inobt_key_t *rkp; /* key pointer for right block */ - xfs_inobt_ptr_t *rpp; /* address pointer for right block */ - xfs_inobt_rec_t *rrp=NULL; /* record pointer for right block */ - xfs_btree_cur_t *tcur; /* temporary cursor */ + XFS_BTREE_STATS_INC(cur, free); - /* - * Set up variables for this block as "left". - */ - lbp = cur->bc_bufs[level]; - left = XFS_BUF_TO_INOBT_BLOCK(lbp); -#ifdef DEBUG - if ((error = xfs_btree_check_sblock(cur, left, level, lbp))) - return error; -#endif - /* - * If we've got no right sibling then we can't shift an entry right. - */ - if (be32_to_cpu(left->bb_rightsib) == NULLAGBLOCK) { - *stat = 0; - return 0; - } - /* - * If the cursor entry is the one that would be moved, don't - * do it... it's too complicated. - */ - if (cur->bc_ptrs[level] >= be16_to_cpu(left->bb_numrecs)) { - *stat = 0; - return 0; - } - /* - * Set up the right neighbor as "right". - */ - if ((error = xfs_btree_read_bufs(cur->bc_mp, cur->bc_tp, - cur->bc_private.a.agno, be32_to_cpu(left->bb_rightsib), - 0, &rbp, XFS_INO_BTREE_REF))) - return error; - right = XFS_BUF_TO_INOBT_BLOCK(rbp); - if ((error = xfs_btree_check_sblock(cur, right, level, rbp))) - return error; - /* - * If it's full, it can't take another entry. - */ - if (be16_to_cpu(right->bb_numrecs) == XFS_INOBT_BLOCK_MAXRECS(level, cur)) { - *stat = 0; - return 0; - } - /* - * Make a hole at the start of the right neighbor block, then - * copy the last left block entry to the hole. - */ - if (level > 0) { - lkp = XFS_INOBT_KEY_ADDR(left, be16_to_cpu(left->bb_numrecs), cur); - lpp = XFS_INOBT_PTR_ADDR(left, be16_to_cpu(left->bb_numrecs), cur); - rkp = XFS_INOBT_KEY_ADDR(right, 1, cur); - rpp = XFS_INOBT_PTR_ADDR(right, 1, cur); -#ifdef DEBUG - for (i = be16_to_cpu(right->bb_numrecs) - 1; i >= 0; i--) { - if ((error = xfs_btree_check_sptr(cur, be32_to_cpu(rpp[i]), level))) - return error; - } -#endif - memmove(rkp + 1, rkp, be16_to_cpu(right->bb_numrecs) * sizeof(*rkp)); - memmove(rpp + 1, rpp, be16_to_cpu(right->bb_numrecs) * sizeof(*rpp)); -#ifdef DEBUG - if ((error = xfs_btree_check_sptr(cur, be32_to_cpu(*lpp), level))) - return error; -#endif - *rkp = *lkp; - *rpp = *lpp; - xfs_inobt_log_keys(cur, rbp, 1, be16_to_cpu(right->bb_numrecs) + 1); - xfs_inobt_log_ptrs(cur, rbp, 1, be16_to_cpu(right->bb_numrecs) + 1); - } else { - lrp = XFS_INOBT_REC_ADDR(left, be16_to_cpu(left->bb_numrecs), cur); - rrp = XFS_INOBT_REC_ADDR(right, 1, cur); - memmove(rrp + 1, rrp, be16_to_cpu(right->bb_numrecs) * sizeof(*rrp)); - *rrp = *lrp; - xfs_inobt_log_recs(cur, rbp, 1, be16_to_cpu(right->bb_numrecs) + 1); - key.ir_startino = rrp->ir_startino; - rkp = &key; - } - /* - * Decrement and log left's numrecs, bump and log right's numrecs. - */ - be16_add_cpu(&left->bb_numrecs, -1); - xfs_inobt_log_block(cur->bc_tp, lbp, XFS_BB_NUMRECS); - be16_add_cpu(&right->bb_numrecs, 1); -#ifdef DEBUG - if (level > 0) - xfs_btree_check_key(cur->bc_btnum, rkp, rkp + 1); - else - xfs_btree_check_rec(cur->bc_btnum, rrp, rrp + 1); -#endif - xfs_inobt_log_block(cur->bc_tp, rbp, XFS_BB_NUMRECS); - /* - * Using a temporary cursor, update the parent key values of the - * block on the right. - */ - if ((error = xfs_btree_dup_cursor(cur, &tcur))) - return error; - xfs_btree_lastrec(tcur, level); - if ((error = xfs_inobt_increment(tcur, level, &i)) || - (error = xfs_inobt_updkey(tcur, rkp, level + 1))) { - xfs_btree_del_cursor(tcur, XFS_BTREE_ERROR); - return error; - } - xfs_btree_del_cursor(tcur, XFS_BTREE_NOERROR); - *stat = 1; + cur->bc_bufs[level] = NULL; + cur->bc_nlevels--; + + XFS_BTREE_TRACE_CURSOR(cur, XBT_EXIT); return 0; } -/* - * Split cur/level block in half. - * Return new block number and its first record (to be inserted into parent). - */ -STATIC int /* error */ -xfs_inobt_split( - xfs_btree_cur_t *cur, /* btree cursor */ - int level, /* level to split */ - xfs_agblock_t *bnop, /* output: block number allocated */ - xfs_inobt_key_t *keyp, /* output: first key of new block */ - xfs_btree_cur_t **curp, /* output: new cursor */ - int *stat) /* success/failure */ -{ - xfs_alloc_arg_t args; /* allocation argument structure */ - int error; /* error return value */ - int i; /* loop index/record number */ - xfs_agblock_t lbno; /* left (current) block number */ - xfs_buf_t *lbp; /* buffer for left block */ - xfs_inobt_block_t *left; /* left (current) btree block */ - xfs_inobt_key_t *lkp; /* left btree key pointer */ - xfs_inobt_ptr_t *lpp; /* left btree address pointer */ - xfs_inobt_rec_t *lrp; /* left btree record pointer */ - xfs_buf_t *rbp; /* buffer for right block */ - xfs_inobt_block_t *right; /* right (new) btree block */ - xfs_inobt_key_t *rkp; /* right btree key pointer */ - xfs_inobt_ptr_t *rpp; /* right btree address pointer */ - xfs_inobt_rec_t *rrp; /* right btree record pointer */ - - /* - * Set up left block (current one). - */ - lbp = cur->bc_bufs[level]; - args.tp = cur->bc_tp; - args.mp = cur->bc_mp; - lbno = XFS_DADDR_TO_AGBNO(args.mp, XFS_BUF_ADDR(lbp)); - /* - * Allocate the new block. - * If we can't do it, we're toast. Give up. - */ - args.fsbno = XFS_AGB_TO_FSB(args.mp, cur->bc_private.a.agno, lbno); - args.mod = args.minleft = args.alignment = args.total = args.wasdel = - args.isfl = args.userdata = args.minalignslop = 0; - args.minlen = args.maxlen = args.prod = 1; - args.type = XFS_ALLOCTYPE_NEAR_BNO; - if ((error = xfs_alloc_vextent(&args))) - return error; - if (args.fsbno == NULLFSBLOCK) { - *stat = 0; - return 0; - } - ASSERT(args.len == 1); - rbp = xfs_btree_get_bufs(args.mp, args.tp, args.agno, args.agbno, 0); - /* - * Set up the new block as "right". - */ - right = XFS_BUF_TO_INOBT_BLOCK(rbp); - /* - * "Left" is the current (according to the cursor) block. - */ - left = XFS_BUF_TO_INOBT_BLOCK(lbp); #ifdef DEBUG - if ((error = xfs_btree_check_sblock(cur, left, level, lbp))) - return error; -#endif - /* - * Fill in the btree header for the new block. - */ - right->bb_magic = cpu_to_be32(xfs_magics[cur->bc_btnum]); - right->bb_level = left->bb_level; - right->bb_numrecs = cpu_to_be16(be16_to_cpu(left->bb_numrecs) / 2); - /* - * Make sure that if there's an odd number of entries now, that - * each new block will have the same number of entries. - */ - if ((be16_to_cpu(left->bb_numrecs) & 1) && - cur->bc_ptrs[level] <= be16_to_cpu(right->bb_numrecs) + 1) - be16_add_cpu(&right->bb_numrecs, 1); - i = be16_to_cpu(left->bb_numrecs) - be16_to_cpu(right->bb_numrecs) + 1; - /* - * For non-leaf blocks, copy keys and addresses over to the new block. - */ - if (level > 0) { - lkp = XFS_INOBT_KEY_ADDR(left, i, cur); - lpp = XFS_INOBT_PTR_ADDR(left, i, cur); - rkp = XFS_INOBT_KEY_ADDR(right, 1, cur); - rpp = XFS_INOBT_PTR_ADDR(right, 1, cur); -#ifdef DEBUG - for (i = 0; i < be16_to_cpu(right->bb_numrecs); i++) { - if ((error = xfs_btree_check_sptr(cur, be32_to_cpu(lpp[i]), level))) - return error; - } -#endif - memcpy(rkp, lkp, be16_to_cpu(right->bb_numrecs) * sizeof(*rkp)); - memcpy(rpp, lpp, be16_to_cpu(right->bb_numrecs) * sizeof(*rpp)); - xfs_inobt_log_keys(cur, rbp, 1, be16_to_cpu(right->bb_numrecs)); - xfs_inobt_log_ptrs(cur, rbp, 1, be16_to_cpu(right->bb_numrecs)); - *keyp = *rkp; - } - /* - * For leaf blocks, copy records over to the new block. - */ - else { - lrp = XFS_INOBT_REC_ADDR(left, i, cur); - rrp = XFS_INOBT_REC_ADDR(right, 1, cur); - memcpy(rrp, lrp, be16_to_cpu(right->bb_numrecs) * sizeof(*rrp)); - xfs_inobt_log_recs(cur, rbp, 1, be16_to_cpu(right->bb_numrecs)); - keyp->ir_startino = rrp->ir_startino; - } - /* - * Find the left block number by looking in the buffer. - * Adjust numrecs, sibling pointers. - */ - be16_add_cpu(&left->bb_numrecs, -(be16_to_cpu(right->bb_numrecs))); - right->bb_rightsib = left->bb_rightsib; - left->bb_rightsib = cpu_to_be32(args.agbno); - right->bb_leftsib = cpu_to_be32(lbno); - xfs_inobt_log_block(args.tp, rbp, XFS_BB_ALL_BITS); - xfs_inobt_log_block(args.tp, lbp, XFS_BB_NUMRECS | XFS_BB_RIGHTSIB); - /* - * If there's a block to the new block's right, make that block - * point back to right instead of to left. - */ - if (be32_to_cpu(right->bb_rightsib) != NULLAGBLOCK) { - xfs_inobt_block_t *rrblock; /* rr btree block */ - xfs_buf_t *rrbp; /* buffer for rrblock */ - - if ((error = xfs_btree_read_bufs(args.mp, args.tp, args.agno, - be32_to_cpu(right->bb_rightsib), 0, &rrbp, - XFS_INO_BTREE_REF))) - return error; - rrblock = XFS_BUF_TO_INOBT_BLOCK(rrbp); - if ((error = xfs_btree_check_sblock(cur, rrblock, level, rrbp))) - return error; - rrblock->bb_leftsib = cpu_to_be32(args.agbno); - xfs_inobt_log_block(args.tp, rrbp, XFS_BB_LEFTSIB); - } - /* - * If the cursor is really in the right block, move it there. - * If it's just pointing past the last entry in left, then we'll - * insert there, so don't change anything in that case. - */ - if (cur->bc_ptrs[level] > be16_to_cpu(left->bb_numrecs) + 1) { - xfs_btree_setbuf(cur, level, rbp); - cur->bc_ptrs[level] -= be16_to_cpu(left->bb_numrecs); - } - /* - * If there are more levels, we'll need another cursor which refers - * the right block, no matter where this cursor was. - */ - if (level + 1 < cur->bc_nlevels) { - if ((error = xfs_btree_dup_cursor(cur, curp))) - return error; - (*curp)->bc_ptrs[level + 1]++; - } - *bnop = args.agbno; - *stat = 1; - return 0; +STATIC int +xfs_inobt_keys_inorder( + struct xfs_btree_cur *cur, + union xfs_btree_key *k1, + union xfs_btree_key *k2) +{ + return be32_to_cpu(k1->inobt.ir_startino) < + be32_to_cpu(k2->inobt.ir_startino); } -/* - * Update keys at all levels from here to the root along the cursor's path. - */ -STATIC int /* error */ -xfs_inobt_updkey( - xfs_btree_cur_t *cur, /* btree cursor */ - xfs_inobt_key_t *keyp, /* new key value to update to */ - int level) /* starting level for update */ +STATIC int +xfs_inobt_recs_inorder( + struct xfs_btree_cur *cur, + union xfs_btree_rec *r1, + union xfs_btree_rec *r2) { - int ptr; /* index of key in block */ - - /* - * Go up the tree from this level toward the root. - * At each level, update the key value to the value input. - * Stop when we reach a level where the cursor isn't pointing - * at the first entry in the block. - */ - for (ptr = 1; ptr == 1 && level < cur->bc_nlevels; level++) { - xfs_buf_t *bp; /* buffer for block */ - xfs_inobt_block_t *block; /* btree block */ -#ifdef DEBUG - int error; /* error return value */ -#endif - xfs_inobt_key_t *kp; /* ptr to btree block keys */ - - bp = cur->bc_bufs[level]; - block = XFS_BUF_TO_INOBT_BLOCK(bp); -#ifdef DEBUG - if ((error = xfs_btree_check_sblock(cur, block, level, bp))) - return error; -#endif - ptr = cur->bc_ptrs[level]; - kp = XFS_INOBT_KEY_ADDR(block, ptr, cur); - *kp = *keyp; - xfs_inobt_log_keys(cur, bp, ptr, ptr); - } - return 0; + return be32_to_cpu(r1->inobt.ir_startino) + XFS_INODES_PER_CHUNK <= + be32_to_cpu(r2->inobt.ir_startino); } +#endif /* DEBUG */ -/* - * Externally visible routines. - */ +#ifdef XFS_BTREE_TRACE +ktrace_t *xfs_inobt_trace_buf; -/* - * Decrement cursor by one record at the level. - * For nonzero levels the leaf-ward information is untouched. - */ -int /* error */ -xfs_inobt_decrement( - xfs_btree_cur_t *cur, /* btree cursor */ - int level, /* level in btree, 0 is leaf */ - int *stat) /* success/failure */ +STATIC void +xfs_inobt_trace_enter( + struct xfs_btree_cur *cur, + const char *func, + char *s, + int type, + int line, + __psunsigned_t a0, + __psunsigned_t a1, + __psunsigned_t a2, + __psunsigned_t a3, + __psunsigned_t a4, + __psunsigned_t a5, + __psunsigned_t a6, + __psunsigned_t a7, + __psunsigned_t a8, + __psunsigned_t a9, + __psunsigned_t a10) { - xfs_inobt_block_t *block; /* btree block */ - int error; - int lev; /* btree level */ - - ASSERT(level < cur->bc_nlevels); - /* - * Read-ahead to the left at this level. - */ - xfs_btree_readahead(cur, level, XFS_BTCUR_LEFTRA); - /* - * Decrement the ptr at this level. If we're still in the block - * then we're done. - */ - if (--cur->bc_ptrs[level] > 0) { - *stat = 1; - return 0; - } - /* - * Get a pointer to the btree block. - */ - block = XFS_BUF_TO_INOBT_BLOCK(cur->bc_bufs[level]); -#ifdef DEBUG - if ((error = xfs_btree_check_sblock(cur, block, level, - cur->bc_bufs[level]))) - return error; -#endif - /* - * If we just went off the left edge of the tree, return failure. - */ - if (be32_to_cpu(block->bb_leftsib) == NULLAGBLOCK) { - *stat = 0; - return 0; - } - /* - * March up the tree decrementing pointers. - * Stop when we don't go off the left edge of a block. - */ - for (lev = level + 1; lev < cur->bc_nlevels; lev++) { - if (--cur->bc_ptrs[lev] > 0) - break; - /* - * Read-ahead the left block, we're going to read it - * in the next loop. - */ - xfs_btree_readahead(cur, lev, XFS_BTCUR_LEFTRA); - } - /* - * If we went off the root then we are seriously confused. - */ - ASSERT(lev < cur->bc_nlevels); - /* - * Now walk back down the tree, fixing up the cursor's buffer - * pointers and key numbers. - */ - for (block = XFS_BUF_TO_INOBT_BLOCK(cur->bc_bufs[lev]); lev > level; ) { - xfs_agblock_t agbno; /* block number of btree block */ - xfs_buf_t *bp; /* buffer containing btree block */ - - agbno = be32_to_cpu(*XFS_INOBT_PTR_ADDR(block, cur->bc_ptrs[lev], cur)); - if ((error = xfs_btree_read_bufs(cur->bc_mp, cur->bc_tp, - cur->bc_private.a.agno, agbno, 0, &bp, - XFS_INO_BTREE_REF))) - return error; - lev--; - xfs_btree_setbuf(cur, lev, bp); - block = XFS_BUF_TO_INOBT_BLOCK(bp); - if ((error = xfs_btree_check_sblock(cur, block, lev, bp))) - return error; - cur->bc_ptrs[lev] = be16_to_cpu(block->bb_numrecs); - } - *stat = 1; - return 0; + ktrace_enter(xfs_inobt_trace_buf, (void *)(__psint_t)type, + (void *)func, (void *)s, NULL, (void *)cur, + (void *)a0, (void *)a1, (void *)a2, (void *)a3, + (void *)a4, (void *)a5, (void *)a6, (void *)a7, + (void *)a8, (void *)a9, (void *)a10); } -/* - * Delete the record pointed to by cur. - * The cursor refers to the place where the record was (could be inserted) - * when the operation returns. - */ -int /* error */ -xfs_inobt_delete( - xfs_btree_cur_t *cur, /* btree cursor */ - int *stat) /* success/failure */ +STATIC void +xfs_inobt_trace_cursor( + struct xfs_btree_cur *cur, + __uint32_t *s0, + __uint64_t *l0, + __uint64_t *l1) { - int error; - int i; /* result code */ - int level; /* btree level */ - - /* - * Go up the tree, starting at leaf level. - * If 2 is returned then a join was done; go to the next level. - * Otherwise we are done. - */ - for (level = 0, i = 2; i == 2; level++) { - if ((error = xfs_inobt_delrec(cur, level, &i))) - return error; - } - if (i == 0) { - for (level = 1; level < cur->bc_nlevels; level++) { - if (cur->bc_ptrs[level] == 0) { - if ((error = xfs_inobt_decrement(cur, level, &i))) - return error; - break; - } - } - } - *stat = i; - return 0; + *s0 = cur->bc_private.a.agno; + *l0 = cur->bc_rec.i.ir_startino; + *l1 = cur->bc_rec.i.ir_free; } - -/* - * Get the data from the pointed-to record. - */ -int /* error */ -xfs_inobt_get_rec( - xfs_btree_cur_t *cur, /* btree cursor */ - xfs_agino_t *ino, /* output: starting inode of chunk */ - __int32_t *fcnt, /* output: number of free inodes */ - xfs_inofree_t *free, /* output: free inode mask */ - int *stat) /* output: success/failure */ +STATIC void +xfs_inobt_trace_key( + struct xfs_btree_cur *cur, + union xfs_btree_key *key, + __uint64_t *l0, + __uint64_t *l1) { - xfs_inobt_block_t *block; /* btree block */ - xfs_buf_t *bp; /* buffer containing btree block */ -#ifdef DEBUG - int error; /* error return value */ -#endif - int ptr; /* record number */ - xfs_inobt_rec_t *rec; /* record data */ - - bp = cur->bc_bufs[0]; - ptr = cur->bc_ptrs[0]; - block = XFS_BUF_TO_INOBT_BLOCK(bp); -#ifdef DEBUG - if ((error = xfs_btree_check_sblock(cur, block, 0, bp))) - return error; -#endif - /* - * Off the right end or left end, return failure. - */ - if (ptr > be16_to_cpu(block->bb_numrecs) || ptr <= 0) { - *stat = 0; - return 0; - } - /* - * Point to the record and extract its data. - */ - rec = XFS_INOBT_REC_ADDR(block, ptr, cur); - *ino = be32_to_cpu(rec->ir_startino); - *fcnt = be32_to_cpu(rec->ir_freecount); - *free = be64_to_cpu(rec->ir_free); - *stat = 1; - return 0; + *l0 = be32_to_cpu(key->inobt.ir_startino); + *l1 = 0; } -/* - * Increment cursor by one record at the level. - * For nonzero levels the leaf-ward information is untouched. - */ -int /* error */ -xfs_inobt_increment( - xfs_btree_cur_t *cur, /* btree cursor */ - int level, /* level in btree, 0 is leaf */ - int *stat) /* success/failure */ +STATIC void +xfs_inobt_trace_record( + struct xfs_btree_cur *cur, + union xfs_btree_rec *rec, + __uint64_t *l0, + __uint64_t *l1, + __uint64_t *l2) { - xfs_inobt_block_t *block; /* btree block */ - xfs_buf_t *bp; /* buffer containing btree block */ - int error; /* error return value */ - int lev; /* btree level */ + *l0 = be32_to_cpu(rec->inobt.ir_startino); + *l1 = be32_to_cpu(rec->inobt.ir_freecount); + *l2 = be64_to_cpu(rec->inobt.ir_free); +} +#endif /* XFS_BTREE_TRACE */ + +static const struct xfs_btree_ops xfs_inobt_ops = { + .rec_len = sizeof(xfs_inobt_rec_t), + .key_len = sizeof(xfs_inobt_key_t), + + .dup_cursor = xfs_inobt_dup_cursor, + .set_root = xfs_inobt_set_root, + .kill_root = xfs_inobt_kill_root, + .alloc_block = xfs_inobt_alloc_block, + .free_block = xfs_inobt_free_block, + .get_minrecs = xfs_inobt_get_minrecs, + .get_maxrecs = xfs_inobt_get_maxrecs, + .init_key_from_rec = xfs_inobt_init_key_from_rec, + .init_rec_from_key = xfs_inobt_init_rec_from_key, + .init_rec_from_cur = xfs_inobt_init_rec_from_cur, + .init_ptr_from_cur = xfs_inobt_init_ptr_from_cur, + .key_diff = xfs_inobt_key_diff, - ASSERT(level < cur->bc_nlevels); - /* - * Read-ahead to the right at this level. - */ - xfs_btree_readahead(cur, level, XFS_BTCUR_RIGHTRA); - /* - * Get a pointer to the btree block. - */ - bp = cur->bc_bufs[level]; - block = XFS_BUF_TO_INOBT_BLOCK(bp); -#ifdef DEBUG - if ((error = xfs_btree_check_sblock(cur, block, level, bp))) - return error; -#endif - /* - * Increment the ptr at this level. If we're still in the block - * then we're done. - */ - if (++cur->bc_ptrs[level] <= be16_to_cpu(block->bb_numrecs)) { - *stat = 1; - return 0; - } - /* - * If we just went off the right edge of the tree, return failure. - */ - if (be32_to_cpu(block->bb_rightsib) == NULLAGBLOCK) { - *stat = 0; - return 0; - } - /* - * March up the tree incrementing pointers. - * Stop when we don't go off the right edge of a block. - */ - for (lev = level + 1; lev < cur->bc_nlevels; lev++) { - bp = cur->bc_bufs[lev]; - block = XFS_BUF_TO_INOBT_BLOCK(bp); #ifdef DEBUG - if ((error = xfs_btree_check_sblock(cur, block, lev, bp))) - return error; + .keys_inorder = xfs_inobt_keys_inorder, + .recs_inorder = xfs_inobt_recs_inorder, #endif - if (++cur->bc_ptrs[lev] <= be16_to_cpu(block->bb_numrecs)) - break; - /* - * Read-ahead the right block, we're going to read it - * in the next loop. - */ - xfs_btree_readahead(cur, lev, XFS_BTCUR_RIGHTRA); - } - /* - * If we went off the root then we are seriously confused. - */ - ASSERT(lev < cur->bc_nlevels); - /* - * Now walk back down the tree, fixing up the cursor's buffer - * pointers and key numbers. - */ - for (bp = cur->bc_bufs[lev], block = XFS_BUF_TO_INOBT_BLOCK(bp); - lev > level; ) { - xfs_agblock_t agbno; /* block number of btree block */ - agbno = be32_to_cpu(*XFS_INOBT_PTR_ADDR(block, cur->bc_ptrs[lev], cur)); - if ((error = xfs_btree_read_bufs(cur->bc_mp, cur->bc_tp, - cur->bc_private.a.agno, agbno, 0, &bp, - XFS_INO_BTREE_REF))) - return error; - lev--; - xfs_btree_setbuf(cur, lev, bp); - block = XFS_BUF_TO_INOBT_BLOCK(bp); - if ((error = xfs_btree_check_sblock(cur, block, lev, bp))) - return error; - cur->bc_ptrs[lev] = 1; - } - *stat = 1; - return 0; -} +#ifdef XFS_BTREE_TRACE + .trace_enter = xfs_inobt_trace_enter, + .trace_cursor = xfs_inobt_trace_cursor, + .trace_key = xfs_inobt_trace_key, + .trace_record = xfs_inobt_trace_record, +#endif +}; /* - * Insert the current record at the point referenced by cur. - * The cursor may be inconsistent on return if splits have been done. + * Allocate a new inode btree cursor. */ -int /* error */ -xfs_inobt_insert( - xfs_btree_cur_t *cur, /* btree cursor */ - int *stat) /* success/failure */ +struct xfs_btree_cur * /* new inode btree cursor */ +xfs_inobt_init_cursor( + struct xfs_mount *mp, /* file system mount point */ + struct xfs_trans *tp, /* transaction pointer */ + struct xfs_buf *agbp, /* buffer for agi structure */ + xfs_agnumber_t agno) /* allocation group number */ { - int error; /* error return value */ - int i; /* result value, 0 for failure */ - int level; /* current level number in btree */ - xfs_agblock_t nbno; /* new block number (split result) */ - xfs_btree_cur_t *ncur; /* new cursor (split result) */ - xfs_inobt_rec_t nrec; /* record being inserted this level */ - xfs_btree_cur_t *pcur; /* previous level's cursor */ + struct xfs_agi *agi = XFS_BUF_TO_AGI(agbp); + struct xfs_btree_cur *cur; - level = 0; - nbno = NULLAGBLOCK; - nrec.ir_startino = cpu_to_be32(cur->bc_rec.i.ir_startino); - nrec.ir_freecount = cpu_to_be32(cur->bc_rec.i.ir_freecount); - nrec.ir_free = cpu_to_be64(cur->bc_rec.i.ir_free); - ncur = NULL; - pcur = cur; - /* - * Loop going up the tree, starting at the leaf level. - * Stop when we don't get a split block, that must mean that - * the insert is finished with this level. - */ - do { - /* - * Insert nrec/nbno into this level of the tree. - * Note if we fail, nbno will be null. - */ - if ((error = xfs_inobt_insrec(pcur, level++, &nbno, &nrec, &ncur, - &i))) { - if (pcur != cur) - xfs_btree_del_cursor(pcur, XFS_BTREE_ERROR); - return error; - } - /* - * See if the cursor we just used is trash. - * Can't trash the caller's cursor, but otherwise we should - * if ncur is a new cursor or we're about to be done. - */ - if (pcur != cur && (ncur || nbno == NULLAGBLOCK)) { - cur->bc_nlevels = pcur->bc_nlevels; - xfs_btree_del_cursor(pcur, XFS_BTREE_NOERROR); - } - /* - * If we got a new cursor, switch to it. - */ - if (ncur) { - pcur = ncur; - ncur = NULL; - } - } while (nbno != NULLAGBLOCK); - *stat = i; - return 0; -} + cur = kmem_zone_zalloc(xfs_btree_cur_zone, KM_SLEEP); -/* - * Lookup the record equal to ino in the btree given by cur. - */ -int /* error */ -xfs_inobt_lookup_eq( - xfs_btree_cur_t *cur, /* btree cursor */ - xfs_agino_t ino, /* starting inode of chunk */ - __int32_t fcnt, /* free inode count */ - xfs_inofree_t free, /* free inode mask */ - int *stat) /* success/failure */ -{ - cur->bc_rec.i.ir_startino = ino; - cur->bc_rec.i.ir_freecount = fcnt; - cur->bc_rec.i.ir_free = free; - return xfs_inobt_lookup(cur, XFS_LOOKUP_EQ, stat); -} + cur->bc_tp = tp; + cur->bc_mp = mp; + cur->bc_nlevels = be32_to_cpu(agi->agi_level); + cur->bc_btnum = XFS_BTNUM_INO; + cur->bc_blocklog = mp->m_sb.sb_blocklog; -/* - * Lookup the first record greater than or equal to ino - * in the btree given by cur. - */ -int /* error */ -xfs_inobt_lookup_ge( - xfs_btree_cur_t *cur, /* btree cursor */ - xfs_agino_t ino, /* starting inode of chunk */ - __int32_t fcnt, /* free inode count */ - xfs_inofree_t free, /* free inode mask */ - int *stat) /* success/failure */ -{ - cur->bc_rec.i.ir_startino = ino; - cur->bc_rec.i.ir_freecount = fcnt; - cur->bc_rec.i.ir_free = free; - return xfs_inobt_lookup(cur, XFS_LOOKUP_GE, stat); -} + cur->bc_ops = &xfs_inobt_ops; -/* - * Lookup the first record less than or equal to ino - * in the btree given by cur. - */ -int /* error */ -xfs_inobt_lookup_le( - xfs_btree_cur_t *cur, /* btree cursor */ - xfs_agino_t ino, /* starting inode of chunk */ - __int32_t fcnt, /* free inode count */ - xfs_inofree_t free, /* free inode mask */ - int *stat) /* success/failure */ -{ - cur->bc_rec.i.ir_startino = ino; - cur->bc_rec.i.ir_freecount = fcnt; - cur->bc_rec.i.ir_free = free; - return xfs_inobt_lookup(cur, XFS_LOOKUP_LE, stat); + cur->bc_private.a.agbp = agbp; + cur->bc_private.a.agno = agno; + + return cur; } /* - * Update the record referred to by cur, to the value given - * by [ino, fcnt, free]. - * This either works (return 0) or gets an EFSCORRUPTED error. + * Calculate number of records in an inobt btree block. */ -int /* error */ -xfs_inobt_update( - xfs_btree_cur_t *cur, /* btree cursor */ - xfs_agino_t ino, /* starting inode of chunk */ - __int32_t fcnt, /* free inode count */ - xfs_inofree_t free) /* free inode mask */ +int +xfs_inobt_maxrecs( + struct xfs_mount *mp, + int blocklen, + int leaf) { - xfs_inobt_block_t *block; /* btree block to update */ - xfs_buf_t *bp; /* buffer containing btree block */ - int error; /* error return value */ - int ptr; /* current record number (updating) */ - xfs_inobt_rec_t *rp; /* pointer to updated record */ + blocklen -= XFS_INOBT_BLOCK_LEN(mp); - /* - * Pick up the current block. - */ - bp = cur->bc_bufs[0]; - block = XFS_BUF_TO_INOBT_BLOCK(bp); -#ifdef DEBUG - if ((error = xfs_btree_check_sblock(cur, block, 0, bp))) - return error; -#endif - /* - * Get the address of the rec to be updated. - */ - ptr = cur->bc_ptrs[0]; - rp = XFS_INOBT_REC_ADDR(block, ptr, cur); - /* - * Fill in the new contents and log them. - */ - rp->ir_startino = cpu_to_be32(ino); - rp->ir_freecount = cpu_to_be32(fcnt); - rp->ir_free = cpu_to_be64(free); - xfs_inobt_log_recs(cur, bp, ptr, ptr); - /* - * Updating first record in leaf. Pass new key value up to our parent. - */ - if (ptr == 1) { - xfs_inobt_key_t key; /* key containing [ino] */ - - key.ir_startino = cpu_to_be32(ino); - if ((error = xfs_inobt_updkey(cur, &key, 1))) - return error; - } - return 0; + if (leaf) + return blocklen / sizeof(xfs_inobt_rec_t); + return blocklen / (sizeof(xfs_inobt_key_t) + sizeof(xfs_inobt_ptr_t)); } diff --git a/fs/xfs/xfs_ialloc_btree.h b/fs/xfs/xfs_ialloc_btree.h index 8efc4a5b8b9..37e5dd01a57 100644 --- a/fs/xfs/xfs_ialloc_btree.h +++ b/fs/xfs/xfs_ialloc_btree.h @@ -24,7 +24,6 @@ struct xfs_buf; struct xfs_btree_cur; -struct xfs_btree_sblock; struct xfs_mount; /* @@ -70,11 +69,6 @@ typedef struct xfs_inobt_key { /* btree pointer type */ typedef __be32 xfs_inobt_ptr_t; -/* btree block header type */ -typedef struct xfs_btree_sblock xfs_inobt_block_t; - -#define XFS_BUF_TO_INOBT_BLOCK(bp) ((xfs_inobt_block_t *)XFS_BUF_PTR(bp)) - /* * Bit manipulations for ir_free. */ @@ -85,14 +79,6 @@ typedef struct xfs_btree_sblock xfs_inobt_block_t; #define XFS_INOBT_CLR_FREE(rp,i) ((rp)->ir_free &= ~XFS_INOBT_MASK(i)) /* - * Real block structures have a size equal to the disk block size. - */ -#define XFS_INOBT_BLOCK_MAXRECS(lev,cur) ((cur)->bc_mp->m_inobt_mxr[lev != 0]) -#define XFS_INOBT_BLOCK_MINRECS(lev,cur) ((cur)->bc_mp->m_inobt_mnr[lev != 0]) -#define XFS_INOBT_IS_LAST_REC(cur) \ - ((cur)->bc_ptrs[0] == be16_to_cpu(XFS_BUF_TO_INOBT_BLOCK((cur)->bc_bufs[0])->bb_numrecs)) - -/* * Maximum number of inode btree levels. */ #define XFS_IN_MAXLEVELS(mp) ((mp)->m_in_maxlevels) @@ -104,75 +90,38 @@ typedef struct xfs_btree_sblock xfs_inobt_block_t; #define XFS_PREALLOC_BLOCKS(mp) ((xfs_agblock_t)(XFS_IBT_BLOCK(mp) + 1)) /* - * Record, key, and pointer address macros for btree blocks. - */ -#define XFS_INOBT_REC_ADDR(bb,i,cur) \ - (XFS_BTREE_REC_ADDR(xfs_inobt, bb, i)) - -#define XFS_INOBT_KEY_ADDR(bb,i,cur) \ - (XFS_BTREE_KEY_ADDR(xfs_inobt, bb, i)) - -#define XFS_INOBT_PTR_ADDR(bb,i,cur) \ - (XFS_BTREE_PTR_ADDR(xfs_inobt, bb, \ - i, XFS_INOBT_BLOCK_MAXRECS(1, cur))) - -/* - * Decrement cursor by one record at the level. - * For nonzero levels the leaf-ward information is untouched. - */ -extern int xfs_inobt_decrement(struct xfs_btree_cur *cur, int level, int *stat); - -/* - * Delete the record pointed to by cur. - * The cursor refers to the place where the record was (could be inserted) - * when the operation returns. - */ -extern int xfs_inobt_delete(struct xfs_btree_cur *cur, int *stat); - -/* - * Get the data from the pointed-to record. - */ -extern int xfs_inobt_get_rec(struct xfs_btree_cur *cur, xfs_agino_t *ino, - __int32_t *fcnt, xfs_inofree_t *free, int *stat); - -/* - * Increment cursor by one record at the level. - * For nonzero levels the leaf-ward information is untouched. - */ -extern int xfs_inobt_increment(struct xfs_btree_cur *cur, int level, int *stat); - -/* - * Insert the current record at the point referenced by cur. - * The cursor may be inconsistent on return if splits have been done. - */ -extern int xfs_inobt_insert(struct xfs_btree_cur *cur, int *stat); - -/* - * Lookup the record equal to ino in the btree given by cur. - */ -extern int xfs_inobt_lookup_eq(struct xfs_btree_cur *cur, xfs_agino_t ino, - __int32_t fcnt, xfs_inofree_t free, int *stat); - -/* - * Lookup the first record greater than or equal to ino - * in the btree given by cur. - */ -extern int xfs_inobt_lookup_ge(struct xfs_btree_cur *cur, xfs_agino_t ino, - __int32_t fcnt, xfs_inofree_t free, int *stat); - -/* - * Lookup the first record less than or equal to ino - * in the btree given by cur. + * Btree block header size depends on a superblock flag. + * + * (not quite yet, but soon) */ -extern int xfs_inobt_lookup_le(struct xfs_btree_cur *cur, xfs_agino_t ino, - __int32_t fcnt, xfs_inofree_t free, int *stat); +#define XFS_INOBT_BLOCK_LEN(mp) XFS_BTREE_SBLOCK_LEN /* - * Update the record referred to by cur, to the value given - * by [ino, fcnt, free]. - * This either works (return 0) or gets an EFSCORRUPTED error. - */ -extern int xfs_inobt_update(struct xfs_btree_cur *cur, xfs_agino_t ino, - __int32_t fcnt, xfs_inofree_t free); + * Record, key, and pointer address macros for btree blocks. + * + * (note that some of these may appear unused, but they are used in userspace) + */ +#define XFS_INOBT_REC_ADDR(mp, block, index) \ + ((xfs_inobt_rec_t *) \ + ((char *)(block) + \ + XFS_INOBT_BLOCK_LEN(mp) + \ + (((index) - 1) * sizeof(xfs_inobt_rec_t)))) + +#define XFS_INOBT_KEY_ADDR(mp, block, index) \ + ((xfs_inobt_key_t *) \ + ((char *)(block) + \ + XFS_INOBT_BLOCK_LEN(mp) + \ + ((index) - 1) * sizeof(xfs_inobt_key_t))) + +#define XFS_INOBT_PTR_ADDR(mp, block, index, maxrecs) \ + ((xfs_inobt_ptr_t *) \ + ((char *)(block) + \ + XFS_INOBT_BLOCK_LEN(mp) + \ + (maxrecs) * sizeof(xfs_inobt_key_t) + \ + ((index) - 1) * sizeof(xfs_inobt_ptr_t))) + +extern struct xfs_btree_cur *xfs_inobt_init_cursor(struct xfs_mount *, + struct xfs_trans *, struct xfs_buf *, xfs_agnumber_t); +extern int xfs_inobt_maxrecs(struct xfs_mount *, int, int); #endif /* __XFS_IALLOC_BTREE_H__ */ diff --git a/fs/xfs/xfs_iget.c b/fs/xfs/xfs_iget.c index e229e9e001c..e2fb6210d4c 100644 --- a/fs/xfs/xfs_iget.c +++ b/fs/xfs/xfs_iget.c @@ -38,281 +38,283 @@ #include "xfs_ialloc.h" #include "xfs_quota.h" #include "xfs_utils.h" +#include "xfs_trans_priv.h" +#include "xfs_inode_item.h" +#include "xfs_bmap.h" +#include "xfs_btree_trace.h" +#include "xfs_dir2_trace.h" + /* - * Look up an inode by number in the given file system. - * The inode is looked up in the cache held in each AG. - * If the inode is found in the cache, attach it to the provided - * vnode. - * - * If it is not in core, read it in from the file system's device, - * add it to the cache and attach the provided vnode. - * - * The inode is locked according to the value of the lock_flags parameter. - * This flag parameter indicates how and if the inode's IO lock and inode lock - * should be taken. - * - * mp -- the mount point structure for the current file system. It points - * to the inode hash table. - * tp -- a pointer to the current transaction if there is one. This is - * simply passed through to the xfs_iread() call. - * ino -- the number of the inode desired. This is the unique identifier - * within the file system for the inode being requested. - * lock_flags -- flags indicating how to lock the inode. See the comment - * for xfs_ilock() for a list of valid values. - * bno -- the block number starting the buffer containing the inode, - * if known (as by bulkstat), else 0. + * Allocate and initialise an xfs_inode. */ -STATIC int -xfs_iget_core( - struct inode *inode, - xfs_mount_t *mp, - xfs_trans_t *tp, - xfs_ino_t ino, - uint flags, - uint lock_flags, - xfs_inode_t **ipp, - xfs_daddr_t bno) +STATIC struct xfs_inode * +xfs_inode_alloc( + struct xfs_mount *mp, + xfs_ino_t ino) { - struct inode *old_inode; - xfs_inode_t *ip; - xfs_inode_t *iq; - int error; - unsigned long first_index, mask; - xfs_perag_t *pag; - xfs_agino_t agino; + struct xfs_inode *ip; - /* the radix tree exists only in inode capable AGs */ - if (XFS_INO_TO_AGNO(mp, ino) >= mp->m_maxagi) - return EINVAL; + /* + * if this didn't occur in transactions, we could use + * KM_MAYFAIL and return NULL here on ENOMEM. Set the + * code up to do this anyway. + */ + ip = kmem_zone_alloc(xfs_inode_zone, KM_SLEEP); + if (!ip) + return NULL; - /* get the perag structure and ensure that it's inode capable */ - pag = xfs_get_perag(mp, ino); - if (!pag->pagi_inodeok) - return EINVAL; - ASSERT(pag->pag_ici_init); - agino = XFS_INO_TO_AGINO(mp, ino); + ASSERT(atomic_read(&ip->i_iocount) == 0); + ASSERT(atomic_read(&ip->i_pincount) == 0); + ASSERT(!spin_is_locked(&ip->i_flags_lock)); + ASSERT(completion_done(&ip->i_flush)); -again: - read_lock(&pag->pag_ici_lock); - ip = radix_tree_lookup(&pag->pag_ici_root, agino); + /* + * initialise the VFS inode here to get failures + * out of the way early. + */ + if (!inode_init_always(mp->m_super, VFS_I(ip))) { + kmem_zone_free(xfs_inode_zone, ip); + return NULL; + } + + /* initialise the xfs inode */ + ip->i_ino = ino; + ip->i_mount = mp; + memset(&ip->i_imap, 0, sizeof(struct xfs_imap)); + ip->i_afp = NULL; + memset(&ip->i_df, 0, sizeof(xfs_ifork_t)); + ip->i_flags = 0; + ip->i_update_core = 0; + ip->i_update_size = 0; + ip->i_delayed_blks = 0; + memset(&ip->i_d, 0, sizeof(xfs_icdinode_t)); + ip->i_size = 0; + ip->i_new_size = 0; + + /* + * Initialize inode's trace buffers. + */ +#ifdef XFS_INODE_TRACE + ip->i_trace = ktrace_alloc(INODE_TRACE_SIZE, KM_NOFS); +#endif +#ifdef XFS_BMAP_TRACE + ip->i_xtrace = ktrace_alloc(XFS_BMAP_KTRACE_SIZE, KM_NOFS); +#endif +#ifdef XFS_BTREE_TRACE + ip->i_btrace = ktrace_alloc(XFS_BMBT_KTRACE_SIZE, KM_NOFS); +#endif +#ifdef XFS_RW_TRACE + ip->i_rwtrace = ktrace_alloc(XFS_RW_KTRACE_SIZE, KM_NOFS); +#endif +#ifdef XFS_ILOCK_TRACE + ip->i_lock_trace = ktrace_alloc(XFS_ILOCK_KTRACE_SIZE, KM_NOFS); +#endif +#ifdef XFS_DIR2_TRACE + ip->i_dir_trace = ktrace_alloc(XFS_DIR2_KTRACE_SIZE, KM_NOFS); +#endif + + return ip; +} + +/* + * Check the validity of the inode we just found it the cache + */ +static int +xfs_iget_cache_hit( + struct xfs_perag *pag, + struct xfs_inode *ip, + int flags, + int lock_flags) __releases(pag->pag_ici_lock) +{ + struct xfs_mount *mp = ip->i_mount; + int error = EAGAIN; + + /* + * If INEW is set this inode is being set up + * If IRECLAIM is set this inode is being torn down + * Pause and try again. + */ + if (xfs_iflags_test(ip, (XFS_INEW|XFS_IRECLAIM))) { + XFS_STATS_INC(xs_ig_frecycle); + goto out_error; + } + + /* If IRECLAIMABLE is set, we've torn down the vfs inode part */ + if (xfs_iflags_test(ip, XFS_IRECLAIMABLE)) { - if (ip != NULL) { /* - * If INEW is set this inode is being set up - * we need to pause and try again. + * If lookup is racing with unlink, then we should return an + * error immediately so we don't remove it from the reclaim + * list and potentially leak the inode. */ - if (xfs_iflags_test(ip, XFS_INEW)) { - read_unlock(&pag->pag_ici_lock); - delay(1); - XFS_STATS_INC(xs_ig_frecycle); - - goto again; + if ((ip->i_d.di_mode == 0) && !(flags & XFS_IGET_CREATE)) { + error = ENOENT; + goto out_error; } - old_inode = ip->i_vnode; - if (old_inode == NULL) { - /* - * If IRECLAIM is set this inode is - * on its way out of the system, - * we need to pause and try again. - */ - if (xfs_iflags_test(ip, XFS_IRECLAIM)) { - read_unlock(&pag->pag_ici_lock); - delay(1); - XFS_STATS_INC(xs_ig_frecycle); - - goto again; - } - ASSERT(xfs_iflags_test(ip, XFS_IRECLAIMABLE)); - - /* - * If lookup is racing with unlink, then we - * should return an error immediately so we - * don't remove it from the reclaim list and - * potentially leak the inode. - */ - if ((ip->i_d.di_mode == 0) && - !(flags & XFS_IGET_CREATE)) { - read_unlock(&pag->pag_ici_lock); - xfs_put_perag(mp, pag); - return ENOENT; - } - - xfs_itrace_exit_tag(ip, "xfs_iget.alloc"); - - XFS_STATS_INC(xs_ig_found); - xfs_iflags_clear(ip, XFS_IRECLAIMABLE); - read_unlock(&pag->pag_ici_lock); - - XFS_MOUNT_ILOCK(mp); - list_del_init(&ip->i_reclaim); - XFS_MOUNT_IUNLOCK(mp); - - goto finish_inode; - - } else if (inode != old_inode) { - /* The inode is being torn down, pause and - * try again. - */ - if (old_inode->i_state & (I_FREEING | I_CLEAR)) { - read_unlock(&pag->pag_ici_lock); - delay(1); - XFS_STATS_INC(xs_ig_frecycle); - - goto again; - } -/* Chances are the other vnode (the one in the inode) is being torn -* down right now, and we landed on top of it. Question is, what do -* we do? Unhook the old inode and hook up the new one? -*/ - cmn_err(CE_PANIC, - "xfs_iget_core: ambiguous vns: vp/0x%p, invp/0x%p", - old_inode, inode); - } + xfs_itrace_exit_tag(ip, "xfs_iget.alloc"); /* - * Inode cache hit + * We need to re-initialise the VFS inode as it has been + * 'freed' by the VFS. Do this here so we can deal with + * errors cleanly, then tag it so it can be set up correctly + * later. */ - read_unlock(&pag->pag_ici_lock); - XFS_STATS_INC(xs_ig_found); - -finish_inode: - if (ip->i_d.di_mode == 0 && !(flags & XFS_IGET_CREATE)) { - xfs_put_perag(mp, pag); - return ENOENT; + if (!inode_init_always(mp->m_super, VFS_I(ip))) { + error = ENOMEM; + goto out_error; } - if (lock_flags != 0) - xfs_ilock(ip, lock_flags); + /* + * We must set the XFS_INEW flag before clearing the + * XFS_IRECLAIMABLE flag so that if a racing lookup does + * not find the XFS_IRECLAIMABLE above but has the igrab() + * below succeed we can safely check XFS_INEW to detect + * that this inode is still being initialised. + */ + xfs_iflags_set(ip, XFS_INEW); + xfs_iflags_clear(ip, XFS_IRECLAIMABLE); + + /* clear the radix tree reclaim flag as well. */ + __xfs_inode_clear_reclaim_tag(mp, pag, ip); + } else if (!igrab(VFS_I(ip))) { + /* If the VFS inode is being torn down, pause and try again. */ + XFS_STATS_INC(xs_ig_frecycle); + goto out_error; + } else if (xfs_iflags_test(ip, XFS_INEW)) { + /* + * We are racing with another cache hit that is + * currently recycling this inode out of the XFS_IRECLAIMABLE + * state. Wait for the initialisation to complete before + * continuing. + */ + wait_on_inode(VFS_I(ip)); + } - xfs_iflags_clear(ip, XFS_ISTALE); - xfs_itrace_exit_tag(ip, "xfs_iget.found"); - goto return_ip; + if (ip->i_d.di_mode == 0 && !(flags & XFS_IGET_CREATE)) { + error = ENOENT; + iput(VFS_I(ip)); + goto out_error; } - /* - * Inode cache miss - */ + /* We've got a live one. */ read_unlock(&pag->pag_ici_lock); - XFS_STATS_INC(xs_ig_missed); - /* - * Read the disk inode attributes into a new inode structure and get - * a new vnode for it. This should also initialize i_ino and i_mount. - */ - error = xfs_iread(mp, tp, ino, &ip, bno, - (flags & XFS_IGET_BULKSTAT) ? XFS_IMAP_BULKSTAT : 0); - if (error) { - xfs_put_perag(mp, pag); - return error; - } + if (lock_flags != 0) + xfs_ilock(ip, lock_flags); - xfs_itrace_exit_tag(ip, "xfs_iget.alloc"); + xfs_iflags_clear(ip, XFS_ISTALE); + xfs_itrace_exit_tag(ip, "xfs_iget.found"); + XFS_STATS_INC(xs_ig_found); + return 0; + +out_error: + read_unlock(&pag->pag_ici_lock); + return error; +} - mrlock_init(&ip->i_lock, MRLOCK_ALLOW_EQUAL_PRI|MRLOCK_BARRIER, - "xfsino", ip->i_ino); - mrlock_init(&ip->i_iolock, MRLOCK_BARRIER, "xfsio", ip->i_ino); - init_waitqueue_head(&ip->i_ipin_wait); - atomic_set(&ip->i_pincount, 0); +static int +xfs_iget_cache_miss( + struct xfs_mount *mp, + struct xfs_perag *pag, + xfs_trans_t *tp, + xfs_ino_t ino, + struct xfs_inode **ipp, + xfs_daddr_t bno, + int flags, + int lock_flags) __releases(pag->pag_ici_lock) +{ + struct xfs_inode *ip; + int error; + unsigned long first_index, mask; + xfs_agino_t agino = XFS_INO_TO_AGINO(mp, ino); - /* - * Because we want to use a counting completion, complete - * the flush completion once to allow a single access to - * the flush completion without blocking. - */ - init_completion(&ip->i_flush); - complete(&ip->i_flush); + ip = xfs_inode_alloc(mp, ino); + if (!ip) + return ENOMEM; - if (lock_flags) - xfs_ilock(ip, lock_flags); + error = xfs_iread(mp, tp, ip, bno, flags); + if (error) + goto out_destroy; + + xfs_itrace_exit_tag(ip, "xfs_iget.alloc"); if ((ip->i_d.di_mode == 0) && !(flags & XFS_IGET_CREATE)) { - xfs_idestroy(ip); - xfs_put_perag(mp, pag); - return ENOENT; + error = ENOENT; + goto out_destroy; } + if (lock_flags) + xfs_ilock(ip, lock_flags); + /* * Preload the radix tree so we can insert safely under the - * write spinlock. + * write spinlock. Note that we cannot sleep inside the preload + * region. */ if (radix_tree_preload(GFP_KERNEL)) { - xfs_idestroy(ip); - delay(1); - goto again; + error = EAGAIN; + goto out_unlock; } + mask = ~(((XFS_INODE_CLUSTER_SIZE(mp) >> mp->m_sb.sb_inodelog)) - 1); first_index = agino & mask; write_lock(&pag->pag_ici_lock); - /* - * insert the new inode - */ + + /* insert the new inode */ error = radix_tree_insert(&pag->pag_ici_root, agino, ip); if (unlikely(error)) { - BUG_ON(error != -EEXIST); - write_unlock(&pag->pag_ici_lock); - radix_tree_preload_end(); - xfs_idestroy(ip); + WARN_ON(error != -EEXIST); XFS_STATS_INC(xs_ig_dup); - goto again; + error = EAGAIN; + goto out_preload_end; } - /* - * These values _must_ be set before releasing the radix tree lock! - */ + /* These values _must_ be set before releasing the radix tree lock! */ ip->i_udquot = ip->i_gdquot = NULL; xfs_iflags_set(ip, XFS_INEW); write_unlock(&pag->pag_ici_lock); radix_tree_preload_end(); - - /* - * Link ip to its mount and thread it on the mount's inode list. - */ - XFS_MOUNT_ILOCK(mp); - if ((iq = mp->m_inodes)) { - ASSERT(iq->i_mprev->i_mnext == iq); - ip->i_mprev = iq->i_mprev; - iq->i_mprev->i_mnext = ip; - iq->i_mprev = ip; - ip->i_mnext = iq; - } else { - ip->i_mnext = ip; - ip->i_mprev = ip; - } - mp->m_inodes = ip; - - XFS_MOUNT_IUNLOCK(mp); - xfs_put_perag(mp, pag); - - return_ip: - ASSERT(ip->i_df.if_ext_max == - XFS_IFORK_DSIZE(ip) / sizeof(xfs_bmbt_rec_t)); - - xfs_iflags_set(ip, XFS_IMODIFIED); *ipp = ip; - - /* - * Set up the Linux with the Linux inode. - */ - ip->i_vnode = inode; - inode->i_private = ip; - - /* - * If we have a real type for an on-disk inode, we can set ops(&unlock) - * now. If it's a new inode being created, xfs_ialloc will handle it. - */ - if (ip->i_d.di_mode != 0) - xfs_setup_inode(ip); return 0; -} +out_preload_end: + write_unlock(&pag->pag_ici_lock); + radix_tree_preload_end(); +out_unlock: + if (lock_flags) + xfs_iunlock(ip, lock_flags); +out_destroy: + xfs_destroy_inode(ip); + return error; +} /* - * The 'normal' internal xfs_iget, if needed it will - * 'allocate', or 'get', the vnode. + * Look up an inode by number in the given file system. + * The inode is looked up in the cache held in each AG. + * If the inode is found in the cache, initialise the vfs inode + * if necessary. + * + * If it is not in core, read it in from the file system's device, + * add it to the cache and initialise the vfs inode. + * + * The inode is locked according to the value of the lock_flags parameter. + * This flag parameter indicates how and if the inode's IO lock and inode lock + * should be taken. + * + * mp -- the mount point structure for the current file system. It points + * to the inode hash table. + * tp -- a pointer to the current transaction if there is one. This is + * simply passed through to the xfs_iread() call. + * ino -- the number of the inode desired. This is the unique identifier + * within the file system for the inode being requested. + * lock_flags -- flags indicating how to lock the inode. See the comment + * for xfs_ilock() for a list of valid values. + * bno -- the block number starting the buffer containing the inode, + * if known (as by bulkstat), else 0. */ int xfs_iget( @@ -324,61 +326,64 @@ xfs_iget( xfs_inode_t **ipp, xfs_daddr_t bno) { - struct inode *inode; xfs_inode_t *ip; int error; + xfs_perag_t *pag; + xfs_agino_t agino; - XFS_STATS_INC(xs_ig_attempts); + /* the radix tree exists only in inode capable AGs */ + if (XFS_INO_TO_AGNO(mp, ino) >= mp->m_maxagi) + return EINVAL; -retry: - inode = iget_locked(mp->m_super, ino); - if (!inode) - /* If we got no inode we are out of memory */ - return ENOMEM; + /* get the perag structure and ensure that it's inode capable */ + pag = xfs_get_perag(mp, ino); + if (!pag->pagi_inodeok) + return EINVAL; + ASSERT(pag->pag_ici_init); + agino = XFS_INO_TO_AGINO(mp, ino); - if (inode->i_state & I_NEW) { - XFS_STATS_INC(vn_active); - XFS_STATS_INC(vn_alloc); - - error = xfs_iget_core(inode, mp, tp, ino, flags, - lock_flags, ipp, bno); - if (error) { - make_bad_inode(inode); - if (inode->i_state & I_NEW) - unlock_new_inode(inode); - iput(inode); - } - return error; +again: + error = 0; + read_lock(&pag->pag_ici_lock); + ip = radix_tree_lookup(&pag->pag_ici_root, agino); + + if (ip) { + error = xfs_iget_cache_hit(pag, ip, flags, lock_flags); + if (error) + goto out_error_or_again; + } else { + read_unlock(&pag->pag_ici_lock); + XFS_STATS_INC(xs_ig_missed); + + error = xfs_iget_cache_miss(mp, pag, tp, ino, &ip, bno, + flags, lock_flags); + if (error) + goto out_error_or_again; } + xfs_put_perag(mp, pag); + *ipp = ip; + + ASSERT(ip->i_df.if_ext_max == + XFS_IFORK_DSIZE(ip) / sizeof(xfs_bmbt_rec_t)); /* - * If the inode is not fully constructed due to - * filehandle mismatches wait for the inode to go - * away and try again. - * - * iget_locked will call __wait_on_freeing_inode - * to wait for the inode to go away. + * If we have a real type for an on-disk inode, we can set ops(&unlock) + * now. If it's a new inode being created, xfs_ialloc will handle it. */ - if (is_bad_inode(inode)) { - iput(inode); - delay(1); - goto retry; - } + if (xfs_iflags_test(ip, XFS_INEW) && ip->i_d.di_mode != 0) + xfs_setup_inode(ip); + return 0; - ip = XFS_I(inode); - if (!ip) { - iput(inode); +out_error_or_again: + if (error == EAGAIN) { delay(1); - goto retry; + goto again; } - - if (lock_flags != 0) - xfs_ilock(ip, lock_flags); - XFS_STATS_INC(xs_ig_found); - *ipp = ip; - return 0; + xfs_put_perag(mp, pag); + return error; } + /* * Look for the inode corresponding to the given ino in the hash table. * If it is there and its i_transp pointer matches tp, return it. @@ -444,99 +449,109 @@ xfs_iput_new( IRELE(ip); } - /* - * This routine embodies the part of the reclaim code that pulls - * the inode from the inode hash table and the mount structure's - * inode list. - * This should only be called from xfs_reclaim(). + * This is called free all the memory associated with an inode. + * It must free the inode itself and any buffers allocated for + * if_extents/if_data and if_broot. It must also free the lock + * associated with the inode. + * + * Note: because we don't initialise everything on reallocation out + * of the zone, we must ensure we nullify everything correctly before + * freeing the structure. */ void -xfs_ireclaim(xfs_inode_t *ip) +xfs_ireclaim( + struct xfs_inode *ip) { - /* - * Remove from old hash list and mount list. - */ - XFS_STATS_INC(xs_ig_reclaims); + struct xfs_mount *mp = ip->i_mount; + struct xfs_perag *pag; - xfs_iextract(ip); - - /* - * Here we do a spurious inode lock in order to coordinate with - * xfs_sync(). This is because xfs_sync() references the inodes - * in the mount list without taking references on the corresponding - * vnodes. We make that OK here by ensuring that we wait until - * the inode is unlocked in xfs_sync() before we go ahead and - * free it. We get both the regular lock and the io lock because - * the xfs_sync() code may need to drop the regular one but will - * still hold the io lock. - */ - xfs_ilock(ip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL); - - /* - * Release dquots (and their references) if any. An inode may escape - * xfs_inactive and get here via vn_alloc->vn_reclaim path. - */ - XFS_QM_DQDETACH(ip->i_mount, ip); - - /* - * Pull our behavior descriptor from the vnode chain. - */ - if (ip->i_vnode) { - ip->i_vnode->i_private = NULL; - ip->i_vnode = NULL; - } + XFS_STATS_INC(xs_ig_reclaims); /* - * Free all memory associated with the inode. + * Remove the inode from the per-AG radix tree. It doesn't matter + * if it was never added to it because radix_tree_delete can deal + * with that case just fine. */ - xfs_iunlock(ip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL); - xfs_idestroy(ip); -} - -/* - * This routine removes an about-to-be-destroyed inode from - * all of the lists in which it is located with the exception - * of the behavior chain. - */ -void -xfs_iextract( - xfs_inode_t *ip) -{ - xfs_mount_t *mp = ip->i_mount; - xfs_perag_t *pag = xfs_get_perag(mp, ip->i_ino); - xfs_inode_t *iq; - + pag = xfs_get_perag(mp, ip->i_ino); write_lock(&pag->pag_ici_lock); radix_tree_delete(&pag->pag_ici_root, XFS_INO_TO_AGINO(mp, ip->i_ino)); write_unlock(&pag->pag_ici_lock); xfs_put_perag(mp, pag); /* - * Remove from mount's inode list. + * Here we do an (almost) spurious inode lock in order to coordinate + * with inode cache radix tree lookups. This is because the lookup + * can reference the inodes in the cache without taking references. + * + * We make that OK here by ensuring that we wait until the inode is + * unlocked after the lookup before we go ahead and free it. We get + * both the ilock and the iolock because the code may need to drop the + * ilock one but will still hold the iolock. */ - XFS_MOUNT_ILOCK(mp); - ASSERT((ip->i_mnext != NULL) && (ip->i_mprev != NULL)); - iq = ip->i_mnext; - iq->i_mprev = ip->i_mprev; - ip->i_mprev->i_mnext = iq; - + xfs_ilock(ip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL); /* - * Fix up the head pointer if it points to the inode being deleted. + * Release dquots (and their references) if any. */ - if (mp->m_inodes == ip) { - if (ip == iq) { - mp->m_inodes = NULL; - } else { - mp->m_inodes = iq; - } + XFS_QM_DQDETACH(ip->i_mount, ip); + xfs_iunlock(ip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL); + + switch (ip->i_d.di_mode & S_IFMT) { + case S_IFREG: + case S_IFDIR: + case S_IFLNK: + xfs_idestroy_fork(ip, XFS_DATA_FORK); + break; } - /* Deal with the deleted inodes list */ - list_del_init(&ip->i_reclaim); + if (ip->i_afp) + xfs_idestroy_fork(ip, XFS_ATTR_FORK); - mp->m_ireclaims++; - XFS_MOUNT_IUNLOCK(mp); +#ifdef XFS_INODE_TRACE + ktrace_free(ip->i_trace); +#endif +#ifdef XFS_BMAP_TRACE + ktrace_free(ip->i_xtrace); +#endif +#ifdef XFS_BTREE_TRACE + ktrace_free(ip->i_btrace); +#endif +#ifdef XFS_RW_TRACE + ktrace_free(ip->i_rwtrace); +#endif +#ifdef XFS_ILOCK_TRACE + ktrace_free(ip->i_lock_trace); +#endif +#ifdef XFS_DIR2_TRACE + ktrace_free(ip->i_dir_trace); +#endif + if (ip->i_itemp) { + /* + * Only if we are shutting down the fs will we see an + * inode still in the AIL. If it is there, we should remove + * it to prevent a use-after-free from occurring. + */ + xfs_log_item_t *lip = &ip->i_itemp->ili_item; + struct xfs_ail *ailp = lip->li_ailp; + + ASSERT(((lip->li_flags & XFS_LI_IN_AIL) == 0) || + XFS_FORCED_SHUTDOWN(ip->i_mount)); + if (lip->li_flags & XFS_LI_IN_AIL) { + spin_lock(&ailp->xa_lock); + if (lip->li_flags & XFS_LI_IN_AIL) + xfs_trans_ail_delete(ailp, lip); + else + spin_unlock(&ailp->xa_lock); + } + xfs_inode_item_destroy(ip); + ip->i_itemp = NULL; + } + /* asserts to verify all state is correct here */ + ASSERT(atomic_read(&ip->i_iocount) == 0); + ASSERT(atomic_read(&ip->i_pincount) == 0); + ASSERT(!spin_is_locked(&ip->i_flags_lock)); + ASSERT(completion_done(&ip->i_flush)); + kmem_zone_free(xfs_inode_zone, ip); } /* @@ -737,7 +752,7 @@ xfs_iunlock( * it is in the AIL and anyone is waiting on it. Don't do * this if the caller has asked us not to. */ - xfs_trans_unlocked_item(ip->i_mount, + xfs_trans_unlocked_item(ip->i_itemp->ili_item.li_ailp, (xfs_log_item_t*)(ip->i_itemp)); } xfs_ilock_trace(ip, 3, lock_flags, (inst_t *)__return_address); @@ -790,3 +805,51 @@ xfs_isilocked( } #endif +#ifdef XFS_INODE_TRACE + +#define KTRACE_ENTER(ip, vk, s, line, ra) \ + ktrace_enter((ip)->i_trace, \ +/* 0 */ (void *)(__psint_t)(vk), \ +/* 1 */ (void *)(s), \ +/* 2 */ (void *)(__psint_t) line, \ +/* 3 */ (void *)(__psint_t)atomic_read(&VFS_I(ip)->i_count), \ +/* 4 */ (void *)(ra), \ +/* 5 */ NULL, \ +/* 6 */ (void *)(__psint_t)current_cpu(), \ +/* 7 */ (void *)(__psint_t)current_pid(), \ +/* 8 */ (void *)__return_address, \ +/* 9 */ NULL, NULL, NULL, NULL, NULL, NULL, NULL) + +/* + * Vnode tracing code. + */ +void +_xfs_itrace_entry(xfs_inode_t *ip, const char *func, inst_t *ra) +{ + KTRACE_ENTER(ip, INODE_KTRACE_ENTRY, func, 0, ra); +} + +void +_xfs_itrace_exit(xfs_inode_t *ip, const char *func, inst_t *ra) +{ + KTRACE_ENTER(ip, INODE_KTRACE_EXIT, func, 0, ra); +} + +void +xfs_itrace_hold(xfs_inode_t *ip, char *file, int line, inst_t *ra) +{ + KTRACE_ENTER(ip, INODE_KTRACE_HOLD, file, line, ra); +} + +void +_xfs_itrace_ref(xfs_inode_t *ip, char *file, int line, inst_t *ra) +{ + KTRACE_ENTER(ip, INODE_KTRACE_REF, file, line, ra); +} + +void +xfs_itrace_rele(xfs_inode_t *ip, char *file, int line, inst_t *ra) +{ + KTRACE_ENTER(ip, INODE_KTRACE_RELE, file, line, ra); +} +#endif /* XFS_INODE_TRACE */ diff --git a/fs/xfs/xfs_imap.h b/fs/xfs/xfs_imap.h deleted file mode 100644 index d3645000398..00000000000 --- a/fs/xfs/xfs_imap.h +++ /dev/null @@ -1,40 +0,0 @@ -/* - * Copyright (c) 2000,2005 Silicon Graphics, Inc. - * All Rights Reserved. - * - * This program is free software; you can redistribute it and/or - * modify it under the terms of the GNU General Public License as - * published by the Free Software Foundation. - * - * This program is distributed in the hope that it would be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write the Free Software Foundation, - * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA - */ -#ifndef __XFS_IMAP_H__ -#define __XFS_IMAP_H__ - -/* - * This is the structure passed to xfs_imap() to map - * an inode number to its on disk location. - */ -typedef struct xfs_imap { - xfs_daddr_t im_blkno; /* starting BB of inode chunk */ - uint im_len; /* length in BBs of inode chunk */ - xfs_agblock_t im_agblkno; /* logical block of inode chunk in ag */ - ushort im_ioffset; /* inode offset in block in "inodes" */ - ushort im_boffset; /* inode offset in block in bytes */ -} xfs_imap_t; - -#ifdef __KERNEL__ -struct xfs_mount; -struct xfs_trans; -int xfs_imap(struct xfs_mount *, struct xfs_trans *, xfs_ino_t, - xfs_imap_t *, uint); -#endif - -#endif /* __XFS_IMAP_H__ */ diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c index dbd9cef852e..5a5e035e5d3 100644 --- a/fs/xfs/xfs_inode.c +++ b/fs/xfs/xfs_inode.c @@ -23,7 +23,6 @@ #include "xfs_bit.h" #include "xfs_log.h" #include "xfs_inum.h" -#include "xfs_imap.h" #include "xfs_trans.h" #include "xfs_trans_priv.h" #include "xfs_sb.h" @@ -41,6 +40,7 @@ #include "xfs_buf_item.h" #include "xfs_inode_item.h" #include "xfs_btree.h" +#include "xfs_btree_trace.h" #include "xfs_alloc.h" #include "xfs_ialloc.h" #include "xfs_bmap.h" @@ -133,10 +133,10 @@ STATIC int xfs_imap_to_bp( xfs_mount_t *mp, xfs_trans_t *tp, - xfs_imap_t *imap, + struct xfs_imap *imap, xfs_buf_t **bpp, uint buf_flags, - uint imap_flags) + uint iget_flags) { int error; int i; @@ -173,12 +173,12 @@ xfs_imap_to_bp( dip = (xfs_dinode_t *)xfs_buf_offset(bp, (i << mp->m_sb.sb_inodelog)); - di_ok = be16_to_cpu(dip->di_core.di_magic) == XFS_DINODE_MAGIC && - XFS_DINODE_GOOD_VERSION(dip->di_core.di_version); + di_ok = be16_to_cpu(dip->di_magic) == XFS_DINODE_MAGIC && + XFS_DINODE_GOOD_VERSION(dip->di_version); if (unlikely(XFS_TEST_ERROR(!di_ok, mp, XFS_ERRTAG_ITOBP_INOTOBP, XFS_RANDOM_ITOBP_INOTOBP))) { - if (imap_flags & XFS_IMAP_BULKSTAT) { + if (iget_flags & XFS_IGET_BULKSTAT) { xfs_trans_brelse(tp, bp); return XFS_ERROR(EINVAL); } @@ -190,7 +190,7 @@ xfs_imap_to_bp( "daddr %lld #%d (magic=%x)", XFS_BUFTARG_NAME(mp->m_ddev_targp), (unsigned long long)imap->im_blkno, i, - be16_to_cpu(dip->di_core.di_magic)); + be16_to_cpu(dip->di_magic)); #endif xfs_trans_brelse(tp, bp); return XFS_ERROR(EFSCORRUPTED); @@ -221,25 +221,26 @@ xfs_imap_to_bp( * Use xfs_imap() to determine the size and location of the * buffer to read from disk. */ -STATIC int +int xfs_inotobp( xfs_mount_t *mp, xfs_trans_t *tp, xfs_ino_t ino, xfs_dinode_t **dipp, xfs_buf_t **bpp, - int *offset) + int *offset, + uint imap_flags) { - xfs_imap_t imap; + struct xfs_imap imap; xfs_buf_t *bp; int error; imap.im_blkno = 0; - error = xfs_imap(mp, tp, ino, &imap, XFS_IMAP_LOOKUP); + error = xfs_imap(mp, tp, ino, &imap, imap_flags); if (error) return error; - error = xfs_imap_to_bp(mp, tp, &imap, &bp, XFS_BUF_LOCK, 0); + error = xfs_imap_to_bp(mp, tp, &imap, &bp, XFS_BUF_LOCK, imap_flags); if (error) return error; @@ -260,15 +261,11 @@ xfs_inotobp( * If a non-zero error is returned, then the contents of bpp and * dipp are undefined. * - * If the inode is new and has not yet been initialized, use xfs_imap() - * to determine the size and location of the buffer to read from disk. - * If the inode has already been mapped to its buffer and read in once, - * then use the mapping information stored in the inode rather than - * calling xfs_imap(). This allows us to avoid the overhead of looking - * at the inode btree for small block file systems (see xfs_dilocate()). - * We can tell whether the inode has been mapped in before by comparing - * its disk block address to 0. Only uninitialized inodes will have - * 0 for the disk block address. + * The inode is expected to already been mapped to its buffer and read + * in once, thus we can use the mapping information stored in the inode + * rather than calling xfs_imap(). This allows us to avoid the overhead + * of looking at the inode btree for small block file systems + * (see xfs_imap()). */ int xfs_itobp( @@ -277,40 +274,14 @@ xfs_itobp( xfs_inode_t *ip, xfs_dinode_t **dipp, xfs_buf_t **bpp, - xfs_daddr_t bno, - uint imap_flags, uint buf_flags) { - xfs_imap_t imap; xfs_buf_t *bp; int error; - if (ip->i_blkno == (xfs_daddr_t)0) { - imap.im_blkno = bno; - error = xfs_imap(mp, tp, ip->i_ino, &imap, - XFS_IMAP_LOOKUP | imap_flags); - if (error) - return error; + ASSERT(ip->i_imap.im_blkno != 0); - /* - * Fill in the fields in the inode that will be used to - * map the inode to its buffer from now on. - */ - ip->i_blkno = imap.im_blkno; - ip->i_len = imap.im_len; - ip->i_boffset = imap.im_boffset; - } else { - /* - * We've already mapped the inode once, so just use the - * mapping that we saved the first time. - */ - imap.im_blkno = ip->i_blkno; - imap.im_len = ip->i_len; - imap.im_boffset = ip->i_boffset; - } - ASSERT(bno == 0 || bno == imap.im_blkno); - - error = xfs_imap_to_bp(mp, tp, &imap, &bp, buf_flags, imap_flags); + error = xfs_imap_to_bp(mp, tp, &ip->i_imap, &bp, buf_flags, 0); if (error) return error; @@ -321,7 +292,7 @@ xfs_itobp( return EAGAIN; } - *dipp = (xfs_dinode_t *)xfs_buf_offset(bp, imap.im_boffset); + *dipp = (xfs_dinode_t *)xfs_buf_offset(bp, ip->i_imap.im_boffset); *bpp = bp; return 0; } @@ -348,26 +319,26 @@ xfs_iformat( XFS_IFORK_DSIZE(ip) / (uint)sizeof(xfs_bmbt_rec_t); error = 0; - if (unlikely(be32_to_cpu(dip->di_core.di_nextents) + - be16_to_cpu(dip->di_core.di_anextents) > - be64_to_cpu(dip->di_core.di_nblocks))) { + if (unlikely(be32_to_cpu(dip->di_nextents) + + be16_to_cpu(dip->di_anextents) > + be64_to_cpu(dip->di_nblocks))) { xfs_fs_repair_cmn_err(CE_WARN, ip->i_mount, "corrupt dinode %Lu, extent total = %d, nblocks = %Lu.", (unsigned long long)ip->i_ino, - (int)(be32_to_cpu(dip->di_core.di_nextents) + - be16_to_cpu(dip->di_core.di_anextents)), + (int)(be32_to_cpu(dip->di_nextents) + + be16_to_cpu(dip->di_anextents)), (unsigned long long) - be64_to_cpu(dip->di_core.di_nblocks)); + be64_to_cpu(dip->di_nblocks)); XFS_CORRUPTION_ERROR("xfs_iformat(1)", XFS_ERRLEVEL_LOW, ip->i_mount, dip); return XFS_ERROR(EFSCORRUPTED); } - if (unlikely(dip->di_core.di_forkoff > ip->i_mount->m_sb.sb_inodesize)) { + if (unlikely(dip->di_forkoff > ip->i_mount->m_sb.sb_inodesize)) { xfs_fs_repair_cmn_err(CE_WARN, ip->i_mount, "corrupt dinode %Lu, forkoff = 0x%x.", (unsigned long long)ip->i_ino, - dip->di_core.di_forkoff); + dip->di_forkoff); XFS_CORRUPTION_ERROR("xfs_iformat(2)", XFS_ERRLEVEL_LOW, ip->i_mount, dip); return XFS_ERROR(EFSCORRUPTED); @@ -378,25 +349,25 @@ xfs_iformat( case S_IFCHR: case S_IFBLK: case S_IFSOCK: - if (unlikely(dip->di_core.di_format != XFS_DINODE_FMT_DEV)) { + if (unlikely(dip->di_format != XFS_DINODE_FMT_DEV)) { XFS_CORRUPTION_ERROR("xfs_iformat(3)", XFS_ERRLEVEL_LOW, ip->i_mount, dip); return XFS_ERROR(EFSCORRUPTED); } ip->i_d.di_size = 0; ip->i_size = 0; - ip->i_df.if_u2.if_rdev = be32_to_cpu(dip->di_u.di_dev); + ip->i_df.if_u2.if_rdev = xfs_dinode_get_rdev(dip); break; case S_IFREG: case S_IFLNK: case S_IFDIR: - switch (dip->di_core.di_format) { + switch (dip->di_format) { case XFS_DINODE_FMT_LOCAL: /* * no local regular files yet */ - if (unlikely((be16_to_cpu(dip->di_core.di_mode) & S_IFMT) == S_IFREG)) { + if (unlikely((be16_to_cpu(dip->di_mode) & S_IFMT) == S_IFREG)) { xfs_fs_repair_cmn_err(CE_WARN, ip->i_mount, "corrupt inode %Lu " "(local format for regular file).", @@ -407,7 +378,7 @@ xfs_iformat( return XFS_ERROR(EFSCORRUPTED); } - di_size = be64_to_cpu(dip->di_core.di_size); + di_size = be64_to_cpu(dip->di_size); if (unlikely(di_size > XFS_DFORK_DSIZE(dip, ip->i_mount))) { xfs_fs_repair_cmn_err(CE_WARN, ip->i_mount, "corrupt inode %Lu " @@ -449,7 +420,7 @@ xfs_iformat( ip->i_afp = kmem_zone_zalloc(xfs_ifork_zone, KM_SLEEP); ip->i_afp->if_ext_max = XFS_IFORK_ASIZE(ip) / (uint)sizeof(xfs_bmbt_rec_t); - switch (dip->di_core.di_aformat) { + switch (dip->di_aformat) { case XFS_DINODE_FMT_LOCAL: atp = (xfs_attr_shortform_t *)XFS_DFORK_APTR(dip); size = be16_to_cpu(atp->hdr.totsize); @@ -621,7 +592,7 @@ xfs_iformat_btree( ifp = XFS_IFORK_PTR(ip, whichfork); dfp = (xfs_bmdr_block_t *)XFS_DFORK_PTR(dip, whichfork); size = XFS_BMAP_BROOT_SPACE(dfp); - nrecs = XFS_BMAP_BROOT_NUMRECS(dfp); + nrecs = be16_to_cpu(dfp->bb_numrecs); /* * blow out if -- fork has less extents than can fit in @@ -649,8 +620,9 @@ xfs_iformat_btree( * Copy and convert from the on-disk structure * to the in-memory structure. */ - xfs_bmdr_to_bmbt(dfp, XFS_DFORK_SIZE(dip, ip->i_mount, whichfork), - ifp->if_broot, size); + xfs_bmdr_to_bmbt(ip->i_mount, dfp, + XFS_DFORK_SIZE(dip, ip->i_mount, whichfork), + ifp->if_broot, size); ifp->if_flags &= ~XFS_IFEXTENTS; ifp->if_flags |= XFS_IFBROOT; @@ -660,7 +632,7 @@ xfs_iformat_btree( void xfs_dinode_from_disk( xfs_icdinode_t *to, - xfs_dinode_core_t *from) + xfs_dinode_t *from) { to->di_magic = be16_to_cpu(from->di_magic); to->di_mode = be16_to_cpu(from->di_mode); @@ -694,7 +666,7 @@ xfs_dinode_from_disk( void xfs_dinode_to_disk( - xfs_dinode_core_t *to, + xfs_dinode_t *to, xfs_icdinode_t *from) { to->di_magic = cpu_to_be16(from->di_magic); @@ -781,93 +753,57 @@ uint xfs_dic2xflags( xfs_dinode_t *dip) { - xfs_dinode_core_t *dic = &dip->di_core; - - return _xfs_dic2xflags(be16_to_cpu(dic->di_flags)) | + return _xfs_dic2xflags(be16_to_cpu(dip->di_flags)) | (XFS_DFORK_Q(dip) ? XFS_XFLAG_HASATTR : 0); } /* - * Given a mount structure and an inode number, return a pointer - * to a newly allocated in-core inode corresponding to the given - * inode number. - * - * Initialize the inode's attributes and extent pointers if it - * already has them (it will not if the inode has no links). + * Read the disk inode attributes into the in-core inode structure. */ int xfs_iread( xfs_mount_t *mp, xfs_trans_t *tp, - xfs_ino_t ino, - xfs_inode_t **ipp, + xfs_inode_t *ip, xfs_daddr_t bno, - uint imap_flags) + uint iget_flags) { xfs_buf_t *bp; xfs_dinode_t *dip; - xfs_inode_t *ip; int error; - ASSERT(xfs_inode_zone != NULL); - - ip = kmem_zone_zalloc(xfs_inode_zone, KM_SLEEP); - ip->i_ino = ino; - ip->i_mount = mp; - atomic_set(&ip->i_iocount, 0); - spin_lock_init(&ip->i_flags_lock); - /* - * Get pointer's to the on-disk inode and the buffer containing it. - * If the inode number refers to a block outside the file system - * then xfs_itobp() will return NULL. In this case we should - * return NULL as well. Set i_blkno to 0 so that xfs_itobp() will - * know that this is a new incore inode. + * Fill in the location information in the in-core inode. */ - error = xfs_itobp(mp, tp, ip, &dip, &bp, bno, imap_flags, XFS_BUF_LOCK); - if (error) { - kmem_zone_free(xfs_inode_zone, ip); + ip->i_imap.im_blkno = bno; + error = xfs_imap(mp, tp, ip->i_ino, &ip->i_imap, iget_flags); + if (error) return error; - } + ASSERT(bno == 0 || bno == ip->i_imap.im_blkno); /* - * Initialize inode's trace buffers. - * Do this before xfs_iformat in case it adds entries. + * Get pointers to the on-disk inode and the buffer containing it. */ -#ifdef XFS_INODE_TRACE - ip->i_trace = ktrace_alloc(INODE_TRACE_SIZE, KM_NOFS); -#endif -#ifdef XFS_BMAP_TRACE - ip->i_xtrace = ktrace_alloc(XFS_BMAP_KTRACE_SIZE, KM_NOFS); -#endif -#ifdef XFS_BMBT_TRACE - ip->i_btrace = ktrace_alloc(XFS_BMBT_KTRACE_SIZE, KM_NOFS); -#endif -#ifdef XFS_RW_TRACE - ip->i_rwtrace = ktrace_alloc(XFS_RW_KTRACE_SIZE, KM_NOFS); -#endif -#ifdef XFS_ILOCK_TRACE - ip->i_lock_trace = ktrace_alloc(XFS_ILOCK_KTRACE_SIZE, KM_NOFS); -#endif -#ifdef XFS_DIR2_TRACE - ip->i_dir_trace = ktrace_alloc(XFS_DIR2_KTRACE_SIZE, KM_NOFS); -#endif + error = xfs_imap_to_bp(mp, tp, &ip->i_imap, &bp, + XFS_BUF_LOCK, iget_flags); + if (error) + return error; + dip = (xfs_dinode_t *)xfs_buf_offset(bp, ip->i_imap.im_boffset); /* * If we got something that isn't an inode it means someone * (nfs or dmi) has a stale handle. */ - if (be16_to_cpu(dip->di_core.di_magic) != XFS_DINODE_MAGIC) { - kmem_zone_free(xfs_inode_zone, ip); - xfs_trans_brelse(tp, bp); + if (be16_to_cpu(dip->di_magic) != XFS_DINODE_MAGIC) { #ifdef DEBUG xfs_fs_cmn_err(CE_ALERT, mp, "xfs_iread: " - "dip->di_core.di_magic (0x%x) != " + "dip->di_magic (0x%x) != " "XFS_DINODE_MAGIC (0x%x)", - be16_to_cpu(dip->di_core.di_magic), + be16_to_cpu(dip->di_magic), XFS_DINODE_MAGIC); #endif /* DEBUG */ - return XFS_ERROR(EINVAL); + error = XFS_ERROR(EINVAL); + goto out_brelse; } /* @@ -877,24 +813,22 @@ xfs_iread( * specific information. * Otherwise, just get the truly permanent information. */ - if (dip->di_core.di_mode) { - xfs_dinode_from_disk(&ip->i_d, &dip->di_core); + if (dip->di_mode) { + xfs_dinode_from_disk(&ip->i_d, dip); error = xfs_iformat(ip, dip); if (error) { - kmem_zone_free(xfs_inode_zone, ip); - xfs_trans_brelse(tp, bp); #ifdef DEBUG xfs_fs_cmn_err(CE_ALERT, mp, "xfs_iread: " "xfs_iformat() returned error %d", error); #endif /* DEBUG */ - return error; + goto out_brelse; } } else { - ip->i_d.di_magic = be16_to_cpu(dip->di_core.di_magic); - ip->i_d.di_version = dip->di_core.di_version; - ip->i_d.di_gen = be32_to_cpu(dip->di_core.di_gen); - ip->i_d.di_flushiter = be16_to_cpu(dip->di_core.di_flushiter); + ip->i_d.di_magic = be16_to_cpu(dip->di_magic); + ip->i_d.di_version = dip->di_version; + ip->i_d.di_gen = be32_to_cpu(dip->di_gen); + ip->i_d.di_flushiter = be16_to_cpu(dip->di_flushiter); /* * Make sure to pull in the mode here as well in * case the inode is released without being used. @@ -911,8 +845,6 @@ xfs_iread( XFS_IFORK_DSIZE(ip) / (uint)sizeof(xfs_bmbt_rec_t); } - INIT_LIST_HEAD(&ip->i_reclaim); - /* * The inode format changed when we moved the link count and * made it 32 bits long. If this is an old format inode, @@ -924,7 +856,7 @@ xfs_iread( * the new format. We don't change the version number so that we * can distinguish this from a real new format inode. */ - if (ip->i_d.di_version == XFS_DINODE_VERSION_1) { + if (ip->i_d.di_version == 1) { ip->i_d.di_nlink = ip->i_d.di_onlink; ip->i_d.di_onlink = 0; ip->i_d.di_projid = 0; @@ -938,7 +870,7 @@ xfs_iread( * around for a while. This helps to keep recently accessed * meta-data in-core longer. */ - XFS_BUF_SET_REF(bp, XFS_INO_REF); + XFS_BUF_SET_REF(bp, XFS_INO_REF); /* * Use xfs_trans_brelse() to release the buffer containing the @@ -953,9 +885,9 @@ xfs_iread( * to worry about the inode being changed just because we released * the buffer. */ + out_brelse: xfs_trans_brelse(tp, bp); - *ipp = ip; - return 0; + return error; } /* @@ -1049,6 +981,7 @@ xfs_ialloc( uint flags; int error; timespec_t tv; + int filestreams = 0; /* * Call the space management code to pick @@ -1056,9 +989,8 @@ xfs_ialloc( */ error = xfs_dialloc(tp, pip ? pip->i_ino : 0, mode, okalloc, ialloc_context, call_again, &ino); - if (error != 0) { + if (error) return error; - } if (*call_again || ino == NULLFSINO) { *ipp = NULL; return 0; @@ -1072,9 +1004,8 @@ xfs_ialloc( */ error = xfs_trans_iget(tp->t_mountp, tp, ino, XFS_IGET_CREATE, XFS_ILOCK_EXCL, &ip); - if (error != 0) { + if (error) return error; - } ASSERT(ip != NULL); ip->i_d.di_mode = (__uint16_t)mode; @@ -1093,8 +1024,8 @@ xfs_ialloc( * here rather than here and in the flush/logging code. */ if (xfs_sb_version_hasnlink(&tp->t_mountp->m_sb) && - ip->i_d.di_version == XFS_DINODE_VERSION_1) { - ip->i_d.di_version = XFS_DINODE_VERSION_2; + ip->i_d.di_version == 1) { + ip->i_d.di_version = 2; /* * We've already zeroed the old link count, the projid field, * and the pad field. @@ -1104,7 +1035,7 @@ xfs_ialloc( /* * Project ids won't be stored on disk if we are using a version 1 inode. */ - if ((prid != 0) && (ip->i_d.di_version == XFS_DINODE_VERSION_1)) + if ((prid != 0) && (ip->i_d.di_version == 1)) xfs_bump_ino_vers2(tp, ip); if (pip && XFS_INHERIT_GID(pip)) { @@ -1155,13 +1086,12 @@ xfs_ialloc( flags |= XFS_ILOG_DEV; break; case S_IFREG: - if (pip && xfs_inode_is_filestream(pip)) { - error = xfs_filestream_associate(pip, ip); - if (error < 0) - return -error; - if (!error) - xfs_iflags_set(ip, XFS_IFILESTREAM); - } + /* + * we can't set up filestreams until after the VFS inode + * is set up properly. + */ + if (pip && xfs_inode_is_filestream(pip)) + filestreams = 1; /* fall through */ case S_IFDIR: if (pip && (pip->i_d.di_flags & XFS_DIFLAG_ANY)) { @@ -1227,6 +1157,15 @@ xfs_ialloc( /* now that we have an i_mode we can setup inode ops and unlock */ xfs_setup_inode(ip); + /* now we have set up the vfs inode we can associate the filestream */ + if (filestreams) { + error = xfs_filestream_associate(pip, ip); + if (error < 0) + return -error; + if (!error) + xfs_iflags_set(ip, XFS_IFILESTREAM); + } + *ipp = ip; return 0; } @@ -1383,8 +1322,8 @@ xfs_itrunc_trace( * direct I/O with the truncate operation. Also, because we hold * the IOLOCK in exclusive mode, we prevent new direct I/Os from being * started until the truncate completes and drops the lock. Essentially, - * the vn_iowait() call forms an I/O barrier that provides strict ordering - * between direct I/Os and the truncate operation. + * the xfs_ioend_wait() call forms an I/O barrier that provides strict + * ordering between direct I/Os and the truncate operation. * * The flags parameter can have either the value XFS_ITRUNC_DEFINITE * or XFS_ITRUNC_MAYBE. The XFS_ITRUNC_MAYBE value should be used @@ -1414,8 +1353,8 @@ xfs_itruncate_start( mp = ip->i_mount; /* wait for the completion of any pending DIOs */ - if (new_size < ip->i_size) - vn_iowait(ip); + if (new_size == 0 || new_size < ip->i_size) + xfs_ioend_wait(ip); /* * Call toss_pages or flushinval_pages to get rid of pages @@ -1726,8 +1665,14 @@ xfs_itruncate_finish( xfs_trans_ijoin(ntp, ip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL); xfs_trans_ihold(ntp, ip); - if (!error) - error = xfs_trans_reserve(ntp, 0, + if (error) + return error; + /* + * transaction commit worked ok so we can drop the extra ticket + * reference that we gained in xfs_trans_dup() + */ + xfs_log_ticket_put(ntp->t_ticket); + error = xfs_trans_reserve(ntp, 0, XFS_ITRUNCATE_LOG_RES(mp), 0, XFS_TRANS_PERM_LOG_RES, XFS_ITRUNCATE_LOG_COUNT); @@ -1781,13 +1726,10 @@ xfs_iunlink( xfs_dinode_t *dip; xfs_buf_t *agibp; xfs_buf_t *ibp; - xfs_agnumber_t agno; - xfs_daddr_t agdaddr; xfs_agino_t agino; short bucket_index; int offset; int error; - int agi_ok; ASSERT(ip->i_d.di_nlink == 0); ASSERT(ip->i_d.di_mode != 0); @@ -1795,31 +1737,15 @@ xfs_iunlink( mp = tp->t_mountp; - agno = XFS_INO_TO_AGNO(mp, ip->i_ino); - agdaddr = XFS_AG_DADDR(mp, agno, XFS_AGI_DADDR(mp)); - /* * Get the agi buffer first. It ensures lock ordering * on the list. */ - error = xfs_trans_read_buf(mp, tp, mp->m_ddev_targp, agdaddr, - XFS_FSS_TO_BB(mp, 1), 0, &agibp); + error = xfs_read_agi(mp, tp, XFS_INO_TO_AGNO(mp, ip->i_ino), &agibp); if (error) return error; - - /* - * Validate the magic number of the agi block. - */ agi = XFS_BUF_TO_AGI(agibp); - agi_ok = - be32_to_cpu(agi->agi_magicnum) == XFS_AGI_MAGIC && - XFS_AGI_GOOD_VERSION(be32_to_cpu(agi->agi_versionnum)); - if (unlikely(XFS_TEST_ERROR(!agi_ok, mp, XFS_ERRTAG_IUNLINK, - XFS_RANDOM_IUNLINK))) { - XFS_CORRUPTION_ERROR("xfs_iunlink", XFS_ERRLEVEL_LOW, mp, agi); - xfs_trans_brelse(tp, agibp); - return XFS_ERROR(EFSCORRUPTED); - } + /* * Get the index into the agi hash table for the * list this inode will go on. @@ -1837,14 +1763,14 @@ xfs_iunlink( * Here we put the head pointer into our next pointer, * and then we fall through to point the head at us. */ - error = xfs_itobp(mp, tp, ip, &dip, &ibp, 0, 0, XFS_BUF_LOCK); + error = xfs_itobp(mp, tp, ip, &dip, &ibp, XFS_BUF_LOCK); if (error) return error; ASSERT(be32_to_cpu(dip->di_next_unlinked) == NULLAGINO); /* both on-disk, don't endian flip twice */ dip->di_next_unlinked = agi->agi_unlinked[bucket_index]; - offset = ip->i_boffset + + offset = ip->i_imap.im_boffset + offsetof(xfs_dinode_t, di_next_unlinked); xfs_trans_inode_buf(tp, ibp); xfs_trans_log_buf(tp, ibp, offset, @@ -1879,7 +1805,6 @@ xfs_iunlink_remove( xfs_buf_t *agibp; xfs_buf_t *ibp; xfs_agnumber_t agno; - xfs_daddr_t agdaddr; xfs_agino_t agino; xfs_agino_t next_agino; xfs_buf_t *last_ibp; @@ -1887,45 +1812,20 @@ xfs_iunlink_remove( short bucket_index; int offset, last_offset = 0; int error; - int agi_ok; - /* - * First pull the on-disk inode from the AGI unlinked list. - */ mp = tp->t_mountp; - agno = XFS_INO_TO_AGNO(mp, ip->i_ino); - agdaddr = XFS_AG_DADDR(mp, agno, XFS_AGI_DADDR(mp)); /* * Get the agi buffer first. It ensures lock ordering * on the list. */ - error = xfs_trans_read_buf(mp, tp, mp->m_ddev_targp, agdaddr, - XFS_FSS_TO_BB(mp, 1), 0, &agibp); - if (error) { - cmn_err(CE_WARN, - "xfs_iunlink_remove: xfs_trans_read_buf() returned an error %d on %s. Returning error.", - error, mp->m_fsname); + error = xfs_read_agi(mp, tp, agno, &agibp); + if (error) return error; - } - /* - * Validate the magic number of the agi block. - */ + agi = XFS_BUF_TO_AGI(agibp); - agi_ok = - be32_to_cpu(agi->agi_magicnum) == XFS_AGI_MAGIC && - XFS_AGI_GOOD_VERSION(be32_to_cpu(agi->agi_versionnum)); - if (unlikely(XFS_TEST_ERROR(!agi_ok, mp, XFS_ERRTAG_IUNLINK_REMOVE, - XFS_RANDOM_IUNLINK_REMOVE))) { - XFS_CORRUPTION_ERROR("xfs_iunlink_remove", XFS_ERRLEVEL_LOW, - mp, agi); - xfs_trans_brelse(tp, agibp); - cmn_err(CE_WARN, - "xfs_iunlink_remove: XFS_TEST_ERROR() returned an error on %s. Returning EFSCORRUPTED.", - mp->m_fsname); - return XFS_ERROR(EFSCORRUPTED); - } + /* * Get the index into the agi hash table for the * list this inode will go on. @@ -1945,7 +1845,7 @@ xfs_iunlink_remove( * of dealing with the buffer when there is no need to * change it. */ - error = xfs_itobp(mp, tp, ip, &dip, &ibp, 0, 0, XFS_BUF_LOCK); + error = xfs_itobp(mp, tp, ip, &dip, &ibp, XFS_BUF_LOCK); if (error) { cmn_err(CE_WARN, "xfs_iunlink_remove: xfs_itobp() returned an error %d on %s. Returning error.", @@ -1956,7 +1856,7 @@ xfs_iunlink_remove( ASSERT(next_agino != 0); if (next_agino != NULLAGINO) { dip->di_next_unlinked = cpu_to_be32(NULLAGINO); - offset = ip->i_boffset + + offset = ip->i_imap.im_boffset + offsetof(xfs_dinode_t, di_next_unlinked); xfs_trans_inode_buf(tp, ibp); xfs_trans_log_buf(tp, ibp, offset, @@ -1992,7 +1892,7 @@ xfs_iunlink_remove( } next_ino = XFS_AGINO_TO_INO(mp, agno, next_agino); error = xfs_inotobp(mp, tp, next_ino, &last_dip, - &last_ibp, &last_offset); + &last_ibp, &last_offset, 0); if (error) { cmn_err(CE_WARN, "xfs_iunlink_remove: xfs_inotobp() returned an error %d on %s. Returning error.", @@ -2007,7 +1907,7 @@ xfs_iunlink_remove( * Now last_ibp points to the buffer previous to us on * the unlinked list. Pull us from the list. */ - error = xfs_itobp(mp, tp, ip, &dip, &ibp, 0, 0, XFS_BUF_LOCK); + error = xfs_itobp(mp, tp, ip, &dip, &ibp, XFS_BUF_LOCK); if (error) { cmn_err(CE_WARN, "xfs_iunlink_remove: xfs_itobp() returned an error %d on %s. Returning error.", @@ -2019,7 +1919,7 @@ xfs_iunlink_remove( ASSERT(next_agino != agino); if (next_agino != NULLAGINO) { dip->di_next_unlinked = cpu_to_be32(NULLAGINO); - offset = ip->i_boffset + + offset = ip->i_imap.im_boffset + offsetof(xfs_dinode_t, di_next_unlinked); xfs_trans_inode_buf(tp, ibp); xfs_trans_log_buf(tp, ibp, offset, @@ -2160,9 +2060,9 @@ xfs_ifree_cluster( iip = (xfs_inode_log_item_t *)lip; ASSERT(iip->ili_logged == 1); lip->li_cb = (void(*)(xfs_buf_t*,xfs_log_item_t*)) xfs_istale_done; - spin_lock(&mp->m_ail_lock); - iip->ili_flush_lsn = iip->ili_item.li_lsn; - spin_unlock(&mp->m_ail_lock); + xfs_trans_ail_copy_lsn(mp->m_ail, + &iip->ili_flush_lsn, + &iip->ili_item.li_lsn); xfs_iflags_set(iip->ili_inode, XFS_ISTALE); pre_flushed++; } @@ -2183,9 +2083,8 @@ xfs_ifree_cluster( iip->ili_last_fields = iip->ili_format.ilf_fields; iip->ili_format.ilf_fields = 0; iip->ili_logged = 1; - spin_lock(&mp->m_ail_lock); - iip->ili_flush_lsn = iip->ili_item.li_lsn; - spin_unlock(&mp->m_ail_lock); + xfs_trans_ail_copy_lsn(mp->m_ail, &iip->ili_flush_lsn, + &iip->ili_item.li_lsn); xfs_buf_attach_iodone(bp, (void(*)(xfs_buf_t*,xfs_log_item_t*)) @@ -2263,7 +2162,7 @@ xfs_ifree( xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); - error = xfs_itobp(ip->i_mount, tp, ip, &dip, &ibp, 0, 0, XFS_BUF_LOCK); + error = xfs_itobp(ip->i_mount, tp, ip, &dip, &ibp, XFS_BUF_LOCK); if (error) return error; @@ -2279,7 +2178,7 @@ xfs_ifree( * This is a temporary hack that would require a proper fix * in the future. */ - dip->di_core.di_mode = 0; + dip->di_mode = 0; if (delete) { xfs_ifree_cluster(ip, tp, first_ino); @@ -2312,9 +2211,10 @@ xfs_iroot_realloc( int rec_diff, int whichfork) { + struct xfs_mount *mp = ip->i_mount; int cur_max; xfs_ifork_t *ifp; - xfs_bmbt_block_t *new_broot; + struct xfs_btree_block *new_broot; int new_max; size_t new_size; char *np; @@ -2335,8 +2235,7 @@ xfs_iroot_realloc( */ if (ifp->if_broot_bytes == 0) { new_size = (size_t)XFS_BMAP_BROOT_SPACE_CALC(rec_diff); - ifp->if_broot = (xfs_bmbt_block_t*)kmem_alloc(new_size, - KM_SLEEP); + ifp->if_broot = kmem_alloc(new_size, KM_SLEEP); ifp->if_broot_bytes = (int)new_size; return; } @@ -2347,18 +2246,16 @@ xfs_iroot_realloc( * location. The records don't change location because * they are kept butted up against the btree block header. */ - cur_max = XFS_BMAP_BROOT_MAXRECS(ifp->if_broot_bytes); + cur_max = xfs_bmbt_maxrecs(mp, ifp->if_broot_bytes, 0); new_max = cur_max + rec_diff; new_size = (size_t)XFS_BMAP_BROOT_SPACE_CALC(new_max); - ifp->if_broot = (xfs_bmbt_block_t *) - kmem_realloc(ifp->if_broot, - new_size, + ifp->if_broot = kmem_realloc(ifp->if_broot, new_size, (size_t)XFS_BMAP_BROOT_SPACE_CALC(cur_max), /* old size */ KM_SLEEP); - op = (char *)XFS_BMAP_BROOT_PTR_ADDR(ifp->if_broot, 1, - ifp->if_broot_bytes); - np = (char *)XFS_BMAP_BROOT_PTR_ADDR(ifp->if_broot, 1, - (int)new_size); + op = (char *)XFS_BMAP_BROOT_PTR_ADDR(mp, ifp->if_broot, 1, + ifp->if_broot_bytes); + np = (char *)XFS_BMAP_BROOT_PTR_ADDR(mp, ifp->if_broot, 1, + (int)new_size); ifp->if_broot_bytes = (int)new_size; ASSERT(ifp->if_broot_bytes <= XFS_IFORK_SIZE(ip, whichfork) + XFS_BROOT_SIZE_ADJ); @@ -2372,7 +2269,7 @@ xfs_iroot_realloc( * records, just get rid of the root and clear the status bit. */ ASSERT((ifp->if_broot != NULL) && (ifp->if_broot_bytes > 0)); - cur_max = XFS_BMAP_BROOT_MAXRECS(ifp->if_broot_bytes); + cur_max = xfs_bmbt_maxrecs(mp, ifp->if_broot_bytes, 0); new_max = cur_max + rec_diff; ASSERT(new_max >= 0); if (new_max > 0) @@ -2380,11 +2277,11 @@ xfs_iroot_realloc( else new_size = 0; if (new_size > 0) { - new_broot = (xfs_bmbt_block_t *)kmem_alloc(new_size, KM_SLEEP); + new_broot = kmem_alloc(new_size, KM_SLEEP); /* * First copy over the btree block header. */ - memcpy(new_broot, ifp->if_broot, sizeof(xfs_bmbt_block_t)); + memcpy(new_broot, ifp->if_broot, XFS_BTREE_LBLOCK_LEN); } else { new_broot = NULL; ifp->if_flags &= ~XFS_IFBROOT; @@ -2397,18 +2294,16 @@ xfs_iroot_realloc( /* * First copy the records. */ - op = (char *)XFS_BMAP_BROOT_REC_ADDR(ifp->if_broot, 1, - ifp->if_broot_bytes); - np = (char *)XFS_BMAP_BROOT_REC_ADDR(new_broot, 1, - (int)new_size); + op = (char *)XFS_BMBT_REC_ADDR(mp, ifp->if_broot, 1); + np = (char *)XFS_BMBT_REC_ADDR(mp, new_broot, 1); memcpy(np, op, new_max * (uint)sizeof(xfs_bmbt_rec_t)); /* * Then copy the pointers. */ - op = (char *)XFS_BMAP_BROOT_PTR_ADDR(ifp->if_broot, 1, + op = (char *)XFS_BMAP_BROOT_PTR_ADDR(mp, ifp->if_broot, 1, ifp->if_broot_bytes); - np = (char *)XFS_BMAP_BROOT_PTR_ADDR(new_broot, 1, + np = (char *)XFS_BMAP_BROOT_PTR_ADDR(mp, new_broot, 1, (int)new_size); memcpy(np, op, new_max * (uint)sizeof(xfs_dfsbno_t)); } @@ -2511,64 +2406,6 @@ xfs_idata_realloc( ASSERT(ifp->if_bytes <= XFS_IFORK_SIZE(ip, whichfork)); } - - - -/* - * Map inode to disk block and offset. - * - * mp -- the mount point structure for the current file system - * tp -- the current transaction - * ino -- the inode number of the inode to be located - * imap -- this structure is filled in with the information necessary - * to retrieve the given inode from disk - * flags -- flags to pass to xfs_dilocate indicating whether or not - * lookups in the inode btree were OK or not - */ -int -xfs_imap( - xfs_mount_t *mp, - xfs_trans_t *tp, - xfs_ino_t ino, - xfs_imap_t *imap, - uint flags) -{ - xfs_fsblock_t fsbno; - int len; - int off; - int error; - - fsbno = imap->im_blkno ? - XFS_DADDR_TO_FSB(mp, imap->im_blkno) : NULLFSBLOCK; - error = xfs_dilocate(mp, tp, ino, &fsbno, &len, &off, flags); - if (error) - return error; - - imap->im_blkno = XFS_FSB_TO_DADDR(mp, fsbno); - imap->im_len = XFS_FSB_TO_BB(mp, len); - imap->im_agblkno = XFS_FSB_TO_AGBNO(mp, fsbno); - imap->im_ioffset = (ushort)off; - imap->im_boffset = (ushort)(off << mp->m_sb.sb_inodelog); - - /* - * If the inode number maps to a block outside the bounds - * of the file system then return NULL rather than calling - * read_buf and panicing when we get an error from the - * driver. - */ - if ((imap->im_blkno + imap->im_len) > - XFS_FSB_TO_BB(mp, mp->m_sb.sb_dblocks)) { - xfs_fs_cmn_err(CE_ALERT, mp, "xfs_imap: " - "(imap->im_blkno (0x%llx) + imap->im_len (0x%llx)) > " - " XFS_FSB_TO_BB(mp, mp->m_sb.sb_dblocks) (0x%llx)", - (unsigned long long) imap->im_blkno, - (unsigned long long) imap->im_len, - XFS_FSB_TO_BB(mp, mp->m_sb.sb_dblocks)); - return EINVAL; - } - return 0; -} - void xfs_idestroy_fork( xfs_inode_t *ip, @@ -2613,70 +2450,6 @@ xfs_idestroy_fork( } /* - * This is called free all the memory associated with an inode. - * It must free the inode itself and any buffers allocated for - * if_extents/if_data and if_broot. It must also free the lock - * associated with the inode. - */ -void -xfs_idestroy( - xfs_inode_t *ip) -{ - switch (ip->i_d.di_mode & S_IFMT) { - case S_IFREG: - case S_IFDIR: - case S_IFLNK: - xfs_idestroy_fork(ip, XFS_DATA_FORK); - break; - } - if (ip->i_afp) - xfs_idestroy_fork(ip, XFS_ATTR_FORK); - mrfree(&ip->i_lock); - mrfree(&ip->i_iolock); - -#ifdef XFS_INODE_TRACE - ktrace_free(ip->i_trace); -#endif -#ifdef XFS_BMAP_TRACE - ktrace_free(ip->i_xtrace); -#endif -#ifdef XFS_BMBT_TRACE - ktrace_free(ip->i_btrace); -#endif -#ifdef XFS_RW_TRACE - ktrace_free(ip->i_rwtrace); -#endif -#ifdef XFS_ILOCK_TRACE - ktrace_free(ip->i_lock_trace); -#endif -#ifdef XFS_DIR2_TRACE - ktrace_free(ip->i_dir_trace); -#endif - if (ip->i_itemp) { - /* - * Only if we are shutting down the fs will we see an - * inode still in the AIL. If it is there, we should remove - * it to prevent a use-after-free from occurring. - */ - xfs_mount_t *mp = ip->i_mount; - xfs_log_item_t *lip = &ip->i_itemp->ili_item; - - ASSERT(((lip->li_flags & XFS_LI_IN_AIL) == 0) || - XFS_FORCED_SHUTDOWN(ip->i_mount)); - if (lip->li_flags & XFS_LI_IN_AIL) { - spin_lock(&mp->m_ail_lock); - if (lip->li_flags & XFS_LI_IN_AIL) - xfs_trans_delete_ail(mp, lip); - else - spin_unlock(&mp->m_ail_lock); - } - xfs_inode_item_destroy(ip); - } - kmem_zone_free(xfs_inode_zone, ip); -} - - -/* * Increment the pin count of the given buffer. * This value is protected by ipinlock spinlock in the mount structure. */ @@ -2880,7 +2653,7 @@ xfs_iflush_fork( ASSERT(ifp->if_broot_bytes <= (XFS_IFORK_SIZE(ip, whichfork) + XFS_BROOT_SIZE_ADJ)); - xfs_bmbt_to_bmdr(ifp->if_broot, ifp->if_broot_bytes, + xfs_bmbt_to_bmdr(mp, ifp->if_broot, ifp->if_broot_bytes, (xfs_bmdr_block_t *)cp, XFS_DFORK_SIZE(dip, mp, whichfork)); } @@ -2889,15 +2662,16 @@ xfs_iflush_fork( case XFS_DINODE_FMT_DEV: if (iip->ili_format.ilf_fields & XFS_ILOG_DEV) { ASSERT(whichfork == XFS_DATA_FORK); - dip->di_u.di_dev = cpu_to_be32(ip->i_df.if_u2.if_rdev); + xfs_dinode_put_rdev(dip, ip->i_df.if_u2.if_rdev); } break; case XFS_DINODE_FMT_UUID: if (iip->ili_format.ilf_fields & XFS_ILOG_UUID) { ASSERT(whichfork == XFS_DATA_FORK); - memcpy(&dip->di_u.di_muuid, &ip->i_df.if_u2.if_uuid, - sizeof(uuid_t)); + memcpy(XFS_DFORK_DPTR(dip), + &ip->i_df.if_u2.if_uuid, + sizeof(uuid_t)); } break; @@ -3030,7 +2804,6 @@ cluster_corrupt_out: XFS_BUF_CLR_BDSTRAT_FUNC(bp); XFS_BUF_UNDONE(bp); XFS_BUF_STALE(bp); - XFS_BUF_SHUT(bp); XFS_BUF_ERROR(bp,EIO); xfs_biodone(bp); } else { @@ -3172,7 +2945,7 @@ xfs_iflush( /* * Get the buffer containing the on-disk inode. */ - error = xfs_itobp(mp, NULL, ip, &dip, &bp, 0, 0, + error = xfs_itobp(mp, NULL, ip, &dip, &bp, noblock ? XFS_BUF_TRYLOCK : XFS_BUF_LOCK); if (error || !bp) { xfs_ifunlock(ip); @@ -3253,7 +3026,7 @@ xfs_iflush_int( } /* set *dip = inode's place in the buffer */ - dip = (xfs_dinode_t *)xfs_buf_offset(bp, ip->i_boffset); + dip = (xfs_dinode_t *)xfs_buf_offset(bp, ip->i_imap.im_boffset); /* * Clear i_update_core before copying out the data. @@ -3275,11 +3048,11 @@ xfs_iflush_int( */ xfs_synchronize_atime(ip); - if (XFS_TEST_ERROR(be16_to_cpu(dip->di_core.di_magic) != XFS_DINODE_MAGIC, + if (XFS_TEST_ERROR(be16_to_cpu(dip->di_magic) != XFS_DINODE_MAGIC, mp, XFS_ERRTAG_IFLUSH_1, XFS_RANDOM_IFLUSH_1)) { xfs_cmn_err(XFS_PTAG_IFLUSH, CE_ALERT, mp, "xfs_iflush: Bad inode %Lu magic number 0x%x, ptr 0x%p", - ip->i_ino, be16_to_cpu(dip->di_core.di_magic), dip); + ip->i_ino, be16_to_cpu(dip->di_magic), dip); goto corrupt_out; } if (XFS_TEST_ERROR(ip->i_d.di_magic != XFS_DINODE_MAGIC, @@ -3342,7 +3115,7 @@ xfs_iflush_int( * because if the inode is dirty at all the core must * be. */ - xfs_dinode_to_disk(&dip->di_core, &ip->i_d); + xfs_dinode_to_disk(dip, &ip->i_d); /* Wrap, we never let the log put out DI_MAX_FLUSH */ if (ip->i_d.di_flushiter == DI_MAX_FLUSH) @@ -3354,28 +3127,27 @@ xfs_iflush_int( * convert back to the old inode format. If the superblock version * has been updated, then make the conversion permanent. */ - ASSERT(ip->i_d.di_version == XFS_DINODE_VERSION_1 || - xfs_sb_version_hasnlink(&mp->m_sb)); - if (ip->i_d.di_version == XFS_DINODE_VERSION_1) { + ASSERT(ip->i_d.di_version == 1 || xfs_sb_version_hasnlink(&mp->m_sb)); + if (ip->i_d.di_version == 1) { if (!xfs_sb_version_hasnlink(&mp->m_sb)) { /* * Convert it back. */ ASSERT(ip->i_d.di_nlink <= XFS_MAXLINK_1); - dip->di_core.di_onlink = cpu_to_be16(ip->i_d.di_nlink); + dip->di_onlink = cpu_to_be16(ip->i_d.di_nlink); } else { /* * The superblock version has already been bumped, * so just make the conversion to the new inode * format permanent. */ - ip->i_d.di_version = XFS_DINODE_VERSION_2; - dip->di_core.di_version = XFS_DINODE_VERSION_2; + ip->i_d.di_version = 2; + dip->di_version = 2; ip->i_d.di_onlink = 0; - dip->di_core.di_onlink = 0; + dip->di_onlink = 0; memset(&(ip->i_d.di_pad[0]), 0, sizeof(ip->i_d.di_pad)); - memset(&(dip->di_core.di_pad[0]), 0, - sizeof(dip->di_core.di_pad)); + memset(&(dip->di_pad[0]), 0, + sizeof(dip->di_pad)); ASSERT(ip->i_d.di_projid == 0); } } @@ -3418,10 +3190,8 @@ xfs_iflush_int( iip->ili_format.ilf_fields = 0; iip->ili_logged = 1; - ASSERT(sizeof(xfs_lsn_t) == 8); /* don't lock if it shrinks */ - spin_lock(&mp->m_ail_lock); - iip->ili_flush_lsn = iip->ili_item.li_lsn; - spin_unlock(&mp->m_ail_lock); + xfs_trans_ail_copy_lsn(mp->m_ail, &iip->ili_flush_lsn, + &iip->ili_item.li_lsn); /* * Attach the function xfs_iflush_done to the inode's @@ -3459,45 +3229,8 @@ corrupt_out: } -/* - * Flush all inactive inodes in mp. - */ -void -xfs_iflush_all( - xfs_mount_t *mp) -{ - xfs_inode_t *ip; - - again: - XFS_MOUNT_ILOCK(mp); - ip = mp->m_inodes; - if (ip == NULL) - goto out; - - do { - /* Make sure we skip markers inserted by sync */ - if (ip->i_mount == NULL) { - ip = ip->i_mnext; - continue; - } - - if (!VFS_I(ip)) { - XFS_MOUNT_IUNLOCK(mp); - xfs_finish_reclaim(ip, 0, XFS_IFLUSH_ASYNC); - goto again; - } - - ASSERT(vn_count(VFS_I(ip)) == 0); - - ip = ip->i_mnext; - } while (ip != mp->m_inodes); - out: - XFS_MOUNT_IUNLOCK(mp); -} #ifdef XFS_ILOCK_TRACE -ktrace_t *xfs_ilock_trace_buf; - void xfs_ilock_trace(xfs_inode_t *ip, int lock, unsigned int lockflags, inst_t *ra) { diff --git a/fs/xfs/xfs_inode.h b/fs/xfs/xfs_inode.h index 1420c49674d..1f175fa34b2 100644 --- a/fs/xfs/xfs_inode.h +++ b/fs/xfs/xfs_inode.h @@ -19,8 +19,7 @@ #define __XFS_INODE_H__ struct xfs_dinode; -struct xfs_dinode_core; - +struct xfs_inode; /* * Fork identifiers. @@ -63,7 +62,7 @@ typedef struct xfs_ext_irec { typedef struct xfs_ifork { int if_bytes; /* bytes in if_u1 */ int if_real_bytes; /* bytes allocated in if_u1 */ - xfs_bmbt_block_t *if_broot; /* file's incore btree root */ + struct xfs_btree_block *if_broot; /* file's incore btree root */ short if_broot_bytes; /* bytes allocated for root */ unsigned char if_flags; /* per-fork flags */ unsigned char if_ext_max; /* max # of extent records */ @@ -84,52 +83,14 @@ typedef struct xfs_ifork { } xfs_ifork_t; /* - * Flags for xfs_ichgtime(). + * Inode location information. Stored in the inode and passed to + * xfs_imap_to_bp() to get a buffer and dinode for a given inode. */ -#define XFS_ICHGTIME_MOD 0x1 /* data fork modification timestamp */ -#define XFS_ICHGTIME_CHG 0x2 /* inode field change timestamp */ - -/* - * Per-fork incore inode flags. - */ -#define XFS_IFINLINE 0x01 /* Inline data is read in */ -#define XFS_IFEXTENTS 0x02 /* All extent pointers are read in */ -#define XFS_IFBROOT 0x04 /* i_broot points to the bmap b-tree root */ -#define XFS_IFEXTIREC 0x08 /* Indirection array of extent blocks */ - -/* - * Flags for xfs_itobp(), xfs_imap() and xfs_dilocate(). - */ -#define XFS_IMAP_LOOKUP 0x1 -#define XFS_IMAP_BULKSTAT 0x2 - -#ifdef __KERNEL__ -struct bhv_desc; -struct cred; -struct ktrace; -struct xfs_buf; -struct xfs_bmap_free; -struct xfs_bmbt_irec; -struct xfs_bmbt_block; -struct xfs_inode; -struct xfs_inode_log_item; -struct xfs_mount; -struct xfs_trans; -struct xfs_dquot; - -#if defined(XFS_ILOCK_TRACE) -#define XFS_ILOCK_KTRACE_SIZE 32 -extern ktrace_t *xfs_ilock_trace_buf; -extern void xfs_ilock_trace(struct xfs_inode *, int, unsigned int, inst_t *); -#else -#define xfs_ilock_trace(i,n,f,ra) -#endif - -typedef struct dm_attrs_s { - __uint32_t da_dmevmask; /* DMIG event mask */ - __uint16_t da_dmstate; /* DMIG state info */ - __uint16_t da_pad; /* DMIG extra padding */ -} dm_attrs_t; +struct xfs_imap { + xfs_daddr_t im_blkno; /* starting BB of inode chunk */ + ushort im_len; /* length in BBs of inode chunk */ + ushort im_boffset; /* inode offset in block in bytes */ +}; /* * This is the xfs in-core inode structure. @@ -160,7 +121,7 @@ typedef struct xfs_ictimestamp { } xfs_ictimestamp_t; /* - * NOTE: This structure must be kept identical to struct xfs_dinode_core + * NOTE: This structure must be kept identical to struct xfs_dinode * in xfs_dinode.h except for the endianess annotations. */ typedef struct xfs_icdinode { @@ -191,27 +152,97 @@ typedef struct xfs_icdinode { __uint32_t di_gen; /* generation number */ } xfs_icdinode_t; -typedef struct { - struct xfs_inode *ip_mnext; /* next inode in mount list */ - struct xfs_inode *ip_mprev; /* ptr to prev inode */ - struct xfs_mount *ip_mount; /* fs mount struct ptr */ -} xfs_iptr_t; +/* + * Flags for xfs_ichgtime(). + */ +#define XFS_ICHGTIME_MOD 0x1 /* data fork modification timestamp */ +#define XFS_ICHGTIME_CHG 0x2 /* inode field change timestamp */ + +/* + * Per-fork incore inode flags. + */ +#define XFS_IFINLINE 0x01 /* Inline data is read in */ +#define XFS_IFEXTENTS 0x02 /* All extent pointers are read in */ +#define XFS_IFBROOT 0x04 /* i_broot points to the bmap b-tree root */ +#define XFS_IFEXTIREC 0x08 /* Indirection array of extent blocks */ + +/* + * Fork handling. + */ + +#define XFS_IFORK_Q(ip) ((ip)->i_d.di_forkoff != 0) +#define XFS_IFORK_BOFF(ip) ((int)((ip)->i_d.di_forkoff << 3)) + +#define XFS_IFORK_PTR(ip,w) \ + ((w) == XFS_DATA_FORK ? \ + &(ip)->i_df : \ + (ip)->i_afp) +#define XFS_IFORK_DSIZE(ip) \ + (XFS_IFORK_Q(ip) ? \ + XFS_IFORK_BOFF(ip) : \ + XFS_LITINO((ip)->i_mount)) +#define XFS_IFORK_ASIZE(ip) \ + (XFS_IFORK_Q(ip) ? \ + XFS_LITINO((ip)->i_mount) - XFS_IFORK_BOFF(ip) : \ + 0) +#define XFS_IFORK_SIZE(ip,w) \ + ((w) == XFS_DATA_FORK ? \ + XFS_IFORK_DSIZE(ip) : \ + XFS_IFORK_ASIZE(ip)) +#define XFS_IFORK_FORMAT(ip,w) \ + ((w) == XFS_DATA_FORK ? \ + (ip)->i_d.di_format : \ + (ip)->i_d.di_aformat) +#define XFS_IFORK_FMT_SET(ip,w,n) \ + ((w) == XFS_DATA_FORK ? \ + ((ip)->i_d.di_format = (n)) : \ + ((ip)->i_d.di_aformat = (n))) +#define XFS_IFORK_NEXTENTS(ip,w) \ + ((w) == XFS_DATA_FORK ? \ + (ip)->i_d.di_nextents : \ + (ip)->i_d.di_anextents) +#define XFS_IFORK_NEXT_SET(ip,w,n) \ + ((w) == XFS_DATA_FORK ? \ + ((ip)->i_d.di_nextents = (n)) : \ + ((ip)->i_d.di_anextents = (n))) + + + +#ifdef __KERNEL__ + +struct bhv_desc; +struct cred; +struct ktrace; +struct xfs_buf; +struct xfs_bmap_free; +struct xfs_bmbt_irec; +struct xfs_inode_log_item; +struct xfs_mount; +struct xfs_trans; +struct xfs_dquot; + +#if defined(XFS_ILOCK_TRACE) +#define XFS_ILOCK_KTRACE_SIZE 32 +extern void xfs_ilock_trace(struct xfs_inode *, int, unsigned int, inst_t *); +#else +#define xfs_ilock_trace(i,n,f,ra) +#endif + +typedef struct dm_attrs_s { + __uint32_t da_dmevmask; /* DMIG event mask */ + __uint16_t da_dmstate; /* DMIG state info */ + __uint16_t da_pad; /* DMIG extra padding */ +} dm_attrs_t; typedef struct xfs_inode { /* Inode linking and identification information. */ - struct xfs_inode *i_mnext; /* next inode in mount list */ - struct xfs_inode *i_mprev; /* ptr to prev inode */ struct xfs_mount *i_mount; /* fs mount struct ptr */ - struct list_head i_reclaim; /* reclaim list */ - struct inode *i_vnode; /* vnode backpointer */ struct xfs_dquot *i_udquot; /* user dquot */ struct xfs_dquot *i_gdquot; /* group dquot */ /* Inode location stuff */ xfs_ino_t i_ino; /* inode number (agno/agino)*/ - xfs_daddr_t i_blkno; /* blkno of inode buffer */ - ushort i_len; /* len of inode buffer */ - ushort i_boffset; /* off of inode in buffer */ + struct xfs_imap i_imap; /* location for xfs_imap() */ /* Extent information. */ xfs_ifork_t *i_afp; /* attribute fork pointer */ @@ -230,7 +261,6 @@ typedef struct xfs_inode { unsigned short i_flags; /* see defined flags below */ unsigned char i_update_core; /* timestamps/size is dirty */ unsigned char i_update_size; /* di_size field is dirty */ - unsigned int i_gen; /* generation count */ unsigned int i_delayed_blks; /* count of delay alloc blks */ xfs_icdinode_t i_d; /* most of ondisk inode */ @@ -238,6 +268,10 @@ typedef struct xfs_inode { xfs_fsize_t i_size; /* in-memory size */ xfs_fsize_t i_new_size; /* size when write completes */ atomic_t i_iocount; /* outstanding I/O count */ + + /* VFS inode */ + struct inode i_vnode; /* embedded VFS inode */ + /* Trace buffers per inode. */ #ifdef XFS_INODE_TRACE struct ktrace *i_trace; /* general inode trace */ @@ -245,7 +279,7 @@ typedef struct xfs_inode { #ifdef XFS_BMAP_TRACE struct ktrace *i_xtrace; /* inode extent list trace */ #endif -#ifdef XFS_BMBT_TRACE +#ifdef XFS_BTREE_TRACE struct ktrace *i_btrace; /* inode bmap btree trace */ #endif #ifdef XFS_RW_TRACE @@ -265,13 +299,30 @@ typedef struct xfs_inode { /* Convert from vfs inode to xfs inode */ static inline struct xfs_inode *XFS_I(struct inode *inode) { - return (struct xfs_inode *)inode->i_private; + return container_of(inode, struct xfs_inode, i_vnode); } /* convert from xfs inode to vfs inode */ static inline struct inode *VFS_I(struct xfs_inode *ip) { - return (struct inode *)ip->i_vnode; + return &ip->i_vnode; +} + +/* + * Get rid of a partially initialized inode. + * + * We have to go through destroy_inode to make sure allocations + * from init_inode_always like the security data are undone. + * + * We mark the inode bad so that it takes the short cut in + * the reclaim path instead of going through the flush path + * which doesn't make sense for an inode that has never seen the + * light of day. + */ +static inline void xfs_destroy_inode(struct xfs_inode *ip) +{ + make_bad_inode(VFS_I(ip)); + return destroy_inode(VFS_I(ip)); } /* @@ -327,65 +378,36 @@ xfs_iflags_test_and_clear(xfs_inode_t *ip, unsigned short flags) spin_unlock(&ip->i_flags_lock); return ret; } -#endif /* __KERNEL__ */ - /* - * Fork handling. + * Manage the i_flush queue embedded in the inode. This completion + * queue synchronizes processes attempting to flush the in-core + * inode back to disk. */ +static inline void xfs_iflock(xfs_inode_t *ip) +{ + wait_for_completion(&ip->i_flush); +} -#define XFS_IFORK_Q(ip) ((ip)->i_d.di_forkoff != 0) -#define XFS_IFORK_BOFF(ip) ((int)((ip)->i_d.di_forkoff << 3)) - -#define XFS_IFORK_PTR(ip,w) \ - ((w) == XFS_DATA_FORK ? \ - &(ip)->i_df : \ - (ip)->i_afp) -#define XFS_IFORK_DSIZE(ip) \ - (XFS_IFORK_Q(ip) ? \ - XFS_IFORK_BOFF(ip) : \ - XFS_LITINO((ip)->i_mount)) -#define XFS_IFORK_ASIZE(ip) \ - (XFS_IFORK_Q(ip) ? \ - XFS_LITINO((ip)->i_mount) - XFS_IFORK_BOFF(ip) : \ - 0) -#define XFS_IFORK_SIZE(ip,w) \ - ((w) == XFS_DATA_FORK ? \ - XFS_IFORK_DSIZE(ip) : \ - XFS_IFORK_ASIZE(ip)) -#define XFS_IFORK_FORMAT(ip,w) \ - ((w) == XFS_DATA_FORK ? \ - (ip)->i_d.di_format : \ - (ip)->i_d.di_aformat) -#define XFS_IFORK_FMT_SET(ip,w,n) \ - ((w) == XFS_DATA_FORK ? \ - ((ip)->i_d.di_format = (n)) : \ - ((ip)->i_d.di_aformat = (n))) -#define XFS_IFORK_NEXTENTS(ip,w) \ - ((w) == XFS_DATA_FORK ? \ - (ip)->i_d.di_nextents : \ - (ip)->i_d.di_anextents) -#define XFS_IFORK_NEXT_SET(ip,w,n) \ - ((w) == XFS_DATA_FORK ? \ - ((ip)->i_d.di_nextents = (n)) : \ - ((ip)->i_d.di_anextents = (n))) +static inline int xfs_iflock_nowait(xfs_inode_t *ip) +{ + return try_wait_for_completion(&ip->i_flush); +} -#ifdef __KERNEL__ +static inline void xfs_ifunlock(xfs_inode_t *ip) +{ + complete(&ip->i_flush); +} /* * In-core inode flags. */ -#define XFS_IGRIO 0x0001 /* inode used for guaranteed rate i/o */ -#define XFS_IUIOSZ 0x0002 /* inode i/o sizes have been explicitly set */ -#define XFS_IQUIESCE 0x0004 /* we have started quiescing for this inode */ -#define XFS_IRECLAIM 0x0008 /* we have started reclaiming this inode */ -#define XFS_ISTALE 0x0010 /* inode has been staled */ -#define XFS_IRECLAIMABLE 0x0020 /* inode can be reclaimed */ -#define XFS_INEW 0x0040 -#define XFS_IFILESTREAM 0x0080 /* inode is in a filestream directory */ -#define XFS_IMODIFIED 0x0100 /* XFS inode state possibly differs */ - /* to the Linux inode state. */ -#define XFS_ITRUNCATED 0x0200 /* truncated down so flush-on-close */ +#define XFS_IRECLAIM 0x0001 /* we have started reclaiming this inode */ +#define XFS_ISTALE 0x0002 /* inode has been staled */ +#define XFS_IRECLAIMABLE 0x0004 /* inode can be reclaimed */ +#define XFS_INEW 0x0008 /* inode has just been allocated */ +#define XFS_IFILESTREAM 0x0010 /* inode is in a filestream directory */ +#define XFS_ITRUNCATED 0x0020 /* truncated down so flush-on-close */ /* * Flags for inode locking. @@ -460,16 +482,8 @@ xfs_iflags_test_and_clear(xfs_inode_t *ip, unsigned short flags) ((pip)->i_d.di_mode & S_ISGID)) /* - * Flags for xfs_iget() - */ -#define XFS_IGET_CREATE 0x1 -#define XFS_IGET_BULKSTAT 0x2 - -/* * xfs_iget.c prototypes. */ -void xfs_ihash_init(struct xfs_mount *); -void xfs_ihash_free(struct xfs_mount *); xfs_inode_t *xfs_inode_incore(struct xfs_mount *, xfs_ino_t, struct xfs_trans *); int xfs_iget(struct xfs_mount *, struct xfs_trans *, xfs_ino_t, @@ -484,25 +498,13 @@ int xfs_isilocked(xfs_inode_t *, uint); uint xfs_ilock_map_shared(xfs_inode_t *); void xfs_iunlock_map_shared(xfs_inode_t *, uint); void xfs_ireclaim(xfs_inode_t *); -int xfs_finish_reclaim(xfs_inode_t *, int, int); -int xfs_finish_reclaim_all(struct xfs_mount *, int); /* * xfs_inode.c prototypes. */ -int xfs_itobp(struct xfs_mount *, struct xfs_trans *, - xfs_inode_t *, struct xfs_dinode **, struct xfs_buf **, - xfs_daddr_t, uint, uint); -int xfs_iread(struct xfs_mount *, struct xfs_trans *, xfs_ino_t, - xfs_inode_t **, xfs_daddr_t, uint); -int xfs_iread_extents(struct xfs_trans *, xfs_inode_t *, int); int xfs_ialloc(struct xfs_trans *, xfs_inode_t *, mode_t, - xfs_nlink_t, xfs_dev_t, struct cred *, xfs_prid_t, + xfs_nlink_t, xfs_dev_t, cred_t *, xfs_prid_t, int, struct xfs_buf **, boolean_t *, xfs_inode_t **); -void xfs_dinode_from_disk(struct xfs_icdinode *, - struct xfs_dinode_core *); -void xfs_dinode_to_disk(struct xfs_dinode_core *, - struct xfs_icdinode *); uint xfs_ip2xflags(struct xfs_inode *); uint xfs_dic2xflags(struct xfs_dinode *); @@ -513,17 +515,10 @@ int xfs_itruncate_finish(struct xfs_trans **, xfs_inode_t *, xfs_fsize_t, int, int); int xfs_iunlink(struct xfs_trans *, xfs_inode_t *); -void xfs_idestroy_fork(xfs_inode_t *, int); -void xfs_idestroy(xfs_inode_t *); -void xfs_idata_realloc(xfs_inode_t *, int, int); -void xfs_iextract(xfs_inode_t *); void xfs_iext_realloc(xfs_inode_t *, int, int); -void xfs_iroot_realloc(xfs_inode_t *, int, int); void xfs_ipin(xfs_inode_t *); void xfs_iunpin(xfs_inode_t *); -int xfs_iextents_copy(xfs_inode_t *, xfs_bmbt_rec_t *, int); int xfs_iflush(xfs_inode_t *, uint); -void xfs_iflush_all(struct xfs_mount *); void xfs_ichgtime(xfs_inode_t *, int); xfs_fsize_t xfs_file_last_byte(xfs_inode_t *); void xfs_lock_inodes(xfs_inode_t **, int, uint); @@ -532,6 +527,77 @@ void xfs_lock_two_inodes(xfs_inode_t *, xfs_inode_t *, uint); void xfs_synchronize_atime(xfs_inode_t *); void xfs_mark_inode_dirty_sync(xfs_inode_t *); +#if defined(XFS_INODE_TRACE) + +#define INODE_TRACE_SIZE 16 /* number of trace entries */ +#define INODE_KTRACE_ENTRY 1 +#define INODE_KTRACE_EXIT 2 +#define INODE_KTRACE_HOLD 3 +#define INODE_KTRACE_REF 4 +#define INODE_KTRACE_RELE 5 + +extern void _xfs_itrace_entry(struct xfs_inode *, const char *, inst_t *); +extern void _xfs_itrace_exit(struct xfs_inode *, const char *, inst_t *); +extern void xfs_itrace_hold(struct xfs_inode *, char *, int, inst_t *); +extern void _xfs_itrace_ref(struct xfs_inode *, char *, int, inst_t *); +extern void xfs_itrace_rele(struct xfs_inode *, char *, int, inst_t *); +#define xfs_itrace_entry(ip) \ + _xfs_itrace_entry(ip, __func__, (inst_t *)__return_address) +#define xfs_itrace_exit(ip) \ + _xfs_itrace_exit(ip, __func__, (inst_t *)__return_address) +#define xfs_itrace_exit_tag(ip, tag) \ + _xfs_itrace_exit(ip, tag, (inst_t *)__return_address) +#define xfs_itrace_ref(ip) \ + _xfs_itrace_ref(ip, __FILE__, __LINE__, (inst_t *)__return_address) + +#else +#define xfs_itrace_entry(a) +#define xfs_itrace_exit(a) +#define xfs_itrace_exit_tag(a, b) +#define xfs_itrace_hold(a, b, c, d) +#define xfs_itrace_ref(a) +#define xfs_itrace_rele(a, b, c, d) +#endif + +#define IHOLD(ip) \ +do { \ + ASSERT(atomic_read(&VFS_I(ip)->i_count) > 0) ; \ + atomic_inc(&(VFS_I(ip)->i_count)); \ + xfs_itrace_hold((ip), __FILE__, __LINE__, (inst_t *)__return_address); \ +} while (0) + +#define IRELE(ip) \ +do { \ + xfs_itrace_rele((ip), __FILE__, __LINE__, (inst_t *)__return_address); \ + iput(VFS_I(ip)); \ +} while (0) + +#endif /* __KERNEL__ */ + +/* + * Flags for xfs_iget() + */ +#define XFS_IGET_CREATE 0x1 +#define XFS_IGET_BULKSTAT 0x2 + +int xfs_inotobp(struct xfs_mount *, struct xfs_trans *, + xfs_ino_t, struct xfs_dinode **, + struct xfs_buf **, int *, uint); +int xfs_itobp(struct xfs_mount *, struct xfs_trans *, + struct xfs_inode *, struct xfs_dinode **, + struct xfs_buf **, uint); +int xfs_iread(struct xfs_mount *, struct xfs_trans *, + struct xfs_inode *, xfs_daddr_t, uint); +void xfs_dinode_from_disk(struct xfs_icdinode *, + struct xfs_dinode *); +void xfs_dinode_to_disk(struct xfs_dinode *, + struct xfs_icdinode *); +void xfs_idestroy_fork(struct xfs_inode *, int); +void xfs_idata_realloc(struct xfs_inode *, int, int); +void xfs_iroot_realloc(struct xfs_inode *, int, int); +int xfs_iread_extents(struct xfs_trans *, struct xfs_inode *, int); +int xfs_iextents_copy(struct xfs_inode *, xfs_bmbt_rec_t *, int); + xfs_bmbt_rec_host_t *xfs_iext_get_ext(xfs_ifork_t *, xfs_extnum_t); void xfs_iext_insert(xfs_ifork_t *, xfs_extnum_t, xfs_extnum_t, xfs_bmbt_irec_t *); @@ -561,7 +627,8 @@ void xfs_iext_irec_update_extoffs(xfs_ifork_t *, int, int); #define xfs_ipincount(ip) ((unsigned int) atomic_read(&ip->i_pincount)) #ifdef DEBUG -void xfs_isize_check(struct xfs_mount *, xfs_inode_t *, xfs_fsize_t); +void xfs_isize_check(struct xfs_mount *, struct xfs_inode *, + xfs_fsize_t); #else /* DEBUG */ #define xfs_isize_check(mp, ip, isize) #endif /* DEBUG */ @@ -576,26 +643,4 @@ extern struct kmem_zone *xfs_ifork_zone; extern struct kmem_zone *xfs_inode_zone; extern struct kmem_zone *xfs_ili_zone; -/* - * Manage the i_flush queue embedded in the inode. This completion - * queue synchronizes processes attempting to flush the in-core - * inode back to disk. - */ -static inline void xfs_iflock(xfs_inode_t *ip) -{ - wait_for_completion(&ip->i_flush); -} - -static inline int xfs_iflock_nowait(xfs_inode_t *ip) -{ - return try_wait_for_completion(&ip->i_flush); -} - -static inline void xfs_ifunlock(xfs_inode_t *ip) -{ - complete(&ip->i_flush); -} - -#endif /* __KERNEL__ */ - #endif /* __XFS_INODE_H__ */ diff --git a/fs/xfs/xfs_inode_item.c b/fs/xfs/xfs_inode_item.c index 97c7452e262..977c4aec587 100644 --- a/fs/xfs/xfs_inode_item.c +++ b/fs/xfs/xfs_inode_item.c @@ -281,7 +281,7 @@ xfs_inode_item_format( xfs_mark_inode_dirty_sync(ip); vecp->i_addr = (xfs_caddr_t)&ip->i_d; - vecp->i_len = sizeof(xfs_dinode_core_t); + vecp->i_len = sizeof(struct xfs_icdinode); XLOG_VEC_SET_TYPE(vecp, XLOG_REG_TYPE_ICORE); vecp++; nvecs++; @@ -296,9 +296,8 @@ xfs_inode_item_format( * has a new version number, then we don't bother converting back. */ mp = ip->i_mount; - ASSERT(ip->i_d.di_version == XFS_DINODE_VERSION_1 || - xfs_sb_version_hasnlink(&mp->m_sb)); - if (ip->i_d.di_version == XFS_DINODE_VERSION_1) { + ASSERT(ip->i_d.di_version == 1 || xfs_sb_version_hasnlink(&mp->m_sb)); + if (ip->i_d.di_version == 1) { if (!xfs_sb_version_hasnlink(&mp->m_sb)) { /* * Convert it back. @@ -311,7 +310,7 @@ xfs_inode_item_format( * so just make the conversion to the new inode * format permanent. */ - ip->i_d.di_version = XFS_DINODE_VERSION_2; + ip->i_d.di_version = 2; ip->i_d.di_onlink = 0; memset(&(ip->i_d.di_pad[0]), 0, sizeof(ip->i_d.di_pad)); } @@ -932,6 +931,7 @@ xfs_inode_item_init( iip->ili_item.li_type = XFS_LI_INODE; iip->ili_item.li_ops = &xfs_inode_item_ops; iip->ili_item.li_mountp = mp; + iip->ili_item.li_ailp = mp->m_ail; iip->ili_inode = ip; /* @@ -942,9 +942,9 @@ xfs_inode_item_init( iip->ili_format.ilf_type = XFS_LI_INODE; iip->ili_format.ilf_ino = ip->i_ino; - iip->ili_format.ilf_blkno = ip->i_blkno; - iip->ili_format.ilf_len = ip->i_len; - iip->ili_format.ilf_boffset = ip->i_boffset; + iip->ili_format.ilf_blkno = ip->i_imap.im_blkno; + iip->ili_format.ilf_len = ip->i_imap.im_len; + iip->ili_format.ilf_boffset = ip->i_imap.im_boffset; } /* @@ -976,9 +976,8 @@ xfs_iflush_done( xfs_buf_t *bp, xfs_inode_log_item_t *iip) { - xfs_inode_t *ip; - - ip = iip->ili_inode; + xfs_inode_t *ip = iip->ili_inode; + struct xfs_ail *ailp = iip->ili_item.li_ailp; /* * We only want to pull the item from the AIL if it is @@ -991,15 +990,12 @@ xfs_iflush_done( */ if (iip->ili_logged && (iip->ili_item.li_lsn == iip->ili_flush_lsn)) { - spin_lock(&ip->i_mount->m_ail_lock); + spin_lock(&ailp->xa_lock); if (iip->ili_item.li_lsn == iip->ili_flush_lsn) { - /* - * xfs_trans_delete_ail() drops the AIL lock. - */ - xfs_trans_delete_ail(ip->i_mount, - (xfs_log_item_t*)iip); + /* xfs_trans_ail_delete() drops the AIL lock. */ + xfs_trans_ail_delete(ailp, (xfs_log_item_t*)iip); } else { - spin_unlock(&ip->i_mount->m_ail_lock); + spin_unlock(&ailp->xa_lock); } } @@ -1031,21 +1027,20 @@ void xfs_iflush_abort( xfs_inode_t *ip) { - xfs_inode_log_item_t *iip; + xfs_inode_log_item_t *iip = ip->i_itemp; xfs_mount_t *mp; iip = ip->i_itemp; mp = ip->i_mount; if (iip) { + struct xfs_ail *ailp = iip->ili_item.li_ailp; if (iip->ili_item.li_flags & XFS_LI_IN_AIL) { - spin_lock(&mp->m_ail_lock); + spin_lock(&ailp->xa_lock); if (iip->ili_item.li_flags & XFS_LI_IN_AIL) { - /* - * xfs_trans_delete_ail() drops the AIL lock. - */ - xfs_trans_delete_ail(mp, (xfs_log_item_t *)iip); + /* xfs_trans_ail_delete() drops the AIL lock. */ + xfs_trans_ail_delete(ailp, (xfs_log_item_t *)iip); } else - spin_unlock(&mp->m_ail_lock); + spin_unlock(&ailp->xa_lock); } iip->ili_logged = 0; /* diff --git a/fs/xfs/xfs_inode_item.h b/fs/xfs/xfs_inode_item.h index 40513077ab3..1ff04cc323a 100644 --- a/fs/xfs/xfs_inode_item.h +++ b/fs/xfs/xfs_inode_item.h @@ -112,6 +112,24 @@ typedef struct xfs_inode_log_format_64 { #define XFS_ILI_IOLOCKED_ANY (XFS_ILI_IOLOCKED_EXCL | XFS_ILI_IOLOCKED_SHARED) +#define XFS_ILOG_FBROOT(w) xfs_ilog_fbroot(w) +static inline int xfs_ilog_fbroot(int w) +{ + return (w == XFS_DATA_FORK ? XFS_ILOG_DBROOT : XFS_ILOG_ABROOT); +} + +#define XFS_ILOG_FEXT(w) xfs_ilog_fext(w) +static inline int xfs_ilog_fext(int w) +{ + return (w == XFS_DATA_FORK ? XFS_ILOG_DEXT : XFS_ILOG_AEXT); +} + +#define XFS_ILOG_FDATA(w) xfs_ilog_fdata(w) +static inline int xfs_ilog_fdata(int w) +{ + return (w == XFS_DATA_FORK ? XFS_ILOG_DDATA : XFS_ILOG_ADATA); +} + #ifdef __KERNEL__ struct xfs_buf; @@ -148,26 +166,6 @@ typedef struct xfs_inode_log_item { } xfs_inode_log_item_t; -#define XFS_ILOG_FDATA(w) xfs_ilog_fdata(w) -static inline int xfs_ilog_fdata(int w) -{ - return (w == XFS_DATA_FORK ? XFS_ILOG_DDATA : XFS_ILOG_ADATA); -} - -#endif /* __KERNEL__ */ - -#define XFS_ILOG_FBROOT(w) xfs_ilog_fbroot(w) -static inline int xfs_ilog_fbroot(int w) -{ - return (w == XFS_DATA_FORK ? XFS_ILOG_DBROOT : XFS_ILOG_ABROOT); -} - -#define XFS_ILOG_FEXT(w) xfs_ilog_fext(w) -static inline int xfs_ilog_fext(int w) -{ - return (w == XFS_DATA_FORK ? XFS_ILOG_DEXT : XFS_ILOG_AEXT); -} - static inline int xfs_inode_clean(xfs_inode_t *ip) { return (!ip->i_itemp || @@ -175,9 +173,6 @@ static inline int xfs_inode_clean(xfs_inode_t *ip) !ip->i_update_core; } - -#ifdef __KERNEL__ - extern void xfs_inode_item_init(struct xfs_inode *, struct xfs_mount *); extern void xfs_inode_item_destroy(struct xfs_inode *); extern void xfs_iflush_done(struct xfs_buf *, xfs_inode_log_item_t *); diff --git a/fs/xfs/xfs_iomap.c b/fs/xfs/xfs_iomap.c index 67f22b2b44b..911062cf73a 100644 --- a/fs/xfs/xfs_iomap.c +++ b/fs/xfs/xfs_iomap.c @@ -290,7 +290,6 @@ STATIC int xfs_iomap_eof_align_last_fsb( xfs_mount_t *mp, xfs_inode_t *ip, - xfs_fsize_t isize, xfs_extlen_t extsize, xfs_fileoff_t *last_fsb) { @@ -306,14 +305,14 @@ xfs_iomap_eof_align_last_fsb( * stripe width and we are allocating past the allocation eof. */ else if (mp->m_swidth && (mp->m_flags & XFS_MOUNT_SWALLOC) && - (isize >= XFS_FSB_TO_B(mp, mp->m_swidth))) + (ip->i_size >= XFS_FSB_TO_B(mp, mp->m_swidth))) new_last_fsb = roundup_64(*last_fsb, mp->m_swidth); /* * Roundup the allocation request to a stripe unit (m_dalign) boundary * if the file size is >= stripe unit size, and we are allocating past * the allocation eof. */ - else if (mp->m_dalign && (isize >= XFS_FSB_TO_B(mp, mp->m_dalign))) + else if (mp->m_dalign && (ip->i_size >= XFS_FSB_TO_B(mp, mp->m_dalign))) new_last_fsb = roundup_64(*last_fsb, mp->m_dalign); /* @@ -403,7 +402,6 @@ xfs_iomap_write_direct( xfs_filblks_t count_fsb, resaligned; xfs_fsblock_t firstfsb; xfs_extlen_t extsz, temp; - xfs_fsize_t isize; int nimaps; int bmapi_flag; int quota_flag; @@ -426,15 +424,10 @@ xfs_iomap_write_direct( rt = XFS_IS_REALTIME_INODE(ip); extsz = xfs_get_extsz_hint(ip); - isize = ip->i_size; - if (ip->i_new_size > isize) - isize = ip->i_new_size; - offset_fsb = XFS_B_TO_FSBT(mp, offset); last_fsb = XFS_B_TO_FSB(mp, ((xfs_ufsize_t)(offset + count))); - if ((offset + count) > isize) { - error = xfs_iomap_eof_align_last_fsb(mp, ip, isize, extsz, - &last_fsb); + if ((offset + count) > ip->i_size) { + error = xfs_iomap_eof_align_last_fsb(mp, ip, extsz, &last_fsb); if (error) goto error_out; } else { @@ -559,7 +552,6 @@ STATIC int xfs_iomap_eof_want_preallocate( xfs_mount_t *mp, xfs_inode_t *ip, - xfs_fsize_t isize, xfs_off_t offset, size_t count, int ioflag, @@ -573,7 +565,7 @@ xfs_iomap_eof_want_preallocate( int n, error, imaps; *prealloc = 0; - if ((ioflag & BMAPI_SYNC) || (offset + count) <= isize) + if ((ioflag & BMAPI_SYNC) || (offset + count) <= ip->i_size) return 0; /* @@ -617,7 +609,6 @@ xfs_iomap_write_delay( xfs_fileoff_t ioalign; xfs_fsblock_t firstblock; xfs_extlen_t extsz; - xfs_fsize_t isize; int nimaps; xfs_bmbt_irec_t imap[XFS_WRITE_IMAPS]; int prealloc, fsynced = 0; @@ -637,11 +628,7 @@ xfs_iomap_write_delay( offset_fsb = XFS_B_TO_FSBT(mp, offset); retry: - isize = ip->i_size; - if (ip->i_new_size > isize) - isize = ip->i_new_size; - - error = xfs_iomap_eof_want_preallocate(mp, ip, isize, offset, count, + error = xfs_iomap_eof_want_preallocate(mp, ip, offset, count, ioflag, imap, XFS_WRITE_IMAPS, &prealloc); if (error) return error; @@ -655,8 +642,7 @@ retry: } if (prealloc || extsz) { - error = xfs_iomap_eof_align_last_fsb(mp, ip, isize, extsz, - &last_fsb); + error = xfs_iomap_eof_align_last_fsb(mp, ip, extsz, &last_fsb); if (error) return error; } diff --git a/fs/xfs/xfs_itable.c b/fs/xfs/xfs_itable.c index cf6754a3c5b..e19d0a8d561 100644 --- a/fs/xfs/xfs_itable.c +++ b/fs/xfs/xfs_itable.c @@ -69,7 +69,7 @@ xfs_bulkstat_one_iget( } ASSERT(ip != NULL); - ASSERT(ip->i_blkno != (xfs_daddr_t)0); + ASSERT(ip->i_imap.im_blkno != 0); dic = &ip->i_d; @@ -125,13 +125,9 @@ STATIC void xfs_bulkstat_one_dinode( xfs_mount_t *mp, /* mount point for filesystem */ xfs_ino_t ino, /* inode number to get data for */ - xfs_dinode_t *dip, /* dinode inode pointer */ + xfs_dinode_t *dic, /* dinode inode pointer */ xfs_bstat_t *buf) /* return buffer */ { - xfs_dinode_core_t *dic; /* dinode core info pointer */ - - dic = &dip->di_core; - /* * The inode format changed when we moved the link count and * made it 32 bits long. If this is an old format inode, @@ -143,7 +139,7 @@ xfs_bulkstat_one_dinode( * the new format. We don't change the version number so that we * can distinguish this from a real new format inode. */ - if (dic->di_version == XFS_DINODE_VERSION_1) { + if (dic->di_version == 1) { buf->bs_nlink = be16_to_cpu(dic->di_onlink); buf->bs_projid = 0; } else { @@ -162,7 +158,7 @@ xfs_bulkstat_one_dinode( buf->bs_mtime.tv_nsec = be32_to_cpu(dic->di_mtime.t_nsec); buf->bs_ctime.tv_sec = be32_to_cpu(dic->di_ctime.t_sec); buf->bs_ctime.tv_nsec = be32_to_cpu(dic->di_ctime.t_nsec); - buf->bs_xflags = xfs_dic2xflags(dip); + buf->bs_xflags = xfs_dic2xflags(dic); buf->bs_extsize = be32_to_cpu(dic->di_extsize) << mp->m_sb.sb_blocklog; buf->bs_extents = be32_to_cpu(dic->di_nextents); buf->bs_gen = be32_to_cpu(dic->di_gen); @@ -173,7 +169,7 @@ xfs_bulkstat_one_dinode( switch (dic->di_format) { case XFS_DINODE_FMT_DEV: - buf->bs_rdev = be32_to_cpu(dip->di_u.di_dev); + buf->bs_rdev = xfs_dinode_get_rdev(dic); buf->bs_blksize = BLKDEV_IOSIZE; buf->bs_blocks = 0; break; @@ -192,27 +188,34 @@ xfs_bulkstat_one_dinode( } } +/* Return 0 on success or positive error */ STATIC int xfs_bulkstat_one_fmt( void __user *ubuffer, + int ubsize, + int *ubused, const xfs_bstat_t *buffer) { + if (ubsize < sizeof(*buffer)) + return XFS_ERROR(ENOMEM); if (copy_to_user(ubuffer, buffer, sizeof(*buffer))) - return -EFAULT; - return sizeof(*buffer); + return XFS_ERROR(EFAULT); + if (ubused) + *ubused = sizeof(*buffer); + return 0; } /* * Return stat information for one inode. * Return 0 if ok, else errno. */ -int /* error status */ -xfs_bulkstat_one( +int /* error status */ +xfs_bulkstat_one_int( xfs_mount_t *mp, /* mount point for filesystem */ xfs_ino_t ino, /* inode number to get data for */ void __user *buffer, /* buffer to place output in */ int ubsize, /* size of buffer */ - void *private_data, /* my private data */ + bulkstat_one_fmt_pf formatter, /* formatter, copy to user */ xfs_daddr_t bno, /* starting bno of inode cluster */ int *ubused, /* bytes used by me */ void *dibuff, /* on-disk inode buffer */ @@ -221,15 +224,12 @@ xfs_bulkstat_one( xfs_bstat_t *buf; /* return buffer */ int error = 0; /* error value */ xfs_dinode_t *dip; /* dinode inode pointer */ - bulkstat_one_fmt_pf formatter = private_data ? : xfs_bulkstat_one_fmt; dip = (xfs_dinode_t *)dibuff; *stat = BULKSTAT_RV_NOTHING; if (!buffer || xfs_internal_inum(mp, ino)) return XFS_ERROR(EINVAL); - if (ubsize < sizeof(*buf)) - return XFS_ERROR(ENOMEM); buf = kmem_alloc(sizeof(*buf), KM_SLEEP); @@ -244,21 +244,34 @@ xfs_bulkstat_one( xfs_bulkstat_one_dinode(mp, ino, dip, buf); } - error = formatter(buffer, buf); - if (error < 0) { - error = EFAULT; + error = formatter(buffer, ubsize, ubused, buf); + if (error) goto out_free; - } *stat = BULKSTAT_RV_DIDONE; - if (ubused) - *ubused = error; out_free: kmem_free(buf); return error; } +int +xfs_bulkstat_one( + xfs_mount_t *mp, /* mount point for filesystem */ + xfs_ino_t ino, /* inode number to get data for */ + void __user *buffer, /* buffer to place output in */ + int ubsize, /* size of buffer */ + void *private_data, /* my private data */ + xfs_daddr_t bno, /* starting bno of inode cluster */ + int *ubused, /* bytes used by me */ + void *dibuff, /* on-disk inode buffer */ + int *stat) /* BULKSTAT_RV_... */ +{ + return xfs_bulkstat_one_int(mp, ino, buffer, ubsize, + xfs_bulkstat_one_fmt, bno, + ubused, dibuff, stat); +} + /* * Test to see whether we can use the ondisk inode directly, based * on the given bulkstat flags, filling in dipp accordingly. @@ -287,19 +300,19 @@ xfs_bulkstat_use_dinode( * to disk yet. This is a temporary hack that would require a proper * fix in the future. */ - if (be16_to_cpu(dip->di_core.di_magic) != XFS_DINODE_MAGIC || - !XFS_DINODE_GOOD_VERSION(dip->di_core.di_version) || - !dip->di_core.di_mode) + if (be16_to_cpu(dip->di_magic) != XFS_DINODE_MAGIC || + !XFS_DINODE_GOOD_VERSION(dip->di_version) || + !dip->di_mode) return 0; if (flags & BULKSTAT_FG_QUICK) { *dipp = dip; return 1; } /* BULKSTAT_FG_INLINE: if attr fork is local, or not there, use it */ - aformat = dip->di_core.di_aformat; + aformat = dip->di_aformat; if ((XFS_DFORK_Q(dip) == 0) || (aformat == XFS_DINODE_FMT_LOCAL) || - (aformat == XFS_DINODE_FMT_EXTENTS && !dip->di_core.di_anextents)) { + (aformat == XFS_DINODE_FMT_EXTENTS && !dip->di_anextents)) { *dipp = dip; return 1; } @@ -359,7 +372,6 @@ xfs_bulkstat( int ubused; /* bytes used by formatter */ xfs_buf_t *bp; /* ptr to on-disk inode cluster buf */ xfs_dinode_t *dip; /* ptr into bp for specific inode */ - xfs_inode_t *ip; /* ptr to in-core inode struct */ /* * Get the last inode value, see if there's nothing to do. @@ -416,8 +428,7 @@ xfs_bulkstat( /* * Allocate and initialize a btree cursor for ialloc btree. */ - cur = xfs_btree_init_cursor(mp, NULL, agbp, agno, XFS_BTNUM_INO, - (xfs_inode_t *)0, 0); + cur = xfs_inobt_init_cursor(mp, NULL, agbp, agno); irbp = irbuf; irbufend = irbuf + nirbuf; end_of_ag = 0; @@ -472,7 +483,7 @@ xfs_bulkstat( * In any case, increment to the next record. */ if (!error) - error = xfs_inobt_increment(cur, 0, &tmp); + error = xfs_btree_increment(cur, 0, &tmp); } else { /* * Start of ag. Lookup the first inode chunk. @@ -539,7 +550,7 @@ xfs_bulkstat( * Set agino to after this chunk and bump the cursor. */ agino = gino + XFS_INODES_PER_CHUNK; - error = xfs_inobt_increment(cur, 0, &tmp); + error = xfs_btree_increment(cur, 0, &tmp); cond_resched(); } /* @@ -586,6 +597,8 @@ xfs_bulkstat( if (flags & (BULKSTAT_FG_QUICK | BULKSTAT_FG_INLINE)) { + int offset; + ino = XFS_AGINO_TO_INO(mp, agno, agino); bno = XFS_AGB_TO_DADDR(mp, agno, @@ -594,21 +607,15 @@ xfs_bulkstat( /* * Get the inode cluster buffer */ - ASSERT(xfs_inode_zone != NULL); - ip = kmem_zone_zalloc(xfs_inode_zone, - KM_SLEEP); - ip->i_ino = ino; - ip->i_mount = mp; - spin_lock_init(&ip->i_flags_lock); if (bp) xfs_buf_relse(bp); - error = xfs_itobp(mp, NULL, ip, - &dip, &bp, bno, - XFS_IMAP_BULKSTAT, - XFS_BUF_LOCK); + + error = xfs_inotobp(mp, NULL, ino, &dip, + &bp, &offset, + XFS_IGET_BULKSTAT); + if (!error) - clustidx = ip->i_boffset / mp->m_sb.sb_inodesize; - kmem_zone_free(xfs_inode_zone, ip); + clustidx = offset / mp->m_sb.sb_inodesize; if (XFS_TEST_ERROR(error != 0, mp, XFS_ERRTAG_BULKSTAT_READ_CHUNK, XFS_RANDOM_BULKSTAT_READ_CHUNK)) { @@ -842,8 +849,7 @@ xfs_inumbers( agino = 0; continue; } - cur = xfs_btree_init_cursor(mp, NULL, agbp, agno, - XFS_BTNUM_INO, (xfs_inode_t *)0, 0); + cur = xfs_inobt_init_cursor(mp, NULL, agbp, agno); error = xfs_inobt_lookup_ge(cur, agino, 0, 0, &tmp); if (error) { xfs_btree_del_cursor(cur, XFS_BTREE_ERROR); @@ -887,7 +893,7 @@ xfs_inumbers( bufidx = 0; } if (left) { - error = xfs_inobt_increment(cur, 0, &tmp); + error = xfs_btree_increment(cur, 0, &tmp); if (error) { xfs_btree_del_cursor(cur, XFS_BTREE_ERROR); cur = NULL; diff --git a/fs/xfs/xfs_itable.h b/fs/xfs/xfs_itable.h index a1f18fce9b7..1fb04e7deb6 100644 --- a/fs/xfs/xfs_itable.h +++ b/fs/xfs/xfs_itable.h @@ -71,9 +71,23 @@ xfs_bulkstat_single( typedef int (*bulkstat_one_fmt_pf)( /* used size in bytes or negative error */ void __user *ubuffer, /* buffer to write to */ + int ubsize, /* remaining user buffer sz */ + int *ubused, /* bytes used by formatter */ const xfs_bstat_t *buffer); /* buffer to read from */ int +xfs_bulkstat_one_int( + xfs_mount_t *mp, + xfs_ino_t ino, + void __user *buffer, + int ubsize, + bulkstat_one_fmt_pf formatter, + xfs_daddr_t bno, + int *ubused, + void *dibuff, + int *stat); + +int xfs_bulkstat_one( xfs_mount_t *mp, xfs_ino_t ino, diff --git a/fs/xfs/xfs_log.c b/fs/xfs/xfs_log.c index 0b02c644355..f4726f702a9 100644 --- a/fs/xfs/xfs_log.c +++ b/fs/xfs/xfs_log.c @@ -100,12 +100,11 @@ STATIC void xlog_ungrant_log_space(xlog_t *log, /* local ticket functions */ -STATIC xlog_ticket_t *xlog_ticket_get(xlog_t *log, +STATIC xlog_ticket_t *xlog_ticket_alloc(xlog_t *log, int unit_bytes, int count, char clientid, uint flags); -STATIC void xlog_ticket_put(xlog_t *log, xlog_ticket_t *ticket); #if defined(DEBUG) STATIC void xlog_verify_dest_ptr(xlog_t *log, __psint_t ptr); @@ -360,7 +359,7 @@ xfs_log_done(xfs_mount_t *mp, */ xlog_trace_loggrant(log, ticket, "xfs_log_done: (non-permanent)"); xlog_ungrant_log_space(log, ticket); - xlog_ticket_put(log, ticket); + xfs_log_ticket_put(ticket); } else { xlog_trace_loggrant(log, ticket, "xfs_log_done: (permanent)"); xlog_regrant_reserve_log_space(log, ticket); @@ -514,7 +513,7 @@ xfs_log_reserve(xfs_mount_t *mp, retval = xlog_regrant_write_log_space(log, internal_ticket); } else { /* may sleep if need to allocate more tickets */ - internal_ticket = xlog_ticket_get(log, unit_bytes, cnt, + internal_ticket = xlog_ticket_alloc(log, unit_bytes, cnt, client, flags); if (!internal_ticket) return XFS_ERROR(ENOMEM); @@ -563,16 +562,21 @@ xfs_log_mount( } mp->m_log = xlog_alloc_log(mp, log_target, blk_offset, num_bblks); + if (!mp->m_log) { + cmn_err(CE_WARN, "XFS: Log allocation failed: No memory!"); + error = ENOMEM; + goto out; + } /* * Initialize the AIL now we have a log. */ - spin_lock_init(&mp->m_ail_lock); error = xfs_trans_ail_init(mp); if (error) { cmn_err(CE_WARN, "XFS: AIL initialisation failed: error %d", error); goto error; } + mp->m_log->l_ailp = mp->m_ail; /* * skip log recovery on a norecovery mount. pretend it all @@ -601,6 +605,7 @@ xfs_log_mount( return 0; error: xfs_log_unmount_dealloc(mp); +out: return error; } /* xfs_log_mount */ @@ -724,8 +729,8 @@ xfs_log_unmount_write(xfs_mount_t *mp) spin_lock(&log->l_icloglock); iclog = log->l_iclog; atomic_inc(&iclog->ic_refcnt); - spin_unlock(&log->l_icloglock); xlog_state_want_sync(log, iclog); + spin_unlock(&log->l_icloglock); error = xlog_state_release_iclog(log, iclog); spin_lock(&log->l_icloglock); @@ -743,7 +748,7 @@ xfs_log_unmount_write(xfs_mount_t *mp) if (tic) { xlog_trace_loggrant(log, tic, "unmount rec"); xlog_ungrant_log_space(log, tic); - xlog_ticket_put(log, tic); + xfs_log_ticket_put(tic); } } else { /* @@ -762,9 +767,9 @@ xfs_log_unmount_write(xfs_mount_t *mp) spin_lock(&log->l_icloglock); iclog = log->l_iclog; atomic_inc(&iclog->ic_refcnt); - spin_unlock(&log->l_icloglock); xlog_state_want_sync(log, iclog); + spin_unlock(&log->l_icloglock); error = xlog_state_release_iclog(log, iclog); spin_lock(&log->l_icloglock); @@ -900,7 +905,7 @@ xfs_log_move_tail(xfs_mount_t *mp, int xfs_log_need_covered(xfs_mount_t *mp) { - int needed = 0, gen; + int needed = 0; xlog_t *log = mp->m_log; if (!xfs_fs_writable(mp)) @@ -909,7 +914,7 @@ xfs_log_need_covered(xfs_mount_t *mp) spin_lock(&log->l_icloglock); if (((log->l_covered_state == XLOG_STATE_COVER_NEED) || (log->l_covered_state == XLOG_STATE_COVER_NEED2)) - && !xfs_trans_first_ail(mp, &gen) + && !xfs_trans_ail_tail(log->l_ailp) && xlog_iclogs_empty(log)) { if (log->l_covered_state == XLOG_STATE_COVER_NEED) log->l_covered_state = XLOG_STATE_COVER_DONE; @@ -946,7 +951,7 @@ xlog_assign_tail_lsn(xfs_mount_t *mp) xfs_lsn_t tail_lsn; xlog_t *log = mp->m_log; - tail_lsn = xfs_trans_tail_ail(mp); + tail_lsn = xfs_trans_ail_tail(mp->m_ail); spin_lock(&log->l_grant_lock); if (tail_lsn != 0) { log->l_tail_lsn = tail_lsn; @@ -1024,12 +1029,6 @@ xlog_iodone(xfs_buf_t *bp) ASSERT(XFS_BUF_FSPRIVATE2(bp, unsigned long) == (unsigned long) 2); XFS_BUF_SET_FSPRIVATE2(bp, (unsigned long)1); aborted = 0; - - /* - * Some versions of cpp barf on the recursive definition of - * ic_log -> hic_fields.ic_log and expand ic_log twice when - * it is passed through two macros. Workaround broken cpp. - */ l = iclog->ic_log; /* @@ -1217,7 +1216,9 @@ xlog_alloc_log(xfs_mount_t *mp, int i; int iclogsize; - log = (xlog_t *)kmem_zalloc(sizeof(xlog_t), KM_SLEEP); + log = kmem_zalloc(sizeof(xlog_t), KM_MAYFAIL); + if (!log) + return NULL; log->l_mp = mp; log->l_targ = log_target; @@ -1249,6 +1250,8 @@ xlog_alloc_log(xfs_mount_t *mp, xlog_get_iclog_buffer_size(mp, log); bp = xfs_buf_get_empty(log->l_iclog_size, mp->m_logdev_targp); + if (!bp) + goto out_free_log; XFS_BUF_SET_IODONE_FUNC(bp, xlog_iodone); XFS_BUF_SET_BDSTRAT_FUNC(bp, xlog_bdstrat_cb); XFS_BUF_SET_FSPRIVATE2(bp, (unsigned long)1); @@ -1275,20 +1278,24 @@ xlog_alloc_log(xfs_mount_t *mp, iclogsize = log->l_iclog_size; ASSERT(log->l_iclog_size >= 4096); for (i=0; i < log->l_iclog_bufs; i++) { - *iclogp = (xlog_in_core_t *) - kmem_zalloc(sizeof(xlog_in_core_t), KM_SLEEP); + *iclogp = kmem_zalloc(sizeof(xlog_in_core_t), KM_MAYFAIL); + if (!*iclogp) + goto out_free_iclog; + iclog = *iclogp; iclog->ic_prev = prev_iclog; prev_iclog = iclog; bp = xfs_buf_get_noaddr(log->l_iclog_size, mp->m_logdev_targp); + if (!bp) + goto out_free_iclog; if (!XFS_BUF_CPSEMA(bp)) ASSERT(0); XFS_BUF_SET_IODONE_FUNC(bp, xlog_iodone); XFS_BUF_SET_BDSTRAT_FUNC(bp, xlog_bdstrat_cb); XFS_BUF_SET_FSPRIVATE2(bp, (unsigned long)1); iclog->ic_bp = bp; - iclog->hic_data = bp->b_addr; + iclog->ic_data = bp->b_addr; #ifdef DEBUG log->l_iclog_bak[i] = (xfs_caddr_t)&(iclog->ic_header); #endif @@ -1308,7 +1315,7 @@ xlog_alloc_log(xfs_mount_t *mp, atomic_set(&iclog->ic_refcnt, 0); spin_lock_init(&iclog->ic_callback_lock); iclog->ic_callback_tail = &(iclog->ic_callback); - iclog->ic_datap = (char *)iclog->hic_data + log->l_iclog_hsize; + iclog->ic_datap = (char *)iclog->ic_data + log->l_iclog_hsize; ASSERT(XFS_BUF_ISBUSY(iclog->ic_bp)); ASSERT(XFS_BUF_VALUSEMA(iclog->ic_bp) <= 0); @@ -1323,6 +1330,25 @@ xlog_alloc_log(xfs_mount_t *mp, log->l_iclog->ic_prev = prev_iclog; /* re-write 1st prev ptr */ return log; + +out_free_iclog: + for (iclog = log->l_iclog; iclog; iclog = prev_iclog) { + prev_iclog = iclog->ic_next; + if (iclog->ic_bp) { + sv_destroy(&iclog->ic_force_wait); + sv_destroy(&iclog->ic_write_wait); + xfs_buf_free(iclog->ic_bp); + xlog_trace_iclog_dealloc(iclog); + } + kmem_free(iclog); + } + spinlock_destroy(&log->l_icloglock); + spinlock_destroy(&log->l_grant_lock); + xlog_trace_loggrant_dealloc(log); + xfs_buf_free(log->l_xbuf); +out_free_log: + kmem_free(log); + return NULL; } /* xlog_alloc_log */ @@ -1413,7 +1439,7 @@ xlog_grant_push_ail(xfs_mount_t *mp, */ if (threshold_lsn && !XLOG_FORCED_SHUTDOWN(log)) - xfs_trans_push_ail(mp, threshold_lsn); + xfs_trans_ail_push(log->l_ailp, threshold_lsn); } /* xlog_grant_push_ail */ @@ -1958,7 +1984,9 @@ xlog_write(xfs_mount_t * mp, if (iclog->ic_size - log_offset <= sizeof(xlog_op_header_t)) { xlog_state_finish_copy(log, iclog, record_cnt, data_cnt); record_cnt = data_cnt = 0; + spin_lock(&log->l_icloglock); xlog_state_want_sync(log, iclog); + spin_unlock(&log->l_icloglock); if (commit_iclog) { ASSERT(flags & XLOG_COMMIT_TRANS); *commit_iclog = iclog; @@ -3167,7 +3195,7 @@ try_again: STATIC void xlog_state_want_sync(xlog_t *log, xlog_in_core_t *iclog) { - spin_lock(&log->l_icloglock); + ASSERT(spin_is_locked(&log->l_icloglock)); if (iclog->ic_state == XLOG_STATE_ACTIVE) { xlog_state_switch_iclogs(log, iclog, 0); @@ -3175,10 +3203,7 @@ xlog_state_want_sync(xlog_t *log, xlog_in_core_t *iclog) ASSERT(iclog->ic_state & (XLOG_STATE_WANT_SYNC|XLOG_STATE_IOERROR)); } - - spin_unlock(&log->l_icloglock); -} /* xlog_state_want_sync */ - +} /***************************************************************************** @@ -3189,22 +3214,33 @@ xlog_state_want_sync(xlog_t *log, xlog_in_core_t *iclog) */ /* - * Free a used ticket. + * Free a used ticket when it's refcount falls to zero. */ -STATIC void -xlog_ticket_put(xlog_t *log, - xlog_ticket_t *ticket) +void +xfs_log_ticket_put( + xlog_ticket_t *ticket) { - sv_destroy(&ticket->t_wait); - kmem_zone_free(xfs_log_ticket_zone, ticket); -} /* xlog_ticket_put */ + ASSERT(atomic_read(&ticket->t_ref) > 0); + if (atomic_dec_and_test(&ticket->t_ref)) { + sv_destroy(&ticket->t_wait); + kmem_zone_free(xfs_log_ticket_zone, ticket); + } +} +xlog_ticket_t * +xfs_log_ticket_get( + xlog_ticket_t *ticket) +{ + ASSERT(atomic_read(&ticket->t_ref) > 0); + atomic_inc(&ticket->t_ref); + return ticket; +} /* * Allocate and initialise a new log ticket. */ STATIC xlog_ticket_t * -xlog_ticket_get(xlog_t *log, +xlog_ticket_alloc(xlog_t *log, int unit_bytes, int cnt, char client, @@ -3275,6 +3311,7 @@ xlog_ticket_get(xlog_t *log, unit_bytes += 2*BBSIZE; } + atomic_set(&tic->t_ref, 1); tic->t_unit_res = unit_bytes; tic->t_curr_res = unit_bytes; tic->t_cnt = cnt; @@ -3290,7 +3327,7 @@ xlog_ticket_get(xlog_t *log, xlog_tic_reset_res(tic); return tic; -} /* xlog_ticket_get */ +} /****************************************************************************** @@ -3419,7 +3456,7 @@ xlog_verify_iclog(xlog_t *log, ptr = iclog->ic_datap; base_ptr = ptr; ophead = (xlog_op_header_t *)ptr; - xhdr = (xlog_in_core_2_t *)&iclog->ic_header; + xhdr = iclog->ic_data; for (i = 0; i < len; i++) { ophead = (xlog_op_header_t *)ptr; @@ -3525,7 +3562,8 @@ xfs_log_force_umount( if (!log || log->l_flags & XLOG_ACTIVE_RECOVERY) { mp->m_flags |= XFS_MOUNT_FS_SHUTDOWN; - XFS_BUF_DONE(mp->m_sb_bp); + if (mp->m_sb_bp) + XFS_BUF_DONE(mp->m_sb_bp); return 0; } @@ -3546,7 +3584,9 @@ xfs_log_force_umount( spin_lock(&log->l_icloglock); spin_lock(&log->l_grant_lock); mp->m_flags |= XFS_MOUNT_FS_SHUTDOWN; - XFS_BUF_DONE(mp->m_sb_bp); + if (mp->m_sb_bp) + XFS_BUF_DONE(mp->m_sb_bp); + /* * This flag is sort of redundant because of the mount flag, but * it's good to maintain the separation between the log and the rest diff --git a/fs/xfs/xfs_log.h b/fs/xfs/xfs_log.h index d47b91f1082..8a3e84e900a 100644 --- a/fs/xfs/xfs_log.h +++ b/fs/xfs/xfs_log.h @@ -134,6 +134,7 @@ typedef struct xfs_log_callback { #ifdef __KERNEL__ /* Log manager interfaces */ struct xfs_mount; +struct xlog_ticket; xfs_lsn_t xfs_log_done(struct xfs_mount *mp, xfs_log_ticket_t ticket, void **iclog, @@ -177,6 +178,9 @@ int xfs_log_need_covered(struct xfs_mount *mp); void xlog_iodone(struct xfs_buf *); +struct xlog_ticket * xfs_log_ticket_get(struct xlog_ticket *ticket); +void xfs_log_ticket_put(struct xlog_ticket *ticket); + #endif diff --git a/fs/xfs/xfs_log_priv.h b/fs/xfs/xfs_log_priv.h index e7d8f84443f..654167be0ef 100644 --- a/fs/xfs/xfs_log_priv.h +++ b/fs/xfs/xfs_log_priv.h @@ -245,6 +245,7 @@ typedef struct xlog_ticket { struct xlog_ticket *t_next; /* :4|8 */ struct xlog_ticket *t_prev; /* :4|8 */ xlog_tid_t t_tid; /* transaction identifier : 4 */ + atomic_t t_ref; /* ticket reference count : 4 */ int t_curr_res; /* current reservation in bytes : 4 */ int t_unit_res; /* unit reservation in bytes : 4 */ char t_ocnt; /* original count : 1 */ @@ -309,6 +310,16 @@ typedef struct xlog_rec_ext_header { } xlog_rec_ext_header_t; #ifdef __KERNEL__ + +/* + * Quite misnamed, because this union lays out the actual on-disk log buffer. + */ +typedef union xlog_in_core2 { + xlog_rec_header_t hic_header; + xlog_rec_ext_header_t hic_xheader; + char hic_sector[XLOG_HEADER_SIZE]; +} xlog_in_core_2_t; + /* * - A log record header is 512 bytes. There is plenty of room to grow the * xlog_rec_header_t into the reserved space. @@ -338,7 +349,7 @@ typedef struct xlog_rec_ext_header { * We'll put all the read-only and l_icloglock fields in the first cacheline, * and move everything else out to subsequent cachelines. */ -typedef struct xlog_iclog_fields { +typedef struct xlog_in_core { sv_t ic_force_wait; sv_t ic_write_wait; struct xlog_in_core *ic_next; @@ -361,41 +372,11 @@ typedef struct xlog_iclog_fields { /* reference counts need their own cacheline */ atomic_t ic_refcnt ____cacheline_aligned_in_smp; -} xlog_iclog_fields_t; - -typedef union xlog_in_core2 { - xlog_rec_header_t hic_header; - xlog_rec_ext_header_t hic_xheader; - char hic_sector[XLOG_HEADER_SIZE]; -} xlog_in_core_2_t; - -typedef struct xlog_in_core { - xlog_iclog_fields_t hic_fields; - xlog_in_core_2_t *hic_data; + xlog_in_core_2_t *ic_data; +#define ic_header ic_data->hic_header } xlog_in_core_t; /* - * Defines to save our code from this glop. - */ -#define ic_force_wait hic_fields.ic_force_wait -#define ic_write_wait hic_fields.ic_write_wait -#define ic_next hic_fields.ic_next -#define ic_prev hic_fields.ic_prev -#define ic_bp hic_fields.ic_bp -#define ic_log hic_fields.ic_log -#define ic_callback hic_fields.ic_callback -#define ic_callback_lock hic_fields.ic_callback_lock -#define ic_callback_tail hic_fields.ic_callback_tail -#define ic_trace hic_fields.ic_trace -#define ic_size hic_fields.ic_size -#define ic_offset hic_fields.ic_offset -#define ic_refcnt hic_fields.ic_refcnt -#define ic_bwritecnt hic_fields.ic_bwritecnt -#define ic_state hic_fields.ic_state -#define ic_datap hic_fields.ic_datap -#define ic_header hic_data->hic_header - -/* * The reservation head lsn is not made up of a cycle number and block number. * Instead, it uses a cycle number and byte number. Logs don't expect to * overflow 31 bits worth of byte offset, so using a byte number will mean @@ -404,6 +385,7 @@ typedef struct xlog_in_core { typedef struct log { /* The following fields don't need locking */ struct xfs_mount *l_mp; /* mount point */ + struct xfs_ail *l_ailp; /* AIL log is working with */ struct xfs_buf *l_xbuf; /* extra buffer for log * wrapping */ struct xfs_buftarg *l_targ; /* buftarg of log */ diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c index 82d46ce69d5..35cca98bd94 100644 --- a/fs/xfs/xfs_log_recover.c +++ b/fs/xfs/xfs_log_recover.c @@ -36,7 +36,6 @@ #include "xfs_dinode.h" #include "xfs_inode.h" #include "xfs_inode_item.h" -#include "xfs_imap.h" #include "xfs_alloc.h" #include "xfs_ialloc.h" #include "xfs_log_priv.h" @@ -54,10 +53,8 @@ STATIC void xlog_recover_insert_item_backq(xlog_recover_item_t **q, xlog_recover_item_t *item); #if defined(DEBUG) STATIC void xlog_recover_check_summary(xlog_t *); -STATIC void xlog_recover_check_ail(xfs_mount_t *, xfs_log_item_t *, int); #else #define xlog_recover_check_summary(log) -#define xlog_recover_check_ail(mp, lip, gen) #endif @@ -270,21 +267,16 @@ STATIC void xlog_recover_iodone( struct xfs_buf *bp) { - xfs_mount_t *mp; - - ASSERT(XFS_BUF_FSPRIVATE(bp, void *)); - if (XFS_BUF_GETERROR(bp)) { /* * We're not going to bother about retrying * this during recovery. One strike! */ - mp = XFS_BUF_FSPRIVATE(bp, xfs_mount_t *); xfs_ioerror_alert("xlog_recover_iodone", - mp, bp, XFS_BUF_ADDR(bp)); - xfs_force_shutdown(mp, SHUTDOWN_META_IO_ERROR); + bp->b_mount, bp, XFS_BUF_ADDR(bp)); + xfs_force_shutdown(bp->b_mount, SHUTDOWN_META_IO_ERROR); } - XFS_BUF_SET_FSPRIVATE(bp, NULL); + bp->b_mount = NULL; XFS_BUF_CLR_IODONE_FUNC(bp); xfs_biodone(bp); } @@ -1419,7 +1411,13 @@ xlog_recover_add_to_trans( return 0; item = trans->r_itemq; if (item == NULL) { - ASSERT(*(uint *)dp == XFS_TRANS_HEADER_MAGIC); + /* we need to catch log corruptions here */ + if (*(uint *)dp != XFS_TRANS_HEADER_MAGIC) { + xlog_warn("XFS: xlog_recover_add_to_trans: " + "bad header magic number"); + ASSERT(0); + return XFS_ERROR(EIO); + } if (len == sizeof(xfs_trans_header_t)) xlog_recover_add_item(&trans->r_itemq); memcpy(&trans->r_theader, dp, len); /* d, s, l */ @@ -2222,9 +2220,8 @@ xlog_recover_do_buffer_trans( XFS_BUF_STALE(bp); error = xfs_bwrite(mp, bp); } else { - ASSERT(XFS_BUF_FSPRIVATE(bp, void *) == NULL || - XFS_BUF_FSPRIVATE(bp, xfs_mount_t *) == mp); - XFS_BUF_SET_FSPRIVATE(bp, mp); + ASSERT(bp->b_mount == NULL || bp->b_mount == mp); + bp->b_mount = mp; XFS_BUF_SET_IODONE_FUNC(bp, xlog_recover_iodone); xfs_bdwrite(mp, bp); } @@ -2241,7 +2238,6 @@ xlog_recover_do_inode_trans( xfs_inode_log_format_t *in_f; xfs_mount_t *mp; xfs_buf_t *bp; - xfs_imap_t imap; xfs_dinode_t *dip; xfs_ino_t ino; int len; @@ -2269,54 +2265,35 @@ xlog_recover_do_inode_trans( } ino = in_f->ilf_ino; mp = log->l_mp; - if (ITEM_TYPE(item) == XFS_LI_INODE) { - imap.im_blkno = (xfs_daddr_t)in_f->ilf_blkno; - imap.im_len = in_f->ilf_len; - imap.im_boffset = in_f->ilf_boffset; - } else { - /* - * It's an old inode format record. We don't know where - * its cluster is located on disk, and we can't allow - * xfs_imap() to figure it out because the inode btrees - * are not ready to be used. Therefore do not pass the - * XFS_IMAP_LOOKUP flag to xfs_imap(). This will give - * us only the single block in which the inode lives - * rather than its cluster, so we must make sure to - * invalidate the buffer when we write it out below. - */ - imap.im_blkno = 0; - error = xfs_imap(log->l_mp, NULL, ino, &imap, 0); - if (error) - goto error; - } /* * Inode buffers can be freed, look out for it, * and do not replay the inode. */ - if (xlog_check_buffer_cancelled(log, imap.im_blkno, imap.im_len, 0)) { + if (xlog_check_buffer_cancelled(log, in_f->ilf_blkno, + in_f->ilf_len, 0)) { error = 0; goto error; } - bp = xfs_buf_read_flags(mp->m_ddev_targp, imap.im_blkno, imap.im_len, - XFS_BUF_LOCK); + bp = xfs_buf_read_flags(mp->m_ddev_targp, in_f->ilf_blkno, + in_f->ilf_len, XFS_BUF_LOCK); if (XFS_BUF_ISERROR(bp)) { xfs_ioerror_alert("xlog_recover_do..(read#2)", mp, - bp, imap.im_blkno); + bp, in_f->ilf_blkno); error = XFS_BUF_GETERROR(bp); xfs_buf_relse(bp); goto error; } error = 0; ASSERT(in_f->ilf_fields & XFS_ILOG_CORE); - dip = (xfs_dinode_t *)xfs_buf_offset(bp, imap.im_boffset); + dip = (xfs_dinode_t *)xfs_buf_offset(bp, in_f->ilf_boffset); /* * Make sure the place we're flushing out to really looks * like an inode! */ - if (unlikely(be16_to_cpu(dip->di_core.di_magic) != XFS_DINODE_MAGIC)) { + if (unlikely(be16_to_cpu(dip->di_magic) != XFS_DINODE_MAGIC)) { xfs_buf_relse(bp); xfs_fs_cmn_err(CE_ALERT, mp, "xfs_inode_recover: Bad inode magic number, dino ptr = 0x%p, dino bp = 0x%p, ino = %Ld", @@ -2339,12 +2316,12 @@ xlog_recover_do_inode_trans( } /* Skip replay when the on disk inode is newer than the log one */ - if (dicp->di_flushiter < be16_to_cpu(dip->di_core.di_flushiter)) { + if (dicp->di_flushiter < be16_to_cpu(dip->di_flushiter)) { /* * Deal with the wrap case, DI_MAX_FLUSH is less * than smaller numbers */ - if (be16_to_cpu(dip->di_core.di_flushiter) == DI_MAX_FLUSH && + if (be16_to_cpu(dip->di_flushiter) == DI_MAX_FLUSH && dicp->di_flushiter < (DI_MAX_FLUSH >> 1)) { /* do nothing */ } else { @@ -2404,7 +2381,7 @@ xlog_recover_do_inode_trans( error = EFSCORRUPTED; goto error; } - if (unlikely(item->ri_buf[1].i_len > sizeof(xfs_dinode_core_t))) { + if (unlikely(item->ri_buf[1].i_len > sizeof(struct xfs_icdinode))) { XFS_CORRUPTION_ERROR("xlog_recover_do_inode_trans(7)", XFS_ERRLEVEL_LOW, mp, dicp); xfs_buf_relse(bp); @@ -2416,23 +2393,24 @@ xlog_recover_do_inode_trans( } /* The core is in in-core format */ - xfs_dinode_to_disk(&dip->di_core, - (xfs_icdinode_t *)item->ri_buf[1].i_addr); + xfs_dinode_to_disk(dip, (xfs_icdinode_t *)item->ri_buf[1].i_addr); /* the rest is in on-disk format */ - if (item->ri_buf[1].i_len > sizeof(xfs_dinode_core_t)) { - memcpy((xfs_caddr_t) dip + sizeof(xfs_dinode_core_t), - item->ri_buf[1].i_addr + sizeof(xfs_dinode_core_t), - item->ri_buf[1].i_len - sizeof(xfs_dinode_core_t)); + if (item->ri_buf[1].i_len > sizeof(struct xfs_icdinode)) { + memcpy((xfs_caddr_t) dip + sizeof(struct xfs_icdinode), + item->ri_buf[1].i_addr + sizeof(struct xfs_icdinode), + item->ri_buf[1].i_len - sizeof(struct xfs_icdinode)); } fields = in_f->ilf_fields; switch (fields & (XFS_ILOG_DEV | XFS_ILOG_UUID)) { case XFS_ILOG_DEV: - dip->di_u.di_dev = cpu_to_be32(in_f->ilf_u.ilfu_rdev); + xfs_dinode_put_rdev(dip, in_f->ilf_u.ilfu_rdev); break; case XFS_ILOG_UUID: - dip->di_u.di_muuid = in_f->ilf_u.ilfu_uuid; + memcpy(XFS_DFORK_DPTR(dip), + &in_f->ilf_u.ilfu_uuid, + sizeof(uuid_t)); break; } @@ -2448,12 +2426,12 @@ xlog_recover_do_inode_trans( switch (fields & XFS_ILOG_DFORK) { case XFS_ILOG_DDATA: case XFS_ILOG_DEXT: - memcpy(&dip->di_u, src, len); + memcpy(XFS_DFORK_DPTR(dip), src, len); break; case XFS_ILOG_DBROOT: - xfs_bmbt_to_bmdr((xfs_bmbt_block_t *)src, len, - &(dip->di_u.di_bmbt), + xfs_bmbt_to_bmdr(mp, (struct xfs_btree_block *)src, len, + (xfs_bmdr_block_t *)XFS_DFORK_DPTR(dip), XFS_DFORK_DSIZE(dip, mp)); break; @@ -2490,8 +2468,8 @@ xlog_recover_do_inode_trans( case XFS_ILOG_ABROOT: dest = XFS_DFORK_APTR(dip); - xfs_bmbt_to_bmdr((xfs_bmbt_block_t *)src, len, - (xfs_bmdr_block_t*)dest, + xfs_bmbt_to_bmdr(mp, (struct xfs_btree_block *)src, + len, (xfs_bmdr_block_t*)dest, XFS_DFORK_ASIZE(dip, mp)); break; @@ -2506,9 +2484,8 @@ xlog_recover_do_inode_trans( write_inode_buffer: if (ITEM_TYPE(item) == XFS_LI_INODE) { - ASSERT(XFS_BUF_FSPRIVATE(bp, void *) == NULL || - XFS_BUF_FSPRIVATE(bp, xfs_mount_t *) == mp); - XFS_BUF_SET_FSPRIVATE(bp, mp); + ASSERT(bp->b_mount == NULL || bp->b_mount == mp); + bp->b_mount = mp; XFS_BUF_SET_IODONE_FUNC(bp, xlog_recover_iodone); xfs_bdwrite(mp, bp); } else { @@ -2639,9 +2616,8 @@ xlog_recover_do_dquot_trans( memcpy(ddq, recddq, item->ri_buf[1].i_len); ASSERT(dq_f->qlf_size == 2); - ASSERT(XFS_BUF_FSPRIVATE(bp, void *) == NULL || - XFS_BUF_FSPRIVATE(bp, xfs_mount_t *) == mp); - XFS_BUF_SET_FSPRIVATE(bp, mp); + ASSERT(bp->b_mount == NULL || bp->b_mount == mp); + bp->b_mount = mp; XFS_BUF_SET_IODONE_FUNC(bp, xlog_recover_iodone); xfs_bdwrite(mp, bp); @@ -2683,11 +2659,11 @@ xlog_recover_do_efi_trans( efip->efi_next_extent = efi_formatp->efi_nextents; efip->efi_flags |= XFS_EFI_COMMITTED; - spin_lock(&mp->m_ail_lock); + spin_lock(&log->l_ailp->xa_lock); /* - * xfs_trans_update_ail() drops the AIL lock. + * xfs_trans_ail_update() drops the AIL lock. */ - xfs_trans_update_ail(mp, (xfs_log_item_t *)efip, lsn); + xfs_trans_ail_update(log->l_ailp, (xfs_log_item_t *)efip, lsn); return 0; } @@ -2706,12 +2682,12 @@ xlog_recover_do_efd_trans( xlog_recover_item_t *item, int pass) { - xfs_mount_t *mp; xfs_efd_log_format_t *efd_formatp; xfs_efi_log_item_t *efip = NULL; xfs_log_item_t *lip; - int gen; __uint64_t efi_id; + struct xfs_ail_cursor cur; + struct xfs_ail *ailp = log->l_ailp; if (pass == XLOG_RECOVER_PASS1) { return; @@ -2728,25 +2704,26 @@ xlog_recover_do_efd_trans( * Search for the efi with the id in the efd format structure * in the AIL. */ - mp = log->l_mp; - spin_lock(&mp->m_ail_lock); - lip = xfs_trans_first_ail(mp, &gen); + spin_lock(&ailp->xa_lock); + lip = xfs_trans_ail_cursor_first(ailp, &cur, 0); while (lip != NULL) { if (lip->li_type == XFS_LI_EFI) { efip = (xfs_efi_log_item_t *)lip; if (efip->efi_format.efi_id == efi_id) { /* - * xfs_trans_delete_ail() drops the + * xfs_trans_ail_delete() drops the * AIL lock. */ - xfs_trans_delete_ail(mp, lip); + xfs_trans_ail_delete(ailp, lip); xfs_efi_item_free(efip); - return; + spin_lock(&ailp->xa_lock); + break; } } - lip = xfs_trans_next_ail(mp, lip, &gen, NULL); + lip = xfs_trans_ail_cursor_next(ailp, &cur); } - spin_unlock(&mp->m_ail_lock); + xfs_trans_ail_cursor_done(ailp, &cur); + spin_unlock(&ailp->xa_lock); } /* @@ -3030,33 +3007,6 @@ abort_error: } /* - * Verify that once we've encountered something other than an EFI - * in the AIL that there are no more EFIs in the AIL. - */ -#if defined(DEBUG) -STATIC void -xlog_recover_check_ail( - xfs_mount_t *mp, - xfs_log_item_t *lip, - int gen) -{ - int orig_gen = gen; - - do { - ASSERT(lip->li_type != XFS_LI_EFI); - lip = xfs_trans_next_ail(mp, lip, &gen, NULL); - /* - * The check will be bogus if we restart from the - * beginning of the AIL, so ASSERT that we don't. - * We never should since we're holding the AIL lock - * the entire time. - */ - ASSERT(gen == orig_gen); - } while (lip != NULL); -} -#endif /* DEBUG */ - -/* * When this is called, all of the EFIs which did not have * corresponding EFDs should be in the AIL. What we do now * is free the extents associated with each one. @@ -3080,20 +3030,23 @@ xlog_recover_process_efis( { xfs_log_item_t *lip; xfs_efi_log_item_t *efip; - int gen; - xfs_mount_t *mp; int error = 0; + struct xfs_ail_cursor cur; + struct xfs_ail *ailp; - mp = log->l_mp; - spin_lock(&mp->m_ail_lock); - - lip = xfs_trans_first_ail(mp, &gen); + ailp = log->l_ailp; + spin_lock(&ailp->xa_lock); + lip = xfs_trans_ail_cursor_first(ailp, &cur, 0); while (lip != NULL) { /* * We're done when we see something other than an EFI. + * There should be no EFIs left in the AIL now. */ if (lip->li_type != XFS_LI_EFI) { - xlog_recover_check_ail(mp, lip, gen); +#ifdef DEBUG + for (; lip; lip = xfs_trans_ail_cursor_next(ailp, &cur)) + ASSERT(lip->li_type != XFS_LI_EFI); +#endif break; } @@ -3102,18 +3055,20 @@ xlog_recover_process_efis( */ efip = (xfs_efi_log_item_t *)lip; if (efip->efi_flags & XFS_EFI_RECOVERED) { - lip = xfs_trans_next_ail(mp, lip, &gen, NULL); + lip = xfs_trans_ail_cursor_next(ailp, &cur); continue; } - spin_unlock(&mp->m_ail_lock); - error = xlog_recover_process_efi(mp, efip); + spin_unlock(&ailp->xa_lock); + error = xlog_recover_process_efi(log->l_mp, efip); + spin_lock(&ailp->xa_lock); if (error) - return error; - spin_lock(&mp->m_ail_lock); - lip = xfs_trans_next_ail(mp, lip, &gen, NULL); + goto out; + lip = xfs_trans_ail_cursor_next(ailp, &cur); } - spin_unlock(&mp->m_ail_lock); +out: + xfs_trans_ail_cursor_done(ailp, &cur); + spin_unlock(&ailp->xa_lock); return error; } @@ -3134,19 +3089,16 @@ xlog_recover_clear_agi_bucket( int error; tp = xfs_trans_alloc(mp, XFS_TRANS_CLEAR_AGI_BUCKET); - error = xfs_trans_reserve(tp, 0, XFS_CLEAR_AGI_BUCKET_LOG_RES(mp), 0, 0, 0); - if (!error) - error = xfs_trans_read_buf(mp, tp, mp->m_ddev_targp, - XFS_AG_DADDR(mp, agno, XFS_AGI_DADDR(mp)), - XFS_FSS_TO_BB(mp, 1), 0, &agibp); + error = xfs_trans_reserve(tp, 0, XFS_CLEAR_AGI_BUCKET_LOG_RES(mp), + 0, 0, 0); if (error) goto out_abort; - error = EINVAL; - agi = XFS_BUF_TO_AGI(agibp); - if (be32_to_cpu(agi->agi_magicnum) != XFS_AGI_MAGIC) + error = xfs_read_agi(mp, tp, agno, &agibp); + if (error) goto out_abort; + agi = XFS_BUF_TO_AGI(agibp); agi->agi_unlinked[bucket] = cpu_to_be32(NULLAGINO); offset = offsetof(xfs_agi_t, agi_unlinked) + (sizeof(xfs_agino_t) * bucket); @@ -3166,6 +3118,62 @@ out_error: return; } +STATIC xfs_agino_t +xlog_recover_process_one_iunlink( + struct xfs_mount *mp, + xfs_agnumber_t agno, + xfs_agino_t agino, + int bucket) +{ + struct xfs_buf *ibp; + struct xfs_dinode *dip; + struct xfs_inode *ip; + xfs_ino_t ino; + int error; + + ino = XFS_AGINO_TO_INO(mp, agno, agino); + error = xfs_iget(mp, NULL, ino, 0, 0, &ip, 0); + if (error) + goto fail; + + /* + * Get the on disk inode to find the next inode in the bucket. + */ + error = xfs_itobp(mp, NULL, ip, &dip, &ibp, XFS_BUF_LOCK); + if (error) + goto fail_iput; + + ASSERT(ip->i_d.di_nlink == 0); + ASSERT(ip->i_d.di_mode != 0); + + /* setup for the next pass */ + agino = be32_to_cpu(dip->di_next_unlinked); + xfs_buf_relse(ibp); + + /* + * Prevent any DMAPI event from being sent when the reference on + * the inode is dropped. + */ + ip->i_d.di_dmevmask = 0; + + IRELE(ip); + return agino; + + fail_iput: + IRELE(ip); + fail: + /* + * We can't read in the inode this bucket points to, or this inode + * is messed up. Just ditch this bucket of inodes. We will lose + * some inodes and space, but at least we won't hang. + * + * Call xlog_recover_clear_agi_bucket() to perform a transaction to + * clear the inode pointer in the bucket. + */ + xlog_recover_clear_agi_bucket(mp, agno, bucket); + return NULLAGINO; +} + /* * xlog_iunlink_recover * @@ -3186,11 +3194,7 @@ xlog_recover_process_iunlinks( xfs_agnumber_t agno; xfs_agi_t *agi; xfs_buf_t *agibp; - xfs_buf_t *ibp; - xfs_dinode_t *dip; - xfs_inode_t *ip; xfs_agino_t agino; - xfs_ino_t ino; int bucket; int error; uint mp_dmevmask; @@ -3207,22 +3211,21 @@ xlog_recover_process_iunlinks( /* * Find the agi for this ag. */ - agibp = xfs_buf_read(mp->m_ddev_targp, - XFS_AG_DADDR(mp, agno, XFS_AGI_DADDR(mp)), - XFS_FSS_TO_BB(mp, 1), 0); - if (XFS_BUF_ISERROR(agibp)) { - xfs_ioerror_alert("xlog_recover_process_iunlinks(#1)", - log->l_mp, agibp, - XFS_AG_DADDR(mp, agno, XFS_AGI_DADDR(mp))); + error = xfs_read_agi(mp, NULL, agno, &agibp); + if (error) { + /* + * AGI is b0rked. Don't process it. + * + * We should probably mark the filesystem as corrupt + * after we've recovered all the ag's we can.... + */ + continue; } agi = XFS_BUF_TO_AGI(agibp); - ASSERT(XFS_AGI_MAGIC == be32_to_cpu(agi->agi_magicnum)); for (bucket = 0; bucket < XFS_AGI_UNLINKED_BUCKETS; bucket++) { - agino = be32_to_cpu(agi->agi_unlinked[bucket]); while (agino != NULLAGINO) { - /* * Release the agi buffer so that it can * be acquired in the normal course of the @@ -3230,87 +3233,17 @@ xlog_recover_process_iunlinks( */ xfs_buf_relse(agibp); - ino = XFS_AGINO_TO_INO(mp, agno, agino); - error = xfs_iget(mp, NULL, ino, 0, 0, &ip, 0); - ASSERT(error || (ip != NULL)); - - if (!error) { - /* - * Get the on disk inode to find the - * next inode in the bucket. - */ - error = xfs_itobp(mp, NULL, ip, &dip, - &ibp, 0, 0, - XFS_BUF_LOCK); - ASSERT(error || (dip != NULL)); - } - - if (!error) { - ASSERT(ip->i_d.di_nlink == 0); - - /* setup for the next pass */ - agino = be32_to_cpu( - dip->di_next_unlinked); - xfs_buf_relse(ibp); - /* - * Prevent any DMAPI event from - * being sent when the - * reference on the inode is - * dropped. - */ - ip->i_d.di_dmevmask = 0; - - /* - * If this is a new inode, handle - * it specially. Otherwise, - * just drop our reference to the - * inode. If there are no - * other references, this will - * send the inode to - * xfs_inactive() which will - * truncate the file and free - * the inode. - */ - if (ip->i_d.di_mode == 0) - xfs_iput_new(ip, 0); - else - IRELE(ip); - } else { - /* - * We can't read in the inode - * this bucket points to, or - * this inode is messed up. Just - * ditch this bucket of inodes. We - * will lose some inodes and space, - * but at least we won't hang. Call - * xlog_recover_clear_agi_bucket() - * to perform a transaction to clear - * the inode pointer in the bucket. - */ - xlog_recover_clear_agi_bucket(mp, agno, - bucket); - - agino = NULLAGINO; - } + agino = xlog_recover_process_one_iunlink(mp, + agno, agino, bucket); /* * Reacquire the agibuffer and continue around - * the loop. + * the loop. This should never fail as we know + * the buffer was good earlier on. */ - agibp = xfs_buf_read(mp->m_ddev_targp, - XFS_AG_DADDR(mp, agno, - XFS_AGI_DADDR(mp)), - XFS_FSS_TO_BB(mp, 1), 0); - if (XFS_BUF_ISERROR(agibp)) { - xfs_ioerror_alert( - "xlog_recover_process_iunlinks(#2)", - log->l_mp, agibp, - XFS_AG_DADDR(mp, agno, - XFS_AGI_DADDR(mp))); - } + error = xfs_read_agi(mp, NULL, agno, &agibp); + ASSERT(error == 0); agi = XFS_BUF_TO_AGI(agibp); - ASSERT(XFS_AGI_MAGIC == be32_to_cpu( - agi->agi_magicnum)); } } @@ -3361,7 +3294,6 @@ xlog_pack_data( int size = iclog->ic_offset + roundoff; __be32 cycle_lsn; xfs_caddr_t dp; - xlog_in_core_2_t *xhdr; xlog_pack_data_checksum(log, iclog, size); @@ -3376,7 +3308,8 @@ xlog_pack_data( } if (xfs_sb_version_haslogv2(&log->l_mp->m_sb)) { - xhdr = (xlog_in_core_2_t *)&iclog->ic_header; + xlog_in_core_2_t *xhdr = iclog->ic_data; + for ( ; i < BTOBB(size); i++) { j = i / (XLOG_HEADER_CYCLE_SIZE / BBSIZE); k = i % (XLOG_HEADER_CYCLE_SIZE / BBSIZE); @@ -3434,7 +3367,6 @@ xlog_unpack_data( xlog_t *log) { int i, j, k; - xlog_in_core_2_t *xhdr; for (i = 0; i < BTOBB(be32_to_cpu(rhead->h_len)) && i < (XLOG_HEADER_CYCLE_SIZE / BBSIZE); i++) { @@ -3443,7 +3375,7 @@ xlog_unpack_data( } if (xfs_sb_version_haslogv2(&log->l_mp->m_sb)) { - xhdr = (xlog_in_core_2_t *)rhead; + xlog_in_core_2_t *xhdr = (xlog_in_core_2_t *)rhead; for ( ; i < BTOBB(be32_to_cpu(rhead->h_len)); i++) { j = i / (XLOG_HEADER_CYCLE_SIZE / BBSIZE); k = i % (XLOG_HEADER_CYCLE_SIZE / BBSIZE); @@ -3997,11 +3929,8 @@ xlog_recover_check_summary( { xfs_mount_t *mp; xfs_agf_t *agfp; - xfs_agi_t *agip; xfs_buf_t *agfbp; xfs_buf_t *agibp; - xfs_daddr_t agfdaddr; - xfs_daddr_t agidaddr; xfs_buf_t *sbbp; #ifdef XFS_LOUD_RECOVERY xfs_sb_t *sbp; @@ -4010,6 +3939,7 @@ xlog_recover_check_summary( __uint64_t freeblks; __uint64_t itotal; __uint64_t ifree; + int error; mp = log->l_mp; @@ -4017,37 +3947,27 @@ xlog_recover_check_summary( itotal = 0LL; ifree = 0LL; for (agno = 0; agno < mp->m_sb.sb_agcount; agno++) { - agfdaddr = XFS_AG_DADDR(mp, agno, XFS_AGF_DADDR(mp)); - agfbp = xfs_buf_read(mp->m_ddev_targp, agfdaddr, - XFS_FSS_TO_BB(mp, 1), 0); - if (XFS_BUF_ISERROR(agfbp)) { - xfs_ioerror_alert("xlog_recover_check_summary(agf)", - mp, agfbp, agfdaddr); - } - agfp = XFS_BUF_TO_AGF(agfbp); - ASSERT(XFS_AGF_MAGIC == be32_to_cpu(agfp->agf_magicnum)); - ASSERT(XFS_AGF_GOOD_VERSION(be32_to_cpu(agfp->agf_versionnum))); - ASSERT(be32_to_cpu(agfp->agf_seqno) == agno); - - freeblks += be32_to_cpu(agfp->agf_freeblks) + - be32_to_cpu(agfp->agf_flcount); - xfs_buf_relse(agfbp); - - agidaddr = XFS_AG_DADDR(mp, agno, XFS_AGI_DADDR(mp)); - agibp = xfs_buf_read(mp->m_ddev_targp, agidaddr, - XFS_FSS_TO_BB(mp, 1), 0); - if (XFS_BUF_ISERROR(agibp)) { - xfs_ioerror_alert("xlog_recover_check_summary(agi)", - mp, agibp, agidaddr); + error = xfs_read_agf(mp, NULL, agno, 0, &agfbp); + if (error) { + xfs_fs_cmn_err(CE_ALERT, mp, + "xlog_recover_check_summary(agf)" + "agf read failed agno %d error %d", + agno, error); + } else { + agfp = XFS_BUF_TO_AGF(agfbp); + freeblks += be32_to_cpu(agfp->agf_freeblks) + + be32_to_cpu(agfp->agf_flcount); + xfs_buf_relse(agfbp); } - agip = XFS_BUF_TO_AGI(agibp); - ASSERT(XFS_AGI_MAGIC == be32_to_cpu(agip->agi_magicnum)); - ASSERT(XFS_AGI_GOOD_VERSION(be32_to_cpu(agip->agi_versionnum))); - ASSERT(be32_to_cpu(agip->agi_seqno) == agno); - itotal += be32_to_cpu(agip->agi_count); - ifree += be32_to_cpu(agip->agi_freecount); - xfs_buf_relse(agibp); + error = xfs_read_agi(mp, NULL, agno, &agibp); + if (!error) { + struct xfs_agi *agi = XFS_BUF_TO_AGI(agibp); + + itotal += be32_to_cpu(agi->agi_count); + ifree += be32_to_cpu(agi->agi_freecount); + xfs_buf_relse(agibp); + } } sbbp = xfs_getsb(mp, 0); diff --git a/fs/xfs/xfs_mount.c b/fs/xfs/xfs_mount.c index a4503f5e949..3c97c6463a4 100644 --- a/fs/xfs/xfs_mount.c +++ b/fs/xfs/xfs_mount.c @@ -567,8 +567,6 @@ xfs_readsb(xfs_mount_t *mp, int flags) STATIC void xfs_mount_common(xfs_mount_t *mp, xfs_sb_t *sbp) { - int i; - mp->m_agfrotor = mp->m_agirotor = 0; spin_lock_init(&mp->m_agirotor_lock); mp->m_maxagi = mp->m_sb.sb_agcount; @@ -577,12 +575,10 @@ xfs_mount_common(xfs_mount_t *mp, xfs_sb_t *sbp) mp->m_sectbb_log = sbp->sb_sectlog - BBSHIFT; mp->m_agno_log = xfs_highbit32(sbp->sb_agcount - 1) + 1; mp->m_agino_log = sbp->sb_inopblog + sbp->sb_agblklog; - mp->m_litino = sbp->sb_inodesize - - ((uint)sizeof(xfs_dinode_core_t) + (uint)sizeof(xfs_agino_t)); + mp->m_litino = sbp->sb_inodesize - sizeof(struct xfs_dinode); mp->m_blockmask = sbp->sb_blocksize - 1; mp->m_blockwsize = sbp->sb_blocksize >> XFS_WORDLOG; mp->m_blockwmask = mp->m_blockwsize - 1; - INIT_LIST_HEAD(&mp->m_del_inodes); /* * Setup for attributes, in case they get created. @@ -605,24 +601,20 @@ xfs_mount_common(xfs_mount_t *mp, xfs_sb_t *sbp) } ASSERT(mp->m_attroffset < XFS_LITINO(mp)); - for (i = 0; i < 2; i++) { - mp->m_alloc_mxr[i] = XFS_BTREE_BLOCK_MAXRECS(sbp->sb_blocksize, - xfs_alloc, i == 0); - mp->m_alloc_mnr[i] = XFS_BTREE_BLOCK_MINRECS(sbp->sb_blocksize, - xfs_alloc, i == 0); - } - for (i = 0; i < 2; i++) { - mp->m_bmap_dmxr[i] = XFS_BTREE_BLOCK_MAXRECS(sbp->sb_blocksize, - xfs_bmbt, i == 0); - mp->m_bmap_dmnr[i] = XFS_BTREE_BLOCK_MINRECS(sbp->sb_blocksize, - xfs_bmbt, i == 0); - } - for (i = 0; i < 2; i++) { - mp->m_inobt_mxr[i] = XFS_BTREE_BLOCK_MAXRECS(sbp->sb_blocksize, - xfs_inobt, i == 0); - mp->m_inobt_mnr[i] = XFS_BTREE_BLOCK_MINRECS(sbp->sb_blocksize, - xfs_inobt, i == 0); - } + mp->m_alloc_mxr[0] = xfs_allocbt_maxrecs(mp, sbp->sb_blocksize, 1); + mp->m_alloc_mxr[1] = xfs_allocbt_maxrecs(mp, sbp->sb_blocksize, 0); + mp->m_alloc_mnr[0] = mp->m_alloc_mxr[0] / 2; + mp->m_alloc_mnr[1] = mp->m_alloc_mxr[1] / 2; + + mp->m_inobt_mxr[0] = xfs_inobt_maxrecs(mp, sbp->sb_blocksize, 1); + mp->m_inobt_mxr[1] = xfs_inobt_maxrecs(mp, sbp->sb_blocksize, 0); + mp->m_inobt_mnr[0] = mp->m_inobt_mxr[0] / 2; + mp->m_inobt_mnr[1] = mp->m_inobt_mxr[1] / 2; + + mp->m_bmap_dmxr[0] = xfs_bmbt_maxrecs(mp, sbp->sb_blocksize, 1); + mp->m_bmap_dmxr[1] = xfs_bmbt_maxrecs(mp, sbp->sb_blocksize, 0); + mp->m_bmap_dmnr[0] = mp->m_bmap_dmxr[0] / 2; + mp->m_bmap_dmnr[1] = mp->m_bmap_dmxr[1] / 2; mp->m_bsize = XFS_FSB_TO_BB(mp, 1); mp->m_ialloc_inos = (int)MAX((__uint16_t)XFS_INODES_PER_CHUNK, @@ -1228,6 +1220,16 @@ xfs_unmountfs( __uint64_t resblks; int error; + /* + * Release dquot that rootinode, rbmino and rsumino might be holding, + * and release the quota inodes. + */ + XFS_QM_UNMOUNT(mp); + + if (mp->m_rbmip) + IRELE(mp->m_rbmip); + if (mp->m_rsumip) + IRELE(mp->m_rsumip); IRELE(mp->m_rootip); /* @@ -1241,10 +1243,13 @@ xfs_unmountfs( * need to force the log first. */ xfs_log_force(mp, (xfs_lsn_t)0, XFS_LOG_FORCE | XFS_LOG_SYNC); - xfs_iflush_all(mp); + xfs_reclaim_inodes(mp, 0, XFS_IFLUSH_ASYNC); XFS_QM_DQPURGEALL(mp, XFS_QMOPT_QUOTALL | XFS_QMOPT_UMOUNTING); + if (mp->m_quotainfo) + XFS_QM_DONE(mp); + /* * Flush out the log synchronously so that we know for sure * that nothing is pinned. This is important because bflush() @@ -1285,11 +1290,6 @@ xfs_unmountfs( xfs_unmountfs_wait(mp); /* wait for async bufs */ xfs_log_unmount(mp); /* Done! No more fs ops. */ - /* - * All inodes from this mount point should be freed. - */ - ASSERT(mp->m_inodes == NULL); - if ((mp->m_flags & XFS_MOUNT_NOUUID) == 0) uuid_table_remove(&mp->m_sb.sb_uuid); @@ -1297,8 +1297,6 @@ xfs_unmountfs( xfs_errortag_clearall(mp, 0); #endif xfs_free_perag(mp); - if (mp->m_quotainfo) - XFS_QM_DONE(mp); } STATIC void @@ -1364,24 +1362,6 @@ xfs_log_sbcount( return error; } -STATIC void -xfs_mark_shared_ro( - xfs_mount_t *mp, - xfs_buf_t *bp) -{ - xfs_dsb_t *sb = XFS_BUF_TO_SBP(bp); - __uint16_t version; - - if (!(sb->sb_flags & XFS_SBF_READONLY)) - sb->sb_flags |= XFS_SBF_READONLY; - - version = be16_to_cpu(sb->sb_versionnum); - if ((version & XFS_SB_VERSION_NUMBITS) != XFS_SB_VERSION_4 || - !(version & XFS_SB_VERSION_SHAREDBIT)) - version |= XFS_SB_VERSION_SHAREDBIT; - sb->sb_versionnum = cpu_to_be16(version); -} - int xfs_unmountfs_writesb(xfs_mount_t *mp) { @@ -1397,12 +1377,6 @@ xfs_unmountfs_writesb(xfs_mount_t *mp) sbp = xfs_getsb(mp, 0); - /* - * mark shared-readonly if desired - */ - if (mp->m_mk_sharedro) - xfs_mark_shared_ro(mp, sbp); - XFS_BUF_UNDONE(sbp); XFS_BUF_UNREAD(sbp); XFS_BUF_UNDELAYWRITE(sbp); @@ -1414,8 +1388,6 @@ xfs_unmountfs_writesb(xfs_mount_t *mp) if (error) xfs_ioerror_alert("xfs_unmountfs_writesb", mp, sbp, XFS_BUF_ADDR(sbp)); - if (error && mp->m_mk_sharedro) - xfs_fs_cmn_err(CE_ALERT, mp, "Superblock write error detected while unmounting. Filesystem may not be marked shared readonly"); xfs_buf_relse(sbp); } return error; diff --git a/fs/xfs/xfs_mount.h b/fs/xfs/xfs_mount.h index f3c1024b124..c1e02846732 100644 --- a/fs/xfs/xfs_mount.h +++ b/fs/xfs/xfs_mount.h @@ -18,7 +18,6 @@ #ifndef __XFS_MOUNT_H__ #define __XFS_MOUNT_H__ - typedef struct xfs_trans_reservations { uint tr_write; /* extent alloc trans */ uint tr_itruncate; /* truncate trans */ @@ -44,14 +43,16 @@ typedef struct xfs_trans_reservations { } xfs_trans_reservations_t; #ifndef __KERNEL__ -/* - * Moved here from xfs_ag.h to avoid reordering header files - */ + #define XFS_DADDR_TO_AGNO(mp,d) \ ((xfs_agnumber_t)(XFS_BB_TO_FSBT(mp, d) / (mp)->m_sb.sb_agblocks)) #define XFS_DADDR_TO_AGBNO(mp,d) \ ((xfs_agblock_t)(XFS_BB_TO_FSBT(mp, d) % (mp)->m_sb.sb_agblocks)) -#else + +#else /* __KERNEL__ */ + +#include "xfs_sync.h" + struct cred; struct log; struct xfs_mount_args; @@ -62,6 +63,7 @@ struct xfs_extdelta; struct xfs_swapext; struct xfs_mru_cache; struct xfs_nameops; +struct xfs_ail; /* * Prototypes and functions for the Data Migration subsystem. @@ -115,7 +117,7 @@ struct xfs_quotainfo; typedef int (*xfs_qminit_t)(struct xfs_mount *, uint *, uint *); typedef int (*xfs_qmmount_t)(struct xfs_mount *, uint, uint); -typedef int (*xfs_qmunmount_t)(struct xfs_mount *); +typedef void (*xfs_qmunmount_t)(struct xfs_mount *); typedef void (*xfs_qmdone_t)(struct xfs_mount *); typedef void (*xfs_dqrele_t)(struct xfs_dquot *); typedef int (*xfs_dqattach_t)(struct xfs_inode *, uint); @@ -132,7 +134,7 @@ typedef struct xfs_dquot * (*xfs_dqvopchown_t)( struct xfs_dquot **, struct xfs_dquot *); typedef int (*xfs_dqvopchownresv_t)(struct xfs_trans *, struct xfs_inode *, struct xfs_dquot *, struct xfs_dquot *, uint); -typedef void (*xfs_dqstatvfs_t)(struct xfs_inode *, bhv_statvfs_t *); +typedef void (*xfs_dqstatvfs_t)(struct xfs_inode *, struct kstatfs *); typedef int (*xfs_dqsync_t)(struct xfs_mount *, int flags); typedef int (*xfs_quotactl_t)(struct xfs_mount *, int, int, xfs_caddr_t); @@ -223,18 +225,10 @@ extern void xfs_icsb_sync_counters_locked(struct xfs_mount *, int); #define xfs_icsb_sync_counters_locked(mp, flags) do { } while (0) #endif -typedef struct xfs_ail { - struct list_head xa_ail; - uint xa_gen; - struct task_struct *xa_task; - xfs_lsn_t xa_target; -} xfs_ail_t; - typedef struct xfs_mount { struct super_block *m_super; xfs_tid_t m_tid; /* next unused tid for fs */ - spinlock_t m_ail_lock; /* fs AIL mutex */ - xfs_ail_t m_ail; /* fs active log item list */ + struct xfs_ail *m_ail; /* fs active log item list */ xfs_sb_t m_sb; /* copy of fs superblock */ spinlock_t m_sb_lock; /* sb counter lock */ struct xfs_buf *m_sb_bp; /* buffer for superblock */ @@ -247,10 +241,6 @@ typedef struct xfs_mount { xfs_agnumber_t m_agirotor; /* last ag dir inode alloced */ spinlock_t m_agirotor_lock;/* .. and lock protecting it */ xfs_agnumber_t m_maxagi; /* highest inode alloc group */ - struct xfs_inode *m_inodes; /* active inode list */ - struct list_head m_del_inodes; /* inodes to reclaim */ - mutex_t m_ilock; /* inode list mutex */ - uint m_ireclaims; /* count of calls to reclaim*/ uint m_readio_log; /* min read size log bytes */ uint m_readio_blocks; /* min read size blocks */ uint m_writeio_log; /* min write size log bytes */ @@ -267,7 +257,6 @@ typedef struct xfs_mount { xfs_buftarg_t *m_ddev_targp; /* saves taking the address */ xfs_buftarg_t *m_logdev_targp;/* ptr to log device */ xfs_buftarg_t *m_rtdev_targp; /* ptr to rt device */ - __uint8_t m_dircook_elog; /* log d-cookie entry bits */ __uint8_t m_blkbit_log; /* blocklog + NBBY */ __uint8_t m_blkbb_log; /* blocklog - BBSHIFT */ __uint8_t m_agno_log; /* log #ag's */ @@ -276,12 +265,12 @@ typedef struct xfs_mount { uint m_blockmask; /* sb_blocksize-1 */ uint m_blockwsize; /* sb_blocksize in words */ uint m_blockwmask; /* blockwsize-1 */ - uint m_alloc_mxr[2]; /* XFS_ALLOC_BLOCK_MAXRECS */ - uint m_alloc_mnr[2]; /* XFS_ALLOC_BLOCK_MINRECS */ - uint m_bmap_dmxr[2]; /* XFS_BMAP_BLOCK_DMAXRECS */ - uint m_bmap_dmnr[2]; /* XFS_BMAP_BLOCK_DMINRECS */ - uint m_inobt_mxr[2]; /* XFS_INOBT_BLOCK_MAXRECS */ - uint m_inobt_mnr[2]; /* XFS_INOBT_BLOCK_MINRECS */ + uint m_alloc_mxr[2]; /* max alloc btree records */ + uint m_alloc_mnr[2]; /* min alloc btree records */ + uint m_bmap_dmxr[2]; /* max bmap btree records */ + uint m_bmap_dmnr[2]; /* min bmap btree records */ + uint m_inobt_mxr[2]; /* max inobt btree records */ + uint m_inobt_mnr[2]; /* min inobt btree records */ uint m_ag_maxlevels; /* XFS_AG_MAXLEVELS */ uint m_bm_maxlevels[2]; /* XFS_BM_MAXLEVELS */ uint m_in_maxlevels; /* XFS_IN_MAXLEVELS */ @@ -312,9 +301,6 @@ typedef struct xfs_mount { int m_sinoalign; /* stripe unit inode alignment */ int m_attr_magicpct;/* 37% of the blocksize */ int m_dir_magicpct; /* 37% of the dir blocksize */ - __uint8_t m_mk_sharedro; /* mark shared ro on unmount */ - __uint8_t m_inode_quiesce;/* call quiesce on new inodes. - field governed by m_ilock */ __uint8_t m_sectbb_log; /* sectlog - BBSHIFT */ const struct xfs_nameops *m_dirnameops; /* vector of dir name ops */ int m_dirblksize; /* directory block sz--bytes */ @@ -362,7 +348,6 @@ typedef struct xfs_mount { #define XFS_MOUNT_ATTR2 (1ULL << 8) /* allow use of attr2 format */ #define XFS_MOUNT_GRPID (1ULL << 9) /* group-ID assigned from directory */ #define XFS_MOUNT_NORECOVERY (1ULL << 10) /* no recovery - dirty fs */ -#define XFS_MOUNT_SHARED (1ULL << 11) /* shared mount */ #define XFS_MOUNT_DFLT_IOSIZE (1ULL << 12) /* set default i/o size */ #define XFS_MOUNT_OSYNCISOSYNC (1ULL << 13) /* o_sync is REALLY o_sync */ /* osyncisdsync is now default*/ @@ -439,6 +424,16 @@ void xfs_do_force_shutdown(struct xfs_mount *mp, int flags, char *fname, #define xfs_force_shutdown(m,f) \ xfs_do_force_shutdown(m, f, __FILE__, __LINE__) +#define SHUTDOWN_META_IO_ERROR 0x0001 /* write attempt to metadata failed */ +#define SHUTDOWN_LOG_IO_ERROR 0x0002 /* write attempt to the log failed */ +#define SHUTDOWN_FORCE_UMOUNT 0x0004 /* shutdown from a forced unmount */ +#define SHUTDOWN_CORRUPT_INCORE 0x0008 /* corrupt in-memory data structures */ +#define SHUTDOWN_REMOTE_REQ 0x0010 /* shutdown came from remote cell */ +#define SHUTDOWN_DEVICE_REQ 0x0020 /* failed all paths to the device */ + +#define xfs_test_for_freeze(mp) ((mp)->m_super->s_frozen) +#define xfs_wait_for_freeze(mp,l) vfs_check_frozen((mp)->m_super, (l)) + /* * Flags for xfs_mountfs */ @@ -508,14 +503,12 @@ typedef struct xfs_mod_sb { #define XFS_MOUNT_ILOCK(mp) mutex_lock(&((mp)->m_ilock)) #define XFS_MOUNT_IUNLOCK(mp) mutex_unlock(&((mp)->m_ilock)) -extern void xfs_mod_sb(xfs_trans_t *, __int64_t); extern int xfs_log_sbcount(xfs_mount_t *, uint); extern int xfs_mountfs(xfs_mount_t *mp); extern void xfs_mountfs_check_barriers(xfs_mount_t *mp); extern void xfs_unmountfs(xfs_mount_t *); extern int xfs_unmountfs_writesb(xfs_mount_t *); -extern int xfs_unmount_flush(xfs_mount_t *, int); extern int xfs_mod_incore_sb(xfs_mount_t *, xfs_sb_field_t, int64_t, int); extern int xfs_mod_incore_sb_unlocked(xfs_mount_t *, xfs_sb_field_t, int64_t, int); @@ -525,20 +518,20 @@ extern struct xfs_buf *xfs_getsb(xfs_mount_t *, int); extern int xfs_readsb(xfs_mount_t *, int); extern void xfs_freesb(xfs_mount_t *); extern int xfs_fs_writable(xfs_mount_t *); -extern int xfs_syncsub(xfs_mount_t *, int, int *); -extern int xfs_sync_inodes(xfs_mount_t *, int, int *); -extern xfs_agnumber_t xfs_initialize_perag(xfs_mount_t *, xfs_agnumber_t); -extern void xfs_sb_from_disk(struct xfs_sb *, struct xfs_dsb *); -extern void xfs_sb_to_disk(struct xfs_dsb *, struct xfs_sb *, __int64_t); extern int xfs_sb_validate_fsb_count(struct xfs_sb *, __uint64_t); -extern int xfs_dmops_get(struct xfs_mount *, struct xfs_mount_args *); +extern int xfs_dmops_get(struct xfs_mount *); extern void xfs_dmops_put(struct xfs_mount *); -extern int xfs_qmops_get(struct xfs_mount *, struct xfs_mount_args *); +extern int xfs_qmops_get(struct xfs_mount *); extern void xfs_qmops_put(struct xfs_mount *); extern struct xfs_dmops xfs_dmcore_xfs; #endif /* __KERNEL__ */ +extern void xfs_mod_sb(struct xfs_trans *, __int64_t); +extern xfs_agnumber_t xfs_initialize_perag(struct xfs_mount *, xfs_agnumber_t); +extern void xfs_sb_from_disk(struct xfs_sb *, struct xfs_dsb *); +extern void xfs_sb_to_disk(struct xfs_dsb *, struct xfs_sb *, __int64_t); + #endif /* __XFS_MOUNT_H__ */ diff --git a/fs/xfs/xfs_qmops.c b/fs/xfs/xfs_qmops.c index a294e58db8d..27f80581520 100644 --- a/fs/xfs/xfs_qmops.c +++ b/fs/xfs/xfs_qmops.c @@ -28,7 +28,6 @@ #include "xfs_mount.h" #include "xfs_quota.h" #include "xfs_error.h" -#include "xfs_clnt.h" STATIC struct xfs_dquot * @@ -131,9 +130,9 @@ static struct xfs_qmops xfs_qmcore_stub = { }; int -xfs_qmops_get(struct xfs_mount *mp, struct xfs_mount_args *args) +xfs_qmops_get(struct xfs_mount *mp) { - if (args->flags & (XFSMNT_UQUOTA | XFSMNT_PQUOTA | XFSMNT_GQUOTA)) { + if (XFS_IS_QUOTA_RUNNING(mp)) { #ifdef CONFIG_XFS_QUOTA mp->m_qm_ops = &xfs_qmcore_xfs; #else diff --git a/fs/xfs/xfs_quota.h b/fs/xfs/xfs_quota.h index 12c4ec775af..48965ecaa15 100644 --- a/fs/xfs/xfs_quota.h +++ b/fs/xfs/xfs_quota.h @@ -84,11 +84,9 @@ typedef struct xfs_dqblk { #define XFS_DQ_USER 0x0001 /* a user quota */ #define XFS_DQ_PROJ 0x0002 /* project quota */ #define XFS_DQ_GROUP 0x0004 /* a group quota */ -#define XFS_DQ_FLOCKED 0x0008 /* flush lock taken */ -#define XFS_DQ_DIRTY 0x0010 /* dquot is dirty */ -#define XFS_DQ_WANT 0x0020 /* for lookup/reclaim race */ -#define XFS_DQ_INACTIVE 0x0040 /* dq off mplist & hashlist */ -#define XFS_DQ_MARKER 0x0080 /* sentinel */ +#define XFS_DQ_DIRTY 0x0008 /* dquot is dirty */ +#define XFS_DQ_WANT 0x0010 /* for lookup/reclaim race */ +#define XFS_DQ_INACTIVE 0x0020 /* dq off mplist & hashlist */ #define XFS_DQ_ALLTYPES (XFS_DQ_USER|XFS_DQ_PROJ|XFS_DQ_GROUP) diff --git a/fs/xfs/xfs_rename.c b/fs/xfs/xfs_rename.c index d700dacdb10..86471bb40fd 100644 --- a/fs/xfs/xfs_rename.c +++ b/fs/xfs/xfs_rename.c @@ -42,31 +42,6 @@ /* - * Given an array of up to 4 inode pointers, unlock the pointed to inodes. - * If there are fewer than 4 entries in the array, the empty entries will - * be at the end and will have NULL pointers in them. - */ -STATIC void -xfs_rename_unlock4( - xfs_inode_t **i_tab, - uint lock_mode) -{ - int i; - - xfs_iunlock(i_tab[0], lock_mode); - for (i = 1; i < 4; i++) { - if (i_tab[i] == NULL) - break; - - /* - * Watch out for duplicate entries in the table. - */ - if (i_tab[i] != i_tab[i-1]) - xfs_iunlock(i_tab[i], lock_mode); - } -} - -/* * Enter all inodes for a rename transaction into a sorted array. */ STATIC void @@ -205,19 +180,6 @@ xfs_rename( xfs_lock_inodes(inodes, num_inodes, XFS_ILOCK_EXCL); /* - * If we are using project inheritance, we only allow renames - * into our tree when the project IDs are the same; else the - * tree quota mechanism would be circumvented. - */ - if (unlikely((target_dp->i_d.di_flags & XFS_DIFLAG_PROJINHERIT) && - (target_dp->i_d.di_projid != src_ip->i_d.di_projid))) { - error = XFS_ERROR(EXDEV); - xfs_rename_unlock4(inodes, XFS_ILOCK_SHARED); - xfs_trans_cancel(tp, cancel_flags); - goto std_return; - } - - /* * Join all the inodes to the transaction. From this point on, * we can rely on either trans_commit or trans_cancel to unlock * them. Note that we need to add a vnode reference to the @@ -242,6 +204,17 @@ xfs_rename( } /* + * If we are using project inheritance, we only allow renames + * into our tree when the project IDs are the same; else the + * tree quota mechanism would be circumvented. + */ + if (unlikely((target_dp->i_d.di_flags & XFS_DIFLAG_PROJINHERIT) && + (target_dp->i_d.di_projid != src_ip->i_d.di_projid))) { + error = XFS_ERROR(EXDEV); + goto error_return; + } + + /* * Set up the target. */ if (target_ip == NULL) { @@ -367,19 +340,11 @@ xfs_rename( &first_block, &free_list, spaceres); if (error) goto abort_return; - xfs_ichgtime(src_dp, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG); - /* - * Update the generation counts on all the directory inodes - * that we're modifying. - */ - src_dp->i_gen++; + xfs_ichgtime(src_dp, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG); xfs_trans_log_inode(tp, src_dp, XFS_ILOG_CORE); - - if (new_parent) { - target_dp->i_gen++; + if (new_parent) xfs_trans_log_inode(tp, target_dp, XFS_ILOG_CORE); - } /* * If this is a synchronous mount, make sure that the diff --git a/fs/xfs/xfs_rtalloc.c b/fs/xfs/xfs_rtalloc.c index e2f68de1615..edf12c7b834 100644 --- a/fs/xfs/xfs_rtalloc.c +++ b/fs/xfs/xfs_rtalloc.c @@ -85,7 +85,6 @@ xfs_growfs_rt_alloc( { xfs_fileoff_t bno; /* block number in file */ xfs_buf_t *bp; /* temporary buffer for zeroing */ - int cancelflags; /* flags for xfs_trans_cancel */ int committed; /* transaction committed flag */ xfs_daddr_t d; /* disk block address */ int error; /* error return value */ @@ -96,15 +95,16 @@ xfs_growfs_rt_alloc( xfs_bmbt_irec_t map; /* block map output */ int nmap; /* number of block maps */ int resblks; /* space reservation */ - xfs_trans_t *tp; /* transaction pointer */ /* * Allocate space to the file, as necessary. */ while (oblocks < nblocks) { + int cancelflags = 0; + xfs_trans_t *tp; + tp = xfs_trans_alloc(mp, XFS_TRANS_GROWFSRT_ALLOC); resblks = XFS_GROWFSRT_SPACE_RES(mp, nblocks - oblocks); - cancelflags = 0; /* * Reserve space & log for one extent added to the file. */ @@ -171,7 +171,9 @@ xfs_growfs_rt_alloc( mp->m_bsize, 0); if (bp == NULL) { error = XFS_ERROR(EIO); - goto error_cancel; +error_cancel: + xfs_trans_cancel(tp, cancelflags); + goto error; } memset(XFS_BUF_PTR(bp), 0, mp->m_sb.sb_blocksize); xfs_trans_log_buf(tp, bp, 0, mp->m_sb.sb_blocksize - 1); @@ -188,8 +190,6 @@ xfs_growfs_rt_alloc( oblocks = map.br_startoff + map.br_blockcount; } return 0; -error_cancel: - xfs_trans_cancel(tp, cancelflags); error: return error; } @@ -1856,7 +1856,6 @@ xfs_growfs_rt( { xfs_rtblock_t bmbno; /* bitmap block number */ xfs_buf_t *bp; /* temporary buffer */ - int cancelflags; /* flags for xfs_trans_cancel */ int error; /* error return value */ xfs_inode_t *ip; /* bitmap inode, used as lock */ xfs_mount_t *nmp; /* new (fake) mount structure */ @@ -1872,13 +1871,13 @@ xfs_growfs_rt( xfs_extlen_t rsumblocks; /* current number of rt summary blks */ xfs_sb_t *sbp; /* old superblock */ xfs_fsblock_t sumbno; /* summary block number */ - xfs_trans_t *tp; /* transaction pointer */ sbp = &mp->m_sb; - cancelflags = 0; /* * Initial error checking. */ + if (!capable(CAP_SYS_ADMIN)) + return XFS_ERROR(EPERM); if (mp->m_rtdev_targp == NULL || mp->m_rbmip == NULL || (nrblocks = in->newblocks) <= sbp->sb_rblocks || (sbp->sb_rblocks && (in->extsize != sbp->sb_rextsize))) @@ -1942,6 +1941,9 @@ xfs_growfs_rt( ((sbp->sb_rextents & ((1 << mp->m_blkbit_log) - 1)) != 0); bmbno < nrbmblocks; bmbno++) { + xfs_trans_t *tp; + int cancelflags = 0; + *nmp = *mp; nsbp = &nmp->m_sb; /* @@ -1967,16 +1969,15 @@ xfs_growfs_rt( * Start a transaction, get the log reservation. */ tp = xfs_trans_alloc(mp, XFS_TRANS_GROWFSRT_FREE); - cancelflags = 0; if ((error = xfs_trans_reserve(tp, 0, XFS_GROWRTFREE_LOG_RES(nmp), 0, 0, 0))) - break; + goto error_cancel; /* * Lock out other callers by grabbing the bitmap inode lock. */ if ((error = xfs_trans_iget(mp, tp, mp->m_sb.sb_rbmino, 0, XFS_ILOCK_EXCL, &ip))) - break; + goto error_cancel; ASSERT(ip == mp->m_rbmip); /* * Update the bitmap inode's size. @@ -1990,7 +1991,7 @@ xfs_growfs_rt( */ if ((error = xfs_trans_iget(mp, tp, mp->m_sb.sb_rsumino, 0, XFS_ILOCK_EXCL, &ip))) - break; + goto error_cancel; ASSERT(ip == mp->m_rsumip); /* * Update the summary inode's size. @@ -2005,7 +2006,7 @@ xfs_growfs_rt( mp->m_rsumlevels != nmp->m_rsumlevels) { error = xfs_rtcopy_summary(mp, nmp, tp); if (error) - break; + goto error_cancel; } /* * Update superblock fields. @@ -2031,8 +2032,11 @@ xfs_growfs_rt( bp = NULL; error = xfs_rtfree_range(nmp, tp, sbp->sb_rextents, nsbp->sb_rextents - sbp->sb_rextents, &bp, &sumbno); - if (error) + if (error) { +error_cancel: + xfs_trans_cancel(tp, cancelflags); break; + } /* * Mark more blocks free in the superblock. */ @@ -2045,15 +2049,10 @@ xfs_growfs_rt( mp->m_rsumsize = nrsumsize; error = xfs_trans_commit(tp, 0); - if (error) { - tp = NULL; + if (error) break; - } } - if (error && tp) - xfs_trans_cancel(tp, cancelflags); - /* * Free the fake mp structure. */ diff --git a/fs/xfs/xfs_rw.c b/fs/xfs/xfs_rw.c index 3a82576dde9..36f3a21c54d 100644 --- a/fs/xfs/xfs_rw.c +++ b/fs/xfs/xfs_rw.c @@ -406,7 +406,7 @@ xfs_bwrite( * XXXsup how does this work for quotas. */ XFS_BUF_SET_BDSTRAT_FUNC(bp, xfs_bdstrat_cb); - XFS_BUF_SET_FSPRIVATE3(bp, mp); + bp->b_mount = mp; XFS_BUF_WRITE(bp); if ((error = XFS_bwrite(bp))) { diff --git a/fs/xfs/xfs_sb.h b/fs/xfs/xfs_sb.h index 3f8cf1587f4..1ed71916e4c 100644 --- a/fs/xfs/xfs_sb.h +++ b/fs/xfs/xfs_sb.h @@ -79,6 +79,7 @@ struct xfs_mount; #define XFS_SB_VERSION2_LAZYSBCOUNTBIT 0x00000002 /* Superblk counters */ #define XFS_SB_VERSION2_RESERVED4BIT 0x00000004 #define XFS_SB_VERSION2_ATTR2BIT 0x00000008 /* Inline attr rework */ +#define XFS_SB_VERSION2_PARENTBIT 0x00000010 /* parent pointers */ #define XFS_SB_VERSION2_OKREALFBITS \ (XFS_SB_VERSION2_LAZYSBCOUNTBIT | \ @@ -296,30 +297,34 @@ typedef enum { #define XFS_SB_VERSION_NUM(sbp) ((sbp)->sb_versionnum & XFS_SB_VERSION_NUMBITS) -#ifdef __KERNEL__ static inline int xfs_sb_good_version(xfs_sb_t *sbp) { - return (((sbp->sb_versionnum >= XFS_SB_VERSION_1) && \ - (sbp->sb_versionnum <= XFS_SB_VERSION_3)) || \ - ((XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4) && \ - !((sbp->sb_versionnum & ~XFS_SB_VERSION_OKREALBITS) || \ - ((sbp->sb_versionnum & XFS_SB_VERSION_MOREBITSBIT) && \ - (sbp->sb_features2 & ~XFS_SB_VERSION2_OKREALBITS))) && \ - (sbp->sb_shared_vn <= XFS_SB_MAX_SHARED_VN))); -} + /* We always support version 1-3 */ + if (sbp->sb_versionnum >= XFS_SB_VERSION_1 && + sbp->sb_versionnum <= XFS_SB_VERSION_3) + return 1; + + /* We support version 4 if all feature bits are supported */ + if (XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4) { + if ((sbp->sb_versionnum & ~XFS_SB_VERSION_OKREALBITS) || + ((sbp->sb_versionnum & XFS_SB_VERSION_MOREBITSBIT) && + (sbp->sb_features2 & ~XFS_SB_VERSION2_OKREALBITS))) + return 0; + +#ifdef __KERNEL__ + if (sbp->sb_shared_vn > XFS_SB_MAX_SHARED_VN) + return 0; #else -static inline int xfs_sb_good_version(xfs_sb_t *sbp) -{ - return (((sbp->sb_versionnum >= XFS_SB_VERSION_1) && \ - (sbp->sb_versionnum <= XFS_SB_VERSION_3)) || \ - ((XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4) && \ - !((sbp->sb_versionnum & ~XFS_SB_VERSION_OKREALBITS) || \ - ((sbp->sb_versionnum & XFS_SB_VERSION_MOREBITSBIT) && \ - (sbp->sb_features2 & ~XFS_SB_VERSION2_OKREALBITS))) && \ - (!(sbp->sb_versionnum & XFS_SB_VERSION_SHAREDBIT) || \ - (sbp->sb_shared_vn <= XFS_SB_MAX_SHARED_VN)))); + if ((sbp->sb_versionnum & XFS_SB_VERSION_SHAREDBIT) && + sbp->sb_shared_vn > XFS_SB_MAX_SHARED_VN) + return 0; +#endif + + return 1; + } + + return 0; } -#endif /* __KERNEL__ */ /* * Detect a mismatched features2 field. Older kernels read/wrote @@ -332,123 +337,127 @@ static inline int xfs_sb_has_mismatched_features2(xfs_sb_t *sbp) static inline unsigned xfs_sb_version_tonew(unsigned v) { - return ((((v) == XFS_SB_VERSION_1) ? \ - 0 : \ - (((v) == XFS_SB_VERSION_2) ? \ - XFS_SB_VERSION_ATTRBIT : \ - (XFS_SB_VERSION_ATTRBIT | XFS_SB_VERSION_NLINKBIT))) | \ - XFS_SB_VERSION_4); + if (v == XFS_SB_VERSION_1) + return XFS_SB_VERSION_4; + + if (v == XFS_SB_VERSION_2) + return XFS_SB_VERSION_4 | XFS_SB_VERSION_ATTRBIT; + + return XFS_SB_VERSION_4 | XFS_SB_VERSION_ATTRBIT | + XFS_SB_VERSION_NLINKBIT; } static inline unsigned xfs_sb_version_toold(unsigned v) { - return (((v) & (XFS_SB_VERSION_QUOTABIT | XFS_SB_VERSION_ALIGNBIT)) ? \ - 0 : \ - (((v) & XFS_SB_VERSION_NLINKBIT) ? \ - XFS_SB_VERSION_3 : \ - (((v) & XFS_SB_VERSION_ATTRBIT) ? \ - XFS_SB_VERSION_2 : \ - XFS_SB_VERSION_1))); + if (v & (XFS_SB_VERSION_QUOTABIT | XFS_SB_VERSION_ALIGNBIT)) + return 0; + if (v & XFS_SB_VERSION_NLINKBIT) + return XFS_SB_VERSION_3; + if (v & XFS_SB_VERSION_ATTRBIT) + return XFS_SB_VERSION_2; + return XFS_SB_VERSION_1; } static inline int xfs_sb_version_hasattr(xfs_sb_t *sbp) { - return ((sbp)->sb_versionnum == XFS_SB_VERSION_2) || \ - ((sbp)->sb_versionnum == XFS_SB_VERSION_3) || \ - ((XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4) && \ - ((sbp)->sb_versionnum & XFS_SB_VERSION_ATTRBIT)); + return sbp->sb_versionnum == XFS_SB_VERSION_2 || + sbp->sb_versionnum == XFS_SB_VERSION_3 || + (XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4 && + (sbp->sb_versionnum & XFS_SB_VERSION_ATTRBIT)); } static inline void xfs_sb_version_addattr(xfs_sb_t *sbp) { - (sbp)->sb_versionnum = (((sbp)->sb_versionnum == XFS_SB_VERSION_1) ? \ - XFS_SB_VERSION_2 : \ - ((XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4) ? \ - ((sbp)->sb_versionnum | XFS_SB_VERSION_ATTRBIT) : \ - (XFS_SB_VERSION_4 | XFS_SB_VERSION_ATTRBIT))); + if (sbp->sb_versionnum == XFS_SB_VERSION_1) + sbp->sb_versionnum = XFS_SB_VERSION_2; + else if (XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4) + sbp->sb_versionnum |= XFS_SB_VERSION_ATTRBIT; + else + sbp->sb_versionnum = XFS_SB_VERSION_4 | XFS_SB_VERSION_ATTRBIT; } static inline int xfs_sb_version_hasnlink(xfs_sb_t *sbp) { - return ((sbp)->sb_versionnum == XFS_SB_VERSION_3) || \ - ((XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4) && \ - ((sbp)->sb_versionnum & XFS_SB_VERSION_NLINKBIT)); + return sbp->sb_versionnum == XFS_SB_VERSION_3 || + (XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4 && + (sbp->sb_versionnum & XFS_SB_VERSION_NLINKBIT)); } static inline void xfs_sb_version_addnlink(xfs_sb_t *sbp) { - (sbp)->sb_versionnum = ((sbp)->sb_versionnum <= XFS_SB_VERSION_2 ? \ - XFS_SB_VERSION_3 : \ - ((sbp)->sb_versionnum | XFS_SB_VERSION_NLINKBIT)); + if (sbp->sb_versionnum <= XFS_SB_VERSION_2) + sbp->sb_versionnum = XFS_SB_VERSION_3; + else + sbp->sb_versionnum |= XFS_SB_VERSION_NLINKBIT; } static inline int xfs_sb_version_hasquota(xfs_sb_t *sbp) { - return (XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4) && \ - ((sbp)->sb_versionnum & XFS_SB_VERSION_QUOTABIT); + return XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4 && + (sbp->sb_versionnum & XFS_SB_VERSION_QUOTABIT); } static inline void xfs_sb_version_addquota(xfs_sb_t *sbp) { - (sbp)->sb_versionnum = \ - (XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4 ? \ - ((sbp)->sb_versionnum | XFS_SB_VERSION_QUOTABIT) : \ - (xfs_sb_version_tonew((sbp)->sb_versionnum) | \ - XFS_SB_VERSION_QUOTABIT)); + if (XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4) + sbp->sb_versionnum |= XFS_SB_VERSION_QUOTABIT; + else + sbp->sb_versionnum = xfs_sb_version_tonew(sbp->sb_versionnum) | + XFS_SB_VERSION_QUOTABIT; } static inline int xfs_sb_version_hasalign(xfs_sb_t *sbp) { - return (XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4) && \ - ((sbp)->sb_versionnum & XFS_SB_VERSION_ALIGNBIT); + return XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4 && + (sbp->sb_versionnum & XFS_SB_VERSION_ALIGNBIT); } static inline int xfs_sb_version_hasdalign(xfs_sb_t *sbp) { - return (XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4) && \ - ((sbp)->sb_versionnum & XFS_SB_VERSION_DALIGNBIT); + return XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4 && + (sbp->sb_versionnum & XFS_SB_VERSION_DALIGNBIT); } static inline int xfs_sb_version_hasshared(xfs_sb_t *sbp) { - return (XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4) && \ - ((sbp)->sb_versionnum & XFS_SB_VERSION_SHAREDBIT); + return XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4 && + (sbp->sb_versionnum & XFS_SB_VERSION_SHAREDBIT); } static inline int xfs_sb_version_hasdirv2(xfs_sb_t *sbp) { - return (XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4) && \ - ((sbp)->sb_versionnum & XFS_SB_VERSION_DIRV2BIT); + return XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4 && + (sbp->sb_versionnum & XFS_SB_VERSION_DIRV2BIT); } static inline int xfs_sb_version_haslogv2(xfs_sb_t *sbp) { - return (XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4) && \ - ((sbp)->sb_versionnum & XFS_SB_VERSION_LOGV2BIT); + return XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4 && + (sbp->sb_versionnum & XFS_SB_VERSION_LOGV2BIT); } static inline int xfs_sb_version_hasextflgbit(xfs_sb_t *sbp) { - return (XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4) && \ - ((sbp)->sb_versionnum & XFS_SB_VERSION_EXTFLGBIT); + return XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4 && + (sbp->sb_versionnum & XFS_SB_VERSION_EXTFLGBIT); } static inline int xfs_sb_version_hassector(xfs_sb_t *sbp) { - return (XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4) && \ - ((sbp)->sb_versionnum & XFS_SB_VERSION_SECTORBIT); + return XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4 && + (sbp->sb_versionnum & XFS_SB_VERSION_SECTORBIT); } static inline int xfs_sb_version_hasasciici(xfs_sb_t *sbp) { - return (XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4) && \ + return XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4 && (sbp->sb_versionnum & XFS_SB_VERSION_BORGBIT); } static inline int xfs_sb_version_hasmorebits(xfs_sb_t *sbp) { - return (XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4) && \ - ((sbp)->sb_versionnum & XFS_SB_VERSION_MOREBITSBIT); + return XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4 && + (sbp->sb_versionnum & XFS_SB_VERSION_MOREBITSBIT); } /* @@ -463,22 +472,20 @@ static inline int xfs_sb_version_hasmorebits(xfs_sb_t *sbp) static inline int xfs_sb_version_haslazysbcount(xfs_sb_t *sbp) { - return (xfs_sb_version_hasmorebits(sbp) && \ - ((sbp)->sb_features2 & XFS_SB_VERSION2_LAZYSBCOUNTBIT)); + return xfs_sb_version_hasmorebits(sbp) && + (sbp->sb_features2 & XFS_SB_VERSION2_LAZYSBCOUNTBIT); } static inline int xfs_sb_version_hasattr2(xfs_sb_t *sbp) { - return (xfs_sb_version_hasmorebits(sbp)) && \ - ((sbp)->sb_features2 & XFS_SB_VERSION2_ATTR2BIT); + return xfs_sb_version_hasmorebits(sbp) && + (sbp->sb_features2 & XFS_SB_VERSION2_ATTR2BIT); } static inline void xfs_sb_version_addattr2(xfs_sb_t *sbp) { - ((sbp)->sb_versionnum = \ - ((sbp)->sb_versionnum | XFS_SB_VERSION_MOREBITSBIT), \ - ((sbp)->sb_features2 = \ - ((sbp)->sb_features2 | XFS_SB_VERSION2_ATTR2BIT))); + sbp->sb_versionnum |= XFS_SB_VERSION_MOREBITSBIT; + sbp->sb_features2 |= XFS_SB_VERSION2_ATTR2BIT; } static inline void xfs_sb_version_removeattr2(xfs_sb_t *sbp) diff --git a/fs/xfs/xfs_trans.c b/fs/xfs/xfs_trans.c index 4e1c22a23be..8570b826fed 100644 --- a/fs/xfs/xfs_trans.c +++ b/fs/xfs/xfs_trans.c @@ -290,7 +290,7 @@ xfs_trans_dup( ASSERT(tp->t_ticket != NULL); ntp->t_flags = XFS_TRANS_PERM_LOG_RES | (tp->t_flags & XFS_TRANS_RESERVE); - ntp->t_ticket = tp->t_ticket; + ntp->t_ticket = xfs_log_ticket_get(tp->t_ticket); ntp->t_blk_res = tp->t_blk_res - tp->t_blk_res_used; tp->t_blk_res = tp->t_blk_res_used; ntp->t_rtx_res = tp->t_rtx_res - tp->t_rtx_res_used; @@ -1260,6 +1260,13 @@ xfs_trans_roll( trans = *tpp; /* + * transaction commit worked ok so we can drop the extra ticket + * reference that we gained in xfs_trans_dup() + */ + xfs_log_ticket_put(trans->t_ticket); + + + /* * Reserve space in the log for th next transaction. * This also pushes items in the "AIL", the list of logged items, * out to disk if they are taking up space at the tail of the log @@ -1383,11 +1390,12 @@ xfs_trans_chunk_committed( xfs_log_item_desc_t *lidp; xfs_log_item_t *lip; xfs_lsn_t item_lsn; - struct xfs_mount *mp; int i; lidp = licp->lic_descs; for (i = 0; i < licp->lic_unused; i++, lidp++) { + struct xfs_ail *ailp; + if (xfs_lic_isfree(licp, i)) { continue; } @@ -1424,19 +1432,19 @@ xfs_trans_chunk_committed( * This would cause the earlier transaction to fail * the test below. */ - mp = lip->li_mountp; - spin_lock(&mp->m_ail_lock); + ailp = lip->li_ailp; + spin_lock(&ailp->xa_lock); if (XFS_LSN_CMP(item_lsn, lip->li_lsn) > 0) { /* * This will set the item's lsn to item_lsn * and update the position of the item in * the AIL. * - * xfs_trans_update_ail() drops the AIL lock. + * xfs_trans_ail_update() drops the AIL lock. */ - xfs_trans_update_ail(mp, lip, item_lsn); + xfs_trans_ail_update(ailp, lip, item_lsn); } else { - spin_unlock(&mp->m_ail_lock); + spin_unlock(&ailp->xa_lock); } /* diff --git a/fs/xfs/xfs_trans.h b/fs/xfs/xfs_trans.h index 74c80bd2b0e..d6fe4a88d79 100644 --- a/fs/xfs/xfs_trans.h +++ b/fs/xfs/xfs_trans.h @@ -18,6 +18,8 @@ #ifndef __XFS_TRANS_H__ #define __XFS_TRANS_H__ +struct xfs_log_item; + /* * This is the structure written in the log at the head of * every transaction. It identifies the type and id of the @@ -98,76 +100,6 @@ typedef struct xfs_trans_header { #define XFS_TRANS_TYPE_MAX 41 /* new transaction types need to be reflected in xfs_logprint(8) */ - -#ifdef __KERNEL__ -struct xfs_buf; -struct xfs_buftarg; -struct xfs_efd_log_item; -struct xfs_efi_log_item; -struct xfs_inode; -struct xfs_item_ops; -struct xfs_log_iovec; -struct xfs_log_item; -struct xfs_log_item_desc; -struct xfs_mount; -struct xfs_trans; -struct xfs_dquot_acct; - -typedef struct xfs_log_item { - struct list_head li_ail; /* AIL pointers */ - xfs_lsn_t li_lsn; /* last on-disk lsn */ - struct xfs_log_item_desc *li_desc; /* ptr to current desc*/ - struct xfs_mount *li_mountp; /* ptr to fs mount */ - uint li_type; /* item type */ - uint li_flags; /* misc flags */ - struct xfs_log_item *li_bio_list; /* buffer item list */ - void (*li_cb)(struct xfs_buf *, - struct xfs_log_item *); - /* buffer item iodone */ - /* callback func */ - struct xfs_item_ops *li_ops; /* function list */ -} xfs_log_item_t; - -#define XFS_LI_IN_AIL 0x1 -#define XFS_LI_ABORTED 0x2 - -typedef struct xfs_item_ops { - uint (*iop_size)(xfs_log_item_t *); - void (*iop_format)(xfs_log_item_t *, struct xfs_log_iovec *); - void (*iop_pin)(xfs_log_item_t *); - void (*iop_unpin)(xfs_log_item_t *, int); - void (*iop_unpin_remove)(xfs_log_item_t *, struct xfs_trans *); - uint (*iop_trylock)(xfs_log_item_t *); - void (*iop_unlock)(xfs_log_item_t *); - xfs_lsn_t (*iop_committed)(xfs_log_item_t *, xfs_lsn_t); - void (*iop_push)(xfs_log_item_t *); - void (*iop_pushbuf)(xfs_log_item_t *); - void (*iop_committing)(xfs_log_item_t *, xfs_lsn_t); -} xfs_item_ops_t; - -#define IOP_SIZE(ip) (*(ip)->li_ops->iop_size)(ip) -#define IOP_FORMAT(ip,vp) (*(ip)->li_ops->iop_format)(ip, vp) -#define IOP_PIN(ip) (*(ip)->li_ops->iop_pin)(ip) -#define IOP_UNPIN(ip, flags) (*(ip)->li_ops->iop_unpin)(ip, flags) -#define IOP_UNPIN_REMOVE(ip,tp) (*(ip)->li_ops->iop_unpin_remove)(ip, tp) -#define IOP_TRYLOCK(ip) (*(ip)->li_ops->iop_trylock)(ip) -#define IOP_UNLOCK(ip) (*(ip)->li_ops->iop_unlock)(ip) -#define IOP_COMMITTED(ip, lsn) (*(ip)->li_ops->iop_committed)(ip, lsn) -#define IOP_PUSH(ip) (*(ip)->li_ops->iop_push)(ip) -#define IOP_PUSHBUF(ip) (*(ip)->li_ops->iop_pushbuf)(ip) -#define IOP_COMMITTING(ip, lsn) (*(ip)->li_ops->iop_committing)(ip, lsn) - -/* - * Return values for the IOP_TRYLOCK() routines. - */ -#define XFS_ITEM_SUCCESS 0 -#define XFS_ITEM_PINNED 1 -#define XFS_ITEM_LOCKED 2 -#define XFS_ITEM_FLUSHING 3 -#define XFS_ITEM_PUSHBUF 4 - -#endif /* __KERNEL__ */ - /* * This structure is used to track log items associated with * a transaction. It points to the log item and keeps some @@ -176,7 +108,7 @@ typedef struct xfs_item_ops { * once we get to commit processing (see xfs_trans_commit()). */ typedef struct xfs_log_item_desc { - xfs_log_item_t *lid_item; + struct xfs_log_item *lid_item; ushort lid_size; unsigned char lid_flags; unsigned char lid_index; @@ -276,94 +208,6 @@ xfs_lic_desc_to_chunk(xfs_log_item_desc_t *dp) (xfs_caddr_t)(((xfs_log_item_chunk_t*)0)->lic_descs)); } -#ifdef __KERNEL__ -/* - * This structure is used to maintain a list of block ranges that have been - * freed in the transaction. The ranges are listed in the perag[] busy list - * between when they're freed and the transaction is committed to disk. - */ - -typedef struct xfs_log_busy_slot { - xfs_agnumber_t lbc_ag; - ushort lbc_idx; /* index in perag.busy[] */ -} xfs_log_busy_slot_t; - -#define XFS_LBC_NUM_SLOTS 31 -typedef struct xfs_log_busy_chunk { - struct xfs_log_busy_chunk *lbc_next; - uint lbc_free; /* free slots bitmask */ - ushort lbc_unused; /* first unused */ - xfs_log_busy_slot_t lbc_busy[XFS_LBC_NUM_SLOTS]; -} xfs_log_busy_chunk_t; - -#define XFS_LBC_MAX_SLOT (XFS_LBC_NUM_SLOTS - 1) -#define XFS_LBC_FREEMASK ((1U << XFS_LBC_NUM_SLOTS) - 1) - -#define XFS_LBC_INIT(cp) ((cp)->lbc_free = XFS_LBC_FREEMASK) -#define XFS_LBC_CLAIM(cp, slot) ((cp)->lbc_free &= ~(1 << (slot))) -#define XFS_LBC_SLOT(cp, slot) (&((cp)->lbc_busy[(slot)])) -#define XFS_LBC_VACANCY(cp) (((cp)->lbc_free) & XFS_LBC_FREEMASK) -#define XFS_LBC_ISFREE(cp, slot) ((cp)->lbc_free & (1 << (slot))) - -/* - * This is the type of function which can be given to xfs_trans_callback() - * to be called upon the transaction's commit to disk. - */ -typedef void (*xfs_trans_callback_t)(struct xfs_trans *, void *); - -/* - * This is the structure maintained for every active transaction. - */ -typedef struct xfs_trans { - unsigned int t_magic; /* magic number */ - xfs_log_callback_t t_logcb; /* log callback struct */ - unsigned int t_type; /* transaction type */ - unsigned int t_log_res; /* amt of log space resvd */ - unsigned int t_log_count; /* count for perm log res */ - unsigned int t_blk_res; /* # of blocks resvd */ - unsigned int t_blk_res_used; /* # of resvd blocks used */ - unsigned int t_rtx_res; /* # of rt extents resvd */ - unsigned int t_rtx_res_used; /* # of resvd rt extents used */ - xfs_log_ticket_t t_ticket; /* log mgr ticket */ - xfs_lsn_t t_lsn; /* log seq num of start of - * transaction. */ - xfs_lsn_t t_commit_lsn; /* log seq num of end of - * transaction. */ - struct xfs_mount *t_mountp; /* ptr to fs mount struct */ - struct xfs_dquot_acct *t_dqinfo; /* acctg info for dquots */ - xfs_trans_callback_t t_callback; /* transaction callback */ - void *t_callarg; /* callback arg */ - unsigned int t_flags; /* misc flags */ - int64_t t_icount_delta; /* superblock icount change */ - int64_t t_ifree_delta; /* superblock ifree change */ - int64_t t_fdblocks_delta; /* superblock fdblocks chg */ - int64_t t_res_fdblocks_delta; /* on-disk only chg */ - int64_t t_frextents_delta;/* superblock freextents chg*/ - int64_t t_res_frextents_delta; /* on-disk only chg */ -#ifdef DEBUG - int64_t t_ag_freeblks_delta; /* debugging counter */ - int64_t t_ag_flist_delta; /* debugging counter */ - int64_t t_ag_btree_delta; /* debugging counter */ -#endif - int64_t t_dblocks_delta;/* superblock dblocks change */ - int64_t t_agcount_delta;/* superblock agcount change */ - int64_t t_imaxpct_delta;/* superblock imaxpct change */ - int64_t t_rextsize_delta;/* superblock rextsize chg */ - int64_t t_rbmblocks_delta;/* superblock rbmblocks chg */ - int64_t t_rblocks_delta;/* superblock rblocks change */ - int64_t t_rextents_delta;/* superblocks rextents chg */ - int64_t t_rextslog_delta;/* superblocks rextslog chg */ - unsigned int t_items_free; /* log item descs free */ - xfs_log_item_chunk_t t_items; /* first log item desc chunk */ - xfs_trans_header_t t_header; /* header for in-log trans */ - unsigned int t_busy_free; /* busy descs free */ - xfs_log_busy_chunk_t t_busy; /* busy/async free blocks */ - unsigned long t_pflags; /* saved process flags state */ -} xfs_trans_t; - -#endif /* __KERNEL__ */ - - #define XFS_TRANS_MAGIC 0x5452414E /* 'TRAN' */ /* * Values for t_flags. @@ -906,6 +750,157 @@ typedef struct xfs_trans { #define XFS_DQUOT_REF 1 #ifdef __KERNEL__ + +struct xfs_buf; +struct xfs_buftarg; +struct xfs_efd_log_item; +struct xfs_efi_log_item; +struct xfs_inode; +struct xfs_item_ops; +struct xfs_log_iovec; +struct xfs_log_item_desc; +struct xfs_mount; +struct xfs_trans; +struct xfs_dquot_acct; + +typedef struct xfs_log_item { + struct list_head li_ail; /* AIL pointers */ + xfs_lsn_t li_lsn; /* last on-disk lsn */ + struct xfs_log_item_desc *li_desc; /* ptr to current desc*/ + struct xfs_mount *li_mountp; /* ptr to fs mount */ + struct xfs_ail *li_ailp; /* ptr to AIL */ + uint li_type; /* item type */ + uint li_flags; /* misc flags */ + struct xfs_log_item *li_bio_list; /* buffer item list */ + void (*li_cb)(struct xfs_buf *, + struct xfs_log_item *); + /* buffer item iodone */ + /* callback func */ + struct xfs_item_ops *li_ops; /* function list */ +} xfs_log_item_t; + +#define XFS_LI_IN_AIL 0x1 +#define XFS_LI_ABORTED 0x2 + +typedef struct xfs_item_ops { + uint (*iop_size)(xfs_log_item_t *); + void (*iop_format)(xfs_log_item_t *, struct xfs_log_iovec *); + void (*iop_pin)(xfs_log_item_t *); + void (*iop_unpin)(xfs_log_item_t *, int); + void (*iop_unpin_remove)(xfs_log_item_t *, struct xfs_trans *); + uint (*iop_trylock)(xfs_log_item_t *); + void (*iop_unlock)(xfs_log_item_t *); + xfs_lsn_t (*iop_committed)(xfs_log_item_t *, xfs_lsn_t); + void (*iop_push)(xfs_log_item_t *); + void (*iop_pushbuf)(xfs_log_item_t *); + void (*iop_committing)(xfs_log_item_t *, xfs_lsn_t); +} xfs_item_ops_t; + +#define IOP_SIZE(ip) (*(ip)->li_ops->iop_size)(ip) +#define IOP_FORMAT(ip,vp) (*(ip)->li_ops->iop_format)(ip, vp) +#define IOP_PIN(ip) (*(ip)->li_ops->iop_pin)(ip) +#define IOP_UNPIN(ip, flags) (*(ip)->li_ops->iop_unpin)(ip, flags) +#define IOP_UNPIN_REMOVE(ip,tp) (*(ip)->li_ops->iop_unpin_remove)(ip, tp) +#define IOP_TRYLOCK(ip) (*(ip)->li_ops->iop_trylock)(ip) +#define IOP_UNLOCK(ip) (*(ip)->li_ops->iop_unlock)(ip) +#define IOP_COMMITTED(ip, lsn) (*(ip)->li_ops->iop_committed)(ip, lsn) +#define IOP_PUSH(ip) (*(ip)->li_ops->iop_push)(ip) +#define IOP_PUSHBUF(ip) (*(ip)->li_ops->iop_pushbuf)(ip) +#define IOP_COMMITTING(ip, lsn) (*(ip)->li_ops->iop_committing)(ip, lsn) + +/* + * Return values for the IOP_TRYLOCK() routines. + */ +#define XFS_ITEM_SUCCESS 0 +#define XFS_ITEM_PINNED 1 +#define XFS_ITEM_LOCKED 2 +#define XFS_ITEM_FLUSHING 3 +#define XFS_ITEM_PUSHBUF 4 + +/* + * This structure is used to maintain a list of block ranges that have been + * freed in the transaction. The ranges are listed in the perag[] busy list + * between when they're freed and the transaction is committed to disk. + */ + +typedef struct xfs_log_busy_slot { + xfs_agnumber_t lbc_ag; + ushort lbc_idx; /* index in perag.busy[] */ +} xfs_log_busy_slot_t; + +#define XFS_LBC_NUM_SLOTS 31 +typedef struct xfs_log_busy_chunk { + struct xfs_log_busy_chunk *lbc_next; + uint lbc_free; /* free slots bitmask */ + ushort lbc_unused; /* first unused */ + xfs_log_busy_slot_t lbc_busy[XFS_LBC_NUM_SLOTS]; +} xfs_log_busy_chunk_t; + +#define XFS_LBC_MAX_SLOT (XFS_LBC_NUM_SLOTS - 1) +#define XFS_LBC_FREEMASK ((1U << XFS_LBC_NUM_SLOTS) - 1) + +#define XFS_LBC_INIT(cp) ((cp)->lbc_free = XFS_LBC_FREEMASK) +#define XFS_LBC_CLAIM(cp, slot) ((cp)->lbc_free &= ~(1 << (slot))) +#define XFS_LBC_SLOT(cp, slot) (&((cp)->lbc_busy[(slot)])) +#define XFS_LBC_VACANCY(cp) (((cp)->lbc_free) & XFS_LBC_FREEMASK) +#define XFS_LBC_ISFREE(cp, slot) ((cp)->lbc_free & (1 << (slot))) + +/* + * This is the type of function which can be given to xfs_trans_callback() + * to be called upon the transaction's commit to disk. + */ +typedef void (*xfs_trans_callback_t)(struct xfs_trans *, void *); + +/* + * This is the structure maintained for every active transaction. + */ +typedef struct xfs_trans { + unsigned int t_magic; /* magic number */ + xfs_log_callback_t t_logcb; /* log callback struct */ + unsigned int t_type; /* transaction type */ + unsigned int t_log_res; /* amt of log space resvd */ + unsigned int t_log_count; /* count for perm log res */ + unsigned int t_blk_res; /* # of blocks resvd */ + unsigned int t_blk_res_used; /* # of resvd blocks used */ + unsigned int t_rtx_res; /* # of rt extents resvd */ + unsigned int t_rtx_res_used; /* # of resvd rt extents used */ + xfs_log_ticket_t t_ticket; /* log mgr ticket */ + xfs_lsn_t t_lsn; /* log seq num of start of + * transaction. */ + xfs_lsn_t t_commit_lsn; /* log seq num of end of + * transaction. */ + struct xfs_mount *t_mountp; /* ptr to fs mount struct */ + struct xfs_dquot_acct *t_dqinfo; /* acctg info for dquots */ + xfs_trans_callback_t t_callback; /* transaction callback */ + void *t_callarg; /* callback arg */ + unsigned int t_flags; /* misc flags */ + int64_t t_icount_delta; /* superblock icount change */ + int64_t t_ifree_delta; /* superblock ifree change */ + int64_t t_fdblocks_delta; /* superblock fdblocks chg */ + int64_t t_res_fdblocks_delta; /* on-disk only chg */ + int64_t t_frextents_delta;/* superblock freextents chg*/ + int64_t t_res_frextents_delta; /* on-disk only chg */ +#ifdef DEBUG + int64_t t_ag_freeblks_delta; /* debugging counter */ + int64_t t_ag_flist_delta; /* debugging counter */ + int64_t t_ag_btree_delta; /* debugging counter */ +#endif + int64_t t_dblocks_delta;/* superblock dblocks change */ + int64_t t_agcount_delta;/* superblock agcount change */ + int64_t t_imaxpct_delta;/* superblock imaxpct change */ + int64_t t_rextsize_delta;/* superblock rextsize chg */ + int64_t t_rbmblocks_delta;/* superblock rbmblocks chg */ + int64_t t_rblocks_delta;/* superblock rblocks change */ + int64_t t_rextents_delta;/* superblocks rextents chg */ + int64_t t_rextslog_delta;/* superblocks rextslog chg */ + unsigned int t_items_free; /* log item descs free */ + xfs_log_item_chunk_t t_items; /* first log item desc chunk */ + xfs_trans_header_t t_header; /* header for in-log trans */ + unsigned int t_busy_free; /* busy descs free */ + xfs_log_busy_chunk_t t_busy; /* busy/async free blocks */ + unsigned long t_pflags; /* saved process flags state */ +} xfs_trans_t; + /* * XFS transaction mechanism exported interfaces that are * actually macros. @@ -928,7 +923,6 @@ typedef struct xfs_trans { /* * XFS transaction mechanism exported interfaces. */ -void xfs_trans_init(struct xfs_mount *); xfs_trans_t *xfs_trans_alloc(struct xfs_mount *, uint); xfs_trans_t *_xfs_trans_alloc(struct xfs_mount *, uint); xfs_trans_t *xfs_trans_dup(xfs_trans_t *); @@ -975,13 +969,8 @@ int _xfs_trans_commit(xfs_trans_t *, int *); #define xfs_trans_commit(tp, flags) _xfs_trans_commit(tp, flags, NULL) void xfs_trans_cancel(xfs_trans_t *, int); -int xfs_trans_roll(struct xfs_trans **, struct xfs_inode *); int xfs_trans_ail_init(struct xfs_mount *); void xfs_trans_ail_destroy(struct xfs_mount *); -void xfs_trans_push_ail(struct xfs_mount *, xfs_lsn_t); -xfs_lsn_t xfs_trans_tail_ail(struct xfs_mount *); -void xfs_trans_unlocked_item(struct xfs_mount *, - xfs_log_item_t *); xfs_log_busy_slot_t *xfs_trans_add_busy(xfs_trans_t *tp, xfs_agnumber_t ag, xfs_extlen_t idx); @@ -990,4 +979,7 @@ extern kmem_zone_t *xfs_trans_zone; #endif /* __KERNEL__ */ +void xfs_trans_init(struct xfs_mount *); +int xfs_trans_roll(struct xfs_trans **, struct xfs_inode *); + #endif /* __XFS_TRANS_H__ */ diff --git a/fs/xfs/xfs_trans_ail.c b/fs/xfs/xfs_trans_ail.c index 1f77c00af56..2d47f10f8be 100644 --- a/fs/xfs/xfs_trans_ail.c +++ b/fs/xfs/xfs_trans_ail.c @@ -1,5 +1,6 @@ /* * Copyright (c) 2000-2002,2005 Silicon Graphics, Inc. + * Copyright (c) 2008 Dave Chinner * All Rights Reserved. * * This program is free software; you can redistribute it and/or @@ -28,13 +29,13 @@ #include "xfs_trans_priv.h" #include "xfs_error.h" -STATIC void xfs_ail_insert(xfs_ail_t *, xfs_log_item_t *); -STATIC xfs_log_item_t * xfs_ail_delete(xfs_ail_t *, xfs_log_item_t *); -STATIC xfs_log_item_t * xfs_ail_min(xfs_ail_t *); -STATIC xfs_log_item_t * xfs_ail_next(xfs_ail_t *, xfs_log_item_t *); +STATIC void xfs_ail_insert(struct xfs_ail *, xfs_log_item_t *); +STATIC xfs_log_item_t * xfs_ail_delete(struct xfs_ail *, xfs_log_item_t *); +STATIC xfs_log_item_t * xfs_ail_min(struct xfs_ail *); +STATIC xfs_log_item_t * xfs_ail_next(struct xfs_ail *, xfs_log_item_t *); #ifdef DEBUG -STATIC void xfs_ail_check(xfs_ail_t *, xfs_log_item_t *); +STATIC void xfs_ail_check(struct xfs_ail *, xfs_log_item_t *); #else #define xfs_ail_check(a,l) #endif /* DEBUG */ @@ -50,20 +51,20 @@ STATIC void xfs_ail_check(xfs_ail_t *, xfs_log_item_t *); * lsn of the last item in the AIL. */ xfs_lsn_t -xfs_trans_tail_ail( - xfs_mount_t *mp) +xfs_trans_ail_tail( + struct xfs_ail *ailp) { xfs_lsn_t lsn; xfs_log_item_t *lip; - spin_lock(&mp->m_ail_lock); - lip = xfs_ail_min(&mp->m_ail); + spin_lock(&ailp->xa_lock); + lip = xfs_ail_min(ailp); if (lip == NULL) { lsn = (xfs_lsn_t)0; } else { lsn = lip->li_lsn; } - spin_unlock(&mp->m_ail_lock); + spin_unlock(&ailp->xa_lock); return lsn; } @@ -85,16 +86,125 @@ xfs_trans_tail_ail( * any of the objects, so the lock is not needed. */ void -xfs_trans_push_ail( - xfs_mount_t *mp, - xfs_lsn_t threshold_lsn) +xfs_trans_ail_push( + struct xfs_ail *ailp, + xfs_lsn_t threshold_lsn) { - xfs_log_item_t *lip; + xfs_log_item_t *lip; + + lip = xfs_ail_min(ailp); + if (lip && !XFS_FORCED_SHUTDOWN(ailp->xa_mount)) { + if (XFS_LSN_CMP(threshold_lsn, ailp->xa_target) > 0) + xfsaild_wakeup(ailp, threshold_lsn); + } +} + +/* + * AIL traversal cursor initialisation. + * + * The cursor keeps track of where our current traversal is up + * to by tracking the next Æ£tem in the list for us. However, for + * this to be safe, removing an object from the AIL needs to invalidate + * any cursor that points to it. hence the traversal cursor needs to + * be linked to the struct xfs_ail so that deletion can search all the + * active cursors for invalidation. + * + * We don't link the push cursor because it is embedded in the struct + * xfs_ail and hence easily findable. + */ +STATIC void +xfs_trans_ail_cursor_init( + struct xfs_ail *ailp, + struct xfs_ail_cursor *cur) +{ + cur->item = NULL; + if (cur == &ailp->xa_cursors) + return; + + cur->next = ailp->xa_cursors.next; + ailp->xa_cursors.next = cur; +} + +/* + * Set the cursor to the next item, because when we look + * up the cursor the current item may have been freed. + */ +STATIC void +xfs_trans_ail_cursor_set( + struct xfs_ail *ailp, + struct xfs_ail_cursor *cur, + struct xfs_log_item *lip) +{ + if (lip) + cur->item = xfs_ail_next(ailp, lip); +} + +/* + * Get the next item in the traversal and advance the cursor. + * If the cursor was invalidated (inidicated by a lip of 1), + * restart the traversal. + */ +struct xfs_log_item * +xfs_trans_ail_cursor_next( + struct xfs_ail *ailp, + struct xfs_ail_cursor *cur) +{ + struct xfs_log_item *lip = cur->item; + + if ((__psint_t)lip & 1) + lip = xfs_ail_min(ailp); + xfs_trans_ail_cursor_set(ailp, cur, lip); + return lip; +} + +/* + * Now that the traversal is complete, we need to remove the cursor + * from the list of traversing cursors. Avoid removing the embedded + * push cursor, but use the fact it is alway present to make the + * list deletion simple. + */ +void +xfs_trans_ail_cursor_done( + struct xfs_ail *ailp, + struct xfs_ail_cursor *done) +{ + struct xfs_ail_cursor *prev = NULL; + struct xfs_ail_cursor *cur; + + done->item = NULL; + if (done == &ailp->xa_cursors) + return; + prev = &ailp->xa_cursors; + for (cur = prev->next; cur; prev = cur, cur = prev->next) { + if (cur == done) { + prev->next = cur->next; + break; + } + } + ASSERT(cur); +} + +/* + * Invalidate any cursor that is pointing to this item. This is + * called when an item is removed from the AIL. Any cursor pointing + * to this object is now invalid and the traversal needs to be + * terminated so it doesn't reference a freed object. We set the + * cursor item to a value of 1 so we can distinguish between an + * invalidation and the end of the list when getting the next item + * from the cursor. + */ +STATIC void +xfs_trans_ail_cursor_clear( + struct xfs_ail *ailp, + struct xfs_log_item *lip) +{ + struct xfs_ail_cursor *cur; - lip = xfs_ail_min(&mp->m_ail); - if (lip && !XFS_FORCED_SHUTDOWN(mp)) { - if (XFS_LSN_CMP(threshold_lsn, mp->m_ail.xa_target) > 0) - xfsaild_wakeup(mp, threshold_lsn); + /* need to search all cursors */ + for (cur = &ailp->xa_cursors; cur; cur = cur->next) { + if (cur->item == lip) + cur->item = (struct xfs_log_item *) + ((__psint_t)cur->item | 1); } } @@ -103,25 +213,27 @@ xfs_trans_push_ail( * Return the current tree generation number for use * in calls to xfs_trans_next_ail(). */ -STATIC xfs_log_item_t * -xfs_trans_first_push_ail( - xfs_mount_t *mp, - int *gen, - xfs_lsn_t lsn) +xfs_log_item_t * +xfs_trans_ail_cursor_first( + struct xfs_ail *ailp, + struct xfs_ail_cursor *cur, + xfs_lsn_t lsn) { - xfs_log_item_t *lip; + xfs_log_item_t *lip; - lip = xfs_ail_min(&mp->m_ail); - *gen = (int)mp->m_ail.xa_gen; + xfs_trans_ail_cursor_init(ailp, cur); + lip = xfs_ail_min(ailp); if (lsn == 0) - return lip; + goto out; - list_for_each_entry(lip, &mp->m_ail.xa_ail, li_ail) { + list_for_each_entry(lip, &ailp->xa_ail, li_ail) { if (XFS_LSN_CMP(lip->li_lsn, lsn) >= 0) - return lip; + goto out; } - - return NULL; + lip = NULL; +out: + xfs_trans_ail_cursor_set(ailp, cur, lip); + return lip; } /* @@ -129,29 +241,29 @@ xfs_trans_first_push_ail( */ long xfsaild_push( - xfs_mount_t *mp, + struct xfs_ail *ailp, xfs_lsn_t *last_lsn) { long tout = 1000; /* milliseconds */ xfs_lsn_t last_pushed_lsn = *last_lsn; - xfs_lsn_t target = mp->m_ail.xa_target; + xfs_lsn_t target = ailp->xa_target; xfs_lsn_t lsn; xfs_log_item_t *lip; - int gen; - int restarts; int flush_log, count, stuck; + xfs_mount_t *mp = ailp->xa_mount; + struct xfs_ail_cursor *cur = &ailp->xa_cursors; -#define XFS_TRANS_PUSH_AIL_RESTARTS 10 - - spin_lock(&mp->m_ail_lock); - lip = xfs_trans_first_push_ail(mp, &gen, *last_lsn); + spin_lock(&ailp->xa_lock); + xfs_trans_ail_cursor_init(ailp, cur); + lip = xfs_trans_ail_cursor_first(ailp, cur, *last_lsn); if (!lip || XFS_FORCED_SHUTDOWN(mp)) { /* * AIL is empty or our push has reached the end. */ - spin_unlock(&mp->m_ail_lock); + xfs_trans_ail_cursor_done(ailp, cur); + spin_unlock(&ailp->xa_lock); last_pushed_lsn = 0; - goto out; + return tout; } XFS_STATS_INC(xs_push_ail); @@ -169,7 +281,7 @@ xfsaild_push( */ tout = 10; lsn = lip->li_lsn; - flush_log = stuck = count = restarts = 0; + flush_log = stuck = count = 0; while ((XFS_LSN_CMP(lip->li_lsn, target) < 0)) { int lock_result; /* @@ -184,7 +296,7 @@ xfsaild_push( * skip to the next item in the list. */ lock_result = IOP_TRYLOCK(lip); - spin_unlock(&mp->m_ail_lock); + spin_unlock(&ailp->xa_lock); switch (lock_result) { case XFS_ITEM_SUCCESS: XFS_STATS_INC(xs_push_ail_success); @@ -221,7 +333,7 @@ xfsaild_push( break; } - spin_lock(&mp->m_ail_lock); + spin_lock(&ailp->xa_lock); /* should we bother continuing? */ if (XFS_FORCED_SHUTDOWN(mp)) break; @@ -244,14 +356,13 @@ xfsaild_push( if (stuck > 100) break; - lip = xfs_trans_next_ail(mp, lip, &gen, &restarts); + lip = xfs_trans_ail_cursor_next(ailp, cur); if (lip == NULL) break; - if (restarts > XFS_TRANS_PUSH_AIL_RESTARTS) - break; lsn = lip->li_lsn; } - spin_unlock(&mp->m_ail_lock); + xfs_trans_ail_cursor_done(ailp, cur); + spin_unlock(&ailp->xa_lock); if (flush_log) { /* @@ -274,8 +385,7 @@ xfsaild_push( */ tout += 20; last_pushed_lsn = 0; - } else if ((restarts > XFS_TRANS_PUSH_AIL_RESTARTS) || - ((stuck * 100) / count > 90)) { + } else if ((stuck * 100) / count > 90) { /* * Either there is a lot of contention on the AIL or we * are stuck due to operations in progress. "Stuck" in this @@ -287,7 +397,6 @@ xfsaild_push( */ tout += 10; } -out: *last_lsn = last_pushed_lsn; return tout; } /* xfsaild_push */ @@ -303,7 +412,7 @@ out: */ void xfs_trans_unlocked_item( - xfs_mount_t *mp, + struct xfs_ail *ailp, xfs_log_item_t *lip) { xfs_log_item_t *min_lip; @@ -315,7 +424,7 @@ xfs_trans_unlocked_item( * over some potentially valid data. */ if (!(lip->li_flags & XFS_LI_IN_AIL) || - XFS_FORCED_SHUTDOWN(mp)) { + XFS_FORCED_SHUTDOWN(ailp->xa_mount)) { return; } @@ -331,10 +440,10 @@ xfs_trans_unlocked_item( * the call to xfs_log_move_tail() doesn't do anything if there's * not enough free space to wake people up so we're safe calling it. */ - min_lip = xfs_ail_min(&mp->m_ail); + min_lip = xfs_ail_min(ailp); if (min_lip == lip) - xfs_log_move_tail(mp, 1); + xfs_log_move_tail(ailp->xa_mount, 1); } /* xfs_trans_unlocked_item */ @@ -347,41 +456,37 @@ xfs_trans_unlocked_item( * we move in the AIL is the minimum one, update the tail lsn in the * log manager. * - * Increment the AIL's generation count to indicate that the tree - * has changed. - * * This function must be called with the AIL lock held. The lock * is dropped before returning. */ void -xfs_trans_update_ail( - xfs_mount_t *mp, +xfs_trans_ail_update( + struct xfs_ail *ailp, xfs_log_item_t *lip, - xfs_lsn_t lsn) __releases(mp->m_ail_lock) + xfs_lsn_t lsn) __releases(ailp->xa_lock) { - xfs_log_item_t *dlip=NULL; + xfs_log_item_t *dlip = NULL; xfs_log_item_t *mlip; /* ptr to minimum lip */ - mlip = xfs_ail_min(&mp->m_ail); + mlip = xfs_ail_min(ailp); if (lip->li_flags & XFS_LI_IN_AIL) { - dlip = xfs_ail_delete(&mp->m_ail, lip); + dlip = xfs_ail_delete(ailp, lip); ASSERT(dlip == lip); + xfs_trans_ail_cursor_clear(ailp, dlip); } else { lip->li_flags |= XFS_LI_IN_AIL; } lip->li_lsn = lsn; - - xfs_ail_insert(&mp->m_ail, lip); - mp->m_ail.xa_gen++; + xfs_ail_insert(ailp, lip); if (mlip == dlip) { - mlip = xfs_ail_min(&mp->m_ail); - spin_unlock(&mp->m_ail_lock); - xfs_log_move_tail(mp, mlip->li_lsn); + mlip = xfs_ail_min(ailp); + spin_unlock(&ailp->xa_lock); + xfs_log_move_tail(ailp->xa_mount, mlip->li_lsn); } else { - spin_unlock(&mp->m_ail_lock); + spin_unlock(&ailp->xa_lock); } @@ -403,29 +508,30 @@ xfs_trans_update_ail( * is dropped before returning. */ void -xfs_trans_delete_ail( - xfs_mount_t *mp, - xfs_log_item_t *lip) __releases(mp->m_ail_lock) +xfs_trans_ail_delete( + struct xfs_ail *ailp, + xfs_log_item_t *lip) __releases(ailp->xa_lock) { xfs_log_item_t *dlip; xfs_log_item_t *mlip; if (lip->li_flags & XFS_LI_IN_AIL) { - mlip = xfs_ail_min(&mp->m_ail); - dlip = xfs_ail_delete(&mp->m_ail, lip); + mlip = xfs_ail_min(ailp); + dlip = xfs_ail_delete(ailp, lip); ASSERT(dlip == lip); + xfs_trans_ail_cursor_clear(ailp, dlip); lip->li_flags &= ~XFS_LI_IN_AIL; lip->li_lsn = 0; - mp->m_ail.xa_gen++; if (mlip == dlip) { - mlip = xfs_ail_min(&mp->m_ail); - spin_unlock(&mp->m_ail_lock); - xfs_log_move_tail(mp, (mlip ? mlip->li_lsn : 0)); + mlip = xfs_ail_min(ailp); + spin_unlock(&ailp->xa_lock); + xfs_log_move_tail(ailp->xa_mount, + (mlip ? mlip->li_lsn : 0)); } else { - spin_unlock(&mp->m_ail_lock); + spin_unlock(&ailp->xa_lock); } } else { @@ -433,13 +539,13 @@ xfs_trans_delete_ail( * If the file system is not being shutdown, we are in * serious trouble if we get to this stage. */ - if (XFS_FORCED_SHUTDOWN(mp)) - spin_unlock(&mp->m_ail_lock); - else { + struct xfs_mount *mp = ailp->xa_mount; + + spin_unlock(&ailp->xa_lock); + if (!XFS_FORCED_SHUTDOWN(mp)) { xfs_cmn_err(XFS_PTAG_AILDELETE, CE_ALERT, mp, "%s: attempting to delete a log item that is not in the AIL", __func__); - spin_unlock(&mp->m_ail_lock); xfs_force_shutdown(mp, SHUTDOWN_CORRUPT_INCORE); } } @@ -448,56 +554,6 @@ xfs_trans_delete_ail( /* - * Return the item in the AIL with the smallest lsn. - * Return the current tree generation number for use - * in calls to xfs_trans_next_ail(). - */ -xfs_log_item_t * -xfs_trans_first_ail( - xfs_mount_t *mp, - int *gen) -{ - xfs_log_item_t *lip; - - lip = xfs_ail_min(&mp->m_ail); - *gen = (int)mp->m_ail.xa_gen; - - return lip; -} - -/* - * If the generation count of the tree has not changed since the - * caller last took something from the AIL, then return the elmt - * in the tree which follows the one given. If the count has changed, - * then return the minimum elmt of the AIL and bump the restarts counter - * if one is given. - */ -xfs_log_item_t * -xfs_trans_next_ail( - xfs_mount_t *mp, - xfs_log_item_t *lip, - int *gen, - int *restarts) -{ - xfs_log_item_t *nlip; - - ASSERT(mp && lip && gen); - if (mp->m_ail.xa_gen == *gen) { - nlip = xfs_ail_next(&mp->m_ail, lip); - } else { - nlip = xfs_ail_min(&mp->m_ail); - *gen = (int)mp->m_ail.xa_gen; - if (restarts != NULL) { - XFS_STATS_INC(xs_push_ail_restarts); - (*restarts)++; - } - } - - return (nlip); -} - - -/* * The active item list (AIL) is a doubly linked list of log * items sorted by ascending lsn. The base of the list is * a forw/back pointer pair embedded in the xfs mount structure. @@ -515,15 +571,35 @@ int xfs_trans_ail_init( xfs_mount_t *mp) { - INIT_LIST_HEAD(&mp->m_ail.xa_ail); - return xfsaild_start(mp); + struct xfs_ail *ailp; + int error; + + ailp = kmem_zalloc(sizeof(struct xfs_ail), KM_MAYFAIL); + if (!ailp) + return ENOMEM; + + ailp->xa_mount = mp; + INIT_LIST_HEAD(&ailp->xa_ail); + spin_lock_init(&ailp->xa_lock); + error = xfsaild_start(ailp); + if (error) + goto out_free_ailp; + mp->m_ail = ailp; + return 0; + +out_free_ailp: + kmem_free(ailp); + return error; } void xfs_trans_ail_destroy( xfs_mount_t *mp) { - xfsaild_stop(mp); + struct xfs_ail *ailp = mp->m_ail; + + xfsaild_stop(ailp); + kmem_free(ailp); } /* @@ -534,7 +610,7 @@ xfs_trans_ail_destroy( */ STATIC void xfs_ail_insert( - xfs_ail_t *ailp, + struct xfs_ail *ailp, xfs_log_item_t *lip) /* ARGSUSED */ { @@ -568,7 +644,7 @@ xfs_ail_insert( /*ARGSUSED*/ STATIC xfs_log_item_t * xfs_ail_delete( - xfs_ail_t *ailp, + struct xfs_ail *ailp, xfs_log_item_t *lip) /* ARGSUSED */ { @@ -585,7 +661,7 @@ xfs_ail_delete( */ STATIC xfs_log_item_t * xfs_ail_min( - xfs_ail_t *ailp) + struct xfs_ail *ailp) /* ARGSUSED */ { if (list_empty(&ailp->xa_ail)) @@ -601,7 +677,7 @@ xfs_ail_min( */ STATIC xfs_log_item_t * xfs_ail_next( - xfs_ail_t *ailp, + struct xfs_ail *ailp, xfs_log_item_t *lip) /* ARGSUSED */ { @@ -617,7 +693,7 @@ xfs_ail_next( */ STATIC void xfs_ail_check( - xfs_ail_t *ailp, + struct xfs_ail *ailp, xfs_log_item_t *lip) { xfs_log_item_t *prev_lip; diff --git a/fs/xfs/xfs_trans_buf.c b/fs/xfs/xfs_trans_buf.c index 4e855b5ced6..8ee2f8c8b0a 100644 --- a/fs/xfs/xfs_trans_buf.c +++ b/fs/xfs/xfs_trans_buf.c @@ -527,9 +527,8 @@ xfs_trans_brelse(xfs_trans_t *tp, lip = XFS_BUF_FSPRIVATE(bp, xfs_log_item_t *); if (lip->li_type == XFS_LI_BUF) { bip = XFS_BUF_FSPRIVATE(bp,xfs_buf_log_item_t*); - xfs_trans_unlocked_item( - bip->bli_item.li_mountp, - lip); + xfs_trans_unlocked_item(bip->bli_item.li_ailp, + lip); } } xfs_buf_relse(bp); @@ -626,7 +625,7 @@ xfs_trans_brelse(xfs_trans_t *tp, * tell the AIL that the buffer is being unlocked. */ if (bip != NULL) { - xfs_trans_unlocked_item(bip->bli_item.li_mountp, + xfs_trans_unlocked_item(bip->bli_item.li_ailp, (xfs_log_item_t*)bip); } diff --git a/fs/xfs/xfs_trans_inode.c b/fs/xfs/xfs_trans_inode.c index 2a1c0f071f9..23d276af2e0 100644 --- a/fs/xfs/xfs_trans_inode.c +++ b/fs/xfs/xfs_trans_inode.c @@ -85,7 +85,6 @@ xfs_trans_iget( { int error; xfs_inode_t *ip; - xfs_inode_log_item_t *iip; /* * If the transaction pointer is NULL, just call the normal @@ -138,34 +137,7 @@ xfs_trans_iget( } ASSERT(ip != NULL); - /* - * Get a log_item_desc to point at the new item. - */ - if (ip->i_itemp == NULL) - xfs_inode_item_init(ip, mp); - iip = ip->i_itemp; - (void) xfs_trans_add_item(tp, (xfs_log_item_t *)(iip)); - - xfs_trans_inode_broot_debug(ip); - - /* - * If the IO lock has been acquired, mark that in - * the inode log item so we'll know to unlock it - * when the transaction commits. - */ - ASSERT(iip->ili_flags == 0); - if (lock_flags & XFS_IOLOCK_EXCL) { - iip->ili_flags |= XFS_ILI_IOLOCKED_EXCL; - } else if (lock_flags & XFS_IOLOCK_SHARED) { - iip->ili_flags |= XFS_ILI_IOLOCKED_SHARED; - } - - /* - * Initialize i_transp so we can find it with xfs_inode_incore() - * above. - */ - ip->i_transp = tp; - + xfs_trans_ijoin(tp, ip, lock_flags); *ipp = ip; return 0; } diff --git a/fs/xfs/xfs_trans_item.c b/fs/xfs/xfs_trans_item.c index 3c666e8317f..e110bf57d7f 100644 --- a/fs/xfs/xfs_trans_item.c +++ b/fs/xfs/xfs_trans_item.c @@ -22,6 +22,14 @@ #include "xfs_inum.h" #include "xfs_trans.h" #include "xfs_trans_priv.h" +/* XXX: from here down needed until struct xfs_trans has it's own ailp */ +#include "xfs_bit.h" +#include "xfs_buf_item.h" +#include "xfs_sb.h" +#include "xfs_ag.h" +#include "xfs_dir2.h" +#include "xfs_dmapi.h" +#include "xfs_mount.h" STATIC int xfs_trans_unlock_chunk(xfs_log_item_chunk_t *, int, int, xfs_lsn_t); @@ -79,6 +87,7 @@ xfs_trans_add_item(xfs_trans_t *tp, xfs_log_item_t *lip) lidp->lid_size = 0; lip->li_desc = lidp; lip->li_mountp = tp->t_mountp; + lip->li_ailp = tp->t_mountp->m_ail; return lidp; } @@ -120,6 +129,7 @@ xfs_trans_add_item(xfs_trans_t *tp, xfs_log_item_t *lip) lidp->lid_size = 0; lip->li_desc = lidp; lip->li_mountp = tp->t_mountp; + lip->li_ailp = tp->t_mountp->m_ail; return lidp; } diff --git a/fs/xfs/xfs_trans_priv.h b/fs/xfs/xfs_trans_priv.h index 3c748c456ed..73e2ad39743 100644 --- a/fs/xfs/xfs_trans_priv.h +++ b/fs/xfs/xfs_trans_priv.h @@ -44,25 +44,93 @@ xfs_log_busy_slot_t *xfs_trans_add_busy(xfs_trans_t *tp, xfs_extlen_t idx); /* - * From xfs_trans_ail.c + * AIL traversal cursor. + * + * Rather than using a generation number for detecting changes in the ail, use + * a cursor that is protected by the ail lock. The aild cursor exists in the + * struct xfs_ail, but other traversals can declare it on the stack and link it + * to the ail list. + * + * When an object is deleted from or moved int the AIL, the cursor list is + * searched to see if the object is a designated cursor item. If it is, it is + * deleted from the cursor so that the next time the cursor is used traversal + * will return to the start. + * + * This means a traversal colliding with a removal will cause a restart of the + * list scan, rather than any insertion or deletion anywhere in the list. The + * low bit of the item pointer is set if the cursor has been invalidated so + * that we can tell the difference between invalidation and reaching the end + * of the list to trigger traversal restarts. */ -void xfs_trans_update_ail(struct xfs_mount *mp, - struct xfs_log_item *lip, xfs_lsn_t lsn) - __releases(mp->m_ail_lock); -void xfs_trans_delete_ail(struct xfs_mount *mp, - struct xfs_log_item *lip) - __releases(mp->m_ail_lock); -struct xfs_log_item *xfs_trans_first_ail(struct xfs_mount *, int *); -struct xfs_log_item *xfs_trans_next_ail(struct xfs_mount *, - struct xfs_log_item *, int *, int *); +struct xfs_ail_cursor { + struct xfs_ail_cursor *next; + struct xfs_log_item *item; +}; +/* + * Private AIL structures. + * + * Eventually we need to drive the locking in here as well. + */ +struct xfs_ail { + struct xfs_mount *xa_mount; + struct list_head xa_ail; + uint xa_gen; + struct task_struct *xa_task; + xfs_lsn_t xa_target; + struct xfs_ail_cursor xa_cursors; + spinlock_t xa_lock; +}; /* - * AIL push thread support + * From xfs_trans_ail.c */ -long xfsaild_push(struct xfs_mount *, xfs_lsn_t *); -void xfsaild_wakeup(struct xfs_mount *, xfs_lsn_t); -int xfsaild_start(struct xfs_mount *); -void xfsaild_stop(struct xfs_mount *); +void xfs_trans_ail_update(struct xfs_ail *ailp, + struct xfs_log_item *lip, xfs_lsn_t lsn) + __releases(ailp->xa_lock); +void xfs_trans_ail_delete(struct xfs_ail *ailp, + struct xfs_log_item *lip) + __releases(ailp->xa_lock); +void xfs_trans_ail_push(struct xfs_ail *, xfs_lsn_t); +void xfs_trans_unlocked_item(struct xfs_ail *, + xfs_log_item_t *); + +xfs_lsn_t xfs_trans_ail_tail(struct xfs_ail *ailp); + +struct xfs_log_item *xfs_trans_ail_cursor_first(struct xfs_ail *ailp, + struct xfs_ail_cursor *cur, + xfs_lsn_t lsn); +struct xfs_log_item *xfs_trans_ail_cursor_next(struct xfs_ail *ailp, + struct xfs_ail_cursor *cur); +void xfs_trans_ail_cursor_done(struct xfs_ail *ailp, + struct xfs_ail_cursor *cur); + +long xfsaild_push(struct xfs_ail *, xfs_lsn_t *); +void xfsaild_wakeup(struct xfs_ail *, xfs_lsn_t); +int xfsaild_start(struct xfs_ail *); +void xfsaild_stop(struct xfs_ail *); +#if BITS_PER_LONG != 64 +static inline void +xfs_trans_ail_copy_lsn( + struct xfs_ail *ailp, + xfs_lsn_t *dst, + xfs_lsn_t *src) +{ + ASSERT(sizeof(xfs_lsn_t) == 8); /* don't lock if it shrinks */ + spin_lock(&ailp->xa_lock); + *dst = *src; + spin_unlock(&ailp->xa_lock); +} +#else +static inline void +xfs_trans_ail_copy_lsn( + struct xfs_ail *ailp, + xfs_lsn_t *dst, + xfs_lsn_t *src) +{ + ASSERT(sizeof(xfs_lsn_t) == 8); + *dst = *src; +} +#endif #endif /* __XFS_TRANS_PRIV_H__ */ diff --git a/fs/xfs/xfs_utils.c b/fs/xfs/xfs_utils.c index 35d4d414bcc..fcc2285d03e 100644 --- a/fs/xfs/xfs_utils.c +++ b/fs/xfs/xfs_utils.c @@ -172,6 +172,12 @@ xfs_dir_ialloc( *ipp = NULL; return code; } + + /* + * transaction commit worked ok so we can drop the extra ticket + * reference that we gained in xfs_trans_dup() + */ + xfs_log_ticket_put(tp->t_ticket); code = xfs_trans_reserve(tp, 0, log_res, 0, XFS_TRANS_PERM_LOG_RES, log_count); /* @@ -268,9 +274,9 @@ xfs_bump_ino_vers2( xfs_mount_t *mp; ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL)); - ASSERT(ip->i_d.di_version == XFS_DINODE_VERSION_1); + ASSERT(ip->i_d.di_version == 1); - ip->i_d.di_version = XFS_DINODE_VERSION_2; + ip->i_d.di_version = 2; ip->i_d.di_onlink = 0; memset(&(ip->i_d.di_pad[0]), 0, sizeof(ip->i_d.di_pad)); mp = tp->t_mountp; @@ -302,7 +308,7 @@ xfs_bumplink( ASSERT(ip->i_d.di_nlink > 0); ip->i_d.di_nlink++; inc_nlink(VFS_I(ip)); - if ((ip->i_d.di_version == XFS_DINODE_VERSION_1) && + if ((ip->i_d.di_version == 1) && (ip->i_d.di_nlink > XFS_MAXLINK_1)) { /* * The inode has increased its number of links beyond diff --git a/fs/xfs/xfs_vfsops.c b/fs/xfs/xfs_vfsops.c deleted file mode 100644 index 439dd3939dd..00000000000 --- a/fs/xfs/xfs_vfsops.c +++ /dev/null @@ -1,757 +0,0 @@ -/* - * Copyright (c) 2000-2005 Silicon Graphics, Inc. - * All Rights Reserved. - * - * This program is free software; you can redistribute it and/or - * modify it under the terms of the GNU General Public License as - * published by the Free Software Foundation. - * - * This program is distributed in the hope that it would be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write the Free Software Foundation, - * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA - */ -#include "xfs.h" -#include "xfs_fs.h" -#include "xfs_types.h" -#include "xfs_bit.h" -#include "xfs_log.h" -#include "xfs_inum.h" -#include "xfs_trans.h" -#include "xfs_sb.h" -#include "xfs_ag.h" -#include "xfs_dir2.h" -#include "xfs_dmapi.h" -#include "xfs_mount.h" -#include "xfs_da_btree.h" -#include "xfs_bmap_btree.h" -#include "xfs_ialloc_btree.h" -#include "xfs_alloc_btree.h" -#include "xfs_dir2_sf.h" -#include "xfs_attr_sf.h" -#include "xfs_dinode.h" -#include "xfs_inode.h" -#include "xfs_inode_item.h" -#include "xfs_btree.h" -#include "xfs_alloc.h" -#include "xfs_ialloc.h" -#include "xfs_quota.h" -#include "xfs_error.h" -#include "xfs_bmap.h" -#include "xfs_rw.h" -#include "xfs_buf_item.h" -#include "xfs_log_priv.h" -#include "xfs_dir2_trace.h" -#include "xfs_extfree_item.h" -#include "xfs_acl.h" -#include "xfs_attr.h" -#include "xfs_clnt.h" -#include "xfs_mru_cache.h" -#include "xfs_filestream.h" -#include "xfs_fsops.h" -#include "xfs_vnodeops.h" -#include "xfs_vfsops.h" -#include "xfs_utils.h" - - -STATIC void -xfs_quiesce_fs( - xfs_mount_t *mp) -{ - int count = 0, pincount; - - xfs_flush_buftarg(mp->m_ddev_targp, 0); - xfs_finish_reclaim_all(mp, 0); - - /* This loop must run at least twice. - * The first instance of the loop will flush - * most meta data but that will generate more - * meta data (typically directory updates). - * Which then must be flushed and logged before - * we can write the unmount record. - */ - do { - xfs_syncsub(mp, SYNC_INODE_QUIESCE, NULL); - pincount = xfs_flush_buftarg(mp->m_ddev_targp, 1); - if (!pincount) { - delay(50); - count++; - } - } while (count < 2); -} - -/* - * Second stage of a quiesce. The data is already synced, now we have to take - * care of the metadata. New transactions are already blocked, so we need to - * wait for any remaining transactions to drain out before proceding. - */ -void -xfs_attr_quiesce( - xfs_mount_t *mp) -{ - int error = 0; - - /* wait for all modifications to complete */ - while (atomic_read(&mp->m_active_trans) > 0) - delay(100); - - /* flush inodes and push all remaining buffers out to disk */ - xfs_quiesce_fs(mp); - - ASSERT_ALWAYS(atomic_read(&mp->m_active_trans) == 0); - - /* Push the superblock and write an unmount record */ - error = xfs_log_sbcount(mp, 1); - if (error) - xfs_fs_cmn_err(CE_WARN, mp, - "xfs_attr_quiesce: failed to log sb changes. " - "Frozen image may not be consistent."); - xfs_log_unmount_write(mp); - xfs_unmountfs_writesb(mp); -} - -/* - * xfs_unmount_flush implements a set of flush operation on special - * inodes, which are needed as a separate set of operations so that - * they can be called as part of relocation process. - */ -int -xfs_unmount_flush( - xfs_mount_t *mp, /* Mount structure we are getting - rid of. */ - int relocation) /* Called from vfs relocation. */ -{ - xfs_inode_t *rip = mp->m_rootip; - xfs_inode_t *rbmip; - xfs_inode_t *rsumip = NULL; - int error; - - xfs_ilock(rip, XFS_ILOCK_EXCL | XFS_ILOCK_PARENT); - xfs_iflock(rip); - - /* - * Flush out the real time inodes. - */ - if ((rbmip = mp->m_rbmip) != NULL) { - xfs_ilock(rbmip, XFS_ILOCK_EXCL); - xfs_iflock(rbmip); - error = xfs_iflush(rbmip, XFS_IFLUSH_SYNC); - xfs_iunlock(rbmip, XFS_ILOCK_EXCL); - - if (error == EFSCORRUPTED) - goto fscorrupt_out; - - ASSERT(vn_count(VFS_I(rbmip)) == 1); - - rsumip = mp->m_rsumip; - xfs_ilock(rsumip, XFS_ILOCK_EXCL); - xfs_iflock(rsumip); - error = xfs_iflush(rsumip, XFS_IFLUSH_SYNC); - xfs_iunlock(rsumip, XFS_ILOCK_EXCL); - - if (error == EFSCORRUPTED) - goto fscorrupt_out; - - ASSERT(vn_count(VFS_I(rsumip)) == 1); - } - - /* - * Synchronously flush root inode to disk - */ - error = xfs_iflush(rip, XFS_IFLUSH_SYNC); - if (error == EFSCORRUPTED) - goto fscorrupt_out2; - - if (vn_count(VFS_I(rip)) != 1 && !relocation) { - xfs_iunlock(rip, XFS_ILOCK_EXCL); - return XFS_ERROR(EBUSY); - } - - /* - * Release dquot that rootinode, rbmino and rsumino might be holding, - * flush and purge the quota inodes. - */ - error = XFS_QM_UNMOUNT(mp); - if (error == EFSCORRUPTED) - goto fscorrupt_out2; - - if (rbmip) { - IRELE(rbmip); - IRELE(rsumip); - } - - xfs_iunlock(rip, XFS_ILOCK_EXCL); - return 0; - -fscorrupt_out: - xfs_ifunlock(rip); - -fscorrupt_out2: - xfs_iunlock(rip, XFS_ILOCK_EXCL); - - return XFS_ERROR(EFSCORRUPTED); -} - -/* - * xfs_sync flushes any pending I/O to file system vfsp. - * - * This routine is called by vfs_sync() to make sure that things make it - * out to disk eventually, on sync() system calls to flush out everything, - * and when the file system is unmounted. For the vfs_sync() case, all - * we really need to do is sync out the log to make all of our meta-data - * updates permanent (except for timestamps). For calls from pflushd(), - * dirty pages are kept moving by calling pdflush() on the inodes - * containing them. We also flush the inodes that we can lock without - * sleeping and the superblock if we can lock it without sleeping from - * vfs_sync() so that items at the tail of the log are always moving out. - * - * Flags: - * SYNC_BDFLUSH - We're being called from vfs_sync() so we don't want - * to sleep if we can help it. All we really need - * to do is ensure that the log is synced at least - * periodically. We also push the inodes and - * superblock if we can lock them without sleeping - * and they are not pinned. - * SYNC_ATTR - We need to flush the inodes. If SYNC_BDFLUSH is not - * set, then we really want to lock each inode and flush - * it. - * SYNC_WAIT - All the flushes that take place in this call should - * be synchronous. - * SYNC_DELWRI - This tells us to push dirty pages associated with - * inodes. SYNC_WAIT and SYNC_BDFLUSH are used to - * determine if they should be flushed sync, async, or - * delwri. - * SYNC_CLOSE - This flag is passed when the system is being - * unmounted. We should sync and invalidate everything. - * SYNC_FSDATA - This indicates that the caller would like to make - * sure the superblock is safe on disk. We can ensure - * this by simply making sure the log gets flushed - * if SYNC_BDFLUSH is set, and by actually writing it - * out otherwise. - * SYNC_IOWAIT - The caller wants us to wait for all data I/O to complete - * before we return (including direct I/O). Forms the drain - * side of the write barrier needed to safely quiesce the - * filesystem. - * - */ -int -xfs_sync( - xfs_mount_t *mp, - int flags) -{ - int error; - - /* - * Get the Quota Manager to flush the dquots. - * - * If XFS quota support is not enabled or this filesystem - * instance does not use quotas XFS_QM_DQSYNC will always - * return zero. - */ - error = XFS_QM_DQSYNC(mp, flags); - if (error) { - /* - * If we got an IO error, we will be shutting down. - * So, there's nothing more for us to do here. - */ - ASSERT(error != EIO || XFS_FORCED_SHUTDOWN(mp)); - if (XFS_FORCED_SHUTDOWN(mp)) - return XFS_ERROR(error); - } - - if (flags & SYNC_IOWAIT) - xfs_filestream_flush(mp); - - return xfs_syncsub(mp, flags, NULL); -} - -/* - * xfs sync routine for internal use - * - * This routine supports all of the flags defined for the generic vfs_sync - * interface as explained above under xfs_sync. - * - */ -int -xfs_sync_inodes( - xfs_mount_t *mp, - int flags, - int *bypassed) -{ - xfs_inode_t *ip = NULL; - struct inode *vp = NULL; - int error; - int last_error; - uint64_t fflag; - uint lock_flags; - uint base_lock_flags; - boolean_t mount_locked; - boolean_t vnode_refed; - int preempt; - xfs_iptr_t *ipointer; -#ifdef DEBUG - boolean_t ipointer_in = B_FALSE; - -#define IPOINTER_SET ipointer_in = B_TRUE -#define IPOINTER_CLR ipointer_in = B_FALSE -#else -#define IPOINTER_SET -#define IPOINTER_CLR -#endif - - -/* Insert a marker record into the inode list after inode ip. The list - * must be locked when this is called. After the call the list will no - * longer be locked. - */ -#define IPOINTER_INSERT(ip, mp) { \ - ASSERT(ipointer_in == B_FALSE); \ - ipointer->ip_mnext = ip->i_mnext; \ - ipointer->ip_mprev = ip; \ - ip->i_mnext = (xfs_inode_t *)ipointer; \ - ipointer->ip_mnext->i_mprev = (xfs_inode_t *)ipointer; \ - preempt = 0; \ - XFS_MOUNT_IUNLOCK(mp); \ - mount_locked = B_FALSE; \ - IPOINTER_SET; \ - } - -/* Remove the marker from the inode list. If the marker was the only item - * in the list then there are no remaining inodes and we should zero out - * the whole list. If we are the current head of the list then move the head - * past us. - */ -#define IPOINTER_REMOVE(ip, mp) { \ - ASSERT(ipointer_in == B_TRUE); \ - if (ipointer->ip_mnext != (xfs_inode_t *)ipointer) { \ - ip = ipointer->ip_mnext; \ - ip->i_mprev = ipointer->ip_mprev; \ - ipointer->ip_mprev->i_mnext = ip; \ - if (mp->m_inodes == (xfs_inode_t *)ipointer) { \ - mp->m_inodes = ip; \ - } \ - } else { \ - ASSERT(mp->m_inodes == (xfs_inode_t *)ipointer); \ - mp->m_inodes = NULL; \ - ip = NULL; \ - } \ - IPOINTER_CLR; \ - } - -#define XFS_PREEMPT_MASK 0x7f - - ASSERT(!(flags & SYNC_BDFLUSH)); - - if (bypassed) - *bypassed = 0; - if (mp->m_flags & XFS_MOUNT_RDONLY) - return 0; - error = 0; - last_error = 0; - preempt = 0; - - /* Allocate a reference marker */ - ipointer = (xfs_iptr_t *)kmem_zalloc(sizeof(xfs_iptr_t), KM_SLEEP); - - fflag = XFS_B_ASYNC; /* default is don't wait */ - if (flags & SYNC_DELWRI) - fflag = XFS_B_DELWRI; - if (flags & SYNC_WAIT) - fflag = 0; /* synchronous overrides all */ - - base_lock_flags = XFS_ILOCK_SHARED; - if (flags & (SYNC_DELWRI | SYNC_CLOSE)) { - /* - * We need the I/O lock if we're going to call any of - * the flush/inval routines. - */ - base_lock_flags |= XFS_IOLOCK_SHARED; - } - - XFS_MOUNT_ILOCK(mp); - - ip = mp->m_inodes; - - mount_locked = B_TRUE; - vnode_refed = B_FALSE; - - IPOINTER_CLR; - - do { - ASSERT(ipointer_in == B_FALSE); - ASSERT(vnode_refed == B_FALSE); - - lock_flags = base_lock_flags; - - /* - * There were no inodes in the list, just break out - * of the loop. - */ - if (ip == NULL) { - break; - } - - /* - * We found another sync thread marker - skip it - */ - if (ip->i_mount == NULL) { - ip = ip->i_mnext; - continue; - } - - vp = VFS_I(ip); - - /* - * If the vnode is gone then this is being torn down, - * call reclaim if it is flushed, else let regular flush - * code deal with it later in the loop. - */ - - if (vp == NULL) { - /* Skip ones already in reclaim */ - if (ip->i_flags & XFS_IRECLAIM) { - ip = ip->i_mnext; - continue; - } - if (xfs_ilock_nowait(ip, XFS_ILOCK_EXCL) == 0) { - ip = ip->i_mnext; - } else if ((xfs_ipincount(ip) == 0) && - xfs_iflock_nowait(ip)) { - IPOINTER_INSERT(ip, mp); - - xfs_finish_reclaim(ip, 1, - XFS_IFLUSH_DELWRI_ELSE_ASYNC); - - XFS_MOUNT_ILOCK(mp); - mount_locked = B_TRUE; - IPOINTER_REMOVE(ip, mp); - } else { - xfs_iunlock(ip, XFS_ILOCK_EXCL); - ip = ip->i_mnext; - } - continue; - } - - if (VN_BAD(vp)) { - ip = ip->i_mnext; - continue; - } - - if (XFS_FORCED_SHUTDOWN(mp) && !(flags & SYNC_CLOSE)) { - XFS_MOUNT_IUNLOCK(mp); - kmem_free(ipointer); - return 0; - } - - /* - * Try to lock without sleeping. We're out of order with - * the inode list lock here, so if we fail we need to drop - * the mount lock and try again. If we're called from - * bdflush() here, then don't bother. - * - * The inode lock here actually coordinates with the - * almost spurious inode lock in xfs_ireclaim() to prevent - * the vnode we handle here without a reference from - * being freed while we reference it. If we lock the inode - * while it's on the mount list here, then the spurious inode - * lock in xfs_ireclaim() after the inode is pulled from - * the mount list will sleep until we release it here. - * This keeps the vnode from being freed while we reference - * it. - */ - if (xfs_ilock_nowait(ip, lock_flags) == 0) { - if (vp == NULL) { - ip = ip->i_mnext; - continue; - } - - vp = vn_grab(vp); - if (vp == NULL) { - ip = ip->i_mnext; - continue; - } - - IPOINTER_INSERT(ip, mp); - xfs_ilock(ip, lock_flags); - - ASSERT(vp == VFS_I(ip)); - ASSERT(ip->i_mount == mp); - - vnode_refed = B_TRUE; - } - - /* From here on in the loop we may have a marker record - * in the inode list. - */ - - /* - * If we have to flush data or wait for I/O completion - * we need to drop the ilock that we currently hold. - * If we need to drop the lock, insert a marker if we - * have not already done so. - */ - if ((flags & (SYNC_CLOSE|SYNC_IOWAIT)) || - ((flags & SYNC_DELWRI) && VN_DIRTY(vp))) { - if (mount_locked) { - IPOINTER_INSERT(ip, mp); - } - xfs_iunlock(ip, XFS_ILOCK_SHARED); - - if (flags & SYNC_CLOSE) { - /* Shutdown case. Flush and invalidate. */ - if (XFS_FORCED_SHUTDOWN(mp)) - xfs_tosspages(ip, 0, -1, - FI_REMAPF); - else - error = xfs_flushinval_pages(ip, - 0, -1, FI_REMAPF); - } else if ((flags & SYNC_DELWRI) && VN_DIRTY(vp)) { - error = xfs_flush_pages(ip, 0, - -1, fflag, FI_NONE); - } - - /* - * When freezing, we need to wait ensure all I/O (including direct - * I/O) is complete to ensure no further data modification can take - * place after this point - */ - if (flags & SYNC_IOWAIT) - vn_iowait(ip); - - xfs_ilock(ip, XFS_ILOCK_SHARED); - } - - if ((flags & SYNC_ATTR) && - (ip->i_update_core || - (ip->i_itemp && ip->i_itemp->ili_format.ilf_fields))) { - if (mount_locked) - IPOINTER_INSERT(ip, mp); - - if (flags & SYNC_WAIT) { - xfs_iflock(ip); - error = xfs_iflush(ip, XFS_IFLUSH_SYNC); - - /* - * If we can't acquire the flush lock, then the inode - * is already being flushed so don't bother waiting. - * - * If we can lock it then do a delwri flush so we can - * combine multiple inode flushes in each disk write. - */ - } else if (xfs_iflock_nowait(ip)) { - error = xfs_iflush(ip, XFS_IFLUSH_DELWRI); - } else if (bypassed) { - (*bypassed)++; - } - } - - if (lock_flags != 0) { - xfs_iunlock(ip, lock_flags); - } - - if (vnode_refed) { - /* - * If we had to take a reference on the vnode - * above, then wait until after we've unlocked - * the inode to release the reference. This is - * because we can be already holding the inode - * lock when IRELE() calls xfs_inactive(). - * - * Make sure to drop the mount lock before calling - * IRELE() so that we don't trip over ourselves if - * we have to go for the mount lock again in the - * inactive code. - */ - if (mount_locked) { - IPOINTER_INSERT(ip, mp); - } - - IRELE(ip); - - vnode_refed = B_FALSE; - } - - if (error) { - last_error = error; - } - - /* - * bail out if the filesystem is corrupted. - */ - if (error == EFSCORRUPTED) { - if (!mount_locked) { - XFS_MOUNT_ILOCK(mp); - IPOINTER_REMOVE(ip, mp); - } - XFS_MOUNT_IUNLOCK(mp); - ASSERT(ipointer_in == B_FALSE); - kmem_free(ipointer); - return XFS_ERROR(error); - } - - /* Let other threads have a chance at the mount lock - * if we have looped many times without dropping the - * lock. - */ - if ((++preempt & XFS_PREEMPT_MASK) == 0) { - if (mount_locked) { - IPOINTER_INSERT(ip, mp); - } - } - - if (mount_locked == B_FALSE) { - XFS_MOUNT_ILOCK(mp); - mount_locked = B_TRUE; - IPOINTER_REMOVE(ip, mp); - continue; - } - - ASSERT(ipointer_in == B_FALSE); - ip = ip->i_mnext; - - } while (ip != mp->m_inodes); - - XFS_MOUNT_IUNLOCK(mp); - - ASSERT(ipointer_in == B_FALSE); - - kmem_free(ipointer); - return XFS_ERROR(last_error); -} - -/* - * xfs sync routine for internal use - * - * This routine supports all of the flags defined for the generic vfs_sync - * interface as explained above under xfs_sync. - * - */ -int -xfs_syncsub( - xfs_mount_t *mp, - int flags, - int *bypassed) -{ - int error = 0; - int last_error = 0; - uint log_flags = XFS_LOG_FORCE; - xfs_buf_t *bp; - xfs_buf_log_item_t *bip; - - /* - * Sync out the log. This ensures that the log is periodically - * flushed even if there is not enough activity to fill it up. - */ - if (flags & SYNC_WAIT) - log_flags |= XFS_LOG_SYNC; - - xfs_log_force(mp, (xfs_lsn_t)0, log_flags); - - if (flags & (SYNC_ATTR|SYNC_DELWRI)) { - if (flags & SYNC_BDFLUSH) - xfs_finish_reclaim_all(mp, 1); - else - error = xfs_sync_inodes(mp, flags, bypassed); - } - - /* - * Flushing out dirty data above probably generated more - * log activity, so if this isn't vfs_sync() then flush - * the log again. - */ - if (flags & SYNC_DELWRI) { - xfs_log_force(mp, (xfs_lsn_t)0, log_flags); - } - - if (flags & SYNC_FSDATA) { - /* - * If this is vfs_sync() then only sync the superblock - * if we can lock it without sleeping and it is not pinned. - */ - if (flags & SYNC_BDFLUSH) { - bp = xfs_getsb(mp, XFS_BUF_TRYLOCK); - if (bp != NULL) { - bip = XFS_BUF_FSPRIVATE(bp,xfs_buf_log_item_t*); - if ((bip != NULL) && - xfs_buf_item_dirty(bip)) { - if (!(XFS_BUF_ISPINNED(bp))) { - XFS_BUF_ASYNC(bp); - error = xfs_bwrite(mp, bp); - } else { - xfs_buf_relse(bp); - } - } else { - xfs_buf_relse(bp); - } - } - } else { - bp = xfs_getsb(mp, 0); - /* - * If the buffer is pinned then push on the log so - * we won't get stuck waiting in the write for - * someone, maybe ourselves, to flush the log. - * Even though we just pushed the log above, we - * did not have the superblock buffer locked at - * that point so it can become pinned in between - * there and here. - */ - if (XFS_BUF_ISPINNED(bp)) - xfs_log_force(mp, (xfs_lsn_t)0, XFS_LOG_FORCE); - if (flags & SYNC_WAIT) - XFS_BUF_UNASYNC(bp); - else - XFS_BUF_ASYNC(bp); - error = xfs_bwrite(mp, bp); - } - if (error) { - last_error = error; - } - } - - /* - * Now check to see if the log needs a "dummy" transaction. - */ - if (!(flags & SYNC_REMOUNT) && xfs_log_need_covered(mp)) { - xfs_trans_t *tp; - xfs_inode_t *ip; - - /* - * Put a dummy transaction in the log to tell - * recovery that all others are OK. - */ - tp = xfs_trans_alloc(mp, XFS_TRANS_DUMMY1); - if ((error = xfs_trans_reserve(tp, 0, - XFS_ICHANGE_LOG_RES(mp), - 0, 0, 0))) { - xfs_trans_cancel(tp, 0); - return error; - } - - ip = mp->m_rootip; - xfs_ilock(ip, XFS_ILOCK_EXCL); - - xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL); - xfs_trans_ihold(tp, ip); - xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); - error = xfs_trans_commit(tp, 0); - xfs_iunlock(ip, XFS_ILOCK_EXCL); - xfs_log_force(mp, (xfs_lsn_t)0, log_flags); - } - - /* - * When shutting down, we need to insure that the AIL is pushed - * to disk or the filesystem can appear corrupt from the PROM. - */ - if ((flags & (SYNC_CLOSE|SYNC_WAIT)) == (SYNC_CLOSE|SYNC_WAIT)) { - XFS_bflush(mp->m_ddev_targp); - if (mp->m_rtdev_targp) { - XFS_bflush(mp->m_rtdev_targp); - } - } - - return XFS_ERROR(last_error); -} diff --git a/fs/xfs/xfs_vfsops.h b/fs/xfs/xfs_vfsops.h deleted file mode 100644 index a74b05087da..00000000000 --- a/fs/xfs/xfs_vfsops.h +++ /dev/null @@ -1,16 +0,0 @@ -#ifndef _XFS_VFSOPS_H -#define _XFS_VFSOPS_H 1 - -struct cred; -struct xfs_fid; -struct inode; -struct kstatfs; -struct xfs_mount; -struct xfs_mount_args; - -int xfs_sync(struct xfs_mount *mp, int flags); -void xfs_do_force_shutdown(struct xfs_mount *mp, int flags, char *fname, - int lnnum); -void xfs_attr_quiesce(struct xfs_mount *mp); - -#endif /* _XFS_VFSOPS_H */ diff --git a/fs/xfs/xfs_vnodeops.c b/fs/xfs/xfs_vnodeops.c index 8b6812f66a1..f07bf8768c3 100644 --- a/fs/xfs/xfs_vnodeops.c +++ b/fs/xfs/xfs_vnodeops.c @@ -54,33 +54,10 @@ #include "xfs_vnodeops.h" int -xfs_open( - xfs_inode_t *ip) -{ - int mode; - - if (XFS_FORCED_SHUTDOWN(ip->i_mount)) - return XFS_ERROR(EIO); - - /* - * If it's a directory with any blocks, read-ahead block 0 - * as we're almost certain to have the next operation be a read there. - */ - if (S_ISDIR(ip->i_d.di_mode) && ip->i_d.di_nextents > 0) { - mode = xfs_ilock_map_shared(ip); - if (ip->i_d.di_nextents > 0) - (void)xfs_da_reada_buf(NULL, ip, 0, XFS_DATA_FORK); - xfs_iunlock(ip, mode); - } - return 0; -} - -int xfs_setattr( struct xfs_inode *ip, struct iattr *iattr, - int flags, - cred_t *credp) + int flags) { xfs_mount_t *mp = ip->i_mount; struct inode *inode = VFS_I(ip); @@ -93,7 +70,6 @@ xfs_setattr( gid_t gid=0, igid=0; int timeflags = 0; struct xfs_dquot *udqp, *gdqp, *olddquot1, *olddquot2; - int file_owner; int need_iolock = 1; xfs_itrace_entry(ip); @@ -104,6 +80,10 @@ xfs_setattr( if (XFS_FORCED_SHUTDOWN(mp)) return XFS_ERROR(EIO); + code = -inode_change_ok(inode, iattr); + if (code) + return code; + olddquot1 = olddquot2 = NULL; udqp = gdqp = NULL; @@ -181,62 +161,8 @@ xfs_setattr( xfs_ilock(ip, lock_flags); - /* boolean: are we the file owner? */ - file_owner = (current_fsuid() == ip->i_d.di_uid); - - /* - * Change various properties of a file. - * Only the owner or users with CAP_FOWNER - * capability may do these things. - */ - if (mask & (ATTR_MODE|ATTR_UID|ATTR_GID)) { - /* - * CAP_FOWNER overrides the following restrictions: - * - * The user ID of the calling process must be equal - * to the file owner ID, except in cases where the - * CAP_FSETID capability is applicable. - */ - if (!file_owner && !capable(CAP_FOWNER)) { - code = XFS_ERROR(EPERM); - goto error_return; - } - - /* - * CAP_FSETID overrides the following restrictions: - * - * The effective user ID of the calling process shall match - * the file owner when setting the set-user-ID and - * set-group-ID bits on that file. - * - * The effective group ID or one of the supplementary group - * IDs of the calling process shall match the group owner of - * the file when setting the set-group-ID bit on that file - */ - if (mask & ATTR_MODE) { - mode_t m = 0; - - if ((iattr->ia_mode & S_ISUID) && !file_owner) - m |= S_ISUID; - if ((iattr->ia_mode & S_ISGID) && - !in_group_p((gid_t)ip->i_d.di_gid)) - m |= S_ISGID; -#if 0 - /* Linux allows this, Irix doesn't. */ - if ((iattr->ia_mode & S_ISVTX) && !S_ISDIR(ip->i_d.di_mode)) - m |= S_ISVTX; -#endif - if (m && !capable(CAP_FSETID)) - iattr->ia_mode &= ~m; - } - } - /* * Change file ownership. Must be the owner or privileged. - * If the system was configured with the "restricted_chown" - * option, the owner is not permitted to give away the file, - * and can change the group id only to a group of which he - * or she is a member. */ if (mask & (ATTR_UID|ATTR_GID)) { /* @@ -251,23 +177,6 @@ xfs_setattr( uid = (mask & ATTR_UID) ? iattr->ia_uid : iuid; /* - * CAP_CHOWN overrides the following restrictions: - * - * If _POSIX_CHOWN_RESTRICTED is defined, this capability - * shall override the restriction that a process cannot - * change the user ID of a file it owns and the restriction - * that the group ID supplied to the chown() function - * shall be equal to either the group ID or one of the - * supplementary group IDs of the calling process. - */ - if (restricted_chown && - (iuid != uid || (igid != gid && - !in_group_p((gid_t)gid))) && - !capable(CAP_CHOWN)) { - code = XFS_ERROR(EPERM); - goto error_return; - } - /* * Do a quota reservation only if uid/gid is actually * going to change. */ @@ -304,36 +213,22 @@ xfs_setattr( code = XFS_ERROR(EINVAL); goto error_return; } + /* * Make sure that the dquots are attached to the inode. */ - if ((code = XFS_QM_DQATTACH(mp, ip, XFS_QMOPT_ILOCKED))) + code = XFS_QM_DQATTACH(mp, ip, XFS_QMOPT_ILOCKED); + if (code) goto error_return; - } - - /* - * Change file access or modified times. - */ - if (mask & (ATTR_ATIME|ATTR_MTIME)) { - if (!file_owner) { - if ((mask & (ATTR_MTIME_SET|ATTR_ATIME_SET)) && - !capable(CAP_FOWNER)) { - code = XFS_ERROR(EPERM); - goto error_return; - } - } - } - /* - * Now we can make the changes. Before we join the inode - * to the transaction, if ATTR_SIZE is set then take care of - * the part of the truncation that must be done without the - * inode lock. This needs to be done before joining the inode - * to the transaction, because the inode cannot be unlocked - * once it is a part of the transaction. - */ - if (mask & ATTR_SIZE) { - code = 0; + /* + * Now we can make the changes. Before we join the inode + * to the transaction, if ATTR_SIZE is set then take care of + * the part of the truncation that must be done without the + * inode lock. This needs to be done before joining the inode + * to the transaction, because the inode cannot be unlocked + * once it is a part of the transaction. + */ if (iattr->ia_size > ip->i_size) { /* * Do the first part of growing a file: zero any data @@ -366,7 +261,7 @@ xfs_setattr( } /* wait for all I/O to complete */ - vn_iowait(ip); + xfs_ioend_wait(ip); if (!code) code = xfs_itruncate_data(ip, iattr->ia_size); @@ -388,17 +283,10 @@ xfs_setattr( } commit_flags = XFS_TRANS_RELEASE_LOG_RES; xfs_ilock(ip, XFS_ILOCK_EXCL); - } - if (tp) { xfs_trans_ijoin(tp, ip, lock_flags); xfs_trans_ihold(tp, ip); - } - /* - * Truncate file. Must have write permission and not be a directory. - */ - if (mask & ATTR_SIZE) { /* * Only change the c/mtime if we are changing the size * or we are explicitly asked to change it. This handles @@ -438,28 +326,13 @@ xfs_setattr( */ xfs_iflags_set(ip, XFS_ITRUNCATED); } - } - - /* - * Change file access modes. - */ - if (mask & ATTR_MODE) { - ip->i_d.di_mode &= S_IFMT; - ip->i_d.di_mode |= iattr->ia_mode & ~S_IFMT; - - inode->i_mode &= S_IFMT; - inode->i_mode |= iattr->ia_mode & ~S_IFMT; - - xfs_trans_log_inode (tp, ip, XFS_ILOG_CORE); - timeflags |= XFS_ICHGTIME_CHG; + } else if (tp) { + xfs_trans_ijoin(tp, ip, lock_flags); + xfs_trans_ihold(tp, ip); } /* * Change file ownership. Must be the owner or privileged. - * If the system was configured with the "restricted_chown" - * option, the owner is not permitted to give away the file, - * and can change the group id only to a group of which he - * or she is a member. */ if (mask & (ATTR_UID|ATTR_GID)) { /* @@ -503,6 +376,24 @@ xfs_setattr( timeflags |= XFS_ICHGTIME_CHG; } + /* + * Change file access modes. + */ + if (mask & ATTR_MODE) { + umode_t mode = iattr->ia_mode; + + if (!in_group_p(inode->i_gid) && !capable(CAP_FSETID)) + mode &= ~S_ISGID; + + ip->i_d.di_mode &= S_IFMT; + ip->i_d.di_mode |= mode & ~S_IFMT; + + inode->i_mode &= S_IFMT; + inode->i_mode |= mode & ~S_IFMT; + + xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); + timeflags |= XFS_ICHGTIME_CHG; + } /* * Change file access or modified times. @@ -713,7 +604,7 @@ xfs_fsync( return XFS_ERROR(EIO); /* capture size updates in I/O completion before writing the inode. */ - error = filemap_fdatawait(VFS_I(ip)->i_mapping); + error = xfs_wait_on_pages(ip, 0, -1); if (error) return XFS_ERROR(error); @@ -1029,6 +920,12 @@ xfs_inactive_symlink_rmt( goto error0; } /* + * transaction commit worked ok so we can drop the extra ticket + * reference that we gained in xfs_trans_dup() + */ + xfs_log_ticket_put(tp->t_ticket); + + /* * Remove the memory for extent descriptions (just bookkeeping). */ if (ip->i_df.if_bytes) @@ -1625,8 +1522,6 @@ xfs_create( xfs_trans_set_sync(tp); } - dp->i_gen++; - /* * Attach the dquot(s) to the inodes and modify them incore. * These ids of the inode couldn't have changed since the new @@ -1993,13 +1888,6 @@ xfs_remove( } xfs_ichgtime(dp, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG); - /* - * Bump the in memory generation count on the parent - * directory so that other can know that it has changed. - */ - dp->i_gen++; - xfs_trans_log_inode(tp, dp, XFS_ILOG_CORE); - if (is_dir) { /* * Drop the link from ip's "..". @@ -2009,7 +1897,7 @@ xfs_remove( goto out_bmap_cancel; /* - * Drop the link from dp to ip. + * Drop the "." link from ip to self. */ error = xfs_droplink(tp, ip); if (error) @@ -2017,14 +1905,14 @@ xfs_remove( } else { /* * When removing a non-directory we need to log the parent - * inode here for the i_gen update. For a directory this is - * done implicitly by the xfs_droplink call for the ".." entry. + * inode here. For a directory this is done implicitly + * by the xfs_droplink call for the ".." entry. */ xfs_trans_log_inode(tp, dp, XFS_ILOG_CORE); } /* - * Drop the "." link from ip to self. + * Drop the link from dp to ip. */ error = xfs_droplink(tp, ip); if (error) @@ -2178,7 +2066,6 @@ xfs_link( if (error) goto abort_return; xfs_ichgtime(tdp, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG); - tdp->i_gen++; xfs_trans_log_inode(tp, tdp, XFS_ILOG_CORE); error = xfs_bumplink(tp, sip); @@ -2355,18 +2242,10 @@ xfs_mkdir( } xfs_ichgtime(dp, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG); - /* - * Bump the in memory version number of the parent directory - * so that other processes accessing it will recognize that - * the directory has changed. - */ - dp->i_gen++; - error = xfs_dir_init(tp, cdp, dp); if (error) goto error2; - cdp->i_gen = 1; error = xfs_bumplink(tp, dp); if (error) goto error2; @@ -2653,13 +2532,6 @@ xfs_symlink( xfs_trans_log_inode(tp, dp, XFS_ILOG_CORE); /* - * Bump the in memory version number of the parent directory - * so that other processes accessing it will recognize that - * the directory has changed. - */ - dp->i_gen++; - - /* * If this is a synchronous mount, make sure that the * symlink transaction goes to disk before returning to * the user. @@ -2809,7 +2681,7 @@ xfs_reclaim( return 0; } - vn_iowait(ip); + xfs_ioend_wait(ip); ASSERT(XFS_FORCED_SHUTDOWN(ip->i_mount) || ip->i_delayed_blks == 0); @@ -2833,122 +2705,10 @@ xfs_reclaim( if (!ip->i_update_core && (ip->i_itemp == NULL)) { xfs_ilock(ip, XFS_ILOCK_EXCL); xfs_iflock(ip); - return xfs_finish_reclaim(ip, 1, XFS_IFLUSH_DELWRI_ELSE_SYNC); - } else { - xfs_mount_t *mp = ip->i_mount; - - /* Protect sync and unpin from us */ - XFS_MOUNT_ILOCK(mp); - spin_lock(&ip->i_flags_lock); - __xfs_iflags_set(ip, XFS_IRECLAIMABLE); - VFS_I(ip)->i_private = NULL; - ip->i_vnode = NULL; - spin_unlock(&ip->i_flags_lock); - list_add_tail(&ip->i_reclaim, &mp->m_del_inodes); - XFS_MOUNT_IUNLOCK(mp); - } - return 0; -} - -int -xfs_finish_reclaim( - xfs_inode_t *ip, - int locked, - int sync_mode) -{ - xfs_perag_t *pag = xfs_get_perag(ip->i_mount, ip->i_ino); - struct inode *vp = VFS_I(ip); - - if (vp && VN_BAD(vp)) - goto reclaim; - - /* The hash lock here protects a thread in xfs_iget_core from - * racing with us on linking the inode back with a vnode. - * Once we have the XFS_IRECLAIM flag set it will not touch - * us. - */ - write_lock(&pag->pag_ici_lock); - spin_lock(&ip->i_flags_lock); - if (__xfs_iflags_test(ip, XFS_IRECLAIM) || - (!__xfs_iflags_test(ip, XFS_IRECLAIMABLE) && vp == NULL)) { - spin_unlock(&ip->i_flags_lock); - write_unlock(&pag->pag_ici_lock); - if (locked) { - xfs_ifunlock(ip); - xfs_iunlock(ip, XFS_ILOCK_EXCL); - } - return 1; - } - __xfs_iflags_set(ip, XFS_IRECLAIM); - spin_unlock(&ip->i_flags_lock); - write_unlock(&pag->pag_ici_lock); - xfs_put_perag(ip->i_mount, pag); - - /* - * If the inode is still dirty, then flush it out. If the inode - * is not in the AIL, then it will be OK to flush it delwri as - * long as xfs_iflush() does not keep any references to the inode. - * We leave that decision up to xfs_iflush() since it has the - * knowledge of whether it's OK to simply do a delwri flush of - * the inode or whether we need to wait until the inode is - * pulled from the AIL. - * We get the flush lock regardless, though, just to make sure - * we don't free it while it is being flushed. - */ - if (!locked) { - xfs_ilock(ip, XFS_ILOCK_EXCL); - xfs_iflock(ip); + xfs_iflags_set(ip, XFS_IRECLAIMABLE); + return xfs_reclaim_inode(ip, 1, XFS_IFLUSH_DELWRI_ELSE_SYNC); } - - /* - * In the case of a forced shutdown we rely on xfs_iflush() to - * wait for the inode to be unpinned before returning an error. - */ - if (xfs_iflush(ip, sync_mode) == 0) { - /* synchronize with xfs_iflush_done */ - xfs_iflock(ip); - xfs_ifunlock(ip); - } - - xfs_iunlock(ip, XFS_ILOCK_EXCL); - - reclaim: - xfs_ireclaim(ip); - return 0; -} - -int -xfs_finish_reclaim_all(xfs_mount_t *mp, int noblock) -{ - int purged; - xfs_inode_t *ip, *n; - int done = 0; - - while (!done) { - purged = 0; - XFS_MOUNT_ILOCK(mp); - list_for_each_entry_safe(ip, n, &mp->m_del_inodes, i_reclaim) { - if (noblock) { - if (xfs_ilock_nowait(ip, XFS_ILOCK_EXCL) == 0) - continue; - if (xfs_ipincount(ip) || - !xfs_iflock_nowait(ip)) { - xfs_iunlock(ip, XFS_ILOCK_EXCL); - continue; - } - } - XFS_MOUNT_IUNLOCK(mp); - if (xfs_finish_reclaim(ip, noblock, - XFS_IFLUSH_DELWRI_ELSE_ASYNC)) - delay(1); - purged = 1; - break; - } - - done = !purged; - } - - XFS_MOUNT_IUNLOCK(mp); + xfs_inode_set_reclaim_tag(ip); return 0; } @@ -3197,6 +2957,8 @@ xfs_zero_remaining_bytes( bp = xfs_buf_get_noaddr(mp->m_sb.sb_blocksize, XFS_IS_REALTIME_INODE(ip) ? mp->m_rtdev_targp : mp->m_ddev_targp); + if (!bp) + return XFS_ERROR(ENOMEM); for (offset = startoff; offset <= endoff; offset = lastoffset + 1) { offset_fsb = XFS_B_TO_FSBT(mp, offset); @@ -3312,7 +3074,8 @@ xfs_free_file_space( need_iolock = 0; if (need_iolock) { xfs_ilock(ip, XFS_IOLOCK_EXCL); - vn_iowait(ip); /* wait for the completion of any pending DIOs */ + /* wait for the completion of any pending DIOs */ + xfs_ioend_wait(ip); } rounding = max_t(uint, 1 << mp->m_sb.sb_blocklog, PAGE_CACHE_SIZE); @@ -3474,7 +3237,6 @@ xfs_change_file_space( int cmd, xfs_flock64_t *bf, xfs_off_t offset, - cred_t *credp, int attr_flags) { xfs_mount_t *mp = ip->i_mount; @@ -3562,7 +3324,7 @@ xfs_change_file_space( iattr.ia_valid = ATTR_SIZE; iattr.ia_size = startoffset; - error = xfs_setattr(ip, &iattr, attr_flags, credp); + error = xfs_setattr(ip, &iattr, attr_flags); if (error) return error; diff --git a/fs/xfs/xfs_vnodeops.h b/fs/xfs/xfs_vnodeops.h index e932a96bec5..76df328c61b 100644 --- a/fs/xfs/xfs_vnodeops.h +++ b/fs/xfs/xfs_vnodeops.h @@ -14,9 +14,7 @@ struct xfs_inode; struct xfs_iomap; -int xfs_open(struct xfs_inode *ip); -int xfs_setattr(struct xfs_inode *ip, struct iattr *vap, int flags, - struct cred *credp); +int xfs_setattr(struct xfs_inode *ip, struct iattr *vap, int flags); #define XFS_ATTR_DMI 0x01 /* invocation from a DMI function */ #define XFS_ATTR_NONBLOCK 0x02 /* return EAGAIN if operation would block */ #define XFS_ATTR_NOLOCK 0x04 /* Don't grab any conflicting locks */ @@ -28,24 +26,23 @@ int xfs_inactive(struct xfs_inode *ip); int xfs_lookup(struct xfs_inode *dp, struct xfs_name *name, struct xfs_inode **ipp, struct xfs_name *ci_name); int xfs_create(struct xfs_inode *dp, struct xfs_name *name, mode_t mode, - xfs_dev_t rdev, struct xfs_inode **ipp, struct cred *credp); + xfs_dev_t rdev, struct xfs_inode **ipp, cred_t *credp); int xfs_remove(struct xfs_inode *dp, struct xfs_name *name, struct xfs_inode *ip); int xfs_link(struct xfs_inode *tdp, struct xfs_inode *sip, struct xfs_name *target_name); int xfs_mkdir(struct xfs_inode *dp, struct xfs_name *dir_name, - mode_t mode, struct xfs_inode **ipp, struct cred *credp); + mode_t mode, struct xfs_inode **ipp, cred_t *credp); int xfs_readdir(struct xfs_inode *dp, void *dirent, size_t bufsize, xfs_off_t *offset, filldir_t filldir); int xfs_symlink(struct xfs_inode *dp, struct xfs_name *link_name, const char *target_path, mode_t mode, struct xfs_inode **ipp, - struct cred *credp); + cred_t *credp); int xfs_inode_flush(struct xfs_inode *ip, int flags); int xfs_set_dmattrs(struct xfs_inode *ip, u_int evmask, u_int16_t state); int xfs_reclaim(struct xfs_inode *ip); int xfs_change_file_space(struct xfs_inode *ip, int cmd, - xfs_flock64_t *bf, xfs_off_t offset, - struct cred *credp, int attr_flags); + xfs_flock64_t *bf, xfs_off_t offset, int attr_flags); int xfs_rename(struct xfs_inode *src_dp, struct xfs_name *src_name, struct xfs_inode *src_ip, struct xfs_inode *target_dp, struct xfs_name *target_name, struct xfs_inode *target_ip); @@ -56,8 +53,6 @@ int xfs_attr_set(struct xfs_inode *dp, const char *name, char *value, int xfs_attr_remove(struct xfs_inode *dp, const char *name, int flags); int xfs_attr_list(struct xfs_inode *dp, char *buffer, int bufsize, int flags, struct attrlist_cursor_kern *cursor); -int xfs_ioctl(struct xfs_inode *ip, struct file *filp, - int ioflags, unsigned int cmd, void __user *arg); ssize_t xfs_read(struct xfs_inode *ip, struct kiocb *iocb, const struct iovec *iovp, unsigned int segs, loff_t *offset, int ioflags); @@ -78,5 +73,6 @@ int xfs_flushinval_pages(struct xfs_inode *ip, xfs_off_t first, xfs_off_t last, int fiopt); int xfs_flush_pages(struct xfs_inode *ip, xfs_off_t first, xfs_off_t last, uint64_t flags, int fiopt); +int xfs_wait_on_pages(struct xfs_inode *ip, xfs_off_t first, xfs_off_t last); #endif /* _XFS_VNODEOPS_H */ |