diff options
Diffstat (limited to 'fs')
276 files changed, 8183 insertions, 7283 deletions
diff --git a/fs/9p/Makefile b/fs/9p/Makefile index 1a940ec7af6..91fba025fcb 100644 --- a/fs/9p/Makefile +++ b/fs/9p/Makefile @@ -8,6 +8,8 @@ obj-$(CONFIG_9P_FS) := 9p.o vfs_dir.o \ vfs_dentry.o \ v9fs.o \ - fid.o + fid.o \ + xattr.o \ + xattr_user.o 9p-$(CONFIG_9P_FSCACHE) += cache.o diff --git a/fs/9p/fid.c b/fs/9p/fid.c index 7317b39b281..35856368906 100644 --- a/fs/9p/fid.c +++ b/fs/9p/fid.c @@ -97,6 +97,34 @@ static struct p9_fid *v9fs_fid_find(struct dentry *dentry, u32 uid, int any) return ret; } +/* + * We need to hold v9ses->rename_sem as long as we hold references + * to returned path array. Array element contain pointers to + * dentry names. + */ +static int build_path_from_dentry(struct v9fs_session_info *v9ses, + struct dentry *dentry, char ***names) +{ + int n = 0, i; + char **wnames; + struct dentry *ds; + + for (ds = dentry; !IS_ROOT(ds); ds = ds->d_parent) + n++; + + wnames = kmalloc(sizeof(char *) * n, GFP_KERNEL); + if (!wnames) + goto err_out; + + for (ds = dentry, i = (n-1); i >= 0; i--, ds = ds->d_parent) + wnames[i] = (char *)ds->d_name.name; + + *names = wnames; + return n; +err_out: + return -ENOMEM; +} + /** * v9fs_fid_lookup - lookup for a fid, try to walk if not found * @dentry: dentry to look for fid in @@ -112,7 +140,7 @@ struct p9_fid *v9fs_fid_lookup(struct dentry *dentry) int i, n, l, clone, any, access; u32 uid; struct p9_fid *fid, *old_fid = NULL; - struct dentry *d, *ds; + struct dentry *ds; struct v9fs_session_info *v9ses; char **wnames, *uname; @@ -139,49 +167,62 @@ struct p9_fid *v9fs_fid_lookup(struct dentry *dentry) fid = v9fs_fid_find(dentry, uid, any); if (fid) return fid; - + /* + * we don't have a matching fid. To do a TWALK we need + * parent fid. We need to prevent rename when we want to + * look at the parent. + */ + down_read(&v9ses->rename_sem); ds = dentry->d_parent; fid = v9fs_fid_find(ds, uid, any); - if (!fid) { /* walk from the root */ - n = 0; - for (ds = dentry; !IS_ROOT(ds); ds = ds->d_parent) - n++; + if (fid) { + /* Found the parent fid do a lookup with that */ + fid = p9_client_walk(fid, 1, (char **)&dentry->d_name.name, 1); + goto fid_out; + } + up_read(&v9ses->rename_sem); - fid = v9fs_fid_find(ds, uid, any); - if (!fid) { /* the user is not attached to the fs yet */ - if (access == V9FS_ACCESS_SINGLE) - return ERR_PTR(-EPERM); + /* start from the root and try to do a lookup */ + fid = v9fs_fid_find(dentry->d_sb->s_root, uid, any); + if (!fid) { + /* the user is not attached to the fs yet */ + if (access == V9FS_ACCESS_SINGLE) + return ERR_PTR(-EPERM); - if (v9fs_proto_dotu(v9ses)) + if (v9fs_proto_dotu(v9ses) || v9fs_proto_dotl(v9ses)) uname = NULL; - else - uname = v9ses->uname; + else + uname = v9ses->uname; - fid = p9_client_attach(v9ses->clnt, NULL, uname, uid, - v9ses->aname); - - if (IS_ERR(fid)) - return fid; - - v9fs_fid_add(ds, fid); - } - } else /* walk from the parent */ - n = 1; + fid = p9_client_attach(v9ses->clnt, NULL, uname, uid, + v9ses->aname); + if (IS_ERR(fid)) + return fid; - if (ds == dentry) + v9fs_fid_add(dentry->d_sb->s_root, fid); + } + /* If we are root ourself just return that */ + if (dentry->d_sb->s_root == dentry) return fid; - - wnames = kmalloc(sizeof(char *) * n, GFP_KERNEL); - if (!wnames) - return ERR_PTR(-ENOMEM); - - for (d = dentry, i = (n-1); i >= 0; i--, d = d->d_parent) - wnames[i] = (char *) d->d_name.name; - + /* + * Do a multipath walk with attached root. + * When walking parent we need to make sure we + * don't have a parallel rename happening + */ + down_read(&v9ses->rename_sem); + n = build_path_from_dentry(v9ses, dentry, &wnames); + if (n < 0) { + fid = ERR_PTR(n); + goto err_out; + } clone = 1; i = 0; while (i < n) { l = min(n - i, P9_MAXWELEM); + /* + * We need to hold rename lock when doing a multipath + * walk to ensure none of the patch component change + */ fid = p9_client_walk(fid, l, &wnames[i], clone); if (IS_ERR(fid)) { if (old_fid) { @@ -193,15 +234,17 @@ struct p9_fid *v9fs_fid_lookup(struct dentry *dentry) p9_client_clunk(old_fid); } kfree(wnames); - return fid; + goto err_out; } old_fid = fid; i += l; clone = 0; } - kfree(wnames); +fid_out: v9fs_fid_add(dentry, fid); +err_out: + up_read(&v9ses->rename_sem); return fid; } diff --git a/fs/9p/v9fs.c b/fs/9p/v9fs.c index f8b86e92cd6..38dc0e06759 100644 --- a/fs/9p/v9fs.c +++ b/fs/9p/v9fs.c @@ -237,6 +237,7 @@ struct p9_fid *v9fs_session_init(struct v9fs_session_info *v9ses, __putname(v9ses->uname); return ERR_PTR(-ENOMEM); } + init_rwsem(&v9ses->rename_sem); rc = bdi_setup_and_register(&v9ses->bdi, "9p", BDI_CAP_MAP_COPY); if (rc) { @@ -278,7 +279,7 @@ struct p9_fid *v9fs_session_init(struct v9fs_session_info *v9ses, v9ses->maxdata = v9ses->clnt->msize - P9_IOHDRSZ; /* for legacy mode, fall back to V9FS_ACCESS_ANY */ - if (!v9fs_proto_dotu(v9ses) && + if (!(v9fs_proto_dotu(v9ses) || v9fs_proto_dotl(v9ses)) && ((v9ses->flags&V9FS_ACCESS_MASK) == V9FS_ACCESS_USER)) { v9ses->flags &= ~V9FS_ACCESS_MASK; diff --git a/fs/9p/v9fs.h b/fs/9p/v9fs.h index bec4d0bcb45..4c963c9fc41 100644 --- a/fs/9p/v9fs.h +++ b/fs/9p/v9fs.h @@ -104,6 +104,7 @@ struct v9fs_session_info { struct p9_client *clnt; /* 9p client */ struct list_head slist; /* list of sessions registered with v9fs */ struct backing_dev_info bdi; + struct rw_semaphore rename_sem; }; struct p9_fid *v9fs_session_init(struct v9fs_session_info *, const char *, diff --git a/fs/9p/v9fs_vfs.h b/fs/9p/v9fs_vfs.h index 32ef4009d03..f47c6bbb01b 100644 --- a/fs/9p/v9fs_vfs.h +++ b/fs/9p/v9fs_vfs.h @@ -55,6 +55,7 @@ struct inode *v9fs_get_inode(struct super_block *sb, int mode); void v9fs_clear_inode(struct inode *inode); ino_t v9fs_qid2ino(struct p9_qid *qid); void v9fs_stat2inode(struct p9_wstat *, struct inode *, struct super_block *); +void v9fs_stat2inode_dotl(struct p9_stat_dotl *, struct inode *); 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_wstat *stat); diff --git a/fs/9p/vfs_dir.c b/fs/9p/vfs_dir.c index 36d961f342a..16c8a2a98c1 100644 --- a/fs/9p/vfs_dir.c +++ b/fs/9p/vfs_dir.c @@ -87,29 +87,19 @@ static void p9stat_init(struct p9_wstat *stbuf) } /** - * v9fs_dir_readdir - read a directory + * v9fs_alloc_rdir_buf - Allocate buffer used for read and readdir * @filp: opened file structure - * @dirent: directory structure ??? - * @filldir: function to populate directory structure ??? + * @buflen: Length in bytes of buffer to allocate * */ -static int v9fs_dir_readdir(struct file *filp, void *dirent, filldir_t filldir) +static int v9fs_alloc_rdir_buf(struct file *filp, int buflen) { - int over; - struct p9_wstat st; - int err = 0; - struct p9_fid *fid; - int buflen; - int reclen = 0; struct p9_rdir *rdir; + struct p9_fid *fid; + int err = 0; - P9_DPRINTK(P9_DEBUG_VFS, "name %s\n", filp->f_path.dentry->d_name.name); fid = filp->private_data; - - buflen = fid->clnt->msize - P9_IOHDRSZ; - - /* allocate rdir on demand */ if (!fid->rdir) { rdir = kmalloc(sizeof(struct p9_rdir) + buflen, GFP_KERNEL); @@ -128,6 +118,36 @@ static int v9fs_dir_readdir(struct file *filp, void *dirent, filldir_t filldir) spin_unlock(&filp->f_dentry->d_lock); kfree(rdir); } +exit: + return err; +} + +/** + * v9fs_dir_readdir - read a directory + * @filp: opened file structure + * @dirent: directory structure ??? + * @filldir: function to populate directory structure ??? + * + */ + +static int v9fs_dir_readdir(struct file *filp, void *dirent, filldir_t filldir) +{ + int over; + struct p9_wstat st; + int err = 0; + struct p9_fid *fid; + int buflen; + int reclen = 0; + struct p9_rdir *rdir; + + P9_DPRINTK(P9_DEBUG_VFS, "name %s\n", filp->f_path.dentry->d_name.name); + fid = filp->private_data; + + buflen = fid->clnt->msize - P9_IOHDRSZ; + + err = v9fs_alloc_rdir_buf(filp, buflen); + if (err) + goto exit; rdir = (struct p9_rdir *) fid->rdir; err = mutex_lock_interruptible(&rdir->mutex); @@ -176,6 +196,88 @@ exit: return err; } +/** + * v9fs_dir_readdir_dotl - read a directory + * @filp: opened file structure + * @dirent: buffer to fill dirent structures + * @filldir: function to populate dirent structures + * + */ +static int v9fs_dir_readdir_dotl(struct file *filp, void *dirent, + filldir_t filldir) +{ + int over; + int err = 0; + struct p9_fid *fid; + int buflen; + struct p9_rdir *rdir; + struct p9_dirent curdirent; + u64 oldoffset = 0; + + P9_DPRINTK(P9_DEBUG_VFS, "name %s\n", filp->f_path.dentry->d_name.name); + fid = filp->private_data; + + buflen = fid->clnt->msize - P9_READDIRHDRSZ; + + err = v9fs_alloc_rdir_buf(filp, buflen); + if (err) + goto exit; + rdir = (struct p9_rdir *) fid->rdir; + + err = mutex_lock_interruptible(&rdir->mutex); + if (err) + return err; + + while (err == 0) { + if (rdir->tail == rdir->head) { + err = p9_client_readdir(fid, rdir->buf, buflen, + filp->f_pos); + if (err <= 0) + goto unlock_and_exit; + + rdir->head = 0; + rdir->tail = err; + } + + while (rdir->head < rdir->tail) { + + err = p9dirent_read(rdir->buf + rdir->head, + buflen - rdir->head, &curdirent, + fid->clnt->proto_version); + if (err < 0) { + P9_DPRINTK(P9_DEBUG_VFS, "returned %d\n", err); + err = -EIO; + goto unlock_and_exit; + } + + /* d_off in dirent structure tracks the offset into + * the next dirent in the dir. However, filldir() + * expects offset into the current dirent. Hence + * while calling filldir send the offset from the + * previous dirent structure. + */ + over = filldir(dirent, curdirent.d_name, + strlen(curdirent.d_name), + oldoffset, v9fs_qid2ino(&curdirent.qid), + curdirent.d_type); + oldoffset = curdirent.d_off; + + if (over) { + err = 0; + goto unlock_and_exit; + } + + filp->f_pos = curdirent.d_off; + rdir->head += err; + } + } + +unlock_and_exit: + mutex_unlock(&rdir->mutex); +exit: + return err; +} + /** * v9fs_dir_release - close a directory @@ -207,7 +309,7 @@ const struct file_operations v9fs_dir_operations = { const struct file_operations v9fs_dir_operations_dotl = { .read = generic_read_dir, .llseek = generic_file_llseek, - .readdir = v9fs_dir_readdir, + .readdir = v9fs_dir_readdir_dotl, .open = v9fs_file_open, .release = v9fs_dir_release, }; diff --git a/fs/9p/vfs_file.c b/fs/9p/vfs_file.c index 2bedc6c94fc..e97c92bd6f1 100644 --- a/fs/9p/vfs_file.c +++ b/fs/9p/vfs_file.c @@ -59,9 +59,13 @@ int v9fs_file_open(struct inode *inode, struct file *file) struct p9_fid *fid; int omode; - P9_DPRINTK(P9_DEBUG_VFS, "inode: %p file: %p \n", inode, file); + P9_DPRINTK(P9_DEBUG_VFS, "inode: %p file: %p\n", inode, file); v9ses = v9fs_inode2v9ses(inode); - omode = v9fs_uflags2omode(file->f_flags, v9fs_proto_dotu(v9ses)); + if (v9fs_proto_dotl(v9ses)) + omode = file->f_flags; + else + omode = v9fs_uflags2omode(file->f_flags, + v9fs_proto_dotu(v9ses)); fid = file->private_data; if (!fid) { fid = v9fs_fid_clone(file->f_path.dentry); @@ -73,11 +77,12 @@ int v9fs_file_open(struct inode *inode, struct file *file) p9_client_clunk(fid); return err; } - if (omode & P9_OTRUNC) { + if (file->f_flags & O_TRUNC) { i_size_write(inode, 0); inode->i_blocks = 0; } - if ((file->f_flags & O_APPEND) && (!v9fs_proto_dotu(v9ses))) + if ((file->f_flags & O_APPEND) && + (!v9fs_proto_dotu(v9ses) && !v9fs_proto_dotl(v9ses))) generic_file_llseek(file, 0, SEEK_END); } @@ -139,7 +144,7 @@ ssize_t v9fs_file_readn(struct file *filp, char *data, char __user *udata, u32 count, u64 offset) { - int n, total; + int n, total, size; struct p9_fid *fid = filp->private_data; P9_DPRINTK(P9_DEBUG_VFS, "fid %d offset %llu count %d\n", fid->fid, @@ -147,6 +152,7 @@ v9fs_file_readn(struct file *filp, char *data, char __user *udata, u32 count, n = 0; total = 0; + size = fid->iounit ? fid->iounit : fid->clnt->msize - P9_IOHDRSZ; do { n = p9_client_read(fid, data, udata, offset, count); if (n <= 0) @@ -160,7 +166,7 @@ v9fs_file_readn(struct file *filp, char *data, char __user *udata, u32 count, offset += n; count -= n; total += n; - } while (count > 0 && n == (fid->clnt->msize - P9_IOHDRSZ)); + } while (count > 0 && n == size); if (n < 0) total = n; @@ -183,11 +189,13 @@ v9fs_file_read(struct file *filp, char __user *udata, size_t count, { int ret; struct p9_fid *fid; + size_t size; P9_DPRINTK(P9_DEBUG_VFS, "count %zu offset %lld\n", count, *offset); fid = filp->private_data; - if (count > (fid->clnt->msize - P9_IOHDRSZ)) + size = fid->iounit ? fid->iounit : fid->clnt->msize - P9_IOHDRSZ; + if (count > size) ret = v9fs_file_readn(filp, NULL, udata, count, *offset); else ret = p9_client_read(fid, NULL, udata, *offset, count); @@ -224,9 +232,7 @@ v9fs_file_write(struct file *filp, const char __user * data, fid = filp->private_data; clnt = fid->clnt; - rsize = fid->iounit; - if (!rsize || rsize > clnt->msize-P9_IOHDRSZ) - rsize = clnt->msize - P9_IOHDRSZ; + rsize = fid->iounit ? fid->iounit : clnt->msize - P9_IOHDRSZ; do { if (count < rsize) diff --git a/fs/9p/vfs_inode.c b/fs/9p/vfs_inode.c index 4331b3b5ee1..6e94f3247ce 100644 --- a/fs/9p/vfs_inode.c +++ b/fs/9p/vfs_inode.c @@ -35,6 +35,7 @@ #include <linux/idr.h> #include <linux/sched.h> #include <linux/slab.h> +#include <linux/xattr.h> #include <net/9p/9p.h> #include <net/9p/client.h> @@ -42,6 +43,7 @@ #include "v9fs_vfs.h" #include "fid.h" #include "cache.h" +#include "xattr.h" static const struct inode_operations v9fs_dir_inode_operations; static const struct inode_operations v9fs_dir_inode_operations_dotu; @@ -236,6 +238,41 @@ void v9fs_destroy_inode(struct inode *inode) #endif /** + * v9fs_get_fsgid_for_create - Helper function to get the gid for creating a + * new file system object. This checks the S_ISGID to determine the owning + * group of the new file system object. + */ + +static gid_t v9fs_get_fsgid_for_create(struct inode *dir_inode) +{ + BUG_ON(dir_inode == NULL); + + if (dir_inode->i_mode & S_ISGID) { + /* set_gid bit is set.*/ + return dir_inode->i_gid; + } + return current_fsgid(); +} + +/** + * v9fs_dentry_from_dir_inode - helper function to get the dentry from + * dir inode. + * + */ + +static struct dentry *v9fs_dentry_from_dir_inode(struct inode *inode) +{ + struct dentry *dentry; + + spin_lock(&dcache_lock); + /* Directory should have only one entry. */ + BUG_ON(S_ISDIR(inode->i_mode) && !list_is_singular(&inode->i_dentry)); + dentry = list_entry(inode->i_dentry.next, struct dentry, d_alias); + spin_unlock(&dcache_lock); + return dentry; +} + +/** * v9fs_get_inode - helper function to setup an inode * @sb: superblock * @mode: mode to setup inode with @@ -267,7 +304,13 @@ struct inode *v9fs_get_inode(struct super_block *sb, int mode) case S_IFBLK: case S_IFCHR: case S_IFSOCK: - if (!v9fs_proto_dotu(v9ses)) { + if (v9fs_proto_dotl(v9ses)) { + inode->i_op = &v9fs_file_inode_operations_dotl; + inode->i_fop = &v9fs_file_operations_dotl; + } else if (v9fs_proto_dotu(v9ses)) { + inode->i_op = &v9fs_file_inode_operations; + inode->i_fop = &v9fs_file_operations; + } else { P9_DPRINTK(P9_DEBUG_ERROR, "special files without extended mode\n"); err = -EINVAL; @@ -396,23 +439,14 @@ void v9fs_clear_inode(struct inode *inode) #endif } -/** - * v9fs_inode_from_fid - populate an inode by issuing a attribute request - * @v9ses: session information - * @fid: fid to issue attribute request for - * @sb: superblock on which to create inode - * - */ - static struct inode * -v9fs_inode_from_fid(struct v9fs_session_info *v9ses, struct p9_fid *fid, +v9fs_inode(struct v9fs_session_info *v9ses, struct p9_fid *fid, struct super_block *sb) { int err, umode; - struct inode *ret; + struct inode *ret = NULL; struct p9_wstat *st; - ret = NULL; st = p9_client_stat(fid); if (IS_ERR(st)) return ERR_CAST(st); @@ -433,15 +467,62 @@ v9fs_inode_from_fid(struct v9fs_session_info *v9ses, struct p9_fid *fid, #endif p9stat_free(st); kfree(st); - return ret; - error: p9stat_free(st); kfree(st); return ERR_PTR(err); } +static struct inode * +v9fs_inode_dotl(struct v9fs_session_info *v9ses, struct p9_fid *fid, + struct super_block *sb) +{ + struct inode *ret = NULL; + int err; + struct p9_stat_dotl *st; + + st = p9_client_getattr_dotl(fid, P9_STATS_BASIC); + if (IS_ERR(st)) + return ERR_CAST(st); + + ret = v9fs_get_inode(sb, st->st_mode); + if (IS_ERR(ret)) { + err = PTR_ERR(ret); + goto error; + } + + v9fs_stat2inode_dotl(st, ret); + ret->i_ino = v9fs_qid2ino(&st->qid); +#ifdef CONFIG_9P_FSCACHE + v9fs_vcookie_set_qid(ret, &st->qid); + v9fs_cache_inode_get_cookie(ret); +#endif + kfree(st); + return ret; +error: + kfree(st); + return ERR_PTR(err); +} + +/** + * v9fs_inode_from_fid - Helper routine to populate an inode by + * issuing a attribute request + * @v9ses: session information + * @fid: fid to issue attribute request for + * @sb: superblock on which to create inode + * + */ +static inline struct inode * +v9fs_inode_from_fid(struct v9fs_session_info *v9ses, struct p9_fid *fid, + struct super_block *sb) +{ + if (v9fs_proto_dotl(v9ses)) + return v9fs_inode_dotl(v9ses, fid, sb); + else + return v9fs_inode(v9ses, fid, sb); +} + /** * v9fs_remove - helper function to remove files and directories * @dir: directory inode that is being deleted @@ -563,6 +644,118 @@ error: } /** + * v9fs_vfs_create_dotl - VFS hook to create files for 9P2000.L protocol. + * @dir: directory inode that is being created + * @dentry: dentry that is being deleted + * @mode: create permissions + * @nd: path information + * + */ + +static int +v9fs_vfs_create_dotl(struct inode *dir, struct dentry *dentry, int mode, + struct nameidata *nd) +{ + int err = 0; + char *name = NULL; + gid_t gid; + int flags; + struct v9fs_session_info *v9ses; + struct p9_fid *fid = NULL; + struct p9_fid *dfid, *ofid; + struct file *filp; + struct p9_qid qid; + struct inode *inode; + + v9ses = v9fs_inode2v9ses(dir); + if (nd && nd->flags & LOOKUP_OPEN) + flags = nd->intent.open.flags - 1; + else + flags = O_RDWR; + + name = (char *) dentry->d_name.name; + P9_DPRINTK(P9_DEBUG_VFS, "v9fs_vfs_create_dotl: name:%s flags:0x%x " + "mode:0x%x\n", name, flags, mode); + + dfid = v9fs_fid_lookup(dentry->d_parent); + if (IS_ERR(dfid)) { + err = PTR_ERR(dfid); + P9_DPRINTK(P9_DEBUG_VFS, "fid lookup failed %d\n", err); + return err; + } + + /* clone a fid to use for creation */ + 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); + return err; + } + + gid = v9fs_get_fsgid_for_create(dir); + err = p9_client_create_dotl(ofid, name, flags, mode, gid, &qid); + if (err < 0) { + P9_DPRINTK(P9_DEBUG_VFS, + "p9_client_open_dotl failed in creat %d\n", + err); + goto error; + } + + /* No need to populate the inode if we are not opening the file AND + * not in cached mode. + */ + if (!v9ses->cache && !(nd && nd->flags & LOOKUP_OPEN)) { + /* Not in cached mode. No need to populate inode with stat */ + dentry->d_op = &v9fs_dentry_operations; + p9_client_clunk(ofid); + d_instantiate(dentry, NULL); + return 0; + } + + /* Now walk from the parent so we can get an unopened fid. */ + fid = p9_client_walk(dfid, 1, &name, 1); + if (IS_ERR(fid)) { + err = PTR_ERR(fid); + P9_DPRINTK(P9_DEBUG_VFS, "p9_client_walk failed %d\n", err); + fid = NULL; + goto error; + } + + /* instantiate inode and assign the unopened fid to dentry */ + 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; + } + dentry->d_op = &v9fs_cached_dentry_operations; + d_instantiate(dentry, inode); + err = v9fs_fid_add(dentry, fid); + if (err < 0) + goto error; + + /* if we are opening a file, assign the open fid to the file */ + if (nd && nd->flags & LOOKUP_OPEN) { + filp = lookup_instantiate_filp(nd, dentry, v9fs_open_created); + if (IS_ERR(filp)) { + p9_client_clunk(ofid); + return PTR_ERR(filp); + } + filp->private_data = ofid; + } else + p9_client_clunk(ofid); + + return 0; + +error: + if (ofid) + p9_client_clunk(ofid); + if (fid) + p9_client_clunk(fid); + return err; +} + +/** * v9fs_vfs_create - VFS hook to create files * @dir: directory inode that is being created * @dentry: dentry that is being deleted @@ -652,6 +845,83 @@ static int v9fs_vfs_mkdir(struct inode *dir, struct dentry *dentry, int mode) return err; } + +/** + * v9fs_vfs_mkdir_dotl - VFS mkdir hook to create a directory + * @dir: inode that is being unlinked + * @dentry: dentry that is being unlinked + * @mode: mode for new directory + * + */ + +static int v9fs_vfs_mkdir_dotl(struct inode *dir, struct dentry *dentry, + int mode) +{ + int err; + struct v9fs_session_info *v9ses; + struct p9_fid *fid = NULL, *dfid = NULL; + gid_t gid; + char *name; + struct inode *inode; + struct p9_qid qid; + struct dentry *dir_dentry; + + P9_DPRINTK(P9_DEBUG_VFS, "name %s\n", dentry->d_name.name); + err = 0; + v9ses = v9fs_inode2v9ses(dir); + + mode |= S_IFDIR; + dir_dentry = v9fs_dentry_from_dir_inode(dir); + dfid = v9fs_fid_lookup(dir_dentry); + if (IS_ERR(dfid)) { + err = PTR_ERR(dfid); + P9_DPRINTK(P9_DEBUG_VFS, "fid lookup failed %d\n", err); + dfid = NULL; + goto error; + } + + gid = v9fs_get_fsgid_for_create(dir); + if (gid < 0) { + P9_DPRINTK(P9_DEBUG_VFS, "v9fs_get_fsgid_for_create failed\n"); + goto error; + } + + name = (char *) dentry->d_name.name; + err = p9_client_mkdir_dotl(dfid, name, mode, gid, &qid); + if (err < 0) + goto error; + + /* instantiate inode and assign the unopened fid to the dentry */ + if (v9ses->cache == CACHE_LOOSE || v9ses->cache == CACHE_FSCACHE) { + fid = p9_client_walk(dfid, 1, &name, 1); + if (IS_ERR(fid)) { + err = PTR_ERR(fid); + P9_DPRINTK(P9_DEBUG_VFS, "p9_client_walk failed %d\n", + err); + fid = NULL; + goto error; + } + + 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; + } + dentry->d_op = &v9fs_cached_dentry_operations; + d_instantiate(dentry, inode); + err = v9fs_fid_add(dentry, fid); + if (err < 0) + goto error; + fid = NULL; + } +error: + if (fid) + p9_client_clunk(fid); + return err; +} + /** * v9fs_vfs_lookup - VFS lookup hook to "walk" to a new inode * @dir: inode that is being walked from @@ -678,6 +948,7 @@ static struct dentry *v9fs_vfs_lookup(struct inode *dir, struct dentry *dentry, sb = dir->i_sb; v9ses = v9fs_inode2v9ses(dir); + /* We can walk d_parent because we hold the dir->i_mutex */ dfid = v9fs_fid_lookup(dentry->d_parent); if (IS_ERR(dfid)) return ERR_CAST(dfid); @@ -785,27 +1056,33 @@ v9fs_vfs_rename(struct inode *old_dir, struct dentry *old_dentry, goto clunk_olddir; } + down_write(&v9ses->rename_sem); if (v9fs_proto_dotl(v9ses)) { retval = p9_client_rename(oldfid, newdirfid, (char *) new_dentry->d_name.name); if (retval != -ENOSYS) goto clunk_newdir; } + if (old_dentry->d_parent != new_dentry->d_parent) { + /* + * 9P .u can only handle file rename in the same directory + */ - /* 9P can only handle file rename in the same directory */ - if (memcmp(&olddirfid->qid, &newdirfid->qid, sizeof(newdirfid->qid))) { P9_DPRINTK(P9_DEBUG_ERROR, "old dir and new dir are different\n"); retval = -EXDEV; goto clunk_newdir; } - v9fs_blank_wstat(&wstat); wstat.muid = v9ses->uname; wstat.name = (char *) new_dentry->d_name.name; retval = p9_client_wstat(oldfid, &wstat); clunk_newdir: + if (!retval) + /* successful rename */ + d_move(old_dentry, new_dentry); + up_write(&v9ses->rename_sem); p9_client_clunk(newdirfid); clunk_olddir: @@ -853,6 +1130,42 @@ v9fs_vfs_getattr(struct vfsmount *mnt, struct dentry *dentry, return 0; } +static int +v9fs_vfs_getattr_dotl(struct vfsmount *mnt, struct dentry *dentry, + struct kstat *stat) +{ + int err; + struct v9fs_session_info *v9ses; + struct p9_fid *fid; + struct p9_stat_dotl *st; + + P9_DPRINTK(P9_DEBUG_VFS, "dentry: %p\n", dentry); + err = -EPERM; + v9ses = v9fs_inode2v9ses(dentry->d_inode); + if (v9ses->cache == CACHE_LOOSE || v9ses->cache == CACHE_FSCACHE) + return simple_getattr(mnt, dentry, stat); + + fid = v9fs_fid_lookup(dentry); + if (IS_ERR(fid)) + return PTR_ERR(fid); + + /* Ask for all the fields in stat structure. Server will return + * whatever it supports + */ + + st = p9_client_getattr_dotl(fid, P9_STATS_ALL); + if (IS_ERR(st)) + return PTR_ERR(st); + + v9fs_stat2inode_dotl(st, dentry->d_inode); + generic_fillattr(dentry->d_inode, stat); + /* Change block size to what the server returned */ + stat->blksize = st->st_blksize; + + kfree(st); + return 0; +} + /** * v9fs_vfs_setattr - set file metadata * @dentry: file whose metadata to set @@ -903,6 +1216,49 @@ static int v9fs_vfs_setattr(struct dentry *dentry, struct iattr *iattr) } /** + * v9fs_vfs_setattr_dotl - set file metadata + * @dentry: file whose metadata to set + * @iattr: metadata assignment structure + * + */ + +static int v9fs_vfs_setattr_dotl(struct dentry *dentry, struct iattr *iattr) +{ + int retval; + struct v9fs_session_info *v9ses; + struct p9_fid *fid; + struct p9_iattr_dotl p9attr; + + P9_DPRINTK(P9_DEBUG_VFS, "\n"); + + retval = inode_change_ok(dentry->d_inode, iattr); + if (retval) + return retval; + + p9attr.valid = iattr->ia_valid; + p9attr.mode = iattr->ia_mode; + p9attr.uid = iattr->ia_uid; + p9attr.gid = iattr->ia_gid; + p9attr.size = iattr->ia_size; + p9attr.atime_sec = iattr->ia_atime.tv_sec; + p9attr.atime_nsec = iattr->ia_atime.tv_nsec; + p9attr.mtime_sec = iattr->ia_mtime.tv_sec; + p9attr.mtime_nsec = iattr->ia_mtime.tv_nsec; + + retval = -EPERM; + v9ses = v9fs_inode2v9ses(dentry->d_inode); + fid = v9fs_fid_lookup(dentry); + if (IS_ERR(fid)) + return PTR_ERR(fid); + + retval = p9_client_setattr(fid, &p9attr); + if (retval >= 0) + retval = inode_setattr(dentry->d_inode, iattr); + + return retval; +} + +/** * v9fs_stat2inode - populate an inode structure with mistat info * @stat: Plan 9 metadata (mistat) structure * @inode: inode to populate @@ -980,6 +1336,77 @@ v9fs_stat2inode(struct p9_wstat *stat, struct inode *inode, } /** + * v9fs_stat2inode_dotl - populate an inode structure with stat info + * @stat: stat structure + * @inode: inode to populate + * @sb: superblock of filesystem + * + */ + +void +v9fs_stat2inode_dotl(struct p9_stat_dotl *stat, struct inode *inode) +{ + + if ((stat->st_result_mask & P9_STATS_BASIC) == P9_STATS_BASIC) { + inode->i_atime.tv_sec = stat->st_atime_sec; + inode->i_atime.tv_nsec = stat->st_atime_nsec; + inode->i_mtime.tv_sec = stat->st_mtime_sec; + inode->i_mtime.tv_nsec = stat->st_mtime_nsec; + inode->i_ctime.tv_sec = stat->st_ctime_sec; + inode->i_ctime.tv_nsec = stat->st_ctime_nsec; + inode->i_uid = stat->st_uid; + inode->i_gid = stat->st_gid; + inode->i_nlink = stat->st_nlink; + inode->i_mode = stat->st_mode; + inode->i_rdev = new_decode_dev(stat->st_rdev); + + if ((S_ISBLK(inode->i_mode)) || (S_ISCHR(inode->i_mode))) + init_special_inode(inode, inode->i_mode, inode->i_rdev); + + i_size_write(inode, stat->st_size); + inode->i_blocks = stat->st_blocks; + } else { + if (stat->st_result_mask & P9_STATS_ATIME) { + inode->i_atime.tv_sec = stat->st_atime_sec; + inode->i_atime.tv_nsec = stat->st_atime_nsec; + } + if (stat->st_result_mask & P9_STATS_MTIME) { + inode->i_mtime.tv_sec = stat->st_mtime_sec; + inode->i_mtime.tv_nsec = stat->st_mtime_nsec; + } + if (stat->st_result_mask & P9_STATS_CTIME) { + inode->i_ctime.tv_sec = stat->st_ctime_sec; + inode->i_ctime.tv_nsec = stat->st_ctime_nsec; + } + if (stat->st_result_mask & P9_STATS_UID) + inode->i_uid = stat->st_uid; + if (stat->st_result_mask & P9_STATS_GID) + inode->i_gid = stat->st_gid; + if (stat->st_result_mask & P9_STATS_NLINK) + inode->i_nlink = stat->st_nlink; + if (stat->st_result_mask & P9_STATS_MODE) { + inode->i_mode = stat->st_mode; + if ((S_ISBLK(inode->i_mode)) || + (S_ISCHR(inode->i_mode))) + init_special_inode(inode, inode->i_mode, + inode->i_rdev); + } + if (stat->st_result_mask & P9_STATS_RDEV) + inode->i_rdev = new_decode_dev(stat->st_rdev); + if (stat->st_result_mask & P9_STATS_SIZE) + i_size_write(inode, stat->st_size); + if (stat->st_result_mask & P9_STATS_BLOCKS) + inode->i_blocks = stat->st_blocks; + } + if (stat->st_result_mask & P9_STATS_GEN) + inode->i_generation = stat->st_gen; + + /* Currently we don't support P9_STATS_BTIME and P9_STATS_DATA_VERSION + * because the inode structure does not have fields for them. + */ +} + +/** * v9fs_qid2ino - convert qid into inode number * @qid: qid to hash * @@ -1022,7 +1449,7 @@ static int v9fs_readlink(struct dentry *dentry, char *buffer, int buflen) if (IS_ERR(fid)) return PTR_ERR(fid); - if (!v9fs_proto_dotu(v9ses)) + if (!v9fs_proto_dotu(v9ses) && !v9fs_proto_dotl(v9ses)) return -EBADF; st = p9_client_stat(fid); @@ -1128,6 +1555,99 @@ static int v9fs_vfs_mkspecial(struct inode *dir, struct dentry *dentry, } /** + * v9fs_vfs_symlink_dotl - helper function to create symlinks + * @dir: directory inode containing symlink + * @dentry: dentry for symlink + * @symname: symlink data + * + * See Also: 9P2000.L RFC for more information + * + */ + +static int +v9fs_vfs_symlink_dotl(struct inode *dir, struct dentry *dentry, + const char *symname) +{ + struct v9fs_session_info *v9ses; + struct p9_fid *dfid; + struct p9_fid *fid = NULL; + struct inode *inode; + struct p9_qid qid; + char *name; + int err; + gid_t gid; + + name = (char *) dentry->d_name.name; + P9_DPRINTK(P9_DEBUG_VFS, "v9fs_vfs_symlink_dotl : %lu,%s,%s\n", + dir->i_ino, name, symname); + v9ses = v9fs_inode2v9ses(dir); + + dfid = v9fs_fid_lookup(dentry->d_parent); + if (IS_ERR(dfid)) { + err = PTR_ERR(dfid); + P9_DPRINTK(P9_DEBUG_VFS, "fid lookup failed %d\n", err); + return err; + } + + gid = v9fs_get_fsgid_for_create(dir); + + if (gid < 0) { + P9_DPRINTK(P9_DEBUG_VFS, "v9fs_get_egid failed %d\n", gid); + goto error; + } + + /* Server doesn't alter fid on TSYMLINK. Hence no need to clone it. */ + err = p9_client_symlink(dfid, name, (char *)symname, gid, &qid); + + if (err < 0) { + P9_DPRINTK(P9_DEBUG_VFS, "p9_client_symlink failed %d\n", err); + goto error; + } + + if (v9ses->cache) { + /* Now walk from the parent so we can get an unopened fid. */ + fid = p9_client_walk(dfid, 1, &name, 1); + if (IS_ERR(fid)) { + err = PTR_ERR(fid); + P9_DPRINTK(P9_DEBUG_VFS, "p9_client_walk failed %d\n", + err); + fid = NULL; + goto error; + } + + /* instantiate inode and assign the unopened fid to dentry */ + 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; + } + dentry->d_op = &v9fs_cached_dentry_operations; + d_instantiate(dentry, inode); + err = v9fs_fid_add(dentry, fid); + if (err < 0) + goto error; + fid = NULL; + } else { + /* Not in cached mode. No need to populate inode with stat */ + inode = v9fs_get_inode(dir->i_sb, S_IFLNK); + if (IS_ERR(inode)) { + err = PTR_ERR(inode); + goto error; + } + dentry->d_op = &v9fs_dentry_operations; + d_instantiate(dentry, inode); + } + +error: + if (fid) + p9_client_clunk(fid); + + return err; +} + +/** * v9fs_vfs_symlink - helper function to create symlinks * @dir: directory inode containing symlink * @dentry: dentry for symlink @@ -1186,6 +1706,76 @@ clunk_fid: } /** + * v9fs_vfs_link_dotl - create a hardlink for dotl + * @old_dentry: dentry for file to link to + * @dir: inode destination for new link + * @dentry: dentry for link + * + */ + +static int +v9fs_vfs_link_dotl(struct dentry *old_dentry, struct inode *dir, + struct dentry *dentry) +{ + int err; + struct p9_fid *dfid, *oldfid; + char *name; + struct v9fs_session_info *v9ses; + struct dentry *dir_dentry; + + P9_DPRINTK(P9_DEBUG_VFS, "dir ino: %lu, old_name: %s, new_name: %s\n", + dir->i_ino, old_dentry->d_name.name, + dentry->d_name.name); + + v9ses = v9fs_inode2v9ses(dir); + dir_dentry = v9fs_dentry_from_dir_inode(dir); + dfid = v9fs_fid_lookup(dir_dentry); + if (IS_ERR(dfid)) + return PTR_ERR(dfid); + + oldfid = v9fs_fid_lookup(old_dentry); + if (IS_ERR(oldfid)) + return PTR_ERR(oldfid); + + name = (char *) dentry->d_name.name; + + err = p9_client_link(dfid, oldfid, (char *)dentry->d_name.name); + + if (err < 0) { + P9_DPRINTK(P9_DEBUG_VFS, "p9_client_link failed %d\n", err); + return err; + } + + if (v9ses->cache == CACHE_LOOSE || v9ses->cache == CACHE_FSCACHE) { + /* Get the latest stat info from server. */ + struct p9_fid *fid; + struct p9_stat_dotl *st; + + fid = v9fs_fid_lookup(old_dentry); + if (IS_ERR(fid)) + return PTR_ERR(fid); + + st = p9_client_getattr_dotl(fid, P9_STATS_BASIC); + if (IS_ERR(st)) + return PTR_ERR(st); + + v9fs_stat2inode_dotl(st, old_dentry->d_inode); + + kfree(st); + } else { + /* Caching disabled. No need to get upto date stat info. + * This dentry will be released immediately. So, just i_count++ + */ + atomic_inc(&old_dentry->d_inode->i_count); + } + + dentry->d_op = old_dentry->d_op; + d_instantiate(dentry, old_dentry->d_inode); + + return err; +} + +/** * v9fs_vfs_mknod - create a special file * @dir: inode destination for new link * @dentry: dentry for file @@ -1230,6 +1820,100 @@ v9fs_vfs_mknod(struct inode *dir, struct dentry *dentry, int mode, dev_t rdev) return retval; } +/** + * v9fs_vfs_mknod_dotl - create a special file + * @dir: inode destination for new link + * @dentry: dentry for file + * @mode: mode for creation + * @rdev: device associated with special file + * + */ +static int +v9fs_vfs_mknod_dotl(struct inode *dir, struct dentry *dentry, int mode, + dev_t rdev) +{ + int err; + char *name; + struct v9fs_session_info *v9ses; + struct p9_fid *fid = NULL, *dfid = NULL; + struct inode *inode; + gid_t gid; + struct p9_qid qid; + struct dentry *dir_dentry; + + P9_DPRINTK(P9_DEBUG_VFS, + " %lu,%s mode: %x MAJOR: %u MINOR: %u\n", dir->i_ino, + dentry->d_name.name, mode, MAJOR(rdev), MINOR(rdev)); + + if (!new_valid_dev(rdev)) + return -EINVAL; + + v9ses = v9fs_inode2v9ses(dir); + dir_dentry = v9fs_dentry_from_dir_inode(dir); + dfid = v9fs_fid_lookup(dir_dentry); + if (IS_ERR(dfid)) { + err = PTR_ERR(dfid); + P9_DPRINTK(P9_DEBUG_VFS, "fid lookup failed %d\n", err); + dfid = NULL; + goto error; + } + + gid = v9fs_get_fsgid_for_create(dir); + if (gid < 0) { + P9_DPRINTK(P9_DEBUG_VFS, "v9fs_get_fsgid_for_create failed\n"); + goto error; + } + + name = (char *) dentry->d_name.name; + + err = p9_client_mknod_dotl(dfid, name, mode, rdev, gid, &qid); + if (err < 0) + goto error; + + /* instantiate inode and assign the unopened fid to the dentry */ + if (v9ses->cache == CACHE_LOOSE || v9ses->cache == CACHE_FSCACHE) { + fid = p9_client_walk(dfid, 1, &name, 1); + if (IS_ERR(fid)) { + err = PTR_ERR(fid); + P9_DPRINTK(P9_DEBUG_VFS, "p9_client_walk failed %d\n", + err); + fid = NULL; + goto error; + } + + 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; + } + dentry->d_op = &v9fs_cached_dentry_operations; + d_instantiate(dentry, inode); + err = v9fs_fid_add(dentry, fid); + if (err < 0) + goto error; + fid = NULL; + } else { + /* + * Not in cached mode. No need to populate inode with stat. + * socket syscall returns a fd, so we need instantiate + */ + inode = v9fs_get_inode(dir->i_sb, mode); + if (IS_ERR(inode)) { + err = PTR_ERR(inode); + goto error; + } + dentry->d_op = &v9fs_dentry_operations; + d_instantiate(dentry, inode); + } + +error: + if (fid) + p9_client_clunk(fid); + return err; +} + static const struct inode_operations v9fs_dir_inode_operations_dotu = { .create = v9fs_vfs_create, .lookup = v9fs_vfs_lookup, @@ -1238,24 +1922,29 @@ static const struct inode_operations v9fs_dir_inode_operations_dotu = { .unlink = v9fs_vfs_unlink, .mkdir = v9fs_vfs_mkdir, .rmdir = v9fs_vfs_rmdir, - .mknod = v9fs_vfs_mknod, + .mknod = v9fs_vfs_mknod_dotl, .rename = v9fs_vfs_rename, .getattr = v9fs_vfs_getattr, .setattr = v9fs_vfs_setattr, }; static const struct inode_operations v9fs_dir_inode_operations_dotl = { - .create = v9fs_vfs_create, + .create = v9fs_vfs_create_dotl, .lookup = v9fs_vfs_lookup, - .symlink = v9fs_vfs_symlink, - .link = v9fs_vfs_link, + .link = v9fs_vfs_link_dotl, + .symlink = v9fs_vfs_symlink_dotl, .unlink = v9fs_vfs_unlink, - .mkdir = v9fs_vfs_mkdir, + .mkdir = v9fs_vfs_mkdir_dotl, .rmdir = v9fs_vfs_rmdir, - .mknod = v9fs_vfs_mknod, + .mknod = v9fs_vfs_mknod_dotl, .rename = v9fs_vfs_rename, - .getattr = v9fs_vfs_getattr, - .setattr = v9fs_vfs_setattr, + .getattr = v9fs_vfs_getattr_dotl, + .setattr = v9fs_vfs_setattr_dotl, + .setxattr = generic_setxattr, + .getxattr = generic_getxattr, + .removexattr = generic_removexattr, + .listxattr = v9fs_listxattr, + }; static const struct inode_operations v9fs_dir_inode_operations = { @@ -1276,8 +1965,12 @@ static const struct inode_operations v9fs_file_inode_operations = { }; static const struct inode_operations v9fs_file_inode_operations_dotl = { - .getattr = v9fs_vfs_getattr, - .setattr = v9fs_vfs_setattr, + .getattr = v9fs_vfs_getattr_dotl, + .setattr = v9fs_vfs_setattr_dotl, + .setxattr = generic_setxattr, + .getxattr = generic_getxattr, + .removexattr = generic_removexattr, + .listxattr = v9fs_listxattr, }; static const struct inode_operations v9fs_symlink_inode_operations = { @@ -1292,6 +1985,10 @@ static const struct inode_operations v9fs_symlink_inode_operations_dotl = { .readlink = generic_readlink, .follow_link = v9fs_vfs_follow_link, .put_link = v9fs_vfs_put_link, - .getattr = v9fs_vfs_getattr, - .setattr = v9fs_vfs_setattr, + .getattr = v9fs_vfs_getattr_dotl, + .setattr = v9fs_vfs_setattr_dotl, + .setxattr = generic_setxattr, + .getxattr = generic_getxattr, + .removexattr = generic_removexattr, + .listxattr = v9fs_listxattr, }; diff --git a/fs/9p/vfs_super.c b/fs/9p/vfs_super.c index be74d020436..4b9ede0b41b 100644 --- a/fs/9p/vfs_super.c +++ b/fs/9p/vfs_super.c @@ -45,6 +45,7 @@ #include "v9fs.h" #include "v9fs_vfs.h" #include "fid.h" +#include "xattr.h" static const struct super_operations v9fs_super_ops, v9fs_super_ops_dotl; @@ -77,9 +78,10 @@ v9fs_fill_super(struct super_block *sb, struct v9fs_session_info *v9ses, sb->s_blocksize_bits = fls(v9ses->maxdata - 1); sb->s_blocksize = 1 << sb->s_blocksize_bits; sb->s_magic = V9FS_MAGIC; - if (v9fs_proto_dotl(v9ses)) + if (v9fs_proto_dotl(v9ses)) { sb->s_op = &v9fs_super_ops_dotl; - else + sb->s_xattr = v9fs_xattr_handlers; + } else sb->s_op = &v9fs_super_ops; sb->s_bdi = &v9ses->bdi; @@ -107,7 +109,6 @@ 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_wstat *st = NULL; int mode = S_IRWXUGO | S_ISVTX; struct p9_fid *fid; int retval = 0; @@ -124,16 +125,10 @@ static int v9fs_get_sb(struct file_system_type *fs_type, int flags, goto close_session; } - st = p9_client_stat(fid); - if (IS_ERR(st)) { - retval = PTR_ERR(st); - goto clunk_fid; - } - sb = sget(fs_type, NULL, v9fs_set_super, v9ses); if (IS_ERR(sb)) { retval = PTR_ERR(sb); - goto free_stat; + goto clunk_fid; } v9fs_fill_super(sb, v9ses, flags, data); @@ -151,22 +146,38 @@ 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); + if (v9fs_proto_dotl(v9ses)) { + struct p9_stat_dotl *st = NULL; + st = p9_client_getattr_dotl(fid, P9_STATS_BASIC); + if (IS_ERR(st)) { + retval = PTR_ERR(st); + goto clunk_fid; + } + + v9fs_stat2inode_dotl(st, root->d_inode); + kfree(st); + } else { + struct p9_wstat *st = NULL; + st = p9_client_stat(fid); + if (IS_ERR(st)) { + retval = PTR_ERR(st); + goto clunk_fid; + } + + root->d_inode->i_ino = v9fs_qid2ino(&st->qid); + v9fs_stat2inode(st, root->d_inode, sb); + + p9stat_free(st); + kfree(st); + } v9fs_fid_add(root, fid); - p9stat_free(st); - kfree(st); P9_DPRINTK(P9_DEBUG_VFS, " simple set mount, return 0\n"); simple_set_mnt(mnt, sb); return 0; -free_stat: - p9stat_free(st); - kfree(st); - clunk_fid: p9_client_clunk(fid); @@ -176,8 +187,6 @@ close_session: return retval; release_sb: - p9stat_free(st); - kfree(st); deactivate_locked_super(sb); return retval; } @@ -278,4 +287,5 @@ struct file_system_type v9fs_fs_type = { .get_sb = v9fs_get_sb, .kill_sb = v9fs_kill_super, .owner = THIS_MODULE, + .fs_flags = FS_RENAME_DOES_D_MOVE, }; diff --git a/fs/9p/xattr.c b/fs/9p/xattr.c new file mode 100644 index 00000000000..f88e5c2dc87 --- /dev/null +++ b/fs/9p/xattr.c @@ -0,0 +1,160 @@ +/* + * Copyright IBM Corporation, 2010 + * Author Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of version 2.1 of the GNU Lesser 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. + * + */ + +#include <linux/module.h> +#include <linux/fs.h> +#include <linux/sched.h> +#include <net/9p/9p.h> +#include <net/9p/client.h> + +#include "fid.h" +#include "xattr.h" + +/* + * v9fs_xattr_get() + * + * Copy an extended attribute into the buffer + * provided, or compute the buffer size required. + * Buffer is NULL to compute the size of the buffer required. + * + * Returns a negative error number on failure, or the number of bytes + * used / required on success. + */ +ssize_t v9fs_xattr_get(struct dentry *dentry, const char *name, + void *buffer, size_t buffer_size) +{ + ssize_t retval; + int msize, read_count; + u64 offset = 0, attr_size; + struct p9_fid *fid, *attr_fid; + + P9_DPRINTK(P9_DEBUG_VFS, "%s: name = %s value_len = %zu\n", + __func__, name, buffer_size); + + fid = v9fs_fid_lookup(dentry); + if (IS_ERR(fid)) + return PTR_ERR(fid); + + attr_fid = p9_client_xattrwalk(fid, name, &attr_size); + if (IS_ERR(attr_fid)) { + retval = PTR_ERR(attr_fid); + P9_DPRINTK(P9_DEBUG_VFS, + "p9_client_attrwalk failed %zd\n", retval); + attr_fid = NULL; + goto error; + } + if (!buffer_size) { + /* request to get the attr_size */ + retval = attr_size; + goto error; + } + if (attr_size > buffer_size) { + retval = -ERANGE; + goto error; + } + msize = attr_fid->clnt->msize; + while (attr_size) { + if (attr_size > (msize - P9_IOHDRSZ)) + read_count = msize - P9_IOHDRSZ; + else + read_count = attr_size; + read_count = p9_client_read(attr_fid, ((char *)buffer)+offset, + NULL, offset, read_count); + if (read_count < 0) { + /* error in xattr read */ + retval = read_count; + goto error; + } + offset += read_count; + attr_size -= read_count; + } + /* Total read xattr bytes */ + retval = offset; +error: + if (attr_fid) + p9_client_clunk(attr_fid); + return retval; + +} + +/* + * v9fs_xattr_set() + * + * Create, replace or remove an extended attribute for this inode. Buffer + * is NULL to remove an existing extended attribute, and non-NULL to + * either replace an existing extended attribute, or create a new extended + * attribute. The flags XATTR_REPLACE and XATTR_CREATE + * specify that an extended attribute must exist and must not exist + * previous to the call, respectively. + * + * Returns 0, or a negative error number on failure. + */ +int v9fs_xattr_set(struct dentry *dentry, const char *name, + const void *value, size_t value_len, int flags) +{ + u64 offset = 0; + int retval, msize, write_count; + struct p9_fid *fid = NULL; + + P9_DPRINTK(P9_DEBUG_VFS, "%s: name = %s value_len = %zu flags = %d\n", + __func__, name, value_len, flags); + + fid = v9fs_fid_clone(dentry); + if (IS_ERR(fid)) { + retval = PTR_ERR(fid); + fid = NULL; + goto error; + } + /* + * On success fid points to xattr + */ + retval = p9_client_xattrcreate(fid, name, value_len, flags); + if (retval < 0) { + P9_DPRINTK(P9_DEBUG_VFS, + "p9_client_xattrcreate failed %d\n", retval); + goto error; + } + msize = fid->clnt->msize;; + while (value_len) { + if (value_len > (msize - P9_IOHDRSZ)) + write_count = msize - P9_IOHDRSZ; + else + write_count = value_len; + write_count = p9_client_write(fid, ((char *)value)+offset, + NULL, offset, write_count); + if (write_count < 0) { + /* error in xattr write */ + retval = write_count; + goto error; + } + offset += write_count; + value_len -= write_count; + } + /* Total read xattr bytes */ + retval = offset; +error: + if (fid) + retval = p9_client_clunk(fid); + return retval; +} + +ssize_t v9fs_listxattr(struct dentry *dentry, char *buffer, size_t buffer_size) +{ + return v9fs_xattr_get(dentry, NULL, buffer, buffer_size); +} + +const struct xattr_handler *v9fs_xattr_handlers[] = { + &v9fs_xattr_user_handler, + NULL +}; diff --git a/fs/9p/xattr.h b/fs/9p/xattr.h new file mode 100644 index 00000000000..9ddf672ae5c --- /dev/null +++ b/fs/9p/xattr.h @@ -0,0 +1,27 @@ +/* + * Copyright IBM Corporation, 2010 + * Author Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of version 2.1 of the GNU Lesser 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. + * + */ +#ifndef FS_9P_XATTR_H +#define FS_9P_XATTR_H + +#include <linux/xattr.h> + +extern const struct xattr_handler *v9fs_xattr_handlers[]; +extern struct xattr_handler v9fs_xattr_user_handler; + +extern ssize_t v9fs_xattr_get(struct dentry *, const char *, + void *, size_t); +extern int v9fs_xattr_set(struct dentry *, const char *, + const void *, size_t, int); +extern ssize_t v9fs_listxattr(struct dentry *, char *, size_t); +#endif /* FS_9P_XATTR_H */ diff --git a/fs/9p/xattr_user.c b/fs/9p/xattr_user.c new file mode 100644 index 00000000000..d0b701b7208 --- /dev/null +++ b/fs/9p/xattr_user.c @@ -0,0 +1,80 @@ +/* + * Copyright IBM Corporation, 2010 + * Author Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of version 2.1 of the GNU Lesser 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. + * + */ + + +#include <linux/module.h> +#include <linux/string.h> +#include <linux/fs.h> +#include <linux/slab.h> +#include "xattr.h" + +static int v9fs_xattr_user_get(struct dentry *dentry, const char *name, + void *buffer, size_t size, int type) +{ + int retval; + char *full_name; + size_t name_len; + size_t prefix_len = XATTR_USER_PREFIX_LEN; + + if (name == NULL) + return -EINVAL; + + if (strcmp(name, "") == 0) + return -EINVAL; + + name_len = strlen(name); + full_name = kmalloc(prefix_len + name_len + 1 , GFP_KERNEL); + if (!full_name) + return -ENOMEM; + memcpy(full_name, XATTR_USER_PREFIX, prefix_len); + memcpy(full_name+prefix_len, name, name_len); + full_name[prefix_len + name_len] = '\0'; + + retval = v9fs_xattr_get(dentry, full_name, buffer, size); + kfree(full_name); + return retval; +} + +static int v9fs_xattr_user_set(struct dentry *dentry, const char *name, + const void *value, size_t size, int flags, int type) +{ + int retval; + char *full_name; + size_t name_len; + size_t prefix_len = XATTR_USER_PREFIX_LEN; + + if (name == NULL) + return -EINVAL; + + if (strcmp(name, "") == 0) + return -EINVAL; + + name_len = strlen(name); + full_name = kmalloc(prefix_len + name_len + 1 , GFP_KERNEL); + if (!full_name) + return -ENOMEM; + memcpy(full_name, XATTR_USER_PREFIX, prefix_len); + memcpy(full_name + prefix_len, name, name_len); + full_name[prefix_len + name_len] = '\0'; + + retval = v9fs_xattr_set(dentry, full_name, value, size, flags); + kfree(full_name); + return retval; +} + +struct xattr_handler v9fs_xattr_user_handler = { + .prefix = XATTR_USER_PREFIX, + .get = v9fs_xattr_user_get, + .set = v9fs_xattr_user_set, +}; diff --git a/fs/Kconfig b/fs/Kconfig index 5f85b594761..3d185308ec8 100644 --- a/fs/Kconfig +++ b/fs/Kconfig @@ -64,7 +64,7 @@ source "fs/autofs4/Kconfig" source "fs/fuse/Kconfig" config CUSE - tristate "Character device in Userpace support" + tristate "Character device in Userspace support" depends on FUSE_FS help This FUSE extension allows character devices to be diff --git a/fs/afs/Kconfig b/fs/afs/Kconfig index 5c4e61d3c77..8f975f25b48 100644 --- a/fs/afs/Kconfig +++ b/fs/afs/Kconfig @@ -2,6 +2,7 @@ config AFS_FS tristate "Andrew File System support (AFS) (EXPERIMENTAL)" depends on INET && EXPERIMENTAL select AF_RXRPC + select DNS_RESOLVER help If you say Y here, you will get an experimental Andrew File System driver. It currently only supports unsecured read-only AFS access. diff --git a/fs/afs/cell.c b/fs/afs/cell.c index e19c13f059e..ffea35c6387 100644 --- a/fs/afs/cell.c +++ b/fs/afs/cell.c @@ -13,6 +13,7 @@ #include <linux/slab.h> #include <linux/key.h> #include <linux/ctype.h> +#include <linux/dns_resolver.h> #include <linux/sched.h> #include <keys/rxrpc-type.h> #include "internal.h" @@ -36,6 +37,8 @@ static struct afs_cell *afs_cell_alloc(const char *name, char *vllist) struct key *key; size_t namelen; char keyname[4 + AFS_MAXCELLNAME + 1], *cp, *dp, *next; + char *dvllist = NULL, *_vllist = NULL; + char delimiter = ':'; int ret; _enter("%s,%s", name, vllist); @@ -43,8 +46,10 @@ static struct afs_cell *afs_cell_alloc(const char *name, char *vllist) BUG_ON(!name); /* TODO: want to look up "this cell" in the cache */ namelen = strlen(name); - if (namelen > AFS_MAXCELLNAME) + if (namelen > AFS_MAXCELLNAME) { + _leave(" = -ENAMETOOLONG"); return ERR_PTR(-ENAMETOOLONG); + } /* allocate and initialise a cell record */ cell = kzalloc(sizeof(struct afs_cell) + namelen + 1, GFP_KERNEL); @@ -64,15 +69,31 @@ static struct afs_cell *afs_cell_alloc(const char *name, char *vllist) INIT_LIST_HEAD(&cell->vl_list); spin_lock_init(&cell->vl_lock); + /* if the ip address is invalid, try dns query */ + if (!vllist || strlen(vllist) < 7) { + ret = dns_query("afsdb", name, namelen, "ipv4", &dvllist, NULL); + if (ret < 0) { + _leave(" = %d", ret); + return ERR_PTR(ret); + } + _vllist = dvllist; + + /* change the delimiter for user-space reply */ + delimiter = ','; + + } else { + _vllist = vllist; + } + /* fill in the VL server list from the rest of the string */ do { unsigned a, b, c, d; - next = strchr(vllist, ':'); + next = strchr(_vllist, delimiter); if (next) *next++ = 0; - if (sscanf(vllist, "%u.%u.%u.%u", &a, &b, &c, &d) != 4) + if (sscanf(_vllist, "%u.%u.%u.%u", &a, &b, &c, &d) != 4) goto bad_address; if (a > 255 || b > 255 || c > 255 || d > 255) @@ -81,7 +102,7 @@ static struct afs_cell *afs_cell_alloc(const char *name, char *vllist) cell->vl_addrs[cell->vl_naddrs++].s_addr = htonl((a << 24) | (b << 16) | (c << 8) | d); - } while (cell->vl_naddrs < AFS_CELL_MAX_ADDRS && (vllist = next)); + } while (cell->vl_naddrs < AFS_CELL_MAX_ADDRS && (_vllist = next)); /* create a key to represent an anonymous user */ memcpy(keyname, "afs@", 4); @@ -110,6 +131,7 @@ bad_address: ret = -EINVAL; error: key_put(cell->anonymous_key); + kfree(dvllist); kfree(cell); _leave(" = %d", ret); return ERR_PTR(ret); @@ -201,14 +223,12 @@ int afs_cell_init(char *rootcell) } cp = strchr(rootcell, ':'); - if (!cp) { - printk(KERN_ERR "kAFS: no VL server IP addresses specified\n"); - _leave(" = -EINVAL"); - return -EINVAL; - } + if (!cp) + _debug("kAFS: no VL server IP addresses specified"); + else + *cp++ = 0; /* allocate a cell record for the root cell */ - *cp++ = 0; new_root = afs_cell_create(rootcell, cp); if (IS_ERR(new_root)) { _leave(" = %ld", PTR_ERR(new_root)); diff --git a/fs/afs/main.c b/fs/afs/main.c index 66d54d348c5..cfd1cbe25b2 100644 --- a/fs/afs/main.c +++ b/fs/afs/main.c @@ -111,6 +111,8 @@ static int __init afs_init(void) /* initialise the callback update process */ ret = afs_callback_update_init(); + if (ret < 0) + goto error_callback_update_init; /* create the RxRPC transport */ ret = afs_open_socket(); @@ -127,15 +129,16 @@ static int __init afs_init(void) error_fs: afs_close_socket(); error_open_socket: + afs_callback_update_kill(); +error_callback_update_init: + afs_vlocation_purge(); error_vl_update_init: + afs_cell_purge(); error_cell_init: #ifdef CONFIG_AFS_FSCACHE fscache_unregister_netfs(&afs_cache_netfs); error_cache: #endif - afs_callback_update_kill(); - afs_vlocation_purge(); - afs_cell_purge(); afs_proc_cleanup(); rcu_barrier(); printk(KERN_ERR "kAFS: failed to register: %d\n", ret); @@ -1277,7 +1277,7 @@ out: /* sys_io_destroy: * Destroy the aio_context specified. May cancel any outstanding * AIOs and block on completion. Will fail with -ENOSYS if not - * implemented. May fail with -EFAULT if the context pointed to + * implemented. May fail with -EINVAL if the context pointed to * is invalid. */ SYSCALL_DEFINE1(io_destroy, aio_context_t, ctx) @@ -1795,15 +1795,16 @@ SYSCALL_DEFINE3(io_cancel, aio_context_t, ctx_id, struct iocb __user *, iocb, /* io_getevents: * Attempts to read at least min_nr events and up to nr events from - * the completion queue for the aio_context specified by ctx_id. May - * fail with -EINVAL if ctx_id is invalid, if min_nr is out of range, - * if nr is out of range, if when is out of range. May fail with - * -EFAULT if any of the memory specified to is invalid. May return - * 0 or < min_nr if no events are available and the timeout specified - * by when has elapsed, where when == NULL specifies an infinite - * timeout. Note that the timeout pointed to by when is relative and - * will be updated if not NULL and the operation blocks. Will fail - * with -ENOSYS if not implemented. + * the completion queue for the aio_context specified by ctx_id. If + * it succeeds, the number of read events is returned. May fail with + * -EINVAL if ctx_id is invalid, if min_nr is out of range, if nr is + * out of range, if timeout is out of range. May fail with -EFAULT + * if any of the memory specified is invalid. May return 0 or + * < min_nr if the timeout specified by timeout has elapsed + * before sufficient events are available, where timeout == NULL + * specifies an infinite timeout. Note that the timeout pointed to by + * timeout is relative and will be updated if not NULL and the + * operation blocks. Will fail with -ENOSYS if not implemented. */ SYSCALL_DEFINE5(io_getevents, aio_context_t, ctx_id, long, min_nr, diff --git a/fs/befs/linuxvfs.c b/fs/befs/linuxvfs.c index 34ddda888e6..dc39d282488 100644 --- a/fs/befs/linuxvfs.c +++ b/fs/befs/linuxvfs.c @@ -436,7 +436,7 @@ befs_init_inodecache(void) init_once); if (befs_inode_cachep == NULL) { printk(KERN_ERR "befs_init_inodecache: " - "Couldn't initalize inode slabcache\n"); + "Couldn't initialize inode slabcache\n"); return -ENOMEM; } diff --git a/fs/block_dev.c b/fs/block_dev.c index 99d6af81174..b3171fb0dc9 100644 --- a/fs/block_dev.c +++ b/fs/block_dev.c @@ -681,8 +681,8 @@ retry: if (!bd_may_claim(bdev, whole, holder)) return -EBUSY; - /* if someone else is claiming, wait for it to finish */ - if (whole->bd_claiming && whole->bd_claiming != holder) { + /* if claiming is already in progress, wait for it to finish */ + if (whole->bd_claiming) { wait_queue_head_t *wq = bit_waitqueue(&whole->bd_claiming, 0); DEFINE_WAIT(wait); diff --git a/fs/cachefiles/namei.c b/fs/cachefiles/namei.c index f4a7840bf42..42c7fafc8bf 100644 --- a/fs/cachefiles/namei.c +++ b/fs/cachefiles/namei.c @@ -37,9 +37,9 @@ void __cachefiles_printk_object(struct cachefiles_object *object, printk(KERN_ERR "%sobject: OBJ%x\n", prefix, object->fscache.debug_id); - printk(KERN_ERR "%sobjstate=%s fl=%lx swfl=%lx ev=%lx[%lx]\n", + printk(KERN_ERR "%sobjstate=%s fl=%lx wbusy=%x ev=%lx[%lx]\n", prefix, fscache_object_states[object->fscache.state], - object->fscache.flags, object->fscache.work.flags, + object->fscache.flags, work_busy(&object->fscache.work), object->fscache.events, object->fscache.event_mask & FSCACHE_OBJECT_EVENTS_MASK); printk(KERN_ERR "%sops=%u inp=%u exc=%u\n", @@ -212,7 +212,7 @@ wait_for_old_object: /* if the object we're waiting for is queued for processing, * then just put ourselves on the queue behind it */ - if (slow_work_is_queued(&xobject->fscache.work)) { + if (work_pending(&xobject->fscache.work)) { _debug("queue OBJ%x behind OBJ%x immediately", object->fscache.debug_id, xobject->fscache.debug_id); @@ -220,8 +220,7 @@ wait_for_old_object: } /* otherwise we sleep until either the object we're waiting for - * is done, or the slow-work facility wants the thread back to - * do other work */ + * is done, or the fscache_object is congested */ wq = bit_waitqueue(&xobject->flags, CACHEFILES_OBJECT_ACTIVE); init_wait(&wait); requeue = false; @@ -229,8 +228,8 @@ wait_for_old_object: prepare_to_wait(wq, &wait, TASK_UNINTERRUPTIBLE); if (!test_bit(CACHEFILES_OBJECT_ACTIVE, &xobject->flags)) break; - requeue = slow_work_sleep_till_thread_needed( - &object->fscache.work, &timeout); + + requeue = fscache_object_sleep_till_congested(&timeout); } while (timeout > 0 && !requeue); finish_wait(wq, &wait); diff --git a/fs/cachefiles/rdwr.c b/fs/cachefiles/rdwr.c index 0f0d41fbb03..0e3c0924cc3 100644 --- a/fs/cachefiles/rdwr.c +++ b/fs/cachefiles/rdwr.c @@ -422,7 +422,7 @@ int cachefiles_read_or_alloc_page(struct fscache_retrieval *op, shift = PAGE_SHIFT - inode->i_sb->s_blocksize_bits; op->op.flags &= FSCACHE_OP_KEEP_FLAGS; - op->op.flags |= FSCACHE_OP_FAST; + op->op.flags |= FSCACHE_OP_ASYNC; op->op.processor = cachefiles_read_copier; pagevec_init(&pagevec, 0); @@ -729,7 +729,7 @@ int cachefiles_read_or_alloc_pages(struct fscache_retrieval *op, pagevec_init(&pagevec, 0); op->op.flags &= FSCACHE_OP_KEEP_FLAGS; - op->op.flags |= FSCACHE_OP_FAST; + op->op.flags |= FSCACHE_OP_ASYNC; op->op.processor = cachefiles_read_copier; INIT_LIST_HEAD(&backpages); diff --git a/fs/char_dev.c b/fs/char_dev.c index d6db933df2b..f80a4f25123 100644 --- a/fs/char_dev.c +++ b/fs/char_dev.c @@ -20,6 +20,7 @@ #include <linux/cdev.h> #include <linux/mutex.h> #include <linux/backing-dev.h> +#include <linux/tty.h> #include "internal.h" diff --git a/fs/cifs/Kconfig b/fs/cifs/Kconfig index 80f35259680..917b7d449bb 100644 --- a/fs/cifs/Kconfig +++ b/fs/cifs/Kconfig @@ -2,7 +2,6 @@ config CIFS tristate "CIFS support (advanced network filesystem, SMBFS successor)" depends on INET select NLS - select SLOW_WORK help This is the client VFS module for the Common Internet File System (CIFS) protocol which is the successor to the Server Message Block @@ -71,14 +70,14 @@ config CIFS_WEAK_PW_HASH 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. + bool "Kerberos/SPNEGO advanced session setup" + depends on CIFS && KEYS + select DNS_RESOLVER + 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" @@ -122,6 +121,7 @@ config CIFS_DEBUG2 config CIFS_DFS_UPCALL bool "DFS feature support" depends on CIFS && KEYS + select DNS_RESOLVER help Distributed File System (DFS) support is used to access shares transparently in an enterprise name space, even if the share @@ -131,6 +131,15 @@ config CIFS_DFS_UPCALL IP addresses) which is needed for implicit mounts of DFS junction points. If unsure, say N. +config CIFS_FSCACHE + bool "Provide CIFS client caching support (EXPERIMENTAL)" + depends on EXPERIMENTAL + depends on CIFS=m && FSCACHE || CIFS=y && FSCACHE=y + help + Makes CIFS FS-Cache capable. Say Y here if you want your CIFS data + to be cached locally on disk through the general filesystem cache + manager. If unsure, say N. + config CIFS_EXPERIMENTAL bool "CIFS Experimental Features (EXPERIMENTAL)" depends on CIFS && EXPERIMENTAL diff --git a/fs/cifs/Makefile b/fs/cifs/Makefile index 9948c0030e8..adefa60a9bd 100644 --- a/fs/cifs/Makefile +++ b/fs/cifs/Makefile @@ -11,3 +11,5 @@ cifs-y := cifsfs.o cifssmb.o cifs_debug.o connect.o dir.o file.o inode.o \ cifs-$(CONFIG_CIFS_UPCALL) += cifs_spnego.o cifs-$(CONFIG_CIFS_DFS_UPCALL) += dns_resolve.o cifs_dfs_ref.o + +cifs-$(CONFIG_CIFS_FSCACHE) += fscache.o cache.o diff --git a/fs/cifs/README b/fs/cifs/README index a727b7cb075..a7081eeeb85 100644 --- a/fs/cifs/README +++ b/fs/cifs/README @@ -568,8 +568,9 @@ module can be displayed via modinfo. Misc /proc/fs/cifs Flags and Debug Info ======================================= Informational pseudo-files: -DebugData Displays information about active CIFS sessions - and shares, as well as the cifs.ko version. +DebugData Displays information about active CIFS sessions and + shares, features enabled as well as the cifs.ko + version. Stats Lists summary resource usage information as well as per share statistics, if CONFIG_CIFS_STATS in enabled in the kernel configuration. diff --git a/fs/cifs/cache.c b/fs/cifs/cache.c new file mode 100644 index 00000000000..224d7bbd1fc --- /dev/null +++ b/fs/cifs/cache.c @@ -0,0 +1,331 @@ +/* + * fs/cifs/cache.c - CIFS filesystem cache index structure definitions + * + * Copyright (c) 2010 Novell, Inc. + * Authors(s): Suresh Jayaraman (sjayaraman@suse.de> + * + * This library is free software; you can redistribute it and/or modify + * it under the terms of the GNU Lesser General Public License as published + * by the Free Software Foundation; either version 2.1 of the License, or + * (at your option) any later version. + * + * This library 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 Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with this library; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ +#include "fscache.h" +#include "cifs_debug.h" + +/* + * CIFS filesystem definition for FS-Cache + */ +struct fscache_netfs cifs_fscache_netfs = { + .name = "cifs", + .version = 0, +}; + +/* + * Register CIFS for caching with FS-Cache + */ +int cifs_fscache_register(void) +{ + return fscache_register_netfs(&cifs_fscache_netfs); +} + +/* + * Unregister CIFS for caching + */ +void cifs_fscache_unregister(void) +{ + fscache_unregister_netfs(&cifs_fscache_netfs); +} + +/* + * Key layout of CIFS server cache index object + */ +struct cifs_server_key { + uint16_t family; /* address family */ + uint16_t port; /* IP port */ + union { + struct in_addr ipv4_addr; + struct in6_addr ipv6_addr; + } addr[0]; +}; + +/* + * Server object keyed by {IPaddress,port,family} tuple + */ +static uint16_t cifs_server_get_key(const void *cookie_netfs_data, + void *buffer, uint16_t maxbuf) +{ + const struct TCP_Server_Info *server = cookie_netfs_data; + const struct sockaddr *sa = (struct sockaddr *) &server->addr.sockAddr; + struct cifs_server_key *key = buffer; + uint16_t key_len = sizeof(struct cifs_server_key); + + memset(key, 0, key_len); + + /* + * Should not be a problem as sin_family/sin6_family overlays + * sa_family field + */ + switch (sa->sa_family) { + case AF_INET: + key->family = server->addr.sockAddr.sin_family; + key->port = server->addr.sockAddr.sin_port; + key->addr[0].ipv4_addr = server->addr.sockAddr.sin_addr; + key_len += sizeof(key->addr[0].ipv4_addr); + break; + + case AF_INET6: + key->family = server->addr.sockAddr6.sin6_family; + key->port = server->addr.sockAddr6.sin6_port; + key->addr[0].ipv6_addr = server->addr.sockAddr6.sin6_addr; + key_len += sizeof(key->addr[0].ipv6_addr); + break; + + default: + cERROR(1, "CIFS: Unknown network family '%d'", sa->sa_family); + key_len = 0; + break; + } + + return key_len; +} + +/* + * Server object for FS-Cache + */ +const struct fscache_cookie_def cifs_fscache_server_index_def = { + .name = "CIFS.server", + .type = FSCACHE_COOKIE_TYPE_INDEX, + .get_key = cifs_server_get_key, +}; + +/* + * Auxiliary data attached to CIFS superblock within the cache + */ +struct cifs_fscache_super_auxdata { + u64 resource_id; /* unique server resource id */ +}; + +static char *extract_sharename(const char *treename) +{ + const char *src; + char *delim, *dst; + int len; + + /* skip double chars at the beginning */ + src = treename + 2; + + /* share name is always preceded by '\\' now */ + delim = strchr(src, '\\'); + if (!delim) + return ERR_PTR(-EINVAL); + delim++; + len = strlen(delim); + + /* caller has to free the memory */ + dst = kstrndup(delim, len, GFP_KERNEL); + if (!dst) + return ERR_PTR(-ENOMEM); + + return dst; +} + +/* + * Superblock object currently keyed by share name + */ +static uint16_t cifs_super_get_key(const void *cookie_netfs_data, void *buffer, + uint16_t maxbuf) +{ + const struct cifsTconInfo *tcon = cookie_netfs_data; + char *sharename; + uint16_t len; + + sharename = extract_sharename(tcon->treeName); + if (IS_ERR(sharename)) { + cFYI(1, "CIFS: couldn't extract sharename\n"); + sharename = NULL; + return 0; + } + + len = strlen(sharename); + if (len > maxbuf) + return 0; + + memcpy(buffer, sharename, len); + + kfree(sharename); + + return len; +} + +static uint16_t +cifs_fscache_super_get_aux(const void *cookie_netfs_data, void *buffer, + uint16_t maxbuf) +{ + struct cifs_fscache_super_auxdata auxdata; + const struct cifsTconInfo *tcon = cookie_netfs_data; + + memset(&auxdata, 0, sizeof(auxdata)); + auxdata.resource_id = tcon->resource_id; + + if (maxbuf > sizeof(auxdata)) + maxbuf = sizeof(auxdata); + + memcpy(buffer, &auxdata, maxbuf); + + return maxbuf; +} + +static enum +fscache_checkaux cifs_fscache_super_check_aux(void *cookie_netfs_data, + const void *data, + uint16_t datalen) +{ + struct cifs_fscache_super_auxdata auxdata; + const struct cifsTconInfo *tcon = cookie_netfs_data; + + if (datalen != sizeof(auxdata)) + return FSCACHE_CHECKAUX_OBSOLETE; + + memset(&auxdata, 0, sizeof(auxdata)); + auxdata.resource_id = tcon->resource_id; + + if (memcmp(data, &auxdata, datalen) != 0) + return FSCACHE_CHECKAUX_OBSOLETE; + + return FSCACHE_CHECKAUX_OKAY; +} + +/* + * Superblock object for FS-Cache + */ +const struct fscache_cookie_def cifs_fscache_super_index_def = { + .name = "CIFS.super", + .type = FSCACHE_COOKIE_TYPE_INDEX, + .get_key = cifs_super_get_key, + .get_aux = cifs_fscache_super_get_aux, + .check_aux = cifs_fscache_super_check_aux, +}; + +/* + * Auxiliary data attached to CIFS inode within the cache + */ +struct cifs_fscache_inode_auxdata { + struct timespec last_write_time; + struct timespec last_change_time; + u64 eof; +}; + +static uint16_t cifs_fscache_inode_get_key(const void *cookie_netfs_data, + void *buffer, uint16_t maxbuf) +{ + const struct cifsInodeInfo *cifsi = cookie_netfs_data; + uint16_t keylen; + + /* use the UniqueId as the key */ + keylen = sizeof(cifsi->uniqueid); + if (keylen > maxbuf) + keylen = 0; + else + memcpy(buffer, &cifsi->uniqueid, keylen); + + return keylen; +} + +static void +cifs_fscache_inode_get_attr(const void *cookie_netfs_data, uint64_t *size) +{ + const struct cifsInodeInfo *cifsi = cookie_netfs_data; + + *size = cifsi->vfs_inode.i_size; +} + +static uint16_t +cifs_fscache_inode_get_aux(const void *cookie_netfs_data, void *buffer, + uint16_t maxbuf) +{ + struct cifs_fscache_inode_auxdata auxdata; + const struct cifsInodeInfo *cifsi = cookie_netfs_data; + + memset(&auxdata, 0, sizeof(auxdata)); + auxdata.eof = cifsi->server_eof; + auxdata.last_write_time = cifsi->vfs_inode.i_mtime; + auxdata.last_change_time = cifsi->vfs_inode.i_ctime; + + if (maxbuf > sizeof(auxdata)) + maxbuf = sizeof(auxdata); + + memcpy(buffer, &auxdata, maxbuf); + + return maxbuf; +} + +static enum +fscache_checkaux cifs_fscache_inode_check_aux(void *cookie_netfs_data, + const void *data, + uint16_t datalen) +{ + struct cifs_fscache_inode_auxdata auxdata; + struct cifsInodeInfo *cifsi = cookie_netfs_data; + + if (datalen != sizeof(auxdata)) + return FSCACHE_CHECKAUX_OBSOLETE; + + memset(&auxdata, 0, sizeof(auxdata)); + auxdata.eof = cifsi->server_eof; + auxdata.last_write_time = cifsi->vfs_inode.i_mtime; + auxdata.last_change_time = cifsi->vfs_inode.i_ctime; + + if (memcmp(data, &auxdata, datalen) != 0) + return FSCACHE_CHECKAUX_OBSOLETE; + + return FSCACHE_CHECKAUX_OKAY; +} + +static void cifs_fscache_inode_now_uncached(void *cookie_netfs_data) +{ + struct cifsInodeInfo *cifsi = cookie_netfs_data; + struct pagevec pvec; + pgoff_t first; + int loop, nr_pages; + + pagevec_init(&pvec, 0); + first = 0; + + cFYI(1, "cifs inode 0x%p now uncached", cifsi); + + for (;;) { + nr_pages = pagevec_lookup(&pvec, + cifsi->vfs_inode.i_mapping, first, + PAGEVEC_SIZE - pagevec_count(&pvec)); + if (!nr_pages) + break; + + for (loop = 0; loop < nr_pages; loop++) + ClearPageFsCache(pvec.pages[loop]); + + first = pvec.pages[nr_pages - 1]->index + 1; + + pvec.nr = nr_pages; + pagevec_release(&pvec); + cond_resched(); + } +} + +const struct fscache_cookie_def cifs_fscache_inode_object_def = { + .name = "CIFS.uniqueid", + .type = FSCACHE_COOKIE_TYPE_DATAFILE, + .get_key = cifs_fscache_inode_get_key, + .get_attr = cifs_fscache_inode_get_attr, + .get_aux = cifs_fscache_inode_get_aux, + .check_aux = cifs_fscache_inode_check_aux, + .now_uncached = cifs_fscache_inode_now_uncached, +}; diff --git a/fs/cifs/cifs_debug.c b/fs/cifs/cifs_debug.c index 4fce6e61b34..eb1ba493489 100644 --- a/fs/cifs/cifs_debug.c +++ b/fs/cifs/cifs_debug.c @@ -119,6 +119,31 @@ static int cifs_debug_data_proc_show(struct seq_file *m, void *v) "Display Internal CIFS Data Structures for Debugging\n" "---------------------------------------------------\n"); seq_printf(m, "CIFS Version %s\n", CIFS_VERSION); + seq_printf(m, "Features: "); +#ifdef CONFIG_CIFS_DFS_UPCALL + seq_printf(m, "dfs"); + seq_putc(m, ' '); +#endif +#ifdef CONFIG_CIFS_FSCACHE + seq_printf(m, "fscache"); + seq_putc(m, ' '); +#endif +#ifdef CONFIG_CIFS_WEAK_PW_HASH + seq_printf(m, "lanman"); + seq_putc(m, ' '); +#endif +#ifdef CONFIG_CIFS_POSIX + seq_printf(m, "posix"); + seq_putc(m, ' '); +#endif +#ifdef CONFIG_CIFS_UPCALL + seq_printf(m, "spnego"); + seq_putc(m, ' '); +#endif +#ifdef CONFIG_CIFS_XATTR + seq_printf(m, "xattr"); +#endif + seq_putc(m, '\n'); seq_printf(m, "Active VFS Requests: %d\n", GlobalTotalActiveXid); seq_printf(m, "Servers:"); diff --git a/fs/cifs/cifs_dfs_ref.c b/fs/cifs/cifs_dfs_ref.c index ac19a6f3dae..d6ced7aa23c 100644 --- a/fs/cifs/cifs_dfs_ref.c +++ b/fs/cifs/cifs_dfs_ref.c @@ -141,7 +141,7 @@ char *cifs_compose_mount_options(const char *sb_mountdata, } rc = dns_resolve_server_name_to_ip(*devname, &srvIP); - if (rc != 0) { + if (rc < 0) { cERROR(1, "%s: Failed to resolve server part of %s to IP: %d", __func__, *devname, rc); goto compose_mount_options_err; @@ -150,8 +150,7 @@ char *cifs_compose_mount_options(const char *sb_mountdata, * assuming that we have 'unc=' and 'ip=' in * the original sb_mountdata */ - md_len = strlen(sb_mountdata) + strlen(srvIP) + - strlen(ref->node_name) + 12; + md_len = strlen(sb_mountdata) + rc + strlen(ref->node_name) + 12; mountdata = kzalloc(md_len+1, GFP_KERNEL); if (mountdata == NULL) { rc = -ENOMEM; @@ -230,28 +229,22 @@ compose_mount_options_err: goto compose_mount_options_out; } - -static struct vfsmount *cifs_dfs_do_refmount(const struct vfsmount *mnt_parent, - struct dentry *dentry, const struct dfs_info3_param *ref) +/** + * cifs_dfs_do_refmount - mounts specified path using provided refferal + * @cifs_sb: parent/root superblock + * @fullpath: full path in UNC format + * @ref: server's referral + */ +static struct vfsmount *cifs_dfs_do_refmount(struct cifs_sb_info *cifs_sb, + const char *fullpath, const struct dfs_info3_param *ref) { - struct cifs_sb_info *cifs_sb; struct vfsmount *mnt; char *mountdata; char *devname = NULL; - char *fullpath; - - cifs_sb = CIFS_SB(dentry->d_inode->i_sb); - /* - * this function gives us a path with a double backslash prefix. We - * require a single backslash for DFS. - */ - fullpath = build_path_from_dentry(dentry); - if (!fullpath) - return ERR_PTR(-ENOMEM); + /* strip first '\' from fullpath */ mountdata = cifs_compose_mount_options(cifs_sb->mountdata, fullpath + 1, ref, &devname); - kfree(fullpath); if (IS_ERR(mountdata)) return (struct vfsmount *)mountdata; @@ -357,8 +350,8 @@ cifs_dfs_follow_mountpoint(struct dentry *dentry, struct nameidata *nd) rc = -EINVAL; goto out_err; } - mnt = cifs_dfs_do_refmount(nd->path.mnt, - nd->path.dentry, referrals + i); + mnt = cifs_dfs_do_refmount(cifs_sb, + full_path, 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 246a167cb91..9e771450c3b 100644 --- a/fs/cifs/cifs_fs_sb.h +++ b/fs/cifs/cifs_fs_sb.h @@ -35,6 +35,7 @@ #define CIFS_MOUNT_DYNPERM 0x1000 /* allow in-memory only mode setting */ #define CIFS_MOUNT_NOPOSIXBRL 0x2000 /* mandatory not posix byte range lock */ #define CIFS_MOUNT_NOSSYNC 0x4000 /* don't do slow SMBflush on every sync*/ +#define CIFS_MOUNT_FSCACHE 0x8000 /* local caching enabled */ struct cifs_sb_info { struct cifsTconInfo *tcon; /* primary mount */ diff --git a/fs/cifs/cifs_spnego.c b/fs/cifs/cifs_spnego.c index 379bd7d9c05..87044906cd1 100644 --- a/fs/cifs/cifs_spnego.c +++ b/fs/cifs/cifs_spnego.c @@ -84,6 +84,9 @@ struct key_type cifs_spnego_key_type = { /* strlen of ";uid=0x" */ #define UID_KEY_LEN 7 +/* strlen of ";creduid=0x" */ +#define CREDUID_KEY_LEN 11 + /* strlen of ";user=" */ #define USER_KEY_LEN 6 @@ -107,6 +110,7 @@ cifs_get_spnego_key(struct cifsSesInfo *sesInfo) IP_KEY_LEN + INET6_ADDRSTRLEN + MAX_MECH_STR_LEN + UID_KEY_LEN + (sizeof(uid_t) * 2) + + CREDUID_KEY_LEN + (sizeof(uid_t) * 2) + USER_KEY_LEN + strlen(sesInfo->userName) + PID_KEY_LEN + (sizeof(pid_t) * 2) + 1; @@ -144,6 +148,9 @@ cifs_get_spnego_key(struct cifsSesInfo *sesInfo) sprintf(dp, ";uid=0x%x", sesInfo->linux_uid); dp = description + strlen(description); + sprintf(dp, ";creduid=0x%x", sesInfo->cred_uid); + + dp = description + strlen(description); sprintf(dp, ";user=%s", sesInfo->userName); dp = description + strlen(description); diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c index 2cb1a70214d..a5ed10c9afe 100644 --- a/fs/cifs/cifsfs.c +++ b/fs/cifs/cifsfs.c @@ -45,8 +45,8 @@ #include "cifs_fs_sb.h" #include <linux/mm.h> #include <linux/key-type.h> -#include "dns_resolve.h" #include "cifs_spnego.h" +#include "fscache.h" #define CIFS_MAGIC_NUMBER 0xFF534D42 /* the first four bytes of SMB PDUs */ int cifsFYI = 0; @@ -329,6 +329,12 @@ cifs_destroy_inode(struct inode *inode) } static void +cifs_clear_inode(struct inode *inode) +{ + cifs_fscache_release_inode_cookie(inode); +} + +static void cifs_show_address(struct seq_file *s, struct TCP_Server_Info *server) { seq_printf(s, ",addr="); @@ -489,6 +495,7 @@ static const struct super_operations cifs_super_ops = { .alloc_inode = cifs_alloc_inode, .destroy_inode = cifs_destroy_inode, .drop_inode = cifs_drop_inode, + .clear_inode = cifs_clear_inode, /* .delete_inode = cifs_delete_inode, */ /* Do not need above function unless later we add lazy close of inodes or unless the kernel forgets to call us with the same number of releases (closes) @@ -902,6 +909,10 @@ init_cifs(void) cFYI(1, "cifs_max_pending set to max of 256"); } + rc = cifs_fscache_register(); + if (rc) + goto out; + rc = cifs_init_inodecache(); if (rc) goto out_clean_proc; @@ -922,27 +933,13 @@ init_cifs(void) if (rc) goto out_unregister_filesystem; #endif -#ifdef CONFIG_CIFS_DFS_UPCALL - rc = cifs_init_dns_resolver(); - if (rc) - goto out_unregister_key_type; -#endif - rc = slow_work_register_user(THIS_MODULE); - if (rc) - goto out_unregister_resolver_key; return 0; - out_unregister_resolver_key: -#ifdef CONFIG_CIFS_DFS_UPCALL - cifs_exit_dns_resolver(); - out_unregister_key_type: -#endif #ifdef CONFIG_CIFS_UPCALL - unregister_key_type(&cifs_spnego_key_type); out_unregister_filesystem: -#endif unregister_filesystem(&cifs_fs_type); +#endif out_destroy_request_bufs: cifs_destroy_request_bufs(); out_destroy_mids: @@ -951,6 +948,8 @@ init_cifs(void) cifs_destroy_inodecache(); out_clean_proc: cifs_proc_clean(); + cifs_fscache_unregister(); + out: return rc; } @@ -959,9 +958,9 @@ exit_cifs(void) { cFYI(DBG2, "exit_cifs"); cifs_proc_clean(); + cifs_fscache_unregister(); #ifdef CONFIG_CIFS_DFS_UPCALL cifs_dfs_release_automount_timer(); - cifs_exit_dns_resolver(); #endif #ifdef CONFIG_CIFS_UPCALL unregister_key_type(&cifs_spnego_key_type); diff --git a/fs/cifs/cifsfs.h b/fs/cifs/cifsfs.h index a7eb65c84b1..d82f5fb4761 100644 --- a/fs/cifs/cifsfs.h +++ b/fs/cifs/cifsfs.h @@ -114,5 +114,5 @@ 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 "1.64" +#define CIFS_VERSION "1.65" #endif /* _CIFSFS_H */ diff --git a/fs/cifs/cifsglob.h b/fs/cifs/cifsglob.h index a88479ceaad..0cdfb8c32ac 100644 --- a/fs/cifs/cifsglob.h +++ b/fs/cifs/cifsglob.h @@ -16,10 +16,13 @@ * the GNU Lesser General Public License for more details. * */ +#ifndef _CIFS_GLOB_H +#define _CIFS_GLOB_H + #include <linux/in.h> #include <linux/in6.h> #include <linux/slab.h> -#include <linux/slow-work.h> +#include <linux/workqueue.h> #include "cifs_fs_sb.h" #include "cifsacl.h" /* @@ -34,7 +37,7 @@ #define MAX_SHARE_SIZE 64 /* used to be 20, this should still be enough */ #define MAX_USERNAME_SIZE 32 /* 32 is to allow for 15 char names + null termination then *2 for unicode versions */ -#define MAX_PASSWORD_SIZE 16 +#define MAX_PASSWORD_SIZE 512 /* max for windows seems to be 256 wide chars */ #define CIFS_MIN_RCV_POOL 4 @@ -80,8 +83,7 @@ enum statusEnum { }; enum securityEnum { - PLAINTXT = 0, /* Legacy with Plaintext passwords */ - LANMAN, /* Legacy LANMAN auth */ + LANMAN = 0, /* Legacy LANMAN auth */ NTLM, /* Legacy NTLM012 auth with NTLM hash */ NTLMv2, /* Legacy NTLM auth with NTLMv2 hash */ RawNTLMSSP, /* NTLMSSP without SPNEGO, NTLMv2 hash */ @@ -142,7 +144,6 @@ struct TCP_Server_Info { struct list_head pending_mid_q; void *Server_NlsInfo; /* BB - placeholder for future NLS info */ unsigned short server_codepage; /* codepage for the server */ - unsigned long ip_address; /* IP addr for the server if known */ enum protocolEnum protocolType; char versionMajor; char versionMinor; @@ -190,19 +191,9 @@ struct TCP_Server_Info { bool sec_mskerberos; /* supports legacy MS Kerberos */ bool sec_kerberosu2u; /* supports U2U Kerberos */ bool sec_ntlmssp; /* supports NTLMSSP */ -}; - -/* - * The following is our shortcut to user information. We surface the uid, - * and name. We always get the password on the fly in case it - * has changed. We also hang a list of sessions owned by this user off here. - */ -struct cifsUidInfo { - struct list_head userList; - struct list_head sessionList; /* SMB sessions for this user */ - uid_t linux_uid; - char user[MAX_USERNAME_SIZE + 1]; /* ascii name of user */ - /* BB may need ptr or callback for PAM or WinBind info */ +#ifdef CONFIG_CIFS_FSCACHE + struct fscache_cookie *fscache; /* client index cache cookie */ +#endif }; /* @@ -212,9 +203,6 @@ struct cifsSesInfo { struct list_head smb_ses_list; struct list_head tcon_list; struct mutex session_mutex; -#if 0 - struct cifsUidInfo *uidInfo; /* pointer to user info */ -#endif struct TCP_Server_Info *server; /* pointer to server info */ int ses_count; /* reference counter */ enum statusEnum status; @@ -226,7 +214,8 @@ struct cifsSesInfo { char *serverNOS; /* name of network operating system of server */ char *serverDomain; /* security realm of server */ int Suid; /* remote smb uid */ - uid_t linux_uid; /* local Linux uid */ + uid_t linux_uid; /* overriding owner of files on the mount */ + uid_t cred_uid; /* owner of credentials */ int capabilities; char serverName[SERVER_NAME_LEN_WITH_NULL * 2]; /* BB make bigger for TCP names - will ipv6 and sctp addresses fit? */ @@ -311,6 +300,10 @@ struct cifsTconInfo { bool local_lease:1; /* check leases (only) on local system not remote */ bool broken_posix_open; /* e.g. Samba server versions < 3.3.2, 3.2.9 */ bool need_reconnect:1; /* connection reset, tid now invalid */ +#ifdef CONFIG_CIFS_FSCACHE + u64 resource_id; /* server resource id */ + struct fscache_cookie *fscache; /* cookie for share */ +#endif /* BB add field for back pointer to sb struct(s)? */ }; @@ -363,7 +356,7 @@ struct cifsFileInfo { atomic_t count; /* reference count */ struct mutex fh_mutex; /* prevents reopen race after dead ses*/ struct cifs_search_info srch_inf; - struct slow_work oplock_break; /* slow_work job for oplock breaks */ + struct work_struct oplock_break; /* work for oplock breaks */ }; /* Take a reference on the file private data */ @@ -398,6 +391,9 @@ struct cifsInodeInfo { bool invalid_mapping:1; /* pagecache is invalid */ u64 server_eof; /* current file size on server */ u64 uniqueid; /* server inode number */ +#ifdef CONFIG_CIFS_FSCACHE + struct fscache_cookie *fscache; +#endif struct inode vfs_inode; }; @@ -732,4 +728,10 @@ GLOBAL_EXTERN unsigned int cifs_min_rcv; /* min size of big ntwrk buf pool */ GLOBAL_EXTERN unsigned int cifs_min_small; /* min size of small buf pool */ GLOBAL_EXTERN unsigned int cifs_max_pending; /* MAX requests at once to server*/ +void cifs_oplock_break(struct work_struct *work); +void cifs_oplock_break_get(struct cifsFileInfo *cfile); +void cifs_oplock_break_put(struct cifsFileInfo *cfile); + extern const struct slow_work_ops cifs_oplock_break_ops; + +#endif /* _CIFS_GLOB_H */ diff --git a/fs/cifs/cifsproto.h b/fs/cifs/cifsproto.h index fb6318b8150..1f545081408 100644 --- a/fs/cifs/cifsproto.h +++ b/fs/cifs/cifsproto.h @@ -86,7 +86,9 @@ extern unsigned int smbCalcSize(struct smb_hdr *ptr); extern unsigned int smbCalcSize_LE(struct smb_hdr *ptr); extern int decode_negTokenInit(unsigned char *security_blob, int length, struct TCP_Server_Info *server); -extern int cifs_convert_address(char *src, void *dst); +extern int cifs_convert_address(struct sockaddr *dst, const char *src, int len); +extern int cifs_fill_sockaddr(struct sockaddr *dst, const char *src, int len, + unsigned short int port); extern int map_smb_to_linux_error(struct smb_hdr *smb, int logErr); extern void header_assemble(struct smb_hdr *, char /* command */ , const struct cifsTconInfo *, int /* length of diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index 2208f06e4c4..95c2ea67edf 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -48,6 +48,7 @@ #include "nterr.h" #include "rfc1002pdu.h" #include "cn_cifs.h" +#include "fscache.h" #define CIFS_PORT 445 #define RFC1001_PORT 139 @@ -66,6 +67,7 @@ struct smb_vol { char *iocharset; /* local code page for mapping to and from Unicode */ char source_rfc1001_name[16]; /* netbios name of client */ char target_rfc1001_name[16]; /* netbios name of server for Win9x/ME */ + uid_t cred_uid; uid_t linux_uid; gid_t linux_gid; mode_t file_mode; @@ -97,6 +99,7 @@ struct smb_vol { bool noblocksnd:1; bool noautotune:1; bool nostrictsync:1; /* do not force expensive SMBflush on every sync */ + bool fsc:1; /* enable fscache */ unsigned int rsize; unsigned int wsize; bool sockopt_tcp_nodelay:1; @@ -830,7 +833,8 @@ 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(); /* use current_euid() instead? */ + vol->cred_uid = current_uid(); + vol->linux_uid = current_uid(); vol->linux_gid = current_gid(); /* default to only allowing write access to owner of the mount */ @@ -1257,6 +1261,12 @@ cifs_parse_mount_options(char *options, const char *devname, } else if ((strnicmp(data, "nocase", 6) == 0) || (strnicmp(data, "ignorecase", 10) == 0)) { vol->nocase = 1; + } else if (strnicmp(data, "mand", 4) == 0) { + /* ignore */ + } else if (strnicmp(data, "nomand", 6) == 0) { + /* ignore */ + } else if (strnicmp(data, "_netdev", 7) == 0) { + /* ignore */ } else if (strnicmp(data, "brl", 3) == 0) { vol->nobrl = 0; } else if ((strnicmp(data, "nobrl", 5) == 0) || @@ -1331,6 +1341,8 @@ cifs_parse_mount_options(char *options, const char *devname, printk(KERN_WARNING "CIFS: Mount option noac not " "supported. Instead set " "/proc/fs/cifs/LookupCacheEnabled to 0\n"); + } else if (strnicmp(data, "fsc", 3) == 0) { + vol->fsc = true; } else printk(KERN_WARNING "CIFS: Unknown mount option %s\n", data); @@ -1380,18 +1392,92 @@ cifs_parse_mount_options(char *options, const char *devname, return 0; } +static bool +match_address(struct TCP_Server_Info *server, struct sockaddr *addr) +{ + struct sockaddr_in *addr4 = (struct sockaddr_in *)addr; + struct sockaddr_in6 *addr6 = (struct sockaddr_in6 *)addr; + + switch (addr->sa_family) { + case AF_INET: + if (addr4->sin_addr.s_addr != + server->addr.sockAddr.sin_addr.s_addr) + return false; + if (addr4->sin_port && + addr4->sin_port != server->addr.sockAddr.sin_port) + return false; + break; + case AF_INET6: + if (!ipv6_addr_equal(&addr6->sin6_addr, + &server->addr.sockAddr6.sin6_addr)) + return false; + if (addr6->sin6_scope_id != + server->addr.sockAddr6.sin6_scope_id) + return false; + if (addr6->sin6_port && + addr6->sin6_port != server->addr.sockAddr6.sin6_port) + return false; + break; + } + + return true; +} + +static bool +match_security(struct TCP_Server_Info *server, struct smb_vol *vol) +{ + unsigned int secFlags; + + if (vol->secFlg & (~(CIFSSEC_MUST_SIGN | CIFSSEC_MUST_SEAL))) + secFlags = vol->secFlg; + else + secFlags = global_secflags | vol->secFlg; + + switch (server->secType) { + case LANMAN: + if (!(secFlags & (CIFSSEC_MAY_LANMAN|CIFSSEC_MAY_PLNTXT))) + return false; + break; + case NTLMv2: + if (!(secFlags & CIFSSEC_MAY_NTLMV2)) + return false; + break; + case NTLM: + if (!(secFlags & CIFSSEC_MAY_NTLM)) + return false; + break; + case Kerberos: + if (!(secFlags & CIFSSEC_MAY_KRB5)) + return false; + break; + case RawNTLMSSP: + if (!(secFlags & CIFSSEC_MAY_NTLMSSP)) + return false; + break; + default: + /* shouldn't happen */ + return false; + } + + /* now check if signing mode is acceptible */ + if ((secFlags & CIFSSEC_MAY_SIGN) == 0 && + (server->secMode & SECMODE_SIGN_REQUIRED)) + return false; + else if (((secFlags & CIFSSEC_MUST_SIGN) == CIFSSEC_MUST_SIGN) && + (server->secMode & + (SECMODE_SIGN_ENABLED|SECMODE_SIGN_REQUIRED)) == 0) + return false; + + return true; +} + static struct TCP_Server_Info * -cifs_find_tcp_session(struct sockaddr_storage *addr, unsigned short int port) +cifs_find_tcp_session(struct sockaddr *addr, struct smb_vol *vol) { - struct list_head *tmp; 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); + list_for_each_entry(server, &cifs_tcp_ses_list, 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 @@ -1401,37 +1487,11 @@ cifs_find_tcp_session(struct sockaddr_storage *addr, unsigned short int port) if (server->tcpStatus == CifsNew) continue; - switch (addr->ss_family) { - case AF_INET: - if (addr4->sin_addr.s_addr == - server->addr.sockAddr.sin_addr.s_addr) { - addr4->sin_port = htons(port); - /* user overrode default port? */ - if (addr4->sin_port) { - if (addr4->sin_port != - server->addr.sockAddr.sin_port) - continue; - } - break; - } else - continue; + if (!match_address(server, addr)) + continue; - case AF_INET6: - if (ipv6_addr_equal(&addr6->sin6_addr, - &server->addr.sockAddr6.sin6_addr) && - (addr6->sin6_scope_id == - server->addr.sockAddr6.sin6_scope_id)) { - addr6->sin6_port = htons(port); - /* user overrode default port? */ - if (addr6->sin6_port) { - if (addr6->sin6_port != - server->addr.sockAddr6.sin6_port) - continue; - } - break; - } else - continue; - } + if (!match_security(server, vol)) + continue; ++server->srv_count; write_unlock(&cifs_tcp_ses_lock); @@ -1460,6 +1520,8 @@ cifs_put_tcp_session(struct TCP_Server_Info *server) server->tcpStatus = CifsExiting; spin_unlock(&GlobalMid_Lock); + cifs_fscache_release_client_cookie(server); + task = xchg(&server->tsk, NULL); if (task) force_sig(SIGKILL, task); @@ -1479,7 +1541,10 @@ cifs_get_tcp_session(struct smb_vol *volume_info) cFYI(1, "UNC: %s ip: %s", volume_info->UNC, volume_info->UNCip); if (volume_info->UNCip && volume_info->UNC) { - rc = cifs_convert_address(volume_info->UNCip, &addr); + rc = cifs_fill_sockaddr((struct sockaddr *)&addr, + volume_info->UNCip, + strlen(volume_info->UNCip), + volume_info->port); if (!rc) { /* we failed translating address */ rc = -EINVAL; @@ -1499,7 +1564,7 @@ cifs_get_tcp_session(struct smb_vol *volume_info) } /* see if we already have a matching tcp_ses */ - tcp_ses = cifs_find_tcp_session(&addr, volume_info->port); + tcp_ses = cifs_find_tcp_session((struct sockaddr *)&addr, volume_info); if (tcp_ses) return tcp_ses; @@ -1543,12 +1608,10 @@ cifs_get_tcp_session(struct smb_vol *volume_info) cFYI(1, "attempting ipv6 connect"); /* BB should we allow ipv6 on port 139? */ /* other OS never observed in Wild doing 139 with v6 */ - sin_server6->sin6_port = htons(volume_info->port); memcpy(&tcp_ses->addr.sockAddr6, sin_server6, sizeof(struct sockaddr_in6)); rc = ipv6_connect(tcp_ses); } else { - sin_server->sin_port = htons(volume_info->port); memcpy(&tcp_ses->addr.sockAddr, sin_server, sizeof(struct sockaddr_in)); rc = ipv4_connect(tcp_ses); @@ -1577,6 +1640,8 @@ cifs_get_tcp_session(struct smb_vol *volume_info) list_add(&tcp_ses->tcp_ses_list, &cifs_tcp_ses_list); write_unlock(&cifs_tcp_ses_lock); + cifs_fscache_get_client_cookie(tcp_ses); + return tcp_ses; out_err: @@ -1591,17 +1656,27 @@ out_err: } static struct cifsSesInfo * -cifs_find_smb_ses(struct TCP_Server_Info *server, char *username) +cifs_find_smb_ses(struct TCP_Server_Info *server, struct smb_vol *vol) { - struct list_head *tmp; struct cifsSesInfo *ses; 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; - + list_for_each_entry(ses, &server->smb_ses_list, smb_ses_list) { + switch (server->secType) { + case Kerberos: + if (vol->cred_uid != ses->cred_uid) + continue; + break; + default: + /* anything else takes username/password */ + if (strncmp(ses->userName, vol->username, + MAX_USERNAME_SIZE)) + continue; + if (strlen(vol->username) != 0 && + strncmp(ses->password, vol->password, + MAX_PASSWORD_SIZE)) + continue; + } ++ses->ses_count; write_unlock(&cifs_tcp_ses_lock); return ses; @@ -1643,7 +1718,7 @@ cifs_get_smb_ses(struct TCP_Server_Info *server, struct smb_vol *volume_info) xid = GetXid(); - ses = cifs_find_smb_ses(server, volume_info->username); + ses = cifs_find_smb_ses(server, volume_info); if (ses) { cFYI(1, "Existing smb sess found (status=%d)", ses->status); @@ -1706,6 +1781,7 @@ cifs_get_smb_ses(struct TCP_Server_Info *server, struct smb_vol *volume_info) if (ses->domainName) strcpy(ses->domainName, volume_info->domainname); } + ses->cred_uid = volume_info->cred_uid; ses->linux_uid = volume_info->linux_uid; ses->overrideSecFlg = volume_info->secFlg; @@ -1773,6 +1849,7 @@ cifs_put_tcon(struct cifsTconInfo *tcon) CIFSSMBTDis(xid, tcon); _FreeXid(xid); + cifs_fscache_release_super_cookie(tcon); tconInfoFree(tcon); cifs_put_smb_ses(ses); } @@ -1843,6 +1920,8 @@ cifs_get_tcon(struct cifsSesInfo *ses, struct smb_vol *volume_info) list_add(&tcon->tcon_list, &ses->tcon_list); write_unlock(&cifs_tcp_ses_lock); + cifs_fscache_get_super_cookie(tcon); + return tcon; out_fail: @@ -2397,6 +2476,8 @@ static void setup_cifs_sb(struct smb_vol *pvolume_info, cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_OVERR_GID; if (pvolume_info->dynperm) cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_DYNPERM; + if (pvolume_info->fsc) + cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_FSCACHE; if (pvolume_info->direct_io) { cFYI(1, "mounting share using direct i/o"); cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_DIRECT_IO; diff --git a/fs/cifs/dir.c b/fs/cifs/dir.c index e7ae78b66fa..578d88c5b46 100644 --- a/fs/cifs/dir.c +++ b/fs/cifs/dir.c @@ -130,12 +130,6 @@ cifs_bp_rename_retry: return full_path; } -/* - * When called with struct file pointer set to NULL, there is no way we could - * update file->private_data, but getting it stuck on openFileList provides a - * way to access it from cifs_fill_filedata and thereby set file->private_data - * from cifs_open. - */ struct cifsFileInfo * cifs_new_fileinfo(struct inode *newinode, __u16 fileHandle, struct file *file, struct vfsmount *mnt, unsigned int oflags) @@ -163,7 +157,7 @@ cifs_new_fileinfo(struct inode *newinode, __u16 fileHandle, mutex_init(&pCifsFile->lock_mutex); INIT_LIST_HEAD(&pCifsFile->llist); atomic_set(&pCifsFile->count, 1); - slow_work_init(&pCifsFile->oplock_break, &cifs_oplock_break_ops); + INIT_WORK(&pCifsFile->oplock_break, cifs_oplock_break); write_lock(&GlobalSMBSeslock); list_add(&pCifsFile->tlist, &cifs_sb->tcon->openFileList); diff --git a/fs/cifs/dns_resolve.c b/fs/cifs/dns_resolve.c index 853a968e82d..0eb87026cad 100644 --- a/fs/cifs/dns_resolve.c +++ b/fs/cifs/dns_resolve.c @@ -4,6 +4,8 @@ * Copyright (c) 2007 Igor Mammedov * Author(s): Igor Mammedov (niallain@gmail.com) * Steve French (sfrench@us.ibm.com) + * Wang Lei (wang840925@gmail.com) + * David Howells (dhowells@redhat.com) * * Contains the CIFS DFS upcall routines used for hostname to * IP address translation. @@ -24,214 +26,73 @@ */ #include <linux/slab.h> -#include <linux/keyctl.h> -#include <linux/key-type.h> -#include <keys/user-type.h> +#include <linux/dns_resolver.h> #include "dns_resolve.h" #include "cifsglob.h" #include "cifsproto.h" #include "cifs_debug.h" -static const struct cred *dns_resolver_cache; - -/* Checks if supplied name is IP address - * returns: - * 1 - name is IP - * 0 - name is not IP - */ -static int -is_ip(char *name) -{ - struct sockaddr_storage ss; - - return cifs_convert_address(name, &ss); -} - -static int -dns_resolver_instantiate(struct key *key, const void *data, - size_t datalen) -{ - int rc = 0; - char *ip; - - ip = kmalloc(datalen + 1, GFP_KERNEL); - if (!ip) - return -ENOMEM; - - memcpy(ip, data, datalen); - ip[datalen] = '\0'; - - /* make sure this looks like an address */ - if (!is_ip(ip)) { - kfree(ip); - return -EINVAL; - } - - key->type_data.x[0] = datalen; - key->payload.data = ip; - - return rc; -} - -static void -dns_resolver_destroy(struct key *key) -{ - kfree(key->payload.data); -} - -struct key_type key_type_dns_resolver = { - .name = "dns_resolver", - .def_datalen = sizeof(struct in_addr), - .describe = user_describe, - .instantiate = dns_resolver_instantiate, - .destroy = dns_resolver_destroy, - .match = user_match, -}; - -/* Resolves server name to ip address. - * input: - * unc - server UNC - * output: - * *ip_addr - pointer to server ip, caller responcible for freeing it. - * return 0 on success +/** + * dns_resolve_server_name_to_ip - Resolve UNC server name to ip address. + * @unc: UNC path specifying the server + * @ip_addr: Where to return the IP address. + * + * The IP address will be returned in string form, and the caller is + * responsible for freeing it. + * + * Returns length of result on success, -ve on error. */ int dns_resolve_server_name_to_ip(const char *unc, char **ip_addr) { - const struct cred *saved_cred; - int rc = -EAGAIN; - struct key *rkey = ERR_PTR(-EAGAIN); + struct sockaddr_storage ss; + const char *hostname, *sep; char *name; - char *data = NULL; - int len; + int len, rc; if (!ip_addr || !unc) return -EINVAL; - /* search for server name delimiter */ len = strlen(unc); if (len < 3) { cFYI(1, "%s: unc is too short: %s", __func__, unc); return -EINVAL; } - len -= 2; - name = memchr(unc+2, '\\', len); - if (!name) { - cFYI(1, "%s: probably server name is whole unc: %s", - __func__, unc); - } else { - len = (name - unc) - 2/* leading // */; - } - - name = kmalloc(len+1, GFP_KERNEL); - if (!name) { - rc = -ENOMEM; - return rc; - } - memcpy(name, unc+2, len); - name[len] = 0; - - if (is_ip(name)) { - cFYI(1, "%s: it is IP, skipping dns upcall: %s", - __func__, name); - data = name; - goto skip_upcall; - } - saved_cred = override_creds(dns_resolver_cache); - rkey = request_key(&key_type_dns_resolver, name, ""); - revert_creds(saved_cred); - if (!IS_ERR(rkey)) { - if (!(rkey->perm & KEY_USR_VIEW)) { - down_read(&rkey->sem); - rkey->perm |= KEY_USR_VIEW; - up_read(&rkey->sem); - } - len = rkey->type_data.x[0]; - data = rkey->payload.data; - } else { - cERROR(1, "%s: unable to resolve: %s", __func__, name); - goto out; - } - -skip_upcall: - if (data) { - *ip_addr = kmalloc(len + 1, GFP_KERNEL); - if (*ip_addr) { - memcpy(*ip_addr, data, len + 1); - if (!IS_ERR(rkey)) - cFYI(1, "%s: resolved: %s to %s", __func__, - name, - *ip_addr - ); - rc = 0; - } else { - rc = -ENOMEM; - } - if (!IS_ERR(rkey)) - key_put(rkey); - } + /* Discount leading slashes for cifs */ + len -= 2; + hostname = unc + 2; -out: - kfree(name); + /* Search for server name delimiter */ + sep = memchr(hostname, '\\', len); + if (sep) + len = sep - unc; + else + cFYI(1, "%s: probably server name is whole unc: %s", + __func__, unc); + + /* Try to interpret hostname as an IPv4 or IPv6 address */ + rc = cifs_convert_address((struct sockaddr *)&ss, hostname, len); + if (rc > 0) + goto name_is_IP_address; + + /* Perform the upcall */ + rc = dns_query(NULL, hostname, len, NULL, ip_addr, NULL); + if (rc < 0) + cERROR(1, "%s: unable to resolve: %*.*s", + __func__, len, len, hostname); + else + cFYI(1, "%s: resolved: %*.*s to %s", + __func__, len, len, hostname, *ip_addr); return rc; -} -int __init cifs_init_dns_resolver(void) -{ - struct cred *cred; - struct key *keyring; - int ret; - - printk(KERN_NOTICE "Registering the %s key type\n", - key_type_dns_resolver.name); - - /* create an override credential set with a special thread keyring in - * which DNS requests are cached - * - * this is used to prevent malicious redirections from being installed - * with add_key(). - */ - cred = prepare_kernel_cred(NULL); - if (!cred) +name_is_IP_address: + name = kmalloc(len + 1, GFP_KERNEL); + if (!name) return -ENOMEM; - - keyring = key_alloc(&key_type_keyring, ".dns_resolver", 0, 0, cred, - (KEY_POS_ALL & ~KEY_POS_SETATTR) | - KEY_USR_VIEW | KEY_USR_READ, - KEY_ALLOC_NOT_IN_QUOTA); - if (IS_ERR(keyring)) { - ret = PTR_ERR(keyring); - goto failed_put_cred; - } - - ret = key_instantiate_and_link(keyring, NULL, 0, NULL, NULL); - if (ret < 0) - goto failed_put_key; - - ret = register_key_type(&key_type_dns_resolver); - if (ret < 0) - goto failed_put_key; - - /* instruct request_key() to use this special keyring as a cache for - * the results it looks up */ - cred->thread_keyring = keyring; - cred->jit_keyring = KEY_REQKEY_DEFL_THREAD_KEYRING; - dns_resolver_cache = cred; + memcpy(name, hostname, len); + name[len] = 0; + cFYI(1, "%s: unc is IP, skipping dns upcall: %s", __func__, name); + *ip_addr = name; return 0; - -failed_put_key: - key_put(keyring); -failed_put_cred: - put_cred(cred); - return ret; -} - -void cifs_exit_dns_resolver(void) -{ - key_revoke(dns_resolver_cache->thread_keyring); - unregister_key_type(&key_type_dns_resolver); - put_cred(dns_resolver_cache); - printk(KERN_NOTICE "Unregistered %s key type\n", - key_type_dns_resolver.name); } diff --git a/fs/cifs/dns_resolve.h b/fs/cifs/dns_resolve.h index 5d7f291df16..d3f5d27f4d0 100644 --- a/fs/cifs/dns_resolve.h +++ b/fs/cifs/dns_resolve.h @@ -24,8 +24,6 @@ #define _DNS_RESOLVE_H #ifdef __KERNEL__ -extern int __init cifs_init_dns_resolver(void); -extern void cifs_exit_dns_resolver(void); extern int dns_resolve_server_name_to_ip(const char *unc, char **ip_addr); #endif /* KERNEL */ diff --git a/fs/cifs/file.c b/fs/cifs/file.c index 409e4f523e6..db11fdef0e9 100644 --- a/fs/cifs/file.c +++ b/fs/cifs/file.c @@ -40,6 +40,7 @@ #include "cifs_unicode.h" #include "cifs_debug.h" #include "cifs_fs_sb.h" +#include "fscache.h" static inline int cifs_convert_flags(unsigned int flags) { @@ -282,6 +283,9 @@ int cifs_open(struct inode *inode, struct file *file) CIFSSMBClose(xid, tcon, netfid); rc = -ENOMEM; } + + cifs_fscache_set_inode_cookie(inode, file); + goto out; } else if ((rc == -EINVAL) || (rc == -EOPNOTSUPP)) { if (tcon->ses->serverNOS) @@ -373,6 +377,8 @@ int cifs_open(struct inode *inode, struct file *file) goto out; } + cifs_fscache_set_inode_cookie(inode, file); + if (oplock & CIFS_CREATE_ACTION) { /* time to set mode which we can not set earlier due to problems creating new read-only files */ @@ -427,7 +433,7 @@ static int cifs_reopen_file(struct file *file, bool can_flush) __u16 netfid; if (file->private_data) - pCifsFile = (struct cifsFileInfo *)file->private_data; + pCifsFile = file->private_data; else return -EBADF; @@ -565,8 +571,7 @@ int cifs_close(struct inode *inode, struct file *file) int xid, timeout; struct cifs_sb_info *cifs_sb; struct cifsTconInfo *pTcon; - struct cifsFileInfo *pSMBFile = - (struct cifsFileInfo *)file->private_data; + struct cifsFileInfo *pSMBFile = file->private_data; xid = GetXid(); @@ -641,8 +646,7 @@ int cifs_closedir(struct inode *inode, struct file *file) { int rc = 0; int xid; - struct cifsFileInfo *pCFileStruct = - (struct cifsFileInfo *)file->private_data; + struct cifsFileInfo *pCFileStruct = file->private_data; char *ptmp; cFYI(1, "Closedir inode = 0x%p", inode); @@ -863,8 +867,7 @@ int cifs_lock(struct file *file, int cmd, struct file_lock *pfLock) length, pfLock, posix_lock_type, wait_flag); } else { - struct cifsFileInfo *fid = - (struct cifsFileInfo *)file->private_data; + struct cifsFileInfo *fid = file->private_data; if (numLock) { rc = CIFSSMBLock(xid, tcon, netfid, length, @@ -965,7 +968,7 @@ ssize_t cifs_user_write(struct file *file, const char __user *write_data, if (file->private_data == NULL) return -EBADF; - open_file = (struct cifsFileInfo *) file->private_data; + open_file = file->private_data; rc = generic_write_checks(file, poffset, &write_size, 0); if (rc) @@ -1067,7 +1070,7 @@ static ssize_t cifs_write(struct file *file, const char *write_data, if (file->private_data == NULL) return -EBADF; - open_file = (struct cifsFileInfo *)file->private_data; + open_file = file->private_data; xid = GetXid(); @@ -1651,8 +1654,7 @@ int cifs_fsync(struct file *file, int datasync) int xid; int rc = 0; struct cifsTconInfo *tcon; - struct cifsFileInfo *smbfile = - (struct cifsFileInfo *)file->private_data; + struct cifsFileInfo *smbfile = file->private_data; struct inode *inode = file->f_path.dentry->d_inode; xid = GetXid(); @@ -1756,7 +1758,7 @@ ssize_t cifs_user_read(struct file *file, char __user *read_data, FreeXid(xid); return rc; } - open_file = (struct cifsFileInfo *)file->private_data; + open_file = file->private_data; if ((file->f_flags & O_ACCMODE) == O_WRONLY) cFYI(1, "attempting read on write only file instance"); @@ -1837,7 +1839,7 @@ static ssize_t cifs_read(struct file *file, char *read_data, size_t read_size, FreeXid(xid); return rc; } - open_file = (struct cifsFileInfo *)file->private_data; + open_file = file->private_data; if ((file->f_flags & O_ACCMODE) == O_WRONLY) cFYI(1, "attempting read on write only file instance"); @@ -1942,6 +1944,9 @@ static void cifs_copy_cache_pages(struct address_space *mapping, SetPageUptodate(page); unlock_page(page); data += PAGE_CACHE_SIZE; + + /* add page to FS-Cache */ + cifs_readpage_to_fscache(mapping->host, page); } return; } @@ -1968,10 +1973,19 @@ static int cifs_readpages(struct file *file, struct address_space *mapping, FreeXid(xid); return rc; } - open_file = (struct cifsFileInfo *)file->private_data; + open_file = file->private_data; cifs_sb = CIFS_SB(file->f_path.dentry->d_sb); pTcon = cifs_sb->tcon; + /* + * Reads as many pages as possible from fscache. Returns -ENOBUFS + * immediately if the cookie is negative + */ + rc = cifs_readpages_from_fscache(mapping->host, mapping, page_list, + &num_pages); + if (rc == 0) + goto read_complete; + cFYI(DBG2, "rpages: num pages %d", num_pages); for (i = 0; i < num_pages; ) { unsigned contig_pages; @@ -2082,6 +2096,7 @@ static int cifs_readpages(struct file *file, struct address_space *mapping, smb_read_data = NULL; } +read_complete: FreeXid(xid); return rc; } @@ -2092,6 +2107,11 @@ static int cifs_readpage_worker(struct file *file, struct page *page, char *read_data; int rc; + /* Is the page cached? */ + rc = cifs_readpage_from_fscache(file->f_path.dentry->d_inode, page); + if (rc == 0) + goto read_complete; + page_cache_get(page); read_data = kmap(page); /* for reads over a certain size could initiate async read ahead */ @@ -2111,11 +2131,17 @@ static int cifs_readpage_worker(struct file *file, struct page *page, flush_dcache_page(page); SetPageUptodate(page); + + /* send this page to the cache */ + cifs_readpage_to_fscache(file->f_path.dentry->d_inode, page); + rc = 0; io_error: kunmap(page); page_cache_release(page); + +read_complete: return rc; } @@ -2265,8 +2291,23 @@ out: return rc; } -static void -cifs_oplock_break(struct slow_work *work) +static int cifs_release_page(struct page *page, gfp_t gfp) +{ + if (PagePrivate(page)) + return 0; + + return cifs_fscache_release_page(page, gfp); +} + +static void cifs_invalidate_page(struct page *page, unsigned long offset) +{ + struct cifsInodeInfo *cifsi = CIFS_I(page->mapping->host); + + if (offset == 0) + cifs_fscache_invalidate_page(page, &cifsi->vfs_inode); +} + +void cifs_oplock_break(struct work_struct *work) { struct cifsFileInfo *cfile = container_of(work, struct cifsFileInfo, oplock_break); @@ -2303,33 +2344,30 @@ cifs_oplock_break(struct slow_work *work) LOCKING_ANDX_OPLOCK_RELEASE, false); cFYI(1, "Oplock release rc = %d", rc); } + + /* + * We might have kicked in before is_valid_oplock_break() + * finished grabbing reference for us. Make sure it's done by + * waiting for GlobalSMSSeslock. + */ + write_lock(&GlobalSMBSeslock); + write_unlock(&GlobalSMBSeslock); + + cifs_oplock_break_put(cfile); } -static int -cifs_oplock_break_get(struct slow_work *work) +void cifs_oplock_break_get(struct cifsFileInfo *cfile) { - struct cifsFileInfo *cfile = container_of(work, struct cifsFileInfo, - oplock_break); mntget(cfile->mnt); cifsFileInfo_get(cfile); - return 0; } -static void -cifs_oplock_break_put(struct slow_work *work) +void cifs_oplock_break_put(struct cifsFileInfo *cfile) { - struct cifsFileInfo *cfile = container_of(work, struct cifsFileInfo, - oplock_break); mntput(cfile->mnt); cifsFileInfo_put(cfile); } -const struct slow_work_ops cifs_oplock_break_ops = { - .get_ref = cifs_oplock_break_get, - .put_ref = cifs_oplock_break_put, - .execute = cifs_oplock_break, -}; - const struct address_space_operations cifs_addr_ops = { .readpage = cifs_readpage, .readpages = cifs_readpages, @@ -2338,6 +2376,8 @@ const struct address_space_operations cifs_addr_ops = { .write_begin = cifs_write_begin, .write_end = cifs_write_end, .set_page_dirty = __set_page_dirty_nobuffers, + .releasepage = cifs_release_page, + .invalidatepage = cifs_invalidate_page, /* .sync_page = cifs_sync_page, */ /* .direct_IO = */ }; @@ -2354,6 +2394,8 @@ const struct address_space_operations cifs_addr_ops_smallbuf = { .write_begin = cifs_write_begin, .write_end = cifs_write_end, .set_page_dirty = __set_page_dirty_nobuffers, + .releasepage = cifs_release_page, + .invalidatepage = cifs_invalidate_page, /* .sync_page = cifs_sync_page, */ /* .direct_IO = */ }; diff --git a/fs/cifs/fscache.c b/fs/cifs/fscache.c new file mode 100644 index 00000000000..9f3f5c4be16 --- /dev/null +++ b/fs/cifs/fscache.c @@ -0,0 +1,236 @@ +/* + * fs/cifs/fscache.c - CIFS filesystem cache interface + * + * Copyright (c) 2010 Novell, Inc. + * Author(s): Suresh Jayaraman (sjayaraman@suse.de> + * + * This library is free software; you can redistribute it and/or modify + * it under the terms of the GNU Lesser General Public License as published + * by the Free Software Foundation; either version 2.1 of the License, or + * (at your option) any later version. + * + * This library 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 Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with this library; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ +#include "fscache.h" +#include "cifsglob.h" +#include "cifs_debug.h" +#include "cifs_fs_sb.h" + +void cifs_fscache_get_client_cookie(struct TCP_Server_Info *server) +{ + server->fscache = + fscache_acquire_cookie(cifs_fscache_netfs.primary_index, + &cifs_fscache_server_index_def, server); + cFYI(1, "CIFS: get client cookie (0x%p/0x%p)", server, + server->fscache); +} + +void cifs_fscache_release_client_cookie(struct TCP_Server_Info *server) +{ + cFYI(1, "CIFS: release client cookie (0x%p/0x%p)", server, + server->fscache); + fscache_relinquish_cookie(server->fscache, 0); + server->fscache = NULL; +} + +void cifs_fscache_get_super_cookie(struct cifsTconInfo *tcon) +{ + struct TCP_Server_Info *server = tcon->ses->server; + + tcon->fscache = + fscache_acquire_cookie(server->fscache, + &cifs_fscache_super_index_def, tcon); + cFYI(1, "CIFS: get superblock cookie (0x%p/0x%p)", + server->fscache, tcon->fscache); +} + +void cifs_fscache_release_super_cookie(struct cifsTconInfo *tcon) +{ + cFYI(1, "CIFS: releasing superblock cookie (0x%p)", tcon->fscache); + fscache_relinquish_cookie(tcon->fscache, 0); + tcon->fscache = NULL; +} + +static void cifs_fscache_enable_inode_cookie(struct inode *inode) +{ + struct cifsInodeInfo *cifsi = CIFS_I(inode); + struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb); + + if (cifsi->fscache) + return; + + cifsi->fscache = fscache_acquire_cookie(cifs_sb->tcon->fscache, + &cifs_fscache_inode_object_def, + cifsi); + cFYI(1, "CIFS: got FH cookie (0x%p/0x%p)", + cifs_sb->tcon->fscache, cifsi->fscache); +} + +void cifs_fscache_release_inode_cookie(struct inode *inode) +{ + struct cifsInodeInfo *cifsi = CIFS_I(inode); + + if (cifsi->fscache) { + cFYI(1, "CIFS releasing inode cookie (0x%p)", + cifsi->fscache); + fscache_relinquish_cookie(cifsi->fscache, 0); + cifsi->fscache = NULL; + } +} + +static void cifs_fscache_disable_inode_cookie(struct inode *inode) +{ + struct cifsInodeInfo *cifsi = CIFS_I(inode); + + if (cifsi->fscache) { + cFYI(1, "CIFS disabling inode cookie (0x%p)", + cifsi->fscache); + fscache_relinquish_cookie(cifsi->fscache, 1); + cifsi->fscache = NULL; + } +} + +void cifs_fscache_set_inode_cookie(struct inode *inode, struct file *filp) +{ + if ((filp->f_flags & O_ACCMODE) != O_RDONLY) + cifs_fscache_disable_inode_cookie(inode); + else { + cifs_fscache_enable_inode_cookie(inode); + cFYI(1, "CIFS: fscache inode cookie set"); + } +} + +void cifs_fscache_reset_inode_cookie(struct inode *inode) +{ + struct cifsInodeInfo *cifsi = CIFS_I(inode); + struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb); + struct fscache_cookie *old = cifsi->fscache; + + if (cifsi->fscache) { + /* retire the current fscache cache and get a new one */ + fscache_relinquish_cookie(cifsi->fscache, 1); + + cifsi->fscache = fscache_acquire_cookie(cifs_sb->tcon->fscache, + &cifs_fscache_inode_object_def, + cifsi); + cFYI(1, "CIFS: new cookie 0x%p oldcookie 0x%p", + cifsi->fscache, old); + } +} + +int cifs_fscache_release_page(struct page *page, gfp_t gfp) +{ + if (PageFsCache(page)) { + struct inode *inode = page->mapping->host; + struct cifsInodeInfo *cifsi = CIFS_I(inode); + + cFYI(1, "CIFS: fscache release page (0x%p/0x%p)", + page, cifsi->fscache); + if (!fscache_maybe_release_page(cifsi->fscache, page, gfp)) + return 0; + } + + return 1; +} + +static void cifs_readpage_from_fscache_complete(struct page *page, void *ctx, + int error) +{ + cFYI(1, "CFS: readpage_from_fscache_complete (0x%p/%d)", + page, error); + if (!error) + SetPageUptodate(page); + unlock_page(page); +} + +/* + * Retrieve a page from FS-Cache + */ +int __cifs_readpage_from_fscache(struct inode *inode, struct page *page) +{ + int ret; + + cFYI(1, "CIFS: readpage_from_fscache(fsc:%p, p:%p, i:0x%p", + CIFS_I(inode)->fscache, page, inode); + ret = fscache_read_or_alloc_page(CIFS_I(inode)->fscache, page, + cifs_readpage_from_fscache_complete, + NULL, + GFP_KERNEL); + switch (ret) { + + case 0: /* page found in fscache, read submitted */ + cFYI(1, "CIFS: readpage_from_fscache: submitted"); + return ret; + case -ENOBUFS: /* page won't be cached */ + case -ENODATA: /* page not in cache */ + cFYI(1, "CIFS: readpage_from_fscache %d", ret); + return 1; + + default: + cERROR(1, "unknown error ret = %d", ret); + } + return ret; +} + +/* + * Retrieve a set of pages from FS-Cache + */ +int __cifs_readpages_from_fscache(struct inode *inode, + struct address_space *mapping, + struct list_head *pages, + unsigned *nr_pages) +{ + int ret; + + cFYI(1, "CIFS: __cifs_readpages_from_fscache (0x%p/%u/0x%p)", + CIFS_I(inode)->fscache, *nr_pages, inode); + ret = fscache_read_or_alloc_pages(CIFS_I(inode)->fscache, mapping, + pages, nr_pages, + cifs_readpage_from_fscache_complete, + NULL, + mapping_gfp_mask(mapping)); + switch (ret) { + case 0: /* read submitted to the cache for all pages */ + cFYI(1, "CIFS: readpages_from_fscache: submitted"); + return ret; + + case -ENOBUFS: /* some pages are not cached and can't be */ + case -ENODATA: /* some pages are not cached */ + cFYI(1, "CIFS: readpages_from_fscache: no page"); + return 1; + + default: + cFYI(1, "unknown error ret = %d", ret); + } + + return ret; +} + +void __cifs_readpage_to_fscache(struct inode *inode, struct page *page) +{ + int ret; + + cFYI(1, "CIFS: readpage_to_fscache(fsc: %p, p: %p, i: %p", + CIFS_I(inode)->fscache, page, inode); + ret = fscache_write_page(CIFS_I(inode)->fscache, page, GFP_KERNEL); + if (ret != 0) + fscache_uncache_page(CIFS_I(inode)->fscache, page); +} + +void __cifs_fscache_invalidate_page(struct page *page, struct inode *inode) +{ + struct cifsInodeInfo *cifsi = CIFS_I(inode); + struct fscache_cookie *cookie = cifsi->fscache; + + cFYI(1, "CIFS: fscache invalidatepage (0x%p/0x%p)", page, cookie); + fscache_wait_on_page_write(cookie, page); + fscache_uncache_page(cookie, page); +} + diff --git a/fs/cifs/fscache.h b/fs/cifs/fscache.h new file mode 100644 index 00000000000..31b88ec2341 --- /dev/null +++ b/fs/cifs/fscache.h @@ -0,0 +1,136 @@ +/* + * fs/cifs/fscache.h - CIFS filesystem cache interface definitions + * + * Copyright (c) 2010 Novell, Inc. + * Authors(s): Suresh Jayaraman (sjayaraman@suse.de> + * + * This library is free software; you can redistribute it and/or modify + * it under the terms of the GNU Lesser General Public License as published + * by the Free Software Foundation; either version 2.1 of the License, or + * (at your option) any later version. + * + * This library 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 Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public License + * along with this library; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ +#ifndef _CIFS_FSCACHE_H +#define _CIFS_FSCACHE_H + +#include <linux/fscache.h> + +#include "cifsglob.h" + +#ifdef CONFIG_CIFS_FSCACHE + +extern struct fscache_netfs cifs_fscache_netfs; +extern const struct fscache_cookie_def cifs_fscache_server_index_def; +extern const struct fscache_cookie_def cifs_fscache_super_index_def; +extern const struct fscache_cookie_def cifs_fscache_inode_object_def; + +extern int cifs_fscache_register(void); +extern void cifs_fscache_unregister(void); + +/* + * fscache.c + */ +extern void cifs_fscache_get_client_cookie(struct TCP_Server_Info *); +extern void cifs_fscache_release_client_cookie(struct TCP_Server_Info *); +extern void cifs_fscache_get_super_cookie(struct cifsTconInfo *); +extern void cifs_fscache_release_super_cookie(struct cifsTconInfo *); + +extern void cifs_fscache_release_inode_cookie(struct inode *); +extern void cifs_fscache_set_inode_cookie(struct inode *, struct file *); +extern void cifs_fscache_reset_inode_cookie(struct inode *); + +extern void __cifs_fscache_invalidate_page(struct page *, struct inode *); +extern int cifs_fscache_release_page(struct page *page, gfp_t gfp); +extern int __cifs_readpage_from_fscache(struct inode *, struct page *); +extern int __cifs_readpages_from_fscache(struct inode *, + struct address_space *, + struct list_head *, + unsigned *); + +extern void __cifs_readpage_to_fscache(struct inode *, struct page *); + +static inline void cifs_fscache_invalidate_page(struct page *page, + struct inode *inode) +{ + if (PageFsCache(page)) + __cifs_fscache_invalidate_page(page, inode); +} + +static inline int cifs_readpage_from_fscache(struct inode *inode, + struct page *page) +{ + if (CIFS_I(inode)->fscache) + return __cifs_readpage_from_fscache(inode, page); + + return -ENOBUFS; +} + +static inline int cifs_readpages_from_fscache(struct inode *inode, + struct address_space *mapping, + struct list_head *pages, + unsigned *nr_pages) +{ + if (CIFS_I(inode)->fscache) + return __cifs_readpages_from_fscache(inode, mapping, pages, + nr_pages); + return -ENOBUFS; +} + +static inline void cifs_readpage_to_fscache(struct inode *inode, + struct page *page) +{ + if (PageFsCache(page)) + __cifs_readpage_to_fscache(inode, page); +} + +#else /* CONFIG_CIFS_FSCACHE */ +static inline int cifs_fscache_register(void) { return 0; } +static inline void cifs_fscache_unregister(void) {} + +static inline void +cifs_fscache_get_client_cookie(struct TCP_Server_Info *server) {} +static inline void +cifs_fscache_release_client_cookie(struct TCP_Server_Info *server) {} +static inline void cifs_fscache_get_super_cookie(struct cifsTconInfo *tcon) {} +static inline void +cifs_fscache_release_super_cookie(struct cifsTconInfo *tcon) {} + +static inline void cifs_fscache_release_inode_cookie(struct inode *inode) {} +static inline void cifs_fscache_set_inode_cookie(struct inode *inode, + struct file *filp) {} +static inline void cifs_fscache_reset_inode_cookie(struct inode *inode) {} +static inline int cifs_fscache_release_page(struct page *page, gfp_t gfp) +{ + return 1; /* May release page */ +} + +static inline void cifs_fscache_invalidate_page(struct page *page, + struct inode *inode) {} +static inline int +cifs_readpage_from_fscache(struct inode *inode, struct page *page) +{ + return -ENOBUFS; +} + +static inline int cifs_readpages_from_fscache(struct inode *inode, + struct address_space *mapping, + struct list_head *pages, + unsigned *nr_pages) +{ + return -ENOBUFS; +} + +static inline void cifs_readpage_to_fscache(struct inode *inode, + struct page *page) {} + +#endif /* CONFIG_CIFS_FSCACHE */ + +#endif /* _CIFS_FSCACHE_H */ diff --git a/fs/cifs/inode.c b/fs/cifs/inode.c index 6f0683c6895..dc4c47ab958 100644 --- a/fs/cifs/inode.c +++ b/fs/cifs/inode.c @@ -29,6 +29,7 @@ #include "cifsproto.h" #include "cifs_debug.h" #include "cifs_fs_sb.h" +#include "fscache.h" static void cifs_set_ops(struct inode *inode, const bool is_dfs_referral) @@ -288,7 +289,7 @@ int cifs_get_file_info_unix(struct file *filp) struct inode *inode = filp->f_path.dentry->d_inode; struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb); struct cifsTconInfo *tcon = cifs_sb->tcon; - struct cifsFileInfo *cfile = (struct cifsFileInfo *) filp->private_data; + struct cifsFileInfo *cfile = filp->private_data; xid = GetXid(); rc = CIFSSMBUnixQFileInfo(xid, tcon, cfile->netfid, &find_data); @@ -515,7 +516,7 @@ int cifs_get_file_info(struct file *filp) struct inode *inode = filp->f_path.dentry->d_inode; struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb); struct cifsTconInfo *tcon = cifs_sb->tcon; - struct cifsFileInfo *cfile = (struct cifsFileInfo *) filp->private_data; + struct cifsFileInfo *cfile = filp->private_data; xid = GetXid(); rc = CIFSSMBQFileInfo(xid, tcon, cfile->netfid, &find_data); @@ -723,18 +724,17 @@ cifs_find_inode(struct inode *inode, void *opaque) { struct cifs_fattr *fattr = (struct cifs_fattr *) opaque; + /* don't match inode with different uniqueid */ if (CIFS_I(inode)->uniqueid != fattr->cf_uniqueid) return 0; - /* - * uh oh -- it's a directory. We can't use it since hardlinked dirs are - * verboten. Disable serverino and return it as if it were found, the - * caller can discard it, generate a uniqueid and retry the find - */ - if (S_ISDIR(inode->i_mode) && !list_empty(&inode->i_dentry)) { + /* don't match inode of different type */ + if ((inode->i_mode & S_IFMT) != (fattr->cf_mode & S_IFMT)) + return 0; + + /* if it's not a directory or has no dentries, then flag it */ + if (S_ISDIR(inode->i_mode) && !list_empty(&inode->i_dentry)) fattr->cf_flags |= CIFS_FATTR_INO_COLLISION; - cifs_autodisable_serverino(CIFS_SB(inode->i_sb)); - } return 1; } @@ -748,6 +748,27 @@ cifs_init_inode(struct inode *inode, void *opaque) return 0; } +/* + * walk dentry list for an inode and report whether it has aliases that + * are hashed. We use this to determine if a directory inode can actually + * be used. + */ +static bool +inode_has_hashed_dentries(struct inode *inode) +{ + struct dentry *dentry; + + spin_lock(&dcache_lock); + list_for_each_entry(dentry, &inode->i_dentry, d_alias) { + if (!d_unhashed(dentry) || IS_ROOT(dentry)) { + spin_unlock(&dcache_lock); + return true; + } + } + spin_unlock(&dcache_lock); + return false; +} + /* Given fattrs, get a corresponding inode */ struct inode * cifs_iget(struct super_block *sb, struct cifs_fattr *fattr) @@ -763,12 +784,16 @@ retry_iget5_locked: inode = iget5_locked(sb, hash, cifs_find_inode, cifs_init_inode, fattr); if (inode) { - /* was there a problematic inode number collision? */ + /* was there a potentially problematic inode collision? */ if (fattr->cf_flags & CIFS_FATTR_INO_COLLISION) { - iput(inode); - fattr->cf_uniqueid = iunique(sb, ROOT_I); fattr->cf_flags &= ~CIFS_FATTR_INO_COLLISION; - goto retry_iget5_locked; + + if (inode_has_hashed_dentries(inode)) { + cifs_autodisable_serverino(CIFS_SB(sb)); + iput(inode); + fattr->cf_uniqueid = iunique(sb, ROOT_I); + goto retry_iget5_locked; + } } cifs_fattr_to_inode(inode, fattr); @@ -776,6 +801,10 @@ retry_iget5_locked: inode->i_flags |= S_NOATIME | S_NOCMTIME; if (inode->i_state & I_NEW) { inode->i_ino = hash; +#ifdef CONFIG_CIFS_FSCACHE + /* initialize per-inode cache cookie pointer */ + CIFS_I(inode)->fscache = NULL; +#endif unlock_new_inode(inode); } } @@ -807,6 +836,11 @@ struct inode *cifs_root_iget(struct super_block *sb, unsigned long ino) if (!inode) return ERR_PTR(-ENOMEM); +#ifdef CONFIG_CIFS_FSCACHE + /* populate tcon->resource_id */ + cifs_sb->tcon->resource_id = CIFS_I(inode)->uniqueid; +#endif + if (rc && cifs_sb->tcon->ipc) { cFYI(1, "ipc connection - fake read inode"); inode->i_mode |= S_IFDIR; @@ -1568,6 +1602,7 @@ cifs_invalidate_mapping(struct inode *inode) cifs_i->write_behind_rc = rc; } invalidate_remote_inode(inode); + cifs_fscache_reset_inode_cookie(inode); } int cifs_revalidate_file(struct file *filp) diff --git a/fs/cifs/ioctl.c b/fs/cifs/ioctl.c index 505926f1ee6..9d38a71c8e1 100644 --- a/fs/cifs/ioctl.c +++ b/fs/cifs/ioctl.c @@ -41,8 +41,7 @@ long cifs_ioctl(struct file *filep, unsigned int command, unsigned long arg) __u64 ExtAttrMask = 0; __u64 caps; struct cifsTconInfo *tcon; - struct cifsFileInfo *pSMBFile = - (struct cifsFileInfo *)filep->private_data; + struct cifsFileInfo *pSMBFile = filep->private_data; #endif /* CONFIG_CIFS_POSIX */ xid = GetXid(); diff --git a/fs/cifs/misc.c b/fs/cifs/misc.c index 1394aa37f26..3ccadc1326d 100644 --- a/fs/cifs/misc.c +++ b/fs/cifs/misc.c @@ -498,7 +498,6 @@ is_valid_oplock_break(struct smb_hdr *buf, struct TCP_Server_Info *srv) struct cifsTconInfo *tcon; struct cifsInodeInfo *pCifsInode; struct cifsFileInfo *netfile; - int rc; cFYI(1, "Checking for oplock break or dnotify response"); if ((pSMB->hdr.Command == SMB_COM_NT_TRANSACT) && @@ -583,13 +582,18 @@ is_valid_oplock_break(struct smb_hdr *buf, struct TCP_Server_Info *srv) pCifsInode->clientCanCacheAll = false; if (pSMB->OplockLevel == 0) pCifsInode->clientCanCacheRead = false; - rc = slow_work_enqueue(&netfile->oplock_break); - if (rc) { - cERROR(1, "failed to enqueue oplock " - "break: %d\n", rc); - } else { - netfile->oplock_break_cancelled = false; - } + + /* + * cifs_oplock_break_put() can't be called + * from here. Get reference after queueing + * succeeded. cifs_oplock_break() will + * synchronize using GlobalSMSSeslock. + */ + if (queue_work(system_nrt_wq, + &netfile->oplock_break)) + cifs_oplock_break_get(netfile); + netfile->oplock_break_cancelled = false; + read_unlock(&GlobalSMBSeslock); read_unlock(&cifs_tcp_ses_lock); return true; diff --git a/fs/cifs/netmisc.c b/fs/cifs/netmisc.c index d35d52889cb..f97851119e6 100644 --- a/fs/cifs/netmisc.c +++ b/fs/cifs/netmisc.c @@ -61,6 +61,7 @@ static const struct smb_to_posix_error mapping_table_ERRDOS[] = { {ERRremcd, -EACCES}, {ERRdiffdevice, -EXDEV}, {ERRnofiles, -ENOENT}, + {ERRwriteprot, -EROFS}, {ERRbadshare, -ETXTBSY}, {ERRlock, -EACCES}, {ERRunsup, -EINVAL}, @@ -139,17 +140,18 @@ static const struct smb_to_posix_error mapping_table_ERRHRD[] = { * Returns 0 on failure. */ static int -cifs_inet_pton(const int address_family, const char *cp, void *dst) +cifs_inet_pton(const int address_family, const char *cp, int len, void *dst) { int ret = 0; /* calculate length by finding first slash or NULL */ if (address_family == AF_INET) - ret = in4_pton(cp, -1 /* len */, dst, '\\', NULL); + ret = in4_pton(cp, len, dst, '\\', NULL); else if (address_family == AF_INET6) - ret = in6_pton(cp, -1 /* len */, dst , '\\', NULL); + ret = in6_pton(cp, len, dst , '\\', NULL); - cFYI(DBG2, "address conversion returned %d for %s", ret, cp); + cFYI(DBG2, "address conversion returned %d for %*.*s", + ret, len, len, cp); if (ret > 0) ret = 1; return ret; @@ -164,43 +166,66 @@ cifs_inet_pton(const int address_family, const char *cp, void *dst) * Returns 0 on failure. */ int -cifs_convert_address(char *src, void *dst) +cifs_convert_address(struct sockaddr *dst, const char *src, int len) { - int rc; - char *pct, *endp; + int rc, alen, slen; + const char *pct; + char *endp, scope_id[13]; struct sockaddr_in *s4 = (struct sockaddr_in *) dst; struct sockaddr_in6 *s6 = (struct sockaddr_in6 *) dst; /* IPv4 address */ - if (cifs_inet_pton(AF_INET, src, &s4->sin_addr.s_addr)) { + if (cifs_inet_pton(AF_INET, src, len, &s4->sin_addr.s_addr)) { s4->sin_family = AF_INET; return 1; } - /* temporarily terminate string */ - pct = strchr(src, '%'); - if (pct) - *pct = '\0'; - - rc = cifs_inet_pton(AF_INET6, src, &s6->sin6_addr.s6_addr); - - /* repair temp termination (if any) and make pct point to scopeid */ - if (pct) - *pct++ = '%'; + /* attempt to exclude the scope ID from the address part */ + pct = memchr(src, '%', len); + alen = pct ? pct - src : len; + rc = cifs_inet_pton(AF_INET6, src, alen, &s6->sin6_addr.s6_addr); if (!rc) return rc; s6->sin6_family = AF_INET6; if (pct) { + /* grab the scope ID */ + slen = len - (alen + 1); + if (slen <= 0 || slen > 12) + return 0; + memcpy(scope_id, pct + 1, slen); + scope_id[slen] = '\0'; + s6->sin6_scope_id = (u32) simple_strtoul(pct, &endp, 0); - if (!*pct || *endp) + if (endp != scope_id + slen) return 0; } return rc; } +int +cifs_fill_sockaddr(struct sockaddr *dst, const char *src, int len, + const unsigned short int port) +{ + if (!cifs_convert_address(dst, src, len)) + return 0; + + switch (dst->sa_family) { + case AF_INET: + ((struct sockaddr_in *)dst)->sin_port = htons(port); + break; + case AF_INET6: + ((struct sockaddr_in6 *)dst)->sin6_port = htons(port); + break; + default: + return 0; + } + + return 1; +} + /***************************************************************************** convert a NT status code to a dos class/code *****************************************************************************/ diff --git a/fs/cifs/readdir.c b/fs/cifs/readdir.c index daf1753af67..d5e591fab47 100644 --- a/fs/cifs/readdir.c +++ b/fs/cifs/readdir.c @@ -847,6 +847,11 @@ int cifs_readdir(struct file *file, void *direntry, filldir_t filldir) end_of_smb = cifsFile->srch_inf.ntwrk_buf_start + max_len; tmp_buf = kmalloc(UNICODE_NAME_MAX, GFP_KERNEL); + if (tmp_buf == NULL) { + rc = -ENOMEM; + break; + } + for (i = 0; (i < num_to_fill) && (rc == 0); i++) { if (current_entry == NULL) { /* evaluate whether this case is an error */ diff --git a/fs/cifs/smberr.h b/fs/cifs/smberr.h index c5084d27db7..7f16cb825fe 100644 --- a/fs/cifs/smberr.h +++ b/fs/cifs/smberr.h @@ -76,6 +76,7 @@ #define ERRnofiles 18 /* A File Search command can find no more files matching the specified criteria. */ +#define ERRwriteprot 19 /* media is write protected */ #define ERRgeneral 31 #define ERRbadshare 32 /* The sharing mode specified for an Open conflicts with existing FIDs on diff --git a/fs/compat.c b/fs/compat.c index 6490d2134ff..c6fda9aeb86 100644 --- a/fs/compat.c +++ b/fs/compat.c @@ -8,7 +8,7 @@ * Copyright (C) 1997-2000 Jakub Jelinek (jakub@redhat.com) * Copyright (C) 1998 Eddie C. Dost (ecd@skynet.be) * Copyright (C) 2001,2002 Andi Kleen, SuSE Labs - * Copyright (C) 2003 Pavel Machek (pavel@suse.cz) + * Copyright (C) 2003 Pavel Machek (pavel@ucw.cz) * * 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 diff --git a/fs/compat_ioctl.c b/fs/compat_ioctl.c index 641640dc7ae..63ae8583146 100644 --- a/fs/compat_ioctl.c +++ b/fs/compat_ioctl.c @@ -4,7 +4,7 @@ * Copyright (C) 1997-2000 Jakub Jelinek (jakub@redhat.com) * Copyright (C) 1998 Eddie C. Dost (ecd@skynet.be) * Copyright (C) 2001,2002 Andi Kleen, SuSE Labs - * Copyright (C) 2003 Pavel Machek (pavel@suse.cz) + * Copyright (C) 2003 Pavel Machek (pavel@ucw.cz) * * These routines maintain argument size conversion between 32bit and 64bit * ioctls. @@ -601,8 +601,11 @@ static int ioc_settimeout(unsigned int fd, unsigned int cmd, } /* Bluetooth ioctls */ -#define HCIUARTSETPROTO _IOW('U', 200, int) -#define HCIUARTGETPROTO _IOR('U', 201, int) +#define HCIUARTSETPROTO _IOW('U', 200, int) +#define HCIUARTGETPROTO _IOR('U', 201, int) +#define HCIUARTGETDEVICE _IOR('U', 202, int) +#define HCIUARTSETFLAGS _IOW('U', 203, int) +#define HCIUARTGETFLAGS _IOR('U', 204, int) #define BNEPCONNADD _IOW('B', 200, int) #define BNEPCONNDEL _IOW('B', 201, int) @@ -1328,6 +1331,8 @@ COMPATIBLE_IOCTL(HCISETLINKPOL) COMPATIBLE_IOCTL(HCISETLINKMODE) COMPATIBLE_IOCTL(HCISETACLMTU) COMPATIBLE_IOCTL(HCISETSCOMTU) +COMPATIBLE_IOCTL(HCIBLOCKADDR) +COMPATIBLE_IOCTL(HCIUNBLOCKADDR) COMPATIBLE_IOCTL(HCIINQUIRY) COMPATIBLE_IOCTL(HCIUARTSETPROTO) COMPATIBLE_IOCTL(HCIUARTGETPROTO) diff --git a/fs/direct-io.c b/fs/direct-io.c index 7600aacf531..a10cb91cade 100644 --- a/fs/direct-io.c +++ b/fs/direct-io.c @@ -218,7 +218,7 @@ static struct page *dio_get_page(struct dio *dio) * filesystems can use it to hold additional state between get_block calls and * dio_complete. */ -static int dio_complete(struct dio *dio, loff_t offset, int ret) +static int dio_complete(struct dio *dio, loff_t offset, int ret, bool is_async) { ssize_t transferred = 0; @@ -239,14 +239,6 @@ static int dio_complete(struct dio *dio, loff_t offset, int ret) transferred = dio->i_size - offset; } - if (dio->end_io && dio->result) - dio->end_io(dio->iocb, offset, transferred, - dio->map_bh.b_private); - - if (dio->flags & DIO_LOCKING) - /* lockdep: non-owner release */ - up_read_non_owner(&dio->inode->i_alloc_sem); - if (ret == 0) ret = dio->page_errors; if (ret == 0) @@ -254,6 +246,17 @@ static int dio_complete(struct dio *dio, loff_t offset, int ret) if (ret == 0) ret = transferred; + if (dio->end_io && dio->result) { + dio->end_io(dio->iocb, offset, transferred, + dio->map_bh.b_private, ret, is_async); + } else if (is_async) { + aio_complete(dio->iocb, ret, 0); + } + + if (dio->flags & DIO_LOCKING) + /* lockdep: non-owner release */ + up_read_non_owner(&dio->inode->i_alloc_sem); + return ret; } @@ -277,8 +280,7 @@ static void dio_bio_end_aio(struct bio *bio, int error) spin_unlock_irqrestore(&dio->bio_lock, flags); if (remaining == 0) { - int ret = dio_complete(dio, dio->iocb->ki_pos, 0); - aio_complete(dio->iocb, ret, 0); + dio_complete(dio, dio->iocb->ki_pos, 0, true); kfree(dio); } } @@ -1126,7 +1128,7 @@ direct_io_worker(int rw, struct kiocb *iocb, struct inode *inode, spin_unlock_irqrestore(&dio->bio_lock, flags); if (ret2 == 0) { - ret = dio_complete(dio, offset, ret); + ret = dio_complete(dio, offset, ret, false); kfree(dio); } else BUG_ON(ret != -EIOCBQUEUED); diff --git a/fs/dlm/lowcomms.c b/fs/dlm/lowcomms.c index c0d35c62052..37a34c2c622 100644 --- a/fs/dlm/lowcomms.c +++ b/fs/dlm/lowcomms.c @@ -248,7 +248,7 @@ static struct connection *assoc2con(int assoc_id) for (i = 0 ; i < CONN_HASH_SIZE; i++) { hlist_for_each_entry(con, h, &connection_hash[i], list) { - if (con && con->sctp_assoc == assoc_id) { + if (con->sctp_assoc == assoc_id) { mutex_unlock(&connections_lock); return con; } diff --git a/fs/dlm/netlink.c b/fs/dlm/netlink.c index 2c6ad518100..ef17e0169da 100644 --- a/fs/dlm/netlink.c +++ b/fs/dlm/netlink.c @@ -81,24 +81,11 @@ static struct genl_ops dlm_nl_ops = { int __init dlm_netlink_init(void) { - int rv; - - rv = genl_register_family(&family); - if (rv) - return rv; - - rv = genl_register_ops(&family, &dlm_nl_ops); - if (rv < 0) - goto err; - return 0; - err: - genl_unregister_family(&family); - return rv; + return genl_register_family_with_ops(&family, &dlm_nl_ops, 1); } void dlm_netlink_exit(void) { - genl_unregister_ops(&family, &dlm_nl_ops); genl_unregister_family(&family); } diff --git a/fs/ecryptfs/crypto.c b/fs/ecryptfs/crypto.c index 1cc087635a5..a2e3b562e65 100644 --- a/fs/ecryptfs/crypto.c +++ b/fs/ecryptfs/crypto.c @@ -762,7 +762,7 @@ ecryptfs_decrypt_page_offset(struct ecryptfs_crypt_stat *crypt_stat, /** * ecryptfs_init_crypt_ctx - * @crypt_stat: Uninitilized crypt stats structure + * @crypt_stat: Uninitialized crypt stats structure * * Initialize the crypto context. * diff --git a/fs/exec.c b/fs/exec.c index e19de6a8033..dab85ecad68 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -28,7 +28,6 @@ #include <linux/mm.h> #include <linux/stat.h> #include <linux/fcntl.h> -#include <linux/smp_lock.h> #include <linux/swap.h> #include <linux/string.h> #include <linux/init.h> @@ -653,6 +652,7 @@ int setup_arg_pages(struct linux_binprm *bprm, else stack_base = vma->vm_start - stack_expand; #endif + current->mm->start_stack = bprm->p; ret = expand_stack(vma, stack_base); if (ret) ret = -EFAULT; @@ -1891,13 +1891,7 @@ void do_coredump(long signr, int exit_code, struct pt_regs *regs) */ clear_thread_flag(TIF_SIGPENDING); - /* - * lock_kernel() because format_corename() is controlled by sysctl, which - * uses lock_kernel() - */ - lock_kernel(); ispipe = format_corename(corename, signr); - unlock_kernel(); if (ispipe) { int dump_count; diff --git a/fs/ext3/Kconfig b/fs/ext3/Kconfig index 522b15498f4..e8c6ba0e4a3 100644 --- a/fs/ext3/Kconfig +++ b/fs/ext3/Kconfig @@ -31,6 +31,7 @@ config EXT3_FS config EXT3_DEFAULTS_TO_ORDERED bool "Default to 'data=ordered' in ext3" depends on EXT3_FS + default y help The journal mode options for ext3 have different tradeoffs between when data is guaranteed to be on disk and diff --git a/fs/ext3/inode.c b/fs/ext3/inode.c index 735f0190ec2..001eb0e2d48 100644 --- a/fs/ext3/inode.c +++ b/fs/ext3/inode.c @@ -1149,9 +1149,25 @@ static int walk_page_buffers( handle_t *handle, static int do_journal_get_write_access(handle_t *handle, struct buffer_head *bh) { + int dirty = buffer_dirty(bh); + int ret; + if (!buffer_mapped(bh) || buffer_freed(bh)) return 0; - return ext3_journal_get_write_access(handle, bh); + /* + * __block_prepare_write() could have dirtied some buffers. Clean + * the dirty bit as jbd2_journal_get_write_access() could complain + * otherwise about fs integrity issues. Setting of the dirty bit + * by __block_prepare_write() isn't a real problem here as we clear + * the bit before releasing a page lock and thus writeback cannot + * ever write the buffer. + */ + if (dirty) + clear_buffer_dirty(bh); + ret = ext3_journal_get_write_access(handle, bh); + if (!ret && dirty) + ret = ext3_journal_dirty_metadata(handle, bh); + return ret; } /* @@ -1625,10 +1641,7 @@ static int ext3_writeback_writepage(struct page *page, goto out_fail; } - if (test_opt(inode->i_sb, NOBH) && ext3_should_writeback_data(inode)) - ret = nobh_writepage(page, ext3_get_block, wbc); - else - ret = block_write_full_page(page, ext3_get_block, wbc); + ret = block_write_full_page(page, ext3_get_block, wbc); err = ext3_journal_stop(handle); if (!ret) @@ -1922,17 +1935,6 @@ static int ext3_block_truncate_page(handle_t *handle, struct page *page, length = blocksize - (offset & (blocksize - 1)); iblock = index << (PAGE_CACHE_SHIFT - inode->i_sb->s_blocksize_bits); - /* - * For "nobh" option, we can only work if we don't need to - * read-in the page - otherwise we create buffers to do the IO. - */ - if (!page_has_buffers(page) && test_opt(inode->i_sb, NOBH) && - ext3_should_writeback_data(inode) && PageUptodate(page)) { - zero_user(page, offset, length); - set_page_dirty(page); - goto unlock; - } - if (!page_has_buffers(page)) create_empty_buffers(page, blocksize, 0); @@ -2284,27 +2286,6 @@ static void ext3_free_branches(handle_t *handle, struct inode *inode, depth); /* - * We've probably journalled the indirect block several - * times during the truncate. But it's no longer - * needed and we now drop it from the transaction via - * journal_revoke(). - * - * That's easy if it's exclusively part of this - * transaction. But if it's part of the committing - * transaction then journal_forget() will simply - * brelse() it. That means that if the underlying - * block is reallocated in ext3_get_block(), - * unmap_underlying_metadata() will find this block - * and will try to get rid of it. damn, damn. - * - * If this block has already been committed to the - * journal, a revoke record will be written. And - * revoke records must be emitted *before* clearing - * this block's bit in the bitmaps. - */ - ext3_forget(handle, 1, inode, bh, bh->b_blocknr); - - /* * Everything below this this pointer has been * released. Now let this top-of-subtree go. * @@ -2327,6 +2308,31 @@ static void ext3_free_branches(handle_t *handle, struct inode *inode, truncate_restart_transaction(handle, inode); } + /* + * We've probably journalled the indirect block several + * times during the truncate. But it's no longer + * needed and we now drop it from the transaction via + * journal_revoke(). + * + * That's easy if it's exclusively part of this + * transaction. But if it's part of the committing + * transaction then journal_forget() will simply + * brelse() it. That means that if the underlying + * block is reallocated in ext3_get_block(), + * unmap_underlying_metadata() will find this block + * and will try to get rid of it. damn, damn. Thus + * we don't allow a block to be reallocated until + * a transaction freeing it has fully committed. + * + * We also have to make sure journal replay after a + * crash does not overwrite non-journaled data blocks + * with old metadata when the block got reallocated for + * data. Thus we have to store a revoke record for a + * block in the same transaction in which we free the + * block. + */ + ext3_forget(handle, 1, inode, bh, bh->b_blocknr); + ext3_free_blocks(handle, inode, nr, 1); if (parent_bh) { diff --git a/fs/ext3/namei.c b/fs/ext3/namei.c index ee184084ca4..2b35ddb70d6 100644 --- a/fs/ext3/namei.c +++ b/fs/ext3/namei.c @@ -1447,7 +1447,6 @@ static int ext3_add_entry (handle_t *handle, struct dentry *dentry, struct inode *inode) { struct inode *dir = dentry->d_parent->d_inode; - unsigned long offset; struct buffer_head * bh; struct ext3_dir_entry_2 *de; struct super_block * sb; @@ -1469,7 +1468,7 @@ static int ext3_add_entry (handle_t *handle, struct dentry *dentry, ext3_mark_inode_dirty(handle, dir); } blocks = dir->i_size >> sb->s_blocksize_bits; - for (block = 0, offset = 0; block < blocks; block++) { + for (block = 0; block < blocks; block++) { bh = ext3_bread(handle, dir, block, 0, &retval); if(!bh) return retval; diff --git a/fs/ext3/resize.c b/fs/ext3/resize.c index 54351ac7cef..0ccd7b12b73 100644 --- a/fs/ext3/resize.c +++ b/fs/ext3/resize.c @@ -964,7 +964,6 @@ int ext3_group_extend(struct super_block *sb, struct ext3_super_block *es, ext3_fsblk_t n_blocks_count) { ext3_fsblk_t o_blocks_count; - unsigned long o_groups_count; ext3_grpblk_t last; ext3_grpblk_t add; struct buffer_head * bh; @@ -976,7 +975,6 @@ int ext3_group_extend(struct super_block *sb, struct ext3_super_block *es, * yet: we're going to revalidate es->s_blocks_count after * taking the s_resize_lock below. */ o_blocks_count = le32_to_cpu(es->s_blocks_count); - o_groups_count = EXT3_SB(sb)->s_groups_count; if (test_opt(sb, DEBUG)) printk(KERN_DEBUG "EXT3-fs: extending last group from "E3FSBLK" uto "E3FSBLK" blocks\n", diff --git a/fs/ext3/super.c b/fs/ext3/super.c index 6c953bb255e..9650a956fd0 100644 --- a/fs/ext3/super.c +++ b/fs/ext3/super.c @@ -661,9 +661,6 @@ static int ext3_show_options(struct seq_file *seq, struct vfsmount *vfs) */ seq_puts(seq, ",barrier="); seq_puts(seq, test_opt(sb, BARRIER) ? "1" : "0"); - if (test_opt(sb, NOBH)) - seq_puts(seq, ",nobh"); - seq_printf(seq, ",data=%s", data_mode_string(test_opt(sb, DATA_FLAGS))); if (test_opt(sb, DATA_ERR_ABORT)) seq_puts(seq, ",data_err=abort"); @@ -1255,10 +1252,12 @@ set_qf_format: *n_blocks_count = option; break; case Opt_nobh: - set_opt(sbi->s_mount_opt, NOBH); + ext3_msg(sb, KERN_WARNING, + "warning: ignoring deprecated nobh option"); break; case Opt_bh: - clear_opt(sbi->s_mount_opt, NOBH); + ext3_msg(sb, KERN_WARNING, + "warning: ignoring deprecated bh option"); break; default: ext3_msg(sb, KERN_ERR, @@ -2001,14 +2000,6 @@ static int ext3_fill_super (struct super_block *sb, void *data, int silent) break; } - if (test_opt(sb, NOBH)) { - if (!(test_opt(sb, DATA_FLAGS) == EXT3_MOUNT_WRITEBACK_DATA)) { - ext3_msg(sb, KERN_WARNING, - "warning: ignoring nobh option - " - "it is supported only with writeback mode"); - clear_opt(sbi->s_mount_opt, NOBH); - } - } /* * The journal_load will have done any necessary log recovery, * so we can safely mount the rest of the filesystem now. diff --git a/fs/ext4/acl.c b/fs/ext4/acl.c index feaf498feaa..5e2ed4504ea 100644 --- a/fs/ext4/acl.c +++ b/fs/ext4/acl.c @@ -204,6 +204,7 @@ ext4_set_acl(handle_t *handle, struct inode *inode, int type, return error; else { inode->i_mode = mode; + inode->i_ctime = ext4_current_time(inode); ext4_mark_inode_dirty(handle, inode); if (error == 0) acl = NULL; diff --git a/fs/ext4/balloc.c b/fs/ext4/balloc.c index 95b7594c76f..bd30799a43e 100644 --- a/fs/ext4/balloc.c +++ b/fs/ext4/balloc.c @@ -377,14 +377,11 @@ void ext4_add_groupblocks(handle_t *handle, struct super_block *sb, ext4_grpblk_t bit; unsigned int i; struct ext4_group_desc *desc; - struct ext4_super_block *es; - struct ext4_sb_info *sbi; + struct ext4_sb_info *sbi = EXT4_SB(sb); int err = 0, ret, blk_free_count; ext4_grpblk_t blocks_freed; struct ext4_group_info *grp; - sbi = EXT4_SB(sb); - es = sbi->s_es; ext4_debug("Adding block(s) %llu-%llu\n", block, block + count - 1); ext4_get_group_no_and_offset(sb, block, &block_group, &bit); @@ -477,7 +474,6 @@ void ext4_add_groupblocks(handle_t *handle, struct super_block *sb, ret = ext4_handle_dirty_metadata(handle, NULL, gd_bh); if (!err) err = ret; - sb->s_dirt = 1; error_return: brelse(bitmap_bh); diff --git a/fs/ext4/block_validity.c b/fs/ext4/block_validity.c index 5b6973fbf1b..3db5084db9b 100644 --- a/fs/ext4/block_validity.c +++ b/fs/ext4/block_validity.c @@ -229,16 +229,20 @@ int ext4_data_block_valid(struct ext4_sb_info *sbi, ext4_fsblk_t start_blk, if ((start_blk <= le32_to_cpu(sbi->s_es->s_first_data_block)) || (start_blk + count < start_blk) || - (start_blk + count > ext4_blocks_count(sbi->s_es))) + (start_blk + count > ext4_blocks_count(sbi->s_es))) { + sbi->s_es->s_last_error_block = cpu_to_le64(start_blk); return 0; + } while (n) { entry = rb_entry(n, struct ext4_system_zone, node); if (start_blk + count - 1 < entry->start_blk) n = n->rb_left; else if (start_blk >= (entry->start_blk + entry->count)) n = n->rb_right; - else + else { + sbi->s_es->s_last_error_block = cpu_to_le64(start_blk); return 0; + } } return 1; } diff --git a/fs/ext4/dir.c b/fs/ext4/dir.c index ea5e6cb7e2a..374510f72ba 100644 --- a/fs/ext4/dir.c +++ b/fs/ext4/dir.c @@ -61,10 +61,11 @@ static unsigned char get_dtype(struct super_block *sb, int filetype) } -int ext4_check_dir_entry(const char *function, struct inode *dir, - struct ext4_dir_entry_2 *de, - struct buffer_head *bh, - unsigned int offset) +int __ext4_check_dir_entry(const char *function, unsigned int line, + struct inode *dir, + struct ext4_dir_entry_2 *de, + struct buffer_head *bh, + unsigned int offset) { const char *error_msg = NULL; const int rlen = ext4_rec_len_from_disk(de->rec_len, @@ -83,11 +84,10 @@ int ext4_check_dir_entry(const char *function, struct inode *dir, error_msg = "inode out of bounds"; if (error_msg != NULL) - ext4_error_inode(function, dir, - "bad entry in directory: %s - block=%llu" + ext4_error_inode(dir, function, line, bh->b_blocknr, + "bad entry in directory: %s - " "offset=%u(%u), inode=%u, rec_len=%d, name_len=%d", - error_msg, (unsigned long long) bh->b_blocknr, - (unsigned) (offset%bh->b_size), offset, + error_msg, (unsigned) (offset%bh->b_size), offset, le32_to_cpu(de->inode), rlen, de->name_len); return error_msg == NULL ? 1 : 0; @@ -121,7 +121,8 @@ static int ext4_readdir(struct file *filp, * We don't set the inode dirty flag since it's not * critical that it get flushed back to the disk. */ - ext4_clear_inode_flag(filp->f_path.dentry->d_inode, EXT4_INODE_INDEX); + ext4_clear_inode_flag(filp->f_path.dentry->d_inode, + EXT4_INODE_INDEX); } stored = 0; offset = filp->f_pos & (sb->s_blocksize - 1); @@ -193,7 +194,7 @@ revalidate: while (!error && filp->f_pos < inode->i_size && offset < sb->s_blocksize) { de = (struct ext4_dir_entry_2 *) (bh->b_data + offset); - if (!ext4_check_dir_entry("ext4_readdir", inode, de, + if (!ext4_check_dir_entry(inode, de, bh, offset)) { /* * On error, skip the f_pos to the next block @@ -343,7 +344,7 @@ int ext4_htree_store_dirent(struct file *dir_file, __u32 hash, struct dir_private_info *info; int len; - info = (struct dir_private_info *) dir_file->private_data; + info = dir_file->private_data; p = &info->root.rb_node; /* Create and allocate the fname structure */ diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 19a4de57128..e03841d9f30 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -57,10 +57,13 @@ #endif #define EXT4_ERROR_INODE(inode, fmt, a...) \ - ext4_error_inode(__func__, (inode), (fmt), ## a) + ext4_error_inode((inode), __func__, __LINE__, 0, (fmt), ## a) + +#define EXT4_ERROR_INODE_BLOCK(inode, block, fmt, a...) \ + ext4_error_inode((inode), __func__, __LINE__, (block), (fmt), ## a) #define EXT4_ERROR_FILE(file, fmt, a...) \ - ext4_error_file(__func__, (file), (fmt), ## a) + ext4_error_file(__func__, __LINE__, (file), (fmt), ## a) /* data type for block offset of block group */ typedef int ext4_grpblk_t; @@ -167,13 +170,15 @@ struct mpage_da_data { }; #define EXT4_IO_UNWRITTEN 0x1 typedef struct ext4_io_end { - struct list_head list; /* per-file finished AIO list */ + struct list_head list; /* per-file finished IO list */ struct inode *inode; /* file being written to */ unsigned int flag; /* unwritten or not */ struct page *page; /* page struct for buffer write */ loff_t offset; /* offset in the file */ ssize_t size; /* size of the extent */ struct work_struct work; /* data work queue */ + struct kiocb *iocb; /* iocb struct for AIO */ + int result; /* error value for AIO */ } ext4_io_end_t; /* @@ -460,7 +465,7 @@ struct ext4_new_group_data { }; /* - * Flags used by ext4_get_blocks() + * Flags used by ext4_map_blocks() */ /* Allocate any needed blocks and/or convert an unitialized extent to be an initialized ext4 */ @@ -873,7 +878,6 @@ struct ext4_inode_info { #define EXT4_MOUNT_POSIX_ACL 0x08000 /* POSIX Access Control Lists */ #define EXT4_MOUNT_NO_AUTO_DA_ALLOC 0x10000 /* No auto delalloc mapping */ #define EXT4_MOUNT_BARRIER 0x20000 /* Use block barriers */ -#define EXT4_MOUNT_NOBH 0x40000 /* No bufferheads */ #define EXT4_MOUNT_QUOTA 0x80000 /* Some quota option set */ #define EXT4_MOUNT_USRQUOTA 0x100000 /* "old" user quota */ #define EXT4_MOUNT_GRPQUOTA 0x200000 /* "old" group quota */ @@ -982,7 +986,7 @@ struct ext4_super_block { __le32 s_last_orphan; /* start of list of inodes to delete */ __le32 s_hash_seed[4]; /* HTREE hash seed */ __u8 s_def_hash_version; /* Default hash version to use */ - __u8 s_reserved_char_pad; + __u8 s_jnl_backup_type; __le16 s_desc_size; /* size of group descriptor */ /*100*/ __le32 s_default_mount_opts; __le32 s_first_meta_bg; /* First metablock block group */ @@ -1000,12 +1004,34 @@ struct ext4_super_block { __le64 s_mmp_block; /* Block for multi-mount protection */ __le32 s_raid_stripe_width; /* blocks on all data disks (N*stride)*/ __u8 s_log_groups_per_flex; /* FLEX_BG group size */ - __u8 s_reserved_char_pad2; + __u8 s_reserved_char_pad; __le16 s_reserved_pad; __le64 s_kbytes_written; /* nr of lifetime kilobytes written */ - __u32 s_reserved[160]; /* Padding to the end of the block */ + __le32 s_snapshot_inum; /* Inode number of active snapshot */ + __le32 s_snapshot_id; /* sequential ID of active snapshot */ + __le64 s_snapshot_r_blocks_count; /* reserved blocks for active + snapshot's future use */ + __le32 s_snapshot_list; /* inode number of the head of the + on-disk snapshot list */ +#define EXT4_S_ERR_START offsetof(struct ext4_super_block, s_error_count) + __le32 s_error_count; /* number of fs errors */ + __le32 s_first_error_time; /* first time an error happened */ + __le32 s_first_error_ino; /* inode involved in first error */ + __le64 s_first_error_block; /* block involved of first error */ + __u8 s_first_error_func[32]; /* function where the error happened */ + __le32 s_first_error_line; /* line number where error happened */ + __le32 s_last_error_time; /* most recent time of an error */ + __le32 s_last_error_ino; /* inode involved in last error */ + __le32 s_last_error_line; /* line number where error happened */ + __le64 s_last_error_block; /* block involved of last error */ + __u8 s_last_error_func[32]; /* function where the error happened */ +#define EXT4_S_ERR_END offsetof(struct ext4_super_block, s_mount_opts) + __u8 s_mount_opts[64]; + __le32 s_reserved[112]; /* Padding to the end of the block */ }; +#define EXT4_S_ERR_LEN (EXT4_S_ERR_END - EXT4_S_ERR_START) + #ifdef __KERNEL__ /* @@ -1143,6 +1169,9 @@ struct ext4_sb_info { /* workqueue for dio unwritten */ struct workqueue_struct *dio_unwritten_wq; + + /* timer for periodic error stats printing */ + struct timer_list s_err_report; }; static inline struct ext4_sb_info *EXT4_SB(struct super_block *sb) @@ -1313,6 +1342,10 @@ EXT4_INODE_BIT_FNS(state, state_flags) #define EXT4_DEFM_JMODE_DATA 0x0020 #define EXT4_DEFM_JMODE_ORDERED 0x0040 #define EXT4_DEFM_JMODE_WBACK 0x0060 +#define EXT4_DEFM_NOBARRIER 0x0100 +#define EXT4_DEFM_BLOCK_VALIDITY 0x0200 +#define EXT4_DEFM_DISCARD 0x0400 +#define EXT4_DEFM_NODELALLOC 0x0800 /* * Default journal batch times @@ -1379,6 +1412,43 @@ struct ext4_dir_entry_2 { #define EXT4_MAX_REC_LEN ((1<<16)-1) /* + * If we ever get support for fs block sizes > page_size, we'll need + * to remove the #if statements in the next two functions... + */ +static inline unsigned int +ext4_rec_len_from_disk(__le16 dlen, unsigned blocksize) +{ + unsigned len = le16_to_cpu(dlen); + +#if (PAGE_CACHE_SIZE >= 65536) + if (len == EXT4_MAX_REC_LEN || len == 0) + return blocksize; + return (len & 65532) | ((len & 3) << 16); +#else + return len; +#endif +} + +static inline __le16 ext4_rec_len_to_disk(unsigned len, unsigned blocksize) +{ + if ((len > blocksize) || (blocksize > (1 << 18)) || (len & 3)) + BUG(); +#if (PAGE_CACHE_SIZE >= 65536) + if (len < 65536) + return cpu_to_le16(len); + if (len == blocksize) { + if (blocksize == 65536) + return cpu_to_le16(EXT4_MAX_REC_LEN); + else + return cpu_to_le16(0); + } + return cpu_to_le16((len & 65532) | ((len >> 16) & 3)); +#else + return cpu_to_le16(len); +#endif +} + +/* * Hash Tree Directory indexing * (c) Daniel Phillips, 2001 */ @@ -1510,9 +1580,11 @@ extern unsigned ext4_init_block_bitmap(struct super_block *sb, ext4_init_block_bitmap(sb, NULL, group, desc) /* dir.c */ -extern int ext4_check_dir_entry(const char *, struct inode *, - struct ext4_dir_entry_2 *, - struct buffer_head *, unsigned int); +extern int __ext4_check_dir_entry(const char *, unsigned int, struct inode *, + struct ext4_dir_entry_2 *, + struct buffer_head *, unsigned int); +#define ext4_check_dir_entry(dir, de, bh, offset) \ + __ext4_check_dir_entry(__func__, __LINE__, (dir), (de), (bh), (offset)) extern int ext4_htree_store_dirent(struct file *dir_file, __u32 hash, __u32 minor_hash, struct ext4_dir_entry_2 *dirent); @@ -1601,8 +1673,6 @@ extern long ext4_compat_ioctl(struct file *, unsigned int, unsigned long); extern int ext4_ext_migrate(struct inode *); /* namei.c */ -extern unsigned int ext4_rec_len_from_disk(__le16 dlen, unsigned blocksize); -extern __le16 ext4_rec_len_to_disk(unsigned len, unsigned blocksize); extern int ext4_orphan_add(handle_t *, struct inode *); extern int ext4_orphan_del(handle_t *, struct inode *); extern int ext4_htree_fill_tree(struct file *dir_file, __u32 start_hash, @@ -1616,25 +1686,38 @@ extern int ext4_group_extend(struct super_block *sb, ext4_fsblk_t n_blocks_count); /* super.c */ -extern void __ext4_error(struct super_block *, const char *, const char *, ...) - __attribute__ ((format (printf, 3, 4))); -#define ext4_error(sb, message...) __ext4_error(sb, __func__, ## message) -extern void ext4_error_inode(const char *, struct inode *, const char *, ...) - __attribute__ ((format (printf, 3, 4))); -extern void ext4_error_file(const char *, struct file *, const char *, ...) - __attribute__ ((format (printf, 3, 4))); -extern void __ext4_std_error(struct super_block *, const char *, int); -extern void ext4_abort(struct super_block *, const char *, const char *, ...) - __attribute__ ((format (printf, 3, 4))); -extern void __ext4_warning(struct super_block *, const char *, +extern void __ext4_error(struct super_block *, const char *, unsigned int, + const char *, ...) + __attribute__ ((format (printf, 4, 5))); +#define ext4_error(sb, message...) __ext4_error(sb, __func__, \ + __LINE__, ## message) +extern void ext4_error_inode(struct inode *, const char *, unsigned int, + ext4_fsblk_t, const char *, ...) + __attribute__ ((format (printf, 5, 6))); +extern void ext4_error_file(struct file *, const char *, unsigned int, + const char *, ...) + __attribute__ ((format (printf, 4, 5))); +extern void __ext4_std_error(struct super_block *, const char *, + unsigned int, int); +extern void __ext4_abort(struct super_block *, const char *, unsigned int, + const char *, ...) + __attribute__ ((format (printf, 4, 5))); +#define ext4_abort(sb, message...) __ext4_abort(sb, __func__, \ + __LINE__, ## message) +extern void __ext4_warning(struct super_block *, const char *, unsigned int, const char *, ...) - __attribute__ ((format (printf, 3, 4))); -#define ext4_warning(sb, message...) __ext4_warning(sb, __func__, ## message) + __attribute__ ((format (printf, 4, 5))); +#define ext4_warning(sb, message...) __ext4_warning(sb, __func__, \ + __LINE__, ## message) extern void ext4_msg(struct super_block *, const char *, const char *, ...) __attribute__ ((format (printf, 3, 4))); -extern void ext4_grp_locked_error(struct super_block *, ext4_group_t, - const char *, const char *, ...) - __attribute__ ((format (printf, 4, 5))); +extern void __ext4_grp_locked_error(const char *, unsigned int, \ + struct super_block *, ext4_group_t, \ + unsigned long, ext4_fsblk_t, \ + const char *, ...) + __attribute__ ((format (printf, 7, 8))); +#define ext4_grp_locked_error(sb, grp, message...) \ + __ext4_grp_locked_error(__func__, __LINE__, (sb), (grp), ## message) extern void ext4_update_dynamic_rev(struct super_block *sb); extern int ext4_update_compat_feature(handle_t *handle, struct super_block *sb, __u32 compat); @@ -1768,7 +1851,7 @@ static inline unsigned int ext4_flex_bg_size(struct ext4_sb_info *sbi) #define ext4_std_error(sb, errno) \ do { \ if ((errno)) \ - __ext4_std_error((sb), __func__, (errno)); \ + __ext4_std_error((sb), __func__, __LINE__, (errno)); \ } while (0) #ifdef CONFIG_SMP @@ -1860,6 +1943,12 @@ static inline void ext4_unlock_group(struct super_block *sb, spin_unlock(ext4_group_lock_ptr(sb, group)); } +static inline void ext4_mark_super_dirty(struct super_block *sb) +{ + if (EXT4_SB(sb)->s_journal == NULL) + sb->s_dirt =1; +} + /* * Inodes and files operations */ @@ -1905,9 +1994,6 @@ extern int ext4_convert_unwritten_extents(struct inode *inode, loff_t offset, ssize_t len); extern int ext4_map_blocks(handle_t *handle, struct inode *inode, struct ext4_map_blocks *map, int flags); -extern int ext4_get_blocks(handle_t *handle, struct inode *inode, - sector_t block, unsigned int max_blocks, - struct buffer_head *bh, int flags); extern int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, __u64 start, __u64 len); /* move_extent.c */ diff --git a/fs/ext4/ext4_jbd2.c b/fs/ext4/ext4_jbd2.c index 53d2764d71c..6e272ef6ba9 100644 --- a/fs/ext4/ext4_jbd2.c +++ b/fs/ext4/ext4_jbd2.c @@ -6,29 +6,29 @@ #include <trace/events/ext4.h> -int __ext4_journal_get_undo_access(const char *where, handle_t *handle, - struct buffer_head *bh) +int __ext4_journal_get_undo_access(const char *where, unsigned int line, + handle_t *handle, struct buffer_head *bh) { int err = 0; if (ext4_handle_valid(handle)) { err = jbd2_journal_get_undo_access(handle, bh); if (err) - ext4_journal_abort_handle(where, __func__, bh, + ext4_journal_abort_handle(where, line, __func__, bh, handle, err); } return err; } -int __ext4_journal_get_write_access(const char *where, handle_t *handle, - struct buffer_head *bh) +int __ext4_journal_get_write_access(const char *where, unsigned int line, + handle_t *handle, struct buffer_head *bh) { int err = 0; if (ext4_handle_valid(handle)) { err = jbd2_journal_get_write_access(handle, bh); if (err) - ext4_journal_abort_handle(where, __func__, bh, + ext4_journal_abort_handle(where, line, __func__, bh, handle, err); } return err; @@ -46,9 +46,9 @@ int __ext4_journal_get_write_access(const char *where, handle_t *handle, * If the handle isn't valid we're not journaling, but we still need to * call into ext4_journal_revoke() to put the buffer head. */ -int __ext4_forget(const char *where, handle_t *handle, int is_metadata, - struct inode *inode, struct buffer_head *bh, - ext4_fsblk_t blocknr) +int __ext4_forget(const char *where, unsigned int line, handle_t *handle, + int is_metadata, struct inode *inode, + struct buffer_head *bh, ext4_fsblk_t blocknr) { int err; @@ -79,8 +79,8 @@ int __ext4_forget(const char *where, handle_t *handle, int is_metadata, BUFFER_TRACE(bh, "call jbd2_journal_forget"); err = jbd2_journal_forget(handle, bh); if (err) - ext4_journal_abort_handle(where, __func__, bh, - handle, err); + ext4_journal_abort_handle(where, line, __func__, + bh, handle, err); return err; } return 0; @@ -92,15 +92,16 @@ int __ext4_forget(const char *where, handle_t *handle, int is_metadata, BUFFER_TRACE(bh, "call jbd2_journal_revoke"); err = jbd2_journal_revoke(handle, blocknr, bh); if (err) { - ext4_journal_abort_handle(where, __func__, bh, handle, err); - ext4_abort(inode->i_sb, __func__, + ext4_journal_abort_handle(where, line, __func__, + bh, handle, err); + __ext4_abort(inode->i_sb, where, line, "error %d when attempting revoke", err); } BUFFER_TRACE(bh, "exit"); return err; } -int __ext4_journal_get_create_access(const char *where, +int __ext4_journal_get_create_access(const char *where, unsigned int line, handle_t *handle, struct buffer_head *bh) { int err = 0; @@ -108,22 +109,23 @@ int __ext4_journal_get_create_access(const char *where, if (ext4_handle_valid(handle)) { err = jbd2_journal_get_create_access(handle, bh); if (err) - ext4_journal_abort_handle(where, __func__, bh, - handle, err); + ext4_journal_abort_handle(where, line, __func__, + bh, handle, err); } return err; } -int __ext4_handle_dirty_metadata(const char *where, handle_t *handle, - struct inode *inode, struct buffer_head *bh) +int __ext4_handle_dirty_metadata(const char *where, unsigned int line, + handle_t *handle, struct inode *inode, + struct buffer_head *bh) { int err = 0; if (ext4_handle_valid(handle)) { err = jbd2_journal_dirty_metadata(handle, bh); if (err) - ext4_journal_abort_handle(where, __func__, bh, - handle, err); + ext4_journal_abort_handle(where, line, __func__, + bh, handle, err); } else { if (inode) mark_buffer_dirty_inode(bh, inode); @@ -132,14 +134,33 @@ int __ext4_handle_dirty_metadata(const char *where, handle_t *handle, if (inode && inode_needs_sync(inode)) { sync_dirty_buffer(bh); if (buffer_req(bh) && !buffer_uptodate(bh)) { - ext4_error(inode->i_sb, - "IO error syncing inode, " - "inode=%lu, block=%llu", - inode->i_ino, - (unsigned long long) bh->b_blocknr); + struct ext4_super_block *es; + + es = EXT4_SB(inode->i_sb)->s_es; + es->s_last_error_block = + cpu_to_le64(bh->b_blocknr); + ext4_error_inode(inode, where, line, + bh->b_blocknr, + "IO error syncing itable block"); err = -EIO; } } } return err; } + +int __ext4_handle_dirty_super(const char *where, unsigned int line, + handle_t *handle, struct super_block *sb) +{ + struct buffer_head *bh = EXT4_SB(sb)->s_sbh; + int err = 0; + + if (ext4_handle_valid(handle)) { + err = jbd2_journal_dirty_metadata(handle, bh); + if (err) + ext4_journal_abort_handle(where, line, __func__, + bh, handle, err); + } else + sb->s_dirt = 1; + return err; +} diff --git a/fs/ext4/ext4_jbd2.h b/fs/ext4/ext4_jbd2.h index dade0c02479..b0bd792c58c 100644 --- a/fs/ext4/ext4_jbd2.h +++ b/fs/ext4/ext4_jbd2.h @@ -122,39 +122,47 @@ int ext4_mark_inode_dirty(handle_t *handle, struct inode *inode); /* * Wrapper functions with which ext4 calls into JBD. */ -void ext4_journal_abort_handle(const char *caller, const char *err_fn, +void ext4_journal_abort_handle(const char *caller, unsigned int line, + const char *err_fn, struct buffer_head *bh, handle_t *handle, int err); -int __ext4_journal_get_undo_access(const char *where, handle_t *handle, - struct buffer_head *bh); +int __ext4_journal_get_undo_access(const char *where, unsigned int line, + handle_t *handle, struct buffer_head *bh); -int __ext4_journal_get_write_access(const char *where, handle_t *handle, - struct buffer_head *bh); +int __ext4_journal_get_write_access(const char *where, unsigned int line, + handle_t *handle, struct buffer_head *bh); -int __ext4_forget(const char *where, handle_t *handle, int is_metadata, - struct inode *inode, struct buffer_head *bh, - ext4_fsblk_t blocknr); +int __ext4_forget(const char *where, unsigned int line, handle_t *handle, + int is_metadata, struct inode *inode, + struct buffer_head *bh, ext4_fsblk_t blocknr); -int __ext4_journal_get_create_access(const char *where, +int __ext4_journal_get_create_access(const char *where, unsigned int line, handle_t *handle, struct buffer_head *bh); -int __ext4_handle_dirty_metadata(const char *where, handle_t *handle, - struct inode *inode, struct buffer_head *bh); +int __ext4_handle_dirty_metadata(const char *where, unsigned int line, + handle_t *handle, struct inode *inode, + struct buffer_head *bh); + +int __ext4_handle_dirty_super(const char *where, unsigned int line, + handle_t *handle, struct super_block *sb); #define ext4_journal_get_undo_access(handle, bh) \ - __ext4_journal_get_undo_access(__func__, (handle), (bh)) + __ext4_journal_get_undo_access(__func__, __LINE__, (handle), (bh)) #define ext4_journal_get_write_access(handle, bh) \ - __ext4_journal_get_write_access(__func__, (handle), (bh)) + __ext4_journal_get_write_access(__func__, __LINE__, (handle), (bh)) #define ext4_forget(handle, is_metadata, inode, bh, block_nr) \ - __ext4_forget(__func__, (handle), (is_metadata), (inode), (bh),\ - (block_nr)) + __ext4_forget(__func__, __LINE__, (handle), (is_metadata), (inode), \ + (bh), (block_nr)) #define ext4_journal_get_create_access(handle, bh) \ - __ext4_journal_get_create_access(__func__, (handle), (bh)) + __ext4_journal_get_create_access(__func__, __LINE__, (handle), (bh)) #define ext4_handle_dirty_metadata(handle, inode, bh) \ - __ext4_handle_dirty_metadata(__func__, (handle), (inode), (bh)) + __ext4_handle_dirty_metadata(__func__, __LINE__, (handle), (inode), \ + (bh)) +#define ext4_handle_dirty_super(handle, sb) \ + __ext4_handle_dirty_super(__func__, __LINE__, (handle), (sb)) handle_t *ext4_journal_start_sb(struct super_block *sb, int nblocks); -int __ext4_journal_stop(const char *where, handle_t *handle); +int __ext4_journal_stop(const char *where, unsigned int line, handle_t *handle); #define EXT4_NOJOURNAL_MAX_REF_COUNT ((unsigned long) 4096) @@ -207,7 +215,7 @@ static inline handle_t *ext4_journal_start(struct inode *inode, int nblocks) } #define ext4_journal_stop(handle) \ - __ext4_journal_stop(__func__, (handle)) + __ext4_journal_stop(__func__, __LINE__, (handle)) static inline handle_t *ext4_journal_current_handle(void) { @@ -308,17 +316,15 @@ static inline int ext4_should_writeback_data(struct inode *inode) * This function controls whether or not we should try to go down the * dioread_nolock code paths, which makes it safe to avoid taking * i_mutex for direct I/O reads. This only works for extent-based - * files, and it doesn't work for nobh or if data journaling is - * enabled, since the dioread_nolock code uses b_private to pass - * information back to the I/O completion handler, and this conflicts - * with the jbd's use of b_private. + * files, and it doesn't work if data journaling is enabled, since the + * dioread_nolock code uses b_private to pass information back to the + * I/O completion handler, and this conflicts with the jbd's use of + * b_private. */ static inline int ext4_should_dioread_nolock(struct inode *inode) { if (!test_opt(inode->i_sb, DIOREAD_NOLOCK)) return 0; - if (test_opt(inode->i_sb, NOBH)) - return 0; if (!S_ISREG(inode->i_mode)) return 0; if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))) diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index 377309c1af6..06328d3e571 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -401,9 +401,9 @@ static int ext4_valid_extent_entries(struct inode *inode, return 1; } -static int __ext4_ext_check(const char *function, struct inode *inode, - struct ext4_extent_header *eh, - int depth) +static int __ext4_ext_check(const char *function, unsigned int line, + struct inode *inode, struct ext4_extent_header *eh, + int depth) { const char *error_msg; int max = 0; @@ -436,7 +436,7 @@ static int __ext4_ext_check(const char *function, struct inode *inode, return 0; corrupted: - ext4_error_inode(function, inode, + ext4_error_inode(inode, function, line, 0, "bad header/extent: %s - magic %x, " "entries %u, max %u(%u), depth %u(%u)", error_msg, le16_to_cpu(eh->eh_magic), @@ -447,7 +447,7 @@ corrupted: } #define ext4_ext_check(inode, eh, depth) \ - __ext4_ext_check(__func__, inode, eh, depth) + __ext4_ext_check(__func__, __LINE__, inode, eh, depth) int ext4_ext_check_inode(struct inode *inode) { @@ -1083,7 +1083,6 @@ static int ext4_ext_grow_indepth(handle_t *handle, struct inode *inode, { struct ext4_ext_path *curp = path; struct ext4_extent_header *neh; - struct ext4_extent_idx *fidx; struct buffer_head *bh; ext4_fsblk_t newblock; int err = 0; @@ -1144,10 +1143,10 @@ static int ext4_ext_grow_indepth(handle_t *handle, struct inode *inode, ext4_idx_store_pblock(curp->p_idx, newblock); neh = ext_inode_hdr(inode); - fidx = EXT_FIRST_INDEX(neh); ext_debug("new root: num %d(%d), lblock %d, ptr %llu\n", le16_to_cpu(neh->eh_entries), le16_to_cpu(neh->eh_max), - le32_to_cpu(fidx->ei_block), idx_pblock(fidx)); + le32_to_cpu(EXT_FIRST_INDEX(neh)->ei_block), + idx_pblock(EXT_FIRST_INDEX(neh))); neh->eh_depth = cpu_to_le16(path->p_depth + 1); err = ext4_ext_dirty(handle, inode, curp); @@ -2937,7 +2936,7 @@ fix_extent_len: * One of more index blocks maybe needed if the extent tree grow after * the unintialized extent split. To prevent ENOSPC occur at the IO * complete, we need to split the uninitialized extent before DIO submit - * the IO. The uninitilized extent called at this time will be split + * the IO. The uninitialized extent called at this time will be split * into three uninitialized extent(at most). After IO complete, the part * being filled will be convert to initialized by the end_io callback function * via ext4_convert_unwritten_extents(). @@ -2954,7 +2953,6 @@ static int ext4_split_unwritten_extents(handle_t *handle, struct ext4_extent *ex1 = NULL; struct ext4_extent *ex2 = NULL; struct ext4_extent *ex3 = NULL; - struct ext4_extent_header *eh; ext4_lblk_t ee_block, eof_block; unsigned int allocated, ee_len, depth; ext4_fsblk_t newblock; @@ -2971,7 +2969,6 @@ static int ext4_split_unwritten_extents(handle_t *handle, eof_block = map->m_lblk + map->m_len; depth = ext_depth(inode); - eh = path[depth].p_hdr; ex = path[depth].p_ext; ee_block = le32_to_cpu(ex->ee_block); ee_len = ext4_ext_get_actual_len(ex); @@ -3058,7 +3055,6 @@ static int ext4_split_unwritten_extents(handle_t *handle, err = PTR_ERR(path); goto out; } - eh = path[depth].p_hdr; ex = path[depth].p_ext; if (ex2 != &newex) ex2 = ex; diff --git a/fs/ext4/file.c b/fs/ext4/file.c index 5313ae4cda2..ee92b66d455 100644 --- a/fs/ext4/file.c +++ b/fs/ext4/file.c @@ -70,7 +70,8 @@ ext4_file_write(struct kiocb *iocb, const struct iovec *iov, struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); size_t length = iov_length(iov, nr_segs); - if (pos > sbi->s_bitmap_maxbytes) + if ((pos > sbi->s_bitmap_maxbytes || + (pos == sbi->s_bitmap_maxbytes && length > 0))) return -EFBIG; if (pos + length > sbi->s_bitmap_maxbytes) { @@ -123,7 +124,7 @@ static int ext4_file_open(struct inode * inode, struct file * filp) if (!IS_ERR(cp)) { memcpy(sbi->s_es->s_last_mounted, cp, sizeof(sbi->s_es->s_last_mounted)); - sb->s_dirt = 1; + ext4_mark_super_dirty(sb); } } return dquot_file_open(inode, filp); diff --git a/fs/ext4/ialloc.c b/fs/ext4/ialloc.c index 25c4b3173fd..ac377505ed5 100644 --- a/fs/ext4/ialloc.c +++ b/fs/ext4/ialloc.c @@ -279,7 +279,7 @@ out: err = ext4_handle_dirty_metadata(handle, NULL, bitmap_bh); if (!fatal) fatal = err; - sb->s_dirt = 1; + ext4_mark_super_dirty(sb); } else ext4_error(sb, "bit already cleared for inode %lu", ino); @@ -965,7 +965,7 @@ got: percpu_counter_dec(&sbi->s_freeinodes_counter); if (S_ISDIR(mode)) percpu_counter_inc(&sbi->s_dirs_counter); - sb->s_dirt = 1; + ext4_mark_super_dirty(sb); if (sbi->s_log_groups_per_flex) { flex_group = ext4_flex_group(sbi, group); diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 42272d67955..a0ab3754d0d 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -221,6 +221,7 @@ void ext4_delete_inode(struct inode *inode) "couldn't extend journal (err %d)", err); stop_handle: ext4_journal_stop(handle); + ext4_orphan_del(NULL, inode); goto no_delete; } } @@ -337,9 +338,11 @@ static int ext4_block_to_path(struct inode *inode, return n; } -static int __ext4_check_blockref(const char *function, struct inode *inode, +static int __ext4_check_blockref(const char *function, unsigned int line, + struct inode *inode, __le32 *p, unsigned int max) { + struct ext4_super_block *es = EXT4_SB(inode->i_sb)->s_es; __le32 *bref = p; unsigned int blk; @@ -348,8 +351,9 @@ static int __ext4_check_blockref(const char *function, struct inode *inode, if (blk && unlikely(!ext4_data_block_valid(EXT4_SB(inode->i_sb), blk, 1))) { - ext4_error_inode(function, inode, - "invalid block reference %u", blk); + es->s_last_error_block = cpu_to_le64(blk); + ext4_error_inode(inode, function, line, blk, + "invalid block"); return -EIO; } } @@ -358,11 +362,13 @@ static int __ext4_check_blockref(const char *function, struct inode *inode, #define ext4_check_indirect_blockref(inode, bh) \ - __ext4_check_blockref(__func__, inode, (__le32 *)(bh)->b_data, \ + __ext4_check_blockref(__func__, __LINE__, inode, \ + (__le32 *)(bh)->b_data, \ EXT4_ADDR_PER_BLOCK((inode)->i_sb)) #define ext4_check_inode_blockref(inode) \ - __ext4_check_blockref(__func__, inode, EXT4_I(inode)->i_data, \ + __ext4_check_blockref(__func__, __LINE__, inode, \ + EXT4_I(inode)->i_data, \ EXT4_NDIR_BLOCKS) /** @@ -1128,20 +1134,24 @@ void ext4_da_update_reserve_space(struct inode *inode, ext4_discard_preallocations(inode); } -static int check_block_validity(struct inode *inode, const char *func, +static int __check_block_validity(struct inode *inode, const char *func, + unsigned int line, struct ext4_map_blocks *map) { if (!ext4_data_block_valid(EXT4_SB(inode->i_sb), map->m_pblk, map->m_len)) { - ext4_error_inode(func, inode, - "lblock %lu mapped to illegal pblock %llu " - "(length %d)", (unsigned long) map->m_lblk, - map->m_pblk, map->m_len); + ext4_error_inode(inode, func, line, map->m_pblk, + "lblock %lu mapped to illegal pblock " + "(length %d)", (unsigned long) map->m_lblk, + map->m_len); return -EIO; } return 0; } +#define check_block_validity(inode, map) \ + __check_block_validity((inode), __func__, __LINE__, (map)) + /* * Return the number of contiguous dirty pages in a given inode * starting at page frame idx. @@ -1244,7 +1254,7 @@ int ext4_map_blocks(handle_t *handle, struct inode *inode, up_read((&EXT4_I(inode)->i_data_sem)); if (retval > 0 && map->m_flags & EXT4_MAP_MAPPED) { - int ret = check_block_validity(inode, __func__, map); + int ret = check_block_validity(inode, map); if (ret != 0) return ret; } @@ -1324,9 +1334,7 @@ int ext4_map_blocks(handle_t *handle, struct inode *inode, up_write((&EXT4_I(inode)->i_data_sem)); if (retval > 0 && map->m_flags & EXT4_MAP_MAPPED) { - int ret = check_block_validity(inode, - "ext4_map_blocks_after_alloc", - map); + int ret = check_block_validity(inode, map); if (ret != 0) return ret; } @@ -1519,9 +1527,25 @@ static int walk_page_buffers(handle_t *handle, static int do_journal_get_write_access(handle_t *handle, struct buffer_head *bh) { + int dirty = buffer_dirty(bh); + int ret; + if (!buffer_mapped(bh) || buffer_freed(bh)) return 0; - return ext4_journal_get_write_access(handle, bh); + /* + * __block_prepare_write() could have dirtied some buffers. Clean + * the dirty bit as jbd2_journal_get_write_access() could complain + * otherwise about fs integrity issues. Setting of the dirty bit + * by __block_prepare_write() isn't a real problem here as we clear + * the bit before releasing a page lock and thus writeback cannot + * ever write the buffer. + */ + if (dirty) + clear_buffer_dirty(bh); + ret = ext4_journal_get_write_access(handle, bh); + if (!ret && dirty) + ret = ext4_handle_dirty_metadata(handle, NULL, bh); + return ret; } /* @@ -2194,7 +2218,7 @@ static int mpage_da_map_blocks(struct mpage_da_data *mpd) BUG_ON(!handle); /* - * Call ext4_get_blocks() to allocate any delayed allocation + * Call ext4_map_blocks() to allocate any delayed allocation * blocks, or to convert an uninitialized extent to be * initialized (in the case where we have written into * one or more preallocated blocks). @@ -2203,7 +2227,7 @@ static int mpage_da_map_blocks(struct mpage_da_data *mpd) * indicate that we are on the delayed allocation path. This * affects functions in many different parts of the allocation * call path. This flag exists primarily because we don't - * want to change *many* call functions, so ext4_get_blocks() + * want to change *many* call functions, so ext4_map_blocks() * will set the magic i_delalloc_reserved_flag once the * inode's allocation semaphore is taken. * @@ -2221,6 +2245,8 @@ static int mpage_da_map_blocks(struct mpage_da_data *mpd) blks = ext4_map_blocks(handle, mpd->inode, &map, get_blocks_flags); if (blks < 0) { + struct super_block *sb = mpd->inode->i_sb; + err = blks; /* * If get block returns with error we simply @@ -2231,7 +2257,7 @@ static int mpage_da_map_blocks(struct mpage_da_data *mpd) return 0; if (err == -ENOSPC && - ext4_count_free_blocks(mpd->inode->i_sb)) { + ext4_count_free_blocks(sb)) { mpd->retval = err; return 0; } @@ -2243,16 +2269,17 @@ static int mpage_da_map_blocks(struct mpage_da_data *mpd) * writepage and writepages will again try to write * the same. */ - ext4_msg(mpd->inode->i_sb, KERN_CRIT, - "delayed block allocation failed for inode %lu at " - "logical offset %llu with max blocks %zd with " - "error %d", mpd->inode->i_ino, - (unsigned long long) next, - mpd->b_size >> mpd->inode->i_blkbits, err); - printk(KERN_CRIT "This should not happen!! " - "Data will be lost\n"); - if (err == -ENOSPC) { - ext4_print_free_blocks(mpd->inode); + if (!(EXT4_SB(sb)->s_mount_flags & EXT4_MF_FS_ABORTED)) { + ext4_msg(sb, KERN_CRIT, + "delayed block allocation failed for inode %lu " + "at logical offset %llu with max blocks %zd " + "with error %d", mpd->inode->i_ino, + (unsigned long long) next, + mpd->b_size >> mpd->inode->i_blkbits, err); + ext4_msg(sb, KERN_CRIT, + "This should not happen!! Data will be lost\n"); + if (err == -ENOSPC) + ext4_print_free_blocks(mpd->inode); } /* invalidate all the pages */ ext4_da_block_invalidatepages(mpd, next, @@ -2320,7 +2347,7 @@ static void mpage_add_bh_to_extent(struct mpage_da_data *mpd, * XXX Don't go larger than mballoc is willing to allocate * This is a stopgap solution. We eventually need to fold * mpage_da_submit_io() into this function and then call - * ext4_get_blocks() multiple times in a loop + * ext4_map_blocks() multiple times in a loop */ if (nrblocks >= 8*1024*1024/mpd->inode->i_sb->s_blocksize) goto flush_it; @@ -2553,18 +2580,16 @@ static int ext4_da_get_block_prep(struct inode *inode, sector_t iblock, /* * This function is used as a standard get_block_t calback function * when there is no desire to allocate any blocks. It is used as a - * callback function for block_prepare_write(), nobh_writepage(), and - * block_write_full_page(). These functions should only try to map a - * single block at a time. + * callback function for block_prepare_write() and block_write_full_page(). + * These functions should only try to map a single block at a time. * * Since this function doesn't do block allocations even if the caller * requests it by passing in create=1, it is critically important that * any caller checks to make sure that any buffer heads are returned * by this function are either all already mapped or marked for - * delayed allocation before calling nobh_writepage() or - * block_write_full_page(). Otherwise, b_blocknr could be left - * unitialized, and the page write functions will be taken by - * surprise. + * delayed allocation before calling block_write_full_page(). Otherwise, + * b_blocknr could be left unitialized, and the page write functions will + * be taken by surprise. */ static int noalloc_get_block_write(struct inode *inode, sector_t iblock, struct buffer_head *bh_result, int create) @@ -2749,9 +2774,7 @@ static int ext4_writepage(struct page *page, return __ext4_journalled_writepage(page, len); } - if (test_opt(inode->i_sb, NOBH) && ext4_should_writeback_data(inode)) - ret = nobh_writepage(page, noalloc_get_block_write, wbc); - else if (page_bufs && buffer_uninit(page_bufs)) { + if (page_bufs && buffer_uninit(page_bufs)) { ext4_set_bh_endio(page_bufs, inode); ret = block_write_full_page_endio(page, noalloc_get_block_write, wbc, ext4_end_io_buffer_write); @@ -3146,13 +3169,10 @@ static int ext4_da_write_begin(struct file *file, struct address_space *mapping, int ret, retries = 0; struct page *page; pgoff_t index; - unsigned from, to; struct inode *inode = mapping->host; handle_t *handle; index = pos >> PAGE_CACHE_SHIFT; - from = pos & (PAGE_CACHE_SIZE - 1); - to = from + len; if (ext4_nonda_switch(inode->i_sb)) { *fsdata = (void *)FALL_BACK_TO_NONDELALLOC; @@ -3668,6 +3688,8 @@ static int ext4_end_io_nolock(ext4_io_end_t *io) return ret; } + if (io->iocb) + aio_complete(io->iocb, io->result, 0); /* clear the DIO AIO unwritten flag */ io->flag = 0; return ret; @@ -3767,6 +3789,8 @@ static ext4_io_end_t *ext4_init_io_end (struct inode *inode, gfp_t flags) io->offset = 0; io->size = 0; io->page = NULL; + io->iocb = NULL; + io->result = 0; INIT_WORK(&io->work, ext4_end_io_work); INIT_LIST_HEAD(&io->list); } @@ -3775,7 +3799,8 @@ static ext4_io_end_t *ext4_init_io_end (struct inode *inode, gfp_t flags) } static void ext4_end_io_dio(struct kiocb *iocb, loff_t offset, - ssize_t size, void *private) + ssize_t size, void *private, int ret, + bool is_async) { ext4_io_end_t *io_end = iocb->private; struct workqueue_struct *wq; @@ -3784,7 +3809,7 @@ static void ext4_end_io_dio(struct kiocb *iocb, loff_t offset, /* if not async direct IO or dio with 0 bytes write, just return */ if (!io_end || !size) - return; + goto out; ext_debug("ext4_end_io_dio(): io_end 0x%p" "for inode %lu, iocb 0x%p, offset %llu, size %llu\n", @@ -3795,12 +3820,18 @@ static void ext4_end_io_dio(struct kiocb *iocb, loff_t offset, if (io_end->flag != EXT4_IO_UNWRITTEN){ ext4_free_io_end(io_end); iocb->private = NULL; +out: + if (is_async) + aio_complete(iocb, ret, 0); return; } io_end->offset = offset; io_end->size = size; - io_end->flag = EXT4_IO_UNWRITTEN; + if (is_async) { + io_end->iocb = iocb; + io_end->result = ret; + } wq = EXT4_SB(io_end->inode->i_sb)->dio_unwritten_wq; /* queue the work to convert unwritten extents to written */ @@ -3937,7 +3968,7 @@ static ssize_t ext4_ext_direct_IO(int rw, struct kiocb *iocb, return -ENOMEM; /* * we save the io structure for current async - * direct IO, so that later ext4_get_blocks() + * direct IO, so that later ext4_map_blocks() * could flag the io structure whether there * is a unwritten extents needs to be converted * when IO is completed. @@ -4128,17 +4159,6 @@ int ext4_block_truncate_page(handle_t *handle, length = blocksize - (offset & (blocksize - 1)); iblock = index << (PAGE_CACHE_SHIFT - inode->i_sb->s_blocksize_bits); - /* - * For "nobh" option, we can only work if we don't need to - * read-in the page - otherwise we create buffers to do the IO. - */ - if (!page_has_buffers(page) && test_opt(inode->i_sb, NOBH) && - ext4_should_writeback_data(inode) && PageUptodate(page)) { - zero_user(page, offset, length); - set_page_dirty(page); - goto unlock; - } - if (!page_has_buffers(page)) create_empty_buffers(page, blocksize, 0); @@ -4488,9 +4508,8 @@ static void ext4_free_branches(handle_t *handle, struct inode *inode, * (should be rare). */ if (!bh) { - EXT4_ERROR_INODE(inode, - "Read failure block=%llu", - (unsigned long long) nr); + EXT4_ERROR_INODE_BLOCK(inode, nr, + "Read failure"); continue; } @@ -4502,27 +4521,6 @@ static void ext4_free_branches(handle_t *handle, struct inode *inode, depth); /* - * We've probably journalled the indirect block several - * times during the truncate. But it's no longer - * needed and we now drop it from the transaction via - * jbd2_journal_revoke(). - * - * That's easy if it's exclusively part of this - * transaction. But if it's part of the committing - * transaction then jbd2_journal_forget() will simply - * brelse() it. That means that if the underlying - * block is reallocated in ext4_get_block(), - * unmap_underlying_metadata() will find this block - * and will try to get rid of it. damn, damn. - * - * If this block has already been committed to the - * journal, a revoke record will be written. And - * revoke records must be emitted *before* clearing - * this block's bit in the bitmaps. - */ - ext4_forget(handle, 1, inode, bh, bh->b_blocknr); - - /* * Everything below this this pointer has been * released. Now let this top-of-subtree go. * @@ -4546,8 +4544,20 @@ static void ext4_free_branches(handle_t *handle, struct inode *inode, blocks_for_truncate(inode)); } + /* + * The forget flag here is critical because if + * we are journaling (and not doing data + * journaling), we have to make sure a revoke + * record is written to prevent the journal + * replay from overwriting the (former) + * indirect block if it gets reallocated as a + * data block. This must happen in the same + * transaction where the data blocks are + * actually freed. + */ ext4_free_blocks(handle, inode, 0, nr, 1, - EXT4_FREE_BLOCKS_METADATA); + EXT4_FREE_BLOCKS_METADATA| + EXT4_FREE_BLOCKS_FORGET); if (parent_bh) { /* @@ -4805,8 +4815,8 @@ static int __ext4_get_inode_loc(struct inode *inode, bh = sb_getblk(sb, block); if (!bh) { - EXT4_ERROR_INODE(inode, "unable to read inode block - " - "block %llu", block); + EXT4_ERROR_INODE_BLOCK(inode, block, + "unable to read itable block"); return -EIO; } if (!buffer_uptodate(bh)) { @@ -4904,8 +4914,8 @@ make_io: submit_bh(READ_META, bh); wait_on_buffer(bh); if (!buffer_uptodate(bh)) { - EXT4_ERROR_INODE(inode, "unable to read inode " - "block %llu", block); + EXT4_ERROR_INODE_BLOCK(inode, block, + "unable to read itable block"); brelse(bh); return -EIO; } @@ -4976,7 +4986,7 @@ static blkcnt_t ext4_inode_blocks(struct ext4_inode *raw_inode, /* we are using combined 48 bit field */ i_blocks = ((u64)le16_to_cpu(raw_inode->i_blocks_high)) << 32 | le32_to_cpu(raw_inode->i_blocks_lo); - if (ei->i_flags & EXT4_HUGE_FILE_FL) { + if (ext4_test_inode_flag(inode, EXT4_INODE_HUGE_FILE)) { /* i_blocks represent file system block size */ return i_blocks << (inode->i_blkbits - 9); } else { @@ -5072,7 +5082,7 @@ struct inode *ext4_iget(struct super_block *sb, unsigned long ino) transaction_t *transaction; tid_t tid; - spin_lock(&journal->j_state_lock); + read_lock(&journal->j_state_lock); if (journal->j_running_transaction) transaction = journal->j_running_transaction; else @@ -5081,7 +5091,7 @@ struct inode *ext4_iget(struct super_block *sb, unsigned long ino) tid = transaction->t_tid; else tid = journal->j_commit_sequence; - spin_unlock(&journal->j_state_lock); + read_unlock(&journal->j_state_lock); ei->i_sync_tid = tid; ei->i_datasync_tid = tid; } @@ -5126,7 +5136,7 @@ struct inode *ext4_iget(struct super_block *sb, unsigned long ino) ei->i_file_acl); ret = -EIO; goto bad_inode; - } else if (ei->i_flags & EXT4_EXTENTS_FL) { + } else if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) { if (S_ISREG(inode->i_mode) || S_ISDIR(inode->i_mode) || (S_ISLNK(inode->i_mode) && !ext4_inode_is_fast_symlink(inode))) @@ -5406,9 +5416,8 @@ int ext4_write_inode(struct inode *inode, struct writeback_control *wbc) if (wbc->sync_mode == WB_SYNC_ALL) sync_dirty_buffer(iloc.bh); if (buffer_req(iloc.bh) && !buffer_uptodate(iloc.bh)) { - EXT4_ERROR_INODE(inode, - "IO error syncing inode (block=%llu)", - (unsigned long long) iloc.bh->b_blocknr); + EXT4_ERROR_INODE_BLOCK(inode, iloc.bh->b_blocknr, + "IO error syncing inode"); err = -EIO; } brelse(iloc.bh); @@ -5483,10 +5492,8 @@ int ext4_setattr(struct dentry *dentry, struct iattr *attr) if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))) { struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); - if (attr->ia_size > sbi->s_bitmap_maxbytes) { - error = -EFBIG; - goto err_out; - } + if (attr->ia_size > sbi->s_bitmap_maxbytes) + return -EFBIG; } } @@ -5688,7 +5695,7 @@ int ext4_writepage_trans_blocks(struct inode *inode) * Calculate the journal credits for a chunk of data modification. * * This is called from DIO, fallocate or whoever calling - * ext4_get_blocks() to map/allocate a chunk of contiguous disk blocks. + * ext4_map_blocks() to map/allocate a chunk of contiguous disk blocks. * * journal buffers for data blocks are not included here, as DIO * and fallocate do no need to journal data buffers. @@ -5754,7 +5761,6 @@ static int ext4_expand_extra_isize(struct inode *inode, { struct ext4_inode *raw_inode; struct ext4_xattr_ibody_header *header; - struct ext4_xattr_entry *entry; if (EXT4_I(inode)->i_extra_isize >= new_extra_isize) return 0; @@ -5762,7 +5768,6 @@ static int ext4_expand_extra_isize(struct inode *inode, raw_inode = ext4_raw_inode(&iloc); header = IHDR(inode, raw_inode); - entry = IFIRST(header); /* No extended attributes present */ if (!ext4_test_inode_state(inode, EXT4_STATE_XATTR) || diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 12b3bc026a6..4b4ad4b7ce5 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -446,10 +446,11 @@ static void mb_free_blocks_double(struct inode *inode, struct ext4_buddy *e4b, blocknr = ext4_group_first_block_no(sb, e4b->bd_group); blocknr += first + i; ext4_grp_locked_error(sb, e4b->bd_group, - __func__, "double-free of inode" - " %lu's block %llu(bit %u in group %u)", - inode ? inode->i_ino : 0, blocknr, - first + i, e4b->bd_group); + inode ? inode->i_ino : 0, + blocknr, + "freeing block already freed " + "(bit %u)", + first + i); } mb_clear_bit(first + i, e4b->bd_info->bb_bitmap); } @@ -712,9 +713,9 @@ void ext4_mb_generate_buddy(struct super_block *sb, grp->bb_fragments = fragments; if (free != grp->bb_free) { - ext4_grp_locked_error(sb, group, __func__, - "EXT4-fs: group %u: %u blocks in bitmap, %u in gd", - group, free, grp->bb_free); + ext4_grp_locked_error(sb, group, 0, 0, + "%u blocks in bitmap, %u in gd", + free, grp->bb_free); /* * If we intent to continue, we consider group descritor * corrupt and update bb_free using bitmap value @@ -1296,10 +1297,10 @@ static void mb_free_blocks(struct inode *inode, struct ext4_buddy *e4b, blocknr = ext4_group_first_block_no(sb, e4b->bd_group); blocknr += block; ext4_grp_locked_error(sb, e4b->bd_group, - __func__, "double-free of inode" - " %lu's block %llu(bit %u in group %u)", - inode ? inode->i_ino : 0, blocknr, block, - e4b->bd_group); + inode ? inode->i_ino : 0, + blocknr, + "freeing already freed block " + "(bit %u)", block); } mb_clear_bit(block, EXT4_MB_BITMAP(e4b)); e4b->bd_info->bb_counters[order]++; @@ -1788,8 +1789,8 @@ void ext4_mb_complex_scan_group(struct ext4_allocation_context *ac, * free blocks even though group info says we * we have free blocks */ - ext4_grp_locked_error(sb, e4b->bd_group, - __func__, "%d free blocks as per " + ext4_grp_locked_error(sb, e4b->bd_group, 0, 0, + "%d free blocks as per " "group info. But bitmap says 0", free); break; @@ -1798,8 +1799,8 @@ void ext4_mb_complex_scan_group(struct ext4_allocation_context *ac, mb_find_extent(e4b, 0, i, ac->ac_g_ex.fe_len, &ex); BUG_ON(ex.fe_len <= 0); if (free < ex.fe_len) { - ext4_grp_locked_error(sb, e4b->bd_group, - __func__, "%d free blocks as per " + ext4_grp_locked_error(sb, e4b->bd_group, 0, 0, + "%d free blocks as per " "group info. But got %d blocks", free, ex.fe_len); /* @@ -1821,8 +1822,7 @@ void ext4_mb_complex_scan_group(struct ext4_allocation_context *ac, /* * This is a special case for storages like raid5 - * we try to find stripe-aligned chunks for stripe-size requests - * XXX should do so at least for multiples of stripe size as well + * we try to find stripe-aligned chunks for stripe-size-multiple requests */ static noinline_for_stack void ext4_mb_scan_aligned(struct ext4_allocation_context *ac, @@ -1999,7 +1999,6 @@ ext4_mb_regular_allocator(struct ext4_allocation_context *ac) ext4_group_t ngroups, group, i; int cr; int err = 0; - int bsbits; struct ext4_sb_info *sbi; struct super_block *sb; struct ext4_buddy e4b; @@ -2041,8 +2040,6 @@ ext4_mb_regular_allocator(struct ext4_allocation_context *ac) ac->ac_2order = i - 1; } - bsbits = ac->ac_sb->s_blocksize_bits; - /* if stream allocation is enabled, use global goal */ if (ac->ac_flags & EXT4_MB_STREAM_ALLOC) { /* TBD: may be hot point */ @@ -2094,8 +2091,8 @@ repeat: ac->ac_groups_scanned++; if (cr == 0) ext4_mb_simple_scan_group(ac, &e4b); - else if (cr == 1 && - ac->ac_g_ex.fe_len == sbi->s_stripe) + else if (cr == 1 && sbi->s_stripe && + !(ac->ac_g_ex.fe_len % sbi->s_stripe)) ext4_mb_scan_aligned(ac, &e4b); else ext4_mb_complex_scan_group(ac, &e4b); @@ -2221,7 +2218,7 @@ static int ext4_mb_seq_groups_open(struct inode *inode, struct file *file) rc = seq_open(file, &ext4_mb_seq_groups_ops); if (rc == 0) { - struct seq_file *m = (struct seq_file *)file->private_data; + struct seq_file *m = file->private_data; m->private = sb; } return rc; @@ -2560,6 +2557,22 @@ int ext4_mb_release(struct super_block *sb) return 0; } +static inline void ext4_issue_discard(struct super_block *sb, + ext4_group_t block_group, ext4_grpblk_t block, int count) +{ + int ret; + ext4_fsblk_t discard_block; + + discard_block = block + ext4_group_first_block_no(sb, block_group); + trace_ext4_discard_blocks(sb, + (unsigned long long) discard_block, count); + ret = sb_issue_discard(sb, discard_block, count); + if (ret == EOPNOTSUPP) { + ext4_warning(sb, "discard not supported, disabling"); + clear_opt(EXT4_SB(sb)->s_mount_opt, DISCARD); + } +} + /* * 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. @@ -2579,22 +2592,9 @@ static void release_blocks_on_commit(journal_t *journal, transaction_t *txn) mb_debug(1, "gonna free %u blocks in group %u (0x%p):", entry->count, entry->group, entry); - if (test_opt(sb, DISCARD)) { - int ret; - ext4_fsblk_t discard_block; - - discard_block = entry->start_blk + - ext4_group_first_block_no(sb, entry->group); - trace_ext4_discard_blocks(sb, - (unsigned long long)discard_block, - entry->count); - ret = sb_issue_discard(sb, discard_block, entry->count); - if (ret == EOPNOTSUPP) { - ext4_warning(sb, - "discard not supported, disabling"); - clear_opt(EXT4_SB(sb)->s_mount_opt, DISCARD); - } - } + if (test_opt(sb, DISCARD)) + ext4_issue_discard(sb, entry->group, + entry->start_blk, entry->count); err = ext4_mb_load_buddy(sb, entry->group, &e4b); /* we expect to find existing buddy because it's pinned */ @@ -2704,7 +2704,7 @@ void exit_ext4_mballoc(void) /* - * Check quota and mark choosed space (ac->ac_b_ex) non-free in bitmaps + * Check quota and mark chosen space (ac->ac_b_ex) non-free in bitmaps * Returns 0 if success or error code */ static noinline_for_stack int @@ -2712,7 +2712,6 @@ ext4_mb_mark_diskspace_used(struct ext4_allocation_context *ac, handle_t *handle, unsigned int reserv_blks) { struct buffer_head *bitmap_bh = NULL; - struct ext4_super_block *es; struct ext4_group_desc *gdp; struct buffer_head *gdp_bh; struct ext4_sb_info *sbi; @@ -2725,8 +2724,6 @@ ext4_mb_mark_diskspace_used(struct ext4_allocation_context *ac, sb = ac->ac_sb; sbi = EXT4_SB(sb); - es = sbi->s_es; - err = -EIO; bitmap_bh = ext4_read_block_bitmap(sb, ac->ac_b_ex.fe_group); @@ -2812,7 +2809,7 @@ ext4_mb_mark_diskspace_used(struct ext4_allocation_context *ac, err = ext4_handle_dirty_metadata(handle, NULL, gdp_bh); out_err: - sb->s_dirt = 1; + ext4_mark_super_dirty(sb); brelse(bitmap_bh); return err; } @@ -2850,7 +2847,7 @@ ext4_mb_normalize_request(struct ext4_allocation_context *ac, int bsbits, max; ext4_lblk_t end; loff_t size, orig_size, start_off; - ext4_lblk_t start, orig_start; + ext4_lblk_t start; struct ext4_inode_info *ei = EXT4_I(ac->ac_inode); struct ext4_prealloc_space *pa; @@ -2881,6 +2878,7 @@ ext4_mb_normalize_request(struct ext4_allocation_context *ac, size = size << bsbits; if (size < i_size_read(ac->ac_inode)) size = i_size_read(ac->ac_inode); + orig_size = size; /* max size of free chunks */ max = 2 << bsbits; @@ -2922,8 +2920,8 @@ ext4_mb_normalize_request(struct ext4_allocation_context *ac, start_off = (loff_t)ac->ac_o_ex.fe_logical << bsbits; size = ac->ac_o_ex.fe_len << bsbits; } - orig_size = size = size >> bsbits; - orig_start = start = start_off >> bsbits; + size = size >> bsbits; + start = start_off >> bsbits; /* don't cover already allocated blocks in selected range */ if (ar->pleft && start <= ar->lleft) { @@ -3547,7 +3545,6 @@ ext4_mb_release_inode_pa(struct ext4_buddy *e4b, struct buffer_head *bitmap_bh, ext4_group_t group; ext4_grpblk_t bit; unsigned long long grp_blk_start; - sector_t start; int err = 0; int free = 0; @@ -3567,10 +3564,9 @@ ext4_mb_release_inode_pa(struct ext4_buddy *e4b, struct buffer_head *bitmap_bh, if (bit >= end) break; next = mb_find_next_bit(bitmap_bh->b_data, end, bit); - start = ext4_group_first_block_no(sb, group) + bit; mb_debug(1, " free preallocated %u/%u in group %u\n", - (unsigned) start, (unsigned) next - bit, - (unsigned) group); + (unsigned) ext4_group_first_block_no(sb, group) + bit, + (unsigned) next - bit, (unsigned) group); free += next - bit; if (ac) { @@ -3581,7 +3577,7 @@ ext4_mb_release_inode_pa(struct ext4_buddy *e4b, struct buffer_head *bitmap_bh, trace_ext4_mballoc_discard(ac); } - trace_ext4_mb_release_inode_pa(ac, pa, grp_blk_start + bit, + trace_ext4_mb_release_inode_pa(sb, ac, pa, grp_blk_start + bit, next - bit); mb_free_blocks(pa->pa_inode, e4b, bit, next - bit); bit = next + 1; @@ -3591,8 +3587,7 @@ ext4_mb_release_inode_pa(struct ext4_buddy *e4b, struct buffer_head *bitmap_bh, pa, (unsigned long) pa->pa_lstart, (unsigned long) pa->pa_pstart, (unsigned long) pa->pa_len); - ext4_grp_locked_error(sb, group, - __func__, "free %u, pa_free %u", + ext4_grp_locked_error(sb, group, 0, 0, "free %u, pa_free %u", free, pa->pa_free); /* * pa is already deleted so we use the value obtained @@ -3613,7 +3608,7 @@ ext4_mb_release_group_pa(struct ext4_buddy *e4b, ext4_group_t group; ext4_grpblk_t bit; - trace_ext4_mb_release_group_pa(ac, pa); + trace_ext4_mb_release_group_pa(sb, ac, pa); BUG_ON(pa->pa_deleted == 0); ext4_get_group_no_and_offset(sb, pa->pa_pstart, &group, &bit); BUG_ON(group != e4b->bd_group && pa->pa_len != 0); @@ -3889,6 +3884,9 @@ static void ext4_mb_show_ac(struct ext4_allocation_context *ac) struct super_block *sb = ac->ac_sb; ext4_group_t ngroups, i; + if (EXT4_SB(sb)->s_mount_flags & EXT4_MF_FS_ABORTED) + return; + printk(KERN_ERR "EXT4-fs: Can't allocate:" " Allocation context details:\n"); printk(KERN_ERR "EXT4-fs: status %d flags %d\n", @@ -4255,7 +4253,7 @@ static int ext4_mb_discard_preallocations(struct super_block *sb, int needed) * to usual allocation */ ext4_fsblk_t ext4_mb_new_blocks(handle_t *handle, - struct ext4_allocation_request *ar, int *errp) + struct ext4_allocation_request *ar, int *errp) { int freed; struct ext4_allocation_context *ac = NULL; @@ -4299,7 +4297,7 @@ ext4_fsblk_t ext4_mb_new_blocks(handle_t *handle, inquota = ar->len; if (ar->len == 0) { *errp = -EDQUOT; - goto out3; + goto out; } } @@ -4307,13 +4305,13 @@ ext4_fsblk_t ext4_mb_new_blocks(handle_t *handle, if (!ac) { ar->len = 0; *errp = -ENOMEM; - goto out1; + goto out; } *errp = ext4_mb_initialize_context(ac, ar); if (*errp) { ar->len = 0; - goto out2; + goto out; } ac->ac_op = EXT4_MB_HISTORY_PREALLOC; @@ -4322,7 +4320,9 @@ ext4_fsblk_t ext4_mb_new_blocks(handle_t *handle, ext4_mb_normalize_request(ac, ar); repeat: /* allocate space in core */ - ext4_mb_regular_allocator(ac); + *errp = ext4_mb_regular_allocator(ac); + if (*errp) + goto errout; /* as we've just preallocated more space than * user requested orinally, we store allocated @@ -4333,7 +4333,7 @@ repeat: } if (likely(ac->ac_status == AC_STATUS_FOUND)) { *errp = ext4_mb_mark_diskspace_used(ac, handle, reserv_blks); - if (*errp == -EAGAIN) { + if (*errp == -EAGAIN) { /* * drop the reference that we took * in ext4_mb_use_best_found @@ -4344,12 +4344,10 @@ repeat: ac->ac_b_ex.fe_len = 0; ac->ac_status = AC_STATUS_CONTINUE; goto repeat; - } else if (*errp) { + } else if (*errp) + errout: ext4_discard_allocated_blocks(ac); - ac->ac_b_ex.fe_len = 0; - ar->len = 0; - ext4_mb_show_ac(ac); - } else { + else { block = ext4_grp_offs_to_block(sb, &ac->ac_b_ex); ar->len = ac->ac_b_ex.fe_len; } @@ -4358,19 +4356,19 @@ repeat: if (freed) goto repeat; *errp = -ENOSPC; + } + + if (*errp) { ac->ac_b_ex.fe_len = 0; ar->len = 0; ext4_mb_show_ac(ac); } - ext4_mb_release_context(ac); - -out2: - kmem_cache_free(ext4_ac_cachep, ac); -out1: +out: + if (ac) + kmem_cache_free(ext4_ac_cachep, ac); if (inquota && ar->len < inquota) dquot_free_block(ar->inode, inquota - ar->len); -out3: if (!ar->len) { if (!EXT4_I(ar->inode)->i_delalloc_reserved_flag) /* release all the reserved blocks if non delalloc */ @@ -4402,6 +4400,7 @@ static noinline_for_stack int ext4_mb_free_metadata(handle_t *handle, struct ext4_buddy *e4b, struct ext4_free_data *new_entry) { + ext4_group_t group = e4b->bd_group; ext4_grpblk_t block; struct ext4_free_data *entry; struct ext4_group_info *db = e4b->bd_info; @@ -4434,9 +4433,9 @@ ext4_mb_free_metadata(handle_t *handle, struct ext4_buddy *e4b, else if (block >= (entry->start_blk + entry->count)) n = &(*n)->rb_right; else { - ext4_grp_locked_error(sb, e4b->bd_group, __func__, - "Double free of blocks %d (%d %d)", - block, entry->start_blk, entry->count); + ext4_grp_locked_error(sb, group, 0, + ext4_group_first_block_no(sb, group) + block, + "Block already on to-be-freed list"); return 0; } } @@ -4494,7 +4493,6 @@ void ext4_free_blocks(handle_t *handle, struct inode *inode, struct super_block *sb = inode->i_sb; struct ext4_allocation_context *ac = NULL; struct ext4_group_desc *gdp; - struct ext4_super_block *es; unsigned long freed = 0; unsigned int overflow; ext4_grpblk_t bit; @@ -4513,7 +4511,6 @@ void ext4_free_blocks(handle_t *handle, struct inode *inode, } sbi = EXT4_SB(sb); - es = EXT4_SB(sb)->s_es; if (!(flags & EXT4_FREE_BLOCKS_VALIDATED) && !ext4_data_block_valid(sbi, block, count)) { ext4_error(sb, "Freeing blocks not in datazone - " @@ -4647,6 +4644,8 @@ do_more: mb_clear_bits(bitmap_bh->b_data, bit, count); mb_free_blocks(inode, &e4b, bit, count); ext4_mb_return_to_preallocation(inode, &e4b, block, count); + if (test_opt(sb, DISCARD)) + ext4_issue_discard(sb, block_group, bit, count); } ret = ext4_free_blks_count(sb, gdp) + count; @@ -4680,7 +4679,7 @@ do_more: put_bh(bitmap_bh); goto do_more; } - sb->s_dirt = 1; + ext4_mark_super_dirty(sb); error_return: if (freed) dquot_free_block(inode, freed); diff --git a/fs/ext4/migrate.c b/fs/ext4/migrate.c index 6f3a27ec30b..1765c2c50a9 100644 --- a/fs/ext4/migrate.c +++ b/fs/ext4/migrate.c @@ -376,7 +376,7 @@ static int ext4_ext_swap_inode_data(handle_t *handle, struct inode *inode, * We have the extent map build with the tmp inode. * Now copy the i_data across */ - ei->i_flags |= EXT4_EXTENTS_FL; + ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS); memcpy(ei->i_data, tmp_ei->i_data, sizeof(ei->i_data)); /* diff --git a/fs/ext4/move_extent.c b/fs/ext4/move_extent.c index 52abfa12762..5f1ed9fc913 100644 --- a/fs/ext4/move_extent.c +++ b/fs/ext4/move_extent.c @@ -148,17 +148,17 @@ mext_next_extent(struct inode *inode, struct ext4_ext_path *path, */ static int mext_check_null_inode(struct inode *inode1, struct inode *inode2, - const char *function) + const char *function, unsigned int line) { int ret = 0; if (inode1 == NULL) { - __ext4_error(inode2->i_sb, function, + __ext4_error(inode2->i_sb, function, line, "Both inodes should not be NULL: " "inode1 NULL inode2 %lu", inode2->i_ino); ret = -EIO; } else if (inode2 == NULL) { - __ext4_error(inode1->i_sb, function, + __ext4_error(inode1->i_sb, function, line, "Both inodes should not be NULL: " "inode1 %lu inode2 NULL", inode1->i_ino); ret = -EIO; @@ -1084,7 +1084,7 @@ mext_inode_double_lock(struct inode *inode1, struct inode *inode2) BUG_ON(inode1 == NULL && inode2 == NULL); - ret = mext_check_null_inode(inode1, inode2, __func__); + ret = mext_check_null_inode(inode1, inode2, __func__, __LINE__); if (ret < 0) goto out; @@ -1121,7 +1121,7 @@ mext_inode_double_unlock(struct inode *inode1, struct inode *inode2) BUG_ON(inode1 == NULL && inode2 == NULL); - ret = mext_check_null_inode(inode1, inode2, __func__); + ret = mext_check_null_inode(inode1, inode2, __func__, __LINE__); if (ret < 0) goto out; diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c index a43e6617b35..314c0d3b3fa 100644 --- a/fs/ext4/namei.c +++ b/fs/ext4/namei.c @@ -179,30 +179,6 @@ static struct buffer_head * ext4_dx_find_entry(struct inode *dir, static int ext4_dx_add_entry(handle_t *handle, struct dentry *dentry, struct inode *inode); -unsigned int ext4_rec_len_from_disk(__le16 dlen, unsigned blocksize) -{ - unsigned len = le16_to_cpu(dlen); - - if (len == EXT4_MAX_REC_LEN || len == 0) - return blocksize; - return (len & 65532) | ((len & 3) << 16); -} - -__le16 ext4_rec_len_to_disk(unsigned len, unsigned blocksize) -{ - if ((len > blocksize) || (blocksize > (1 << 18)) || (len & 3)) - BUG(); - if (len < 65536) - return cpu_to_le16(len); - if (len == blocksize) { - if (blocksize == 65536) - return cpu_to_le16(EXT4_MAX_REC_LEN); - else - return cpu_to_le16(0); - } - return cpu_to_le16((len & 65532) | ((len >> 16) & 3)); -} - /* * p is at least 6 bytes before the end of page */ @@ -605,7 +581,7 @@ static int htree_dirblock_to_tree(struct file *dir_file, dir->i_sb->s_blocksize - EXT4_DIR_REC_LEN(0)); for (; de < top; de = ext4_next_entry(de, dir->i_sb->s_blocksize)) { - if (!ext4_check_dir_entry("htree_dirblock_to_tree", dir, de, bh, + if (!ext4_check_dir_entry(dir, de, bh, (block<<EXT4_BLOCK_SIZE_BITS(dir->i_sb)) +((char *)de - bh->b_data))) { /* On error, skip the f_pos to the next block. */ @@ -844,8 +820,7 @@ static inline int search_dirblock(struct buffer_head *bh, if ((char *) de + namelen <= dlimit && ext4_match (namelen, name, de)) { /* found a match - just to be sure, do a full check */ - if (!ext4_check_dir_entry("ext4_find_entry", - dir, de, bh, offset)) + if (!ext4_check_dir_entry(dir, de, bh, offset)) return -1; *res_dir = de; return 1; @@ -1019,7 +994,7 @@ static struct buffer_head * ext4_dx_find_entry(struct inode *dir, const struct q int off = (block << EXT4_BLOCK_SIZE_BITS(sb)) + ((char *) de - bh->b_data); - if (!ext4_check_dir_entry(__func__, dir, de, bh, off)) { + if (!ext4_check_dir_entry(dir, de, bh, off)) { brelse(bh); *err = ERR_BAD_DX_DIR; goto errout; @@ -1088,7 +1063,6 @@ static struct dentry *ext4_lookup(struct inode *dir, struct dentry *dentry, stru struct dentry *ext4_get_parent(struct dentry *child) { __u32 ino; - struct inode *inode; static const struct qstr dotdot = { .name = "..", .len = 2, @@ -1097,7 +1071,6 @@ struct dentry *ext4_get_parent(struct dentry *child) struct buffer_head *bh; bh = ext4_find_entry(child->d_inode, &dotdot, &de); - inode = NULL; if (!bh) return ERR_PTR(-ENOENT); ino = le32_to_cpu(de->inode); @@ -1305,8 +1278,7 @@ static int add_dirent_to_buf(handle_t *handle, struct dentry *dentry, de = (struct ext4_dir_entry_2 *)bh->b_data; top = bh->b_data + blocksize - reclen; while ((char *) de <= top) { - if (!ext4_check_dir_entry("ext4_add_entry", dir, de, - bh, offset)) + if (!ext4_check_dir_entry(dir, de, bh, offset)) return -EIO; if (ext4_match(namelen, name, de)) return -EEXIST; @@ -1673,7 +1645,7 @@ static int ext4_delete_entry(handle_t *handle, pde = NULL; de = (struct ext4_dir_entry_2 *) bh->b_data; while (i < bh->b_size) { - if (!ext4_check_dir_entry("ext4_delete_entry", dir, de, bh, i)) + if (!ext4_check_dir_entry(dir, de, bh, i)) return -EIO; if (de == de_del) { BUFFER_TRACE(bh, "get_write_access"); @@ -1956,7 +1928,7 @@ static int empty_dir(struct inode *inode) } de = (struct ext4_dir_entry_2 *) bh->b_data; } - if (!ext4_check_dir_entry("empty_dir", inode, de, bh, offset)) { + if (!ext4_check_dir_entry(inode, de, bh, offset)) { de = (struct ext4_dir_entry_2 *)(bh->b_data + sb->s_blocksize); offset = (offset | (sb->s_blocksize - 1)) + 1; diff --git a/fs/ext4/resize.c b/fs/ext4/resize.c index 6df797eb9ae..ca5c8aa00a2 100644 --- a/fs/ext4/resize.c +++ b/fs/ext4/resize.c @@ -921,8 +921,7 @@ int ext4_group_add(struct super_block *sb, struct ext4_new_group_data *input) &sbi->s_flex_groups[flex_group].free_inodes); } - ext4_handle_dirty_metadata(handle, NULL, sbi->s_sbh); - sb->s_dirt = 1; + ext4_handle_dirty_super(handle, sb); exit_journal: mutex_unlock(&sbi->s_resize_lock); @@ -953,7 +952,6 @@ int ext4_group_extend(struct super_block *sb, struct ext4_super_block *es, ext4_fsblk_t n_blocks_count) { ext4_fsblk_t o_blocks_count; - ext4_group_t o_groups_count; ext4_grpblk_t last; ext4_grpblk_t add; struct buffer_head *bh; @@ -965,7 +963,6 @@ int ext4_group_extend(struct super_block *sb, struct ext4_super_block *es, * yet: we're going to revalidate es->s_blocks_count after * taking the s_resize_lock below. */ o_blocks_count = ext4_blocks_count(es); - o_groups_count = EXT4_SB(sb)->s_groups_count; if (test_opt(sb, DEBUG)) printk(KERN_DEBUG "EXT4-fs: extending last group from %llu uto %llu blocks\n", @@ -1045,13 +1042,12 @@ int ext4_group_extend(struct super_block *sb, struct ext4_super_block *es, goto exit_put; } ext4_blocks_count_set(es, o_blocks_count + add); - ext4_handle_dirty_metadata(handle, NULL, EXT4_SB(sb)->s_sbh); - sb->s_dirt = 1; mutex_unlock(&EXT4_SB(sb)->s_resize_lock); ext4_debug("freeing blocks %llu through %llu\n", o_blocks_count, o_blocks_count + add); /* We add the blocks to the bitmap and set the group need init bit */ ext4_add_groupblocks(handle, sb, o_blocks_count, add); + ext4_handle_dirty_super(handle, sb); ext4_debug("freed blocks %llu through %llu\n", o_blocks_count, o_blocks_count + add); if ((err = ext4_journal_stop(handle))) diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 4e8983a9811..8d65575f8c8 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -241,14 +241,14 @@ handle_t *ext4_journal_start_sb(struct super_block *sb, int nblocks) if (sb->s_flags & MS_RDONLY) return ERR_PTR(-EROFS); - vfs_check_frozen(sb, SB_FREEZE_WRITE); + vfs_check_frozen(sb, SB_FREEZE_TRANS); /* Special case here: if the journal has aborted behind our * backs (eg. EIO in the commit thread), then we still need to * take the FS itself readonly cleanly. */ journal = EXT4_SB(sb)->s_journal; if (journal) { if (is_journal_aborted(journal)) { - ext4_abort(sb, __func__, "Detected aborted journal"); + ext4_abort(sb, "Detected aborted journal"); return ERR_PTR(-EROFS); } return jbd2_journal_start(journal, nblocks); @@ -262,7 +262,7 @@ handle_t *ext4_journal_start_sb(struct super_block *sb, int nblocks) * that sync() will call the filesystem's write_super callback if * appropriate. */ -int __ext4_journal_stop(const char *where, handle_t *handle) +int __ext4_journal_stop(const char *where, unsigned int line, handle_t *handle) { struct super_block *sb; int err; @@ -279,12 +279,13 @@ int __ext4_journal_stop(const char *where, handle_t *handle) if (!err) err = rc; if (err) - __ext4_std_error(sb, where, err); + __ext4_std_error(sb, where, line, err); return err; } -void ext4_journal_abort_handle(const char *caller, const char *err_fn, - struct buffer_head *bh, handle_t *handle, int err) +void ext4_journal_abort_handle(const char *caller, unsigned int line, + const char *err_fn, struct buffer_head *bh, + handle_t *handle, int err) { char nbuf[16]; const char *errstr = ext4_decode_error(NULL, err, nbuf); @@ -300,12 +301,47 @@ void ext4_journal_abort_handle(const char *caller, const char *err_fn, if (is_handle_aborted(handle)) return; - printk(KERN_ERR "%s: aborting transaction: %s in %s\n", - caller, errstr, err_fn); + printk(KERN_ERR "%s:%d: aborting transaction: %s in %s\n", + caller, line, errstr, err_fn); jbd2_journal_abort_handle(handle); } +static void __save_error_info(struct super_block *sb, const char *func, + unsigned int line) +{ + struct ext4_super_block *es = EXT4_SB(sb)->s_es; + + EXT4_SB(sb)->s_mount_state |= EXT4_ERROR_FS; + es->s_state |= cpu_to_le16(EXT4_ERROR_FS); + es->s_last_error_time = cpu_to_le32(get_seconds()); + strncpy(es->s_last_error_func, func, sizeof(es->s_last_error_func)); + es->s_last_error_line = cpu_to_le32(line); + if (!es->s_first_error_time) { + es->s_first_error_time = es->s_last_error_time; + strncpy(es->s_first_error_func, func, + sizeof(es->s_first_error_func)); + es->s_first_error_line = cpu_to_le32(line); + es->s_first_error_ino = es->s_last_error_ino; + es->s_first_error_block = es->s_last_error_block; + } + /* + * Start the daily error reporting function if it hasn't been + * started already + */ + if (!es->s_error_count) + mod_timer(&EXT4_SB(sb)->s_err_report, jiffies + 24*60*60*HZ); + es->s_error_count = cpu_to_le32(le32_to_cpu(es->s_error_count) + 1); +} + +static void save_error_info(struct super_block *sb, const char *func, + unsigned int line) +{ + __save_error_info(sb, func, line); + ext4_commit_super(sb, 1); +} + + /* Deal with the reporting of failure conditions on a filesystem such as * inconsistencies detected or read IO failures. * @@ -323,11 +359,6 @@ void ext4_journal_abort_handle(const char *caller, const char *err_fn, static void ext4_handle_error(struct super_block *sb) { - struct ext4_super_block *es = EXT4_SB(sb)->s_es; - - EXT4_SB(sb)->s_mount_state |= EXT4_ERROR_FS; - es->s_state |= cpu_to_le16(EXT4_ERROR_FS); - if (sb->s_flags & MS_RDONLY) return; @@ -342,19 +373,19 @@ static void ext4_handle_error(struct super_block *sb) ext4_msg(sb, KERN_CRIT, "Remounting filesystem read-only"); sb->s_flags |= MS_RDONLY; } - ext4_commit_super(sb, 1); if (test_opt(sb, ERRORS_PANIC)) panic("EXT4-fs (device %s): panic forced after error\n", sb->s_id); } void __ext4_error(struct super_block *sb, const char *function, - const char *fmt, ...) + unsigned int line, const char *fmt, ...) { va_list args; va_start(args, fmt); - printk(KERN_CRIT "EXT4-fs error (device %s): %s: ", sb->s_id, function); + printk(KERN_CRIT "EXT4-fs error (device %s): %s:%d: comm %s: ", + sb->s_id, function, line, current->comm); vprintk(fmt, args); printk("\n"); va_end(args); @@ -362,14 +393,22 @@ void __ext4_error(struct super_block *sb, const char *function, ext4_handle_error(sb); } -void ext4_error_inode(const char *function, struct inode *inode, +void ext4_error_inode(struct inode *inode, const char *function, + unsigned int line, ext4_fsblk_t block, const char *fmt, ...) { va_list args; + struct ext4_super_block *es = EXT4_SB(inode->i_sb)->s_es; + es->s_last_error_ino = cpu_to_le32(inode->i_ino); + es->s_last_error_block = cpu_to_le64(block); + save_error_info(inode->i_sb, function, line); va_start(args, fmt); - printk(KERN_CRIT "EXT4-fs error (device %s): %s: inode #%lu: (comm %s) ", - inode->i_sb->s_id, function, inode->i_ino, current->comm); + printk(KERN_CRIT "EXT4-fs error (device %s): %s:%d: inode #%lu: ", + inode->i_sb->s_id, function, line, inode->i_ino); + if (block) + printk("block %llu: ", block); + printk("comm %s: ", current->comm); vprintk(fmt, args); printk("\n"); va_end(args); @@ -377,20 +416,26 @@ void ext4_error_inode(const char *function, struct inode *inode, ext4_handle_error(inode->i_sb); } -void ext4_error_file(const char *function, struct file *file, - const char *fmt, ...) +void ext4_error_file(struct file *file, const char *function, + unsigned int line, const char *fmt, ...) { va_list args; + struct ext4_super_block *es; struct inode *inode = file->f_dentry->d_inode; char pathname[80], *path; + es = EXT4_SB(inode->i_sb)->s_es; + es->s_last_error_ino = cpu_to_le32(inode->i_ino); + save_error_info(inode->i_sb, function, line); va_start(args, fmt); path = d_path(&(file->f_path), pathname, sizeof(pathname)); if (!path) path = "(unknown)"; printk(KERN_CRIT - "EXT4-fs error (device %s): %s: inode #%lu (comm %s path %s): ", - inode->i_sb->s_id, function, inode->i_ino, current->comm, path); + "EXT4-fs error (device %s): %s:%d: inode #%lu " + "(comm %s path %s): ", + inode->i_sb->s_id, function, line, inode->i_ino, + current->comm, path); vprintk(fmt, args); printk("\n"); va_end(args); @@ -435,7 +480,8 @@ static const char *ext4_decode_error(struct super_block *sb, int errno, /* __ext4_std_error decodes expected errors from journaling functions * automatically and invokes the appropriate error response. */ -void __ext4_std_error(struct super_block *sb, const char *function, int errno) +void __ext4_std_error(struct super_block *sb, const char *function, + unsigned int line, int errno) { char nbuf[16]; const char *errstr; @@ -448,8 +494,9 @@ void __ext4_std_error(struct super_block *sb, const char *function, int errno) return; errstr = ext4_decode_error(sb, errno, nbuf); - printk(KERN_CRIT "EXT4-fs error (device %s) in %s: %s\n", - sb->s_id, function, errstr); + printk(KERN_CRIT "EXT4-fs error (device %s) in %s:%d: %s\n", + sb->s_id, function, line, errstr); + save_error_info(sb, function, line); ext4_handle_error(sb); } @@ -464,29 +511,29 @@ void __ext4_std_error(struct super_block *sb, const char *function, int errno) * case we take the easy way out and panic immediately. */ -void ext4_abort(struct super_block *sb, const char *function, - const char *fmt, ...) +void __ext4_abort(struct super_block *sb, const char *function, + unsigned int line, const char *fmt, ...) { va_list args; + save_error_info(sb, function, line); va_start(args, fmt); - printk(KERN_CRIT "EXT4-fs error (device %s): %s: ", sb->s_id, function); + printk(KERN_CRIT "EXT4-fs error (device %s): %s:%d: ", sb->s_id, + function, line); vprintk(fmt, args); printk("\n"); va_end(args); + if ((sb->s_flags & MS_RDONLY) == 0) { + ext4_msg(sb, KERN_CRIT, "Remounting filesystem read-only"); + sb->s_flags |= MS_RDONLY; + EXT4_SB(sb)->s_mount_flags |= EXT4_MF_FS_ABORTED; + if (EXT4_SB(sb)->s_journal) + jbd2_journal_abort(EXT4_SB(sb)->s_journal, -EIO); + save_error_info(sb, function, line); + } if (test_opt(sb, ERRORS_PANIC)) panic("EXT4-fs panic from previous error\n"); - - if (sb->s_flags & MS_RDONLY) - return; - - ext4_msg(sb, KERN_CRIT, "Remounting filesystem read-only"); - EXT4_SB(sb)->s_mount_state |= EXT4_ERROR_FS; - sb->s_flags |= MS_RDONLY; - EXT4_SB(sb)->s_mount_flags |= EXT4_MF_FS_ABORTED; - if (EXT4_SB(sb)->s_journal) - jbd2_journal_abort(EXT4_SB(sb)->s_journal, -EIO); } void ext4_msg (struct super_block * sb, const char *prefix, @@ -502,38 +549,47 @@ void ext4_msg (struct super_block * sb, const char *prefix, } void __ext4_warning(struct super_block *sb, const char *function, - const char *fmt, ...) + unsigned int line, const char *fmt, ...) { va_list args; va_start(args, fmt); - printk(KERN_WARNING "EXT4-fs warning (device %s): %s: ", - sb->s_id, function); + printk(KERN_WARNING "EXT4-fs warning (device %s): %s:%d: ", + sb->s_id, function, line); vprintk(fmt, args); printk("\n"); va_end(args); } -void ext4_grp_locked_error(struct super_block *sb, ext4_group_t grp, - const char *function, const char *fmt, ...) +void __ext4_grp_locked_error(const char *function, unsigned int line, + struct super_block *sb, ext4_group_t grp, + unsigned long ino, ext4_fsblk_t block, + const char *fmt, ...) __releases(bitlock) __acquires(bitlock) { va_list args; struct ext4_super_block *es = EXT4_SB(sb)->s_es; + es->s_last_error_ino = cpu_to_le32(ino); + es->s_last_error_block = cpu_to_le64(block); + __save_error_info(sb, function, line); va_start(args, fmt); - printk(KERN_CRIT "EXT4-fs error (device %s): %s: ", sb->s_id, function); + printk(KERN_CRIT "EXT4-fs error (device %s): %s:%d: group %u", + sb->s_id, function, line, grp); + if (ino) + printk("inode %lu: ", ino); + if (block) + printk("block %llu:", (unsigned long long) block); vprintk(fmt, args); printk("\n"); va_end(args); if (test_opt(sb, ERRORS_CONT)) { - EXT4_SB(sb)->s_mount_state |= EXT4_ERROR_FS; - es->s_state |= cpu_to_le16(EXT4_ERROR_FS); ext4_commit_super(sb, 0); return; } + ext4_unlock_group(sb, grp); ext4_handle_error(sb); /* @@ -660,8 +716,7 @@ static void ext4_put_super(struct super_block *sb) err = jbd2_journal_destroy(sbi->s_journal); sbi->s_journal = NULL; if (err < 0) - ext4_abort(sb, __func__, - "Couldn't clean up the journal"); + ext4_abort(sb, "Couldn't clean up the journal"); } ext4_release_system_zone(sb); @@ -946,14 +1001,12 @@ static int ext4_show_options(struct seq_file *seq, struct vfsmount *vfs) seq_puts(seq, ",journal_async_commit"); else if (test_opt(sb, JOURNAL_CHECKSUM)) seq_puts(seq, ",journal_checksum"); - if (test_opt(sb, NOBH)) - seq_puts(seq, ",nobh"); if (test_opt(sb, I_VERSION)) seq_puts(seq, ",i_version"); - if (!test_opt(sb, DELALLOC)) + if (!test_opt(sb, DELALLOC) && + !(def_mount_opts & EXT4_DEFM_NODELALLOC)) seq_puts(seq, ",nodelalloc"); - if (sbi->s_stripe) seq_printf(seq, ",stripe=%lu", sbi->s_stripe); /* @@ -977,7 +1030,7 @@ static int ext4_show_options(struct seq_file *seq, struct vfsmount *vfs) if (test_opt(sb, NO_AUTO_DA_ALLOC)) seq_puts(seq, ",noauto_da_alloc"); - if (test_opt(sb, DISCARD)) + if (test_opt(sb, DISCARD) && !(def_mount_opts & EXT4_DEFM_DISCARD)) seq_puts(seq, ",discard"); if (test_opt(sb, NOLOAD)) @@ -986,6 +1039,10 @@ static int ext4_show_options(struct seq_file *seq, struct vfsmount *vfs) if (test_opt(sb, DIOREAD_NOLOCK)) seq_puts(seq, ",dioread_nolock"); + if (test_opt(sb, BLOCK_VALIDITY) && + !(def_mount_opts & EXT4_DEFM_BLOCK_VALIDITY)) + seq_puts(seq, ",block_validity"); + ext4_show_quota_options(seq, sb); return 0; @@ -1065,6 +1122,7 @@ static int ext4_mark_dquot_dirty(struct dquot *dquot); static int ext4_write_info(struct super_block *sb, int type); static int ext4_quota_on(struct super_block *sb, int type, int format_id, char *path); +static int ext4_quota_off(struct super_block *sb, int type); static int ext4_quota_on_mount(struct super_block *sb, int type); static ssize_t ext4_quota_read(struct super_block *sb, int type, char *data, size_t len, loff_t off); @@ -1086,7 +1144,7 @@ static const struct dquot_operations ext4_quota_operations = { static const struct quotactl_ops ext4_qctl_operations = { .quota_on = ext4_quota_on, - .quota_off = dquot_quota_off, + .quota_off = ext4_quota_off, .quota_sync = dquot_quota_sync, .get_info = dquot_get_dqinfo, .set_info = dquot_set_dqinfo, @@ -1624,10 +1682,12 @@ set_qf_format: *n_blocks_count = option; break; case Opt_nobh: - set_opt(sbi->s_mount_opt, NOBH); + ext4_msg(sb, KERN_WARNING, + "Ignoring deprecated nobh option"); break; case Opt_bh: - clear_opt(sbi->s_mount_opt, NOBH); + ext4_msg(sb, KERN_WARNING, + "Ignoring deprecated bh option"); break; case Opt_i_version: set_opt(sbi->s_mount_opt, I_VERSION); @@ -2249,6 +2309,8 @@ static ssize_t session_write_kbytes_show(struct ext4_attr *a, { struct super_block *sb = sbi->s_buddy_cache->i_sb; + if (!sb->s_bdev->bd_part) + return snprintf(buf, PAGE_SIZE, "0\n"); return snprintf(buf, PAGE_SIZE, "%lu\n", (part_stat_read(sb->s_bdev->bd_part, sectors[1]) - sbi->s_sectors_written_start) >> 1); @@ -2259,6 +2321,8 @@ static ssize_t lifetime_write_kbytes_show(struct ext4_attr *a, { struct super_block *sb = sbi->s_buddy_cache->i_sb; + if (!sb->s_bdev->bd_part) + return snprintf(buf, PAGE_SIZE, "0\n"); return snprintf(buf, PAGE_SIZE, "%llu\n", (unsigned long long)(sbi->s_kbytes_written + ((part_stat_read(sb->s_bdev->bd_part, sectors[1]) - @@ -2431,6 +2495,53 @@ static int ext4_feature_set_ok(struct super_block *sb, int readonly) return 1; } +/* + * This function is called once a day if we have errors logged + * on the file system + */ +static void print_daily_error_info(unsigned long arg) +{ + struct super_block *sb = (struct super_block *) arg; + struct ext4_sb_info *sbi; + struct ext4_super_block *es; + + sbi = EXT4_SB(sb); + es = sbi->s_es; + + if (es->s_error_count) + ext4_msg(sb, KERN_NOTICE, "error count: %u", + le32_to_cpu(es->s_error_count)); + if (es->s_first_error_time) { + printk(KERN_NOTICE "EXT4-fs (%s): initial error at %u: %.*s:%d", + sb->s_id, le32_to_cpu(es->s_first_error_time), + (int) sizeof(es->s_first_error_func), + es->s_first_error_func, + le32_to_cpu(es->s_first_error_line)); + if (es->s_first_error_ino) + printk(": inode %u", + le32_to_cpu(es->s_first_error_ino)); + if (es->s_first_error_block) + printk(": block %llu", (unsigned long long) + le64_to_cpu(es->s_first_error_block)); + printk("\n"); + } + if (es->s_last_error_time) { + printk(KERN_NOTICE "EXT4-fs (%s): last error at %u: %.*s:%d", + sb->s_id, le32_to_cpu(es->s_last_error_time), + (int) sizeof(es->s_last_error_func), + es->s_last_error_func, + le32_to_cpu(es->s_last_error_line)); + if (es->s_last_error_ino) + printk(": inode %u", + le32_to_cpu(es->s_last_error_ino)); + if (es->s_last_error_block) + printk(": block %llu", (unsigned long long) + le64_to_cpu(es->s_last_error_block)); + printk("\n"); + } + mod_timer(&sbi->s_err_report, jiffies + 24*60*60*HZ); /* Once a day */ +} + static int ext4_fill_super(struct super_block *sb, void *data, int silent) __releases(kernel_lock) __acquires(kernel_lock) @@ -2448,7 +2559,7 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) struct inode *root; char *cp; const char *descr; - int ret = -EINVAL; + int ret = -ENOMEM; int blocksize; unsigned int db_count; unsigned int i; @@ -2459,13 +2570,13 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) sbi = kzalloc(sizeof(*sbi), GFP_KERNEL); if (!sbi) - return -ENOMEM; + goto out_free_orig; sbi->s_blockgroup_lock = kzalloc(sizeof(struct blockgroup_lock), GFP_KERNEL); if (!sbi->s_blockgroup_lock) { kfree(sbi); - return -ENOMEM; + goto out_free_orig; } sb->s_fs_info = sbi; sbi->s_mount_opt = 0; @@ -2473,8 +2584,9 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) sbi->s_resgid = EXT4_DEF_RESGID; sbi->s_inode_readahead_blks = EXT4_DEF_INODE_READAHEAD_BLKS; sbi->s_sb_block = sb_block; - sbi->s_sectors_written_start = part_stat_read(sb->s_bdev->bd_part, - sectors[1]); + if (sb->s_bdev->bd_part) + sbi->s_sectors_written_start = + part_stat_read(sb->s_bdev->bd_part, sectors[1]); unlock_kernel(); @@ -2482,6 +2594,7 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) for (cp = sb->s_id; (cp = strchr(cp, '/'));) *cp = '!'; + ret = -EINVAL; blocksize = sb_min_blocksize(sb, EXT4_MIN_BLOCK_SIZE); if (!blocksize) { ext4_msg(sb, KERN_ERR, "unable to set blocksize"); @@ -2546,6 +2659,10 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) set_opt(sbi->s_mount_opt, ERRORS_CONT); else set_opt(sbi->s_mount_opt, ERRORS_RO); + if (def_mount_opts & EXT4_DEFM_BLOCK_VALIDITY) + set_opt(sbi->s_mount_opt, BLOCK_VALIDITY); + if (def_mount_opts & EXT4_DEFM_DISCARD) + set_opt(sbi->s_mount_opt, DISCARD); sbi->s_resuid = le16_to_cpu(es->s_def_resuid); sbi->s_resgid = le16_to_cpu(es->s_def_resgid); @@ -2553,15 +2670,23 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) sbi->s_min_batch_time = EXT4_DEF_MIN_BATCH_TIME; sbi->s_max_batch_time = EXT4_DEF_MAX_BATCH_TIME; - set_opt(sbi->s_mount_opt, BARRIER); + if ((def_mount_opts & EXT4_DEFM_NOBARRIER) == 0) + set_opt(sbi->s_mount_opt, BARRIER); /* * enable delayed allocation by default * Use -o nodelalloc to turn it off */ - if (!IS_EXT3_SB(sb)) + if (!IS_EXT3_SB(sb) && + ((def_mount_opts & EXT4_DEFM_NODELALLOC) == 0)) set_opt(sbi->s_mount_opt, DELALLOC); + if (!parse_options((char *) sbi->s_es->s_mount_opts, sb, + &journal_devnum, &journal_ioprio, NULL, 0)) { + ext4_msg(sb, KERN_WARNING, + "failed to parse options in superblock: %s", + sbi->s_es->s_mount_opts); + } if (!parse_options((char *) data, sb, &journal_devnum, &journal_ioprio, NULL, 0)) goto failed_mount; @@ -2912,18 +3037,7 @@ no_journal: ext4_msg(sb, KERN_ERR, "insufficient memory"); goto failed_mount_wq; } - if (test_opt(sb, NOBH)) { - if (!(test_opt(sb, DATA_FLAGS) == EXT4_MOUNT_WRITEBACK_DATA)) { - ext4_msg(sb, KERN_WARNING, "Ignoring nobh option - " - "its supported only with writeback mode"); - clear_opt(sbi->s_mount_opt, NOBH); - } - if (test_opt(sb, DIOREAD_NOLOCK)) { - ext4_msg(sb, KERN_WARNING, "dioread_nolock option is " - "not supported with nobh mode"); - goto failed_mount_wq; - } - } + EXT4_SB(sb)->dio_unwritten_wq = create_workqueue("ext4-dio-unwritten"); if (!EXT4_SB(sb)->dio_unwritten_wq) { printk(KERN_ERR "EXT4-fs: failed to create DIO workqueue\n"); @@ -3010,7 +3124,7 @@ no_journal: ext4_ext_init(sb); err = ext4_mb_init(sb, needs_recovery); if (err) { - ext4_msg(sb, KERN_ERR, "failed to initalize mballoc (%d)", + ext4_msg(sb, KERN_ERR, "failed to initialize mballoc (%d)", err); goto failed_mount4; } @@ -3043,7 +3157,14 @@ no_journal: descr = "out journal"; ext4_msg(sb, KERN_INFO, "mounted filesystem with%s. " - "Opts: %s", descr, orig_data); + "Opts: %s%s%s", descr, sbi->s_es->s_mount_opts, + *sbi->s_es->s_mount_opts ? "; " : "", orig_data); + + init_timer(&sbi->s_err_report); + sbi->s_err_report.function = print_daily_error_info; + sbi->s_err_report.data = (unsigned long) sb; + if (es->s_error_count) + mod_timer(&sbi->s_err_report, jiffies + 300*HZ); /* 5 minutes */ lock_kernel(); kfree(orig_data); @@ -3093,6 +3214,7 @@ out_fail: kfree(sbi->s_blockgroup_lock); kfree(sbi); lock_kernel(); +out_free_orig: kfree(orig_data); return ret; } @@ -3110,7 +3232,7 @@ static void ext4_init_journal_params(struct super_block *sb, journal_t *journal) journal->j_min_batch_time = sbi->s_min_batch_time; journal->j_max_batch_time = sbi->s_max_batch_time; - spin_lock(&journal->j_state_lock); + write_lock(&journal->j_state_lock); if (test_opt(sb, BARRIER)) journal->j_flags |= JBD2_BARRIER; else @@ -3119,7 +3241,7 @@ static void ext4_init_journal_params(struct super_block *sb, journal_t *journal) journal->j_flags |= JBD2_ABORT_ON_SYNCDATA_ERR; else journal->j_flags &= ~JBD2_ABORT_ON_SYNCDATA_ERR; - spin_unlock(&journal->j_state_lock); + write_unlock(&journal->j_state_lock); } static journal_t *ext4_get_journal(struct super_block *sb, @@ -3327,8 +3449,17 @@ static int ext4_load_journal(struct super_block *sb, if (!EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_RECOVER)) err = jbd2_journal_wipe(journal, !really_read_only); - if (!err) + if (!err) { + char *save = kmalloc(EXT4_S_ERR_LEN, GFP_KERNEL); + if (save) + memcpy(save, ((char *) es) + + EXT4_S_ERR_START, EXT4_S_ERR_LEN); err = jbd2_journal_load(journal); + if (save) + memcpy(((char *) es) + EXT4_S_ERR_START, + save, EXT4_S_ERR_LEN); + kfree(save); + } if (err) { ext4_msg(sb, KERN_ERR, "error loading journal"); @@ -3384,10 +3515,14 @@ static int ext4_commit_super(struct super_block *sb, int sync) */ if (!(sb->s_flags & MS_RDONLY)) es->s_wtime = cpu_to_le32(get_seconds()); - es->s_kbytes_written = - cpu_to_le64(EXT4_SB(sb)->s_kbytes_written + + if (sb->s_bdev->bd_part) + es->s_kbytes_written = + cpu_to_le64(EXT4_SB(sb)->s_kbytes_written + ((part_stat_read(sb->s_bdev->bd_part, sectors[1]) - EXT4_SB(sb)->s_sectors_written_start) >> 1)); + else + es->s_kbytes_written = + cpu_to_le64(EXT4_SB(sb)->s_kbytes_written); ext4_free_blocks_count_set(es, percpu_counter_sum_positive( &EXT4_SB(sb)->s_freeblocks_counter)); es->s_free_inodes_count = cpu_to_le32(percpu_counter_sum_positive( @@ -3491,7 +3626,7 @@ int ext4_force_commit(struct super_block *sb) journal = EXT4_SB(sb)->s_journal; if (journal) { - vfs_check_frozen(sb, SB_FREEZE_WRITE); + vfs_check_frozen(sb, SB_FREEZE_TRANS); ret = ext4_journal_force_commit(journal); } @@ -3616,7 +3751,7 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) } if (sbi->s_mount_flags & EXT4_MF_FS_ABORTED) - ext4_abort(sb, __func__, "Abort forced by user"); + ext4_abort(sb, "Abort forced by user"); sb->s_flags = (sb->s_flags & ~MS_POSIXACL) | (test_opt(sb, POSIX_ACL) ? MS_POSIXACL : 0); @@ -3981,6 +4116,18 @@ static int ext4_quota_on(struct super_block *sb, int type, int format_id, return err; } +static int ext4_quota_off(struct super_block *sb, int type) +{ + /* Force all delayed allocation blocks to be allocated */ + if (test_opt(sb, DELALLOC)) { + down_read(&sb->s_umount); + sync_filesystem(sb); + up_read(&sb->s_umount); + } + + return dquot_quota_off(sb, type); +} + /* Read data from quotafile - avoid pagecache and such because we cannot afford * acquiring the locks... As quota files are never truncated and quota code * itself serializes the operations (and noone else should touch the files) @@ -4030,7 +4177,6 @@ static ssize_t ext4_quota_write(struct super_block *sb, int type, ext4_lblk_t blk = off >> EXT4_BLOCK_SIZE_BITS(sb); int err = 0; int offset = off & (sb->s_blocksize - 1); - int journal_quota = EXT4_SB(sb)->s_qf_names[type] != NULL; struct buffer_head *bh; handle_t *handle = journal_current_handle(); @@ -4055,24 +4201,16 @@ static ssize_t ext4_quota_write(struct super_block *sb, int type, bh = ext4_bread(handle, inode, blk, 1, &err); if (!bh) goto out; - if (journal_quota) { - err = ext4_journal_get_write_access(handle, bh); - if (err) { - brelse(bh); - goto out; - } + err = ext4_journal_get_write_access(handle, bh); + if (err) { + brelse(bh); + goto out; } lock_buffer(bh); memcpy(bh->b_data+offset, data, len); flush_dcache_page(bh->b_page); unlock_buffer(bh); - if (journal_quota) - err = ext4_handle_dirty_metadata(handle, NULL, bh); - else { - /* Always do at least ordered writes for quotas */ - err = ext4_jbd2_file_inode(handle, inode); - mark_buffer_dirty(bh); - } + err = ext4_handle_dirty_metadata(handle, NULL, bh); brelse(bh); out: if (err) { diff --git a/fs/ext4/xattr.c b/fs/ext4/xattr.c index 04338009793..a6f31424957 100644 --- a/fs/ext4/xattr.c +++ b/fs/ext4/xattr.c @@ -458,8 +458,7 @@ static void ext4_xattr_update_super_block(handle_t *handle, if (ext4_journal_get_write_access(handle, EXT4_SB(sb)->s_sbh) == 0) { EXT4_SET_COMPAT_FEATURE(sb, EXT4_FEATURE_COMPAT_EXT_ATTR); - sb->s_dirt = 1; - ext4_handle_dirty_metadata(handle, NULL, EXT4_SB(sb)->s_sbh); + ext4_handle_dirty_super(handle, sb); } } diff --git a/fs/file.c b/fs/file.c index 34bb7f71d99..cccaead962c 100644 --- a/fs/file.c +++ b/fs/file.c @@ -178,7 +178,6 @@ static struct fdtable * alloc_fdtable(unsigned int nr) fdt->open_fds = (fd_set *)data; data += nr / BITS_PER_BYTE; fdt->close_on_exec = (fd_set *)data; - INIT_RCU_HEAD(&fdt->rcu); fdt->next = NULL; return fdt; @@ -312,7 +311,6 @@ struct files_struct *dup_fd(struct files_struct *oldf, int *errorp) new_fdt->close_on_exec = (fd_set *)&newf->close_on_exec_init; new_fdt->open_fds = (fd_set *)&newf->open_fds_init; new_fdt->fd = &newf->fd_array[0]; - INIT_RCU_HEAD(&new_fdt->rcu); new_fdt->next = NULL; spin_lock(&oldf->file_lock); @@ -430,7 +428,6 @@ struct files_struct init_files = { .fd = &init_files.fd_array[0], .close_on_exec = (fd_set *)&init_files.close_on_exec_init, .open_fds = (fd_set *)&init_files.open_fds_init, - .rcu = RCU_HEAD_INIT, }, .file_lock = __SPIN_LOCK_UNLOCKED(init_task.file_lock), }; diff --git a/fs/freevxfs/vxfs_super.c b/fs/freevxfs/vxfs_super.c index 1e8af939b3e..5132c99b1ca 100644 --- a/fs/freevxfs/vxfs_super.c +++ b/fs/freevxfs/vxfs_super.c @@ -135,7 +135,7 @@ static int vxfs_remount(struct super_block *sb, int *flags, char *data) } /** - * vxfs_read_super - read superblock into memory and initalize filesystem + * vxfs_read_super - read superblock into memory and initialize filesystem * @sbp: VFS superblock (to fill) * @dp: fs private mount data * @silent: do not complain loudly when sth is wrong diff --git a/fs/fscache/Kconfig b/fs/fscache/Kconfig index cc94bb9563f..3f6dfa98988 100644 --- a/fs/fscache/Kconfig +++ b/fs/fscache/Kconfig @@ -1,7 +1,6 @@ config FSCACHE tristate "General filesystem local caching manager" - select SLOW_WORK help This option enables a generic filesystem caching manager that can be used by various network and other filesystems to cache data locally. diff --git a/fs/fscache/internal.h b/fs/fscache/internal.h index edd7434ab6e..6a026441c5a 100644 --- a/fs/fscache/internal.h +++ b/fs/fscache/internal.h @@ -82,6 +82,14 @@ extern unsigned fscache_defer_lookup; extern unsigned fscache_defer_create; extern unsigned fscache_debug; extern struct kobject *fscache_root; +extern struct workqueue_struct *fscache_object_wq; +extern struct workqueue_struct *fscache_op_wq; +DECLARE_PER_CPU(wait_queue_head_t, fscache_object_cong_wait); + +static inline bool fscache_object_congested(void) +{ + return workqueue_congested(WORK_CPU_UNBOUND, fscache_object_wq); +} extern int fscache_wait_bit(void *); extern int fscache_wait_bit_interruptible(void *); diff --git a/fs/fscache/main.c b/fs/fscache/main.c index add6bdb53f0..f9d856773f7 100644 --- a/fs/fscache/main.c +++ b/fs/fscache/main.c @@ -15,6 +15,7 @@ #include <linux/sched.h> #include <linux/completion.h> #include <linux/slab.h> +#include <linux/seq_file.h> #include "internal.h" MODULE_DESCRIPTION("FS Cache Manager"); @@ -40,22 +41,105 @@ MODULE_PARM_DESC(fscache_debug, "FS-Cache debugging mask"); struct kobject *fscache_root; +struct workqueue_struct *fscache_object_wq; +struct workqueue_struct *fscache_op_wq; + +DEFINE_PER_CPU(wait_queue_head_t, fscache_object_cong_wait); + +/* these values serve as lower bounds, will be adjusted in fscache_init() */ +static unsigned fscache_object_max_active = 4; +static unsigned fscache_op_max_active = 2; + +#ifdef CONFIG_SYSCTL +static struct ctl_table_header *fscache_sysctl_header; + +static int fscache_max_active_sysctl(struct ctl_table *table, int write, + void __user *buffer, + size_t *lenp, loff_t *ppos) +{ + struct workqueue_struct **wqp = table->extra1; + unsigned int *datap = table->data; + int ret; + + ret = proc_dointvec(table, write, buffer, lenp, ppos); + if (ret == 0) + workqueue_set_max_active(*wqp, *datap); + return ret; +} + +ctl_table fscache_sysctls[] = { + { + .procname = "object_max_active", + .data = &fscache_object_max_active, + .maxlen = sizeof(unsigned), + .mode = 0644, + .proc_handler = fscache_max_active_sysctl, + .extra1 = &fscache_object_wq, + }, + { + .procname = "operation_max_active", + .data = &fscache_op_max_active, + .maxlen = sizeof(unsigned), + .mode = 0644, + .proc_handler = fscache_max_active_sysctl, + .extra1 = &fscache_op_wq, + }, + {} +}; + +ctl_table fscache_sysctls_root[] = { + { + .procname = "fscache", + .mode = 0555, + .child = fscache_sysctls, + }, + {} +}; +#endif /* * initialise the fs caching module */ static int __init fscache_init(void) { + unsigned int nr_cpus = num_possible_cpus(); + unsigned int cpu; int ret; - ret = slow_work_register_user(THIS_MODULE); - if (ret < 0) - goto error_slow_work; + fscache_object_max_active = + clamp_val(nr_cpus, + fscache_object_max_active, WQ_UNBOUND_MAX_ACTIVE); + + ret = -ENOMEM; + fscache_object_wq = alloc_workqueue("fscache_object", WQ_UNBOUND, + fscache_object_max_active); + if (!fscache_object_wq) + goto error_object_wq; + + fscache_op_max_active = + clamp_val(fscache_object_max_active / 2, + fscache_op_max_active, WQ_UNBOUND_MAX_ACTIVE); + + ret = -ENOMEM; + fscache_op_wq = alloc_workqueue("fscache_operation", WQ_UNBOUND, + fscache_op_max_active); + if (!fscache_op_wq) + goto error_op_wq; + + for_each_possible_cpu(cpu) + init_waitqueue_head(&per_cpu(fscache_object_cong_wait, cpu)); ret = fscache_proc_init(); if (ret < 0) goto error_proc; +#ifdef CONFIG_SYSCTL + ret = -ENOMEM; + fscache_sysctl_header = register_sysctl_table(fscache_sysctls_root); + if (!fscache_sysctl_header) + goto error_sysctl; +#endif + fscache_cookie_jar = kmem_cache_create("fscache_cookie_jar", sizeof(struct fscache_cookie), 0, @@ -78,10 +162,16 @@ static int __init fscache_init(void) error_kobj: kmem_cache_destroy(fscache_cookie_jar); error_cookie_jar: +#ifdef CONFIG_SYSCTL + unregister_sysctl_table(fscache_sysctl_header); +error_sysctl: +#endif fscache_proc_cleanup(); error_proc: - slow_work_unregister_user(THIS_MODULE); -error_slow_work: + destroy_workqueue(fscache_op_wq); +error_op_wq: + destroy_workqueue(fscache_object_wq); +error_object_wq: return ret; } @@ -96,8 +186,12 @@ static void __exit fscache_exit(void) kobject_put(fscache_root); kmem_cache_destroy(fscache_cookie_jar); +#ifdef CONFIG_SYSCTL + unregister_sysctl_table(fscache_sysctl_header); +#endif fscache_proc_cleanup(); - slow_work_unregister_user(THIS_MODULE); + destroy_workqueue(fscache_op_wq); + destroy_workqueue(fscache_object_wq); printk(KERN_NOTICE "FS-Cache: Unloaded\n"); } diff --git a/fs/fscache/object-list.c b/fs/fscache/object-list.c index 4a8eb31c533..ebe29c58138 100644 --- a/fs/fscache/object-list.c +++ b/fs/fscache/object-list.c @@ -34,8 +34,8 @@ struct fscache_objlist_data { #define FSCACHE_OBJLIST_CONFIG_NOREADS 0x00000200 /* show objects without active reads */ #define FSCACHE_OBJLIST_CONFIG_EVENTS 0x00000400 /* show objects with events */ #define FSCACHE_OBJLIST_CONFIG_NOEVENTS 0x00000800 /* show objects without no events */ -#define FSCACHE_OBJLIST_CONFIG_WORK 0x00001000 /* show objects with slow work */ -#define FSCACHE_OBJLIST_CONFIG_NOWORK 0x00002000 /* show objects without slow work */ +#define FSCACHE_OBJLIST_CONFIG_WORK 0x00001000 /* show objects with work */ +#define FSCACHE_OBJLIST_CONFIG_NOWORK 0x00002000 /* show objects without work */ u8 buf[512]; /* key and aux data buffer */ }; @@ -231,12 +231,11 @@ static int fscache_objlist_show(struct seq_file *m, void *v) READS, NOREADS); FILTER(obj->events & obj->event_mask, EVENTS, NOEVENTS); - FILTER(obj->work.flags & ~(1UL << SLOW_WORK_VERY_SLOW), - WORK, NOWORK); + FILTER(work_busy(&obj->work), WORK, NOWORK); } seq_printf(m, - "%8x %8x %s %5u %3u %3u %3u %2u %5u %2lx %2lx %1lx %1lx | ", + "%8x %8x %s %5u %3u %3u %3u %2u %5u %2lx %2lx %1lx %1x | ", obj->debug_id, obj->parent ? obj->parent->debug_id : -1, fscache_object_states_short[obj->state], @@ -249,7 +248,7 @@ static int fscache_objlist_show(struct seq_file *m, void *v) obj->event_mask & FSCACHE_OBJECT_EVENTS_MASK, obj->events, obj->flags, - obj->work.flags); + work_busy(&obj->work)); no_cookie = true; keylen = auxlen = 0; diff --git a/fs/fscache/object.c b/fs/fscache/object.c index 0b589a9b4ff..b6b897c550a 100644 --- a/fs/fscache/object.c +++ b/fs/fscache/object.c @@ -14,7 +14,6 @@ #define FSCACHE_DEBUG_LEVEL COOKIE #include <linux/module.h> -#include <linux/seq_file.h> #include "internal.h" const char *fscache_object_states[FSCACHE_OBJECT__NSTATES] = { @@ -50,12 +49,8 @@ const char fscache_object_states_short[FSCACHE_OBJECT__NSTATES][5] = { [FSCACHE_OBJECT_DEAD] = "DEAD", }; -static void fscache_object_slow_work_put_ref(struct slow_work *); -static int fscache_object_slow_work_get_ref(struct slow_work *); -static void fscache_object_slow_work_execute(struct slow_work *); -#ifdef CONFIG_SLOW_WORK_DEBUG -static void fscache_object_slow_work_desc(struct slow_work *, struct seq_file *); -#endif +static int fscache_get_object(struct fscache_object *); +static void fscache_put_object(struct fscache_object *); static void fscache_initialise_object(struct fscache_object *); static void fscache_lookup_object(struct fscache_object *); static void fscache_object_available(struct fscache_object *); @@ -64,17 +59,6 @@ static void fscache_withdraw_object(struct fscache_object *); static void fscache_enqueue_dependents(struct fscache_object *); static void fscache_dequeue_object(struct fscache_object *); -const struct slow_work_ops fscache_object_slow_work_ops = { - .owner = THIS_MODULE, - .get_ref = fscache_object_slow_work_get_ref, - .put_ref = fscache_object_slow_work_put_ref, - .execute = fscache_object_slow_work_execute, -#ifdef CONFIG_SLOW_WORK_DEBUG - .desc = fscache_object_slow_work_desc, -#endif -}; -EXPORT_SYMBOL(fscache_object_slow_work_ops); - /* * we need to notify the parent when an op completes that we had outstanding * upon it @@ -345,7 +329,7 @@ unsupported_event: /* * execute an object */ -static void fscache_object_slow_work_execute(struct slow_work *work) +void fscache_object_work_func(struct work_struct *work) { struct fscache_object *object = container_of(work, struct fscache_object, work); @@ -359,23 +343,9 @@ static void fscache_object_slow_work_execute(struct slow_work *work) if (object->events & object->event_mask) fscache_enqueue_object(object); clear_bit(FSCACHE_OBJECT_EV_REQUEUE, &object->events); + fscache_put_object(object); } - -/* - * describe an object for slow-work debugging - */ -#ifdef CONFIG_SLOW_WORK_DEBUG -static void fscache_object_slow_work_desc(struct slow_work *work, - struct seq_file *m) -{ - struct fscache_object *object = - container_of(work, struct fscache_object, work); - - seq_printf(m, "FSC: OBJ%x: %s", - object->debug_id, - fscache_object_states_short[object->state]); -} -#endif +EXPORT_SYMBOL(fscache_object_work_func); /* * initialise an object @@ -393,7 +363,6 @@ static void fscache_initialise_object(struct fscache_object *object) _enter(""); ASSERT(object->cookie != NULL); ASSERT(object->cookie->parent != NULL); - ASSERT(list_empty(&object->work.link)); if (object->events & ((1 << FSCACHE_OBJECT_EV_ERROR) | (1 << FSCACHE_OBJECT_EV_RELEASE) | @@ -671,10 +640,8 @@ static void fscache_drop_object(struct fscache_object *object) object->parent = NULL; } - /* this just shifts the object release to the slow work processor */ - fscache_stat(&fscache_n_cop_put_object); - object->cache->ops->put_object(object); - fscache_stat_d(&fscache_n_cop_put_object); + /* this just shifts the object release to the work processor */ + fscache_put_object(object); _leave(""); } @@ -758,12 +725,10 @@ void fscache_withdrawing_object(struct fscache_cache *cache, } /* - * allow the slow work item processor to get a ref on an object + * get a ref on an object */ -static int fscache_object_slow_work_get_ref(struct slow_work *work) +static int fscache_get_object(struct fscache_object *object) { - struct fscache_object *object = - container_of(work, struct fscache_object, work); int ret; fscache_stat(&fscache_n_cop_grab_object); @@ -773,13 +738,10 @@ static int fscache_object_slow_work_get_ref(struct slow_work *work) } /* - * allow the slow work item processor to discard a ref on a work item + * discard a ref on a work item */ -static void fscache_object_slow_work_put_ref(struct slow_work *work) +static void fscache_put_object(struct fscache_object *object) { - struct fscache_object *object = - container_of(work, struct fscache_object, work); - fscache_stat(&fscache_n_cop_put_object); object->cache->ops->put_object(object); fscache_stat_d(&fscache_n_cop_put_object); @@ -792,8 +754,48 @@ void fscache_enqueue_object(struct fscache_object *object) { _enter("{OBJ%x}", object->debug_id); - slow_work_enqueue(&object->work); + if (fscache_get_object(object) >= 0) { + wait_queue_head_t *cong_wq = + &get_cpu_var(fscache_object_cong_wait); + + if (queue_work(fscache_object_wq, &object->work)) { + if (fscache_object_congested()) + wake_up(cong_wq); + } else + fscache_put_object(object); + + put_cpu_var(fscache_object_cong_wait); + } +} + +/** + * fscache_object_sleep_till_congested - Sleep until object wq is congested + * @timoutp: Scheduler sleep timeout + * + * Allow an object handler to sleep until the object workqueue is congested. + * + * The caller must set up a wake up event before calling this and must have set + * the appropriate sleep mode (such as TASK_UNINTERRUPTIBLE) and tested its own + * condition before calling this function as no test is made here. + * + * %true is returned if the object wq is congested, %false otherwise. + */ +bool fscache_object_sleep_till_congested(signed long *timeoutp) +{ + wait_queue_head_t *cong_wq = &__get_cpu_var(fscache_object_cong_wait); + DEFINE_WAIT(wait); + + if (fscache_object_congested()) + return true; + + add_wait_queue_exclusive(cong_wq, &wait); + if (!fscache_object_congested()) + *timeoutp = schedule_timeout(*timeoutp); + finish_wait(cong_wq, &wait); + + return fscache_object_congested(); } +EXPORT_SYMBOL_GPL(fscache_object_sleep_till_congested); /* * enqueue the dependents of an object for metadata-type processing @@ -819,9 +821,7 @@ static void fscache_enqueue_dependents(struct fscache_object *object) /* sort onto appropriate lists */ fscache_enqueue_object(dep); - fscache_stat(&fscache_n_cop_put_object); - dep->cache->ops->put_object(dep); - fscache_stat_d(&fscache_n_cop_put_object); + fscache_put_object(dep); if (!list_empty(&object->dependents)) cond_resched_lock(&object->lock); diff --git a/fs/fscache/operation.c b/fs/fscache/operation.c index f17cecafae4..b9f34eaede0 100644 --- a/fs/fscache/operation.c +++ b/fs/fscache/operation.c @@ -42,16 +42,12 @@ void fscache_enqueue_operation(struct fscache_operation *op) fscache_stat(&fscache_n_op_enqueue); switch (op->flags & FSCACHE_OP_TYPE) { - case FSCACHE_OP_FAST: - _debug("queue fast"); + case FSCACHE_OP_ASYNC: + _debug("queue async"); atomic_inc(&op->usage); - if (!schedule_work(&op->fast_work)) + if (!queue_work(fscache_op_wq, &op->work)) fscache_put_operation(op); break; - case FSCACHE_OP_SLOW: - _debug("queue slow"); - slow_work_enqueue(&op->slow_work); - break; case FSCACHE_OP_MYTHREAD: _debug("queue for caller's attention"); break; @@ -455,36 +451,13 @@ void fscache_operation_gc(struct work_struct *work) } /* - * allow the slow work item processor to get a ref on an operation - */ -static int fscache_op_get_ref(struct slow_work *work) -{ - struct fscache_operation *op = - container_of(work, struct fscache_operation, slow_work); - - atomic_inc(&op->usage); - return 0; -} - -/* - * allow the slow work item processor to discard a ref on an operation - */ -static void fscache_op_put_ref(struct slow_work *work) -{ - struct fscache_operation *op = - container_of(work, struct fscache_operation, slow_work); - - fscache_put_operation(op); -} - -/* - * execute an operation using the slow thread pool to provide processing context - * - the caller holds a ref to this object, so we don't need to hold one + * execute an operation using fs_op_wq to provide processing context - + * the caller holds a ref to this object, so we don't need to hold one */ -static void fscache_op_execute(struct slow_work *work) +void fscache_op_work_func(struct work_struct *work) { struct fscache_operation *op = - container_of(work, struct fscache_operation, slow_work); + container_of(work, struct fscache_operation, work); unsigned long start; _enter("{OBJ%x OP%x,%d}", @@ -494,31 +467,7 @@ static void fscache_op_execute(struct slow_work *work) start = jiffies; op->processor(op); fscache_hist(fscache_ops_histogram, start); + fscache_put_operation(op); _leave(""); } - -/* - * describe an operation for slow-work debugging - */ -#ifdef CONFIG_SLOW_WORK_DEBUG -static void fscache_op_desc(struct slow_work *work, struct seq_file *m) -{ - struct fscache_operation *op = - container_of(work, struct fscache_operation, slow_work); - - seq_printf(m, "FSC: OBJ%x OP%x: %s/%s fl=%lx", - op->object->debug_id, op->debug_id, - op->name, op->state, op->flags); -} -#endif - -const struct slow_work_ops fscache_op_slow_work_ops = { - .owner = THIS_MODULE, - .get_ref = fscache_op_get_ref, - .put_ref = fscache_op_put_ref, - .execute = fscache_op_execute, -#ifdef CONFIG_SLOW_WORK_DEBUG - .desc = fscache_op_desc, -#endif -}; diff --git a/fs/fscache/page.c b/fs/fscache/page.c index 723b889fd21..41c441c2058 100644 --- a/fs/fscache/page.c +++ b/fs/fscache/page.c @@ -105,7 +105,7 @@ bool __fscache_maybe_release_page(struct fscache_cookie *cookie, page_busy: /* we might want to wait here, but that could deadlock the allocator as - * the slow-work threads writing to the cache may all end up sleeping + * the work threads writing to the cache may all end up sleeping * on memory allocation */ fscache_stat(&fscache_n_store_vmscan_busy); return false; @@ -188,9 +188,8 @@ int __fscache_attr_changed(struct fscache_cookie *cookie) return -ENOMEM; } - fscache_operation_init(op, NULL); - fscache_operation_init_slow(op, fscache_attr_changed_op); - op->flags = FSCACHE_OP_SLOW | (1 << FSCACHE_OP_EXCLUSIVE); + fscache_operation_init(op, fscache_attr_changed_op, NULL); + op->flags = FSCACHE_OP_ASYNC | (1 << FSCACHE_OP_EXCLUSIVE); fscache_set_op_name(op, "Attr"); spin_lock(&cookie->lock); @@ -218,24 +217,6 @@ nobufs: EXPORT_SYMBOL(__fscache_attr_changed); /* - * handle secondary execution given to a retrieval op on behalf of the - * cache - */ -static void fscache_retrieval_work(struct work_struct *work) -{ - struct fscache_retrieval *op = - container_of(work, struct fscache_retrieval, op.fast_work); - unsigned long start; - - _enter("{OP%x}", op->op.debug_id); - - start = jiffies; - op->op.processor(&op->op); - fscache_hist(fscache_ops_histogram, start); - fscache_put_operation(&op->op); -} - -/* * release a retrieval op reference */ static void fscache_release_retrieval_op(struct fscache_operation *_op) @@ -269,13 +250,12 @@ static struct fscache_retrieval *fscache_alloc_retrieval( return NULL; } - fscache_operation_init(&op->op, fscache_release_retrieval_op); + fscache_operation_init(&op->op, NULL, fscache_release_retrieval_op); op->op.flags = FSCACHE_OP_MYTHREAD | (1 << FSCACHE_OP_WAITING); op->mapping = mapping; op->end_io_func = end_io_func; op->context = context; op->start_time = jiffies; - INIT_WORK(&op->op.fast_work, fscache_retrieval_work); INIT_LIST_HEAD(&op->to_do); fscache_set_op_name(&op->op, "Retr"); return op; @@ -795,9 +775,9 @@ int __fscache_write_page(struct fscache_cookie *cookie, if (!op) goto nomem; - fscache_operation_init(&op->op, fscache_release_write_op); - fscache_operation_init_slow(&op->op, fscache_write_op); - op->op.flags = FSCACHE_OP_SLOW | (1 << FSCACHE_OP_WAITING); + fscache_operation_init(&op->op, fscache_write_op, + fscache_release_write_op); + op->op.flags = FSCACHE_OP_ASYNC | (1 << FSCACHE_OP_WAITING); fscache_set_op_name(&op->op, "Write1"); ret = radix_tree_preload(gfp & ~__GFP_HIGHMEM); @@ -852,7 +832,7 @@ int __fscache_write_page(struct fscache_cookie *cookie, fscache_stat(&fscache_n_store_ops); fscache_stat(&fscache_n_stores_ok); - /* the slow work queue now carries its own ref on the object */ + /* the work queue now carries its own ref on the object */ fscache_put_operation(&op->op); _leave(" = 0"); return 0; diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index 9424796d663..69ad053ffd7 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -239,7 +239,6 @@ static u64 fuse_get_unique(struct fuse_conn *fc) static void queue_request(struct fuse_conn *fc, struct fuse_req *req) { - req->in.h.unique = fuse_get_unique(fc); req->in.h.len = sizeof(struct fuse_in_header) + len_args(req->in.numargs, (struct fuse_arg *) req->in.args); list_add_tail(&req->list, &fc->pending); @@ -261,6 +260,7 @@ static void flush_bg_queue(struct fuse_conn *fc) req = list_entry(fc->bg_queue.next, struct fuse_req, list); list_del(&req->list); fc->active_background++; + req->in.h.unique = fuse_get_unique(fc); queue_request(fc, req); } } @@ -398,6 +398,7 @@ void fuse_request_send(struct fuse_conn *fc, struct fuse_req *req) else if (fc->conn_error) req->out.h.error = -ECONNREFUSED; else { + req->in.h.unique = fuse_get_unique(fc); queue_request(fc, req); /* acquire extra reference, since request is still needed after request_end() */ @@ -450,6 +451,23 @@ void fuse_request_send_background(struct fuse_conn *fc, struct fuse_req *req) } EXPORT_SYMBOL_GPL(fuse_request_send_background); +static int fuse_request_send_notify_reply(struct fuse_conn *fc, + struct fuse_req *req, u64 unique) +{ + int err = -ENODEV; + + req->isreply = 0; + req->in.h.unique = unique; + spin_lock(&fc->lock); + if (fc->connected) { + queue_request(fc, req); + err = 0; + } + spin_unlock(&fc->lock); + + return err; +} + /* * Called under fc->lock * @@ -535,13 +553,13 @@ static void fuse_copy_finish(struct fuse_copy_state *cs) if (!cs->write) { buf->ops->unmap(cs->pipe, buf, cs->mapaddr); } else { - kunmap_atomic(cs->mapaddr, KM_USER0); + kunmap(buf->page); buf->len = PAGE_SIZE - cs->len; } cs->currbuf = NULL; cs->mapaddr = NULL; } else if (cs->mapaddr) { - kunmap_atomic(cs->mapaddr, KM_USER0); + kunmap(cs->pg); if (cs->write) { flush_dcache_page(cs->pg); set_page_dirty_lock(cs->pg); @@ -572,7 +590,7 @@ static int fuse_copy_fill(struct fuse_copy_state *cs) BUG_ON(!cs->nr_segs); cs->currbuf = buf; - cs->mapaddr = buf->ops->map(cs->pipe, buf, 1); + cs->mapaddr = buf->ops->map(cs->pipe, buf, 0); cs->len = buf->len; cs->buf = cs->mapaddr + buf->offset; cs->pipebufs++; @@ -592,7 +610,7 @@ static int fuse_copy_fill(struct fuse_copy_state *cs) buf->len = 0; cs->currbuf = buf; - cs->mapaddr = kmap_atomic(page, KM_USER0); + cs->mapaddr = kmap(page); cs->buf = cs->mapaddr; cs->len = PAGE_SIZE; cs->pipebufs++; @@ -611,7 +629,7 @@ static int fuse_copy_fill(struct fuse_copy_state *cs) return err; BUG_ON(err != 1); offset = cs->addr % PAGE_SIZE; - cs->mapaddr = kmap_atomic(cs->pg, KM_USER0); + cs->mapaddr = kmap(cs->pg); cs->buf = cs->mapaddr + offset; cs->len = min(PAGE_SIZE - offset, cs->seglen); cs->seglen -= cs->len; @@ -1231,6 +1249,199 @@ err: return err; } +static int fuse_notify_store(struct fuse_conn *fc, unsigned int size, + struct fuse_copy_state *cs) +{ + struct fuse_notify_store_out outarg; + struct inode *inode; + struct address_space *mapping; + u64 nodeid; + int err; + pgoff_t index; + unsigned int offset; + unsigned int num; + loff_t file_size; + loff_t end; + + err = -EINVAL; + if (size < sizeof(outarg)) + goto out_finish; + + err = fuse_copy_one(cs, &outarg, sizeof(outarg)); + if (err) + goto out_finish; + + err = -EINVAL; + if (size - sizeof(outarg) != outarg.size) + goto out_finish; + + nodeid = outarg.nodeid; + + down_read(&fc->killsb); + + err = -ENOENT; + if (!fc->sb) + goto out_up_killsb; + + inode = ilookup5(fc->sb, nodeid, fuse_inode_eq, &nodeid); + if (!inode) + goto out_up_killsb; + + mapping = inode->i_mapping; + index = outarg.offset >> PAGE_CACHE_SHIFT; + offset = outarg.offset & ~PAGE_CACHE_MASK; + file_size = i_size_read(inode); + end = outarg.offset + outarg.size; + if (end > file_size) { + file_size = end; + fuse_write_update_size(inode, file_size); + } + + num = outarg.size; + while (num) { + struct page *page; + unsigned int this_num; + + err = -ENOMEM; + page = find_or_create_page(mapping, index, + mapping_gfp_mask(mapping)); + if (!page) + goto out_iput; + + this_num = min_t(unsigned, num, PAGE_CACHE_SIZE - offset); + err = fuse_copy_page(cs, &page, offset, this_num, 0); + if (!err && offset == 0 && (num != 0 || file_size == end)) + SetPageUptodate(page); + unlock_page(page); + page_cache_release(page); + + if (err) + goto out_iput; + + num -= this_num; + offset = 0; + index++; + } + + err = 0; + +out_iput: + iput(inode); +out_up_killsb: + up_read(&fc->killsb); +out_finish: + fuse_copy_finish(cs); + return err; +} + +static void fuse_retrieve_end(struct fuse_conn *fc, struct fuse_req *req) +{ + int i; + + for (i = 0; i < req->num_pages; i++) { + struct page *page = req->pages[i]; + page_cache_release(page); + } +} + +static int fuse_retrieve(struct fuse_conn *fc, struct inode *inode, + struct fuse_notify_retrieve_out *outarg) +{ + int err; + struct address_space *mapping = inode->i_mapping; + struct fuse_req *req; + pgoff_t index; + loff_t file_size; + unsigned int num; + unsigned int offset; + size_t total_len; + + req = fuse_get_req(fc); + if (IS_ERR(req)) + return PTR_ERR(req); + + offset = outarg->offset & ~PAGE_CACHE_MASK; + + req->in.h.opcode = FUSE_NOTIFY_REPLY; + req->in.h.nodeid = outarg->nodeid; + req->in.numargs = 2; + req->in.argpages = 1; + req->page_offset = offset; + req->end = fuse_retrieve_end; + + index = outarg->offset >> PAGE_CACHE_SHIFT; + file_size = i_size_read(inode); + num = outarg->size; + if (outarg->offset > file_size) + num = 0; + else if (outarg->offset + num > file_size) + num = file_size - outarg->offset; + + while (num) { + struct page *page; + unsigned int this_num; + + page = find_get_page(mapping, index); + if (!page) + break; + + this_num = min_t(unsigned, num, PAGE_CACHE_SIZE - offset); + req->pages[req->num_pages] = page; + req->num_pages++; + + num -= this_num; + total_len += this_num; + } + req->misc.retrieve_in.offset = outarg->offset; + req->misc.retrieve_in.size = total_len; + req->in.args[0].size = sizeof(req->misc.retrieve_in); + req->in.args[0].value = &req->misc.retrieve_in; + req->in.args[1].size = total_len; + + err = fuse_request_send_notify_reply(fc, req, outarg->notify_unique); + if (err) + fuse_retrieve_end(fc, req); + + return err; +} + +static int fuse_notify_retrieve(struct fuse_conn *fc, unsigned int size, + struct fuse_copy_state *cs) +{ + struct fuse_notify_retrieve_out outarg; + struct inode *inode; + int err; + + err = -EINVAL; + if (size != sizeof(outarg)) + goto copy_finish; + + err = fuse_copy_one(cs, &outarg, sizeof(outarg)); + if (err) + goto copy_finish; + + fuse_copy_finish(cs); + + down_read(&fc->killsb); + err = -ENOENT; + if (fc->sb) { + u64 nodeid = outarg.nodeid; + + inode = ilookup5(fc->sb, nodeid, fuse_inode_eq, &nodeid); + if (inode) { + err = fuse_retrieve(fc, inode, &outarg); + iput(inode); + } + } + up_read(&fc->killsb); + + return err; + +copy_finish: + fuse_copy_finish(cs); + return err; +} + static int fuse_notify(struct fuse_conn *fc, enum fuse_notify_code code, unsigned int size, struct fuse_copy_state *cs) { @@ -1244,6 +1455,12 @@ static int fuse_notify(struct fuse_conn *fc, enum fuse_notify_code code, case FUSE_NOTIFY_INVAL_ENTRY: return fuse_notify_inval_entry(fc, size, cs); + case FUSE_NOTIFY_STORE: + return fuse_notify_store(fc, size, cs); + + case FUSE_NOTIFY_RETRIEVE: + return fuse_notify_retrieve(fc, size, cs); + default: fuse_copy_finish(cs); return -EINVAL; diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c index 3cdc5f78a40..431be0795b6 100644 --- a/fs/fuse/dir.c +++ b/fs/fuse/dir.c @@ -1016,7 +1016,7 @@ static int fuse_permission(struct inode *inode, int mask) exist. So if permissions are revoked this won't be noticed immediately, only after the attribute timeout has expired */ - } else if (mask & MAY_ACCESS) { + } else if (mask & (MAY_ACCESS | MAY_CHDIR)) { err = fuse_access(inode, mask); } else if ((mask & MAY_EXEC) && S_ISREG(inode->i_mode)) { if (!(inode->i_mode & S_IXUGO)) { diff --git a/fs/fuse/file.c b/fs/fuse/file.c index ada0adeb3bb..147c1f71bdb 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -706,7 +706,7 @@ static int fuse_write_begin(struct file *file, struct address_space *mapping, return 0; } -static void fuse_write_update_size(struct inode *inode, loff_t pos) +void fuse_write_update_size(struct inode *inode, loff_t pos) { struct fuse_conn *fc = get_fuse_conn(inode); struct fuse_inode *fi = get_fuse_inode(inode); diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h index 8f309f04064..57d4a3a0f10 100644 --- a/fs/fuse/fuse_i.h +++ b/fs/fuse/fuse_i.h @@ -272,6 +272,7 @@ struct fuse_req { struct fuse_write_in in; struct fuse_write_out out; } write; + struct fuse_notify_retrieve_in retrieve_in; struct fuse_lk_in lk_in; } misc; @@ -748,4 +749,6 @@ long fuse_do_ioctl(struct file *file, unsigned int cmd, unsigned long arg, unsigned fuse_file_poll(struct file *file, poll_table *wait); int fuse_dev_release(struct inode *inode, struct file *file); +void fuse_write_update_size(struct inode *inode, loff_t pos); + #endif /* _FS_FUSE_I_H */ diff --git a/fs/gfs2/Kconfig b/fs/gfs2/Kconfig index a47b4310711..cc966552214 100644 --- a/fs/gfs2/Kconfig +++ b/fs/gfs2/Kconfig @@ -7,7 +7,6 @@ config GFS2_FS select IP_SCTP if DLM_SCTP select FS_POSIX_ACL select CRC32 - select SLOW_WORK select QUOTACTL help A cluster filesystem. diff --git a/fs/gfs2/aops.c b/fs/gfs2/aops.c index 9f8b52500d6..5e96cbd8a45 100644 --- a/fs/gfs2/aops.c +++ b/fs/gfs2/aops.c @@ -136,10 +136,7 @@ static int gfs2_writeback_writepage(struct page *page, if (ret <= 0) return ret; - ret = mpage_writepage(page, gfs2_get_block_noalloc, wbc); - if (ret == -EAGAIN) - ret = block_write_full_page(page, gfs2_get_block_noalloc, wbc); - return ret; + return nobh_writepage(page, gfs2_get_block_noalloc, wbc); } /** @@ -637,9 +634,7 @@ static int gfs2_write_begin(struct file *file, struct address_space *mapping, } } - error = gfs2_write_alloc_required(ip, pos, len, &alloc_required); - if (error) - goto out_unlock; + alloc_required = gfs2_write_alloc_required(ip, pos, len); if (alloc_required || gfs2_is_jdata(ip)) gfs2_write_calc_reserv(ip, len, &data_blocks, &ind_blocks); diff --git a/fs/gfs2/bmap.c b/fs/gfs2/bmap.c index 84da64b551b..6f482809d1a 100644 --- a/fs/gfs2/bmap.c +++ b/fs/gfs2/bmap.c @@ -1040,7 +1040,7 @@ static int trunc_start(struct gfs2_inode *ip, u64 size) goto out; if (gfs2_is_stuffed(ip)) { - u64 dsize = size + sizeof(struct gfs2_inode); + u64 dsize = size + sizeof(struct gfs2_dinode); ip->i_disksize = size; ip->i_inode.i_mtime = ip->i_inode.i_ctime = CURRENT_TIME; gfs2_trans_add_bh(ip->i_gl, dibh, 1); @@ -1244,13 +1244,12 @@ int gfs2_file_dealloc(struct gfs2_inode *ip) * @ip: the file being written to * @offset: the offset to write to * @len: the number of bytes being written - * @alloc_required: set to 1 if an alloc is required, 0 otherwise * - * Returns: errno + * Returns: 1 if an alloc is required, 0 otherwise */ int gfs2_write_alloc_required(struct gfs2_inode *ip, u64 offset, - unsigned int len, int *alloc_required) + unsigned int len) { struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); struct buffer_head bh; @@ -1258,26 +1257,23 @@ int gfs2_write_alloc_required(struct gfs2_inode *ip, u64 offset, u64 lblock, lblock_stop, size; u64 end_of_file; - *alloc_required = 0; - if (!len) return 0; if (gfs2_is_stuffed(ip)) { if (offset + len > sdp->sd_sb.sb_bsize - sizeof(struct gfs2_dinode)) - *alloc_required = 1; + return 1; return 0; } - *alloc_required = 1; shift = sdp->sd_sb.sb_bsize_shift; BUG_ON(gfs2_is_dir(ip)); end_of_file = (ip->i_disksize + sdp->sd_sb.sb_bsize - 1) >> shift; lblock = offset >> shift; lblock_stop = (offset + len + sdp->sd_sb.sb_bsize - 1) >> shift; if (lblock_stop > end_of_file) - return 0; + return 1; size = (lblock_stop - lblock) << shift; do { @@ -1285,12 +1281,11 @@ int gfs2_write_alloc_required(struct gfs2_inode *ip, u64 offset, bh.b_size = size; gfs2_block_map(&ip->i_inode, lblock, &bh, 0); if (!buffer_mapped(&bh)) - return 0; + return 1; size -= bh.b_size; lblock += (bh.b_size >> ip->i_inode.i_blkbits); } while(size > 0); - *alloc_required = 0; return 0; } diff --git a/fs/gfs2/bmap.h b/fs/gfs2/bmap.h index c983177e05a..a20a5213135 100644 --- a/fs/gfs2/bmap.h +++ b/fs/gfs2/bmap.h @@ -52,6 +52,6 @@ int gfs2_truncatei(struct gfs2_inode *ip, u64 size); int gfs2_truncatei_resume(struct gfs2_inode *ip); int gfs2_file_dealloc(struct gfs2_inode *ip); int gfs2_write_alloc_required(struct gfs2_inode *ip, u64 offset, - unsigned int len, int *alloc_required); + unsigned int len); #endif /* __BMAP_DOT_H__ */ diff --git a/fs/gfs2/dir.c b/fs/gfs2/dir.c index 6b48d7c268b..b9dd88a78dd 100644 --- a/fs/gfs2/dir.c +++ b/fs/gfs2/dir.c @@ -955,7 +955,12 @@ static int dir_split_leaf(struct inode *inode, const struct qstr *name) /* Change the pointers. Don't bother distinguishing stuffed from non-stuffed. This code is complicated enough already. */ - lp = kmalloc(half_len * sizeof(__be64), GFP_NOFS | __GFP_NOFAIL); + lp = kmalloc(half_len * sizeof(__be64), GFP_NOFS); + if (!lp) { + error = -ENOMEM; + goto fail_brelse; + } + /* Change the pointers */ for (x = 0; x < half_len; x++) lp[x] = cpu_to_be64(bn); @@ -1063,7 +1068,9 @@ static int dir_double_exhash(struct gfs2_inode *dip) /* Allocate both the "from" and "to" buffers in one big chunk */ - buf = kcalloc(3, sdp->sd_hash_bsize, GFP_NOFS | __GFP_NOFAIL); + buf = kcalloc(3, sdp->sd_hash_bsize, GFP_NOFS); + if (!buf) + return -ENOMEM; for (block = dip->i_disksize >> sdp->sd_hash_bsize_shift; block--;) { error = gfs2_dir_read_data(dip, (char *)buf, diff --git a/fs/gfs2/file.c b/fs/gfs2/file.c index ed9a94f0ef1..4edd662c823 100644 --- a/fs/gfs2/file.c +++ b/fs/gfs2/file.c @@ -351,7 +351,6 @@ static int gfs2_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) unsigned long last_index; u64 pos = page->index << PAGE_CACHE_SHIFT; unsigned int data_blocks, ind_blocks, rblocks; - int alloc_required = 0; struct gfs2_holder gh; struct gfs2_alloc *al; int ret; @@ -364,8 +363,7 @@ static int gfs2_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) set_bit(GLF_DIRTY, &ip->i_gl->gl_flags); set_bit(GIF_SW_PAGED, &ip->i_flags); - ret = gfs2_write_alloc_required(ip, pos, PAGE_CACHE_SIZE, &alloc_required); - if (ret || !alloc_required) + if (!gfs2_write_alloc_required(ip, pos, PAGE_CACHE_SIZE)) goto out_unlock; ret = -ENOMEM; al = gfs2_alloc_get(ip); diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c index 0898f3ec821..9adf8f924e0 100644 --- a/fs/gfs2/glock.c +++ b/fs/gfs2/glock.c @@ -328,6 +328,30 @@ static void gfs2_holder_wake(struct gfs2_holder *gh) } /** + * do_error - Something unexpected has happened during a lock request + * + */ + +static inline void do_error(struct gfs2_glock *gl, const int ret) +{ + struct gfs2_holder *gh, *tmp; + + list_for_each_entry_safe(gh, tmp, &gl->gl_holders, gh_list) { + if (test_bit(HIF_HOLDER, &gh->gh_iflags)) + continue; + if (ret & LM_OUT_ERROR) + gh->gh_error = -EIO; + else if (gh->gh_flags & (LM_FLAG_TRY | LM_FLAG_TRY_1CB)) + gh->gh_error = GLR_TRYFAILED; + else + continue; + list_del_init(&gh->gh_list); + trace_gfs2_glock_queue(gh, 0); + gfs2_holder_wake(gh); + } +} + +/** * do_promote - promote as many requests as possible on the current queue * @gl: The glock * @@ -375,36 +399,13 @@ restart: } if (gh->gh_list.prev == &gl->gl_holders) return 1; + do_error(gl, 0); break; } return 0; } /** - * do_error - Something unexpected has happened during a lock request - * - */ - -static inline void do_error(struct gfs2_glock *gl, const int ret) -{ - struct gfs2_holder *gh, *tmp; - - list_for_each_entry_safe(gh, tmp, &gl->gl_holders, gh_list) { - if (test_bit(HIF_HOLDER, &gh->gh_iflags)) - continue; - if (ret & LM_OUT_ERROR) - gh->gh_error = -EIO; - else if (gh->gh_flags & (LM_FLAG_TRY | LM_FLAG_TRY_1CB)) - gh->gh_error = GLR_TRYFAILED; - else - continue; - list_del_init(&gh->gh_list); - trace_gfs2_glock_queue(gh, 0); - gfs2_holder_wake(gh); - } -} - -/** * find_first_waiter - find the first gh that's waiting for the glock * @gl: the glock */ @@ -706,18 +707,8 @@ static void glock_work_func(struct work_struct *work) { unsigned long delay = 0; struct gfs2_glock *gl = container_of(work, struct gfs2_glock, gl_work.work); - struct gfs2_holder *gh; int drop_ref = 0; - if (unlikely(test_bit(GLF_FROZEN, &gl->gl_flags))) { - spin_lock(&gl->gl_spin); - gh = find_first_waiter(gl); - if (gh && (gh->gh_flags & LM_FLAG_NOEXP) && - test_and_clear_bit(GLF_FROZEN, &gl->gl_flags)) - set_bit(GLF_REPLY_PENDING, &gl->gl_flags); - spin_unlock(&gl->gl_spin); - } - if (test_and_clear_bit(GLF_REPLY_PENDING, &gl->gl_flags)) { finish_xmote(gl, gl->gl_reply); drop_ref = 1; @@ -1072,6 +1063,9 @@ int gfs2_glock_nq(struct gfs2_holder *gh) spin_lock(&gl->gl_spin); add_to_queue(gh); + if ((LM_FLAG_NOEXP & gh->gh_flags) && + test_and_clear_bit(GLF_FROZEN, &gl->gl_flags)) + set_bit(GLF_REPLY_PENDING, &gl->gl_flags); run_queue(gl, 1); spin_unlock(&gl->gl_spin); @@ -1329,6 +1323,36 @@ void gfs2_glock_cb(struct gfs2_glock *gl, unsigned int state) } /** + * gfs2_should_freeze - Figure out if glock should be frozen + * @gl: The glock in question + * + * Glocks are not frozen if (a) the result of the dlm operation is + * an error, (b) the locking operation was an unlock operation or + * (c) if there is a "noexp" flagged request anywhere in the queue + * + * Returns: 1 if freezing should occur, 0 otherwise + */ + +static int gfs2_should_freeze(const struct gfs2_glock *gl) +{ + const struct gfs2_holder *gh; + + if (gl->gl_reply & ~LM_OUT_ST_MASK) + return 0; + if (gl->gl_target == LM_ST_UNLOCKED) + return 0; + + list_for_each_entry(gh, &gl->gl_holders, gh_list) { + if (test_bit(HIF_HOLDER, &gh->gh_iflags)) + continue; + if (LM_FLAG_NOEXP & gh->gh_flags) + return 0; + } + + return 1; +} + +/** * gfs2_glock_complete - Callback used by locking * @gl: Pointer to the glock * @ret: The return value from the dlm @@ -1338,18 +1362,17 @@ void gfs2_glock_cb(struct gfs2_glock *gl, unsigned int state) void gfs2_glock_complete(struct gfs2_glock *gl, int ret) { struct lm_lockstruct *ls = &gl->gl_sbd->sd_lockstruct; + gl->gl_reply = ret; + if (unlikely(test_bit(DFL_BLOCK_LOCKS, &ls->ls_flags))) { - struct gfs2_holder *gh; spin_lock(&gl->gl_spin); - gh = find_first_waiter(gl); - if ((!(gh && (gh->gh_flags & LM_FLAG_NOEXP)) && - (gl->gl_target != LM_ST_UNLOCKED)) || - ((ret & ~LM_OUT_ST_MASK) != 0)) + if (gfs2_should_freeze(gl)) { set_bit(GLF_FROZEN, &gl->gl_flags); - spin_unlock(&gl->gl_spin); - if (test_bit(GLF_FROZEN, &gl->gl_flags)) + spin_unlock(&gl->gl_spin); return; + } + spin_unlock(&gl->gl_spin); } set_bit(GLF_REPLY_PENDING, &gl->gl_flags); gfs2_glock_hold(gl); diff --git a/fs/gfs2/incore.h b/fs/gfs2/incore.h index b5d7363b22d..fdbf4b366fa 100644 --- a/fs/gfs2/incore.h +++ b/fs/gfs2/incore.h @@ -12,7 +12,6 @@ #include <linux/fs.h> #include <linux/workqueue.h> -#include <linux/slow-work.h> #include <linux/dlm.h> #include <linux/buffer_head.h> @@ -383,7 +382,7 @@ struct gfs2_journal_extent { struct gfs2_jdesc { struct list_head jd_list; struct list_head extent_list; - struct slow_work jd_work; + struct work_struct jd_work; struct inode *jd_inode; unsigned long jd_flags; #define JDF_RECOVERY 1 @@ -460,6 +459,7 @@ enum { SDF_NOBARRIERS = 3, SDF_NORECOVERY = 4, SDF_DEMOTE = 5, + SDF_NOJOURNALID = 6, }; #define GFS2_FSNAME_LEN 256 diff --git a/fs/gfs2/main.c b/fs/gfs2/main.c index fb2a5f93b7c..b1e9630eb46 100644 --- a/fs/gfs2/main.c +++ b/fs/gfs2/main.c @@ -15,7 +15,6 @@ #include <linux/init.h> #include <linux/gfs2_ondisk.h> #include <asm/atomic.h> -#include <linux/slow-work.h> #include "gfs2.h" #include "incore.h" @@ -24,6 +23,7 @@ #include "util.h" #include "glock.h" #include "quota.h" +#include "recovery.h" static struct shrinker qd_shrinker = { .shrink = gfs2_shrink_qd_memory, @@ -138,9 +138,11 @@ static int __init init_gfs2_fs(void) if (error) goto fail_unregister; - error = slow_work_register_user(THIS_MODULE); - if (error) - goto fail_slow; + error = -ENOMEM; + gfs_recovery_wq = alloc_workqueue("gfs_recovery", + WQ_NON_REENTRANT | WQ_RESCUER, 0); + if (!gfs_recovery_wq) + goto fail_wq; gfs2_register_debugfs(); @@ -148,7 +150,7 @@ static int __init init_gfs2_fs(void) return 0; -fail_slow: +fail_wq: unregister_filesystem(&gfs2meta_fs_type); fail_unregister: unregister_filesystem(&gfs2_fs_type); @@ -190,7 +192,7 @@ static void __exit exit_gfs2_fs(void) gfs2_unregister_debugfs(); unregister_filesystem(&gfs2_fs_type); unregister_filesystem(&gfs2meta_fs_type); - slow_work_unregister_user(THIS_MODULE); + destroy_workqueue(gfs_recovery_wq); kmem_cache_destroy(gfs2_quotad_cachep); kmem_cache_destroy(gfs2_rgrpd_cachep); diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c index 3593b3a7290..4f44bdeb2f0 100644 --- a/fs/gfs2/ops_fstype.c +++ b/fs/gfs2/ops_fstype.c @@ -17,7 +17,6 @@ #include <linux/namei.h> #include <linux/mount.h> #include <linux/gfs2_ondisk.h> -#include <linux/slow-work.h> #include <linux/quotaops.h> #include "gfs2.h" @@ -76,7 +75,7 @@ static struct gfs2_sbd *init_sbd(struct super_block *sb) sb->s_fs_info = sdp; sdp->sd_vfs = sb; - + set_bit(SDF_NOJOURNALID, &sdp->sd_flags); gfs2_tune_init(&sdp->sd_tune); init_waitqueue_head(&sdp->sd_glock_wait); @@ -673,7 +672,7 @@ static int gfs2_jindex_hold(struct gfs2_sbd *sdp, struct gfs2_holder *ji_gh) break; INIT_LIST_HEAD(&jd->extent_list); - slow_work_init(&jd->jd_work, &gfs2_recover_ops); + INIT_WORK(&jd->jd_work, gfs2_recover_func); jd->jd_inode = gfs2_lookupi(sdp->sd_jindex, &name, 1); if (!jd->jd_inode || IS_ERR(jd->jd_inode)) { if (!jd->jd_inode) @@ -782,7 +781,8 @@ static int init_journal(struct gfs2_sbd *sdp, int undo) if (sdp->sd_lockstruct.ls_first) { unsigned int x; for (x = 0; x < sdp->sd_journals; x++) { - error = gfs2_recover_journal(gfs2_jdesc_find(sdp, x)); + error = gfs2_recover_journal(gfs2_jdesc_find(sdp, x), + true); if (error) { fs_err(sdp, "error recovering journal %u: %d\n", x, error); @@ -792,7 +792,7 @@ static int init_journal(struct gfs2_sbd *sdp, int undo) gfs2_others_may_mount(sdp); } else if (!sdp->sd_args.ar_spectator) { - error = gfs2_recover_journal(sdp->sd_jdesc); + error = gfs2_recover_journal(sdp->sd_jdesc, true); if (error) { fs_err(sdp, "error recovering my journal: %d\n", error); goto fail_jinode_gh; @@ -1050,7 +1050,8 @@ static int gfs2_lm_mount(struct gfs2_sbd *sdp, int silent) ret = match_int(&tmp[0], &option); if (ret || option < 0) goto hostdata_error; - ls->ls_jid = option; + if (test_and_clear_bit(SDF_NOJOURNALID, &sdp->sd_flags)) + ls->ls_jid = option; break; case Opt_id: /* Obsolete, but left for backward compat purposes */ @@ -1102,6 +1103,24 @@ void gfs2_lm_unmount(struct gfs2_sbd *sdp) lm->lm_unmount(sdp); } +static int gfs2_journalid_wait(void *word) +{ + if (signal_pending(current)) + return -EINTR; + schedule(); + return 0; +} + +static int wait_on_journal(struct gfs2_sbd *sdp) +{ + if (sdp->sd_args.ar_spectator) + return 0; + if (sdp->sd_lockstruct.ls_ops->lm_mount == NULL) + return 0; + + return wait_on_bit(&sdp->sd_flags, SDF_NOJOURNALID, gfs2_journalid_wait, TASK_INTERRUPTIBLE); +} + void gfs2_online_uevent(struct gfs2_sbd *sdp) { struct super_block *sb = sdp->sd_vfs; @@ -1194,6 +1213,10 @@ static int fill_super(struct super_block *sb, struct gfs2_args *args, int silent if (error) goto fail_locking; + error = wait_on_journal(sdp); + if (error) + goto fail_sb; + error = init_inodes(sdp, DO); if (error) goto fail_sb; diff --git a/fs/gfs2/quota.c b/fs/gfs2/quota.c index 8f02d3db8f4..1bc6b5695e6 100644 --- a/fs/gfs2/quota.c +++ b/fs/gfs2/quota.c @@ -787,15 +787,9 @@ static int do_sync(unsigned int num_qd, struct gfs2_quota_data **qda) goto out; for (x = 0; x < num_qd; x++) { - int alloc_required; - offset = qd2offset(qda[x]); - error = gfs2_write_alloc_required(ip, offset, - sizeof(struct gfs2_quota), - &alloc_required); - if (error) - goto out_gunlock; - if (alloc_required) + if (gfs2_write_alloc_required(ip, offset, + sizeof(struct gfs2_quota))) nalloc++; } @@ -1455,10 +1449,10 @@ static int gfs2_quota_get_xstate(struct super_block *sb, switch (sdp->sd_args.ar_quota) { case GFS2_QUOTA_ON: - fqs->qs_flags |= (XFS_QUOTA_UDQ_ENFD | XFS_QUOTA_GDQ_ENFD); + fqs->qs_flags |= (FS_QUOTA_UDQ_ENFD | FS_QUOTA_GDQ_ENFD); /*FALLTHRU*/ case GFS2_QUOTA_ACCOUNT: - fqs->qs_flags |= (XFS_QUOTA_UDQ_ACCT | XFS_QUOTA_GDQ_ACCT); + fqs->qs_flags |= (FS_QUOTA_UDQ_ACCT | FS_QUOTA_GDQ_ACCT); break; case GFS2_QUOTA_OFF: break; @@ -1504,7 +1498,7 @@ static int gfs2_get_dqblk(struct super_block *sb, int type, qid_t id, qlvb = (struct gfs2_quota_lvb *)qd->qd_gl->gl_lvb; fdq->d_version = FS_DQUOT_VERSION; - fdq->d_flags = (type == QUOTA_USER) ? XFS_USER_QUOTA : XFS_GROUP_QUOTA; + fdq->d_flags = (type == QUOTA_USER) ? FS_USER_QUOTA : FS_GROUP_QUOTA; fdq->d_id = id; fdq->d_blk_hardlimit = be64_to_cpu(qlvb->qb_limit); fdq->d_blk_softlimit = be64_to_cpu(qlvb->qb_warn); @@ -1539,12 +1533,12 @@ static int gfs2_set_dqblk(struct super_block *sb, int type, qid_t id, switch(type) { case USRQUOTA: type = QUOTA_USER; - if (fdq->d_flags != XFS_USER_QUOTA) + if (fdq->d_flags != FS_USER_QUOTA) return -EINVAL; break; case GRPQUOTA: type = QUOTA_GROUP; - if (fdq->d_flags != XFS_GROUP_QUOTA) + if (fdq->d_flags != FS_GROUP_QUOTA) return -EINVAL; break; default: @@ -1584,10 +1578,7 @@ static int gfs2_set_dqblk(struct super_block *sb, int type, qid_t id, goto out_i; offset = qd2offset(qd); - error = gfs2_write_alloc_required(ip, offset, sizeof(struct gfs2_quota), - &alloc_required); - if (error) - goto out_i; + alloc_required = gfs2_write_alloc_required(ip, offset, sizeof(struct gfs2_quota)); if (alloc_required) { al = gfs2_alloc_get(ip); if (al == NULL) diff --git a/fs/gfs2/recovery.c b/fs/gfs2/recovery.c index 4b9bece3d43..f7f89a94a5a 100644 --- a/fs/gfs2/recovery.c +++ b/fs/gfs2/recovery.c @@ -14,7 +14,6 @@ #include <linux/buffer_head.h> #include <linux/gfs2_ondisk.h> #include <linux/crc32.h> -#include <linux/slow-work.h> #include "gfs2.h" #include "incore.h" @@ -28,6 +27,8 @@ #include "util.h" #include "dir.h" +struct workqueue_struct *gfs_recovery_wq; + int gfs2_replay_read_block(struct gfs2_jdesc *jd, unsigned int blk, struct buffer_head **bh) { @@ -443,23 +444,7 @@ static void gfs2_recovery_done(struct gfs2_sbd *sdp, unsigned int jid, kobject_uevent_env(&sdp->sd_kobj, KOBJ_CHANGE, envp); } -static int gfs2_recover_get_ref(struct slow_work *work) -{ - struct gfs2_jdesc *jd = container_of(work, struct gfs2_jdesc, jd_work); - if (test_and_set_bit(JDF_RECOVERY, &jd->jd_flags)) - return -EBUSY; - return 0; -} - -static void gfs2_recover_put_ref(struct slow_work *work) -{ - struct gfs2_jdesc *jd = container_of(work, struct gfs2_jdesc, jd_work); - clear_bit(JDF_RECOVERY, &jd->jd_flags); - smp_mb__after_clear_bit(); - wake_up_bit(&jd->jd_flags, JDF_RECOVERY); -} - -static void gfs2_recover_work(struct slow_work *work) +void gfs2_recover_func(struct work_struct *work) { struct gfs2_jdesc *jd = container_of(work, struct gfs2_jdesc, jd_work); struct gfs2_inode *ip = GFS2_I(jd->jd_inode); @@ -578,7 +563,7 @@ static void gfs2_recover_work(struct slow_work *work) gfs2_glock_dq_uninit(&j_gh); fs_info(sdp, "jid=%u: Done\n", jd->jd_jid); - return; + goto done; fail_gunlock_tr: gfs2_glock_dq_uninit(&t_gh); @@ -590,32 +575,35 @@ fail_gunlock_j: } fs_info(sdp, "jid=%u: %s\n", jd->jd_jid, (error) ? "Failed" : "Done"); - fail: gfs2_recovery_done(sdp, jd->jd_jid, LM_RD_GAVEUP); +done: + clear_bit(JDF_RECOVERY, &jd->jd_flags); + smp_mb__after_clear_bit(); + wake_up_bit(&jd->jd_flags, JDF_RECOVERY); } -struct slow_work_ops gfs2_recover_ops = { - .owner = THIS_MODULE, - .get_ref = gfs2_recover_get_ref, - .put_ref = gfs2_recover_put_ref, - .execute = gfs2_recover_work, -}; - - static int gfs2_recovery_wait(void *word) { schedule(); return 0; } -int gfs2_recover_journal(struct gfs2_jdesc *jd) +int gfs2_recover_journal(struct gfs2_jdesc *jd, bool wait) { int rv; - rv = slow_work_enqueue(&jd->jd_work); - if (rv) - return rv; - wait_on_bit(&jd->jd_flags, JDF_RECOVERY, gfs2_recovery_wait, TASK_UNINTERRUPTIBLE); + + if (test_and_set_bit(JDF_RECOVERY, &jd->jd_flags)) + return -EBUSY; + + /* we have JDF_RECOVERY, queue should always succeed */ + rv = queue_work(gfs_recovery_wq, &jd->jd_work); + BUG_ON(!rv); + + if (wait) + wait_on_bit(&jd->jd_flags, JDF_RECOVERY, gfs2_recovery_wait, + TASK_UNINTERRUPTIBLE); + return 0; } diff --git a/fs/gfs2/recovery.h b/fs/gfs2/recovery.h index 1616ac22569..2226136c764 100644 --- a/fs/gfs2/recovery.h +++ b/fs/gfs2/recovery.h @@ -12,6 +12,8 @@ #include "incore.h" +extern struct workqueue_struct *gfs_recovery_wq; + static inline void gfs2_replay_incr_blk(struct gfs2_sbd *sdp, unsigned int *blk) { if (++*blk == sdp->sd_jdesc->jd_blocks) @@ -27,8 +29,8 @@ extern void gfs2_revoke_clean(struct gfs2_sbd *sdp); extern int gfs2_find_jhead(struct gfs2_jdesc *jd, struct gfs2_log_header_host *head); -extern int gfs2_recover_journal(struct gfs2_jdesc *gfs2_jd); -extern struct slow_work_ops gfs2_recover_ops; +extern int gfs2_recover_journal(struct gfs2_jdesc *gfs2_jd, bool wait); +extern void gfs2_recover_func(struct work_struct *work); #endif /* __RECOVERY_DOT_H__ */ diff --git a/fs/gfs2/super.c b/fs/gfs2/super.c index 4d1aad38f1b..4140811a921 100644 --- a/fs/gfs2/super.c +++ b/fs/gfs2/super.c @@ -342,8 +342,6 @@ int gfs2_jdesc_check(struct gfs2_jdesc *jd) { struct gfs2_inode *ip = GFS2_I(jd->jd_inode); struct gfs2_sbd *sdp = GFS2_SB(jd->jd_inode); - int ar; - int error; if (ip->i_disksize < (8 << 20) || ip->i_disksize > (1 << 30) || (ip->i_disksize & (sdp->sd_sb.sb_bsize - 1))) { @@ -352,13 +350,12 @@ int gfs2_jdesc_check(struct gfs2_jdesc *jd) } jd->jd_blocks = ip->i_disksize >> sdp->sd_sb.sb_bsize_shift; - error = gfs2_write_alloc_required(ip, 0, ip->i_disksize, &ar); - if (!error && ar) { + if (gfs2_write_alloc_required(ip, 0, ip->i_disksize)) { gfs2_consist_inode(ip); - error = -EIO; + return -EIO; } - return error; + return 0; } /** diff --git a/fs/gfs2/sys.c b/fs/gfs2/sys.c index 37f5393e68e..ccacffd2faa 100644 --- a/fs/gfs2/sys.c +++ b/fs/gfs2/sys.c @@ -25,6 +25,7 @@ #include "quota.h" #include "util.h" #include "glops.h" +#include "recovery.h" struct gfs2_attr { struct attribute attr; @@ -325,6 +326,30 @@ static ssize_t lkfirst_show(struct gfs2_sbd *sdp, char *buf) return sprintf(buf, "%d\n", ls->ls_first); } +static ssize_t lkfirst_store(struct gfs2_sbd *sdp, const char *buf, size_t len) +{ + unsigned first; + int rv; + + rv = sscanf(buf, "%u", &first); + if (rv != 1 || first > 1) + return -EINVAL; + spin_lock(&sdp->sd_jindex_spin); + rv = -EBUSY; + if (test_bit(SDF_NOJOURNALID, &sdp->sd_flags) == 0) + goto out; + rv = -EINVAL; + if (sdp->sd_args.ar_spectator) + goto out; + if (sdp->sd_lockstruct.ls_ops->lm_mount == NULL) + goto out; + sdp->sd_lockstruct.ls_first = first; + rv = 0; +out: + spin_unlock(&sdp->sd_jindex_spin); + return rv ? rv : len; +} + static ssize_t first_done_show(struct gfs2_sbd *sdp, char *buf) { struct lm_lockstruct *ls = &sdp->sd_lockstruct; @@ -352,7 +377,7 @@ static ssize_t recover_store(struct gfs2_sbd *sdp, const char *buf, size_t len) list_for_each_entry(jd, &sdp->sd_jindex_list, jd_list) { if (jd->jd_jid != jid) continue; - rv = slow_work_enqueue(&jd->jd_work); + rv = gfs2_recover_journal(jd, false); break; } out: @@ -377,14 +402,41 @@ static ssize_t jid_show(struct gfs2_sbd *sdp, char *buf) return sprintf(buf, "%u\n", sdp->sd_lockstruct.ls_jid); } +static ssize_t jid_store(struct gfs2_sbd *sdp, const char *buf, size_t len) +{ + unsigned jid; + int rv; + + rv = sscanf(buf, "%u", &jid); + if (rv != 1) + return -EINVAL; + + spin_lock(&sdp->sd_jindex_spin); + rv = -EINVAL; + if (sdp->sd_args.ar_spectator) + goto out; + if (sdp->sd_lockstruct.ls_ops->lm_mount == NULL) + goto out; + rv = -EBUSY; + if (test_and_clear_bit(SDF_NOJOURNALID, &sdp->sd_flags) == 0) + goto out; + sdp->sd_lockstruct.ls_jid = jid; + smp_mb__after_clear_bit(); + wake_up_bit(&sdp->sd_flags, SDF_NOJOURNALID); + rv = 0; +out: + spin_unlock(&sdp->sd_jindex_spin); + return rv ? rv : len; +} + #define GDLM_ATTR(_name,_mode,_show,_store) \ static struct gfs2_attr gdlm_attr_##_name = __ATTR(_name,_mode,_show,_store) GDLM_ATTR(proto_name, 0444, proto_name_show, NULL); GDLM_ATTR(block, 0644, block_show, block_store); GDLM_ATTR(withdraw, 0644, withdraw_show, withdraw_store); -GDLM_ATTR(jid, 0444, jid_show, NULL); -GDLM_ATTR(first, 0444, lkfirst_show, NULL); +GDLM_ATTR(jid, 0644, jid_show, jid_store); +GDLM_ATTR(first, 0644, lkfirst_show, lkfirst_store); GDLM_ATTR(first_done, 0444, first_done_show, NULL); GDLM_ATTR(recover, 0600, NULL, recover_store); GDLM_ATTR(recover_done, 0444, recover_done_show, NULL); @@ -564,7 +616,7 @@ static int gfs2_uevent(struct kset *kset, struct kobject *kobj, add_uevent_var(env, "LOCKTABLE=%s", sdp->sd_table_name); add_uevent_var(env, "LOCKPROTO=%s", sdp->sd_proto_name); - if (!sdp->sd_args.ar_spectator) + if (!test_bit(SDF_NOJOURNALID, &sdp->sd_flags)) add_uevent_var(env, "JOURNALID=%u", sdp->sd_lockstruct.ls_jid); if (gfs2_uuid_valid(uuid)) add_uevent_var(env, "UUID=%pUB", uuid); diff --git a/fs/jbd/journal.c b/fs/jbd/journal.c index 93d1e47647b..f19ce94693d 100644 --- a/fs/jbd/journal.c +++ b/fs/jbd/journal.c @@ -1281,13 +1281,9 @@ int journal_check_used_features (journal_t *journal, unsigned long compat, int journal_check_available_features (journal_t *journal, unsigned long compat, unsigned long ro, unsigned long incompat) { - journal_superblock_t *sb; - if (!compat && !ro && !incompat) return 1; - sb = journal->j_superblock; - /* We can support any known requested features iff the * superblock is in version 2. Otherwise we fail to support any * extended sb features. */ @@ -1481,7 +1477,6 @@ int journal_flush(journal_t *journal) int journal_wipe(journal_t *journal, int write) { - journal_superblock_t *sb; int err = 0; J_ASSERT (!(journal->j_flags & JFS_LOADED)); @@ -1490,8 +1485,6 @@ int journal_wipe(journal_t *journal, int write) if (err) return err; - sb = journal->j_superblock; - if (!journal->j_tail) goto no_recovery; diff --git a/fs/jbd/recovery.c b/fs/jbd/recovery.c index 54c9bc9e1b1..81051dafebf 100644 --- a/fs/jbd/recovery.c +++ b/fs/jbd/recovery.c @@ -283,12 +283,9 @@ int journal_recover(journal_t *journal) int journal_skip_recovery(journal_t *journal) { int err; - journal_superblock_t * sb; - struct recovery_info info; memset (&info, 0, sizeof(info)); - sb = journal->j_superblock; err = do_one_pass(journal, &info, PASS_SCAN); @@ -297,7 +294,8 @@ int journal_skip_recovery(journal_t *journal) ++journal->j_transaction_sequence; } else { #ifdef CONFIG_JBD_DEBUG - int dropped = info.end_transaction - be32_to_cpu(sb->s_sequence); + int dropped = info.end_transaction - + be32_to_cpu(journal->j_superblock->s_sequence); #endif jbd_debug(1, "JBD: ignoring %d transaction%s from the journal.\n", @@ -321,11 +319,6 @@ static int do_one_pass(journal_t *journal, unsigned int sequence; int blocktype; - /* Precompute the maximum metadata descriptors in a descriptor block */ - int MAX_BLOCKS_PER_DESC; - MAX_BLOCKS_PER_DESC = ((journal->j_blocksize-sizeof(journal_header_t)) - / sizeof(journal_block_tag_t)); - /* * First thing is to establish what we expect to find in the log * (in terms of transaction IDs), and where (in terms of log diff --git a/fs/jbd2/checkpoint.c b/fs/jbd2/checkpoint.c index 076d1cc44f9..1c23a0f4e8a 100644 --- a/fs/jbd2/checkpoint.c +++ b/fs/jbd2/checkpoint.c @@ -118,13 +118,13 @@ static int __try_to_free_cp_buf(struct journal_head *jh) void __jbd2_log_wait_for_space(journal_t *journal) { int nblocks, space_left; - assert_spin_locked(&journal->j_state_lock); + /* assert_spin_locked(&journal->j_state_lock); */ nblocks = jbd_space_needed(journal); while (__jbd2_log_space_left(journal) < nblocks) { if (journal->j_flags & JBD2_ABORT) return; - spin_unlock(&journal->j_state_lock); + write_unlock(&journal->j_state_lock); mutex_lock(&journal->j_checkpoint_mutex); /* @@ -138,7 +138,7 @@ void __jbd2_log_wait_for_space(journal_t *journal) * filesystem, so abort the journal and leave a stack * trace for forensic evidence. */ - spin_lock(&journal->j_state_lock); + write_lock(&journal->j_state_lock); spin_lock(&journal->j_list_lock); nblocks = jbd_space_needed(journal); space_left = __jbd2_log_space_left(journal); @@ -149,7 +149,7 @@ void __jbd2_log_wait_for_space(journal_t *journal) if (journal->j_committing_transaction) tid = journal->j_committing_transaction->t_tid; spin_unlock(&journal->j_list_lock); - spin_unlock(&journal->j_state_lock); + write_unlock(&journal->j_state_lock); if (chkpt) { jbd2_log_do_checkpoint(journal); } else if (jbd2_cleanup_journal_tail(journal) == 0) { @@ -167,7 +167,7 @@ void __jbd2_log_wait_for_space(journal_t *journal) WARN_ON(1); jbd2_journal_abort(journal, 0); } - spin_lock(&journal->j_state_lock); + write_lock(&journal->j_state_lock); } else { spin_unlock(&journal->j_list_lock); } @@ -474,7 +474,7 @@ int jbd2_cleanup_journal_tail(journal_t *journal) * next transaction ID we will write, and where it will * start. */ - spin_lock(&journal->j_state_lock); + write_lock(&journal->j_state_lock); spin_lock(&journal->j_list_lock); transaction = journal->j_checkpoint_transactions; if (transaction) { @@ -496,7 +496,7 @@ int jbd2_cleanup_journal_tail(journal_t *journal) /* If the oldest pinned transaction is at the tail of the log already then there's not much we can do right now. */ if (journal->j_tail_sequence == first_tid) { - spin_unlock(&journal->j_state_lock); + write_unlock(&journal->j_state_lock); return 1; } @@ -516,7 +516,7 @@ int jbd2_cleanup_journal_tail(journal_t *journal) journal->j_free += freed; journal->j_tail_sequence = first_tid; journal->j_tail = blocknr; - spin_unlock(&journal->j_state_lock); + write_unlock(&journal->j_state_lock); /* * If there is an external journal, we need to make sure that @@ -775,7 +775,7 @@ void __jbd2_journal_drop_transaction(journal_t *journal, transaction_t *transact J_ASSERT(transaction->t_log_list == NULL); J_ASSERT(transaction->t_checkpoint_list == NULL); J_ASSERT(transaction->t_checkpoint_io_list == NULL); - J_ASSERT(transaction->t_updates == 0); + J_ASSERT(atomic_read(&transaction->t_updates) == 0); J_ASSERT(journal->j_committing_transaction != transaction); J_ASSERT(journal->j_running_transaction != transaction); diff --git a/fs/jbd2/commit.c b/fs/jbd2/commit.c index 75716d3d2be..f52e5e8049f 100644 --- a/fs/jbd2/commit.c +++ b/fs/jbd2/commit.c @@ -150,11 +150,11 @@ static int journal_submit_commit_record(journal_t *journal, */ if (ret == -EOPNOTSUPP && barrier_done) { printk(KERN_WARNING - "JBD: barrier-based sync failed on %s - " - "disabling barriers\n", journal->j_devname); - spin_lock(&journal->j_state_lock); + "JBD2: Disabling barriers on %s, " + "not supported by device\n", journal->j_devname); + write_lock(&journal->j_state_lock); journal->j_flags &= ~JBD2_BARRIER; - spin_unlock(&journal->j_state_lock); + write_unlock(&journal->j_state_lock); /* And try again, without the barrier */ lock_buffer(bh); @@ -180,11 +180,11 @@ retry: wait_on_buffer(bh); if (buffer_eopnotsupp(bh) && (journal->j_flags & JBD2_BARRIER)) { printk(KERN_WARNING - "JBD2: wait_on_commit_record: sync failed on %s - " - "disabling barriers\n", journal->j_devname); - spin_lock(&journal->j_state_lock); + "JBD2: %s: disabling barries on %s - not supported " + "by device\n", __func__, journal->j_devname); + write_lock(&journal->j_state_lock); journal->j_flags &= ~JBD2_BARRIER; - spin_unlock(&journal->j_state_lock); + write_unlock(&journal->j_state_lock); lock_buffer(bh); clear_buffer_dirty(bh); @@ -400,7 +400,7 @@ void jbd2_journal_commit_transaction(journal_t *journal) jbd_debug(1, "JBD: starting commit of transaction %d\n", commit_transaction->t_tid); - spin_lock(&journal->j_state_lock); + write_lock(&journal->j_state_lock); commit_transaction->t_state = T_LOCKED; /* @@ -417,23 +417,23 @@ void jbd2_journal_commit_transaction(journal_t *journal) stats.run.rs_locked); spin_lock(&commit_transaction->t_handle_lock); - while (commit_transaction->t_updates) { + while (atomic_read(&commit_transaction->t_updates)) { DEFINE_WAIT(wait); prepare_to_wait(&journal->j_wait_updates, &wait, TASK_UNINTERRUPTIBLE); - if (commit_transaction->t_updates) { + if (atomic_read(&commit_transaction->t_updates)) { spin_unlock(&commit_transaction->t_handle_lock); - spin_unlock(&journal->j_state_lock); + write_unlock(&journal->j_state_lock); schedule(); - spin_lock(&journal->j_state_lock); + write_lock(&journal->j_state_lock); spin_lock(&commit_transaction->t_handle_lock); } finish_wait(&journal->j_wait_updates, &wait); } spin_unlock(&commit_transaction->t_handle_lock); - J_ASSERT (commit_transaction->t_outstanding_credits <= + J_ASSERT (atomic_read(&commit_transaction->t_outstanding_credits) <= journal->j_max_transaction_buffers); /* @@ -497,7 +497,7 @@ void jbd2_journal_commit_transaction(journal_t *journal) start_time = ktime_get(); commit_transaction->t_log_start = journal->j_head; wake_up(&journal->j_wait_transaction_locked); - spin_unlock(&journal->j_state_lock); + write_unlock(&journal->j_state_lock); jbd_debug (3, "JBD: commit phase 2\n"); @@ -519,19 +519,20 @@ void jbd2_journal_commit_transaction(journal_t *journal) * transaction! Now comes the tricky part: we need to write out * metadata. Loop over the transaction's entire buffer list: */ - spin_lock(&journal->j_state_lock); + write_lock(&journal->j_state_lock); commit_transaction->t_state = T_COMMIT; - spin_unlock(&journal->j_state_lock); + write_unlock(&journal->j_state_lock); trace_jbd2_commit_logging(journal, commit_transaction); stats.run.rs_logging = jiffies; stats.run.rs_flushing = jbd2_time_diff(stats.run.rs_flushing, stats.run.rs_logging); - stats.run.rs_blocks = commit_transaction->t_outstanding_credits; + stats.run.rs_blocks = + atomic_read(&commit_transaction->t_outstanding_credits); stats.run.rs_blocks_logged = 0; J_ASSERT(commit_transaction->t_nr_buffers <= - commit_transaction->t_outstanding_credits); + atomic_read(&commit_transaction->t_outstanding_credits)); err = 0; descriptor = NULL; @@ -616,7 +617,7 @@ void jbd2_journal_commit_transaction(journal_t *journal) * the free space in the log, but this counter is changed * by jbd2_journal_next_log_block() also. */ - commit_transaction->t_outstanding_credits--; + atomic_dec(&commit_transaction->t_outstanding_credits); /* Bump b_count to prevent truncate from stumbling over the shadowed buffer! @@@ This can go if we ever get @@ -977,7 +978,7 @@ restart_loop: * __jbd2_journal_drop_transaction(). Otherwise we could race with * other checkpointing code processing the transaction... */ - spin_lock(&journal->j_state_lock); + write_lock(&journal->j_state_lock); spin_lock(&journal->j_list_lock); /* * Now recheck if some buffers did not get attached to the transaction @@ -985,7 +986,7 @@ restart_loop: */ if (commit_transaction->t_forget) { spin_unlock(&journal->j_list_lock); - spin_unlock(&journal->j_state_lock); + write_unlock(&journal->j_state_lock); goto restart_loop; } @@ -1003,7 +1004,8 @@ restart_loop: * File the transaction statistics */ stats.ts_tid = commit_transaction->t_tid; - stats.run.rs_handle_count = commit_transaction->t_handle_count; + stats.run.rs_handle_count = + atomic_read(&commit_transaction->t_handle_count); trace_jbd2_run_stats(journal->j_fs_dev->bd_dev, commit_transaction->t_tid, &stats.run); @@ -1037,7 +1039,7 @@ restart_loop: journal->j_average_commit_time*3) / 4; else journal->j_average_commit_time = commit_time; - spin_unlock(&journal->j_state_lock); + write_unlock(&journal->j_state_lock); if (commit_transaction->t_checkpoint_list == NULL && commit_transaction->t_checkpoint_io_list == NULL) { diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c index 036880895bf..ad5866aaf0f 100644 --- a/fs/jbd2/journal.c +++ b/fs/jbd2/journal.c @@ -41,6 +41,7 @@ #include <linux/hash.h> #include <linux/log2.h> #include <linux/vmalloc.h> +#include <linux/backing-dev.h> #define CREATE_TRACE_POINTS #include <trace/events/jbd2.h> @@ -48,8 +49,6 @@ #include <asm/uaccess.h> #include <asm/page.h> -EXPORT_SYMBOL(jbd2_journal_start); -EXPORT_SYMBOL(jbd2_journal_restart); EXPORT_SYMBOL(jbd2_journal_extend); EXPORT_SYMBOL(jbd2_journal_stop); EXPORT_SYMBOL(jbd2_journal_lock_updates); @@ -143,7 +142,7 @@ static int kjournald2(void *arg) /* * And now, wait forever for commit wakeup events. */ - spin_lock(&journal->j_state_lock); + write_lock(&journal->j_state_lock); loop: if (journal->j_flags & JBD2_UNMOUNT) @@ -154,10 +153,10 @@ loop: if (journal->j_commit_sequence != journal->j_commit_request) { jbd_debug(1, "OK, requests differ\n"); - spin_unlock(&journal->j_state_lock); + write_unlock(&journal->j_state_lock); del_timer_sync(&journal->j_commit_timer); jbd2_journal_commit_transaction(journal); - spin_lock(&journal->j_state_lock); + write_lock(&journal->j_state_lock); goto loop; } @@ -169,9 +168,9 @@ loop: * be already stopped. */ jbd_debug(1, "Now suspending kjournald2\n"); - spin_unlock(&journal->j_state_lock); + write_unlock(&journal->j_state_lock); refrigerator(); - spin_lock(&journal->j_state_lock); + write_lock(&journal->j_state_lock); } else { /* * We assume on resume that commits are already there, @@ -191,9 +190,9 @@ loop: if (journal->j_flags & JBD2_UNMOUNT) should_sleep = 0; if (should_sleep) { - spin_unlock(&journal->j_state_lock); + write_unlock(&journal->j_state_lock); schedule(); - spin_lock(&journal->j_state_lock); + write_lock(&journal->j_state_lock); } finish_wait(&journal->j_wait_commit, &wait); } @@ -211,7 +210,7 @@ loop: goto loop; end_loop: - spin_unlock(&journal->j_state_lock); + write_unlock(&journal->j_state_lock); del_timer_sync(&journal->j_commit_timer); journal->j_task = NULL; wake_up(&journal->j_wait_done_commit); @@ -234,16 +233,16 @@ static int jbd2_journal_start_thread(journal_t *journal) static void journal_kill_thread(journal_t *journal) { - spin_lock(&journal->j_state_lock); + write_lock(&journal->j_state_lock); journal->j_flags |= JBD2_UNMOUNT; while (journal->j_task) { wake_up(&journal->j_wait_commit); - spin_unlock(&journal->j_state_lock); + write_unlock(&journal->j_state_lock); wait_event(journal->j_wait_done_commit, journal->j_task == NULL); - spin_lock(&journal->j_state_lock); + write_lock(&journal->j_state_lock); } - spin_unlock(&journal->j_state_lock); + write_unlock(&journal->j_state_lock); } /* @@ -310,7 +309,17 @@ int jbd2_journal_write_metadata_buffer(transaction_t *transaction, */ J_ASSERT_BH(bh_in, buffer_jbddirty(bh_in)); - new_bh = alloc_buffer_head(GFP_NOFS|__GFP_NOFAIL); +retry_alloc: + new_bh = alloc_buffer_head(GFP_NOFS); + if (!new_bh) { + /* + * Failure is not an option, but __GFP_NOFAIL is going + * away; so we retry ourselves here. + */ + congestion_wait(BLK_RW_ASYNC, HZ/50); + goto retry_alloc; + } + /* keep subsequent assertions sane */ new_bh->b_state = 0; init_buffer(new_bh, NULL, NULL); @@ -442,7 +451,7 @@ int __jbd2_log_space_left(journal_t *journal) { int left = journal->j_free; - assert_spin_locked(&journal->j_state_lock); + /* assert_spin_locked(&journal->j_state_lock); */ /* * Be pessimistic here about the number of those free blocks which @@ -487,9 +496,9 @@ int jbd2_log_start_commit(journal_t *journal, tid_t tid) { int ret; - spin_lock(&journal->j_state_lock); + write_lock(&journal->j_state_lock); ret = __jbd2_log_start_commit(journal, tid); - spin_unlock(&journal->j_state_lock); + write_unlock(&journal->j_state_lock); return ret; } @@ -508,7 +517,7 @@ int jbd2_journal_force_commit_nested(journal_t *journal) transaction_t *transaction = NULL; tid_t tid; - spin_lock(&journal->j_state_lock); + read_lock(&journal->j_state_lock); if (journal->j_running_transaction && !current->journal_info) { transaction = journal->j_running_transaction; __jbd2_log_start_commit(journal, transaction->t_tid); @@ -516,12 +525,12 @@ int jbd2_journal_force_commit_nested(journal_t *journal) transaction = journal->j_committing_transaction; if (!transaction) { - spin_unlock(&journal->j_state_lock); + read_unlock(&journal->j_state_lock); return 0; /* Nothing to retry */ } tid = transaction->t_tid; - spin_unlock(&journal->j_state_lock); + read_unlock(&journal->j_state_lock); jbd2_log_wait_commit(journal, tid); return 1; } @@ -535,7 +544,7 @@ int jbd2_journal_start_commit(journal_t *journal, tid_t *ptid) { int ret = 0; - spin_lock(&journal->j_state_lock); + write_lock(&journal->j_state_lock); if (journal->j_running_transaction) { tid_t tid = journal->j_running_transaction->t_tid; @@ -554,7 +563,7 @@ int jbd2_journal_start_commit(journal_t *journal, tid_t *ptid) *ptid = journal->j_committing_transaction->t_tid; ret = 1; } - spin_unlock(&journal->j_state_lock); + write_unlock(&journal->j_state_lock); return ret; } @@ -566,26 +575,24 @@ int jbd2_log_wait_commit(journal_t *journal, tid_t tid) { int err = 0; + read_lock(&journal->j_state_lock); #ifdef CONFIG_JBD2_DEBUG - spin_lock(&journal->j_state_lock); if (!tid_geq(journal->j_commit_request, tid)) { printk(KERN_EMERG "%s: error: j_commit_request=%d, tid=%d\n", __func__, journal->j_commit_request, tid); } - spin_unlock(&journal->j_state_lock); #endif - spin_lock(&journal->j_state_lock); while (tid_gt(tid, journal->j_commit_sequence)) { jbd_debug(1, "JBD: want %d, j_commit_sequence=%d\n", tid, journal->j_commit_sequence); wake_up(&journal->j_wait_commit); - spin_unlock(&journal->j_state_lock); + read_unlock(&journal->j_state_lock); wait_event(journal->j_wait_done_commit, !tid_gt(tid, journal->j_commit_sequence)); - spin_lock(&journal->j_state_lock); + read_lock(&journal->j_state_lock); } - spin_unlock(&journal->j_state_lock); + read_unlock(&journal->j_state_lock); if (unlikely(is_journal_aborted(journal))) { printk(KERN_EMERG "journal commit I/O error\n"); @@ -602,7 +609,7 @@ int jbd2_journal_next_log_block(journal_t *journal, unsigned long long *retp) { unsigned long blocknr; - spin_lock(&journal->j_state_lock); + write_lock(&journal->j_state_lock); J_ASSERT(journal->j_free > 1); blocknr = journal->j_head; @@ -610,7 +617,7 @@ int jbd2_journal_next_log_block(journal_t *journal, unsigned long long *retp) journal->j_free--; if (journal->j_head == journal->j_last) journal->j_head = journal->j_first; - spin_unlock(&journal->j_state_lock); + write_unlock(&journal->j_state_lock); return jbd2_journal_bmap(journal, blocknr, retp); } @@ -830,7 +837,7 @@ static journal_t * journal_init_common (void) mutex_init(&journal->j_checkpoint_mutex); spin_lock_init(&journal->j_revoke_lock); spin_lock_init(&journal->j_list_lock); - spin_lock_init(&journal->j_state_lock); + rwlock_init(&journal->j_state_lock); journal->j_commit_interval = (HZ * JBD2_DEFAULT_MAX_COMMIT_AGE); journal->j_min_batch_time = 0; @@ -1096,14 +1103,14 @@ void jbd2_journal_update_superblock(journal_t *journal, int wait) set_buffer_uptodate(bh); } - spin_lock(&journal->j_state_lock); + read_lock(&journal->j_state_lock); jbd_debug(1,"JBD: updating superblock (start %ld, seq %d, errno %d)\n", journal->j_tail, journal->j_tail_sequence, journal->j_errno); sb->s_sequence = cpu_to_be32(journal->j_tail_sequence); sb->s_start = cpu_to_be32(journal->j_tail); sb->s_errno = cpu_to_be32(journal->j_errno); - spin_unlock(&journal->j_state_lock); + read_unlock(&journal->j_state_lock); BUFFER_TRACE(bh, "marking dirty"); mark_buffer_dirty(bh); @@ -1124,12 +1131,12 @@ out: * any future commit will have to be careful to update the * superblock again to re-record the true start of the log. */ - spin_lock(&journal->j_state_lock); + write_lock(&journal->j_state_lock); if (sb->s_start) journal->j_flags &= ~JBD2_FLUSHED; else journal->j_flags |= JBD2_FLUSHED; - spin_unlock(&journal->j_state_lock); + write_unlock(&journal->j_state_lock); } /* @@ -1391,13 +1398,9 @@ int jbd2_journal_check_used_features (journal_t *journal, unsigned long compat, int jbd2_journal_check_available_features (journal_t *journal, unsigned long compat, unsigned long ro, unsigned long incompat) { - journal_superblock_t *sb; - if (!compat && !ro && !incompat) return 1; - sb = journal->j_superblock; - /* We can support any known requested features iff the * superblock is in version 2. Otherwise we fail to support any * extended sb features. */ @@ -1545,7 +1548,7 @@ int jbd2_journal_flush(journal_t *journal) transaction_t *transaction = NULL; unsigned long old_tail; - spin_lock(&journal->j_state_lock); + write_lock(&journal->j_state_lock); /* Force everything buffered to the log... */ if (journal->j_running_transaction) { @@ -1558,10 +1561,10 @@ int jbd2_journal_flush(journal_t *journal) if (transaction) { tid_t tid = transaction->t_tid; - spin_unlock(&journal->j_state_lock); + write_unlock(&journal->j_state_lock); jbd2_log_wait_commit(journal, tid); } else { - spin_unlock(&journal->j_state_lock); + write_unlock(&journal->j_state_lock); } /* ...and flush everything in the log out to disk. */ @@ -1585,12 +1588,12 @@ int jbd2_journal_flush(journal_t *journal) * the magic code for a fully-recovered superblock. Any future * commits of data to the journal will restore the current * s_start value. */ - spin_lock(&journal->j_state_lock); + write_lock(&journal->j_state_lock); old_tail = journal->j_tail; journal->j_tail = 0; - spin_unlock(&journal->j_state_lock); + write_unlock(&journal->j_state_lock); jbd2_journal_update_superblock(journal, 1); - spin_lock(&journal->j_state_lock); + write_lock(&journal->j_state_lock); journal->j_tail = old_tail; J_ASSERT(!journal->j_running_transaction); @@ -1598,7 +1601,7 @@ int jbd2_journal_flush(journal_t *journal) J_ASSERT(!journal->j_checkpoint_transactions); J_ASSERT(journal->j_head == journal->j_tail); J_ASSERT(journal->j_tail_sequence == journal->j_transaction_sequence); - spin_unlock(&journal->j_state_lock); + write_unlock(&journal->j_state_lock); return 0; } @@ -1617,7 +1620,6 @@ int jbd2_journal_flush(journal_t *journal) int jbd2_journal_wipe(journal_t *journal, int write) { - journal_superblock_t *sb; int err = 0; J_ASSERT (!(journal->j_flags & JBD2_LOADED)); @@ -1626,8 +1628,6 @@ int jbd2_journal_wipe(journal_t *journal, int write) if (err) return err; - sb = journal->j_superblock; - if (!journal->j_tail) goto no_recovery; @@ -1665,12 +1665,12 @@ void __jbd2_journal_abort_hard(journal_t *journal) printk(KERN_ERR "Aborting journal on device %s.\n", journal->j_devname); - spin_lock(&journal->j_state_lock); + write_lock(&journal->j_state_lock); journal->j_flags |= JBD2_ABORT; transaction = journal->j_running_transaction; if (transaction) __jbd2_log_start_commit(journal, transaction->t_tid); - spin_unlock(&journal->j_state_lock); + write_unlock(&journal->j_state_lock); } /* Soft abort: record the abort error status in the journal superblock, @@ -1755,12 +1755,12 @@ int jbd2_journal_errno(journal_t *journal) { int err; - spin_lock(&journal->j_state_lock); + read_lock(&journal->j_state_lock); if (journal->j_flags & JBD2_ABORT) err = -EROFS; else err = journal->j_errno; - spin_unlock(&journal->j_state_lock); + read_unlock(&journal->j_state_lock); return err; } @@ -1775,12 +1775,12 @@ int jbd2_journal_clear_err(journal_t *journal) { int err = 0; - spin_lock(&journal->j_state_lock); + write_lock(&journal->j_state_lock); if (journal->j_flags & JBD2_ABORT) err = -EROFS; else journal->j_errno = 0; - spin_unlock(&journal->j_state_lock); + write_unlock(&journal->j_state_lock); return err; } @@ -1793,10 +1793,10 @@ int jbd2_journal_clear_err(journal_t *journal) */ void jbd2_journal_ack_err(journal_t *journal) { - spin_lock(&journal->j_state_lock); + write_lock(&journal->j_state_lock); if (journal->j_errno) journal->j_flags |= JBD2_ACK_ERR; - spin_unlock(&journal->j_state_lock); + write_unlock(&journal->j_state_lock); } int jbd2_journal_blocks_per_page(struct inode *inode) @@ -2201,8 +2201,6 @@ void jbd2_journal_init_jbd_inode(struct jbd2_inode *jinode, struct inode *inode) void jbd2_journal_release_jbd_inode(journal_t *journal, struct jbd2_inode *jinode) { - int writeout = 0; - if (!journal) return; restart: @@ -2219,9 +2217,6 @@ restart: goto restart; } - /* Do we need to wait for data writeback? */ - if (journal->j_committing_transaction == jinode->i_transaction) - writeout = 1; if (jinode->i_transaction) { list_del(&jinode->i_list); jinode->i_transaction = NULL; diff --git a/fs/jbd2/recovery.c b/fs/jbd2/recovery.c index 049281b7cb8..2bc4d5f116f 100644 --- a/fs/jbd2/recovery.c +++ b/fs/jbd2/recovery.c @@ -285,12 +285,10 @@ int jbd2_journal_recover(journal_t *journal) int jbd2_journal_skip_recovery(journal_t *journal) { int err; - journal_superblock_t * sb; struct recovery_info info; memset (&info, 0, sizeof(info)); - sb = journal->j_superblock; err = do_one_pass(journal, &info, PASS_SCAN); @@ -299,7 +297,8 @@ int jbd2_journal_skip_recovery(journal_t *journal) ++journal->j_transaction_sequence; } else { #ifdef CONFIG_JBD2_DEBUG - int dropped = info.end_transaction - be32_to_cpu(sb->s_sequence); + int dropped = info.end_transaction - + be32_to_cpu(journal->j_superblock->s_sequence); #endif jbd_debug(1, "JBD: ignoring %d transaction%s from the journal.\n", @@ -365,11 +364,6 @@ static int do_one_pass(journal_t *journal, int tag_bytes = journal_tag_bytes(journal); __u32 crc32_sum = ~0; /* Transactional Checksums */ - /* Precompute the maximum metadata descriptors in a descriptor block */ - int MAX_BLOCKS_PER_DESC; - MAX_BLOCKS_PER_DESC = ((journal->j_blocksize-sizeof(journal_header_t)) - / tag_bytes); - /* * First thing is to establish what we expect to find in the log * (in terms of transaction IDs), and where (in terms of log diff --git a/fs/jbd2/transaction.c b/fs/jbd2/transaction.c index b8e0806681b..d95cc9d0401 100644 --- a/fs/jbd2/transaction.c +++ b/fs/jbd2/transaction.c @@ -26,6 +26,8 @@ #include <linux/mm.h> #include <linux/highmem.h> #include <linux/hrtimer.h> +#include <linux/backing-dev.h> +#include <linux/module.h> static void __jbd2_journal_temp_unlink_buffer(struct journal_head *jh); @@ -53,6 +55,9 @@ jbd2_get_transaction(journal_t *journal, transaction_t *transaction) transaction->t_tid = journal->j_transaction_sequence++; transaction->t_expires = jiffies + journal->j_commit_interval; spin_lock_init(&transaction->t_handle_lock); + atomic_set(&transaction->t_updates, 0); + atomic_set(&transaction->t_outstanding_credits, 0); + atomic_set(&transaction->t_handle_count, 0); INIT_LIST_HEAD(&transaction->t_inode_list); INIT_LIST_HEAD(&transaction->t_private_list); @@ -83,65 +88,75 @@ jbd2_get_transaction(journal_t *journal, transaction_t *transaction) * transaction's buffer credits. */ -static int start_this_handle(journal_t *journal, handle_t *handle) +static int start_this_handle(journal_t *journal, handle_t *handle, + int gfp_mask) { transaction_t *transaction; int needed; int nblocks = handle->h_buffer_credits; transaction_t *new_transaction = NULL; - int ret = 0; unsigned long ts = jiffies; if (nblocks > journal->j_max_transaction_buffers) { printk(KERN_ERR "JBD: %s wants too many credits (%d > %d)\n", current->comm, nblocks, journal->j_max_transaction_buffers); - ret = -ENOSPC; - goto out; + return -ENOSPC; } alloc_transaction: if (!journal->j_running_transaction) { - new_transaction = kzalloc(sizeof(*new_transaction), - GFP_NOFS|__GFP_NOFAIL); + new_transaction = kzalloc(sizeof(*new_transaction), gfp_mask); if (!new_transaction) { - ret = -ENOMEM; - goto out; + /* + * If __GFP_FS is not present, then we may be + * being called from inside the fs writeback + * layer, so we MUST NOT fail. Since + * __GFP_NOFAIL is going away, we will arrange + * to retry the allocation ourselves. + */ + if ((gfp_mask & __GFP_FS) == 0) { + congestion_wait(BLK_RW_ASYNC, HZ/50); + goto alloc_transaction; + } + return -ENOMEM; } } jbd_debug(3, "New handle %p going live.\n", handle); -repeat: - /* * We need to hold j_state_lock until t_updates has been incremented, * for proper journal barrier handling */ - spin_lock(&journal->j_state_lock); -repeat_locked: +repeat: + read_lock(&journal->j_state_lock); if (is_journal_aborted(journal) || (journal->j_errno != 0 && !(journal->j_flags & JBD2_ACK_ERR))) { - spin_unlock(&journal->j_state_lock); - ret = -EROFS; - goto out; + read_unlock(&journal->j_state_lock); + kfree(new_transaction); + return -EROFS; } /* Wait on the journal's transaction barrier if necessary */ if (journal->j_barrier_count) { - spin_unlock(&journal->j_state_lock); + read_unlock(&journal->j_state_lock); wait_event(journal->j_wait_transaction_locked, journal->j_barrier_count == 0); goto repeat; } if (!journal->j_running_transaction) { - if (!new_transaction) { - spin_unlock(&journal->j_state_lock); + read_unlock(&journal->j_state_lock); + if (!new_transaction) goto alloc_transaction; + write_lock(&journal->j_state_lock); + if (!journal->j_running_transaction) { + jbd2_get_transaction(journal, new_transaction); + new_transaction = NULL; } - jbd2_get_transaction(journal, new_transaction); - new_transaction = NULL; + write_unlock(&journal->j_state_lock); + goto repeat; } transaction = journal->j_running_transaction; @@ -155,7 +170,7 @@ repeat_locked: prepare_to_wait(&journal->j_wait_transaction_locked, &wait, TASK_UNINTERRUPTIBLE); - spin_unlock(&journal->j_state_lock); + read_unlock(&journal->j_state_lock); schedule(); finish_wait(&journal->j_wait_transaction_locked, &wait); goto repeat; @@ -166,8 +181,8 @@ repeat_locked: * buffers requested by this operation, we need to stall pending a log * checkpoint to free some more log space. */ - spin_lock(&transaction->t_handle_lock); - needed = transaction->t_outstanding_credits + nblocks; + needed = atomic_add_return(nblocks, + &transaction->t_outstanding_credits); if (needed > journal->j_max_transaction_buffers) { /* @@ -178,11 +193,11 @@ repeat_locked: DEFINE_WAIT(wait); jbd_debug(2, "Handle %p starting new commit...\n", handle); - spin_unlock(&transaction->t_handle_lock); + atomic_sub(nblocks, &transaction->t_outstanding_credits); prepare_to_wait(&journal->j_wait_transaction_locked, &wait, TASK_UNINTERRUPTIBLE); __jbd2_log_start_commit(journal, transaction->t_tid); - spin_unlock(&journal->j_state_lock); + read_unlock(&journal->j_state_lock); schedule(); finish_wait(&journal->j_wait_transaction_locked, &wait); goto repeat; @@ -215,35 +230,48 @@ repeat_locked: */ if (__jbd2_log_space_left(journal) < jbd_space_needed(journal)) { jbd_debug(2, "Handle %p waiting for checkpoint...\n", handle); - spin_unlock(&transaction->t_handle_lock); - __jbd2_log_wait_for_space(journal); - goto repeat_locked; + atomic_sub(nblocks, &transaction->t_outstanding_credits); + read_unlock(&journal->j_state_lock); + write_lock(&journal->j_state_lock); + if (__jbd2_log_space_left(journal) < jbd_space_needed(journal)) + __jbd2_log_wait_for_space(journal); + write_unlock(&journal->j_state_lock); + goto repeat; } /* OK, account for the buffers that this operation expects to - * use and add the handle to the running transaction. */ - - if (time_after(transaction->t_start, ts)) { + * use and add the handle to the running transaction. + * + * In order for t_max_wait to be reliable, it must be + * protected by a lock. But doing so will mean that + * start_this_handle() can not be run in parallel on SMP + * systems, which limits our scalability. So we only enable + * it when debugging is enabled. We may want to use a + * separate flag, eventually, so we can enable this + * independently of debugging. + */ +#ifdef CONFIG_JBD2_DEBUG + if (jbd2_journal_enable_debug && + time_after(transaction->t_start, ts)) { ts = jbd2_time_diff(ts, transaction->t_start); + spin_lock(&transaction->t_handle_lock); if (ts > transaction->t_max_wait) transaction->t_max_wait = ts; + spin_unlock(&transaction->t_handle_lock); } - +#endif handle->h_transaction = transaction; - transaction->t_outstanding_credits += nblocks; - transaction->t_updates++; - transaction->t_handle_count++; + atomic_inc(&transaction->t_updates); + atomic_inc(&transaction->t_handle_count); jbd_debug(4, "Handle %p given %d credits (total %d, free %d)\n", - handle, nblocks, transaction->t_outstanding_credits, + handle, nblocks, + atomic_read(&transaction->t_outstanding_credits), __jbd2_log_space_left(journal)); - spin_unlock(&transaction->t_handle_lock); - spin_unlock(&journal->j_state_lock); + read_unlock(&journal->j_state_lock); lock_map_acquire(&handle->h_lockdep_map); -out: - if (unlikely(new_transaction)) /* It's usually NULL */ - kfree(new_transaction); - return ret; + kfree(new_transaction); + return 0; } static struct lock_class_key jbd2_handle_key; @@ -278,7 +306,7 @@ static handle_t *new_handle(int nblocks) * * Return a pointer to a newly allocated handle, or NULL on failure */ -handle_t *jbd2_journal_start(journal_t *journal, int nblocks) +handle_t *jbd2__journal_start(journal_t *journal, int nblocks, int gfp_mask) { handle_t *handle = journal_current_handle(); int err; @@ -298,7 +326,7 @@ handle_t *jbd2_journal_start(journal_t *journal, int nblocks) current->journal_info = handle; - err = start_this_handle(journal, handle); + err = start_this_handle(journal, handle, gfp_mask); if (err < 0) { jbd2_free_handle(handle); current->journal_info = NULL; @@ -308,6 +336,15 @@ handle_t *jbd2_journal_start(journal_t *journal, int nblocks) out: return handle; } +EXPORT_SYMBOL(jbd2__journal_start); + + +handle_t *jbd2_journal_start(journal_t *journal, int nblocks) +{ + return jbd2__journal_start(journal, nblocks, GFP_NOFS); +} +EXPORT_SYMBOL(jbd2_journal_start); + /** * int jbd2_journal_extend() - extend buffer credits. @@ -342,7 +379,7 @@ int jbd2_journal_extend(handle_t *handle, int nblocks) result = 1; - spin_lock(&journal->j_state_lock); + read_lock(&journal->j_state_lock); /* Don't extend a locked-down transaction! */ if (handle->h_transaction->t_state != T_RUNNING) { @@ -352,7 +389,7 @@ int jbd2_journal_extend(handle_t *handle, int nblocks) } spin_lock(&transaction->t_handle_lock); - wanted = transaction->t_outstanding_credits + nblocks; + wanted = atomic_read(&transaction->t_outstanding_credits) + nblocks; if (wanted > journal->j_max_transaction_buffers) { jbd_debug(3, "denied handle %p %d blocks: " @@ -367,14 +404,14 @@ int jbd2_journal_extend(handle_t *handle, int nblocks) } handle->h_buffer_credits += nblocks; - transaction->t_outstanding_credits += nblocks; + atomic_add(nblocks, &transaction->t_outstanding_credits); result = 0; jbd_debug(3, "extended handle %p by %d\n", handle, nblocks); unlock: spin_unlock(&transaction->t_handle_lock); error_out: - spin_unlock(&journal->j_state_lock); + read_unlock(&journal->j_state_lock); out: return result; } @@ -394,8 +431,7 @@ out: * transaction capabable of guaranteeing the requested number of * credits. */ - -int jbd2_journal_restart(handle_t *handle, int nblocks) +int jbd2__journal_restart(handle_t *handle, int nblocks, int gfp_mask) { transaction_t *transaction = handle->h_transaction; journal_t *journal = transaction->t_journal; @@ -410,29 +446,35 @@ int jbd2_journal_restart(handle_t *handle, int nblocks) * First unlink the handle from its current transaction, and start the * commit on that. */ - J_ASSERT(transaction->t_updates > 0); + J_ASSERT(atomic_read(&transaction->t_updates) > 0); J_ASSERT(journal_current_handle() == handle); - spin_lock(&journal->j_state_lock); + read_lock(&journal->j_state_lock); spin_lock(&transaction->t_handle_lock); - transaction->t_outstanding_credits -= handle->h_buffer_credits; - transaction->t_updates--; - - if (!transaction->t_updates) + atomic_sub(handle->h_buffer_credits, + &transaction->t_outstanding_credits); + if (atomic_dec_and_test(&transaction->t_updates)) wake_up(&journal->j_wait_updates); spin_unlock(&transaction->t_handle_lock); jbd_debug(2, "restarting handle %p\n", handle); __jbd2_log_start_commit(journal, transaction->t_tid); - spin_unlock(&journal->j_state_lock); + read_unlock(&journal->j_state_lock); lock_map_release(&handle->h_lockdep_map); handle->h_buffer_credits = nblocks; - ret = start_this_handle(journal, handle); + ret = start_this_handle(journal, handle, gfp_mask); return ret; } +EXPORT_SYMBOL(jbd2__journal_restart); +int jbd2_journal_restart(handle_t *handle, int nblocks) +{ + return jbd2__journal_restart(handle, nblocks, GFP_NOFS); +} +EXPORT_SYMBOL(jbd2_journal_restart); + /** * void jbd2_journal_lock_updates () - establish a transaction barrier. * @journal: Journal to establish a barrier on. @@ -447,7 +489,7 @@ void jbd2_journal_lock_updates(journal_t *journal) { DEFINE_WAIT(wait); - spin_lock(&journal->j_state_lock); + write_lock(&journal->j_state_lock); ++journal->j_barrier_count; /* Wait until there are no running updates */ @@ -458,19 +500,19 @@ void jbd2_journal_lock_updates(journal_t *journal) break; spin_lock(&transaction->t_handle_lock); - if (!transaction->t_updates) { + if (!atomic_read(&transaction->t_updates)) { spin_unlock(&transaction->t_handle_lock); break; } prepare_to_wait(&journal->j_wait_updates, &wait, TASK_UNINTERRUPTIBLE); spin_unlock(&transaction->t_handle_lock); - spin_unlock(&journal->j_state_lock); + write_unlock(&journal->j_state_lock); schedule(); finish_wait(&journal->j_wait_updates, &wait); - spin_lock(&journal->j_state_lock); + write_lock(&journal->j_state_lock); } - spin_unlock(&journal->j_state_lock); + write_unlock(&journal->j_state_lock); /* * We have now established a barrier against other normal updates, but @@ -494,9 +536,9 @@ void jbd2_journal_unlock_updates (journal_t *journal) J_ASSERT(journal->j_barrier_count != 0); mutex_unlock(&journal->j_barrier); - spin_lock(&journal->j_state_lock); + write_lock(&journal->j_state_lock); --journal->j_barrier_count; - spin_unlock(&journal->j_state_lock); + write_unlock(&journal->j_state_lock); wake_up(&journal->j_wait_transaction_locked); } @@ -1238,7 +1280,8 @@ int jbd2_journal_stop(handle_t *handle) { transaction_t *transaction = handle->h_transaction; journal_t *journal = transaction->t_journal; - int err; + int err, wait_for_commit = 0; + tid_t tid; pid_t pid; J_ASSERT(journal_current_handle() == handle); @@ -1246,7 +1289,7 @@ int jbd2_journal_stop(handle_t *handle) if (is_handle_aborted(handle)) err = -EIO; else { - J_ASSERT(transaction->t_updates > 0); + J_ASSERT(atomic_read(&transaction->t_updates) > 0); err = 0; } @@ -1291,9 +1334,9 @@ int jbd2_journal_stop(handle_t *handle) journal->j_last_sync_writer = pid; - spin_lock(&journal->j_state_lock); + read_lock(&journal->j_state_lock); commit_time = journal->j_average_commit_time; - spin_unlock(&journal->j_state_lock); + read_unlock(&journal->j_state_lock); trans_time = ktime_to_ns(ktime_sub(ktime_get(), transaction->t_start_time)); @@ -1314,14 +1357,8 @@ int jbd2_journal_stop(handle_t *handle) if (handle->h_sync) transaction->t_synchronous_commit = 1; current->journal_info = NULL; - spin_lock(&transaction->t_handle_lock); - transaction->t_outstanding_credits -= handle->h_buffer_credits; - transaction->t_updates--; - if (!transaction->t_updates) { - wake_up(&journal->j_wait_updates); - if (journal->j_barrier_count) - wake_up(&journal->j_wait_transaction_locked); - } + atomic_sub(handle->h_buffer_credits, + &transaction->t_outstanding_credits); /* * If the handle is marked SYNC, we need to set another commit @@ -1330,15 +1367,13 @@ int jbd2_journal_stop(handle_t *handle) * transaction is too old now. */ if (handle->h_sync || - transaction->t_outstanding_credits > - journal->j_max_transaction_buffers || - time_after_eq(jiffies, transaction->t_expires)) { + (atomic_read(&transaction->t_outstanding_credits) > + journal->j_max_transaction_buffers) || + time_after_eq(jiffies, transaction->t_expires)) { /* Do this even for aborted journals: an abort still * completes the commit thread, it just doesn't write * anything to disk. */ - tid_t tid = transaction->t_tid; - spin_unlock(&transaction->t_handle_lock); jbd_debug(2, "transaction too old, requesting commit for " "handle %p\n", handle); /* This is non-blocking */ @@ -1349,11 +1384,25 @@ int jbd2_journal_stop(handle_t *handle) * to wait for the commit to complete. */ if (handle->h_sync && !(current->flags & PF_MEMALLOC)) - err = jbd2_log_wait_commit(journal, tid); - } else { - spin_unlock(&transaction->t_handle_lock); + wait_for_commit = 1; } + /* + * Once we drop t_updates, if it goes to zero the transaction + * could start commiting on us and eventually disappear. So + * once we do this, we must not dereference transaction + * pointer again. + */ + tid = transaction->t_tid; + if (atomic_dec_and_test(&transaction->t_updates)) { + wake_up(&journal->j_wait_updates); + if (journal->j_barrier_count) + wake_up(&journal->j_wait_transaction_locked); + } + + if (wait_for_commit) + err = jbd2_log_wait_commit(journal, tid); + lock_map_release(&handle->h_lockdep_map); jbd2_free_handle(handle); @@ -1719,7 +1768,7 @@ static int journal_unmap_buffer(journal_t *journal, struct buffer_head *bh) goto zap_buffer_unlocked; /* OK, we have data buffer in journaled mode */ - spin_lock(&journal->j_state_lock); + write_lock(&journal->j_state_lock); jbd_lock_bh_state(bh); spin_lock(&journal->j_list_lock); @@ -1772,7 +1821,7 @@ static int journal_unmap_buffer(journal_t *journal, struct buffer_head *bh) jbd2_journal_put_journal_head(jh); spin_unlock(&journal->j_list_lock); jbd_unlock_bh_state(bh); - spin_unlock(&journal->j_state_lock); + write_unlock(&journal->j_state_lock); return ret; } else { /* There is no currently-running transaction. So the @@ -1786,7 +1835,7 @@ static int journal_unmap_buffer(journal_t *journal, struct buffer_head *bh) jbd2_journal_put_journal_head(jh); spin_unlock(&journal->j_list_lock); jbd_unlock_bh_state(bh); - spin_unlock(&journal->j_state_lock); + write_unlock(&journal->j_state_lock); return ret; } else { /* The orphan record's transaction has @@ -1810,7 +1859,7 @@ static int journal_unmap_buffer(journal_t *journal, struct buffer_head *bh) jbd2_journal_put_journal_head(jh); spin_unlock(&journal->j_list_lock); jbd_unlock_bh_state(bh); - spin_unlock(&journal->j_state_lock); + write_unlock(&journal->j_state_lock); return 0; } else { /* Good, the buffer belongs to the running transaction. @@ -1829,7 +1878,7 @@ zap_buffer: zap_buffer_no_jh: spin_unlock(&journal->j_list_lock); jbd_unlock_bh_state(bh); - spin_unlock(&journal->j_state_lock); + write_unlock(&journal->j_state_lock); zap_buffer_unlocked: clear_buffer_dirty(bh); J_ASSERT_BH(bh, !buffer_jbddirty(bh)); @@ -2136,9 +2185,9 @@ int jbd2_journal_begin_ordered_truncate(journal_t *journal, /* Locks are here just to force reading of recent values, it is * enough that the transaction was not committing before we started * a transaction adding the inode to orphan list */ - spin_lock(&journal->j_state_lock); + read_lock(&journal->j_state_lock); commit_trans = journal->j_committing_transaction; - spin_unlock(&journal->j_state_lock); + read_unlock(&journal->j_state_lock); spin_lock(&journal->j_list_lock); inode_trans = jinode->i_transaction; spin_unlock(&journal->j_list_lock); diff --git a/fs/namei.c b/fs/namei.c index 868d0cb9d47..42d2d28fb82 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -282,8 +282,7 @@ int inode_permission(struct inode *inode, int mask) if (retval) return retval; - return security_inode_permission(inode, - mask & (MAY_READ|MAY_WRITE|MAY_EXEC|MAY_APPEND)); + return security_inode_permission(inode, mask); } /** @@ -1484,8 +1483,7 @@ static int handle_truncate(struct path *path) */ error = locks_verify_locked(inode); if (!error) - error = security_path_truncate(path, 0, - ATTR_MTIME|ATTR_CTIME|ATTR_OPEN); + error = security_path_truncate(path); if (!error) { error = do_truncate(path->dentry, 0, ATTR_MTIME|ATTR_CTIME|ATTR_OPEN, diff --git a/fs/ncpfs/inode.c b/fs/ncpfs/inode.c index fa338515402..1e634deff94 100644 --- a/fs/ncpfs/inode.c +++ b/fs/ncpfs/inode.c @@ -728,8 +728,8 @@ out_fput: out_bdi: /* 23/12/1998 Marcin Dalecki <dalecki@cs.net.pl>: * - * The previously used put_filp(ncp_filp); was bogous, since - * it doesn't proper unlocking. + * The previously used put_filp(ncp_filp); was bogus, since + * it doesn't perform proper unlocking. */ fput(ncp_filp); out: diff --git a/fs/nfs/Kconfig b/fs/nfs/Kconfig index a43d07e7b92..cc1bb33b59b 100644 --- a/fs/nfs/Kconfig +++ b/fs/nfs/Kconfig @@ -61,8 +61,8 @@ config NFS_V3_ACL If unsure, say N. config NFS_V4 - bool "NFS client support for NFS version 4 (EXPERIMENTAL)" - depends on NFS_FS && EXPERIMENTAL + bool "NFS client support for NFS version 4" + depends on NFS_FS select RPCSEC_GSS_KRB5 help This option enables support for version 4 of the NFS protocol @@ -72,16 +72,16 @@ config NFS_V4 space programs which can be found in the Linux nfs-utils package, available from http://linux-nfs.org/. - If unsure, say N. + If unsure, say Y. config NFS_V4_1 - bool "NFS client support for NFSv4.1 (DEVELOPER ONLY)" + bool "NFS client support for NFSv4.1 (EXPERIMENTAL)" depends on NFS_V4 && EXPERIMENTAL help This option enables support for minor version 1 of the NFSv4 protocol (draft-ietf-nfsv4-minorversion1) in the kernel's NFS client. - Unless you're an NFS developer, say N. + If unsure, say N. config ROOT_NFS bool "Root file system on NFS" diff --git a/fs/nfs/callback_proc.c b/fs/nfs/callback_proc.c index a08770a7e85..930d10fecda 100644 --- a/fs/nfs/callback_proc.c +++ b/fs/nfs/callback_proc.c @@ -37,8 +37,8 @@ __be32 nfs4_callback_getattr(struct cb_getattrargs *args, struct cb_getattrres * if (inode == NULL) goto out_putclient; nfsi = NFS_I(inode); - down_read(&nfsi->rwsem); - delegation = nfsi->delegation; + rcu_read_lock(); + delegation = rcu_dereference(nfsi->delegation); if (delegation == NULL || (delegation->type & FMODE_WRITE) == 0) goto out_iput; res->size = i_size_read(inode); @@ -53,7 +53,7 @@ __be32 nfs4_callback_getattr(struct cb_getattrargs *args, struct cb_getattrres * args->bitmap[1]; res->status = 0; out_iput: - up_read(&nfsi->rwsem); + rcu_read_unlock(); iput(inode); out_putclient: nfs_put_client(clp); @@ -62,16 +62,6 @@ out: return res->status; } -static int (*nfs_validate_delegation_stateid(struct nfs_client *clp))(struct nfs_delegation *, const nfs4_stateid *) -{ -#if defined(CONFIG_NFS_V4_1) - if (clp->cl_minorversion > 0) - return nfs41_validate_delegation_stateid; -#endif - return nfs4_validate_delegation_stateid; -} - - __be32 nfs4_callback_recall(struct cb_recallargs *args, void *dummy) { struct nfs_client *clp; @@ -92,8 +82,7 @@ __be32 nfs4_callback_recall(struct cb_recallargs *args, void *dummy) inode = nfs_delegation_find_inode(clp, &args->fh); if (inode != NULL) { /* Set up a helper thread to actually return the delegation */ - switch (nfs_async_inode_return_delegation(inode, &args->stateid, - nfs_validate_delegation_stateid(clp))) { + switch (nfs_async_inode_return_delegation(inode, &args->stateid)) { case 0: res = 0; break; diff --git a/fs/nfs/client.c b/fs/nfs/client.c index d25b5257b7a..4e7df2adb21 100644 --- a/fs/nfs/client.c +++ b/fs/nfs/client.c @@ -150,6 +150,7 @@ static struct nfs_client *nfs_alloc_client(const struct nfs_client_initdata *cl_ clp->cl_boot_time = CURRENT_TIME; clp->cl_state = 1 << NFS4CLNT_LEASE_EXPIRED; clp->cl_minorversion = cl_init->minorversion; + clp->cl_mvops = nfs_v4_minor_ops[cl_init->minorversion]; #endif cred = rpc_lookup_machine_cred(); if (!IS_ERR(cred)) @@ -178,7 +179,7 @@ static void nfs4_clear_client_minor_version(struct nfs_client *clp) clp->cl_session = NULL; } - clp->cl_call_sync = _nfs4_call_sync; + clp->cl_mvops = nfs_v4_minor_ops[0]; #endif /* CONFIG_NFS_V4_1 */ } @@ -188,7 +189,7 @@ static void nfs4_clear_client_minor_version(struct nfs_client *clp) static void nfs4_destroy_callback(struct nfs_client *clp) { if (__test_and_clear_bit(NFS_CS_CALLBACK, &clp->cl_res_state)) - nfs_callback_down(clp->cl_minorversion); + nfs_callback_down(clp->cl_mvops->minor_version); } static void nfs4_shutdown_client(struct nfs_client *clp) @@ -1126,7 +1127,7 @@ static int nfs4_init_callback(struct nfs_client *clp) return error; } - error = nfs_callback_up(clp->cl_minorversion, + error = nfs_callback_up(clp->cl_mvops->minor_version, clp->cl_rpcclient->cl_xprt); if (error < 0) { dprintk("%s: failed to start callback. Error = %d\n", @@ -1143,10 +1144,8 @@ static int nfs4_init_callback(struct nfs_client *clp) */ static int nfs4_init_client_minor_version(struct nfs_client *clp) { - clp->cl_call_sync = _nfs4_call_sync; - #if defined(CONFIG_NFS_V4_1) - if (clp->cl_minorversion) { + if (clp->cl_mvops->minor_version) { struct nfs4_session *session = NULL; /* * Create the session and mark it expired. @@ -1158,7 +1157,13 @@ static int nfs4_init_client_minor_version(struct nfs_client *clp) return -ENOMEM; clp->cl_session = session; - clp->cl_call_sync = _nfs4_call_sync_session; + /* + * The create session reply races with the server back + * channel probe. Mark the client NFS_CS_SESSION_INITING + * so that the client back channel can find the + * nfs_client struct + */ + clp->cl_cons_state = NFS_CS_SESSION_INITING; } #endif /* CONFIG_NFS_V4_1 */ @@ -1454,7 +1459,7 @@ struct nfs_server *nfs4_create_referral_server(struct nfs_clone_mount *data, data->authflavor, parent_server->client->cl_xprt->prot, parent_server->client->cl_timeout, - parent_client->cl_minorversion); + parent_client->cl_mvops->minor_version); if (error < 0) goto error; diff --git a/fs/nfs/delegation.c b/fs/nfs/delegation.c index 30163454397..b9c3c43cea1 100644 --- a/fs/nfs/delegation.c +++ b/fs/nfs/delegation.c @@ -268,14 +268,6 @@ out: return status; } -/* Sync all data to disk upon delegation return */ -static void nfs_msync_inode(struct inode *inode) -{ - filemap_fdatawrite(inode->i_mapping); - nfs_wb_all(inode); - filemap_fdatawait(inode->i_mapping); -} - /* * Basic procedure for returning a delegation to the server */ @@ -367,7 +359,7 @@ int nfs_inode_return_delegation(struct inode *inode) delegation = nfs_detach_delegation_locked(nfsi, NULL, clp); spin_unlock(&clp->cl_lock); if (delegation != NULL) { - nfs_msync_inode(inode); + nfs_wb_all(inode); err = __nfs_inode_return_delegation(inode, delegation, 1); } } @@ -471,9 +463,7 @@ void nfs_expire_unreferenced_delegations(struct nfs_client *clp) /* * Asynchronous delegation recall! */ -int nfs_async_inode_return_delegation(struct inode *inode, const nfs4_stateid *stateid, - int (*validate_stateid)(struct nfs_delegation *delegation, - const nfs4_stateid *stateid)) +int nfs_async_inode_return_delegation(struct inode *inode, const nfs4_stateid *stateid) { struct nfs_client *clp = NFS_SERVER(inode)->nfs_client; struct nfs_delegation *delegation; @@ -481,7 +471,7 @@ int nfs_async_inode_return_delegation(struct inode *inode, const nfs4_stateid *s rcu_read_lock(); delegation = rcu_dereference(NFS_I(inode)->delegation); - if (!validate_stateid(delegation, stateid)) { + if (!clp->cl_mvops->validate_stateid(delegation, stateid)) { rcu_read_unlock(); return -ENOENT; } diff --git a/fs/nfs/delegation.h b/fs/nfs/delegation.h index 69e7b814012..2026304bda1 100644 --- a/fs/nfs/delegation.h +++ b/fs/nfs/delegation.h @@ -34,9 +34,7 @@ enum { 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); -int nfs_async_inode_return_delegation(struct inode *inode, const nfs4_stateid *stateid, - int (*validate_stateid)(struct nfs_delegation *delegation, - const nfs4_stateid *stateid)); +int nfs_async_inode_return_delegation(struct inode *inode, const nfs4_stateid *stateid); void nfs_inode_return_delegation_noreclaim(struct inode *inode); struct inode *nfs_delegation_find_inode(struct nfs_client *clp, const struct nfs_fh *fhandle); diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c index e60416d3f81..29539ceeb74 100644 --- a/fs/nfs/dir.c +++ b/fs/nfs/dir.c @@ -1652,16 +1652,7 @@ static int nfs_rename(struct inode *old_dir, struct dentry *old_dentry, } } - /* - * ... prune child dentries and writebacks if needed. - */ - if (atomic_read(&old_dentry->d_count) > 1) { - if (S_ISREG(old_inode->i_mode)) - nfs_wb_all(old_inode); - shrink_dcache_parent(old_dentry); - } nfs_inode_return_delegation(old_inode); - if (new_inode != NULL) nfs_inode_return_delegation(new_inode); @@ -1953,7 +1944,7 @@ int nfs_permission(struct inode *inode, int mask) if ((mask & (MAY_READ | MAY_WRITE | MAY_EXEC)) == 0) goto out; /* Is this sys_access() ? */ - if (mask & MAY_ACCESS) + if (mask & (MAY_ACCESS | MAY_CHDIR)) goto force_lookup; switch (inode->i_mode & S_IFMT) { diff --git a/fs/nfs/direct.c b/fs/nfs/direct.c index ad4cd31d605..064a8096167 100644 --- a/fs/nfs/direct.c +++ b/fs/nfs/direct.c @@ -69,6 +69,7 @@ struct nfs_direct_req { /* I/O parameters */ struct nfs_open_context *ctx; /* file open context info */ + struct nfs_lock_context *l_ctx; /* Lock context info */ struct kiocb * iocb; /* controlling i/o request */ struct inode * inode; /* target file of i/o */ @@ -160,6 +161,7 @@ static inline struct nfs_direct_req *nfs_direct_req_alloc(void) INIT_LIST_HEAD(&dreq->rewrite_list); dreq->iocb = NULL; dreq->ctx = NULL; + dreq->l_ctx = NULL; spin_lock_init(&dreq->lock); atomic_set(&dreq->io_count, 0); dreq->count = 0; @@ -173,6 +175,8 @@ static void nfs_direct_req_free(struct kref *kref) { struct nfs_direct_req *dreq = container_of(kref, struct nfs_direct_req, kref); + if (dreq->l_ctx != NULL) + nfs_put_lock_context(dreq->l_ctx); if (dreq->ctx != NULL) put_nfs_open_context(dreq->ctx); kmem_cache_free(nfs_direct_cachep, dreq); @@ -336,6 +340,7 @@ static ssize_t nfs_direct_read_schedule_segment(struct nfs_direct_req *dreq, data->cred = msg.rpc_cred; data->args.fh = NFS_FH(inode); data->args.context = ctx; + data->args.lock_context = dreq->l_ctx; data->args.offset = pos; data->args.pgbase = pgbase; data->args.pages = data->pagevec; @@ -416,24 +421,28 @@ static ssize_t nfs_direct_read_schedule_iovec(struct nfs_direct_req *dreq, static ssize_t nfs_direct_read(struct kiocb *iocb, const struct iovec *iov, unsigned long nr_segs, loff_t pos) { - ssize_t result = 0; + ssize_t result = -ENOMEM; struct inode *inode = iocb->ki_filp->f_mapping->host; struct nfs_direct_req *dreq; dreq = nfs_direct_req_alloc(); - if (!dreq) - return -ENOMEM; + if (dreq == NULL) + goto out; dreq->inode = inode; dreq->ctx = get_nfs_open_context(nfs_file_open_context(iocb->ki_filp)); + dreq->l_ctx = nfs_get_lock_context(dreq->ctx); + if (dreq->l_ctx == NULL) + goto out_release; if (!is_sync_kiocb(iocb)) dreq->iocb = iocb; result = nfs_direct_read_schedule_iovec(dreq, iov, nr_segs, pos); if (!result) result = nfs_direct_wait(dreq); +out_release: nfs_direct_req_release(dreq); - +out: return result; } @@ -574,6 +583,7 @@ static void nfs_direct_commit_schedule(struct nfs_direct_req *dreq) data->args.offset = 0; data->args.count = 0; data->args.context = dreq->ctx; + data->args.lock_context = dreq->l_ctx; data->res.count = 0; data->res.fattr = &data->fattr; data->res.verf = &data->verf; @@ -761,6 +771,7 @@ static ssize_t nfs_direct_write_schedule_segment(struct nfs_direct_req *dreq, data->cred = msg.rpc_cred; data->args.fh = NFS_FH(inode); data->args.context = ctx; + data->args.lock_context = dreq->l_ctx; data->args.offset = pos; data->args.pgbase = pgbase; data->args.pages = data->pagevec; @@ -845,7 +856,7 @@ static ssize_t nfs_direct_write(struct kiocb *iocb, const struct iovec *iov, unsigned long nr_segs, loff_t pos, size_t count) { - ssize_t result = 0; + ssize_t result = -ENOMEM; struct inode *inode = iocb->ki_filp->f_mapping->host; struct nfs_direct_req *dreq; size_t wsize = NFS_SERVER(inode)->wsize; @@ -853,7 +864,7 @@ static ssize_t nfs_direct_write(struct kiocb *iocb, const struct iovec *iov, dreq = nfs_direct_req_alloc(); if (!dreq) - return -ENOMEM; + goto out; nfs_alloc_commit_data(dreq); if (dreq->commit_data == NULL || count < wsize) @@ -861,14 +872,18 @@ static ssize_t nfs_direct_write(struct kiocb *iocb, const struct iovec *iov, dreq->inode = inode; dreq->ctx = get_nfs_open_context(nfs_file_open_context(iocb->ki_filp)); + dreq->l_ctx = nfs_get_lock_context(dreq->ctx); + if (dreq->l_ctx != NULL) + goto out_release; if (!is_sync_kiocb(iocb)) dreq->iocb = iocb; result = nfs_direct_write_schedule_iovec(dreq, iov, nr_segs, pos, sync); if (!result) result = nfs_direct_wait(dreq); +out_release: nfs_direct_req_release(dreq); - +out: return result; } diff --git a/fs/nfs/file.c b/fs/nfs/file.c index f036153d9f5..2d141a74ae8 100644 --- a/fs/nfs/file.c +++ b/fs/nfs/file.c @@ -203,37 +203,11 @@ static loff_t nfs_file_llseek(struct file *filp, loff_t offset, int origin) } /* - * Helper for nfs_file_flush() and nfs_file_fsync() - * - * Notice that it clears the NFS_CONTEXT_ERROR_WRITE before synching to - * disk, but it retrieves and clears ctx->error after synching, despite - * the two being set at the same time in nfs_context_set_write_error(). - * This is because the former is used to notify the _next_ call to - * nfs_file_write() that a write error occured, and hence cause it to - * fall back to doing a synchronous write. - */ -static int nfs_do_fsync(struct nfs_open_context *ctx, struct inode *inode) -{ - int have_error, status; - int ret = 0; - - have_error = test_and_clear_bit(NFS_CONTEXT_ERROR_WRITE, &ctx->flags); - status = nfs_wb_all(inode); - have_error |= test_bit(NFS_CONTEXT_ERROR_WRITE, &ctx->flags); - if (have_error) - ret = xchg(&ctx->error, 0); - if (!ret) - ret = status; - return ret; -} - -/* * Flush all dirty pages, and check for write errors. */ static int nfs_file_flush(struct file *file, fl_owner_t id) { - struct nfs_open_context *ctx = nfs_file_open_context(file); struct dentry *dentry = file->f_path.dentry; struct inode *inode = dentry->d_inode; @@ -246,7 +220,7 @@ nfs_file_flush(struct file *file, fl_owner_t id) return 0; /* Flush writes to the server and return any errors */ - return nfs_do_fsync(ctx, inode); + return vfs_fsync(file, 0); } static ssize_t @@ -321,6 +295,13 @@ nfs_file_mmap(struct file * file, struct vm_area_struct * vma) * Flush any dirty pages for this process, and check for write errors. * The return status from this call provides a reliable indication of * whether any write errors occurred for this process. + * + * Notice that it clears the NFS_CONTEXT_ERROR_WRITE before synching to + * disk, but it retrieves and clears ctx->error after synching, despite + * the two being set at the same time in nfs_context_set_write_error(). + * This is because the former is used to notify the _next_ call to + * nfs_file_write() that a write error occured, and hence cause it to + * fall back to doing a synchronous write. */ static int nfs_file_fsync(struct file *file, int datasync) @@ -328,13 +309,23 @@ nfs_file_fsync(struct file *file, int datasync) struct dentry *dentry = file->f_path.dentry; struct nfs_open_context *ctx = nfs_file_open_context(file); struct inode *inode = dentry->d_inode; + int have_error, status; + int ret = 0; + dprintk("NFS: fsync file(%s/%s) datasync %d\n", dentry->d_parent->d_name.name, dentry->d_name.name, datasync); nfs_inc_stats(inode, NFSIOS_VFSFSYNC); - return nfs_do_fsync(ctx, inode); + have_error = test_and_clear_bit(NFS_CONTEXT_ERROR_WRITE, &ctx->flags); + status = nfs_commit_inode(inode, FLUSH_SYNC); + have_error |= test_bit(NFS_CONTEXT_ERROR_WRITE, &ctx->flags); + if (have_error) + ret = xchg(&ctx->error, 0); + if (!ret) + ret = status; + return ret; } /* @@ -648,7 +639,7 @@ static ssize_t nfs_file_write(struct kiocb *iocb, const struct iovec *iov, /* Return error values for O_DSYNC and IS_SYNC() */ if (result >= 0 && nfs_need_sync_write(iocb->ki_filp, inode)) { - int err = nfs_do_fsync(nfs_file_open_context(iocb->ki_filp), inode); + int err = vfs_fsync(iocb->ki_filp, 0); if (err < 0) result = err; } @@ -684,7 +675,7 @@ static ssize_t nfs_file_splice_write(struct pipe_inode_info *pipe, written = ret; if (ret >= 0 && nfs_need_sync_write(filp, inode)) { - int err = nfs_do_fsync(nfs_file_open_context(filp), inode); + int err = vfs_fsync(filp, 0); if (err < 0) ret = err; } diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c index 099b3518fee..581d8f081e6 100644 --- a/fs/nfs/inode.c +++ b/fs/nfs/inode.c @@ -413,10 +413,8 @@ nfs_setattr(struct dentry *dentry, struct iattr *attr) return 0; /* Write all dirty data */ - if (S_ISREG(inode->i_mode)) { - filemap_write_and_wait(inode->i_mapping); + if (S_ISREG(inode->i_mode)) nfs_wb_all(inode); - } fattr = nfs_alloc_fattr(); if (fattr == NULL) @@ -530,6 +528,68 @@ out: return err; } +static void nfs_init_lock_context(struct nfs_lock_context *l_ctx) +{ + atomic_set(&l_ctx->count, 1); + l_ctx->lockowner = current->files; + l_ctx->pid = current->tgid; + INIT_LIST_HEAD(&l_ctx->list); +} + +static struct nfs_lock_context *__nfs_find_lock_context(struct nfs_open_context *ctx) +{ + struct nfs_lock_context *pos; + + list_for_each_entry(pos, &ctx->lock_context.list, list) { + if (pos->lockowner != current->files) + continue; + if (pos->pid != current->tgid) + continue; + atomic_inc(&pos->count); + return pos; + } + return NULL; +} + +struct nfs_lock_context *nfs_get_lock_context(struct nfs_open_context *ctx) +{ + struct nfs_lock_context *res, *new = NULL; + struct inode *inode = ctx->path.dentry->d_inode; + + spin_lock(&inode->i_lock); + res = __nfs_find_lock_context(ctx); + if (res == NULL) { + spin_unlock(&inode->i_lock); + new = kmalloc(sizeof(*new), GFP_KERNEL); + if (new == NULL) + return NULL; + nfs_init_lock_context(new); + spin_lock(&inode->i_lock); + res = __nfs_find_lock_context(ctx); + if (res == NULL) { + list_add_tail(&new->list, &ctx->lock_context.list); + new->open_context = ctx; + res = new; + new = NULL; + } + } + spin_unlock(&inode->i_lock); + kfree(new); + return res; +} + +void nfs_put_lock_context(struct nfs_lock_context *l_ctx) +{ + struct nfs_open_context *ctx = l_ctx->open_context; + struct inode *inode = ctx->path.dentry->d_inode; + + if (!atomic_dec_and_lock(&l_ctx->count, &inode->i_lock)) + return; + list_del(&l_ctx->list); + spin_unlock(&inode->i_lock); + kfree(l_ctx); +} + /** * nfs_close_context - Common close_context() routine NFSv2/v3 * @ctx: pointer to context @@ -566,11 +626,11 @@ static struct nfs_open_context *alloc_nfs_open_context(struct path *path, struct path_get(&ctx->path); ctx->cred = get_rpccred(cred); ctx->state = NULL; - ctx->lockowner = current->files; ctx->flags = 0; ctx->error = 0; ctx->dir_cookie = 0; - atomic_set(&ctx->count, 1); + nfs_init_lock_context(&ctx->lock_context); + ctx->lock_context.open_context = ctx; } return ctx; } @@ -578,7 +638,7 @@ static struct nfs_open_context *alloc_nfs_open_context(struct path *path, struct struct nfs_open_context *get_nfs_open_context(struct nfs_open_context *ctx) { if (ctx != NULL) - atomic_inc(&ctx->count); + atomic_inc(&ctx->lock_context.count); return ctx; } @@ -586,7 +646,7 @@ static void __put_nfs_open_context(struct nfs_open_context *ctx, int is_sync) { struct inode *inode = ctx->path.dentry->d_inode; - if (!atomic_dec_and_lock(&ctx->count, &inode->i_lock)) + if (!atomic_dec_and_lock(&ctx->lock_context.count, &inode->i_lock)) return; list_del(&ctx->list); spin_unlock(&inode->i_lock); diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h index e70f44b9b3f..4c2150d8671 100644 --- a/fs/nfs/internal.h +++ b/fs/nfs/internal.h @@ -370,10 +370,9 @@ unsigned int nfs_page_array_len(unsigned int base, size_t len) * Helper for restarting RPC calls in the possible presence of NFSv4.1 * sessions. */ -static inline void nfs_restart_rpc(struct rpc_task *task, const struct nfs_client *clp) +static inline int nfs_restart_rpc(struct rpc_task *task, const struct nfs_client *clp) { if (nfs4_has_session(clp)) - rpc_restart_call_prepare(task); - else - rpc_restart_call(task); + return rpc_restart_call_prepare(task); + return rpc_restart_call(task); } diff --git a/fs/nfs/nfs2xdr.c b/fs/nfs/nfs2xdr.c index 81cf1425791..db8846a0e82 100644 --- a/fs/nfs/nfs2xdr.c +++ b/fs/nfs/nfs2xdr.c @@ -233,7 +233,7 @@ nfs_xdr_removeargs(struct rpc_rqst *req, __be32 *p, const struct nfs_removeargs static int nfs_xdr_readargs(struct rpc_rqst *req, __be32 *p, struct nfs_readargs *args) { - struct rpc_auth *auth = req->rq_task->tk_msg.rpc_cred->cr_auth; + struct rpc_auth *auth = req->rq_cred->cr_auth; unsigned int replen; u32 offset = (u32)args->offset; u32 count = args->count; @@ -393,8 +393,7 @@ nfs_xdr_symlinkargs(struct rpc_rqst *req, __be32 *p, struct nfs_symlinkargs *arg static int nfs_xdr_readdirargs(struct rpc_rqst *req, __be32 *p, struct nfs_readdirargs *args) { - struct rpc_task *task = req->rq_task; - struct rpc_auth *auth = task->tk_msg.rpc_cred->cr_auth; + struct rpc_auth *auth = req->rq_cred->cr_auth; unsigned int replen; u32 count = args->count; @@ -575,7 +574,7 @@ nfs_xdr_diropres(struct rpc_rqst *req, __be32 *p, struct nfs_diropok *res) static int nfs_xdr_readlinkargs(struct rpc_rqst *req, __be32 *p, struct nfs_readlinkargs *args) { - struct rpc_auth *auth = req->rq_task->tk_msg.rpc_cred->cr_auth; + struct rpc_auth *auth = req->rq_cred->cr_auth; unsigned int replen; p = xdr_encode_fhandle(p, args->fh); diff --git a/fs/nfs/nfs3xdr.c b/fs/nfs/nfs3xdr.c index 75dcfc7da36..9769704f8ce 100644 --- a/fs/nfs/nfs3xdr.c +++ b/fs/nfs/nfs3xdr.c @@ -330,7 +330,7 @@ nfs3_xdr_accessargs(struct rpc_rqst *req, __be32 *p, struct nfs3_accessargs *arg static int nfs3_xdr_readargs(struct rpc_rqst *req, __be32 *p, struct nfs_readargs *args) { - struct rpc_auth *auth = req->rq_task->tk_msg.rpc_cred->cr_auth; + struct rpc_auth *auth = req->rq_cred->cr_auth; unsigned int replen; u32 count = args->count; @@ -471,7 +471,7 @@ nfs3_xdr_linkargs(struct rpc_rqst *req, __be32 *p, struct nfs3_linkargs *args) static int nfs3_xdr_readdirargs(struct rpc_rqst *req, __be32 *p, struct nfs3_readdirargs *args) { - struct rpc_auth *auth = req->rq_task->tk_msg.rpc_cred->cr_auth; + struct rpc_auth *auth = req->rq_cred->cr_auth; unsigned int replen; u32 count = args->count; @@ -675,7 +675,7 @@ static int nfs3_xdr_getaclargs(struct rpc_rqst *req, __be32 *p, struct nfs3_getaclargs *args) { - struct rpc_auth *auth = req->rq_task->tk_msg.rpc_cred->cr_auth; + struct rpc_auth *auth = req->rq_cred->cr_auth; unsigned int replen; p = xdr_encode_fhandle(p, args->fh); @@ -802,7 +802,7 @@ nfs3_xdr_accessres(struct rpc_rqst *req, __be32 *p, struct nfs3_accessres *res) static int nfs3_xdr_readlinkargs(struct rpc_rqst *req, __be32 *p, struct nfs3_readlinkargs *args) { - struct rpc_auth *auth = req->rq_task->tk_msg.rpc_cred->cr_auth; + struct rpc_auth *auth = req->rq_cred->cr_auth; unsigned int replen; p = xdr_encode_fhandle(p, args->fh); diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h index c538c6106e1..311e15cc8af 100644 --- a/fs/nfs/nfs4_fs.h +++ b/fs/nfs/nfs4_fs.h @@ -45,10 +45,29 @@ enum nfs4_client_state { NFS4CLNT_RECLAIM_NOGRACE, NFS4CLNT_DELEGRETURN, NFS4CLNT_SESSION_RESET, - NFS4CLNT_SESSION_DRAINING, NFS4CLNT_RECALL_SLOT, }; +enum nfs4_session_state { + NFS4_SESSION_INITING, + NFS4_SESSION_DRAINING, +}; + +struct nfs4_minor_version_ops { + u32 minor_version; + + int (*call_sync)(struct nfs_server *server, + struct rpc_message *msg, + struct nfs4_sequence_args *args, + struct nfs4_sequence_res *res, + int cache_reply); + int (*validate_stateid)(struct nfs_delegation *, + const nfs4_stateid *); + const struct nfs4_state_recovery_ops *reboot_recovery_ops; + const struct nfs4_state_recovery_ops *nograce_recovery_ops; + const struct nfs4_state_maintenance_ops *state_renewal_ops; +}; + /* * struct rpc_sequence ensures that RPC calls are sent in the exact * order that they appear on the list. @@ -89,7 +108,6 @@ struct nfs_unique_id { */ struct nfs4_state_owner { struct nfs_unique_id so_owner_id; - struct nfs_client *so_client; struct nfs_server *so_server; struct rb_node so_client_node; @@ -99,7 +117,6 @@ struct nfs4_state_owner { 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; }; @@ -125,10 +142,20 @@ enum { * LOCK: one nfs4_state (LOCK) to hold the lock stateid nfs4_state(OPEN) */ +struct nfs4_lock_owner { + unsigned int lo_type; +#define NFS4_ANY_LOCK_TYPE (0U) +#define NFS4_FLOCK_LOCK_TYPE (1U << 0) +#define NFS4_POSIX_LOCK_TYPE (1U << 1) + union { + fl_owner_t posix_owner; + pid_t flock_owner; + } lo_u; +}; + struct nfs4_lock_state { struct list_head ls_locks; /* Other lock stateids */ struct nfs4_state * ls_state; /* Pointer to open state */ - fl_owner_t ls_owner; /* POSIX lock owner */ #define NFS_LOCK_INITIALIZED 1 int ls_flags; struct nfs_seqid_counter ls_seqid; @@ -136,6 +163,7 @@ struct nfs4_lock_state { struct nfs_unique_id ls_id; nfs4_stateid ls_stateid; atomic_t ls_count; + struct nfs4_lock_owner ls_owner; }; /* bits for nfs4_state->flags */ @@ -219,11 +247,15 @@ extern int nfs4_open_revalidate(struct inode *, struct dentry *, int, struct nam extern int nfs4_server_capabilities(struct nfs_server *server, struct nfs_fh *fhandle); extern int nfs4_proc_fs_locations(struct inode *dir, const struct qstr *name, struct nfs4_fs_locations *fs_locations, struct page *page); +extern void nfs4_release_lockowner(const struct nfs4_lock_state *); -extern struct nfs4_state_recovery_ops *nfs4_reboot_recovery_ops[]; -extern struct nfs4_state_recovery_ops *nfs4_nograce_recovery_ops[]; #if defined(CONFIG_NFS_V4_1) -extern int nfs4_setup_sequence(struct nfs_client *clp, +static inline struct nfs4_session *nfs4_get_session(const struct nfs_server *server) +{ + return server->nfs_client->cl_session; +} + +extern int nfs4_setup_sequence(const struct nfs_server *server, struct nfs4_sequence_args *args, struct nfs4_sequence_res *res, int cache_reply, struct rpc_task *task); extern void nfs4_destroy_session(struct nfs4_session *session); @@ -234,7 +266,12 @@ extern int nfs4_init_session(struct nfs_server *server); extern int nfs4_proc_get_lease_time(struct nfs_client *clp, struct nfs_fsinfo *fsinfo); #else /* CONFIG_NFS_v4_1 */ -static inline int nfs4_setup_sequence(struct nfs_client *clp, +static inline struct nfs4_session *nfs4_get_session(const struct nfs_server *server) +{ + return NULL; +} + +static inline int nfs4_setup_sequence(const struct nfs_server *server, struct nfs4_sequence_args *args, struct nfs4_sequence_res *res, int cache_reply, struct rpc_task *task) { @@ -247,7 +284,7 @@ static inline int nfs4_init_session(struct nfs_server *server) } #endif /* CONFIG_NFS_V4_1 */ -extern struct nfs4_state_maintenance_ops *nfs4_state_renewal_ops[]; +extern const struct nfs4_minor_version_ops *nfs_v4_minor_ops[]; extern const u32 nfs4_fattr_bitmap[2]; extern const u32 nfs4_statfs_bitmap[2]; @@ -284,7 +321,7 @@ extern void nfs41_handle_sequence_flag_errors(struct nfs_client *clp, u32 flags) extern void nfs41_handle_recall_slot(struct nfs_client *clp); 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); +extern void nfs4_copy_stateid(nfs4_stateid *, struct nfs4_state *, fl_owner_t, pid_t); extern struct nfs_seqid *nfs_alloc_seqid(struct nfs_seqid_counter *counter, gfp_t gfp_mask); extern int nfs_wait_on_sequence(struct nfs_seqid *seqid, struct rpc_task *task); diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 70015dd60a9..7ffbb98ddec 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -303,15 +303,19 @@ do_state_recovery: } -static void renew_lease(const struct nfs_server *server, unsigned long timestamp) +static void do_renew_lease(struct nfs_client *clp, unsigned long timestamp) { - struct nfs_client *clp = server->nfs_client; spin_lock(&clp->cl_lock); if (time_before(clp->cl_last_renewal,timestamp)) clp->cl_last_renewal = timestamp; spin_unlock(&clp->cl_lock); } +static void renew_lease(const struct nfs_server *server, unsigned long timestamp) +{ + do_renew_lease(server->nfs_client, timestamp); +} + #if defined(CONFIG_NFS_V4_1) /* @@ -356,7 +360,7 @@ static void nfs41_check_drain_session_complete(struct nfs4_session *ses) { struct rpc_task *task; - if (!test_bit(NFS4CLNT_SESSION_DRAINING, &ses->clp->cl_state)) { + if (!test_bit(NFS4_SESSION_DRAINING, &ses->session_state)) { task = rpc_wake_up_next(&ses->fc_slot_table.slot_tbl_waitq); if (task) rpc_task_set_priority(task, RPC_PRIORITY_PRIVILEGED); @@ -370,12 +374,11 @@ static void nfs41_check_drain_session_complete(struct nfs4_session *ses) complete(&ses->complete); } -static void nfs41_sequence_free_slot(const struct nfs_client *clp, - struct nfs4_sequence_res *res) +static void nfs41_sequence_free_slot(struct nfs4_sequence_res *res) { struct nfs4_slot_table *tbl; - tbl = &clp->cl_session->fc_slot_table; + tbl = &res->sr_session->fc_slot_table; if (res->sr_slotid == NFS4_MAX_SLOT_TABLE) { /* just wake up the next guy waiting since * we may have not consumed a slot after all */ @@ -385,18 +388,17 @@ static void nfs41_sequence_free_slot(const struct nfs_client *clp, spin_lock(&tbl->slot_tbl_lock); nfs4_free_slot(tbl, res->sr_slotid); - nfs41_check_drain_session_complete(clp->cl_session); + nfs41_check_drain_session_complete(res->sr_session); spin_unlock(&tbl->slot_tbl_lock); res->sr_slotid = NFS4_MAX_SLOT_TABLE; } -static void nfs41_sequence_done(struct nfs_client *clp, - struct nfs4_sequence_res *res, - int rpc_status) +static int nfs41_sequence_done(struct rpc_task *task, struct nfs4_sequence_res *res) { unsigned long timestamp; struct nfs4_slot_table *tbl; struct nfs4_slot *slot; + struct nfs_client *clp; /* * sr_status remains 1 if an RPC level error occurred. The server @@ -411,25 +413,51 @@ static void nfs41_sequence_done(struct nfs_client *clp, if (res->sr_slotid == NFS4_MAX_SLOT_TABLE) goto out; + tbl = &res->sr_session->fc_slot_table; + slot = tbl->slots + res->sr_slotid; + /* Check the SEQUENCE operation status */ - if (res->sr_status == 0) { - tbl = &clp->cl_session->fc_slot_table; - slot = tbl->slots + res->sr_slotid; + switch (res->sr_status) { + case 0: /* Update the slot's sequence and clientid lease timer */ ++slot->seq_nr; timestamp = res->sr_renewal_time; - spin_lock(&clp->cl_lock); - if (time_before(clp->cl_last_renewal, timestamp)) - clp->cl_last_renewal = timestamp; - spin_unlock(&clp->cl_lock); + clp = res->sr_session->clp; + do_renew_lease(clp, timestamp); /* Check sequence flags */ if (atomic_read(&clp->cl_count) > 1) nfs41_handle_sequence_flag_errors(clp, res->sr_status_flags); + break; + case -NFS4ERR_DELAY: + /* The server detected a resend of the RPC call and + * returned NFS4ERR_DELAY as per Section 2.10.6.2 + * of RFC5661. + */ + dprintk("%s: slot=%d seq=%d: Operation in progress\n", + __func__, res->sr_slotid, slot->seq_nr); + goto out_retry; + default: + /* Just update the slot sequence no. */ + ++slot->seq_nr; } out: /* The session may be reset by one of the error handlers. */ dprintk("%s: Error %d free the slot \n", __func__, res->sr_status); - nfs41_sequence_free_slot(clp, res); + nfs41_sequence_free_slot(res); + return 1; +out_retry: + if (!rpc_restart_call(task)) + goto out; + rpc_delay(task, NFS4_POLL_RETRY_MAX); + return 0; +} + +static int nfs4_sequence_done(struct rpc_task *task, + struct nfs4_sequence_res *res) +{ + if (res->sr_session == NULL) + return 1; + return nfs41_sequence_done(task, res); } /* @@ -480,12 +508,11 @@ static int nfs41_setup_sequence(struct nfs4_session *session, if (res->sr_slotid != NFS4_MAX_SLOT_TABLE) return 0; - memset(res, 0, sizeof(*res)); res->sr_slotid = NFS4_MAX_SLOT_TABLE; tbl = &session->fc_slot_table; spin_lock(&tbl->slot_tbl_lock); - if (test_bit(NFS4CLNT_SESSION_DRAINING, &session->clp->cl_state) && + if (test_bit(NFS4_SESSION_DRAINING, &session->session_state) && !rpc_task_has_priority(task, RPC_PRIORITY_PRIVILEGED)) { /* * The state manager will wait until the slot table is empty. @@ -525,6 +552,7 @@ static int nfs41_setup_sequence(struct nfs4_session *session, res->sr_session = session; res->sr_slotid = slotid; res->sr_renewal_time = jiffies; + res->sr_status_flags = 0; /* * sr_status is only set in decode_sequence, and so will remain * set to 1 if an rpc level failure occurs. @@ -533,33 +561,33 @@ static int nfs41_setup_sequence(struct nfs4_session *session, return 0; } -int nfs4_setup_sequence(struct nfs_client *clp, +int nfs4_setup_sequence(const struct nfs_server *server, struct nfs4_sequence_args *args, struct nfs4_sequence_res *res, int cache_reply, struct rpc_task *task) { + struct nfs4_session *session = nfs4_get_session(server); int ret = 0; + if (session == NULL) { + args->sa_session = NULL; + res->sr_session = NULL; + goto out; + } + dprintk("--> %s clp %p session %p sr_slotid %d\n", - __func__, clp, clp->cl_session, res->sr_slotid); + __func__, session->clp, session, res->sr_slotid); - if (!nfs4_has_session(clp)) - goto out; - ret = nfs41_setup_sequence(clp->cl_session, args, res, cache_reply, + ret = nfs41_setup_sequence(session, args, res, cache_reply, task); - if (ret && ret != -EAGAIN) { - /* terminate rpc task */ - task->tk_status = ret; - task->tk_action = NULL; - } out: dprintk("<-- %s status=%d\n", __func__, ret); return ret; } struct nfs41_call_sync_data { - struct nfs_client *clp; + const struct nfs_server *seq_server; struct nfs4_sequence_args *seq_args; struct nfs4_sequence_res *seq_res; int cache_reply; @@ -569,9 +597,9 @@ static void nfs41_call_sync_prepare(struct rpc_task *task, void *calldata) { struct nfs41_call_sync_data *data = calldata; - dprintk("--> %s data->clp->cl_session %p\n", __func__, - data->clp->cl_session); - if (nfs4_setup_sequence(data->clp, data->seq_args, + dprintk("--> %s data->seq_server %p\n", __func__, data->seq_server); + + if (nfs4_setup_sequence(data->seq_server, data->seq_args, data->seq_res, data->cache_reply, task)) return; rpc_call_start(task); @@ -587,7 +615,7 @@ static void nfs41_call_sync_done(struct rpc_task *task, void *calldata) { struct nfs41_call_sync_data *data = calldata; - nfs41_sequence_done(data->clp, data->seq_res, task->tk_status); + nfs41_sequence_done(task, data->seq_res); } struct rpc_call_ops nfs41_call_sync_ops = { @@ -600,8 +628,7 @@ struct rpc_call_ops nfs41_call_priv_sync_ops = { .rpc_call_done = nfs41_call_sync_done, }; -static int nfs4_call_sync_sequence(struct nfs_client *clp, - struct rpc_clnt *clnt, +static int nfs4_call_sync_sequence(struct nfs_server *server, struct rpc_message *msg, struct nfs4_sequence_args *args, struct nfs4_sequence_res *res, @@ -611,13 +638,13 @@ static int nfs4_call_sync_sequence(struct nfs_client *clp, int ret; struct rpc_task *task; struct nfs41_call_sync_data data = { - .clp = clp, + .seq_server = server, .seq_args = args, .seq_res = res, .cache_reply = cache_reply, }; struct rpc_task_setup task_setup = { - .rpc_client = clnt, + .rpc_client = server->client, .rpc_message = msg, .callback_ops = &nfs41_call_sync_ops, .callback_data = &data @@ -642,10 +669,15 @@ int _nfs4_call_sync_session(struct nfs_server *server, struct nfs4_sequence_res *res, int cache_reply) { - return nfs4_call_sync_sequence(server->nfs_client, server->client, - msg, args, res, cache_reply, 0); + return nfs4_call_sync_sequence(server, msg, args, res, cache_reply, 0); } +#else +static int nfs4_sequence_done(struct rpc_task *task, + struct nfs4_sequence_res *res) +{ + return 1; +} #endif /* CONFIG_NFS_V4_1 */ int _nfs4_call_sync(struct nfs_server *server, @@ -659,18 +691,9 @@ int _nfs4_call_sync(struct nfs_server *server, } #define nfs4_call_sync(server, msg, args, res, cache_reply) \ - (server)->nfs_client->cl_call_sync((server), (msg), &(args)->seq_args, \ + (server)->nfs_client->cl_mvops->call_sync((server), (msg), &(args)->seq_args, \ &(res)->seq_res, (cache_reply)) -static void nfs4_sequence_done(const struct nfs_server *server, - struct nfs4_sequence_res *res, int rpc_status) -{ -#ifdef CONFIG_NFS_V4_1 - if (nfs4_has_session(server->nfs_client)) - nfs41_sequence_done(server->nfs_client, res, rpc_status); -#endif /* CONFIG_NFS_V4_1 */ -} - static void update_changeattr(struct inode *dir, struct nfs4_change_info *cinfo) { struct nfs_inode *nfsi = NFS_I(dir); @@ -745,19 +768,14 @@ static struct nfs4_opendata *nfs4_opendata_alloc(struct path *path, p->o_arg.server = server; p->o_arg.bitmask = server->attr_bitmask; p->o_arg.claim = NFS4_OPEN_CLAIM_NULL; - if (flags & O_EXCL) { - if (nfs4_has_persistent_session(server->nfs_client)) { - /* GUARDED */ - p->o_arg.u.attrs = &p->attrs; - memcpy(&p->attrs, attrs, sizeof(p->attrs)); - } else { /* EXCLUSIVE4_1 */ - u32 *s = (u32 *) p->o_arg.u.verifier.data; - s[0] = jiffies; - s[1] = current->pid; - } - } else if (flags & O_CREAT) { + if (flags & O_CREAT) { + u32 *s; + p->o_arg.u.attrs = &p->attrs; memcpy(&p->attrs, attrs, sizeof(p->attrs)); + s = (u32 *) p->o_arg.u.verifier.data; + s[0] = jiffies; + s[1] = current->pid; } p->c_arg.fh = &p->o_res.fh; p->c_arg.stateid = &p->o_res.stateid; @@ -1255,8 +1273,6 @@ static void nfs4_open_confirm_done(struct rpc_task *task, void *calldata) struct nfs4_opendata *data = calldata; data->rpc_status = task->tk_status; - if (RPC_ASSASSINATED(task)) - return; if (data->rpc_status == 0) { memcpy(data->o_res.stateid.data, data->c_res.stateid.data, sizeof(data->o_res.stateid.data)); @@ -1356,13 +1372,13 @@ static void nfs4_open_prepare(struct rpc_task *task, void *calldata) } /* Update sequence id. */ data->o_arg.id = sp->so_owner_id.id; - data->o_arg.clientid = sp->so_client->cl_clientid; + data->o_arg.clientid = sp->so_server->nfs_client->cl_clientid; if (data->o_arg.claim == NFS4_OPEN_CLAIM_PREVIOUS) { task->tk_msg.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_OPEN_NOATTR]; nfs_copy_fh(&data->o_res.fh, data->o_arg.fh); } data->timestamp = jiffies; - if (nfs4_setup_sequence(data->o_arg.server->nfs_client, + if (nfs4_setup_sequence(data->o_arg.server, &data->o_arg.seq_args, &data->o_res.seq_res, 1, task)) return; @@ -1385,11 +1401,9 @@ static void nfs4_open_done(struct rpc_task *task, void *calldata) data->rpc_status = task->tk_status; - nfs4_sequence_done(data->o_arg.server, &data->o_res.seq_res, - task->tk_status); - - if (RPC_ASSASSINATED(task)) + if (!nfs4_sequence_done(task, &data->o_res.seq_res)) return; + if (task->tk_status == 0) { switch (data->o_res.f_attr->mode & S_IFMT) { case S_IFREG: @@ -1773,7 +1787,7 @@ static int _nfs4_do_setattr(struct inode *inode, struct rpc_cred *cred, if (nfs4_copy_delegation_stateid(&arg.stateid, inode)) { /* Use that stateid */ } else if (state != NULL) { - nfs4_copy_stateid(&arg.stateid, state, current->files); + nfs4_copy_stateid(&arg.stateid, state, current->files, current->tgid); } else memcpy(&arg.stateid, &zero_stateid, sizeof(arg.stateid)); @@ -1838,8 +1852,7 @@ static void nfs4_close_done(struct rpc_task *task, void *data) struct nfs4_state *state = calldata->state; struct nfs_server *server = NFS_SERVER(calldata->inode); - nfs4_sequence_done(server, &calldata->res.seq_res, task->tk_status); - if (RPC_ASSASSINATED(task)) + if (!nfs4_sequence_done(task, &calldata->res.seq_res)) return; /* hmm. we are done with the inode, and in the process of freeing * the state_owner. we keep this around to process errors @@ -1903,7 +1916,7 @@ static void nfs4_close_prepare(struct rpc_task *task, void *data) nfs_fattr_init(calldata->res.fattr); calldata->timestamp = jiffies; - if (nfs4_setup_sequence((NFS_SERVER(calldata->inode))->nfs_client, + if (nfs4_setup_sequence(NFS_SERVER(calldata->inode), &calldata->arg.seq_args, &calldata->res.seq_res, 1, task)) return; @@ -2648,7 +2661,8 @@ static int nfs4_proc_unlink_done(struct rpc_task *task, struct inode *dir) { struct nfs_removeres *res = task->tk_msg.rpc_resp; - nfs4_sequence_done(res->server, &res->seq_res, task->tk_status); + if (!nfs4_sequence_done(task, &res->seq_res)) + return 0; if (nfs4_async_handle_error(task, res->server, NULL) == -EAGAIN) return 0; update_changeattr(dir, &res->cinfo); @@ -3093,7 +3107,8 @@ static int nfs4_read_done(struct rpc_task *task, struct nfs_read_data *data) dprintk("--> %s\n", __func__); - nfs4_sequence_done(server, &data->res.seq_res, task->tk_status); + if (!nfs4_sequence_done(task, &data->res.seq_res)) + return -EAGAIN; if (nfs4_async_handle_error(task, server, data->args.context->state) == -EAGAIN) { nfs_restart_rpc(task, server->nfs_client); @@ -3116,8 +3131,8 @@ static int nfs4_write_done(struct rpc_task *task, struct nfs_write_data *data) { struct inode *inode = data->inode; - nfs4_sequence_done(NFS_SERVER(inode), &data->res.seq_res, - task->tk_status); + if (!nfs4_sequence_done(task, &data->res.seq_res)) + return -EAGAIN; if (nfs4_async_handle_error(task, NFS_SERVER(inode), data->args.context->state) == -EAGAIN) { nfs_restart_rpc(task, NFS_SERVER(inode)->nfs_client); @@ -3145,8 +3160,9 @@ static int nfs4_commit_done(struct rpc_task *task, struct nfs_write_data *data) { struct inode *inode = data->inode; - nfs4_sequence_done(NFS_SERVER(inode), &data->res.seq_res, - task->tk_status); + if (!nfs4_sequence_done(task, &data->res.seq_res)) + return -EAGAIN; + if (nfs4_async_handle_error(task, NFS_SERVER(inode), NULL) == -EAGAIN) { nfs_restart_rpc(task, NFS_SERVER(inode)->nfs_client); return -EAGAIN; @@ -3196,10 +3212,7 @@ static void nfs4_renew_done(struct rpc_task *task, void *calldata) nfs4_schedule_state_recovery(clp); return; } - spin_lock(&clp->cl_lock); - if (time_before(clp->cl_last_renewal,timestamp)) - clp->cl_last_renewal = timestamp; - spin_unlock(&clp->cl_lock); + do_renew_lease(clp, timestamp); } static const struct rpc_call_ops nfs4_renew_ops = { @@ -3240,10 +3253,7 @@ int nfs4_proc_renew(struct nfs_client *clp, struct rpc_cred *cred) status = rpc_call_sync(clp->cl_rpcclient, &msg, 0); if (status < 0) return status; - spin_lock(&clp->cl_lock); - if (time_before(clp->cl_last_renewal,now)) - clp->cl_last_renewal = now; - spin_unlock(&clp->cl_lock); + do_renew_lease(clp, now); return 0; } @@ -3464,9 +3474,11 @@ 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, struct nfs_client *clp, struct nfs4_state *state) +nfs4_async_handle_error(struct rpc_task *task, const struct nfs_server *server, struct nfs4_state *state) { - if (!clp || task->tk_status >= 0) + struct nfs_client *clp = server->nfs_client; + + if (task->tk_status >= 0) return 0; switch(task->tk_status) { case -NFS4ERR_ADMIN_REVOKED: @@ -3498,8 +3510,7 @@ _nfs4_async_handle_error(struct rpc_task *task, const struct nfs_server *server, return -EAGAIN; #endif /* CONFIG_NFS_V4_1 */ case -NFS4ERR_DELAY: - if (server) - nfs_inc_server_stats(server, NFSIOS_DELAY); + nfs_inc_server_stats(server, NFSIOS_DELAY); case -NFS4ERR_GRACE: case -EKEYEXPIRED: rpc_delay(task, NFS4_POLL_RETRY_MAX); @@ -3520,12 +3531,6 @@ do_state_recovery: return -EAGAIN; } -static int -nfs4_async_handle_error(struct rpc_task *task, const struct nfs_server *server, struct nfs4_state *state) -{ - return _nfs4_async_handle_error(task, server, server->nfs_client, state); -} - int nfs4_proc_setclientid(struct nfs_client *clp, u32 program, unsigned short port, struct rpc_cred *cred, struct nfs4_setclientid_res *res) @@ -3641,8 +3646,8 @@ static void nfs4_delegreturn_done(struct rpc_task *task, void *calldata) { struct nfs4_delegreturndata *data = calldata; - nfs4_sequence_done(data->res.server, &data->res.seq_res, - task->tk_status); + if (!nfs4_sequence_done(task, &data->res.seq_res)) + return; switch (task->tk_status) { case -NFS4ERR_STALE_STATEID: @@ -3672,7 +3677,7 @@ static void nfs4_delegreturn_prepare(struct rpc_task *task, void *data) d_data = (struct nfs4_delegreturndata *)data; - if (nfs4_setup_sequence(d_data->res.server->nfs_client, + if (nfs4_setup_sequence(d_data->res.server, &d_data->args.seq_args, &d_data->res.seq_res, 1, task)) return; @@ -3892,9 +3897,7 @@ static void nfs4_locku_done(struct rpc_task *task, void *data) { struct nfs4_unlockdata *calldata = data; - nfs4_sequence_done(calldata->server, &calldata->res.seq_res, - task->tk_status); - if (RPC_ASSASSINATED(task)) + if (!nfs4_sequence_done(task, &calldata->res.seq_res)) return; switch (task->tk_status) { case 0: @@ -3927,7 +3930,7 @@ static void nfs4_locku_prepare(struct rpc_task *task, void *data) return; } calldata->timestamp = jiffies; - if (nfs4_setup_sequence(calldata->server->nfs_client, + if (nfs4_setup_sequence(calldata->server, &calldata->arg.seq_args, &calldata->res.seq_res, 1, task)) return; @@ -4082,7 +4085,8 @@ static void nfs4_lock_prepare(struct rpc_task *task, void *calldata) } else data->arg.new_lock_owner = 0; data->timestamp = jiffies; - if (nfs4_setup_sequence(data->server->nfs_client, &data->arg.seq_args, + if (nfs4_setup_sequence(data->server, + &data->arg.seq_args, &data->res.seq_res, 1, task)) return; rpc_call_start(task); @@ -4101,12 +4105,10 @@ static void nfs4_lock_done(struct rpc_task *task, void *calldata) dprintk("%s: begin!\n", __func__); - nfs4_sequence_done(data->server, &data->res.seq_res, - task->tk_status); + if (!nfs4_sequence_done(task, &data->res.seq_res)) + return; data->rpc_status = task->tk_status; - if (RPC_ASSASSINATED(task)) - goto out; if (data->arg.new_lock_owner != 0) { if (data->rpc_status == 0) nfs_confirm_seqid(&data->lsp->ls_seqid, 0); @@ -4424,6 +4426,34 @@ out: return err; } +static void nfs4_release_lockowner_release(void *calldata) +{ + kfree(calldata); +} + +const struct rpc_call_ops nfs4_release_lockowner_ops = { + .rpc_release = nfs4_release_lockowner_release, +}; + +void nfs4_release_lockowner(const struct nfs4_lock_state *lsp) +{ + struct nfs_server *server = lsp->ls_state->owner->so_server; + struct nfs_release_lockowner_args *args; + struct rpc_message msg = { + .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_RELEASE_LOCKOWNER], + }; + + if (server->nfs_client->cl_mvops->minor_version != 0) + return; + args = kmalloc(sizeof(*args), GFP_NOFS); + if (!args) + return; + args->lock_owner.clientid = server->nfs_client->cl_clientid; + args->lock_owner.id = lsp->ls_id.id; + msg.rpc_argp = args; + rpc_call_async(server->client, &msg, 0, &nfs4_release_lockowner_ops, args); +} + #define XATTR_NAME_NFSV4_ACL "system.nfs4_acl" int nfs4_setxattr(struct dentry *dentry, const char *key, const void *buf, @@ -4611,7 +4641,8 @@ static void nfs4_get_lease_time_done(struct rpc_task *task, void *calldata) (struct nfs4_get_lease_time_data *)calldata; dprintk("--> %s\n", __func__); - nfs41_sequence_done(data->clp, &data->res->lr_seq_res, task->tk_status); + if (!nfs41_sequence_done(task, &data->res->lr_seq_res)) + return; switch (task->tk_status) { case -NFS4ERR_DELAY: case -NFS4ERR_GRACE: @@ -4805,13 +4836,6 @@ struct nfs4_session *nfs4_alloc_session(struct nfs_client *clp) if (!session) return NULL; - /* - * The create session reply races with the server back - * channel probe. Mark the client NFS_CS_SESSION_INITING - * so that the client back channel can find the - * nfs_client struct - */ - clp->cl_cons_state = NFS_CS_SESSION_INITING; init_completion(&session->complete); tbl = &session->fc_slot_table; @@ -4824,6 +4848,8 @@ struct nfs4_session *nfs4_alloc_session(struct nfs_client *clp) spin_lock_init(&tbl->slot_tbl_lock); rpc_init_wait_queue(&tbl->slot_tbl_waitq, "BackChannel Slot table"); + session->session_state = 1<<NFS4_SESSION_INITING; + session->clp = clp; return session; } @@ -5040,6 +5066,10 @@ int nfs4_init_session(struct nfs_server *server) if (!nfs4_has_session(clp)) return 0; + session = clp->cl_session; + if (!test_and_clear_bit(NFS4_SESSION_INITING, &session->session_state)) + return 0; + rsize = server->rsize; if (rsize == 0) rsize = NFS_MAX_FILE_IO_SIZE; @@ -5047,7 +5077,6 @@ int nfs4_init_session(struct nfs_server *server) if (wsize == 0) wsize = NFS_MAX_FILE_IO_SIZE; - session = clp->cl_session; session->fc_attrs.max_rqst_sz = wsize + nfs41_maxwrite_overhead; session->fc_attrs.max_resp_sz = rsize + nfs41_maxread_overhead; @@ -5060,69 +5089,70 @@ int nfs4_init_session(struct nfs_server *server) /* * Renew the cl_session lease. */ -static int nfs4_proc_sequence(struct nfs_client *clp, struct rpc_cred *cred) -{ +struct nfs4_sequence_data { + struct nfs_client *clp; struct nfs4_sequence_args args; struct nfs4_sequence_res res; - - struct rpc_message msg = { - .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_SEQUENCE], - .rpc_argp = &args, - .rpc_resp = &res, - .rpc_cred = cred, - }; - - args.sa_cache_this = 0; - - return nfs4_call_sync_sequence(clp, clp->cl_rpcclient, &msg, &args, - &res, args.sa_cache_this, 1); -} +}; static void nfs41_sequence_release(void *data) { - struct nfs_client *clp = (struct nfs_client *)data; + struct nfs4_sequence_data *calldata = data; + struct nfs_client *clp = calldata->clp; if (atomic_read(&clp->cl_count) > 1) nfs4_schedule_state_renewal(clp); nfs_put_client(clp); + kfree(calldata); +} + +static int nfs41_sequence_handle_errors(struct rpc_task *task, struct nfs_client *clp) +{ + switch(task->tk_status) { + case -NFS4ERR_DELAY: + case -EKEYEXPIRED: + rpc_delay(task, NFS4_POLL_RETRY_MAX); + return -EAGAIN; + default: + nfs4_schedule_state_recovery(clp); + } + return 0; } static void nfs41_sequence_call_done(struct rpc_task *task, void *data) { - struct nfs_client *clp = (struct nfs_client *)data; + struct nfs4_sequence_data *calldata = data; + struct nfs_client *clp = calldata->clp; - nfs41_sequence_done(clp, task->tk_msg.rpc_resp, task->tk_status); + if (!nfs41_sequence_done(task, task->tk_msg.rpc_resp)) + return; if (task->tk_status < 0) { dprintk("%s ERROR %d\n", __func__, task->tk_status); if (atomic_read(&clp->cl_count) == 1) goto out; - if (_nfs4_async_handle_error(task, NULL, clp, NULL) - == -EAGAIN) { - nfs_restart_rpc(task, clp); + if (nfs41_sequence_handle_errors(task, clp) == -EAGAIN) { + rpc_restart_call_prepare(task); return; } } dprintk("%s rpc_cred %p\n", __func__, task->tk_msg.rpc_cred); out: - kfree(task->tk_msg.rpc_argp); - kfree(task->tk_msg.rpc_resp); - dprintk("<-- %s\n", __func__); } static void nfs41_sequence_prepare(struct rpc_task *task, void *data) { - struct nfs_client *clp; + struct nfs4_sequence_data *calldata = data; + struct nfs_client *clp = calldata->clp; struct nfs4_sequence_args *args; struct nfs4_sequence_res *res; - clp = (struct nfs_client *)data; args = task->tk_msg.rpc_argp; res = task->tk_msg.rpc_resp; - if (nfs4_setup_sequence(clp, args, res, 0, task)) + if (nfs41_setup_sequence(clp->cl_session, args, res, 0, task)) return; rpc_call_start(task); } @@ -5133,32 +5163,67 @@ static const struct rpc_call_ops nfs41_sequence_ops = { .rpc_release = nfs41_sequence_release, }; -static int nfs41_proc_async_sequence(struct nfs_client *clp, - struct rpc_cred *cred) +static struct rpc_task *_nfs41_proc_sequence(struct nfs_client *clp, struct rpc_cred *cred) { - struct nfs4_sequence_args *args; - struct nfs4_sequence_res *res; + struct nfs4_sequence_data *calldata; struct rpc_message msg = { .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_SEQUENCE], .rpc_cred = cred, }; + struct rpc_task_setup task_setup_data = { + .rpc_client = clp->cl_rpcclient, + .rpc_message = &msg, + .callback_ops = &nfs41_sequence_ops, + .flags = RPC_TASK_ASYNC | RPC_TASK_SOFT, + }; if (!atomic_inc_not_zero(&clp->cl_count)) - return -EIO; - args = kzalloc(sizeof(*args), GFP_NOFS); - res = kzalloc(sizeof(*res), GFP_NOFS); - if (!args || !res) { - kfree(args); - kfree(res); + return ERR_PTR(-EIO); + calldata = kmalloc(sizeof(*calldata), GFP_NOFS); + if (calldata == NULL) { nfs_put_client(clp); - return -ENOMEM; + return ERR_PTR(-ENOMEM); } - res->sr_slotid = NFS4_MAX_SLOT_TABLE; - msg.rpc_argp = args; - msg.rpc_resp = res; + calldata->res.sr_slotid = NFS4_MAX_SLOT_TABLE; + msg.rpc_argp = &calldata->args; + msg.rpc_resp = &calldata->res; + calldata->clp = clp; + task_setup_data.callback_data = calldata; - return rpc_call_async(clp->cl_rpcclient, &msg, RPC_TASK_SOFT, - &nfs41_sequence_ops, (void *)clp); + return rpc_run_task(&task_setup_data); +} + +static int nfs41_proc_async_sequence(struct nfs_client *clp, struct rpc_cred *cred) +{ + struct rpc_task *task; + int ret = 0; + + task = _nfs41_proc_sequence(clp, cred); + if (IS_ERR(task)) + ret = PTR_ERR(task); + else + rpc_put_task(task); + dprintk("<-- %s status=%d\n", __func__, ret); + return ret; +} + +static int nfs4_proc_sequence(struct nfs_client *clp, struct rpc_cred *cred) +{ + struct rpc_task *task; + int ret; + + task = _nfs41_proc_sequence(clp, cred); + if (IS_ERR(task)) { + ret = PTR_ERR(task); + goto out; + } + ret = rpc_wait_for_completion_task(task); + if (!ret) + ret = task->tk_status; + rpc_put_task(task); +out: + dprintk("<-- %s status=%d\n", __func__, ret); + return ret; } struct nfs4_reclaim_complete_data { @@ -5172,13 +5237,31 @@ static void nfs4_reclaim_complete_prepare(struct rpc_task *task, void *data) struct nfs4_reclaim_complete_data *calldata = data; rpc_task_set_priority(task, RPC_PRIORITY_PRIVILEGED); - if (nfs4_setup_sequence(calldata->clp, &calldata->arg.seq_args, + if (nfs41_setup_sequence(calldata->clp->cl_session, + &calldata->arg.seq_args, &calldata->res.seq_res, 0, task)) return; rpc_call_start(task); } +static int nfs41_reclaim_complete_handle_errors(struct rpc_task *task, struct nfs_client *clp) +{ + switch(task->tk_status) { + case 0: + case -NFS4ERR_COMPLETE_ALREADY: + case -NFS4ERR_WRONG_CRED: /* What to do here? */ + break; + case -NFS4ERR_DELAY: + case -EKEYEXPIRED: + rpc_delay(task, NFS4_POLL_RETRY_MAX); + return -EAGAIN; + default: + nfs4_schedule_state_recovery(clp); + } + return 0; +} + static void nfs4_reclaim_complete_done(struct rpc_task *task, void *data) { struct nfs4_reclaim_complete_data *calldata = data; @@ -5186,32 +5269,13 @@ static void nfs4_reclaim_complete_done(struct rpc_task *task, void *data) struct nfs4_sequence_res *res = &calldata->res.seq_res; dprintk("--> %s\n", __func__); - nfs41_sequence_done(clp, res, task->tk_status); - switch (task->tk_status) { - case 0: - case -NFS4ERR_COMPLETE_ALREADY: - break; - case -NFS4ERR_BADSESSION: - case -NFS4ERR_DEADSESSION: - /* - * Handle the session error, but do not retry the operation, as - * we have no way of telling whether the clientid had to be - * reset before we got our reply. If reset, a new wave of - * reclaim operations will follow, containing their own reclaim - * complete. We don't want our retry to get on the way of - * recovery by incorrectly indicating to the server that we're - * done reclaiming state since the process had to be restarted. - */ - _nfs4_async_handle_error(task, NULL, clp, NULL); - break; - default: - if (_nfs4_async_handle_error( - task, NULL, clp, NULL) == -EAGAIN) { - rpc_restart_call_prepare(task); - return; - } - } + if (!nfs41_sequence_done(task, res)) + return; + if (nfs41_reclaim_complete_handle_errors(task, clp) == -EAGAIN) { + rpc_restart_call_prepare(task); + return; + } dprintk("<-- %s\n", __func__); } @@ -5325,28 +5389,30 @@ struct nfs4_state_maintenance_ops nfs41_state_renewal_ops = { }; #endif -/* - * Per minor version reboot and network partition recovery ops - */ - -struct nfs4_state_recovery_ops *nfs4_reboot_recovery_ops[] = { - &nfs40_reboot_recovery_ops, -#if defined(CONFIG_NFS_V4_1) - &nfs41_reboot_recovery_ops, -#endif +static const struct nfs4_minor_version_ops nfs_v4_0_minor_ops = { + .minor_version = 0, + .call_sync = _nfs4_call_sync, + .validate_stateid = nfs4_validate_delegation_stateid, + .reboot_recovery_ops = &nfs40_reboot_recovery_ops, + .nograce_recovery_ops = &nfs40_nograce_recovery_ops, + .state_renewal_ops = &nfs40_state_renewal_ops, }; -struct nfs4_state_recovery_ops *nfs4_nograce_recovery_ops[] = { - &nfs40_nograce_recovery_ops, #if defined(CONFIG_NFS_V4_1) - &nfs41_nograce_recovery_ops, -#endif +static const struct nfs4_minor_version_ops nfs_v4_1_minor_ops = { + .minor_version = 1, + .call_sync = _nfs4_call_sync_session, + .validate_stateid = nfs41_validate_delegation_stateid, + .reboot_recovery_ops = &nfs41_reboot_recovery_ops, + .nograce_recovery_ops = &nfs41_nograce_recovery_ops, + .state_renewal_ops = &nfs41_state_renewal_ops, }; +#endif -struct nfs4_state_maintenance_ops *nfs4_state_renewal_ops[] = { - &nfs40_state_renewal_ops, +const struct nfs4_minor_version_ops *nfs_v4_minor_ops[] = { + [0] = &nfs_v4_0_minor_ops, #if defined(CONFIG_NFS_V4_1) - &nfs41_state_renewal_ops, + [1] = &nfs_v4_1_minor_ops, #endif }; diff --git a/fs/nfs/nfs4renewd.c b/fs/nfs/nfs4renewd.c index d87f10327b7..72b6c580af1 100644 --- a/fs/nfs/nfs4renewd.c +++ b/fs/nfs/nfs4renewd.c @@ -54,14 +54,14 @@ void nfs4_renew_state(struct work_struct *work) { - struct nfs4_state_maintenance_ops *ops; + const struct nfs4_state_maintenance_ops *ops; struct nfs_client *clp = container_of(work, struct nfs_client, cl_renewd.work); struct rpc_cred *cred; long lease; unsigned long last, now; - ops = nfs4_state_renewal_ops[clp->cl_minorversion]; + ops = clp->cl_mvops->state_renewal_ops; dprintk("%s: start\n", __func__); /* Are there any active superblocks? */ if (list_empty(&clp->cl_superblocks)) diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c index 34acf5926fd..3e2f19b04c0 100644 --- a/fs/nfs/nfs4state.c +++ b/fs/nfs/nfs4state.c @@ -145,7 +145,9 @@ static void nfs4_end_drain_session(struct nfs_client *clp) struct nfs4_session *ses = clp->cl_session; int max_slots; - if (test_and_clear_bit(NFS4CLNT_SESSION_DRAINING, &clp->cl_state)) { + if (ses == NULL) + return; + if (test_and_clear_bit(NFS4_SESSION_DRAINING, &ses->session_state)) { spin_lock(&ses->fc_slot_table.slot_tbl_lock); max_slots = ses->fc_slot_table.max_slots; while (max_slots--) { @@ -167,7 +169,7 @@ static int nfs4_begin_drain_session(struct nfs_client *clp) struct nfs4_slot_table *tbl = &ses->fc_slot_table; spin_lock(&tbl->slot_tbl_lock); - set_bit(NFS4CLNT_SESSION_DRAINING, &clp->cl_state); + set_bit(NFS4_SESSION_DRAINING, &ses->session_state); if (tbl->highest_used_slotid != -1) { INIT_COMPLETION(ses->complete); spin_unlock(&tbl->slot_tbl_lock); @@ -371,7 +373,6 @@ nfs4_alloc_state_owner(void) return NULL; spin_lock_init(&sp->so_lock); INIT_LIST_HEAD(&sp->so_states); - INIT_LIST_HEAD(&sp->so_delegations); rpc_init_wait_queue(&sp->so_sequence.wait, "Seqid_waitqueue"); sp->so_seqid.sequence = &sp->so_sequence; spin_lock_init(&sp->so_sequence.lock); @@ -384,7 +385,7 @@ static void nfs4_drop_state_owner(struct nfs4_state_owner *sp) { if (!RB_EMPTY_NODE(&sp->so_client_node)) { - struct nfs_client *clp = sp->so_client; + struct nfs_client *clp = sp->so_server->nfs_client; spin_lock(&clp->cl_lock); rb_erase(&sp->so_client_node, &clp->cl_state_owners); @@ -406,7 +407,6 @@ struct nfs4_state_owner *nfs4_get_state_owner(struct nfs_server *server, struct new = nfs4_alloc_state_owner(); if (new == NULL) return NULL; - new->so_client = clp; new->so_server = server; new->so_cred = cred; spin_lock(&clp->cl_lock); @@ -423,7 +423,7 @@ struct nfs4_state_owner *nfs4_get_state_owner(struct nfs_server *server, struct void nfs4_put_state_owner(struct nfs4_state_owner *sp) { - struct nfs_client *clp = sp->so_client; + struct nfs_client *clp = sp->so_server->nfs_client; struct rpc_cred *cred = sp->so_cred; if (!atomic_dec_and_lock(&sp->so_count, &clp->cl_lock)) @@ -602,12 +602,21 @@ void nfs4_close_sync(struct path *path, struct nfs4_state *state, fmode_t fmode) * that is compatible with current->files */ static struct nfs4_lock_state * -__nfs4_find_lock_state(struct nfs4_state *state, fl_owner_t fl_owner) +__nfs4_find_lock_state(struct nfs4_state *state, fl_owner_t fl_owner, pid_t fl_pid, unsigned int type) { struct nfs4_lock_state *pos; list_for_each_entry(pos, &state->lock_states, ls_locks) { - if (pos->ls_owner != fl_owner) + if (type != NFS4_ANY_LOCK_TYPE && pos->ls_owner.lo_type != type) continue; + switch (pos->ls_owner.lo_type) { + case NFS4_POSIX_LOCK_TYPE: + if (pos->ls_owner.lo_u.posix_owner != fl_owner) + continue; + break; + case NFS4_FLOCK_LOCK_TYPE: + if (pos->ls_owner.lo_u.flock_owner != fl_pid) + continue; + } atomic_inc(&pos->ls_count); return pos; } @@ -619,10 +628,10 @@ __nfs4_find_lock_state(struct nfs4_state *state, fl_owner_t fl_owner) * exists, return an uninitialized one. * */ -static struct nfs4_lock_state *nfs4_alloc_lock_state(struct nfs4_state *state, fl_owner_t fl_owner) +static struct nfs4_lock_state *nfs4_alloc_lock_state(struct nfs4_state *state, fl_owner_t fl_owner, pid_t fl_pid, unsigned int type) { struct nfs4_lock_state *lsp; - struct nfs_client *clp = state->owner->so_client; + struct nfs_client *clp = state->owner->so_server->nfs_client; lsp = kzalloc(sizeof(*lsp), GFP_NOFS); if (lsp == NULL) @@ -633,7 +642,18 @@ static struct nfs4_lock_state *nfs4_alloc_lock_state(struct nfs4_state *state, f lsp->ls_seqid.sequence = &lsp->ls_sequence; atomic_set(&lsp->ls_count, 1); lsp->ls_state = state; - lsp->ls_owner = fl_owner; + lsp->ls_owner.lo_type = type; + switch (lsp->ls_owner.lo_type) { + case NFS4_FLOCK_LOCK_TYPE: + lsp->ls_owner.lo_u.flock_owner = fl_pid; + break; + case NFS4_POSIX_LOCK_TYPE: + lsp->ls_owner.lo_u.posix_owner = fl_owner; + break; + default: + kfree(lsp); + return NULL; + } spin_lock(&clp->cl_lock); nfs_alloc_unique_id(&clp->cl_lockowner_id, &lsp->ls_id, 1, 64); spin_unlock(&clp->cl_lock); @@ -643,7 +663,7 @@ static struct nfs4_lock_state *nfs4_alloc_lock_state(struct nfs4_state *state, f static void nfs4_free_lock_state(struct nfs4_lock_state *lsp) { - struct nfs_client *clp = lsp->ls_state->owner->so_client; + struct nfs_client *clp = lsp->ls_state->owner->so_server->nfs_client; spin_lock(&clp->cl_lock); nfs_free_unique_id(&clp->cl_lockowner_id, &lsp->ls_id); @@ -657,13 +677,13 @@ static void nfs4_free_lock_state(struct nfs4_lock_state *lsp) * exists, return an uninitialized one. * */ -static struct nfs4_lock_state *nfs4_get_lock_state(struct nfs4_state *state, fl_owner_t owner) +static struct nfs4_lock_state *nfs4_get_lock_state(struct nfs4_state *state, fl_owner_t owner, pid_t pid, unsigned int type) { struct nfs4_lock_state *lsp, *new = NULL; for(;;) { spin_lock(&state->state_lock); - lsp = __nfs4_find_lock_state(state, owner); + lsp = __nfs4_find_lock_state(state, owner, pid, type); if (lsp != NULL) break; if (new != NULL) { @@ -674,7 +694,7 @@ static struct nfs4_lock_state *nfs4_get_lock_state(struct nfs4_state *state, fl_ break; } spin_unlock(&state->state_lock); - new = nfs4_alloc_lock_state(state, owner); + new = nfs4_alloc_lock_state(state, owner, pid, type); if (new == NULL) return NULL; } @@ -701,6 +721,8 @@ void nfs4_put_lock_state(struct nfs4_lock_state *lsp) if (list_empty(&state->lock_states)) clear_bit(LK_STATE_IN_USE, &state->flags); spin_unlock(&state->state_lock); + if (lsp->ls_flags & NFS_LOCK_INITIALIZED) + nfs4_release_lockowner(lsp); nfs4_free_lock_state(lsp); } @@ -728,7 +750,12 @@ int nfs4_set_lock_state(struct nfs4_state *state, struct file_lock *fl) if (fl->fl_ops != NULL) return 0; - lsp = nfs4_get_lock_state(state, fl->fl_owner); + if (fl->fl_flags & FL_POSIX) + lsp = nfs4_get_lock_state(state, fl->fl_owner, 0, NFS4_POSIX_LOCK_TYPE); + else if (fl->fl_flags & FL_FLOCK) + lsp = nfs4_get_lock_state(state, 0, fl->fl_pid, NFS4_FLOCK_LOCK_TYPE); + else + return -EINVAL; if (lsp == NULL) return -ENOMEM; fl->fl_u.nfs4_fl.owner = lsp; @@ -740,7 +767,7 @@ int nfs4_set_lock_state(struct nfs4_state *state, struct file_lock *fl) * Byte-range lock aware utility to initialize the stateid of read/write * requests. */ -void nfs4_copy_stateid(nfs4_stateid *dst, struct nfs4_state *state, fl_owner_t fl_owner) +void nfs4_copy_stateid(nfs4_stateid *dst, struct nfs4_state *state, fl_owner_t fl_owner, pid_t fl_pid) { struct nfs4_lock_state *lsp; int seq; @@ -753,7 +780,7 @@ void nfs4_copy_stateid(nfs4_stateid *dst, struct nfs4_state *state, fl_owner_t f return; spin_lock(&state->state_lock); - lsp = __nfs4_find_lock_state(state, fl_owner); + lsp = __nfs4_find_lock_state(state, fl_owner, fl_pid, NFS4_ANY_LOCK_TYPE); if (lsp != NULL && (lsp->ls_flags & NFS_LOCK_INITIALIZED) != 0) memcpy(dst, &lsp->ls_stateid, sizeof(*dst)); spin_unlock(&state->state_lock); @@ -1041,11 +1068,11 @@ restart: case -NFS4ERR_BAD_STATEID: case -NFS4ERR_RECLAIM_BAD: case -NFS4ERR_RECLAIM_CONFLICT: - nfs4_state_mark_reclaim_nograce(sp->so_client, state); + nfs4_state_mark_reclaim_nograce(sp->so_server->nfs_client, state); break; case -NFS4ERR_EXPIRED: case -NFS4ERR_NO_GRACE: - nfs4_state_mark_reclaim_nograce(sp->so_client, state); + nfs4_state_mark_reclaim_nograce(sp->so_server->nfs_client, state); case -NFS4ERR_STALE_CLIENTID: case -NFS4ERR_BADSESSION: case -NFS4ERR_BADSLOT: @@ -1120,8 +1147,7 @@ static void nfs4_state_end_reclaim_reboot(struct nfs_client *clp) if (!test_and_clear_bit(NFS4CLNT_RECLAIM_REBOOT, &clp->cl_state)) return; - nfs4_reclaim_complete(clp, - nfs4_reboot_recovery_ops[clp->cl_minorversion]); + nfs4_reclaim_complete(clp, clp->cl_mvops->reboot_recovery_ops); 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); @@ -1211,8 +1237,8 @@ restart: static int nfs4_check_lease(struct nfs_client *clp) { struct rpc_cred *cred; - struct nfs4_state_maintenance_ops *ops = - nfs4_state_renewal_ops[clp->cl_minorversion]; + const struct nfs4_state_maintenance_ops *ops = + clp->cl_mvops->state_renewal_ops; int status = -NFS4ERR_EXPIRED; /* Is the client already known to have an expired lease? */ @@ -1235,8 +1261,8 @@ out: static int nfs4_reclaim_lease(struct nfs_client *clp) { struct rpc_cred *cred; - struct nfs4_state_recovery_ops *ops = - nfs4_reboot_recovery_ops[clp->cl_minorversion]; + const struct nfs4_state_recovery_ops *ops = + clp->cl_mvops->reboot_recovery_ops; int status = -ENOENT; cred = ops->get_clid_cred(clp); @@ -1444,7 +1470,7 @@ static void nfs4_state_manager(struct nfs_client *clp) /* First recover reboot state... */ if (test_bit(NFS4CLNT_RECLAIM_REBOOT, &clp->cl_state)) { status = nfs4_do_reclaim(clp, - nfs4_reboot_recovery_ops[clp->cl_minorversion]); + clp->cl_mvops->reboot_recovery_ops); if (test_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state) || test_bit(NFS4CLNT_SESSION_RESET, &clp->cl_state)) continue; @@ -1458,7 +1484,7 @@ static void nfs4_state_manager(struct nfs_client *clp) /* Now recover expired state... */ if (test_and_clear_bit(NFS4CLNT_RECLAIM_NOGRACE, &clp->cl_state)) { status = nfs4_do_reclaim(clp, - nfs4_nograce_recovery_ops[clp->cl_minorversion]); + clp->cl_mvops->nograce_recovery_ops); if (test_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state) || test_bit(NFS4CLNT_SESSION_RESET, &clp->cl_state) || test_bit(NFS4CLNT_RECLAIM_REBOOT, &clp->cl_state)) diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index 65c8dae4b26..08ef9129113 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -202,14 +202,17 @@ static int nfs4_stat_to_errno(int); #define encode_link_maxsz (op_encode_hdr_maxsz + \ nfs4_name_maxsz) #define decode_link_maxsz (op_decode_hdr_maxsz + decode_change_info_maxsz) +#define encode_lockowner_maxsz (7) #define encode_lock_maxsz (op_encode_hdr_maxsz + \ 7 + \ - 1 + encode_stateid_maxsz + 8) + 1 + encode_stateid_maxsz + 1 + \ + encode_lockowner_maxsz) #define decode_lock_denied_maxsz \ (8 + decode_lockowner_maxsz) #define decode_lock_maxsz (op_decode_hdr_maxsz + \ decode_lock_denied_maxsz) -#define encode_lockt_maxsz (op_encode_hdr_maxsz + 12) +#define encode_lockt_maxsz (op_encode_hdr_maxsz + 5 + \ + encode_lockowner_maxsz) #define decode_lockt_maxsz (op_decode_hdr_maxsz + \ decode_lock_denied_maxsz) #define encode_locku_maxsz (op_encode_hdr_maxsz + 3 + \ @@ -217,6 +220,11 @@ static int nfs4_stat_to_errno(int); 4) #define decode_locku_maxsz (op_decode_hdr_maxsz + \ decode_stateid_maxsz) +#define encode_release_lockowner_maxsz \ + (op_encode_hdr_maxsz + \ + encode_lockowner_maxsz) +#define decode_release_lockowner_maxsz \ + (op_decode_hdr_maxsz) #define encode_access_maxsz (op_encode_hdr_maxsz + 1) #define decode_access_maxsz (op_decode_hdr_maxsz + 2) #define encode_symlink_maxsz (op_encode_hdr_maxsz + \ @@ -471,6 +479,12 @@ static int nfs4_stat_to_errno(int); decode_sequence_maxsz + \ decode_putfh_maxsz + \ decode_locku_maxsz) +#define NFS4_enc_release_lockowner_sz \ + (compound_encode_hdr_maxsz + \ + encode_lockowner_maxsz) +#define NFS4_dec_release_lockowner_sz \ + (compound_decode_hdr_maxsz + \ + decode_lockowner_maxsz) #define NFS4_enc_access_sz (compound_encode_hdr_maxsz + \ encode_sequence_maxsz + \ encode_putfh_maxsz + \ @@ -744,7 +758,7 @@ static void encode_compound_hdr(struct xdr_stream *xdr, struct compound_hdr *hdr) { __be32 *p; - struct rpc_auth *auth = req->rq_task->tk_msg.rpc_cred->cr_auth; + struct rpc_auth *auth = req->rq_cred->cr_auth; /* initialize running count of expected bytes in reply. * NOTE: the replied tag SHOULD be the same is the one sent, @@ -1042,6 +1056,17 @@ static inline uint64_t nfs4_lock_length(struct file_lock *fl) return fl->fl_end - fl->fl_start + 1; } +static void encode_lockowner(struct xdr_stream *xdr, const struct nfs_lowner *lowner) +{ + __be32 *p; + + p = reserve_space(xdr, 28); + p = xdr_encode_hyper(p, lowner->clientid); + *p++ = cpu_to_be32(16); + p = xdr_encode_opaque_fixed(p, "lock id:", 8); + xdr_encode_hyper(p, lowner->id); +} + /* * opcode,type,reclaim,offset,length,new_lock_owner = 32 * open_seqid,open_stateid,lock_seqid,lock_owner.clientid, lock_owner.id = 40 @@ -1058,14 +1083,11 @@ static void encode_lock(struct xdr_stream *xdr, const struct nfs_lock_args *args p = xdr_encode_hyper(p, nfs4_lock_length(args->fl)); *p = cpu_to_be32(args->new_lock_owner); if (args->new_lock_owner){ - p = reserve_space(xdr, 4+NFS4_STATEID_SIZE+32); + p = reserve_space(xdr, 4+NFS4_STATEID_SIZE+4); *p++ = cpu_to_be32(args->open_seqid->sequence->counter); p = xdr_encode_opaque_fixed(p, args->open_stateid->data, NFS4_STATEID_SIZE); *p++ = cpu_to_be32(args->lock_seqid->sequence->counter); - p = xdr_encode_hyper(p, args->lock_owner.clientid); - *p++ = cpu_to_be32(16); - p = xdr_encode_opaque_fixed(p, "lock id:", 8); - xdr_encode_hyper(p, args->lock_owner.id); + encode_lockowner(xdr, &args->lock_owner); } else { p = reserve_space(xdr, NFS4_STATEID_SIZE+4); @@ -1080,15 +1102,12 @@ static void encode_lockt(struct xdr_stream *xdr, const struct nfs_lockt_args *ar { __be32 *p; - p = reserve_space(xdr, 52); + p = reserve_space(xdr, 24); *p++ = cpu_to_be32(OP_LOCKT); *p++ = cpu_to_be32(nfs4_lock_type(args->fl, 0)); p = xdr_encode_hyper(p, args->fl->fl_start); p = xdr_encode_hyper(p, nfs4_lock_length(args->fl)); - p = xdr_encode_hyper(p, args->lock_owner.clientid); - *p++ = cpu_to_be32(16); - p = xdr_encode_opaque_fixed(p, "lock id:", 8); - xdr_encode_hyper(p, args->lock_owner.id); + encode_lockowner(xdr, &args->lock_owner); hdr->nops++; hdr->replen += decode_lockt_maxsz; } @@ -1108,6 +1127,17 @@ static void encode_locku(struct xdr_stream *xdr, const struct nfs_locku_args *ar hdr->replen += decode_locku_maxsz; } +static void encode_release_lockowner(struct xdr_stream *xdr, const struct nfs_lowner *lowner, struct compound_hdr *hdr) +{ + __be32 *p; + + p = reserve_space(xdr, 4); + *p = cpu_to_be32(OP_RELEASE_LOCKOWNER); + encode_lockowner(xdr, lowner); + hdr->nops++; + hdr->replen += decode_release_lockowner_maxsz; +} + static void encode_lookup(struct xdr_stream *xdr, const struct qstr *name, struct compound_hdr *hdr) { int len = name->len; @@ -1172,7 +1202,7 @@ static inline void encode_createmode(struct xdr_stream *xdr, const struct nfs_op break; default: clp = arg->server->nfs_client; - if (clp->cl_minorversion > 0) { + if (clp->cl_mvops->minor_version > 0) { if (nfs4_has_persistent_session(clp)) { *p = cpu_to_be32(NFS4_CREATE_GUARDED); encode_attrs(xdr, arg->u.attrs, arg->server); @@ -1324,14 +1354,14 @@ static void encode_putrootfh(struct xdr_stream *xdr, struct compound_hdr *hdr) hdr->replen += decode_putrootfh_maxsz; } -static void encode_stateid(struct xdr_stream *xdr, const struct nfs_open_context *ctx) +static void encode_stateid(struct xdr_stream *xdr, const struct nfs_open_context *ctx, const struct nfs_lock_context *l_ctx) { nfs4_stateid stateid; __be32 *p; p = reserve_space(xdr, NFS4_STATEID_SIZE); if (ctx->state != NULL) { - nfs4_copy_stateid(&stateid, ctx->state, ctx->lockowner); + nfs4_copy_stateid(&stateid, ctx->state, l_ctx->lockowner, l_ctx->pid); xdr_encode_opaque_fixed(p, stateid.data, NFS4_STATEID_SIZE); } else xdr_encode_opaque_fixed(p, zero_stateid.data, NFS4_STATEID_SIZE); @@ -1344,7 +1374,7 @@ static void encode_read(struct xdr_stream *xdr, const struct nfs_readargs *args, p = reserve_space(xdr, 4); *p = cpu_to_be32(OP_READ); - encode_stateid(xdr, args->context); + encode_stateid(xdr, args->context, args->lock_context); p = reserve_space(xdr, 12); p = xdr_encode_hyper(p, args->offset); @@ -1523,7 +1553,7 @@ static void encode_write(struct xdr_stream *xdr, const struct nfs_writeargs *arg p = reserve_space(xdr, 4); *p = cpu_to_be32(OP_WRITE); - encode_stateid(xdr, args->context); + encode_stateid(xdr, args->context, args->lock_context); p = reserve_space(xdr, 16); p = xdr_encode_hyper(p, args->offset); @@ -1704,7 +1734,7 @@ static u32 nfs4_xdr_minorversion(const struct nfs4_sequence_args *args) { #if defined(CONFIG_NFS_V4_1) if (args->sa_session) - return args->sa_session->clp->cl_minorversion; + return args->sa_session->clp->cl_mvops->minor_version; #endif /* CONFIG_NFS_V4_1 */ return 0; } @@ -2048,6 +2078,20 @@ static int nfs4_xdr_enc_locku(struct rpc_rqst *req, __be32 *p, struct nfs_locku_ return 0; } +static int nfs4_xdr_enc_release_lockowner(struct rpc_rqst *req, __be32 *p, struct nfs_release_lockowner_args *args) +{ + struct xdr_stream xdr; + struct compound_hdr hdr = { + .minorversion = 0, + }; + + xdr_init_encode(&xdr, &req->rq_snd_buf, p); + encode_compound_hdr(&xdr, req, &hdr); + encode_release_lockowner(&xdr, &args->lock_owner, &hdr); + encode_nops(&hdr); + return 0; +} + /* * Encode a READLINK request */ @@ -2395,7 +2439,7 @@ static int nfs4_xdr_enc_exchange_id(struct rpc_rqst *req, uint32_t *p, { struct xdr_stream xdr; struct compound_hdr hdr = { - .minorversion = args->client->cl_minorversion, + .minorversion = args->client->cl_mvops->minor_version, }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); @@ -2413,7 +2457,7 @@ static int nfs4_xdr_enc_create_session(struct rpc_rqst *req, uint32_t *p, { struct xdr_stream xdr; struct compound_hdr hdr = { - .minorversion = args->client->cl_minorversion, + .minorversion = args->client->cl_mvops->minor_version, }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); @@ -2431,7 +2475,7 @@ static int nfs4_xdr_enc_destroy_session(struct rpc_rqst *req, uint32_t *p, { struct xdr_stream xdr; struct compound_hdr hdr = { - .minorversion = session->clp->cl_minorversion, + .minorversion = session->clp->cl_mvops->minor_version, }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); @@ -3973,6 +4017,11 @@ static int decode_locku(struct xdr_stream *xdr, struct nfs_locku_res *res) return status; } +static int decode_release_lockowner(struct xdr_stream *xdr) +{ + return decode_op_hdr(xdr, OP_RELEASE_LOCKOWNER); +} + static int decode_lookup(struct xdr_stream *xdr) { return decode_op_hdr(xdr, OP_LOOKUP); @@ -5259,6 +5308,19 @@ out: return status; } +static int nfs4_xdr_dec_release_lockowner(struct rpc_rqst *rqstp, __be32 *p, void *dummy) +{ + 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) + status = decode_release_lockowner(&xdr); + return status; +} + /* * Decode READLINK response */ @@ -5866,6 +5928,7 @@ struct rpc_procinfo nfs4_procedures[] = { PROC(GETACL, enc_getacl, dec_getacl), PROC(SETACL, enc_setacl, dec_setacl), PROC(FS_LOCATIONS, enc_fs_locations, dec_fs_locations), + PROC(RELEASE_LOCKOWNER, enc_release_lockowner, dec_release_lockowner), #if defined(CONFIG_NFS_V4_1) PROC(EXCHANGE_ID, enc_exchange_id, dec_exchange_id), PROC(CREATE_SESSION, enc_create_session, dec_create_session), diff --git a/fs/nfs/pagelist.c b/fs/nfs/pagelist.c index a3654e57b58..919490232e1 100644 --- a/fs/nfs/pagelist.c +++ b/fs/nfs/pagelist.c @@ -79,6 +79,7 @@ nfs_create_request(struct nfs_open_context *ctx, struct inode *inode, req->wb_pgbase = offset; req->wb_bytes = count; req->wb_context = get_nfs_open_context(ctx); + req->wb_lock_context = nfs_get_lock_context(ctx); kref_init(&req->wb_kref); return req; } @@ -141,11 +142,16 @@ void nfs_clear_request(struct nfs_page *req) { struct page *page = req->wb_page; struct nfs_open_context *ctx = req->wb_context; + struct nfs_lock_context *l_ctx = req->wb_lock_context; if (page != NULL) { page_cache_release(page); req->wb_page = NULL; } + if (l_ctx != NULL) { + nfs_put_lock_context(l_ctx); + req->wb_lock_context = NULL; + } if (ctx != NULL) { put_nfs_open_context(ctx); req->wb_context = NULL; @@ -235,7 +241,7 @@ static int nfs_can_coalesce_requests(struct nfs_page *prev, { if (req->wb_context->cred != prev->wb_context->cred) return 0; - if (req->wb_context->lockowner != prev->wb_context->lockowner) + if (req->wb_lock_context->lockowner != prev->wb_lock_context->lockowner) return 0; if (req->wb_context->state != prev->wb_context->state) return 0; diff --git a/fs/nfs/read.c b/fs/nfs/read.c index 6e2b06e6ca7..87adc274424 100644 --- a/fs/nfs/read.c +++ b/fs/nfs/read.c @@ -190,6 +190,7 @@ static int nfs_read_rpcsetup(struct nfs_page *req, struct nfs_read_data *data, data->args.pages = data->pagevec; data->args.count = count; data->args.context = get_nfs_open_context(req->wb_context); + data->args.lock_context = req->wb_lock_context; data->res.fattr = &data->fattr; data->res.count = count; @@ -410,7 +411,7 @@ void nfs_read_prepare(struct rpc_task *task, void *calldata) { struct nfs_read_data *data = calldata; - if (nfs4_setup_sequence(NFS_SERVER(data->inode)->nfs_client, + if (nfs4_setup_sequence(NFS_SERVER(data->inode), &data->args.seq_args, &data->res.seq_res, 0, task)) return; diff --git a/fs/nfs/super.c b/fs/nfs/super.c index f9df16de4a5..f1ae39f6cb0 100644 --- a/fs/nfs/super.c +++ b/fs/nfs/super.c @@ -546,6 +546,9 @@ static void nfs_show_mountd_options(struct seq_file *m, struct nfs_server *nfss, { struct sockaddr *sap = (struct sockaddr *)&nfss->mountd_address; + if (nfss->flags & NFS_MOUNT_LEGACY_INTERFACE) + return; + switch (sap->sa_family) { case AF_INET: { struct sockaddr_in *sin = (struct sockaddr_in *)sap; @@ -1780,6 +1783,7 @@ static int nfs_validate_mount_data(void *options, * can deal with. */ args->flags = data->flags & NFS_MOUNT_FLAGMASK; + args->flags |= NFS_MOUNT_LEGACY_INTERFACE; args->rsize = data->rsize; args->wsize = data->wsize; args->timeo = data->timeo; diff --git a/fs/nfs/unlink.c b/fs/nfs/unlink.c index a2242af6a17..2f84adaad42 100644 --- a/fs/nfs/unlink.c +++ b/fs/nfs/unlink.c @@ -110,7 +110,7 @@ void nfs_unlink_prepare(struct rpc_task *task, void *calldata) struct nfs_unlinkdata *data = calldata; struct nfs_server *server = NFS_SERVER(data->dir); - if (nfs4_setup_sequence(server->nfs_client, &data->args.seq_args, + if (nfs4_setup_sequence(server, &data->args.seq_args, &data->res.seq_res, 1, task)) return; rpc_call_start(task); diff --git a/fs/nfs/write.c b/fs/nfs/write.c index 9f81bdd91c5..874972d9427 100644 --- a/fs/nfs/write.c +++ b/fs/nfs/write.c @@ -700,7 +700,9 @@ int nfs_flush_incompatible(struct file *file, struct page *page) req = nfs_page_find_request(page); if (req == NULL) return 0; - do_flush = req->wb_page != page || req->wb_context != ctx; + do_flush = req->wb_page != page || req->wb_context != ctx || + req->wb_lock_context->lockowner != current->files || + req->wb_lock_context->pid != current->tgid; nfs_release_request(req); if (!do_flush) return 0; @@ -824,6 +826,7 @@ static int nfs_write_rpcsetup(struct nfs_page *req, data->args.pages = data->pagevec; data->args.count = count; data->args.context = get_nfs_open_context(req->wb_context); + data->args.lock_context = req->wb_lock_context; data->args.stable = NFS_UNSTABLE; if (how & FLUSH_STABLE) { data->args.stable = NFS_DATA_SYNC; @@ -1047,9 +1050,9 @@ out: void nfs_write_prepare(struct rpc_task *task, void *calldata) { struct nfs_write_data *data = calldata; - struct nfs_client *clp = (NFS_SERVER(data->inode))->nfs_client; - if (nfs4_setup_sequence(clp, &data->args.seq_args, + if (nfs4_setup_sequence(NFS_SERVER(data->inode), + &data->args.seq_args, &data->res.seq_res, 1, task)) return; rpc_call_start(task); diff --git a/fs/nfsd/nfs3proc.c b/fs/nfsd/nfs3proc.c index 3d68f45a37b..5b7e3021e06 100644 --- a/fs/nfsd/nfs3proc.c +++ b/fs/nfsd/nfs3proc.c @@ -168,7 +168,7 @@ nfsd3_proc_read(struct svc_rqst *rqstp, struct nfsd3_readargs *argp, svc_reserve_auth(rqstp, ((1 + NFS3_POST_OP_ATTR_WORDS + 3)<<2) + resp->count +4); fh_copy(&resp->fh, &argp->fh); - nfserr = nfsd_read(rqstp, &resp->fh, NULL, + nfserr = nfsd_read(rqstp, &resp->fh, argp->offset, rqstp->rq_vec, argp->vlen, &resp->count); @@ -271,7 +271,7 @@ nfsd3_proc_mkdir(struct svc_rqst *rqstp, struct nfsd3_createargs *argp, fh_init(&resp->fh, NFS3_FHSIZE); nfserr = nfsd_create(rqstp, &resp->dirfh, argp->name, argp->len, &argp->attrs, S_IFDIR, 0, &resp->fh); - + fh_unlock(&resp->dirfh); RETURN_STATUS(nfserr); } @@ -327,7 +327,7 @@ nfsd3_proc_mknod(struct svc_rqst *rqstp, struct nfsd3_mknodargs *argp, type = nfs3_ftypes[argp->ftype]; nfserr = nfsd_create(rqstp, &resp->dirfh, argp->name, argp->len, &argp->attrs, type, rdev, &resp->fh); - + fh_unlock(&resp->dirfh); RETURN_STATUS(nfserr); } @@ -348,6 +348,7 @@ nfsd3_proc_remove(struct svc_rqst *rqstp, struct nfsd3_diropargs *argp, /* Unlink. -S_IFDIR means file must not be a directory */ fh_copy(&resp->fh, &argp->fh); nfserr = nfsd_unlink(rqstp, &resp->fh, -S_IFDIR, argp->name, argp->len); + fh_unlock(&resp->fh); RETURN_STATUS(nfserr); } @@ -367,6 +368,7 @@ nfsd3_proc_rmdir(struct svc_rqst *rqstp, struct nfsd3_diropargs *argp, fh_copy(&resp->fh, &argp->fh); nfserr = nfsd_unlink(rqstp, &resp->fh, S_IFDIR, argp->name, argp->len); + fh_unlock(&resp->fh); RETURN_STATUS(nfserr); } diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c index eb78e7e2207..988cbb3a19b 100644 --- a/fs/nfsd/nfs4callback.c +++ b/fs/nfsd/nfs4callback.c @@ -143,8 +143,6 @@ struct nfs4_cb_compound_hdr { u32 minorversion; /* res */ int status; - u32 taglen; - char *tag; }; static struct { @@ -205,6 +203,16 @@ nfs_cb_stat_to_errno(int stat) */ static void +encode_stateid(struct xdr_stream *xdr, stateid_t *sid) +{ + __be32 *p; + + RESERVE_SPACE(sizeof(stateid_t)); + WRITE32(sid->si_generation); + WRITEMEM(&sid->si_opaque, sizeof(stateid_opaque_t)); +} + +static void encode_cb_compound_hdr(struct xdr_stream *xdr, struct nfs4_cb_compound_hdr *hdr) { __be32 * p; @@ -229,10 +237,10 @@ encode_cb_recall(struct xdr_stream *xdr, struct nfs4_delegation *dp, __be32 *p; int len = dp->dl_fh.fh_size; - RESERVE_SPACE(12+sizeof(dp->dl_stateid) + len); + RESERVE_SPACE(4); WRITE32(OP_CB_RECALL); - WRITE32(dp->dl_stateid.si_generation); - WRITEMEM(&dp->dl_stateid.si_opaque, sizeof(stateid_opaque_t)); + encode_stateid(xdr, &dp->dl_stateid); + RESERVE_SPACE(8 + (XDR_QUADLEN(len) << 2)); WRITE32(0); /* truncate optimization not implemented */ WRITE32(len); WRITEMEM(&dp->dl_fh.fh_base, len); @@ -293,13 +301,14 @@ nfs4_xdr_enc_cb_recall(struct rpc_rqst *req, __be32 *p, static int decode_cb_compound_hdr(struct xdr_stream *xdr, struct nfs4_cb_compound_hdr *hdr){ __be32 *p; + u32 taglen; READ_BUF(8); READ32(hdr->status); - READ32(hdr->taglen); - READ_BUF(hdr->taglen + 4); - hdr->tag = (char *)p; - p += XDR_QUADLEN(hdr->taglen); + /* We've got no use for the tag; ignore it: */ + READ32(taglen); + READ_BUF(taglen + 4); + p += XDR_QUADLEN(taglen); READ32(hdr->nops); return 0; } @@ -667,28 +676,28 @@ static void nfsd4_cb_recall_done(struct rpc_task *task, void *calldata) } switch (task->tk_status) { - case -EIO: + case 0: + return; + case -EBADHANDLE: + case -NFS4ERR_BAD_STATEID: + /* Race: client probably got cb_recall + * before open reply granting delegation */ + break; + default: /* Network partition? */ atomic_set(&clp->cl_cb_set, 0); warn_no_callback_path(clp, task->tk_status); if (current_rpc_client != task->tk_client) { /* queue a callback on the new connection: */ + atomic_inc(&dp->dl_count); nfsd4_cb_recall(dp); return; } - case -EBADHANDLE: - case -NFS4ERR_BAD_STATEID: - /* Race: client probably got cb_recall - * before open reply granting delegation */ - break; - default: - /* success, or error we can't handle */ - return; } if (dp->dl_retries--) { rpc_delay(task, 2*HZ); task->tk_status = 0; - rpc_restart_call(task); + rpc_restart_call_prepare(task); return; } else { atomic_set(&clp->cl_cb_set, 0); @@ -752,18 +761,16 @@ static void _nfsd4_cb_recall(struct nfs4_delegation *dp) .rpc_proc = &nfs4_cb_procedures[NFSPROC4_CLNT_CB_RECALL], .rpc_cred = callback_cred }; - int status; - if (clnt == NULL) + if (clnt == NULL) { + nfs4_put_delegation(dp); return; /* Client is shutting down; give up. */ + } args->args_op = dp; msg.rpc_argp = args; dp->dl_retries = 1; - status = rpc_call_async(clnt, &msg, RPC_TASK_SOFT, - &nfsd4_cb_recall_ops, dp); - if (status) - nfs4_put_delegation(dp); + rpc_call_async(clnt, &msg, RPC_TASK_SOFT, &nfsd4_cb_recall_ops, dp); } void nfsd4_do_callback_rpc(struct work_struct *w) diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 4a273475877..2e7357104cf 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -51,7 +51,6 @@ static time_t boot_time; static u32 current_ownerid = 1; static u32 current_fileid = 1; static u32 current_delegid = 1; -static u32 nfs4_init; static stateid_t zerostateid; /* bits all 0 */ static stateid_t onestateid; /* bits all 1 */ static u64 current_sessionid = 1; @@ -163,6 +162,46 @@ static struct list_head ownerstr_hashtbl[OWNER_HASH_SIZE]; static struct list_head file_hashtbl[FILE_HASH_SIZE]; static struct list_head stateid_hashtbl[STATEID_HASH_SIZE]; +static void __nfs4_file_get_access(struct nfs4_file *fp, int oflag) +{ + BUG_ON(!(fp->fi_fds[oflag] || fp->fi_fds[O_RDWR])); + atomic_inc(&fp->fi_access[oflag]); +} + +static void nfs4_file_get_access(struct nfs4_file *fp, int oflag) +{ + if (oflag == O_RDWR) { + __nfs4_file_get_access(fp, O_RDONLY); + __nfs4_file_get_access(fp, O_WRONLY); + } else + __nfs4_file_get_access(fp, oflag); +} + +static void nfs4_file_put_fd(struct nfs4_file *fp, int oflag) +{ + if (fp->fi_fds[oflag]) { + fput(fp->fi_fds[oflag]); + fp->fi_fds[oflag] = NULL; + } +} + +static void __nfs4_file_put_access(struct nfs4_file *fp, int oflag) +{ + if (atomic_dec_and_test(&fp->fi_access[oflag])) { + nfs4_file_put_fd(fp, O_RDWR); + nfs4_file_put_fd(fp, oflag); + } +} + +static void nfs4_file_put_access(struct nfs4_file *fp, int oflag) +{ + if (oflag == O_RDWR) { + __nfs4_file_put_access(fp, O_RDONLY); + __nfs4_file_put_access(fp, O_WRONLY); + } else + __nfs4_file_put_access(fp, oflag); +} + static struct nfs4_delegation * alloc_init_deleg(struct nfs4_client *clp, struct nfs4_stateid *stp, struct svc_fh *current_fh, u32 type) { @@ -171,6 +210,13 @@ alloc_init_deleg(struct nfs4_client *clp, struct nfs4_stateid *stp, struct svc_f struct nfs4_cb_conn *cb = &stp->st_stateowner->so_client->cl_cb_conn; dprintk("NFSD alloc_init_deleg\n"); + /* + * Major work on the lease subsystem (for example, to support + * calbacks on stat) will be required before we can support + * write delegations properly. + */ + if (type != NFS4_OPEN_DELEGATE_READ) + return NULL; if (fp->fi_had_conflict) return NULL; if (num_delegations > max_delegations) @@ -185,9 +231,8 @@ alloc_init_deleg(struct nfs4_client *clp, struct nfs4_stateid *stp, struct svc_f dp->dl_client = clp; get_nfs4_file(fp); dp->dl_file = fp; + nfs4_file_get_access(fp, O_RDONLY); dp->dl_flock = NULL; - get_file(stp->st_vfs_file); - dp->dl_vfs_file = stp->st_vfs_file; dp->dl_type = type; dp->dl_ident = cb->cb_ident; dp->dl_stateid.si_boot = boot_time; @@ -222,15 +267,12 @@ nfs4_put_delegation(struct nfs4_delegation *dp) static void nfs4_close_delegation(struct nfs4_delegation *dp) { - struct file *filp = dp->dl_vfs_file; + struct file *filp = find_readable_file(dp->dl_file); dprintk("NFSD: close_delegation dp %p\n",dp); - dp->dl_vfs_file = NULL; - /* The following nfsd_close may not actually close the file, - * but we want to remove the lease in any case. */ if (dp->dl_flock) vfs_setlease(filp, F_UNLCK, &dp->dl_flock); - nfsd_close(filp); + nfs4_file_put_access(dp->dl_file, O_RDONLY); } /* Called under the state lock. */ @@ -302,8 +344,12 @@ static void free_generic_stateid(struct nfs4_stateid *stp) static void release_lock_stateid(struct nfs4_stateid *stp) { + struct file *file; + unhash_generic_stateid(stp); - locks_remove_posix(stp->st_vfs_file, (fl_owner_t)stp->st_stateowner); + file = find_any_file(stp->st_file); + if (file) + locks_remove_posix(file, (fl_owner_t)stp->st_stateowner); free_generic_stateid(stp); } @@ -341,11 +387,85 @@ release_stateid_lockowners(struct nfs4_stateid *open_stp) } } +/* + * We store the NONE, READ, WRITE, and BOTH bits separately in the + * st_{access,deny}_bmap field of the stateid, in order to track not + * only what share bits are currently in force, but also what + * combinations of share bits previous opens have used. This allows us + * to enforce the recommendation of rfc 3530 14.2.19 that the server + * return an error if the client attempt to downgrade to a combination + * of share bits not explicable by closing some of its previous opens. + * + * XXX: This enforcement is actually incomplete, since we don't keep + * track of access/deny bit combinations; so, e.g., we allow: + * + * OPEN allow read, deny write + * OPEN allow both, deny none + * DOWNGRADE allow read, deny none + * + * which we should reject. + */ +static void +set_access(unsigned int *access, unsigned long bmap) { + int i; + + *access = 0; + for (i = 1; i < 4; i++) { + if (test_bit(i, &bmap)) + *access |= i; + } +} + +static void +set_deny(unsigned int *deny, unsigned long bmap) { + int i; + + *deny = 0; + for (i = 0; i < 4; i++) { + if (test_bit(i, &bmap)) + *deny |= i ; + } +} + +static int +test_share(struct nfs4_stateid *stp, struct nfsd4_open *open) { + unsigned int access, deny; + + set_access(&access, stp->st_access_bmap); + set_deny(&deny, stp->st_deny_bmap); + if ((access & open->op_share_deny) || (deny & open->op_share_access)) + return 0; + return 1; +} + +static int nfs4_access_to_omode(u32 access) +{ + switch (access) { + case NFS4_SHARE_ACCESS_READ: + return O_RDONLY; + case NFS4_SHARE_ACCESS_WRITE: + return O_WRONLY; + case NFS4_SHARE_ACCESS_BOTH: + return O_RDWR; + } + BUG(); +} + +static int nfs4_access_bmap_to_omode(struct nfs4_stateid *stp) +{ + unsigned int access; + + set_access(&access, stp->st_access_bmap); + return nfs4_access_to_omode(access); +} + static void release_open_stateid(struct nfs4_stateid *stp) { + int oflag = nfs4_access_bmap_to_omode(stp); + unhash_generic_stateid(stp); release_stateid_lockowners(stp); - nfsd_close(stp->st_vfs_file); + nfs4_file_put_access(stp->st_file, oflag); free_generic_stateid(stp); } @@ -457,7 +577,7 @@ static int set_forechannel_drc_size(struct nfsd4_channel_attrs *fchan) spin_unlock(&nfsd_drc_lock); if (fchan->maxreqs == 0) - return nfserr_serverfault; + return nfserr_jukebox; fchan->maxresp_cached = size + NFSD_MIN_HDR_SEQ_SZ; return 0; @@ -542,7 +662,7 @@ alloc_init_session(struct svc_rqst *rqstp, struct nfs4_client *clp, BUILD_BUG_ON(NFSD_MAX_SLOTS_PER_SESSION * sizeof(struct nfsd4_slot) + sizeof(struct nfsd4_session) > PAGE_SIZE); - status = nfserr_serverfault; + status = nfserr_jukebox; /* allocate struct nfsd4_session and slot table pointers in one piece */ slotsize = tmp.se_fchannel.maxreqs * sizeof(struct nfsd4_slot *); new = kzalloc(sizeof(*new) + slotsize, GFP_KERNEL); @@ -591,10 +711,8 @@ find_in_sessionid_hashtbl(struct nfs4_sessionid *sessionid) dump_sessionid(__func__, sessionid); idx = hash_sessionid(sessionid); - dprintk("%s: idx is %d\n", __func__, idx); /* Search in the appropriate list */ list_for_each_entry(elem, &sessionid_hashtbl[idx], se_hash) { - dump_sessionid("list traversal", &elem->se_sessionid); if (!memcmp(elem->se_sessionid.data, sessionid->data, NFS4_MAX_SESSIONID_LEN)) { return elem; @@ -714,7 +832,6 @@ release_session_client(struct nfsd4_session *session) } else renew_client_locked(clp); spin_unlock(&client_lock); - nfsd4_put_session(session); } /* must be called under the client_lock */ @@ -1220,7 +1337,7 @@ out_new: /* Normal case */ new = create_client(exid->clname, dname, rqstp, &verf); if (new == NULL) { - status = nfserr_serverfault; + status = nfserr_jukebox; goto out; } @@ -1760,6 +1877,8 @@ alloc_init_file(struct inode *ino) fp->fi_inode = igrab(ino); fp->fi_id = current_fileid++; fp->fi_had_conflict = false; + memset(fp->fi_fds, 0, sizeof(fp->fi_fds)); + memset(fp->fi_access, 0, sizeof(fp->fi_access)); spin_lock(&recall_lock); list_add(&fp->fi_hash, &file_hashtbl[hashval]); spin_unlock(&recall_lock); @@ -1971,57 +2090,6 @@ static inline int deny_valid(u32 x) } /* - * We store the NONE, READ, WRITE, and BOTH bits separately in the - * st_{access,deny}_bmap field of the stateid, in order to track not - * only what share bits are currently in force, but also what - * combinations of share bits previous opens have used. This allows us - * to enforce the recommendation of rfc 3530 14.2.19 that the server - * return an error if the client attempt to downgrade to a combination - * of share bits not explicable by closing some of its previous opens. - * - * XXX: This enforcement is actually incomplete, since we don't keep - * track of access/deny bit combinations; so, e.g., we allow: - * - * OPEN allow read, deny write - * OPEN allow both, deny none - * DOWNGRADE allow read, deny none - * - * which we should reject. - */ -static void -set_access(unsigned int *access, unsigned long bmap) { - int i; - - *access = 0; - for (i = 1; i < 4; i++) { - if (test_bit(i, &bmap)) - *access |= i; - } -} - -static void -set_deny(unsigned int *deny, unsigned long bmap) { - int i; - - *deny = 0; - for (i = 0; i < 4; i++) { - if (test_bit(i, &bmap)) - *deny |= i ; - } -} - -static int -test_share(struct nfs4_stateid *stp, struct nfsd4_open *open) { - unsigned int access, deny; - - set_access(&access, stp->st_access_bmap); - set_deny(&deny, stp->st_deny_bmap); - if ((access & open->op_share_deny) || (deny & open->op_share_access)) - return 0; - return 1; -} - -/* * Called to check deny when READ with all zero stateid or * WRITE with all zero or all one stateid */ @@ -2052,14 +2120,12 @@ out: } static inline void -nfs4_file_downgrade(struct file *filp, unsigned int share_access) +nfs4_file_downgrade(struct nfs4_file *fp, unsigned int share_access) { - if (share_access & NFS4_SHARE_ACCESS_WRITE) { - drop_file_write_access(filp); - spin_lock(&filp->f_lock); - filp->f_mode = (filp->f_mode | FMODE_READ) & ~FMODE_WRITE; - spin_unlock(&filp->f_lock); - } + if (share_access & NFS4_SHARE_ACCESS_WRITE) + nfs4_file_put_access(fp, O_WRONLY); + if (share_access & NFS4_SHARE_ACCESS_READ) + nfs4_file_put_access(fp, O_RDONLY); } /* @@ -2255,6 +2321,13 @@ find_delegation_file(struct nfs4_file *fp, stateid_t *stid) return NULL; } +int share_access_to_flags(u32 share_access) +{ + share_access &= ~NFS4_SHARE_WANT_MASK; + + return share_access == NFS4_SHARE_ACCESS_READ ? RD_STATE : WR_STATE; +} + static __be32 nfs4_check_deleg(struct nfs4_file *fp, struct nfsd4_open *open, struct nfs4_delegation **dp) @@ -2265,8 +2338,7 @@ nfs4_check_deleg(struct nfs4_file *fp, struct nfsd4_open *open, *dp = find_delegation_file(fp, &open->op_delegate_stateid); if (*dp == NULL) goto out; - flags = open->op_share_access == NFS4_SHARE_ACCESS_READ ? - RD_STATE : WR_STATE; + flags = share_access_to_flags(open->op_share_access); status = nfs4_check_delegmode(*dp, flags); if (status) *dp = NULL; @@ -2308,30 +2380,53 @@ nfs4_alloc_stateid(void) return kmem_cache_alloc(stateid_slab, GFP_KERNEL); } +static inline int nfs4_access_to_access(u32 nfs4_access) +{ + int flags = 0; + + if (nfs4_access & NFS4_SHARE_ACCESS_READ) + flags |= NFSD_MAY_READ; + if (nfs4_access & NFS4_SHARE_ACCESS_WRITE) + flags |= NFSD_MAY_WRITE; + return flags; +} + +static __be32 nfs4_get_vfs_file(struct svc_rqst *rqstp, struct nfs4_file +*fp, struct svc_fh *cur_fh, u32 nfs4_access) +{ + __be32 status; + int oflag = nfs4_access_to_omode(nfs4_access); + int access = nfs4_access_to_access(nfs4_access); + + if (!fp->fi_fds[oflag]) { + status = nfsd_open(rqstp, cur_fh, S_IFREG, access, + &fp->fi_fds[oflag]); + if (status == nfserr_dropit) + status = nfserr_jukebox; + if (status) + return status; + } + nfs4_file_get_access(fp, oflag); + + return nfs_ok; +} + static __be32 nfs4_new_open(struct svc_rqst *rqstp, struct nfs4_stateid **stpp, - struct nfs4_delegation *dp, - struct svc_fh *cur_fh, int flags) + struct nfs4_file *fp, struct svc_fh *cur_fh, + struct nfsd4_open *open) { struct nfs4_stateid *stp; + __be32 status; stp = nfs4_alloc_stateid(); if (stp == NULL) return nfserr_resource; - if (dp) { - get_file(dp->dl_vfs_file); - stp->st_vfs_file = dp->dl_vfs_file; - } else { - __be32 status; - status = nfsd_open(rqstp, cur_fh, S_IFREG, flags, - &stp->st_vfs_file); - if (status) { - if (status == nfserr_dropit) - status = nfserr_jukebox; - kmem_cache_free(stateid_slab, stp); - return status; - } + status = nfs4_get_vfs_file(rqstp, fp, cur_fh, open->op_share_access); + if (status) { + kmem_cache_free(stateid_slab, stp); + return status; } *stpp = stp; return 0; @@ -2353,35 +2448,30 @@ nfsd4_truncate(struct svc_rqst *rqstp, struct svc_fh *fh, } static __be32 -nfs4_upgrade_open(struct svc_rqst *rqstp, struct svc_fh *cur_fh, struct nfs4_stateid *stp, struct nfsd4_open *open) +nfs4_upgrade_open(struct svc_rqst *rqstp, struct nfs4_file *fp, struct svc_fh *cur_fh, struct nfs4_stateid *stp, struct nfsd4_open *open) { - struct file *filp = stp->st_vfs_file; - struct inode *inode = filp->f_path.dentry->d_inode; - unsigned int share_access, new_writer; + u32 op_share_access, new_access; __be32 status; - set_access(&share_access, stp->st_access_bmap); - new_writer = (~share_access) & open->op_share_access - & NFS4_SHARE_ACCESS_WRITE; - - if (new_writer) { - int err = get_write_access(inode); - if (err) - return nfserrno(err); - err = mnt_want_write(cur_fh->fh_export->ex_path.mnt); - if (err) - return nfserrno(err); - file_take_write(filp); + set_access(&new_access, stp->st_access_bmap); + new_access = (~new_access) & open->op_share_access & ~NFS4_SHARE_WANT_MASK; + + if (new_access) { + status = nfs4_get_vfs_file(rqstp, fp, cur_fh, new_access); + if (status) + return status; } status = nfsd4_truncate(rqstp, cur_fh, open); if (status) { - if (new_writer) - put_write_access(inode); + if (new_access) { + int oflag = nfs4_access_to_omode(new_access); + nfs4_file_put_access(fp, oflag); + } return status; } /* remember the open */ - filp->f_mode |= open->op_share_access; - __set_bit(open->op_share_access, &stp->st_access_bmap); + op_share_access = open->op_share_access & ~NFS4_SHARE_WANT_MASK; + __set_bit(op_share_access, &stp->st_access_bmap); __set_bit(open->op_share_deny, &stp->st_deny_bmap); return nfs_ok; @@ -2444,13 +2534,14 @@ nfs4_open_delegation(struct svc_fh *fh, struct nfsd4_open *open, struct nfs4_sta fl.fl_type = flag == NFS4_OPEN_DELEGATE_READ? F_RDLCK: F_WRLCK; fl.fl_end = OFFSET_MAX; fl.fl_owner = (fl_owner_t)dp; - fl.fl_file = stp->st_vfs_file; + fl.fl_file = find_readable_file(stp->st_file); + BUG_ON(!fl.fl_file); fl.fl_pid = current->tgid; /* vfs_setlease checks to see if delegation should be handed out. * the lock_manager callbacks fl_mylease and fl_change are used */ - if ((status = vfs_setlease(stp->st_vfs_file, fl.fl_type, &flp))) { + if ((status = vfs_setlease(fl.fl_file, fl.fl_type, &flp))) { dprintk("NFSD: setlease failed [%d], no delegation\n", status); unhash_delegation(dp); flag = NFS4_OPEN_DELEGATE_NONE; @@ -2514,18 +2605,12 @@ nfsd4_process_open2(struct svc_rqst *rqstp, struct svc_fh *current_fh, struct nf */ if (stp) { /* Stateid was found, this is an OPEN upgrade */ - status = nfs4_upgrade_open(rqstp, current_fh, stp, open); + status = nfs4_upgrade_open(rqstp, fp, current_fh, stp, open); if (status) goto out; update_stateid(&stp->st_stateid); } else { - /* Stateid was not found, this is a new OPEN */ - int flags = 0; - if (open->op_share_access & NFS4_SHARE_ACCESS_READ) - flags |= NFSD_MAY_READ; - if (open->op_share_access & NFS4_SHARE_ACCESS_WRITE) - flags |= NFSD_MAY_WRITE; - status = nfs4_new_open(rqstp, &stp, dp, current_fh, flags); + status = nfs4_new_open(rqstp, &stp, fp, current_fh, open); if (status) goto out; init_stateid(stp, fp, open); @@ -2727,7 +2812,7 @@ search_close_lru(u32 st_id, int flags) static inline int nfs4_check_fh(struct svc_fh *fhp, struct nfs4_stateid *stp) { - return fhp->fh_dentry->d_inode != stp->st_vfs_file->f_path.dentry->d_inode; + return fhp->fh_dentry->d_inode != stp->st_file->fi_inode; } static int @@ -2760,6 +2845,9 @@ __be32 nfs4_check_openmode(struct nfs4_stateid *stp, int flags) { __be32 status = nfserr_openmode; + /* For lock stateid's, we test the parent open, not the lock: */ + if (stp->st_openstp) + stp = stp->st_openstp; if ((flags & WR_STATE) && (!access_permit_write(stp->st_access_bmap))) goto out; if ((flags & RD_STATE) && (!access_permit_read(stp->st_access_bmap))) @@ -2872,7 +2960,8 @@ nfs4_preprocess_stateid_op(struct nfsd4_compound_state *cstate, goto out; renew_client(dp->dl_client); if (filpp) - *filpp = dp->dl_vfs_file; + *filpp = find_readable_file(dp->dl_file); + BUG_ON(!*filpp); } else { /* open or lock stateid */ stp = find_stateid(stateid, flags); if (!stp) @@ -2889,8 +2978,13 @@ nfs4_preprocess_stateid_op(struct nfsd4_compound_state *cstate, if (status) goto out; renew_client(stp->st_stateowner->so_client); - if (filpp) - *filpp = stp->st_vfs_file; + if (filpp) { + if (flags & RD_STATE) + *filpp = find_readable_file(stp->st_file); + else + *filpp = find_writeable_file(stp->st_file); + BUG_ON(!*filpp); /* assured by check_openmode */ + } } status = nfs_ok; out: @@ -3126,8 +3220,7 @@ nfsd4_open_downgrade(struct svc_rqst *rqstp, goto out; } set_access(&share_access, stp->st_access_bmap); - nfs4_file_downgrade(stp->st_vfs_file, - share_access & ~od->od_share_access); + nfs4_file_downgrade(stp->st_file, share_access & ~od->od_share_access); reset_union_bmap_access(od->od_share_access, &stp->st_access_bmap); reset_union_bmap_deny(od->od_share_deny, &stp->st_deny_bmap); @@ -3346,11 +3439,9 @@ static inline void nfs4_set_lock_denied(struct file_lock *fl, struct nfsd4_lock_denied *deny) { struct nfs4_stateowner *sop; - unsigned int hval; if (fl->fl_lmops == &nfsd_posix_mng_ops) { sop = (struct nfs4_stateowner *) fl->fl_owner; - hval = lockownerid_hashval(sop->so_id); kref_get(&sop->so_ref); deny->ld_sop = sop; deny->ld_clientid = sop->so_client->cl_clientid; @@ -3446,8 +3537,6 @@ alloc_init_lock_stateid(struct nfs4_stateowner *sop, struct nfs4_file *fp, struc stp->st_stateid.si_stateownerid = sop->so_id; stp->st_stateid.si_fileid = fp->fi_id; stp->st_stateid.si_generation = 0; - stp->st_vfs_file = open_stp->st_vfs_file; /* FIXME refcount?? */ - stp->st_access_bmap = open_stp->st_access_bmap; stp->st_deny_bmap = open_stp->st_deny_bmap; stp->st_openstp = open_stp; @@ -3547,7 +3636,6 @@ nfsd4_lock(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, lock_sop = lock->lk_replay_owner; } /* lock->lk_replay_owner and lock_stp have been created or found */ - filp = lock_stp->st_vfs_file; status = nfserr_grace; if (locks_in_grace() && !lock->lk_reclaim) @@ -3560,11 +3648,13 @@ nfsd4_lock(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, switch (lock->lk_type) { case NFS4_READ_LT: case NFS4_READW_LT: + filp = find_readable_file(lock_stp->st_file); file_lock.fl_type = F_RDLCK; cmd = F_SETLK; break; case NFS4_WRITE_LT: case NFS4_WRITEW_LT: + filp = find_writeable_file(lock_stp->st_file); file_lock.fl_type = F_WRLCK; cmd = F_SETLK; break; @@ -3572,6 +3662,10 @@ nfsd4_lock(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, status = nfserr_inval; goto out; } + if (!filp) { + status = nfserr_openmode; + goto out; + } file_lock.fl_owner = (fl_owner_t)lock_sop; file_lock.fl_pid = current->tgid; file_lock.fl_file = filp; @@ -3740,7 +3834,11 @@ nfsd4_locku(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, &locku->lu_stateowner, &stp, NULL))) goto out; - filp = stp->st_vfs_file; + filp = find_any_file(stp->st_file); + if (!filp) { + status = nfserr_lock_range; + goto out; + } BUG_ON(!filp); locks_init_lock(&file_lock); file_lock.fl_type = F_UNLCK; @@ -3787,10 +3885,10 @@ out_nfserr: * 0: no locks held by lockowner */ static int -check_for_locks(struct file *filp, struct nfs4_stateowner *lowner) +check_for_locks(struct nfs4_file *filp, struct nfs4_stateowner *lowner) { struct file_lock **flpp; - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = filp->fi_inode; int status = 0; lock_kernel(); @@ -3841,7 +3939,7 @@ nfsd4_release_lockowner(struct svc_rqst *rqstp, continue; list_for_each_entry(stp, &sop->so_stateids, st_perstateowner) { - if (check_for_locks(stp->st_vfs_file, sop)) + if (check_for_locks(stp->st_file, sop)) goto out; /* Note: so_perclient unused for lockowners, * so it's OK to fool with here. */ @@ -4066,16 +4164,8 @@ out_free_laundry: int nfs4_state_start(void) { - int ret; - - if (nfs4_init) - return 0; nfsd4_load_reboot_recovery_data(); - ret = __nfs4_state_start(); - if (ret) - return ret; - nfs4_init = 1; - return 0; + return __nfs4_state_start(); } static void @@ -4110,7 +4200,6 @@ __nfs4_state_shutdown(void) } nfsd4_shutdown_recdir(); - nfs4_init = 0; } void diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index ac17a708023..f8931acb05f 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -2630,7 +2630,7 @@ nfsd4_encode_read(struct nfsd4_compoundres *resp, __be32 nfserr, } read->rd_vlen = v; - nfserr = nfsd_read(read->rd_rqstp, read->rd_fhp, read->rd_filp, + nfserr = nfsd_read_file(read->rd_rqstp, read->rd_fhp, read->rd_filp, read->rd_offset, resp->rqstp->rq_vec, read->rd_vlen, &maxcount); @@ -3325,6 +3325,7 @@ nfs4svc_encode_compoundres(struct svc_rqst *rqstp, __be32 *p, struct nfsd4_compo } /* Renew the clientid on success and on replay */ release_session_client(cs->session); + nfsd4_put_session(cs->session); } return 1; } diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index 508941c23af..b53b1d042f1 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -949,15 +949,12 @@ static ssize_t __write_ports_addfd(char *buf) if (err != 0) return err; - err = lockd_up(); - if (err != 0) - goto out; - err = svc_addsock(nfsd_serv, fd, buf, SIMPLE_TRANSACTION_LIMIT); - if (err < 0) - lockd_down(); + if (err < 0) { + svc_destroy(nfsd_serv); + return err; + } -out: /* Decrease the count, but don't shut down the service */ nfsd_serv->sv_nrthreads--; return err; @@ -978,9 +975,6 @@ static ssize_t __write_ports_delfd(char *buf) if (nfsd_serv != NULL) len = svc_sock_names(nfsd_serv, buf, SIMPLE_TRANSACTION_LIMIT, toclose); - if (len >= 0) - lockd_down(); - kfree(toclose); return len; } @@ -1014,6 +1008,9 @@ static ssize_t __write_ports_addxprt(char *buf) PF_INET6, port, SVC_SOCK_ANONYMOUS); if (err < 0 && err != -EAFNOSUPPORT) goto out_close; + + /* Decrease the count, but don't shut down the service */ + nfsd_serv->sv_nrthreads--; return 0; out_close: xprt = svc_find_xprt(nfsd_serv, transport, PF_INET, port); @@ -1022,8 +1019,7 @@ out_close: svc_xprt_put(xprt); } out_err: - /* Decrease the count, but don't shut down the service */ - nfsd_serv->sv_nrthreads--; + svc_destroy(nfsd_serv); return err; } @@ -1194,7 +1190,7 @@ static ssize_t write_maxblksize(struct file *file, char *buf, size_t size) bsize = NFSSVC_MAXBLKSIZE; bsize &= ~(1024-1); mutex_lock(&nfsd_mutex); - if (nfsd_serv && nfsd_serv->sv_nrthreads) { + if (nfsd_serv) { mutex_unlock(&nfsd_mutex); return -EBUSY; } @@ -1310,6 +1306,8 @@ static ssize_t __write_recoverydir(struct file *file, char *buf, size_t size) return -EINVAL; status = nfs4_reset_recoverydir(recdir); + if (status) + return status; } return scnprintf(buf, SIMPLE_TRANSACTION_LIMIT, "%s\n", diff --git a/fs/nfsd/nfsd.h b/fs/nfsd/nfsd.h index 72377761270..b76ac3a82e3 100644 --- a/fs/nfsd/nfsd.h +++ b/fs/nfsd/nfsd.h @@ -153,6 +153,7 @@ void nfsd_lockd_shutdown(void); #define nfserr_bad_seqid cpu_to_be32(NFSERR_BAD_SEQID) #define nfserr_symlink cpu_to_be32(NFSERR_SYMLINK) #define nfserr_not_same cpu_to_be32(NFSERR_NOT_SAME) +#define nfserr_lock_range cpu_to_be32(NFSERR_LOCK_RANGE) #define nfserr_restorefh cpu_to_be32(NFSERR_RESTOREFH) #define nfserr_attrnotsupp cpu_to_be32(NFSERR_ATTRNOTSUPP) #define nfserr_bad_xdr cpu_to_be32(NFSERR_BAD_XDR) diff --git a/fs/nfsd/nfsproc.c b/fs/nfsd/nfsproc.c index a047ad6111e..08e17264784 100644 --- a/fs/nfsd/nfsproc.c +++ b/fs/nfsd/nfsproc.c @@ -144,7 +144,7 @@ nfsd_proc_read(struct svc_rqst *rqstp, struct nfsd_readargs *argp, svc_reserve_auth(rqstp, (19<<2) + argp->count + 4); resp->count = argp->count; - nfserr = nfsd_read(rqstp, fh_copy(&resp->fh, &argp->fh), NULL, + nfserr = nfsd_read(rqstp, fh_copy(&resp->fh, &argp->fh), argp->offset, rqstp->rq_vec, argp->vlen, &resp->count); @@ -290,7 +290,6 @@ nfsd_proc_create(struct svc_rqst *rqstp, struct nfsd_createargs *argp, * gospel of sun micro */ if (type != S_IFREG) { - int is_borc = 0; if (type != S_IFBLK && type != S_IFCHR) { rdev = 0; } else if (type == S_IFCHR && !(attr->ia_valid & ATTR_SIZE)) { @@ -298,7 +297,6 @@ nfsd_proc_create(struct svc_rqst *rqstp, struct nfsd_createargs *argp, type = S_IFIFO; } else { /* Okay, char or block special */ - is_borc = 1; if (!rdev) rdev = wanted; } diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c index 06b2a26edfe..e2c43464f23 100644 --- a/fs/nfsd/nfssvc.c +++ b/fs/nfsd/nfssvc.c @@ -180,15 +180,80 @@ int nfsd_nrthreads(void) return rv; } +static int nfsd_init_socks(int port) +{ + int error; + if (!list_empty(&nfsd_serv->sv_permsocks)) + return 0; + + error = svc_create_xprt(nfsd_serv, "udp", PF_INET, port, + SVC_SOCK_DEFAULTS); + if (error < 0) + return error; + + error = svc_create_xprt(nfsd_serv, "tcp", PF_INET, port, + SVC_SOCK_DEFAULTS); + if (error < 0) + return error; + + return 0; +} + +static bool nfsd_up = false; + +static int nfsd_startup(unsigned short port, int nrservs) +{ + int ret; + + if (nfsd_up) + return 0; + /* + * Readahead param cache - will no-op if it already exists. + * (Note therefore results will be suboptimal if number of + * threads is modified after nfsd start.) + */ + ret = nfsd_racache_init(2*nrservs); + if (ret) + return ret; + ret = nfsd_init_socks(port); + if (ret) + goto out_racache; + ret = lockd_up(); + if (ret) + goto out_racache; + ret = nfs4_state_start(); + if (ret) + goto out_lockd; + nfsd_up = true; + return 0; +out_lockd: + lockd_down(); +out_racache: + nfsd_racache_shutdown(); + return ret; +} + +static void nfsd_shutdown(void) +{ + /* + * write_ports can create the server without actually starting + * any threads--if we get shut down before any threads are + * started, then nfsd_last_thread will be run before any of this + * other initialization has been done. + */ + if (!nfsd_up) + return; + nfs4_state_shutdown(); + lockd_down(); + nfsd_racache_shutdown(); + nfsd_up = false; +} + static void nfsd_last_thread(struct svc_serv *serv) { /* When last nfsd thread exits we need to do some clean-up */ - struct svc_xprt *xprt; - list_for_each_entry(xprt, &serv->sv_permsocks, xpt_list) - lockd_down(); nfsd_serv = NULL; - nfsd_racache_shutdown(); - nfs4_state_shutdown(); + nfsd_shutdown(); printk(KERN_WARNING "nfsd: last server has exited, flushing export " "cache\n"); @@ -263,45 +328,18 @@ int nfsd_create_serv(void) nfsd_max_blksize >= 8*1024*2) nfsd_max_blksize /= 2; } + nfsd_reset_versions(); nfsd_serv = svc_create_pooled(&nfsd_program, nfsd_max_blksize, nfsd_last_thread, nfsd, THIS_MODULE); if (nfsd_serv == NULL) - err = -ENOMEM; - else - set_max_drc(); + return -ENOMEM; + set_max_drc(); do_gettimeofday(&nfssvc_boot); /* record boot time */ return err; } -static int nfsd_init_socks(int port) -{ - int error; - if (!list_empty(&nfsd_serv->sv_permsocks)) - return 0; - - error = svc_create_xprt(nfsd_serv, "udp", PF_INET, port, - SVC_SOCK_DEFAULTS); - if (error < 0) - return error; - - error = lockd_up(); - if (error < 0) - return error; - - error = svc_create_xprt(nfsd_serv, "tcp", PF_INET, port, - SVC_SOCK_DEFAULTS); - if (error < 0) - return error; - - error = lockd_up(); - if (error < 0) - return error; - - return 0; -} - int nfsd_nrpools(void) { if (nfsd_serv == NULL) @@ -376,10 +414,16 @@ int nfsd_set_nrthreads(int n, int *nthreads) return err; } +/* + * Adjust the number of threads and return the new number of threads. + * This is also the function that starts the server if necessary, if + * this is the first time nrservs is nonzero. + */ int nfsd_svc(unsigned short port, int nrservs) { int error; + bool nfsd_up_before; mutex_lock(&nfsd_mutex); dprintk("nfsd: creating service\n"); @@ -391,34 +435,29 @@ nfsd_svc(unsigned short port, int nrservs) if (nrservs == 0 && nfsd_serv == NULL) goto out; - /* Readahead param cache - will no-op if it already exists */ - error = nfsd_racache_init(2*nrservs); - if (error<0) - goto out; - error = nfs4_state_start(); + error = nfsd_create_serv(); if (error) goto out; - nfsd_reset_versions(); - - error = nfsd_create_serv(); + nfsd_up_before = nfsd_up; + error = nfsd_startup(port, nrservs); if (error) - goto out; - error = nfsd_init_socks(port); - if (error) - goto failure; - + goto out_destroy; error = svc_set_num_threads(nfsd_serv, NULL, nrservs); - if (error == 0) - /* We are holding a reference to nfsd_serv which - * we don't want to count in the return value, - * so subtract 1 - */ - error = nfsd_serv->sv_nrthreads - 1; - failure: + if (error) + goto out_shutdown; + /* We are holding a reference to nfsd_serv which + * we don't want to count in the return value, + * so subtract 1 + */ + error = nfsd_serv->sv_nrthreads - 1; +out_shutdown: + if (error < 0 && !nfsd_up_before) + nfsd_shutdown(); +out_destroy: svc_destroy(nfsd_serv); /* Release server */ - out: +out: mutex_unlock(&nfsd_mutex); return error; } diff --git a/fs/nfsd/state.h b/fs/nfsd/state.h index 006c84230c7..7731a75971d 100644 --- a/fs/nfsd/state.h +++ b/fs/nfsd/state.h @@ -88,7 +88,6 @@ struct nfs4_delegation { struct nfs4_client *dl_client; struct nfs4_file *dl_file; struct file_lock *dl_flock; - struct file *dl_vfs_file; u32 dl_type; time_t dl_time; /* For recall: */ @@ -342,12 +341,50 @@ struct nfs4_file { struct list_head fi_hash; /* hash by "struct inode *" */ struct list_head fi_stateids; struct list_head fi_delegations; + /* One each for O_RDONLY, O_WRONLY, O_RDWR: */ + struct file * fi_fds[3]; + /* One each for O_RDONLY, O_WRONLY: */ + atomic_t fi_access[2]; + /* + * Each open stateid contributes 1 to either fi_readers or + * fi_writers, or both, depending on the open mode. A + * delegation also takes an fi_readers reference. Lock + * stateid's take none. + */ + atomic_t fi_readers; + atomic_t fi_writers; struct inode *fi_inode; u32 fi_id; /* used with stateowner->so_id * for stateid_hashtbl hash */ bool fi_had_conflict; }; +/* XXX: for first cut may fall back on returning file that doesn't work + * at all? */ +static inline struct file *find_writeable_file(struct nfs4_file *f) +{ + if (f->fi_fds[O_RDWR]) + return f->fi_fds[O_RDWR]; + return f->fi_fds[O_WRONLY]; +} + +static inline struct file *find_readable_file(struct nfs4_file *f) +{ + if (f->fi_fds[O_RDWR]) + return f->fi_fds[O_RDWR]; + return f->fi_fds[O_RDONLY]; +} + +static inline struct file *find_any_file(struct nfs4_file *f) +{ + if (f->fi_fds[O_RDWR]) + return f->fi_fds[O_RDWR]; + else if (f->fi_fds[O_RDWR]) + return f->fi_fds[O_WRONLY]; + else + return f->fi_fds[O_RDONLY]; +} + /* * nfs4_stateid can either be an open stateid or (eventually) a lock stateid * @@ -373,7 +410,6 @@ struct nfs4_stateid { struct nfs4_stateowner * st_stateowner; struct nfs4_file * st_file; stateid_t st_stateid; - struct file * st_vfs_file; unsigned long st_access_bmap; unsigned long st_deny_bmap; struct nfs4_stateid * st_openstp; diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c index 3c111120b61..9df85a13af2 100644 --- a/fs/nfsd/vfs.c +++ b/fs/nfsd/vfs.c @@ -604,7 +604,7 @@ nfsd4_get_nfs4_acl(struct svc_rqst *rqstp, struct dentry *dentry, struct nfs4_ac return error; } -#endif /* defined(CONFIG_NFS_V4) */ +#endif /* defined(CONFIG_NFSD_V4) */ #ifdef CONFIG_NFSD_V3 /* @@ -903,7 +903,6 @@ nfsd_vfs_read(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file, loff_t offset, struct kvec *vec, int vlen, unsigned long *count) { struct inode *inode; - struct raparms *ra; mm_segment_t oldfs; __be32 err; int host_err; @@ -914,12 +913,6 @@ nfsd_vfs_read(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file, if (svc_msnfs(fhp) && !lock_may_read(inode, offset, *count)) goto out; - /* Get readahead parameters */ - ra = nfsd_get_raparms(inode->i_sb->s_dev, inode->i_ino); - - if (ra && ra->p_set) - file->f_ra = ra->p_ra; - if (file->f_op->splice_read && rqstp->rq_splice_ok) { struct splice_desc sd = { .len = 0, @@ -937,16 +930,6 @@ nfsd_vfs_read(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file, set_fs(oldfs); } - /* Write back readahead params */ - if (ra) { - struct raparm_hbucket *rab = &raparm_hash[ra->p_hindex]; - spin_lock(&rab->pb_lock); - ra->p_ra = file->f_ra; - ra->p_set = 1; - ra->p_count--; - spin_unlock(&rab->pb_lock); - } - if (host_err >= 0) { nfsdstats.io_read += host_err; *count = host_err; @@ -1086,8 +1069,45 @@ out: * on entry. On return, *count contains the number of bytes actually read. * N.B. After this call fhp needs an fh_put */ +__be32 nfsd_read(struct svc_rqst *rqstp, struct svc_fh *fhp, + loff_t offset, struct kvec *vec, int vlen, unsigned long *count) +{ + struct file *file; + struct inode *inode; + struct raparms *ra; + __be32 err; + + err = nfsd_open(rqstp, fhp, S_IFREG, NFSD_MAY_READ, &file); + if (err) + return err; + + inode = file->f_path.dentry->d_inode; + + /* Get readahead parameters */ + ra = nfsd_get_raparms(inode->i_sb->s_dev, inode->i_ino); + + if (ra && ra->p_set) + file->f_ra = ra->p_ra; + + err = nfsd_vfs_read(rqstp, fhp, file, offset, vec, vlen, count); + + /* Write back readahead params */ + if (ra) { + struct raparm_hbucket *rab = &raparm_hash[ra->p_hindex]; + spin_lock(&rab->pb_lock); + ra->p_ra = file->f_ra; + ra->p_set = 1; + ra->p_count--; + spin_unlock(&rab->pb_lock); + } + + nfsd_close(file); + return err; +} + +/* As above, but use the provided file descriptor. */ __be32 -nfsd_read(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file, +nfsd_read_file(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file, loff_t offset, struct kvec *vec, int vlen, unsigned long *count) { @@ -1099,13 +1119,8 @@ nfsd_read(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file, if (err) goto out; err = nfsd_vfs_read(rqstp, fhp, file, offset, vec, vlen, count); - } else { - err = nfsd_open(rqstp, fhp, S_IFREG, NFSD_MAY_READ, &file); - if (err) - goto out; - err = nfsd_vfs_read(rqstp, fhp, file, offset, vec, vlen, count); - nfsd_close(file); - } + } else /* Note file may still be NULL in NFSv4 special stateid case: */ + err = nfsd_read(rqstp, fhp, offset, vec, vlen, count); out: return err; } @@ -1631,7 +1646,7 @@ nfsd_link(struct svc_rqst *rqstp, struct svc_fh *ffhp, char *name, int len, struct svc_fh *tfhp) { struct dentry *ddir, *dnew, *dold; - struct inode *dirp, *dest; + struct inode *dirp; __be32 err; int host_err; @@ -1659,7 +1674,6 @@ nfsd_link(struct svc_rqst *rqstp, struct svc_fh *ffhp, goto out_nfserr; dold = tfhp->fh_dentry; - dest = dold->d_inode; host_err = mnt_want_write(tfhp->fh_export->ex_path.mnt); if (host_err) { @@ -2038,7 +2052,6 @@ nfsd_permission(struct svc_rqst *rqstp, struct svc_export *exp, struct dentry *dentry, int acc) { struct inode *inode = dentry->d_inode; - struct path path; int err; if (acc == NFSD_MAY_NOP) @@ -2111,15 +2124,7 @@ nfsd_permission(struct svc_rqst *rqstp, struct svc_export *exp, if (err == -EACCES && S_ISREG(inode->i_mode) && acc == (NFSD_MAY_READ | NFSD_MAY_OWNER_OVERRIDE)) err = inode_permission(inode, MAY_EXEC); - if (err) - goto nfsd_out; - /* Do integrity (permission) checking now, but defer incrementing - * IMA counts to the actual file open. - */ - path.mnt = exp->ex_path.mnt; - path.dentry = dentry; -nfsd_out: return err? nfserrno(err) : 0; } diff --git a/fs/nfsd/vfs.h b/fs/nfsd/vfs.h index 217a62c2a35..9a370a5e36b 100644 --- a/fs/nfsd/vfs.h +++ b/fs/nfsd/vfs.h @@ -64,7 +64,9 @@ __be32 nfsd_commit(struct svc_rqst *, struct svc_fh *, __be32 nfsd_open(struct svc_rqst *, struct svc_fh *, int, int, struct file **); void nfsd_close(struct file *); -__be32 nfsd_read(struct svc_rqst *, struct svc_fh *, struct file *, +__be32 nfsd_read(struct svc_rqst *, struct svc_fh *, + loff_t, struct kvec *, int, unsigned long *); +__be32 nfsd_read_file(struct svc_rqst *, struct svc_fh *, struct file *, loff_t, struct kvec *, int, unsigned long *); __be32 nfsd_write(struct svc_rqst *, struct svc_fh *,struct file *, loff_t, struct kvec *,int, unsigned long *, int *); diff --git a/fs/nilfs2/bmap.c b/fs/nilfs2/bmap.c index effdbdbe6c1..3dbdc1d356b 100644 --- a/fs/nilfs2/bmap.c +++ b/fs/nilfs2/bmap.c @@ -26,6 +26,8 @@ #include "nilfs.h" #include "bmap.h" #include "sb.h" +#include "btree.h" +#include "direct.h" #include "btnode.h" #include "mdt.h" #include "dat.h" @@ -533,7 +535,7 @@ void nilfs_bmap_init_gc(struct nilfs_bmap *bmap) void nilfs_bmap_init_gcdat(struct nilfs_bmap *gcbmap, struct nilfs_bmap *bmap) { - memcpy(gcbmap, bmap, sizeof(union nilfs_bmap_union)); + memcpy(gcbmap, bmap, sizeof(*bmap)); init_rwsem(&gcbmap->b_sem); lockdep_set_class(&bmap->b_sem, &nilfs_bmap_dat_lock_key); gcbmap->b_inode = &NILFS_BMAP_I(gcbmap)->vfs_inode; @@ -541,7 +543,7 @@ void nilfs_bmap_init_gcdat(struct nilfs_bmap *gcbmap, struct nilfs_bmap *bmap) void nilfs_bmap_commit_gcdat(struct nilfs_bmap *gcbmap, struct nilfs_bmap *bmap) { - memcpy(bmap, gcbmap, sizeof(union nilfs_bmap_union)); + memcpy(bmap, gcbmap, sizeof(*bmap)); init_rwsem(&bmap->b_sem); lockdep_set_class(&bmap->b_sem, &nilfs_bmap_dat_lock_key); bmap->b_inode = &NILFS_BMAP_I(bmap)->vfs_inode; diff --git a/fs/nilfs2/bmap.h b/fs/nilfs2/bmap.h index 9980d7dbab9..a20569b1992 100644 --- a/fs/nilfs2/bmap.h +++ b/fs/nilfs2/bmap.h @@ -32,11 +32,6 @@ #define NILFS_BMAP_INVALID_PTR 0 -#define nilfs_bmap_dkey_to_key(dkey) le64_to_cpu(dkey) -#define nilfs_bmap_key_to_dkey(key) cpu_to_le64(key) -#define nilfs_bmap_dptr_to_ptr(dptr) le64_to_cpu(dptr) -#define nilfs_bmap_ptr_to_dptr(ptr) cpu_to_le64(ptr) - #define nilfs_bmap_keydiff_abs(diff) ((diff) < 0 ? -(diff) : (diff)) @@ -71,7 +66,7 @@ struct nilfs_bmap_operations { int (*bop_delete)(struct nilfs_bmap *, __u64); void (*bop_clear)(struct nilfs_bmap *); - int (*bop_propagate)(const struct nilfs_bmap *, struct buffer_head *); + int (*bop_propagate)(struct nilfs_bmap *, struct buffer_head *); void (*bop_lookup_dirty_buffers)(struct nilfs_bmap *, struct list_head *); @@ -110,6 +105,7 @@ static inline int nilfs_bmap_is_new_ptr(unsigned long ptr) * @b_last_allocated_ptr: last allocated ptr for data block * @b_ptr_type: pointer type * @b_state: state + * @b_nchildren_per_block: maximum number of child nodes for non-root nodes */ struct nilfs_bmap { union { @@ -123,6 +119,7 @@ struct nilfs_bmap { __u64 b_last_allocated_ptr; int b_ptr_type; int b_state; + __u16 b_nchildren_per_block; }; /* pointer type */ @@ -224,6 +221,13 @@ static inline void nilfs_bmap_abort_end_ptr(struct nilfs_bmap *bmap, nilfs_dat_abort_end(dat, &req->bpr_req); } +static inline void nilfs_bmap_set_target_v(struct nilfs_bmap *bmap, __u64 key, + __u64 ptr) +{ + bmap->b_last_allocated_key = key; + bmap->b_last_allocated_ptr = ptr; +} + __u64 nilfs_bmap_data_get_key(const struct nilfs_bmap *, const struct buffer_head *); diff --git a/fs/nilfs2/bmap_union.h b/fs/nilfs2/bmap_union.h deleted file mode 100644 index d41509bff47..00000000000 --- a/fs/nilfs2/bmap_union.h +++ /dev/null @@ -1,42 +0,0 @@ -/* - * bmap_union.h - NILFS block mapping. - * - * Copyright (C) 2006-2008 Nippon Telegraph and Telephone Corporation. - * - * 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. - * - * 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA - * - * Written by Koji Sato <koji@osrg.net>. - */ - -#ifndef _NILFS_BMAP_UNION_H -#define _NILFS_BMAP_UNION_H - -#include "bmap.h" -#include "direct.h" -#include "btree.h" - -/** - * nilfs_bmap_union - - * @bi_bmap: bmap structure - * @bi_btree: direct map structure - * @bi_direct: B-tree structure - */ -union nilfs_bmap_union { - struct nilfs_bmap bi_bmap; - struct nilfs_direct bi_direct; - struct nilfs_btree bi_btree; -}; - -#endif /* _NILFS_BMAP_UNION_H */ diff --git a/fs/nilfs2/btnode.c b/fs/nilfs2/btnode.c index 447ce47a330..f78ab1044d1 100644 --- a/fs/nilfs2/btnode.c +++ b/fs/nilfs2/btnode.c @@ -96,10 +96,12 @@ nilfs_btnode_create_block(struct address_space *btnc, __u64 blocknr) } int nilfs_btnode_submit_block(struct address_space *btnc, __u64 blocknr, - sector_t pblocknr, struct buffer_head **pbh) + sector_t pblocknr, int mode, + struct buffer_head **pbh, sector_t *submit_ptr) { struct buffer_head *bh; struct inode *inode = NILFS_BTNC_I(btnc); + struct page *page; int err; bh = nilfs_grab_buffer(inode, btnc, blocknr, 1 << BH_NILFS_Node); @@ -107,6 +109,7 @@ int nilfs_btnode_submit_block(struct address_space *btnc, __u64 blocknr, return -ENOMEM; err = -EEXIST; /* internal code */ + page = bh->b_page; if (buffer_uptodate(bh) || buffer_dirty(bh)) goto found; @@ -125,7 +128,16 @@ int nilfs_btnode_submit_block(struct address_space *btnc, __u64 blocknr, } } } - lock_buffer(bh); + + if (mode == READA) { + if (pblocknr != *submit_ptr + 1 || !trylock_buffer(bh)) { + err = -EBUSY; /* internal code */ + brelse(bh); + goto out_locked; + } + } else { /* mode == READ */ + lock_buffer(bh); + } if (buffer_uptodate(bh)) { unlock_buffer(bh); err = -EEXIST; /* internal code */ @@ -136,15 +148,16 @@ int nilfs_btnode_submit_block(struct address_space *btnc, __u64 blocknr, bh->b_blocknr = pblocknr; /* set block address for read */ bh->b_end_io = end_buffer_read_sync; get_bh(bh); - submit_bh(READ, bh); + submit_bh(mode, bh); bh->b_blocknr = blocknr; /* set back to the given block address */ + *submit_ptr = pblocknr; err = 0; found: *pbh = bh; out_locked: - unlock_page(bh->b_page); - page_cache_release(bh->b_page); + unlock_page(page); + page_cache_release(page); return err; } diff --git a/fs/nilfs2/btnode.h b/fs/nilfs2/btnode.h index 07da83f0771..79037494f1e 100644 --- a/fs/nilfs2/btnode.h +++ b/fs/nilfs2/btnode.h @@ -42,8 +42,8 @@ void nilfs_btnode_cache_init(struct address_space *, struct backing_dev_info *); void nilfs_btnode_cache_clear(struct address_space *); struct buffer_head *nilfs_btnode_create_block(struct address_space *btnc, __u64 blocknr); -int nilfs_btnode_submit_block(struct address_space *, __u64, sector_t, - struct buffer_head **); +int nilfs_btnode_submit_block(struct address_space *, __u64, sector_t, int, + struct buffer_head **, sector_t *); void nilfs_btnode_delete(struct buffer_head *); int nilfs_btnode_prepare_change_key(struct address_space *, struct nilfs_btnode_chkey_ctxt *); diff --git a/fs/nilfs2/btree.c b/fs/nilfs2/btree.c index b27a342c5af..300c2bc00c3 100644 --- a/fs/nilfs2/btree.c +++ b/fs/nilfs2/btree.c @@ -66,30 +66,10 @@ static void nilfs_btree_free_path(struct nilfs_btree_path *path) /* * B-tree node operations */ -static int nilfs_btree_get_block(const struct nilfs_btree *btree, __u64 ptr, - struct buffer_head **bhp) -{ - struct address_space *btnc = - &NILFS_BMAP_I((struct nilfs_bmap *)btree)->i_btnode_cache; - int err; - - err = nilfs_btnode_submit_block(btnc, ptr, 0, bhp); - if (err) - return err == -EEXIST ? 0 : err; - - wait_on_buffer(*bhp); - if (!buffer_uptodate(*bhp)) { - brelse(*bhp); - return -EIO; - } - return 0; -} - -static int nilfs_btree_get_new_block(const struct nilfs_btree *btree, +static int nilfs_btree_get_new_block(const struct nilfs_bmap *btree, __u64 ptr, struct buffer_head **bhp) { - struct address_space *btnc = - &NILFS_BMAP_I((struct nilfs_bmap *)btree)->i_btnode_cache; + struct address_space *btnc = &NILFS_BMAP_I(btree)->i_btnode_cache; struct buffer_head *bh; bh = nilfs_btnode_create_block(btnc, ptr); @@ -101,71 +81,55 @@ static int nilfs_btree_get_new_block(const struct nilfs_btree *btree, return 0; } -static inline int -nilfs_btree_node_get_flags(const struct nilfs_btree_node *node) +static int nilfs_btree_node_get_flags(const struct nilfs_btree_node *node) { return node->bn_flags; } -static inline void +static void nilfs_btree_node_set_flags(struct nilfs_btree_node *node, int flags) { node->bn_flags = flags; } -static inline int nilfs_btree_node_root(const struct nilfs_btree_node *node) +static int nilfs_btree_node_root(const struct nilfs_btree_node *node) { return nilfs_btree_node_get_flags(node) & NILFS_BTREE_NODE_ROOT; } -static inline int -nilfs_btree_node_get_level(const struct nilfs_btree_node *node) +static int nilfs_btree_node_get_level(const struct nilfs_btree_node *node) { return node->bn_level; } -static inline void +static void nilfs_btree_node_set_level(struct nilfs_btree_node *node, int level) { node->bn_level = level; } -static inline int -nilfs_btree_node_get_nchildren(const struct nilfs_btree_node *node) +static int nilfs_btree_node_get_nchildren(const struct nilfs_btree_node *node) { return le16_to_cpu(node->bn_nchildren); } -static inline void +static void nilfs_btree_node_set_nchildren(struct nilfs_btree_node *node, int nchildren) { node->bn_nchildren = cpu_to_le16(nchildren); } -static inline int nilfs_btree_node_size(const struct nilfs_btree *btree) +static int nilfs_btree_node_size(const struct nilfs_bmap *btree) { - return 1 << btree->bt_bmap.b_inode->i_blkbits; + return 1 << btree->b_inode->i_blkbits; } -static inline int -nilfs_btree_node_nchildren_min(const struct nilfs_btree_node *node, - const struct nilfs_btree *btree) +static int nilfs_btree_nchildren_per_block(const struct nilfs_bmap *btree) { - return nilfs_btree_node_root(node) ? - NILFS_BTREE_ROOT_NCHILDREN_MIN : - NILFS_BTREE_NODE_NCHILDREN_MIN(nilfs_btree_node_size(btree)); + return btree->b_nchildren_per_block; } -static inline int -nilfs_btree_node_nchildren_max(const struct nilfs_btree_node *node, - const struct nilfs_btree *btree) -{ - return nilfs_btree_node_root(node) ? - NILFS_BTREE_ROOT_NCHILDREN_MAX : - NILFS_BTREE_NODE_NCHILDREN_MAX(nilfs_btree_node_size(btree)); -} - -static inline __le64 * +static __le64 * nilfs_btree_node_dkeys(const struct nilfs_btree_node *node) { return (__le64 *)((char *)(node + 1) + @@ -173,45 +137,40 @@ nilfs_btree_node_dkeys(const struct nilfs_btree_node *node) 0 : NILFS_BTREE_NODE_EXTRA_PAD_SIZE)); } -static inline __le64 * -nilfs_btree_node_dptrs(const struct nilfs_btree_node *node, - const struct nilfs_btree *btree) +static __le64 * +nilfs_btree_node_dptrs(const struct nilfs_btree_node *node, int ncmax) { - return (__le64 *)(nilfs_btree_node_dkeys(node) + - nilfs_btree_node_nchildren_max(node, btree)); + return (__le64 *)(nilfs_btree_node_dkeys(node) + ncmax); } -static inline __u64 +static __u64 nilfs_btree_node_get_key(const struct nilfs_btree_node *node, int index) { - return nilfs_bmap_dkey_to_key(*(nilfs_btree_node_dkeys(node) + index)); + return le64_to_cpu(*(nilfs_btree_node_dkeys(node) + index)); } -static inline void +static void nilfs_btree_node_set_key(struct nilfs_btree_node *node, int index, __u64 key) { - *(nilfs_btree_node_dkeys(node) + index) = nilfs_bmap_key_to_dkey(key); + *(nilfs_btree_node_dkeys(node) + index) = cpu_to_le64(key); } -static inline __u64 -nilfs_btree_node_get_ptr(const struct nilfs_btree *btree, - const struct nilfs_btree_node *node, int index) +static __u64 +nilfs_btree_node_get_ptr(const struct nilfs_btree_node *node, int index, + int ncmax) { - return nilfs_bmap_dptr_to_ptr(*(nilfs_btree_node_dptrs(node, btree) + - index)); + return le64_to_cpu(*(nilfs_btree_node_dptrs(node, ncmax) + index)); } -static inline void -nilfs_btree_node_set_ptr(struct nilfs_btree *btree, - struct nilfs_btree_node *node, int index, __u64 ptr) +static void +nilfs_btree_node_set_ptr(struct nilfs_btree_node *node, int index, __u64 ptr, + int ncmax) { - *(nilfs_btree_node_dptrs(node, btree) + index) = - nilfs_bmap_ptr_to_dptr(ptr); + *(nilfs_btree_node_dptrs(node, ncmax) + index) = cpu_to_le64(ptr); } -static void nilfs_btree_node_init(struct nilfs_btree *btree, - struct nilfs_btree_node *node, - int flags, int level, int nchildren, +static void nilfs_btree_node_init(struct nilfs_btree_node *node, int flags, + int level, int nchildren, int ncmax, const __u64 *keys, const __u64 *ptrs) { __le64 *dkeys; @@ -223,29 +182,28 @@ static void nilfs_btree_node_init(struct nilfs_btree *btree, nilfs_btree_node_set_nchildren(node, nchildren); dkeys = nilfs_btree_node_dkeys(node); - dptrs = nilfs_btree_node_dptrs(node, btree); + dptrs = nilfs_btree_node_dptrs(node, ncmax); for (i = 0; i < nchildren; i++) { - dkeys[i] = nilfs_bmap_key_to_dkey(keys[i]); - dptrs[i] = nilfs_bmap_ptr_to_dptr(ptrs[i]); + dkeys[i] = cpu_to_le64(keys[i]); + dptrs[i] = cpu_to_le64(ptrs[i]); } } /* Assume the buffer heads corresponding to left and right are locked. */ -static void nilfs_btree_node_move_left(struct nilfs_btree *btree, - struct nilfs_btree_node *left, +static void nilfs_btree_node_move_left(struct nilfs_btree_node *left, struct nilfs_btree_node *right, - int n) + int n, int lncmax, int rncmax) { __le64 *ldkeys, *rdkeys; __le64 *ldptrs, *rdptrs; int lnchildren, rnchildren; ldkeys = nilfs_btree_node_dkeys(left); - ldptrs = nilfs_btree_node_dptrs(left, btree); + ldptrs = nilfs_btree_node_dptrs(left, lncmax); lnchildren = nilfs_btree_node_get_nchildren(left); rdkeys = nilfs_btree_node_dkeys(right); - rdptrs = nilfs_btree_node_dptrs(right, btree); + rdptrs = nilfs_btree_node_dptrs(right, rncmax); rnchildren = nilfs_btree_node_get_nchildren(right); memcpy(ldkeys + lnchildren, rdkeys, n * sizeof(*rdkeys)); @@ -260,21 +218,20 @@ static void nilfs_btree_node_move_left(struct nilfs_btree *btree, } /* Assume that the buffer heads corresponding to left and right are locked. */ -static void nilfs_btree_node_move_right(struct nilfs_btree *btree, - struct nilfs_btree_node *left, +static void nilfs_btree_node_move_right(struct nilfs_btree_node *left, struct nilfs_btree_node *right, - int n) + int n, int lncmax, int rncmax) { __le64 *ldkeys, *rdkeys; __le64 *ldptrs, *rdptrs; int lnchildren, rnchildren; ldkeys = nilfs_btree_node_dkeys(left); - ldptrs = nilfs_btree_node_dptrs(left, btree); + ldptrs = nilfs_btree_node_dptrs(left, lncmax); lnchildren = nilfs_btree_node_get_nchildren(left); rdkeys = nilfs_btree_node_dkeys(right); - rdptrs = nilfs_btree_node_dptrs(right, btree); + rdptrs = nilfs_btree_node_dptrs(right, rncmax); rnchildren = nilfs_btree_node_get_nchildren(right); memmove(rdkeys + n, rdkeys, rnchildren * sizeof(*rdkeys)); @@ -289,16 +246,15 @@ static void nilfs_btree_node_move_right(struct nilfs_btree *btree, } /* Assume that the buffer head corresponding to node is locked. */ -static void nilfs_btree_node_insert(struct nilfs_btree *btree, - struct nilfs_btree_node *node, - __u64 key, __u64 ptr, int index) +static void nilfs_btree_node_insert(struct nilfs_btree_node *node, int index, + __u64 key, __u64 ptr, int ncmax) { __le64 *dkeys; __le64 *dptrs; int nchildren; dkeys = nilfs_btree_node_dkeys(node); - dptrs = nilfs_btree_node_dptrs(node, btree); + dptrs = nilfs_btree_node_dptrs(node, ncmax); nchildren = nilfs_btree_node_get_nchildren(node); if (index < nchildren) { memmove(dkeys + index + 1, dkeys + index, @@ -306,16 +262,15 @@ static void nilfs_btree_node_insert(struct nilfs_btree *btree, memmove(dptrs + index + 1, dptrs + index, (nchildren - index) * sizeof(*dptrs)); } - dkeys[index] = nilfs_bmap_key_to_dkey(key); - dptrs[index] = nilfs_bmap_ptr_to_dptr(ptr); + dkeys[index] = cpu_to_le64(key); + dptrs[index] = cpu_to_le64(ptr); nchildren++; nilfs_btree_node_set_nchildren(node, nchildren); } /* Assume that the buffer head corresponding to node is locked. */ -static void nilfs_btree_node_delete(struct nilfs_btree *btree, - struct nilfs_btree_node *node, - __u64 *keyp, __u64 *ptrp, int index) +static void nilfs_btree_node_delete(struct nilfs_btree_node *node, int index, + __u64 *keyp, __u64 *ptrp, int ncmax) { __u64 key; __u64 ptr; @@ -324,9 +279,9 @@ static void nilfs_btree_node_delete(struct nilfs_btree *btree, int nchildren; dkeys = nilfs_btree_node_dkeys(node); - dptrs = nilfs_btree_node_dptrs(node, btree); - key = nilfs_bmap_dkey_to_key(dkeys[index]); - ptr = nilfs_bmap_dptr_to_ptr(dptrs[index]); + dptrs = nilfs_btree_node_dptrs(node, ncmax); + key = le64_to_cpu(dkeys[index]); + ptr = le64_to_cpu(dptrs[index]); nchildren = nilfs_btree_node_get_nchildren(node); if (keyp != NULL) *keyp = key; @@ -382,40 +337,92 @@ static int nilfs_btree_node_lookup(const struct nilfs_btree_node *node, return s == 0; } -static inline struct nilfs_btree_node * -nilfs_btree_get_root(const struct nilfs_btree *btree) +/** + * nilfs_btree_node_broken - verify consistency of btree node + * @node: btree node block to be examined + * @size: node size (in bytes) + * @blocknr: block number + * + * Return Value: If node is broken, 1 is returned. Otherwise, 0 is returned. + */ +static int nilfs_btree_node_broken(const struct nilfs_btree_node *node, + size_t size, sector_t blocknr) { - return (struct nilfs_btree_node *)btree->bt_bmap.b_u.u_data; + int level, flags, nchildren; + int ret = 0; + + level = nilfs_btree_node_get_level(node); + flags = nilfs_btree_node_get_flags(node); + nchildren = nilfs_btree_node_get_nchildren(node); + + if (unlikely(level < NILFS_BTREE_LEVEL_NODE_MIN || + level >= NILFS_BTREE_LEVEL_MAX || + (flags & NILFS_BTREE_NODE_ROOT) || + nchildren < 0 || + nchildren > NILFS_BTREE_NODE_NCHILDREN_MAX(size))) { + printk(KERN_CRIT "NILFS: bad btree node (blocknr=%llu): " + "level = %d, flags = 0x%x, nchildren = %d\n", + (unsigned long long)blocknr, level, flags, nchildren); + ret = 1; + } + return ret; } -static inline struct nilfs_btree_node * +int nilfs_btree_broken_node_block(struct buffer_head *bh) +{ + int ret; + + if (buffer_nilfs_checked(bh)) + return 0; + + ret = nilfs_btree_node_broken((struct nilfs_btree_node *)bh->b_data, + bh->b_size, bh->b_blocknr); + if (likely(!ret)) + set_buffer_nilfs_checked(bh); + return ret; +} + +static struct nilfs_btree_node * +nilfs_btree_get_root(const struct nilfs_bmap *btree) +{ + return (struct nilfs_btree_node *)btree->b_u.u_data; +} + +static struct nilfs_btree_node * nilfs_btree_get_nonroot_node(const struct nilfs_btree_path *path, int level) { return (struct nilfs_btree_node *)path[level].bp_bh->b_data; } -static inline struct nilfs_btree_node * +static struct nilfs_btree_node * nilfs_btree_get_sib_node(const struct nilfs_btree_path *path, int level) { return (struct nilfs_btree_node *)path[level].bp_sib_bh->b_data; } -static inline int nilfs_btree_height(const struct nilfs_btree *btree) +static int nilfs_btree_height(const struct nilfs_bmap *btree) { return nilfs_btree_node_get_level(nilfs_btree_get_root(btree)) + 1; } -static inline struct nilfs_btree_node * -nilfs_btree_get_node(const struct nilfs_btree *btree, +static struct nilfs_btree_node * +nilfs_btree_get_node(const struct nilfs_bmap *btree, const struct nilfs_btree_path *path, - int level) + int level, int *ncmaxp) { - return (level == nilfs_btree_height(btree) - 1) ? - nilfs_btree_get_root(btree) : - nilfs_btree_get_nonroot_node(path, level); + struct nilfs_btree_node *node; + + if (level == nilfs_btree_height(btree) - 1) { + node = nilfs_btree_get_root(btree); + *ncmaxp = NILFS_BTREE_ROOT_NCHILDREN_MAX; + } else { + node = nilfs_btree_get_nonroot_node(path, level); + *ncmaxp = nilfs_btree_nchildren_per_block(btree); + } + return node; } -static inline int +static int nilfs_btree_bad_node(struct nilfs_btree_node *node, int level) { if (unlikely(nilfs_btree_node_get_level(node) != level)) { @@ -427,13 +434,83 @@ nilfs_btree_bad_node(struct nilfs_btree_node *node, int level) return 0; } -static int nilfs_btree_do_lookup(const struct nilfs_btree *btree, +struct nilfs_btree_readahead_info { + struct nilfs_btree_node *node; /* parent node */ + int max_ra_blocks; /* max nof blocks to read ahead */ + int index; /* current index on the parent node */ + int ncmax; /* nof children in the parent node */ +}; + +static int __nilfs_btree_get_block(const struct nilfs_bmap *btree, __u64 ptr, + struct buffer_head **bhp, + const struct nilfs_btree_readahead_info *ra) +{ + struct address_space *btnc = &NILFS_BMAP_I(btree)->i_btnode_cache; + struct buffer_head *bh, *ra_bh; + sector_t submit_ptr = 0; + int ret; + + ret = nilfs_btnode_submit_block(btnc, ptr, 0, READ, &bh, &submit_ptr); + if (ret) { + if (ret != -EEXIST) + return ret; + goto out_check; + } + + if (ra) { + int i, n; + __u64 ptr2; + + /* read ahead sibling nodes */ + for (n = ra->max_ra_blocks, i = ra->index + 1; + n > 0 && i < ra->ncmax; n--, i++) { + ptr2 = nilfs_btree_node_get_ptr(ra->node, i, ra->ncmax); + + ret = nilfs_btnode_submit_block(btnc, ptr2, 0, READA, + &ra_bh, &submit_ptr); + if (likely(!ret || ret == -EEXIST)) + brelse(ra_bh); + else if (ret != -EBUSY) + break; + if (!buffer_locked(bh)) + goto out_no_wait; + } + } + + wait_on_buffer(bh); + + out_no_wait: + if (!buffer_uptodate(bh)) { + brelse(bh); + return -EIO; + } + + out_check: + if (nilfs_btree_broken_node_block(bh)) { + clear_buffer_uptodate(bh); + brelse(bh); + return -EINVAL; + } + + *bhp = bh; + return 0; +} + +static int nilfs_btree_get_block(const struct nilfs_bmap *btree, __u64 ptr, + struct buffer_head **bhp) +{ + return __nilfs_btree_get_block(btree, ptr, bhp, NULL); +} + +static int nilfs_btree_do_lookup(const struct nilfs_bmap *btree, struct nilfs_btree_path *path, - __u64 key, __u64 *ptrp, int minlevel) + __u64 key, __u64 *ptrp, int minlevel, + int readahead) { struct nilfs_btree_node *node; + struct nilfs_btree_readahead_info p, *ra; __u64 ptr; - int level, index, found, ret; + int level, index, found, ncmax, ret; node = nilfs_btree_get_root(btree); level = nilfs_btree_node_get_level(node); @@ -441,14 +518,27 @@ static int nilfs_btree_do_lookup(const struct nilfs_btree *btree, return -ENOENT; found = nilfs_btree_node_lookup(node, key, &index); - ptr = nilfs_btree_node_get_ptr(btree, node, index); + ptr = nilfs_btree_node_get_ptr(node, index, + NILFS_BTREE_ROOT_NCHILDREN_MAX); path[level].bp_bh = NULL; path[level].bp_index = index; - for (level--; level >= minlevel; level--) { - ret = nilfs_btree_get_block(btree, ptr, &path[level].bp_bh); + ncmax = nilfs_btree_nchildren_per_block(btree); + + while (--level >= minlevel) { + ra = NULL; + if (level == NILFS_BTREE_LEVEL_NODE_MIN && readahead) { + p.node = nilfs_btree_get_node(btree, path, level + 1, + &p.ncmax); + p.index = index; + p.max_ra_blocks = 7; + ra = &p; + } + ret = __nilfs_btree_get_block(btree, ptr, &path[level].bp_bh, + ra); if (ret < 0) return ret; + node = nilfs_btree_get_nonroot_node(path, level); if (nilfs_btree_bad_node(node, level)) return -EINVAL; @@ -456,9 +546,9 @@ static int nilfs_btree_do_lookup(const struct nilfs_btree *btree, found = nilfs_btree_node_lookup(node, key, &index); else index = 0; - if (index < nilfs_btree_node_nchildren_max(node, btree)) - ptr = nilfs_btree_node_get_ptr(btree, node, index); - else { + if (index < ncmax) { + ptr = nilfs_btree_node_get_ptr(node, index, ncmax); + } else { WARN_ON(found || level != NILFS_BTREE_LEVEL_NODE_MIN); /* insert */ ptr = NILFS_BMAP_INVALID_PTR; @@ -474,22 +564,24 @@ static int nilfs_btree_do_lookup(const struct nilfs_btree *btree, return 0; } -static int nilfs_btree_do_lookup_last(const struct nilfs_btree *btree, +static int nilfs_btree_do_lookup_last(const struct nilfs_bmap *btree, struct nilfs_btree_path *path, __u64 *keyp, __u64 *ptrp) { struct nilfs_btree_node *node; __u64 ptr; - int index, level, ret; + int index, level, ncmax, ret; node = nilfs_btree_get_root(btree); index = nilfs_btree_node_get_nchildren(node) - 1; if (index < 0) return -ENOENT; level = nilfs_btree_node_get_level(node); - ptr = nilfs_btree_node_get_ptr(btree, node, index); + ptr = nilfs_btree_node_get_ptr(node, index, + NILFS_BTREE_ROOT_NCHILDREN_MAX); path[level].bp_bh = NULL; path[level].bp_index = index; + ncmax = nilfs_btree_nchildren_per_block(btree); for (level--; level > 0; level--) { ret = nilfs_btree_get_block(btree, ptr, &path[level].bp_bh); @@ -499,7 +591,7 @@ static int nilfs_btree_do_lookup_last(const struct nilfs_btree *btree, if (nilfs_btree_bad_node(node, level)) return -EINVAL; index = nilfs_btree_node_get_nchildren(node) - 1; - ptr = nilfs_btree_node_get_ptr(btree, node, index); + ptr = nilfs_btree_node_get_ptr(node, index, ncmax); path[level].bp_index = index; } @@ -511,51 +603,45 @@ static int nilfs_btree_do_lookup_last(const struct nilfs_btree *btree, return 0; } -static int nilfs_btree_lookup(const struct nilfs_bmap *bmap, +static int nilfs_btree_lookup(const struct nilfs_bmap *btree, __u64 key, int level, __u64 *ptrp) { - struct nilfs_btree *btree; struct nilfs_btree_path *path; - __u64 ptr; int ret; - btree = (struct nilfs_btree *)bmap; path = nilfs_btree_alloc_path(); if (path == NULL) return -ENOMEM; - ret = nilfs_btree_do_lookup(btree, path, key, &ptr, level); - - if (ptrp != NULL) - *ptrp = ptr; + ret = nilfs_btree_do_lookup(btree, path, key, ptrp, level, 0); nilfs_btree_free_path(path); return ret; } -static int nilfs_btree_lookup_contig(const struct nilfs_bmap *bmap, +static int nilfs_btree_lookup_contig(const struct nilfs_bmap *btree, __u64 key, __u64 *ptrp, unsigned maxblocks) { - struct nilfs_btree *btree = (struct nilfs_btree *)bmap; struct nilfs_btree_path *path; struct nilfs_btree_node *node; struct inode *dat = NULL; __u64 ptr, ptr2; sector_t blocknr; int level = NILFS_BTREE_LEVEL_NODE_MIN; - int ret, cnt, index, maxlevel; + int ret, cnt, index, maxlevel, ncmax; + struct nilfs_btree_readahead_info p; path = nilfs_btree_alloc_path(); if (path == NULL) return -ENOMEM; - ret = nilfs_btree_do_lookup(btree, path, key, &ptr, level); + ret = nilfs_btree_do_lookup(btree, path, key, &ptr, level, 1); if (ret < 0) goto out; - if (NILFS_BMAP_USE_VBN(bmap)) { - dat = nilfs_bmap_get_dat(bmap); + if (NILFS_BMAP_USE_VBN(btree)) { + dat = nilfs_bmap_get_dat(btree); ret = nilfs_dat_translate(dat, ptr, &blocknr); if (ret < 0) goto out; @@ -566,14 +652,14 @@ static int nilfs_btree_lookup_contig(const struct nilfs_bmap *bmap, goto end; maxlevel = nilfs_btree_height(btree) - 1; - node = nilfs_btree_get_node(btree, path, level); + node = nilfs_btree_get_node(btree, path, level, &ncmax); index = path[level].bp_index + 1; for (;;) { while (index < nilfs_btree_node_get_nchildren(node)) { if (nilfs_btree_node_get_key(node, index) != key + cnt) goto end; - ptr2 = nilfs_btree_node_get_ptr(btree, node, index); + ptr2 = nilfs_btree_node_get_ptr(node, index, ncmax); if (dat) { ret = nilfs_dat_translate(dat, ptr2, &blocknr); if (ret < 0) @@ -589,20 +675,24 @@ static int nilfs_btree_lookup_contig(const struct nilfs_bmap *bmap, break; /* look-up right sibling node */ - node = nilfs_btree_get_node(btree, path, level + 1); - index = path[level + 1].bp_index + 1; - if (index >= nilfs_btree_node_get_nchildren(node) || - nilfs_btree_node_get_key(node, index) != key + cnt) + p.node = nilfs_btree_get_node(btree, path, level + 1, &p.ncmax); + p.index = path[level + 1].bp_index + 1; + p.max_ra_blocks = 7; + if (p.index >= nilfs_btree_node_get_nchildren(p.node) || + nilfs_btree_node_get_key(p.node, p.index) != key + cnt) break; - ptr2 = nilfs_btree_node_get_ptr(btree, node, index); - path[level + 1].bp_index = index; + ptr2 = nilfs_btree_node_get_ptr(p.node, p.index, p.ncmax); + path[level + 1].bp_index = p.index; brelse(path[level].bp_bh); path[level].bp_bh = NULL; - ret = nilfs_btree_get_block(btree, ptr2, &path[level].bp_bh); + + ret = __nilfs_btree_get_block(btree, ptr2, &path[level].bp_bh, + &p); if (ret < 0) goto out; node = nilfs_btree_get_nonroot_node(path, level); + ncmax = nilfs_btree_nchildren_per_block(btree); index = 0; path[level].bp_index = index; } @@ -614,7 +704,7 @@ static int nilfs_btree_lookup_contig(const struct nilfs_bmap *bmap, return ret; } -static void nilfs_btree_promote_key(struct nilfs_btree *btree, +static void nilfs_btree_promote_key(struct nilfs_bmap *btree, struct nilfs_btree_path *path, int level, __u64 key) { @@ -636,16 +726,18 @@ static void nilfs_btree_promote_key(struct nilfs_btree *btree, } } -static void nilfs_btree_do_insert(struct nilfs_btree *btree, +static void nilfs_btree_do_insert(struct nilfs_bmap *btree, struct nilfs_btree_path *path, int level, __u64 *keyp, __u64 *ptrp) { struct nilfs_btree_node *node; + int ncblk; if (level < nilfs_btree_height(btree) - 1) { node = nilfs_btree_get_nonroot_node(path, level); - nilfs_btree_node_insert(btree, node, *keyp, *ptrp, - path[level].bp_index); + ncblk = nilfs_btree_nchildren_per_block(btree); + nilfs_btree_node_insert(node, path[level].bp_index, + *keyp, *ptrp, ncblk); if (!buffer_dirty(path[level].bp_bh)) nilfs_btnode_mark_dirty(path[level].bp_bh); @@ -655,22 +747,24 @@ static void nilfs_btree_do_insert(struct nilfs_btree *btree, 0)); } else { node = nilfs_btree_get_root(btree); - nilfs_btree_node_insert(btree, node, *keyp, *ptrp, - path[level].bp_index); + nilfs_btree_node_insert(node, path[level].bp_index, + *keyp, *ptrp, + NILFS_BTREE_ROOT_NCHILDREN_MAX); } } -static void nilfs_btree_carry_left(struct nilfs_btree *btree, +static void nilfs_btree_carry_left(struct nilfs_bmap *btree, struct nilfs_btree_path *path, int level, __u64 *keyp, __u64 *ptrp) { struct nilfs_btree_node *node, *left; - int nchildren, lnchildren, n, move; + int nchildren, lnchildren, n, move, ncblk; node = nilfs_btree_get_nonroot_node(path, level); left = nilfs_btree_get_sib_node(path, level); nchildren = nilfs_btree_node_get_nchildren(node); lnchildren = nilfs_btree_node_get_nchildren(left); + ncblk = nilfs_btree_nchildren_per_block(btree); move = 0; n = (nchildren + lnchildren + 1) / 2 - lnchildren; @@ -680,7 +774,7 @@ static void nilfs_btree_carry_left(struct nilfs_btree *btree, move = 1; } - nilfs_btree_node_move_left(btree, left, node, n); + nilfs_btree_node_move_left(left, node, n, ncblk, ncblk); if (!buffer_dirty(path[level].bp_bh)) nilfs_btnode_mark_dirty(path[level].bp_bh); @@ -705,17 +799,18 @@ static void nilfs_btree_carry_left(struct nilfs_btree *btree, nilfs_btree_do_insert(btree, path, level, keyp, ptrp); } -static void nilfs_btree_carry_right(struct nilfs_btree *btree, +static void nilfs_btree_carry_right(struct nilfs_bmap *btree, struct nilfs_btree_path *path, int level, __u64 *keyp, __u64 *ptrp) { struct nilfs_btree_node *node, *right; - int nchildren, rnchildren, n, move; + int nchildren, rnchildren, n, move, ncblk; node = nilfs_btree_get_nonroot_node(path, level); right = nilfs_btree_get_sib_node(path, level); nchildren = nilfs_btree_node_get_nchildren(node); rnchildren = nilfs_btree_node_get_nchildren(right); + ncblk = nilfs_btree_nchildren_per_block(btree); move = 0; n = (nchildren + rnchildren + 1) / 2 - rnchildren; @@ -725,7 +820,7 @@ static void nilfs_btree_carry_right(struct nilfs_btree *btree, move = 1; } - nilfs_btree_node_move_right(btree, node, right, n); + nilfs_btree_node_move_right(node, right, n, ncblk, ncblk); if (!buffer_dirty(path[level].bp_bh)) nilfs_btnode_mark_dirty(path[level].bp_bh); @@ -751,18 +846,19 @@ static void nilfs_btree_carry_right(struct nilfs_btree *btree, nilfs_btree_do_insert(btree, path, level, keyp, ptrp); } -static void nilfs_btree_split(struct nilfs_btree *btree, +static void nilfs_btree_split(struct nilfs_bmap *btree, struct nilfs_btree_path *path, int level, __u64 *keyp, __u64 *ptrp) { struct nilfs_btree_node *node, *right; __u64 newkey; __u64 newptr; - int nchildren, n, move; + int nchildren, n, move, ncblk; node = nilfs_btree_get_nonroot_node(path, level); right = nilfs_btree_get_sib_node(path, level); nchildren = nilfs_btree_node_get_nchildren(node); + ncblk = nilfs_btree_nchildren_per_block(btree); move = 0; n = (nchildren + 1) / 2; @@ -771,7 +867,7 @@ static void nilfs_btree_split(struct nilfs_btree *btree, move = 1; } - nilfs_btree_node_move_right(btree, node, right, n); + nilfs_btree_node_move_right(node, right, n, ncblk, ncblk); if (!buffer_dirty(path[level].bp_bh)) nilfs_btnode_mark_dirty(path[level].bp_bh); @@ -783,8 +879,8 @@ static void nilfs_btree_split(struct nilfs_btree *btree, if (move) { path[level].bp_index -= nilfs_btree_node_get_nchildren(node); - nilfs_btree_node_insert(btree, right, *keyp, *ptrp, - path[level].bp_index); + nilfs_btree_node_insert(right, path[level].bp_index, + *keyp, *ptrp, ncblk); *keyp = nilfs_btree_node_get_key(right, 0); *ptrp = path[level].bp_newreq.bpr_ptr; @@ -805,19 +901,21 @@ static void nilfs_btree_split(struct nilfs_btree *btree, path[level + 1].bp_index++; } -static void nilfs_btree_grow(struct nilfs_btree *btree, +static void nilfs_btree_grow(struct nilfs_bmap *btree, struct nilfs_btree_path *path, int level, __u64 *keyp, __u64 *ptrp) { struct nilfs_btree_node *root, *child; - int n; + int n, ncblk; root = nilfs_btree_get_root(btree); child = nilfs_btree_get_sib_node(path, level); + ncblk = nilfs_btree_nchildren_per_block(btree); n = nilfs_btree_node_get_nchildren(root); - nilfs_btree_node_move_right(btree, root, child, n); + nilfs_btree_node_move_right(root, child, n, + NILFS_BTREE_ROOT_NCHILDREN_MAX, ncblk); nilfs_btree_node_set_level(root, level + 1); if (!buffer_dirty(path[level].bp_sib_bh)) @@ -832,11 +930,11 @@ static void nilfs_btree_grow(struct nilfs_btree *btree, *ptrp = path[level].bp_newreq.bpr_ptr; } -static __u64 nilfs_btree_find_near(const struct nilfs_btree *btree, +static __u64 nilfs_btree_find_near(const struct nilfs_bmap *btree, const struct nilfs_btree_path *path) { struct nilfs_btree_node *node; - int level; + int level, ncmax; if (path == NULL) return NILFS_BMAP_INVALID_PTR; @@ -844,29 +942,30 @@ static __u64 nilfs_btree_find_near(const struct nilfs_btree *btree, /* left sibling */ level = NILFS_BTREE_LEVEL_NODE_MIN; if (path[level].bp_index > 0) { - node = nilfs_btree_get_node(btree, path, level); - return nilfs_btree_node_get_ptr(btree, node, - path[level].bp_index - 1); + node = nilfs_btree_get_node(btree, path, level, &ncmax); + return nilfs_btree_node_get_ptr(node, + path[level].bp_index - 1, + ncmax); } /* parent */ level = NILFS_BTREE_LEVEL_NODE_MIN + 1; if (level <= nilfs_btree_height(btree) - 1) { - node = nilfs_btree_get_node(btree, path, level); - return nilfs_btree_node_get_ptr(btree, node, - path[level].bp_index); + node = nilfs_btree_get_node(btree, path, level, &ncmax); + return nilfs_btree_node_get_ptr(node, path[level].bp_index, + ncmax); } return NILFS_BMAP_INVALID_PTR; } -static __u64 nilfs_btree_find_target_v(const struct nilfs_btree *btree, +static __u64 nilfs_btree_find_target_v(const struct nilfs_bmap *btree, const struct nilfs_btree_path *path, __u64 key) { __u64 ptr; - ptr = nilfs_bmap_find_target_seq(&btree->bt_bmap, key); + ptr = nilfs_bmap_find_target_seq(btree, key); if (ptr != NILFS_BMAP_INVALID_PTR) /* sequential access */ return ptr; @@ -877,17 +976,10 @@ static __u64 nilfs_btree_find_target_v(const struct nilfs_btree *btree, return ptr; } /* block group */ - return nilfs_bmap_find_target_in_group(&btree->bt_bmap); -} - -static void nilfs_btree_set_target_v(struct nilfs_btree *btree, __u64 key, - __u64 ptr) -{ - btree->bt_bmap.b_last_allocated_key = key; - btree->bt_bmap.b_last_allocated_ptr = ptr; + return nilfs_bmap_find_target_in_group(btree); } -static int nilfs_btree_prepare_insert(struct nilfs_btree *btree, +static int nilfs_btree_prepare_insert(struct nilfs_bmap *btree, struct nilfs_btree_path *path, int *levelp, __u64 key, __u64 ptr, struct nilfs_bmap_stats *stats) @@ -895,79 +987,78 @@ static int nilfs_btree_prepare_insert(struct nilfs_btree *btree, struct buffer_head *bh; struct nilfs_btree_node *node, *parent, *sib; __u64 sibptr; - int pindex, level, ret; + int pindex, level, ncmax, ncblk, ret; struct inode *dat = NULL; stats->bs_nblocks = 0; level = NILFS_BTREE_LEVEL_DATA; /* allocate a new ptr for data block */ - if (NILFS_BMAP_USE_VBN(&btree->bt_bmap)) { + if (NILFS_BMAP_USE_VBN(btree)) { path[level].bp_newreq.bpr_ptr = nilfs_btree_find_target_v(btree, path, key); - dat = nilfs_bmap_get_dat(&btree->bt_bmap); + dat = nilfs_bmap_get_dat(btree); } - ret = nilfs_bmap_prepare_alloc_ptr(&btree->bt_bmap, - &path[level].bp_newreq, dat); + ret = nilfs_bmap_prepare_alloc_ptr(btree, &path[level].bp_newreq, dat); if (ret < 0) goto err_out_data; + ncblk = nilfs_btree_nchildren_per_block(btree); + for (level = NILFS_BTREE_LEVEL_NODE_MIN; level < nilfs_btree_height(btree) - 1; level++) { node = nilfs_btree_get_nonroot_node(path, level); - if (nilfs_btree_node_get_nchildren(node) < - nilfs_btree_node_nchildren_max(node, btree)) { + if (nilfs_btree_node_get_nchildren(node) < ncblk) { path[level].bp_op = nilfs_btree_do_insert; stats->bs_nblocks++; goto out; } - parent = nilfs_btree_get_node(btree, path, level + 1); + parent = nilfs_btree_get_node(btree, path, level + 1, &ncmax); pindex = path[level + 1].bp_index; /* left sibling */ if (pindex > 0) { - sibptr = nilfs_btree_node_get_ptr(btree, parent, - pindex - 1); + sibptr = nilfs_btree_node_get_ptr(parent, pindex - 1, + ncmax); ret = nilfs_btree_get_block(btree, sibptr, &bh); if (ret < 0) goto err_out_child_node; sib = (struct nilfs_btree_node *)bh->b_data; - if (nilfs_btree_node_get_nchildren(sib) < - nilfs_btree_node_nchildren_max(sib, btree)) { + if (nilfs_btree_node_get_nchildren(sib) < ncblk) { path[level].bp_sib_bh = bh; path[level].bp_op = nilfs_btree_carry_left; stats->bs_nblocks++; goto out; - } else + } else { brelse(bh); + } } /* right sibling */ - if (pindex < - nilfs_btree_node_get_nchildren(parent) - 1) { - sibptr = nilfs_btree_node_get_ptr(btree, parent, - pindex + 1); + if (pindex < nilfs_btree_node_get_nchildren(parent) - 1) { + sibptr = nilfs_btree_node_get_ptr(parent, pindex + 1, + ncmax); ret = nilfs_btree_get_block(btree, sibptr, &bh); if (ret < 0) goto err_out_child_node; sib = (struct nilfs_btree_node *)bh->b_data; - if (nilfs_btree_node_get_nchildren(sib) < - nilfs_btree_node_nchildren_max(sib, btree)) { + if (nilfs_btree_node_get_nchildren(sib) < ncblk) { path[level].bp_sib_bh = bh; path[level].bp_op = nilfs_btree_carry_right; stats->bs_nblocks++; goto out; - } else + } else { brelse(bh); + } } /* split */ path[level].bp_newreq.bpr_ptr = path[level - 1].bp_newreq.bpr_ptr + 1; - ret = nilfs_bmap_prepare_alloc_ptr(&btree->bt_bmap, + ret = nilfs_bmap_prepare_alloc_ptr(btree, &path[level].bp_newreq, dat); if (ret < 0) goto err_out_child_node; @@ -979,9 +1070,8 @@ static int nilfs_btree_prepare_insert(struct nilfs_btree *btree, stats->bs_nblocks++; - nilfs_btree_node_init(btree, - (struct nilfs_btree_node *)bh->b_data, - 0, level, 0, NULL, NULL); + sib = (struct nilfs_btree_node *)bh->b_data; + nilfs_btree_node_init(sib, 0, level, 0, ncblk, NULL, NULL); path[level].bp_sib_bh = bh; path[level].bp_op = nilfs_btree_split; } @@ -989,7 +1079,7 @@ static int nilfs_btree_prepare_insert(struct nilfs_btree *btree, /* root */ node = nilfs_btree_get_root(btree); if (nilfs_btree_node_get_nchildren(node) < - nilfs_btree_node_nchildren_max(node, btree)) { + NILFS_BTREE_ROOT_NCHILDREN_MAX) { path[level].bp_op = nilfs_btree_do_insert; stats->bs_nblocks++; goto out; @@ -997,8 +1087,7 @@ static int nilfs_btree_prepare_insert(struct nilfs_btree *btree, /* grow */ path[level].bp_newreq.bpr_ptr = path[level - 1].bp_newreq.bpr_ptr + 1; - ret = nilfs_bmap_prepare_alloc_ptr(&btree->bt_bmap, - &path[level].bp_newreq, dat); + ret = nilfs_bmap_prepare_alloc_ptr(btree, &path[level].bp_newreq, dat); if (ret < 0) goto err_out_child_node; ret = nilfs_btree_get_new_block(btree, path[level].bp_newreq.bpr_ptr, @@ -1006,8 +1095,8 @@ static int nilfs_btree_prepare_insert(struct nilfs_btree *btree, if (ret < 0) goto err_out_curr_node; - nilfs_btree_node_init(btree, (struct nilfs_btree_node *)bh->b_data, - 0, level, 0, NULL, NULL); + nilfs_btree_node_init((struct nilfs_btree_node *)bh->b_data, + 0, level, 0, ncblk, NULL, NULL); path[level].bp_sib_bh = bh; path[level].bp_op = nilfs_btree_grow; @@ -1024,25 +1113,22 @@ static int nilfs_btree_prepare_insert(struct nilfs_btree *btree, /* error */ err_out_curr_node: - nilfs_bmap_abort_alloc_ptr(&btree->bt_bmap, &path[level].bp_newreq, - dat); + nilfs_bmap_abort_alloc_ptr(btree, &path[level].bp_newreq, dat); err_out_child_node: for (level--; level > NILFS_BTREE_LEVEL_DATA; level--) { nilfs_btnode_delete(path[level].bp_sib_bh); - nilfs_bmap_abort_alloc_ptr(&btree->bt_bmap, - &path[level].bp_newreq, dat); + nilfs_bmap_abort_alloc_ptr(btree, &path[level].bp_newreq, dat); } - nilfs_bmap_abort_alloc_ptr(&btree->bt_bmap, &path[level].bp_newreq, - dat); + nilfs_bmap_abort_alloc_ptr(btree, &path[level].bp_newreq, dat); err_out_data: *levelp = level; stats->bs_nblocks = 0; return ret; } -static void nilfs_btree_commit_insert(struct nilfs_btree *btree, +static void nilfs_btree_commit_insert(struct nilfs_bmap *btree, struct nilfs_btree_path *path, int maxlevel, __u64 key, __u64 ptr) { @@ -1051,35 +1137,33 @@ static void nilfs_btree_commit_insert(struct nilfs_btree *btree, set_buffer_nilfs_volatile((struct buffer_head *)((unsigned long)ptr)); ptr = path[NILFS_BTREE_LEVEL_DATA].bp_newreq.bpr_ptr; - if (NILFS_BMAP_USE_VBN(&btree->bt_bmap)) { - nilfs_btree_set_target_v(btree, key, ptr); - dat = nilfs_bmap_get_dat(&btree->bt_bmap); + if (NILFS_BMAP_USE_VBN(btree)) { + nilfs_bmap_set_target_v(btree, key, ptr); + dat = nilfs_bmap_get_dat(btree); } for (level = NILFS_BTREE_LEVEL_NODE_MIN; level <= maxlevel; level++) { - nilfs_bmap_commit_alloc_ptr(&btree->bt_bmap, + nilfs_bmap_commit_alloc_ptr(btree, &path[level - 1].bp_newreq, dat); path[level].bp_op(btree, path, level, &key, &ptr); } - if (!nilfs_bmap_dirty(&btree->bt_bmap)) - nilfs_bmap_set_dirty(&btree->bt_bmap); + if (!nilfs_bmap_dirty(btree)) + nilfs_bmap_set_dirty(btree); } -static int nilfs_btree_insert(struct nilfs_bmap *bmap, __u64 key, __u64 ptr) +static int nilfs_btree_insert(struct nilfs_bmap *btree, __u64 key, __u64 ptr) { - struct nilfs_btree *btree; struct nilfs_btree_path *path; struct nilfs_bmap_stats stats; int level, ret; - btree = (struct nilfs_btree *)bmap; path = nilfs_btree_alloc_path(); if (path == NULL) return -ENOMEM; ret = nilfs_btree_do_lookup(btree, path, key, NULL, - NILFS_BTREE_LEVEL_NODE_MIN); + NILFS_BTREE_LEVEL_NODE_MIN, 0); if (ret != -ENOENT) { if (ret == 0) ret = -EEXIST; @@ -1090,23 +1174,25 @@ static int nilfs_btree_insert(struct nilfs_bmap *bmap, __u64 key, __u64 ptr) if (ret < 0) goto out; nilfs_btree_commit_insert(btree, path, level, key, ptr); - nilfs_bmap_add_blocks(bmap, stats.bs_nblocks); + nilfs_bmap_add_blocks(btree, stats.bs_nblocks); out: nilfs_btree_free_path(path); return ret; } -static void nilfs_btree_do_delete(struct nilfs_btree *btree, +static void nilfs_btree_do_delete(struct nilfs_bmap *btree, struct nilfs_btree_path *path, int level, __u64 *keyp, __u64 *ptrp) { struct nilfs_btree_node *node; + int ncblk; if (level < nilfs_btree_height(btree) - 1) { node = nilfs_btree_get_nonroot_node(path, level); - nilfs_btree_node_delete(btree, node, keyp, ptrp, - path[level].bp_index); + ncblk = nilfs_btree_nchildren_per_block(btree); + nilfs_btree_node_delete(node, path[level].bp_index, + keyp, ptrp, ncblk); if (!buffer_dirty(path[level].bp_bh)) nilfs_btnode_mark_dirty(path[level].bp_bh); if (path[level].bp_index == 0) @@ -1114,17 +1200,18 @@ static void nilfs_btree_do_delete(struct nilfs_btree *btree, nilfs_btree_node_get_key(node, 0)); } else { node = nilfs_btree_get_root(btree); - nilfs_btree_node_delete(btree, node, keyp, ptrp, - path[level].bp_index); + nilfs_btree_node_delete(node, path[level].bp_index, + keyp, ptrp, + NILFS_BTREE_ROOT_NCHILDREN_MAX); } } -static void nilfs_btree_borrow_left(struct nilfs_btree *btree, +static void nilfs_btree_borrow_left(struct nilfs_bmap *btree, struct nilfs_btree_path *path, int level, __u64 *keyp, __u64 *ptrp) { struct nilfs_btree_node *node, *left; - int nchildren, lnchildren, n; + int nchildren, lnchildren, n, ncblk; nilfs_btree_do_delete(btree, path, level, keyp, ptrp); @@ -1132,10 +1219,11 @@ static void nilfs_btree_borrow_left(struct nilfs_btree *btree, left = nilfs_btree_get_sib_node(path, level); nchildren = nilfs_btree_node_get_nchildren(node); lnchildren = nilfs_btree_node_get_nchildren(left); + ncblk = nilfs_btree_nchildren_per_block(btree); n = (nchildren + lnchildren) / 2 - nchildren; - nilfs_btree_node_move_right(btree, left, node, n); + nilfs_btree_node_move_right(left, node, n, ncblk, ncblk); if (!buffer_dirty(path[level].bp_bh)) nilfs_btnode_mark_dirty(path[level].bp_bh); @@ -1150,12 +1238,12 @@ static void nilfs_btree_borrow_left(struct nilfs_btree *btree, path[level].bp_index += n; } -static void nilfs_btree_borrow_right(struct nilfs_btree *btree, +static void nilfs_btree_borrow_right(struct nilfs_bmap *btree, struct nilfs_btree_path *path, int level, __u64 *keyp, __u64 *ptrp) { struct nilfs_btree_node *node, *right; - int nchildren, rnchildren, n; + int nchildren, rnchildren, n, ncblk; nilfs_btree_do_delete(btree, path, level, keyp, ptrp); @@ -1163,10 +1251,11 @@ static void nilfs_btree_borrow_right(struct nilfs_btree *btree, right = nilfs_btree_get_sib_node(path, level); nchildren = nilfs_btree_node_get_nchildren(node); rnchildren = nilfs_btree_node_get_nchildren(right); + ncblk = nilfs_btree_nchildren_per_block(btree); n = (nchildren + rnchildren) / 2 - nchildren; - nilfs_btree_node_move_left(btree, node, right, n); + nilfs_btree_node_move_left(node, right, n, ncblk, ncblk); if (!buffer_dirty(path[level].bp_bh)) nilfs_btnode_mark_dirty(path[level].bp_bh); @@ -1182,21 +1271,22 @@ static void nilfs_btree_borrow_right(struct nilfs_btree *btree, path[level].bp_sib_bh = NULL; } -static void nilfs_btree_concat_left(struct nilfs_btree *btree, +static void nilfs_btree_concat_left(struct nilfs_bmap *btree, struct nilfs_btree_path *path, int level, __u64 *keyp, __u64 *ptrp) { struct nilfs_btree_node *node, *left; - int n; + int n, ncblk; nilfs_btree_do_delete(btree, path, level, keyp, ptrp); node = nilfs_btree_get_nonroot_node(path, level); left = nilfs_btree_get_sib_node(path, level); + ncblk = nilfs_btree_nchildren_per_block(btree); n = nilfs_btree_node_get_nchildren(node); - nilfs_btree_node_move_left(btree, left, node, n); + nilfs_btree_node_move_left(left, node, n, ncblk, ncblk); if (!buffer_dirty(path[level].bp_sib_bh)) nilfs_btnode_mark_dirty(path[level].bp_sib_bh); @@ -1207,21 +1297,22 @@ static void nilfs_btree_concat_left(struct nilfs_btree *btree, path[level].bp_index += nilfs_btree_node_get_nchildren(left); } -static void nilfs_btree_concat_right(struct nilfs_btree *btree, +static void nilfs_btree_concat_right(struct nilfs_bmap *btree, struct nilfs_btree_path *path, int level, __u64 *keyp, __u64 *ptrp) { struct nilfs_btree_node *node, *right; - int n; + int n, ncblk; nilfs_btree_do_delete(btree, path, level, keyp, ptrp); node = nilfs_btree_get_nonroot_node(path, level); right = nilfs_btree_get_sib_node(path, level); + ncblk = nilfs_btree_nchildren_per_block(btree); n = nilfs_btree_node_get_nchildren(right); - nilfs_btree_node_move_left(btree, node, right, n); + nilfs_btree_node_move_left(node, right, n, ncblk, ncblk); if (!buffer_dirty(path[level].bp_bh)) nilfs_btnode_mark_dirty(path[level].bp_bh); @@ -1231,29 +1322,32 @@ static void nilfs_btree_concat_right(struct nilfs_btree *btree, path[level + 1].bp_index++; } -static void nilfs_btree_shrink(struct nilfs_btree *btree, +static void nilfs_btree_shrink(struct nilfs_bmap *btree, struct nilfs_btree_path *path, int level, __u64 *keyp, __u64 *ptrp) { struct nilfs_btree_node *root, *child; - int n; + int n, ncblk; nilfs_btree_do_delete(btree, path, level, keyp, ptrp); root = nilfs_btree_get_root(btree); child = nilfs_btree_get_nonroot_node(path, level); + ncblk = nilfs_btree_nchildren_per_block(btree); - nilfs_btree_node_delete(btree, root, NULL, NULL, 0); + nilfs_btree_node_delete(root, 0, NULL, NULL, + NILFS_BTREE_ROOT_NCHILDREN_MAX); nilfs_btree_node_set_level(root, level); n = nilfs_btree_node_get_nchildren(child); - nilfs_btree_node_move_left(btree, root, child, n); + nilfs_btree_node_move_left(root, child, n, + NILFS_BTREE_ROOT_NCHILDREN_MAX, ncblk); nilfs_btnode_delete(path[level].bp_bh); path[level].bp_bh = NULL; } -static int nilfs_btree_prepare_delete(struct nilfs_btree *btree, +static int nilfs_btree_prepare_delete(struct nilfs_bmap *btree, struct nilfs_btree_path *path, int *levelp, struct nilfs_bmap_stats *stats, @@ -1262,42 +1356,43 @@ static int nilfs_btree_prepare_delete(struct nilfs_btree *btree, struct buffer_head *bh; struct nilfs_btree_node *node, *parent, *sib; __u64 sibptr; - int pindex, level, ret; + int pindex, level, ncmin, ncmax, ncblk, ret; ret = 0; stats->bs_nblocks = 0; + ncmin = NILFS_BTREE_NODE_NCHILDREN_MIN(nilfs_btree_node_size(btree)); + ncblk = nilfs_btree_nchildren_per_block(btree); + for (level = NILFS_BTREE_LEVEL_NODE_MIN; level < nilfs_btree_height(btree) - 1; level++) { node = nilfs_btree_get_nonroot_node(path, level); path[level].bp_oldreq.bpr_ptr = - nilfs_btree_node_get_ptr(btree, node, - path[level].bp_index); - ret = nilfs_bmap_prepare_end_ptr(&btree->bt_bmap, + nilfs_btree_node_get_ptr(node, path[level].bp_index, + ncblk); + ret = nilfs_bmap_prepare_end_ptr(btree, &path[level].bp_oldreq, dat); if (ret < 0) goto err_out_child_node; - if (nilfs_btree_node_get_nchildren(node) > - nilfs_btree_node_nchildren_min(node, btree)) { + if (nilfs_btree_node_get_nchildren(node) > ncmin) { path[level].bp_op = nilfs_btree_do_delete; stats->bs_nblocks++; goto out; } - parent = nilfs_btree_get_node(btree, path, level + 1); + parent = nilfs_btree_get_node(btree, path, level + 1, &ncmax); pindex = path[level + 1].bp_index; if (pindex > 0) { /* left sibling */ - sibptr = nilfs_btree_node_get_ptr(btree, parent, - pindex - 1); + sibptr = nilfs_btree_node_get_ptr(parent, pindex - 1, + ncmax); ret = nilfs_btree_get_block(btree, sibptr, &bh); if (ret < 0) goto err_out_curr_node; sib = (struct nilfs_btree_node *)bh->b_data; - if (nilfs_btree_node_get_nchildren(sib) > - nilfs_btree_node_nchildren_min(sib, btree)) { + if (nilfs_btree_node_get_nchildren(sib) > ncmin) { path[level].bp_sib_bh = bh; path[level].bp_op = nilfs_btree_borrow_left; stats->bs_nblocks++; @@ -1311,14 +1406,13 @@ static int nilfs_btree_prepare_delete(struct nilfs_btree *btree, } else if (pindex < nilfs_btree_node_get_nchildren(parent) - 1) { /* right sibling */ - sibptr = nilfs_btree_node_get_ptr(btree, parent, - pindex + 1); + sibptr = nilfs_btree_node_get_ptr(parent, pindex + 1, + ncmax); ret = nilfs_btree_get_block(btree, sibptr, &bh); if (ret < 0) goto err_out_curr_node; sib = (struct nilfs_btree_node *)bh->b_data; - if (nilfs_btree_node_get_nchildren(sib) > - nilfs_btree_node_nchildren_min(sib, btree)) { + if (nilfs_btree_node_get_nchildren(sib) > ncmin) { path[level].bp_sib_bh = bh; path[level].bp_op = nilfs_btree_borrow_right; stats->bs_nblocks++; @@ -1349,10 +1443,10 @@ static int nilfs_btree_prepare_delete(struct nilfs_btree *btree, node = nilfs_btree_get_root(btree); path[level].bp_oldreq.bpr_ptr = - nilfs_btree_node_get_ptr(btree, node, path[level].bp_index); + nilfs_btree_node_get_ptr(node, path[level].bp_index, + NILFS_BTREE_ROOT_NCHILDREN_MAX); - ret = nilfs_bmap_prepare_end_ptr(&btree->bt_bmap, - &path[level].bp_oldreq, dat); + ret = nilfs_bmap_prepare_end_ptr(btree, &path[level].bp_oldreq, dat); if (ret < 0) goto err_out_child_node; @@ -1367,75 +1461,68 @@ static int nilfs_btree_prepare_delete(struct nilfs_btree *btree, /* error */ err_out_curr_node: - nilfs_bmap_abort_end_ptr(&btree->bt_bmap, &path[level].bp_oldreq, dat); + nilfs_bmap_abort_end_ptr(btree, &path[level].bp_oldreq, dat); err_out_child_node: for (level--; level >= NILFS_BTREE_LEVEL_NODE_MIN; level--) { brelse(path[level].bp_sib_bh); - nilfs_bmap_abort_end_ptr(&btree->bt_bmap, - &path[level].bp_oldreq, dat); + nilfs_bmap_abort_end_ptr(btree, &path[level].bp_oldreq, dat); } *levelp = level; stats->bs_nblocks = 0; return ret; } -static void nilfs_btree_commit_delete(struct nilfs_btree *btree, +static void nilfs_btree_commit_delete(struct nilfs_bmap *btree, struct nilfs_btree_path *path, int maxlevel, struct inode *dat) { int level; for (level = NILFS_BTREE_LEVEL_NODE_MIN; level <= maxlevel; level++) { - nilfs_bmap_commit_end_ptr(&btree->bt_bmap, - &path[level].bp_oldreq, dat); + nilfs_bmap_commit_end_ptr(btree, &path[level].bp_oldreq, dat); path[level].bp_op(btree, path, level, NULL, NULL); } - if (!nilfs_bmap_dirty(&btree->bt_bmap)) - nilfs_bmap_set_dirty(&btree->bt_bmap); + if (!nilfs_bmap_dirty(btree)) + nilfs_bmap_set_dirty(btree); } -static int nilfs_btree_delete(struct nilfs_bmap *bmap, __u64 key) +static int nilfs_btree_delete(struct nilfs_bmap *btree, __u64 key) { - struct nilfs_btree *btree; struct nilfs_btree_path *path; struct nilfs_bmap_stats stats; struct inode *dat; int level, ret; - btree = (struct nilfs_btree *)bmap; path = nilfs_btree_alloc_path(); if (path == NULL) return -ENOMEM; ret = nilfs_btree_do_lookup(btree, path, key, NULL, - NILFS_BTREE_LEVEL_NODE_MIN); + NILFS_BTREE_LEVEL_NODE_MIN, 0); if (ret < 0) goto out; - dat = NILFS_BMAP_USE_VBN(&btree->bt_bmap) ? - nilfs_bmap_get_dat(&btree->bt_bmap) : NULL; + dat = NILFS_BMAP_USE_VBN(btree) ? nilfs_bmap_get_dat(btree) : NULL; ret = nilfs_btree_prepare_delete(btree, path, &level, &stats, dat); if (ret < 0) goto out; nilfs_btree_commit_delete(btree, path, level, dat); - nilfs_bmap_sub_blocks(bmap, stats.bs_nblocks); + nilfs_bmap_sub_blocks(btree, stats.bs_nblocks); out: nilfs_btree_free_path(path); return ret; } -static int nilfs_btree_last_key(const struct nilfs_bmap *bmap, __u64 *keyp) +static int nilfs_btree_last_key(const struct nilfs_bmap *btree, __u64 *keyp) { - struct nilfs_btree *btree; struct nilfs_btree_path *path; int ret; - btree = (struct nilfs_btree *)bmap; path = nilfs_btree_alloc_path(); if (path == NULL) return -ENOMEM; @@ -1447,16 +1534,14 @@ static int nilfs_btree_last_key(const struct nilfs_bmap *bmap, __u64 *keyp) return ret; } -static int nilfs_btree_check_delete(struct nilfs_bmap *bmap, __u64 key) +static int nilfs_btree_check_delete(struct nilfs_bmap *btree, __u64 key) { struct buffer_head *bh; - struct nilfs_btree *btree; struct nilfs_btree_node *root, *node; __u64 maxkey, nextmaxkey; __u64 ptr; int nchildren, ret; - btree = (struct nilfs_btree *)bmap; root = nilfs_btree_get_root(btree); switch (nilfs_btree_height(btree)) { case 2: @@ -1467,7 +1552,8 @@ static int nilfs_btree_check_delete(struct nilfs_bmap *bmap, __u64 key) nchildren = nilfs_btree_node_get_nchildren(root); if (nchildren > 1) return 0; - ptr = nilfs_btree_node_get_ptr(btree, root, nchildren - 1); + ptr = nilfs_btree_node_get_ptr(root, nchildren - 1, + NILFS_BTREE_ROOT_NCHILDREN_MAX); ret = nilfs_btree_get_block(btree, ptr, &bh); if (ret < 0) return ret; @@ -1487,32 +1573,33 @@ static int nilfs_btree_check_delete(struct nilfs_bmap *bmap, __u64 key) return (maxkey == key) && (nextmaxkey < NILFS_BMAP_LARGE_LOW); } -static int nilfs_btree_gather_data(struct nilfs_bmap *bmap, +static int nilfs_btree_gather_data(struct nilfs_bmap *btree, __u64 *keys, __u64 *ptrs, int nitems) { struct buffer_head *bh; - struct nilfs_btree *btree; struct nilfs_btree_node *node, *root; __le64 *dkeys; __le64 *dptrs; __u64 ptr; - int nchildren, i, ret; + int nchildren, ncmax, i, ret; - btree = (struct nilfs_btree *)bmap; root = nilfs_btree_get_root(btree); switch (nilfs_btree_height(btree)) { case 2: bh = NULL; node = root; + ncmax = NILFS_BTREE_ROOT_NCHILDREN_MAX; break; case 3: nchildren = nilfs_btree_node_get_nchildren(root); WARN_ON(nchildren > 1); - ptr = nilfs_btree_node_get_ptr(btree, root, nchildren - 1); + ptr = nilfs_btree_node_get_ptr(root, nchildren - 1, + NILFS_BTREE_ROOT_NCHILDREN_MAX); ret = nilfs_btree_get_block(btree, ptr, &bh); if (ret < 0) return ret; node = (struct nilfs_btree_node *)bh->b_data; + ncmax = nilfs_btree_nchildren_per_block(btree); break; default: node = NULL; @@ -1523,10 +1610,10 @@ static int nilfs_btree_gather_data(struct nilfs_bmap *bmap, if (nchildren < nitems) nitems = nchildren; dkeys = nilfs_btree_node_dkeys(node); - dptrs = nilfs_btree_node_dptrs(node, btree); + dptrs = nilfs_btree_node_dptrs(node, ncmax); for (i = 0; i < nitems; i++) { - keys[i] = nilfs_bmap_dkey_to_key(dkeys[i]); - ptrs[i] = nilfs_bmap_dptr_to_ptr(dptrs[i]); + keys[i] = le64_to_cpu(dkeys[i]); + ptrs[i] = le64_to_cpu(dptrs[i]); } if (bh != NULL) @@ -1536,14 +1623,13 @@ static int nilfs_btree_gather_data(struct nilfs_bmap *bmap, } static int -nilfs_btree_prepare_convert_and_insert(struct nilfs_bmap *bmap, __u64 key, +nilfs_btree_prepare_convert_and_insert(struct nilfs_bmap *btree, __u64 key, union nilfs_bmap_ptr_req *dreq, union nilfs_bmap_ptr_req *nreq, struct buffer_head **bhp, struct nilfs_bmap_stats *stats) { struct buffer_head *bh; - struct nilfs_btree *btree = (struct nilfs_btree *)bmap; struct inode *dat = NULL; int ret; @@ -1551,12 +1637,12 @@ nilfs_btree_prepare_convert_and_insert(struct nilfs_bmap *bmap, __u64 key, /* for data */ /* cannot find near ptr */ - if (NILFS_BMAP_USE_VBN(bmap)) { + if (NILFS_BMAP_USE_VBN(btree)) { dreq->bpr_ptr = nilfs_btree_find_target_v(btree, NULL, key); - dat = nilfs_bmap_get_dat(bmap); + dat = nilfs_bmap_get_dat(btree); } - ret = nilfs_bmap_prepare_alloc_ptr(bmap, dreq, dat); + ret = nilfs_bmap_prepare_alloc_ptr(btree, dreq, dat); if (ret < 0) return ret; @@ -1564,7 +1650,7 @@ nilfs_btree_prepare_convert_and_insert(struct nilfs_bmap *bmap, __u64 key, stats->bs_nblocks++; if (nreq != NULL) { nreq->bpr_ptr = dreq->bpr_ptr + 1; - ret = nilfs_bmap_prepare_alloc_ptr(bmap, nreq, dat); + ret = nilfs_bmap_prepare_alloc_ptr(btree, nreq, dat); if (ret < 0) goto err_out_dreq; @@ -1581,16 +1667,16 @@ nilfs_btree_prepare_convert_and_insert(struct nilfs_bmap *bmap, __u64 key, /* error */ err_out_nreq: - nilfs_bmap_abort_alloc_ptr(bmap, nreq, dat); + nilfs_bmap_abort_alloc_ptr(btree, nreq, dat); err_out_dreq: - nilfs_bmap_abort_alloc_ptr(bmap, dreq, dat); + nilfs_bmap_abort_alloc_ptr(btree, dreq, dat); stats->bs_nblocks = 0; return ret; } static void -nilfs_btree_commit_convert_and_insert(struct nilfs_bmap *bmap, +nilfs_btree_commit_convert_and_insert(struct nilfs_bmap *btree, __u64 key, __u64 ptr, const __u64 *keys, const __u64 *ptrs, int n, @@ -1598,57 +1684,59 @@ nilfs_btree_commit_convert_and_insert(struct nilfs_bmap *bmap, union nilfs_bmap_ptr_req *nreq, struct buffer_head *bh) { - struct nilfs_btree *btree = (struct nilfs_btree *)bmap; struct nilfs_btree_node *node; struct inode *dat; __u64 tmpptr; + int ncblk; /* free resources */ - if (bmap->b_ops->bop_clear != NULL) - bmap->b_ops->bop_clear(bmap); + if (btree->b_ops->bop_clear != NULL) + btree->b_ops->bop_clear(btree); /* ptr must be a pointer to a buffer head. */ set_buffer_nilfs_volatile((struct buffer_head *)((unsigned long)ptr)); /* convert and insert */ - dat = NILFS_BMAP_USE_VBN(bmap) ? nilfs_bmap_get_dat(bmap) : NULL; - nilfs_btree_init(bmap); + dat = NILFS_BMAP_USE_VBN(btree) ? nilfs_bmap_get_dat(btree) : NULL; + nilfs_btree_init(btree); if (nreq != NULL) { - nilfs_bmap_commit_alloc_ptr(bmap, dreq, dat); - nilfs_bmap_commit_alloc_ptr(bmap, nreq, dat); + nilfs_bmap_commit_alloc_ptr(btree, dreq, dat); + nilfs_bmap_commit_alloc_ptr(btree, nreq, dat); /* create child node at level 1 */ node = (struct nilfs_btree_node *)bh->b_data; - nilfs_btree_node_init(btree, node, 0, 1, n, keys, ptrs); - nilfs_btree_node_insert(btree, node, - key, dreq->bpr_ptr, n); + ncblk = nilfs_btree_nchildren_per_block(btree); + nilfs_btree_node_init(node, 0, 1, n, ncblk, keys, ptrs); + nilfs_btree_node_insert(node, n, key, dreq->bpr_ptr, ncblk); if (!buffer_dirty(bh)) nilfs_btnode_mark_dirty(bh); - if (!nilfs_bmap_dirty(bmap)) - nilfs_bmap_set_dirty(bmap); + if (!nilfs_bmap_dirty(btree)) + nilfs_bmap_set_dirty(btree); brelse(bh); /* create root node at level 2 */ node = nilfs_btree_get_root(btree); tmpptr = nreq->bpr_ptr; - nilfs_btree_node_init(btree, node, NILFS_BTREE_NODE_ROOT, - 2, 1, &keys[0], &tmpptr); + nilfs_btree_node_init(node, NILFS_BTREE_NODE_ROOT, 2, 1, + NILFS_BTREE_ROOT_NCHILDREN_MAX, + &keys[0], &tmpptr); } else { - nilfs_bmap_commit_alloc_ptr(bmap, dreq, dat); + nilfs_bmap_commit_alloc_ptr(btree, dreq, dat); /* create root node at level 1 */ node = nilfs_btree_get_root(btree); - nilfs_btree_node_init(btree, node, NILFS_BTREE_NODE_ROOT, - 1, n, keys, ptrs); - nilfs_btree_node_insert(btree, node, - key, dreq->bpr_ptr, n); - if (!nilfs_bmap_dirty(bmap)) - nilfs_bmap_set_dirty(bmap); + nilfs_btree_node_init(node, NILFS_BTREE_NODE_ROOT, 1, n, + NILFS_BTREE_ROOT_NCHILDREN_MAX, + keys, ptrs); + nilfs_btree_node_insert(node, n, key, dreq->bpr_ptr, + NILFS_BTREE_ROOT_NCHILDREN_MAX); + if (!nilfs_bmap_dirty(btree)) + nilfs_bmap_set_dirty(btree); } - if (NILFS_BMAP_USE_VBN(bmap)) - nilfs_btree_set_target_v(btree, key, dreq->bpr_ptr); + if (NILFS_BMAP_USE_VBN(btree)) + nilfs_bmap_set_target_v(btree, key, dreq->bpr_ptr); } /** @@ -1660,7 +1748,7 @@ nilfs_btree_commit_convert_and_insert(struct nilfs_bmap *bmap, * @ptrs: * @n: */ -int nilfs_btree_convert_and_insert(struct nilfs_bmap *bmap, +int nilfs_btree_convert_and_insert(struct nilfs_bmap *btree, __u64 key, __u64 ptr, const __u64 *keys, const __u64 *ptrs, int n) { @@ -1673,7 +1761,7 @@ int nilfs_btree_convert_and_insert(struct nilfs_bmap *bmap, di = &dreq; ni = NULL; } else if ((n + 1) <= NILFS_BTREE_NODE_NCHILDREN_MAX( - 1 << bmap->b_inode->i_blkbits)) { + 1 << btree->b_inode->i_blkbits)) { di = &dreq; ni = &nreq; } else { @@ -1682,17 +1770,17 @@ int nilfs_btree_convert_and_insert(struct nilfs_bmap *bmap, BUG(); } - ret = nilfs_btree_prepare_convert_and_insert(bmap, key, di, ni, &bh, + ret = nilfs_btree_prepare_convert_and_insert(btree, key, di, ni, &bh, &stats); if (ret < 0) return ret; - nilfs_btree_commit_convert_and_insert(bmap, key, ptr, keys, ptrs, n, + nilfs_btree_commit_convert_and_insert(btree, key, ptr, keys, ptrs, n, di, ni, bh); - nilfs_bmap_add_blocks(bmap, stats.bs_nblocks); + nilfs_bmap_add_blocks(btree, stats.bs_nblocks); return 0; } -static int nilfs_btree_propagate_p(struct nilfs_btree *btree, +static int nilfs_btree_propagate_p(struct nilfs_bmap *btree, struct nilfs_btree_path *path, int level, struct buffer_head *bh) @@ -1704,17 +1792,17 @@ static int nilfs_btree_propagate_p(struct nilfs_btree *btree, return 0; } -static int nilfs_btree_prepare_update_v(struct nilfs_btree *btree, +static int nilfs_btree_prepare_update_v(struct nilfs_bmap *btree, struct nilfs_btree_path *path, int level, struct inode *dat) { struct nilfs_btree_node *parent; - int ret; + int ncmax, ret; - parent = nilfs_btree_get_node(btree, path, level + 1); + parent = nilfs_btree_get_node(btree, path, level + 1, &ncmax); path[level].bp_oldreq.bpr_ptr = - nilfs_btree_node_get_ptr(btree, parent, - path[level + 1].bp_index); + nilfs_btree_node_get_ptr(parent, path[level + 1].bp_index, + ncmax); path[level].bp_newreq.bpr_ptr = path[level].bp_oldreq.bpr_ptr + 1; ret = nilfs_dat_prepare_update(dat, &path[level].bp_oldreq.bpr_req, &path[level].bp_newreq.bpr_req); @@ -1726,7 +1814,7 @@ static int nilfs_btree_prepare_update_v(struct nilfs_btree *btree, path[level].bp_ctxt.newkey = path[level].bp_newreq.bpr_ptr; path[level].bp_ctxt.bh = path[level].bp_bh; ret = nilfs_btnode_prepare_change_key( - &NILFS_BMAP_I(&btree->bt_bmap)->i_btnode_cache, + &NILFS_BMAP_I(btree)->i_btnode_cache, &path[level].bp_ctxt); if (ret < 0) { nilfs_dat_abort_update(dat, @@ -1739,30 +1827,31 @@ static int nilfs_btree_prepare_update_v(struct nilfs_btree *btree, return 0; } -static void nilfs_btree_commit_update_v(struct nilfs_btree *btree, +static void nilfs_btree_commit_update_v(struct nilfs_bmap *btree, struct nilfs_btree_path *path, int level, struct inode *dat) { struct nilfs_btree_node *parent; + int ncmax; nilfs_dat_commit_update(dat, &path[level].bp_oldreq.bpr_req, &path[level].bp_newreq.bpr_req, - btree->bt_bmap.b_ptr_type == NILFS_BMAP_PTR_VS); + btree->b_ptr_type == NILFS_BMAP_PTR_VS); if (buffer_nilfs_node(path[level].bp_bh)) { nilfs_btnode_commit_change_key( - &NILFS_BMAP_I(&btree->bt_bmap)->i_btnode_cache, + &NILFS_BMAP_I(btree)->i_btnode_cache, &path[level].bp_ctxt); path[level].bp_bh = path[level].bp_ctxt.bh; } set_buffer_nilfs_volatile(path[level].bp_bh); - parent = nilfs_btree_get_node(btree, path, level + 1); - nilfs_btree_node_set_ptr(btree, parent, path[level + 1].bp_index, - path[level].bp_newreq.bpr_ptr); + parent = nilfs_btree_get_node(btree, path, level + 1, &ncmax); + nilfs_btree_node_set_ptr(parent, path[level + 1].bp_index, + path[level].bp_newreq.bpr_ptr, ncmax); } -static void nilfs_btree_abort_update_v(struct nilfs_btree *btree, +static void nilfs_btree_abort_update_v(struct nilfs_bmap *btree, struct nilfs_btree_path *path, int level, struct inode *dat) { @@ -1770,11 +1859,11 @@ static void nilfs_btree_abort_update_v(struct nilfs_btree *btree, &path[level].bp_newreq.bpr_req); if (buffer_nilfs_node(path[level].bp_bh)) nilfs_btnode_abort_change_key( - &NILFS_BMAP_I(&btree->bt_bmap)->i_btnode_cache, + &NILFS_BMAP_I(btree)->i_btnode_cache, &path[level].bp_ctxt); } -static int nilfs_btree_prepare_propagate_v(struct nilfs_btree *btree, +static int nilfs_btree_prepare_propagate_v(struct nilfs_bmap *btree, struct nilfs_btree_path *path, int minlevel, int *maxlevelp, struct inode *dat) @@ -1809,7 +1898,7 @@ static int nilfs_btree_prepare_propagate_v(struct nilfs_btree *btree, return ret; } -static void nilfs_btree_commit_propagate_v(struct nilfs_btree *btree, +static void nilfs_btree_commit_propagate_v(struct nilfs_bmap *btree, struct nilfs_btree_path *path, int minlevel, int maxlevel, struct buffer_head *bh, @@ -1824,14 +1913,15 @@ static void nilfs_btree_commit_propagate_v(struct nilfs_btree *btree, nilfs_btree_commit_update_v(btree, path, level, dat); } -static int nilfs_btree_propagate_v(struct nilfs_btree *btree, +static int nilfs_btree_propagate_v(struct nilfs_bmap *btree, struct nilfs_btree_path *path, int level, struct buffer_head *bh) { int maxlevel = 0, ret; struct nilfs_btree_node *parent; - struct inode *dat = nilfs_bmap_get_dat(&btree->bt_bmap); + struct inode *dat = nilfs_bmap_get_dat(btree); __u64 ptr; + int ncmax; get_bh(bh); path[level].bp_bh = bh; @@ -1841,9 +1931,10 @@ static int nilfs_btree_propagate_v(struct nilfs_btree *btree, goto out; if (buffer_nilfs_volatile(path[level].bp_bh)) { - parent = nilfs_btree_get_node(btree, path, level + 1); - ptr = nilfs_btree_node_get_ptr(btree, parent, - path[level + 1].bp_index); + parent = nilfs_btree_get_node(btree, path, level + 1, &ncmax); + ptr = nilfs_btree_node_get_ptr(parent, + path[level + 1].bp_index, + ncmax); ret = nilfs_dat_mark_dirty(dat, ptr); if (ret < 0) goto out; @@ -1857,10 +1948,9 @@ static int nilfs_btree_propagate_v(struct nilfs_btree *btree, return ret; } -static int nilfs_btree_propagate(const struct nilfs_bmap *bmap, +static int nilfs_btree_propagate(struct nilfs_bmap *btree, struct buffer_head *bh) { - struct nilfs_btree *btree; struct nilfs_btree_path *path; struct nilfs_btree_node *node; __u64 key; @@ -1868,7 +1958,6 @@ static int nilfs_btree_propagate(const struct nilfs_bmap *bmap, WARN_ON(!buffer_dirty(bh)); - btree = (struct nilfs_btree *)bmap; path = nilfs_btree_alloc_path(); if (path == NULL) return -ENOMEM; @@ -1878,11 +1967,11 @@ static int nilfs_btree_propagate(const struct nilfs_bmap *bmap, key = nilfs_btree_node_get_key(node, 0); level = nilfs_btree_node_get_level(node); } else { - key = nilfs_bmap_data_get_key(bmap, bh); + key = nilfs_bmap_data_get_key(btree, bh); level = NILFS_BTREE_LEVEL_DATA; } - ret = nilfs_btree_do_lookup(btree, path, key, NULL, level + 1); + ret = nilfs_btree_do_lookup(btree, path, key, NULL, level + 1, 0); if (ret < 0) { if (unlikely(ret == -ENOENT)) printk(KERN_CRIT "%s: key = %llu, level == %d\n", @@ -1890,7 +1979,7 @@ static int nilfs_btree_propagate(const struct nilfs_bmap *bmap, goto out; } - ret = NILFS_BMAP_USE_VBN(bmap) ? + ret = NILFS_BMAP_USE_VBN(btree) ? nilfs_btree_propagate_v(btree, path, level, bh) : nilfs_btree_propagate_p(btree, path, level, bh); @@ -1900,13 +1989,13 @@ static int nilfs_btree_propagate(const struct nilfs_bmap *bmap, return ret; } -static int nilfs_btree_propagate_gc(const struct nilfs_bmap *bmap, +static int nilfs_btree_propagate_gc(struct nilfs_bmap *btree, struct buffer_head *bh) { - return nilfs_dat_mark_dirty(nilfs_bmap_get_dat(bmap), bh->b_blocknr); + return nilfs_dat_mark_dirty(nilfs_bmap_get_dat(btree), bh->b_blocknr); } -static void nilfs_btree_add_dirty_buffer(struct nilfs_btree *btree, +static void nilfs_btree_add_dirty_buffer(struct nilfs_bmap *btree, struct list_head *lists, struct buffer_head *bh) { @@ -1920,6 +2009,18 @@ static void nilfs_btree_add_dirty_buffer(struct nilfs_btree *btree, node = (struct nilfs_btree_node *)bh->b_data; key = nilfs_btree_node_get_key(node, 0); level = nilfs_btree_node_get_level(node); + if (level < NILFS_BTREE_LEVEL_NODE_MIN || + level >= NILFS_BTREE_LEVEL_MAX) { + dump_stack(); + printk(KERN_WARNING + "%s: invalid btree level: %d (key=%llu, ino=%lu, " + "blocknr=%llu)\n", + __func__, level, (unsigned long long)key, + NILFS_BMAP_I(btree)->vfs_inode.i_ino, + (unsigned long long)bh->b_blocknr); + return; + } + list_for_each(head, &lists[level]) { cbh = list_entry(head, struct buffer_head, b_assoc_buffers); cnode = (struct nilfs_btree_node *)cbh->b_data; @@ -1930,11 +2031,10 @@ static void nilfs_btree_add_dirty_buffer(struct nilfs_btree *btree, list_add_tail(&bh->b_assoc_buffers, head); } -static void nilfs_btree_lookup_dirty_buffers(struct nilfs_bmap *bmap, +static void nilfs_btree_lookup_dirty_buffers(struct nilfs_bmap *btree, struct list_head *listp) { - struct nilfs_btree *btree = (struct nilfs_btree *)bmap; - struct address_space *btcache = &NILFS_BMAP_I(bmap)->i_btnode_cache; + struct address_space *btcache = &NILFS_BMAP_I(btree)->i_btnode_cache; struct list_head lists[NILFS_BTREE_LEVEL_MAX]; struct pagevec pvec; struct buffer_head *bh, *head; @@ -1968,7 +2068,7 @@ static void nilfs_btree_lookup_dirty_buffers(struct nilfs_bmap *bmap, list_splice_tail(&lists[level], listp); } -static int nilfs_btree_assign_p(struct nilfs_btree *btree, +static int nilfs_btree_assign_p(struct nilfs_bmap *btree, struct nilfs_btree_path *path, int level, struct buffer_head **bh, @@ -1978,38 +2078,38 @@ static int nilfs_btree_assign_p(struct nilfs_btree *btree, struct nilfs_btree_node *parent; __u64 key; __u64 ptr; - int ret; + int ncmax, ret; - parent = nilfs_btree_get_node(btree, path, level + 1); - ptr = nilfs_btree_node_get_ptr(btree, parent, - path[level + 1].bp_index); + parent = nilfs_btree_get_node(btree, path, level + 1, &ncmax); + ptr = nilfs_btree_node_get_ptr(parent, path[level + 1].bp_index, + ncmax); if (buffer_nilfs_node(*bh)) { path[level].bp_ctxt.oldkey = ptr; path[level].bp_ctxt.newkey = blocknr; path[level].bp_ctxt.bh = *bh; ret = nilfs_btnode_prepare_change_key( - &NILFS_BMAP_I(&btree->bt_bmap)->i_btnode_cache, + &NILFS_BMAP_I(btree)->i_btnode_cache, &path[level].bp_ctxt); if (ret < 0) return ret; nilfs_btnode_commit_change_key( - &NILFS_BMAP_I(&btree->bt_bmap)->i_btnode_cache, + &NILFS_BMAP_I(btree)->i_btnode_cache, &path[level].bp_ctxt); *bh = path[level].bp_ctxt.bh; } - nilfs_btree_node_set_ptr(btree, parent, - path[level + 1].bp_index, blocknr); + nilfs_btree_node_set_ptr(parent, path[level + 1].bp_index, blocknr, + ncmax); key = nilfs_btree_node_get_key(parent, path[level + 1].bp_index); /* on-disk format */ - binfo->bi_dat.bi_blkoff = nilfs_bmap_key_to_dkey(key); + binfo->bi_dat.bi_blkoff = cpu_to_le64(key); binfo->bi_dat.bi_level = level; return 0; } -static int nilfs_btree_assign_v(struct nilfs_btree *btree, +static int nilfs_btree_assign_v(struct nilfs_bmap *btree, struct nilfs_btree_path *path, int level, struct buffer_head **bh, @@ -2017,15 +2117,15 @@ static int nilfs_btree_assign_v(struct nilfs_btree *btree, union nilfs_binfo *binfo) { struct nilfs_btree_node *parent; - struct inode *dat = nilfs_bmap_get_dat(&btree->bt_bmap); + struct inode *dat = nilfs_bmap_get_dat(btree); __u64 key; __u64 ptr; union nilfs_bmap_ptr_req req; - int ret; + int ncmax, ret; - parent = nilfs_btree_get_node(btree, path, level + 1); - ptr = nilfs_btree_node_get_ptr(btree, parent, - path[level + 1].bp_index); + parent = nilfs_btree_get_node(btree, path, level + 1, &ncmax); + ptr = nilfs_btree_node_get_ptr(parent, path[level + 1].bp_index, + ncmax); req.bpr_ptr = ptr; ret = nilfs_dat_prepare_start(dat, &req.bpr_req); if (ret < 0) @@ -2034,24 +2134,22 @@ static int nilfs_btree_assign_v(struct nilfs_btree *btree, key = nilfs_btree_node_get_key(parent, path[level + 1].bp_index); /* on-disk format */ - binfo->bi_v.bi_vblocknr = nilfs_bmap_ptr_to_dptr(ptr); - binfo->bi_v.bi_blkoff = nilfs_bmap_key_to_dkey(key); + binfo->bi_v.bi_vblocknr = cpu_to_le64(ptr); + binfo->bi_v.bi_blkoff = cpu_to_le64(key); return 0; } -static int nilfs_btree_assign(struct nilfs_bmap *bmap, +static int nilfs_btree_assign(struct nilfs_bmap *btree, struct buffer_head **bh, sector_t blocknr, union nilfs_binfo *binfo) { - struct nilfs_btree *btree; struct nilfs_btree_path *path; struct nilfs_btree_node *node; __u64 key; int level, ret; - btree = (struct nilfs_btree *)bmap; path = nilfs_btree_alloc_path(); if (path == NULL) return -ENOMEM; @@ -2061,17 +2159,17 @@ static int nilfs_btree_assign(struct nilfs_bmap *bmap, key = nilfs_btree_node_get_key(node, 0); level = nilfs_btree_node_get_level(node); } else { - key = nilfs_bmap_data_get_key(bmap, *bh); + key = nilfs_bmap_data_get_key(btree, *bh); level = NILFS_BTREE_LEVEL_DATA; } - ret = nilfs_btree_do_lookup(btree, path, key, NULL, level + 1); + ret = nilfs_btree_do_lookup(btree, path, key, NULL, level + 1, 0); if (ret < 0) { WARN_ON(ret == -ENOENT); goto out; } - ret = NILFS_BMAP_USE_VBN(bmap) ? + ret = NILFS_BMAP_USE_VBN(btree) ? nilfs_btree_assign_v(btree, path, level, bh, blocknr, binfo) : nilfs_btree_assign_p(btree, path, level, bh, blocknr, binfo); @@ -2081,7 +2179,7 @@ static int nilfs_btree_assign(struct nilfs_bmap *bmap, return ret; } -static int nilfs_btree_assign_gc(struct nilfs_bmap *bmap, +static int nilfs_btree_assign_gc(struct nilfs_bmap *btree, struct buffer_head **bh, sector_t blocknr, union nilfs_binfo *binfo) @@ -2090,7 +2188,7 @@ static int nilfs_btree_assign_gc(struct nilfs_bmap *bmap, __u64 key; int ret; - ret = nilfs_dat_move(nilfs_bmap_get_dat(bmap), (*bh)->b_blocknr, + ret = nilfs_dat_move(nilfs_bmap_get_dat(btree), (*bh)->b_blocknr, blocknr); if (ret < 0) return ret; @@ -2099,29 +2197,27 @@ static int nilfs_btree_assign_gc(struct nilfs_bmap *bmap, node = (struct nilfs_btree_node *)(*bh)->b_data; key = nilfs_btree_node_get_key(node, 0); } else - key = nilfs_bmap_data_get_key(bmap, *bh); + key = nilfs_bmap_data_get_key(btree, *bh); /* on-disk format */ binfo->bi_v.bi_vblocknr = cpu_to_le64((*bh)->b_blocknr); - binfo->bi_v.bi_blkoff = nilfs_bmap_key_to_dkey(key); + binfo->bi_v.bi_blkoff = cpu_to_le64(key); return 0; } -static int nilfs_btree_mark(struct nilfs_bmap *bmap, __u64 key, int level) +static int nilfs_btree_mark(struct nilfs_bmap *btree, __u64 key, int level) { struct buffer_head *bh; - struct nilfs_btree *btree; struct nilfs_btree_path *path; __u64 ptr; int ret; - btree = (struct nilfs_btree *)bmap; path = nilfs_btree_alloc_path(); if (path == NULL) return -ENOMEM; - ret = nilfs_btree_do_lookup(btree, path, key, &ptr, level + 1); + ret = nilfs_btree_do_lookup(btree, path, key, &ptr, level + 1, 0); if (ret < 0) { WARN_ON(ret == -ENOENT); goto out; @@ -2135,8 +2231,8 @@ static int nilfs_btree_mark(struct nilfs_bmap *bmap, __u64 key, int level) if (!buffer_dirty(bh)) nilfs_btnode_mark_dirty(bh); brelse(bh); - if (!nilfs_bmap_dirty(&btree->bt_bmap)) - nilfs_bmap_set_dirty(&btree->bt_bmap); + if (!nilfs_bmap_dirty(btree)) + nilfs_bmap_set_dirty(btree); out: nilfs_btree_free_path(path); @@ -2186,10 +2282,14 @@ static const struct nilfs_bmap_operations nilfs_btree_ops_gc = { int nilfs_btree_init(struct nilfs_bmap *bmap) { bmap->b_ops = &nilfs_btree_ops; + bmap->b_nchildren_per_block = + NILFS_BTREE_NODE_NCHILDREN_MAX(nilfs_btree_node_size(bmap)); return 0; } void nilfs_btree_init_gc(struct nilfs_bmap *bmap) { bmap->b_ops = &nilfs_btree_ops_gc; + bmap->b_nchildren_per_block = + NILFS_BTREE_NODE_NCHILDREN_MAX(nilfs_btree_node_size(bmap)); } diff --git a/fs/nilfs2/btree.h b/fs/nilfs2/btree.h index 43c8c5b541f..22c02e35b6e 100644 --- a/fs/nilfs2/btree.h +++ b/fs/nilfs2/btree.h @@ -31,14 +31,6 @@ #include "bmap.h" /** - * struct nilfs_btree - B-tree structure - * @bt_bmap: bmap base structure - */ -struct nilfs_btree { - struct nilfs_bmap bt_bmap; -}; - -/** * struct nilfs_btree_path - A path on which B-tree operations are executed * @bp_bh: buffer head of node block * @bp_sib_bh: buffer head of sibling node block @@ -54,7 +46,7 @@ struct nilfs_btree_path { union nilfs_bmap_ptr_req bp_oldreq; union nilfs_bmap_ptr_req bp_newreq; struct nilfs_btnode_chkey_ctxt bp_ctxt; - void (*bp_op)(struct nilfs_btree *, struct nilfs_btree_path *, + void (*bp_op)(struct nilfs_bmap *, struct nilfs_btree_path *, int, __u64 *, __u64 *); }; @@ -80,4 +72,6 @@ int nilfs_btree_convert_and_insert(struct nilfs_bmap *, __u64, __u64, const __u64 *, const __u64 *, int); void nilfs_btree_init_gc(struct nilfs_bmap *); +int nilfs_btree_broken_node_block(struct buffer_head *bh); + #endif /* _NILFS_BTREE_H */ diff --git a/fs/nilfs2/dir.c b/fs/nilfs2/dir.c index 85c89dfc71f..b60277b4446 100644 --- a/fs/nilfs2/dir.c +++ b/fs/nilfs2/dir.c @@ -141,7 +141,7 @@ static void nilfs_check_page(struct page *page) } for (offs = 0; offs <= limit - NILFS_DIR_REC_LEN(1); offs += rec_len) { p = (struct nilfs_dir_entry *)(kaddr + offs); - rec_len = le16_to_cpu(p->rec_len); + rec_len = nilfs_rec_len_from_disk(p->rec_len); if (rec_len < NILFS_DIR_REC_LEN(1)) goto Eshort; @@ -199,13 +199,10 @@ fail: static struct page *nilfs_get_page(struct inode *dir, unsigned long n) { struct address_space *mapping = dir->i_mapping; - struct page *page = read_cache_page(mapping, n, - (filler_t *)mapping->a_ops->readpage, NULL); + struct page *page = read_mapping_page(mapping, n, NULL); + if (!IS_ERR(page)) { - wait_on_page_locked(page); kmap(page); - if (!PageUptodate(page)) - goto fail; if (!PageChecked(page)) nilfs_check_page(page); if (PageError(page)) @@ -238,7 +235,8 @@ nilfs_match(int len, const unsigned char *name, struct nilfs_dir_entry *de) */ static struct nilfs_dir_entry *nilfs_next_entry(struct nilfs_dir_entry *p) { - return (struct nilfs_dir_entry *)((char *)p + le16_to_cpu(p->rec_len)); + return (struct nilfs_dir_entry *)((char *)p + + nilfs_rec_len_from_disk(p->rec_len)); } static unsigned char @@ -329,7 +327,7 @@ static int nilfs_readdir(struct file *filp, void *dirent, filldir_t filldir) goto success; } } - filp->f_pos += le16_to_cpu(de->rec_len); + filp->f_pos += nilfs_rec_len_from_disk(de->rec_len); } nilfs_put_page(page); } @@ -444,7 +442,7 @@ void nilfs_set_link(struct inode *dir, struct nilfs_dir_entry *de, struct page *page, struct inode *inode) { unsigned from = (char *) de - (char *) page_address(page); - unsigned to = from + le16_to_cpu(de->rec_len); + unsigned to = from + nilfs_rec_len_from_disk(de->rec_len); struct address_space *mapping = page->mapping; int err; @@ -500,7 +498,7 @@ int nilfs_add_link(struct dentry *dentry, struct inode *inode) /* We hit i_size */ name_len = 0; rec_len = chunk_size; - de->rec_len = cpu_to_le16(chunk_size); + de->rec_len = nilfs_rec_len_to_disk(chunk_size); de->inode = 0; goto got_it; } @@ -514,7 +512,7 @@ int nilfs_add_link(struct dentry *dentry, struct inode *inode) if (nilfs_match(namelen, name, de)) goto out_unlock; name_len = NILFS_DIR_REC_LEN(de->name_len); - rec_len = le16_to_cpu(de->rec_len); + rec_len = nilfs_rec_len_from_disk(de->rec_len); if (!de->inode && rec_len >= reclen) goto got_it; if (rec_len >= name_len + reclen) @@ -537,8 +535,8 @@ got_it: struct nilfs_dir_entry *de1; de1 = (struct nilfs_dir_entry *)((char *)de + name_len); - de1->rec_len = cpu_to_le16(rec_len - name_len); - de->rec_len = cpu_to_le16(name_len); + de1->rec_len = nilfs_rec_len_to_disk(rec_len - name_len); + de->rec_len = nilfs_rec_len_to_disk(name_len); de = de1; } de->name_len = namelen; @@ -569,7 +567,8 @@ int nilfs_delete_entry(struct nilfs_dir_entry *dir, struct page *page) struct inode *inode = mapping->host; char *kaddr = page_address(page); unsigned from = ((char *)dir - kaddr) & ~(nilfs_chunk_size(inode) - 1); - unsigned to = ((char *)dir - kaddr) + le16_to_cpu(dir->rec_len); + unsigned to = ((char *)dir - kaddr) + + nilfs_rec_len_from_disk(dir->rec_len); struct nilfs_dir_entry *pde = NULL; struct nilfs_dir_entry *de = (struct nilfs_dir_entry *)(kaddr + from); int err; @@ -590,7 +589,7 @@ int nilfs_delete_entry(struct nilfs_dir_entry *dir, struct page *page) err = nilfs_prepare_chunk(page, mapping, from, to); BUG_ON(err); if (pde) - pde->rec_len = cpu_to_le16(to - from); + pde->rec_len = nilfs_rec_len_to_disk(to - from); dir->inode = 0; nilfs_commit_chunk(page, mapping, from, to); inode->i_ctime = inode->i_mtime = CURRENT_TIME; @@ -624,14 +623,14 @@ int nilfs_make_empty(struct inode *inode, struct inode *parent) memset(kaddr, 0, chunk_size); de = (struct nilfs_dir_entry *)kaddr; de->name_len = 1; - de->rec_len = cpu_to_le16(NILFS_DIR_REC_LEN(1)); + de->rec_len = nilfs_rec_len_to_disk(NILFS_DIR_REC_LEN(1)); memcpy(de->name, ".\0\0", 4); de->inode = cpu_to_le64(inode->i_ino); nilfs_set_de_type(de, inode); de = (struct nilfs_dir_entry *)(kaddr + NILFS_DIR_REC_LEN(1)); de->name_len = 2; - de->rec_len = cpu_to_le16(chunk_size - NILFS_DIR_REC_LEN(1)); + de->rec_len = nilfs_rec_len_to_disk(chunk_size - NILFS_DIR_REC_LEN(1)); de->inode = cpu_to_le64(parent->i_ino); memcpy(de->name, "..\0", 4); nilfs_set_de_type(de, inode); diff --git a/fs/nilfs2/direct.c b/fs/nilfs2/direct.c index 236753df5cd..324d80c5751 100644 --- a/fs/nilfs2/direct.c +++ b/fs/nilfs2/direct.c @@ -27,47 +27,43 @@ #include "alloc.h" #include "dat.h" -static inline __le64 *nilfs_direct_dptrs(const struct nilfs_direct *direct) +static inline __le64 *nilfs_direct_dptrs(const struct nilfs_bmap *direct) { return (__le64 *) - ((struct nilfs_direct_node *)direct->d_bmap.b_u.u_data + 1); + ((struct nilfs_direct_node *)direct->b_u.u_data + 1); } static inline __u64 -nilfs_direct_get_ptr(const struct nilfs_direct *direct, __u64 key) +nilfs_direct_get_ptr(const struct nilfs_bmap *direct, __u64 key) { - return nilfs_bmap_dptr_to_ptr(*(nilfs_direct_dptrs(direct) + key)); + return le64_to_cpu(*(nilfs_direct_dptrs(direct) + key)); } -static inline void nilfs_direct_set_ptr(struct nilfs_direct *direct, +static inline void nilfs_direct_set_ptr(struct nilfs_bmap *direct, __u64 key, __u64 ptr) { - *(nilfs_direct_dptrs(direct) + key) = nilfs_bmap_ptr_to_dptr(ptr); + *(nilfs_direct_dptrs(direct) + key) = cpu_to_le64(ptr); } -static int nilfs_direct_lookup(const struct nilfs_bmap *bmap, +static int nilfs_direct_lookup(const struct nilfs_bmap *direct, __u64 key, int level, __u64 *ptrp) { - struct nilfs_direct *direct; __u64 ptr; - direct = (struct nilfs_direct *)bmap; /* XXX: use macro for level 1 */ if (key > NILFS_DIRECT_KEY_MAX || level != 1) return -ENOENT; ptr = nilfs_direct_get_ptr(direct, key); if (ptr == NILFS_BMAP_INVALID_PTR) return -ENOENT; - if (ptrp != NULL) - *ptrp = ptr; + *ptrp = ptr; return 0; } -static int nilfs_direct_lookup_contig(const struct nilfs_bmap *bmap, +static int nilfs_direct_lookup_contig(const struct nilfs_bmap *direct, __u64 key, __u64 *ptrp, unsigned maxblocks) { - struct nilfs_direct *direct = (struct nilfs_direct *)bmap; struct inode *dat = NULL; __u64 ptr, ptr2; sector_t blocknr; @@ -79,8 +75,8 @@ static int nilfs_direct_lookup_contig(const struct nilfs_bmap *bmap, if (ptr == NILFS_BMAP_INVALID_PTR) return -ENOENT; - if (NILFS_BMAP_USE_VBN(bmap)) { - dat = nilfs_bmap_get_dat(bmap); + if (NILFS_BMAP_USE_VBN(direct)) { + dat = nilfs_bmap_get_dat(direct); ret = nilfs_dat_translate(dat, ptr, &blocknr); if (ret < 0) return ret; @@ -106,29 +102,21 @@ static int nilfs_direct_lookup_contig(const struct nilfs_bmap *bmap, } static __u64 -nilfs_direct_find_target_v(const struct nilfs_direct *direct, __u64 key) +nilfs_direct_find_target_v(const struct nilfs_bmap *direct, __u64 key) { __u64 ptr; - ptr = nilfs_bmap_find_target_seq(&direct->d_bmap, key); + ptr = nilfs_bmap_find_target_seq(direct, key); if (ptr != NILFS_BMAP_INVALID_PTR) /* sequential access */ return ptr; else /* block group */ - return nilfs_bmap_find_target_in_group(&direct->d_bmap); -} - -static void nilfs_direct_set_target_v(struct nilfs_direct *direct, - __u64 key, __u64 ptr) -{ - direct->d_bmap.b_last_allocated_key = key; - direct->d_bmap.b_last_allocated_ptr = ptr; + return nilfs_bmap_find_target_in_group(direct); } static int nilfs_direct_insert(struct nilfs_bmap *bmap, __u64 key, __u64 ptr) { - struct nilfs_direct *direct = (struct nilfs_direct *)bmap; union nilfs_bmap_ptr_req req; struct inode *dat = NULL; struct buffer_head *bh; @@ -136,11 +124,11 @@ static int nilfs_direct_insert(struct nilfs_bmap *bmap, __u64 key, __u64 ptr) if (key > NILFS_DIRECT_KEY_MAX) return -ENOENT; - if (nilfs_direct_get_ptr(direct, key) != NILFS_BMAP_INVALID_PTR) + if (nilfs_direct_get_ptr(bmap, key) != NILFS_BMAP_INVALID_PTR) return -EEXIST; if (NILFS_BMAP_USE_VBN(bmap)) { - req.bpr_ptr = nilfs_direct_find_target_v(direct, key); + req.bpr_ptr = nilfs_direct_find_target_v(bmap, key); dat = nilfs_bmap_get_dat(bmap); } ret = nilfs_bmap_prepare_alloc_ptr(bmap, &req, dat); @@ -150,13 +138,13 @@ static int nilfs_direct_insert(struct nilfs_bmap *bmap, __u64 key, __u64 ptr) set_buffer_nilfs_volatile(bh); nilfs_bmap_commit_alloc_ptr(bmap, &req, dat); - nilfs_direct_set_ptr(direct, key, req.bpr_ptr); + nilfs_direct_set_ptr(bmap, key, req.bpr_ptr); if (!nilfs_bmap_dirty(bmap)) nilfs_bmap_set_dirty(bmap); if (NILFS_BMAP_USE_VBN(bmap)) - nilfs_direct_set_target_v(direct, key, req.bpr_ptr); + nilfs_bmap_set_target_v(bmap, key, req.bpr_ptr); nilfs_bmap_add_blocks(bmap, 1); } @@ -165,33 +153,30 @@ static int nilfs_direct_insert(struct nilfs_bmap *bmap, __u64 key, __u64 ptr) static int nilfs_direct_delete(struct nilfs_bmap *bmap, __u64 key) { - struct nilfs_direct *direct = (struct nilfs_direct *)bmap; union nilfs_bmap_ptr_req req; struct inode *dat; int ret; if (key > NILFS_DIRECT_KEY_MAX || - nilfs_direct_get_ptr(direct, key) == NILFS_BMAP_INVALID_PTR) + nilfs_direct_get_ptr(bmap, key) == NILFS_BMAP_INVALID_PTR) return -ENOENT; dat = NILFS_BMAP_USE_VBN(bmap) ? nilfs_bmap_get_dat(bmap) : NULL; - req.bpr_ptr = nilfs_direct_get_ptr(direct, key); + req.bpr_ptr = nilfs_direct_get_ptr(bmap, key); ret = nilfs_bmap_prepare_end_ptr(bmap, &req, dat); if (!ret) { nilfs_bmap_commit_end_ptr(bmap, &req, dat); - nilfs_direct_set_ptr(direct, key, NILFS_BMAP_INVALID_PTR); + nilfs_direct_set_ptr(bmap, key, NILFS_BMAP_INVALID_PTR); nilfs_bmap_sub_blocks(bmap, 1); } return ret; } -static int nilfs_direct_last_key(const struct nilfs_bmap *bmap, __u64 *keyp) +static int nilfs_direct_last_key(const struct nilfs_bmap *direct, __u64 *keyp) { - struct nilfs_direct *direct; __u64 key, lastkey; - direct = (struct nilfs_direct *)bmap; lastkey = NILFS_DIRECT_KEY_MAX + 1; for (key = NILFS_DIRECT_KEY_MIN; key <= NILFS_DIRECT_KEY_MAX; key++) if (nilfs_direct_get_ptr(direct, key) != @@ -211,15 +196,13 @@ static int nilfs_direct_check_insert(const struct nilfs_bmap *bmap, __u64 key) return key > NILFS_DIRECT_KEY_MAX; } -static int nilfs_direct_gather_data(struct nilfs_bmap *bmap, +static int nilfs_direct_gather_data(struct nilfs_bmap *direct, __u64 *keys, __u64 *ptrs, int nitems) { - struct nilfs_direct *direct; __u64 key; __u64 ptr; int n; - direct = (struct nilfs_direct *)bmap; if (nitems > NILFS_DIRECT_NBLOCKS) nitems = NILFS_DIRECT_NBLOCKS; n = 0; @@ -237,7 +220,6 @@ static int nilfs_direct_gather_data(struct nilfs_bmap *bmap, int nilfs_direct_delete_and_convert(struct nilfs_bmap *bmap, __u64 key, __u64 *keys, __u64 *ptrs, int n) { - struct nilfs_direct *direct; __le64 *dptrs; int ret, i, j; @@ -253,12 +235,11 @@ int nilfs_direct_delete_and_convert(struct nilfs_bmap *bmap, bmap->b_ops->bop_clear(bmap); /* convert */ - direct = (struct nilfs_direct *)bmap; - dptrs = nilfs_direct_dptrs(direct); + dptrs = nilfs_direct_dptrs(bmap); for (i = 0, j = 0; i < NILFS_DIRECT_NBLOCKS; i++) { if ((j < n) && (i == keys[j])) { dptrs[i] = (i != key) ? - nilfs_bmap_ptr_to_dptr(ptrs[j]) : + cpu_to_le64(ptrs[j]) : NILFS_BMAP_INVALID_PTR; j++; } else @@ -269,10 +250,9 @@ int nilfs_direct_delete_and_convert(struct nilfs_bmap *bmap, return 0; } -static int nilfs_direct_propagate(const struct nilfs_bmap *bmap, +static int nilfs_direct_propagate(struct nilfs_bmap *bmap, struct buffer_head *bh) { - struct nilfs_direct *direct = (struct nilfs_direct *)bmap; struct nilfs_palloc_req oldreq, newreq; struct inode *dat; __u64 key; @@ -284,7 +264,7 @@ static int nilfs_direct_propagate(const struct nilfs_bmap *bmap, dat = nilfs_bmap_get_dat(bmap); key = nilfs_bmap_data_get_key(bmap, bh); - ptr = nilfs_direct_get_ptr(direct, key); + ptr = nilfs_direct_get_ptr(bmap, key); if (!buffer_nilfs_volatile(bh)) { oldreq.pr_entry_nr = ptr; newreq.pr_entry_nr = ptr; @@ -294,20 +274,20 @@ static int nilfs_direct_propagate(const struct nilfs_bmap *bmap, nilfs_dat_commit_update(dat, &oldreq, &newreq, bmap->b_ptr_type == NILFS_BMAP_PTR_VS); set_buffer_nilfs_volatile(bh); - nilfs_direct_set_ptr(direct, key, newreq.pr_entry_nr); + nilfs_direct_set_ptr(bmap, key, newreq.pr_entry_nr); } else ret = nilfs_dat_mark_dirty(dat, ptr); return ret; } -static int nilfs_direct_assign_v(struct nilfs_direct *direct, +static int nilfs_direct_assign_v(struct nilfs_bmap *direct, __u64 key, __u64 ptr, struct buffer_head **bh, sector_t blocknr, union nilfs_binfo *binfo) { - struct inode *dat = nilfs_bmap_get_dat(&direct->d_bmap); + struct inode *dat = nilfs_bmap_get_dat(direct); union nilfs_bmap_ptr_req req; int ret; @@ -315,13 +295,13 @@ static int nilfs_direct_assign_v(struct nilfs_direct *direct, ret = nilfs_dat_prepare_start(dat, &req.bpr_req); if (!ret) { nilfs_dat_commit_start(dat, &req.bpr_req, blocknr); - binfo->bi_v.bi_vblocknr = nilfs_bmap_ptr_to_dptr(ptr); - binfo->bi_v.bi_blkoff = nilfs_bmap_key_to_dkey(key); + binfo->bi_v.bi_vblocknr = cpu_to_le64(ptr); + binfo->bi_v.bi_blkoff = cpu_to_le64(key); } return ret; } -static int nilfs_direct_assign_p(struct nilfs_direct *direct, +static int nilfs_direct_assign_p(struct nilfs_bmap *direct, __u64 key, __u64 ptr, struct buffer_head **bh, sector_t blocknr, @@ -329,7 +309,7 @@ static int nilfs_direct_assign_p(struct nilfs_direct *direct, { nilfs_direct_set_ptr(direct, key, blocknr); - binfo->bi_dat.bi_blkoff = nilfs_bmap_key_to_dkey(key); + binfo->bi_dat.bi_blkoff = cpu_to_le64(key); binfo->bi_dat.bi_level = 0; return 0; @@ -340,18 +320,16 @@ static int nilfs_direct_assign(struct nilfs_bmap *bmap, sector_t blocknr, union nilfs_binfo *binfo) { - struct nilfs_direct *direct; __u64 key; __u64 ptr; - direct = (struct nilfs_direct *)bmap; key = nilfs_bmap_data_get_key(bmap, *bh); if (unlikely(key > NILFS_DIRECT_KEY_MAX)) { printk(KERN_CRIT "%s: invalid key: %llu\n", __func__, (unsigned long long)key); return -EINVAL; } - ptr = nilfs_direct_get_ptr(direct, key); + ptr = nilfs_direct_get_ptr(bmap, key); if (unlikely(ptr == NILFS_BMAP_INVALID_PTR)) { printk(KERN_CRIT "%s: invalid pointer: %llu\n", __func__, (unsigned long long)ptr); @@ -359,8 +337,8 @@ static int nilfs_direct_assign(struct nilfs_bmap *bmap, } return NILFS_BMAP_USE_VBN(bmap) ? - nilfs_direct_assign_v(direct, key, ptr, bh, blocknr, binfo) : - nilfs_direct_assign_p(direct, key, ptr, bh, blocknr, binfo); + nilfs_direct_assign_v(bmap, key, ptr, bh, blocknr, binfo) : + nilfs_direct_assign_p(bmap, key, ptr, bh, blocknr, binfo); } static const struct nilfs_bmap_operations nilfs_direct_ops = { diff --git a/fs/nilfs2/direct.h b/fs/nilfs2/direct.h index a5ffd66e25d..dc643de20a2 100644 --- a/fs/nilfs2/direct.h +++ b/fs/nilfs2/direct.h @@ -28,8 +28,6 @@ #include "bmap.h" -struct nilfs_direct; - /** * struct nilfs_direct_node - direct node * @dn_flags: flags @@ -40,15 +38,6 @@ struct nilfs_direct_node { __u8 pad[7]; }; -/** - * struct nilfs_direct - direct mapping - * @d_bmap: bmap structure - */ -struct nilfs_direct { - struct nilfs_bmap d_bmap; -}; - - #define NILFS_DIRECT_NBLOCKS (NILFS_BMAP_SIZE / sizeof(__le64) - 1) #define NILFS_DIRECT_KEY_MIN 0 #define NILFS_DIRECT_KEY_MAX (NILFS_DIRECT_NBLOCKS - 1) diff --git a/fs/nilfs2/gcinode.c b/fs/nilfs2/gcinode.c index 145f03cd7d3..bed3a783129 100644 --- a/fs/nilfs2/gcinode.c +++ b/fs/nilfs2/gcinode.c @@ -48,6 +48,8 @@ #include <linux/slab.h> #include <linux/swap.h> #include "nilfs.h" +#include "btree.h" +#include "btnode.h" #include "page.h" #include "mdt.h" #include "dat.h" @@ -149,8 +151,10 @@ int nilfs_gccache_submit_read_data(struct inode *inode, sector_t blkoff, int nilfs_gccache_submit_read_node(struct inode *inode, sector_t pbn, __u64 vbn, struct buffer_head **out_bh) { - int ret = nilfs_btnode_submit_block(&NILFS_I(inode)->i_btnode_cache, - vbn ? : pbn, pbn, out_bh); + int ret; + + ret = nilfs_btnode_submit_block(&NILFS_I(inode)->i_btnode_cache, + vbn ? : pbn, pbn, READ, out_bh, &pbn); if (ret == -EEXIST) /* internal code (cache hit) */ ret = 0; return ret; @@ -164,10 +168,15 @@ int nilfs_gccache_wait_and_mark_dirty(struct buffer_head *bh) if (buffer_dirty(bh)) return -EEXIST; - if (buffer_nilfs_node(bh)) + if (buffer_nilfs_node(bh)) { + if (nilfs_btree_broken_node_block(bh)) { + clear_buffer_uptodate(bh); + return -EIO; + } nilfs_btnode_mark_dirty(bh); - else + } else { nilfs_mdt_mark_buffer_dirty(bh); + } return 0; } diff --git a/fs/nilfs2/mdt.c b/fs/nilfs2/mdt.c index 024be8c35bb..d01aff4957d 100644 --- a/fs/nilfs2/mdt.c +++ b/fs/nilfs2/mdt.c @@ -28,6 +28,7 @@ #include <linux/swap.h> #include <linux/slab.h> #include "nilfs.h" +#include "btnode.h" #include "segment.h" #include "page.h" #include "mdt.h" diff --git a/fs/nilfs2/nilfs.h b/fs/nilfs2/nilfs.h index 47d6d792812..0842d775b3e 100644 --- a/fs/nilfs2/nilfs.h +++ b/fs/nilfs2/nilfs.h @@ -32,7 +32,6 @@ #include "the_nilfs.h" #include "sb.h" #include "bmap.h" -#include "bmap_union.h" /* * nilfs inode data in memory @@ -41,7 +40,7 @@ struct nilfs_inode_info { __u32 i_flags; unsigned long i_state; /* Dynamic state flags */ struct nilfs_bmap *i_bmap; - union nilfs_bmap_union i_bmap_union; + struct nilfs_bmap i_bmap_data; __u64 i_xattr; /* sector_t ??? */ __u32 i_dir_start_lookup; __u64 i_cno; /* check point number for GC inode */ @@ -71,9 +70,7 @@ static inline struct nilfs_inode_info *NILFS_I(const struct inode *inode) static inline struct nilfs_inode_info * NILFS_BMAP_I(const struct nilfs_bmap *bmap) { - return container_of((union nilfs_bmap_union *)bmap, - struct nilfs_inode_info, - i_bmap_union); + return container_of(bmap, struct nilfs_inode_info, i_bmap_data); } static inline struct inode *NILFS_BTNC_I(struct address_space *btnc) @@ -107,6 +104,14 @@ enum { }; /* + * commit flags for nilfs_commit_super and nilfs_sync_super + */ +enum { + NILFS_SB_COMMIT = 0, /* Commit a super block alternately */ + NILFS_SB_COMMIT_ALL /* Commit both super blocks */ +}; + +/* * Macros to check inode numbers */ #define NILFS_MDT_INO_BITS \ @@ -270,7 +275,14 @@ extern struct nilfs_super_block * nilfs_read_super_block(struct super_block *, u64, int, struct buffer_head **); extern int nilfs_store_magic_and_option(struct super_block *, struct nilfs_super_block *, char *); +extern int nilfs_check_feature_compatibility(struct super_block *, + struct nilfs_super_block *); +extern void nilfs_set_log_cursor(struct nilfs_super_block *, + struct the_nilfs *); +extern struct nilfs_super_block **nilfs_prepare_super(struct nilfs_sb_info *, + int flip); extern int nilfs_commit_super(struct nilfs_sb_info *, int); +extern int nilfs_cleanup_super(struct nilfs_sb_info *); extern int nilfs_attach_checkpoint(struct nilfs_sb_info *, __u64); extern void nilfs_detach_checkpoint(struct nilfs_sb_info *); diff --git a/fs/nilfs2/page.c b/fs/nilfs2/page.c index 8de3e1e4813..aab11db2cb0 100644 --- a/fs/nilfs2/page.c +++ b/fs/nilfs2/page.c @@ -37,7 +37,8 @@ #define NILFS_BUFFER_INHERENT_BITS \ ((1UL << BH_Uptodate) | (1UL << BH_Mapped) | (1UL << BH_NILFS_Node) | \ - (1UL << BH_NILFS_Volatile) | (1UL << BH_NILFS_Allocated)) + (1UL << BH_NILFS_Volatile) | (1UL << BH_NILFS_Allocated) | \ + (1UL << BH_NILFS_Checked)) static struct buffer_head * __nilfs_get_page_block(struct page *page, unsigned long block, pgoff_t index, @@ -129,6 +130,7 @@ void nilfs_forget_buffer(struct buffer_head *bh) lock_buffer(bh); clear_buffer_nilfs_volatile(bh); + clear_buffer_nilfs_checked(bh); clear_buffer_dirty(bh); if (nilfs_page_buffers_clean(page)) __nilfs_clear_page_dirty(page); @@ -480,6 +482,7 @@ void nilfs_clear_dirty_pages(struct address_space *mapping) lock_buffer(bh); clear_buffer_dirty(bh); clear_buffer_nilfs_volatile(bh); + clear_buffer_nilfs_checked(bh); clear_buffer_uptodate(bh); clear_buffer_mapped(bh); unlock_buffer(bh); diff --git a/fs/nilfs2/page.h b/fs/nilfs2/page.h index 8abca4d1c1f..f53d8da41ed 100644 --- a/fs/nilfs2/page.h +++ b/fs/nilfs2/page.h @@ -34,11 +34,13 @@ enum { BH_NILFS_Allocated = BH_PrivateStart, BH_NILFS_Node, BH_NILFS_Volatile, + BH_NILFS_Checked, }; BUFFER_FNS(NILFS_Allocated, nilfs_allocated) /* nilfs private buffers */ BUFFER_FNS(NILFS_Node, nilfs_node) /* nilfs node buffers */ BUFFER_FNS(NILFS_Volatile, nilfs_volatile) +BUFFER_FNS(NILFS_Checked, nilfs_checked) /* buffer is verified */ void nilfs_mark_buffer_dirty(struct buffer_head *bh); diff --git a/fs/nilfs2/recovery.c b/fs/nilfs2/recovery.c index bae2a516b4e..83e3d8c61a0 100644 --- a/fs/nilfs2/recovery.c +++ b/fs/nilfs2/recovery.c @@ -91,27 +91,9 @@ static int nilfs_warn_segment_error(int err) return -EINVAL; } -static void store_segsum_info(struct nilfs_segsum_info *ssi, - struct nilfs_segment_summary *sum, - unsigned int blocksize) -{ - ssi->flags = le16_to_cpu(sum->ss_flags); - ssi->seg_seq = le64_to_cpu(sum->ss_seq); - ssi->ctime = le64_to_cpu(sum->ss_create); - ssi->next = le64_to_cpu(sum->ss_next); - ssi->nblocks = le32_to_cpu(sum->ss_nblocks); - ssi->nfinfo = le32_to_cpu(sum->ss_nfinfo); - ssi->sumbytes = le32_to_cpu(sum->ss_sumbytes); - - ssi->nsumblk = DIV_ROUND_UP(ssi->sumbytes, blocksize); - ssi->nfileblk = ssi->nblocks - ssi->nsumblk - !!NILFS_SEG_HAS_SR(ssi); - - /* need to verify ->ss_bytes field if read ->ss_cno */ -} - /** - * calc_crc_cont - check CRC of blocks continuously - * @sbi: nilfs_sb_info + * nilfs_compute_checksum - compute checksum of blocks continuously + * @nilfs: nilfs object * @bhs: buffer head of start block * @sum: place to store result * @offset: offset bytes in the first block @@ -119,23 +101,25 @@ static void store_segsum_info(struct nilfs_segsum_info *ssi, * @start: DBN of start block * @nblock: number of blocks to be checked */ -static int calc_crc_cont(struct nilfs_sb_info *sbi, struct buffer_head *bhs, - u32 *sum, unsigned long offset, u64 check_bytes, - sector_t start, unsigned long nblock) +static int nilfs_compute_checksum(struct the_nilfs *nilfs, + struct buffer_head *bhs, u32 *sum, + unsigned long offset, u64 check_bytes, + sector_t start, unsigned long nblock) { - unsigned long blocksize = sbi->s_super->s_blocksize; + unsigned int blocksize = nilfs->ns_blocksize; unsigned long size; u32 crc; BUG_ON(offset >= blocksize); check_bytes -= offset; size = min_t(u64, check_bytes, blocksize - offset); - crc = crc32_le(sbi->s_nilfs->ns_crc_seed, + crc = crc32_le(nilfs->ns_crc_seed, (unsigned char *)bhs->b_data + offset, size); if (--nblock > 0) { do { - struct buffer_head *bh - = sb_bread(sbi->s_super, ++start); + struct buffer_head *bh; + + bh = __bread(nilfs->ns_bdev, ++start, blocksize); if (!bh) return -EIO; check_bytes -= size; @@ -150,12 +134,12 @@ static int calc_crc_cont(struct nilfs_sb_info *sbi, struct buffer_head *bhs, /** * nilfs_read_super_root_block - read super root block - * @sb: super_block + * @nilfs: nilfs object * @sr_block: disk block number of the super root block * @pbh: address of a buffer_head pointer to return super root buffer * @check: CRC check flag */ -int nilfs_read_super_root_block(struct super_block *sb, sector_t sr_block, +int nilfs_read_super_root_block(struct the_nilfs *nilfs, sector_t sr_block, struct buffer_head **pbh, int check) { struct buffer_head *bh_sr; @@ -164,7 +148,7 @@ int nilfs_read_super_root_block(struct super_block *sb, sector_t sr_block, int ret; *pbh = NULL; - bh_sr = sb_bread(sb, sr_block); + bh_sr = __bread(nilfs->ns_bdev, sr_block, nilfs->ns_blocksize); if (unlikely(!bh_sr)) { ret = NILFS_SEG_FAIL_IO; goto failed; @@ -174,12 +158,13 @@ int nilfs_read_super_root_block(struct super_block *sb, sector_t sr_block, if (check) { unsigned bytes = le16_to_cpu(sr->sr_bytes); - if (bytes == 0 || bytes > sb->s_blocksize) { + if (bytes == 0 || bytes > nilfs->ns_blocksize) { ret = NILFS_SEG_FAIL_CHECKSUM_SUPER_ROOT; goto failed_bh; } - if (calc_crc_cont(NILFS_SB(sb), bh_sr, &crc, - sizeof(sr->sr_sum), bytes, sr_block, 1)) { + if (nilfs_compute_checksum( + nilfs, bh_sr, &crc, sizeof(sr->sr_sum), bytes, + sr_block, 1)) { ret = NILFS_SEG_FAIL_IO; goto failed_bh; } @@ -199,64 +184,76 @@ int nilfs_read_super_root_block(struct super_block *sb, sector_t sr_block, } /** - * load_segment_summary - read segment summary of the specified partial segment - * @sbi: nilfs_sb_info - * @pseg_start: start disk block number of partial segment - * @seg_seq: sequence number requested - * @ssi: pointer to nilfs_segsum_info struct to store information + * nilfs_read_log_header - read summary header of the specified log + * @nilfs: nilfs object + * @start_blocknr: start block number of the log + * @sum: pointer to return segment summary structure */ -static int -load_segment_summary(struct nilfs_sb_info *sbi, sector_t pseg_start, - u64 seg_seq, struct nilfs_segsum_info *ssi) +static struct buffer_head * +nilfs_read_log_header(struct the_nilfs *nilfs, sector_t start_blocknr, + struct nilfs_segment_summary **sum) { struct buffer_head *bh_sum; - struct nilfs_segment_summary *sum; + + bh_sum = __bread(nilfs->ns_bdev, start_blocknr, nilfs->ns_blocksize); + if (bh_sum) + *sum = (struct nilfs_segment_summary *)bh_sum->b_data; + return bh_sum; +} + +/** + * nilfs_validate_log - verify consistency of log + * @nilfs: nilfs object + * @seg_seq: sequence number of segment + * @bh_sum: buffer head of summary block + * @sum: segment summary struct + */ +static int nilfs_validate_log(struct the_nilfs *nilfs, u64 seg_seq, + struct buffer_head *bh_sum, + struct nilfs_segment_summary *sum) +{ unsigned long nblock; u32 crc; - int ret = NILFS_SEG_FAIL_IO; + int ret; - bh_sum = sb_bread(sbi->s_super, pseg_start); - if (!bh_sum) + ret = NILFS_SEG_FAIL_MAGIC; + if (le32_to_cpu(sum->ss_magic) != NILFS_SEGSUM_MAGIC) goto out; - sum = (struct nilfs_segment_summary *)bh_sum->b_data; - - /* Check consistency of segment summary */ - if (le32_to_cpu(sum->ss_magic) != NILFS_SEGSUM_MAGIC) { - ret = NILFS_SEG_FAIL_MAGIC; - goto failed; - } - store_segsum_info(ssi, sum, sbi->s_super->s_blocksize); - if (seg_seq != ssi->seg_seq) { - ret = NILFS_SEG_FAIL_SEQ; - goto failed; - } + ret = NILFS_SEG_FAIL_SEQ; + if (le64_to_cpu(sum->ss_seq) != seg_seq) + goto out; - nblock = ssi->nblocks; - if (unlikely(nblock == 0 || - nblock > sbi->s_nilfs->ns_blocks_per_segment)) { + nblock = le32_to_cpu(sum->ss_nblocks); + ret = NILFS_SEG_FAIL_CONSISTENCY; + if (unlikely(nblock == 0 || nblock > nilfs->ns_blocks_per_segment)) /* This limits the number of blocks read in the CRC check */ - ret = NILFS_SEG_FAIL_CONSISTENCY; - goto failed; - } - if (calc_crc_cont(sbi, bh_sum, &crc, sizeof(sum->ss_datasum), - ((u64)nblock << sbi->s_super->s_blocksize_bits), - pseg_start, nblock)) { - ret = NILFS_SEG_FAIL_IO; - goto failed; - } - if (crc == le32_to_cpu(sum->ss_datasum)) - ret = 0; - else - ret = NILFS_SEG_FAIL_CHECKSUM_FULL; - failed: - brelse(bh_sum); - out: + goto out; + + ret = NILFS_SEG_FAIL_IO; + if (nilfs_compute_checksum(nilfs, bh_sum, &crc, sizeof(sum->ss_datasum), + ((u64)nblock << nilfs->ns_blocksize_bits), + bh_sum->b_blocknr, nblock)) + goto out; + + ret = NILFS_SEG_FAIL_CHECKSUM_FULL; + if (crc != le32_to_cpu(sum->ss_datasum)) + goto out; + ret = 0; +out: return ret; } -static void *segsum_get(struct super_block *sb, struct buffer_head **pbh, - unsigned int *offset, unsigned int bytes) +/** + * nilfs_read_summary_info - read an item on summary blocks of a log + * @nilfs: nilfs object + * @pbh: the current buffer head on summary blocks [in, out] + * @offset: the current byte offset on summary blocks [in, out] + * @bytes: byte size of the item to be read + */ +static void *nilfs_read_summary_info(struct the_nilfs *nilfs, + struct buffer_head **pbh, + unsigned int *offset, unsigned int bytes) { void *ptr; sector_t blocknr; @@ -265,7 +262,8 @@ static void *segsum_get(struct super_block *sb, struct buffer_head **pbh, if (bytes > (*pbh)->b_size - *offset) { blocknr = (*pbh)->b_blocknr; brelse(*pbh); - *pbh = sb_bread(sb, blocknr + 1); + *pbh = __bread(nilfs->ns_bdev, blocknr + 1, + nilfs->ns_blocksize); if (unlikely(!*pbh)) return NULL; *offset = 0; @@ -275,9 +273,18 @@ static void *segsum_get(struct super_block *sb, struct buffer_head **pbh, return ptr; } -static void segsum_skip(struct super_block *sb, struct buffer_head **pbh, - unsigned int *offset, unsigned int bytes, - unsigned long count) +/** + * nilfs_skip_summary_info - skip items on summary blocks of a log + * @nilfs: nilfs object + * @pbh: the current buffer head on summary blocks [in, out] + * @offset: the current byte offset on summary blocks [in, out] + * @bytes: byte size of the item to be skipped + * @count: number of items to be skipped + */ +static void nilfs_skip_summary_info(struct the_nilfs *nilfs, + struct buffer_head **pbh, + unsigned int *offset, unsigned int bytes, + unsigned long count) { unsigned int rest_item_in_current_block = ((*pbh)->b_size - *offset) / bytes; @@ -294,36 +301,46 @@ static void segsum_skip(struct super_block *sb, struct buffer_head **pbh, *offset = bytes * (count - (bcnt - 1) * nitem_per_block); brelse(*pbh); - *pbh = sb_bread(sb, blocknr + bcnt); + *pbh = __bread(nilfs->ns_bdev, blocknr + bcnt, + nilfs->ns_blocksize); } } -static int -collect_blocks_from_segsum(struct nilfs_sb_info *sbi, sector_t sum_blocknr, - struct nilfs_segsum_info *ssi, - struct list_head *head) +/** + * nilfs_scan_dsync_log - get block information of a log written for data sync + * @nilfs: nilfs object + * @start_blocknr: start block number of the log + * @sum: log summary information + * @head: list head to add nilfs_recovery_block struct + */ +static int nilfs_scan_dsync_log(struct the_nilfs *nilfs, sector_t start_blocknr, + struct nilfs_segment_summary *sum, + struct list_head *head) { struct buffer_head *bh; unsigned int offset; - unsigned long nfinfo = ssi->nfinfo; - sector_t blocknr = sum_blocknr + ssi->nsumblk; + u32 nfinfo, sumbytes; + sector_t blocknr; ino_t ino; int err = -EIO; + nfinfo = le32_to_cpu(sum->ss_nfinfo); if (!nfinfo) return 0; - bh = sb_bread(sbi->s_super, sum_blocknr); + sumbytes = le32_to_cpu(sum->ss_sumbytes); + blocknr = start_blocknr + DIV_ROUND_UP(sumbytes, nilfs->ns_blocksize); + bh = __bread(nilfs->ns_bdev, start_blocknr, nilfs->ns_blocksize); if (unlikely(!bh)) goto out; - offset = le16_to_cpu( - ((struct nilfs_segment_summary *)bh->b_data)->ss_bytes); + offset = le16_to_cpu(sum->ss_bytes); for (;;) { unsigned long nblocks, ndatablk, nnodeblk; struct nilfs_finfo *finfo; - finfo = segsum_get(sbi->s_super, &bh, &offset, sizeof(*finfo)); + finfo = nilfs_read_summary_info(nilfs, &bh, &offset, + sizeof(*finfo)); if (unlikely(!finfo)) goto out; @@ -336,8 +353,8 @@ collect_blocks_from_segsum(struct nilfs_sb_info *sbi, sector_t sum_blocknr, struct nilfs_recovery_block *rb; struct nilfs_binfo_v *binfo; - binfo = segsum_get(sbi->s_super, &bh, &offset, - sizeof(*binfo)); + binfo = nilfs_read_summary_info(nilfs, &bh, &offset, + sizeof(*binfo)); if (unlikely(!binfo)) goto out; @@ -355,9 +372,9 @@ collect_blocks_from_segsum(struct nilfs_sb_info *sbi, sector_t sum_blocknr, } if (--nfinfo == 0) break; - blocknr += nnodeblk; /* always 0 for the data sync segments */ - segsum_skip(sbi->s_super, &bh, &offset, sizeof(__le64), - nnodeblk); + blocknr += nnodeblk; /* always 0 for data sync logs */ + nilfs_skip_summary_info(nilfs, &bh, &offset, sizeof(__le64), + nnodeblk); if (unlikely(!bh)) goto out; } @@ -467,14 +484,14 @@ static int nilfs_prepare_segment_for_recovery(struct the_nilfs *nilfs, return err; } -static int nilfs_recovery_copy_block(struct nilfs_sb_info *sbi, +static int nilfs_recovery_copy_block(struct the_nilfs *nilfs, struct nilfs_recovery_block *rb, struct page *page) { struct buffer_head *bh_org; void *kaddr; - bh_org = sb_bread(sbi->s_super, rb->blocknr); + bh_org = __bread(nilfs->ns_bdev, rb->blocknr, nilfs->ns_blocksize); if (unlikely(!bh_org)) return -EIO; @@ -485,13 +502,14 @@ static int nilfs_recovery_copy_block(struct nilfs_sb_info *sbi, return 0; } -static int recover_dsync_blocks(struct nilfs_sb_info *sbi, - struct list_head *head, - unsigned long *nr_salvaged_blocks) +static int nilfs_recover_dsync_blocks(struct the_nilfs *nilfs, + struct nilfs_sb_info *sbi, + struct list_head *head, + unsigned long *nr_salvaged_blocks) { struct inode *inode; struct nilfs_recovery_block *rb, *n; - unsigned blocksize = sbi->s_super->s_blocksize; + unsigned blocksize = nilfs->ns_blocksize; struct page *page; loff_t pos; int err = 0, err2 = 0; @@ -511,7 +529,7 @@ static int recover_dsync_blocks(struct nilfs_sb_info *sbi, if (unlikely(err)) goto failed_inode; - err = nilfs_recovery_copy_block(sbi, rb, page); + err = nilfs_recovery_copy_block(nilfs, rb, page); if (unlikely(err)) goto failed_page; @@ -551,18 +569,20 @@ static int recover_dsync_blocks(struct nilfs_sb_info *sbi, /** * nilfs_do_roll_forward - salvage logical segments newer than the latest * checkpoint + * @nilfs: nilfs object * @sbi: nilfs_sb_info - * @nilfs: the_nilfs * @ri: pointer to a nilfs_recovery_info */ static int nilfs_do_roll_forward(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi, struct nilfs_recovery_info *ri) { - struct nilfs_segsum_info ssi; + struct buffer_head *bh_sum = NULL; + struct nilfs_segment_summary *sum; sector_t pseg_start; sector_t seg_start, seg_end; /* Starting/ending DBN of full segment */ unsigned long nsalvaged_blocks = 0; + unsigned int flags; u64 seg_seq; __u64 segnum, nextnum = 0; int empty_seg = 0; @@ -581,8 +601,14 @@ static int nilfs_do_roll_forward(struct the_nilfs *nilfs, nilfs_get_segment_range(nilfs, segnum, &seg_start, &seg_end); while (segnum != ri->ri_segnum || pseg_start <= ri->ri_pseg_start) { + brelse(bh_sum); + bh_sum = nilfs_read_log_header(nilfs, pseg_start, &sum); + if (!bh_sum) { + err = -EIO; + goto failed; + } - ret = load_segment_summary(sbi, pseg_start, seg_seq, &ssi); + ret = nilfs_validate_log(nilfs, seg_seq, bh_sum, sum); if (ret) { if (ret == NILFS_SEG_FAIL_IO) { err = -EIO; @@ -590,33 +616,38 @@ static int nilfs_do_roll_forward(struct the_nilfs *nilfs, } goto strayed; } - if (unlikely(NILFS_SEG_HAS_SR(&ssi))) + + flags = le16_to_cpu(sum->ss_flags); + if (flags & NILFS_SS_SR) goto confused; /* Found a valid partial segment; do recovery actions */ - nextnum = nilfs_get_segnum_of_block(nilfs, ssi.next); + nextnum = nilfs_get_segnum_of_block(nilfs, + le64_to_cpu(sum->ss_next)); empty_seg = 0; - nilfs->ns_ctime = ssi.ctime; - if (!(ssi.flags & NILFS_SS_GC)) - nilfs->ns_nongc_ctime = ssi.ctime; + nilfs->ns_ctime = le64_to_cpu(sum->ss_create); + if (!(flags & NILFS_SS_GC)) + nilfs->ns_nongc_ctime = nilfs->ns_ctime; switch (state) { case RF_INIT_ST: - if (!NILFS_SEG_LOGBGN(&ssi) || !NILFS_SEG_DSYNC(&ssi)) + if (!(flags & NILFS_SS_LOGBGN) || + !(flags & NILFS_SS_SYNDT)) goto try_next_pseg; state = RF_DSYNC_ST; /* Fall through */ case RF_DSYNC_ST: - if (!NILFS_SEG_DSYNC(&ssi)) + if (!(flags & NILFS_SS_SYNDT)) goto confused; - err = collect_blocks_from_segsum( - sbi, pseg_start, &ssi, &dsync_blocks); + err = nilfs_scan_dsync_log(nilfs, pseg_start, sum, + &dsync_blocks); if (unlikely(err)) goto failed; - if (NILFS_SEG_LOGEND(&ssi)) { - err = recover_dsync_blocks( - sbi, &dsync_blocks, &nsalvaged_blocks); + if (flags & NILFS_SS_LOGEND) { + err = nilfs_recover_dsync_blocks( + nilfs, sbi, &dsync_blocks, + &nsalvaged_blocks); if (unlikely(err)) goto failed; state = RF_INIT_ST; @@ -627,7 +658,7 @@ static int nilfs_do_roll_forward(struct the_nilfs *nilfs, try_next_pseg: if (pseg_start == ri->ri_lsegs_end) break; - pseg_start += ssi.nblocks; + pseg_start += le32_to_cpu(sum->ss_nblocks); if (pseg_start < seg_end) continue; goto feed_segment; @@ -652,8 +683,9 @@ static int nilfs_do_roll_forward(struct the_nilfs *nilfs, ri->ri_need_recovery = NILFS_RECOVERY_ROLLFORWARD_DONE; } out: + brelse(bh_sum); dispose_recovery_list(&dsync_blocks); - nilfs_detach_writer(sbi->s_nilfs, sbi); + nilfs_detach_writer(nilfs, sbi); return err; confused: @@ -667,7 +699,6 @@ static int nilfs_do_roll_forward(struct the_nilfs *nilfs, } static void nilfs_finish_roll_forward(struct the_nilfs *nilfs, - struct nilfs_sb_info *sbi, struct nilfs_recovery_info *ri) { struct buffer_head *bh; @@ -677,7 +708,7 @@ static void nilfs_finish_roll_forward(struct the_nilfs *nilfs, nilfs_get_segnum_of_block(nilfs, ri->ri_super_root)) return; - bh = sb_getblk(sbi->s_super, ri->ri_lsegs_start); + bh = __getblk(nilfs->ns_bdev, ri->ri_lsegs_start, nilfs->ns_blocksize); BUG_ON(!bh); memset(bh->b_data, 0, bh->b_size); set_buffer_dirty(bh); @@ -690,9 +721,8 @@ static void nilfs_finish_roll_forward(struct the_nilfs *nilfs, } /** - * nilfs_recover_logical_segments - salvage logical segments written after - * the latest super root - * @nilfs: the_nilfs + * nilfs_salvage_orphan_logs - salvage logs written after the latest checkpoint + * @nilfs: nilfs object * @sbi: nilfs_sb_info * @ri: pointer to a nilfs_recovery_info struct to store search results. * @@ -709,9 +739,9 @@ static void nilfs_finish_roll_forward(struct the_nilfs *nilfs, * * %-ENOMEM - Insufficient memory available. */ -int nilfs_recover_logical_segments(struct the_nilfs *nilfs, - struct nilfs_sb_info *sbi, - struct nilfs_recovery_info *ri) +int nilfs_salvage_orphan_logs(struct the_nilfs *nilfs, + struct nilfs_sb_info *sbi, + struct nilfs_recovery_info *ri) { int err; @@ -751,7 +781,7 @@ int nilfs_recover_logical_segments(struct the_nilfs *nilfs, goto failed; } - nilfs_finish_roll_forward(nilfs, sbi, ri); + nilfs_finish_roll_forward(nilfs, ri); } failed: @@ -762,7 +792,6 @@ int nilfs_recover_logical_segments(struct the_nilfs *nilfs, /** * nilfs_search_super_root - search the latest valid super root * @nilfs: the_nilfs - * @sbi: nilfs_sb_info * @ri: pointer to a nilfs_recovery_info struct to store search results. * * nilfs_search_super_root() looks for the latest super-root from a partial @@ -775,14 +804,19 @@ int nilfs_recover_logical_segments(struct the_nilfs *nilfs, * %-EINVAL - No valid segment found * * %-EIO - I/O error + * + * %-ENOMEM - Insufficient memory available. */ -int nilfs_search_super_root(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi, +int nilfs_search_super_root(struct the_nilfs *nilfs, struct nilfs_recovery_info *ri) { - struct nilfs_segsum_info ssi; + struct buffer_head *bh_sum = NULL; + struct nilfs_segment_summary *sum; sector_t pseg_start, pseg_end, sr_pseg_start = 0; sector_t seg_start, seg_end; /* range of full segment (block number) */ sector_t b, end; + unsigned long nblocks; + unsigned int flags; u64 seg_seq; __u64 segnum, nextnum = 0; __u64 cno; @@ -801,17 +835,24 @@ int nilfs_search_super_root(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi, /* Read ahead segment */ b = seg_start; while (b <= seg_end) - sb_breadahead(sbi->s_super, b++); + __breadahead(nilfs->ns_bdev, b++, nilfs->ns_blocksize); for (;;) { - /* Load segment summary */ - ret = load_segment_summary(sbi, pseg_start, seg_seq, &ssi); + brelse(bh_sum); + ret = NILFS_SEG_FAIL_IO; + bh_sum = nilfs_read_log_header(nilfs, pseg_start, &sum); + if (!bh_sum) + goto failed; + + ret = nilfs_validate_log(nilfs, seg_seq, bh_sum, sum); if (ret) { if (ret == NILFS_SEG_FAIL_IO) goto failed; goto strayed; } - pseg_end = pseg_start + ssi.nblocks - 1; + + nblocks = le32_to_cpu(sum->ss_nblocks); + pseg_end = pseg_start + nblocks - 1; if (unlikely(pseg_end > seg_end)) { ret = NILFS_SEG_FAIL_CONSISTENCY; goto strayed; @@ -821,11 +862,13 @@ int nilfs_search_super_root(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi, ri->ri_pseg_start = pseg_start; ri->ri_seq = seg_seq; ri->ri_segnum = segnum; - nextnum = nilfs_get_segnum_of_block(nilfs, ssi.next); + nextnum = nilfs_get_segnum_of_block(nilfs, + le64_to_cpu(sum->ss_next)); ri->ri_nextnum = nextnum; empty_seg = 0; - if (!NILFS_SEG_HAS_SR(&ssi) && !scan_newer) { + flags = le16_to_cpu(sum->ss_flags); + if (!(flags & NILFS_SS_SR) && !scan_newer) { /* This will never happen because a superblock (last_segment) always points to a pseg having a super root. */ @@ -836,14 +879,15 @@ int nilfs_search_super_root(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi, if (pseg_start == seg_start) { nilfs_get_segment_range(nilfs, nextnum, &b, &end); while (b <= end) - sb_breadahead(sbi->s_super, b++); + __breadahead(nilfs->ns_bdev, b++, + nilfs->ns_blocksize); } - if (!NILFS_SEG_HAS_SR(&ssi)) { - if (!ri->ri_lsegs_start && NILFS_SEG_LOGBGN(&ssi)) { + if (!(flags & NILFS_SS_SR)) { + if (!ri->ri_lsegs_start && (flags & NILFS_SS_LOGBGN)) { ri->ri_lsegs_start = pseg_start; ri->ri_lsegs_start_seq = seg_seq; } - if (NILFS_SEG_LOGEND(&ssi)) + if (flags & NILFS_SS_LOGEND) ri->ri_lsegs_end = pseg_start; goto try_next_pseg; } @@ -854,12 +898,12 @@ int nilfs_search_super_root(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi, ri->ri_lsegs_start = ri->ri_lsegs_end = 0; nilfs_dispose_segment_list(&segments); - nilfs->ns_pseg_offset = (sr_pseg_start = pseg_start) - + ssi.nblocks - seg_start; + sr_pseg_start = pseg_start; + nilfs->ns_pseg_offset = pseg_start + nblocks - seg_start; nilfs->ns_seg_seq = seg_seq; nilfs->ns_segnum = segnum; nilfs->ns_cno = cno; /* nilfs->ns_cno = ri->ri_cno + 1 */ - nilfs->ns_ctime = ssi.ctime; + nilfs->ns_ctime = le64_to_cpu(sum->ss_create); nilfs->ns_nextnum = nextnum; if (scan_newer) @@ -870,15 +914,9 @@ int nilfs_search_super_root(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi, scan_newer = 1; } - /* reset region for roll-forward */ - pseg_start += ssi.nblocks; - if (pseg_start < seg_end) - continue; - goto feed_segment; - try_next_pseg: /* Standing on a course, or met an inconsistent state */ - pseg_start += ssi.nblocks; + pseg_start += nblocks; if (pseg_start < seg_end) continue; goto feed_segment; @@ -909,6 +947,7 @@ int nilfs_search_super_root(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi, super_root_found: /* Updating pointers relating to the latest checkpoint */ + brelse(bh_sum); list_splice_tail(&segments, &ri->ri_used_segments); nilfs->ns_last_pseg = sr_pseg_start; nilfs->ns_last_seq = nilfs->ns_seg_seq; @@ -916,6 +955,7 @@ int nilfs_search_super_root(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi, return 0; failed: + brelse(bh_sum); nilfs_dispose_segment_list(&segments); return (ret < 0) ? ret : nilfs_warn_segment_error(ret); } diff --git a/fs/nilfs2/segbuf.h b/fs/nilfs2/segbuf.h index 85fbb66455e..b04f08cc239 100644 --- a/fs/nilfs2/segbuf.h +++ b/fs/nilfs2/segbuf.h @@ -54,17 +54,6 @@ struct nilfs_segsum_info { sector_t next; }; -/* macro for the flags */ -#define NILFS_SEG_HAS_SR(sum) ((sum)->flags & NILFS_SS_SR) -#define NILFS_SEG_LOGBGN(sum) ((sum)->flags & NILFS_SS_LOGBGN) -#define NILFS_SEG_LOGEND(sum) ((sum)->flags & NILFS_SS_LOGEND) -#define NILFS_SEG_DSYNC(sum) ((sum)->flags & NILFS_SS_SYNDT) -#define NILFS_SEG_SIMPLEX(sum) \ - (((sum)->flags & (NILFS_SS_LOGBGN | NILFS_SS_LOGEND)) == \ - (NILFS_SS_LOGBGN | NILFS_SS_LOGEND)) - -#define NILFS_SEG_EMPTY(sum) ((sum)->nblocks == (sum)->nsumblk) - /** * struct nilfs_segment_buffer - Segment buffer * @sb_super: back pointer to a superblock struct @@ -141,6 +130,19 @@ int nilfs_segbuf_extend_payload(struct nilfs_segment_buffer *, struct buffer_head **); void nilfs_segbuf_fill_in_segsum(struct nilfs_segment_buffer *); +static inline int nilfs_segbuf_simplex(struct nilfs_segment_buffer *segbuf) +{ + unsigned int flags = segbuf->sb_sum.flags; + + return (flags & (NILFS_SS_LOGBGN | NILFS_SS_LOGEND)) == + (NILFS_SS_LOGBGN | NILFS_SS_LOGEND); +} + +static inline int nilfs_segbuf_empty(struct nilfs_segment_buffer *segbuf) +{ + return segbuf->sb_sum.nblocks == segbuf->sb_sum.nsumblk; +} + static inline void nilfs_segbuf_add_segsum_buffer(struct nilfs_segment_buffer *segbuf, struct buffer_head *bh) diff --git a/fs/nilfs2/segment.c b/fs/nilfs2/segment.c index c9201649cc4..9fd051a33c4 100644 --- a/fs/nilfs2/segment.c +++ b/fs/nilfs2/segment.c @@ -1914,12 +1914,12 @@ static void nilfs_segctor_complete_write(struct nilfs_sc_info *sci) } } - if (!NILFS_SEG_SIMPLEX(&segbuf->sb_sum)) { - if (NILFS_SEG_LOGBGN(&segbuf->sb_sum)) { + if (!nilfs_segbuf_simplex(segbuf)) { + if (segbuf->sb_sum.flags & NILFS_SS_LOGBGN) { set_bit(NILFS_SC_UNCLOSED, &sci->sc_flags); sci->sc_lseg_stime = jiffies; } - if (NILFS_SEG_LOGEND(&segbuf->sb_sum)) + if (segbuf->sb_sum.flags & NILFS_SS_LOGEND) clear_bit(NILFS_SC_UNCLOSED, &sci->sc_flags); } } @@ -1951,7 +1951,6 @@ static void nilfs_segctor_complete_write(struct nilfs_sc_info *sci) if (update_sr) { nilfs_set_last_segment(nilfs, segbuf->sb_pseg_start, segbuf->sb_sum.seg_seq, nilfs->ns_cno++); - set_nilfs_sb_dirty(nilfs); clear_bit(NILFS_SC_HAVE_DELTA, &sci->sc_flags); clear_bit(NILFS_SC_DIRTY, &sci->sc_flags); @@ -2082,7 +2081,7 @@ static int nilfs_segctor_do_construct(struct nilfs_sc_info *sci, int mode) /* Avoid empty segment */ if (sci->sc_stage.scnt == NILFS_ST_DONE && - NILFS_SEG_EMPTY(&sci->sc_curseg->sb_sum)) { + nilfs_segbuf_empty(sci->sc_curseg)) { nilfs_segctor_abort_construction(sci, nilfs, 1); goto out; } @@ -2408,6 +2407,7 @@ static int nilfs_segctor_construct(struct nilfs_sc_info *sci, int mode) { struct nilfs_sb_info *sbi = sci->sc_sbi; struct the_nilfs *nilfs = sbi->s_nilfs; + struct nilfs_super_block **sbp; int err = 0; nilfs_segctor_accept(sci); @@ -2423,8 +2423,13 @@ static int nilfs_segctor_construct(struct nilfs_sc_info *sci, int mode) if (test_bit(NILFS_SC_SUPER_ROOT, &sci->sc_flags) && nilfs_discontinued(nilfs)) { down_write(&nilfs->ns_sem); - err = nilfs_commit_super( - sbi, nilfs_altsb_need_update(nilfs)); + err = -EIO; + sbp = nilfs_prepare_super(sbi, + nilfs_sb_will_flip(nilfs)); + if (likely(sbp)) { + nilfs_set_log_cursor(sbp[0], nilfs); + err = nilfs_commit_super(sbi, NILFS_SB_COMMIT); + } up_write(&nilfs->ns_sem); } } diff --git a/fs/nilfs2/segment.h b/fs/nilfs2/segment.h index 01e20dbb217..17c487bd815 100644 --- a/fs/nilfs2/segment.h +++ b/fs/nilfs2/segment.h @@ -234,13 +234,13 @@ extern int nilfs_attach_segment_constructor(struct nilfs_sb_info *); extern void nilfs_detach_segment_constructor(struct nilfs_sb_info *); /* recovery.c */ -extern int nilfs_read_super_root_block(struct super_block *, sector_t, +extern int nilfs_read_super_root_block(struct the_nilfs *, sector_t, struct buffer_head **, int); -extern int nilfs_search_super_root(struct the_nilfs *, struct nilfs_sb_info *, +extern int nilfs_search_super_root(struct the_nilfs *, struct nilfs_recovery_info *); -extern int nilfs_recover_logical_segments(struct the_nilfs *, - struct nilfs_sb_info *, - struct nilfs_recovery_info *); +extern int nilfs_salvage_orphan_logs(struct the_nilfs *, + struct nilfs_sb_info *, + struct nilfs_recovery_info *); extern void nilfs_dispose_segment_list(struct list_head *); #endif /* _NILFS_SEGMENT_H */ diff --git a/fs/nilfs2/super.c b/fs/nilfs2/super.c index 414ef68931c..26078b3407c 100644 --- a/fs/nilfs2/super.c +++ b/fs/nilfs2/super.c @@ -55,6 +55,8 @@ #include "nilfs.h" #include "mdt.h" #include "alloc.h" +#include "btree.h" +#include "btnode.h" #include "page.h" #include "cpfile.h" #include "ifile.h" @@ -74,6 +76,25 @@ struct kmem_cache *nilfs_btree_path_cache; static int nilfs_remount(struct super_block *sb, int *flags, char *data); +static void nilfs_set_error(struct nilfs_sb_info *sbi) +{ + struct the_nilfs *nilfs = sbi->s_nilfs; + struct nilfs_super_block **sbp; + + down_write(&nilfs->ns_sem); + if (!(nilfs->ns_mount_state & NILFS_ERROR_FS)) { + nilfs->ns_mount_state |= NILFS_ERROR_FS; + sbp = nilfs_prepare_super(sbi, 0); + if (likely(sbp)) { + sbp[0]->s_state |= cpu_to_le16(NILFS_ERROR_FS); + if (sbp[1]) + sbp[1]->s_state |= cpu_to_le16(NILFS_ERROR_FS); + nilfs_commit_super(sbi, NILFS_SB_COMMIT_ALL); + } + } + up_write(&nilfs->ns_sem); +} + /** * nilfs_error() - report failure condition on a filesystem * @@ -99,16 +120,7 @@ void nilfs_error(struct super_block *sb, const char *function, va_end(args); if (!(sb->s_flags & MS_RDONLY)) { - struct the_nilfs *nilfs = sbi->s_nilfs; - - down_write(&nilfs->ns_sem); - if (!(nilfs->ns_mount_state & NILFS_ERROR_FS)) { - nilfs->ns_mount_state |= NILFS_ERROR_FS; - nilfs->ns_sbp[0]->s_state |= - cpu_to_le16(NILFS_ERROR_FS); - nilfs_commit_super(sbi, 1); - } - up_write(&nilfs->ns_sem); + nilfs_set_error(sbi); if (nilfs_test_opt(sbi, ERRORS_RO)) { printk(KERN_CRIT "Remounting filesystem read-only\n"); @@ -176,7 +188,7 @@ static void nilfs_clear_inode(struct inode *inode) nilfs_btnode_cache_clear(&ii->i_btnode_cache); } -static int nilfs_sync_super(struct nilfs_sb_info *sbi, int dupsb) +static int nilfs_sync_super(struct nilfs_sb_info *sbi, int flag) { struct the_nilfs *nilfs = sbi->s_nilfs; int err; @@ -202,12 +214,20 @@ static int nilfs_sync_super(struct nilfs_sb_info *sbi, int dupsb) printk(KERN_ERR "NILFS: unable to write superblock (err=%d)\n", err); if (err == -EIO && nilfs->ns_sbh[1]) { + /* + * sbp[0] points to newer log than sbp[1], + * so copy sbp[0] to sbp[1] to take over sbp[0]. + */ + memcpy(nilfs->ns_sbp[1], nilfs->ns_sbp[0], + nilfs->ns_sbsize); nilfs_fall_back_super_block(nilfs); goto retry; } } else { struct nilfs_super_block *sbp = nilfs->ns_sbp[0]; + nilfs->ns_sbwcount++; + /* * The latest segment becomes trailable from the position * written in superblock. @@ -216,66 +236,122 @@ static int nilfs_sync_super(struct nilfs_sb_info *sbi, int dupsb) /* update GC protection for recent segments */ if (nilfs->ns_sbh[1]) { - sbp = NULL; - if (dupsb) { + if (flag == NILFS_SB_COMMIT_ALL) { set_buffer_dirty(nilfs->ns_sbh[1]); - if (!sync_dirty_buffer(nilfs->ns_sbh[1])) - sbp = nilfs->ns_sbp[1]; + if (sync_dirty_buffer(nilfs->ns_sbh[1]) < 0) + goto out; } + if (le64_to_cpu(nilfs->ns_sbp[1]->s_last_cno) < + le64_to_cpu(nilfs->ns_sbp[0]->s_last_cno)) + sbp = nilfs->ns_sbp[1]; } - if (sbp) { - spin_lock(&nilfs->ns_last_segment_lock); - nilfs->ns_prot_seq = le64_to_cpu(sbp->s_last_seq); - spin_unlock(&nilfs->ns_last_segment_lock); - } - } + spin_lock(&nilfs->ns_last_segment_lock); + nilfs->ns_prot_seq = le64_to_cpu(sbp->s_last_seq); + spin_unlock(&nilfs->ns_last_segment_lock); + } + out: return err; } -int nilfs_commit_super(struct nilfs_sb_info *sbi, int dupsb) +void nilfs_set_log_cursor(struct nilfs_super_block *sbp, + struct the_nilfs *nilfs) +{ + sector_t nfreeblocks; + + /* nilfs->ns_sem must be locked by the caller. */ + nilfs_count_free_blocks(nilfs, &nfreeblocks); + sbp->s_free_blocks_count = cpu_to_le64(nfreeblocks); + + spin_lock(&nilfs->ns_last_segment_lock); + sbp->s_last_seq = cpu_to_le64(nilfs->ns_last_seq); + sbp->s_last_pseg = cpu_to_le64(nilfs->ns_last_pseg); + sbp->s_last_cno = cpu_to_le64(nilfs->ns_last_cno); + spin_unlock(&nilfs->ns_last_segment_lock); +} + +struct nilfs_super_block **nilfs_prepare_super(struct nilfs_sb_info *sbi, + int flip) { struct the_nilfs *nilfs = sbi->s_nilfs; struct nilfs_super_block **sbp = nilfs->ns_sbp; - sector_t nfreeblocks; - time_t t; - int err; - /* nilfs->sem must be locked by the caller. */ + /* nilfs->ns_sem must be locked by the caller. */ if (sbp[0]->s_magic != cpu_to_le16(NILFS_SUPER_MAGIC)) { - if (sbp[1] && sbp[1]->s_magic == cpu_to_le16(NILFS_SUPER_MAGIC)) - nilfs_swap_super_block(nilfs); - else { + if (sbp[1] && + sbp[1]->s_magic == cpu_to_le16(NILFS_SUPER_MAGIC)) { + memcpy(sbp[0], sbp[1], nilfs->ns_sbsize); + } else { printk(KERN_CRIT "NILFS: superblock broke on dev %s\n", sbi->s_super->s_id); - return -EIO; + return NULL; } + } else if (sbp[1] && + sbp[1]->s_magic != cpu_to_le16(NILFS_SUPER_MAGIC)) { + memcpy(sbp[1], sbp[0], nilfs->ns_sbsize); } - err = nilfs_count_free_blocks(nilfs, &nfreeblocks); - if (unlikely(err)) { - printk(KERN_ERR "NILFS: failed to count free blocks\n"); - return err; - } - spin_lock(&nilfs->ns_last_segment_lock); - sbp[0]->s_last_seq = cpu_to_le64(nilfs->ns_last_seq); - sbp[0]->s_last_pseg = cpu_to_le64(nilfs->ns_last_pseg); - sbp[0]->s_last_cno = cpu_to_le64(nilfs->ns_last_cno); - spin_unlock(&nilfs->ns_last_segment_lock); + if (flip && sbp[1]) + nilfs_swap_super_block(nilfs); + + return sbp; +} + +int nilfs_commit_super(struct nilfs_sb_info *sbi, int flag) +{ + struct the_nilfs *nilfs = sbi->s_nilfs; + struct nilfs_super_block **sbp = nilfs->ns_sbp; + time_t t; + + /* nilfs->ns_sem must be locked by the caller. */ t = get_seconds(); - nilfs->ns_sbwtime[0] = t; - sbp[0]->s_free_blocks_count = cpu_to_le64(nfreeblocks); + nilfs->ns_sbwtime = t; sbp[0]->s_wtime = cpu_to_le64(t); sbp[0]->s_sum = 0; sbp[0]->s_sum = cpu_to_le32(crc32_le(nilfs->ns_crc_seed, (unsigned char *)sbp[0], nilfs->ns_sbsize)); - if (dupsb && sbp[1]) { - memcpy(sbp[1], sbp[0], nilfs->ns_sbsize); - nilfs->ns_sbwtime[1] = t; + if (flag == NILFS_SB_COMMIT_ALL && sbp[1]) { + sbp[1]->s_wtime = sbp[0]->s_wtime; + sbp[1]->s_sum = 0; + sbp[1]->s_sum = cpu_to_le32(crc32_le(nilfs->ns_crc_seed, + (unsigned char *)sbp[1], + nilfs->ns_sbsize)); } clear_nilfs_sb_dirty(nilfs); - return nilfs_sync_super(sbi, dupsb); + return nilfs_sync_super(sbi, flag); +} + +/** + * nilfs_cleanup_super() - write filesystem state for cleanup + * @sbi: nilfs_sb_info to be unmounted or degraded to read-only + * + * This function restores state flags in the on-disk super block. + * This will set "clean" flag (i.e. NILFS_VALID_FS) unless the + * filesystem was not clean previously. + */ +int nilfs_cleanup_super(struct nilfs_sb_info *sbi) +{ + struct nilfs_super_block **sbp; + int flag = NILFS_SB_COMMIT; + int ret = -EIO; + + sbp = nilfs_prepare_super(sbi, 0); + if (sbp) { + sbp[0]->s_state = cpu_to_le16(sbi->s_nilfs->ns_mount_state); + nilfs_set_log_cursor(sbp[0], sbi->s_nilfs); + if (sbp[1] && sbp[0]->s_last_cno == sbp[1]->s_last_cno) { + /* + * make the "clean" flag also to the opposite + * super block if both super blocks point to + * the same checkpoint. + */ + sbp[1]->s_state = sbp[0]->s_state; + flag = NILFS_SB_COMMIT_ALL; + } + ret = nilfs_commit_super(sbi, flag); + } + return ret; } static void nilfs_put_super(struct super_block *sb) @@ -289,8 +365,7 @@ static void nilfs_put_super(struct super_block *sb) if (!(sb->s_flags & MS_RDONLY)) { down_write(&nilfs->ns_sem); - nilfs->ns_sbp[0]->s_state = cpu_to_le16(nilfs->ns_mount_state); - nilfs_commit_super(sbi, 1); + nilfs_cleanup_super(sbi); up_write(&nilfs->ns_sem); } down_write(&nilfs->ns_super_sem); @@ -311,6 +386,7 @@ static int nilfs_sync_fs(struct super_block *sb, int wait) { struct nilfs_sb_info *sbi = NILFS_SB(sb); struct the_nilfs *nilfs = sbi->s_nilfs; + struct nilfs_super_block **sbp; int err = 0; /* This function is called when super block should be written back */ @@ -318,8 +394,13 @@ static int nilfs_sync_fs(struct super_block *sb, int wait) err = nilfs_construct_segment(sb); down_write(&nilfs->ns_sem); - if (nilfs_sb_dirty(nilfs)) - nilfs_commit_super(sbi, 1); + if (nilfs_sb_dirty(nilfs)) { + sbp = nilfs_prepare_super(sbi, nilfs_sb_will_flip(nilfs)); + if (likely(sbp)) { + nilfs_set_log_cursor(sbp[0], nilfs); + nilfs_commit_super(sbi, NILFS_SB_COMMIT); + } + } up_write(&nilfs->ns_sem); return err; @@ -442,20 +523,20 @@ static int nilfs_show_options(struct seq_file *seq, struct vfsmount *vfs) struct nilfs_sb_info *sbi = NILFS_SB(sb); if (!nilfs_test_opt(sbi, BARRIER)) - seq_printf(seq, ",nobarrier"); + seq_puts(seq, ",nobarrier"); if (nilfs_test_opt(sbi, SNAPSHOT)) seq_printf(seq, ",cp=%llu", (unsigned long long int)sbi->s_snapshot_cno); if (nilfs_test_opt(sbi, ERRORS_PANIC)) - seq_printf(seq, ",errors=panic"); + seq_puts(seq, ",errors=panic"); if (nilfs_test_opt(sbi, ERRORS_CONT)) - seq_printf(seq, ",errors=continue"); + seq_puts(seq, ",errors=continue"); if (nilfs_test_opt(sbi, STRICT_ORDER)) - seq_printf(seq, ",order=strict"); + seq_puts(seq, ",order=strict"); if (nilfs_test_opt(sbi, NORECOVERY)) - seq_printf(seq, ",norecovery"); + seq_puts(seq, ",norecovery"); if (nilfs_test_opt(sbi, DISCARD)) - seq_printf(seq, ",discard"); + seq_puts(seq, ",discard"); return 0; } @@ -524,23 +605,25 @@ static const struct export_operations nilfs_export_ops = { enum { Opt_err_cont, Opt_err_panic, Opt_err_ro, - Opt_nobarrier, Opt_snapshot, Opt_order, Opt_norecovery, - Opt_discard, Opt_err, + Opt_barrier, Opt_nobarrier, Opt_snapshot, Opt_order, Opt_norecovery, + Opt_discard, Opt_nodiscard, Opt_err, }; static match_table_t tokens = { {Opt_err_cont, "errors=continue"}, {Opt_err_panic, "errors=panic"}, {Opt_err_ro, "errors=remount-ro"}, + {Opt_barrier, "barrier"}, {Opt_nobarrier, "nobarrier"}, {Opt_snapshot, "cp=%u"}, {Opt_order, "order=%s"}, {Opt_norecovery, "norecovery"}, {Opt_discard, "discard"}, + {Opt_nodiscard, "nodiscard"}, {Opt_err, NULL} }; -static int parse_options(char *options, struct super_block *sb) +static int parse_options(char *options, struct super_block *sb, int is_remount) { struct nilfs_sb_info *sbi = NILFS_SB(sb); char *p; @@ -557,6 +640,9 @@ static int parse_options(char *options, struct super_block *sb) token = match_token(p, tokens, args); switch (token) { + case Opt_barrier: + nilfs_set_opt(sbi, BARRIER); + break; case Opt_nobarrier: nilfs_clear_opt(sbi, BARRIER); break; @@ -582,8 +668,26 @@ static int parse_options(char *options, struct super_block *sb) case Opt_snapshot: if (match_int(&args[0], &option) || option <= 0) return 0; - if (!(sb->s_flags & MS_RDONLY)) + if (is_remount) { + if (!nilfs_test_opt(sbi, SNAPSHOT)) { + printk(KERN_ERR + "NILFS: cannot change regular " + "mount to snapshot.\n"); + return 0; + } else if (option != sbi->s_snapshot_cno) { + printk(KERN_ERR + "NILFS: cannot remount to a " + "different snapshot.\n"); + return 0; + } + break; + } + if (!(sb->s_flags & MS_RDONLY)) { + printk(KERN_ERR "NILFS: cannot mount snapshot " + "read/write. A read-only option is " + "required.\n"); return 0; + } sbi->s_snapshot_cno = option; nilfs_set_opt(sbi, SNAPSHOT); break; @@ -593,6 +697,9 @@ static int parse_options(char *options, struct super_block *sb) case Opt_discard: nilfs_set_opt(sbi, DISCARD); break; + case Opt_nodiscard: + nilfs_clear_opt(sbi, DISCARD); + break; default: printk(KERN_ERR "NILFS: Unrecognized mount option \"%s\"\n", p); @@ -613,11 +720,18 @@ nilfs_set_default_options(struct nilfs_sb_info *sbi, static int nilfs_setup_super(struct nilfs_sb_info *sbi) { struct the_nilfs *nilfs = sbi->s_nilfs; - struct nilfs_super_block *sbp = nilfs->ns_sbp[0]; - int max_mnt_count = le16_to_cpu(sbp->s_max_mnt_count); - int mnt_count = le16_to_cpu(sbp->s_mnt_count); + struct nilfs_super_block **sbp; + int max_mnt_count; + int mnt_count; + + /* nilfs->ns_sem must be locked by the caller. */ + sbp = nilfs_prepare_super(sbi, 0); + if (!sbp) + return -EIO; + + max_mnt_count = le16_to_cpu(sbp[0]->s_max_mnt_count); + mnt_count = le16_to_cpu(sbp[0]->s_mnt_count); - /* nilfs->sem must be locked by the caller. */ if (nilfs->ns_mount_state & NILFS_ERROR_FS) { printk(KERN_WARNING "NILFS warning: mounting fs with errors\n"); @@ -628,12 +742,15 @@ static int nilfs_setup_super(struct nilfs_sb_info *sbi) #endif } if (!max_mnt_count) - sbp->s_max_mnt_count = cpu_to_le16(NILFS_DFL_MAX_MNT_COUNT); - - sbp->s_mnt_count = cpu_to_le16(mnt_count + 1); - sbp->s_state = cpu_to_le16(le16_to_cpu(sbp->s_state) & ~NILFS_VALID_FS); - sbp->s_mtime = cpu_to_le64(get_seconds()); - return nilfs_commit_super(sbi, 1); + sbp[0]->s_max_mnt_count = cpu_to_le16(NILFS_DFL_MAX_MNT_COUNT); + + sbp[0]->s_mnt_count = cpu_to_le16(mnt_count + 1); + sbp[0]->s_state = + cpu_to_le16(le16_to_cpu(sbp[0]->s_state) & ~NILFS_VALID_FS); + sbp[0]->s_mtime = cpu_to_le64(get_seconds()); + /* synchronize sbp[1] with sbp[0] */ + memcpy(sbp[1], sbp[0], nilfs->ns_sbsize); + return nilfs_commit_super(sbi, NILFS_SB_COMMIT_ALL); } struct nilfs_super_block *nilfs_read_super_block(struct super_block *sb, @@ -670,7 +787,31 @@ int nilfs_store_magic_and_option(struct super_block *sb, sbi->s_interval = le32_to_cpu(sbp->s_c_interval); sbi->s_watermark = le32_to_cpu(sbp->s_c_block_max); - return !parse_options(data, sb) ? -EINVAL : 0 ; + return !parse_options(data, sb, 0) ? -EINVAL : 0 ; +} + +int nilfs_check_feature_compatibility(struct super_block *sb, + struct nilfs_super_block *sbp) +{ + __u64 features; + + features = le64_to_cpu(sbp->s_feature_incompat) & + ~NILFS_FEATURE_INCOMPAT_SUPP; + if (features) { + printk(KERN_ERR "NILFS: couldn't mount because of unsupported " + "optional features (%llx)\n", + (unsigned long long)features); + return -EINVAL; + } + features = le64_to_cpu(sbp->s_feature_compat_ro) & + ~NILFS_FEATURE_COMPAT_RO_SUPP; + if (!(sb->s_flags & MS_RDONLY) && features) { + printk(KERN_ERR "NILFS: couldn't mount RDWR because of " + "unsupported optional features (%llx)\n", + (unsigned long long)features); + return -EINVAL; + } + return 0; } /** @@ -819,7 +960,6 @@ nilfs_fill_super(struct super_block *sb, void *data, int silent, static int nilfs_remount(struct super_block *sb, int *flags, char *data) { struct nilfs_sb_info *sbi = NILFS_SB(sb); - struct nilfs_super_block *sbp; struct the_nilfs *nilfs = sbi->s_nilfs; unsigned long old_sb_flags; struct nilfs_mount_options old_opts; @@ -833,32 +973,17 @@ static int nilfs_remount(struct super_block *sb, int *flags, char *data) old_opts.snapshot_cno = sbi->s_snapshot_cno; was_snapshot = nilfs_test_opt(sbi, SNAPSHOT); - if (!parse_options(data, sb)) { + if (!parse_options(data, sb, 1)) { err = -EINVAL; goto restore_opts; } sb->s_flags = (sb->s_flags & ~MS_POSIXACL); err = -EINVAL; - if (was_snapshot) { - if (!(*flags & MS_RDONLY)) { - printk(KERN_ERR "NILFS (device %s): cannot remount " - "snapshot read/write.\n", - sb->s_id); - goto restore_opts; - } else if (sbi->s_snapshot_cno != old_opts.snapshot_cno) { - printk(KERN_ERR "NILFS (device %s): cannot " - "remount to a different snapshot.\n", - sb->s_id); - goto restore_opts; - } - } else { - if (nilfs_test_opt(sbi, SNAPSHOT)) { - printk(KERN_ERR "NILFS (device %s): cannot change " - "a regular mount to a snapshot.\n", - sb->s_id); - goto restore_opts; - } + if (was_snapshot && !(*flags & MS_RDONLY)) { + printk(KERN_ERR "NILFS (device %s): cannot remount snapshot " + "read/write.\n", sb->s_id); + goto restore_opts; } if (!nilfs_valid_fs(nilfs)) { @@ -880,19 +1005,29 @@ static int nilfs_remount(struct super_block *sb, int *flags, char *data) * the RDONLY flag and then mark the partition as valid again. */ down_write(&nilfs->ns_sem); - sbp = nilfs->ns_sbp[0]; - if (!(sbp->s_state & le16_to_cpu(NILFS_VALID_FS)) && - (nilfs->ns_mount_state & NILFS_VALID_FS)) - sbp->s_state = cpu_to_le16(nilfs->ns_mount_state); - sbp->s_mtime = cpu_to_le64(get_seconds()); - nilfs_commit_super(sbi, 1); + nilfs_cleanup_super(sbi); up_write(&nilfs->ns_sem); } else { + __u64 features; + /* * Mounting a RDONLY partition read-write, so reread and * store the current valid flag. (It may have been changed * by fsck since we originally mounted the partition.) */ + down_read(&nilfs->ns_sem); + features = le64_to_cpu(nilfs->ns_sbp[0]->s_feature_compat_ro) & + ~NILFS_FEATURE_COMPAT_RO_SUPP; + up_read(&nilfs->ns_sem); + if (features) { + printk(KERN_WARNING "NILFS (device %s): couldn't " + "remount RDWR because of unsupported optional " + "features (%llx)\n", + sb->s_id, (unsigned long long)features); + err = -EROFS; + goto restore_opts; + } + sb->s_flags &= ~MS_RDONLY; err = nilfs_attach_segment_constructor(sbi); @@ -1119,7 +1254,7 @@ static void nilfs_inode_init_once(void *obj) init_rwsem(&ii->xattr_sem); #endif nilfs_btnode_cache_init_once(&ii->i_btnode_cache); - ii->i_bmap = (struct nilfs_bmap *)&ii->i_bmap_union; + ii->i_bmap = &ii->i_bmap_data; inode_init_once(&ii->vfs_inode); } diff --git a/fs/nilfs2/the_nilfs.c b/fs/nilfs2/the_nilfs.c index 8c1097327ab..37de1f062d8 100644 --- a/fs/nilfs2/the_nilfs.c +++ b/fs/nilfs2/the_nilfs.c @@ -38,6 +38,8 @@ static LIST_HEAD(nilfs_objects); static DEFINE_SPINLOCK(nilfs_lock); +static int nilfs_valid_sb(struct nilfs_super_block *sbp); + void nilfs_set_last_segment(struct the_nilfs *nilfs, sector_t start_blocknr, u64 seq, __u64 cno) { @@ -45,6 +47,16 @@ void nilfs_set_last_segment(struct the_nilfs *nilfs, nilfs->ns_last_pseg = start_blocknr; nilfs->ns_last_seq = seq; nilfs->ns_last_cno = cno; + + if (!nilfs_sb_dirty(nilfs)) { + if (nilfs->ns_prev_seq == nilfs->ns_last_seq) + goto stay_cursor; + + set_nilfs_sb_dirty(nilfs); + } + nilfs->ns_prev_seq = nilfs->ns_last_seq; + + stay_cursor: spin_unlock(&nilfs->ns_last_segment_lock); } @@ -159,8 +171,7 @@ void put_nilfs(struct the_nilfs *nilfs) kfree(nilfs); } -static int nilfs_load_super_root(struct the_nilfs *nilfs, - struct nilfs_sb_info *sbi, sector_t sr_block) +static int nilfs_load_super_root(struct the_nilfs *nilfs, sector_t sr_block) { struct buffer_head *bh_sr; struct nilfs_super_root *raw_sr; @@ -169,7 +180,7 @@ static int nilfs_load_super_root(struct the_nilfs *nilfs, unsigned inode_size; int err; - err = nilfs_read_super_root_block(sbi->s_super, sr_block, &bh_sr, 1); + err = nilfs_read_super_root_block(nilfs, sr_block, &bh_sr, 1); if (unlikely(err)) return err; @@ -248,6 +259,37 @@ static void nilfs_clear_recovery_info(struct nilfs_recovery_info *ri) } /** + * nilfs_store_log_cursor - load log cursor from a super block + * @nilfs: nilfs object + * @sbp: buffer storing super block to be read + * + * nilfs_store_log_cursor() reads the last position of the log + * containing a super root from a given super block, and initializes + * relevant information on the nilfs object preparatory for log + * scanning and recovery. + */ +static int nilfs_store_log_cursor(struct the_nilfs *nilfs, + struct nilfs_super_block *sbp) +{ + int ret = 0; + + nilfs->ns_last_pseg = le64_to_cpu(sbp->s_last_pseg); + nilfs->ns_last_cno = le64_to_cpu(sbp->s_last_cno); + nilfs->ns_last_seq = le64_to_cpu(sbp->s_last_seq); + + nilfs->ns_prev_seq = nilfs->ns_last_seq; + nilfs->ns_seg_seq = nilfs->ns_last_seq; + nilfs->ns_segnum = + nilfs_get_segnum_of_block(nilfs, nilfs->ns_last_pseg); + nilfs->ns_cno = nilfs->ns_last_cno + 1; + if (nilfs->ns_segnum >= nilfs->ns_nsegments) { + printk(KERN_ERR "NILFS invalid last segment number.\n"); + ret = -EINVAL; + } + return ret; +} + +/** * load_nilfs - load and recover the nilfs * @nilfs: the_nilfs structure to be released * @sbi: nilfs_sb_info used to recover past segment @@ -285,13 +327,55 @@ int load_nilfs(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi) nilfs_init_recovery_info(&ri); - err = nilfs_search_super_root(nilfs, sbi, &ri); + err = nilfs_search_super_root(nilfs, &ri); if (unlikely(err)) { - printk(KERN_ERR "NILFS: error searching super root.\n"); - goto failed; + struct nilfs_super_block **sbp = nilfs->ns_sbp; + int blocksize; + + if (err != -EINVAL) + goto scan_error; + + if (!nilfs_valid_sb(sbp[1])) { + printk(KERN_WARNING + "NILFS warning: unable to fall back to spare" + "super block\n"); + goto scan_error; + } + printk(KERN_INFO + "NILFS: try rollback from an earlier position\n"); + + /* + * restore super block with its spare and reconfigure + * relevant states of the nilfs object. + */ + memcpy(sbp[0], sbp[1], nilfs->ns_sbsize); + nilfs->ns_crc_seed = le32_to_cpu(sbp[0]->s_crc_seed); + nilfs->ns_sbwtime = le64_to_cpu(sbp[0]->s_wtime); + + /* verify consistency between two super blocks */ + blocksize = BLOCK_SIZE << le32_to_cpu(sbp[0]->s_log_block_size); + if (blocksize != nilfs->ns_blocksize) { + printk(KERN_WARNING + "NILFS warning: blocksize differs between " + "two super blocks (%d != %d)\n", + blocksize, nilfs->ns_blocksize); + goto scan_error; + } + + err = nilfs_store_log_cursor(nilfs, sbp[0]); + if (err) + goto scan_error; + + /* drop clean flag to allow roll-forward and recovery */ + nilfs->ns_mount_state &= ~NILFS_VALID_FS; + valid_fs = 0; + + err = nilfs_search_super_root(nilfs, &ri); + if (err) + goto scan_error; } - err = nilfs_load_super_root(nilfs, sbi, ri.ri_super_root); + err = nilfs_load_super_root(nilfs, ri.ri_super_root); if (unlikely(err)) { printk(KERN_ERR "NILFS: error loading super root.\n"); goto failed; @@ -301,11 +385,23 @@ int load_nilfs(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi) goto skip_recovery; if (s_flags & MS_RDONLY) { + __u64 features; + if (nilfs_test_opt(sbi, NORECOVERY)) { printk(KERN_INFO "NILFS: norecovery option specified. " "skipping roll-forward recovery\n"); goto skip_recovery; } + features = le64_to_cpu(nilfs->ns_sbp[0]->s_feature_compat_ro) & + ~NILFS_FEATURE_COMPAT_RO_SUPP; + if (features) { + printk(KERN_ERR "NILFS: couldn't proceed with " + "recovery because of unsupported optional " + "features (%llx)\n", + (unsigned long long)features); + err = -EROFS; + goto failed_unload; + } if (really_read_only) { printk(KERN_ERR "NILFS: write access " "unavailable, cannot proceed.\n"); @@ -320,14 +416,13 @@ int load_nilfs(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi) goto failed_unload; } - err = nilfs_recover_logical_segments(nilfs, sbi, &ri); + err = nilfs_salvage_orphan_logs(nilfs, sbi, &ri); if (err) goto failed_unload; down_write(&nilfs->ns_sem); - nilfs->ns_mount_state |= NILFS_VALID_FS; - nilfs->ns_sbp[0]->s_state = cpu_to_le16(nilfs->ns_mount_state); - err = nilfs_commit_super(sbi, 1); + nilfs->ns_mount_state |= NILFS_VALID_FS; /* set "clean" flag */ + err = nilfs_cleanup_super(sbi); up_write(&nilfs->ns_sem); if (err) { @@ -343,6 +438,10 @@ int load_nilfs(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi) sbi->s_super->s_flags = s_flags; return 0; + scan_error: + printk(KERN_ERR "NILFS: error searching super root.\n"); + goto failed; + failed_unload: nilfs_mdt_destroy(nilfs->ns_cpfile); nilfs_mdt_destroy(nilfs->ns_sufile); @@ -515,8 +614,8 @@ static int nilfs_load_super_block(struct the_nilfs *nilfs, nilfs_swap_super_block(nilfs); } - nilfs->ns_sbwtime[0] = le64_to_cpu(sbp[0]->s_wtime); - nilfs->ns_sbwtime[1] = valid[!swp] ? le64_to_cpu(sbp[1]->s_wtime) : 0; + nilfs->ns_sbwcount = 0; + nilfs->ns_sbwtime = le64_to_cpu(sbp[0]->s_wtime); nilfs->ns_prot_seq = le64_to_cpu(sbp[valid[1] & !swp]->s_last_seq); *sbpp = sbp[0]; return 0; @@ -557,6 +656,10 @@ int init_nilfs(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi, char *data) if (err) goto out; + err = nilfs_check_feature_compatibility(sb, sbp); + if (err) + goto out; + blocksize = BLOCK_SIZE << le32_to_cpu(sbp->s_log_block_size); if (sb->s_blocksize != blocksize && !sb_set_blocksize(sb, blocksize)) { @@ -568,7 +671,7 @@ int init_nilfs(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi, char *data) goto out; } - blocksize = sb_min_blocksize(sb, BLOCK_SIZE); + blocksize = sb_min_blocksize(sb, NILFS_MIN_BLOCK_SIZE); if (!blocksize) { printk(KERN_ERR "NILFS: unable to set blocksize\n"); err = -EINVAL; @@ -582,7 +685,18 @@ int init_nilfs(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi, char *data) if (err) goto failed_sbh; + err = nilfs_check_feature_compatibility(sb, sbp); + if (err) + goto failed_sbh; + blocksize = BLOCK_SIZE << le32_to_cpu(sbp->s_log_block_size); + if (blocksize < NILFS_MIN_BLOCK_SIZE || + blocksize > NILFS_MAX_BLOCK_SIZE) { + printk(KERN_ERR "NILFS: couldn't mount because of unsupported " + "filesystem blocksize %d\n", blocksize); + err = -EINVAL; + goto failed_sbh; + } if (sb->s_blocksize != blocksize) { int hw_blocksize = bdev_logical_block_size(sb->s_bdev); @@ -604,6 +718,7 @@ int init_nilfs(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi, char *data) when reloading fails. */ } nilfs->ns_blocksize_bits = sb->s_blocksize_bits; + nilfs->ns_blocksize = blocksize; err = nilfs_store_disk_layout(nilfs, sbp); if (err) @@ -616,23 +731,9 @@ int init_nilfs(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi, char *data) bdi = nilfs->ns_bdev->bd_inode->i_mapping->backing_dev_info; nilfs->ns_bdi = bdi ? : &default_backing_dev_info; - /* Finding last segment */ - nilfs->ns_last_pseg = le64_to_cpu(sbp->s_last_pseg); - nilfs->ns_last_cno = le64_to_cpu(sbp->s_last_cno); - nilfs->ns_last_seq = le64_to_cpu(sbp->s_last_seq); - - nilfs->ns_seg_seq = nilfs->ns_last_seq; - nilfs->ns_segnum = - nilfs_get_segnum_of_block(nilfs, nilfs->ns_last_pseg); - nilfs->ns_cno = nilfs->ns_last_cno + 1; - if (nilfs->ns_segnum >= nilfs->ns_nsegments) { - printk(KERN_ERR "NILFS invalid last segment number.\n"); - err = -EINVAL; + err = nilfs_store_log_cursor(nilfs, sbp); + if (err) goto failed_sbh; - } - /* Dummy values */ - nilfs->ns_free_segments_count = - nilfs->ns_nsegments - (nilfs->ns_segnum + 1); /* Initialize gcinode cache */ err = nilfs_init_gccache(nilfs); diff --git a/fs/nilfs2/the_nilfs.h b/fs/nilfs2/the_nilfs.h index 1ab97453369..f785a7b0ab9 100644 --- a/fs/nilfs2/the_nilfs.h +++ b/fs/nilfs2/the_nilfs.h @@ -57,7 +57,8 @@ enum { * @ns_current: back pointer to current mount * @ns_sbh: buffer heads of on-disk super blocks * @ns_sbp: pointers to super block data - * @ns_sbwtime: previous write time of super blocks + * @ns_sbwtime: previous write time of super block + * @ns_sbwcount: write count of super block * @ns_sbsize: size of valid data in super block * @ns_supers: list of nilfs super block structs * @ns_seg_seq: segment sequence counter @@ -73,7 +74,7 @@ enum { * @ns_last_seq: sequence value of the latest segment * @ns_last_cno: checkpoint number of the latest segment * @ns_prot_seq: least sequence number of segments which must not be reclaimed - * @ns_free_segments_count: counter of free segments + * @ns_prev_seq: base sequence number used to decide if advance log cursor * @ns_segctor_sem: segment constructor semaphore * @ns_dat: DAT file inode * @ns_cpfile: checkpoint file inode @@ -82,6 +83,7 @@ enum { * @ns_gc_inodes: dummy inodes to keep live blocks * @ns_gc_inodes_h: hash list to keep dummy inode holding live blocks * @ns_blocksize_bits: bit length of block size + * @ns_blocksize: block size * @ns_nsegments: number of segments in filesystem * @ns_blocks_per_segment: number of blocks per segment * @ns_r_segments_percentage: reserved segments percentage @@ -119,7 +121,8 @@ struct the_nilfs { */ struct buffer_head *ns_sbh[2]; struct nilfs_super_block *ns_sbp[2]; - time_t ns_sbwtime[2]; + time_t ns_sbwtime; + unsigned ns_sbwcount; unsigned ns_sbsize; unsigned ns_mount_state; @@ -149,7 +152,7 @@ struct the_nilfs { u64 ns_last_seq; __u64 ns_last_cno; u64 ns_prot_seq; - unsigned long ns_free_segments_count; + u64 ns_prev_seq; struct rw_semaphore ns_segctor_sem; @@ -168,6 +171,7 @@ struct the_nilfs { /* Disk layout information (static) */ unsigned int ns_blocksize_bits; + unsigned int ns_blocksize; unsigned long ns_nsegments; unsigned long ns_blocks_per_segment; unsigned long ns_r_segments_percentage; @@ -203,20 +207,17 @@ THE_NILFS_FNS(SB_DIRTY, sb_dirty) /* Minimum interval of periodical update of superblocks (in seconds) */ #define NILFS_SB_FREQ 10 -#define NILFS_ALTSB_FREQ 60 /* spare superblock */ static inline int nilfs_sb_need_update(struct the_nilfs *nilfs) { u64 t = get_seconds(); - return t < nilfs->ns_sbwtime[0] || - t > nilfs->ns_sbwtime[0] + NILFS_SB_FREQ; + return t < nilfs->ns_sbwtime || t > nilfs->ns_sbwtime + NILFS_SB_FREQ; } -static inline int nilfs_altsb_need_update(struct the_nilfs *nilfs) +static inline int nilfs_sb_will_flip(struct the_nilfs *nilfs) { - u64 t = get_seconds(); - struct nilfs_super_block **sbp = nilfs->ns_sbp; - return sbp[1] && t > nilfs->ns_sbwtime[1] + NILFS_ALTSB_FREQ; + int flip_bits = nilfs->ns_sbwcount & 0x0FL; + return (flip_bits != 0x08 && flip_bits != 0x0F); } void nilfs_set_last_segment(struct the_nilfs *, sector_t, u64, __u64); diff --git a/fs/ocfs2/aops.c b/fs/ocfs2/aops.c index 356e976772b..96337a4fbbd 100644 --- a/fs/ocfs2/aops.c +++ b/fs/ocfs2/aops.c @@ -578,7 +578,9 @@ bail: static void ocfs2_dio_end_io(struct kiocb *iocb, loff_t offset, ssize_t bytes, - void *private) + void *private, + int ret, + bool is_async) { struct inode *inode = iocb->ki_filp->f_path.dentry->d_inode; int level; @@ -592,6 +594,9 @@ static void ocfs2_dio_end_io(struct kiocb *iocb, if (!level) up_read(&inode->i_alloc_sem); ocfs2_rw_unlock(inode, level); + + if (is_async) + aio_complete(iocb, ret, 0); } /* diff --git a/fs/ocfs2/dlm/dlmdebug.c b/fs/ocfs2/dlm/dlmdebug.c index 0cd24cf5439..5efdd37dfe4 100644 --- a/fs/ocfs2/dlm/dlmdebug.c +++ b/fs/ocfs2/dlm/dlmdebug.c @@ -419,7 +419,7 @@ static loff_t debug_buffer_llseek(struct file *file, loff_t off, int whence) static int debug_buffer_release(struct inode *inode, struct file *file) { - struct debug_buffer *db = (struct debug_buffer *)file->private_data; + struct debug_buffer *db = file->private_data; if (db) kfree(db->buf); @@ -715,7 +715,7 @@ static int debug_lockres_open(struct inode *inode, struct file *file) goto bail; } - seq = (struct seq_file *) file->private_data; + seq = file->private_data; seq->private = dl; dlm_grab(dlm); @@ -731,7 +731,7 @@ bail: static int debug_lockres_release(struct inode *inode, struct file *file) { - struct seq_file *seq = (struct seq_file *)file->private_data; + struct seq_file *seq = file->private_data; struct debug_lockres *dl = (struct debug_lockres *)seq->private; if (dl->dl_res) diff --git a/fs/ocfs2/dlmfs/dlmfs.c b/fs/ocfs2/dlmfs/dlmfs.c index b83d6107a1f..bef34d0528d 100644 --- a/fs/ocfs2/dlmfs/dlmfs.c +++ b/fs/ocfs2/dlmfs/dlmfs.c @@ -182,8 +182,7 @@ static int dlmfs_file_release(struct inode *inode, { int level, status; struct dlmfs_inode_private *ip = DLMFS_I(inode); - struct dlmfs_filp_private *fp = - (struct dlmfs_filp_private *) file->private_data; + struct dlmfs_filp_private *fp = file->private_data; if (S_ISDIR(inode->i_mode)) BUG(); diff --git a/fs/ocfs2/dlmglue.c b/fs/ocfs2/dlmglue.c index 39eb16ac5f9..5e02a893f46 100644 --- a/fs/ocfs2/dlmglue.c +++ b/fs/ocfs2/dlmglue.c @@ -2966,7 +2966,7 @@ static const struct seq_operations ocfs2_dlm_seq_ops = { static int ocfs2_dlm_debug_release(struct inode *inode, struct file *file) { - struct seq_file *seq = (struct seq_file *) file->private_data; + struct seq_file *seq = file->private_data; struct ocfs2_dlm_seq_priv *priv = seq->private; struct ocfs2_lock_res *res = &priv->p_iter_res; @@ -3000,7 +3000,7 @@ static int ocfs2_dlm_debug_open(struct inode *inode, struct file *file) goto out; } - seq = (struct seq_file *) file->private_data; + seq = file->private_data; seq->private = priv; ocfs2_add_lockres_tracking(&priv->p_iter_res, diff --git a/fs/ocfs2/journal.c b/fs/ocfs2/journal.c index 625de9d7088..9b57c0350ff 100644 --- a/fs/ocfs2/journal.c +++ b/fs/ocfs2/journal.c @@ -760,13 +760,13 @@ void ocfs2_set_journal_params(struct ocfs2_super *osb) if (osb->osb_commit_interval) commit_interval = osb->osb_commit_interval; - spin_lock(&journal->j_state_lock); + write_lock(&journal->j_state_lock); journal->j_commit_interval = commit_interval; if (osb->s_mount_opt & OCFS2_MOUNT_BARRIER) journal->j_flags |= JBD2_BARRIER; else journal->j_flags &= ~JBD2_BARRIER; - spin_unlock(&journal->j_state_lock); + write_unlock(&journal->j_state_lock); } int ocfs2_journal_init(struct ocfs2_journal *journal, int *dirty) diff --git a/fs/ocfs2/super.c b/fs/ocfs2/super.c index 0eaa929a4db..03a799fdd74 100644 --- a/fs/ocfs2/super.c +++ b/fs/ocfs2/super.c @@ -2472,7 +2472,7 @@ static void ocfs2_delete_osb(struct ocfs2_super *osb) kfree(osb->slot_recovery_generations); /* FIXME * This belongs in journal shutdown, but because we have to - * allocate osb->journal at the start of ocfs2_initalize_osb(), + * allocate osb->journal at the start of ocfs2_initialize_osb(), * we free it here. */ kfree(osb->journal); diff --git a/fs/open.c b/fs/open.c index 5463266db9e..0d1fa3dc0ef 100644 --- a/fs/open.c +++ b/fs/open.c @@ -110,7 +110,7 @@ static long do_sys_truncate(const char __user *pathname, loff_t length) error = locks_verify_truncate(inode, NULL, length); if (!error) - error = security_path_truncate(&path, length, 0); + error = security_path_truncate(&path); if (!error) error = do_truncate(path.dentry, length, 0, NULL); @@ -165,8 +165,7 @@ static long do_sys_ftruncate(unsigned int fd, loff_t length, int small) error = locks_verify_truncate(inode, file, length); if (!error) - error = security_path_truncate(&file->f_path, length, - ATTR_MTIME|ATTR_CTIME); + error = security_path_truncate(&file->f_path); if (!error) error = do_truncate(dentry, length, ATTR_MTIME|ATTR_CTIME, file); out_putf: @@ -367,7 +366,7 @@ SYSCALL_DEFINE1(chdir, const char __user *, filename) if (error) goto out; - error = inode_permission(path.dentry->d_inode, MAY_EXEC | MAY_ACCESS); + error = inode_permission(path.dentry->d_inode, MAY_EXEC | MAY_CHDIR); if (error) goto dput_and_out; @@ -396,7 +395,7 @@ SYSCALL_DEFINE1(fchdir, unsigned int, fd) if (!S_ISDIR(inode->i_mode)) goto out_putf; - error = inode_permission(inode, MAY_EXEC | MAY_ACCESS); + error = inode_permission(inode, MAY_EXEC | MAY_CHDIR); if (!error) set_fs_pwd(current->fs, &file->f_path); out_putf: @@ -414,7 +413,7 @@ SYSCALL_DEFINE1(chroot, const char __user *, filename) if (error) goto out; - error = inode_permission(path.dentry->d_inode, MAY_EXEC | MAY_ACCESS); + error = inode_permission(path.dentry->d_inode, MAY_EXEC | MAY_CHDIR); if (error) goto dput_and_out; diff --git a/fs/partitions/check.c b/fs/partitions/check.c index 5dcd4b0c553..72c52656dc2 100644 --- a/fs/partitions/check.c +++ b/fs/partitions/check.c @@ -459,7 +459,6 @@ struct hd_struct *add_partition(struct gendisk *disk, int partno, } /* everything is up and running, commence */ - INIT_RCU_HEAD(&p->rcu_head); rcu_assign_pointer(ptbl->part[partno], p); /* suppress uevent if the disk supresses it */ diff --git a/fs/quota/dquot.c b/fs/quota/dquot.c index 437d2ca2de9..ef72b169942 100644 --- a/fs/quota/dquot.c +++ b/fs/quota/dquot.c @@ -132,6 +132,22 @@ static __cacheline_aligned_in_smp DEFINE_SPINLOCK(dq_state_lock); __cacheline_aligned_in_smp DEFINE_SPINLOCK(dq_data_lock); EXPORT_SYMBOL(dq_data_lock); +void __quota_error(struct super_block *sb, const char *func, + const char *fmt, ...) +{ + va_list args; + + if (printk_ratelimit()) { + va_start(args, fmt); + printk(KERN_ERR "Quota error (device %s): %s: ", + sb->s_id, func); + vprintk(fmt, args); + printk("\n"); + va_end(args); + } +} +EXPORT_SYMBOL(__quota_error); + #if defined(CONFIG_QUOTA_DEBUG) || defined(CONFIG_PRINT_QUOTA_WARNING) static char *quotatypes[] = INITQFNAMES; #endif @@ -705,11 +721,8 @@ void dqput(struct dquot *dquot) return; #ifdef CONFIG_QUOTA_DEBUG if (!atomic_read(&dquot->dq_count)) { - printk("VFS: dqput: trying to free free dquot\n"); - printk("VFS: device %s, dquot of %s %d\n", - dquot->dq_sb->s_id, - quotatypes[dquot->dq_type], - dquot->dq_id); + quota_error(dquot->dq_sb, "trying to free free dquot of %s %d", + quotatypes[dquot->dq_type], dquot->dq_id); BUG(); } #endif @@ -732,9 +745,9 @@ we_slept: /* Commit dquot before releasing */ ret = dquot->dq_sb->dq_op->write_dquot(dquot); if (ret < 0) { - printk(KERN_ERR "VFS: cannot write quota structure on " - "device %s (error %d). Quota may get out of " - "sync!\n", dquot->dq_sb->s_id, ret); + quota_error(dquot->dq_sb, "Can't write quota structure" + " (error %d). Quota may get out of sync!", + ret); /* * We clear dirty bit anyway, so that we avoid * infinite loop here @@ -914,9 +927,9 @@ static void add_dquot_ref(struct super_block *sb, int type) #ifdef CONFIG_QUOTA_DEBUG if (reserved) { - printk(KERN_WARNING "VFS (%s): Writes happened before quota" - " was turned on thus quota information is probably " - "inconsistent. Please run quotacheck(8).\n", sb->s_id); + quota_error(sb, "Writes happened before quota was turned on " + "thus quota information is probably inconsistent. " + "Please run quotacheck(8)"); } #endif } @@ -947,7 +960,9 @@ static int remove_inode_dquot_ref(struct inode *inode, int type, if (dqput_blocks(dquot)) { #ifdef CONFIG_QUOTA_DEBUG if (atomic_read(&dquot->dq_count) != 1) - printk(KERN_WARNING "VFS: Adding dquot with dq_count %d to dispose list.\n", atomic_read(&dquot->dq_count)); + quota_error(inode->i_sb, "Adding dquot with " + "dq_count %d to dispose list", + atomic_read(&dquot->dq_count)); #endif spin_lock(&dq_list_lock); /* As dquot must have currently users it can't be on @@ -986,6 +1001,7 @@ static void remove_dquot_ref(struct super_block *sb, int type, struct list_head *tofree_head) { struct inode *inode; + int reserved = 0; spin_lock(&inode_lock); list_for_each_entry(inode, &sb->s_inodes, i_sb_list) { @@ -995,10 +1011,20 @@ static void remove_dquot_ref(struct super_block *sb, int type, * only quota pointers and these have separate locking * (dqptr_sem). */ - if (!IS_NOQUOTA(inode)) + if (!IS_NOQUOTA(inode)) { + if (unlikely(inode_get_rsv_space(inode) > 0)) + reserved = 1; remove_inode_dquot_ref(inode, type, tofree_head); + } } spin_unlock(&inode_lock); +#ifdef CONFIG_QUOTA_DEBUG + if (reserved) { + printk(KERN_WARNING "VFS (%s): Writes happened after quota" + " was disabled thus quota information is probably " + "inconsistent. Please run quotacheck(8).\n", sb->s_id); + } +#endif } /* Gather all references from inodes and drop them */ @@ -1304,6 +1330,15 @@ static int info_bdq_free(struct dquot *dquot, qsize_t space) return QUOTA_NL_NOWARN; } +static int dquot_active(const struct inode *inode) +{ + struct super_block *sb = inode->i_sb; + + if (IS_NOQUOTA(inode)) + return 0; + return sb_any_quota_loaded(sb) & ~sb_any_quota_suspended(sb); +} + /* * Initialize quota pointers in inode * @@ -1323,7 +1358,7 @@ static void __dquot_initialize(struct inode *inode, int type) /* First test before acquiring mutex - solves deadlocks when we * re-enter the quota code and are already holding the mutex */ - if (!sb_any_quota_active(inode->i_sb) || IS_NOQUOTA(inode)) + if (!dquot_active(inode)) return; /* First get references to structures we might need. */ @@ -1507,7 +1542,7 @@ int __dquot_alloc_space(struct inode *inode, qsize_t number, int flags) * First test before acquiring mutex - solves deadlocks when we * re-enter the quota code and are already holding the mutex */ - if (!sb_any_quota_active(inode->i_sb) || IS_NOQUOTA(inode)) { + if (!dquot_active(inode)) { inode_incr_space(inode, number, reserve); goto out; } @@ -1559,7 +1594,7 @@ int dquot_alloc_inode(const struct inode *inode) /* First test before acquiring mutex - solves deadlocks when we * re-enter the quota code and are already holding the mutex */ - if (!sb_any_quota_active(inode->i_sb) || IS_NOQUOTA(inode)) + if (!dquot_active(inode)) return 0; for (cnt = 0; cnt < MAXQUOTAS; cnt++) warntype[cnt] = QUOTA_NL_NOWARN; @@ -1596,7 +1631,7 @@ int dquot_claim_space_nodirty(struct inode *inode, qsize_t number) { int cnt; - if (!sb_any_quota_active(inode->i_sb) || IS_NOQUOTA(inode)) { + if (!dquot_active(inode)) { inode_claim_rsv_space(inode, number); return 0; } @@ -1629,7 +1664,7 @@ void __dquot_free_space(struct inode *inode, qsize_t number, int flags) /* First test before acquiring mutex - solves deadlocks when we * re-enter the quota code and are already holding the mutex */ - if (!sb_any_quota_active(inode->i_sb) || IS_NOQUOTA(inode)) { + if (!dquot_active(inode)) { inode_decr_space(inode, number, reserve); return; } @@ -1667,7 +1702,7 @@ void dquot_free_inode(const struct inode *inode) /* First test before acquiring mutex - solves deadlocks when we * re-enter the quota code and are already holding the mutex */ - if (!sb_any_quota_active(inode->i_sb) || IS_NOQUOTA(inode)) + if (!dquot_active(inode)) return; down_read(&sb_dqopt(inode->i_sb)->dqptr_sem); @@ -1790,7 +1825,7 @@ int dquot_transfer(struct inode *inode, struct iattr *iattr) struct super_block *sb = inode->i_sb; int ret; - if (!sb_any_quota_active(sb) || IS_NOQUOTA(inode)) + if (!dquot_active(inode)) return 0; if (iattr->ia_valid & ATTR_UID && iattr->ia_uid != inode->i_uid) @@ -1957,7 +1992,7 @@ int dquot_disable(struct super_block *sb, int type, unsigned int flags) truncate_inode_pages(&toputinode[cnt]->i_data, 0); mutex_unlock(&toputinode[cnt]->i_mutex); - mark_inode_dirty(toputinode[cnt]); + mark_inode_dirty_sync(toputinode[cnt]); } mutex_unlock(&dqopt->dqonoff_mutex); } @@ -2270,7 +2305,7 @@ static void do_get_dqblk(struct dquot *dquot, struct fs_disk_quota *di) memset(di, 0, sizeof(*di)); di->d_version = FS_DQUOT_VERSION; di->d_flags = dquot->dq_type == USRQUOTA ? - XFS_USER_QUOTA : XFS_GROUP_QUOTA; + FS_USER_QUOTA : FS_GROUP_QUOTA; di->d_id = dquot->dq_id; spin_lock(&dq_data_lock); diff --git a/fs/quota/quota_tree.c b/fs/quota/quota_tree.c index 24f03407eeb..9e48874eabc 100644 --- a/fs/quota/quota_tree.c +++ b/fs/quota/quota_tree.c @@ -65,8 +65,7 @@ static ssize_t write_blk(struct qtree_mem_dqinfo *info, uint blk, char *buf) ret = sb->s_op->quota_write(sb, info->dqi_type, buf, info->dqi_usable_bs, blk << info->dqi_blocksize_bits); if (ret != info->dqi_usable_bs) { - q_warn(KERN_WARNING "VFS: dquota write failed on " - "dev %s\n", sb->s_id); + quota_error(sb, "dquota write failed"); if (ret >= 0) ret = -EIO; } @@ -160,9 +159,8 @@ static int remove_free_dqentry(struct qtree_mem_dqinfo *info, char *buf, dh->dqdh_next_free = dh->dqdh_prev_free = cpu_to_le32(0); /* No matter whether write succeeds block is out of list */ if (write_blk(info, blk, buf) < 0) - q_warn(KERN_ERR - "VFS: Can't write block (%u) with free entries.\n", - blk); + quota_error(info->dqi_sb, "Can't write block (%u) " + "with free entries", blk); return 0; out_buf: kfree(tmpbuf); @@ -252,9 +250,8 @@ static uint find_free_dqentry(struct qtree_mem_dqinfo *info, if (le16_to_cpu(dh->dqdh_entries) + 1 >= qtree_dqstr_in_blk(info)) { *err = remove_free_dqentry(info, buf, blk); if (*err < 0) { - q_warn(KERN_ERR "VFS: find_free_dqentry(): Can't " - "remove block (%u) from entry free list.\n", - blk); + quota_error(dquot->dq_sb, "Can't remove block (%u) " + "from entry free list", blk); goto out_buf; } } @@ -268,16 +265,15 @@ static uint find_free_dqentry(struct qtree_mem_dqinfo *info, } #ifdef __QUOTA_QT_PARANOIA if (i == qtree_dqstr_in_blk(info)) { - printk(KERN_ERR "VFS: find_free_dqentry(): Data block full " - "but it shouldn't.\n"); + quota_error(dquot->dq_sb, "Data block full but it shouldn't"); *err = -EIO; goto out_buf; } #endif *err = write_blk(info, blk, buf); if (*err < 0) { - q_warn(KERN_ERR "VFS: find_free_dqentry(): Can't write quota " - "data block %u.\n", blk); + quota_error(dquot->dq_sb, "Can't write quota data block %u", + blk); goto out_buf; } dquot->dq_off = (blk << info->dqi_blocksize_bits) + @@ -311,8 +307,8 @@ static int do_insert_tree(struct qtree_mem_dqinfo *info, struct dquot *dquot, } else { ret = read_blk(info, *treeblk, buf); if (ret < 0) { - q_warn(KERN_ERR "VFS: Can't read tree quota block " - "%u.\n", *treeblk); + quota_error(dquot->dq_sb, "Can't read tree quota " + "block %u", *treeblk); goto out_buf; } } @@ -323,9 +319,9 @@ static int do_insert_tree(struct qtree_mem_dqinfo *info, struct dquot *dquot, if (depth == info->dqi_qtree_depth - 1) { #ifdef __QUOTA_QT_PARANOIA if (newblk) { - printk(KERN_ERR "VFS: Inserting already present quota " - "entry (block %u).\n", - le32_to_cpu(ref[get_index(info, + quota_error(dquot->dq_sb, "Inserting already present " + "quota entry (block %u)", + le32_to_cpu(ref[get_index(info, dquot->dq_id, depth)])); ret = -EIO; goto out_buf; @@ -373,8 +369,8 @@ int qtree_write_dquot(struct qtree_mem_dqinfo *info, struct dquot *dquot) if (!dquot->dq_off) { ret = dq_insert_tree(info, dquot); if (ret < 0) { - q_warn(KERN_ERR "VFS: Error %zd occurred while " - "creating quota.\n", ret); + quota_error(sb, "Error %zd occurred while creating " + "quota", ret); kfree(ddquot); return ret; } @@ -385,8 +381,7 @@ int qtree_write_dquot(struct qtree_mem_dqinfo *info, struct dquot *dquot) ret = sb->s_op->quota_write(sb, type, ddquot, info->dqi_entry_size, dquot->dq_off); if (ret != info->dqi_entry_size) { - q_warn(KERN_WARNING "VFS: dquota write failed on dev %s\n", - sb->s_id); + quota_error(sb, "dquota write failed"); if (ret >= 0) ret = -ENOSPC; } else { @@ -410,14 +405,15 @@ static int free_dqentry(struct qtree_mem_dqinfo *info, struct dquot *dquot, if (!buf) return -ENOMEM; if (dquot->dq_off >> info->dqi_blocksize_bits != blk) { - q_warn(KERN_ERR "VFS: Quota structure has offset to other " - "block (%u) than it should (%u).\n", blk, - (uint)(dquot->dq_off >> info->dqi_blocksize_bits)); + quota_error(dquot->dq_sb, "Quota structure has offset to " + "other block (%u) than it should (%u)", blk, + (uint)(dquot->dq_off >> info->dqi_blocksize_bits)); goto out_buf; } ret = read_blk(info, blk, buf); if (ret < 0) { - q_warn(KERN_ERR "VFS: Can't read quota data block %u\n", blk); + quota_error(dquot->dq_sb, "Can't read quota data block %u", + blk); goto out_buf; } dh = (struct qt_disk_dqdbheader *)buf; @@ -427,8 +423,8 @@ static int free_dqentry(struct qtree_mem_dqinfo *info, struct dquot *dquot, if (ret >= 0) ret = put_free_dqblk(info, buf, blk); if (ret < 0) { - q_warn(KERN_ERR "VFS: Can't move quota data block (%u) " - "to free list.\n", blk); + quota_error(dquot->dq_sb, "Can't move quota data block " + "(%u) to free list", blk); goto out_buf; } } else { @@ -440,15 +436,15 @@ static int free_dqentry(struct qtree_mem_dqinfo *info, struct dquot *dquot, /* Insert will write block itself */ ret = insert_free_dqentry(info, buf, blk); if (ret < 0) { - q_warn(KERN_ERR "VFS: Can't insert quota data " - "block (%u) to free entry list.\n", blk); + quota_error(dquot->dq_sb, "Can't insert quota " + "data block (%u) to free entry list", blk); goto out_buf; } } else { ret = write_blk(info, blk, buf); if (ret < 0) { - q_warn(KERN_ERR "VFS: Can't write quota data " - "block %u\n", blk); + quota_error(dquot->dq_sb, "Can't write quota " + "data block %u", blk); goto out_buf; } } @@ -472,7 +468,8 @@ static int remove_tree(struct qtree_mem_dqinfo *info, struct dquot *dquot, return -ENOMEM; ret = read_blk(info, *blk, buf); if (ret < 0) { - q_warn(KERN_ERR "VFS: Can't read quota data block %u\n", *blk); + quota_error(dquot->dq_sb, "Can't read quota data " + "block %u", blk); goto out_buf; } newblk = le32_to_cpu(ref[get_index(info, dquot->dq_id, depth)]); @@ -496,8 +493,8 @@ static int remove_tree(struct qtree_mem_dqinfo *info, struct dquot *dquot, } else { ret = write_blk(info, *blk, buf); if (ret < 0) - q_warn(KERN_ERR "VFS: Can't write quota tree " - "block %u.\n", *blk); + quota_error(dquot->dq_sb, "Can't write quota " + "tree block %u", blk); } } out_buf: @@ -529,7 +526,8 @@ static loff_t find_block_dqentry(struct qtree_mem_dqinfo *info, return -ENOMEM; ret = read_blk(info, blk, buf); if (ret < 0) { - q_warn(KERN_ERR "VFS: Can't read quota tree block %u.\n", blk); + quota_error(dquot->dq_sb, "Can't read quota tree " + "block %u", blk); goto out_buf; } ddquot = buf + sizeof(struct qt_disk_dqdbheader); @@ -539,8 +537,8 @@ static loff_t find_block_dqentry(struct qtree_mem_dqinfo *info, ddquot += info->dqi_entry_size; } if (i == qtree_dqstr_in_blk(info)) { - q_warn(KERN_ERR "VFS: Quota for id %u referenced " - "but not present.\n", dquot->dq_id); + quota_error(dquot->dq_sb, "Quota for id %u referenced " + "but not present", dquot->dq_id); ret = -EIO; goto out_buf; } else { @@ -564,7 +562,8 @@ static loff_t find_tree_dqentry(struct qtree_mem_dqinfo *info, return -ENOMEM; ret = read_blk(info, blk, buf); if (ret < 0) { - q_warn(KERN_ERR "VFS: Can't read quota tree block %u.\n", blk); + quota_error(dquot->dq_sb, "Can't read quota tree block %u", + blk); goto out_buf; } ret = 0; @@ -598,7 +597,7 @@ int qtree_read_dquot(struct qtree_mem_dqinfo *info, struct dquot *dquot) #ifdef __QUOTA_QT_PARANOIA /* Invalidated quota? */ if (!sb_dqopt(dquot->dq_sb)->files[type]) { - printk(KERN_ERR "VFS: Quota invalidated while reading!\n"); + quota_error(sb, "Quota invalidated while reading!"); return -EIO; } #endif @@ -607,8 +606,8 @@ int qtree_read_dquot(struct qtree_mem_dqinfo *info, struct dquot *dquot) offset = find_dqentry(info, dquot); if (offset <= 0) { /* Entry not present? */ if (offset < 0) - q_warn(KERN_ERR "VFS: Can't read quota " - "structure for id %u.\n", dquot->dq_id); + quota_error(sb, "Can't read quota structure " + "for id %u", dquot->dq_id); dquot->dq_off = 0; set_bit(DQ_FAKE_B, &dquot->dq_flags); memset(&dquot->dq_dqb, 0, sizeof(struct mem_dqblk)); @@ -625,8 +624,8 @@ int qtree_read_dquot(struct qtree_mem_dqinfo *info, struct dquot *dquot) if (ret != info->dqi_entry_size) { if (ret >= 0) ret = -EIO; - q_warn(KERN_ERR "VFS: Error while reading quota " - "structure for id %u.\n", dquot->dq_id); + quota_error(sb, "Error while reading quota structure for id %u", + dquot->dq_id); set_bit(DQ_FAKE_B, &dquot->dq_flags); memset(&dquot->dq_dqb, 0, sizeof(struct mem_dqblk)); kfree(ddquot); diff --git a/fs/quota/quota_tree.h b/fs/quota/quota_tree.h index ccc3e71fb1d..a1ab8db81a5 100644 --- a/fs/quota/quota_tree.h +++ b/fs/quota/quota_tree.h @@ -22,10 +22,4 @@ struct qt_disk_dqdbheader { #define QT_TREEOFF 1 /* Offset of tree in file in blocks */ -#define q_warn(fmt, args...) \ -do { \ - if (printk_ratelimit()) \ - printk(fmt, ## args); \ -} while(0) - #endif /* _LINUX_QUOTAIO_TREE_H */ diff --git a/fs/quota/quota_v1.c b/fs/quota/quota_v1.c index 4af344c5852..34b37a67bb1 100644 --- a/fs/quota/quota_v1.c +++ b/fs/quota/quota_v1.c @@ -95,8 +95,7 @@ static int v1_commit_dqblk(struct dquot *dquot) (char *)&dqblk, sizeof(struct v1_disk_dqblk), v1_dqoff(dquot->dq_id)); if (ret != sizeof(struct v1_disk_dqblk)) { - printk(KERN_WARNING "VFS: dquota write failed on dev %s\n", - dquot->dq_sb->s_id); + quota_error(dquot->dq_sb, "dquota write failed"); if (ret >= 0) ret = -EIO; goto out; diff --git a/fs/quota/quota_v2.c b/fs/quota/quota_v2.c index 135206af145..65444d29406 100644 --- a/fs/quota/quota_v2.c +++ b/fs/quota/quota_v2.c @@ -63,9 +63,8 @@ static int v2_read_header(struct super_block *sb, int type, size = sb->s_op->quota_read(sb, type, (char *)dqhead, sizeof(struct v2_disk_dqheader), 0); if (size != sizeof(struct v2_disk_dqheader)) { - q_warn(KERN_WARNING "quota_v2: Failed header read:" - " expected=%zd got=%zd\n", - sizeof(struct v2_disk_dqheader), size); + quota_error(sb, "Failed header read: expected=%zd got=%zd", + sizeof(struct v2_disk_dqheader), size); return 0; } return 1; @@ -106,8 +105,7 @@ static int v2_read_file_info(struct super_block *sb, int type) size = sb->s_op->quota_read(sb, type, (char *)&dinfo, sizeof(struct v2_disk_dqinfo), V2_DQINFOOFF); if (size != sizeof(struct v2_disk_dqinfo)) { - q_warn(KERN_WARNING "quota_v2: Can't read info structure on device %s.\n", - sb->s_id); + quota_error(sb, "Can't read info structure"); return -1; } info->dqi_priv = kmalloc(sizeof(struct qtree_mem_dqinfo), GFP_NOFS); @@ -167,8 +165,7 @@ static int v2_write_file_info(struct super_block *sb, int type) size = sb->s_op->quota_write(sb, type, (char *)&dinfo, sizeof(struct v2_disk_dqinfo), V2_DQINFOOFF); if (size != sizeof(struct v2_disk_dqinfo)) { - q_warn(KERN_WARNING "Can't write info structure on device %s.\n", - sb->s_id); + quota_error(sb, "Can't write info structure"); return -1; } return 0; diff --git a/fs/reiserfs/inode.c b/fs/reiserfs/inode.c index 0f22fdaf54a..29db72203bd 100644 --- a/fs/reiserfs/inode.c +++ b/fs/reiserfs/inode.c @@ -1221,7 +1221,7 @@ static void init_inode(struct inode *inode, struct treepath *path) inode_set_bytes(inode, to_real_used_space(inode, inode->i_blocks, SD_V2_SIZE)); - /* read persistent inode attributes from sd and initalise + /* read persistent inode attributes from sd and initialise generic inode flags from them */ REISERFS_I(inode)->i_attrs = sd_v2_attrs(sd); sd_attrs_to_i_attrs(sd_v2_attrs(sd), inode); diff --git a/fs/sysfs/file.c b/fs/sysfs/file.c index 1beaa739d0a..1b27b5688f6 100644 --- a/fs/sysfs/file.c +++ b/fs/sysfs/file.c @@ -593,7 +593,8 @@ EXPORT_SYMBOL_GPL(sysfs_add_file_to_group); * @mode: file permissions. * */ -int sysfs_chmod_file(struct kobject *kobj, struct attribute *attr, mode_t mode) +int sysfs_chmod_file(struct kobject *kobj, const struct attribute *attr, + mode_t mode) { struct sysfs_dirent *sd; struct iattr newattrs; diff --git a/fs/ubifs/lpt.c b/fs/ubifs/lpt.c index ad7f67b827e..0084a33c4c6 100644 --- a/fs/ubifs/lpt.c +++ b/fs/ubifs/lpt.c @@ -1457,13 +1457,13 @@ struct ubifs_lprops *ubifs_lpt_lookup(struct ubifs_info *c, int lnum) shft -= UBIFS_LPT_FANOUT_SHIFT; nnode = ubifs_get_nnode(c, nnode, iip); if (IS_ERR(nnode)) - return ERR_PTR(PTR_ERR(nnode)); + return ERR_CAST(nnode); } iip = ((i >> shft) & (UBIFS_LPT_FANOUT - 1)); shft -= UBIFS_LPT_FANOUT_SHIFT; pnode = ubifs_get_pnode(c, nnode, iip); if (IS_ERR(pnode)) - return ERR_PTR(PTR_ERR(pnode)); + return ERR_CAST(pnode); iip = (i & (UBIFS_LPT_FANOUT - 1)); dbg_lp("LEB %d, free %d, dirty %d, flags %d", lnum, pnode->lprops[iip].free, pnode->lprops[iip].dirty, @@ -1586,7 +1586,7 @@ struct ubifs_lprops *ubifs_lpt_lookup_dirty(struct ubifs_info *c, int lnum) nnode = c->nroot; nnode = dirty_cow_nnode(c, nnode); if (IS_ERR(nnode)) - return ERR_PTR(PTR_ERR(nnode)); + return ERR_CAST(nnode); i = lnum - c->main_first; shft = c->lpt_hght * UBIFS_LPT_FANOUT_SHIFT; for (h = 1; h < c->lpt_hght; h++) { @@ -1594,19 +1594,19 @@ struct ubifs_lprops *ubifs_lpt_lookup_dirty(struct ubifs_info *c, int lnum) shft -= UBIFS_LPT_FANOUT_SHIFT; nnode = ubifs_get_nnode(c, nnode, iip); if (IS_ERR(nnode)) - return ERR_PTR(PTR_ERR(nnode)); + return ERR_CAST(nnode); nnode = dirty_cow_nnode(c, nnode); if (IS_ERR(nnode)) - return ERR_PTR(PTR_ERR(nnode)); + return ERR_CAST(nnode); } iip = ((i >> shft) & (UBIFS_LPT_FANOUT - 1)); shft -= UBIFS_LPT_FANOUT_SHIFT; pnode = ubifs_get_pnode(c, nnode, iip); if (IS_ERR(pnode)) - return ERR_PTR(PTR_ERR(pnode)); + return ERR_CAST(pnode); pnode = dirty_cow_pnode(c, pnode); if (IS_ERR(pnode)) - return ERR_PTR(PTR_ERR(pnode)); + return ERR_CAST(pnode); iip = (i & (UBIFS_LPT_FANOUT - 1)); dbg_lp("LEB %d, free %d, dirty %d, flags %d", lnum, pnode->lprops[iip].free, pnode->lprops[iip].dirty, diff --git a/fs/ubifs/lpt_commit.c b/fs/ubifs/lpt_commit.c index 13cb7a4237b..d12535b7fc7 100644 --- a/fs/ubifs/lpt_commit.c +++ b/fs/ubifs/lpt_commit.c @@ -646,7 +646,7 @@ static struct ubifs_pnode *pnode_lookup(struct ubifs_info *c, int i) shft -= UBIFS_LPT_FANOUT_SHIFT; nnode = ubifs_get_nnode(c, nnode, iip); if (IS_ERR(nnode)) - return ERR_PTR(PTR_ERR(nnode)); + return ERR_CAST(nnode); } iip = ((i >> shft) & (UBIFS_LPT_FANOUT - 1)); return ubifs_get_pnode(c, nnode, iip); diff --git a/fs/ubifs/recovery.c b/fs/ubifs/recovery.c index 109c6ea03bb..daae9e1f538 100644 --- a/fs/ubifs/recovery.c +++ b/fs/ubifs/recovery.c @@ -24,7 +24,7 @@ * This file implements functions needed to recover from unclean un-mounts. * When UBIFS is mounted, it checks a flag on the master node to determine if * an un-mount was completed successfully. If not, the process of mounting - * incorparates additional checking and fixing of on-flash data structures. + * incorporates additional checking and fixing of on-flash data structures. * UBIFS always cleans away all remnants of an unclean un-mount, so that * errors do not accumulate. However UBIFS defers recovery if it is mounted * read-only, and the flash is not modified in that case. @@ -1063,8 +1063,21 @@ int ubifs_rcvry_gc_commit(struct ubifs_info *c) } err = ubifs_find_dirty_leb(c, &lp, wbuf->offs, 2); if (err) { - if (err == -ENOSPC) - dbg_err("could not find a dirty LEB"); + /* + * There are no dirty or empty LEBs subject to here being + * enough for the index. Try to use + * 'ubifs_find_free_leb_for_idx()', which will return any empty + * LEBs (ignoring index requirements). If the index then + * doesn't have enough LEBs the recovery commit will fail - + * which is the same result anyway i.e. recovery fails. So + * there is no problem ignoring index requirements and just + * grabbing a free LEB since we have already established there + * is not a dirty LEB we could have used instead. + */ + if (err == -ENOSPC) { + dbg_rcvry("could not find a dirty LEB"); + goto find_free; + } return err; } ubifs_assert(!(lp.flags & LPROPS_INDEX)); @@ -1139,8 +1152,8 @@ int ubifs_rcvry_gc_commit(struct ubifs_info *c) find_free: /* * There is no GC head LEB or the free space in the GC head LEB is too - * small. Allocate gc_lnum by calling 'ubifs_find_free_leb_for_idx()' so - * GC is not run. + * small, or there are not dirty LEBs. Allocate gc_lnum by calling + * 'ubifs_find_free_leb_for_idx()' so GC is not run. */ lnum = ubifs_find_free_leb_for_idx(c); if (lnum < 0) { diff --git a/fs/ubifs/super.c b/fs/ubifs/super.c index 4d2f2157dd3..5fc5a098897 100644 --- a/fs/ubifs/super.c +++ b/fs/ubifs/super.c @@ -1307,6 +1307,8 @@ static int mount_ubifs(struct ubifs_info *c) if (err) goto out_orphans; err = ubifs_rcvry_gc_commit(c); + if (err) + goto out_orphans; } else { err = take_gc_lnum(c); if (err) @@ -1318,7 +1320,7 @@ static int mount_ubifs(struct ubifs_info *c) */ err = ubifs_leb_unmap(c, c->gc_lnum); if (err) - return err; + goto out_orphans; } err = dbg_check_lprops(c); diff --git a/fs/udf/file.c b/fs/udf/file.c index 94e06d6bddb..6e450e01a1b 100644 --- a/fs/udf/file.c +++ b/fs/udf/file.c @@ -36,7 +36,6 @@ #include <linux/pagemap.h> #include <linux/buffer_head.h> #include <linux/aio.h> -#include <linux/smp_lock.h> #include "udf_i.h" #include "udf_sb.h" diff --git a/fs/udf/super.c b/fs/udf/super.c index 612d1e2e285..12bb651e540 100644 --- a/fs/udf/super.c +++ b/fs/udf/super.c @@ -1579,9 +1579,7 @@ static int udf_load_sequence(struct super_block *sb, struct buffer_head *bh, { struct anchorVolDescPtr *anchor; long main_s, main_e, reserve_s, reserve_e; - struct udf_sb_info *sbi; - sbi = UDF_SB(sb); anchor = (struct anchorVolDescPtr *)bh->b_data; /* Locate the main sequence */ diff --git a/fs/xfs/Makefile b/fs/xfs/Makefile index c8fb13f83b3..0dce969d6ca 100644 --- a/fs/xfs/Makefile +++ b/fs/xfs/Makefile @@ -87,11 +87,9 @@ xfs-y += xfs_alloc.o \ xfs_trans_buf.o \ xfs_trans_extfree.o \ xfs_trans_inode.o \ - xfs_trans_item.o \ xfs_utils.o \ xfs_vnodeops.o \ - xfs_rw.o \ - xfs_dmops.o + xfs_rw.o xfs-$(CONFIG_XFS_TRACE) += xfs_btree_trace.o diff --git a/fs/xfs/linux-2.6/xfs_acl.c b/fs/xfs/linux-2.6/xfs_acl.c index 9f769b5b38f..b2771862fd3 100644 --- a/fs/xfs/linux-2.6/xfs_acl.c +++ b/fs/xfs/linux-2.6/xfs_acl.c @@ -225,7 +225,7 @@ xfs_check_acl(struct inode *inode, int mask) struct posix_acl *acl; int error = -EAGAIN; - xfs_itrace_entry(ip); + trace_xfs_check_acl(ip); /* * If there is no attribute fork no ACL exists on this inode and diff --git a/fs/xfs/linux-2.6/xfs_aops.c b/fs/xfs/linux-2.6/xfs_aops.c index 34640d6dbdc..d24e78f32f3 100644 --- a/fs/xfs/linux-2.6/xfs_aops.c +++ b/fs/xfs/linux-2.6/xfs_aops.c @@ -21,19 +21,12 @@ #include "xfs_inum.h" #include "xfs_sb.h" #include "xfs_ag.h" -#include "xfs_dir2.h" #include "xfs_trans.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_dir2_sf.h" -#include "xfs_attr_sf.h" #include "xfs_dinode.h" #include "xfs_inode.h" #include "xfs_alloc.h" -#include "xfs_btree.h" #include "xfs_error.h" #include "xfs_rw.h" #include "xfs_iomap.h" @@ -92,18 +85,15 @@ void xfs_count_page_state( struct page *page, int *delalloc, - int *unmapped, int *unwritten) { struct buffer_head *bh, *head; - *delalloc = *unmapped = *unwritten = 0; + *delalloc = *unwritten = 0; bh = head = page_buffers(page); do { - if (buffer_uptodate(bh) && !buffer_mapped(bh)) - (*unmapped) = 1; - else if (buffer_unwritten(bh)) + if (buffer_unwritten(bh)) (*unwritten) = 1; else if (buffer_delay(bh)) (*delalloc) = 1; @@ -212,23 +202,17 @@ xfs_setfilesize( } /* - * Schedule IO completion handling on a xfsdatad if this was - * the final hold on this ioend. If we are asked to wait, - * flush the workqueue. + * Schedule IO completion handling on the final put of an ioend. */ STATIC void xfs_finish_ioend( - xfs_ioend_t *ioend, - int wait) + struct xfs_ioend *ioend) { if (atomic_dec_and_test(&ioend->io_remaining)) { - struct workqueue_struct *wq; - - wq = (ioend->io_type == IO_UNWRITTEN) ? - xfsconvertd_workqueue : xfsdatad_workqueue; - queue_work(wq, &ioend->io_work); - if (wait) - flush_workqueue(wq); + if (ioend->io_type == IO_UNWRITTEN) + queue_work(xfsconvertd_workqueue, &ioend->io_work); + else + queue_work(xfsdatad_workqueue, &ioend->io_work); } } @@ -272,11 +256,25 @@ xfs_end_io( */ if (error == EAGAIN) { atomic_inc(&ioend->io_remaining); - xfs_finish_ioend(ioend, 0); + xfs_finish_ioend(ioend); /* ensure we don't spin on blocked ioends */ delay(1); - } else + } else { + if (ioend->io_iocb) + aio_complete(ioend->io_iocb, ioend->io_result, 0); xfs_destroy_ioend(ioend); + } +} + +/* + * Call IO completion handling in caller context on the final put of an ioend. + */ +STATIC void +xfs_finish_ioend_sync( + struct xfs_ioend *ioend) +{ + if (atomic_dec_and_test(&ioend->io_remaining)) + xfs_end_io(&ioend->io_work); } /* @@ -309,6 +307,8 @@ xfs_alloc_ioend( atomic_inc(&XFS_I(ioend->io_inode)->i_iocount); ioend->io_offset = 0; ioend->io_size = 0; + ioend->io_iocb = NULL; + ioend->io_result = 0; INIT_WORK(&ioend->io_work, xfs_end_io); return ioend; @@ -358,7 +358,7 @@ xfs_end_bio( bio->bi_end_io = NULL; bio_put(bio); - xfs_finish_ioend(ioend, 0); + xfs_finish_ioend(ioend); } STATIC void @@ -500,7 +500,7 @@ xfs_submit_ioend( } if (bio) xfs_submit_ioend_bio(wbc, ioend, bio); - xfs_finish_ioend(ioend, 0); + xfs_finish_ioend(ioend); } while ((ioend = next) != NULL); } @@ -614,31 +614,30 @@ xfs_map_at_offset( STATIC unsigned int xfs_probe_page( struct page *page, - unsigned int pg_offset, - int mapped) + unsigned int pg_offset) { + struct buffer_head *bh, *head; int ret = 0; if (PageWriteback(page)) return 0; + if (!PageDirty(page)) + return 0; + if (!page->mapping) + return 0; + if (!page_has_buffers(page)) + return 0; - if (page->mapping && PageDirty(page)) { - if (page_has_buffers(page)) { - struct buffer_head *bh, *head; - - bh = head = page_buffers(page); - do { - if (!buffer_uptodate(bh)) - break; - if (mapped != buffer_mapped(bh)) - break; - ret += bh->b_size; - if (ret >= pg_offset) - break; - } while ((bh = bh->b_this_page) != head); - } else - ret = mapped ? 0 : PAGE_CACHE_SIZE; - } + bh = head = page_buffers(page); + do { + if (!buffer_uptodate(bh)) + break; + if (!buffer_mapped(bh)) + break; + ret += bh->b_size; + if (ret >= pg_offset) + break; + } while ((bh = bh->b_this_page) != head); return ret; } @@ -648,8 +647,7 @@ xfs_probe_cluster( struct inode *inode, struct page *startpage, struct buffer_head *bh, - struct buffer_head *head, - int mapped) + struct buffer_head *head) { struct pagevec pvec; pgoff_t tindex, tlast, tloff; @@ -658,7 +656,7 @@ xfs_probe_cluster( /* First sum forwards in this page */ do { - if (!buffer_uptodate(bh) || (mapped != buffer_mapped(bh))) + if (!buffer_uptodate(bh) || !buffer_mapped(bh)) return total; total += bh->b_size; } while ((bh = bh->b_this_page) != head); @@ -692,7 +690,7 @@ xfs_probe_cluster( pg_offset = PAGE_CACHE_SIZE; if (page->index == tindex && trylock_page(page)) { - pg_len = xfs_probe_page(page, pg_offset, mapped); + pg_len = xfs_probe_page(page, pg_offset); unlock_page(page); } @@ -761,7 +759,6 @@ xfs_convert_page( struct xfs_bmbt_irec *imap, xfs_ioend_t **ioendp, struct writeback_control *wbc, - int startio, int all_bh) { struct buffer_head *bh, *head; @@ -832,19 +829,14 @@ xfs_convert_page( ASSERT(imap->br_startblock != DELAYSTARTBLOCK); xfs_map_at_offset(inode, bh, imap, offset); - if (startio) { - xfs_add_to_ioend(inode, bh, offset, - type, ioendp, done); - } else { - set_buffer_dirty(bh); - unlock_buffer(bh); - mark_buffer_dirty(bh); - } + xfs_add_to_ioend(inode, bh, offset, type, + ioendp, done); + page_dirty--; count++; } else { type = IO_NEW; - if (buffer_mapped(bh) && all_bh && startio) { + if (buffer_mapped(bh) && all_bh) { lock_buffer(bh); xfs_add_to_ioend(inode, bh, offset, type, ioendp, done); @@ -859,14 +851,12 @@ xfs_convert_page( if (uptodate && bh == head) SetPageUptodate(page); - if (startio) { - if (count) { - wbc->nr_to_write--; - if (wbc->nr_to_write <= 0) - done = 1; - } - xfs_start_page_writeback(page, !page_dirty, count); + if (count) { + wbc->nr_to_write--; + if (wbc->nr_to_write <= 0) + done = 1; } + xfs_start_page_writeback(page, !page_dirty, count); return done; fail_unlock_page: @@ -886,7 +876,6 @@ xfs_cluster_write( struct xfs_bmbt_irec *imap, xfs_ioend_t **ioendp, struct writeback_control *wbc, - int startio, int all_bh, pgoff_t tlast) { @@ -902,7 +891,7 @@ xfs_cluster_write( for (i = 0; i < pagevec_count(&pvec); i++) { done = xfs_convert_page(inode, pvec.pages[i], tindex++, - imap, ioendp, wbc, startio, all_bh); + imap, ioendp, wbc, all_bh); if (done) break; } @@ -981,7 +970,7 @@ xfs_aops_discard_page( */ error = xfs_bmapi(NULL, ip, offset_fsb, 1, XFS_BMAPI_ENTIRE, NULL, 0, &imap, - &nimaps, NULL, NULL); + &nimaps, NULL); if (error) { /* something screwed, just bail */ @@ -1009,7 +998,7 @@ xfs_aops_discard_page( */ xfs_bmap_init(&flist, &firstblock); error = xfs_bunmapi(NULL, ip, offset_fsb, 1, 0, 1, &firstblock, - &flist, NULL, &done); + &flist, &done); ASSERT(!flist.xbf_count && !flist.xbf_first); if (error) { @@ -1032,50 +1021,66 @@ out_invalidate: } /* - * Calling this without startio set means we are being asked to make a dirty - * page ready for freeing it's buffers. When called with startio set then - * we are coming from writepage. + * Write out a dirty page. + * + * For delalloc space on the page we need to allocate space and flush it. + * For unwritten space on the page we need to start the conversion to + * regular allocated space. + * For any other dirty buffer heads on the page we should flush them. * - * When called with startio set it is important that we write the WHOLE - * page if possible. - * The bh->b_state's cannot know if any of the blocks or which block for - * that matter are dirty due to mmap writes, and therefore bh uptodate is - * only valid if the page itself isn't completely uptodate. Some layers - * may clear the page dirty flag prior to calling write page, under the - * assumption the entire page will be written out; by not writing out the - * whole page the page can be reused before all valid dirty data is - * written out. Note: in the case of a page that has been dirty'd by - * mapwrite and but partially setup by block_prepare_write the - * bh->b_states's will not agree and only ones setup by BPW/BCW will have - * valid state, thus the whole page must be written out thing. + * If we detect that a transaction would be required to flush the page, we + * have to check the process flags first, if we are already in a transaction + * or disk I/O during allocations is off, we need to fail the writepage and + * redirty the page. */ - STATIC int -xfs_page_state_convert( - struct inode *inode, - struct page *page, - struct writeback_control *wbc, - int startio, - int unmapped) /* also implies page uptodate */ +xfs_vm_writepage( + struct page *page, + struct writeback_control *wbc) { + struct inode *inode = page->mapping->host; + int delalloc, unwritten; struct buffer_head *bh, *head; struct xfs_bmbt_irec imap; xfs_ioend_t *ioend = NULL, *iohead = NULL; loff_t offset; - unsigned long p_offset = 0; unsigned int type; __uint64_t end_offset; pgoff_t end_index, last_index; ssize_t size, len; int flags, err, imap_valid = 0, uptodate = 1; - int page_dirty, count = 0; - int trylock = 0; - int all_bh = unmapped; + int count = 0; + int all_bh = 0; - if (startio) { - if (wbc->sync_mode == WB_SYNC_NONE && wbc->nonblocking) - trylock |= BMAPI_TRYLOCK; - } + trace_xfs_writepage(inode, page, 0); + + ASSERT(page_has_buffers(page)); + + /* + * Refuse to write the page out if we are called from reclaim context. + * + * This avoids stack overflows when called from deeply used stacks in + * random callers for direct reclaim or memcg reclaim. We explicitly + * allow reclaim from kswapd as the stack usage there is relatively low. + * + * This should really be done by the core VM, but until that happens + * filesystems like XFS, btrfs and ext4 have to take care of this + * by themselves. + */ + if ((current->flags & (PF_MEMALLOC|PF_KSWAPD)) == PF_MEMALLOC) + goto out_fail; + + /* + * We need a transaction if there are delalloc or unwritten buffers + * on the page. + * + * If we need a transaction and the process flags say we are already + * in a transaction, or no IO is allowed then mark the page dirty + * again and leave the page as is. + */ + xfs_count_page_state(page, &delalloc, &unwritten); + if ((current->flags & PF_FSTRANS) && (delalloc || unwritten)) + goto out_fail; /* Is this page beyond the end of the file? */ offset = i_size_read(inode); @@ -1084,50 +1089,33 @@ xfs_page_state_convert( if (page->index >= end_index) { if ((page->index >= end_index + 1) || !(i_size_read(inode) & (PAGE_CACHE_SIZE - 1))) { - if (startio) - unlock_page(page); + unlock_page(page); return 0; } } - /* - * page_dirty is initially a count of buffers on the page before - * EOF and is decremented as we move each into a cleanable state. - * - * Derivation: - * - * End offset is the highest offset that this page should represent. - * If we are on the last page, (end_offset & (PAGE_CACHE_SIZE - 1)) - * will evaluate non-zero and be less than PAGE_CACHE_SIZE and - * hence give us the correct page_dirty count. On any other page, - * it will be zero and in that case we need page_dirty to be the - * count of buffers on the page. - */ end_offset = min_t(unsigned long long, - (xfs_off_t)(page->index + 1) << PAGE_CACHE_SHIFT, offset); + (xfs_off_t)(page->index + 1) << PAGE_CACHE_SHIFT, + offset); len = 1 << inode->i_blkbits; - p_offset = min_t(unsigned long, end_offset & (PAGE_CACHE_SIZE - 1), - PAGE_CACHE_SIZE); - p_offset = p_offset ? roundup(p_offset, len) : PAGE_CACHE_SIZE; - page_dirty = p_offset / len; bh = head = page_buffers(page); offset = page_offset(page); flags = BMAPI_READ; type = IO_NEW; - /* TODO: cleanup count and page_dirty */ - do { if (offset >= end_offset) break; if (!buffer_uptodate(bh)) uptodate = 0; - if (!(PageUptodate(page) || buffer_uptodate(bh)) && !startio) { - /* - * the iomap is actually still valid, but the ioend - * isn't. shouldn't happen too often. - */ + + /* + * A hole may still be marked uptodate because discard_buffer + * leaves the flag set. + */ + if (!buffer_mapped(bh) && buffer_uptodate(bh)) { + ASSERT(!buffer_dirty(bh)); imap_valid = 0; continue; } @@ -1135,19 +1123,7 @@ xfs_page_state_convert( if (imap_valid) imap_valid = xfs_imap_valid(inode, &imap, offset); - /* - * First case, map an unwritten extent and prepare for - * extent state conversion transaction on completion. - * - * Second case, allocate space for a delalloc buffer. - * We can return EAGAIN here in the release page case. - * - * Third case, an unmapped buffer was found, and we are - * in a path where we need to write the whole page out. - */ - if (buffer_unwritten(bh) || buffer_delay(bh) || - ((buffer_uptodate(bh) || PageUptodate(page)) && - !buffer_mapped(bh) && (unmapped || startio))) { + if (buffer_unwritten(bh) || buffer_delay(bh)) { int new_ioend = 0; /* @@ -1161,15 +1137,16 @@ xfs_page_state_convert( flags = BMAPI_WRITE | BMAPI_IGNSTATE; } else if (buffer_delay(bh)) { type = IO_DELAY; - flags = BMAPI_ALLOCATE | trylock; - } else { - type = IO_NEW; - flags = BMAPI_WRITE | BMAPI_MMAP; + flags = BMAPI_ALLOCATE; + + if (wbc->sync_mode == WB_SYNC_NONE && + wbc->nonblocking) + flags |= BMAPI_TRYLOCK; } if (!imap_valid) { /* - * if we didn't have a valid mapping then we + * If we didn't have a valid mapping then we * need to ensure that we put the new mapping * in a new ioend structure. This needs to be * done to ensure that the ioends correctly @@ -1177,14 +1154,7 @@ xfs_page_state_convert( * for unwritten extent conversion. */ new_ioend = 1; - if (type == IO_NEW) { - size = xfs_probe_cluster(inode, - page, bh, head, 0); - } else { - size = len; - } - - err = xfs_map_blocks(inode, offset, size, + err = xfs_map_blocks(inode, offset, len, &imap, flags); if (err) goto error; @@ -1193,19 +1163,11 @@ xfs_page_state_convert( } if (imap_valid) { xfs_map_at_offset(inode, bh, &imap, offset); - if (startio) { - xfs_add_to_ioend(inode, bh, offset, - type, &ioend, - new_ioend); - } else { - set_buffer_dirty(bh); - unlock_buffer(bh); - mark_buffer_dirty(bh); - } - page_dirty--; + xfs_add_to_ioend(inode, bh, offset, type, + &ioend, new_ioend); count++; } - } else if (buffer_uptodate(bh) && startio) { + } else if (buffer_uptodate(bh)) { /* * we got here because the buffer is already mapped. * That means it must already have extents allocated @@ -1213,8 +1175,7 @@ xfs_page_state_convert( */ if (!imap_valid || flags != BMAPI_READ) { flags = BMAPI_READ; - size = xfs_probe_cluster(inode, page, bh, - head, 1); + size = xfs_probe_cluster(inode, page, bh, head); err = xfs_map_blocks(inode, offset, size, &imap, flags); if (err) @@ -1233,18 +1194,16 @@ xfs_page_state_convert( */ type = IO_NEW; if (trylock_buffer(bh)) { - ASSERT(buffer_mapped(bh)); if (imap_valid) all_bh = 1; xfs_add_to_ioend(inode, bh, offset, type, &ioend, !imap_valid); - page_dirty--; count++; } else { imap_valid = 0; } - } else if ((buffer_uptodate(bh) || PageUptodate(page)) && - (unmapped || startio)) { + } else if (PageUptodate(page)) { + ASSERT(buffer_mapped(bh)); imap_valid = 0; } @@ -1256,8 +1215,7 @@ xfs_page_state_convert( if (uptodate && bh == head) SetPageUptodate(page); - if (startio) - xfs_start_page_writeback(page, 1, count); + xfs_start_page_writeback(page, 1, count); if (ioend && imap_valid) { xfs_off_t end_index; @@ -1275,131 +1233,27 @@ xfs_page_state_convert( end_index = last_index; xfs_cluster_write(inode, page->index + 1, &imap, &ioend, - wbc, startio, all_bh, end_index); + wbc, all_bh, end_index); } if (iohead) xfs_submit_ioend(wbc, iohead); - return page_dirty; + return 0; error: if (iohead) xfs_cancel_ioend(iohead); - /* - * If it's delalloc and we have nowhere to put it, - * throw it away, unless the lower layers told - * us to try again. - */ - if (err != -EAGAIN) { - if (!unmapped) - xfs_aops_discard_page(page); - ClearPageUptodate(page); - } + xfs_aops_discard_page(page); + ClearPageUptodate(page); + unlock_page(page); return err; -} - -/* - * writepage: Called from one of two places: - * - * 1. we are flushing a delalloc buffer head. - * - * 2. we are writing out a dirty page. Typically the page dirty - * state is cleared before we get here. In this case is it - * conceivable we have no buffer heads. - * - * For delalloc space on the page we need to allocate space and - * flush it. For unmapped buffer heads on the page we should - * allocate space if the page is uptodate. For any other dirty - * buffer heads on the page we should flush them. - * - * If we detect that a transaction would be required to flush - * the page, we have to check the process flags first, if we - * are already in a transaction or disk I/O during allocations - * is off, we need to fail the writepage and redirty the page. - */ - -STATIC int -xfs_vm_writepage( - struct page *page, - struct writeback_control *wbc) -{ - int error; - int need_trans; - int delalloc, unmapped, unwritten; - struct inode *inode = page->mapping->host; - - trace_xfs_writepage(inode, page, 0); - - /* - * Refuse to write the page out if we are called from reclaim context. - * - * This is primarily to avoid stack overflows when called from deep - * used stacks in random callers for direct reclaim, but disabling - * reclaim for kswap is a nice side-effect as kswapd causes rather - * suboptimal I/O patters, too. - * - * This should really be done by the core VM, but until that happens - * filesystems like XFS, btrfs and ext4 have to take care of this - * by themselves. - */ - if (current->flags & PF_MEMALLOC) - goto out_fail; - - /* - * We need a transaction if: - * 1. There are delalloc buffers on the page - * 2. The page is uptodate and we have unmapped buffers - * 3. The page is uptodate and we have no buffers - * 4. There are unwritten buffers on the page - */ - - if (!page_has_buffers(page)) { - unmapped = 1; - need_trans = 1; - } else { - xfs_count_page_state(page, &delalloc, &unmapped, &unwritten); - if (!PageUptodate(page)) - unmapped = 0; - need_trans = delalloc + unmapped + unwritten; - } - - /* - * If we need a transaction and the process flags say - * we are already in a transaction, or no IO is allowed - * then mark the page dirty again and leave the page - * as is. - */ - if (current_test_flags(PF_FSTRANS) && need_trans) - goto out_fail; - - /* - * Delay hooking up buffer heads until we have - * made our go/no-go decision. - */ - if (!page_has_buffers(page)) - create_empty_buffers(page, 1 << inode->i_blkbits, 0); - - /* - * Convert delayed allocate, unwritten or unmapped space - * to real space and flush out to disk. - */ - error = xfs_page_state_convert(inode, page, wbc, 1, unmapped); - if (error == -EAGAIN) - goto out_fail; - if (unlikely(error < 0)) - goto out_unlock; - - return 0; out_fail: redirty_page_for_writepage(wbc, page); unlock_page(page); return 0; -out_unlock: - unlock_page(page); - return error; } STATIC int @@ -1413,65 +1267,27 @@ xfs_vm_writepages( /* * Called to move a page into cleanable state - and from there - * to be released. Possibly the page is already clean. We always + * to be released. The page should already be clean. We always * have buffer heads in this call. * - * Returns 0 if the page is ok to release, 1 otherwise. - * - * Possible scenarios are: - * - * 1. We are being called to release a page which has been written - * to via regular I/O. buffer heads will be dirty and possibly - * delalloc. If no delalloc buffer heads in this case then we - * can just return zero. - * - * 2. We are called to release a page which has been written via - * mmap, all we need to do is ensure there is no delalloc - * state in the buffer heads, if not we can let the caller - * free them and we should come back later via writepage. + * Returns 1 if the page is ok to release, 0 otherwise. */ STATIC int xfs_vm_releasepage( struct page *page, gfp_t gfp_mask) { - struct inode *inode = page->mapping->host; - int dirty, delalloc, unmapped, unwritten; - struct writeback_control wbc = { - .sync_mode = WB_SYNC_ALL, - .nr_to_write = 1, - }; + int delalloc, unwritten; - trace_xfs_releasepage(inode, page, 0); - - if (!page_has_buffers(page)) - return 0; + trace_xfs_releasepage(page->mapping->host, page, 0); - xfs_count_page_state(page, &delalloc, &unmapped, &unwritten); - if (!delalloc && !unwritten) - goto free_buffers; + xfs_count_page_state(page, &delalloc, &unwritten); - if (!(gfp_mask & __GFP_FS)) + if (WARN_ON(delalloc)) return 0; - - /* If we are already inside a transaction or the thread cannot - * do I/O, we cannot release this page. - */ - if (current_test_flags(PF_FSTRANS)) + if (WARN_ON(unwritten)) return 0; - /* - * Convert delalloc space to real space, do not flush the - * data out to disk, that will be done by the caller. - * Never need to allocate space here - we will always - * come back to writepage in that case. - */ - dirty = xfs_page_state_convert(inode, page, &wbc, 0, 0); - if (dirty == 0 && !unwritten) - goto free_buffers; - return 0; - -free_buffers: return try_to_free_buffers(page); } @@ -1481,9 +1297,9 @@ __xfs_get_blocks( sector_t iblock, struct buffer_head *bh_result, int create, - int direct, - bmapi_flags_t flags) + int direct) { + int flags = create ? BMAPI_WRITE : BMAPI_READ; struct xfs_bmbt_irec imap; xfs_off_t offset; ssize_t size; @@ -1498,8 +1314,11 @@ __xfs_get_blocks( if (!create && direct && offset >= i_size_read(inode)) return 0; - error = xfs_iomap(XFS_I(inode), offset, size, - create ? flags : BMAPI_READ, &imap, &nimap, &new); + if (direct && create) + flags |= BMAPI_DIRECT; + + error = xfs_iomap(XFS_I(inode), offset, size, flags, &imap, &nimap, + &new); if (error) return -error; if (nimap == 0) @@ -1579,8 +1398,7 @@ xfs_get_blocks( struct buffer_head *bh_result, int create) { - return __xfs_get_blocks(inode, iblock, - bh_result, create, 0, BMAPI_WRITE); + return __xfs_get_blocks(inode, iblock, bh_result, create, 0); } STATIC int @@ -1590,61 +1408,59 @@ xfs_get_blocks_direct( struct buffer_head *bh_result, int create) { - return __xfs_get_blocks(inode, iblock, - bh_result, create, 1, BMAPI_WRITE|BMAPI_DIRECT); + return __xfs_get_blocks(inode, iblock, bh_result, create, 1); } +/* + * Complete a direct I/O write request. + * + * If the private argument is non-NULL __xfs_get_blocks signals us that we + * need to issue a transaction to convert the range from unwritten to written + * extents. In case this is regular synchronous I/O we just call xfs_end_io + * to do this and we are done. But in case this was a successfull AIO + * request this handler is called from interrupt context, from which we + * can't start transactions. In that case offload the I/O completion to + * the workqueues we also use for buffered I/O completion. + */ STATIC void -xfs_end_io_direct( - struct kiocb *iocb, - loff_t offset, - ssize_t size, - void *private) +xfs_end_io_direct_write( + struct kiocb *iocb, + loff_t offset, + ssize_t size, + void *private, + int ret, + bool is_async) { - xfs_ioend_t *ioend = iocb->private; + struct xfs_ioend *ioend = iocb->private; /* - * Non-NULL private data means we need to issue a transaction to - * convert a range from unwritten to written extents. This needs - * to happen from process context but aio+dio I/O completion - * happens from irq context so we need to defer it to a workqueue. - * This is not necessary for synchronous direct I/O, but we do - * it anyway to keep the code uniform and simpler. - * - * Well, if only it were that simple. Because synchronous direct I/O - * requires extent conversion to occur *before* we return to userspace, - * we have to wait for extent conversion to complete. Look at the - * iocb that has been passed to us to determine if this is AIO or - * not. If it is synchronous, tell xfs_finish_ioend() to kick the - * workqueue and wait for it to complete. - * - * The core direct I/O code might be changed to always call the - * completion handler in the future, in which case all this can - * go away. + * blockdev_direct_IO can return an error even after the I/O + * completion handler was called. Thus we need to protect + * against double-freeing. */ + iocb->private = NULL; + ioend->io_offset = offset; ioend->io_size = size; - if (ioend->io_type == IO_READ) { - xfs_finish_ioend(ioend, 0); - } else if (private && size > 0) { - xfs_finish_ioend(ioend, is_sync_kiocb(iocb)); - } else { + if (private && size > 0) + ioend->io_type = IO_UNWRITTEN; + + if (is_async) { /* - * A direct I/O write ioend starts it's life in unwritten - * state in case they map an unwritten extent. This write - * didn't map an unwritten extent so switch it's completion - * handler. + * If we are converting an unwritten extent we need to delay + * the AIO completion until after the unwrittent extent + * conversion has completed, otherwise do it ASAP. */ - ioend->io_type = IO_NEW; - xfs_finish_ioend(ioend, 0); + if (ioend->io_type == IO_UNWRITTEN) { + ioend->io_iocb = iocb; + ioend->io_result = ret; + } else { + aio_complete(iocb, ret, 0); + } + xfs_finish_ioend(ioend); + } else { + xfs_finish_ioend_sync(ioend); } - - /* - * blockdev_direct_IO can return an error even after the I/O - * completion handler was called. Thus we need to protect - * against double-freeing. - */ - iocb->private = NULL; } STATIC ssize_t @@ -1655,23 +1471,26 @@ xfs_vm_direct_IO( loff_t offset, unsigned long nr_segs) { - struct file *file = iocb->ki_filp; - struct inode *inode = file->f_mapping->host; - struct block_device *bdev; - ssize_t ret; - - bdev = xfs_find_bdev_for_inode(inode); - - iocb->private = xfs_alloc_ioend(inode, rw == WRITE ? - IO_UNWRITTEN : IO_READ); - - ret = blockdev_direct_IO_no_locking(rw, iocb, inode, bdev, iov, - offset, nr_segs, - xfs_get_blocks_direct, - xfs_end_io_direct); + struct inode *inode = iocb->ki_filp->f_mapping->host; + struct block_device *bdev = xfs_find_bdev_for_inode(inode); + ssize_t ret; + + if (rw & WRITE) { + iocb->private = xfs_alloc_ioend(inode, IO_NEW); + + ret = blockdev_direct_IO_no_locking(rw, iocb, inode, bdev, iov, + offset, nr_segs, + xfs_get_blocks_direct, + xfs_end_io_direct_write); + if (ret != -EIOCBQUEUED && iocb->private) + xfs_destroy_ioend(iocb->private); + } else { + ret = blockdev_direct_IO_no_locking(rw, iocb, inode, bdev, iov, + offset, nr_segs, + xfs_get_blocks_direct, + NULL); + } - if (unlikely(ret != -EIOCBQUEUED && iocb->private)) - xfs_destroy_ioend(iocb->private); return ret; } @@ -1686,8 +1505,8 @@ xfs_vm_write_begin( void **fsdata) { *pagep = NULL; - return block_write_begin(file, mapping, pos, len, flags, pagep, fsdata, - xfs_get_blocks); + return block_write_begin(file, mapping, pos, len, flags | AOP_FLAG_NOFS, + pagep, fsdata, xfs_get_blocks); } STATIC sector_t @@ -1698,7 +1517,7 @@ xfs_vm_bmap( struct inode *inode = (struct inode *)mapping->host; struct xfs_inode *ip = XFS_I(inode); - xfs_itrace_entry(XFS_I(inode)); + trace_xfs_vm_bmap(XFS_I(inode)); xfs_ilock(ip, XFS_IOLOCK_SHARED); xfs_flush_pages(ip, (xfs_off_t)0, -1, 0, FI_REMAPF); xfs_iunlock(ip, XFS_IOLOCK_SHARED); diff --git a/fs/xfs/linux-2.6/xfs_aops.h b/fs/xfs/linux-2.6/xfs_aops.h index 4cfc6ea87df..c5057fb6237 100644 --- a/fs/xfs/linux-2.6/xfs_aops.h +++ b/fs/xfs/linux-2.6/xfs_aops.h @@ -37,6 +37,8 @@ typedef struct xfs_ioend { size_t io_size; /* size of the extent */ xfs_off_t io_offset; /* offset in the file */ struct work_struct io_work; /* xfsdatad work queue */ + struct kiocb *io_iocb; + int io_result; } xfs_ioend_t; extern const struct address_space_operations xfs_address_space_operations; @@ -45,6 +47,6 @@ 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 *); -extern void xfs_count_page_state(struct page *, int *, int *, int *); +extern void xfs_count_page_state(struct page *, int *, int *); #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 2ee3f7a6016..ea79072f521 100644 --- a/fs/xfs/linux-2.6/xfs_buf.c +++ b/fs/xfs/linux-2.6/xfs_buf.c @@ -39,7 +39,6 @@ #include "xfs_inum.h" #include "xfs_log.h" #include "xfs_ag.h" -#include "xfs_dmapi.h" #include "xfs_mount.h" #include "xfs_trace.h" @@ -579,9 +578,9 @@ _xfs_buf_read( XBF_READ_AHEAD | _XBF_RUN_QUEUES); status = xfs_buf_iorequest(bp); - if (!status && !(flags & XBF_ASYNC)) - status = xfs_buf_iowait(bp); - return status; + if (status || XFS_BUF_ISERROR(bp) || (flags & XBF_ASYNC)) + return status; + return xfs_buf_iowait(bp); } xfs_buf_t * @@ -897,36 +896,6 @@ xfs_buf_unlock( trace_xfs_buf_unlock(bp, _RET_IP_); } - -/* - * Pinning Buffer Storage in Memory - * Ensure that no attempt to force a buffer to disk will succeed. - */ -void -xfs_buf_pin( - xfs_buf_t *bp) -{ - trace_xfs_buf_pin(bp, _RET_IP_); - atomic_inc(&bp->b_pin_count); -} - -void -xfs_buf_unpin( - xfs_buf_t *bp) -{ - trace_xfs_buf_unpin(bp, _RET_IP_); - - if (atomic_dec_and_test(&bp->b_pin_count)) - wake_up_all(&bp->b_waiters); -} - -int -xfs_buf_ispin( - xfs_buf_t *bp) -{ - return atomic_read(&bp->b_pin_count); -} - STATIC void xfs_buf_wait_unpin( xfs_buf_t *bp) @@ -1018,13 +987,12 @@ xfs_bwrite( { int error; - bp->b_strat = xfs_bdstrat_cb; bp->b_mount = mp; bp->b_flags |= XBF_WRITE; bp->b_flags &= ~(XBF_ASYNC | XBF_READ); xfs_buf_delwri_dequeue(bp); - xfs_buf_iostrategy(bp); + xfs_bdstrat_cb(bp); error = xfs_buf_iowait(bp); if (error) @@ -1040,7 +1008,6 @@ xfs_bdwrite( { trace_xfs_buf_bdwrite(bp, _RET_IP_); - bp->b_strat = xfs_bdstrat_cb; bp->b_mount = mp; bp->b_flags &= ~XBF_READ; @@ -1075,7 +1042,6 @@ xfs_bioerror( XFS_BUF_UNDONE(bp); XFS_BUF_STALE(bp); - XFS_BUF_CLR_BDSTRAT_FUNC(bp); xfs_biodone(bp); return EIO; @@ -1105,7 +1071,6 @@ xfs_bioerror_relse( XFS_BUF_DONE(bp); XFS_BUF_STALE(bp); XFS_BUF_CLR_IODONE_FUNC(bp); - XFS_BUF_CLR_BDSTRAT_FUNC(bp); if (!(fl & XBF_ASYNC)) { /* * Mark b_error and B_ERROR _both_. @@ -1311,8 +1276,19 @@ submit_io: if (size) goto next_chunk; } else { - bio_put(bio); + /* + * if we get here, no pages were added to the bio. However, + * we can't just error out here - if the pages are locked then + * we have to unlock them otherwise we can hang on a later + * access to the page. + */ xfs_buf_ioerror(bp, EIO); + if (bp->b_flags & _XBF_PAGE_LOCKED) { + int i; + for (i = 0; i < bp->b_page_count; i++) + unlock_page(bp->b_pages[i]); + } + bio_put(bio); } } @@ -1804,7 +1780,7 @@ xfs_buf_delwri_split( trace_xfs_buf_delwri_split(bp, _RET_IP_); ASSERT(bp->b_flags & XBF_DELWRI); - if (!xfs_buf_ispin(bp) && !xfs_buf_cond_lock(bp)) { + if (!XFS_BUF_ISPINNED(bp) && !xfs_buf_cond_lock(bp)) { if (!force && time_before(jiffies, bp->b_queuetime + age)) { xfs_buf_unlock(bp); @@ -1889,7 +1865,7 @@ xfsbufd( struct xfs_buf *bp; bp = list_first_entry(&tmp, struct xfs_buf, b_list); list_del_init(&bp->b_list); - xfs_buf_iostrategy(bp); + xfs_bdstrat_cb(bp); count++; } if (count) @@ -1936,7 +1912,7 @@ xfs_flush_buftarg( bp->b_flags &= ~XBF_ASYNC; list_add(&bp->b_list, &wait_list); } - xfs_buf_iostrategy(bp); + xfs_bdstrat_cb(bp); } if (wait) { diff --git a/fs/xfs/linux-2.6/xfs_buf.h b/fs/xfs/linux-2.6/xfs_buf.h index 5fbecefa5df..d072e5ff923 100644 --- a/fs/xfs/linux-2.6/xfs_buf.h +++ b/fs/xfs/linux-2.6/xfs_buf.h @@ -44,57 +44,57 @@ typedef enum { XBRW_ZERO = 3, /* Zero target memory */ } xfs_buf_rw_t; -typedef enum { - XBF_READ = (1 << 0), /* buffer intended for reading from device */ - XBF_WRITE = (1 << 1), /* buffer intended for writing to device */ - XBF_MAPPED = (1 << 2), /* buffer mapped (b_addr valid) */ - XBF_ASYNC = (1 << 4), /* initiator will not wait for completion */ - XBF_DONE = (1 << 5), /* all pages in the buffer uptodate */ - XBF_DELWRI = (1 << 6), /* buffer has dirty pages */ - XBF_STALE = (1 << 7), /* buffer has been staled, do not find it */ - XBF_FS_MANAGED = (1 << 8), /* filesystem controls freeing memory */ - XBF_ORDERED = (1 << 11), /* use ordered writes */ - XBF_READ_AHEAD = (1 << 12), /* asynchronous read-ahead */ - XBF_LOG_BUFFER = (1 << 13), /* this is a buffer used for the log */ - - /* flags used only as arguments to access routines */ - XBF_LOCK = (1 << 14), /* lock requested */ - XBF_TRYLOCK = (1 << 15), /* lock requested, but do not wait */ - XBF_DONT_BLOCK = (1 << 16), /* do not block in current thread */ - - /* flags used only internally */ - _XBF_PAGE_CACHE = (1 << 17),/* backed by pagecache */ - _XBF_PAGES = (1 << 18), /* backed by refcounted pages */ - _XBF_RUN_QUEUES = (1 << 19),/* run block device task queue */ - _XBF_DELWRI_Q = (1 << 21), /* buffer on delwri queue */ - - /* - * Special flag for supporting metadata blocks smaller than a FSB. - * - * In this case we can have multiple xfs_buf_t on a single page and - * need to lock out concurrent xfs_buf_t readers as they only - * serialise access to the buffer. - * - * If the FSB size >= PAGE_CACHE_SIZE case, we have no serialisation - * between reads of the page. Hence we can have one thread read the - * page and modify it, but then race with another thread that thinks - * the page is not up-to-date and hence reads it again. - * - * The result is that the first modifcation to the page is lost. - * This sort of AGF/AGI reading race can happen when unlinking inodes - * that require truncation and results in the AGI unlinked list - * modifications being lost. - */ - _XBF_PAGE_LOCKED = (1 << 22), - - /* - * If we try a barrier write, but it fails we have to communicate - * this to the upper layers. Unfortunately b_error gets overwritten - * when the buffer is re-issued so we have to add another flag to - * keep this information. - */ - _XFS_BARRIER_FAILED = (1 << 23), -} xfs_buf_flags_t; +#define XBF_READ (1 << 0) /* buffer intended for reading from device */ +#define XBF_WRITE (1 << 1) /* buffer intended for writing to device */ +#define XBF_MAPPED (1 << 2) /* buffer mapped (b_addr valid) */ +#define XBF_ASYNC (1 << 4) /* initiator will not wait for completion */ +#define XBF_DONE (1 << 5) /* all pages in the buffer uptodate */ +#define XBF_DELWRI (1 << 6) /* buffer has dirty pages */ +#define XBF_STALE (1 << 7) /* buffer has been staled, do not find it */ +#define XBF_FS_MANAGED (1 << 8) /* filesystem controls freeing memory */ +#define XBF_ORDERED (1 << 11)/* use ordered writes */ +#define XBF_READ_AHEAD (1 << 12)/* asynchronous read-ahead */ +#define XBF_LOG_BUFFER (1 << 13)/* this is a buffer used for the log */ + +/* flags used only as arguments to access routines */ +#define XBF_LOCK (1 << 14)/* lock requested */ +#define XBF_TRYLOCK (1 << 15)/* lock requested, but do not wait */ +#define XBF_DONT_BLOCK (1 << 16)/* do not block in current thread */ + +/* flags used only internally */ +#define _XBF_PAGE_CACHE (1 << 17)/* backed by pagecache */ +#define _XBF_PAGES (1 << 18)/* backed by refcounted pages */ +#define _XBF_RUN_QUEUES (1 << 19)/* run block device task queue */ +#define _XBF_DELWRI_Q (1 << 21)/* buffer on delwri queue */ + +/* + * Special flag for supporting metadata blocks smaller than a FSB. + * + * In this case we can have multiple xfs_buf_t on a single page and + * need to lock out concurrent xfs_buf_t readers as they only + * serialise access to the buffer. + * + * If the FSB size >= PAGE_CACHE_SIZE case, we have no serialisation + * between reads of the page. Hence we can have one thread read the + * page and modify it, but then race with another thread that thinks + * the page is not up-to-date and hence reads it again. + * + * The result is that the first modifcation to the page is lost. + * This sort of AGF/AGI reading race can happen when unlinking inodes + * that require truncation and results in the AGI unlinked list + * modifications being lost. + */ +#define _XBF_PAGE_LOCKED (1 << 22) + +/* + * If we try a barrier write, but it fails we have to communicate + * this to the upper layers. Unfortunately b_error gets overwritten + * when the buffer is re-issued so we have to add another flag to + * keep this information. + */ +#define _XFS_BARRIER_FAILED (1 << 23) + +typedef unsigned int xfs_buf_flags_t; #define XFS_BUF_FLAGS \ { XBF_READ, "READ" }, \ @@ -187,7 +187,6 @@ typedef struct xfs_buf { atomic_t b_io_remaining; /* #outstanding I/O requests */ xfs_buf_iodone_t b_iodone; /* I/O completion function */ xfs_buf_relse_t b_relse; /* releasing function */ - xfs_buf_bdstrat_t b_strat; /* pre-write function */ struct completion b_iowait; /* queue for I/O waiters */ void *b_fspriv; void *b_fspriv2; @@ -245,11 +244,6 @@ extern int xfs_buf_iowait(xfs_buf_t *); extern void xfs_buf_iomove(xfs_buf_t *, size_t, size_t, void *, xfs_buf_rw_t); -static inline int xfs_buf_iostrategy(xfs_buf_t *bp) -{ - return bp->b_strat ? bp->b_strat(bp) : xfs_buf_iorequest(bp); -} - static inline int xfs_buf_geterror(xfs_buf_t *bp) { return bp ? bp->b_error : ENOMEM; @@ -258,11 +252,6 @@ static inline int xfs_buf_geterror(xfs_buf_t *bp) /* Buffer Utility Routines */ extern xfs_caddr_t xfs_buf_offset(xfs_buf_t *, size_t); -/* Pinning Buffer Storage in Memory */ -extern void xfs_buf_pin(xfs_buf_t *); -extern void xfs_buf_unpin(xfs_buf_t *); -extern int xfs_buf_ispin(xfs_buf_t *); - /* Delayed Write Buffer Routines */ extern void xfs_buf_delwri_dequeue(xfs_buf_t *); extern void xfs_buf_delwri_promote(xfs_buf_t *); @@ -326,8 +315,6 @@ extern void xfs_buf_terminate(void); #define XFS_BUF_IODONE_FUNC(bp) ((bp)->b_iodone) #define XFS_BUF_SET_IODONE_FUNC(bp, func) ((bp)->b_iodone = (func)) #define XFS_BUF_CLR_IODONE_FUNC(bp) ((bp)->b_iodone = NULL) -#define XFS_BUF_SET_BDSTRAT_FUNC(bp, func) ((bp)->b_strat = (func)) -#define XFS_BUF_CLR_BDSTRAT_FUNC(bp) ((bp)->b_strat = NULL) #define XFS_BUF_FSPRIVATE(bp, type) ((type)(bp)->b_fspriv) #define XFS_BUF_SET_FSPRIVATE(bp, val) ((bp)->b_fspriv = (void*)(val)) @@ -351,7 +338,7 @@ extern void xfs_buf_terminate(void); #define XFS_BUF_SET_VTYPE(bp, type) do { } while (0) #define XFS_BUF_SET_REF(bp, ref) do { } while (0) -#define XFS_BUF_ISPINNED(bp) xfs_buf_ispin(bp) +#define XFS_BUF_ISPINNED(bp) atomic_read(&((bp)->b_pin_count)) #define XFS_BUF_VALUSEMA(bp) xfs_buf_lock_value(bp) #define XFS_BUF_CPSEMA(bp) (xfs_buf_cond_lock(bp) == 0) @@ -370,8 +357,6 @@ static inline void xfs_buf_relse(xfs_buf_t *bp) xfs_buf_rele(bp); } -#define xfs_bpin(bp) xfs_buf_pin(bp) -#define xfs_bunpin(bp) xfs_buf_unpin(bp) #define xfs_biodone(bp) xfs_buf_ioend(bp, 0) #define xfs_biomove(bp, off, len, data, rw) \ diff --git a/fs/xfs/linux-2.6/xfs_dmapi_priv.h b/fs/xfs/linux-2.6/xfs_dmapi_priv.h deleted file mode 100644 index a8b0b1685ee..00000000000 --- a/fs/xfs/linux-2.6/xfs_dmapi_priv.h +++ /dev/null @@ -1,28 +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_DMAPI_PRIV_H__ -#define __XFS_DMAPI_PRIV_H__ - -/* - * Based on IO_ISDIRECT, decide which i_ flag is set. - */ -#define DM_SEM_FLAG_RD(ioflags) (((ioflags) & IO_ISDIRECT) ? \ - DM_FLAGS_IMUX : 0) -#define DM_SEM_FLAG_WR (DM_FLAGS_IALLOCSEM_WR | DM_FLAGS_IMUX) - -#endif /*__XFS_DMAPI_PRIV_H__*/ diff --git a/fs/xfs/linux-2.6/xfs_export.c b/fs/xfs/linux-2.6/xfs_export.c index e7839ee49e4..3764d74790e 100644 --- a/fs/xfs/linux-2.6/xfs_export.c +++ b/fs/xfs/linux-2.6/xfs_export.c @@ -23,13 +23,13 @@ #include "xfs_sb.h" #include "xfs_ag.h" #include "xfs_dir2.h" -#include "xfs_dmapi.h" #include "xfs_mount.h" #include "xfs_export.h" #include "xfs_vnodeops.h" #include "xfs_bmap_btree.h" #include "xfs_inode.h" #include "xfs_inode_item.h" +#include "xfs_trace.h" /* * Note that we only accept fileids which are long enough rather than allow @@ -132,8 +132,7 @@ xfs_nfs_get_inode( * fine and not an indication of a corrupted filesystem as clients can * send invalid file handles and we have to handle it gracefully.. */ - error = xfs_iget(mp, NULL, ino, XFS_IGET_UNTRUSTED, - XFS_ILOCK_SHARED, &ip); + error = xfs_iget(mp, NULL, ino, XFS_IGET_UNTRUSTED, 0, &ip); if (error) { /* * EINVAL means the inode cluster doesn't exist anymore. @@ -148,11 +147,10 @@ xfs_nfs_get_inode( } if (ip->i_d.di_gen != generation) { - xfs_iput_new(ip, XFS_ILOCK_SHARED); + IRELE(ip); return ERR_PTR(-ENOENT); } - xfs_iunlock(ip, XFS_ILOCK_SHARED); return VFS_I(ip); } diff --git a/fs/xfs/linux-2.6/xfs_file.c b/fs/xfs/linux-2.6/xfs_file.c index 257a56b127c..ba8ad422a16 100644 --- a/fs/xfs/linux-2.6/xfs_file.c +++ b/fs/xfs/linux-2.6/xfs_file.c @@ -22,23 +22,15 @@ #include "xfs_inum.h" #include "xfs_sb.h" #include "xfs_ag.h" -#include "xfs_dir2.h" #include "xfs_trans.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_alloc.h" -#include "xfs_btree.h" -#include "xfs_attr_sf.h" -#include "xfs_dir2_sf.h" #include "xfs_dinode.h" #include "xfs_inode.h" #include "xfs_inode_item.h" #include "xfs_bmap.h" #include "xfs_error.h" -#include "xfs_rw.h" #include "xfs_vnodeops.h" #include "xfs_da_btree.h" #include "xfs_ioctl.h" @@ -108,7 +100,7 @@ xfs_file_fsync( int error = 0; int log_flushed = 0; - xfs_itrace_entry(ip); + trace_xfs_file_fsync(ip); if (XFS_FORCED_SHUTDOWN(ip->i_mount)) return -XFS_ERROR(EIO); @@ -166,8 +158,7 @@ xfs_file_fsync( * transaction. So we play it safe and fire off the * transaction anyway. */ - xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL); - xfs_trans_ihold(tp, ip); + xfs_trans_ijoin(tp, ip); xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); xfs_trans_set_sync(tp); error = _xfs_trans_commit(tp, 0, &log_flushed); @@ -275,20 +266,6 @@ xfs_file_aio_read( mutex_lock(&inode->i_mutex); xfs_ilock(ip, XFS_IOLOCK_SHARED); - if (DM_EVENT_ENABLED(ip, DM_EVENT_READ) && !(ioflags & IO_INVIS)) { - int dmflags = FILP_DELAY_FLAG(file) | DM_SEM_FLAG_RD(ioflags); - int iolock = XFS_IOLOCK_SHARED; - - ret = -XFS_SEND_DATA(mp, DM_EVENT_READ, ip, iocb->ki_pos, size, - dmflags, &iolock); - if (ret) { - xfs_iunlock(ip, XFS_IOLOCK_SHARED); - if (unlikely(ioflags & IO_ISDIRECT)) - mutex_unlock(&inode->i_mutex); - return ret; - } - } - if (unlikely(ioflags & IO_ISDIRECT)) { if (inode->i_mapping->nrpages) { ret = -xfs_flushinval_pages(ip, @@ -321,7 +298,6 @@ xfs_file_splice_read( unsigned int flags) { struct xfs_inode *ip = XFS_I(infilp->f_mapping->host); - struct xfs_mount *mp = ip->i_mount; int ioflags = 0; ssize_t ret; @@ -335,18 +311,6 @@ xfs_file_splice_read( xfs_ilock(ip, XFS_IOLOCK_SHARED); - if (DM_EVENT_ENABLED(ip, DM_EVENT_READ) && !(ioflags & IO_INVIS)) { - int iolock = XFS_IOLOCK_SHARED; - int error; - - error = XFS_SEND_DATA(mp, DM_EVENT_READ, ip, *ppos, count, - FILP_DELAY_FLAG(infilp), &iolock); - if (error) { - xfs_iunlock(ip, XFS_IOLOCK_SHARED); - return -error; - } - } - trace_xfs_file_splice_read(ip, count, *ppos, ioflags); ret = generic_file_splice_read(infilp, ppos, pipe, count, flags); @@ -367,7 +331,6 @@ xfs_file_splice_write( { struct inode *inode = outfilp->f_mapping->host; struct xfs_inode *ip = XFS_I(inode); - struct xfs_mount *mp = ip->i_mount; xfs_fsize_t isize, new_size; int ioflags = 0; ssize_t ret; @@ -382,18 +345,6 @@ xfs_file_splice_write( xfs_ilock(ip, XFS_IOLOCK_EXCL); - if (DM_EVENT_ENABLED(ip, DM_EVENT_WRITE) && !(ioflags & IO_INVIS)) { - int iolock = XFS_IOLOCK_EXCL; - int error; - - error = XFS_SEND_DATA(mp, DM_EVENT_WRITE, ip, *ppos, count, - FILP_DELAY_FLAG(outfilp), &iolock); - if (error) { - xfs_iunlock(ip, XFS_IOLOCK_EXCL); - return -error; - } - } - new_size = *ppos + count; xfs_ilock(ip, XFS_ILOCK_EXCL); @@ -463,7 +414,7 @@ xfs_zero_last_block( last_fsb = XFS_B_TO_FSBT(mp, isize); nimaps = 1; error = xfs_bmapi(NULL, ip, last_fsb, 1, 0, NULL, 0, &imap, - &nimaps, NULL, NULL); + &nimaps, NULL); if (error) { return error; } @@ -558,7 +509,7 @@ xfs_zero_eof( nimaps = 1; zero_count_fsb = end_zero_fsb - start_zero_fsb + 1; error = xfs_bmapi(NULL, ip, start_zero_fsb, zero_count_fsb, - 0, NULL, 0, &imap, &nimaps, NULL, NULL); + 0, NULL, 0, &imap, &nimaps, NULL); if (error) { ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL|XFS_IOLOCK_EXCL)); return error; @@ -627,7 +578,6 @@ xfs_file_aio_write( int ioflags = 0; xfs_fsize_t isize, new_size; int iolock; - int eventsent = 0; size_t ocount = 0, count; int need_i_mutex; @@ -673,33 +623,6 @@ start: goto out_unlock_mutex; } - if ((DM_EVENT_ENABLED(ip, DM_EVENT_WRITE) && - !(ioflags & IO_INVIS) && !eventsent)) { - int dmflags = FILP_DELAY_FLAG(file); - - if (need_i_mutex) - dmflags |= DM_FLAGS_IMUX; - - xfs_iunlock(ip, XFS_ILOCK_EXCL); - error = XFS_SEND_DATA(ip->i_mount, DM_EVENT_WRITE, ip, - pos, count, dmflags, &iolock); - if (error) { - goto out_unlock_internal; - } - xfs_ilock(ip, XFS_ILOCK_EXCL); - eventsent = 1; - - /* - * The iolock was dropped and reacquired in XFS_SEND_DATA - * so we have to recheck the size when appending. - * We will only "goto start;" once, since having sent the - * event prevents another call to XFS_SEND_DATA, which is - * what allows the size to change in the first place. - */ - if ((file->f_flags & O_APPEND) && pos != ip->i_size) - goto start; - } - if (ioflags & IO_ISDIRECT) { xfs_buftarg_t *target = XFS_IS_REALTIME_INODE(ip) ? @@ -830,22 +753,6 @@ write_retry: xfs_iunlock(ip, XFS_ILOCK_EXCL); } - if (ret == -ENOSPC && - DM_EVENT_ENABLED(ip, DM_EVENT_NOSPACE) && !(ioflags & IO_INVIS)) { - xfs_iunlock(ip, iolock); - if (need_i_mutex) - mutex_unlock(&inode->i_mutex); - error = XFS_SEND_NAMESP(ip->i_mount, DM_EVENT_NOSPACE, ip, - DM_RIGHT_NULL, ip, DM_RIGHT_NULL, NULL, NULL, - 0, 0, 0); /* Delay flag intentionally unused */ - if (need_i_mutex) - mutex_lock(&inode->i_mutex); - xfs_ilock(ip, iolock); - if (error) - goto out_unlock_internal; - goto start; - } - error = -ret; if (ret <= 0) goto out_unlock_internal; @@ -1014,9 +921,6 @@ const struct file_operations xfs_file_operations = { .open = xfs_file_open, .release = xfs_file_release, .fsync = xfs_file_fsync, -#ifdef HAVE_FOP_OPEN_EXEC - .open_exec = xfs_file_open_exec, -#endif }; const struct file_operations xfs_dir_file_operations = { diff --git a/fs/xfs/linux-2.6/xfs_fs_subr.c b/fs/xfs/linux-2.6/xfs_fs_subr.c index b6918d76bc7..1f279b012f9 100644 --- a/fs/xfs/linux-2.6/xfs_fs_subr.c +++ b/fs/xfs/linux-2.6/xfs_fs_subr.c @@ -21,10 +21,6 @@ #include "xfs_inode.h" #include "xfs_trace.h" -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. diff --git a/fs/xfs/linux-2.6/xfs_fs_subr.h b/fs/xfs/linux-2.6/xfs_fs_subr.h deleted file mode 100644 index 82bb19b2599..00000000000 --- a/fs/xfs/linux-2.6/xfs_fs_subr.h +++ /dev/null @@ -1,25 +0,0 @@ -/* - * Copyright (c) 2000,2002,2005 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_FS_SUBR_H__ -#define __XFS_FS_SUBR_H__ - -extern int fs_noerr(void); -extern int fs_nosys(void); -extern void fs_noval(void); - -#endif /* __XFS_FS_SUBR_H__ */ diff --git a/fs/xfs/linux-2.6/xfs_ioctl.c b/fs/xfs/linux-2.6/xfs_ioctl.c index e59a8106283..237f5ffb2ee 100644 --- a/fs/xfs/linux-2.6/xfs_ioctl.c +++ b/fs/xfs/linux-2.6/xfs_ioctl.c @@ -23,24 +23,15 @@ #include "xfs_trans.h" #include "xfs_sb.h" #include "xfs_ag.h" -#include "xfs_dir2.h" #include "xfs_alloc.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_attr_sf.h" -#include "xfs_dir2_sf.h" #include "xfs_dinode.h" #include "xfs_inode.h" #include "xfs_ioctl.h" -#include "xfs_btree.h" -#include "xfs_ialloc.h" #include "xfs_rtalloc.h" #include "xfs_itable.h" #include "xfs_error.h" -#include "xfs_rw.h" #include "xfs_attr.h" #include "xfs_bmap.h" #include "xfs_buf_item.h" @@ -908,7 +899,7 @@ xfs_ioctl_setattr( struct xfs_dquot *olddquot = NULL; int code; - xfs_itrace_entry(ip); + trace_xfs_ioctl_setattr(ip); if (mp->m_flags & XFS_MOUNT_RDONLY) return XFS_ERROR(EROFS); @@ -1043,8 +1034,7 @@ xfs_ioctl_setattr( } } - xfs_trans_ijoin(tp, ip, lock_flags); - xfs_trans_ihold(tp, ip); + xfs_trans_ijoin(tp, ip); /* * Change file ownership. Must be the owner or privileged. @@ -1116,16 +1106,7 @@ xfs_ioctl_setattr( xfs_qm_dqrele(udqp); xfs_qm_dqrele(gdqp); - if (code) - return code; - - if (DM_EVENT_ENABLED(ip, DM_EVENT_ATTRIBUTE)) { - XFS_SEND_NAMESP(mp, DM_EVENT_ATTRIBUTE, ip, DM_RIGHT_NULL, - NULL, DM_RIGHT_NULL, NULL, NULL, 0, 0, - (mask & FSX_NONBLOCK) ? DM_FLAGS_NDELAY : 0); - } - - return 0; + return code; error_return: xfs_qm_dqrele(udqp); @@ -1301,7 +1282,7 @@ xfs_file_ioctl( if (filp->f_mode & FMODE_NOCMTIME) ioflags |= IO_INVIS; - xfs_itrace_entry(ip); + trace_xfs_file_ioctl(ip); switch (cmd) { case XFS_IOC_ALLOCSP: diff --git a/fs/xfs/linux-2.6/xfs_ioctl32.c b/fs/xfs/linux-2.6/xfs_ioctl32.c index 52ed49e6465..6c83f7f62dc 100644 --- a/fs/xfs/linux-2.6/xfs_ioctl32.c +++ b/fs/xfs/linux-2.6/xfs_ioctl32.c @@ -28,12 +28,8 @@ #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_attr_sf.h" -#include "xfs_dir2_sf.h" #include "xfs_vnode.h" #include "xfs_dinode.h" #include "xfs_inode.h" @@ -544,7 +540,7 @@ xfs_file_compat_ioctl( if (filp->f_mode & FMODE_NOCMTIME) ioflags |= IO_INVIS; - xfs_itrace_entry(ip); + trace_xfs_file_compat_ioctl(ip); switch (cmd) { /* No size or alignment issues on any arch */ diff --git a/fs/xfs/linux-2.6/xfs_iops.c b/fs/xfs/linux-2.6/xfs_iops.c index 44f0b2de153..536b81e63a3 100644 --- a/fs/xfs/linux-2.6/xfs_iops.c +++ b/fs/xfs/linux-2.6/xfs_iops.c @@ -24,21 +24,13 @@ #include "xfs_trans.h" #include "xfs_sb.h" #include "xfs_ag.h" -#include "xfs_dir2.h" #include "xfs_alloc.h" -#include "xfs_dmapi.h" #include "xfs_quota.h" #include "xfs_mount.h" #include "xfs_bmap_btree.h" -#include "xfs_alloc_btree.h" -#include "xfs_ialloc_btree.h" -#include "xfs_dir2_sf.h" -#include "xfs_attr_sf.h" #include "xfs_dinode.h" #include "xfs_inode.h" #include "xfs_bmap.h" -#include "xfs_btree.h" -#include "xfs_ialloc.h" #include "xfs_rtalloc.h" #include "xfs_error.h" #include "xfs_itable.h" @@ -496,7 +488,7 @@ xfs_vn_getattr( struct xfs_inode *ip = XFS_I(inode); struct xfs_mount *mp = ip->i_mount; - xfs_itrace_entry(ip); + trace_xfs_getattr(ip); if (XFS_FORCED_SHUTDOWN(mp)) return XFS_ERROR(EIO); diff --git a/fs/xfs/linux-2.6/xfs_linux.h b/fs/xfs/linux-2.6/xfs_linux.h index facfb323a70..998a9d7fb9c 100644 --- a/fs/xfs/linux-2.6/xfs_linux.h +++ b/fs/xfs/linux-2.6/xfs_linux.h @@ -87,7 +87,6 @@ #include <xfs_aops.h> #include <xfs_super.h> #include <xfs_globals.h> -#include <xfs_fs_subr.h> #include <xfs_buf.h> /* diff --git a/fs/xfs/linux-2.6/xfs_quotaops.c b/fs/xfs/linux-2.6/xfs_quotaops.c index 067cafbfc63..29b9d642e93 100644 --- a/fs/xfs/linux-2.6/xfs_quotaops.c +++ b/fs/xfs/linux-2.6/xfs_quotaops.c @@ -16,7 +16,6 @@ * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA */ #include "xfs.h" -#include "xfs_dmapi.h" #include "xfs_sb.h" #include "xfs_inum.h" #include "xfs_log.h" @@ -69,15 +68,15 @@ xfs_fs_set_xstate( if (op != Q_XQUOTARM && !XFS_IS_QUOTA_RUNNING(mp)) return -ENOSYS; - if (uflags & XFS_QUOTA_UDQ_ACCT) + if (uflags & FS_QUOTA_UDQ_ACCT) flags |= XFS_UQUOTA_ACCT; - if (uflags & XFS_QUOTA_PDQ_ACCT) + if (uflags & FS_QUOTA_PDQ_ACCT) flags |= XFS_PQUOTA_ACCT; - if (uflags & XFS_QUOTA_GDQ_ACCT) + if (uflags & FS_QUOTA_GDQ_ACCT) flags |= XFS_GQUOTA_ACCT; - if (uflags & XFS_QUOTA_UDQ_ENFD) + if (uflags & FS_QUOTA_UDQ_ENFD) flags |= XFS_UQUOTA_ENFD; - if (uflags & (XFS_QUOTA_PDQ_ENFD|XFS_QUOTA_GDQ_ENFD)) + if (uflags & (FS_QUOTA_PDQ_ENFD|FS_QUOTA_GDQ_ENFD)) flags |= XFS_OQUOTA_ENFD; switch (op) { diff --git a/fs/xfs/linux-2.6/xfs_super.c b/fs/xfs/linux-2.6/xfs_super.c index 80938c736c2..758df94690e 100644 --- a/fs/xfs/linux-2.6/xfs_super.c +++ b/fs/xfs/linux-2.6/xfs_super.c @@ -25,14 +25,11 @@ #include "xfs_ag.h" #include "xfs_dir2.h" #include "xfs_alloc.h" -#include "xfs_dmapi.h" #include "xfs_quota.h" #include "xfs_mount.h" #include "xfs_bmap_btree.h" #include "xfs_alloc_btree.h" #include "xfs_ialloc_btree.h" -#include "xfs_dir2_sf.h" -#include "xfs_attr_sf.h" #include "xfs_dinode.h" #include "xfs_inode.h" #include "xfs_btree.h" @@ -43,7 +40,6 @@ #include "xfs_error.h" #include "xfs_itable.h" #include "xfs_fsops.h" -#include "xfs_rw.h" #include "xfs_attr.h" #include "xfs_buf_item.h" #include "xfs_utils.h" @@ -94,7 +90,6 @@ mempool_t *xfs_ioend_pool; #define MNTOPT_BARRIER "barrier" /* use writer barriers for log write and * unwritten extent conversion */ #define MNTOPT_NOBARRIER "nobarrier" /* .. disable */ -#define MNTOPT_OSYNCISOSYNC "osyncisosync" /* o_sync is REALLY o_sync */ #define MNTOPT_64BITINODE "inode64" /* inodes can be allocated anywhere */ #define MNTOPT_IKEEP "ikeep" /* do not free empty inode clusters */ #define MNTOPT_NOIKEEP "noikeep" /* free empty inode clusters */ @@ -116,9 +111,6 @@ mempool_t *xfs_ioend_pool; #define MNTOPT_GQUOTANOENF "gqnoenforce"/* group quota limit enforcement */ #define MNTOPT_PQUOTANOENF "pqnoenforce"/* project quota limit enforcement */ #define MNTOPT_QUOTANOENF "qnoenforce" /* same as uqnoenforce */ -#define MNTOPT_DMAPI "dmapi" /* DMI enabled (DMAPI / XDSM) */ -#define MNTOPT_XDSM "xdsm" /* DMI enabled (DMAPI / XDSM) */ -#define MNTOPT_DMI "dmi" /* DMI enabled (DMAPI / XDSM) */ #define MNTOPT_DELAYLOG "delaylog" /* Delayed loging enabled */ #define MNTOPT_NODELAYLOG "nodelaylog" /* Delayed loging disabled */ @@ -172,15 +164,13 @@ suffix_strtoul(char *s, char **endp, unsigned int base) STATIC int xfs_parseargs( struct xfs_mount *mp, - char *options, - char **mtpt) + char *options) { struct super_block *sb = mp->m_super; char *this_char, *value, *eov; int dsunit = 0; int dswidth = 0; int iosize = 0; - int dmapi_implies_ikeep = 1; __uint8_t iosizelog = 0; /* @@ -243,15 +233,10 @@ xfs_parseargs( if (!mp->m_logname) return ENOMEM; } else if (!strcmp(this_char, MNTOPT_MTPT)) { - if (!value || !*value) { - cmn_err(CE_WARN, - "XFS: %s option requires an argument", - this_char); - return EINVAL; - } - *mtpt = kstrndup(value, MAXNAMELEN, GFP_KERNEL); - if (!*mtpt) - return ENOMEM; + cmn_err(CE_WARN, + "XFS: %s option not allowed on this system", + this_char); + return EINVAL; } else if (!strcmp(this_char, MNTOPT_RTDEV)) { if (!value || !*value) { cmn_err(CE_WARN, @@ -288,8 +273,6 @@ xfs_parseargs( mp->m_flags &= ~XFS_MOUNT_GRPID; } else if (!strcmp(this_char, MNTOPT_WSYNC)) { mp->m_flags |= XFS_MOUNT_WSYNC; - } else if (!strcmp(this_char, MNTOPT_OSYNCISOSYNC)) { - mp->m_flags |= XFS_MOUNT_OSYNCISOSYNC; } else if (!strcmp(this_char, MNTOPT_NORECOVERY)) { mp->m_flags |= XFS_MOUNT_NORECOVERY; } else if (!strcmp(this_char, MNTOPT_NOALIGN)) { @@ -329,7 +312,6 @@ xfs_parseargs( } else if (!strcmp(this_char, MNTOPT_IKEEP)) { mp->m_flags |= XFS_MOUNT_IKEEP; } else if (!strcmp(this_char, MNTOPT_NOIKEEP)) { - dmapi_implies_ikeep = 0; mp->m_flags &= ~XFS_MOUNT_IKEEP; } else if (!strcmp(this_char, MNTOPT_LARGEIO)) { mp->m_flags &= ~XFS_MOUNT_COMPAT_IOSIZE; @@ -370,12 +352,6 @@ xfs_parseargs( } else if (!strcmp(this_char, MNTOPT_GQUOTANOENF)) { mp->m_qflags |= (XFS_GQUOTA_ACCT | XFS_GQUOTA_ACTIVE); mp->m_qflags &= ~XFS_OQUOTA_ENFD; - } else if (!strcmp(this_char, MNTOPT_DMAPI)) { - mp->m_flags |= XFS_MOUNT_DMAPI; - } else if (!strcmp(this_char, MNTOPT_XDSM)) { - mp->m_flags |= XFS_MOUNT_DMAPI; - } else if (!strcmp(this_char, MNTOPT_DMI)) { - mp->m_flags |= XFS_MOUNT_DMAPI; } else if (!strcmp(this_char, MNTOPT_DELAYLOG)) { mp->m_flags |= XFS_MOUNT_DELAYLOG; cmn_err(CE_WARN, @@ -387,9 +363,11 @@ xfs_parseargs( cmn_err(CE_WARN, "XFS: ihashsize no longer used, option is deprecated."); } else if (!strcmp(this_char, "osyncisdsync")) { - /* no-op, this is now the default */ cmn_err(CE_WARN, - "XFS: osyncisdsync is now the default, option is deprecated."); + "XFS: osyncisdsync has no effect, option is deprecated."); + } else if (!strcmp(this_char, "osyncisosync")) { + cmn_err(CE_WARN, + "XFS: osyncisosync has no effect, option is deprecated."); } else if (!strcmp(this_char, "irixsgid")) { cmn_err(CE_WARN, "XFS: irixsgid is now a sysctl(2) variable, option is deprecated."); @@ -430,12 +408,6 @@ xfs_parseargs( return EINVAL; } - 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; - } - if ((dsunit && !dswidth) || (!dsunit && dswidth)) { cmn_err(CE_WARN, "XFS: sunit and swidth must be specified together"); @@ -449,18 +421,6 @@ xfs_parseargs( return EINVAL; } - /* - * Applications using DMI filesystems often expect the - * inode generation number to be monotonically increasing. - * If we delete inode chunks we break this assumption, so - * keep unused inode chunks on disk for DMI filesystems - * until we come up with a better solution. - * Note that if "ikeep" or "noikeep" mount options are - * supplied, then they are honored. - */ - if ((mp->m_flags & XFS_MOUNT_DMAPI) && dmapi_implies_ikeep) - mp->m_flags |= XFS_MOUNT_IKEEP; - done: if (!(mp->m_flags & XFS_MOUNT_NOALIGN)) { /* @@ -539,10 +499,8 @@ xfs_showargs( { XFS_MOUNT_SWALLOC, "," MNTOPT_SWALLOC }, { XFS_MOUNT_NOUUID, "," MNTOPT_NOUUID }, { XFS_MOUNT_NORECOVERY, "," MNTOPT_NORECOVERY }, - { XFS_MOUNT_OSYNCISOSYNC, "," MNTOPT_OSYNCISOSYNC }, { XFS_MOUNT_ATTR2, "," MNTOPT_ATTR2 }, { XFS_MOUNT_FILESTREAMS, "," MNTOPT_FILESTREAM }, - { XFS_MOUNT_DMAPI, "," MNTOPT_DMAPI }, { XFS_MOUNT_GRPID, "," MNTOPT_GRPID }, { XFS_MOUNT_DELAYLOG, "," MNTOPT_DELAYLOG }, { 0, NULL } @@ -947,7 +905,7 @@ xfs_fs_destroy_inode( { struct xfs_inode *ip = XFS_I(inode); - xfs_itrace_entry(ip); + trace_xfs_destroy_inode(ip); XFS_STATS_INC(vn_reclaim); @@ -1063,10 +1021,8 @@ xfs_log_inode( * an inode in another recent transaction. So we play it safe and * fire off the transaction anyway. */ - xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL); - xfs_trans_ihold(tp, ip); + xfs_trans_ijoin(tp, ip); xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); - xfs_trans_set_sync(tp); error = xfs_trans_commit(tp, 0); xfs_ilock_demote(ip, XFS_ILOCK_EXCL); @@ -1082,27 +1038,18 @@ xfs_fs_write_inode( struct xfs_mount *mp = ip->i_mount; int error = EAGAIN; - xfs_itrace_entry(ip); + trace_xfs_write_inode(ip); if (XFS_FORCED_SHUTDOWN(mp)) return XFS_ERROR(EIO); if (wbc->sync_mode == WB_SYNC_ALL) { /* - * Make sure the inode has hit stable storage. By using the - * log and the fsync transactions we reduce the IOs we have - * to do here from two (log and inode) to just the log. - * - * Note: We still need to do a delwri write of the inode after - * this to flush it to the backing buffer so that bulkstat - * works properly if this is the first time the inode has been - * written. Because we hold the ilock atomically over the - * transaction commit and the inode flush we are guaranteed - * that the inode is not pinned when it returns. If the flush - * lock is already held, then the inode has already been - * flushed once and we don't need to flush it again. Hence - * the code will only flush the inode if it isn't already - * being flushed. + * Make sure the inode has made it it into the log. Instead + * of forcing it all the way to stable storage using a + * synchronous transaction we let the log force inside the + * ->sync_fs call do that for thus, which reduces the number + * of synchronous log foces dramatically. */ xfs_ioend_wait(ip); xfs_ilock(ip, XFS_ILOCK_SHARED); @@ -1116,27 +1063,29 @@ xfs_fs_write_inode( * We make this non-blocking if the inode is contended, return * EAGAIN to indicate to the caller that they did not succeed. * This prevents the flush path from blocking on inodes inside - * another operation right now, they get caught later by xfs_sync. + * another operation right now, they get caught later by + * xfs_sync. */ if (!xfs_ilock_nowait(ip, XFS_ILOCK_SHARED)) goto out; - } - if (xfs_ipincount(ip) || !xfs_iflock_nowait(ip)) - goto out_unlock; + if (xfs_ipincount(ip) || !xfs_iflock_nowait(ip)) + goto out_unlock; - /* - * Now we have the flush lock and the inode is not pinned, we can check - * if the inode is really clean as we know that there are no pending - * transaction completions, it is not waiting on the delayed write - * queue and there is no IO in progress. - */ - if (xfs_inode_clean(ip)) { - xfs_ifunlock(ip); - error = 0; - goto out_unlock; + /* + * Now we have the flush lock and the inode is not pinned, we + * can check if the inode is really clean as we know that + * there are no pending transaction completions, it is not + * waiting on the delayed write queue and there is no IO in + * progress. + */ + if (xfs_inode_clean(ip)) { + xfs_ifunlock(ip); + error = 0; + goto out_unlock; + } + error = xfs_iflush(ip, 0); } - error = xfs_iflush(ip, 0); out_unlock: xfs_iunlock(ip, XFS_ILOCK_SHARED); @@ -1156,7 +1105,8 @@ xfs_fs_clear_inode( { xfs_inode_t *ip = XFS_I(inode); - xfs_itrace_entry(ip); + trace_xfs_clear_inode(ip); + XFS_STATS_INC(vn_rele); XFS_STATS_INC(vn_remove); XFS_STATS_DEC(vn_active); @@ -1193,22 +1143,13 @@ xfs_fs_put_super( { struct xfs_mount *mp = XFS_M(sb); + /* + * Unregister the memory shrinker before we tear down the mount + * structure so we don't have memory reclaim racing with us here. + */ + xfs_inode_shrinker_unregister(mp); xfs_syncd_stop(mp); - if (!(sb->s_flags & MS_RDONLY)) { - /* - * XXX(hch): this should be SYNC_WAIT. - * - * Or more likely not needed at all because the VFS is already - * calling ->sync_fs after shutting down all filestem - * operations and just before calling ->put_super. - */ - xfs_sync_data(mp, 0); - xfs_sync_attr(mp, 0); - } - - XFS_SEND_PREUNMOUNT(mp); - /* * Blow away any referenced inode in the filestreams cache. * This can and will cause log traffic as inodes go inactive @@ -1218,14 +1159,10 @@ xfs_fs_put_super( XFS_bflush(mp->m_ddev_targp); - XFS_SEND_UNMOUNT(mp); - xfs_unmountfs(mp); xfs_freesb(mp); - xfs_inode_shrinker_unregister(mp); xfs_icsb_destroy_counters(mp); xfs_close_devices(mp); - xfs_dmops_put(mp); xfs_free_fsname(mp); kfree(mp); } @@ -1543,7 +1480,6 @@ xfs_fs_fill_super( struct inode *root; struct xfs_mount *mp = NULL; int flags = 0, error = ENOMEM; - char *mtpt = NULL; mp = kzalloc(sizeof(struct xfs_mount), GFP_KERNEL); if (!mp) @@ -1559,7 +1495,7 @@ xfs_fs_fill_super( mp->m_super = sb; sb->s_fs_info = mp; - error = xfs_parseargs(mp, (char *)data, &mtpt); + error = xfs_parseargs(mp, (char *)data); if (error) goto out_free_fsname; @@ -1571,16 +1507,12 @@ xfs_fs_fill_super( #endif sb->s_op = &xfs_super_operations; - error = xfs_dmops_get(mp); - if (error) - goto out_free_fsname; - if (silent) flags |= XFS_MFSI_QUIET; error = xfs_open_devices(mp); if (error) - goto out_put_dmops; + goto out_free_fsname; if (xfs_icsb_init_counters(mp)) mp->m_flags |= XFS_MOUNT_NO_PERCPU_SB; @@ -1608,8 +1540,6 @@ xfs_fs_fill_super( if (error) goto out_filestream_unmount; - XFS_SEND_MOUNT(mp, DM_RIGHT_NULL, mtpt, mp->m_fsname); - sb->s_magic = XFS_SB_MAGIC; sb->s_blocksize = mp->m_sb.sb_blocksize; sb->s_blocksize_bits = ffs(sb->s_blocksize) - 1; @@ -1638,7 +1568,6 @@ xfs_fs_fill_super( xfs_inode_shrinker_register(mp); - kfree(mtpt); return 0; out_filestream_unmount: @@ -1648,11 +1577,8 @@ xfs_fs_fill_super( out_destroy_counters: xfs_icsb_destroy_counters(mp); xfs_close_devices(mp); - out_put_dmops: - xfs_dmops_put(mp); out_free_fsname: xfs_free_fsname(mp); - kfree(mtpt); kfree(mp); out: return -error; @@ -1759,6 +1685,12 @@ xfs_init_zones(void) if (!xfs_trans_zone) goto out_destroy_ifork_zone; + xfs_log_item_desc_zone = + kmem_zone_init(sizeof(struct xfs_log_item_desc), + "xfs_log_item_desc"); + if (!xfs_log_item_desc_zone) + goto out_destroy_trans_zone; + /* * The size of the zone allocated buf log item is the maximum * size possible under XFS. This wastes a little bit of memory, @@ -1768,7 +1700,7 @@ xfs_init_zones(void) (((XFS_MAX_BLOCKSIZE / XFS_BLF_CHUNK) / NBWORD) * sizeof(int))), "xfs_buf_item"); if (!xfs_buf_item_zone) - goto out_destroy_trans_zone; + goto out_destroy_log_item_desc_zone; xfs_efd_zone = kmem_zone_init((sizeof(xfs_efd_log_item_t) + ((XFS_EFD_MAX_FAST_EXTENTS - 1) * @@ -1805,6 +1737,8 @@ xfs_init_zones(void) kmem_zone_destroy(xfs_efd_zone); out_destroy_buf_item_zone: kmem_zone_destroy(xfs_buf_item_zone); + out_destroy_log_item_desc_zone: + kmem_zone_destroy(xfs_log_item_desc_zone); out_destroy_trans_zone: kmem_zone_destroy(xfs_trans_zone); out_destroy_ifork_zone: @@ -1835,6 +1769,7 @@ xfs_destroy_zones(void) kmem_zone_destroy(xfs_efi_zone); kmem_zone_destroy(xfs_efd_zone); kmem_zone_destroy(xfs_buf_item_zone); + kmem_zone_destroy(xfs_log_item_desc_zone); kmem_zone_destroy(xfs_trans_zone); kmem_zone_destroy(xfs_ifork_zone); kmem_zone_destroy(xfs_dabuf_zone); diff --git a/fs/xfs/linux-2.6/xfs_super.h b/fs/xfs/linux-2.6/xfs_super.h index 519618e9279..1ef4a4d2d99 100644 --- a/fs/xfs/linux-2.6/xfs_super.h +++ b/fs/xfs/linux-2.6/xfs_super.h @@ -56,12 +56,6 @@ extern void xfs_qm_exit(void); # define XFS_BIGFS_STRING #endif -#ifdef CONFIG_XFS_DMAPI -# define XFS_DMAPI_STRING "dmapi support, " -#else -# define XFS_DMAPI_STRING -#endif - #ifdef DEBUG # define XFS_DBG_STRING "debug" #else @@ -72,7 +66,6 @@ extern void xfs_qm_exit(void); XFS_SECURITY_STRING \ XFS_REALTIME_STRING \ XFS_BIGFS_STRING \ - XFS_DMAPI_STRING \ XFS_DBG_STRING /* DBG must be last */ struct xfs_inode; diff --git a/fs/xfs/linux-2.6/xfs_sync.c b/fs/xfs/linux-2.6/xfs_sync.c index a51a07c3a70..dfcbd98d159 100644 --- a/fs/xfs/linux-2.6/xfs_sync.c +++ b/fs/xfs/linux-2.6/xfs_sync.c @@ -24,25 +24,14 @@ #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 "xfs_quota.h" #include "xfs_trace.h" @@ -319,7 +308,7 @@ xfs_sync_inode_attr( /* * Write out pagecache data for the whole filesystem. */ -int +STATIC int xfs_sync_data( struct xfs_mount *mp, int flags) @@ -340,7 +329,7 @@ xfs_sync_data( /* * Write out inode metadata (attributes) for the whole filesystem. */ -int +STATIC int xfs_sync_attr( struct xfs_mount *mp, int flags) @@ -373,8 +362,7 @@ xfs_commit_dummy_trans( xfs_ilock(ip, XFS_ILOCK_EXCL); - xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL); - xfs_trans_ihold(tp, ip); + xfs_trans_ijoin(tp, ip); xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); error = xfs_trans_commit(tp, 0); xfs_iunlock(ip, XFS_ILOCK_EXCL); @@ -867,7 +855,36 @@ out: reclaim: xfs_ifunlock(ip); xfs_iunlock(ip, XFS_ILOCK_EXCL); - xfs_ireclaim(ip); + + XFS_STATS_INC(xs_ig_reclaims); + /* + * Remove the inode from the per-AG radix tree. + * + * Because radix_tree_delete won't complain even if the item was never + * added to the tree assert that it's been there before to catch + * problems with the inode life time early on. + */ + write_lock(&pag->pag_ici_lock); + if (!radix_tree_delete(&pag->pag_ici_root, + XFS_INO_TO_AGINO(ip->i_mount, ip->i_ino))) + ASSERT(0); + write_unlock(&pag->pag_ici_lock); + + /* + * 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_ilock(ip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL); + xfs_qm_dqdetach(ip); + xfs_iunlock(ip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL); + + xfs_inode_free(ip); return error; } diff --git a/fs/xfs/linux-2.6/xfs_sync.h b/fs/xfs/linux-2.6/xfs_sync.h index e28139aaa4a..fe78726196f 100644 --- a/fs/xfs/linux-2.6/xfs_sync.h +++ b/fs/xfs/linux-2.6/xfs_sync.h @@ -35,9 +35,6 @@ typedef struct xfs_sync_work { int xfs_syncd_init(struct xfs_mount *mp); void xfs_syncd_stop(struct xfs_mount *mp); -int xfs_sync_attr(struct xfs_mount *mp, int flags); -int xfs_sync_data(struct xfs_mount *mp, int flags); - int xfs_quiesce_data(struct xfs_mount *mp); void xfs_quiesce_attr(struct xfs_mount *mp); diff --git a/fs/xfs/linux-2.6/xfs_trace.c b/fs/xfs/linux-2.6/xfs_trace.c index d12be8470cb..88d25d4aa56 100644 --- a/fs/xfs/linux-2.6/xfs_trace.c +++ b/fs/xfs/linux-2.6/xfs_trace.c @@ -24,17 +24,13 @@ #include "xfs_trans.h" #include "xfs_sb.h" #include "xfs_ag.h" -#include "xfs_dir2.h" #include "xfs_da_btree.h" #include "xfs_bmap_btree.h" #include "xfs_alloc_btree.h" #include "xfs_ialloc_btree.h" -#include "xfs_dir2_sf.h" -#include "xfs_attr_sf.h" #include "xfs_dinode.h" #include "xfs_inode.h" #include "xfs_btree.h" -#include "xfs_dmapi.h" #include "xfs_mount.h" #include "xfs_ialloc.h" #include "xfs_itable.h" diff --git a/fs/xfs/linux-2.6/xfs_trace.h b/fs/xfs/linux-2.6/xfs_trace.h index 30282069090..c657cdca2cd 100644 --- a/fs/xfs/linux-2.6/xfs_trace.h +++ b/fs/xfs/linux-2.6/xfs_trace.h @@ -317,8 +317,6 @@ DEFINE_BUF_EVENT(xfs_buf_init); DEFINE_BUF_EVENT(xfs_buf_free); DEFINE_BUF_EVENT(xfs_buf_hold); DEFINE_BUF_EVENT(xfs_buf_rele); -DEFINE_BUF_EVENT(xfs_buf_pin); -DEFINE_BUF_EVENT(xfs_buf_unpin); DEFINE_BUF_EVENT(xfs_buf_iodone); DEFINE_BUF_EVENT(xfs_buf_iorequest); DEFINE_BUF_EVENT(xfs_buf_bawrite); @@ -541,7 +539,7 @@ DEFINE_LOCK_EVENT(xfs_ilock_nowait); DEFINE_LOCK_EVENT(xfs_ilock_demote); DEFINE_LOCK_EVENT(xfs_iunlock); -DECLARE_EVENT_CLASS(xfs_iget_class, +DECLARE_EVENT_CLASS(xfs_inode_class, TP_PROTO(struct xfs_inode *ip), TP_ARGS(ip), TP_STRUCT__entry( @@ -557,16 +555,38 @@ DECLARE_EVENT_CLASS(xfs_iget_class, __entry->ino) ) -#define DEFINE_IGET_EVENT(name) \ -DEFINE_EVENT(xfs_iget_class, name, \ +#define DEFINE_INODE_EVENT(name) \ +DEFINE_EVENT(xfs_inode_class, name, \ TP_PROTO(struct xfs_inode *ip), \ TP_ARGS(ip)) -DEFINE_IGET_EVENT(xfs_iget_skip); -DEFINE_IGET_EVENT(xfs_iget_reclaim); -DEFINE_IGET_EVENT(xfs_iget_found); -DEFINE_IGET_EVENT(xfs_iget_alloc); - -DECLARE_EVENT_CLASS(xfs_inode_class, +DEFINE_INODE_EVENT(xfs_iget_skip); +DEFINE_INODE_EVENT(xfs_iget_reclaim); +DEFINE_INODE_EVENT(xfs_iget_reclaim_fail); +DEFINE_INODE_EVENT(xfs_iget_hit); +DEFINE_INODE_EVENT(xfs_iget_miss); + +DEFINE_INODE_EVENT(xfs_getattr); +DEFINE_INODE_EVENT(xfs_setattr); +DEFINE_INODE_EVENT(xfs_readlink); +DEFINE_INODE_EVENT(xfs_alloc_file_space); +DEFINE_INODE_EVENT(xfs_free_file_space); +DEFINE_INODE_EVENT(xfs_readdir); +#ifdef CONFIG_XFS_POSIX_ACL +DEFINE_INODE_EVENT(xfs_check_acl); +#endif +DEFINE_INODE_EVENT(xfs_vm_bmap); +DEFINE_INODE_EVENT(xfs_file_ioctl); +DEFINE_INODE_EVENT(xfs_file_compat_ioctl); +DEFINE_INODE_EVENT(xfs_ioctl_setattr); +DEFINE_INODE_EVENT(xfs_file_fsync); +DEFINE_INODE_EVENT(xfs_destroy_inode); +DEFINE_INODE_EVENT(xfs_write_inode); +DEFINE_INODE_EVENT(xfs_clear_inode); + +DEFINE_INODE_EVENT(xfs_dquot_dqalloc); +DEFINE_INODE_EVENT(xfs_dquot_dqdetach); + +DECLARE_EVENT_CLASS(xfs_iref_class, TP_PROTO(struct xfs_inode *ip, unsigned long caller_ip), TP_ARGS(ip, caller_ip), TP_STRUCT__entry( @@ -591,20 +611,71 @@ DECLARE_EVENT_CLASS(xfs_inode_class, (char *)__entry->caller_ip) ) -#define DEFINE_INODE_EVENT(name) \ -DEFINE_EVENT(xfs_inode_class, name, \ +#define DEFINE_IREF_EVENT(name) \ +DEFINE_EVENT(xfs_iref_class, name, \ TP_PROTO(struct xfs_inode *ip, unsigned long caller_ip), \ TP_ARGS(ip, caller_ip)) -DEFINE_INODE_EVENT(xfs_ihold); -DEFINE_INODE_EVENT(xfs_irele); -DEFINE_INODE_EVENT(xfs_inode_pin); -DEFINE_INODE_EVENT(xfs_inode_unpin); -DEFINE_INODE_EVENT(xfs_inode_unpin_nowait); +DEFINE_IREF_EVENT(xfs_ihold); +DEFINE_IREF_EVENT(xfs_irele); +DEFINE_IREF_EVENT(xfs_inode_pin); +DEFINE_IREF_EVENT(xfs_inode_unpin); +DEFINE_IREF_EVENT(xfs_inode_unpin_nowait); + +DECLARE_EVENT_CLASS(xfs_namespace_class, + TP_PROTO(struct xfs_inode *dp, struct xfs_name *name), + TP_ARGS(dp, name), + TP_STRUCT__entry( + __field(dev_t, dev) + __field(xfs_ino_t, dp_ino) + __dynamic_array(char, name, name->len) + ), + TP_fast_assign( + __entry->dev = VFS_I(dp)->i_sb->s_dev; + __entry->dp_ino = dp->i_ino; + memcpy(__get_str(name), name->name, name->len); + ), + TP_printk("dev %d:%d dp ino 0x%llx name %s", + MAJOR(__entry->dev), MINOR(__entry->dev), + __entry->dp_ino, + __get_str(name)) +) -/* the old xfs_itrace_entry tracer - to be replaced by s.th. in the VFS */ -DEFINE_INODE_EVENT(xfs_inode); -#define xfs_itrace_entry(ip) \ - trace_xfs_inode(ip, _THIS_IP_) +#define DEFINE_NAMESPACE_EVENT(name) \ +DEFINE_EVENT(xfs_namespace_class, name, \ + TP_PROTO(struct xfs_inode *dp, struct xfs_name *name), \ + TP_ARGS(dp, name)) +DEFINE_NAMESPACE_EVENT(xfs_remove); +DEFINE_NAMESPACE_EVENT(xfs_link); +DEFINE_NAMESPACE_EVENT(xfs_lookup); +DEFINE_NAMESPACE_EVENT(xfs_create); +DEFINE_NAMESPACE_EVENT(xfs_symlink); + +TRACE_EVENT(xfs_rename, + TP_PROTO(struct xfs_inode *src_dp, struct xfs_inode *target_dp, + struct xfs_name *src_name, struct xfs_name *target_name), + TP_ARGS(src_dp, target_dp, src_name, target_name), + TP_STRUCT__entry( + __field(dev_t, dev) + __field(xfs_ino_t, src_dp_ino) + __field(xfs_ino_t, target_dp_ino) + __dynamic_array(char, src_name, src_name->len) + __dynamic_array(char, target_name, target_name->len) + ), + TP_fast_assign( + __entry->dev = VFS_I(src_dp)->i_sb->s_dev; + __entry->src_dp_ino = src_dp->i_ino; + __entry->target_dp_ino = target_dp->i_ino; + memcpy(__get_str(src_name), src_name->name, src_name->len); + memcpy(__get_str(target_name), target_name->name, target_name->len); + ), + TP_printk("dev %d:%d src dp ino 0x%llx target dp ino 0x%llx" + " src name %s target name %s", + MAJOR(__entry->dev), MINOR(__entry->dev), + __entry->src_dp_ino, + __entry->target_dp_ino, + __get_str(src_name), + __get_str(target_name)) +) DECLARE_EVENT_CLASS(xfs_dquot_class, TP_PROTO(struct xfs_dquot *dqp), @@ -684,9 +755,6 @@ DEFINE_DQUOT_EVENT(xfs_dqrele); DEFINE_DQUOT_EVENT(xfs_dqflush); DEFINE_DQUOT_EVENT(xfs_dqflush_force); DEFINE_DQUOT_EVENT(xfs_dqflush_done); -/* not really iget events, but we re-use the format */ -DEFINE_IGET_EVENT(xfs_dquot_dqalloc); -DEFINE_IGET_EVENT(xfs_dquot_dqdetach); DECLARE_EVENT_CLASS(xfs_loggrant_class, TP_PROTO(struct log *log, struct xlog_ticket *tic), @@ -834,33 +902,29 @@ DECLARE_EVENT_CLASS(xfs_page_class, __field(loff_t, size) __field(unsigned long, offset) __field(int, delalloc) - __field(int, unmapped) __field(int, unwritten) ), TP_fast_assign( - int delalloc = -1, unmapped = -1, unwritten = -1; + int delalloc = -1, unwritten = -1; if (page_has_buffers(page)) - xfs_count_page_state(page, &delalloc, - &unmapped, &unwritten); + xfs_count_page_state(page, &delalloc, &unwritten); __entry->dev = inode->i_sb->s_dev; __entry->ino = XFS_I(inode)->i_ino; __entry->pgoff = page_offset(page); __entry->size = i_size_read(inode); __entry->offset = off; __entry->delalloc = delalloc; - __entry->unmapped = unmapped; __entry->unwritten = unwritten; ), TP_printk("dev %d:%d ino 0x%llx pgoff 0x%lx size 0x%llx offset %lx " - "delalloc %d unmapped %d unwritten %d", + "delalloc %d unwritten %d", MAJOR(__entry->dev), MINOR(__entry->dev), __entry->ino, __entry->pgoff, __entry->size, __entry->offset, __entry->delalloc, - __entry->unmapped, __entry->unwritten) ) diff --git a/fs/xfs/quota/xfs_dquot.c b/fs/xfs/quota/xfs_dquot.c index 585e7633dfc..e1a2f6800e0 100644 --- a/fs/xfs/quota/xfs_dquot.c +++ b/fs/xfs/quota/xfs_dquot.c @@ -23,25 +23,15 @@ #include "xfs_trans.h" #include "xfs_sb.h" #include "xfs_ag.h" -#include "xfs_dir2.h" #include "xfs_alloc.h" -#include "xfs_dmapi.h" #include "xfs_quota.h" #include "xfs_mount.h" #include "xfs_bmap_btree.h" -#include "xfs_alloc_btree.h" -#include "xfs_ialloc_btree.h" -#include "xfs_dir2_sf.h" -#include "xfs_attr_sf.h" -#include "xfs_dinode.h" #include "xfs_inode.h" -#include "xfs_btree.h" -#include "xfs_ialloc.h" #include "xfs_bmap.h" #include "xfs_rtalloc.h" #include "xfs_error.h" #include "xfs_itable.h" -#include "xfs_rw.h" #include "xfs_attr.h" #include "xfs_buf_item.h" #include "xfs_trans_space.h" @@ -64,8 +54,6 @@ flush lock - ditto. */ -STATIC void xfs_qm_dqflush_done(xfs_buf_t *, xfs_dq_logitem_t *); - #ifdef DEBUG xfs_buftarg_t *xfs_dqerror_target; int xfs_do_dqerror; @@ -390,21 +378,14 @@ xfs_qm_dqalloc( return (ESRCH); } - /* - * xfs_trans_commit normally decrements the vnode ref count - * when it unlocks the inode. Since we want to keep the quota - * inode around, we bump the vnode ref count now. - */ - IHOLD(quotip); - - xfs_trans_ijoin(tp, quotip, XFS_ILOCK_EXCL); + xfs_trans_ijoin_ref(tp, quotip, XFS_ILOCK_EXCL); nmaps = 1; if ((error = xfs_bmapi(tp, quotip, offset_fsb, XFS_DQUOT_CLUSTER_SIZE_FSB, XFS_BMAPI_METADATA | XFS_BMAPI_WRITE, &firstblock, XFS_QM_DQALLOC_SPACE_RES(mp), - &map, &nmaps, &flist, NULL))) { + &map, &nmaps, &flist))) { goto error0; } ASSERT(map.br_blockcount == XFS_DQUOT_CLUSTER_SIZE_FSB); @@ -520,7 +501,7 @@ xfs_qm_dqtobp( error = xfs_bmapi(NULL, quotip, dqp->q_fileoffset, XFS_DQUOT_CLUSTER_SIZE_FSB, XFS_BMAPI_METADATA, - NULL, 0, &map, &nmaps, NULL, NULL); + NULL, 0, &map, &nmaps, NULL); xfs_iunlock(quotip, XFS_ILOCK_SHARED); if (error) @@ -1141,6 +1122,46 @@ xfs_qm_dqrele( xfs_qm_dqput(dqp); } +/* + * This is the dquot flushing I/O completion routine. It is called + * from interrupt level when the buffer containing the dquot is + * flushed to disk. It is responsible for removing the dquot logitem + * from the AIL if it has not been re-logged, and unlocking the dquot's + * flush lock. This behavior is very similar to that of inodes.. + */ +STATIC void +xfs_qm_dqflush_done( + struct xfs_buf *bp, + struct xfs_log_item *lip) +{ + xfs_dq_logitem_t *qip = (struct xfs_dq_logitem *)lip; + xfs_dquot_t *dqp = qip->qli_dquot; + struct xfs_ail *ailp = lip->li_ailp; + + /* + * We only want to pull the item from the AIL if its + * location in the log has not changed since we started the flush. + * Thus, we only bother if the dquot's lsn has + * not changed. First we check the lsn outside the lock + * since it's cheaper, and then we recheck while + * holding the lock before removing the dquot from the AIL. + */ + if ((lip->li_flags & XFS_LI_IN_AIL) && + lip->li_lsn == qip->qli_flush_lsn) { + + /* xfs_trans_ail_delete() drops the AIL lock. */ + spin_lock(&ailp->xa_lock); + if (lip->li_lsn == qip->qli_flush_lsn) + xfs_trans_ail_delete(ailp, lip); + else + spin_unlock(&ailp->xa_lock); + } + + /* + * Release the dq's flush lock since we're done with it. + */ + xfs_dqfunlock(dqp); +} /* * Write a modified dquot to disk. @@ -1222,8 +1243,9 @@ xfs_qm_dqflush( * Attach an iodone routine so that we can remove this dquot from the * AIL and release the flush lock once the dquot is synced to disk. */ - xfs_buf_attach_iodone(bp, (void(*)(xfs_buf_t *, xfs_log_item_t *)) - xfs_qm_dqflush_done, &(dqp->q_logitem.qli_item)); + xfs_buf_attach_iodone(bp, xfs_qm_dqflush_done, + &dqp->q_logitem.qli_item); + /* * If the buffer is pinned then push on the log so we won't * get stuck waiting in the write for too long. @@ -1247,50 +1269,6 @@ xfs_qm_dqflush( } -/* - * This is the dquot flushing I/O completion routine. It is called - * from interrupt level when the buffer containing the dquot is - * flushed to disk. It is responsible for removing the dquot logitem - * from the AIL if it has not been re-logged, and unlocking the dquot's - * flush lock. This behavior is very similar to that of inodes.. - */ -/*ARGSUSED*/ -STATIC void -xfs_qm_dqflush_done( - xfs_buf_t *bp, - 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 - * location in the log has not changed since we started the flush. - * Thus, we only bother if the dquot's lsn has - * not changed. First we check the lsn outside the lock - * since it's cheaper, and then we recheck while - * holding the lock before removing the dquot from the AIL. - */ - if ((qip->qli_item.li_flags & XFS_LI_IN_AIL) && - qip->qli_item.li_lsn == qip->qli_flush_lsn) { - - /* 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_ail_delete(ailp, (xfs_log_item_t*)qip); - else - spin_unlock(&ailp->xa_lock); - } - - /* - * Release the dq's flush lock since we're done with it. - */ - xfs_dqfunlock(dqp); -} - int xfs_qm_dqlock_nowait( xfs_dquot_t *dqp) diff --git a/fs/xfs/quota/xfs_dquot_item.c b/fs/xfs/quota/xfs_dquot_item.c index 8d89a24ae32..2a1f3dc10a0 100644 --- a/fs/xfs/quota/xfs_dquot_item.c +++ b/fs/xfs/quota/xfs_dquot_item.c @@ -23,42 +23,36 @@ #include "xfs_trans.h" #include "xfs_sb.h" #include "xfs_ag.h" -#include "xfs_dir2.h" #include "xfs_alloc.h" -#include "xfs_dmapi.h" #include "xfs_quota.h" #include "xfs_mount.h" #include "xfs_bmap_btree.h" -#include "xfs_alloc_btree.h" -#include "xfs_ialloc_btree.h" -#include "xfs_dir2_sf.h" -#include "xfs_attr_sf.h" -#include "xfs_dinode.h" #include "xfs_inode.h" #include "xfs_bmap.h" -#include "xfs_btree.h" -#include "xfs_ialloc.h" #include "xfs_rtalloc.h" #include "xfs_error.h" #include "xfs_itable.h" -#include "xfs_rw.h" #include "xfs_attr.h" #include "xfs_buf_item.h" #include "xfs_trans_priv.h" #include "xfs_qm.h" +static inline struct xfs_dq_logitem *DQUOT_ITEM(struct xfs_log_item *lip) +{ + return container_of(lip, struct xfs_dq_logitem, qli_item); +} + /* * returns the number of iovecs needed to log the given dquot item. */ -/* ARGSUSED */ STATIC uint xfs_qm_dquot_logitem_size( - xfs_dq_logitem_t *logitem) + struct xfs_log_item *lip) { /* * we need only two iovecs, one for the format, one for the real thing */ - return (2); + return 2; } /* @@ -66,22 +60,21 @@ xfs_qm_dquot_logitem_size( */ STATIC void xfs_qm_dquot_logitem_format( - xfs_dq_logitem_t *logitem, - xfs_log_iovec_t *logvec) + struct xfs_log_item *lip, + struct xfs_log_iovec *logvec) { - ASSERT(logitem); - ASSERT(logitem->qli_dquot); + struct xfs_dq_logitem *qlip = DQUOT_ITEM(lip); - logvec->i_addr = (xfs_caddr_t)&logitem->qli_format; + logvec->i_addr = &qlip->qli_format; logvec->i_len = sizeof(xfs_dq_logformat_t); logvec->i_type = XLOG_REG_TYPE_QFORMAT; logvec++; - logvec->i_addr = (xfs_caddr_t)&logitem->qli_dquot->q_core; + logvec->i_addr = &qlip->qli_dquot->q_core; logvec->i_len = sizeof(xfs_disk_dquot_t); logvec->i_type = XLOG_REG_TYPE_DQUOT; - ASSERT(2 == logitem->qli_item.li_desc->lid_size); - logitem->qli_format.qlf_size = 2; + ASSERT(2 == lip->li_desc->lid_size); + qlip->qli_format.qlf_size = 2; } @@ -90,9 +83,9 @@ xfs_qm_dquot_logitem_format( */ STATIC void xfs_qm_dquot_logitem_pin( - xfs_dq_logitem_t *logitem) + struct xfs_log_item *lip) { - xfs_dquot_t *dqp = logitem->qli_dquot; + struct xfs_dquot *dqp = DQUOT_ITEM(lip)->qli_dquot; ASSERT(XFS_DQ_IS_LOCKED(dqp)); atomic_inc(&dqp->q_pincount); @@ -104,27 +97,18 @@ xfs_qm_dquot_logitem_pin( * dquot must have been previously pinned with a call to * xfs_qm_dquot_logitem_pin(). */ -/* ARGSUSED */ STATIC void xfs_qm_dquot_logitem_unpin( - xfs_dq_logitem_t *logitem) + struct xfs_log_item *lip, + int remove) { - xfs_dquot_t *dqp = logitem->qli_dquot; + struct xfs_dquot *dqp = DQUOT_ITEM(lip)->qli_dquot; ASSERT(atomic_read(&dqp->q_pincount) > 0); if (atomic_dec_and_test(&dqp->q_pincount)) wake_up(&dqp->q_pinwait); } -/* ARGSUSED */ -STATIC void -xfs_qm_dquot_logitem_unpin_remove( - xfs_dq_logitem_t *logitem, - xfs_trans_t *tp) -{ - xfs_qm_dquot_logitem_unpin(logitem); -} - /* * Given the logitem, this writes the corresponding dquot entry to disk * asynchronously. This is called with the dquot entry securely locked; @@ -133,12 +117,10 @@ xfs_qm_dquot_logitem_unpin_remove( */ STATIC void xfs_qm_dquot_logitem_push( - xfs_dq_logitem_t *logitem) + struct xfs_log_item *lip) { - xfs_dquot_t *dqp; - int error; - - dqp = logitem->qli_dquot; + struct xfs_dquot *dqp = DQUOT_ITEM(lip)->qli_dquot; + int error; ASSERT(XFS_DQ_IS_LOCKED(dqp)); ASSERT(!completion_done(&dqp->q_flush)); @@ -160,27 +142,25 @@ xfs_qm_dquot_logitem_push( xfs_dqunlock(dqp); } -/*ARGSUSED*/ STATIC xfs_lsn_t xfs_qm_dquot_logitem_committed( - xfs_dq_logitem_t *l, + struct xfs_log_item *lip, xfs_lsn_t lsn) { /* * We always re-log the entire dquot when it becomes dirty, * so, the latest copy _is_ the only one that matters. */ - return (lsn); + return lsn; } - /* * This is called to wait for the given dquot to be unpinned. * Most of these pin/unpin routines are plagiarized from inode code. */ void xfs_qm_dqunpin_wait( - xfs_dquot_t *dqp) + struct xfs_dquot *dqp) { ASSERT(XFS_DQ_IS_LOCKED(dqp)); if (atomic_read(&dqp->q_pincount) == 0) @@ -206,13 +186,12 @@ xfs_qm_dqunpin_wait( */ STATIC void xfs_qm_dquot_logitem_pushbuf( - xfs_dq_logitem_t *qip) + struct xfs_log_item *lip) { - xfs_dquot_t *dqp; - xfs_mount_t *mp; - xfs_buf_t *bp; + struct xfs_dq_logitem *qlip = DQUOT_ITEM(lip); + struct xfs_dquot *dqp = qlip->qli_dquot; + struct xfs_buf *bp; - dqp = qip->qli_dquot; ASSERT(XFS_DQ_IS_LOCKED(dqp)); /* @@ -220,22 +199,20 @@ xfs_qm_dquot_logitem_pushbuf( * inode flush completed and the inode was taken off the AIL. * So, just get out. */ - if (completion_done(&dqp->q_flush) || - ((qip->qli_item.li_flags & XFS_LI_IN_AIL) == 0)) { + if (completion_done(&dqp->q_flush) || + !(lip->li_flags & XFS_LI_IN_AIL)) { xfs_dqunlock(dqp); return; } - mp = dqp->q_mount; - bp = xfs_incore(mp->m_ddev_targp, qip->qli_format.qlf_blkno, - mp->m_quotainfo->qi_dqchunklen, XBF_TRYLOCK); + + bp = xfs_incore(dqp->q_mount->m_ddev_targp, qlip->qli_format.qlf_blkno, + dqp->q_mount->m_quotainfo->qi_dqchunklen, XBF_TRYLOCK); xfs_dqunlock(dqp); if (!bp) return; if (XFS_BUF_ISDELAYWRITE(bp)) xfs_buf_delwri_promote(bp); xfs_buf_relse(bp); - return; - } /* @@ -250,15 +227,14 @@ xfs_qm_dquot_logitem_pushbuf( */ STATIC uint xfs_qm_dquot_logitem_trylock( - xfs_dq_logitem_t *qip) + struct xfs_log_item *lip) { - xfs_dquot_t *dqp; + struct xfs_dquot *dqp = DQUOT_ITEM(lip)->qli_dquot; - dqp = qip->qli_dquot; if (atomic_read(&dqp->q_pincount) > 0) return XFS_ITEM_PINNED; - if (! xfs_qm_dqlock_nowait(dqp)) + if (!xfs_qm_dqlock_nowait(dqp)) return XFS_ITEM_LOCKED; if (!xfs_dqflock_nowait(dqp)) { @@ -269,11 +245,10 @@ xfs_qm_dquot_logitem_trylock( return XFS_ITEM_PUSHBUF; } - ASSERT(qip->qli_item.li_flags & XFS_LI_IN_AIL); + ASSERT(lip->li_flags & XFS_LI_IN_AIL); return XFS_ITEM_SUCCESS; } - /* * Unlock the dquot associated with the log item. * Clear the fields of the dquot and dquot log item that @@ -282,12 +257,10 @@ xfs_qm_dquot_logitem_trylock( */ STATIC void xfs_qm_dquot_logitem_unlock( - xfs_dq_logitem_t *ql) + struct xfs_log_item *lip) { - xfs_dquot_t *dqp; + struct xfs_dquot *dqp = DQUOT_ITEM(lip)->qli_dquot; - ASSERT(ql != NULL); - dqp = ql->qli_dquot; ASSERT(XFS_DQ_IS_LOCKED(dqp)); /* @@ -304,43 +277,32 @@ xfs_qm_dquot_logitem_unlock( xfs_dqunlock(dqp); } - /* * this needs to stamp an lsn into the dquot, I think. * rpc's that look at user dquot's would then have to * push on the dependency recorded in the dquot */ -/* ARGSUSED */ STATIC void xfs_qm_dquot_logitem_committing( - xfs_dq_logitem_t *l, + struct xfs_log_item *lip, xfs_lsn_t lsn) { - return; } - /* * This is the ops vector for dquots */ static struct xfs_item_ops xfs_dquot_item_ops = { - .iop_size = (uint(*)(xfs_log_item_t*))xfs_qm_dquot_logitem_size, - .iop_format = (void(*)(xfs_log_item_t*, xfs_log_iovec_t*)) - xfs_qm_dquot_logitem_format, - .iop_pin = (void(*)(xfs_log_item_t*))xfs_qm_dquot_logitem_pin, - .iop_unpin = (void(*)(xfs_log_item_t*))xfs_qm_dquot_logitem_unpin, - .iop_unpin_remove = (void(*)(xfs_log_item_t*, xfs_trans_t*)) - xfs_qm_dquot_logitem_unpin_remove, - .iop_trylock = (uint(*)(xfs_log_item_t*)) - xfs_qm_dquot_logitem_trylock, - .iop_unlock = (void(*)(xfs_log_item_t*))xfs_qm_dquot_logitem_unlock, - .iop_committed = (xfs_lsn_t(*)(xfs_log_item_t*, xfs_lsn_t)) - xfs_qm_dquot_logitem_committed, - .iop_push = (void(*)(xfs_log_item_t*))xfs_qm_dquot_logitem_push, - .iop_pushbuf = (void(*)(xfs_log_item_t*)) - xfs_qm_dquot_logitem_pushbuf, - .iop_committing = (void(*)(xfs_log_item_t*, xfs_lsn_t)) - xfs_qm_dquot_logitem_committing + .iop_size = xfs_qm_dquot_logitem_size, + .iop_format = xfs_qm_dquot_logitem_format, + .iop_pin = xfs_qm_dquot_logitem_pin, + .iop_unpin = xfs_qm_dquot_logitem_unpin, + .iop_trylock = xfs_qm_dquot_logitem_trylock, + .iop_unlock = xfs_qm_dquot_logitem_unlock, + .iop_committed = xfs_qm_dquot_logitem_committed, + .iop_push = xfs_qm_dquot_logitem_push, + .iop_pushbuf = xfs_qm_dquot_logitem_pushbuf, + .iop_committing = xfs_qm_dquot_logitem_committing }; /* @@ -350,10 +312,9 @@ static struct xfs_item_ops xfs_dquot_item_ops = { */ void xfs_qm_dquot_logitem_init( - struct xfs_dquot *dqp) + struct xfs_dquot *dqp) { - xfs_dq_logitem_t *lp; - lp = &dqp->q_logitem; + struct xfs_dq_logitem *lp = &dqp->q_logitem; xfs_log_item_init(dqp->q_mount, &lp->qli_item, XFS_LI_DQUOT, &xfs_dquot_item_ops); @@ -374,16 +335,22 @@ xfs_qm_dquot_logitem_init( /*------------------ QUOTAOFF LOG ITEMS -------------------*/ +static inline struct xfs_qoff_logitem *QOFF_ITEM(struct xfs_log_item *lip) +{ + return container_of(lip, struct xfs_qoff_logitem, qql_item); +} + + /* * This returns the number of iovecs needed to log the given quotaoff item. * We only need 1 iovec for an quotaoff item. It just logs the * quotaoff_log_format structure. */ -/*ARGSUSED*/ STATIC uint -xfs_qm_qoff_logitem_size(xfs_qoff_logitem_t *qf) +xfs_qm_qoff_logitem_size( + struct xfs_log_item *lip) { - return (1); + return 1; } /* @@ -394,53 +361,46 @@ xfs_qm_qoff_logitem_size(xfs_qoff_logitem_t *qf) * slots in the quotaoff item have been filled. */ STATIC void -xfs_qm_qoff_logitem_format(xfs_qoff_logitem_t *qf, - xfs_log_iovec_t *log_vector) +xfs_qm_qoff_logitem_format( + struct xfs_log_item *lip, + struct xfs_log_iovec *log_vector) { - ASSERT(qf->qql_format.qf_type == XFS_LI_QUOTAOFF); + struct xfs_qoff_logitem *qflip = QOFF_ITEM(lip); + + ASSERT(qflip->qql_format.qf_type == XFS_LI_QUOTAOFF); - log_vector->i_addr = (xfs_caddr_t)&(qf->qql_format); + log_vector->i_addr = &qflip->qql_format; log_vector->i_len = sizeof(xfs_qoff_logitem_t); log_vector->i_type = XLOG_REG_TYPE_QUOTAOFF; - qf->qql_format.qf_size = 1; + qflip->qql_format.qf_size = 1; } - /* * Pinning has no meaning for an quotaoff item, so just return. */ -/*ARGSUSED*/ STATIC void -xfs_qm_qoff_logitem_pin(xfs_qoff_logitem_t *qf) +xfs_qm_qoff_logitem_pin( + struct xfs_log_item *lip) { - return; } - /* * Since pinning has no meaning for an quotaoff item, unpinning does * not either. */ -/*ARGSUSED*/ STATIC void -xfs_qm_qoff_logitem_unpin(xfs_qoff_logitem_t *qf) +xfs_qm_qoff_logitem_unpin( + struct xfs_log_item *lip, + int remove) { - return; -} - -/*ARGSUSED*/ -STATIC void -xfs_qm_qoff_logitem_unpin_remove(xfs_qoff_logitem_t *qf, xfs_trans_t *tp) -{ - return; } /* * Quotaoff items have no locking, so just return success. */ -/*ARGSUSED*/ STATIC uint -xfs_qm_qoff_logitem_trylock(xfs_qoff_logitem_t *qf) +xfs_qm_qoff_logitem_trylock( + struct xfs_log_item *lip) { return XFS_ITEM_LOCKED; } @@ -449,53 +409,51 @@ xfs_qm_qoff_logitem_trylock(xfs_qoff_logitem_t *qf) * Quotaoff items have no locking or pushing, so return failure * so that the caller doesn't bother with us. */ -/*ARGSUSED*/ STATIC void -xfs_qm_qoff_logitem_unlock(xfs_qoff_logitem_t *qf) +xfs_qm_qoff_logitem_unlock( + struct xfs_log_item *lip) { - return; } /* * The quotaoff-start-item is logged only once and cannot be moved in the log, * so simply return the lsn at which it's been logged. */ -/*ARGSUSED*/ STATIC xfs_lsn_t -xfs_qm_qoff_logitem_committed(xfs_qoff_logitem_t *qf, xfs_lsn_t lsn) +xfs_qm_qoff_logitem_committed( + struct xfs_log_item *lip, + xfs_lsn_t lsn) { - return (lsn); + return lsn; } /* * There isn't much you can do to push on an quotaoff item. It is simply * stuck waiting for the log to be flushed to disk. */ -/*ARGSUSED*/ STATIC void -xfs_qm_qoff_logitem_push(xfs_qoff_logitem_t *qf) +xfs_qm_qoff_logitem_push( + struct xfs_log_item *lip) { - return; } -/*ARGSUSED*/ STATIC xfs_lsn_t xfs_qm_qoffend_logitem_committed( - xfs_qoff_logitem_t *qfe, - xfs_lsn_t lsn) + struct xfs_log_item *lip, + xfs_lsn_t lsn) { - xfs_qoff_logitem_t *qfs; - struct xfs_ail *ailp; + struct xfs_qoff_logitem *qfe = QOFF_ITEM(lip); + struct xfs_qoff_logitem *qfs = qfe->qql_start_lip; + struct xfs_ail *ailp = qfs->qql_item.li_ailp; - qfs = qfe->qql_start_lip; - ailp = qfs->qql_item.li_ailp; - spin_lock(&ailp->xa_lock); /* * Delete the qoff-start logitem from the AIL. * xfs_trans_ail_delete() drops the AIL lock. */ + spin_lock(&ailp->xa_lock); xfs_trans_ail_delete(ailp, (xfs_log_item_t *)qfs); + kmem_free(qfs); kmem_free(qfe); return (xfs_lsn_t)-1; @@ -515,71 +473,52 @@ xfs_qm_qoffend_logitem_committed( * (truly makes the quotaoff irrevocable). If we do something else, * then maybe we don't need two. */ -/* ARGSUSED */ -STATIC void -xfs_qm_qoff_logitem_committing(xfs_qoff_logitem_t *qip, xfs_lsn_t commit_lsn) -{ - return; -} - -/* ARGSUSED */ STATIC void -xfs_qm_qoffend_logitem_committing(xfs_qoff_logitem_t *qip, xfs_lsn_t commit_lsn) +xfs_qm_qoff_logitem_committing( + struct xfs_log_item *lip, + xfs_lsn_t commit_lsn) { - return; } static struct xfs_item_ops xfs_qm_qoffend_logitem_ops = { - .iop_size = (uint(*)(xfs_log_item_t*))xfs_qm_qoff_logitem_size, - .iop_format = (void(*)(xfs_log_item_t*, xfs_log_iovec_t*)) - xfs_qm_qoff_logitem_format, - .iop_pin = (void(*)(xfs_log_item_t*))xfs_qm_qoff_logitem_pin, - .iop_unpin = (void(*)(xfs_log_item_t*))xfs_qm_qoff_logitem_unpin, - .iop_unpin_remove = (void(*)(xfs_log_item_t*,xfs_trans_t*)) - xfs_qm_qoff_logitem_unpin_remove, - .iop_trylock = (uint(*)(xfs_log_item_t*))xfs_qm_qoff_logitem_trylock, - .iop_unlock = (void(*)(xfs_log_item_t*))xfs_qm_qoff_logitem_unlock, - .iop_committed = (xfs_lsn_t(*)(xfs_log_item_t*, xfs_lsn_t)) - xfs_qm_qoffend_logitem_committed, - .iop_push = (void(*)(xfs_log_item_t*))xfs_qm_qoff_logitem_push, - .iop_pushbuf = NULL, - .iop_committing = (void(*)(xfs_log_item_t*, xfs_lsn_t)) - xfs_qm_qoffend_logitem_committing + .iop_size = xfs_qm_qoff_logitem_size, + .iop_format = xfs_qm_qoff_logitem_format, + .iop_pin = xfs_qm_qoff_logitem_pin, + .iop_unpin = xfs_qm_qoff_logitem_unpin, + .iop_trylock = xfs_qm_qoff_logitem_trylock, + .iop_unlock = xfs_qm_qoff_logitem_unlock, + .iop_committed = xfs_qm_qoffend_logitem_committed, + .iop_push = xfs_qm_qoff_logitem_push, + .iop_committing = xfs_qm_qoff_logitem_committing }; /* * This is the ops vector shared by all quotaoff-start log items. */ static struct xfs_item_ops xfs_qm_qoff_logitem_ops = { - .iop_size = (uint(*)(xfs_log_item_t*))xfs_qm_qoff_logitem_size, - .iop_format = (void(*)(xfs_log_item_t*, xfs_log_iovec_t*)) - xfs_qm_qoff_logitem_format, - .iop_pin = (void(*)(xfs_log_item_t*))xfs_qm_qoff_logitem_pin, - .iop_unpin = (void(*)(xfs_log_item_t*))xfs_qm_qoff_logitem_unpin, - .iop_unpin_remove = (void(*)(xfs_log_item_t*,xfs_trans_t*)) - xfs_qm_qoff_logitem_unpin_remove, - .iop_trylock = (uint(*)(xfs_log_item_t*))xfs_qm_qoff_logitem_trylock, - .iop_unlock = (void(*)(xfs_log_item_t*))xfs_qm_qoff_logitem_unlock, - .iop_committed = (xfs_lsn_t(*)(xfs_log_item_t*, xfs_lsn_t)) - xfs_qm_qoff_logitem_committed, - .iop_push = (void(*)(xfs_log_item_t*))xfs_qm_qoff_logitem_push, - .iop_pushbuf = NULL, - .iop_committing = (void(*)(xfs_log_item_t*, xfs_lsn_t)) - xfs_qm_qoff_logitem_committing + .iop_size = xfs_qm_qoff_logitem_size, + .iop_format = xfs_qm_qoff_logitem_format, + .iop_pin = xfs_qm_qoff_logitem_pin, + .iop_unpin = xfs_qm_qoff_logitem_unpin, + .iop_trylock = xfs_qm_qoff_logitem_trylock, + .iop_unlock = xfs_qm_qoff_logitem_unlock, + .iop_committed = xfs_qm_qoff_logitem_committed, + .iop_push = xfs_qm_qoff_logitem_push, + .iop_committing = xfs_qm_qoff_logitem_committing }; /* * Allocate and initialize an quotaoff item of the correct quota type(s). */ -xfs_qoff_logitem_t * +struct xfs_qoff_logitem * xfs_qm_qoff_logitem_init( - struct xfs_mount *mp, - xfs_qoff_logitem_t *start, - uint flags) + struct xfs_mount *mp, + struct xfs_qoff_logitem *start, + uint flags) { - xfs_qoff_logitem_t *qf; + struct xfs_qoff_logitem *qf; - qf = (xfs_qoff_logitem_t*) kmem_zalloc(sizeof(xfs_qoff_logitem_t), KM_SLEEP); + qf = kmem_zalloc(sizeof(struct xfs_qoff_logitem), KM_SLEEP); xfs_log_item_init(mp, &qf->qql_item, XFS_LI_QUOTAOFF, start ? &xfs_qm_qoffend_logitem_ops : &xfs_qm_qoff_logitem_ops); @@ -587,5 +526,5 @@ xfs_qm_qoff_logitem_init( qf->qql_format.qf_type = XFS_LI_QUOTAOFF; qf->qql_format.qf_flags = flags; qf->qql_start_lip = start; - return (qf); + return qf; } diff --git a/fs/xfs/quota/xfs_qm.c b/fs/xfs/quota/xfs_qm.c index 67c018392d6..9a92407109a 100644 --- a/fs/xfs/quota/xfs_qm.c +++ b/fs/xfs/quota/xfs_qm.c @@ -23,25 +23,18 @@ #include "xfs_trans.h" #include "xfs_sb.h" #include "xfs_ag.h" -#include "xfs_dir2.h" #include "xfs_alloc.h" -#include "xfs_dmapi.h" #include "xfs_quota.h" #include "xfs_mount.h" #include "xfs_bmap_btree.h" -#include "xfs_alloc_btree.h" #include "xfs_ialloc_btree.h" -#include "xfs_dir2_sf.h" -#include "xfs_attr_sf.h" #include "xfs_dinode.h" #include "xfs_inode.h" -#include "xfs_btree.h" #include "xfs_ialloc.h" #include "xfs_itable.h" #include "xfs_rtalloc.h" #include "xfs_error.h" #include "xfs_bmap.h" -#include "xfs_rw.h" #include "xfs_attr.h" #include "xfs_buf_item.h" #include "xfs_trans_space.h" @@ -1497,7 +1490,7 @@ xfs_qm_dqiterate( maxlblkcnt - lblkno, XFS_BMAPI_METADATA, NULL, - 0, map, &nmaps, NULL, NULL); + 0, map, &nmaps, NULL); xfs_iunlock(qip, XFS_ILOCK_SHARED); if (error) break; @@ -1669,7 +1662,8 @@ xfs_qm_dqusage_adjust( * making us disable quotas for the file system. */ if ((error = xfs_qm_dqget_noattach(ip, &udqp, &gdqp))) { - xfs_iput(ip, XFS_ILOCK_EXCL); + xfs_iunlock(ip, XFS_ILOCK_EXCL); + IRELE(ip); *res = BULKSTAT_RV_GIVEUP; return error; } @@ -1682,7 +1676,8 @@ xfs_qm_dqusage_adjust( * Walk thru the extent list and count the realtime blocks. */ if ((error = xfs_qm_get_rtblks(ip, &rtblks))) { - xfs_iput(ip, XFS_ILOCK_EXCL); + xfs_iunlock(ip, XFS_ILOCK_EXCL); + IRELE(ip); if (udqp) xfs_qm_dqput(udqp); if (gdqp) diff --git a/fs/xfs/quota/xfs_qm_bhv.c b/fs/xfs/quota/xfs_qm_bhv.c index 97b410c1279..bea02d786c5 100644 --- a/fs/xfs/quota/xfs_qm_bhv.c +++ b/fs/xfs/quota/xfs_qm_bhv.c @@ -23,25 +23,15 @@ #include "xfs_trans.h" #include "xfs_sb.h" #include "xfs_ag.h" -#include "xfs_dir2.h" #include "xfs_alloc.h" -#include "xfs_dmapi.h" #include "xfs_quota.h" #include "xfs_mount.h" #include "xfs_bmap_btree.h" -#include "xfs_alloc_btree.h" -#include "xfs_ialloc_btree.h" -#include "xfs_dir2_sf.h" -#include "xfs_attr_sf.h" -#include "xfs_dinode.h" #include "xfs_inode.h" -#include "xfs_ialloc.h" #include "xfs_itable.h" -#include "xfs_btree.h" #include "xfs_bmap.h" #include "xfs_rtalloc.h" #include "xfs_error.h" -#include "xfs_rw.h" #include "xfs_attr.h" #include "xfs_buf_item.h" #include "xfs_qm.h" diff --git a/fs/xfs/quota/xfs_qm_stats.c b/fs/xfs/quota/xfs_qm_stats.c index 3d1fc79532e..8671a0b3264 100644 --- a/fs/xfs/quota/xfs_qm_stats.c +++ b/fs/xfs/quota/xfs_qm_stats.c @@ -23,25 +23,15 @@ #include "xfs_trans.h" #include "xfs_sb.h" #include "xfs_ag.h" -#include "xfs_dir2.h" #include "xfs_alloc.h" -#include "xfs_dmapi.h" #include "xfs_quota.h" #include "xfs_mount.h" #include "xfs_bmap_btree.h" -#include "xfs_alloc_btree.h" -#include "xfs_ialloc_btree.h" -#include "xfs_dir2_sf.h" -#include "xfs_attr_sf.h" -#include "xfs_dinode.h" #include "xfs_inode.h" -#include "xfs_ialloc.h" #include "xfs_itable.h" #include "xfs_bmap.h" -#include "xfs_btree.h" #include "xfs_rtalloc.h" #include "xfs_error.h" -#include "xfs_rw.h" #include "xfs_attr.h" #include "xfs_buf_item.h" #include "xfs_qm.h" diff --git a/fs/xfs/quota/xfs_qm_syscalls.c b/fs/xfs/quota/xfs_qm_syscalls.c index b4487764e92..45e5849df23 100644 --- a/fs/xfs/quota/xfs_qm_syscalls.c +++ b/fs/xfs/quota/xfs_qm_syscalls.c @@ -26,25 +26,15 @@ #include "xfs_trans.h" #include "xfs_sb.h" #include "xfs_ag.h" -#include "xfs_dir2.h" #include "xfs_alloc.h" -#include "xfs_dmapi.h" #include "xfs_quota.h" #include "xfs_mount.h" #include "xfs_bmap_btree.h" -#include "xfs_alloc_btree.h" -#include "xfs_ialloc_btree.h" -#include "xfs_dir2_sf.h" -#include "xfs_attr_sf.h" -#include "xfs_dinode.h" #include "xfs_inode.h" -#include "xfs_ialloc.h" #include "xfs_itable.h" #include "xfs_bmap.h" -#include "xfs_btree.h" #include "xfs_rtalloc.h" #include "xfs_error.h" -#include "xfs_rw.h" #include "xfs_attr.h" #include "xfs_buf_item.h" #include "xfs_utils.h" @@ -248,40 +238,74 @@ out_unlock: return error; } +STATIC int +xfs_qm_scall_trunc_qfile( + struct xfs_mount *mp, + xfs_ino_t ino) +{ + struct xfs_inode *ip; + struct xfs_trans *tp; + int error; + + if (ino == NULLFSINO) + return 0; + + error = xfs_iget(mp, NULL, ino, 0, 0, &ip); + if (error) + return error; + + xfs_ilock(ip, XFS_IOLOCK_EXCL); + + tp = xfs_trans_alloc(mp, XFS_TRANS_TRUNCATE_FILE); + error = xfs_trans_reserve(tp, 0, XFS_ITRUNCATE_LOG_RES(mp), 0, + XFS_TRANS_PERM_LOG_RES, + XFS_ITRUNCATE_LOG_COUNT); + if (error) { + xfs_trans_cancel(tp, 0); + xfs_iunlock(ip, XFS_IOLOCK_EXCL); + goto out_put; + } + + xfs_ilock(ip, XFS_ILOCK_EXCL); + xfs_trans_ijoin(tp, ip); + + error = xfs_itruncate_finish(&tp, ip, 0, XFS_DATA_FORK, 1); + if (error) { + xfs_trans_cancel(tp, XFS_TRANS_RELEASE_LOG_RES | + XFS_TRANS_ABORT); + goto out_unlock; + } + + xfs_ichgtime(ip, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG); + error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES); + +out_unlock: + xfs_iunlock(ip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL); +out_put: + IRELE(ip); + return error; +} + int xfs_qm_scall_trunc_qfiles( xfs_mount_t *mp, uint flags) { int error = 0, error2 = 0; - xfs_inode_t *qip; if (!xfs_sb_version_hasquota(&mp->m_sb) || flags == 0) { qdprintk("qtrunc flags=%x m_qflags=%x\n", flags, mp->m_qflags); return XFS_ERROR(EINVAL); } - if ((flags & XFS_DQ_USER) && mp->m_sb.sb_uquotino != NULLFSINO) { - error = xfs_iget(mp, NULL, mp->m_sb.sb_uquotino, 0, 0, &qip); - if (!error) { - error = xfs_truncate_file(mp, qip); - IRELE(qip); - } - } - - if ((flags & (XFS_DQ_GROUP|XFS_DQ_PROJ)) && - mp->m_sb.sb_gquotino != NULLFSINO) { - error2 = xfs_iget(mp, NULL, mp->m_sb.sb_gquotino, 0, 0, &qip); - if (!error2) { - error2 = xfs_truncate_file(mp, qip); - IRELE(qip); - } - } + if (flags & XFS_DQ_USER) + error = xfs_qm_scall_trunc_qfile(mp, mp->m_sb.sb_uquotino); + if (flags & (XFS_DQ_GROUP|XFS_DQ_PROJ)) + error2 = xfs_qm_scall_trunc_qfile(mp, mp->m_sb.sb_gquotino); return error ? error : error2; } - /* * Switch on (a given) quota enforcement for a filesystem. This takes * effect immediately. @@ -786,9 +810,9 @@ xfs_qm_export_dquot( } #ifdef DEBUG - if (((XFS_IS_UQUOTA_ENFORCED(mp) && dst->d_flags == XFS_USER_QUOTA) || + if (((XFS_IS_UQUOTA_ENFORCED(mp) && dst->d_flags == FS_USER_QUOTA) || (XFS_IS_OQUOTA_ENFORCED(mp) && - (dst->d_flags & (XFS_PROJ_QUOTA | XFS_GROUP_QUOTA)))) && + (dst->d_flags & (FS_PROJ_QUOTA | FS_GROUP_QUOTA)))) && dst->d_id != 0) { if (((int) dst->d_bcount >= (int) dst->d_blk_softlimit) && (dst->d_blk_softlimit > 0)) { @@ -809,17 +833,17 @@ xfs_qm_export_qtype_flags( /* * Can't be more than one, or none. */ - ASSERT((flags & (XFS_PROJ_QUOTA | XFS_USER_QUOTA)) != - (XFS_PROJ_QUOTA | XFS_USER_QUOTA)); - ASSERT((flags & (XFS_PROJ_QUOTA | XFS_GROUP_QUOTA)) != - (XFS_PROJ_QUOTA | XFS_GROUP_QUOTA)); - ASSERT((flags & (XFS_USER_QUOTA | XFS_GROUP_QUOTA)) != - (XFS_USER_QUOTA | XFS_GROUP_QUOTA)); - ASSERT((flags & (XFS_PROJ_QUOTA|XFS_USER_QUOTA|XFS_GROUP_QUOTA)) != 0); + ASSERT((flags & (FS_PROJ_QUOTA | FS_USER_QUOTA)) != + (FS_PROJ_QUOTA | FS_USER_QUOTA)); + ASSERT((flags & (FS_PROJ_QUOTA | FS_GROUP_QUOTA)) != + (FS_PROJ_QUOTA | FS_GROUP_QUOTA)); + ASSERT((flags & (FS_USER_QUOTA | FS_GROUP_QUOTA)) != + (FS_USER_QUOTA | FS_GROUP_QUOTA)); + ASSERT((flags & (FS_PROJ_QUOTA|FS_USER_QUOTA|FS_GROUP_QUOTA)) != 0); return (flags & XFS_DQ_USER) ? - XFS_USER_QUOTA : (flags & XFS_DQ_PROJ) ? - XFS_PROJ_QUOTA : XFS_GROUP_QUOTA; + FS_USER_QUOTA : (flags & XFS_DQ_PROJ) ? + FS_PROJ_QUOTA : FS_GROUP_QUOTA; } STATIC uint @@ -830,16 +854,16 @@ xfs_qm_export_flags( uflags = 0; if (flags & XFS_UQUOTA_ACCT) - uflags |= XFS_QUOTA_UDQ_ACCT; + uflags |= FS_QUOTA_UDQ_ACCT; if (flags & XFS_PQUOTA_ACCT) - uflags |= XFS_QUOTA_PDQ_ACCT; + uflags |= FS_QUOTA_PDQ_ACCT; if (flags & XFS_GQUOTA_ACCT) - uflags |= XFS_QUOTA_GDQ_ACCT; + uflags |= FS_QUOTA_GDQ_ACCT; if (flags & XFS_UQUOTA_ENFD) - uflags |= XFS_QUOTA_UDQ_ENFD; + uflags |= FS_QUOTA_UDQ_ENFD; if (flags & (XFS_OQUOTA_ENFD)) { uflags |= (flags & XFS_GQUOTA_ACCT) ? - XFS_QUOTA_GDQ_ENFD : XFS_QUOTA_PDQ_ENFD; + FS_QUOTA_GDQ_ENFD : FS_QUOTA_PDQ_ENFD; } return (uflags); } @@ -875,8 +899,9 @@ xfs_dqrele_inode( xfs_qm_dqrele(ip->i_gdquot); ip->i_gdquot = NULL; } - xfs_iput(ip, XFS_ILOCK_EXCL); + xfs_iunlock(ip, XFS_ILOCK_EXCL); + IRELE(ip); return 0; } @@ -1143,7 +1168,8 @@ xfs_qm_internalqcheck_adjust( * of those now. */ if (! ipreleased) { - xfs_iput(ip, lock_flags); + xfs_iunlock(ip, lock_flags); + IRELE(ip); ipreleased = B_TRUE; goto again; } @@ -1160,7 +1186,8 @@ xfs_qm_internalqcheck_adjust( ASSERT(gd); xfs_qm_internalqcheck_dqadjust(ip, gd); } - xfs_iput(ip, lock_flags); + xfs_iunlock(ip, lock_flags); + IRELE(ip); *res = BULKSTAT_RV_DIDONE; return (0); } diff --git a/fs/xfs/quota/xfs_trans_dquot.c b/fs/xfs/quota/xfs_trans_dquot.c index 061d827da33..7de91d1b75c 100644 --- a/fs/xfs/quota/xfs_trans_dquot.c +++ b/fs/xfs/quota/xfs_trans_dquot.c @@ -23,25 +23,15 @@ #include "xfs_trans.h" #include "xfs_sb.h" #include "xfs_ag.h" -#include "xfs_dir2.h" #include "xfs_alloc.h" -#include "xfs_dmapi.h" #include "xfs_quota.h" #include "xfs_mount.h" #include "xfs_bmap_btree.h" -#include "xfs_alloc_btree.h" -#include "xfs_ialloc_btree.h" -#include "xfs_attr_sf.h" -#include "xfs_dir2_sf.h" -#include "xfs_dinode.h" #include "xfs_inode.h" -#include "xfs_ialloc.h" #include "xfs_itable.h" -#include "xfs_btree.h" #include "xfs_bmap.h" #include "xfs_rtalloc.h" #include "xfs_error.h" -#include "xfs_rw.h" #include "xfs_attr.h" #include "xfs_buf_item.h" #include "xfs_trans_priv.h" @@ -59,16 +49,14 @@ xfs_trans_dqjoin( xfs_trans_t *tp, xfs_dquot_t *dqp) { - xfs_dq_logitem_t *lp = &dqp->q_logitem; - ASSERT(dqp->q_transp != tp); ASSERT(XFS_DQ_IS_LOCKED(dqp)); - ASSERT(lp->qli_dquot == dqp); + ASSERT(dqp->q_logitem.qli_dquot == dqp); /* * Get a log_item_desc to point at the new item. */ - (void) xfs_trans_add_item(tp, (xfs_log_item_t*)(lp)); + xfs_trans_add_item(tp, &dqp->q_logitem.qli_item); /* * Initialize i_transp so we can later determine if this dquot is @@ -93,16 +81,11 @@ xfs_trans_log_dquot( xfs_trans_t *tp, xfs_dquot_t *dqp) { - xfs_log_item_desc_t *lidp; - ASSERT(dqp->q_transp == tp); ASSERT(XFS_DQ_IS_LOCKED(dqp)); - lidp = xfs_trans_find_item(tp, (xfs_log_item_t*)(&dqp->q_logitem)); - ASSERT(lidp != NULL); - tp->t_flags |= XFS_TRANS_DIRTY; - lidp->lid_flags |= XFS_LID_DIRTY; + dqp->q_logitem.qli_item.li_desc->lid_flags |= XFS_LID_DIRTY; } /* @@ -874,9 +857,8 @@ xfs_trans_get_qoff_item( /* * Get a log_item_desc to point at the new item. */ - (void) xfs_trans_add_item(tp, (xfs_log_item_t*)q); - - return (q); + xfs_trans_add_item(tp, &q->qql_item); + return q; } @@ -890,13 +872,8 @@ xfs_trans_log_quotaoff_item( xfs_trans_t *tp, xfs_qoff_logitem_t *qlp) { - xfs_log_item_desc_t *lidp; - - lidp = xfs_trans_find_item(tp, (xfs_log_item_t *)qlp); - ASSERT(lidp != NULL); - tp->t_flags |= XFS_TRANS_DIRTY; - lidp->lid_flags |= XFS_LID_DIRTY; + qlp->qql_item.li_desc->lid_flags |= XFS_LID_DIRTY; } STATIC void diff --git a/fs/xfs/support/debug.c b/fs/xfs/support/debug.c index 3f3610a7ee0..975aa10e1a4 100644 --- a/fs/xfs/support/debug.c +++ b/fs/xfs/support/debug.c @@ -22,7 +22,6 @@ #include "xfs_sb.h" #include "xfs_inum.h" #include "xfs_ag.h" -#include "xfs_dmapi.h" #include "xfs_mount.h" #include "xfs_error.h" diff --git a/fs/xfs/xfs_alloc.c b/fs/xfs/xfs_alloc.c index a7fbe8a99b1..af168faccc7 100644 --- a/fs/xfs/xfs_alloc.c +++ b/fs/xfs/xfs_alloc.c @@ -24,18 +24,13 @@ #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_dir2_sf.h" -#include "xfs_attr_sf.h" #include "xfs_dinode.h" #include "xfs_inode.h" #include "xfs_btree.h" -#include "xfs_ialloc.h" #include "xfs_alloc.h" #include "xfs_error.h" #include "xfs_trace.h" @@ -688,8 +683,6 @@ xfs_alloc_ag_vextent_near( xfs_agblock_t ltbno; /* start bno of left side entry */ xfs_agblock_t ltbnoa; /* aligned ... */ xfs_extlen_t ltdiff; /* difference to left side entry */ - /*REFERENCED*/ - xfs_agblock_t ltend; /* end bno of left side entry */ xfs_extlen_t ltlen; /* length of left side entry */ xfs_extlen_t ltlena; /* aligned ... */ xfs_agblock_t ltnew; /* useful start bno of left side */ @@ -814,8 +807,7 @@ xfs_alloc_ag_vextent_near( if ((error = xfs_alloc_get_rec(cnt_cur, <bno, <len, &i))) goto error0; XFS_WANT_CORRUPTED_GOTO(i == 1, error0); - ltend = ltbno + ltlen; - ASSERT(ltend <= be32_to_cpu(XFS_BUF_TO_AGF(args->agbp)->agf_length)); + ASSERT(ltbno + ltlen <= be32_to_cpu(XFS_BUF_TO_AGF(args->agbp)->agf_length)); args->len = blen; if (!xfs_alloc_fix_minleft(args)) { xfs_btree_del_cursor(cnt_cur, XFS_BTREE_NOERROR); @@ -828,7 +820,7 @@ xfs_alloc_ag_vextent_near( */ args->agbno = bnew; ASSERT(bnew >= ltbno); - ASSERT(bnew + blen <= ltend); + ASSERT(bnew + blen <= ltbno + ltlen); /* * Set up a cursor for the by-bno tree. */ @@ -1157,7 +1149,6 @@ xfs_alloc_ag_vextent_near( /* * Fix up the length and compute the useful address. */ - ltend = ltbno + ltlen; args->len = XFS_EXTLEN_MIN(ltlena, args->maxlen); xfs_alloc_fix_len(args); if (!xfs_alloc_fix_minleft(args)) { @@ -1170,7 +1161,7 @@ xfs_alloc_ag_vextent_near( (void)xfs_alloc_compute_diff(args->agbno, rlen, args->alignment, ltbno, ltlen, <new); ASSERT(ltnew >= ltbno); - ASSERT(ltnew + rlen <= ltend); + ASSERT(ltnew + rlen <= ltbno + ltlen); ASSERT(ltnew + rlen <= be32_to_cpu(XFS_BUF_TO_AGF(args->agbp)->agf_length)); args->agbno = ltnew; if ((error = xfs_alloc_fixup_trees(cnt_cur, bno_cur_lt, ltbno, ltlen, diff --git a/fs/xfs/xfs_alloc.h b/fs/xfs/xfs_alloc.h index 6d05199b667..895009a9727 100644 --- a/fs/xfs/xfs_alloc.h +++ b/fs/xfs/xfs_alloc.h @@ -27,16 +27,16 @@ struct xfs_busy_extent; /* * Freespace allocation types. Argument to xfs_alloc_[v]extent. */ -typedef enum xfs_alloctype -{ - XFS_ALLOCTYPE_ANY_AG, /* allocate anywhere, use rotor */ - XFS_ALLOCTYPE_FIRST_AG, /* ... start at ag 0 */ - XFS_ALLOCTYPE_START_AG, /* anywhere, start in this a.g. */ - XFS_ALLOCTYPE_THIS_AG, /* anywhere in this a.g. */ - XFS_ALLOCTYPE_START_BNO, /* near this block else anywhere */ - XFS_ALLOCTYPE_NEAR_BNO, /* in this a.g. and near this block */ - XFS_ALLOCTYPE_THIS_BNO /* at exactly this block */ -} xfs_alloctype_t; +#define XFS_ALLOCTYPE_ANY_AG 0x01 /* allocate anywhere, use rotor */ +#define XFS_ALLOCTYPE_FIRST_AG 0x02 /* ... start at ag 0 */ +#define XFS_ALLOCTYPE_START_AG 0x04 /* anywhere, start in this a.g. */ +#define XFS_ALLOCTYPE_THIS_AG 0x08 /* anywhere in this a.g. */ +#define XFS_ALLOCTYPE_START_BNO 0x10 /* near this block else anywhere */ +#define XFS_ALLOCTYPE_NEAR_BNO 0x20 /* in this a.g. and near this block */ +#define XFS_ALLOCTYPE_THIS_BNO 0x40 /* at exactly this block */ + +/* this should become an enum again when the tracing code is fixed */ +typedef unsigned int xfs_alloctype_t; #define XFS_ALLOC_TYPES \ { XFS_ALLOCTYPE_ANY_AG, "ANY_AG" }, \ diff --git a/fs/xfs/xfs_alloc_btree.c b/fs/xfs/xfs_alloc_btree.c index 83f49421875..97f7328967f 100644 --- a/fs/xfs/xfs_alloc_btree.c +++ b/fs/xfs/xfs_alloc_btree.c @@ -24,19 +24,14 @@ #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_dir2_sf.h" -#include "xfs_attr_sf.h" #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" #include "xfs_trace.h" diff --git a/fs/xfs/xfs_attr.c b/fs/xfs/xfs_attr.c index b9c196a53c4..c2568242a90 100644 --- a/fs/xfs/xfs_attr.c +++ b/fs/xfs/xfs_attr.c @@ -25,19 +25,13 @@ #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_alloc_btree.h" -#include "xfs_ialloc_btree.h" -#include "xfs_dir2_sf.h" #include "xfs_attr_sf.h" #include "xfs_dinode.h" #include "xfs_inode.h" #include "xfs_alloc.h" -#include "xfs_btree.h" #include "xfs_inode_item.h" #include "xfs_bmap.h" #include "xfs_attr.h" @@ -325,8 +319,7 @@ xfs_attr_set_int( return (error); } - xfs_trans_ijoin(args.trans, dp, XFS_ILOCK_EXCL); - xfs_trans_ihold(args.trans, dp); + xfs_trans_ijoin(args.trans, dp); /* * If the attribute list is non-existent or a shortform list, @@ -396,10 +389,8 @@ xfs_attr_set_int( * bmap_finish() may have committed the last trans and started * a new one. We need the inode to be in all transactions. */ - if (committed) { - xfs_trans_ijoin(args.trans, dp, XFS_ILOCK_EXCL); - xfs_trans_ihold(args.trans, dp); - } + if (committed) + xfs_trans_ijoin(args.trans, dp); /* * Commit the leaf transformation. We'll need another (linked) @@ -544,8 +535,7 @@ xfs_attr_remove_int(xfs_inode_t *dp, struct xfs_name *name, int flags) * No need to make quota reservations here. We expect to release some * blocks not allocate in the common case. */ - xfs_trans_ijoin(args.trans, dp, XFS_ILOCK_EXCL); - xfs_trans_ihold(args.trans, dp); + xfs_trans_ijoin(args.trans, dp); /* * Decide on what work routines to call based on the inode size. @@ -821,8 +811,7 @@ xfs_attr_inactive(xfs_inode_t *dp) * No need to make quota reservations here. We expect to release some * blocks, not allocate, in the common case. */ - xfs_trans_ijoin(trans, dp, XFS_ILOCK_EXCL); - xfs_trans_ihold(trans, dp); + xfs_trans_ijoin(trans, dp); /* * Decide on what work routines to call based on the inode size. @@ -981,10 +970,8 @@ xfs_attr_leaf_addname(xfs_da_args_t *args) * bmap_finish() may have committed the last trans and started * a new one. We need the inode to be in all transactions. */ - if (committed) { - xfs_trans_ijoin(args->trans, dp, XFS_ILOCK_EXCL); - xfs_trans_ihold(args->trans, dp); - } + if (committed) + xfs_trans_ijoin(args->trans, dp); /* * Commit the current trans (including the inode) and start @@ -1085,10 +1072,8 @@ xfs_attr_leaf_addname(xfs_da_args_t *args) * and started a new one. We need the inode to be * in all transactions. */ - if (committed) { - xfs_trans_ijoin(args->trans, dp, XFS_ILOCK_EXCL); - xfs_trans_ihold(args->trans, dp); - } + if (committed) + xfs_trans_ijoin(args->trans, dp); } else xfs_da_buf_done(bp); @@ -1161,10 +1146,8 @@ xfs_attr_leaf_removename(xfs_da_args_t *args) * bmap_finish() may have committed the last trans and started * a new one. We need the inode to be in all transactions. */ - if (committed) { - xfs_trans_ijoin(args->trans, dp, XFS_ILOCK_EXCL); - xfs_trans_ihold(args->trans, dp); - } + if (committed) + xfs_trans_ijoin(args->trans, dp); } else xfs_da_buf_done(bp); return(0); @@ -1317,10 +1300,8 @@ restart: * and started a new one. We need the inode to be * in all transactions. */ - if (committed) { - xfs_trans_ijoin(args->trans, dp, XFS_ILOCK_EXCL); - xfs_trans_ihold(args->trans, dp); - } + if (committed) + xfs_trans_ijoin(args->trans, dp); /* * Commit the node conversion and start the next @@ -1356,10 +1337,8 @@ restart: * bmap_finish() may have committed the last trans and started * a new one. We need the inode to be in all transactions. */ - if (committed) { - xfs_trans_ijoin(args->trans, dp, XFS_ILOCK_EXCL); - xfs_trans_ihold(args->trans, dp); - } + if (committed) + xfs_trans_ijoin(args->trans, dp); } else { /* * Addition succeeded, update Btree hashvals. @@ -1470,10 +1449,8 @@ restart: * and started a new one. We need the inode to be * in all transactions. */ - if (committed) { - xfs_trans_ijoin(args->trans, dp, XFS_ILOCK_EXCL); - xfs_trans_ihold(args->trans, dp); - } + if (committed) + xfs_trans_ijoin(args->trans, dp); } /* @@ -1604,10 +1581,8 @@ xfs_attr_node_removename(xfs_da_args_t *args) * bmap_finish() may have committed the last trans and started * a new one. We need the inode to be in all transactions. */ - if (committed) { - xfs_trans_ijoin(args->trans, dp, XFS_ILOCK_EXCL); - xfs_trans_ihold(args->trans, dp); - } + if (committed) + xfs_trans_ijoin(args->trans, dp); /* * Commit the Btree join operation and start a new trans. @@ -1658,10 +1633,8 @@ xfs_attr_node_removename(xfs_da_args_t *args) * and started a new one. We need the inode to be * in all transactions. */ - if (committed) { - xfs_trans_ijoin(args->trans, dp, XFS_ILOCK_EXCL); - xfs_trans_ihold(args->trans, dp); - } + if (committed) + xfs_trans_ijoin(args->trans, dp); } else xfs_da_brelse(args->trans, bp); } @@ -2004,7 +1977,7 @@ xfs_attr_rmtval_get(xfs_da_args_t *args) error = xfs_bmapi(args->trans, args->dp, (xfs_fileoff_t)lblkno, args->rmtblkcnt, XFS_BMAPI_ATTRFORK | XFS_BMAPI_METADATA, - NULL, 0, map, &nmap, NULL, NULL); + NULL, 0, map, &nmap, NULL); if (error) return(error); ASSERT(nmap >= 1); @@ -2083,7 +2056,7 @@ xfs_attr_rmtval_set(xfs_da_args_t *args) XFS_BMAPI_ATTRFORK | XFS_BMAPI_METADATA | XFS_BMAPI_WRITE, args->firstblock, args->total, &map, &nmap, - args->flist, NULL); + args->flist); if (!error) { error = xfs_bmap_finish(&args->trans, args->flist, &committed); @@ -2099,10 +2072,8 @@ xfs_attr_rmtval_set(xfs_da_args_t *args) * bmap_finish() may have committed the last trans and started * a new one. We need the inode to be in all transactions. */ - if (committed) { - xfs_trans_ijoin(args->trans, dp, XFS_ILOCK_EXCL); - xfs_trans_ihold(args->trans, dp); - } + if (committed) + xfs_trans_ijoin(args->trans, dp); ASSERT(nmap == 1); ASSERT((map.br_startblock != DELAYSTARTBLOCK) && @@ -2136,7 +2107,7 @@ xfs_attr_rmtval_set(xfs_da_args_t *args) args->rmtblkcnt, XFS_BMAPI_ATTRFORK | XFS_BMAPI_METADATA, args->firstblock, 0, &map, &nmap, - NULL, NULL); + NULL); if (error) { return(error); } @@ -2201,7 +2172,7 @@ xfs_attr_rmtval_remove(xfs_da_args_t *args) args->rmtblkcnt, XFS_BMAPI_ATTRFORK | XFS_BMAPI_METADATA, args->firstblock, 0, &map, &nmap, - args->flist, NULL); + args->flist); if (error) { return(error); } @@ -2239,7 +2210,7 @@ xfs_attr_rmtval_remove(xfs_da_args_t *args) error = xfs_bunmapi(args->trans, args->dp, lblkno, blkcnt, XFS_BMAPI_ATTRFORK | XFS_BMAPI_METADATA, 1, args->firstblock, args->flist, - NULL, &done); + &done); if (!error) { error = xfs_bmap_finish(&args->trans, args->flist, &committed); @@ -2255,10 +2226,8 @@ xfs_attr_rmtval_remove(xfs_da_args_t *args) * bmap_finish() may have committed the last trans and started * a new one. We need the inode to be in all transactions. */ - if (committed) { - xfs_trans_ijoin(args->trans, args->dp, XFS_ILOCK_EXCL); - xfs_trans_ihold(args->trans, args->dp); - } + if (committed) + xfs_trans_ijoin(args->trans, args->dp); /* * Close out trans and start the next one in the chain. diff --git a/fs/xfs/xfs_attr_leaf.c b/fs/xfs/xfs_attr_leaf.c index a90ce74fc25..a6cff8edcdb 100644 --- a/fs/xfs/xfs_attr_leaf.c +++ b/fs/xfs/xfs_attr_leaf.c @@ -24,8 +24,6 @@ #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" @@ -33,7 +31,6 @@ #include "xfs_ialloc_btree.h" #include "xfs_alloc.h" #include "xfs_btree.h" -#include "xfs_dir2_sf.h" #include "xfs_attr_sf.h" #include "xfs_dinode.h" #include "xfs_inode.h" @@ -2931,7 +2928,7 @@ xfs_attr_leaf_freextent(xfs_trans_t **trans, xfs_inode_t *dp, nmap = 1; error = xfs_bmapi(*trans, dp, (xfs_fileoff_t)tblkno, tblkcnt, XFS_BMAPI_ATTRFORK | XFS_BMAPI_METADATA, - NULL, 0, &map, &nmap, NULL, NULL); + NULL, 0, &map, &nmap, NULL); if (error) { return(error); } diff --git a/fs/xfs/xfs_bmap.c b/fs/xfs/xfs_bmap.c index 99587ded043..23f14e595c1 100644 --- a/fs/xfs/xfs_bmap.c +++ b/fs/xfs/xfs_bmap.c @@ -30,13 +30,10 @@ #include "xfs_alloc_btree.h" #include "xfs_ialloc_btree.h" #include "xfs_dir2_sf.h" -#include "xfs_attr_sf.h" #include "xfs_dinode.h" #include "xfs_inode.h" #include "xfs_btree.h" -#include "xfs_dmapi.h" #include "xfs_mount.h" -#include "xfs_ialloc.h" #include "xfs_itable.h" #include "xfs_dir2_data.h" #include "xfs_dir2_leaf.h" @@ -104,7 +101,6 @@ xfs_bmap_add_extent( xfs_fsblock_t *first, /* pointer to firstblock variable */ xfs_bmap_free_t *flist, /* list of extents to be freed */ int *logflagsp, /* inode logging flags */ - xfs_extdelta_t *delta, /* Change made to incore extents */ int whichfork, /* data or attr fork */ int rsvd); /* OK to allocate reserved blocks */ @@ -122,7 +118,6 @@ xfs_bmap_add_extent_delay_real( xfs_fsblock_t *first, /* pointer to firstblock variable */ xfs_bmap_free_t *flist, /* list of extents to be freed */ int *logflagsp, /* inode logging flags */ - xfs_extdelta_t *delta, /* Change made to incore extents */ int rsvd); /* OK to allocate reserved blocks */ /* @@ -135,7 +130,6 @@ xfs_bmap_add_extent_hole_delay( xfs_extnum_t idx, /* extent number to update/insert */ xfs_bmbt_irec_t *new, /* new data to add to file extents */ int *logflagsp,/* inode logging flags */ - xfs_extdelta_t *delta, /* Change made to incore extents */ int rsvd); /* OK to allocate reserved blocks */ /* @@ -149,7 +143,6 @@ xfs_bmap_add_extent_hole_real( xfs_btree_cur_t *cur, /* if null, not a btree */ xfs_bmbt_irec_t *new, /* new data to add to file extents */ int *logflagsp, /* inode logging flags */ - xfs_extdelta_t *delta, /* Change made to incore extents */ int whichfork); /* data or attr fork */ /* @@ -162,8 +155,7 @@ xfs_bmap_add_extent_unwritten_real( xfs_extnum_t idx, /* extent number to update/insert */ xfs_btree_cur_t **curp, /* if *curp is null, not a btree */ xfs_bmbt_irec_t *new, /* new data to add to file extents */ - int *logflagsp, /* inode logging flags */ - xfs_extdelta_t *delta); /* Change made to incore extents */ + int *logflagsp); /* inode logging flags */ /* * xfs_bmap_alloc is called by xfs_bmapi to allocate an extent for a file. @@ -200,7 +192,6 @@ xfs_bmap_del_extent( xfs_btree_cur_t *cur, /* if null, not a btree */ xfs_bmbt_irec_t *new, /* new data to add to file extents */ int *logflagsp,/* inode logging flags */ - xfs_extdelta_t *delta, /* Change made to incore extents */ int whichfork, /* data or attr fork */ int rsvd); /* OK to allocate reserved blocks */ @@ -489,7 +480,6 @@ xfs_bmap_add_extent( xfs_fsblock_t *first, /* pointer to firstblock variable */ xfs_bmap_free_t *flist, /* list of extents to be freed */ int *logflagsp, /* inode logging flags */ - xfs_extdelta_t *delta, /* Change made to incore extents */ int whichfork, /* data or attr fork */ int rsvd) /* OK to use reserved data blocks */ { @@ -524,15 +514,6 @@ xfs_bmap_add_extent( logflags = XFS_ILOG_CORE | xfs_ilog_fext(whichfork); } else logflags = 0; - /* DELTA: single new extent */ - if (delta) { - if (delta->xed_startoff > new->br_startoff) - delta->xed_startoff = new->br_startoff; - if (delta->xed_blockcount < - new->br_startoff + new->br_blockcount) - delta->xed_blockcount = new->br_startoff + - new->br_blockcount; - } } /* * Any kind of new delayed allocation goes here. @@ -542,7 +523,7 @@ xfs_bmap_add_extent( ASSERT((cur->bc_private.b.flags & XFS_BTCUR_BPRV_WASDEL) == 0); if ((error = xfs_bmap_add_extent_hole_delay(ip, idx, new, - &logflags, delta, rsvd))) + &logflags, rsvd))) goto done; } /* @@ -553,7 +534,7 @@ xfs_bmap_add_extent( ASSERT((cur->bc_private.b.flags & XFS_BTCUR_BPRV_WASDEL) == 0); if ((error = xfs_bmap_add_extent_hole_real(ip, idx, cur, new, - &logflags, delta, whichfork))) + &logflags, whichfork))) goto done; } else { xfs_bmbt_irec_t prev; /* old extent at offset idx */ @@ -578,17 +559,17 @@ xfs_bmap_add_extent( XFS_BTCUR_BPRV_WASDEL); if ((error = xfs_bmap_add_extent_delay_real(ip, idx, &cur, new, &da_new, first, flist, - &logflags, delta, rsvd))) + &logflags, rsvd))) goto done; } else if (new->br_state == XFS_EXT_NORM) { ASSERT(new->br_state == XFS_EXT_NORM); if ((error = xfs_bmap_add_extent_unwritten_real( - ip, idx, &cur, new, &logflags, delta))) + ip, idx, &cur, new, &logflags))) goto done; } else { ASSERT(new->br_state == XFS_EXT_UNWRITTEN); if ((error = xfs_bmap_add_extent_unwritten_real( - ip, idx, &cur, new, &logflags, delta))) + ip, idx, &cur, new, &logflags))) goto done; } ASSERT(*curp == cur || *curp == NULL); @@ -601,7 +582,7 @@ xfs_bmap_add_extent( ASSERT((cur->bc_private.b.flags & XFS_BTCUR_BPRV_WASDEL) == 0); if ((error = xfs_bmap_add_extent_hole_real(ip, idx, cur, - new, &logflags, delta, whichfork))) + new, &logflags, whichfork))) goto done; } } @@ -666,7 +647,6 @@ xfs_bmap_add_extent_delay_real( xfs_fsblock_t *first, /* pointer to firstblock variable */ xfs_bmap_free_t *flist, /* list of extents to be freed */ int *logflagsp, /* inode logging flags */ - xfs_extdelta_t *delta, /* Change made to incore extents */ int rsvd) /* OK to use reserved data block allocation */ { xfs_btree_cur_t *cur; /* btree cursor */ @@ -797,11 +777,6 @@ xfs_bmap_add_extent_delay_real( goto done; } *dnew = 0; - /* DELTA: Three in-core extents are replaced by one. */ - temp = LEFT.br_startoff; - temp2 = LEFT.br_blockcount + - PREV.br_blockcount + - RIGHT.br_blockcount; break; case BMAP_LEFT_FILLING | BMAP_RIGHT_FILLING | BMAP_LEFT_CONTIG: @@ -832,10 +807,6 @@ xfs_bmap_add_extent_delay_real( goto done; } *dnew = 0; - /* DELTA: Two in-core extents are replaced by one. */ - temp = LEFT.br_startoff; - temp2 = LEFT.br_blockcount + - PREV.br_blockcount; break; case BMAP_LEFT_FILLING | BMAP_RIGHT_FILLING | BMAP_RIGHT_CONTIG: @@ -867,10 +838,6 @@ xfs_bmap_add_extent_delay_real( goto done; } *dnew = 0; - /* DELTA: Two in-core extents are replaced by one. */ - temp = PREV.br_startoff; - temp2 = PREV.br_blockcount + - RIGHT.br_blockcount; break; case BMAP_LEFT_FILLING | BMAP_RIGHT_FILLING: @@ -900,9 +867,6 @@ xfs_bmap_add_extent_delay_real( XFS_WANT_CORRUPTED_GOTO(i == 1, done); } *dnew = 0; - /* DELTA: The in-core extent described by new changed type. */ - temp = new->br_startoff; - temp2 = new->br_blockcount; break; case BMAP_LEFT_FILLING | BMAP_LEFT_CONTIG: @@ -942,10 +906,6 @@ xfs_bmap_add_extent_delay_real( xfs_bmbt_set_startblock(ep, nullstartblock((int)temp)); trace_xfs_bmap_post_update(ip, idx, state, _THIS_IP_); *dnew = temp; - /* DELTA: The boundary between two in-core extents moved. */ - temp = LEFT.br_startoff; - temp2 = LEFT.br_blockcount + - PREV.br_blockcount; break; case BMAP_LEFT_FILLING: @@ -990,9 +950,6 @@ xfs_bmap_add_extent_delay_real( xfs_bmbt_set_startblock(ep, nullstartblock((int)temp)); trace_xfs_bmap_post_update(ip, idx + 1, state, _THIS_IP_); *dnew = temp; - /* DELTA: One in-core extent is split in two. */ - temp = PREV.br_startoff; - temp2 = PREV.br_blockcount; break; case BMAP_RIGHT_FILLING | BMAP_RIGHT_CONTIG: @@ -1031,10 +988,6 @@ xfs_bmap_add_extent_delay_real( xfs_bmbt_set_startblock(ep, nullstartblock((int)temp)); trace_xfs_bmap_post_update(ip, idx, state, _THIS_IP_); *dnew = temp; - /* DELTA: The boundary between two in-core extents moved. */ - temp = PREV.br_startoff; - temp2 = PREV.br_blockcount + - RIGHT.br_blockcount; break; case BMAP_RIGHT_FILLING: @@ -1078,9 +1031,6 @@ xfs_bmap_add_extent_delay_real( xfs_bmbt_set_startblock(ep, nullstartblock((int)temp)); trace_xfs_bmap_post_update(ip, idx, state, _THIS_IP_); *dnew = temp; - /* DELTA: One in-core extent is split in two. */ - temp = PREV.br_startoff; - temp2 = PREV.br_blockcount; break; case 0: @@ -1161,9 +1111,6 @@ xfs_bmap_add_extent_delay_real( nullstartblock((int)temp2)); trace_xfs_bmap_post_update(ip, idx + 2, state, _THIS_IP_); *dnew = temp + temp2; - /* DELTA: One in-core extent is split in three. */ - temp = PREV.br_startoff; - temp2 = PREV.br_blockcount; break; case BMAP_LEFT_FILLING | BMAP_LEFT_CONTIG | BMAP_RIGHT_CONTIG: @@ -1179,13 +1126,6 @@ xfs_bmap_add_extent_delay_real( ASSERT(0); } *curp = cur; - if (delta) { - temp2 += temp; - if (delta->xed_startoff > temp) - delta->xed_startoff = temp; - if (delta->xed_blockcount < temp2) - delta->xed_blockcount = temp2; - } done: *logflagsp = rval; return error; @@ -1204,8 +1144,7 @@ xfs_bmap_add_extent_unwritten_real( xfs_extnum_t idx, /* extent number to update/insert */ xfs_btree_cur_t **curp, /* if *curp is null, not a btree */ xfs_bmbt_irec_t *new, /* new data to add to file extents */ - int *logflagsp, /* inode logging flags */ - xfs_extdelta_t *delta) /* Change made to incore extents */ + int *logflagsp) /* inode logging flags */ { xfs_btree_cur_t *cur; /* btree cursor */ xfs_bmbt_rec_host_t *ep; /* extent entry for idx */ @@ -1219,8 +1158,6 @@ xfs_bmap_add_extent_unwritten_real( /* left is 0, right is 1, prev is 2 */ int rval=0; /* return value (logging flags) */ int state = 0;/* state bits, accessed thru macros */ - xfs_filblks_t temp=0; - xfs_filblks_t temp2=0; #define LEFT r[0] #define RIGHT r[1] @@ -1341,11 +1278,6 @@ xfs_bmap_add_extent_unwritten_real( RIGHT.br_blockcount, LEFT.br_state))) goto done; } - /* DELTA: Three in-core extents are replaced by one. */ - temp = LEFT.br_startoff; - temp2 = LEFT.br_blockcount + - PREV.br_blockcount + - RIGHT.br_blockcount; break; case BMAP_LEFT_FILLING | BMAP_RIGHT_FILLING | BMAP_LEFT_CONTIG: @@ -1382,10 +1314,6 @@ xfs_bmap_add_extent_unwritten_real( LEFT.br_state))) goto done; } - /* DELTA: Two in-core extents are replaced by one. */ - temp = LEFT.br_startoff; - temp2 = LEFT.br_blockcount + - PREV.br_blockcount; break; case BMAP_LEFT_FILLING | BMAP_RIGHT_FILLING | BMAP_RIGHT_CONTIG: @@ -1422,10 +1350,6 @@ xfs_bmap_add_extent_unwritten_real( newext))) goto done; } - /* DELTA: Two in-core extents are replaced by one. */ - temp = PREV.br_startoff; - temp2 = PREV.br_blockcount + - RIGHT.br_blockcount; break; case BMAP_LEFT_FILLING | BMAP_RIGHT_FILLING: @@ -1453,9 +1377,6 @@ xfs_bmap_add_extent_unwritten_real( newext))) goto done; } - /* DELTA: The in-core extent described by new changed type. */ - temp = new->br_startoff; - temp2 = new->br_blockcount; break; case BMAP_LEFT_FILLING | BMAP_LEFT_CONTIG: @@ -1501,10 +1422,6 @@ xfs_bmap_add_extent_unwritten_real( LEFT.br_state)) goto done; } - /* DELTA: The boundary between two in-core extents moved. */ - temp = LEFT.br_startoff; - temp2 = LEFT.br_blockcount + - PREV.br_blockcount; break; case BMAP_LEFT_FILLING: @@ -1544,9 +1461,6 @@ xfs_bmap_add_extent_unwritten_real( goto done; XFS_WANT_CORRUPTED_GOTO(i == 1, done); } - /* DELTA: One in-core extent is split in two. */ - temp = PREV.br_startoff; - temp2 = PREV.br_blockcount; break; case BMAP_RIGHT_FILLING | BMAP_RIGHT_CONTIG: @@ -1587,10 +1501,6 @@ xfs_bmap_add_extent_unwritten_real( newext))) goto done; } - /* DELTA: The boundary between two in-core extents moved. */ - temp = PREV.br_startoff; - temp2 = PREV.br_blockcount + - RIGHT.br_blockcount; break; case BMAP_RIGHT_FILLING: @@ -1630,9 +1540,6 @@ xfs_bmap_add_extent_unwritten_real( goto done; XFS_WANT_CORRUPTED_GOTO(i == 1, done); } - /* DELTA: One in-core extent is split in two. */ - temp = PREV.br_startoff; - temp2 = PREV.br_blockcount; break; case 0: @@ -1692,9 +1599,6 @@ xfs_bmap_add_extent_unwritten_real( goto done; XFS_WANT_CORRUPTED_GOTO(i == 1, done); } - /* DELTA: One in-core extent is split in three. */ - temp = PREV.br_startoff; - temp2 = PREV.br_blockcount; break; case BMAP_LEFT_FILLING | BMAP_LEFT_CONTIG | BMAP_RIGHT_CONTIG: @@ -1710,13 +1614,6 @@ xfs_bmap_add_extent_unwritten_real( ASSERT(0); } *curp = cur; - if (delta) { - temp2 += temp; - if (delta->xed_startoff > temp) - delta->xed_startoff = temp; - if (delta->xed_blockcount < temp2) - delta->xed_blockcount = temp2; - } done: *logflagsp = rval; return error; @@ -1736,7 +1633,6 @@ xfs_bmap_add_extent_hole_delay( xfs_extnum_t idx, /* extent number to update/insert */ xfs_bmbt_irec_t *new, /* new data to add to file extents */ int *logflagsp, /* inode logging flags */ - xfs_extdelta_t *delta, /* Change made to incore extents */ int rsvd) /* OK to allocate reserved blocks */ { xfs_bmbt_rec_host_t *ep; /* extent record for idx */ @@ -1747,7 +1643,6 @@ xfs_bmap_add_extent_hole_delay( xfs_bmbt_irec_t right; /* right neighbor extent entry */ int state; /* state bits, accessed thru macros */ xfs_filblks_t temp=0; /* temp for indirect calculations */ - xfs_filblks_t temp2=0; ifp = XFS_IFORK_PTR(ip, XFS_DATA_FORK); ep = xfs_iext_get_ext(ifp, idx); @@ -1819,9 +1714,6 @@ xfs_bmap_add_extent_hole_delay( xfs_iext_remove(ip, idx, 1, state); ip->i_df.if_lastex = idx - 1; - /* DELTA: Two in-core extents were replaced by one. */ - temp2 = temp; - temp = left.br_startoff; break; case BMAP_LEFT_CONTIG: @@ -1841,9 +1733,6 @@ xfs_bmap_add_extent_hole_delay( trace_xfs_bmap_post_update(ip, idx - 1, state, _THIS_IP_); ip->i_df.if_lastex = idx - 1; - /* DELTA: One in-core extent grew into a hole. */ - temp2 = temp; - temp = left.br_startoff; break; case BMAP_RIGHT_CONTIG: @@ -1862,9 +1751,6 @@ xfs_bmap_add_extent_hole_delay( trace_xfs_bmap_post_update(ip, idx, state, _THIS_IP_); ip->i_df.if_lastex = idx; - /* DELTA: One in-core extent grew into a hole. */ - temp2 = temp; - temp = new->br_startoff; break; case 0: @@ -1876,9 +1762,6 @@ xfs_bmap_add_extent_hole_delay( oldlen = newlen = 0; xfs_iext_insert(ip, idx, 1, new, state); ip->i_df.if_lastex = idx; - /* DELTA: A new in-core extent was added in a hole. */ - temp2 = new->br_blockcount; - temp = new->br_startoff; break; } if (oldlen != newlen) { @@ -1889,13 +1772,6 @@ xfs_bmap_add_extent_hole_delay( * Nothing to do for disk quota accounting here. */ } - if (delta) { - temp2 += temp; - if (delta->xed_startoff > temp) - delta->xed_startoff = temp; - if (delta->xed_blockcount < temp2) - delta->xed_blockcount = temp2; - } *logflagsp = 0; return 0; } @@ -1911,7 +1787,6 @@ xfs_bmap_add_extent_hole_real( xfs_btree_cur_t *cur, /* if null, not a btree */ xfs_bmbt_irec_t *new, /* new data to add to file extents */ int *logflagsp, /* inode logging flags */ - xfs_extdelta_t *delta, /* Change made to incore extents */ int whichfork) /* data or attr fork */ { xfs_bmbt_rec_host_t *ep; /* pointer to extent entry ins. point */ @@ -1922,8 +1797,6 @@ xfs_bmap_add_extent_hole_real( xfs_bmbt_irec_t right; /* right neighbor extent entry */ int rval=0; /* return value (logging flags) */ int state; /* state bits, accessed thru macros */ - xfs_filblks_t temp=0; - xfs_filblks_t temp2=0; ifp = XFS_IFORK_PTR(ip, whichfork); ASSERT(idx <= ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t)); @@ -2020,11 +1893,6 @@ xfs_bmap_add_extent_hole_real( left.br_state))) goto done; } - /* DELTA: Two in-core extents were replaced by one. */ - temp = left.br_startoff; - temp2 = left.br_blockcount + - new->br_blockcount + - right.br_blockcount; break; case BMAP_LEFT_CONTIG: @@ -2056,10 +1924,6 @@ xfs_bmap_add_extent_hole_real( left.br_state))) goto done; } - /* DELTA: One in-core extent grew. */ - temp = left.br_startoff; - temp2 = left.br_blockcount + - new->br_blockcount; break; case BMAP_RIGHT_CONTIG: @@ -2092,10 +1956,6 @@ xfs_bmap_add_extent_hole_real( right.br_state))) goto done; } - /* DELTA: One in-core extent grew. */ - temp = new->br_startoff; - temp2 = new->br_blockcount + - right.br_blockcount; break; case 0: @@ -2123,18 +1983,8 @@ xfs_bmap_add_extent_hole_real( goto done; XFS_WANT_CORRUPTED_GOTO(i == 1, done); } - /* DELTA: A new extent was added in a hole. */ - temp = new->br_startoff; - temp2 = new->br_blockcount; break; } - if (delta) { - temp2 += temp; - if (delta->xed_startoff > temp) - delta->xed_startoff = temp; - if (delta->xed_blockcount < temp2) - delta->xed_blockcount = temp2; - } done: *logflagsp = rval; return error; @@ -2959,7 +2809,6 @@ xfs_bmap_del_extent( xfs_btree_cur_t *cur, /* if null, not a btree */ xfs_bmbt_irec_t *del, /* data to remove from extents */ int *logflagsp, /* inode logging flags */ - xfs_extdelta_t *delta, /* Change made to incore extents */ int whichfork, /* data or attr fork */ int rsvd) /* OK to allocate reserved blocks */ { @@ -3265,14 +3114,6 @@ xfs_bmap_del_extent( if (da_old > da_new) xfs_mod_incore_sb(mp, XFS_SBS_FDBLOCKS, (int64_t)(da_old - da_new), rsvd); - if (delta) { - /* DELTA: report the original extent. */ - if (delta->xed_startoff > got.br_startoff) - delta->xed_startoff = got.br_startoff; - if (delta->xed_blockcount < got.br_startoff+got.br_blockcount) - delta->xed_blockcount = got.br_startoff + - got.br_blockcount; - } done: *logflagsp = flags; return error; @@ -3754,9 +3595,10 @@ xfs_bmap_add_attrfork( ip->i_d.di_aformat = XFS_DINODE_FMT_EXTENTS; } ASSERT(ip->i_d.di_anextents == 0); - IHOLD(ip); - xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL); + + xfs_trans_ijoin_ref(tp, ip, XFS_ILOCK_EXCL); xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); + switch (ip->i_d.di_format) { case XFS_DINODE_FMT_DEV: ip->i_d.di_forkoff = roundup(sizeof(xfs_dev_t), 8) >> 3; @@ -4483,8 +4325,7 @@ xfs_bmapi( xfs_extlen_t total, /* total blocks needed */ xfs_bmbt_irec_t *mval, /* output: map values */ int *nmap, /* i/o: mval size/count */ - xfs_bmap_free_t *flist, /* i/o: list extents to free */ - xfs_extdelta_t *delta) /* o: change made to incore extents */ + xfs_bmap_free_t *flist) /* i/o: list extents to free */ { xfs_fsblock_t abno; /* allocated block number */ xfs_extlen_t alen; /* allocated extent length */ @@ -4596,10 +4437,7 @@ xfs_bmapi( end = bno + len; obno = bno; bma.ip = NULL; - if (delta) { - delta->xed_startoff = NULLFILEOFF; - delta->xed_blockcount = 0; - } + while (bno < end && n < *nmap) { /* * Reading past eof, act as though there's a hole @@ -4620,19 +4458,13 @@ xfs_bmapi( * allocate the stuff asked for in this bmap call * but that wouldn't be as good. */ - if (wasdelay && !(flags & XFS_BMAPI_EXACT)) { + if (wasdelay) { alen = (xfs_extlen_t)got.br_blockcount; aoff = got.br_startoff; if (lastx != NULLEXTNUM && lastx) { ep = xfs_iext_get_ext(ifp, lastx - 1); xfs_bmbt_get_all(ep, &prev); } - } else if (wasdelay) { - alen = (xfs_extlen_t) - XFS_FILBLKS_MIN(len, - (got.br_startoff + - got.br_blockcount) - bno); - aoff = bno; } else { alen = (xfs_extlen_t) XFS_FILBLKS_MIN(len, MAXEXTLEN); @@ -4831,7 +4663,7 @@ xfs_bmapi( got.br_state = XFS_EXT_UNWRITTEN; } error = xfs_bmap_add_extent(ip, lastx, &cur, &got, - firstblock, flist, &tmp_logflags, delta, + firstblock, flist, &tmp_logflags, whichfork, (flags & XFS_BMAPI_RSVBLOCKS)); logflags |= tmp_logflags; if (error) @@ -4927,7 +4759,7 @@ xfs_bmapi( } mval->br_state = XFS_EXT_NORM; error = xfs_bmap_add_extent(ip, lastx, &cur, mval, - firstblock, flist, &tmp_logflags, delta, + firstblock, flist, &tmp_logflags, whichfork, (flags & XFS_BMAPI_RSVBLOCKS)); logflags |= tmp_logflags; if (error) @@ -5017,14 +4849,6 @@ xfs_bmapi( ASSERT(XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE || XFS_IFORK_NEXTENTS(ip, whichfork) > ifp->if_ext_max); error = 0; - if (delta && delta->xed_startoff != NULLFILEOFF) { - /* A change was actually made. - * Note that delta->xed_blockount is an offset at this - * point and needs to be converted to a block count. - */ - ASSERT(delta->xed_blockcount > delta->xed_startoff); - delta->xed_blockcount -= delta->xed_startoff; - } error0: /* * Log everything. Do this after conversion, there's no point in @@ -5136,8 +4960,6 @@ xfs_bunmapi( xfs_fsblock_t *firstblock, /* first allocated block controls a.g. for allocs */ xfs_bmap_free_t *flist, /* i/o: list extents to free */ - xfs_extdelta_t *delta, /* o: change made to incore - extents */ int *done) /* set if not done yet */ { xfs_btree_cur_t *cur; /* bmap btree cursor */ @@ -5196,10 +5018,7 @@ xfs_bunmapi( bno = start + len - 1; ep = xfs_bmap_search_extents(ip, bno, whichfork, &eof, &lastx, &got, &prev); - if (delta) { - delta->xed_startoff = NULLFILEOFF; - delta->xed_blockcount = 0; - } + /* * Check to see if the given block number is past the end of the * file, back up to the last block if so... @@ -5297,7 +5116,7 @@ xfs_bunmapi( } del.br_state = XFS_EXT_UNWRITTEN; error = xfs_bmap_add_extent(ip, lastx, &cur, &del, - firstblock, flist, &logflags, delta, + firstblock, flist, &logflags, XFS_DATA_FORK, 0); if (error) goto error0; @@ -5352,7 +5171,7 @@ xfs_bunmapi( prev.br_state = XFS_EXT_UNWRITTEN; error = xfs_bmap_add_extent(ip, lastx - 1, &cur, &prev, firstblock, flist, &logflags, - delta, XFS_DATA_FORK, 0); + XFS_DATA_FORK, 0); if (error) goto error0; goto nodelete; @@ -5361,7 +5180,7 @@ xfs_bunmapi( del.br_state = XFS_EXT_UNWRITTEN; error = xfs_bmap_add_extent(ip, lastx, &cur, &del, firstblock, flist, &logflags, - delta, XFS_DATA_FORK, 0); + XFS_DATA_FORK, 0); if (error) goto error0; goto nodelete; @@ -5414,7 +5233,7 @@ xfs_bunmapi( goto error0; } error = xfs_bmap_del_extent(ip, tp, lastx, flist, cur, &del, - &tmp_logflags, delta, whichfork, rsvd); + &tmp_logflags, whichfork, rsvd); logflags |= tmp_logflags; if (error) goto error0; @@ -5471,14 +5290,6 @@ nodelete: ASSERT(ifp->if_ext_max == XFS_IFORK_SIZE(ip, whichfork) / (uint)sizeof(xfs_bmbt_rec_t)); error = 0; - if (delta && delta->xed_startoff != NULLFILEOFF) { - /* A change was actually made. - * Note that delta->xed_blockount is an offset at this - * point and needs to be converted to a block count. - */ - ASSERT(delta->xed_blockcount > delta->xed_startoff); - delta->xed_blockcount -= delta->xed_startoff; - } error0: /* * Log everything. Do this after conversion, there's no point in @@ -5605,28 +5416,6 @@ xfs_getbmap( prealloced = 0; fixlen = 1LL << 32; } else { - /* - * If the BMV_IF_NO_DMAPI_READ interface bit specified, do - * not generate a DMAPI read event. Otherwise, if the - * DM_EVENT_READ bit is set for the file, generate a read - * event in order that the DMAPI application may do its thing - * before we return the extents. Usually this means restoring - * user file data to regions of the file that look like holes. - * - * The "old behavior" (from XFS_IOC_GETBMAP) is to not specify - * BMV_IF_NO_DMAPI_READ so that read events are generated. - * If this were not true, callers of ioctl(XFS_IOC_GETBMAP) - * could misinterpret holes in a DMAPI file as true holes, - * when in fact they may represent offline user data. - */ - if (DM_EVENT_ENABLED(ip, DM_EVENT_READ) && - !(iflags & BMV_IF_NO_DMAPI_READ)) { - error = XFS_SEND_DATA(mp, DM_EVENT_READ, ip, - 0, 0, 0, NULL); - if (error) - return XFS_ERROR(error); - } - if (ip->i_d.di_format != XFS_DINODE_FMT_EXTENTS && ip->i_d.di_format != XFS_DINODE_FMT_BTREE && ip->i_d.di_format != XFS_DINODE_FMT_LOCAL) @@ -5713,7 +5502,7 @@ xfs_getbmap( error = xfs_bmapi(NULL, ip, XFS_BB_TO_FSBT(mp, bmv->bmv_offset), XFS_BB_TO_FSB(mp, bmv->bmv_length), bmapi_flags, NULL, 0, map, &nmap, - NULL, NULL); + NULL); if (error) goto out_free_map; ASSERT(nmap <= subnex); @@ -5859,66 +5648,34 @@ xfs_bmap_eof( } #ifdef DEBUG -STATIC -xfs_buf_t * +STATIC struct xfs_buf * xfs_bmap_get_bp( - xfs_btree_cur_t *cur, + struct xfs_btree_cur *cur, xfs_fsblock_t bno) { - int i; - xfs_buf_t *bp; + struct xfs_log_item_desc *lidp; + int i; if (!cur) - return(NULL); - - bp = NULL; - for(i = 0; i < XFS_BTREE_MAXLEVELS; i++) { - bp = cur->bc_bufs[i]; - if (!bp) break; - if (XFS_BUF_ADDR(bp) == bno) - break; /* Found it */ - } - if (i == XFS_BTREE_MAXLEVELS) - bp = NULL; - - if (!bp) { /* Chase down all the log items to see if the bp is there */ - xfs_log_item_chunk_t *licp; - xfs_trans_t *tp; - - tp = cur->bc_tp; - licp = &tp->t_items; - while (!bp && licp != NULL) { - if (xfs_lic_are_all_free(licp)) { - licp = licp->lic_next; - continue; - } - for (i = 0; i < licp->lic_unused; i++) { - xfs_log_item_desc_t *lidp; - xfs_log_item_t *lip; - xfs_buf_log_item_t *bip; - xfs_buf_t *lbp; - - if (xfs_lic_isfree(licp, i)) { - continue; - } - - lidp = xfs_lic_slot(licp, i); - lip = lidp->lid_item; - if (lip->li_type != XFS_LI_BUF) - continue; + return NULL; - bip = (xfs_buf_log_item_t *)lip; - lbp = bip->bli_buf; + for (i = 0; i < XFS_BTREE_MAXLEVELS; i++) { + if (!cur->bc_bufs[i]) + break; + if (XFS_BUF_ADDR(cur->bc_bufs[i]) == bno) + return cur->bc_bufs[i]; + } - if (XFS_BUF_ADDR(lbp) == bno) { - bp = lbp; - break; /* Found it */ - } - } - licp = licp->lic_next; - } + /* Chase down all the log items to see if the bp is there */ + list_for_each_entry(lidp, &cur->bc_tp->t_items, lid_trans) { + struct xfs_buf_log_item *bip; + bip = (struct xfs_buf_log_item *)lidp->lid_item; + if (bip->bli_item.li_type == XFS_LI_BUF && + XFS_BUF_ADDR(bip->bli_buf) == bno) + return bip->bli_buf; } - return(bp); + + return NULL; } STATIC void diff --git a/fs/xfs/xfs_bmap.h b/fs/xfs/xfs_bmap.h index 419dafb9d87..b13569a6179 100644 --- a/fs/xfs/xfs_bmap.h +++ b/fs/xfs/xfs_bmap.h @@ -28,20 +28,6 @@ struct xfs_trans; extern kmem_zone_t *xfs_bmap_free_item_zone; /* - * DELTA: describe a change to the in-core extent list. - * - * Internally the use of xed_blockount is somewhat funky. - * xed_blockcount contains an offset much of the time because this - * makes merging changes easier. (xfs_fileoff_t and xfs_filblks_t are - * the same underlying type). - */ -typedef struct xfs_extdelta -{ - xfs_fileoff_t xed_startoff; /* offset of range */ - xfs_filblks_t xed_blockcount; /* blocks in range */ -} xfs_extdelta_t; - -/* * List of extents to be free "later". * The list is kept sorted on xbf_startblock. */ @@ -82,16 +68,13 @@ typedef struct xfs_bmap_free #define XFS_BMAPI_DELAY 0x002 /* delayed write operation */ #define XFS_BMAPI_ENTIRE 0x004 /* return entire extent, not trimmed */ #define XFS_BMAPI_METADATA 0x008 /* mapping metadata not user data */ -#define XFS_BMAPI_EXACT 0x010 /* allocate only to spec'd bounds */ -#define XFS_BMAPI_ATTRFORK 0x020 /* use attribute fork not data */ -#define XFS_BMAPI_ASYNC 0x040 /* bunmapi xactions can be async */ -#define XFS_BMAPI_RSVBLOCKS 0x080 /* OK to alloc. reserved data blocks */ -#define XFS_BMAPI_PREALLOC 0x100 /* preallocation op: unwritten space */ -#define XFS_BMAPI_IGSTATE 0x200 /* Ignore state - */ +#define XFS_BMAPI_ATTRFORK 0x010 /* use attribute fork not data */ +#define XFS_BMAPI_RSVBLOCKS 0x020 /* OK to alloc. reserved data blocks */ +#define XFS_BMAPI_PREALLOC 0x040 /* preallocation op: unwritten space */ +#define XFS_BMAPI_IGSTATE 0x080 /* Ignore state - */ /* combine contig. space */ -#define XFS_BMAPI_CONTIG 0x400 /* must allocate only one extent */ -/* XFS_BMAPI_DIRECT_IO 0x800 */ -#define XFS_BMAPI_CONVERT 0x1000 /* unwritten extent conversion - */ +#define XFS_BMAPI_CONTIG 0x100 /* must allocate only one extent */ +#define XFS_BMAPI_CONVERT 0x200 /* unwritten extent conversion - */ /* need write cache flushing and no */ /* additional allocation alignments */ @@ -100,9 +83,7 @@ typedef struct xfs_bmap_free { XFS_BMAPI_DELAY, "DELAY" }, \ { XFS_BMAPI_ENTIRE, "ENTIRE" }, \ { XFS_BMAPI_METADATA, "METADATA" }, \ - { XFS_BMAPI_EXACT, "EXACT" }, \ { XFS_BMAPI_ATTRFORK, "ATTRFORK" }, \ - { XFS_BMAPI_ASYNC, "ASYNC" }, \ { XFS_BMAPI_RSVBLOCKS, "RSVBLOCKS" }, \ { XFS_BMAPI_PREALLOC, "PREALLOC" }, \ { XFS_BMAPI_IGSTATE, "IGSTATE" }, \ @@ -310,9 +291,7 @@ xfs_bmapi( xfs_extlen_t total, /* total blocks needed */ struct xfs_bmbt_irec *mval, /* output: map values */ int *nmap, /* i/o: mval size/count */ - xfs_bmap_free_t *flist, /* i/o: list extents to free */ - xfs_extdelta_t *delta); /* o: change made to incore - extents */ + xfs_bmap_free_t *flist); /* i/o: list extents to free */ /* * Map file blocks to filesystem blocks, simple version. @@ -346,8 +325,6 @@ xfs_bunmapi( xfs_fsblock_t *firstblock, /* first allocated block controls a.g. for allocs */ xfs_bmap_free_t *flist, /* i/o: list extents to free */ - xfs_extdelta_t *delta, /* o: change made to incore - extents */ int *done); /* set if not done yet */ /* diff --git a/fs/xfs/xfs_bmap_btree.c b/fs/xfs/xfs_bmap_btree.c index 416e47e54b8..87d3c10b695 100644 --- a/fs/xfs/xfs_bmap_btree.c +++ b/fs/xfs/xfs_bmap_btree.c @@ -24,21 +24,16 @@ #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_dir2_sf.h" -#include "xfs_attr_sf.h" #include "xfs_dinode.h" #include "xfs_inode.h" #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" diff --git a/fs/xfs/xfs_btree.c b/fs/xfs/xfs_btree.c index 96be4b0f249..829af92f0fb 100644 --- a/fs/xfs/xfs_btree.c +++ b/fs/xfs/xfs_btree.c @@ -24,20 +24,15 @@ #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_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_btree_trace.h" -#include "xfs_ialloc.h" #include "xfs_error.h" #include "xfs_trace.h" diff --git a/fs/xfs/xfs_buf_item.c b/fs/xfs/xfs_buf_item.c index 02a80984aa0..1b09d7a280d 100644 --- a/fs/xfs/xfs_buf_item.c +++ b/fs/xfs/xfs_buf_item.c @@ -24,7 +24,6 @@ #include "xfs_trans.h" #include "xfs_sb.h" #include "xfs_ag.h" -#include "xfs_dmapi.h" #include "xfs_mount.h" #include "xfs_buf_item.h" #include "xfs_trans_priv.h" @@ -34,6 +33,12 @@ kmem_zone_t *xfs_buf_item_zone; +static inline struct xfs_buf_log_item *BUF_ITEM(struct xfs_log_item *lip) +{ + return container_of(lip, struct xfs_buf_log_item, bli_item); +} + + #ifdef XFS_TRANS_DEBUG /* * This function uses an alternate strategy for tracking the bytes @@ -151,12 +156,13 @@ STATIC void xfs_buf_do_callbacks(xfs_buf_t *bp, xfs_log_item_t *lip); */ STATIC uint xfs_buf_item_size( - xfs_buf_log_item_t *bip) + struct xfs_log_item *lip) { - uint nvecs; - int next_bit; - int last_bit; - xfs_buf_t *bp; + struct xfs_buf_log_item *bip = BUF_ITEM(lip); + struct xfs_buf *bp = bip->bli_buf; + uint nvecs; + int next_bit; + int last_bit; ASSERT(atomic_read(&bip->bli_refcount) > 0); if (bip->bli_flags & XFS_BLI_STALE) { @@ -170,7 +176,6 @@ xfs_buf_item_size( return 1; } - bp = bip->bli_buf; ASSERT(bip->bli_flags & XFS_BLI_LOGGED); nvecs = 1; last_bit = xfs_next_bit(bip->bli_format.blf_data_map, @@ -219,13 +224,13 @@ xfs_buf_item_size( */ STATIC void xfs_buf_item_format( - xfs_buf_log_item_t *bip, - xfs_log_iovec_t *log_vector) + struct xfs_log_item *lip, + struct xfs_log_iovec *vecp) { + struct xfs_buf_log_item *bip = BUF_ITEM(lip); + struct xfs_buf *bp = bip->bli_buf; uint base_size; uint nvecs; - xfs_log_iovec_t *vecp; - xfs_buf_t *bp; int first_bit; int last_bit; int next_bit; @@ -235,8 +240,6 @@ xfs_buf_item_format( ASSERT(atomic_read(&bip->bli_refcount) > 0); ASSERT((bip->bli_flags & XFS_BLI_LOGGED) || (bip->bli_flags & XFS_BLI_STALE)); - bp = bip->bli_buf; - vecp = log_vector; /* * The size of the base structure is the size of the @@ -248,7 +251,7 @@ xfs_buf_item_format( base_size = (uint)(sizeof(xfs_buf_log_format_t) + ((bip->bli_format.blf_map_size - 1) * sizeof(uint))); - vecp->i_addr = (xfs_caddr_t)&bip->bli_format; + vecp->i_addr = &bip->bli_format; vecp->i_len = base_size; vecp->i_type = XLOG_REG_TYPE_BFORMAT; vecp++; @@ -263,7 +266,7 @@ xfs_buf_item_format( */ if (bip->bli_flags & XFS_BLI_INODE_BUF) { if (!((bip->bli_flags & XFS_BLI_INODE_ALLOC_BUF) && - xfs_log_item_in_current_chkpt(&bip->bli_item))) + xfs_log_item_in_current_chkpt(lip))) bip->bli_format.blf_flags |= XFS_BLF_INODE_BUF; bip->bli_flags &= ~XFS_BLI_INODE_BUF; } @@ -356,66 +359,90 @@ xfs_buf_item_format( /* * This is called to pin the buffer associated with the buf log item in memory - * so it cannot be written out. Simply call bpin() on the buffer to do this. + * so it cannot be written out. * * We also always take a reference to the buffer log item here so that the bli * is held while the item is pinned in memory. This means that we can * unconditionally drop the reference count a transaction holds when the * transaction is completed. */ - STATIC void xfs_buf_item_pin( - xfs_buf_log_item_t *bip) + struct xfs_log_item *lip) { - xfs_buf_t *bp; + struct xfs_buf_log_item *bip = BUF_ITEM(lip); - bp = bip->bli_buf; - ASSERT(XFS_BUF_ISBUSY(bp)); + ASSERT(XFS_BUF_ISBUSY(bip->bli_buf)); ASSERT(atomic_read(&bip->bli_refcount) > 0); ASSERT((bip->bli_flags & XFS_BLI_LOGGED) || (bip->bli_flags & XFS_BLI_STALE)); - atomic_inc(&bip->bli_refcount); + trace_xfs_buf_item_pin(bip); - xfs_bpin(bp); -} + atomic_inc(&bip->bli_refcount); + atomic_inc(&bip->bli_buf->b_pin_count); +} /* * This is called to unpin the buffer associated with the buf log * item which was previously pinned with a call to xfs_buf_item_pin(). - * Just call bunpin() on the buffer to do this. * * Also drop the reference to the buf item for the current transaction. * If the XFS_BLI_STALE flag is set and we are the last reference, * then free up the buf log item and unlock the buffer. + * + * If the remove flag is set we are called from uncommit in the + * forced-shutdown path. If that is true and the reference count on + * the log item is going to drop to zero we need to free the item's + * descriptor in the transaction. */ STATIC void xfs_buf_item_unpin( - xfs_buf_log_item_t *bip) + struct xfs_log_item *lip, + int remove) { - struct xfs_ail *ailp; - xfs_buf_t *bp; - int freed; + struct xfs_buf_log_item *bip = BUF_ITEM(lip); + xfs_buf_t *bp = bip->bli_buf; + struct xfs_ail *ailp = lip->li_ailp; int stale = bip->bli_flags & XFS_BLI_STALE; + int freed; - bp = bip->bli_buf; - ASSERT(bp != NULL); ASSERT(XFS_BUF_FSPRIVATE(bp, xfs_buf_log_item_t *) == bip); ASSERT(atomic_read(&bip->bli_refcount) > 0); + trace_xfs_buf_item_unpin(bip); freed = atomic_dec_and_test(&bip->bli_refcount); - ailp = bip->bli_item.li_ailp; - xfs_bunpin(bp); + + if (atomic_dec_and_test(&bp->b_pin_count)) + wake_up_all(&bp->b_waiters); + if (freed && stale) { ASSERT(bip->bli_flags & XFS_BLI_STALE); ASSERT(XFS_BUF_VALUSEMA(bp) <= 0); ASSERT(!(XFS_BUF_ISDELAYWRITE(bp))); ASSERT(XFS_BUF_ISSTALE(bp)); ASSERT(bip->bli_format.blf_flags & XFS_BLF_CANCEL); + trace_xfs_buf_item_unpin_stale(bip); + if (remove) { + /* + * We have to remove the log item from the transaction + * as we are about to release our reference to the + * buffer. If we don't, the unlock that occurs later + * in xfs_trans_uncommit() will ry to reference the + * buffer which we no longer have a hold on. + */ + xfs_trans_del_item(lip); + + /* + * Since the transaction no longer refers to the buffer, + * the buffer should no longer refer to the transaction. + */ + XFS_BUF_SET_FSPRIVATE2(bp, NULL); + } + /* * If we get called here because of an IO error, we may * or may not have the item on the AIL. xfs_trans_ail_delete() @@ -437,48 +464,6 @@ xfs_buf_item_unpin( } /* - * this is called from uncommit in the forced-shutdown path. - * we need to check to see if the reference count on the log item - * is going to drop to zero. If so, unpin will free the log item - * so we need to free the item's descriptor (that points to the item) - * in the transaction. - */ -STATIC void -xfs_buf_item_unpin_remove( - xfs_buf_log_item_t *bip, - xfs_trans_t *tp) -{ - /* will xfs_buf_item_unpin() call xfs_buf_item_relse()? */ - if ((atomic_read(&bip->bli_refcount) == 1) && - (bip->bli_flags & XFS_BLI_STALE)) { - /* - * yes -- We can safely do some work here and then call - * buf_item_unpin to do the rest because we are - * are holding the buffer locked so no one else will be - * able to bump up the refcount. We have to remove the - * log item from the transaction as we are about to release - * our reference to the buffer. If we don't, the unlock that - * occurs later in the xfs_trans_uncommit() will try to - * reference the buffer which we no longer have a hold on. - */ - struct xfs_log_item_desc *lidp; - - ASSERT(XFS_BUF_VALUSEMA(bip->bli_buf) <= 0); - trace_xfs_buf_item_unpin_stale(bip); - - lidp = xfs_trans_find_item(tp, (xfs_log_item_t *)bip); - xfs_trans_free_item(tp, lidp); - - /* - * Since the transaction no longer refers to the buffer, the - * buffer should no longer refer to the transaction. - */ - XFS_BUF_SET_FSPRIVATE2(bip->bli_buf, NULL); - } - xfs_buf_item_unpin(bip); -} - -/* * This is called to attempt to lock the buffer associated with this * buf log item. Don't sleep on the buffer lock. If we can't get * the lock right away, return 0. If we can get the lock, take a @@ -488,11 +473,11 @@ xfs_buf_item_unpin_remove( */ STATIC uint xfs_buf_item_trylock( - xfs_buf_log_item_t *bip) + struct xfs_log_item *lip) { - xfs_buf_t *bp; + struct xfs_buf_log_item *bip = BUF_ITEM(lip); + struct xfs_buf *bp = bip->bli_buf; - bp = bip->bli_buf; if (XFS_BUF_ISPINNED(bp)) return XFS_ITEM_PINNED; if (!XFS_BUF_CPSEMA(bp)) @@ -529,13 +514,12 @@ xfs_buf_item_trylock( */ STATIC void xfs_buf_item_unlock( - xfs_buf_log_item_t *bip) + struct xfs_log_item *lip) { - int aborted; - xfs_buf_t *bp; - uint hold; - - bp = bip->bli_buf; + struct xfs_buf_log_item *bip = BUF_ITEM(lip); + struct xfs_buf *bp = bip->bli_buf; + int aborted; + uint hold; /* Clear the buffer's association with this transaction. */ XFS_BUF_SET_FSPRIVATE2(bp, NULL); @@ -546,7 +530,7 @@ xfs_buf_item_unlock( * (cancelled) buffers at unpin time, but we'll never go through the * pin/unpin cycle if we abort inside commit. */ - aborted = (bip->bli_item.li_flags & XFS_LI_ABORTED) != 0; + aborted = (lip->li_flags & XFS_LI_ABORTED) != 0; /* * Before possibly freeing the buf item, determine if we should @@ -607,16 +591,16 @@ xfs_buf_item_unlock( */ STATIC xfs_lsn_t xfs_buf_item_committed( - xfs_buf_log_item_t *bip, + struct xfs_log_item *lip, xfs_lsn_t lsn) { + struct xfs_buf_log_item *bip = BUF_ITEM(lip); + trace_xfs_buf_item_committed(bip); - if ((bip->bli_flags & XFS_BLI_INODE_ALLOC_BUF) && - (bip->bli_item.li_lsn != 0)) { - return bip->bli_item.li_lsn; - } - return (lsn); + if ((bip->bli_flags & XFS_BLI_INODE_ALLOC_BUF) && lip->li_lsn != 0) + return lip->li_lsn; + return lsn; } /* @@ -626,15 +610,16 @@ xfs_buf_item_committed( */ STATIC void xfs_buf_item_push( - xfs_buf_log_item_t *bip) + struct xfs_log_item *lip) { - xfs_buf_t *bp; + struct xfs_buf_log_item *bip = BUF_ITEM(lip); + struct xfs_buf *bp = bip->bli_buf; ASSERT(!(bip->bli_flags & XFS_BLI_STALE)); + ASSERT(!XFS_BUF_ISDELAYWRITE(bp)); + trace_xfs_buf_item_push(bip); - bp = bip->bli_buf; - ASSERT(!XFS_BUF_ISDELAYWRITE(bp)); xfs_buf_relse(bp); } @@ -646,22 +631,24 @@ xfs_buf_item_push( */ STATIC void xfs_buf_item_pushbuf( - xfs_buf_log_item_t *bip) + struct xfs_log_item *lip) { - xfs_buf_t *bp; + struct xfs_buf_log_item *bip = BUF_ITEM(lip); + struct xfs_buf *bp = bip->bli_buf; ASSERT(!(bip->bli_flags & XFS_BLI_STALE)); + ASSERT(XFS_BUF_ISDELAYWRITE(bp)); + trace_xfs_buf_item_pushbuf(bip); - bp = bip->bli_buf; - ASSERT(XFS_BUF_ISDELAYWRITE(bp)); xfs_buf_delwri_promote(bp); xfs_buf_relse(bp); } -/* ARGSUSED */ STATIC void -xfs_buf_item_committing(xfs_buf_log_item_t *bip, xfs_lsn_t commit_lsn) +xfs_buf_item_committing( + struct xfs_log_item *lip, + xfs_lsn_t commit_lsn) { } @@ -669,21 +656,16 @@ xfs_buf_item_committing(xfs_buf_log_item_t *bip, xfs_lsn_t commit_lsn) * This is the ops vector shared by all buf log items. */ static struct xfs_item_ops xfs_buf_item_ops = { - .iop_size = (uint(*)(xfs_log_item_t*))xfs_buf_item_size, - .iop_format = (void(*)(xfs_log_item_t*, xfs_log_iovec_t*)) - xfs_buf_item_format, - .iop_pin = (void(*)(xfs_log_item_t*))xfs_buf_item_pin, - .iop_unpin = (void(*)(xfs_log_item_t*))xfs_buf_item_unpin, - .iop_unpin_remove = (void(*)(xfs_log_item_t*, xfs_trans_t *)) - xfs_buf_item_unpin_remove, - .iop_trylock = (uint(*)(xfs_log_item_t*))xfs_buf_item_trylock, - .iop_unlock = (void(*)(xfs_log_item_t*))xfs_buf_item_unlock, - .iop_committed = (xfs_lsn_t(*)(xfs_log_item_t*, xfs_lsn_t)) - xfs_buf_item_committed, - .iop_push = (void(*)(xfs_log_item_t*))xfs_buf_item_push, - .iop_pushbuf = (void(*)(xfs_log_item_t*))xfs_buf_item_pushbuf, - .iop_committing = (void(*)(xfs_log_item_t*, xfs_lsn_t)) - xfs_buf_item_committing + .iop_size = xfs_buf_item_size, + .iop_format = xfs_buf_item_format, + .iop_pin = xfs_buf_item_pin, + .iop_unpin = xfs_buf_item_unpin, + .iop_trylock = xfs_buf_item_trylock, + .iop_unlock = xfs_buf_item_unlock, + .iop_committed = xfs_buf_item_committed, + .iop_push = xfs_buf_item_push, + .iop_pushbuf = xfs_buf_item_pushbuf, + .iop_committing = xfs_buf_item_committing }; @@ -712,7 +694,6 @@ xfs_buf_item_init( */ 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 *); if (lip->li_type == XFS_LI_BUF) { @@ -1098,15 +1079,14 @@ xfs_buf_error_relse( * It is called by xfs_buf_iodone_callbacks() above which will take * care of cleaning up the buffer itself. */ -/* ARGSUSED */ void xfs_buf_iodone( - xfs_buf_t *bp, - xfs_buf_log_item_t *bip) + struct xfs_buf *bp, + struct xfs_log_item *lip) { - struct xfs_ail *ailp = bip->bli_item.li_ailp; + struct xfs_ail *ailp = lip->li_ailp; - ASSERT(bip->bli_buf == bp); + ASSERT(BUF_ITEM(lip)->bli_buf == bp); xfs_buf_rele(bp); @@ -1120,6 +1100,6 @@ xfs_buf_iodone( * Either way, AIL is useless if we're forcing a shutdown. */ spin_lock(&ailp->xa_lock); - xfs_trans_ail_delete(ailp, (xfs_log_item_t *)bip); - xfs_buf_item_free(bip); + xfs_trans_ail_delete(ailp, lip); + xfs_buf_item_free(BUF_ITEM(lip)); } diff --git a/fs/xfs/xfs_buf_item.h b/fs/xfs/xfs_buf_item.h index f20bb472d58..0e2ed43f16c 100644 --- a/fs/xfs/xfs_buf_item.h +++ b/fs/xfs/xfs_buf_item.h @@ -124,7 +124,7 @@ void xfs_buf_attach_iodone(struct xfs_buf *, void(*)(struct xfs_buf *, xfs_log_item_t *), xfs_log_item_t *); void xfs_buf_iodone_callbacks(struct xfs_buf *); -void xfs_buf_iodone(struct xfs_buf *, xfs_buf_log_item_t *); +void xfs_buf_iodone(struct xfs_buf *, struct xfs_log_item *); #ifdef XFS_TRANS_DEBUG void diff --git a/fs/xfs/xfs_da_btree.c b/fs/xfs/xfs_da_btree.c index 0ca556b4bf3..30fa0e206fb 100644 --- a/fs/xfs/xfs_da_btree.c +++ b/fs/xfs/xfs_da_btree.c @@ -25,19 +25,14 @@ #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_alloc_btree.h" -#include "xfs_ialloc_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_alloc.h" -#include "xfs_btree.h" #include "xfs_bmap.h" #include "xfs_attr.h" #include "xfs_attr_leaf.h" @@ -581,16 +576,14 @@ xfs_da_node_add(xfs_da_state_t *state, xfs_da_state_blk_t *oldblk, xfs_da_intnode_t *node; xfs_da_node_entry_t *btree; int tmp; - xfs_mount_t *mp; node = oldblk->bp->data; - mp = state->mp; ASSERT(be16_to_cpu(node->hdr.info.magic) == XFS_DA_NODE_MAGIC); ASSERT((oldblk->index >= 0) && (oldblk->index <= be16_to_cpu(node->hdr.count))); ASSERT(newblk->blkno != 0); if (state->args->whichfork == XFS_DATA_FORK) - ASSERT(newblk->blkno >= mp->m_dirleafblk && - newblk->blkno < mp->m_dirfreeblk); + ASSERT(newblk->blkno >= state->mp->m_dirleafblk && + newblk->blkno < state->mp->m_dirfreeblk); /* * We may need to make some room before we insert the new node. @@ -1601,7 +1594,7 @@ xfs_da_grow_inode(xfs_da_args_t *args, xfs_dablk_t *new_blkno) xfs_bmapi_aflag(w)|XFS_BMAPI_WRITE|XFS_BMAPI_METADATA| XFS_BMAPI_CONTIG, args->firstblock, args->total, &map, &nmap, - args->flist, NULL))) { + args->flist))) { return error; } ASSERT(nmap <= 1); @@ -1622,8 +1615,7 @@ xfs_da_grow_inode(xfs_da_args_t *args, xfs_dablk_t *new_blkno) xfs_bmapi_aflag(w)|XFS_BMAPI_WRITE| XFS_BMAPI_METADATA, args->firstblock, args->total, - &mapp[mapi], &nmap, args->flist, - NULL))) { + &mapp[mapi], &nmap, args->flist))) { kmem_free(mapp); return error; } @@ -1884,7 +1876,7 @@ xfs_da_shrink_inode(xfs_da_args_t *args, xfs_dablk_t dead_blkno, */ if ((error = xfs_bunmapi(tp, dp, dead_blkno, count, xfs_bmapi_aflag(w)|XFS_BMAPI_METADATA, - 0, args->firstblock, args->flist, NULL, + 0, args->firstblock, args->flist, &done)) == ENOSPC) { if (w != XFS_DATA_FORK) break; @@ -1989,7 +1981,7 @@ xfs_da_do_buf( nfsb, XFS_BMAPI_METADATA | xfs_bmapi_aflag(whichfork), - NULL, 0, mapp, &nmap, NULL, NULL))) + NULL, 0, mapp, &nmap, NULL))) goto exit0; } } else { diff --git a/fs/xfs/xfs_dfrag.c b/fs/xfs/xfs_dfrag.c index 7f159d2a429..3b9582c60a2 100644 --- a/fs/xfs/xfs_dfrag.c +++ b/fs/xfs/xfs_dfrag.c @@ -24,24 +24,15 @@ #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_dir2_sf.h" -#include "xfs_attr_sf.h" #include "xfs_dinode.h" #include "xfs_inode.h" #include "xfs_inode_item.h" #include "xfs_bmap.h" -#include "xfs_btree.h" -#include "xfs_ialloc.h" #include "xfs_itable.h" #include "xfs_dfrag.h" #include "xfs_error.h" -#include "xfs_rw.h" #include "xfs_vnodeops.h" #include "xfs_trace.h" @@ -425,11 +416,8 @@ xfs_swap_extents( } - IHOLD(ip); - xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL); - - IHOLD(tip); - xfs_trans_ijoin(tp, tip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL); + xfs_trans_ijoin_ref(tp, ip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL); + xfs_trans_ijoin_ref(tp, tip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL); xfs_trans_log_inode(tp, ip, ilf_fields); xfs_trans_log_inode(tp, tip, tilf_fields); diff --git a/fs/xfs/xfs_dir2.c b/fs/xfs/xfs_dir2.c index 42520f04126..a1321bc7f19 100644 --- a/fs/xfs/xfs_dir2.c +++ b/fs/xfs/xfs_dir2.c @@ -25,13 +25,11 @@ #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_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" @@ -382,7 +380,7 @@ xfs_readdir( int rval; /* return value */ int v; /* type-checking value */ - xfs_itrace_entry(dp); + trace_xfs_readdir(dp); if (XFS_FORCED_SHUTDOWN(dp->i_mount)) return XFS_ERROR(EIO); @@ -549,7 +547,7 @@ xfs_dir2_grow_inode( if ((error = xfs_bmapi(tp, dp, bno, count, XFS_BMAPI_WRITE|XFS_BMAPI_METADATA|XFS_BMAPI_CONTIG, args->firstblock, args->total, &map, &nmap, - args->flist, NULL))) + args->flist))) return error; ASSERT(nmap <= 1); if (nmap == 1) { @@ -581,8 +579,7 @@ xfs_dir2_grow_inode( if ((error = xfs_bmapi(tp, dp, b, c, XFS_BMAPI_WRITE|XFS_BMAPI_METADATA, args->firstblock, args->total, - &mapp[mapi], &nmap, args->flist, - NULL))) { + &mapp[mapi], &nmap, args->flist))) { kmem_free(mapp); return error; } @@ -715,7 +712,7 @@ xfs_dir2_shrink_inode( */ if ((error = xfs_bunmapi(tp, dp, da, mp->m_dirblkfsbs, XFS_BMAPI_METADATA, 0, args->firstblock, args->flist, - NULL, &done))) { + &done))) { /* * ENOSPC actually can happen if we're in a removename with * no space reservation, and the resulting block removal diff --git a/fs/xfs/xfs_dir2_block.c b/fs/xfs/xfs_dir2_block.c index 779a267b0a8..580d99cef9e 100644 --- a/fs/xfs/xfs_dir2_block.c +++ b/fs/xfs/xfs_dir2_block.c @@ -24,12 +24,10 @@ #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_dir2_sf.h" -#include "xfs_attr_sf.h" #include "xfs_dinode.h" #include "xfs_inode.h" #include "xfs_inode_item.h" @@ -1073,10 +1071,10 @@ xfs_dir2_sf_to_block( */ buf_len = dp->i_df.if_bytes; - buf = kmem_alloc(dp->i_df.if_bytes, KM_SLEEP); + buf = kmem_alloc(buf_len, KM_SLEEP); - memcpy(buf, sfp, dp->i_df.if_bytes); - xfs_idata_realloc(dp, -dp->i_df.if_bytes, XFS_DATA_FORK); + memcpy(buf, sfp, buf_len); + xfs_idata_realloc(dp, -buf_len, XFS_DATA_FORK); dp->i_d.di_size = 0; xfs_trans_log_inode(tp, dp, XFS_ILOG_CORE); /* diff --git a/fs/xfs/xfs_dir2_data.c b/fs/xfs/xfs_dir2_data.c index 498f8d69433..921595b84f5 100644 --- a/fs/xfs/xfs_dir2_data.c +++ b/fs/xfs/xfs_dir2_data.c @@ -24,12 +24,10 @@ #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_dir2_sf.h" -#include "xfs_attr_sf.h" #include "xfs_dinode.h" #include "xfs_inode.h" #include "xfs_dir2_data.h" diff --git a/fs/xfs/xfs_dir2_leaf.c b/fs/xfs/xfs_dir2_leaf.c index e2d89854ec9..504be8640e9 100644 --- a/fs/xfs/xfs_dir2_leaf.c +++ b/fs/xfs/xfs_dir2_leaf.c @@ -25,11 +25,9 @@ #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_attr_sf.h" #include "xfs_dir2_sf.h" #include "xfs_dinode.h" #include "xfs_inode.h" @@ -875,7 +873,7 @@ xfs_dir2_leaf_getdents( xfs_dir2_byte_to_da(mp, XFS_DIR2_LEAF_OFFSET) - map_off, XFS_BMAPI_METADATA, NULL, 0, - &map[map_valid], &nmap, NULL, NULL); + &map[map_valid], &nmap, NULL); /* * Don't know if we should ignore this or * try to return an error. diff --git a/fs/xfs/xfs_dir2_node.c b/fs/xfs/xfs_dir2_node.c index 78fc4d9ae75..f9a0864b696 100644 --- a/fs/xfs/xfs_dir2_node.c +++ b/fs/xfs/xfs_dir2_node.c @@ -24,12 +24,10 @@ #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_dir2_sf.h" -#include "xfs_attr_sf.h" #include "xfs_dinode.h" #include "xfs_inode.h" #include "xfs_bmap.h" diff --git a/fs/xfs/xfs_dir2_sf.c b/fs/xfs/xfs_dir2_sf.c index c1a5945d463..b1bae6b1eed 100644 --- a/fs/xfs/xfs_dir2_sf.c +++ b/fs/xfs/xfs_dir2_sf.c @@ -24,12 +24,10 @@ #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_dir2_sf.h" -#include "xfs_attr_sf.h" #include "xfs_dinode.h" #include "xfs_inode.h" #include "xfs_inode_item.h" diff --git a/fs/xfs/xfs_dmapi.h b/fs/xfs/xfs_dmapi.h deleted file mode 100644 index 2813cdd7237..00000000000 --- a/fs/xfs/xfs_dmapi.h +++ /dev/null @@ -1,170 +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_DMAPI_H__ -#define __XFS_DMAPI_H__ - -/* Values used to define the on-disk version of dm_attrname_t. All - * on-disk attribute names start with the 8-byte string "SGI_DMI_". - * - * In the on-disk inode, DMAPI attribute names consist of the user-provided - * name with the DMATTR_PREFIXSTRING pre-pended. This string must NEVER be - * changed. - */ - -#define DMATTR_PREFIXLEN 8 -#define DMATTR_PREFIXSTRING "SGI_DMI_" - -typedef enum { - DM_EVENT_INVALID = -1, - DM_EVENT_CANCEL = 0, /* not supported */ - DM_EVENT_MOUNT = 1, - DM_EVENT_PREUNMOUNT = 2, - DM_EVENT_UNMOUNT = 3, - DM_EVENT_DEBUT = 4, /* not supported */ - DM_EVENT_CREATE = 5, - DM_EVENT_CLOSE = 6, /* not supported */ - DM_EVENT_POSTCREATE = 7, - DM_EVENT_REMOVE = 8, - DM_EVENT_POSTREMOVE = 9, - DM_EVENT_RENAME = 10, - DM_EVENT_POSTRENAME = 11, - DM_EVENT_LINK = 12, - DM_EVENT_POSTLINK = 13, - DM_EVENT_SYMLINK = 14, - DM_EVENT_POSTSYMLINK = 15, - DM_EVENT_READ = 16, - DM_EVENT_WRITE = 17, - DM_EVENT_TRUNCATE = 18, - DM_EVENT_ATTRIBUTE = 19, - DM_EVENT_DESTROY = 20, - DM_EVENT_NOSPACE = 21, - DM_EVENT_USER = 22, - DM_EVENT_MAX = 23 -} dm_eventtype_t; -#define HAVE_DM_EVENTTYPE_T - -typedef enum { - DM_RIGHT_NULL, - DM_RIGHT_SHARED, - DM_RIGHT_EXCL -} dm_right_t; -#define HAVE_DM_RIGHT_T - -/* Defines for determining if an event message should be sent. */ -#ifdef HAVE_DMAPI -#define DM_EVENT_ENABLED(ip, event) ( \ - unlikely ((ip)->i_mount->m_flags & XFS_MOUNT_DMAPI) && \ - ( ((ip)->i_d.di_dmevmask & (1 << event)) || \ - ((ip)->i_mount->m_dmevmask & (1 << event)) ) \ - ) -#else -#define DM_EVENT_ENABLED(ip, event) (0) -#endif - -#define DM_XFS_VALID_FS_EVENTS ( \ - (1 << DM_EVENT_PREUNMOUNT) | \ - (1 << DM_EVENT_UNMOUNT) | \ - (1 << DM_EVENT_NOSPACE) | \ - (1 << DM_EVENT_DEBUT) | \ - (1 << DM_EVENT_CREATE) | \ - (1 << DM_EVENT_POSTCREATE) | \ - (1 << DM_EVENT_REMOVE) | \ - (1 << DM_EVENT_POSTREMOVE) | \ - (1 << DM_EVENT_RENAME) | \ - (1 << DM_EVENT_POSTRENAME) | \ - (1 << DM_EVENT_LINK) | \ - (1 << DM_EVENT_POSTLINK) | \ - (1 << DM_EVENT_SYMLINK) | \ - (1 << DM_EVENT_POSTSYMLINK) | \ - (1 << DM_EVENT_ATTRIBUTE) | \ - (1 << DM_EVENT_DESTROY) ) - -/* Events valid in dm_set_eventlist() when called with a file handle for - a regular file or a symlink. These events are persistent. -*/ - -#define DM_XFS_VALID_FILE_EVENTS ( \ - (1 << DM_EVENT_ATTRIBUTE) | \ - (1 << DM_EVENT_DESTROY) ) - -/* Events valid in dm_set_eventlist() when called with a file handle for - a directory. These events are persistent. -*/ - -#define DM_XFS_VALID_DIRECTORY_EVENTS ( \ - (1 << DM_EVENT_CREATE) | \ - (1 << DM_EVENT_POSTCREATE) | \ - (1 << DM_EVENT_REMOVE) | \ - (1 << DM_EVENT_POSTREMOVE) | \ - (1 << DM_EVENT_RENAME) | \ - (1 << DM_EVENT_POSTRENAME) | \ - (1 << DM_EVENT_LINK) | \ - (1 << DM_EVENT_POSTLINK) | \ - (1 << DM_EVENT_SYMLINK) | \ - (1 << DM_EVENT_POSTSYMLINK) | \ - (1 << DM_EVENT_ATTRIBUTE) | \ - (1 << DM_EVENT_DESTROY) ) - -/* Events supported by the XFS filesystem. */ -#define DM_XFS_SUPPORTED_EVENTS ( \ - (1 << DM_EVENT_MOUNT) | \ - (1 << DM_EVENT_PREUNMOUNT) | \ - (1 << DM_EVENT_UNMOUNT) | \ - (1 << DM_EVENT_NOSPACE) | \ - (1 << DM_EVENT_CREATE) | \ - (1 << DM_EVENT_POSTCREATE) | \ - (1 << DM_EVENT_REMOVE) | \ - (1 << DM_EVENT_POSTREMOVE) | \ - (1 << DM_EVENT_RENAME) | \ - (1 << DM_EVENT_POSTRENAME) | \ - (1 << DM_EVENT_LINK) | \ - (1 << DM_EVENT_POSTLINK) | \ - (1 << DM_EVENT_SYMLINK) | \ - (1 << DM_EVENT_POSTSYMLINK) | \ - (1 << DM_EVENT_READ) | \ - (1 << DM_EVENT_WRITE) | \ - (1 << DM_EVENT_TRUNCATE) | \ - (1 << DM_EVENT_ATTRIBUTE) | \ - (1 << DM_EVENT_DESTROY) ) - - -/* - * Definitions used for the flags field on dm_send_*_event(). - */ - -#define DM_FLAGS_NDELAY 0x001 /* return EAGAIN after dm_pending() */ -#define DM_FLAGS_UNWANTED 0x002 /* event not in fsys dm_eventset_t */ -#define DM_FLAGS_IMUX 0x004 /* thread holds i_mutex */ -#define DM_FLAGS_IALLOCSEM_RD 0x010 /* thread holds i_alloc_sem rd */ -#define DM_FLAGS_IALLOCSEM_WR 0x020 /* thread holds i_alloc_sem wr */ - -/* - * Pull in platform specific event flags defines - */ -#include "xfs_dmapi_priv.h" - -/* - * Macros to turn caller specified delay/block flags into - * dm_send_xxxx_event flag DM_FLAGS_NDELAY. - */ - -#define FILP_DELAY_FLAG(filp) ((filp->f_flags&(O_NDELAY|O_NONBLOCK)) ? \ - DM_FLAGS_NDELAY : 0) -#define AT_DELAY_FLAG(f) ((f & XFS_ATTR_NONBLOCK) ? DM_FLAGS_NDELAY : 0) - -#endif /* __XFS_DMAPI_H__ */ diff --git a/fs/xfs/xfs_dmops.c b/fs/xfs/xfs_dmops.c deleted file mode 100644 index e71e2581c0c..00000000000 --- a/fs/xfs/xfs_dmops.c +++ /dev/null @@ -1,55 +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_fs.h" -#include "xfs_types.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" - - -static struct xfs_dmops xfs_dmcore_stub = { - .xfs_send_data = (xfs_send_data_t)fs_nosys, - .xfs_send_mmap = (xfs_send_mmap_t)fs_noerr, - .xfs_send_destroy = (xfs_send_destroy_t)fs_nosys, - .xfs_send_namesp = (xfs_send_namesp_t)fs_nosys, - .xfs_send_mount = (xfs_send_mount_t)fs_nosys, - .xfs_send_unmount = (xfs_send_unmount_t)fs_noerr, -}; - -int -xfs_dmops_get(struct xfs_mount *mp) -{ - if (mp->m_flags & XFS_MOUNT_DMAPI) { - cmn_err(CE_WARN, - "XFS: dmapi support not available in this kernel."); - return EINVAL; - } - - mp->m_dm_ops = &xfs_dmcore_stub; - return 0; -} - -void -xfs_dmops_put(struct xfs_mount *mp) -{ -} diff --git a/fs/xfs/xfs_error.c b/fs/xfs/xfs_error.c index 047b8a8e5c2..ed999026766 100644 --- a/fs/xfs/xfs_error.c +++ b/fs/xfs/xfs_error.c @@ -23,12 +23,8 @@ #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_dir2_sf.h" -#include "xfs_attr_sf.h" #include "xfs_dinode.h" #include "xfs_inode.h" #include "xfs_utils.h" diff --git a/fs/xfs/xfs_extfree_item.c b/fs/xfs/xfs_extfree_item.c index 409fe81585f..a55e687bf56 100644 --- a/fs/xfs/xfs_extfree_item.c +++ b/fs/xfs/xfs_extfree_item.c @@ -24,7 +24,6 @@ #include "xfs_buf_item.h" #include "xfs_sb.h" #include "xfs_ag.h" -#include "xfs_dmapi.h" #include "xfs_mount.h" #include "xfs_trans_priv.h" #include "xfs_extfree_item.h" @@ -33,18 +32,19 @@ kmem_zone_t *xfs_efi_zone; kmem_zone_t *xfs_efd_zone; -STATIC void xfs_efi_item_unlock(xfs_efi_log_item_t *); +static inline struct xfs_efi_log_item *EFI_ITEM(struct xfs_log_item *lip) +{ + return container_of(lip, struct xfs_efi_log_item, efi_item); +} void -xfs_efi_item_free(xfs_efi_log_item_t *efip) +xfs_efi_item_free( + struct xfs_efi_log_item *efip) { - int nexts = efip->efi_format.efi_nextents; - - if (nexts > XFS_EFI_MAX_FAST_EXTENTS) { + if (efip->efi_format.efi_nextents > XFS_EFI_MAX_FAST_EXTENTS) kmem_free(efip); - } else { + else kmem_zone_free(xfs_efi_zone, efip); - } } /* @@ -52,9 +52,9 @@ xfs_efi_item_free(xfs_efi_log_item_t *efip) * We only need 1 iovec for an efi item. It just logs the efi_log_format * structure. */ -/*ARGSUSED*/ STATIC uint -xfs_efi_item_size(xfs_efi_log_item_t *efip) +xfs_efi_item_size( + struct xfs_log_item *lip) { return 1; } @@ -67,10 +67,12 @@ xfs_efi_item_size(xfs_efi_log_item_t *efip) * slots in the efi item have been filled. */ STATIC void -xfs_efi_item_format(xfs_efi_log_item_t *efip, - xfs_log_iovec_t *log_vector) +xfs_efi_item_format( + struct xfs_log_item *lip, + struct xfs_log_iovec *log_vector) { - uint size; + struct xfs_efi_log_item *efip = EFI_ITEM(lip); + uint size; ASSERT(efip->efi_next_extent == efip->efi_format.efi_nextents); @@ -80,7 +82,7 @@ xfs_efi_item_format(xfs_efi_log_item_t *efip, size += (efip->efi_format.efi_nextents - 1) * sizeof(xfs_extent_t); efip->efi_format.efi_size = 1; - log_vector->i_addr = (xfs_caddr_t)&(efip->efi_format); + log_vector->i_addr = &efip->efi_format; log_vector->i_len = size; log_vector->i_type = XLOG_REG_TYPE_EFI_FORMAT; ASSERT(size >= sizeof(xfs_efi_log_format_t)); @@ -90,60 +92,33 @@ xfs_efi_item_format(xfs_efi_log_item_t *efip, /* * Pinning has no meaning for an efi item, so just return. */ -/*ARGSUSED*/ STATIC void -xfs_efi_item_pin(xfs_efi_log_item_t *efip) +xfs_efi_item_pin( + struct xfs_log_item *lip) { - return; } - /* * While EFIs cannot really be pinned, the unpin operation is the * last place at which the EFI is manipulated during a transaction. * Here we coordinate with xfs_efi_cancel() to determine who gets to * free the EFI. */ -/*ARGSUSED*/ -STATIC void -xfs_efi_item_unpin(xfs_efi_log_item_t *efip) -{ - struct xfs_ail *ailp = efip->efi_item.li_ailp; - - spin_lock(&ailp->xa_lock); - if (efip->efi_flags & XFS_EFI_CANCELED) { - /* 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(&ailp->xa_lock); - } -} - -/* - * like unpin only we have to also clear the xaction descriptor - * pointing the log item if we free the item. This routine duplicates - * unpin because efi_flags is protected by the AIL lock. Freeing - * the descriptor and then calling unpin would force us to drop the AIL - * lock which would open up a race condition. - */ STATIC void -xfs_efi_item_unpin_remove(xfs_efi_log_item_t *efip, xfs_trans_t *tp) +xfs_efi_item_unpin( + struct xfs_log_item *lip, + int remove) { - struct xfs_ail *ailp = efip->efi_item.li_ailp; - xfs_log_item_desc_t *lidp; + struct xfs_efi_log_item *efip = EFI_ITEM(lip); + struct xfs_ail *ailp = lip->li_ailp; 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); + if (remove) + xfs_trans_del_item(lip); /* xfs_trans_ail_delete() drops the AIL lock. */ - xfs_trans_ail_delete(ailp, (xfs_log_item_t *)efip); + xfs_trans_ail_delete(ailp, lip); xfs_efi_item_free(efip); } else { efip->efi_flags |= XFS_EFI_COMMITTED; @@ -158,9 +133,9 @@ xfs_efi_item_unpin_remove(xfs_efi_log_item_t *efip, xfs_trans_t *tp) * XFS_ITEM_PINNED so that the caller will eventually flush the log. * This should help in getting the EFI out of the AIL. */ -/*ARGSUSED*/ STATIC uint -xfs_efi_item_trylock(xfs_efi_log_item_t *efip) +xfs_efi_item_trylock( + struct xfs_log_item *lip) { return XFS_ITEM_PINNED; } @@ -168,13 +143,12 @@ xfs_efi_item_trylock(xfs_efi_log_item_t *efip) /* * Efi items have no locking, so just return. */ -/*ARGSUSED*/ STATIC void -xfs_efi_item_unlock(xfs_efi_log_item_t *efip) +xfs_efi_item_unlock( + struct xfs_log_item *lip) { - if (efip->efi_item.li_flags & XFS_LI_ABORTED) - xfs_efi_item_free(efip); - return; + if (lip->li_flags & XFS_LI_ABORTED) + xfs_efi_item_free(EFI_ITEM(lip)); } /* @@ -183,9 +157,10 @@ xfs_efi_item_unlock(xfs_efi_log_item_t *efip) * flag is not paid any attention here. Checking for that is delayed * until the EFI is unpinned. */ -/*ARGSUSED*/ STATIC xfs_lsn_t -xfs_efi_item_committed(xfs_efi_log_item_t *efip, xfs_lsn_t lsn) +xfs_efi_item_committed( + struct xfs_log_item *lip, + xfs_lsn_t lsn) { return lsn; } @@ -195,11 +170,10 @@ xfs_efi_item_committed(xfs_efi_log_item_t *efip, xfs_lsn_t lsn) * stuck waiting for all of its corresponding efd items to be * committed to disk. */ -/*ARGSUSED*/ STATIC void -xfs_efi_item_push(xfs_efi_log_item_t *efip) +xfs_efi_item_push( + struct xfs_log_item *lip) { - return; } /* @@ -209,61 +183,55 @@ xfs_efi_item_push(xfs_efi_log_item_t *efip) * example, for inodes, the inode is locked throughout the extent freeing * so the dependency should be recorded there. */ -/*ARGSUSED*/ STATIC void -xfs_efi_item_committing(xfs_efi_log_item_t *efip, xfs_lsn_t lsn) +xfs_efi_item_committing( + struct xfs_log_item *lip, + xfs_lsn_t lsn) { - return; } /* * This is the ops vector shared by all efi log items. */ static struct xfs_item_ops xfs_efi_item_ops = { - .iop_size = (uint(*)(xfs_log_item_t*))xfs_efi_item_size, - .iop_format = (void(*)(xfs_log_item_t*, xfs_log_iovec_t*)) - xfs_efi_item_format, - .iop_pin = (void(*)(xfs_log_item_t*))xfs_efi_item_pin, - .iop_unpin = (void(*)(xfs_log_item_t*))xfs_efi_item_unpin, - .iop_unpin_remove = (void(*)(xfs_log_item_t*, xfs_trans_t *)) - xfs_efi_item_unpin_remove, - .iop_trylock = (uint(*)(xfs_log_item_t*))xfs_efi_item_trylock, - .iop_unlock = (void(*)(xfs_log_item_t*))xfs_efi_item_unlock, - .iop_committed = (xfs_lsn_t(*)(xfs_log_item_t*, xfs_lsn_t)) - xfs_efi_item_committed, - .iop_push = (void(*)(xfs_log_item_t*))xfs_efi_item_push, - .iop_pushbuf = NULL, - .iop_committing = (void(*)(xfs_log_item_t*, xfs_lsn_t)) - xfs_efi_item_committing + .iop_size = xfs_efi_item_size, + .iop_format = xfs_efi_item_format, + .iop_pin = xfs_efi_item_pin, + .iop_unpin = xfs_efi_item_unpin, + .iop_trylock = xfs_efi_item_trylock, + .iop_unlock = xfs_efi_item_unlock, + .iop_committed = xfs_efi_item_committed, + .iop_push = xfs_efi_item_push, + .iop_committing = xfs_efi_item_committing }; /* * Allocate and initialize an efi item with the given number of extents. */ -xfs_efi_log_item_t * -xfs_efi_init(xfs_mount_t *mp, - uint nextents) +struct xfs_efi_log_item * +xfs_efi_init( + struct xfs_mount *mp, + uint nextents) { - xfs_efi_log_item_t *efip; + struct xfs_efi_log_item *efip; uint size; ASSERT(nextents > 0); if (nextents > XFS_EFI_MAX_FAST_EXTENTS) { size = (uint)(sizeof(xfs_efi_log_item_t) + ((nextents - 1) * sizeof(xfs_extent_t))); - efip = (xfs_efi_log_item_t*)kmem_zalloc(size, KM_SLEEP); + efip = kmem_zalloc(size, KM_SLEEP); } else { - efip = (xfs_efi_log_item_t*)kmem_zone_zalloc(xfs_efi_zone, - KM_SLEEP); + efip = kmem_zone_zalloc(xfs_efi_zone, KM_SLEEP); } xfs_log_item_init(mp, &efip->efi_item, XFS_LI_EFI, &xfs_efi_item_ops); efip->efi_format.efi_nextents = nextents; efip->efi_format.efi_id = (__psint_t)(void*)efip; - return (efip); + return efip; } /* @@ -276,7 +244,7 @@ xfs_efi_init(xfs_mount_t *mp, int xfs_efi_copy_format(xfs_log_iovec_t *buf, xfs_efi_log_format_t *dst_efi_fmt) { - xfs_efi_log_format_t *src_efi_fmt = (xfs_efi_log_format_t *)buf->i_addr; + xfs_efi_log_format_t *src_efi_fmt = buf->i_addr; uint i; uint len = sizeof(xfs_efi_log_format_t) + (src_efi_fmt->efi_nextents - 1) * sizeof(xfs_extent_t); @@ -289,8 +257,7 @@ xfs_efi_copy_format(xfs_log_iovec_t *buf, xfs_efi_log_format_t *dst_efi_fmt) memcpy((char *)dst_efi_fmt, (char*)src_efi_fmt, len); return 0; } else if (buf->i_len == len32) { - xfs_efi_log_format_32_t *src_efi_fmt_32 = - (xfs_efi_log_format_32_t *)buf->i_addr; + xfs_efi_log_format_32_t *src_efi_fmt_32 = buf->i_addr; dst_efi_fmt->efi_type = src_efi_fmt_32->efi_type; dst_efi_fmt->efi_size = src_efi_fmt_32->efi_size; @@ -304,8 +271,7 @@ xfs_efi_copy_format(xfs_log_iovec_t *buf, xfs_efi_log_format_t *dst_efi_fmt) } return 0; } else if (buf->i_len == len64) { - xfs_efi_log_format_64_t *src_efi_fmt_64 = - (xfs_efi_log_format_64_t *)buf->i_addr; + xfs_efi_log_format_64_t *src_efi_fmt_64 = buf->i_addr; dst_efi_fmt->efi_type = src_efi_fmt_64->efi_type; dst_efi_fmt->efi_size = src_efi_fmt_64->efi_size; @@ -356,16 +322,18 @@ xfs_efi_release(xfs_efi_log_item_t *efip, } } -STATIC void -xfs_efd_item_free(xfs_efd_log_item_t *efdp) +static inline struct xfs_efd_log_item *EFD_ITEM(struct xfs_log_item *lip) { - int nexts = efdp->efd_format.efd_nextents; + return container_of(lip, struct xfs_efd_log_item, efd_item); +} - if (nexts > XFS_EFD_MAX_FAST_EXTENTS) { +STATIC void +xfs_efd_item_free(struct xfs_efd_log_item *efdp) +{ + if (efdp->efd_format.efd_nextents > XFS_EFD_MAX_FAST_EXTENTS) kmem_free(efdp); - } else { + else kmem_zone_free(xfs_efd_zone, efdp); - } } /* @@ -373,9 +341,9 @@ xfs_efd_item_free(xfs_efd_log_item_t *efdp) * We only need 1 iovec for an efd item. It just logs the efd_log_format * structure. */ -/*ARGSUSED*/ STATIC uint -xfs_efd_item_size(xfs_efd_log_item_t *efdp) +xfs_efd_item_size( + struct xfs_log_item *lip) { return 1; } @@ -388,10 +356,12 @@ xfs_efd_item_size(xfs_efd_log_item_t *efdp) * slots in the efd item have been filled. */ STATIC void -xfs_efd_item_format(xfs_efd_log_item_t *efdp, - xfs_log_iovec_t *log_vector) +xfs_efd_item_format( + struct xfs_log_item *lip, + struct xfs_log_iovec *log_vector) { - uint size; + struct xfs_efd_log_item *efdp = EFD_ITEM(lip); + uint size; ASSERT(efdp->efd_next_extent == efdp->efd_format.efd_nextents); @@ -401,48 +371,38 @@ xfs_efd_item_format(xfs_efd_log_item_t *efdp, size += (efdp->efd_format.efd_nextents - 1) * sizeof(xfs_extent_t); efdp->efd_format.efd_size = 1; - log_vector->i_addr = (xfs_caddr_t)&(efdp->efd_format); + log_vector->i_addr = &efdp->efd_format; log_vector->i_len = size; log_vector->i_type = XLOG_REG_TYPE_EFD_FORMAT; ASSERT(size >= sizeof(xfs_efd_log_format_t)); } - /* * Pinning has no meaning for an efd item, so just return. */ -/*ARGSUSED*/ STATIC void -xfs_efd_item_pin(xfs_efd_log_item_t *efdp) +xfs_efd_item_pin( + struct xfs_log_item *lip) { - return; } - /* * Since pinning has no meaning for an efd item, unpinning does * not either. */ -/*ARGSUSED*/ -STATIC void -xfs_efd_item_unpin(xfs_efd_log_item_t *efdp) -{ - return; -} - -/*ARGSUSED*/ STATIC void -xfs_efd_item_unpin_remove(xfs_efd_log_item_t *efdp, xfs_trans_t *tp) +xfs_efd_item_unpin( + struct xfs_log_item *lip, + int remove) { - return; } /* * Efd items have no locking, so just return success. */ -/*ARGSUSED*/ STATIC uint -xfs_efd_item_trylock(xfs_efd_log_item_t *efdp) +xfs_efd_item_trylock( + struct xfs_log_item *lip) { return XFS_ITEM_LOCKED; } @@ -451,13 +411,12 @@ xfs_efd_item_trylock(xfs_efd_log_item_t *efdp) * Efd items have no locking or pushing, so return failure * so that the caller doesn't bother with us. */ -/*ARGSUSED*/ STATIC void -xfs_efd_item_unlock(xfs_efd_log_item_t *efdp) +xfs_efd_item_unlock( + struct xfs_log_item *lip) { - if (efdp->efd_item.li_flags & XFS_LI_ABORTED) - xfs_efd_item_free(efdp); - return; + if (lip->li_flags & XFS_LI_ABORTED) + xfs_efd_item_free(EFD_ITEM(lip)); } /* @@ -467,15 +426,18 @@ xfs_efd_item_unlock(xfs_efd_log_item_t *efdp) * return -1 to keep the transaction code from further referencing * this item. */ -/*ARGSUSED*/ STATIC xfs_lsn_t -xfs_efd_item_committed(xfs_efd_log_item_t *efdp, xfs_lsn_t lsn) +xfs_efd_item_committed( + struct xfs_log_item *lip, + xfs_lsn_t lsn) { + struct xfs_efd_log_item *efdp = EFD_ITEM(lip); + /* * If we got a log I/O error, it's always the case that the LR with the * EFI got unpinned and freed before the EFD got aborted. */ - if ((efdp->efd_item.li_flags & XFS_LI_ABORTED) == 0) + if (!(lip->li_flags & XFS_LI_ABORTED)) xfs_efi_release(efdp->efd_efip, efdp->efd_format.efd_nextents); xfs_efd_item_free(efdp); @@ -486,11 +448,10 @@ xfs_efd_item_committed(xfs_efd_log_item_t *efdp, xfs_lsn_t lsn) * There isn't much you can do to push on an efd item. It is simply * stuck waiting for the log to be flushed to disk. */ -/*ARGSUSED*/ STATIC void -xfs_efd_item_push(xfs_efd_log_item_t *efdp) +xfs_efd_item_push( + struct xfs_log_item *lip) { - return; } /* @@ -500,55 +461,48 @@ xfs_efd_item_push(xfs_efd_log_item_t *efdp) * example, for inodes, the inode is locked throughout the extent freeing * so the dependency should be recorded there. */ -/*ARGSUSED*/ STATIC void -xfs_efd_item_committing(xfs_efd_log_item_t *efip, xfs_lsn_t lsn) +xfs_efd_item_committing( + struct xfs_log_item *lip, + xfs_lsn_t lsn) { - return; } /* * This is the ops vector shared by all efd log items. */ static struct xfs_item_ops xfs_efd_item_ops = { - .iop_size = (uint(*)(xfs_log_item_t*))xfs_efd_item_size, - .iop_format = (void(*)(xfs_log_item_t*, xfs_log_iovec_t*)) - xfs_efd_item_format, - .iop_pin = (void(*)(xfs_log_item_t*))xfs_efd_item_pin, - .iop_unpin = (void(*)(xfs_log_item_t*))xfs_efd_item_unpin, - .iop_unpin_remove = (void(*)(xfs_log_item_t*, xfs_trans_t*)) - xfs_efd_item_unpin_remove, - .iop_trylock = (uint(*)(xfs_log_item_t*))xfs_efd_item_trylock, - .iop_unlock = (void(*)(xfs_log_item_t*))xfs_efd_item_unlock, - .iop_committed = (xfs_lsn_t(*)(xfs_log_item_t*, xfs_lsn_t)) - xfs_efd_item_committed, - .iop_push = (void(*)(xfs_log_item_t*))xfs_efd_item_push, - .iop_pushbuf = NULL, - .iop_committing = (void(*)(xfs_log_item_t*, xfs_lsn_t)) - xfs_efd_item_committing + .iop_size = xfs_efd_item_size, + .iop_format = xfs_efd_item_format, + .iop_pin = xfs_efd_item_pin, + .iop_unpin = xfs_efd_item_unpin, + .iop_trylock = xfs_efd_item_trylock, + .iop_unlock = xfs_efd_item_unlock, + .iop_committed = xfs_efd_item_committed, + .iop_push = xfs_efd_item_push, + .iop_committing = xfs_efd_item_committing }; - /* * Allocate and initialize an efd item with the given number of extents. */ -xfs_efd_log_item_t * -xfs_efd_init(xfs_mount_t *mp, - xfs_efi_log_item_t *efip, - uint nextents) +struct xfs_efd_log_item * +xfs_efd_init( + struct xfs_mount *mp, + struct xfs_efi_log_item *efip, + uint nextents) { - xfs_efd_log_item_t *efdp; + struct xfs_efd_log_item *efdp; uint size; ASSERT(nextents > 0); if (nextents > XFS_EFD_MAX_FAST_EXTENTS) { size = (uint)(sizeof(xfs_efd_log_item_t) + ((nextents - 1) * sizeof(xfs_extent_t))); - efdp = (xfs_efd_log_item_t*)kmem_zalloc(size, KM_SLEEP); + efdp = kmem_zalloc(size, KM_SLEEP); } else { - efdp = (xfs_efd_log_item_t*)kmem_zone_zalloc(xfs_efd_zone, - KM_SLEEP); + efdp = kmem_zone_zalloc(xfs_efd_zone, KM_SLEEP); } xfs_log_item_init(mp, &efdp->efd_item, XFS_LI_EFD, &xfs_efd_item_ops); @@ -556,5 +510,5 @@ xfs_efd_init(xfs_mount_t *mp, efdp->efd_format.efd_nextents = nextents; efdp->efd_format.efd_efi_id = efip->efi_format.efi_id; - return (efdp); + return efdp; } diff --git a/fs/xfs/xfs_filestream.c b/fs/xfs/xfs_filestream.c index 390850ee660..9b715dce569 100644 --- a/fs/xfs/xfs_filestream.c +++ b/fs/xfs/xfs_filestream.c @@ -18,13 +18,9 @@ #include "xfs.h" #include "xfs_bmap_btree.h" #include "xfs_inum.h" -#include "xfs_dir2.h" -#include "xfs_dir2_sf.h" -#include "xfs_attr_sf.h" #include "xfs_dinode.h" #include "xfs_inode.h" #include "xfs_ag.h" -#include "xfs_dmapi.h" #include "xfs_log.h" #include "xfs_trans.h" #include "xfs_sb.h" @@ -127,6 +123,82 @@ typedef struct fstrm_item xfs_inode_t *pip; /* Parent directory inode pointer. */ } fstrm_item_t; +/* + * Allocation group filestream associations are tracked with per-ag atomic + * counters. These counters allow _xfs_filestream_pick_ag() to tell whether a + * particular AG already has active filestreams associated with it. The mount + * point's m_peraglock is used to protect these counters from per-ag array + * re-allocation during a growfs operation. When xfs_growfs_data_private() is + * about to reallocate the array, it calls xfs_filestream_flush() with the + * m_peraglock held in write mode. + * + * Since xfs_mru_cache_flush() guarantees that all the free functions for all + * the cache elements have finished executing before it returns, it's safe for + * the free functions to use the atomic counters without m_peraglock protection. + * This allows the implementation of xfs_fstrm_free_func() to be agnostic about + * whether it was called with the m_peraglock held in read mode, write mode or + * not held at all. The race condition this addresses is the following: + * + * - The work queue scheduler fires and pulls a filestream directory cache + * element off the LRU end of the cache for deletion, then gets pre-empted. + * - A growfs operation grabs the m_peraglock in write mode, flushes all the + * remaining items from the cache and reallocates the mount point's per-ag + * array, resetting all the counters to zero. + * - The work queue thread resumes and calls the free function for the element + * it started cleaning up earlier. In the process it decrements the + * filestreams counter for an AG that now has no references. + * + * With a shrinkfs feature, the above scenario could panic the system. + * + * All other uses of the following macros should be protected by either the + * m_peraglock held in read mode, or the cache's internal locking exposed by the + * interval between a call to xfs_mru_cache_lookup() and a call to + * xfs_mru_cache_done(). In addition, the m_peraglock must be held in read mode + * when new elements are added to the cache. + * + * Combined, these locking rules ensure that no associations will ever exist in + * the cache that reference per-ag array elements that have since been + * reallocated. + */ +static int +xfs_filestream_peek_ag( + xfs_mount_t *mp, + xfs_agnumber_t agno) +{ + struct xfs_perag *pag; + int ret; + + pag = xfs_perag_get(mp, agno); + ret = atomic_read(&pag->pagf_fstrms); + xfs_perag_put(pag); + return ret; +} + +static int +xfs_filestream_get_ag( + xfs_mount_t *mp, + xfs_agnumber_t agno) +{ + struct xfs_perag *pag; + int ret; + + pag = xfs_perag_get(mp, agno); + ret = atomic_inc_return(&pag->pagf_fstrms); + xfs_perag_put(pag); + return ret; +} + +static void +xfs_filestream_put_ag( + xfs_mount_t *mp, + xfs_agnumber_t agno) +{ + struct xfs_perag *pag; + + pag = xfs_perag_get(mp, agno); + atomic_dec(&pag->pagf_fstrms); + xfs_perag_put(pag); +} /* * Scan the AGs starting at startag looking for an AG that isn't in use and has @@ -355,16 +427,14 @@ xfs_fstrm_free_func( { fstrm_item_t *item = (fstrm_item_t *)data; xfs_inode_t *ip = item->ip; - int ref; ASSERT(ip->i_ino == ino); xfs_iflags_clear(ip, XFS_IFILESTREAM); /* Drop the reference taken on the AG when the item was added. */ - ref = xfs_filestream_put_ag(ip->i_mount, item->ag); + xfs_filestream_put_ag(ip->i_mount, item->ag); - ASSERT(ref >= 0); TRACE_FREE(ip->i_mount, ip, item->pip, item->ag, xfs_filestream_peek_ag(ip->i_mount, item->ag)); diff --git a/fs/xfs/xfs_filestream.h b/fs/xfs/xfs_filestream.h index 260f757bbc5..09dd9af4543 100644 --- a/fs/xfs/xfs_filestream.h +++ b/fs/xfs/xfs_filestream.h @@ -42,88 +42,6 @@ extern ktrace_t *xfs_filestreams_trace_buf; #endif -/* - * Allocation group filestream associations are tracked with per-ag atomic - * counters. These counters allow _xfs_filestream_pick_ag() to tell whether a - * particular AG already has active filestreams associated with it. The mount - * point's m_peraglock is used to protect these counters from per-ag array - * re-allocation during a growfs operation. When xfs_growfs_data_private() is - * about to reallocate the array, it calls xfs_filestream_flush() with the - * m_peraglock held in write mode. - * - * Since xfs_mru_cache_flush() guarantees that all the free functions for all - * the cache elements have finished executing before it returns, it's safe for - * the free functions to use the atomic counters without m_peraglock protection. - * This allows the implementation of xfs_fstrm_free_func() to be agnostic about - * whether it was called with the m_peraglock held in read mode, write mode or - * not held at all. The race condition this addresses is the following: - * - * - The work queue scheduler fires and pulls a filestream directory cache - * element off the LRU end of the cache for deletion, then gets pre-empted. - * - A growfs operation grabs the m_peraglock in write mode, flushes all the - * remaining items from the cache and reallocates the mount point's per-ag - * array, resetting all the counters to zero. - * - The work queue thread resumes and calls the free function for the element - * it started cleaning up earlier. In the process it decrements the - * filestreams counter for an AG that now has no references. - * - * With a shrinkfs feature, the above scenario could panic the system. - * - * All other uses of the following macros should be protected by either the - * m_peraglock held in read mode, or the cache's internal locking exposed by the - * interval between a call to xfs_mru_cache_lookup() and a call to - * xfs_mru_cache_done(). In addition, the m_peraglock must be held in read mode - * when new elements are added to the cache. - * - * Combined, these locking rules ensure that no associations will ever exist in - * the cache that reference per-ag array elements that have since been - * reallocated. - */ -/* - * xfs_filestream_peek_ag is only used in tracing code - */ -static inline int -xfs_filestream_peek_ag( - xfs_mount_t *mp, - xfs_agnumber_t agno) -{ - struct xfs_perag *pag; - int ret; - - pag = xfs_perag_get(mp, agno); - ret = atomic_read(&pag->pagf_fstrms); - xfs_perag_put(pag); - return ret; -} - -static inline int -xfs_filestream_get_ag( - xfs_mount_t *mp, - xfs_agnumber_t agno) -{ - struct xfs_perag *pag; - int ret; - - pag = xfs_perag_get(mp, agno); - ret = atomic_inc_return(&pag->pagf_fstrms); - xfs_perag_put(pag); - return ret; -} - -static inline int -xfs_filestream_put_ag( - xfs_mount_t *mp, - xfs_agnumber_t agno) -{ - struct xfs_perag *pag; - int ret; - - pag = xfs_perag_get(mp, agno); - ret = atomic_dec_return(&pag->pagf_fstrms); - xfs_perag_put(pag); - return ret; -} - /* allocation selection flags */ typedef enum xfs_fstrm_alloc { XFS_PICK_USERDATA = 1, diff --git a/fs/xfs/xfs_fsops.c b/fs/xfs/xfs_fsops.c index 37a6f62c57b..dbca5f5c37b 100644 --- a/fs/xfs/xfs_fsops.c +++ b/fs/xfs/xfs_fsops.c @@ -24,14 +24,10 @@ #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_dir2_sf.h" -#include "xfs_attr_sf.h" #include "xfs_dinode.h" #include "xfs_inode.h" #include "xfs_inode_item.h" @@ -626,8 +622,7 @@ xfs_fs_log_dummy( 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_ijoin(tp, ip); xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); xfs_trans_set_sync(tp); error = xfs_trans_commit(tp, 0); diff --git a/fs/xfs/xfs_ialloc.c b/fs/xfs/xfs_ialloc.c index c7142a064c4..abf80ae1e95 100644 --- a/fs/xfs/xfs_ialloc.c +++ b/fs/xfs/xfs_ialloc.c @@ -24,14 +24,10 @@ #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_dir2_sf.h" -#include "xfs_attr_sf.h" #include "xfs_dinode.h" #include "xfs_inode.h" #include "xfs_btree.h" diff --git a/fs/xfs/xfs_ialloc_btree.c b/fs/xfs/xfs_ialloc_btree.c index c282a9af539..d352862cefa 100644 --- a/fs/xfs/xfs_ialloc_btree.c +++ b/fs/xfs/xfs_ialloc_btree.c @@ -24,14 +24,10 @@ #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_dir2_sf.h" -#include "xfs_attr_sf.h" #include "xfs_dinode.h" #include "xfs_inode.h" #include "xfs_btree.h" diff --git a/fs/xfs/xfs_iget.c b/fs/xfs/xfs_iget.c index 8f8b91be2c9..b1ecc6f97ad 100644 --- a/fs/xfs/xfs_iget.c +++ b/fs/xfs/xfs_iget.c @@ -25,14 +25,10 @@ #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_dir2_sf.h" -#include "xfs_attr_sf.h" #include "xfs_dinode.h" #include "xfs_inode.h" #include "xfs_btree.h" @@ -95,7 +91,7 @@ xfs_inode_alloc( return ip; } -STATIC void +void xfs_inode_free( struct xfs_inode *ip) { @@ -212,7 +208,7 @@ xfs_iget_cache_hit( ip->i_flags &= ~XFS_INEW; ip->i_flags |= XFS_IRECLAIMABLE; __xfs_inode_set_reclaim_tag(pag, ip); - trace_xfs_iget_reclaim(ip); + trace_xfs_iget_reclaim_fail(ip); goto out_error; } @@ -227,6 +223,7 @@ xfs_iget_cache_hit( } else { /* If the VFS inode is being torn down, pause and try again. */ if (!igrab(inode)) { + trace_xfs_iget_skip(ip); error = EAGAIN; goto out_error; } @@ -234,6 +231,7 @@ xfs_iget_cache_hit( /* We've got a live one. */ spin_unlock(&ip->i_flags_lock); read_unlock(&pag->pag_ici_lock); + trace_xfs_iget_hit(ip); } if (lock_flags != 0) @@ -242,7 +240,6 @@ xfs_iget_cache_hit( xfs_iflags_clear(ip, XFS_ISTALE); XFS_STATS_INC(xs_ig_found); - trace_xfs_iget_found(ip); return 0; out_error: @@ -264,7 +261,6 @@ xfs_iget_cache_miss( { struct xfs_inode *ip; int error; - unsigned long first_index, mask; xfs_agino_t agino = XFS_INO_TO_AGINO(mp, ino); ip = xfs_inode_alloc(mp, ino); @@ -275,7 +271,7 @@ xfs_iget_cache_miss( if (error) goto out_destroy; - xfs_itrace_entry(ip); + trace_xfs_iget_miss(ip); if ((ip->i_d.di_mode == 0) && !(flags & XFS_IGET_CREATE)) { error = ENOENT; @@ -301,8 +297,6 @@ xfs_iget_cache_miss( BUG(); } - 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 */ @@ -321,7 +315,6 @@ xfs_iget_cache_miss( write_unlock(&pag->pag_ici_lock); radix_tree_preload_end(); - trace_xfs_iget_alloc(ip); *ipp = ip; return 0; @@ -422,97 +415,6 @@ out_error_or_again: } /* - * Decrement reference count of an inode structure and unlock it. - * - * ip -- the inode being released - * lock_flags -- this parameter indicates the inode's locks to be - * to be released. See the comment on xfs_iunlock() for a list - * of valid values. - */ -void -xfs_iput(xfs_inode_t *ip, - uint lock_flags) -{ - xfs_itrace_entry(ip); - xfs_iunlock(ip, lock_flags); - IRELE(ip); -} - -/* - * Special iput for brand-new inodes that are still locked - */ -void -xfs_iput_new( - xfs_inode_t *ip, - uint lock_flags) -{ - struct inode *inode = VFS_I(ip); - - xfs_itrace_entry(ip); - - if ((ip->i_d.di_mode == 0)) { - ASSERT(!xfs_iflags_test(ip, XFS_IRECLAIMABLE)); - make_bad_inode(inode); - } - if (inode->i_state & I_NEW) - unlock_new_inode(inode); - if (lock_flags) - xfs_iunlock(ip, lock_flags); - IRELE(ip); -} - -/* - * 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( - struct xfs_inode *ip) -{ - struct xfs_mount *mp = ip->i_mount; - struct xfs_perag *pag; - xfs_agino_t agino = XFS_INO_TO_AGINO(mp, ip->i_ino); - - XFS_STATS_INC(xs_ig_reclaims); - - /* - * Remove the inode from the per-AG radix tree. - * - * Because radix_tree_delete won't complain even if the item was never - * added to the tree assert that it's been there before to catch - * problems with the inode life time early on. - */ - pag = xfs_perag_get(mp, XFS_INO_TO_AGNO(mp, ip->i_ino)); - write_lock(&pag->pag_ici_lock); - if (!radix_tree_delete(&pag->pag_ici_root, agino)) - ASSERT(0); - write_unlock(&pag->pag_ici_lock); - xfs_perag_put(pag); - - /* - * 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_ilock(ip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL); - xfs_qm_dqdetach(ip); - xfs_iunlock(ip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL); - - xfs_inode_free(ip); -} - -/* * This is a wrapper routine around the xfs_ilock() routine * used to centralize some grungy code. It is used in places * that wish to lock the inode solely for reading the extents. diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c index b76a829d7e2..68415cb4f23 100644 --- a/fs/xfs/xfs_inode.c +++ b/fs/xfs/xfs_inode.c @@ -27,13 +27,10 @@ #include "xfs_trans_priv.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_dir2_sf.h" #include "xfs_attr_sf.h" #include "xfs_dinode.h" #include "xfs_inode.h" @@ -44,7 +41,6 @@ #include "xfs_alloc.h" #include "xfs_ialloc.h" #include "xfs_bmap.h" -#include "xfs_rw.h" #include "xfs_error.h" #include "xfs_utils.h" #include "xfs_quota.h" @@ -426,7 +422,7 @@ xfs_iformat( if (!XFS_DFORK_Q(dip)) return 0; ASSERT(ip->i_afp == NULL); - ip->i_afp = kmem_zone_zalloc(xfs_ifork_zone, KM_SLEEP); + ip->i_afp = kmem_zone_zalloc(xfs_ifork_zone, KM_SLEEP | KM_NOFS); ip->i_afp->if_ext_max = XFS_IFORK_ASIZE(ip) / (uint)sizeof(xfs_bmbt_rec_t); switch (dip->di_aformat) { @@ -509,7 +505,7 @@ xfs_iformat_local( ifp->if_u1.if_data = ifp->if_u2.if_inline_data; else { real_size = roundup(size, 4); - ifp->if_u1.if_data = kmem_alloc(real_size, KM_SLEEP); + ifp->if_u1.if_data = kmem_alloc(real_size, KM_SLEEP | KM_NOFS); } ifp->if_bytes = size; ifp->if_real_bytes = real_size; @@ -636,7 +632,7 @@ xfs_iformat_btree( } ifp->if_broot_bytes = size; - ifp->if_broot = kmem_alloc(size, KM_SLEEP); + ifp->if_broot = kmem_alloc(size, KM_SLEEP | KM_NOFS); ASSERT(ifp->if_broot != NULL); /* * Copy and convert from the on-disk structure @@ -922,7 +918,6 @@ xfs_iread_extents( int error; xfs_ifork_t *ifp; xfs_extnum_t nextents; - size_t size; if (unlikely(XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE)) { XFS_ERROR_REPORT("xfs_iread_extents", XFS_ERRLEVEL_LOW, @@ -930,7 +925,6 @@ xfs_iread_extents( return XFS_ERROR(EFSCORRUPTED); } nextents = XFS_IFORK_NEXTENTS(ip, whichfork); - size = nextents * sizeof(xfs_bmbt_rec_t); ifp = XFS_IFORK_PTR(ip, whichfork); /* @@ -1226,7 +1220,7 @@ xfs_isize_check( (xfs_ufsize_t)XFS_MAXIOFFSET(mp)) - map_first), XFS_BMAPI_ENTIRE, NULL, 0, imaps, &nimaps, - NULL, NULL)) + NULL)) return; ASSERT(nimaps == 1); ASSERT(imaps[0].br_startblock == HOLESTARTBLOCK); @@ -1460,7 +1454,7 @@ xfs_itruncate_finish( ASSERT((*tp)->t_flags & XFS_TRANS_PERM_LOG_RES); ASSERT(ip->i_transp == *tp); ASSERT(ip->i_itemp != NULL); - ASSERT(ip->i_itemp->ili_flags & XFS_ILI_HOLD); + ASSERT(ip->i_itemp->ili_lock_flags == 0); ntp = *tp; @@ -1589,11 +1583,10 @@ xfs_itruncate_finish( xfs_bmap_init(&free_list, &first_block); error = xfs_bunmapi(ntp, ip, first_unmap_block, unmap_len, - xfs_bmapi_aflag(fork) | - (sync ? 0 : XFS_BMAPI_ASYNC), + xfs_bmapi_aflag(fork), XFS_ITRUNC_MAX_EXTENTS, &first_block, &free_list, - NULL, &done); + &done); if (error) { /* * If the bunmapi call encounters an error, @@ -1612,12 +1605,8 @@ xfs_itruncate_finish( */ error = xfs_bmap_finish(tp, &free_list, &committed); ntp = *tp; - if (committed) { - /* link the inode into the next xact in the chain */ - xfs_trans_ijoin(ntp, ip, - XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL); - xfs_trans_ihold(ntp, ip); - } + if (committed) + xfs_trans_ijoin(ntp, ip); if (error) { /* @@ -1646,9 +1635,7 @@ xfs_itruncate_finish( error = xfs_trans_commit(*tp, 0); *tp = ntp; - /* link the inode into the next transaction in the chain */ - xfs_trans_ijoin(ntp, ip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL); - xfs_trans_ihold(ntp, ip); + xfs_trans_ijoin(ntp, ip); if (error) return error; @@ -1985,7 +1972,7 @@ xfs_ifree_cluster( if (lip->li_type == XFS_LI_INODE) { 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; + lip->li_cb = xfs_istale_done; xfs_trans_ail_copy_lsn(mp->m_ail, &iip->ili_flush_lsn, &iip->ili_item.li_lsn); @@ -2055,9 +2042,8 @@ xfs_ifree_cluster( 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*)) - xfs_istale_done, (xfs_log_item_t *)iip); + xfs_buf_attach_iodone(bp, xfs_istale_done, + &iip->ili_item); if (ip != free_ip) xfs_iunlock(ip, XFS_ILOCK_EXCL); @@ -2203,7 +2189,7 @@ xfs_iroot_realloc( */ if (ifp->if_broot_bytes == 0) { new_size = (size_t)XFS_BMAP_BROOT_SPACE_CALC(rec_diff); - ifp->if_broot = kmem_alloc(new_size, KM_SLEEP); + ifp->if_broot = kmem_alloc(new_size, KM_SLEEP | KM_NOFS); ifp->if_broot_bytes = (int)new_size; return; } @@ -2219,7 +2205,7 @@ xfs_iroot_realloc( new_size = (size_t)XFS_BMAP_BROOT_SPACE_CALC(new_max); ifp->if_broot = kmem_realloc(ifp->if_broot, new_size, (size_t)XFS_BMAP_BROOT_SPACE_CALC(cur_max), /* old size */ - KM_SLEEP); + KM_SLEEP | KM_NOFS); 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, @@ -2245,7 +2231,7 @@ xfs_iroot_realloc( else new_size = 0; if (new_size > 0) { - new_broot = kmem_alloc(new_size, KM_SLEEP); + new_broot = kmem_alloc(new_size, KM_SLEEP | KM_NOFS); /* * First copy over the btree block header. */ @@ -2349,7 +2335,8 @@ xfs_idata_realloc( real_size = roundup(new_size, 4); if (ifp->if_u1.if_data == NULL) { ASSERT(ifp->if_real_bytes == 0); - ifp->if_u1.if_data = kmem_alloc(real_size, KM_SLEEP); + ifp->if_u1.if_data = kmem_alloc(real_size, + KM_SLEEP | KM_NOFS); } else if (ifp->if_u1.if_data != ifp->if_u2.if_inline_data) { /* * Only do the realloc if the underlying size @@ -2360,11 +2347,12 @@ xfs_idata_realloc( kmem_realloc(ifp->if_u1.if_data, real_size, ifp->if_real_bytes, - KM_SLEEP); + KM_SLEEP | KM_NOFS); } } else { ASSERT(ifp->if_real_bytes == 0); - ifp->if_u1.if_data = kmem_alloc(real_size, KM_SLEEP); + ifp->if_u1.if_data = kmem_alloc(real_size, + KM_SLEEP | KM_NOFS); memcpy(ifp->if_u1.if_data, ifp->if_u2.if_inline_data, ifp->if_bytes); } @@ -2731,7 +2719,6 @@ cluster_corrupt_out: * mark it as stale and brelse. */ if (XFS_BUF_IODONE_FUNC(bp)) { - XFS_BUF_CLR_BDSTRAT_FUNC(bp); XFS_BUF_UNDONE(bp); XFS_BUF_STALE(bp); XFS_BUF_ERROR(bp,EIO); @@ -3069,8 +3056,7 @@ xfs_iflush_int( * and unlock the inode's flush lock when the inode is * completely written to disk. */ - xfs_buf_attach_iodone(bp, (void(*)(xfs_buf_t*,xfs_log_item_t*)) - xfs_iflush_done, (xfs_log_item_t *)iip); + xfs_buf_attach_iodone(bp, xfs_iflush_done, &iip->ili_item); ASSERT(XFS_BUF_FSPRIVATE(bp, void *) != NULL); ASSERT(XFS_BUF_IODONE_FUNC(bp) != NULL); @@ -3514,13 +3500,11 @@ xfs_iext_remove_indirect( xfs_extnum_t ext_diff; /* extents to remove in current list */ xfs_extnum_t nex1; /* number of extents before idx */ xfs_extnum_t nex2; /* extents after idx + count */ - int nlists; /* entries in indirection array */ int page_idx = idx; /* index in target extent list */ ASSERT(ifp->if_flags & XFS_IFEXTIREC); erp = xfs_iext_idx_to_irec(ifp, &page_idx, &erp_idx, 0); ASSERT(erp != NULL); - nlists = ifp->if_real_bytes / XFS_IEXT_BUFSZ; nex1 = page_idx; ext_cnt = count; while (ext_cnt) { diff --git a/fs/xfs/xfs_inode.h b/fs/xfs/xfs_inode.h index 78550df13cd..0898c5417d1 100644 --- a/fs/xfs/xfs_inode.h +++ b/fs/xfs/xfs_inode.h @@ -443,8 +443,6 @@ static inline void xfs_ifunlock(xfs_inode_t *ip) */ int xfs_iget(struct xfs_mount *, struct xfs_trans *, xfs_ino_t, uint, uint, xfs_inode_t **); -void xfs_iput(xfs_inode_t *, uint); -void xfs_iput_new(xfs_inode_t *, uint); void xfs_ilock(xfs_inode_t *, uint); int xfs_ilock_nowait(xfs_inode_t *, uint); void xfs_iunlock(xfs_inode_t *, uint); @@ -452,7 +450,7 @@ void xfs_ilock_demote(xfs_inode_t *, uint); 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 *); +void xfs_inode_free(struct xfs_inode *ip); /* * xfs_inode.c prototypes. diff --git a/fs/xfs/xfs_inode_item.c b/fs/xfs/xfs_inode_item.c index cf8249a6000..fe00777e279 100644 --- a/fs/xfs/xfs_inode_item.c +++ b/fs/xfs/xfs_inode_item.c @@ -22,30 +22,26 @@ #include "xfs_log.h" #include "xfs_inum.h" #include "xfs_trans.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" #include "xfs_trans_priv.h" #include "xfs_bmap_btree.h" -#include "xfs_alloc_btree.h" -#include "xfs_ialloc_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_ialloc.h" -#include "xfs_rw.h" #include "xfs_error.h" #include "xfs_trace.h" kmem_zone_t *xfs_ili_zone; /* inode log item zone */ +static inline struct xfs_inode_log_item *INODE_ITEM(struct xfs_log_item *lip) +{ + return container_of(lip, struct xfs_inode_log_item, ili_item); +} + + /* * This returns the number of iovecs needed to log the given inode item. * @@ -55,13 +51,11 @@ kmem_zone_t *xfs_ili_zone; /* inode log item zone */ */ STATIC uint xfs_inode_item_size( - xfs_inode_log_item_t *iip) + struct xfs_log_item *lip) { - uint nvecs; - xfs_inode_t *ip; - - ip = iip->ili_inode; - nvecs = 2; + struct xfs_inode_log_item *iip = INODE_ITEM(lip); + struct xfs_inode *ip = iip->ili_inode; + uint nvecs = 2; /* * Only log the data/extents/b-tree root if there is something @@ -212,21 +206,17 @@ xfs_inode_item_size( */ STATIC void xfs_inode_item_format( - xfs_inode_log_item_t *iip, - xfs_log_iovec_t *log_vector) + struct xfs_log_item *lip, + struct xfs_log_iovec *vecp) { + struct xfs_inode_log_item *iip = INODE_ITEM(lip); + struct xfs_inode *ip = iip->ili_inode; uint nvecs; - xfs_log_iovec_t *vecp; - xfs_inode_t *ip; size_t data_bytes; xfs_bmbt_rec_t *ext_buffer; - int nrecs; xfs_mount_t *mp; - ip = iip->ili_inode; - vecp = log_vector; - - vecp->i_addr = (xfs_caddr_t)&iip->ili_format; + vecp->i_addr = &iip->ili_format; vecp->i_len = sizeof(xfs_inode_log_format_t); vecp->i_type = XLOG_REG_TYPE_IFORMAT; vecp++; @@ -277,7 +267,7 @@ xfs_inode_item_format( */ xfs_synchronize_times(ip); - vecp->i_addr = (xfs_caddr_t)&ip->i_d; + vecp->i_addr = &ip->i_d; vecp->i_len = sizeof(struct xfs_icdinode); vecp->i_type = XLOG_REG_TYPE_ICORE; vecp++; @@ -323,18 +313,17 @@ xfs_inode_item_format( ASSERT(ip->i_df.if_u1.if_extents != NULL); ASSERT(ip->i_d.di_nextents > 0); ASSERT(iip->ili_extents_buf == NULL); - nrecs = ip->i_df.if_bytes / - (uint)sizeof(xfs_bmbt_rec_t); - ASSERT(nrecs > 0); + ASSERT((ip->i_df.if_bytes / + (uint)sizeof(xfs_bmbt_rec_t)) > 0); #ifdef XFS_NATIVE_HOST - if (nrecs == ip->i_d.di_nextents) { + if (ip->i_d.di_nextents == ip->i_df.if_bytes / + (uint)sizeof(xfs_bmbt_rec_t)) { /* * There are no delayed allocation * extents, so just point to the * real extents array. */ - vecp->i_addr = - (char *)(ip->i_df.if_u1.if_extents); + vecp->i_addr = ip->i_df.if_u1.if_extents; vecp->i_len = ip->i_df.if_bytes; vecp->i_type = XLOG_REG_TYPE_IEXT; } else @@ -352,7 +341,7 @@ xfs_inode_item_format( ext_buffer = kmem_alloc(ip->i_df.if_bytes, KM_SLEEP); iip->ili_extents_buf = ext_buffer; - vecp->i_addr = (xfs_caddr_t)ext_buffer; + vecp->i_addr = ext_buffer; vecp->i_len = xfs_iextents_copy(ip, ext_buffer, XFS_DATA_FORK); vecp->i_type = XLOG_REG_TYPE_IEXT; @@ -371,7 +360,7 @@ xfs_inode_item_format( if (iip->ili_format.ilf_fields & XFS_ILOG_DBROOT) { ASSERT(ip->i_df.if_broot_bytes > 0); ASSERT(ip->i_df.if_broot != NULL); - vecp->i_addr = (xfs_caddr_t)ip->i_df.if_broot; + vecp->i_addr = ip->i_df.if_broot; vecp->i_len = ip->i_df.if_broot_bytes; vecp->i_type = XLOG_REG_TYPE_IBROOT; vecp++; @@ -389,7 +378,7 @@ xfs_inode_item_format( ASSERT(ip->i_df.if_u1.if_data != NULL); ASSERT(ip->i_d.di_size > 0); - vecp->i_addr = (xfs_caddr_t)ip->i_df.if_u1.if_data; + vecp->i_addr = ip->i_df.if_u1.if_data; /* * Round i_bytes up to a word boundary. * The underlying memory is guaranteed to @@ -437,7 +426,7 @@ xfs_inode_item_format( * Assert that no attribute-related log flags are set. */ if (!XFS_IFORK_Q(ip)) { - ASSERT(nvecs == iip->ili_item.li_desc->lid_size); + ASSERT(nvecs == lip->li_desc->lid_size); iip->ili_format.ilf_size = nvecs; ASSERT(!(iip->ili_format.ilf_fields & (XFS_ILOG_ADATA | XFS_ILOG_ABROOT | XFS_ILOG_AEXT))); @@ -449,21 +438,21 @@ xfs_inode_item_format( ASSERT(!(iip->ili_format.ilf_fields & (XFS_ILOG_ADATA | XFS_ILOG_ABROOT))); if (iip->ili_format.ilf_fields & XFS_ILOG_AEXT) { - ASSERT(ip->i_afp->if_bytes > 0); - ASSERT(ip->i_afp->if_u1.if_extents != NULL); - ASSERT(ip->i_d.di_anextents > 0); #ifdef DEBUG - nrecs = ip->i_afp->if_bytes / + int nrecs = ip->i_afp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t); -#endif ASSERT(nrecs > 0); ASSERT(nrecs == ip->i_d.di_anextents); + ASSERT(ip->i_afp->if_bytes > 0); + ASSERT(ip->i_afp->if_u1.if_extents != NULL); + ASSERT(ip->i_d.di_anextents > 0); +#endif #ifdef XFS_NATIVE_HOST /* * There are not delayed allocation extents * for attributes, so just point at the array. */ - vecp->i_addr = (char *)(ip->i_afp->if_u1.if_extents); + vecp->i_addr = ip->i_afp->if_u1.if_extents; vecp->i_len = ip->i_afp->if_bytes; #else ASSERT(iip->ili_aextents_buf == NULL); @@ -473,7 +462,7 @@ xfs_inode_item_format( ext_buffer = kmem_alloc(ip->i_afp->if_bytes, KM_SLEEP); iip->ili_aextents_buf = ext_buffer; - vecp->i_addr = (xfs_caddr_t)ext_buffer; + vecp->i_addr = ext_buffer; vecp->i_len = xfs_iextents_copy(ip, ext_buffer, XFS_ATTR_FORK); #endif @@ -490,7 +479,7 @@ xfs_inode_item_format( if (iip->ili_format.ilf_fields & XFS_ILOG_ABROOT) { ASSERT(ip->i_afp->if_broot_bytes > 0); ASSERT(ip->i_afp->if_broot != NULL); - vecp->i_addr = (xfs_caddr_t)ip->i_afp->if_broot; + vecp->i_addr = ip->i_afp->if_broot; vecp->i_len = ip->i_afp->if_broot_bytes; vecp->i_type = XLOG_REG_TYPE_IATTR_BROOT; vecp++; @@ -506,7 +495,7 @@ xfs_inode_item_format( ASSERT(ip->i_afp->if_bytes > 0); ASSERT(ip->i_afp->if_u1.if_data != NULL); - vecp->i_addr = (xfs_caddr_t)ip->i_afp->if_u1.if_data; + vecp->i_addr = ip->i_afp->if_u1.if_data; /* * Round i_bytes up to a word boundary. * The underlying memory is guaranteed to @@ -528,7 +517,7 @@ xfs_inode_item_format( break; } - ASSERT(nvecs == iip->ili_item.li_desc->lid_size); + ASSERT(nvecs == lip->li_desc->lid_size); iip->ili_format.ilf_size = nvecs; } @@ -539,12 +528,14 @@ xfs_inode_item_format( */ STATIC void xfs_inode_item_pin( - xfs_inode_log_item_t *iip) + struct xfs_log_item *lip) { - ASSERT(xfs_isilocked(iip->ili_inode, XFS_ILOCK_EXCL)); + struct xfs_inode *ip = INODE_ITEM(lip)->ili_inode; + + ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL)); - trace_xfs_inode_pin(iip->ili_inode, _RET_IP_); - atomic_inc(&iip->ili_inode->i_pincount); + trace_xfs_inode_pin(ip, _RET_IP_); + atomic_inc(&ip->i_pincount); } @@ -554,12 +545,12 @@ xfs_inode_item_pin( * * Also wake up anyone in xfs_iunpin_wait() if the count goes to 0. */ -/* ARGSUSED */ STATIC void xfs_inode_item_unpin( - xfs_inode_log_item_t *iip) + struct xfs_log_item *lip, + int remove) { - struct xfs_inode *ip = iip->ili_inode; + struct xfs_inode *ip = INODE_ITEM(lip)->ili_inode; trace_xfs_inode_unpin(ip, _RET_IP_); ASSERT(atomic_read(&ip->i_pincount) > 0); @@ -567,15 +558,6 @@ xfs_inode_item_unpin( wake_up(&ip->i_ipin_wait); } -/* ARGSUSED */ -STATIC void -xfs_inode_item_unpin_remove( - xfs_inode_log_item_t *iip, - xfs_trans_t *tp) -{ - xfs_inode_item_unpin(iip); -} - /* * This is called to attempt to lock the inode associated with this * inode log item, in preparation for the push routine which does the actual @@ -591,19 +573,16 @@ xfs_inode_item_unpin_remove( */ STATIC uint xfs_inode_item_trylock( - xfs_inode_log_item_t *iip) + struct xfs_log_item *lip) { - register xfs_inode_t *ip; - - ip = iip->ili_inode; + struct xfs_inode_log_item *iip = INODE_ITEM(lip); + struct xfs_inode *ip = iip->ili_inode; - if (xfs_ipincount(ip) > 0) { + if (xfs_ipincount(ip) > 0) return XFS_ITEM_PINNED; - } - if (!xfs_ilock_nowait(ip, XFS_ILOCK_SHARED)) { + if (!xfs_ilock_nowait(ip, XFS_ILOCK_SHARED)) return XFS_ITEM_LOCKED; - } if (!xfs_iflock_nowait(ip)) { /* @@ -629,7 +608,7 @@ xfs_inode_item_trylock( if (!XFS_FORCED_SHUTDOWN(ip->i_mount)) { ASSERT(iip->ili_format.ilf_fields != 0); ASSERT(iip->ili_logged == 0); - ASSERT(iip->ili_item.li_flags & XFS_LI_IN_AIL); + ASSERT(lip->li_flags & XFS_LI_IN_AIL); } #endif return XFS_ITEM_SUCCESS; @@ -643,26 +622,18 @@ xfs_inode_item_trylock( */ STATIC void xfs_inode_item_unlock( - xfs_inode_log_item_t *iip) + struct xfs_log_item *lip) { - uint hold; - uint iolocked; - uint lock_flags; - xfs_inode_t *ip; + struct xfs_inode_log_item *iip = INODE_ITEM(lip); + struct xfs_inode *ip = iip->ili_inode; + unsigned short lock_flags; - ASSERT(iip != NULL); ASSERT(iip->ili_inode->i_itemp != NULL); ASSERT(xfs_isilocked(iip->ili_inode, XFS_ILOCK_EXCL)); - ASSERT((!(iip->ili_inode->i_itemp->ili_flags & - XFS_ILI_IOLOCKED_EXCL)) || - xfs_isilocked(iip->ili_inode, XFS_IOLOCK_EXCL)); - ASSERT((!(iip->ili_inode->i_itemp->ili_flags & - XFS_ILI_IOLOCKED_SHARED)) || - xfs_isilocked(iip->ili_inode, XFS_IOLOCK_SHARED)); + /* * Clear the transaction pointer in the inode. */ - ip = iip->ili_inode; ip->i_transp = NULL; /* @@ -686,34 +657,11 @@ xfs_inode_item_unlock( iip->ili_aextents_buf = NULL; } - /* - * Figure out if we should unlock the inode or not. - */ - hold = iip->ili_flags & XFS_ILI_HOLD; - - /* - * Before clearing out the flags, remember whether we - * are holding the inode's IO lock. - */ - iolocked = iip->ili_flags & XFS_ILI_IOLOCKED_ANY; - - /* - * Clear out the fields of the inode log item particular - * to the current transaction. - */ - iip->ili_flags = 0; - - /* - * Unlock the inode if XFS_ILI_HOLD was not set. - */ - if (!hold) { - lock_flags = XFS_ILOCK_EXCL; - if (iolocked & XFS_ILI_IOLOCKED_EXCL) { - lock_flags |= XFS_IOLOCK_EXCL; - } else if (iolocked & XFS_ILI_IOLOCKED_SHARED) { - lock_flags |= XFS_IOLOCK_SHARED; - } - xfs_iput(iip->ili_inode, lock_flags); + lock_flags = iip->ili_lock_flags; + iip->ili_lock_flags = 0; + if (lock_flags) { + xfs_iunlock(iip->ili_inode, lock_flags); + IRELE(iip->ili_inode); } } @@ -725,13 +673,12 @@ xfs_inode_item_unlock( * is the only one that matters. Therefore, simply return the * given lsn. */ -/*ARGSUSED*/ STATIC xfs_lsn_t xfs_inode_item_committed( - xfs_inode_log_item_t *iip, + struct xfs_log_item *lip, xfs_lsn_t lsn) { - return (lsn); + return lsn; } /* @@ -743,13 +690,12 @@ xfs_inode_item_committed( */ STATIC void xfs_inode_item_pushbuf( - xfs_inode_log_item_t *iip) + struct xfs_log_item *lip) { - xfs_inode_t *ip; - xfs_mount_t *mp; - xfs_buf_t *bp; + struct xfs_inode_log_item *iip = INODE_ITEM(lip); + struct xfs_inode *ip = iip->ili_inode; + struct xfs_buf *bp; - ip = iip->ili_inode; ASSERT(xfs_isilocked(ip, XFS_ILOCK_SHARED)); /* @@ -757,14 +703,13 @@ xfs_inode_item_pushbuf( * inode was taken off the AIL. So, just get out. */ if (completion_done(&ip->i_flush) || - ((iip->ili_item.li_flags & XFS_LI_IN_AIL) == 0)) { + !(lip->li_flags & XFS_LI_IN_AIL)) { xfs_iunlock(ip, XFS_ILOCK_SHARED); return; } - mp = ip->i_mount; - bp = xfs_incore(mp->m_ddev_targp, iip->ili_format.ilf_blkno, - iip->ili_format.ilf_len, XBF_TRYLOCK); + bp = xfs_incore(ip->i_mount->m_ddev_targp, iip->ili_format.ilf_blkno, + iip->ili_format.ilf_len, XBF_TRYLOCK); xfs_iunlock(ip, XFS_ILOCK_SHARED); if (!bp) @@ -772,10 +717,8 @@ xfs_inode_item_pushbuf( if (XFS_BUF_ISDELAYWRITE(bp)) xfs_buf_delwri_promote(bp); xfs_buf_relse(bp); - return; } - /* * This is called to asynchronously write the inode associated with this * inode log item out to disk. The inode will already have been locked by @@ -783,14 +726,14 @@ xfs_inode_item_pushbuf( */ STATIC void xfs_inode_item_push( - xfs_inode_log_item_t *iip) + struct xfs_log_item *lip) { - xfs_inode_t *ip; - - ip = iip->ili_inode; + struct xfs_inode_log_item *iip = INODE_ITEM(lip); + struct xfs_inode *ip = iip->ili_inode; ASSERT(xfs_isilocked(ip, XFS_ILOCK_SHARED)); ASSERT(!completion_done(&ip->i_flush)); + /* * Since we were able to lock the inode's flush lock and * we found it on the AIL, the inode must be dirty. This @@ -813,43 +756,34 @@ xfs_inode_item_push( */ (void) xfs_iflush(ip, 0); xfs_iunlock(ip, XFS_ILOCK_SHARED); - - return; } /* * XXX rcc - this one really has to do something. Probably needs * to stamp in a new field in the incore inode. */ -/* ARGSUSED */ STATIC void xfs_inode_item_committing( - xfs_inode_log_item_t *iip, + struct xfs_log_item *lip, xfs_lsn_t lsn) { - iip->ili_last_lsn = lsn; - return; + INODE_ITEM(lip)->ili_last_lsn = lsn; } /* * This is the ops vector shared by all buf log items. */ static struct xfs_item_ops xfs_inode_item_ops = { - .iop_size = (uint(*)(xfs_log_item_t*))xfs_inode_item_size, - .iop_format = (void(*)(xfs_log_item_t*, xfs_log_iovec_t*)) - xfs_inode_item_format, - .iop_pin = (void(*)(xfs_log_item_t*))xfs_inode_item_pin, - .iop_unpin = (void(*)(xfs_log_item_t*))xfs_inode_item_unpin, - .iop_unpin_remove = (void(*)(xfs_log_item_t*, xfs_trans_t*)) - xfs_inode_item_unpin_remove, - .iop_trylock = (uint(*)(xfs_log_item_t*))xfs_inode_item_trylock, - .iop_unlock = (void(*)(xfs_log_item_t*))xfs_inode_item_unlock, - .iop_committed = (xfs_lsn_t(*)(xfs_log_item_t*, xfs_lsn_t)) - xfs_inode_item_committed, - .iop_push = (void(*)(xfs_log_item_t*))xfs_inode_item_push, - .iop_pushbuf = (void(*)(xfs_log_item_t*))xfs_inode_item_pushbuf, - .iop_committing = (void(*)(xfs_log_item_t*, xfs_lsn_t)) - xfs_inode_item_committing + .iop_size = xfs_inode_item_size, + .iop_format = xfs_inode_item_format, + .iop_pin = xfs_inode_item_pin, + .iop_unpin = xfs_inode_item_unpin, + .iop_trylock = xfs_inode_item_trylock, + .iop_unlock = xfs_inode_item_unlock, + .iop_committed = xfs_inode_item_committed, + .iop_push = xfs_inode_item_push, + .iop_pushbuf = xfs_inode_item_pushbuf, + .iop_committing = xfs_inode_item_committing }; @@ -858,10 +792,10 @@ static struct xfs_item_ops xfs_inode_item_ops = { */ void xfs_inode_item_init( - xfs_inode_t *ip, - xfs_mount_t *mp) + struct xfs_inode *ip, + struct xfs_mount *mp) { - xfs_inode_log_item_t *iip; + struct xfs_inode_log_item *iip; ASSERT(ip->i_itemp == NULL); iip = ip->i_itemp = kmem_zone_zalloc(xfs_ili_zone, KM_SLEEP); @@ -899,14 +833,14 @@ xfs_inode_item_destroy( * from the AIL if it has not been re-logged, and unlocking the inode's * flush lock. */ -/*ARGSUSED*/ void xfs_iflush_done( - xfs_buf_t *bp, - xfs_inode_log_item_t *iip) + struct xfs_buf *bp, + struct xfs_log_item *lip) { + struct xfs_inode_log_item *iip = INODE_ITEM(lip); xfs_inode_t *ip = iip->ili_inode; - struct xfs_ail *ailp = iip->ili_item.li_ailp; + struct xfs_ail *ailp = lip->li_ailp; /* * We only want to pull the item from the AIL if it is @@ -917,12 +851,11 @@ xfs_iflush_done( * the lock since it's cheaper, and then we recheck while * holding the lock before removing the inode from the AIL. */ - if (iip->ili_logged && - (iip->ili_item.li_lsn == iip->ili_flush_lsn)) { + if (iip->ili_logged && lip->li_lsn == iip->ili_flush_lsn) { spin_lock(&ailp->xa_lock); - if (iip->ili_item.li_lsn == iip->ili_flush_lsn) { + if (lip->li_lsn == iip->ili_flush_lsn) { /* xfs_trans_ail_delete() drops the AIL lock. */ - xfs_trans_ail_delete(ailp, (xfs_log_item_t*)iip); + xfs_trans_ail_delete(ailp, lip); } else { spin_unlock(&ailp->xa_lock); } @@ -940,8 +873,6 @@ xfs_iflush_done( * Release the inode's flush lock since we're done with it. */ xfs_ifunlock(ip); - - return; } /* @@ -957,10 +888,8 @@ xfs_iflush_abort( xfs_inode_t *ip) { 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) { @@ -991,10 +920,10 @@ xfs_iflush_abort( void xfs_istale_done( - xfs_buf_t *bp, - xfs_inode_log_item_t *iip) + struct xfs_buf *bp, + struct xfs_log_item *lip) { - xfs_iflush_abort(iip->ili_inode); + xfs_iflush_abort(INODE_ITEM(lip)->ili_inode); } /* @@ -1007,9 +936,8 @@ xfs_inode_item_format_convert( xfs_inode_log_format_t *in_f) { if (buf->i_len == sizeof(xfs_inode_log_format_32_t)) { - xfs_inode_log_format_32_t *in_f32; + xfs_inode_log_format_32_t *in_f32 = buf->i_addr; - in_f32 = (xfs_inode_log_format_32_t *)buf->i_addr; in_f->ilf_type = in_f32->ilf_type; in_f->ilf_size = in_f32->ilf_size; in_f->ilf_fields = in_f32->ilf_fields; @@ -1025,9 +953,8 @@ xfs_inode_item_format_convert( in_f->ilf_boffset = in_f32->ilf_boffset; return 0; } else if (buf->i_len == sizeof(xfs_inode_log_format_64_t)){ - xfs_inode_log_format_64_t *in_f64; + xfs_inode_log_format_64_t *in_f64 = buf->i_addr; - in_f64 = (xfs_inode_log_format_64_t *)buf->i_addr; in_f->ilf_type = in_f64->ilf_type; in_f->ilf_size = in_f64->ilf_size; in_f->ilf_fields = in_f64->ilf_fields; diff --git a/fs/xfs/xfs_inode_item.h b/fs/xfs/xfs_inode_item.h index 9a467958ecd..d3dee61e6d9 100644 --- a/fs/xfs/xfs_inode_item.h +++ b/fs/xfs/xfs_inode_item.h @@ -103,12 +103,6 @@ typedef struct xfs_inode_log_format_64 { XFS_ILOG_ADATA | XFS_ILOG_AEXT | \ XFS_ILOG_ABROOT) -#define XFS_ILI_HOLD 0x1 -#define XFS_ILI_IOLOCKED_EXCL 0x2 -#define XFS_ILI_IOLOCKED_SHARED 0x4 - -#define XFS_ILI_IOLOCKED_ANY (XFS_ILI_IOLOCKED_EXCL | XFS_ILI_IOLOCKED_SHARED) - static inline int xfs_ilog_fbroot(int w) { return (w == XFS_DATA_FORK ? XFS_ILOG_DBROOT : XFS_ILOG_ABROOT); @@ -137,7 +131,7 @@ typedef struct xfs_inode_log_item { struct xfs_inode *ili_inode; /* inode ptr */ xfs_lsn_t ili_flush_lsn; /* lsn at last flush */ xfs_lsn_t ili_last_lsn; /* lsn at last transaction */ - unsigned short ili_flags; /* misc flags */ + unsigned short ili_lock_flags; /* lock flags */ unsigned short ili_logged; /* flushed logged data */ unsigned int ili_last_fields; /* fields when flushed */ struct xfs_bmbt_rec *ili_extents_buf; /* array of logged @@ -161,8 +155,8 @@ static inline int xfs_inode_clean(xfs_inode_t *ip) 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 *); -extern void xfs_istale_done(struct xfs_buf *, xfs_inode_log_item_t *); +extern void xfs_iflush_done(struct xfs_buf *, struct xfs_log_item *); +extern void xfs_istale_done(struct xfs_buf *, struct xfs_log_item *); extern void xfs_iflush_abort(struct xfs_inode *); extern int xfs_inode_item_format_convert(xfs_log_iovec_t *, xfs_inode_log_format_t *); diff --git a/fs/xfs/xfs_iomap.c b/fs/xfs/xfs_iomap.c index ef14943829d..20576146369 100644 --- a/fs/xfs/xfs_iomap.c +++ b/fs/xfs/xfs_iomap.c @@ -23,19 +23,14 @@ #include "xfs_trans.h" #include "xfs_sb.h" #include "xfs_ag.h" -#include "xfs_dir2.h" #include "xfs_alloc.h" -#include "xfs_dmapi.h" #include "xfs_quota.h" #include "xfs_mount.h" #include "xfs_bmap_btree.h" #include "xfs_alloc_btree.h" #include "xfs_ialloc_btree.h" -#include "xfs_dir2_sf.h" -#include "xfs_attr_sf.h" #include "xfs_dinode.h" #include "xfs_inode.h" -#include "xfs_ialloc.h" #include "xfs_btree.h" #include "xfs_bmap.h" #include "xfs_rtalloc.h" @@ -123,7 +118,7 @@ xfs_iomap( error = xfs_bmapi(NULL, ip, offset_fsb, (xfs_filblks_t)(end_fsb - offset_fsb), bmapi_flags, NULL, 0, imap, - nimaps, NULL, NULL); + nimaps, NULL); if (error) goto out; @@ -138,7 +133,7 @@ xfs_iomap( break; } - if (flags & (BMAPI_DIRECT|BMAPI_MMAP)) { + if (flags & BMAPI_DIRECT) { error = xfs_iomap_write_direct(ip, offset, count, flags, imap, nimaps); } else { @@ -247,7 +242,7 @@ xfs_iomap_write_direct( xfs_off_t offset, size_t count, int flags, - xfs_bmbt_irec_t *ret_imap, + xfs_bmbt_irec_t *imap, int *nmaps) { xfs_mount_t *mp = ip->i_mount; @@ -261,7 +256,6 @@ xfs_iomap_write_direct( int quota_flag; int rt; xfs_trans_t *tp; - xfs_bmbt_irec_t imap; xfs_bmap_free_t free_list; uint qblocks, resblks, resrtextents; int committed; @@ -285,10 +279,10 @@ xfs_iomap_write_direct( if (error) goto error_out; } else { - if (*nmaps && (ret_imap->br_startblock == HOLESTARTBLOCK)) + if (*nmaps && (imap->br_startblock == HOLESTARTBLOCK)) last_fsb = MIN(last_fsb, (xfs_fileoff_t) - ret_imap->br_blockcount + - ret_imap->br_startoff); + imap->br_blockcount + + imap->br_startoff); } count_fsb = last_fsb - offset_fsb; ASSERT(count_fsb > 0); @@ -334,20 +328,22 @@ xfs_iomap_write_direct( if (error) goto error1; - xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL); - xfs_trans_ihold(tp, ip); + xfs_trans_ijoin(tp, ip); bmapi_flag = XFS_BMAPI_WRITE; if ((flags & BMAPI_DIRECT) && (offset < ip->i_size || extsz)) bmapi_flag |= XFS_BMAPI_PREALLOC; /* - * Issue the xfs_bmapi() call to allocate the blocks + * Issue the xfs_bmapi() call to allocate the blocks. + * + * From this point onwards we overwrite the imap pointer that the + * caller gave to us. */ xfs_bmap_init(&free_list, &firstfsb); nimaps = 1; error = xfs_bmapi(tp, ip, offset_fsb, count_fsb, bmapi_flag, - &firstfsb, 0, &imap, &nimaps, &free_list, NULL); + &firstfsb, 0, imap, &nimaps, &free_list); if (error) goto error0; @@ -369,12 +365,11 @@ xfs_iomap_write_direct( goto error_out; } - if (!(imap.br_startblock || XFS_IS_REALTIME_INODE(ip))) { - error = xfs_cmn_err_fsblock_zero(ip, &imap); + if (!(imap->br_startblock || XFS_IS_REALTIME_INODE(ip))) { + error = xfs_cmn_err_fsblock_zero(ip, imap); goto error_out; } - *ret_imap = imap; *nmaps = 1; return 0; @@ -425,7 +420,7 @@ xfs_iomap_eof_want_preallocate( imaps = nimaps; firstblock = NULLFSBLOCK; error = xfs_bmapi(NULL, ip, start_fsb, count_fsb, 0, - &firstblock, 0, imap, &imaps, NULL, NULL); + &firstblock, 0, imap, &imaps, NULL); if (error) return error; for (n = 0; n < imaps; n++) { @@ -500,7 +495,7 @@ retry: (xfs_filblks_t)(last_fsb - offset_fsb), XFS_BMAPI_DELAY | XFS_BMAPI_WRITE | XFS_BMAPI_ENTIRE, &firstblock, 1, imap, - &nimaps, NULL, NULL); + &nimaps, NULL); if (error && (error != ENOSPC)) return XFS_ERROR(error); @@ -548,7 +543,7 @@ xfs_iomap_write_allocate( xfs_inode_t *ip, xfs_off_t offset, size_t count, - xfs_bmbt_irec_t *map, + xfs_bmbt_irec_t *imap, int *retmap) { xfs_mount_t *mp = ip->i_mount; @@ -557,7 +552,6 @@ xfs_iomap_write_allocate( xfs_fsblock_t first_block; xfs_bmap_free_t free_list; xfs_filblks_t count_fsb; - xfs_bmbt_irec_t imap; xfs_trans_t *tp; int nimaps, committed; int error = 0; @@ -573,8 +567,8 @@ xfs_iomap_write_allocate( return XFS_ERROR(error); offset_fsb = XFS_B_TO_FSBT(mp, offset); - count_fsb = map->br_blockcount; - map_start_fsb = map->br_startoff; + count_fsb = imap->br_blockcount; + map_start_fsb = imap->br_startoff; XFS_STATS_ADD(xs_xstrat_bytes, XFS_FSB_TO_B(mp, count_fsb)); @@ -602,8 +596,7 @@ xfs_iomap_write_allocate( return XFS_ERROR(error); } xfs_ilock(ip, XFS_ILOCK_EXCL); - xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL); - xfs_trans_ihold(tp, ip); + xfs_trans_ijoin(tp, ip); xfs_bmap_init(&free_list, &first_block); @@ -654,10 +647,15 @@ xfs_iomap_write_allocate( } } - /* Go get the actual blocks */ + /* + * Go get the actual blocks. + * + * From this point onwards we overwrite the imap + * pointer that the caller gave to us. + */ error = xfs_bmapi(tp, ip, map_start_fsb, count_fsb, XFS_BMAPI_WRITE, &first_block, 1, - &imap, &nimaps, &free_list, NULL); + imap, &nimaps, &free_list); if (error) goto trans_cancel; @@ -676,13 +674,12 @@ xfs_iomap_write_allocate( * See if we were able to allocate an extent that * covers at least part of the callers request */ - if (!(imap.br_startblock || XFS_IS_REALTIME_INODE(ip))) - return xfs_cmn_err_fsblock_zero(ip, &imap); + if (!(imap->br_startblock || XFS_IS_REALTIME_INODE(ip))) + return xfs_cmn_err_fsblock_zero(ip, imap); - if ((offset_fsb >= imap.br_startoff) && - (offset_fsb < (imap.br_startoff + - imap.br_blockcount))) { - *map = imap; + if ((offset_fsb >= imap->br_startoff) && + (offset_fsb < (imap->br_startoff + + imap->br_blockcount))) { *retmap = 1; XFS_STATS_INC(xs_xstrat_quick); return 0; @@ -692,8 +689,8 @@ xfs_iomap_write_allocate( * So far we have not mapped the requested part of the * file, just surrounding data, try again. */ - count_fsb -= imap.br_blockcount; - map_start_fsb = imap.br_startoff + imap.br_blockcount; + count_fsb -= imap->br_blockcount; + map_start_fsb = imap->br_startoff + imap->br_blockcount; } trans_cancel: @@ -766,8 +763,7 @@ xfs_iomap_write_unwritten( } xfs_ilock(ip, XFS_ILOCK_EXCL); - xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL); - xfs_trans_ihold(tp, ip); + xfs_trans_ijoin(tp, ip); /* * Modify the unwritten extent state of the buffer. @@ -776,7 +772,7 @@ xfs_iomap_write_unwritten( nimaps = 1; error = xfs_bmapi(tp, ip, offset_fsb, count_fsb, XFS_BMAPI_WRITE|XFS_BMAPI_CONVERT, &firstfsb, - 1, &imap, &nimaps, &free_list, NULL); + 1, &imap, &nimaps, &free_list); if (error) goto error_on_bmapi_transaction; diff --git a/fs/xfs/xfs_iomap.h b/fs/xfs/xfs_iomap.h index 81ac4afd45b..7748a430f50 100644 --- a/fs/xfs/xfs_iomap.h +++ b/fs/xfs/xfs_iomap.h @@ -18,17 +18,16 @@ #ifndef __XFS_IOMAP_H__ #define __XFS_IOMAP_H__ -typedef enum { - /* base extent manipulation calls */ - BMAPI_READ = (1 << 0), /* read extents */ - BMAPI_WRITE = (1 << 1), /* create extents */ - BMAPI_ALLOCATE = (1 << 2), /* delayed allocate to real extents */ - /* modifiers */ - BMAPI_IGNSTATE = (1 << 4), /* ignore unwritten state on read */ - BMAPI_DIRECT = (1 << 5), /* direct instead of buffered write */ - BMAPI_MMAP = (1 << 6), /* allocate for mmap write */ - BMAPI_TRYLOCK = (1 << 7), /* non-blocking request */ -} bmapi_flags_t; +/* base extent manipulation calls */ +#define BMAPI_READ (1 << 0) /* read extents */ +#define BMAPI_WRITE (1 << 1) /* create extents */ +#define BMAPI_ALLOCATE (1 << 2) /* delayed allocate to real extents */ + +/* modifiers */ +#define BMAPI_IGNSTATE (1 << 4) /* ignore unwritten state on read */ +#define BMAPI_DIRECT (1 << 5) /* direct instead of buffered write */ +#define BMAPI_MMA (1 << 6) /* allocate for mmap write */ +#define BMAPI_TRYLOCK (1 << 7) /* non-blocking request */ #define BMAPI_FLAGS \ { BMAPI_READ, "READ" }, \ @@ -36,7 +35,6 @@ typedef enum { { BMAPI_ALLOCATE, "ALLOCATE" }, \ { BMAPI_IGNSTATE, "IGNSTATE" }, \ { BMAPI_DIRECT, "DIRECT" }, \ - { BMAPI_MMAP, "MMAP" }, \ { BMAPI_TRYLOCK, "TRYLOCK" } struct xfs_inode; diff --git a/fs/xfs/xfs_itable.c b/fs/xfs/xfs_itable.c index 2b86f861051..7e3626e5925 100644 --- a/fs/xfs/xfs_itable.c +++ b/fs/xfs/xfs_itable.c @@ -24,20 +24,17 @@ #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_dir2_sf.h" -#include "xfs_attr_sf.h" #include "xfs_dinode.h" #include "xfs_inode.h" #include "xfs_ialloc.h" #include "xfs_itable.h" #include "xfs_error.h" #include "xfs_btree.h" +#include "xfs_trace.h" STATIC int xfs_internal_inum( @@ -143,7 +140,8 @@ xfs_bulkstat_one_int( buf->bs_blocks = dic->di_nblocks + ip->i_delayed_blks; break; } - xfs_iput(ip, XFS_ILOCK_SHARED); + xfs_iunlock(ip, XFS_ILOCK_SHARED); + IRELE(ip); error = formatter(buffer, ubsize, ubused, buf); diff --git a/fs/xfs/xfs_log.c b/fs/xfs/xfs_log.c index 5215abc8023..925d572bf0f 100644 --- a/fs/xfs/xfs_log.c +++ b/fs/xfs/xfs_log.c @@ -24,8 +24,6 @@ #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_error.h" #include "xfs_log_priv.h" @@ -35,8 +33,6 @@ #include "xfs_ialloc_btree.h" #include "xfs_log_recover.h" #include "xfs_trans_priv.h" -#include "xfs_dir2_sf.h" -#include "xfs_attr_sf.h" #include "xfs_dinode.h" #include "xfs_inode.h" #include "xfs_rw.h" @@ -337,7 +333,6 @@ xfs_log_reserve( int retval = 0; ASSERT(client == XFS_TRANSACTION || client == XFS_LOG); - ASSERT((flags & XFS_LOG_NOSLEEP) == 0); if (XLOG_FORCED_SHUTDOWN(log)) return XFS_ERROR(EIO); @@ -552,7 +547,7 @@ xfs_log_unmount_write(xfs_mount_t *mp) .magic = XLOG_UNMOUNT_TYPE, }; struct xfs_log_iovec reg = { - .i_addr = (void *)&magic, + .i_addr = &magic, .i_len = sizeof(magic), .i_type = XLOG_REG_TYPE_UNMOUNT, }; @@ -1047,7 +1042,6 @@ xlog_alloc_log(xfs_mount_t *mp, xlog_in_core_t *iclog, *prev_iclog=NULL; xfs_buf_t *bp; int i; - int iclogsize; int error = ENOMEM; uint log2_size = 0; @@ -1127,7 +1121,6 @@ xlog_alloc_log(xfs_mount_t *mp, * with different amounts of memory. See the definition of * xlog_in_core_t in xfs_log_priv.h for details. */ - iclogsize = log->l_iclog_size; ASSERT(log->l_iclog_size >= 4096); for (i=0; i < log->l_iclog_bufs; i++) { *iclogp = kmem_zalloc(sizeof(xlog_in_core_t), KM_MAYFAIL); @@ -1428,11 +1421,8 @@ xlog_sync(xlog_t *log, XFS_BUF_BUSY(bp); XFS_BUF_ASYNC(bp); bp->b_flags |= XBF_LOG_BUFFER; - /* - * Do an ordered write for the log block. - * Its unnecessary to flush the first split block in the log wrap case. - */ - if (!split && (log->l_mp->m_flags & XFS_MOUNT_BARRIER)) + + if (log->l_mp->m_flags & XFS_MOUNT_BARRIER) XFS_BUF_ORDERED(bp); ASSERT(XFS_BUF_ADDR(bp) <= log->l_logBBsize-1); diff --git a/fs/xfs/xfs_log.h b/fs/xfs/xfs_log.h index 04c78e642cc..916eb7db14d 100644 --- a/fs/xfs/xfs_log.h +++ b/fs/xfs/xfs_log.h @@ -55,14 +55,10 @@ static inline xfs_lsn_t _lsn_cmp(xfs_lsn_t lsn1, xfs_lsn_t lsn2) /* * Flags to xfs_log_reserve() * - * XFS_LOG_SLEEP: If space is not available, sleep (default) - * XFS_LOG_NOSLEEP: If space is not available, return error * XFS_LOG_PERM_RESERV: Permanent reservation. When writes are * performed against this type of reservation, the reservation * is not decreased. Long running transactions should use this. */ -#define XFS_LOG_SLEEP 0x0 -#define XFS_LOG_NOSLEEP 0x1 #define XFS_LOG_PERM_RESERV 0x2 /* @@ -104,7 +100,7 @@ static inline xfs_lsn_t _lsn_cmp(xfs_lsn_t lsn1, xfs_lsn_t lsn2) #define XLOG_REG_TYPE_MAX 19 typedef struct xfs_log_iovec { - xfs_caddr_t i_addr; /* beginning address of region */ + void *i_addr; /* beginning address of region */ int i_len; /* length in bytes of region */ uint i_type; /* type of region */ } xfs_log_iovec_t; @@ -201,9 +197,4 @@ int xfs_log_commit_cil(struct xfs_mount *mp, struct xfs_trans *tp, bool xfs_log_item_in_current_chkpt(struct xfs_log_item *lip); #endif - - -extern int xlog_debug; /* set to 1 to enable real log */ - - #endif /* __XFS_LOG_H__ */ diff --git a/fs/xfs/xfs_log_cil.c b/fs/xfs/xfs_log_cil.c index bb17cc044bf..31e4ea2d19a 100644 --- a/fs/xfs/xfs_log_cil.c +++ b/fs/xfs/xfs_log_cil.c @@ -26,8 +26,6 @@ #include "xfs_log_priv.h" #include "xfs_sb.h" #include "xfs_ag.h" -#include "xfs_dir2.h" -#include "xfs_dmapi.h" #include "xfs_mount.h" #include "xfs_error.h" #include "xfs_alloc.h" @@ -554,7 +552,7 @@ xlog_cil_push( thdr.th_type = XFS_TRANS_CHECKPOINT; thdr.th_tid = tic->t_tid; thdr.th_num_items = num_iovecs; - lhdr.i_addr = (xfs_caddr_t)&thdr; + lhdr.i_addr = &thdr; lhdr.i_len = sizeof(xfs_trans_header_t); lhdr.i_type = XLOG_REG_TYPE_TRANSHDR; tic->t_curr_res -= lhdr.i_len + sizeof(xlog_op_header_t); diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c index 9ac5cfab27b..6f3f5fa37ac 100644 --- a/fs/xfs/xfs_log_recover.c +++ b/fs/xfs/xfs_log_recover.c @@ -24,15 +24,11 @@ #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_error.h" #include "xfs_bmap_btree.h" #include "xfs_alloc_btree.h" #include "xfs_ialloc_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" @@ -1565,9 +1561,7 @@ xlog_recover_reorder_trans( list_splice_init(&trans->r_itemq, &sort_list); list_for_each_entry_safe(item, n, &sort_list, ri_list) { - xfs_buf_log_format_t *buf_f; - - buf_f = (xfs_buf_log_format_t *)item->ri_buf[0].i_addr; + xfs_buf_log_format_t *buf_f = item->ri_buf[0].i_addr; switch (ITEM_TYPE(item)) { case XFS_LI_BUF: @@ -1892,9 +1886,8 @@ xlog_recover_do_inode_buffer( * current di_next_unlinked field. Extract its value * and copy it to the buffer copy. */ - logged_nextp = (xfs_agino_t *) - ((char *)(item->ri_buf[item_index].i_addr) + - (next_unlinked_offset - reg_buf_offset)); + logged_nextp = item->ri_buf[item_index].i_addr + + next_unlinked_offset - reg_buf_offset; if (unlikely(*logged_nextp == 0)) { xfs_fs_cmn_err(CE_ALERT, mp, "bad inode buffer log record (ptr = 0x%p, bp = 0x%p). XFS trying to replay bad (0) inode di_next_unlinked field", @@ -1973,8 +1966,7 @@ xlog_recover_do_reg_buffer( item->ri_buf[i].i_len, __func__); goto next; } - error = xfs_qm_dqcheck((xfs_disk_dquot_t *) - item->ri_buf[i].i_addr, + error = xfs_qm_dqcheck(item->ri_buf[i].i_addr, -1, 0, XFS_QMOPT_DOWARN, "dquot_buf_recover"); if (error) @@ -2187,7 +2179,7 @@ xlog_recover_do_buffer_trans( xlog_recover_item_t *item, int pass) { - xfs_buf_log_format_t *buf_f; + xfs_buf_log_format_t *buf_f = item->ri_buf[0].i_addr; xfs_mount_t *mp; xfs_buf_t *bp; int error; @@ -2197,8 +2189,6 @@ xlog_recover_do_buffer_trans( ushort flags; uint buf_flags; - buf_f = (xfs_buf_log_format_t *)item->ri_buf[0].i_addr; - if (pass == XLOG_RECOVER_PASS1) { /* * In this pass we're only looking for buf items @@ -2319,10 +2309,9 @@ xlog_recover_do_inode_trans( } if (item->ri_buf[0].i_len == sizeof(xfs_inode_log_format_t)) { - in_f = (xfs_inode_log_format_t *)item->ri_buf[0].i_addr; + in_f = item->ri_buf[0].i_addr; } else { - in_f = (xfs_inode_log_format_t *)kmem_alloc( - sizeof(xfs_inode_log_format_t), KM_SLEEP); + in_f = kmem_alloc(sizeof(xfs_inode_log_format_t), KM_SLEEP); need_free = 1; error = xfs_inode_item_format_convert(&item->ri_buf[0], in_f); if (error) @@ -2370,7 +2359,7 @@ xlog_recover_do_inode_trans( error = EFSCORRUPTED; goto error; } - dicp = (xfs_icdinode_t *)(item->ri_buf[1].i_addr); + dicp = item->ri_buf[1].i_addr; if (unlikely(dicp->di_magic != XFS_DINODE_MAGIC)) { xfs_buf_relse(bp); xfs_fs_cmn_err(CE_ALERT, mp, @@ -2461,7 +2450,7 @@ xlog_recover_do_inode_trans( } /* The core is in in-core format */ - xfs_dinode_to_disk(dip, (xfs_icdinode_t *)item->ri_buf[1].i_addr); + xfs_dinode_to_disk(dip, item->ri_buf[1].i_addr); /* the rest is in on-disk format */ if (item->ri_buf[1].i_len > sizeof(struct xfs_icdinode)) { @@ -2578,7 +2567,7 @@ xlog_recover_do_quotaoff_trans( return (0); } - qoff_f = (xfs_qoff_logformat_t *)item->ri_buf[0].i_addr; + qoff_f = item->ri_buf[0].i_addr; ASSERT(qoff_f); /* @@ -2622,9 +2611,8 @@ xlog_recover_do_dquot_trans( if (mp->m_qflags == 0) return (0); - recddq = (xfs_disk_dquot_t *)item->ri_buf[1].i_addr; - - if (item->ri_buf[1].i_addr == NULL) { + recddq = item->ri_buf[1].i_addr; + if (recddq == NULL) { cmn_err(CE_ALERT, "XFS: NULL dquot in %s.", __func__); return XFS_ERROR(EIO); @@ -2654,7 +2642,7 @@ xlog_recover_do_dquot_trans( * The other possibility, of course, is that the quota subsystem was * removed since the last mount - ENOSYS. */ - dq_f = (xfs_dq_logformat_t *)item->ri_buf[0].i_addr; + dq_f = item->ri_buf[0].i_addr; ASSERT(dq_f); if ((error = xfs_qm_dqcheck(recddq, dq_f->qlf_id, @@ -2721,7 +2709,7 @@ xlog_recover_do_efi_trans( return 0; } - efi_formatp = (xfs_efi_log_format_t *)item->ri_buf[0].i_addr; + efi_formatp = item->ri_buf[0].i_addr; mp = log->l_mp; efip = xfs_efi_init(mp, efi_formatp->efi_nextents); @@ -2767,7 +2755,7 @@ xlog_recover_do_efd_trans( return; } - efd_formatp = (xfs_efd_log_format_t *)item->ri_buf[0].i_addr; + efd_formatp = item->ri_buf[0].i_addr; ASSERT((item->ri_buf[0].i_len == (sizeof(xfs_efd_log_format_32_t) + ((efd_formatp->efd_nextents - 1) * sizeof(xfs_extent_32_t)))) || (item->ri_buf[0].i_len == (sizeof(xfs_efd_log_format_64_t) + diff --git a/fs/xfs/xfs_mount.c b/fs/xfs/xfs_mount.c index 69f62d8b281..aeb9d72ebf6 100644 --- a/fs/xfs/xfs_mount.c +++ b/fs/xfs/xfs_mount.c @@ -25,13 +25,10 @@ #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_dir2_sf.h" -#include "xfs_attr_sf.h" #include "xfs_dinode.h" #include "xfs_inode.h" #include "xfs_btree.h" diff --git a/fs/xfs/xfs_mount.h b/fs/xfs/xfs_mount.h index 5761087ee8e..622da2179a5 100644 --- a/fs/xfs/xfs_mount.h +++ b/fs/xfs/xfs_mount.h @@ -66,65 +66,6 @@ struct xfs_nameops; struct xfs_ail; struct xfs_quotainfo; - -/* - * Prototypes and functions for the Data Migration subsystem. - */ - -typedef int (*xfs_send_data_t)(int, struct xfs_inode *, - xfs_off_t, size_t, int, int *); -typedef int (*xfs_send_mmap_t)(struct vm_area_struct *, uint); -typedef int (*xfs_send_destroy_t)(struct xfs_inode *, dm_right_t); -typedef int (*xfs_send_namesp_t)(dm_eventtype_t, struct xfs_mount *, - struct xfs_inode *, dm_right_t, - struct xfs_inode *, dm_right_t, - const unsigned char *, const unsigned char *, - mode_t, int, int); -typedef int (*xfs_send_mount_t)(struct xfs_mount *, dm_right_t, - char *, char *); -typedef void (*xfs_send_unmount_t)(struct xfs_mount *, struct xfs_inode *, - dm_right_t, mode_t, int, int); - -typedef struct xfs_dmops { - xfs_send_data_t xfs_send_data; - xfs_send_mmap_t xfs_send_mmap; - xfs_send_destroy_t xfs_send_destroy; - xfs_send_namesp_t xfs_send_namesp; - xfs_send_mount_t xfs_send_mount; - xfs_send_unmount_t xfs_send_unmount; -} xfs_dmops_t; - -#define XFS_DMAPI_UNMOUNT_FLAGS(mp) \ - (((mp)->m_dmevmask & (1 << DM_EVENT_UNMOUNT)) ? 0 : DM_FLAGS_UNWANTED) - -#define XFS_SEND_DATA(mp, ev,ip,off,len,fl,lock) \ - (*(mp)->m_dm_ops->xfs_send_data)(ev,ip,off,len,fl,lock) -#define XFS_SEND_MMAP(mp, vma,fl) \ - (*(mp)->m_dm_ops->xfs_send_mmap)(vma,fl) -#define XFS_SEND_DESTROY(mp, ip,right) \ - (*(mp)->m_dm_ops->xfs_send_destroy)(ip,right) -#define XFS_SEND_NAMESP(mp, ev,b1,r1,b2,r2,n1,n2,mode,rval,fl) \ - (*(mp)->m_dm_ops->xfs_send_namesp)(ev,NULL,b1,r1,b2,r2,n1,n2,mode,rval,fl) -#define XFS_SEND_MOUNT(mp,right,path,name) \ - (*(mp)->m_dm_ops->xfs_send_mount)(mp,right,path,name) -#define XFS_SEND_PREUNMOUNT(mp) \ -do { \ - if (mp->m_flags & XFS_MOUNT_DMAPI) { \ - (*(mp)->m_dm_ops->xfs_send_namesp)(DM_EVENT_PREUNMOUNT, mp, \ - (mp)->m_rootip, DM_RIGHT_NULL, \ - (mp)->m_rootip, DM_RIGHT_NULL, \ - NULL, NULL, 0, 0, XFS_DMAPI_UNMOUNT_FLAGS(mp)); \ - } \ -} while (0) -#define XFS_SEND_UNMOUNT(mp) \ -do { \ - if (mp->m_flags & XFS_MOUNT_DMAPI) { \ - (*(mp)->m_dm_ops->xfs_send_unmount)(mp, (mp)->m_rootip, \ - DM_RIGHT_NULL, 0, 0, XFS_DMAPI_UNMOUNT_FLAGS(mp)); \ - } \ -} while (0) - - #ifdef HAVE_PERCPU_SB /* @@ -241,8 +182,6 @@ typedef struct xfs_mount { uint m_chsize; /* size of next field */ struct xfs_chash *m_chash; /* fs private inode per-cluster * hash table */ - struct xfs_dmops *m_dm_ops; /* vector of DMI ops */ - struct xfs_qmops *m_qm_ops; /* vector of XQM ops */ atomic_t m_active_trans; /* number trans frozen */ #ifdef HAVE_PERCPU_SB xfs_icsb_cnts_t __percpu *m_sb_cnts; /* per-cpu superblock counters */ @@ -269,7 +208,6 @@ typedef struct xfs_mount { must be synchronous except for space allocations */ #define XFS_MOUNT_DELAYLOG (1ULL << 1) /* delayed logging is enabled */ -#define XFS_MOUNT_DMAPI (1ULL << 2) /* dmapi is enabled */ #define XFS_MOUNT_WAS_CLEAN (1ULL << 3) #define XFS_MOUNT_FS_SHUTDOWN (1ULL << 4) /* atomic stop of all filesystem operations, typically for @@ -282,8 +220,6 @@ typedef struct xfs_mount { #define XFS_MOUNT_GRPID (1ULL << 9) /* group-ID assigned from directory */ #define XFS_MOUNT_NORECOVERY (1ULL << 10) /* no recovery - dirty fs */ #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*/ #define XFS_MOUNT_32BITINODES (1ULL << 14) /* do not create inodes above * 32 bits in size */ #define XFS_MOUNT_SMALL_INUMS (1ULL << 15) /* users wants 32bit inodes */ @@ -440,11 +376,6 @@ extern int xfs_sb_validate_fsb_count(struct xfs_sb *, __uint64_t); extern int xfs_dev_is_read_only(struct xfs_mount *, char *); -extern int xfs_dmops_get(struct xfs_mount *); -extern void xfs_dmops_put(struct xfs_mount *); - -extern struct xfs_dmops xfs_dmcore_xfs; - #endif /* __KERNEL__ */ extern void xfs_mod_sb(struct xfs_trans *, __int64_t); diff --git a/fs/xfs/xfs_rename.c b/fs/xfs/xfs_rename.c index fc1cda23b81..8fca957200d 100644 --- a/fs/xfs/xfs_rename.c +++ b/fs/xfs/xfs_rename.c @@ -24,12 +24,9 @@ #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_dir2_sf.h" -#include "xfs_attr_sf.h" #include "xfs_dinode.h" #include "xfs_inode.h" #include "xfs_inode_item.h" @@ -116,20 +113,7 @@ xfs_rename( int spaceres; int num_inodes; - xfs_itrace_entry(src_dp); - xfs_itrace_entry(target_dp); - - if (DM_EVENT_ENABLED(src_dp, DM_EVENT_RENAME) || - DM_EVENT_ENABLED(target_dp, DM_EVENT_RENAME)) { - error = XFS_SEND_NAMESP(mp, DM_EVENT_RENAME, - src_dp, DM_RIGHT_NULL, - target_dp, DM_RIGHT_NULL, - src_name->name, target_name->name, - 0, 0, 0); - if (error) - return error; - } - /* Return through std_return after this point. */ + trace_xfs_rename(src_dp, target_dp, src_name, target_name); new_parent = (src_dp != target_dp); src_is_directory = ((src_ip->i_d.di_mode & S_IFMT) == S_IFDIR); @@ -184,26 +168,14 @@ xfs_rename( /* * 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 - * directories since trans_commit & trans_cancel will decrement - * them when they unlock the inodes. Also, we need to be careful - * not to add an inode to the transaction more than once. + * them. */ - IHOLD(src_dp); - xfs_trans_ijoin(tp, src_dp, XFS_ILOCK_EXCL); - - if (new_parent) { - IHOLD(target_dp); - xfs_trans_ijoin(tp, target_dp, XFS_ILOCK_EXCL); - } - - IHOLD(src_ip); - xfs_trans_ijoin(tp, src_ip, XFS_ILOCK_EXCL); - - if (target_ip) { - IHOLD(target_ip); - xfs_trans_ijoin(tp, target_ip, XFS_ILOCK_EXCL); - } + xfs_trans_ijoin_ref(tp, src_dp, XFS_ILOCK_EXCL); + if (new_parent) + xfs_trans_ijoin_ref(tp, target_dp, XFS_ILOCK_EXCL); + xfs_trans_ijoin_ref(tp, src_ip, XFS_ILOCK_EXCL); + if (target_ip) + xfs_trans_ijoin_ref(tp, target_ip, XFS_ILOCK_EXCL); /* * If we are using project inheritance, we only allow renames @@ -369,26 +341,13 @@ xfs_rename( * trans_commit will unlock src_ip, target_ip & decrement * the vnode references. */ - error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES); - - /* Fall through to std_return with error = 0 or errno from - * xfs_trans_commit */ -std_return: - if (DM_EVENT_ENABLED(src_dp, DM_EVENT_POSTRENAME) || - DM_EVENT_ENABLED(target_dp, DM_EVENT_POSTRENAME)) { - (void) XFS_SEND_NAMESP (mp, DM_EVENT_POSTRENAME, - src_dp, DM_RIGHT_NULL, - target_dp, DM_RIGHT_NULL, - src_name->name, target_name->name, - 0, error, 0); - } - return error; + return xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES); abort_return: cancel_flags |= XFS_TRANS_ABORT; - /* FALLTHROUGH */ error_return: xfs_bmap_cancel(&free_list); xfs_trans_cancel(tp, cancel_flags); - goto std_return; + std_return: + return error; } diff --git a/fs/xfs/xfs_rtalloc.c b/fs/xfs/xfs_rtalloc.c index a2d32ce335a..891260fea11 100644 --- a/fs/xfs/xfs_rtalloc.c +++ b/fs/xfs/xfs_rtalloc.c @@ -25,17 +25,10 @@ #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_dir2_sf.h" -#include "xfs_attr_sf.h" #include "xfs_dinode.h" #include "xfs_inode.h" -#include "xfs_btree.h" -#include "xfs_ialloc.h" #include "xfs_alloc.h" #include "xfs_bmap.h" #include "xfs_rtalloc.h" @@ -129,7 +122,7 @@ xfs_growfs_rt_alloc( cancelflags |= XFS_TRANS_ABORT; error = xfs_bmapi(tp, ip, oblocks, nblocks - oblocks, XFS_BMAPI_WRITE | XFS_BMAPI_METADATA, &firstblock, - resblks, &map, &nmap, &flist, NULL); + resblks, &map, &nmap, &flist); if (!error && nmap < 1) error = XFS_ERROR(ENOSPC); if (error) diff --git a/fs/xfs/xfs_rw.c b/fs/xfs/xfs_rw.c index e336742a58a..56861d5daae 100644 --- a/fs/xfs/xfs_rw.c +++ b/fs/xfs/xfs_rw.c @@ -24,27 +24,12 @@ #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_dir2_sf.h" -#include "xfs_attr_sf.h" #include "xfs_dinode.h" #include "xfs_inode.h" -#include "xfs_inode_item.h" -#include "xfs_itable.h" -#include "xfs_btree.h" -#include "xfs_alloc.h" -#include "xfs_ialloc.h" -#include "xfs_attr.h" -#include "xfs_bmap.h" #include "xfs_error.h" -#include "xfs_buf_item.h" #include "xfs_rw.h" -#include "xfs_trace.h" /* * Force a shutdown of the filesystem instantly while keeping diff --git a/fs/xfs/xfs_trans.c b/fs/xfs/xfs_trans.c index 28547dfce03..fdca7416c75 100644 --- a/fs/xfs/xfs_trans.c +++ b/fs/xfs/xfs_trans.c @@ -1,5 +1,6 @@ /* * Copyright (c) 2000-2003,2005 Silicon Graphics, Inc. + * Copyright (C) 2010 Red Hat, Inc. * All Rights Reserved. * * This program is free software; you can redistribute it and/or @@ -24,16 +25,12 @@ #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_error.h" #include "xfs_da_btree.h" #include "xfs_bmap_btree.h" #include "xfs_alloc_btree.h" #include "xfs_ialloc_btree.h" -#include "xfs_dir2_sf.h" -#include "xfs_attr_sf.h" #include "xfs_dinode.h" #include "xfs_inode.h" #include "xfs_btree.h" @@ -47,6 +44,7 @@ #include "xfs_trace.h" kmem_zone_t *xfs_trans_zone; +kmem_zone_t *xfs_log_item_desc_zone; /* @@ -597,8 +595,7 @@ _xfs_trans_alloc( tp->t_magic = XFS_TRANS_MAGIC; tp->t_type = type; tp->t_mountp = mp; - tp->t_items_free = XFS_LIC_NUM_SLOTS; - xfs_lic_init(&(tp->t_items)); + INIT_LIST_HEAD(&tp->t_items); INIT_LIST_HEAD(&tp->t_busy); return tp; } @@ -643,8 +640,7 @@ xfs_trans_dup( ntp->t_magic = XFS_TRANS_MAGIC; ntp->t_type = tp->t_type; ntp->t_mountp = tp->t_mountp; - ntp->t_items_free = XFS_LIC_NUM_SLOTS; - xfs_lic_init(&(ntp->t_items)); + INIT_LIST_HEAD(&ntp->t_items); INIT_LIST_HEAD(&ntp->t_busy); ASSERT(tp->t_flags & XFS_TRANS_PERM_LOG_RES); @@ -1124,6 +1120,108 @@ xfs_trans_unreserve_and_mod_sb( } /* + * Add the given log item to the transaction's list of log items. + * + * The log item will now point to its new descriptor with its li_desc field. + */ +void +xfs_trans_add_item( + struct xfs_trans *tp, + struct xfs_log_item *lip) +{ + struct xfs_log_item_desc *lidp; + + ASSERT(lip->li_mountp = tp->t_mountp); + ASSERT(lip->li_ailp = tp->t_mountp->m_ail); + + lidp = kmem_zone_zalloc(xfs_log_item_desc_zone, KM_SLEEP | KM_NOFS); + + lidp->lid_item = lip; + lidp->lid_flags = 0; + lidp->lid_size = 0; + list_add_tail(&lidp->lid_trans, &tp->t_items); + + lip->li_desc = lidp; +} + +STATIC void +xfs_trans_free_item_desc( + struct xfs_log_item_desc *lidp) +{ + list_del_init(&lidp->lid_trans); + kmem_zone_free(xfs_log_item_desc_zone, lidp); +} + +/* + * Unlink and free the given descriptor. + */ +void +xfs_trans_del_item( + struct xfs_log_item *lip) +{ + xfs_trans_free_item_desc(lip->li_desc); + lip->li_desc = NULL; +} + +/* + * Unlock all of the items of a transaction and free all the descriptors + * of that transaction. + */ +STATIC void +xfs_trans_free_items( + struct xfs_trans *tp, + xfs_lsn_t commit_lsn, + int flags) +{ + struct xfs_log_item_desc *lidp, *next; + + list_for_each_entry_safe(lidp, next, &tp->t_items, lid_trans) { + struct xfs_log_item *lip = lidp->lid_item; + + lip->li_desc = NULL; + + if (commit_lsn != NULLCOMMITLSN) + IOP_COMMITTING(lip, commit_lsn); + if (flags & XFS_TRANS_ABORT) + lip->li_flags |= XFS_LI_ABORTED; + IOP_UNLOCK(lip); + + xfs_trans_free_item_desc(lidp); + } +} + +/* + * Unlock the items associated with a transaction. + * + * Items which were not logged should be freed. Those which were logged must + * still be tracked so they can be unpinned when the transaction commits. + */ +STATIC void +xfs_trans_unlock_items( + struct xfs_trans *tp, + xfs_lsn_t commit_lsn) +{ + struct xfs_log_item_desc *lidp, *next; + + list_for_each_entry_safe(lidp, next, &tp->t_items, lid_trans) { + struct xfs_log_item *lip = lidp->lid_item; + + lip->li_desc = NULL; + + if (commit_lsn != NULLCOMMITLSN) + IOP_COMMITTING(lip, commit_lsn); + IOP_UNLOCK(lip); + + /* + * Free the descriptor if the item is not dirty + * within this transaction. + */ + if (!(lidp->lid_flags & XFS_LID_DIRTY)) + xfs_trans_free_item_desc(lidp); + } +} + +/* * Total up the number of log iovecs needed to commit this * transaction. The transaction itself needs one for the * transaction header. Ask each dirty item in turn how many @@ -1134,30 +1232,27 @@ xfs_trans_count_vecs( struct xfs_trans *tp) { int nvecs; - xfs_log_item_desc_t *lidp; + struct xfs_log_item_desc *lidp; nvecs = 1; - lidp = xfs_trans_first_item(tp); - ASSERT(lidp != NULL); /* In the non-debug case we need to start bailing out if we * didn't find a log_item here, return zero and let trans_commit * deal with it. */ - if (lidp == NULL) + if (list_empty(&tp->t_items)) { + ASSERT(0); return 0; + } - while (lidp != NULL) { + list_for_each_entry(lidp, &tp->t_items, lid_trans) { /* * Skip items which aren't dirty in this transaction. */ - if (!(lidp->lid_flags & XFS_LID_DIRTY)) { - lidp = xfs_trans_next_item(tp, lidp); + if (!(lidp->lid_flags & XFS_LID_DIRTY)) continue; - } lidp->lid_size = IOP_SIZE(lidp->lid_item); nvecs += lidp->lid_size; - lidp = xfs_trans_next_item(tp, lidp); } return nvecs; @@ -1177,7 +1272,7 @@ xfs_trans_fill_vecs( struct xfs_trans *tp, struct xfs_log_iovec *log_vector) { - xfs_log_item_desc_t *lidp; + struct xfs_log_item_desc *lidp; struct xfs_log_iovec *vecp; uint nitems; @@ -1188,14 +1283,11 @@ xfs_trans_fill_vecs( vecp = log_vector + 1; nitems = 0; - lidp = xfs_trans_first_item(tp); - ASSERT(lidp); - while (lidp) { + ASSERT(!list_empty(&tp->t_items)); + list_for_each_entry(lidp, &tp->t_items, lid_trans) { /* Skip items which aren't dirty in this transaction. */ - if (!(lidp->lid_flags & XFS_LID_DIRTY)) { - lidp = xfs_trans_next_item(tp, lidp); + if (!(lidp->lid_flags & XFS_LID_DIRTY)) continue; - } /* * The item may be marked dirty but not log anything. This can @@ -1206,7 +1298,6 @@ xfs_trans_fill_vecs( IOP_FORMAT(lidp->lid_item, vecp); vecp += lidp->lid_size; IOP_PIN(lidp->lid_item); - lidp = xfs_trans_next_item(tp, lidp); } /* @@ -1284,7 +1375,7 @@ xfs_trans_item_committed( * log item flags, if anyone else stales the buffer we do not want to * pay any attention to it. */ - IOP_UNPIN(lip); + IOP_UNPIN(lip, 0); } /* @@ -1301,24 +1392,15 @@ xfs_trans_committed( struct xfs_trans *tp, int abortflag) { - xfs_log_item_desc_t *lidp; - xfs_log_item_chunk_t *licp; - xfs_log_item_chunk_t *next_licp; + struct xfs_log_item_desc *lidp, *next; /* Call the transaction's completion callback if there is one. */ if (tp->t_callback != NULL) tp->t_callback(tp, tp->t_callarg); - for (lidp = xfs_trans_first_item(tp); - lidp != NULL; - lidp = xfs_trans_next_item(tp, lidp)) { + list_for_each_entry_safe(lidp, next, &tp->t_items, lid_trans) { xfs_trans_item_committed(lidp->lid_item, tp->t_lsn, abortflag); - } - - /* free the item chunks, ignoring the embedded chunk */ - for (licp = tp->t_items.lic_next; licp != NULL; licp = next_licp) { - next_licp = licp->lic_next; - kmem_free(licp); + xfs_trans_free_item_desc(lidp); } xfs_trans_free(tp); @@ -1333,16 +1415,14 @@ xfs_trans_uncommit( struct xfs_trans *tp, uint flags) { - xfs_log_item_desc_t *lidp; + struct xfs_log_item_desc *lidp; - for (lidp = xfs_trans_first_item(tp); - lidp != NULL; - lidp = xfs_trans_next_item(tp, lidp)) { + list_for_each_entry(lidp, &tp->t_items, lid_trans) { /* * Unpin all but those that aren't dirty. */ if (lidp->lid_flags & XFS_LID_DIRTY) - IOP_UNPIN_REMOVE(lidp->lid_item, tp); + IOP_UNPIN(lidp->lid_item, 1); } xfs_trans_unreserve_and_mod_sb(tp); @@ -1508,33 +1588,28 @@ STATIC struct xfs_log_vec * xfs_trans_alloc_log_vecs( xfs_trans_t *tp) { - xfs_log_item_desc_t *lidp; + struct xfs_log_item_desc *lidp; struct xfs_log_vec *lv = NULL; struct xfs_log_vec *ret_lv = NULL; - lidp = xfs_trans_first_item(tp); /* Bail out if we didn't find a log item. */ - if (!lidp) { + if (list_empty(&tp->t_items)) { ASSERT(0); return NULL; } - while (lidp != NULL) { + list_for_each_entry(lidp, &tp->t_items, lid_trans) { struct xfs_log_vec *new_lv; /* Skip items which aren't dirty in this transaction. */ - if (!(lidp->lid_flags & XFS_LID_DIRTY)) { - lidp = xfs_trans_next_item(tp, lidp); + if (!(lidp->lid_flags & XFS_LID_DIRTY)) continue; - } /* Skip items that do not have any vectors for writing */ lidp->lid_size = IOP_SIZE(lidp->lid_item); - if (!lidp->lid_size) { - lidp = xfs_trans_next_item(tp, lidp); + if (!lidp->lid_size) continue; - } new_lv = kmem_zalloc(sizeof(*new_lv) + lidp->lid_size * sizeof(struct xfs_log_iovec), @@ -1549,7 +1624,6 @@ xfs_trans_alloc_log_vecs( else lv->lv_next = new_lv; lv = new_lv; - lidp = xfs_trans_next_item(tp, lidp); } return ret_lv; @@ -1708,12 +1782,6 @@ xfs_trans_cancel( int flags) { int log_flags; -#ifdef DEBUG - xfs_log_item_chunk_t *licp; - xfs_log_item_desc_t *lidp; - xfs_log_item_t *lip; - int i; -#endif xfs_mount_t *mp = tp->t_mountp; /* @@ -1732,21 +1800,11 @@ xfs_trans_cancel( xfs_force_shutdown(mp, SHUTDOWN_CORRUPT_INCORE); } #ifdef DEBUG - if (!(flags & XFS_TRANS_ABORT)) { - licp = &(tp->t_items); - while (licp != NULL) { - lidp = licp->lic_descs; - for (i = 0; i < licp->lic_unused; i++, lidp++) { - if (xfs_lic_isfree(licp, i)) { - continue; - } - - lip = lidp->lid_item; - if (!XFS_FORCED_SHUTDOWN(mp)) - ASSERT(!(lip->li_type == XFS_LI_EFD)); - } - licp = licp->lic_next; - } + if (!(flags & XFS_TRANS_ABORT) && !XFS_FORCED_SHUTDOWN(mp)) { + struct xfs_log_item_desc *lidp; + + list_for_each_entry(lidp, &tp->t_items, lid_trans) + ASSERT(!(lidp->lid_item->li_type == XFS_LI_EFD)); } #endif xfs_trans_unreserve_and_mod_sb(tp); @@ -1834,7 +1892,6 @@ xfs_trans_roll( if (error) return error; - xfs_trans_ijoin(trans, dp, XFS_ILOCK_EXCL); - xfs_trans_ihold(trans, dp); + xfs_trans_ijoin(trans, dp); return 0; } diff --git a/fs/xfs/xfs_trans.h b/fs/xfs/xfs_trans.h index e639e8e9a2a..c13c0f97b49 100644 --- a/fs/xfs/xfs_trans.h +++ b/fs/xfs/xfs_trans.h @@ -161,105 +161,14 @@ typedef struct xfs_trans_header { * the amount of space needed to log the item it describes * once we get to commit processing (see xfs_trans_commit()). */ -typedef struct xfs_log_item_desc { +struct xfs_log_item_desc { struct xfs_log_item *lid_item; - ushort lid_size; - unsigned char lid_flags; - unsigned char lid_index; -} xfs_log_item_desc_t; + ushort lid_size; + unsigned char lid_flags; + struct list_head lid_trans; +}; #define XFS_LID_DIRTY 0x1 -#define XFS_LID_PINNED 0x2 - -/* - * This structure is used to maintain a chunk list of log_item_desc - * structures. The free field is a bitmask indicating which descriptors - * in this chunk's array are free. The unused field is the first value - * not used since this chunk was allocated. - */ -#define XFS_LIC_NUM_SLOTS 15 -typedef struct xfs_log_item_chunk { - struct xfs_log_item_chunk *lic_next; - ushort lic_free; - ushort lic_unused; - xfs_log_item_desc_t lic_descs[XFS_LIC_NUM_SLOTS]; -} xfs_log_item_chunk_t; - -#define XFS_LIC_MAX_SLOT (XFS_LIC_NUM_SLOTS - 1) -#define XFS_LIC_FREEMASK ((1 << XFS_LIC_NUM_SLOTS) - 1) - - -/* - * Initialize the given chunk. Set the chunk's free descriptor mask - * to indicate that all descriptors are free. The caller gets to set - * lic_unused to the right value (0 matches all free). The - * lic_descs.lid_index values are set up as each desc is allocated. - */ -static inline void xfs_lic_init(xfs_log_item_chunk_t *cp) -{ - cp->lic_free = XFS_LIC_FREEMASK; -} - -static inline void xfs_lic_init_slot(xfs_log_item_chunk_t *cp, int slot) -{ - cp->lic_descs[slot].lid_index = (unsigned char)(slot); -} - -static inline int xfs_lic_vacancy(xfs_log_item_chunk_t *cp) -{ - return cp->lic_free & XFS_LIC_FREEMASK; -} - -static inline void xfs_lic_all_free(xfs_log_item_chunk_t *cp) -{ - cp->lic_free = XFS_LIC_FREEMASK; -} - -static inline int xfs_lic_are_all_free(xfs_log_item_chunk_t *cp) -{ - return ((cp->lic_free & XFS_LIC_FREEMASK) == XFS_LIC_FREEMASK); -} - -static inline int xfs_lic_isfree(xfs_log_item_chunk_t *cp, int slot) -{ - return (cp->lic_free & (1 << slot)); -} - -static inline void xfs_lic_claim(xfs_log_item_chunk_t *cp, int slot) -{ - cp->lic_free &= ~(1 << slot); -} - -static inline void xfs_lic_relse(xfs_log_item_chunk_t *cp, int slot) -{ - cp->lic_free |= 1 << slot; -} - -static inline xfs_log_item_desc_t * -xfs_lic_slot(xfs_log_item_chunk_t *cp, int slot) -{ - return &(cp->lic_descs[slot]); -} - -static inline int xfs_lic_desc_to_slot(xfs_log_item_desc_t *dp) -{ - return (uint)dp->lid_index; -} - -/* - * Calculate the address of a chunk given a descriptor pointer: - * dp - dp->lid_index give the address of the start of the lic_descs array. - * From this we subtract the offset of the lic_descs field in a chunk. - * All of this yields the address of the chunk, which is - * cast to a chunk pointer. - */ -static inline xfs_log_item_chunk_t * -xfs_lic_desc_to_chunk(xfs_log_item_desc_t *dp) -{ - return (xfs_log_item_chunk_t*) \ - (((xfs_caddr_t)((dp) - (dp)->lid_index)) - \ - (xfs_caddr_t)(((xfs_log_item_chunk_t*)0)->lic_descs)); -} #define XFS_TRANS_MAGIC 0x5452414E /* 'TRAN' */ /* @@ -275,8 +184,6 @@ xfs_lic_desc_to_chunk(xfs_log_item_desc_t *dp) /* * Values for call flags parameter. */ -#define XFS_TRANS_NOSLEEP 0x1 -#define XFS_TRANS_WAIT 0x2 #define XFS_TRANS_RELEASE_LOG_RES 0x4 #define XFS_TRANS_ABORT 0x8 @@ -438,8 +345,7 @@ 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 *); - void (*iop_unpin_remove)(xfs_log_item_t *, struct xfs_trans *); + void (*iop_unpin)(xfs_log_item_t *, int remove); 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); @@ -451,8 +357,7 @@ typedef struct xfs_item_ops { #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) (*(ip)->li_ops->iop_unpin)(ip) -#define IOP_UNPIN_REMOVE(ip,tp) (*(ip)->li_ops->iop_unpin_remove)(ip, tp) +#define IOP_UNPIN(ip, remove) (*(ip)->li_ops->iop_unpin)(ip, remove) #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) @@ -516,8 +421,7 @@ typedef struct xfs_trans { 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 */ + struct list_head t_items; /* log item descriptors */ xfs_trans_header_t t_header; /* header for in-log trans */ struct list_head t_busy; /* list of busy extents */ unsigned long t_pflags; /* saved process flags state */ @@ -569,8 +473,8 @@ void xfs_trans_dquot_buf(xfs_trans_t *, struct xfs_buf *, uint); void xfs_trans_inode_alloc_buf(xfs_trans_t *, struct xfs_buf *); int xfs_trans_iget(struct xfs_mount *, xfs_trans_t *, xfs_ino_t , uint, uint, struct xfs_inode **); -void xfs_trans_ijoin(xfs_trans_t *, struct xfs_inode *, uint); -void xfs_trans_ihold(xfs_trans_t *, struct xfs_inode *); +void xfs_trans_ijoin_ref(struct xfs_trans *, struct xfs_inode *, uint); +void xfs_trans_ijoin(struct xfs_trans *, struct xfs_inode *); void xfs_trans_log_buf(xfs_trans_t *, struct xfs_buf *, uint, uint); void xfs_trans_log_inode(xfs_trans_t *, struct xfs_inode *, uint); struct xfs_efi_log_item *xfs_trans_get_efi(xfs_trans_t *, uint); @@ -595,6 +499,7 @@ int xfs_trans_ail_init(struct xfs_mount *); void xfs_trans_ail_destroy(struct xfs_mount *); extern kmem_zone_t *xfs_trans_zone; +extern kmem_zone_t *xfs_log_item_desc_zone; #endif /* __KERNEL__ */ diff --git a/fs/xfs/xfs_trans_ail.c b/fs/xfs/xfs_trans_ail.c index e799824f724..dc9069568ff 100644 --- a/fs/xfs/xfs_trans_ail.c +++ b/fs/xfs/xfs_trans_ail.c @@ -24,7 +24,6 @@ #include "xfs_trans.h" #include "xfs_sb.h" #include "xfs_ag.h" -#include "xfs_dmapi.h" #include "xfs_mount.h" #include "xfs_trans_priv.h" #include "xfs_error.h" diff --git a/fs/xfs/xfs_trans_buf.c b/fs/xfs/xfs_trans_buf.c index 63d81a22f4f..90af025e683 100644 --- a/fs/xfs/xfs_trans_buf.c +++ b/fs/xfs/xfs_trans_buf.c @@ -24,14 +24,10 @@ #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_dir2_sf.h" -#include "xfs_attr_sf.h" #include "xfs_dinode.h" #include "xfs_inode.h" #include "xfs_buf_item.h" @@ -51,36 +47,17 @@ xfs_trans_buf_item_match( xfs_daddr_t blkno, int len) { - xfs_log_item_chunk_t *licp; - xfs_log_item_desc_t *lidp; - xfs_buf_log_item_t *blip; - int i; + struct xfs_log_item_desc *lidp; + struct xfs_buf_log_item *blip; len = BBTOB(len); - for (licp = &tp->t_items; licp != NULL; licp = licp->lic_next) { - if (xfs_lic_are_all_free(licp)) { - ASSERT(licp == &tp->t_items); - ASSERT(licp->lic_next == NULL); - return NULL; - } - - for (i = 0; i < licp->lic_unused; i++) { - /* - * Skip unoccupied slots. - */ - if (xfs_lic_isfree(licp, i)) - continue; - - lidp = xfs_lic_slot(licp, i); - blip = (xfs_buf_log_item_t *)lidp->lid_item; - if (blip->bli_item.li_type != XFS_LI_BUF) - continue; - - if (XFS_BUF_TARGET(blip->bli_buf) == target && - XFS_BUF_ADDR(blip->bli_buf) == blkno && - XFS_BUF_COUNT(blip->bli_buf) == len) - return blip->bli_buf; - } + list_for_each_entry(lidp, &tp->t_items, lid_trans) { + blip = (struct xfs_buf_log_item *)lidp->lid_item; + if (blip->bli_item.li_type == XFS_LI_BUF && + XFS_BUF_TARGET(blip->bli_buf) == target && + XFS_BUF_ADDR(blip->bli_buf) == blkno && + XFS_BUF_COUNT(blip->bli_buf) == len) + return blip->bli_buf; } return NULL; @@ -127,7 +104,7 @@ _xfs_trans_bjoin( /* * Get a log_item_desc to point at the new item. */ - (void) xfs_trans_add_item(tp, (xfs_log_item_t *)bip); + xfs_trans_add_item(tp, &bip->bli_item); /* * Initialize b_fsprivate2 so we can find it with incore_match() @@ -483,7 +460,6 @@ xfs_trans_brelse(xfs_trans_t *tp, { xfs_buf_log_item_t *bip; xfs_log_item_t *lip; - xfs_log_item_desc_t *lidp; /* * Default to a normal brelse() call if the tp is NULL. @@ -514,13 +490,6 @@ xfs_trans_brelse(xfs_trans_t *tp, ASSERT(!(bip->bli_format.blf_flags & XFS_BLF_CANCEL)); ASSERT(atomic_read(&bip->bli_refcount) > 0); - /* - * Find the item descriptor pointing to this buffer's - * log item. It must be there. - */ - lidp = xfs_trans_find_item(tp, (xfs_log_item_t*)bip); - ASSERT(lidp != NULL); - trace_xfs_trans_brelse(bip); /* @@ -536,7 +505,7 @@ xfs_trans_brelse(xfs_trans_t *tp, * If the buffer is dirty within this transaction, we can't * release it until we commit. */ - if (lidp->lid_flags & XFS_LID_DIRTY) + if (bip->bli_item.li_desc->lid_flags & XFS_LID_DIRTY) return; /* @@ -553,7 +522,7 @@ xfs_trans_brelse(xfs_trans_t *tp, /* * Free up the log item descriptor tracking the released item. */ - xfs_trans_free_item(tp, lidp); + xfs_trans_del_item(&bip->bli_item); /* * Clear the hold flag in the buf log item if it is set. @@ -665,7 +634,6 @@ xfs_trans_log_buf(xfs_trans_t *tp, uint last) { xfs_buf_log_item_t *bip; - xfs_log_item_desc_t *lidp; ASSERT(XFS_BUF_ISBUSY(bp)); ASSERT(XFS_BUF_FSPRIVATE2(bp, xfs_trans_t *) == tp); @@ -690,7 +658,7 @@ xfs_trans_log_buf(xfs_trans_t *tp, bip = XFS_BUF_FSPRIVATE(bp, xfs_buf_log_item_t *); ASSERT(atomic_read(&bip->bli_refcount) > 0); XFS_BUF_SET_IODONE_FUNC(bp, xfs_buf_iodone_callbacks); - bip->bli_item.li_cb = (void(*)(xfs_buf_t*,xfs_log_item_t*))xfs_buf_iodone; + bip->bli_item.li_cb = xfs_buf_iodone; trace_xfs_trans_log_buf(bip); @@ -707,11 +675,8 @@ xfs_trans_log_buf(xfs_trans_t *tp, bip->bli_format.blf_flags &= ~XFS_BLF_CANCEL; } - lidp = xfs_trans_find_item(tp, (xfs_log_item_t*)bip); - ASSERT(lidp != NULL); - tp->t_flags |= XFS_TRANS_DIRTY; - lidp->lid_flags |= XFS_LID_DIRTY; + bip->bli_item.li_desc->lid_flags |= XFS_LID_DIRTY; bip->bli_flags |= XFS_BLI_LOGGED; xfs_buf_item_log(bip, first, last); } @@ -740,7 +705,6 @@ xfs_trans_binval( xfs_trans_t *tp, xfs_buf_t *bp) { - xfs_log_item_desc_t *lidp; xfs_buf_log_item_t *bip; ASSERT(XFS_BUF_ISBUSY(bp)); @@ -748,8 +712,6 @@ xfs_trans_binval( ASSERT(XFS_BUF_FSPRIVATE(bp, void *) != NULL); bip = XFS_BUF_FSPRIVATE(bp, xfs_buf_log_item_t *); - lidp = xfs_trans_find_item(tp, (xfs_log_item_t*)bip); - ASSERT(lidp != NULL); ASSERT(atomic_read(&bip->bli_refcount) > 0); trace_xfs_trans_binval(bip); @@ -764,7 +726,7 @@ xfs_trans_binval( ASSERT(!(bip->bli_flags & (XFS_BLI_LOGGED | XFS_BLI_DIRTY))); ASSERT(!(bip->bli_format.blf_flags & XFS_BLF_INODE_BUF)); ASSERT(bip->bli_format.blf_flags & XFS_BLF_CANCEL); - ASSERT(lidp->lid_flags & XFS_LID_DIRTY); + ASSERT(bip->bli_item.li_desc->lid_flags & XFS_LID_DIRTY); ASSERT(tp->t_flags & XFS_TRANS_DIRTY); return; } @@ -797,7 +759,7 @@ xfs_trans_binval( bip->bli_format.blf_flags |= XFS_BLF_CANCEL; memset((char *)(bip->bli_format.blf_data_map), 0, (bip->bli_format.blf_map_size * sizeof(uint))); - lidp->lid_flags |= XFS_LID_DIRTY; + bip->bli_item.li_desc->lid_flags |= XFS_LID_DIRTY; tp->t_flags |= XFS_TRANS_DIRTY; } @@ -853,12 +815,9 @@ xfs_trans_stale_inode_buf( ASSERT(atomic_read(&bip->bli_refcount) > 0); bip->bli_flags |= XFS_BLI_STALE_INODE; - bip->bli_item.li_cb = (void(*)(xfs_buf_t*,xfs_log_item_t*)) - xfs_buf_iodone; + bip->bli_item.li_cb = xfs_buf_iodone; } - - /* * Mark the buffer as being one which contains newly allocated * inodes. We need to make sure that even if this buffer is diff --git a/fs/xfs/xfs_trans_extfree.c b/fs/xfs/xfs_trans_extfree.c index 27cce2a9c7e..f783d5e9fa7 100644 --- a/fs/xfs/xfs_trans_extfree.c +++ b/fs/xfs/xfs_trans_extfree.c @@ -23,7 +23,6 @@ #include "xfs_trans.h" #include "xfs_sb.h" #include "xfs_ag.h" -#include "xfs_dmapi.h" #include "xfs_mount.h" #include "xfs_trans_priv.h" #include "xfs_extfree_item.h" @@ -49,9 +48,8 @@ xfs_trans_get_efi(xfs_trans_t *tp, /* * Get a log_item_desc to point at the new item. */ - (void) xfs_trans_add_item(tp, (xfs_log_item_t*)efip); - - return (efip); + xfs_trans_add_item(tp, &efip->efi_item); + return efip; } /* @@ -65,15 +63,11 @@ xfs_trans_log_efi_extent(xfs_trans_t *tp, xfs_fsblock_t start_block, xfs_extlen_t ext_len) { - xfs_log_item_desc_t *lidp; uint next_extent; xfs_extent_t *extp; - lidp = xfs_trans_find_item(tp, (xfs_log_item_t*)efip); - ASSERT(lidp != NULL); - tp->t_flags |= XFS_TRANS_DIRTY; - lidp->lid_flags |= XFS_LID_DIRTY; + efip->efi_item.li_desc->lid_flags |= XFS_LID_DIRTY; next_extent = efip->efi_next_extent; ASSERT(next_extent < efip->efi_format.efi_nextents); @@ -106,9 +100,8 @@ xfs_trans_get_efd(xfs_trans_t *tp, /* * Get a log_item_desc to point at the new item. */ - (void) xfs_trans_add_item(tp, (xfs_log_item_t*)efdp); - - return (efdp); + xfs_trans_add_item(tp, &efdp->efd_item); + return efdp; } /* @@ -122,15 +115,11 @@ xfs_trans_log_efd_extent(xfs_trans_t *tp, xfs_fsblock_t start_block, xfs_extlen_t ext_len) { - xfs_log_item_desc_t *lidp; uint next_extent; xfs_extent_t *extp; - lidp = xfs_trans_find_item(tp, (xfs_log_item_t*)efdp); - ASSERT(lidp != NULL); - tp->t_flags |= XFS_TRANS_DIRTY; - lidp->lid_flags |= XFS_LID_DIRTY; + efdp->efd_item.li_desc->lid_flags |= XFS_LID_DIRTY; next_extent = efdp->efd_next_extent; ASSERT(next_extent < efdp->efd_format.efd_nextents); diff --git a/fs/xfs/xfs_trans_inode.c b/fs/xfs/xfs_trans_inode.c index 2559dfec946..cdc53a1050c 100644 --- a/fs/xfs/xfs_trans_inode.c +++ b/fs/xfs/xfs_trans_inode.c @@ -24,20 +24,16 @@ #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_dir2_sf.h" -#include "xfs_attr_sf.h" #include "xfs_dinode.h" #include "xfs_inode.h" #include "xfs_btree.h" -#include "xfs_ialloc.h" #include "xfs_trans_priv.h" #include "xfs_inode_item.h" +#include "xfs_trace.h" #ifdef XFS_TRANS_DEBUG STATIC void @@ -47,7 +43,6 @@ xfs_trans_inode_broot_debug( #define xfs_trans_inode_broot_debug(ip) #endif - /* * Get an inode and join it to the transaction. */ @@ -63,77 +58,65 @@ xfs_trans_iget( int error; error = xfs_iget(mp, tp, ino, flags, lock_flags, ipp); - if (!error && tp) - xfs_trans_ijoin(tp, *ipp, lock_flags); + if (!error && tp) { + xfs_trans_ijoin(tp, *ipp); + (*ipp)->i_itemp->ili_lock_flags = lock_flags; + } return error; } /* - * Add the locked inode to the transaction. - * The inode must be locked, and it cannot be associated with any - * transaction. The caller must specify the locks already held - * on the inode. + * Add a locked inode to the transaction. + * + * The inode must be locked, and it cannot be associated with any transaction. */ void xfs_trans_ijoin( - xfs_trans_t *tp, - xfs_inode_t *ip, - uint lock_flags) + struct xfs_trans *tp, + struct xfs_inode *ip) { xfs_inode_log_item_t *iip; ASSERT(ip->i_transp == NULL); ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL)); - ASSERT(lock_flags & XFS_ILOCK_EXCL); if (ip->i_itemp == NULL) xfs_inode_item_init(ip, ip->i_mount); iip = ip->i_itemp; - ASSERT(iip->ili_flags == 0); + ASSERT(iip->ili_lock_flags == 0); /* * Get a log_item_desc to point at the new item. */ - (void) xfs_trans_add_item(tp, (xfs_log_item_t*)(iip)); + xfs_trans_add_item(tp, &iip->ili_item); xfs_trans_inode_broot_debug(ip); /* - * If the IO lock is already held, mark that in the inode log item. - */ - 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() * in xfs_trans_iget() above. */ ip->i_transp = tp; } - - /* - * Mark the inode as not needing to be unlocked when the inode item's - * IOP_UNLOCK() routine is called. The inode must already be locked - * and associated with the given transaction. + * Add a locked inode to the transaction. + * + * + * Grabs a reference to the inode which will be dropped when the transaction + * is commited. The inode will also be unlocked at that point. The inode + * must be locked, and it cannot be associated with any transaction. */ -/*ARGSUSED*/ void -xfs_trans_ihold( - xfs_trans_t *tp, - xfs_inode_t *ip) +xfs_trans_ijoin_ref( + struct xfs_trans *tp, + struct xfs_inode *ip, + uint lock_flags) { - ASSERT(ip->i_transp == tp); - ASSERT(ip->i_itemp != NULL); - ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL)); - - ip->i_itemp->ili_flags |= XFS_ILI_HOLD; + xfs_trans_ijoin(tp, ip); + IHOLD(ip); + ip->i_itemp->ili_lock_flags = lock_flags; } - /* * This is called to mark the fields indicated in fieldmask as needing * to be logged when the transaction is committed. The inode must @@ -149,17 +132,12 @@ xfs_trans_log_inode( xfs_inode_t *ip, uint flags) { - xfs_log_item_desc_t *lidp; - ASSERT(ip->i_transp == tp); ASSERT(ip->i_itemp != NULL); ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL)); - lidp = xfs_trans_find_item(tp, (xfs_log_item_t*)(ip->i_itemp)); - ASSERT(lidp != NULL); - tp->t_flags |= XFS_TRANS_DIRTY; - lidp->lid_flags |= XFS_LID_DIRTY; + ip->i_itemp->ili_item.li_desc->lid_flags |= XFS_LID_DIRTY; /* * Always OR in the bits from the ili_last_fields field. diff --git a/fs/xfs/xfs_trans_item.c b/fs/xfs/xfs_trans_item.c deleted file mode 100644 index f11d37d06dc..00000000000 --- a/fs/xfs/xfs_trans_item.c +++ /dev/null @@ -1,441 +0,0 @@ -/* - * Copyright (c) 2000-2002,2005 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_log.h" -#include "xfs_inum.h" -#include "xfs_trans.h" -#include "xfs_trans_priv.h" -/* XXX: from here down needed until struct xfs_trans has its 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); - -/* - * This is called to add the given log item to the transaction's - * list of log items. It must find a free log item descriptor - * or allocate a new one and add the item to that descriptor. - * The function returns a pointer to item descriptor used to point - * to the new item. The log item will now point to its new descriptor - * with its li_desc field. - */ -xfs_log_item_desc_t * -xfs_trans_add_item(xfs_trans_t *tp, xfs_log_item_t *lip) -{ - xfs_log_item_desc_t *lidp; - xfs_log_item_chunk_t *licp; - int i=0; - - /* - * If there are no free descriptors, allocate a new chunk - * of them and put it at the front of the chunk list. - */ - if (tp->t_items_free == 0) { - licp = (xfs_log_item_chunk_t*) - kmem_alloc(sizeof(xfs_log_item_chunk_t), KM_SLEEP); - ASSERT(licp != NULL); - /* - * Initialize the chunk, and then - * claim the first slot in the newly allocated chunk. - */ - xfs_lic_init(licp); - xfs_lic_claim(licp, 0); - licp->lic_unused = 1; - xfs_lic_init_slot(licp, 0); - lidp = xfs_lic_slot(licp, 0); - - /* - * Link in the new chunk and update the free count. - */ - licp->lic_next = tp->t_items.lic_next; - tp->t_items.lic_next = licp; - tp->t_items_free = XFS_LIC_NUM_SLOTS - 1; - - /* - * Initialize the descriptor and the generic portion - * of the log item. - * - * Point the new slot at this item and return it. - * Also point the log item at its currently active - * descriptor and set the item's mount pointer. - */ - lidp->lid_item = lip; - lidp->lid_flags = 0; - lidp->lid_size = 0; - lip->li_desc = lidp; - lip->li_mountp = tp->t_mountp; - lip->li_ailp = tp->t_mountp->m_ail; - return lidp; - } - - /* - * Find the free descriptor. It is somewhere in the chunklist - * of descriptors. - */ - licp = &tp->t_items; - while (licp != NULL) { - if (xfs_lic_vacancy(licp)) { - if (licp->lic_unused <= XFS_LIC_MAX_SLOT) { - i = licp->lic_unused; - ASSERT(xfs_lic_isfree(licp, i)); - break; - } - for (i = 0; i <= XFS_LIC_MAX_SLOT; i++) { - if (xfs_lic_isfree(licp, i)) - break; - } - ASSERT(i <= XFS_LIC_MAX_SLOT); - break; - } - licp = licp->lic_next; - } - ASSERT(licp != NULL); - /* - * If we find a free descriptor, claim it, - * initialize it, and return it. - */ - xfs_lic_claim(licp, i); - if (licp->lic_unused <= i) { - licp->lic_unused = i + 1; - xfs_lic_init_slot(licp, i); - } - lidp = xfs_lic_slot(licp, i); - tp->t_items_free--; - lidp->lid_item = lip; - lidp->lid_flags = 0; - lidp->lid_size = 0; - lip->li_desc = lidp; - lip->li_mountp = tp->t_mountp; - lip->li_ailp = tp->t_mountp->m_ail; - return lidp; -} - -/* - * Free the given descriptor. - * - * This requires setting the bit in the chunk's free mask corresponding - * to the given slot. - */ -void -xfs_trans_free_item(xfs_trans_t *tp, xfs_log_item_desc_t *lidp) -{ - uint slot; - xfs_log_item_chunk_t *licp; - xfs_log_item_chunk_t **licpp; - - slot = xfs_lic_desc_to_slot(lidp); - licp = xfs_lic_desc_to_chunk(lidp); - xfs_lic_relse(licp, slot); - lidp->lid_item->li_desc = NULL; - tp->t_items_free++; - - /* - * If there are no more used items in the chunk and this is not - * the chunk embedded in the transaction structure, then free - * the chunk. First pull it from the chunk list and then - * free it back to the heap. We didn't bother with a doubly - * linked list here because the lists should be very short - * and this is not a performance path. It's better to save - * the memory of the extra pointer. - * - * Also decrement the transaction structure's count of free items - * by the number in a chunk since we are freeing an empty chunk. - */ - if (xfs_lic_are_all_free(licp) && (licp != &(tp->t_items))) { - licpp = &(tp->t_items.lic_next); - while (*licpp != licp) { - ASSERT(*licpp != NULL); - licpp = &((*licpp)->lic_next); - } - *licpp = licp->lic_next; - kmem_free(licp); - tp->t_items_free -= XFS_LIC_NUM_SLOTS; - } -} - -/* - * This is called to find the descriptor corresponding to the given - * log item. It returns a pointer to the descriptor. - * The log item MUST have a corresponding descriptor in the given - * transaction. This routine does not return NULL, it panics. - * - * The descriptor pointer is kept in the log item's li_desc field. - * Just return it. - */ -/*ARGSUSED*/ -xfs_log_item_desc_t * -xfs_trans_find_item(xfs_trans_t *tp, xfs_log_item_t *lip) -{ - ASSERT(lip->li_desc != NULL); - - return lip->li_desc; -} - - -/* - * Return a pointer to the first descriptor in the chunk list. - * This does not return NULL if there are none, it panics. - * - * The first descriptor must be in either the first or second chunk. - * This is because the only chunk allowed to be empty is the first. - * All others are freed when they become empty. - * - * At some point this and xfs_trans_next_item() should be optimized - * to quickly look at the mask to determine if there is anything to - * look at. - */ -xfs_log_item_desc_t * -xfs_trans_first_item(xfs_trans_t *tp) -{ - xfs_log_item_chunk_t *licp; - int i; - - licp = &tp->t_items; - /* - * If it's not in the first chunk, skip to the second. - */ - if (xfs_lic_are_all_free(licp)) { - licp = licp->lic_next; - } - - /* - * Return the first non-free descriptor in the chunk. - */ - ASSERT(!xfs_lic_are_all_free(licp)); - for (i = 0; i < licp->lic_unused; i++) { - if (xfs_lic_isfree(licp, i)) { - continue; - } - - return xfs_lic_slot(licp, i); - } - cmn_err(CE_WARN, "xfs_trans_first_item() -- no first item"); - return NULL; -} - - -/* - * Given a descriptor, return the next descriptor in the chunk list. - * This returns NULL if there are no more used descriptors in the list. - * - * We do this by first locating the chunk in which the descriptor resides, - * and then scanning forward in the chunk and the list for the next - * used descriptor. - */ -/*ARGSUSED*/ -xfs_log_item_desc_t * -xfs_trans_next_item(xfs_trans_t *tp, xfs_log_item_desc_t *lidp) -{ - xfs_log_item_chunk_t *licp; - int i; - - licp = xfs_lic_desc_to_chunk(lidp); - - /* - * First search the rest of the chunk. The for loop keeps us - * from referencing things beyond the end of the chunk. - */ - for (i = (int)xfs_lic_desc_to_slot(lidp) + 1; i < licp->lic_unused; i++) { - if (xfs_lic_isfree(licp, i)) { - continue; - } - - return xfs_lic_slot(licp, i); - } - - /* - * Now search the next chunk. It must be there, because the - * next chunk would have been freed if it were empty. - * If there is no next chunk, return NULL. - */ - if (licp->lic_next == NULL) { - return NULL; - } - - licp = licp->lic_next; - ASSERT(!xfs_lic_are_all_free(licp)); - for (i = 0; i < licp->lic_unused; i++) { - if (xfs_lic_isfree(licp, i)) { - continue; - } - - return xfs_lic_slot(licp, i); - } - ASSERT(0); - /* NOTREACHED */ - return NULL; /* keep gcc quite */ -} - -/* - * This is called to unlock all of the items of a transaction and to free - * all the descriptors of that transaction. - * - * It walks the list of descriptors and unlocks each item. It frees - * each chunk except that embedded in the transaction as it goes along. - */ -void -xfs_trans_free_items( - xfs_trans_t *tp, - xfs_lsn_t commit_lsn, - int flags) -{ - xfs_log_item_chunk_t *licp; - xfs_log_item_chunk_t *next_licp; - int abort; - - abort = flags & XFS_TRANS_ABORT; - licp = &tp->t_items; - /* - * Special case the embedded chunk so we don't free it below. - */ - if (!xfs_lic_are_all_free(licp)) { - (void) xfs_trans_unlock_chunk(licp, 1, abort, commit_lsn); - xfs_lic_all_free(licp); - licp->lic_unused = 0; - } - licp = licp->lic_next; - - /* - * Unlock each item in each chunk and free the chunks. - */ - while (licp != NULL) { - ASSERT(!xfs_lic_are_all_free(licp)); - (void) xfs_trans_unlock_chunk(licp, 1, abort, commit_lsn); - next_licp = licp->lic_next; - kmem_free(licp); - licp = next_licp; - } - - /* - * Reset the transaction structure's free item count. - */ - tp->t_items_free = XFS_LIC_NUM_SLOTS; - tp->t_items.lic_next = NULL; -} - - - -/* - * This is called to unlock the items associated with a transaction. - * Items which were not logged should be freed. - * Those which were logged must still be tracked so they can be unpinned - * when the transaction commits. - */ -void -xfs_trans_unlock_items(xfs_trans_t *tp, xfs_lsn_t commit_lsn) -{ - xfs_log_item_chunk_t *licp; - xfs_log_item_chunk_t *next_licp; - xfs_log_item_chunk_t **licpp; - int freed; - - freed = 0; - licp = &tp->t_items; - - /* - * Special case the embedded chunk so we don't free. - */ - if (!xfs_lic_are_all_free(licp)) { - freed = xfs_trans_unlock_chunk(licp, 0, 0, commit_lsn); - } - licpp = &(tp->t_items.lic_next); - licp = licp->lic_next; - - /* - * Unlock each item in each chunk, free non-dirty descriptors, - * and free empty chunks. - */ - while (licp != NULL) { - ASSERT(!xfs_lic_are_all_free(licp)); - freed += xfs_trans_unlock_chunk(licp, 0, 0, commit_lsn); - next_licp = licp->lic_next; - if (xfs_lic_are_all_free(licp)) { - *licpp = next_licp; - kmem_free(licp); - freed -= XFS_LIC_NUM_SLOTS; - } else { - licpp = &(licp->lic_next); - } - ASSERT(*licpp == next_licp); - licp = next_licp; - } - - /* - * Fix the free descriptor count in the transaction. - */ - tp->t_items_free += freed; -} - -/* - * Unlock each item pointed to by a descriptor in the given chunk. - * Stamp the commit lsn into each item if necessary. - * Free descriptors pointing to items which are not dirty if freeing_chunk - * is zero. If freeing_chunk is non-zero, then we need to unlock all - * items in the chunk. - * - * Return the number of descriptors freed. - */ -STATIC int -xfs_trans_unlock_chunk( - xfs_log_item_chunk_t *licp, - int freeing_chunk, - int abort, - xfs_lsn_t commit_lsn) -{ - xfs_log_item_desc_t *lidp; - xfs_log_item_t *lip; - int i; - int freed; - - freed = 0; - lidp = licp->lic_descs; - for (i = 0; i < licp->lic_unused; i++, lidp++) { - if (xfs_lic_isfree(licp, i)) { - continue; - } - lip = lidp->lid_item; - lip->li_desc = NULL; - - if (commit_lsn != NULLCOMMITLSN) - IOP_COMMITTING(lip, commit_lsn); - if (abort) - lip->li_flags |= XFS_LI_ABORTED; - IOP_UNLOCK(lip); - - /* - * Free the descriptor if the item is not dirty - * within this transaction and the caller is not - * going to just free the entire thing regardless. - */ - if (!(freeing_chunk) && - (!(lidp->lid_flags & XFS_LID_DIRTY) || abort)) { - xfs_lic_relse(licp, i); - freed++; - } - } - - return freed; -} diff --git a/fs/xfs/xfs_trans_priv.h b/fs/xfs/xfs_trans_priv.h index c6e4f2c8de6..e2d93d8ead7 100644 --- a/fs/xfs/xfs_trans_priv.h +++ b/fs/xfs/xfs_trans_priv.h @@ -23,22 +23,8 @@ struct xfs_log_item_desc; struct xfs_mount; struct xfs_trans; -/* - * From xfs_trans_item.c - */ -struct xfs_log_item_desc *xfs_trans_add_item(struct xfs_trans *, - struct xfs_log_item *); -void xfs_trans_free_item(struct xfs_trans *, - struct xfs_log_item_desc *); -struct xfs_log_item_desc *xfs_trans_find_item(struct xfs_trans *, - struct xfs_log_item *); -struct xfs_log_item_desc *xfs_trans_first_item(struct xfs_trans *); -struct xfs_log_item_desc *xfs_trans_next_item(struct xfs_trans *, - struct xfs_log_item_desc *); - -void xfs_trans_unlock_items(struct xfs_trans *tp, xfs_lsn_t commit_lsn); -void xfs_trans_free_items(struct xfs_trans *tp, xfs_lsn_t commit_lsn, - int flags); +void xfs_trans_add_item(struct xfs_trans *, struct xfs_log_item *); +void xfs_trans_del_item(struct xfs_log_item *); void xfs_trans_item_committed(struct xfs_log_item *lip, xfs_lsn_t commit_lsn, int aborted); diff --git a/fs/xfs/xfs_utils.c b/fs/xfs/xfs_utils.c index 4d88616bde9..b7d5769d2df 100644 --- a/fs/xfs/xfs_utils.c +++ b/fs/xfs/xfs_utils.c @@ -25,18 +25,14 @@ #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_dir2_sf.h" -#include "xfs_attr_sf.h" #include "xfs_dinode.h" #include "xfs_inode.h" #include "xfs_inode_item.h" #include "xfs_bmap.h" #include "xfs_error.h" #include "xfs_quota.h" -#include "xfs_rw.h" #include "xfs_itable.h" #include "xfs_utils.h" @@ -324,86 +320,3 @@ xfs_bumplink( xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); return 0; } - -/* - * Try to truncate the given file to 0 length. Currently called - * only out of xfs_remove when it has to truncate a file to free - * up space for the remove to proceed. - */ -int -xfs_truncate_file( - xfs_mount_t *mp, - xfs_inode_t *ip) -{ - xfs_trans_t *tp; - int error; - -#ifdef QUOTADEBUG - /* - * This is called to truncate the quotainodes too. - */ - if (XFS_IS_UQUOTA_ON(mp)) { - if (ip->i_ino != mp->m_sb.sb_uquotino) - ASSERT(ip->i_udquot); - } - if (XFS_IS_OQUOTA_ON(mp)) { - if (ip->i_ino != mp->m_sb.sb_gquotino) - ASSERT(ip->i_gdquot); - } -#endif - /* - * Make the call to xfs_itruncate_start before starting the - * transaction, because we cannot make the call while we're - * in a transaction. - */ - xfs_ilock(ip, XFS_IOLOCK_EXCL); - error = xfs_itruncate_start(ip, XFS_ITRUNC_DEFINITE, (xfs_fsize_t)0); - if (error) { - xfs_iunlock(ip, XFS_IOLOCK_EXCL); - return error; - } - - tp = xfs_trans_alloc(mp, XFS_TRANS_TRUNCATE_FILE); - if ((error = xfs_trans_reserve(tp, 0, XFS_ITRUNCATE_LOG_RES(mp), 0, - XFS_TRANS_PERM_LOG_RES, - XFS_ITRUNCATE_LOG_COUNT))) { - xfs_trans_cancel(tp, 0); - xfs_iunlock(ip, XFS_IOLOCK_EXCL); - return error; - } - - /* - * Follow the normal truncate locking protocol. Since we - * hold the inode in the transaction, we know that its number - * of references will stay constant. - */ - xfs_ilock(ip, XFS_ILOCK_EXCL); - xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL); - xfs_trans_ihold(tp, ip); - /* - * Signal a sync xaction. The only case where that isn't - * the case is if we're truncating an already unlinked file - * on a wsync fs. In that case, we know the blocks can't - * reappear in the file because the links to file are - * permanently toast. Currently, we're always going to - * want a sync transaction because this code is being - * called from places where nlink is guaranteed to be 1 - * but I'm leaving the tests in to protect against future - * changes -- rcc. - */ - error = xfs_itruncate_finish(&tp, ip, (xfs_fsize_t)0, - XFS_DATA_FORK, - ((ip->i_d.di_nlink != 0 || - !(mp->m_flags & XFS_MOUNT_WSYNC)) - ? 1 : 0)); - if (error) { - xfs_trans_cancel(tp, XFS_TRANS_RELEASE_LOG_RES | - XFS_TRANS_ABORT); - } else { - xfs_ichgtime(ip, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG); - error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES); - } - xfs_iunlock(ip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL); - - return error; -} diff --git a/fs/xfs/xfs_utils.h b/fs/xfs/xfs_utils.h index ef321225d26..f55b9678264 100644 --- a/fs/xfs/xfs_utils.h +++ b/fs/xfs/xfs_utils.h @@ -18,7 +18,6 @@ #ifndef __XFS_UTILS_H__ #define __XFS_UTILS_H__ -extern int xfs_truncate_file(xfs_mount_t *, xfs_inode_t *); extern int xfs_dir_ialloc(xfs_trans_t **, xfs_inode_t *, mode_t, xfs_nlink_t, xfs_dev_t, cred_t *, prid_t, int, xfs_inode_t **, int *); diff --git a/fs/xfs/xfs_vnodeops.c b/fs/xfs/xfs_vnodeops.c index c1646838898..3ac137dd531 100644 --- a/fs/xfs/xfs_vnodeops.c +++ b/fs/xfs/xfs_vnodeops.c @@ -26,19 +26,14 @@ #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_alloc_btree.h" #include "xfs_ialloc_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_itable.h" -#include "xfs_btree.h" #include "xfs_ialloc.h" #include "xfs_alloc.h" #include "xfs_bmap.h" @@ -73,7 +68,7 @@ xfs_setattr( struct xfs_dquot *udqp, *gdqp, *olddquot1, *olddquot2; int need_iolock = 1; - xfs_itrace_entry(ip); + trace_xfs_setattr(ip); if (mp->m_flags & XFS_MOUNT_RDONLY) return XFS_ERROR(EROFS); @@ -143,16 +138,6 @@ xfs_setattr( goto error_return; } } else { - if (DM_EVENT_ENABLED(ip, DM_EVENT_TRUNCATE) && - !(flags & XFS_ATTR_DMI)) { - int dmflags = AT_DELAY_FLAG(flags) | DM_SEM_FLAG_WR; - code = XFS_SEND_DATA(mp, DM_EVENT_TRUNCATE, ip, - iattr->ia_size, 0, dmflags, NULL); - if (code) { - lock_flags = 0; - goto error_return; - } - } if (need_iolock) lock_flags |= XFS_IOLOCK_EXCL; } @@ -283,8 +268,7 @@ xfs_setattr( commit_flags = XFS_TRANS_RELEASE_LOG_RES; xfs_ilock(ip, XFS_ILOCK_EXCL); - xfs_trans_ijoin(tp, ip, lock_flags); - xfs_trans_ihold(tp, ip); + xfs_trans_ijoin(tp, ip); /* * Only change the c/mtime if we are changing the size @@ -334,8 +318,7 @@ xfs_setattr( xfs_iflags_set(ip, XFS_ITRUNCATED); } } else if (tp) { - xfs_trans_ijoin(tp, ip, lock_flags); - xfs_trans_ihold(tp, ip); + xfs_trans_ijoin(tp, ip); } /* @@ -470,17 +453,10 @@ xfs_setattr( return XFS_ERROR(code); } - if (DM_EVENT_ENABLED(ip, DM_EVENT_ATTRIBUTE) && - !(flags & XFS_ATTR_DMI)) { - (void) XFS_SEND_NAMESP(mp, DM_EVENT_ATTRIBUTE, ip, DM_RIGHT_NULL, - NULL, DM_RIGHT_NULL, NULL, NULL, - 0, 0, AT_DELAY_FLAG(flags)); - } return 0; abort_return: commit_flags |= XFS_TRANS_ABORT; - /* FALLTHROUGH */ error_return: xfs_qm_dqrele(udqp); xfs_qm_dqrele(gdqp); @@ -516,7 +492,7 @@ xfs_readlink_bmap( int error = 0; error = xfs_bmapi(NULL, ip, 0, XFS_B_TO_FSB(mp, pathlen), 0, NULL, 0, - mval, &nmaps, NULL, NULL); + mval, &nmaps, NULL); if (error) goto out; @@ -557,7 +533,7 @@ xfs_readlink( int pathlen; int error = 0; - xfs_itrace_entry(ip); + trace_xfs_readlink(ip); if (XFS_FORCED_SHUTDOWN(mp)) return XFS_ERROR(EIO); @@ -613,14 +589,14 @@ xfs_free_eofblocks( */ end_fsb = XFS_B_TO_FSB(mp, ((xfs_ufsize_t)ip->i_size)); last_fsb = XFS_B_TO_FSB(mp, (xfs_ufsize_t)XFS_MAXIOFFSET(mp)); - map_len = last_fsb - end_fsb; - if (map_len <= 0) + if (last_fsb <= end_fsb) return 0; + map_len = last_fsb - end_fsb; nimaps = 1; xfs_ilock(ip, XFS_ILOCK_SHARED); error = xfs_bmapi(NULL, ip, end_fsb, map_len, 0, - NULL, 0, &imap, &nimaps, NULL, NULL); + NULL, 0, &imap, &nimaps, NULL); xfs_iunlock(ip, XFS_ILOCK_SHARED); if (!error && (nimaps != 0) && @@ -675,10 +651,7 @@ xfs_free_eofblocks( } xfs_ilock(ip, XFS_ILOCK_EXCL); - xfs_trans_ijoin(tp, ip, - XFS_IOLOCK_EXCL | - XFS_ILOCK_EXCL); - xfs_trans_ihold(tp, ip); + xfs_trans_ijoin(tp, ip); error = xfs_itruncate_finish(&tp, ip, ip->i_size, @@ -750,8 +723,7 @@ xfs_inactive_symlink_rmt( xfs_ilock(ip, XFS_IOLOCK_EXCL | XFS_ILOCK_EXCL); size = (int)ip->i_d.di_size; ip->i_d.di_size = 0; - xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL); - xfs_trans_ihold(tp, ip); + xfs_trans_ijoin(tp, ip); xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); /* * Find the block(s) so we can inval and unmap them. @@ -761,7 +733,7 @@ xfs_inactive_symlink_rmt( nmaps = ARRAY_SIZE(mval); if ((error = xfs_bmapi(tp, ip, 0, XFS_B_TO_FSB(mp, size), XFS_BMAPI_METADATA, &first_block, 0, mval, &nmaps, - &free_list, NULL))) + &free_list))) goto error0; /* * Invalidate the block(s). @@ -776,7 +748,7 @@ xfs_inactive_symlink_rmt( * Unmap the dead block(s) to the free_list. */ if ((error = xfs_bunmapi(tp, ip, 0, size, XFS_BMAPI_METADATA, nmaps, - &first_block, &free_list, NULL, &done))) + &first_block, &free_list, &done))) goto error1; ASSERT(done); /* @@ -795,8 +767,7 @@ xfs_inactive_symlink_rmt( * Mark it dirty so it will be logged and moved forward in the log as * part of every commit. */ - xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL); - xfs_trans_ihold(tp, ip); + xfs_trans_ijoin(tp, ip); xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); /* * Get a new, empty transaction to return to our caller. @@ -929,8 +900,7 @@ xfs_inactive_attrs( goto error_cancel; xfs_ilock(ip, XFS_ILOCK_EXCL); - xfs_trans_ijoin(tp, ip, XFS_IOLOCK_EXCL | XFS_ILOCK_EXCL); - xfs_trans_ihold(tp, ip); + xfs_trans_ijoin(tp, ip); xfs_idestroy_fork(ip, XFS_ATTR_FORK); ASSERT(ip->i_d.di_anextents == 0); @@ -1035,8 +1005,6 @@ xfs_inactive( int error; int truncate; - xfs_itrace_entry(ip); - /* * If the inode is already free, then there can be nothing * to clean up here. @@ -1060,9 +1028,6 @@ xfs_inactive( mp = ip->i_mount; - if (ip->i_d.di_nlink == 0 && DM_EVENT_ENABLED(ip, DM_EVENT_DESTROY)) - XFS_SEND_DESTROY(mp, ip, DM_RIGHT_NULL); - error = 0; /* If this is a read-only mount, don't do this (would generate I/O) */ @@ -1120,8 +1085,7 @@ xfs_inactive( } xfs_ilock(ip, XFS_ILOCK_EXCL); - xfs_trans_ijoin(tp, ip, XFS_IOLOCK_EXCL | XFS_ILOCK_EXCL); - xfs_trans_ihold(tp, ip); + xfs_trans_ijoin(tp, ip); /* * normally, we have to run xfs_itruncate_finish sync. @@ -1154,8 +1118,7 @@ xfs_inactive( return VN_INACTIVE_CACHE; } - xfs_trans_ijoin(tp, ip, XFS_IOLOCK_EXCL | XFS_ILOCK_EXCL); - xfs_trans_ihold(tp, ip); + xfs_trans_ijoin(tp, ip); } else { error = xfs_trans_reserve(tp, 0, XFS_IFREE_LOG_RES(mp), @@ -1168,8 +1131,7 @@ xfs_inactive( } xfs_ilock(ip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL); - xfs_trans_ijoin(tp, ip, XFS_IOLOCK_EXCL | XFS_ILOCK_EXCL); - xfs_trans_ihold(tp, ip); + xfs_trans_ijoin(tp, ip); } /* @@ -1257,7 +1219,7 @@ xfs_lookup( int error; uint lock_mode; - xfs_itrace_entry(dp); + trace_xfs_lookup(dp, name); if (XFS_FORCED_SHUTDOWN(dp->i_mount)) return XFS_ERROR(EIO); @@ -1309,21 +1271,11 @@ xfs_create( uint log_res; uint log_count; - xfs_itrace_entry(dp); + trace_xfs_create(dp, name); if (XFS_FORCED_SHUTDOWN(mp)) return XFS_ERROR(EIO); - if (DM_EVENT_ENABLED(dp, DM_EVENT_CREATE)) { - error = XFS_SEND_NAMESP(mp, DM_EVENT_CREATE, - dp, DM_RIGHT_NULL, NULL, - DM_RIGHT_NULL, name->name, NULL, - mode, 0, 0); - - if (error) - return error; - } - if (dp->i_d.di_flags & XFS_DIFLAG_PROJINHERIT) prid = dp->i_d.di_projid; else @@ -1427,8 +1379,7 @@ xfs_create( * the transaction cancel unlocking dp so don't do it explicitly in the * error path. */ - IHOLD(dp); - xfs_trans_ijoin(tp, dp, XFS_ILOCK_EXCL); + xfs_trans_ijoin_ref(tp, dp, XFS_ILOCK_EXCL); unlock_dp_on_error = B_FALSE; error = xfs_dir_createname(tp, dp, name, ip->i_ino, @@ -1487,16 +1438,7 @@ xfs_create( xfs_qm_dqrele(gdqp); *ipp = ip; - - /* Fallthrough to std_return with error = 0 */ - std_return: - if (DM_EVENT_ENABLED(dp, DM_EVENT_POSTCREATE)) { - XFS_SEND_NAMESP(mp, DM_EVENT_POSTCREATE, dp, DM_RIGHT_NULL, - ip, DM_RIGHT_NULL, name->name, NULL, mode, - error, 0); - } - - return error; + return 0; out_bmap_cancel: xfs_bmap_cancel(&free_list); @@ -1510,8 +1452,8 @@ xfs_create( if (unlock_dp_on_error) xfs_iunlock(dp, XFS_ILOCK_EXCL); - - goto std_return; + std_return: + return error; out_abort_rele: /* @@ -1726,20 +1668,11 @@ xfs_remove( uint resblks; uint log_count; - xfs_itrace_entry(dp); - xfs_itrace_entry(ip); + trace_xfs_remove(dp, name); if (XFS_FORCED_SHUTDOWN(mp)) return XFS_ERROR(EIO); - if (DM_EVENT_ENABLED(dp, DM_EVENT_REMOVE)) { - error = XFS_SEND_NAMESP(mp, DM_EVENT_REMOVE, dp, DM_RIGHT_NULL, - NULL, DM_RIGHT_NULL, name->name, NULL, - ip->i_d.di_mode, 0, 0); - if (error) - return error; - } - error = xfs_qm_dqattach(dp, 0); if (error) goto std_return; @@ -1782,15 +1715,8 @@ xfs_remove( xfs_lock_two_inodes(dp, ip, XFS_ILOCK_EXCL); - /* - * At this point, we've gotten both the directory and the entry - * inodes locked. - */ - IHOLD(ip); - xfs_trans_ijoin(tp, dp, XFS_ILOCK_EXCL); - - IHOLD(dp); - xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL); + xfs_trans_ijoin_ref(tp, dp, XFS_ILOCK_EXCL); + xfs_trans_ijoin_ref(tp, ip, XFS_ILOCK_EXCL); /* * If we're removing a directory perform some additional validation. @@ -1877,21 +1803,15 @@ xfs_remove( if (!is_dir && link_zero && xfs_inode_is_filestream(ip)) xfs_filestream_deassociate(ip); - std_return: - if (DM_EVENT_ENABLED(dp, DM_EVENT_POSTREMOVE)) { - XFS_SEND_NAMESP(mp, DM_EVENT_POSTREMOVE, dp, DM_RIGHT_NULL, - NULL, DM_RIGHT_NULL, name->name, NULL, - ip->i_d.di_mode, error, 0); - } - - return error; + return 0; out_bmap_cancel: xfs_bmap_cancel(&free_list); cancel_flags |= XFS_TRANS_ABORT; out_trans_cancel: xfs_trans_cancel(tp, cancel_flags); - goto std_return; + std_return: + return error; } int @@ -1909,25 +1829,13 @@ xfs_link( int committed; int resblks; - xfs_itrace_entry(tdp); - xfs_itrace_entry(sip); + trace_xfs_link(tdp, target_name); ASSERT(!S_ISDIR(sip->i_d.di_mode)); if (XFS_FORCED_SHUTDOWN(mp)) return XFS_ERROR(EIO); - if (DM_EVENT_ENABLED(tdp, DM_EVENT_LINK)) { - error = XFS_SEND_NAMESP(mp, DM_EVENT_LINK, - tdp, DM_RIGHT_NULL, - sip, DM_RIGHT_NULL, - target_name->name, NULL, 0, 0, 0); - if (error) - return error; - } - - /* Return through std_return after this point. */ - error = xfs_qm_dqattach(sip, 0); if (error) goto std_return; @@ -1953,15 +1861,8 @@ xfs_link( xfs_lock_two_inodes(sip, tdp, XFS_ILOCK_EXCL); - /* - * Increment vnode ref counts since xfs_trans_commit & - * xfs_trans_cancel will both unlock the inodes and - * decrement the associated ref counts. - */ - IHOLD(sip); - IHOLD(tdp); - xfs_trans_ijoin(tp, sip, XFS_ILOCK_EXCL); - xfs_trans_ijoin(tp, tdp, XFS_ILOCK_EXCL); + xfs_trans_ijoin_ref(tp, sip, XFS_ILOCK_EXCL); + xfs_trans_ijoin_ref(tp, tdp, XFS_ILOCK_EXCL); /* * If the source has too many links, we can't make any more to it. @@ -2014,27 +1915,14 @@ xfs_link( goto abort_return; } - error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES); - if (error) - goto std_return; - - /* Fall through to std_return with error = 0. */ -std_return: - if (DM_EVENT_ENABLED(sip, DM_EVENT_POSTLINK)) { - (void) XFS_SEND_NAMESP(mp, DM_EVENT_POSTLINK, - tdp, DM_RIGHT_NULL, - sip, DM_RIGHT_NULL, - target_name->name, NULL, 0, error, 0); - } - return error; + return xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES); abort_return: cancel_flags |= XFS_TRANS_ABORT; - /* FALLTHROUGH */ - error_return: xfs_trans_cancel(tp, cancel_flags); - goto std_return; + std_return: + return error; } int @@ -2074,7 +1962,7 @@ xfs_symlink( ip = NULL; tp = NULL; - xfs_itrace_entry(dp); + trace_xfs_symlink(dp, link_name); if (XFS_FORCED_SHUTDOWN(mp)) return XFS_ERROR(EIO); @@ -2086,17 +1974,6 @@ xfs_symlink( if (pathlen >= MAXPATHLEN) /* total string too long */ return XFS_ERROR(ENAMETOOLONG); - if (DM_EVENT_ENABLED(dp, DM_EVENT_SYMLINK)) { - error = XFS_SEND_NAMESP(mp, DM_EVENT_SYMLINK, dp, - DM_RIGHT_NULL, NULL, DM_RIGHT_NULL, - link_name->name, - (unsigned char *)target_path, 0, 0, 0); - if (error) - return error; - } - - /* Return through std_return after this point. */ - udqp = gdqp = NULL; if (dp->i_d.di_flags & XFS_DIFLAG_PROJINHERIT) prid = dp->i_d.di_projid; @@ -2180,8 +2057,7 @@ xfs_symlink( * transaction cancel unlocking dp so don't do it explicitly in the * error path. */ - IHOLD(dp); - xfs_trans_ijoin(tp, dp, XFS_ILOCK_EXCL); + xfs_trans_ijoin_ref(tp, dp, XFS_ILOCK_EXCL); unlock_dp_on_error = B_FALSE; /* @@ -2215,7 +2091,7 @@ xfs_symlink( error = xfs_bmapi(tp, ip, first_fsb, fs_blocks, XFS_BMAPI_WRITE | XFS_BMAPI_METADATA, &first_block, resblks, mval, &nmaps, - &free_list, NULL); + &free_list); if (error) { goto error1; } @@ -2278,21 +2154,8 @@ xfs_symlink( xfs_qm_dqrele(udqp); xfs_qm_dqrele(gdqp); - /* Fall through to std_return with error = 0 or errno from - * xfs_trans_commit */ -std_return: - if (DM_EVENT_ENABLED(dp, DM_EVENT_POSTSYMLINK)) { - (void) XFS_SEND_NAMESP(mp, DM_EVENT_POSTSYMLINK, - dp, DM_RIGHT_NULL, - error ? NULL : ip, - DM_RIGHT_NULL, link_name->name, - (unsigned char *)target_path, - 0, error, 0); - } - - if (!error) - *ipp = ip; - return error; + *ipp = ip; + return 0; error2: IRELE(ip); @@ -2306,8 +2169,8 @@ std_return: if (unlock_dp_on_error) xfs_iunlock(dp, XFS_ILOCK_EXCL); - - goto std_return; + std_return: + return error; } int @@ -2333,13 +2196,12 @@ xfs_set_dmattrs( return error; } xfs_ilock(ip, XFS_ILOCK_EXCL); - xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL); + xfs_trans_ijoin_ref(tp, ip, XFS_ILOCK_EXCL); ip->i_d.di_dmevmask = evmask; ip->i_d.di_dmstate = state; xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); - IHOLD(ip); error = xfs_trans_commit(tp, 0); return error; @@ -2390,7 +2252,7 @@ xfs_alloc_file_space( int committed; int error; - xfs_itrace_entry(ip); + trace_xfs_alloc_file_space(ip); if (XFS_FORCED_SHUTDOWN(mp)) return XFS_ERROR(EIO); @@ -2412,25 +2274,9 @@ xfs_alloc_file_space( startoffset_fsb = XFS_B_TO_FSBT(mp, offset); allocatesize_fsb = XFS_B_TO_FSB(mp, count); - /* Generate a DMAPI event if needed. */ - if (alloc_type != 0 && offset < ip->i_size && - (attr_flags & XFS_ATTR_DMI) == 0 && - DM_EVENT_ENABLED(ip, DM_EVENT_WRITE)) { - xfs_off_t end_dmi_offset; - - end_dmi_offset = offset+len; - if (end_dmi_offset > ip->i_size) - end_dmi_offset = ip->i_size; - error = XFS_SEND_DATA(mp, DM_EVENT_WRITE, ip, offset, - end_dmi_offset - offset, 0, NULL); - if (error) - return error; - } - /* * Allocate file space until done or until there is an error */ -retry: while (allocatesize_fsb && !error) { xfs_fileoff_t s, e; @@ -2488,8 +2334,7 @@ retry: if (error) goto error1; - xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL); - xfs_trans_ihold(tp, ip); + xfs_trans_ijoin(tp, ip); /* * Issue the xfs_bmapi() call to allocate the blocks @@ -2498,7 +2343,7 @@ retry: error = xfs_bmapi(tp, ip, startoffset_fsb, allocatesize_fsb, bmapi_flag, &firstfsb, 0, imapp, &nimaps, - &free_list, NULL); + &free_list); if (error) { goto error0; } @@ -2527,17 +2372,6 @@ retry: startoffset_fsb += allocated_fsb; allocatesize_fsb -= allocated_fsb; } -dmapi_enospc_check: - if (error == ENOSPC && (attr_flags & XFS_ATTR_DMI) == 0 && - DM_EVENT_ENABLED(ip, DM_EVENT_NOSPACE)) { - error = XFS_SEND_NAMESP(mp, DM_EVENT_NOSPACE, - ip, DM_RIGHT_NULL, - ip, DM_RIGHT_NULL, - NULL, NULL, 0, 0, 0); /* Delay flag intentionally unused */ - if (error == 0) - goto retry; /* Maybe DMAPI app. has made space */ - /* else fall through with error from XFS_SEND_DATA */ - } return error; @@ -2548,7 +2382,7 @@ error0: /* Cancel bmap, unlock inode, unreserve quota blocks, cancel trans */ error1: /* Just cancel transaction */ xfs_trans_cancel(tp, XFS_TRANS_RELEASE_LOG_RES | XFS_TRANS_ABORT); xfs_iunlock(ip, XFS_ILOCK_EXCL); - goto dmapi_enospc_check; + return error; } /* @@ -2598,7 +2432,7 @@ xfs_zero_remaining_bytes( offset_fsb = XFS_B_TO_FSBT(mp, offset); nimap = 1; error = xfs_bmapi(NULL, ip, offset_fsb, 1, 0, - NULL, 0, &imap, &nimap, NULL, NULL); + NULL, 0, &imap, &nimap, NULL); if (error || nimap < 1) break; ASSERT(imap.br_blockcount >= 1); @@ -2661,7 +2495,6 @@ xfs_free_file_space( { int committed; int done; - xfs_off_t end_dmi_offset; xfs_fileoff_t endoffset_fsb; int error; xfs_fsblock_t firstfsb; @@ -2680,7 +2513,7 @@ xfs_free_file_space( mp = ip->i_mount; - xfs_itrace_entry(ip); + trace_xfs_free_file_space(ip); error = xfs_qm_dqattach(ip, 0); if (error) @@ -2691,19 +2524,7 @@ xfs_free_file_space( return error; rt = XFS_IS_REALTIME_INODE(ip); startoffset_fsb = XFS_B_TO_FSB(mp, offset); - end_dmi_offset = offset + len; - endoffset_fsb = XFS_B_TO_FSBT(mp, end_dmi_offset); - - if (offset < ip->i_size && (attr_flags & XFS_ATTR_DMI) == 0 && - DM_EVENT_ENABLED(ip, DM_EVENT_WRITE)) { - if (end_dmi_offset > ip->i_size) - end_dmi_offset = ip->i_size; - error = XFS_SEND_DATA(mp, DM_EVENT_WRITE, ip, - offset, end_dmi_offset - offset, - AT_DELAY_FLAG(attr_flags), NULL); - if (error) - return error; - } + endoffset_fsb = XFS_B_TO_FSBT(mp, offset + len); if (attr_flags & XFS_ATTR_NOLOCK) need_iolock = 0; @@ -2731,7 +2552,7 @@ xfs_free_file_space( if (rt && !xfs_sb_version_hasextflgbit(&mp->m_sb)) { nimap = 1; error = xfs_bmapi(NULL, ip, startoffset_fsb, - 1, 0, NULL, 0, &imap, &nimap, NULL, NULL); + 1, 0, NULL, 0, &imap, &nimap, NULL); if (error) goto out_unlock_iolock; ASSERT(nimap == 0 || nimap == 1); @@ -2746,7 +2567,7 @@ xfs_free_file_space( } nimap = 1; error = xfs_bmapi(NULL, ip, endoffset_fsb - 1, - 1, 0, NULL, 0, &imap, &nimap, NULL, NULL); + 1, 0, NULL, 0, &imap, &nimap, NULL); if (error) goto out_unlock_iolock; ASSERT(nimap == 0 || nimap == 1); @@ -2814,8 +2635,7 @@ xfs_free_file_space( if (error) goto error1; - xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL); - xfs_trans_ihold(tp, ip); + xfs_trans_ijoin(tp, ip); /* * issue the bunmapi() call to free the blocks @@ -2823,7 +2643,7 @@ xfs_free_file_space( xfs_bmap_init(&free_list, &firstfsb); error = xfs_bunmapi(tp, ip, startoffset_fsb, endoffset_fsb - startoffset_fsb, - 0, 2, &firstfsb, &free_list, NULL, &done); + 0, 2, &firstfsb, &free_list, &done); if (error) { goto error0; } @@ -2883,8 +2703,6 @@ xfs_change_file_space( xfs_trans_t *tp; struct iattr iattr; - xfs_itrace_entry(ip); - if (!S_ISREG(ip->i_d.di_mode)) return XFS_ERROR(EINVAL); @@ -2985,8 +2803,7 @@ xfs_change_file_space( xfs_ilock(ip, XFS_ILOCK_EXCL); - xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL); - xfs_trans_ihold(tp, ip); + xfs_trans_ijoin(tp, ip); if ((attr_flags & XFS_ATTR_DMI) == 0) { ip->i_d.di_mode &= ~S_ISUID; |