diff options
Diffstat (limited to 'fs')
546 files changed, 20658 insertions, 15608 deletions
diff --git a/fs/9p/conv.c b/fs/9p/conv.c index a767e05b60b..56d88c1a09c 100644 --- a/fs/9p/conv.c +++ b/fs/9p/conv.c @@ -24,7 +24,6 @@ * */ -#include <linux/config.h> #include <linux/module.h> #include <linux/errno.h> #include <linux/fs.h> @@ -674,8 +673,10 @@ struct v9fs_fcall *v9fs_create_tcreate(u32 fid, char *name, u32 perm, u8 mode, struct cbuf *bufp = &buffer; size = 4 + 2 + strlen(name) + 4 + 1; /* fid[4] name[s] perm[4] mode[1] */ - if (extended && extension!=NULL) - size += 2 + strlen(extension); /* extension[s] */ + if (extended) { + size += 2 + /* extension[s] */ + (extension == NULL ? 0 : strlen(extension)); + } fc = v9fs_create_common(bufp, size, TCREATE); if (IS_ERR(fc)) diff --git a/fs/9p/error.c b/fs/9p/error.c index 981fe8ecd78..ae91555c155 100644 --- a/fs/9p/error.c +++ b/fs/9p/error.c @@ -27,7 +27,6 @@ * */ -#include <linux/config.h> #include <linux/module.h> #include <linux/list.h> diff --git a/fs/9p/fcall.c b/fs/9p/fcall.c index 6f2617820a4..8556097fcda 100644 --- a/fs/9p/fcall.c +++ b/fs/9p/fcall.c @@ -24,7 +24,6 @@ * */ -#include <linux/config.h> #include <linux/module.h> #include <linux/errno.h> #include <linux/fs.h> diff --git a/fs/9p/fcprint.c b/fs/9p/fcprint.c index 583e827baeb..34b96114a28 100644 --- a/fs/9p/fcprint.c +++ b/fs/9p/fcprint.c @@ -21,7 +21,6 @@ * Boston, MA 02111-1301 USA * */ -#include <linux/config.h> #include <linux/module.h> #include <linux/errno.h> #include <linux/fs.h> diff --git a/fs/9p/fid.c b/fs/9p/fid.c index b7608af07ce..70492ccb438 100644 --- a/fs/9p/fid.c +++ b/fs/9p/fid.c @@ -20,7 +20,6 @@ * */ -#include <linux/config.h> #include <linux/module.h> #include <linux/errno.h> #include <linux/fs.h> diff --git a/fs/9p/mux.c b/fs/9p/mux.c index f4407eb276c..90a79c78454 100644 --- a/fs/9p/mux.c +++ b/fs/9p/mux.c @@ -23,7 +23,6 @@ * */ -#include <linux/config.h> #include <linux/module.h> #include <linux/errno.h> #include <linux/fs.h> @@ -712,7 +711,7 @@ static void v9fs_read_work(void *a) * v9fs_send_request - send 9P request * The function can sleep until the request is scheduled for sending. * The function can be interrupted. Return from the function is not - * a guarantee that the request is sent succesfully. Can return errors + * a guarantee that the request is sent successfully. Can return errors * that can be retrieved by PTR_ERR macros. * * @m: mux data @@ -932,6 +931,8 @@ v9fs_mux_rpc(struct v9fs_mux_data *m, struct v9fs_fcall *tc, r.rcall || r.err); } while (!r.rcall && !r.err && err==-ERESTARTSYS && m->trans->status==Connected && !m->err); + + err = -ERESTARTSYS; } sigpending = 1; } diff --git a/fs/9p/trans_fd.c b/fs/9p/trans_fd.c index 94e0a7fd9fc..34d43355beb 100644 --- a/fs/9p/trans_fd.c +++ b/fs/9p/trans_fd.c @@ -25,7 +25,6 @@ * */ -#include <linux/config.h> #include <linux/in.h> #include <linux/module.h> #include <linux/net.h> diff --git a/fs/9p/v9fs.c b/fs/9p/v9fs.c index d37416eb579..0f628041e3f 100644 --- a/fs/9p/v9fs.c +++ b/fs/9p/v9fs.c @@ -23,7 +23,6 @@ * */ -#include <linux/config.h> #include <linux/module.h> #include <linux/errno.h> #include <linux/fs.h> @@ -461,8 +460,10 @@ static int __init init_v9fs(void) ret = v9fs_mux_global_init(); if (!ret) - ret = register_filesystem(&v9fs_fs_type); - + return ret; + ret = register_filesystem(&v9fs_fs_type); + if (!ret) + v9fs_mux_global_exit(); return ret; } diff --git a/fs/9p/v9fs_vfs.h b/fs/9p/v9fs_vfs.h index f867b8d3e97..450b0c1b385 100644 --- a/fs/9p/v9fs_vfs.h +++ b/fs/9p/v9fs_vfs.h @@ -38,7 +38,7 @@ */ extern struct file_system_type v9fs_fs_type; -extern struct address_space_operations v9fs_addr_operations; +extern const struct address_space_operations v9fs_addr_operations; extern const struct file_operations v9fs_file_operations; extern const struct file_operations v9fs_dir_operations; extern struct dentry_operations v9fs_dentry_operations; diff --git a/fs/9p/vfs_addr.c b/fs/9p/vfs_addr.c index efda46fb64d..9dfd259a70b 100644 --- a/fs/9p/vfs_addr.c +++ b/fs/9p/vfs_addr.c @@ -31,7 +31,6 @@ #include <linux/string.h> #include <linux/smp_lock.h> #include <linux/inet.h> -#include <linux/version.h> #include <linux/pagemap.h> #include <linux/idr.h> @@ -103,6 +102,6 @@ UnmapAndUnlock: return retval; } -struct address_space_operations v9fs_addr_operations = { +const struct address_space_operations v9fs_addr_operations = { .readpage = v9fs_vfs_readpage, }; diff --git a/fs/9p/vfs_file.c b/fs/9p/vfs_file.c index 1a8e46084f0..c3c47eda757 100644 --- a/fs/9p/vfs_file.c +++ b/fs/9p/vfs_file.c @@ -31,7 +31,6 @@ #include <linux/string.h> #include <linux/smp_lock.h> #include <linux/inet.h> -#include <linux/version.h> #include <linux/list.h> #include <asm/uaccess.h> #include <linux/idr.h> diff --git a/fs/9p/vfs_inode.c b/fs/9p/vfs_inode.c index 2cb87ba4b1c..7a7ec2d1d2f 100644 --- a/fs/9p/vfs_inode.c +++ b/fs/9p/vfs_inode.c @@ -204,7 +204,6 @@ struct inode *v9fs_get_inode(struct super_block *sb, int mode) inode->i_mode = mode; inode->i_uid = current->fsuid; inode->i_gid = current->fsgid; - inode->i_blksize = sb->s_blocksize; inode->i_blocks = 0; inode->i_rdev = 0; inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME; @@ -300,7 +299,7 @@ clunk_fid: fid = V9FS_NOFID; put_fid: - if (fid >= 0) + if (fid != V9FS_NOFID) v9fs_put_idpool(fid, &v9ses->fidpool); kfree(fcall); @@ -434,11 +433,11 @@ static int v9fs_remove(struct inode *dir, struct dentry *file, int rmdir) result = v9fs_t_remove(v9ses, fid, &fcall); if (result < 0) { PRINT_FCALL_ERROR("remove fails", fcall); - } else { - v9fs_put_idpool(fid, &v9ses->fidpool); - v9fs_fid_destroy(v9fid); } + v9fs_put_idpool(fid, &v9ses->fidpool); + v9fs_fid_destroy(v9fid); + kfree(fcall); return result; } @@ -530,9 +529,6 @@ error: if (vfid) v9fs_fid_destroy(vfid); - if (inode) - iput(inode); - return err; } @@ -953,9 +949,8 @@ v9fs_stat2inode(struct v9fs_stat *stat, struct inode *inode, inode->i_size = stat->length; - inode->i_blksize = sb->s_blocksize; inode->i_blocks = - (inode->i_size + inode->i_blksize - 1) >> sb->s_blocksize_bits; + (inode->i_size + sb->s_blocksize - 1) >> sb->s_blocksize_bits; } /** @@ -1054,6 +1049,9 @@ static int v9fs_vfs_readlink(struct dentry *dentry, char __user * buffer, int ret; char *link = __getname(); + if (unlikely(!link)) + return -ENOMEM; + if (buflen > PATH_MAX) buflen = PATH_MAX; @@ -1171,9 +1169,6 @@ error: if (vfid) v9fs_fid_destroy(vfid); - if (inode) - iput(inode); - return err; } @@ -1227,6 +1222,9 @@ v9fs_vfs_link(struct dentry *old_dentry, struct inode *dir, } name = __getname(); + if (unlikely(!name)) + return -ENOMEM; + sprintf(name, "%d\n", oldfid->fid); retval = v9fs_vfs_mkspecial(dir, dentry, V9FS_DMLINK, name); __putname(name); diff --git a/fs/9p/vfs_super.c b/fs/9p/vfs_super.c index 872943004e5..63320d4e15d 100644 --- a/fs/9p/vfs_super.c +++ b/fs/9p/vfs_super.c @@ -25,7 +25,6 @@ */ #include <linux/kernel.h> -#include <linux/config.h> #include <linux/module.h> #include <linux/errno.h> #include <linux/fs.h> @@ -256,11 +255,12 @@ static int v9fs_show_options(struct seq_file *m, struct vfsmount *mnt) } static void -v9fs_umount_begin(struct super_block *sb) +v9fs_umount_begin(struct vfsmount *vfsmnt, int flags) { - struct v9fs_session_info *v9ses = sb->s_fs_info; + struct v9fs_session_info *v9ses = vfsmnt->mnt_sb->s_fs_info; - v9fs_session_cancel(v9ses); + if (flags & MNT_FORCE) + v9fs_session_cancel(v9ses); } static struct super_operations v9fs_super_ops = { diff --git a/fs/Kconfig b/fs/Kconfig index 2aa4624cc01..4fd9efac29a 100644 --- a/fs/Kconfig +++ b/fs/Kconfig @@ -325,8 +325,8 @@ config FS_POSIX_ACL source "fs/xfs/Kconfig" config OCFS2_FS - tristate "OCFS2 file system support (EXPERIMENTAL)" - depends on NET && EXPERIMENTAL + tristate "OCFS2 file system support" + depends on NET && SYSFS select CONFIGFS_FS select JBD select CRC32 @@ -356,6 +356,16 @@ config OCFS2_FS - POSIX ACLs - readpages / writepages (not user visible) +config OCFS2_DEBUG_MASKLOG + bool "OCFS2 logging support" + depends on OCFS2_FS + default y + help + The ocfs2 filesystem has an extensive logging system. The system + allows selection of events to log via files in /sys/o2cb/logmask/. + This option will enlarge your kernel, but it allows debugging of + ocfs2 filesystem issues. + config MINIX_FS tristate "Minix fs support" help @@ -776,7 +786,8 @@ endmenu menu "Pseudo filesystems" config PROC_FS - bool "/proc file system support" + bool "/proc file system support" if EMBEDDED + default y help This is a virtual file system providing information about the status of the system. "Virtual" means that it doesn't take up any space on @@ -815,6 +826,25 @@ config PROC_VMCORE help Exports the dump image of crashed kernel in ELF format. +config PROC_SYSCTL + bool "Sysctl support (/proc/sys)" if EMBEDDED + depends on PROC_FS + select SYSCTL + default y + ---help--- + The sysctl interface provides a means of dynamically changing + certain kernel parameters and variables on the fly without requiring + a recompile of the kernel or reboot of the system. The primary + interface is through /proc/sys. If you say Y here a tree of + modifiable sysctl entries will be generated beneath the + /proc/sys directory. They are explained in the files + in <file:Documentation/sysctl/>. Note that enabling this + option will enlarge the kernel by at least 8 KB. + + As it is generally a good thing, you should say Y here unless + building a kernel for install/rescue disks or your system is very + limited in memory. + config SYSFS bool "sysfs file system support" if EMBEDDED default y @@ -851,6 +881,19 @@ config TMPFS See <file:Documentation/filesystems/tmpfs.txt> for details. +config TMPFS_POSIX_ACL + bool "Tmpfs POSIX Access Control Lists" + depends on TMPFS + select GENERIC_ACL + help + POSIX Access Control Lists (ACLs) support permissions for users and + groups beyond the owner/group/world scheme. + + To learn more about Access Control Lists, visit the POSIX ACLs for + Linux website <http://acl.bestbits.at/>. + + If you don't know what Access Control Lists are, say N. + config HUGETLBFS bool "HugeTLB file system support" depends X86 || IA64 || PPC64 || SPARC64 || SUPERH || BROKEN @@ -1115,7 +1158,7 @@ config JFFS2_SUMMARY config JFFS2_FS_XATTR bool "JFFS2 XATTR support (EXPERIMENTAL)" - depends on JFFS2_FS && EXPERIMENTAL && !JFFS2_FS_WRITEBUFFER + depends on JFFS2_FS && EXPERIMENTAL default n help Extended attributes are name:value pairs associated with inodes by @@ -1370,11 +1413,19 @@ config UFS_FS config UFS_FS_WRITE bool "UFS file system write support (DANGEROUS)" - depends on UFS_FS && EXPERIMENTAL && BROKEN + depends on UFS_FS && EXPERIMENTAL help Say Y here if you want to try writing to UFS partitions. This is experimental, so you should back up your UFS partitions beforehand. +config UFS_DEBUG + bool "UFS debugging" + depends on UFS_FS + help + If you are experiencing any problems with the UFS filesystem, say + Y here. This will result in _many_ additional debugging messages to be + written to the system log. + endmenu menu "Network File Systems" @@ -1452,8 +1503,8 @@ config NFS_V4 If unsure, say N. config NFS_DIRECTIO - bool "Allow direct I/O on NFS files (EXPERIMENTAL)" - depends on NFS_FS && EXPERIMENTAL + bool "Allow direct I/O on NFS files" + depends on NFS_FS help This option enables applications to perform uncached I/O on files in NFS file systems using the O_DIRECT open() flag. When O_DIRECT @@ -1481,7 +1532,12 @@ config NFSD select LOCKD select SUNRPC select EXPORTFS - select NFS_ACL_SUPPORT if NFSD_V3_ACL || NFSD_V2_ACL + select NFSD_V2_ACL if NFSD_V3_ACL + select NFS_ACL_SUPPORT if NFSD_V2_ACL + select NFSD_TCP if NFSD_V4 + select CRYPTO_MD5 if NFSD_V4 + select CRYPTO if NFSD_V4 + select FS_POSIX_ACL if NFSD_V4 help If you want your Linux box to act as an NFS *server*, so that other computers on your local network which support NFS can access certain @@ -1519,7 +1575,6 @@ config NFSD_V3 config NFSD_V3_ACL bool "Provide server support for the NFSv3 ACL protocol extension" depends on NFSD_V3 - select NFSD_V2_ACL help Implement the NFSv3 ACL protocol extension for manipulating POSIX Access Control Lists on exported file systems. NFS clients should @@ -1529,10 +1584,6 @@ config NFSD_V3_ACL config NFSD_V4 bool "Provide NFSv4 server support (EXPERIMENTAL)" depends on NFSD_V3 && EXPERIMENTAL - select NFSD_TCP - select CRYPTO_MD5 - select CRYPTO - select FS_POSIX_ACL help If you would like to include the NFSv4 server as well as the NFSv2 and NFSv3 servers, say Y here. This feature is experimental, and @@ -1713,7 +1764,7 @@ config CIFS_STATS mounted by the cifs client to be displayed in /proc/fs/cifs/Stats config CIFS_STATS2 - bool "CIFS extended statistics" + bool "Extended statistics" depends on CIFS_STATS help Enabling this option will allow more detailed statistics on SMB @@ -1726,6 +1777,32 @@ config CIFS_STATS2 Unless you are a developer or are doing network performance analysis or tuning, say N. +config CIFS_WEAK_PW_HASH + bool "Support legacy servers which use weaker LANMAN security" + depends on CIFS + help + Modern CIFS servers including Samba and most Windows versions + (since 1997) support stronger NTLM (and even NTLMv2 and Kerberos) + security mechanisms. These hash the password more securely + than the mechanisms used in the older LANMAN version of the + SMB protocol needed to establish sessions with old SMB servers. + + Enabling this option allows the cifs module to mount to older + LANMAN based servers such as OS/2 and Windows 95, but such + mounts may be less secure than mounts using NTLM or more recent + security mechanisms if you are on a public network. Unless you + have a need to access old SMB servers (and are on a private + network) you probably want to say N. Even if this support + is enabled in the kernel build, they will not be used + automatically. At runtime LANMAN mounts are disabled but + can be set to required (or optional) either in + /proc/fs/cifs (see fs/cifs/README for more detail) or via an + option on the mount command. This support is disabled by + default in order to reduce the possibility of a downgrade + attack. + + If unsure, say N. + config CIFS_XATTR bool "CIFS extended attributes" depends on CIFS @@ -1754,6 +1831,17 @@ config CIFS_POSIX (such as Samba 3.10 and later) which can negotiate CIFS POSIX ACL support. If unsure, say N. +config CIFS_DEBUG2 + bool "Enable additional CIFS debugging routines" + depends on CIFS + help + Enabling this option adds a few more debugging routines + to the cifs code which slightly increases the size of + the cifs module and can cause additional logging of debug + messages in some error paths, slowing performance. This + option can be turned off unless you are debugging + cifs problems. If unsure, say N. + config CIFS_EXPERIMENTAL bool "CIFS Experimental Features (EXPERIMENTAL)" depends on CIFS && EXPERIMENTAL @@ -1769,7 +1857,7 @@ config CIFS_EXPERIMENTAL If unsure, say N. config CIFS_UPCALL - bool "CIFS Kerberos/SPNEGO advanced session setup (EXPERIMENTAL)" + bool "Kerberos/SPNEGO advanced session setup (EXPERIMENTAL)" depends on CIFS_EXPERIMENTAL select CONNECTOR help @@ -1865,6 +1953,10 @@ config 9P_FS If unsure, say N. +config GENERIC_ACL + bool + select FS_POSIX_ACL + endmenu menu "Partition Types" diff --git a/fs/Makefile b/fs/Makefile index d0ea6bfccf2..46b8cfe497b 100644 --- a/fs/Makefile +++ b/fs/Makefile @@ -35,6 +35,7 @@ obj-$(CONFIG_BINFMT_FLAT) += binfmt_flat.o obj-$(CONFIG_FS_MBCACHE) += mbcache.o obj-$(CONFIG_FS_POSIX_ACL) += posix_acl.o xattr_acl.o obj-$(CONFIG_NFS_COMMON) += nfs_common/ +obj-$(CONFIG_GENERIC_ACL) += generic_acl.o obj-$(CONFIG_QUOTA) += dquot.o obj-$(CONFIG_QFMT_V1) += quota_v1.o @@ -66,7 +67,6 @@ obj-$(CONFIG_MSDOS_FS) += msdos/ obj-$(CONFIG_VFAT_FS) += vfat/ obj-$(CONFIG_BFS_FS) += bfs/ obj-$(CONFIG_ISO9660_FS) += isofs/ -obj-$(CONFIG_DEVFS_FS) += devfs/ obj-$(CONFIG_HFSPLUS_FS) += hfsplus/ # Before hfs to find wrapped HFS+ obj-$(CONFIG_HFS_FS) += hfs/ obj-$(CONFIG_VXFS_FS) += freevxfs/ diff --git a/fs/adfs/dir.c b/fs/adfs/dir.c index 7b075fc397d..d3c7905b2dd 100644 --- a/fs/adfs/dir.c +++ b/fs/adfs/dir.c @@ -9,7 +9,6 @@ * * Common directory handling for ADFS */ -#include <linux/config.h> #include <linux/errno.h> #include <linux/fs.h> #include <linux/adfs_fs.h> diff --git a/fs/adfs/inode.c b/fs/adfs/inode.c index a02802a3079..7e7a04be127 100644 --- a/fs/adfs/inode.c +++ b/fs/adfs/inode.c @@ -72,7 +72,7 @@ static sector_t _adfs_bmap(struct address_space *mapping, sector_t block) return generic_block_bmap(mapping, block, adfs_get_block); } -static struct address_space_operations adfs_aops = { +static const struct address_space_operations adfs_aops = { .readpage = adfs_readpage, .writepage = adfs_writepage, .sync_page = block_sync_page, @@ -269,7 +269,6 @@ adfs_iget(struct super_block *sb, struct object_info *obj) inode->i_ino = obj->file_id; inode->i_size = obj->size; inode->i_nlink = 2; - inode->i_blksize = PAGE_SIZE; inode->i_blocks = (inode->i_size + sb->s_blocksize - 1) >> sb->s_blocksize_bits; diff --git a/fs/adfs/super.c b/fs/adfs/super.c index ba1c88af49f..9ade139086f 100644 --- a/fs/adfs/super.c +++ b/fs/adfs/super.c @@ -251,8 +251,7 @@ static int init_inodecache(void) static void destroy_inodecache(void) { - if (kmem_cache_destroy(adfs_inode_cachep)) - printk(KERN_INFO "adfs_inode_cache: not all structures were freed\n"); + kmem_cache_destroy(adfs_inode_cachep); } static struct super_operations adfs_sops = { @@ -308,7 +307,7 @@ static struct adfs_discmap *adfs_read_map(struct super_block *sb, struct adfs_di if (adfs_checkmap(sb, dm)) return dm; - adfs_error(sb, NULL, "map corrupted"); + adfs_error(sb, "map corrupted"); error_free: while (--zone >= 0) @@ -339,11 +338,10 @@ static int adfs_fill_super(struct super_block *sb, void *data, int silent) sb->s_flags |= MS_NODIRATIME; - asb = kmalloc(sizeof(*asb), GFP_KERNEL); + asb = kzalloc(sizeof(*asb), GFP_KERNEL); if (!asb) return -ENOMEM; sb->s_fs_info = asb; - memset(asb, 0, sizeof(*asb)); /* set default options */ asb->s_uid = 0; diff --git a/fs/affs/affs.h b/fs/affs/affs.h index a43a876742b..1dc8438ef38 100644 --- a/fs/affs/affs.h +++ b/fs/affs/affs.h @@ -1,7 +1,6 @@ #include <linux/types.h> #include <linux/fs.h> #include <linux/buffer_head.h> -#include <linux/affs_fs.h> #include <linux/amigaffs.h> /* AmigaOS allows file names with up to 30 characters length. @@ -195,9 +194,9 @@ extern struct inode_operations affs_symlink_inode_operations; extern const struct file_operations affs_file_operations; extern const struct file_operations affs_file_operations_ofs; extern const struct file_operations affs_dir_operations; -extern struct address_space_operations affs_symlink_aops; -extern struct address_space_operations affs_aops; -extern struct address_space_operations affs_aops_ofs; +extern const struct address_space_operations affs_symlink_aops; +extern const struct address_space_operations affs_aops; +extern const struct address_space_operations affs_aops_ofs; extern struct dentry_operations affs_dentry_operations; extern struct dentry_operations affs_dentry_operations_intl; diff --git a/fs/affs/file.c b/fs/affs/file.c index 7076262af39..3de8590e4f6 100644 --- a/fs/affs/file.c +++ b/fs/affs/file.c @@ -406,7 +406,7 @@ static sector_t _affs_bmap(struct address_space *mapping, sector_t block) { return generic_block_bmap(mapping,block,affs_get_block); } -struct address_space_operations affs_aops = { +const struct address_space_operations affs_aops = { .readpage = affs_readpage, .writepage = affs_writepage, .sync_page = block_sync_page, @@ -759,7 +759,7 @@ out: goto done; } -struct address_space_operations affs_aops_ofs = { +const struct address_space_operations affs_aops_ofs = { .readpage = affs_readpage_ofs, //.writepage = affs_writepage_ofs, //.sync_page = affs_sync_page_ofs, diff --git a/fs/affs/super.c b/fs/affs/super.c index 8765cba35bb..5ea72c3a16c 100644 --- a/fs/affs/super.c +++ b/fs/affs/super.c @@ -14,6 +14,7 @@ #include <linux/init.h> #include <linux/statfs.h> #include <linux/parser.h> +#include <linux/magic.h> #include "affs.h" extern struct timezone sys_tz; @@ -108,8 +109,7 @@ static int init_inodecache(void) static void destroy_inodecache(void) { - if (kmem_cache_destroy(affs_inode_cachep)) - printk(KERN_INFO "affs_inode_cache: not all structures were freed\n"); + kmem_cache_destroy(affs_inode_cachep); } static struct super_operations affs_sops = { @@ -271,6 +271,7 @@ static int affs_fill_super(struct super_block *sb, void *data, int silent) int reserved; unsigned long mount_flags; int tmp_flags; /* fix remount prototype... */ + u8 sig[4]; pr_debug("AFFS: read_super(%s)\n",data ? (const char *)data : "no options"); @@ -278,11 +279,10 @@ static int affs_fill_super(struct super_block *sb, void *data, int silent) sb->s_op = &affs_sops; sb->s_flags |= MS_NODIRATIME; - sbi = kmalloc(sizeof(struct affs_sb_info), GFP_KERNEL); + sbi = kzalloc(sizeof(struct affs_sb_info), GFP_KERNEL); if (!sbi) return -ENOMEM; sb->s_fs_info = sbi; - memset(sbi, 0, sizeof(*sbi)); init_MUTEX(&sbi->s_bmlock); if (!parse_options(data,&uid,&gid,&i,&reserved,&root_block, @@ -370,8 +370,9 @@ got_root: printk(KERN_ERR "AFFS: Cannot read boot block\n"); goto out_error; } - chksum = be32_to_cpu(*(__be32 *)boot_bh->b_data); + memcpy(sig, boot_bh->b_data, 4); brelse(boot_bh); + chksum = be32_to_cpu(*(__be32 *)sig); /* Dircache filesystems are compatible with non-dircache ones * when reading. As long as they aren't supported, writing is @@ -420,11 +421,11 @@ got_root: } if (mount_flags & SF_VERBOSE) { - chksum = cpu_to_be32(chksum); - printk(KERN_NOTICE "AFFS: Mounting volume \"%*s\": Type=%.3s\\%c, Blocksize=%d\n", - AFFS_ROOT_TAIL(sb, root_bh)->disk_name[0], + u8 len = AFFS_ROOT_TAIL(sb, root_bh)->disk_name[0]; + printk(KERN_NOTICE "AFFS: Mounting volume \"%.*s\": Type=%.3s\\%c, Blocksize=%d\n", + len > 31 ? 31 : len, AFFS_ROOT_TAIL(sb, root_bh)->disk_name + 1, - (char *)&chksum,((char *)&chksum)[3] + '0',blocksize); + sig, sig[3] + '0', blocksize); } sb->s_flags |= MS_NODEV | MS_NOSUID; diff --git a/fs/affs/symlink.c b/fs/affs/symlink.c index 426f0f094f2..f802256a593 100644 --- a/fs/affs/symlink.c +++ b/fs/affs/symlink.c @@ -66,7 +66,7 @@ fail: return err; } -struct address_space_operations affs_symlink_aops = { +const struct address_space_operations affs_symlink_aops = { .readpage = affs_symlink_readpage, }; diff --git a/fs/afs/cell.c b/fs/afs/cell.c index 009a9ae88d6..bfc1fd22d5b 100644 --- a/fs/afs/cell.c +++ b/fs/afs/cell.c @@ -413,8 +413,7 @@ int afs_server_find_by_peer(const struct rxrpc_peer *peer, /* we found it in the graveyard - resurrect it */ found_dead_server: - list_del(&server->link); - list_add_tail(&server->link, &cell->sv_list); + list_move_tail(&server->link, &cell->sv_list); afs_get_server(server); afs_kafstimod_del_timer(&server->timeout); spin_unlock(&cell->sv_gylock); diff --git a/fs/afs/file.c b/fs/afs/file.c index 7bb716887e2..67d6634101f 100644 --- a/fs/afs/file.c +++ b/fs/afs/file.c @@ -35,7 +35,7 @@ struct inode_operations afs_file_inode_operations = { .getattr = afs_inode_getattr, }; -struct address_space_operations afs_fs_aops = { +const struct address_space_operations afs_fs_aops = { .readpage = afs_file_readpage, .sync_page = block_sync_page, .set_page_dirty = __set_page_dirty_nobuffers, diff --git a/fs/afs/inode.c b/fs/afs/inode.c index 4ebb30a50ed..6f37754906c 100644 --- a/fs/afs/inode.c +++ b/fs/afs/inode.c @@ -72,7 +72,6 @@ static int afs_inode_map_status(struct afs_vnode *vnode) inode->i_ctime.tv_sec = vnode->status.mtime_server; inode->i_ctime.tv_nsec = 0; inode->i_atime = inode->i_mtime = inode->i_ctime; - inode->i_blksize = PAGE_CACHE_SIZE; inode->i_blocks = 0; inode->i_version = vnode->fid.unique; inode->i_mapping->a_ops = &afs_fs_aops; diff --git a/fs/afs/internal.h b/fs/afs/internal.h index 72febdf9a35..e88b3b65ae4 100644 --- a/fs/afs/internal.h +++ b/fs/afs/internal.h @@ -69,7 +69,7 @@ extern const struct file_operations afs_dir_file_operations; /* * file.c */ -extern struct address_space_operations afs_fs_aops; +extern const struct address_space_operations afs_fs_aops; extern struct inode_operations afs_file_inode_operations; #ifdef AFS_CACHING_SUPPORT diff --git a/fs/afs/kafsasyncd.c b/fs/afs/kafsasyncd.c index 7ac07d0d47b..f09a794f248 100644 --- a/fs/afs/kafsasyncd.c +++ b/fs/afs/kafsasyncd.c @@ -136,8 +136,7 @@ static int kafsasyncd(void *arg) if (!list_empty(&kafsasyncd_async_attnq)) { op = list_entry(kafsasyncd_async_attnq.next, struct afs_async_op, link); - list_del(&op->link); - list_add_tail(&op->link, + list_move_tail(&op->link, &kafsasyncd_async_busyq); } @@ -204,8 +203,7 @@ void afs_kafsasyncd_begin_op(struct afs_async_op *op) init_waitqueue_entry(&op->waiter, kafsasyncd_task); add_wait_queue(&op->call->waitq, &op->waiter); - list_del(&op->link); - list_add_tail(&op->link, &kafsasyncd_async_busyq); + list_move_tail(&op->link, &kafsasyncd_async_busyq); spin_unlock(&kafsasyncd_async_lock); @@ -223,8 +221,7 @@ void afs_kafsasyncd_attend_op(struct afs_async_op *op) spin_lock(&kafsasyncd_async_lock); - list_del(&op->link); - list_add_tail(&op->link, &kafsasyncd_async_attnq); + list_move_tail(&op->link, &kafsasyncd_async_attnq); spin_unlock(&kafsasyncd_async_lock); diff --git a/fs/afs/mntpt.c b/fs/afs/mntpt.c index b5cf9e1205a..99785a79d04 100644 --- a/fs/afs/mntpt.c +++ b/fs/afs/mntpt.c @@ -203,7 +203,7 @@ static struct vfsmount *afs_mntpt_do_automount(struct dentry *mntpt) /* try and do the mount */ kdebug("--- attempting mount %s -o %s ---", devname, options); - mnt = do_kern_mount("afs", 0, devname, options); + mnt = vfs_kern_mount(&afs_fs_type, 0, devname, options); kdebug("--- mount result %p ---", mnt); free_page((unsigned long) devname); diff --git a/fs/afs/proc.c b/fs/afs/proc.c index 101d21b6c03..86463ec9ccb 100644 --- a/fs/afs/proc.c +++ b/fs/afs/proc.c @@ -775,6 +775,7 @@ static int afs_proc_cell_servers_release(struct inode *inode, * first item */ static void *afs_proc_cell_servers_start(struct seq_file *m, loff_t *_pos) + __acquires(m->private->sv_lock) { struct list_head *_p; struct afs_cell *cell = m->private; @@ -823,6 +824,7 @@ static void *afs_proc_cell_servers_next(struct seq_file *p, void *v, * clean up after reading from the cells list */ static void afs_proc_cell_servers_stop(struct seq_file *p, void *v) + __releases(p->private->sv_lock) { struct afs_cell *cell = p->private; diff --git a/fs/afs/server.c b/fs/afs/server.c index 62b093aa41c..22afaae1a4c 100644 --- a/fs/afs/server.c +++ b/fs/afs/server.c @@ -123,8 +123,7 @@ int afs_server_lookup(struct afs_cell *cell, const struct in_addr *addr, resurrect_server: _debug("resurrecting server"); - list_del(&zombie->link); - list_add_tail(&zombie->link, &cell->sv_list); + list_move_tail(&zombie->link, &cell->sv_list); afs_get_server(zombie); afs_kafstimod_del_timer(&zombie->timeout); spin_unlock(&cell->sv_gylock); @@ -168,8 +167,7 @@ void afs_put_server(struct afs_server *server) } spin_lock(&cell->sv_gylock); - list_del(&server->link); - list_add_tail(&server->link, &cell->sv_graveyard); + list_move_tail(&server->link, &cell->sv_graveyard); /* time out in 10 secs */ afs_kafstimod_add_timer(&server->timeout, 10 * HZ); diff --git a/fs/afs/super.c b/fs/afs/super.c index 82468df0ba5..67d1f5c819e 100644 --- a/fs/afs/super.c +++ b/fs/afs/super.c @@ -48,7 +48,7 @@ static void afs_put_super(struct super_block *sb); static void afs_destroy_inode(struct inode *inode); -static struct file_system_type afs_fs_type = { +struct file_system_type afs_fs_type = { .owner = THIS_MODULE, .name = "afs", .get_sb = afs_get_sb, diff --git a/fs/afs/super.h b/fs/afs/super.h index ac11362f4e9..32de8cc6fae 100644 --- a/fs/afs/super.h +++ b/fs/afs/super.h @@ -38,6 +38,8 @@ static inline struct afs_super_info *AFS_FS_S(struct super_block *sb) return sb->s_fs_info; } +extern struct file_system_type afs_fs_type; + #endif /* __KERNEL__ */ #endif /* _LINUX_AFS_SUPER_H */ diff --git a/fs/afs/vlocation.c b/fs/afs/vlocation.c index eced20618ec..782ee7c600c 100644 --- a/fs/afs/vlocation.c +++ b/fs/afs/vlocation.c @@ -281,11 +281,10 @@ int afs_vlocation_lookup(struct afs_cell *cell, spin_unlock(&cell->vl_gylock); /* not in the cell's in-memory lists - create a new record */ - vlocation = kmalloc(sizeof(struct afs_vlocation), GFP_KERNEL); + vlocation = kzalloc(sizeof(struct afs_vlocation), GFP_KERNEL); if (!vlocation) return -ENOMEM; - memset(vlocation, 0, sizeof(struct afs_vlocation)); atomic_set(&vlocation->usage, 1); INIT_LIST_HEAD(&vlocation->link); rwlock_init(&vlocation->lock); @@ -326,8 +325,7 @@ int afs_vlocation_lookup(struct afs_cell *cell, /* found in the graveyard - resurrect */ _debug("found in graveyard"); atomic_inc(&vlocation->usage); - list_del(&vlocation->link); - list_add_tail(&vlocation->link, &cell->vl_list); + list_move_tail(&vlocation->link, &cell->vl_list); spin_unlock(&cell->vl_gylock); afs_kafstimod_del_timer(&vlocation->timeout); @@ -478,8 +476,7 @@ static void __afs_put_vlocation(struct afs_vlocation *vlocation) } /* move to graveyard queue */ - list_del(&vlocation->link); - list_add_tail(&vlocation->link,&cell->vl_graveyard); + list_move_tail(&vlocation->link,&cell->vl_graveyard); /* remove from pending timeout queue (refcounted if actually being * updated) */ diff --git a/fs/afs/vnode.c b/fs/afs/vnode.c index 9867fef3261..cf62da5d782 100644 --- a/fs/afs/vnode.c +++ b/fs/afs/vnode.c @@ -104,8 +104,7 @@ static void afs_vnode_finalise_status_update(struct afs_vnode *vnode, vnode->cb_expiry * HZ); spin_lock(&afs_cb_hash_lock); - list_del(&vnode->cb_hash_link); - list_add_tail(&vnode->cb_hash_link, + list_move_tail(&vnode->cb_hash_link, &afs_cb_hash(server, &vnode->fid)); spin_unlock(&afs_cb_hash_lock); diff --git a/fs/afs/volume.c b/fs/afs/volume.c index 0ff4b86476e..768c6dbd323 100644 --- a/fs/afs/volume.c +++ b/fs/afs/volume.c @@ -186,11 +186,10 @@ int afs_volume_lookup(const char *name, struct afs_cell *cell, int rwpath, _debug("creating new volume record"); ret = -ENOMEM; - volume = kmalloc(sizeof(struct afs_volume), GFP_KERNEL); + volume = kzalloc(sizeof(struct afs_volume), GFP_KERNEL); if (!volume) goto error_up; - memset(volume, 0, sizeof(struct afs_volume)); atomic_set(&volume->usage, 1); volume->type = type; volume->type_force = force; @@ -641,7 +641,7 @@ static inline int __queue_kicked_iocb(struct kiocb *iocb) * invoked both for initial i/o submission and * subsequent retries via the aio_kick_handler. * Expects to be invoked with iocb->ki_ctx->lock - * already held. The lock is released and reaquired + * already held. The lock is released and reacquired * as needed during processing. * * Calls the iocb retry method (already setup for the diff --git a/fs/autofs/autofs_i.h b/fs/autofs/autofs_i.h index a62327f1bdf..c7700d9b3f9 100644 --- a/fs/autofs/autofs_i.h +++ b/fs/autofs/autofs_i.h @@ -37,8 +37,6 @@ #define DPRINTK(D) ((void)0) #endif -#define AUTOFS_SUPER_MAGIC 0x0187 - /* * If the daemon returns a negative response (AUTOFS_IOC_FAIL) then the * kernel will keep the negative response cached for up to the time given diff --git a/fs/autofs/inode.c b/fs/autofs/inode.c index 65e5ed42190..2c9759baad6 100644 --- a/fs/autofs/inode.c +++ b/fs/autofs/inode.c @@ -16,6 +16,7 @@ #include <linux/file.h> #include <linux/parser.h> #include <linux/bitops.h> +#include <linux/magic.h> #include "autofs_i.h" #include <linux/module.h> @@ -128,10 +129,9 @@ int autofs_fill_super(struct super_block *s, void *data, int silent) struct autofs_sb_info *sbi; int minproto, maxproto; - sbi = kmalloc(sizeof(*sbi), GFP_KERNEL); + sbi = kzalloc(sizeof(*sbi), GFP_KERNEL); if ( !sbi ) goto fail_unlock; - memset(sbi, 0, sizeof(*sbi)); DPRINTK(("autofs: starting up, sbi = %p\n",sbi)); s->s_fs_info = sbi; @@ -216,7 +216,6 @@ static void autofs_read_inode(struct inode *inode) inode->i_nlink = 2; inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME; inode->i_blocks = 0; - inode->i_blksize = 1024; if ( ino == AUTOFS_ROOT_INO ) { inode->i_mode = S_IFDIR | S_IRUGO | S_IXUGO | S_IWUSR; @@ -241,7 +240,7 @@ static void autofs_read_inode(struct inode *inode) inode->i_op = &autofs_symlink_inode_operations; sl = &sbi->symlink[n]; - inode->u.generic_ip = sl; + inode->i_private = sl; inode->i_mode = S_IFLNK | S_IRWXUGO; inode->i_mtime.tv_sec = inode->i_ctime.tv_sec = sl->mtime; inode->i_mtime.tv_nsec = inode->i_ctime.tv_nsec = 0; diff --git a/fs/autofs/symlink.c b/fs/autofs/symlink.c index 52e8772b066..c74f2eb6577 100644 --- a/fs/autofs/symlink.c +++ b/fs/autofs/symlink.c @@ -15,7 +15,7 @@ /* Nothing to release.. */ static void *autofs_follow_link(struct dentry *dentry, struct nameidata *nd) { - char *s=((struct autofs_symlink *)dentry->d_inode->u.generic_ip)->data; + char *s=((struct autofs_symlink *)dentry->d_inode->i_private)->data; nd_set_link(nd, s); return NULL; } diff --git a/fs/autofs4/autofs_i.h b/fs/autofs4/autofs_i.h index d6603d02304..480ab178cba 100644 --- a/fs/autofs4/autofs_i.h +++ b/fs/autofs4/autofs_i.h @@ -40,8 +40,6 @@ #define DPRINTK(fmt,args...) do {} while(0) #endif -#define AUTOFS_SUPER_MAGIC 0x0187 - /* Unified info structure. This is pointed to by both the dentry and inode structures. Each file in the filesystem has an instance of this structure. It holds a reference to the dentry, so dentries are never diff --git a/fs/autofs4/expire.c b/fs/autofs4/expire.c index b8ce02607d6..d96e5c14a9c 100644 --- a/fs/autofs4/expire.c +++ b/fs/autofs4/expire.c @@ -32,7 +32,7 @@ static inline int autofs4_can_expire(struct dentry *dentry, if (!do_now) { /* Too young to die */ - if (time_after(ino->last_used + timeout, now)) + if (!timeout || time_after(ino->last_used + timeout, now)) return 0; /* update last_used here :- @@ -174,6 +174,12 @@ static int autofs4_tree_busy(struct vfsmount *mnt, struct autofs_info *ino = autofs4_dentry_ino(p); unsigned int ino_count = atomic_read(&ino->count); + /* + * Clean stale dentries below that have not been + * invalidated after a mount fail during lookup + */ + d_invalidate(p); + /* allow for dget above and top is already dgot */ if (p == top) ino_count += 2; @@ -247,7 +253,7 @@ static struct dentry *autofs4_expire_direct(struct super_block *sb, struct dentry *root = dget(sb->s_root); int do_now = how & AUTOFS_EXP_IMMEDIATE; - if (!sbi->exp_timeout || !root) + if (!root) return NULL; now = jiffies; @@ -287,7 +293,7 @@ static struct dentry *autofs4_expire_indirect(struct super_block *sb, int do_now = how & AUTOFS_EXP_IMMEDIATE; int exp_leaves = how & AUTOFS_EXP_LEAVES; - if ( !sbi->exp_timeout || !root ) + if (!root) return NULL; now = jiffies; @@ -370,8 +376,7 @@ next: DPRINTK("returning %p %.*s", expired, (int)expired->d_name.len, expired->d_name.name); spin_lock(&dcache_lock); - list_del(&expired->d_parent->d_subdirs); - list_add(&expired->d_parent->d_subdirs, &expired->d_u.d_child); + list_move(&expired->d_parent->d_subdirs, &expired->d_u.d_child); spin_unlock(&dcache_lock); return expired; } diff --git a/fs/autofs4/inode.c b/fs/autofs4/inode.c index fde78b110dd..800ce876cae 100644 --- a/fs/autofs4/inode.c +++ b/fs/autofs4/inode.c @@ -19,6 +19,7 @@ #include <linux/parser.h> #include <linux/bitops.h> #include <linux/smp_lock.h> +#include <linux/magic.h> #include "autofs_i.h" #include <linux/module.h> @@ -446,7 +447,6 @@ struct inode *autofs4_get_inode(struct super_block *sb, inode->i_uid = 0; inode->i_gid = 0; } - inode->i_blksize = PAGE_CACHE_SIZE; inode->i_blocks = 0; inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME; diff --git a/fs/autofs4/root.c b/fs/autofs4/root.c index 5100f984783..563ef9d7da9 100644 --- a/fs/autofs4/root.c +++ b/fs/autofs4/root.c @@ -137,7 +137,9 @@ static int autofs4_dir_open(struct inode *inode, struct file *file) nd.flags = LOOKUP_DIRECTORY; ret = (dentry->d_op->d_revalidate)(dentry, &nd); - if (!ret) { + if (ret <= 0) { + if (ret < 0) + status = ret; dcache_dir_close(inode, file); goto out; } @@ -279,9 +281,6 @@ static int try_to_fill_dentry(struct dentry *dentry, int flags) DPRINTK("mount done status=%d", status); - if (status && dentry->d_inode) - return status; /* Try to get the kernel to invalidate this dentry */ - /* Turn this into a real negative dentry? */ if (status == -ENOENT) { spin_lock(&dentry->d_lock); @@ -357,7 +356,7 @@ static void *autofs4_follow_link(struct dentry *dentry, struct nameidata *nd) * don't try to mount it again. */ spin_lock(&dcache_lock); - if (!d_mountpoint(dentry) && list_empty(&dentry->d_subdirs)) { + if (!d_mountpoint(dentry) && __simple_empty(dentry)) { spin_unlock(&dcache_lock); status = try_to_fill_dentry(dentry, 0); @@ -400,13 +399,23 @@ static int autofs4_revalidate(struct dentry *dentry, struct nameidata *nd) struct autofs_sb_info *sbi = autofs4_sbi(dir->i_sb); int oz_mode = autofs4_oz_mode(sbi); int flags = nd ? nd->flags : 0; - int status = 0; + int status = 1; /* Pending dentry */ if (autofs4_ispending(dentry)) { - if (!oz_mode) - status = try_to_fill_dentry(dentry, flags); - return !status; + /* The daemon never causes a mount to trigger */ + if (oz_mode) + return 1; + + /* + * A zero status is success otherwise we have a + * negative error code. + */ + status = try_to_fill_dentry(dentry, flags); + if (status == 0) + return 1; + + return status; } /* Negative dentry.. invalidate if "old" */ @@ -421,9 +430,19 @@ static int autofs4_revalidate(struct dentry *dentry, struct nameidata *nd) DPRINTK("dentry=%p %.*s, emptydir", dentry, dentry->d_name.len, dentry->d_name.name); spin_unlock(&dcache_lock); - if (!oz_mode) - status = try_to_fill_dentry(dentry, flags); - return !status; + /* The daemon never causes a mount to trigger */ + if (oz_mode) + return 1; + + /* + * A zero status is success otherwise we have a + * negative error code. + */ + status = try_to_fill_dentry(dentry, flags); + if (status == 0) + return 1; + + return status; } spin_unlock(&dcache_lock); @@ -518,6 +537,9 @@ static struct dentry *autofs4_lookup(struct inode *dir, struct dentry *dentry, s return ERR_PTR(-ERESTARTNOINTR); } } + spin_lock(&dentry->d_lock); + dentry->d_flags &= ~DCACHE_AUTOFS_PENDING; + spin_unlock(&dentry->d_lock); } /* diff --git a/fs/befs/linuxvfs.c b/fs/befs/linuxvfs.c index 08201fab26c..57020c7a7e6 100644 --- a/fs/befs/linuxvfs.c +++ b/fs/befs/linuxvfs.c @@ -73,7 +73,7 @@ static struct inode_operations befs_dir_inode_operations = { .lookup = befs_lookup, }; -static struct address_space_operations befs_aops = { +static const struct address_space_operations befs_aops = { .readpage = befs_readpage, .sync_page = block_sync_page, .bmap = befs_bmap, @@ -325,7 +325,7 @@ befs_read_inode(struct inode *inode) if (!bh) { befs_error(sb, "unable to read inode block - " "inode = %lu", inode->i_ino); - goto unaquire_none; + goto unacquire_none; } raw_inode = (befs_inode *) bh->b_data; @@ -334,7 +334,7 @@ befs_read_inode(struct inode *inode) if (befs_check_inode(sb, raw_inode, inode->i_ino) != BEFS_OK) { befs_error(sb, "Bad inode: %lu", inode->i_ino); - goto unaquire_bh; + goto unacquire_bh; } inode->i_mode = (umode_t) fs32_to_cpu(sb, raw_inode->mode); @@ -365,7 +365,6 @@ befs_read_inode(struct inode *inode) inode->i_mtime.tv_nsec = 0; /* lower 16 bits are not a time */ inode->i_ctime = inode->i_mtime; inode->i_atime = inode->i_mtime; - inode->i_blksize = befs_sb->block_size; befs_ino->i_inode_num = fsrun_to_cpu(sb, raw_inode->inode_num); befs_ino->i_parent = fsrun_to_cpu(sb, raw_inode->parent); @@ -402,17 +401,17 @@ befs_read_inode(struct inode *inode) befs_error(sb, "Inode %lu is not a regular file, " "directory or symlink. THAT IS WRONG! BeFS has no " "on disk special files", inode->i_ino); - goto unaquire_bh; + goto unacquire_bh; } brelse(bh); befs_debug(sb, "<--- befs_read_inode()"); return; - unaquire_bh: + unacquire_bh: brelse(bh); - unaquire_none: + unacquire_none: make_bad_inode(inode); befs_debug(sb, "<--- befs_read_inode() - Bad inode"); return; @@ -446,9 +445,7 @@ befs_init_inodecache(void) static void befs_destroy_inodecache(void) { - if (kmem_cache_destroy(befs_inode_cachep)) - printk(KERN_ERR "befs_destroy_inodecache: " - "not all structures were freed\n"); + kmem_cache_destroy(befs_inode_cachep); } /* @@ -512,7 +509,11 @@ befs_utf2nls(struct super_block *sb, const char *in, wchar_t uni; int unilen, utflen; char *result; - int maxlen = in_len; /* The utf8->nls conversion can't make more chars */ + /* The utf8->nls conversion won't make the final nls string bigger + * than the utf one, but if the string is pure ascii they'll have the + * same width and an extra char is needed to save the additional \0 + */ + int maxlen = in_len + 1; befs_debug(sb, "---> utf2nls()"); @@ -588,7 +589,10 @@ befs_nls2utf(struct super_block *sb, const char *in, wchar_t uni; int unilen, utflen; char *result; - int maxlen = 3 * in_len; + /* There're nls characters that will translate to 3-chars-wide UTF-8 + * characters, a additional byte is needed to save the final \0 + * in special cases */ + int maxlen = (3 * in_len) + 1; befs_debug(sb, "---> nls2utf()\n"); @@ -761,14 +765,14 @@ befs_fill_super(struct super_block *sb, void *data, int silent) printk(KERN_ERR "BeFS(%s): Unable to allocate memory for private " "portion of superblock. Bailing.\n", sb->s_id); - goto unaquire_none; + goto unacquire_none; } befs_sb = BEFS_SB(sb); memset(befs_sb, 0, sizeof(befs_sb_info)); if (!parse_options((char *) data, &befs_sb->mount_opts)) { befs_error(sb, "cannot parse mount options"); - goto unaquire_priv_sbp; + goto unacquire_priv_sbp; } befs_debug(sb, "---> befs_fill_super()"); @@ -794,7 +798,7 @@ befs_fill_super(struct super_block *sb, void *data, int silent) if (!(bh = sb_bread(sb, sb_block))) { befs_error(sb, "unable to read superblock"); - goto unaquire_priv_sbp; + goto unacquire_priv_sbp; } /* account for offset of super block on x86 */ @@ -809,20 +813,20 @@ befs_fill_super(struct super_block *sb, void *data, int silent) } if (befs_load_sb(sb, disk_sb) != BEFS_OK) - goto unaquire_bh; + goto unacquire_bh; befs_dump_super_block(sb, disk_sb); brelse(bh); if (befs_check_sb(sb) != BEFS_OK) - goto unaquire_priv_sbp; + goto unacquire_priv_sbp; if( befs_sb->num_blocks > ~((sector_t)0) ) { befs_error(sb, "blocks count: %Lu " "is larger than the host can use", befs_sb->num_blocks); - goto unaquire_priv_sbp; + goto unacquire_priv_sbp; } /* @@ -838,7 +842,7 @@ befs_fill_super(struct super_block *sb, void *data, int silent) if (!sb->s_root) { iput(root); befs_error(sb, "get root inode failed"); - goto unaquire_priv_sbp; + goto unacquire_priv_sbp; } /* load nls library */ @@ -860,13 +864,13 @@ befs_fill_super(struct super_block *sb, void *data, int silent) return 0; /*****************/ - unaquire_bh: + unacquire_bh: brelse(bh); - unaquire_priv_sbp: + unacquire_priv_sbp: kfree(sb->s_fs_info); - unaquire_none: + unacquire_none: sb->s_fs_info = NULL; return -EINVAL; } @@ -925,18 +929,18 @@ init_befs_fs(void) err = befs_init_inodecache(); if (err) - goto unaquire_none; + goto unacquire_none; err = register_filesystem(&befs_fs_type); if (err) - goto unaquire_inodecache; + goto unacquire_inodecache; return 0; -unaquire_inodecache: +unacquire_inodecache: befs_destroy_inodecache(); -unaquire_none: +unacquire_none: return err; } diff --git a/fs/bfs/bfs.h b/fs/bfs/bfs.h index 9d791004b21..31973bbbf05 100644 --- a/fs/bfs/bfs.h +++ b/fs/bfs/bfs.h @@ -50,7 +50,7 @@ static inline struct bfs_inode_info *BFS_I(struct inode *inode) /* file.c */ extern struct inode_operations bfs_file_inops; extern const struct file_operations bfs_file_operations; -extern struct address_space_operations bfs_aops; +extern const struct address_space_operations bfs_aops; /* dir.c */ extern struct inode_operations bfs_dir_inops; diff --git a/fs/bfs/dir.c b/fs/bfs/dir.c index 26fad962173..dcf04cb1328 100644 --- a/fs/bfs/dir.c +++ b/fs/bfs/dir.c @@ -102,7 +102,7 @@ static int bfs_create(struct inode * dir, struct dentry * dentry, int mode, inode->i_uid = current->fsuid; inode->i_gid = (dir->i_mode & S_ISGID) ? dir->i_gid : current->fsgid; inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME_SEC; - inode->i_blocks = inode->i_blksize = 0; + inode->i_blocks = 0; inode->i_op = &bfs_file_inops; inode->i_fop = &bfs_file_operations; inode->i_mapping->a_ops = &bfs_aops; diff --git a/fs/bfs/file.c b/fs/bfs/file.c index d83cd74a2e4..3d5aca28a0a 100644 --- a/fs/bfs/file.c +++ b/fs/bfs/file.c @@ -153,7 +153,7 @@ static sector_t bfs_bmap(struct address_space *mapping, sector_t block) return generic_block_bmap(mapping, block, bfs_get_block); } -struct address_space_operations bfs_aops = { +const struct address_space_operations bfs_aops = { .readpage = bfs_readpage, .writepage = bfs_writepage, .sync_page = block_sync_page, diff --git a/fs/bfs/inode.c b/fs/bfs/inode.c index cf74f3d4d96..ed27ffb3459 100644 --- a/fs/bfs/inode.c +++ b/fs/bfs/inode.c @@ -76,7 +76,6 @@ static void bfs_read_inode(struct inode * inode) inode->i_size = BFS_FILESIZE(di); inode->i_blocks = BFS_FILEBLOCKS(di); if (inode->i_size || inode->i_blocks) dprintf("Registered inode with %lld size, %ld blocks\n", inode->i_size, inode->i_blocks); - inode->i_blksize = PAGE_SIZE; inode->i_atime.tv_sec = le32_to_cpu(di->i_atime); inode->i_mtime.tv_sec = le32_to_cpu(di->i_mtime); inode->i_ctime.tv_sec = le32_to_cpu(di->i_ctime); @@ -268,8 +267,7 @@ static int init_inodecache(void) static void destroy_inodecache(void) { - if (kmem_cache_destroy(bfs_inode_cachep)) - printk(KERN_INFO "bfs_inode_cache: not all structures were freed\n"); + kmem_cache_destroy(bfs_inode_cachep); } static struct super_operations bfs_sops = { @@ -311,11 +309,10 @@ static int bfs_fill_super(struct super_block *s, void *data, int silent) unsigned i, imap_len; struct bfs_sb_info * info; - info = kmalloc(sizeof(*info), GFP_KERNEL); + info = kzalloc(sizeof(*info), GFP_KERNEL); if (!info) return -ENOMEM; s->s_fs_info = info; - memset(info, 0, sizeof(*info)); sb_set_blocksize(s, BFS_BSIZE); @@ -338,10 +335,9 @@ static int bfs_fill_super(struct super_block *s, void *data, int silent) + BFS_ROOT_INO - 1; imap_len = info->si_lasti/8 + 1; - info->si_imap = kmalloc(imap_len, GFP_KERNEL); + info->si_imap = kzalloc(imap_len, GFP_KERNEL); if (!info->si_imap) goto out; - memset(info->si_imap, 0, imap_len); for (i=0; i<BFS_ROOT_INO; i++) set_bit(i, info->si_imap); diff --git a/fs/binfmt_aout.c b/fs/binfmt_aout.c index f312103434d..517e111bb7e 100644 --- a/fs/binfmt_aout.c +++ b/fs/binfmt_aout.c @@ -278,6 +278,13 @@ static int load_aout_binary(struct linux_binprm * bprm, struct pt_regs * regs) return -ENOEXEC; } + /* + * Requires a mmap handler. This prevents people from using a.out + * as part of an exploit attack against /proc-related vulnerabilities. + */ + if (!bprm->file->f_op || !bprm->file->f_op->mmap) + return -ENOEXEC; + fd_offset = N_TXTOFF(ex); /* Check initial limits. This avoids letting people circumvent @@ -476,6 +483,13 @@ static int load_aout_library(struct file *file) goto out; } + /* + * Requires a mmap handler. This prevents people from using a.out + * as part of an exploit attack against /proc-related vulnerabilities. + */ + if (!file->f_op || !file->f_op->mmap) + goto out; + if (N_FLAGS(ex)) goto out; diff --git a/fs/binfmt_elf.c b/fs/binfmt_elf.c index d0434406eae..6eb48e1446e 100644 --- a/fs/binfmt_elf.c +++ b/fs/binfmt_elf.c @@ -84,7 +84,7 @@ static struct linux_binfmt elf_format = { .min_coredump = ELF_EXEC_PAGESIZE }; -#define BAD_ADDR(x) ((unsigned long)(x) > TASK_SIZE) +#define BAD_ADDR(x) ((unsigned long)(x) >= TASK_SIZE) static int set_brk(unsigned long start, unsigned long end) { @@ -394,7 +394,7 @@ static unsigned long load_elf_interp(struct elfhdr *interp_elf_ex, * <= p_memsize so it's only necessary to check p_memsz. */ k = load_addr + eppnt->p_vaddr; - if (k > TASK_SIZE || + if (BAD_ADDR(k) || eppnt->p_filesz > eppnt->p_memsz || eppnt->p_memsz > TASK_SIZE || TASK_SIZE - eppnt->p_memsz < k) { @@ -515,7 +515,8 @@ static unsigned long randomize_stack_top(unsigned long stack_top) { unsigned int random_variable = 0; - if (current->flags & PF_RANDOMIZE) { + if ((current->flags & PF_RANDOMIZE) && + !(current->personality & ADDR_NO_RANDOMIZE)) { random_variable = get_random_int() & STACK_RND_MASK; random_variable <<= PAGE_SHIFT; } @@ -887,7 +888,7 @@ static int load_elf_binary(struct linux_binprm *bprm, struct pt_regs *regs) * allowed task size. Note that p_filesz must always be * <= p_memsz so it is only necessary to check p_memsz. */ - if (k > TASK_SIZE || elf_ppnt->p_filesz > elf_ppnt->p_memsz || + if (BAD_ADDR(k) || elf_ppnt->p_filesz > elf_ppnt->p_memsz || elf_ppnt->p_memsz > TASK_SIZE || TASK_SIZE - elf_ppnt->p_memsz < k) { /* set_brk can never work. Avoid overflows. */ @@ -941,10 +942,9 @@ static int load_elf_binary(struct linux_binprm *bprm, struct pt_regs *regs) interpreter, &interp_load_addr); if (BAD_ADDR(elf_entry)) { - printk(KERN_ERR "Unable to load interpreter %.128s\n", - elf_interpreter); force_sig(SIGSEGV, current); - retval = -ENOEXEC; /* Nobody gets to see this, but.. */ + retval = IS_ERR((void *)elf_entry) ? + (int)elf_entry : -EINVAL; goto out_free_dentry; } reloc_func_desc = interp_load_addr; @@ -955,8 +955,8 @@ static int load_elf_binary(struct linux_binprm *bprm, struct pt_regs *regs) } else { elf_entry = loc->elf_ex.e_entry; if (BAD_ADDR(elf_entry)) { - send_sig(SIGSEGV, current, 0); - retval = -ENOEXEC; /* Nobody gets to see this, but.. */ + force_sig(SIGSEGV, current); + retval = -EINVAL; goto out_free_dentry; } } @@ -1038,10 +1038,8 @@ out_free_interp: out_free_file: sys_close(elf_exec_fileno); out_free_fh: - if (files) { - put_files_struct(current->files); - current->files = files; - } + if (files) + reset_files_struct(current, files); out_free_ph: kfree(elf_phdata); goto out; @@ -1186,8 +1184,6 @@ static int maydump(struct vm_area_struct *vma) return 1; } -#define roundup(x, y) ((((x) + ((y) - 1)) / (y)) * (y)) - /* An ELF note in memory */ struct memelfnote { @@ -1265,7 +1261,7 @@ static void fill_elf_header(struct elfhdr *elf, int segs) return; } -static void fill_elf_note_phdr(struct elf_phdr *phdr, int sz, off_t offset) +static void fill_elf_note_phdr(struct elf_phdr *phdr, int sz, loff_t offset) { phdr->p_type = PT_NOTE; phdr->p_offset = offset; @@ -1431,7 +1427,7 @@ static int elf_core_dump(long signr, struct pt_regs *regs, struct file *file) int i; struct vm_area_struct *vma; struct elfhdr *elf = NULL; - off_t offset = 0, dataoff; + loff_t offset = 0, dataoff; unsigned long limit = current->signal->rlim[RLIMIT_CORE].rlim_cur; int numnote; struct memelfnote *notes = NULL; @@ -1483,20 +1479,19 @@ static int elf_core_dump(long signr, struct pt_regs *regs, struct file *file) if (signr) { struct elf_thread_status *tmp; - read_lock(&tasklist_lock); + rcu_read_lock(); do_each_thread(g,p) if (current->mm == p->mm && current != p) { tmp = kzalloc(sizeof(*tmp), GFP_ATOMIC); if (!tmp) { - read_unlock(&tasklist_lock); + rcu_read_unlock(); goto cleanup; } - INIT_LIST_HEAD(&tmp->list); tmp->thread = p; list_add(&tmp->list, &thread_list); } while_each_thread(g,p); - read_unlock(&tasklist_lock); + rcu_read_unlock(); list_for_each(t, &thread_list) { struct elf_thread_status *tmp; int sz; @@ -1664,11 +1659,11 @@ static int elf_core_dump(long signr, struct pt_regs *regs, struct file *file) ELF_CORE_WRITE_EXTRA_DATA; #endif - if ((off_t)file->f_pos != offset) { + if (file->f_pos != offset) { /* Sanity check */ printk(KERN_WARNING - "elf_core_dump: file->f_pos (%ld) != offset (%ld)\n", - (off_t)file->f_pos, offset); + "elf_core_dump: file->f_pos (%Ld) != offset (%Ld)\n", + file->f_pos, offset); } end_coredump: diff --git a/fs/binfmt_elf_fdpic.c b/fs/binfmt_elf_fdpic.c index eba4e23b9ca..f86d5c9ce5e 100644 --- a/fs/binfmt_elf_fdpic.c +++ b/fs/binfmt_elf_fdpic.c @@ -1,6 +1,6 @@ /* binfmt_elf_fdpic.c: FDPIC ELF binary format * - * Copyright (C) 2003, 2004 Red Hat, Inc. All Rights Reserved. + * Copyright (C) 2003, 2004, 2006 Red Hat, Inc. All Rights Reserved. * Written by David Howells (dhowells@redhat.com) * Derived from binfmt_elf.c * @@ -24,7 +24,9 @@ #include <linux/file.h> #include <linux/fcntl.h> #include <linux/slab.h> +#include <linux/pagemap.h> #include <linux/highmem.h> +#include <linux/highuid.h> #include <linux/personality.h> #include <linux/ptrace.h> #include <linux/init.h> @@ -48,45 +50,59 @@ typedef char *elf_caddr_t; #define kdebug(fmt, ...) do {} while(0) #endif +#if 0 +#define kdcore(fmt, ...) printk("FDPIC "fmt"\n" ,##__VA_ARGS__ ) +#else +#define kdcore(fmt, ...) do {} while(0) +#endif + MODULE_LICENSE("GPL"); -static int load_elf_fdpic_binary(struct linux_binprm *bprm, struct pt_regs *regs); -//static int load_elf_fdpic_library(struct file *); -static int elf_fdpic_fetch_phdrs(struct elf_fdpic_params *params, struct file *file); -static int elf_fdpic_map_file(struct elf_fdpic_params *params, - struct file *file, - struct mm_struct *mm, - const char *what); +static int load_elf_fdpic_binary(struct linux_binprm *, struct pt_regs *); +static int elf_fdpic_fetch_phdrs(struct elf_fdpic_params *, struct file *); +static int elf_fdpic_map_file(struct elf_fdpic_params *, struct file *, + struct mm_struct *, const char *); -static int create_elf_fdpic_tables(struct linux_binprm *bprm, - struct mm_struct *mm, - struct elf_fdpic_params *exec_params, - struct elf_fdpic_params *interp_params); +static int create_elf_fdpic_tables(struct linux_binprm *, struct mm_struct *, + struct elf_fdpic_params *, + struct elf_fdpic_params *); #ifndef CONFIG_MMU -static int elf_fdpic_transfer_args_to_stack(struct linux_binprm *bprm, unsigned long *_sp); -static int elf_fdpic_map_file_constdisp_on_uclinux(struct elf_fdpic_params *params, - struct file *file, - struct mm_struct *mm); +static int elf_fdpic_transfer_args_to_stack(struct linux_binprm *, + unsigned long *); +static int elf_fdpic_map_file_constdisp_on_uclinux(struct elf_fdpic_params *, + struct file *, + struct mm_struct *); #endif -static int elf_fdpic_map_file_by_direct_mmap(struct elf_fdpic_params *params, - struct file *file, - struct mm_struct *mm); +static int elf_fdpic_map_file_by_direct_mmap(struct elf_fdpic_params *, + struct file *, struct mm_struct *); + +#if defined(USE_ELF_CORE_DUMP) && defined(CONFIG_ELF_CORE) +static int elf_fdpic_core_dump(long, struct pt_regs *, struct file *); +#endif static struct linux_binfmt elf_fdpic_format = { .module = THIS_MODULE, .load_binary = load_elf_fdpic_binary, -// .load_shlib = load_elf_fdpic_library, -// .core_dump = elf_fdpic_core_dump, +#if defined(USE_ELF_CORE_DUMP) && defined(CONFIG_ELF_CORE) + .core_dump = elf_fdpic_core_dump, +#endif .min_coredump = ELF_EXEC_PAGESIZE, }; -static int __init init_elf_fdpic_binfmt(void) { return register_binfmt(&elf_fdpic_format); } -static void __exit exit_elf_fdpic_binfmt(void) { unregister_binfmt(&elf_fdpic_format); } +static int __init init_elf_fdpic_binfmt(void) +{ + return register_binfmt(&elf_fdpic_format); +} + +static void __exit exit_elf_fdpic_binfmt(void) +{ + unregister_binfmt(&elf_fdpic_format); +} -module_init(init_elf_fdpic_binfmt) -module_exit(exit_elf_fdpic_binfmt) +core_initcall(init_elf_fdpic_binfmt); +module_exit(exit_elf_fdpic_binfmt); static int is_elf_fdpic(struct elfhdr *hdr, struct file *file) { @@ -105,7 +121,8 @@ static int is_elf_fdpic(struct elfhdr *hdr, struct file *file) /* * read the program headers table into memory */ -static int elf_fdpic_fetch_phdrs(struct elf_fdpic_params *params, struct file *file) +static int elf_fdpic_fetch_phdrs(struct elf_fdpic_params *params, + struct file *file) { struct elf32_phdr *phdr; unsigned long size; @@ -121,7 +138,8 @@ static int elf_fdpic_fetch_phdrs(struct elf_fdpic_params *params, struct file *f if (!params->phdrs) return -ENOMEM; - retval = kernel_read(file, params->hdr.e_phoff, (char *) params->phdrs, size); + retval = kernel_read(file, params->hdr.e_phoff, + (char *) params->phdrs, size); if (retval < 0) return retval; @@ -141,17 +159,24 @@ static int elf_fdpic_fetch_phdrs(struct elf_fdpic_params *params, struct file *f } return 0; -} /* end elf_fdpic_fetch_phdrs() */ +} /*****************************************************************************/ /* * load an fdpic binary into various bits of memory */ -static int load_elf_fdpic_binary(struct linux_binprm *bprm, struct pt_regs *regs) +static int load_elf_fdpic_binary(struct linux_binprm *bprm, + struct pt_regs *regs) { struct elf_fdpic_params exec_params, interp_params; struct elf_phdr *phdr; - unsigned long stack_size; + unsigned long stack_size, entryaddr; +#ifndef CONFIG_MMU + unsigned long fullsize; +#endif +#ifdef ELF_FDPIC_PLAT_INIT + unsigned long dynaddr; +#endif struct file *interpreter = NULL; /* to shut gcc up */ char *interpreter_name = NULL; int executable_stack; @@ -212,7 +237,8 @@ static int load_elf_fdpic_binary(struct linux_binprm *bprm, struct pt_regs *regs goto error; } - retval = kernel_read(interpreter, 0, bprm->buf, BINPRM_BUF_SIZE); + retval = kernel_read(interpreter, 0, bprm->buf, + BINPRM_BUF_SIZE); if (retval < 0) goto error; @@ -295,7 +321,8 @@ static int load_elf_fdpic_binary(struct linux_binprm *bprm, struct pt_regs *regs ¤t->mm->start_stack, ¤t->mm->start_brk); - retval = setup_arg_pages(bprm, current->mm->start_stack, executable_stack); + retval = setup_arg_pages(bprm, current->mm->start_stack, + executable_stack); if (retval < 0) { send_sig(SIGKILL, current, 0); goto error_kill; @@ -303,7 +330,8 @@ static int load_elf_fdpic_binary(struct linux_binprm *bprm, struct pt_regs *regs #endif /* load the executable and interpreter into memory */ - retval = elf_fdpic_map_file(&exec_params, bprm->file, current->mm, "executable"); + retval = elf_fdpic_map_file(&exec_params, bprm->file, current->mm, + "executable"); if (retval < 0) goto error_kill; @@ -324,7 +352,8 @@ static int load_elf_fdpic_binary(struct linux_binprm *bprm, struct pt_regs *regs if (!current->mm->start_brk) current->mm->start_brk = current->mm->end_data; - current->mm->brk = current->mm->start_brk = PAGE_ALIGN(current->mm->start_brk); + current->mm->brk = current->mm->start_brk = + PAGE_ALIGN(current->mm->start_brk); #else /* create a stack and brk area big enough for everyone @@ -336,47 +365,45 @@ static int load_elf_fdpic_binary(struct linux_binprm *bprm, struct pt_regs *regs stack_size = PAGE_SIZE * 2; down_write(¤t->mm->mmap_sem); - current->mm->start_brk = do_mmap(NULL, - 0, - stack_size, + current->mm->start_brk = do_mmap(NULL, 0, stack_size, PROT_READ | PROT_WRITE | PROT_EXEC, MAP_PRIVATE | MAP_ANON | MAP_GROWSDOWN, 0); - if (IS_ERR((void *) current->mm->start_brk)) { + if (IS_ERR_VALUE(current->mm->start_brk)) { up_write(¤t->mm->mmap_sem); retval = current->mm->start_brk; current->mm->start_brk = 0; goto error_kill; } - if (do_mremap(current->mm->start_brk, - stack_size, - ksize((char *) current->mm->start_brk), - 0, 0 - ) == current->mm->start_brk - ) - stack_size = ksize((char *) current->mm->start_brk); + /* expand the stack mapping to use up the entire allocation granule */ + fullsize = ksize((char *) current->mm->start_brk); + if (!IS_ERR_VALUE(do_mremap(current->mm->start_brk, stack_size, + fullsize, 0, 0))) + stack_size = fullsize; up_write(¤t->mm->mmap_sem); current->mm->brk = current->mm->start_brk; current->mm->context.end_brk = current->mm->start_brk; - current->mm->context.end_brk += (stack_size > PAGE_SIZE) ? (stack_size - PAGE_SIZE) : 0; + current->mm->context.end_brk += + (stack_size > PAGE_SIZE) ? (stack_size - PAGE_SIZE) : 0; current->mm->start_stack = current->mm->start_brk + stack_size; #endif compute_creds(bprm); current->flags &= ~PF_FORKNOEXEC; - if (create_elf_fdpic_tables(bprm, current->mm, &exec_params, &interp_params) < 0) + if (create_elf_fdpic_tables(bprm, current->mm, + &exec_params, &interp_params) < 0) goto error_kill; - kdebug("- start_code %lx", (long) current->mm->start_code); - kdebug("- end_code %lx", (long) current->mm->end_code); - kdebug("- start_data %lx", (long) current->mm->start_data); - kdebug("- end_data %lx", (long) current->mm->end_data); - kdebug("- start_brk %lx", (long) current->mm->start_brk); - kdebug("- brk %lx", (long) current->mm->brk); - kdebug("- start_stack %lx", (long) current->mm->start_stack); + kdebug("- start_code %lx", current->mm->start_code); + kdebug("- end_code %lx", current->mm->end_code); + kdebug("- start_data %lx", current->mm->start_data); + kdebug("- end_data %lx", current->mm->end_data); + kdebug("- start_brk %lx", current->mm->start_brk); + kdebug("- brk %lx", current->mm->brk); + kdebug("- start_stack %lx", current->mm->start_stack); #ifdef ELF_FDPIC_PLAT_INIT /* @@ -385,21 +412,18 @@ static int load_elf_fdpic_binary(struct linux_binprm *bprm, struct pt_regs *regs * example. This macro performs whatever initialization to * the regs structure is required. */ - ELF_FDPIC_PLAT_INIT(regs, - exec_params.map_addr, - interp_params.map_addr, - interp_params.dynamic_addr ?: exec_params.dynamic_addr - ); + dynaddr = interp_params.dynamic_addr ?: exec_params.dynamic_addr; + ELF_FDPIC_PLAT_INIT(regs, exec_params.map_addr, interp_params.map_addr, + dynaddr); #endif /* everything is now ready... get the userspace context ready to roll */ - start_thread(regs, - interp_params.entry_addr ?: exec_params.entry_addr, - current->mm->start_stack); + entryaddr = interp_params.entry_addr ?: exec_params.entry_addr; + start_thread(regs, entryaddr, current->mm->start_stack); if (unlikely(current->ptrace & PT_PTRACED)) { if (current->ptrace & PT_TRACE_EXEC) - ptrace_notify ((PTRACE_EVENT_EXEC << 8) | SIGTRAP); + ptrace_notify((PTRACE_EVENT_EXEC << 8) | SIGTRAP); else send_sig(SIGTRAP, current, 0); } @@ -419,11 +443,11 @@ error: return retval; /* unrecoverable error - kill the process */ - error_kill: +error_kill: send_sig(SIGSEGV, current, 0); goto error; -} /* end load_elf_fdpic_binary() */ +} /*****************************************************************************/ /* @@ -459,6 +483,7 @@ static int create_elf_fdpic_tables(struct linux_binprm *bprm, */ hwcap = ELF_HWCAP; k_platform = ELF_PLATFORM; + u_platform = NULL; if (k_platform) { platform_len = strlen(k_platform) + 1; @@ -470,11 +495,11 @@ static int create_elf_fdpic_tables(struct linux_binprm *bprm, #if defined(__i386__) && defined(CONFIG_SMP) /* in some cases (e.g. Hyper-Threading), we want to avoid L1 evictions - * by the processes running on the same package. One thing we can do - * is to shuffle the initial stack for them. + * by the processes running on the same package. One thing we can do is + * to shuffle the initial stack for them. * - * the conditionals here are unneeded, but kept in to make the - * code behaviour the same as pre change unless we have hyperthreaded + * the conditionals here are unneeded, but kept in to make the code + * behaviour the same as pre change unless we have hyperthreaded * processors. This keeps Mr Marcelo Person happier but should be * removed for 2.5 */ @@ -497,11 +522,13 @@ static int create_elf_fdpic_tables(struct linux_binprm *bprm, if (interp_params->loadmap) { len = sizeof(struct elf32_fdpic_loadmap); - len += sizeof(struct elf32_fdpic_loadseg) * interp_params->loadmap->nsegs; + len += sizeof(struct elf32_fdpic_loadseg) * + interp_params->loadmap->nsegs; sp = (sp - len) & ~7UL; interp_params->map_addr = sp; - if (copy_to_user((void __user *) sp, interp_params->loadmap, len) != 0) + if (copy_to_user((void __user *) sp, interp_params->loadmap, + len) != 0) return -EFAULT; current->mm->context.interp_fdpic_loadmap = (unsigned long) sp; @@ -525,34 +552,37 @@ static int create_elf_fdpic_tables(struct linux_binprm *bprm, sp -= sp & 15UL; /* put the ELF interpreter info on the stack */ -#define NEW_AUX_ENT(nr, id, val) \ - do { \ - struct { unsigned long _id, _val; } __user *ent = (void __user *) csp; \ - __put_user((id), &ent[nr]._id); \ - __put_user((val), &ent[nr]._val); \ +#define NEW_AUX_ENT(nr, id, val) \ + do { \ + struct { unsigned long _id, _val; } __user *ent; \ + \ + ent = (void __user *) csp; \ + __put_user((id), &ent[nr]._id); \ + __put_user((val), &ent[nr]._val); \ } while (0) csp -= 2 * sizeof(unsigned long); NEW_AUX_ENT(0, AT_NULL, 0); if (k_platform) { csp -= 2 * sizeof(unsigned long); - NEW_AUX_ENT(0, AT_PLATFORM, (elf_addr_t)(unsigned long) u_platform); + NEW_AUX_ENT(0, AT_PLATFORM, + (elf_addr_t) (unsigned long) u_platform); } csp -= DLINFO_ITEMS * 2 * sizeof(unsigned long); - NEW_AUX_ENT( 0, AT_HWCAP, hwcap); - NEW_AUX_ENT( 1, AT_PAGESZ, PAGE_SIZE); - NEW_AUX_ENT( 2, AT_CLKTCK, CLOCKS_PER_SEC); - NEW_AUX_ENT( 3, AT_PHDR, exec_params->ph_addr); - NEW_AUX_ENT( 4, AT_PHENT, sizeof(struct elf_phdr)); - NEW_AUX_ENT( 5, AT_PHNUM, exec_params->hdr.e_phnum); - NEW_AUX_ENT( 6, AT_BASE, interp_params->elfhdr_addr); - NEW_AUX_ENT( 7, AT_FLAGS, 0); - NEW_AUX_ENT( 8, AT_ENTRY, exec_params->entry_addr); - NEW_AUX_ENT( 9, AT_UID, (elf_addr_t) current->uid); - NEW_AUX_ENT(10, AT_EUID, (elf_addr_t) current->euid); - NEW_AUX_ENT(11, AT_GID, (elf_addr_t) current->gid); - NEW_AUX_ENT(12, AT_EGID, (elf_addr_t) current->egid); + NEW_AUX_ENT( 0, AT_HWCAP, hwcap); + NEW_AUX_ENT( 1, AT_PAGESZ, PAGE_SIZE); + NEW_AUX_ENT( 2, AT_CLKTCK, CLOCKS_PER_SEC); + NEW_AUX_ENT( 3, AT_PHDR, exec_params->ph_addr); + NEW_AUX_ENT( 4, AT_PHENT, sizeof(struct elf_phdr)); + NEW_AUX_ENT( 5, AT_PHNUM, exec_params->hdr.e_phnum); + NEW_AUX_ENT( 6, AT_BASE, interp_params->elfhdr_addr); + NEW_AUX_ENT( 7, AT_FLAGS, 0); + NEW_AUX_ENT( 8, AT_ENTRY, exec_params->entry_addr); + NEW_AUX_ENT( 9, AT_UID, (elf_addr_t) current->uid); + NEW_AUX_ENT(10, AT_EUID, (elf_addr_t) current->euid); + NEW_AUX_ENT(11, AT_GID, (elf_addr_t) current->gid); + NEW_AUX_ENT(12, AT_EGID, (elf_addr_t) current->egid); #ifdef ARCH_DLINFO /* ARCH_DLINFO must come last so platform specific code can enforce @@ -578,7 +608,8 @@ static int create_elf_fdpic_tables(struct linux_binprm *bprm, #ifdef CONFIG_MMU current->mm->arg_start = bprm->p; #else - current->mm->arg_start = current->mm->start_stack - (MAX_ARG_PAGES * PAGE_SIZE - bprm->p); + current->mm->arg_start = current->mm->start_stack - + (MAX_ARG_PAGES * PAGE_SIZE - bprm->p); #endif p = (char __user *) current->mm->arg_start; @@ -606,7 +637,7 @@ static int create_elf_fdpic_tables(struct linux_binprm *bprm, mm->start_stack = (unsigned long) sp; return 0; -} /* end create_elf_fdpic_tables() */ +} /*****************************************************************************/ /* @@ -614,7 +645,8 @@ static int create_elf_fdpic_tables(struct linux_binprm *bprm, * the stack */ #ifndef CONFIG_MMU -static int elf_fdpic_transfer_args_to_stack(struct linux_binprm *bprm, unsigned long *_sp) +static int elf_fdpic_transfer_args_to_stack(struct linux_binprm *bprm, + unsigned long *_sp) { unsigned long index, stop, sp; char *src; @@ -635,9 +667,9 @@ static int elf_fdpic_transfer_args_to_stack(struct linux_binprm *bprm, unsigned *_sp = (*_sp - (MAX_ARG_PAGES * PAGE_SIZE - bprm->p)) & ~15; - out: +out: return ret; -} /* end elf_fdpic_transfer_args_to_stack() */ +} #endif /*****************************************************************************/ @@ -712,17 +744,18 @@ static int elf_fdpic_map_file(struct elf_fdpic_params *params, seg = loadmap->segs; for (loop = loadmap->nsegs; loop > 0; loop--, seg++) { if (params->hdr.e_entry >= seg->p_vaddr && - params->hdr.e_entry < seg->p_vaddr + seg->p_memsz - ) { + params->hdr.e_entry < seg->p_vaddr + seg->p_memsz) { params->entry_addr = - (params->hdr.e_entry - seg->p_vaddr) + seg->addr; + (params->hdr.e_entry - seg->p_vaddr) + + seg->addr; break; } } } /* determine where the program header table has wound up if mapped */ - stop = params->hdr.e_phoff + params->hdr.e_phnum * sizeof (struct elf_phdr); + stop = params->hdr.e_phoff; + stop += params->hdr.e_phnum * sizeof (struct elf_phdr); phdr = params->phdrs; for (loop = 0; loop < params->hdr.e_phnum; loop++, phdr++) { @@ -736,9 +769,11 @@ static int elf_fdpic_map_file(struct elf_fdpic_params *params, seg = loadmap->segs; for (loop = loadmap->nsegs; loop > 0; loop--, seg++) { if (phdr->p_vaddr >= seg->p_vaddr && - phdr->p_vaddr + phdr->p_filesz <= seg->p_vaddr + seg->p_memsz - ) { - params->ph_addr = (phdr->p_vaddr - seg->p_vaddr) + seg->addr + + phdr->p_vaddr + phdr->p_filesz <= + seg->p_vaddr + seg->p_memsz) { + params->ph_addr = + (phdr->p_vaddr - seg->p_vaddr) + + seg->addr + params->hdr.e_phoff - phdr->p_offset; break; } @@ -755,18 +790,22 @@ static int elf_fdpic_map_file(struct elf_fdpic_params *params, seg = loadmap->segs; for (loop = loadmap->nsegs; loop > 0; loop--, seg++) { if (phdr->p_vaddr >= seg->p_vaddr && - phdr->p_vaddr + phdr->p_memsz <= seg->p_vaddr + seg->p_memsz - ) { - params->dynamic_addr = (phdr->p_vaddr - seg->p_vaddr) + seg->addr; - - /* check the dynamic section contains at least one item, and that - * the last item is a NULL entry */ + phdr->p_vaddr + phdr->p_memsz <= + seg->p_vaddr + seg->p_memsz) { + params->dynamic_addr = + (phdr->p_vaddr - seg->p_vaddr) + + seg->addr; + + /* check the dynamic section contains at least + * one item, and that the last item is a NULL + * entry */ if (phdr->p_memsz == 0 || phdr->p_memsz % sizeof(Elf32_Dyn) != 0) goto dynamic_error; tmp = phdr->p_memsz / sizeof(Elf32_Dyn); - if (((Elf32_Dyn *) params->dynamic_addr)[tmp - 1].d_tag != 0) + if (((Elf32_Dyn *) + params->dynamic_addr)[tmp - 1].d_tag != 0) goto dynamic_error; break; } @@ -775,8 +814,8 @@ static int elf_fdpic_map_file(struct elf_fdpic_params *params, } /* now elide adjacent segments in the load map on MMU linux - * - on uClinux the holes between may actually be filled with system stuff or stuff from - * other processes + * - on uClinux the holes between may actually be filled with system + * stuff or stuff from other processes */ #ifdef CONFIG_MMU nloads = loadmap->nsegs; @@ -787,7 +826,9 @@ static int elf_fdpic_map_file(struct elf_fdpic_params *params, if (seg->p_vaddr - mseg->p_vaddr == seg->addr - mseg->addr) { load_addr = PAGE_ALIGN(mseg->addr + mseg->p_memsz); if (load_addr == (seg->addr & PAGE_MASK)) { - mseg->p_memsz += load_addr - (mseg->addr + mseg->p_memsz); + mseg->p_memsz += + load_addr - + (mseg->addr + mseg->p_memsz); mseg->p_memsz += seg->addr & ~PAGE_MASK; mseg->p_memsz += seg->p_memsz; loadmap->nsegs--; @@ -815,20 +856,21 @@ static int elf_fdpic_map_file(struct elf_fdpic_params *params, return 0; - dynamic_error: +dynamic_error: printk("ELF FDPIC %s with invalid DYNAMIC section (inode=%lu)\n", what, file->f_dentry->d_inode->i_ino); return -ELIBBAD; -} /* end elf_fdpic_map_file() */ +} /*****************************************************************************/ /* * map a file with constant displacement under uClinux */ #ifndef CONFIG_MMU -static int elf_fdpic_map_file_constdisp_on_uclinux(struct elf_fdpic_params *params, - struct file *file, - struct mm_struct *mm) +static int elf_fdpic_map_file_constdisp_on_uclinux( + struct elf_fdpic_params *params, + struct file *file, + struct mm_struct *mm) { struct elf32_fdpic_loadseg *seg; struct elf32_phdr *phdr; @@ -839,7 +881,8 @@ static int elf_fdpic_map_file_constdisp_on_uclinux(struct elf_fdpic_params *para load_addr = params->load_addr; seg = params->loadmap->segs; - /* determine the bounds of the contiguous overall allocation we must make */ + /* determine the bounds of the contiguous overall allocation we must + * make */ phdr = params->phdrs; for (loop = 0; loop < params->hdr.e_phnum; loop++, phdr++) { if (params->phdrs[loop].p_type != PT_LOAD) @@ -860,7 +903,7 @@ static int elf_fdpic_map_file_constdisp_on_uclinux(struct elf_fdpic_params *para maddr = do_mmap(NULL, load_addr, top - base, PROT_READ | PROT_WRITE | PROT_EXEC, mflags, 0); up_write(&mm->mmap_sem); - if (IS_ERR((void *) maddr)) + if (IS_ERR_VALUE(maddr)) return (int) maddr; if (load_addr != 0) @@ -878,7 +921,8 @@ static int elf_fdpic_map_file_constdisp_on_uclinux(struct elf_fdpic_params *para seg->p_vaddr = phdr->p_vaddr; seg->p_memsz = phdr->p_memsz; - ret = file->f_op->read(file, (void *) seg->addr, phdr->p_filesz, &fpos); + ret = file->f_op->read(file, (void *) seg->addr, + phdr->p_filesz, &fpos); if (ret < 0) return ret; @@ -895,8 +939,7 @@ static int elf_fdpic_map_file_constdisp_on_uclinux(struct elf_fdpic_params *para if (phdr->p_flags & PF_X) { mm->start_code = seg->addr; mm->end_code = seg->addr + phdr->p_memsz; - } - else if (!mm->start_data) { + } else if (!mm->start_data) { mm->start_data = seg->addr; #ifndef CONFIG_MMU mm->end_data = seg->addr + phdr->p_memsz; @@ -913,7 +956,7 @@ static int elf_fdpic_map_file_constdisp_on_uclinux(struct elf_fdpic_params *para } return 0; -} /* end elf_fdpic_map_file_constdisp_on_uclinux() */ +} #endif /*****************************************************************************/ @@ -974,14 +1017,14 @@ static int elf_fdpic_map_file_by_direct_mmap(struct elf_fdpic_params *params, case ELF_FDPIC_FLAG_CONSTDISP: /* constant displacement - * - can be mapped anywhere, but must be mapped as a unit + * - can be mapped anywhere, but must be mapped as a + * unit */ if (!dvset) { maddr = load_addr; delta_vaddr = phdr->p_vaddr; dvset = 1; - } - else { + } else { maddr = load_addr + phdr->p_vaddr - delta_vaddr; flags |= MAP_FIXED; } @@ -1005,13 +1048,14 @@ static int elf_fdpic_map_file_by_direct_mmap(struct elf_fdpic_params *params, up_write(&mm->mmap_sem); kdebug("mmap[%d] <file> sz=%lx pr=%x fl=%x of=%lx --> %08lx", - loop, phdr->p_memsz + disp, prot, flags, phdr->p_offset - disp, - maddr); + loop, phdr->p_memsz + disp, prot, flags, + phdr->p_offset - disp, maddr); - if (IS_ERR((void *) maddr)) + if (IS_ERR_VALUE(maddr)) return (int) maddr; - if ((params->flags & ELF_FDPIC_FLAG_ARRANGEMENT) == ELF_FDPIC_FLAG_CONTIGUOUS) + if ((params->flags & ELF_FDPIC_FLAG_ARRANGEMENT) == + ELF_FDPIC_FLAG_CONTIGUOUS) load_addr += PAGE_ALIGN(phdr->p_memsz + disp); seg->addr = maddr + disp; @@ -1022,7 +1066,8 @@ static int elf_fdpic_map_file_by_direct_mmap(struct elf_fdpic_params *params, if (phdr->p_offset == 0) params->elfhdr_addr = seg->addr; - /* clear the bit between beginning of mapping and beginning of PT_LOAD */ + /* clear the bit between beginning of mapping and beginning of + * PT_LOAD */ if (prot & PROT_WRITE && disp > 0) { kdebug("clear[%d] ad=%lx sz=%lx", loop, maddr, disp); clear_user((void __user *) maddr, disp); @@ -1038,19 +1083,20 @@ static int elf_fdpic_map_file_by_direct_mmap(struct elf_fdpic_params *params, excess1 = PAGE_SIZE - ((maddr + phdr->p_filesz) & ~PAGE_MASK); #ifdef CONFIG_MMU - if (excess > excess1) { unsigned long xaddr = maddr + phdr->p_filesz + excess1; unsigned long xmaddr; flags |= MAP_FIXED | MAP_ANONYMOUS; down_write(&mm->mmap_sem); - xmaddr = do_mmap(NULL, xaddr, excess - excess1, prot, flags, 0); + xmaddr = do_mmap(NULL, xaddr, excess - excess1, + prot, flags, 0); up_write(&mm->mmap_sem); kdebug("mmap[%d] <anon>" " ad=%lx sz=%lx pr=%x fl=%x of=0 --> %08lx", - loop, xaddr, excess - excess1, prot, flags, xmaddr); + loop, xaddr, excess - excess1, prot, flags, + xmaddr); if (xmaddr != xaddr) return -ENOMEM; @@ -1059,7 +1105,8 @@ static int elf_fdpic_map_file_by_direct_mmap(struct elf_fdpic_params *params, if (prot & PROT_WRITE && excess1 > 0) { kdebug("clear[%d] ad=%lx sz=%lx", loop, maddr + phdr->p_filesz, excess1); - clear_user((void __user *) maddr + phdr->p_filesz, excess1); + clear_user((void __user *) maddr + phdr->p_filesz, + excess1); } #else @@ -1074,8 +1121,7 @@ static int elf_fdpic_map_file_by_direct_mmap(struct elf_fdpic_params *params, if (phdr->p_flags & PF_X) { mm->start_code = maddr; mm->end_code = maddr + phdr->p_memsz; - } - else if (!mm->start_data) { + } else if (!mm->start_data) { mm->start_data = maddr; mm->end_data = maddr + phdr->p_memsz; } @@ -1085,4 +1131,661 @@ static int elf_fdpic_map_file_by_direct_mmap(struct elf_fdpic_params *params, } return 0; -} /* end elf_fdpic_map_file_by_direct_mmap() */ +} + +/*****************************************************************************/ +/* + * ELF-FDPIC core dumper + * + * Modelled on fs/exec.c:aout_core_dump() + * Jeremy Fitzhardinge <jeremy@sw.oz.au> + * + * Modelled on fs/binfmt_elf.c core dumper + */ +#if defined(USE_ELF_CORE_DUMP) && defined(CONFIG_ELF_CORE) + +/* + * These are the only things you should do on a core-file: use only these + * functions to write out all the necessary info. + */ +static int dump_write(struct file *file, const void *addr, int nr) +{ + return file->f_op->write(file, addr, nr, &file->f_pos) == nr; +} + +static int dump_seek(struct file *file, loff_t off) +{ + if (file->f_op->llseek) { + if (file->f_op->llseek(file, off, SEEK_SET) != off) + return 0; + } else { + file->f_pos = off; + } + return 1; +} + +/* + * Decide whether a segment is worth dumping; default is yes to be + * sure (missing info is worse than too much; etc). + * Personally I'd include everything, and use the coredump limit... + * + * I think we should skip something. But I am not sure how. H.J. + */ +static int maydump(struct vm_area_struct *vma) +{ + /* Do not dump I/O mapped devices or special mappings */ + if (vma->vm_flags & (VM_IO | VM_RESERVED)) { + kdcore("%08lx: %08lx: no (IO)", vma->vm_start, vma->vm_flags); + return 0; + } + + /* If we may not read the contents, don't allow us to dump + * them either. "dump_write()" can't handle it anyway. + */ + if (!(vma->vm_flags & VM_READ)) { + kdcore("%08lx: %08lx: no (!read)", vma->vm_start, vma->vm_flags); + return 0; + } + + /* Dump shared memory only if mapped from an anonymous file. */ + if (vma->vm_flags & VM_SHARED) { + if (vma->vm_file->f_dentry->d_inode->i_nlink == 0) { + kdcore("%08lx: %08lx: no (share)", vma->vm_start, vma->vm_flags); + return 1; + } + + kdcore("%08lx: %08lx: no (share)", vma->vm_start, vma->vm_flags); + return 0; + } + +#ifdef CONFIG_MMU + /* If it hasn't been written to, don't write it out */ + if (!vma->anon_vma) { + kdcore("%08lx: %08lx: no (!anon)", vma->vm_start, vma->vm_flags); + return 0; + } +#endif + + kdcore("%08lx: %08lx: yes", vma->vm_start, vma->vm_flags); + return 1; +} + +/* An ELF note in memory */ +struct memelfnote +{ + const char *name; + int type; + unsigned int datasz; + void *data; +}; + +static int notesize(struct memelfnote *en) +{ + int sz; + + sz = sizeof(struct elf_note); + sz += roundup(strlen(en->name) + 1, 4); + sz += roundup(en->datasz, 4); + + return sz; +} + +/* #define DEBUG */ + +#define DUMP_WRITE(addr, nr) \ + do { if (!dump_write(file, (addr), (nr))) return 0; } while(0) +#define DUMP_SEEK(off) \ + do { if (!dump_seek(file, (off))) return 0; } while(0) + +static int writenote(struct memelfnote *men, struct file *file) +{ + struct elf_note en; + + en.n_namesz = strlen(men->name) + 1; + en.n_descsz = men->datasz; + en.n_type = men->type; + + DUMP_WRITE(&en, sizeof(en)); + DUMP_WRITE(men->name, en.n_namesz); + /* XXX - cast from long long to long to avoid need for libgcc.a */ + DUMP_SEEK(roundup((unsigned long)file->f_pos, 4)); /* XXX */ + DUMP_WRITE(men->data, men->datasz); + DUMP_SEEK(roundup((unsigned long)file->f_pos, 4)); /* XXX */ + + return 1; +} +#undef DUMP_WRITE +#undef DUMP_SEEK + +#define DUMP_WRITE(addr, nr) \ + if ((size += (nr)) > limit || !dump_write(file, (addr), (nr))) \ + goto end_coredump; +#define DUMP_SEEK(off) \ + if (!dump_seek(file, (off))) \ + goto end_coredump; + +static inline void fill_elf_fdpic_header(struct elfhdr *elf, int segs) +{ + memcpy(elf->e_ident, ELFMAG, SELFMAG); + elf->e_ident[EI_CLASS] = ELF_CLASS; + elf->e_ident[EI_DATA] = ELF_DATA; + elf->e_ident[EI_VERSION] = EV_CURRENT; + elf->e_ident[EI_OSABI] = ELF_OSABI; + memset(elf->e_ident+EI_PAD, 0, EI_NIDENT-EI_PAD); + + elf->e_type = ET_CORE; + elf->e_machine = ELF_ARCH; + elf->e_version = EV_CURRENT; + elf->e_entry = 0; + elf->e_phoff = sizeof(struct elfhdr); + elf->e_shoff = 0; + elf->e_flags = ELF_FDPIC_CORE_EFLAGS; + elf->e_ehsize = sizeof(struct elfhdr); + elf->e_phentsize = sizeof(struct elf_phdr); + elf->e_phnum = segs; + elf->e_shentsize = 0; + elf->e_shnum = 0; + elf->e_shstrndx = 0; + return; +} + +static inline void fill_elf_note_phdr(struct elf_phdr *phdr, int sz, loff_t offset) +{ + phdr->p_type = PT_NOTE; + phdr->p_offset = offset; + phdr->p_vaddr = 0; + phdr->p_paddr = 0; + phdr->p_filesz = sz; + phdr->p_memsz = 0; + phdr->p_flags = 0; + phdr->p_align = 0; + return; +} + +static inline void fill_note(struct memelfnote *note, const char *name, int type, + unsigned int sz, void *data) +{ + note->name = name; + note->type = type; + note->datasz = sz; + note->data = data; + return; +} + +/* + * fill up all the fields in prstatus from the given task struct, except + * registers which need to be filled up seperately. + */ +static void fill_prstatus(struct elf_prstatus *prstatus, + struct task_struct *p, long signr) +{ + prstatus->pr_info.si_signo = prstatus->pr_cursig = signr; + prstatus->pr_sigpend = p->pending.signal.sig[0]; + prstatus->pr_sighold = p->blocked.sig[0]; + prstatus->pr_pid = p->pid; + prstatus->pr_ppid = p->parent->pid; + prstatus->pr_pgrp = process_group(p); + prstatus->pr_sid = p->signal->session; + if (thread_group_leader(p)) { + /* + * This is the record for the group leader. Add in the + * cumulative times of previous dead threads. This total + * won't include the time of each live thread whose state + * is included in the core dump. The final total reported + * to our parent process when it calls wait4 will include + * those sums as well as the little bit more time it takes + * this and each other thread to finish dying after the + * core dump synchronization phase. + */ + cputime_to_timeval(cputime_add(p->utime, p->signal->utime), + &prstatus->pr_utime); + cputime_to_timeval(cputime_add(p->stime, p->signal->stime), + &prstatus->pr_stime); + } else { + cputime_to_timeval(p->utime, &prstatus->pr_utime); + cputime_to_timeval(p->stime, &prstatus->pr_stime); + } + cputime_to_timeval(p->signal->cutime, &prstatus->pr_cutime); + cputime_to_timeval(p->signal->cstime, &prstatus->pr_cstime); + + prstatus->pr_exec_fdpic_loadmap = p->mm->context.exec_fdpic_loadmap; + prstatus->pr_interp_fdpic_loadmap = p->mm->context.interp_fdpic_loadmap; +} + +static int fill_psinfo(struct elf_prpsinfo *psinfo, struct task_struct *p, + struct mm_struct *mm) +{ + unsigned int i, len; + + /* first copy the parameters from user space */ + memset(psinfo, 0, sizeof(struct elf_prpsinfo)); + + len = mm->arg_end - mm->arg_start; + if (len >= ELF_PRARGSZ) + len = ELF_PRARGSZ - 1; + if (copy_from_user(&psinfo->pr_psargs, + (const char __user *) mm->arg_start, len)) + return -EFAULT; + for (i = 0; i < len; i++) + if (psinfo->pr_psargs[i] == 0) + psinfo->pr_psargs[i] = ' '; + psinfo->pr_psargs[len] = 0; + + psinfo->pr_pid = p->pid; + psinfo->pr_ppid = p->parent->pid; + psinfo->pr_pgrp = process_group(p); + psinfo->pr_sid = p->signal->session; + + i = p->state ? ffz(~p->state) + 1 : 0; + psinfo->pr_state = i; + psinfo->pr_sname = (i > 5) ? '.' : "RSDTZW"[i]; + psinfo->pr_zomb = psinfo->pr_sname == 'Z'; + psinfo->pr_nice = task_nice(p); + psinfo->pr_flag = p->flags; + SET_UID(psinfo->pr_uid, p->uid); + SET_GID(psinfo->pr_gid, p->gid); + strncpy(psinfo->pr_fname, p->comm, sizeof(psinfo->pr_fname)); + + return 0; +} + +/* Here is the structure in which status of each thread is captured. */ +struct elf_thread_status +{ + struct list_head list; + struct elf_prstatus prstatus; /* NT_PRSTATUS */ + elf_fpregset_t fpu; /* NT_PRFPREG */ + struct task_struct *thread; +#ifdef ELF_CORE_COPY_XFPREGS + elf_fpxregset_t xfpu; /* NT_PRXFPREG */ +#endif + struct memelfnote notes[3]; + int num_notes; +}; + +/* + * In order to add the specific thread information for the elf file format, + * we need to keep a linked list of every thread's pr_status and then create + * a single section for them in the final core file. + */ +static int elf_dump_thread_status(long signr, struct elf_thread_status *t) +{ + struct task_struct *p = t->thread; + int sz = 0; + + t->num_notes = 0; + + fill_prstatus(&t->prstatus, p, signr); + elf_core_copy_task_regs(p, &t->prstatus.pr_reg); + + fill_note(&t->notes[0], "CORE", NT_PRSTATUS, sizeof(t->prstatus), + &t->prstatus); + t->num_notes++; + sz += notesize(&t->notes[0]); + + t->prstatus.pr_fpvalid = elf_core_copy_task_fpregs(p, NULL, &t->fpu); + if (t->prstatus.pr_fpvalid) { + fill_note(&t->notes[1], "CORE", NT_PRFPREG, sizeof(t->fpu), + &t->fpu); + t->num_notes++; + sz += notesize(&t->notes[1]); + } + +#ifdef ELF_CORE_COPY_XFPREGS + if (elf_core_copy_task_xfpregs(p, &t->xfpu)) { + fill_note(&t->notes[2], "LINUX", NT_PRXFPREG, sizeof(t->xfpu), + &t->xfpu); + t->num_notes++; + sz += notesize(&t->notes[2]); + } +#endif + return sz; +} + +/* + * dump the segments for an MMU process + */ +#ifdef CONFIG_MMU +static int elf_fdpic_dump_segments(struct file *file, struct mm_struct *mm, + size_t *size, unsigned long *limit) +{ + struct vm_area_struct *vma; + + for (vma = current->mm->mmap; vma; vma = vma->vm_next) { + unsigned long addr; + + if (!maydump(vma)) + continue; + + for (addr = vma->vm_start; + addr < vma->vm_end; + addr += PAGE_SIZE + ) { + struct vm_area_struct *vma; + struct page *page; + + if (get_user_pages(current, current->mm, addr, 1, 0, 1, + &page, &vma) <= 0) { + DUMP_SEEK(file->f_pos + PAGE_SIZE); + } + else if (page == ZERO_PAGE(addr)) { + DUMP_SEEK(file->f_pos + PAGE_SIZE); + page_cache_release(page); + } + else { + void *kaddr; + + flush_cache_page(vma, addr, page_to_pfn(page)); + kaddr = kmap(page); + if ((*size += PAGE_SIZE) > *limit || + !dump_write(file, kaddr, PAGE_SIZE) + ) { + kunmap(page); + page_cache_release(page); + return -EIO; + } + kunmap(page); + page_cache_release(page); + } + } + } + + return 0; + +end_coredump: + return -EFBIG; +} +#endif + +/* + * dump the segments for a NOMMU process + */ +#ifndef CONFIG_MMU +static int elf_fdpic_dump_segments(struct file *file, struct mm_struct *mm, + size_t *size, unsigned long *limit) +{ + struct vm_list_struct *vml; + + for (vml = current->mm->context.vmlist; vml; vml = vml->next) { + struct vm_area_struct *vma = vml->vma; + + if (!maydump(vma)) + continue; + + if ((*size += PAGE_SIZE) > *limit) + return -EFBIG; + + if (!dump_write(file, (void *) vma->vm_start, + vma->vm_end - vma->vm_start)) + return -EIO; + } + + return 0; +} +#endif + +/* + * Actual dumper + * + * This is a two-pass process; first we find the offsets of the bits, + * and then they are actually written out. If we run out of core limit + * we just truncate. + */ +static int elf_fdpic_core_dump(long signr, struct pt_regs *regs, + struct file *file) +{ +#define NUM_NOTES 6 + int has_dumped = 0; + mm_segment_t fs; + int segs; + size_t size = 0; + int i; + struct vm_area_struct *vma; + struct elfhdr *elf = NULL; + loff_t offset = 0, dataoff; + unsigned long limit = current->signal->rlim[RLIMIT_CORE].rlim_cur; + int numnote; + struct memelfnote *notes = NULL; + struct elf_prstatus *prstatus = NULL; /* NT_PRSTATUS */ + struct elf_prpsinfo *psinfo = NULL; /* NT_PRPSINFO */ + struct task_struct *g, *p; + LIST_HEAD(thread_list); + struct list_head *t; + elf_fpregset_t *fpu = NULL; +#ifdef ELF_CORE_COPY_XFPREGS + elf_fpxregset_t *xfpu = NULL; +#endif + int thread_status_size = 0; +#ifndef CONFIG_MMU + struct vm_list_struct *vml; +#endif + elf_addr_t *auxv; + + /* + * We no longer stop all VM operations. + * + * This is because those proceses that could possibly change map_count + * or the mmap / vma pages are now blocked in do_exit on current + * finishing this core dump. + * + * Only ptrace can touch these memory addresses, but it doesn't change + * the map_count or the pages allocated. So no possibility of crashing + * exists while dumping the mm->vm_next areas to the core file. + */ + + /* alloc memory for large data structures: too large to be on stack */ + elf = kmalloc(sizeof(*elf), GFP_KERNEL); + if (!elf) + goto cleanup; + prstatus = kzalloc(sizeof(*prstatus), GFP_KERNEL); + if (!prstatus) + goto cleanup; + psinfo = kmalloc(sizeof(*psinfo), GFP_KERNEL); + if (!psinfo) + goto cleanup; + notes = kmalloc(NUM_NOTES * sizeof(struct memelfnote), GFP_KERNEL); + if (!notes) + goto cleanup; + fpu = kmalloc(sizeof(*fpu), GFP_KERNEL); + if (!fpu) + goto cleanup; +#ifdef ELF_CORE_COPY_XFPREGS + xfpu = kmalloc(sizeof(*xfpu), GFP_KERNEL); + if (!xfpu) + goto cleanup; +#endif + + if (signr) { + struct elf_thread_status *tmp; + rcu_read_lock(); + do_each_thread(g,p) + if (current->mm == p->mm && current != p) { + tmp = kzalloc(sizeof(*tmp), GFP_ATOMIC); + if (!tmp) { + rcu_read_unlock(); + goto cleanup; + } + tmp->thread = p; + list_add(&tmp->list, &thread_list); + } + while_each_thread(g,p); + rcu_read_unlock(); + list_for_each(t, &thread_list) { + struct elf_thread_status *tmp; + int sz; + + tmp = list_entry(t, struct elf_thread_status, list); + sz = elf_dump_thread_status(signr, tmp); + thread_status_size += sz; + } + } + + /* now collect the dump for the current */ + fill_prstatus(prstatus, current, signr); + elf_core_copy_regs(&prstatus->pr_reg, regs); + +#ifdef CONFIG_MMU + segs = current->mm->map_count; +#else + segs = 0; + for (vml = current->mm->context.vmlist; vml; vml = vml->next) + segs++; +#endif +#ifdef ELF_CORE_EXTRA_PHDRS + segs += ELF_CORE_EXTRA_PHDRS; +#endif + + /* Set up header */ + fill_elf_fdpic_header(elf, segs + 1); /* including notes section */ + + has_dumped = 1; + current->flags |= PF_DUMPCORE; + + /* + * Set up the notes in similar form to SVR4 core dumps made + * with info from their /proc. + */ + + fill_note(notes + 0, "CORE", NT_PRSTATUS, sizeof(*prstatus), prstatus); + fill_psinfo(psinfo, current->group_leader, current->mm); + fill_note(notes + 1, "CORE", NT_PRPSINFO, sizeof(*psinfo), psinfo); + + numnote = 2; + + auxv = (elf_addr_t *) current->mm->saved_auxv; + + i = 0; + do + i += 2; + while (auxv[i - 2] != AT_NULL); + fill_note(¬es[numnote++], "CORE", NT_AUXV, + i * sizeof(elf_addr_t), auxv); + + /* Try to dump the FPU. */ + if ((prstatus->pr_fpvalid = + elf_core_copy_task_fpregs(current, regs, fpu))) + fill_note(notes + numnote++, + "CORE", NT_PRFPREG, sizeof(*fpu), fpu); +#ifdef ELF_CORE_COPY_XFPREGS + if (elf_core_copy_task_xfpregs(current, xfpu)) + fill_note(notes + numnote++, + "LINUX", NT_PRXFPREG, sizeof(*xfpu), xfpu); +#endif + + fs = get_fs(); + set_fs(KERNEL_DS); + + DUMP_WRITE(elf, sizeof(*elf)); + offset += sizeof(*elf); /* Elf header */ + offset += (segs+1) * sizeof(struct elf_phdr); /* Program headers */ + + /* Write notes phdr entry */ + { + struct elf_phdr phdr; + int sz = 0; + + for (i = 0; i < numnote; i++) + sz += notesize(notes + i); + + sz += thread_status_size; + + fill_elf_note_phdr(&phdr, sz, offset); + offset += sz; + DUMP_WRITE(&phdr, sizeof(phdr)); + } + + /* Page-align dumped data */ + dataoff = offset = roundup(offset, ELF_EXEC_PAGESIZE); + + /* write program headers for segments dump */ + for ( +#ifdef CONFIG_MMU + vma = current->mm->mmap; vma; vma = vma->vm_next +#else + vml = current->mm->context.vmlist; vml; vml = vml->next +#endif + ) { + struct elf_phdr phdr; + size_t sz; + +#ifndef CONFIG_MMU + vma = vml->vma; +#endif + + sz = vma->vm_end - vma->vm_start; + + phdr.p_type = PT_LOAD; + phdr.p_offset = offset; + phdr.p_vaddr = vma->vm_start; + phdr.p_paddr = 0; + phdr.p_filesz = maydump(vma) ? sz : 0; + phdr.p_memsz = sz; + offset += phdr.p_filesz; + phdr.p_flags = vma->vm_flags & VM_READ ? PF_R : 0; + if (vma->vm_flags & VM_WRITE) + phdr.p_flags |= PF_W; + if (vma->vm_flags & VM_EXEC) + phdr.p_flags |= PF_X; + phdr.p_align = ELF_EXEC_PAGESIZE; + + DUMP_WRITE(&phdr, sizeof(phdr)); + } + +#ifdef ELF_CORE_WRITE_EXTRA_PHDRS + ELF_CORE_WRITE_EXTRA_PHDRS; +#endif + + /* write out the notes section */ + for (i = 0; i < numnote; i++) + if (!writenote(notes + i, file)) + goto end_coredump; + + /* write out the thread status notes section */ + list_for_each(t, &thread_list) { + struct elf_thread_status *tmp = + list_entry(t, struct elf_thread_status, list); + + for (i = 0; i < tmp->num_notes; i++) + if (!writenote(&tmp->notes[i], file)) + goto end_coredump; + } + + DUMP_SEEK(dataoff); + + if (elf_fdpic_dump_segments(file, current->mm, &size, &limit) < 0) + goto end_coredump; + +#ifdef ELF_CORE_WRITE_EXTRA_DATA + ELF_CORE_WRITE_EXTRA_DATA; +#endif + + if (file->f_pos != offset) { + /* Sanity check */ + printk(KERN_WARNING + "elf_core_dump: file->f_pos (%lld) != offset (%lld)\n", + file->f_pos, offset); + } + +end_coredump: + set_fs(fs); + +cleanup: + while (!list_empty(&thread_list)) { + struct list_head *tmp = thread_list.next; + list_del(tmp); + kfree(list_entry(tmp, struct elf_thread_status, list)); + } + + kfree(elf); + kfree(prstatus); + kfree(psinfo); + kfree(notes); + kfree(fpu); +#ifdef ELF_CORE_COPY_XFPREGS + kfree(xfpu); +#endif + return has_dumped; +#undef NUM_NOTES +} + +#endif /* USE_ELF_CORE_DUMP */ diff --git a/fs/binfmt_flat.c b/fs/binfmt_flat.c index b1c902e319c..a62fd4018a2 100644 --- a/fs/binfmt_flat.c +++ b/fs/binfmt_flat.c @@ -16,7 +16,6 @@ */ #include <linux/module.h> -#include <linux/config.h> #include <linux/kernel.h> #include <linux/sched.h> #include <linux/mm.h> @@ -510,7 +509,7 @@ static int load_flat_file(struct linux_binprm * bprm, } /* OK, This is the point of no return */ - set_personality(PER_LINUX); + set_personality(PER_LINUX_32BIT); } /* diff --git a/fs/binfmt_misc.c b/fs/binfmt_misc.c index 07a4996cca3..1713c48fef5 100644 --- a/fs/binfmt_misc.c +++ b/fs/binfmt_misc.c @@ -55,6 +55,7 @@ typedef struct { } Node; static DEFINE_RWLOCK(entries_lock); +static struct file_system_type bm_fs_type; static struct vfsmount *bm_mnt; static int entry_count; @@ -214,10 +215,8 @@ _error: bprm->interp_flags = 0; bprm->interp_data = 0; _unshare: - if (files) { - put_files_struct(current->files); - current->files = files; - } + if (files) + reset_files_struct(current, files); goto _ret; } @@ -506,7 +505,6 @@ static struct inode *bm_get_inode(struct super_block *sb, int mode) inode->i_mode = mode; inode->i_uid = 0; inode->i_gid = 0; - inode->i_blksize = PAGE_CACHE_SIZE; inode->i_blocks = 0; inode->i_atime = inode->i_mtime = inode->i_ctime = current_fs_time(inode->i_sb); @@ -516,7 +514,7 @@ static struct inode *bm_get_inode(struct super_block *sb, int mode) static void bm_clear_inode(struct inode *inode) { - kfree(inode->u.generic_ip); + kfree(inode->i_private); } static void kill_node(Node *e) @@ -544,7 +542,7 @@ static void kill_node(Node *e) static ssize_t bm_entry_read(struct file * file, char __user * buf, size_t nbytes, loff_t *ppos) { - Node *e = file->f_dentry->d_inode->u.generic_ip; + Node *e = file->f_dentry->d_inode->i_private; loff_t pos = *ppos; ssize_t res; char *page; @@ -578,7 +576,7 @@ static ssize_t bm_entry_write(struct file *file, const char __user *buffer, size_t count, loff_t *ppos) { struct dentry *root; - Node *e = file->f_dentry->d_inode->u.generic_ip; + Node *e = file->f_dentry->d_inode->i_private; int res = parse_command(buffer, count); switch (res) { @@ -637,7 +635,7 @@ static ssize_t bm_register_write(struct file *file, const char __user *buffer, if (!inode) goto out2; - err = simple_pin_fs("binfmt_misc", &bm_mnt, &entry_count); + err = simple_pin_fs(&bm_fs_type, &bm_mnt, &entry_count); if (err) { iput(inode); inode = NULL; @@ -645,7 +643,7 @@ static ssize_t bm_register_write(struct file *file, const char __user *buffer, } e->dentry = dget(dentry); - inode->u.generic_ip = e; + inode->i_private = e; inode->i_fop = &bm_entry_operations; d_instantiate(dentry, inode); diff --git a/fs/binfmt_som.c b/fs/binfmt_som.c index 00a91dc25d1..32b5d625ce9 100644 --- a/fs/binfmt_som.c +++ b/fs/binfmt_som.c @@ -32,7 +32,6 @@ #include <asm/uaccess.h> #include <asm/pgtable.h> -#include <linux/config.h> #include <linux/elf.h> diff --git a/fs/block_dev.c b/fs/block_dev.c index 028d9fb9c2d..4346468139e 100644 --- a/fs/block_dev.c +++ b/fs/block_dev.c @@ -5,14 +5,12 @@ * Copyright (C) 2001 Andrea Arcangeli <andrea@suse.de> SuSE */ -#include <linux/config.h> #include <linux/init.h> #include <linux/mm.h> #include <linux/fcntl.h> #include <linux/slab.h> #include <linux/kmod.h> #include <linux/major.h> -#include <linux/devfs_fs_kernel.h> #include <linux/smp_lock.h> #include <linux/highmem.h> #include <linux/blkdev.h> @@ -545,11 +543,11 @@ static struct kobject *bdev_get_holder(struct block_device *bdev) return kobject_get(bdev->bd_disk->holder_dir); } -static void add_symlink(struct kobject *from, struct kobject *to) +static int add_symlink(struct kobject *from, struct kobject *to) { if (!from || !to) - return; - sysfs_create_link(from, to, kobject_name(to)); + return 0; + return sysfs_create_link(from, to, kobject_name(to)); } static void del_symlink(struct kobject *from, struct kobject *to) @@ -650,30 +648,38 @@ static void free_bd_holder(struct bd_holder *bo) * If there is no matching entry with @bo in @bdev->bd_holder_list, * add @bo to the list, create symlinks. * - * Returns 1 if @bo was added to the list. - * Returns 0 if @bo wasn't used by any reason and should be freed. + * Returns 0 if symlinks are created or already there. + * Returns -ve if something fails and @bo can be freed. */ static int add_bd_holder(struct block_device *bdev, struct bd_holder *bo) { struct bd_holder *tmp; + int ret; if (!bo) - return 0; + return -EINVAL; list_for_each_entry(tmp, &bdev->bd_holder_list, list) { if (tmp->sdir == bo->sdir) { tmp->count++; + /* We've already done what we need to do here. */ + free_bd_holder(bo); return 0; } } if (!bd_holder_grab_dirs(bdev, bo)) - return 0; + return -EBUSY; - add_symlink(bo->sdir, bo->sdev); - add_symlink(bo->hdir, bo->hdev); - list_add_tail(&bo->list, &bdev->bd_holder_list); - return 1; + ret = add_symlink(bo->sdir, bo->sdev); + if (ret == 0) { + ret = add_symlink(bo->hdir, bo->hdev); + if (ret) + del_symlink(bo->sdir, bo->sdev); + } + if (ret == 0) + list_add_tail(&bo->list, &bdev->bd_holder_list); + return ret; } /** @@ -741,9 +747,11 @@ static int bd_claim_by_kobject(struct block_device *bdev, void *holder, if (!bo) return -ENOMEM; - mutex_lock(&bdev->bd_mutex); + mutex_lock_nested(&bdev->bd_mutex, BD_MUTEX_PARTITION); res = bd_claim(bdev, holder); - if (res || !add_bd_holder(bdev, bo)) + if (res == 0) + res = add_bd_holder(bdev, bo); + if (res) free_bd_holder(bo); mutex_unlock(&bdev->bd_mutex); @@ -766,7 +774,7 @@ static void bd_release_from_kobject(struct block_device *bdev, if (!kobj) return; - mutex_lock(&bdev->bd_mutex); + mutex_lock_nested(&bdev->bd_mutex, BD_MUTEX_PARTITION); bd_release(bdev); if ((bo = del_bd_holder(bdev, kobj))) free_bd_holder(bo); @@ -824,6 +832,22 @@ struct block_device *open_by_devnum(dev_t dev, unsigned mode) EXPORT_SYMBOL(open_by_devnum); +static int +blkdev_get_partition(struct block_device *bdev, mode_t mode, unsigned flags); + +struct block_device *open_partition_by_devnum(dev_t dev, unsigned mode) +{ + struct block_device *bdev = bdget(dev); + int err = -ENOMEM; + int flags = mode & FMODE_WRITE ? O_RDWR : O_RDONLY; + if (bdev) + err = blkdev_get_partition(bdev, mode, flags); + return err ? ERR_PTR(err) : bdev; +} + +EXPORT_SYMBOL(open_partition_by_devnum); + + /* * This routine checks whether a removable media has been changed, * and invalidates all buffer-cache-entries in that case. This @@ -870,7 +894,66 @@ void bd_set_size(struct block_device *bdev, loff_t size) } EXPORT_SYMBOL(bd_set_size); -static int do_open(struct block_device *bdev, struct file *file) +static int __blkdev_put(struct block_device *bdev, unsigned int subclass) +{ + int ret = 0; + struct inode *bd_inode = bdev->bd_inode; + struct gendisk *disk = bdev->bd_disk; + + mutex_lock_nested(&bdev->bd_mutex, subclass); + lock_kernel(); + if (!--bdev->bd_openers) { + sync_blockdev(bdev); + kill_bdev(bdev); + } + if (bdev->bd_contains == bdev) { + if (disk->fops->release) + ret = disk->fops->release(bd_inode, NULL); + } else { + mutex_lock_nested(&bdev->bd_contains->bd_mutex, + subclass + 1); + bdev->bd_contains->bd_part_count--; + mutex_unlock(&bdev->bd_contains->bd_mutex); + } + if (!bdev->bd_openers) { + struct module *owner = disk->fops->owner; + + put_disk(disk); + module_put(owner); + + if (bdev->bd_contains != bdev) { + kobject_put(&bdev->bd_part->kobj); + bdev->bd_part = NULL; + } + bdev->bd_disk = NULL; + bdev->bd_inode->i_data.backing_dev_info = &default_backing_dev_info; + if (bdev != bdev->bd_contains) + __blkdev_put(bdev->bd_contains, subclass + 1); + bdev->bd_contains = NULL; + } + unlock_kernel(); + mutex_unlock(&bdev->bd_mutex); + bdput(bdev); + return ret; +} + +int blkdev_put(struct block_device *bdev) +{ + return __blkdev_put(bdev, BD_MUTEX_NORMAL); +} +EXPORT_SYMBOL(blkdev_put); + +int blkdev_put_partition(struct block_device *bdev) +{ + return __blkdev_put(bdev, BD_MUTEX_PARTITION); +} +EXPORT_SYMBOL(blkdev_put_partition); + +static int +blkdev_get_whole(struct block_device *bdev, mode_t mode, unsigned flags); + +static int +do_open(struct block_device *bdev, struct file *file, unsigned int subclass) { struct module *owner = NULL; struct gendisk *disk; @@ -887,7 +970,8 @@ static int do_open(struct block_device *bdev, struct file *file) } owner = disk->fops->owner; - mutex_lock(&bdev->bd_mutex); + mutex_lock_nested(&bdev->bd_mutex, subclass); + if (!bdev->bd_openers) { bdev->bd_disk = disk; bdev->bd_contains = bdev; @@ -914,11 +998,11 @@ static int do_open(struct block_device *bdev, struct file *file) ret = -ENOMEM; if (!whole) goto out_first; - ret = blkdev_get(whole, file->f_mode, file->f_flags); + ret = blkdev_get_whole(whole, file->f_mode, file->f_flags); if (ret) goto out_first; bdev->bd_contains = whole; - mutex_lock(&whole->bd_mutex); + mutex_lock_nested(&whole->bd_mutex, BD_MUTEX_WHOLE); whole->bd_part_count++; p = disk->part[part - 1]; bdev->bd_inode->i_data.backing_dev_info = @@ -946,7 +1030,8 @@ static int do_open(struct block_device *bdev, struct file *file) if (bdev->bd_invalidated) rescan_partitions(bdev->bd_disk, bdev); } else { - mutex_lock(&bdev->bd_contains->bd_mutex); + mutex_lock_nested(&bdev->bd_contains->bd_mutex, + BD_MUTEX_WHOLE); bdev->bd_contains->bd_part_count++; mutex_unlock(&bdev->bd_contains->bd_mutex); } @@ -960,7 +1045,7 @@ out_first: bdev->bd_disk = NULL; bdev->bd_inode->i_data.backing_dev_info = &default_backing_dev_info; if (bdev != bdev->bd_contains) - blkdev_put(bdev->bd_contains); + __blkdev_put(bdev->bd_contains, BD_MUTEX_WHOLE); bdev->bd_contains = NULL; put_disk(disk); module_put(owner); @@ -987,11 +1072,49 @@ int blkdev_get(struct block_device *bdev, mode_t mode, unsigned flags) fake_file.f_dentry = &fake_dentry; fake_dentry.d_inode = bdev->bd_inode; - return do_open(bdev, &fake_file); + return do_open(bdev, &fake_file, BD_MUTEX_NORMAL); } EXPORT_SYMBOL(blkdev_get); +static int +blkdev_get_whole(struct block_device *bdev, mode_t mode, unsigned flags) +{ + /* + * This crockload is due to bad choice of ->open() type. + * It will go away. + * For now, block device ->open() routine must _not_ + * examine anything in 'inode' argument except ->i_rdev. + */ + struct file fake_file = {}; + struct dentry fake_dentry = {}; + fake_file.f_mode = mode; + fake_file.f_flags = flags; + fake_file.f_dentry = &fake_dentry; + fake_dentry.d_inode = bdev->bd_inode; + + return do_open(bdev, &fake_file, BD_MUTEX_WHOLE); +} + +static int +blkdev_get_partition(struct block_device *bdev, mode_t mode, unsigned flags) +{ + /* + * This crockload is due to bad choice of ->open() type. + * It will go away. + * For now, block device ->open() routine must _not_ + * examine anything in 'inode' argument except ->i_rdev. + */ + struct file fake_file = {}; + struct dentry fake_dentry = {}; + fake_file.f_mode = mode; + fake_file.f_flags = flags; + fake_file.f_dentry = &fake_dentry; + fake_dentry.d_inode = bdev->bd_inode; + + return do_open(bdev, &fake_file, BD_MUTEX_PARTITION); +} + static int blkdev_open(struct inode * inode, struct file * filp) { struct block_device *bdev; @@ -1007,7 +1130,7 @@ static int blkdev_open(struct inode * inode, struct file * filp) bdev = bd_acquire(inode); - res = do_open(bdev, filp); + res = do_open(bdev, filp, BD_MUTEX_NORMAL); if (res) return res; @@ -1021,51 +1144,6 @@ static int blkdev_open(struct inode * inode, struct file * filp) return res; } -int blkdev_put(struct block_device *bdev) -{ - int ret = 0; - struct inode *bd_inode = bdev->bd_inode; - struct gendisk *disk = bdev->bd_disk; - - mutex_lock(&bdev->bd_mutex); - lock_kernel(); - if (!--bdev->bd_openers) { - sync_blockdev(bdev); - kill_bdev(bdev); - } - if (bdev->bd_contains == bdev) { - if (disk->fops->release) - ret = disk->fops->release(bd_inode, NULL); - } else { - mutex_lock(&bdev->bd_contains->bd_mutex); - bdev->bd_contains->bd_part_count--; - mutex_unlock(&bdev->bd_contains->bd_mutex); - } - if (!bdev->bd_openers) { - struct module *owner = disk->fops->owner; - - put_disk(disk); - module_put(owner); - - if (bdev->bd_contains != bdev) { - kobject_put(&bdev->bd_part->kobj); - bdev->bd_part = NULL; - } - bdev->bd_disk = NULL; - bdev->bd_inode->i_data.backing_dev_info = &default_backing_dev_info; - if (bdev != bdev->bd_contains) { - blkdev_put(bdev->bd_contains); - } - bdev->bd_contains = NULL; - } - unlock_kernel(); - mutex_unlock(&bdev->bd_mutex); - bdput(bdev); - return ret; -} - -EXPORT_SYMBOL(blkdev_put); - static int blkdev_close(struct inode * inode, struct file * filp) { struct block_device *bdev = I_BDEV(filp->f_mapping->host); @@ -1095,7 +1173,7 @@ static long block_ioctl(struct file *file, unsigned cmd, unsigned long arg) return blkdev_ioctl(file->f_mapping->host, file, cmd, arg); } -struct address_space_operations def_blk_aops = { +const struct address_space_operations def_blk_aops = { .readpage = blkdev_readpage, .writepage = blkdev_writepage, .sync_page = block_sync_page, diff --git a/fs/buffer.c b/fs/buffer.c index 373bb6292bd..3b6d701073e 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -18,7 +18,6 @@ * async buffer flushing, 1999 Andrea Arcangeli <andrea@suse.de> */ -#include <linux/config.h> #include <linux/kernel.h> #include <linux/syscalls.h> #include <linux/fs.h> @@ -471,13 +470,18 @@ out: pass does the actual I/O. */ void invalidate_bdev(struct block_device *bdev, int destroy_dirty_buffers) { + struct address_space *mapping = bdev->bd_inode->i_mapping; + + if (mapping->nrpages == 0) + return; + invalidate_bh_lrus(); /* * FIXME: what about destroy_dirty_buffers? * We really want to use invalidate_inode_pages2() for * that, but not until that's cleaned up. */ - invalidate_inode_pages(bdev->bd_inode->i_mapping); + invalidate_inode_pages(mapping); } /* @@ -564,7 +568,7 @@ still_busy: * Completion handler for block_write_full_page() - pages which are unlocked * during I/O, and which have PageWriteback cleared upon I/O completion. */ -void end_buffer_async_write(struct buffer_head *bh, int uptodate) +static void end_buffer_async_write(struct buffer_head *bh, int uptodate) { char b[BDEVNAME_SIZE]; unsigned long flags; @@ -852,7 +856,7 @@ int __set_page_dirty_buffers(struct page *page) write_lock_irq(&mapping->tree_lock); if (page->mapping) { /* Race with truncate? */ if (mapping_cap_account_dirty(mapping)) - inc_page_state(nr_dirty); + __inc_zone_page_state(page, NR_FILE_DIRTY); radix_tree_tag_set(&mapping->page_tree, page_index(page), PAGECACHE_TAG_DIRTY); @@ -2598,7 +2602,7 @@ int nobh_truncate_page(struct address_space *mapping, loff_t from) unsigned offset = from & (PAGE_CACHE_SIZE-1); unsigned to; struct page *page; - struct address_space_operations *a_ops = mapping->a_ops; + const struct address_space_operations *a_ops = mapping->a_ops; char *kaddr; int ret = 0; @@ -2983,6 +2987,7 @@ int try_to_free_buffers(struct page *page) spin_lock(&mapping->private_lock); ret = drop_buffers(page, &buffers_to_free); + spin_unlock(&mapping->private_lock); if (ret) { /* * If the filesystem writes its buffers by hand (eg ext3) @@ -2994,7 +2999,6 @@ int try_to_free_buffers(struct page *page) */ clear_page_dirty(page); } - spin_unlock(&mapping->private_lock); out: if (buffers_to_free) { struct buffer_head *bh = buffers_to_free; @@ -3166,7 +3170,6 @@ EXPORT_SYMBOL(block_sync_page); EXPORT_SYMBOL(block_truncate_page); EXPORT_SYMBOL(block_write_full_page); EXPORT_SYMBOL(cont_prepare_write); -EXPORT_SYMBOL(end_buffer_async_write); EXPORT_SYMBOL(end_buffer_read_sync); EXPORT_SYMBOL(end_buffer_write_sync); EXPORT_SYMBOL(file_fsync); diff --git a/fs/char_dev.c b/fs/char_dev.c index f3418f7a6e9..1f3285affa3 100644 --- a/fs/char_dev.c +++ b/fs/char_dev.c @@ -4,7 +4,6 @@ * Copyright (C) 1991, 1992 Linus Torvalds */ -#include <linux/config.h> #include <linux/init.h> #include <linux/fs.h> #include <linux/slab.h> @@ -14,18 +13,36 @@ #include <linux/errno.h> #include <linux/module.h> #include <linux/smp_lock.h> -#include <linux/devfs_fs_kernel.h> #include <linux/seq_file.h> #include <linux/kobject.h> #include <linux/kobj_map.h> #include <linux/cdev.h> #include <linux/mutex.h> +#include <linux/backing-dev.h> #ifdef CONFIG_KMOD #include <linux/kmod.h> #endif +/* + * capabilities for /dev/mem, /dev/kmem and similar directly mappable character + * devices + * - permits shared-mmap for read, write and/or exec + * - does not permit private mmap in NOMMU mode (can't do COW) + * - no readahead or I/O queue unplugging required + */ +struct backing_dev_info directly_mappable_cdev_bdi = { + .capabilities = ( +#ifdef CONFIG_MMU + /* permit private copies of the data to be taken */ + BDI_CAP_MAP_COPY | +#endif + /* permit direct mmap, for read, write or exec */ + BDI_CAP_MAP_DIRECT | + BDI_CAP_READ_MAP | BDI_CAP_WRITE_MAP | BDI_CAP_EXEC_MAP), +}; + static struct kobj_map *cdev_map; static DEFINE_MUTEX(chrdevs_lock); @@ -111,13 +128,31 @@ __register_chrdev_region(unsigned int major, unsigned int baseminor, for (cp = &chrdevs[i]; *cp; cp = &(*cp)->next) if ((*cp)->major > major || - ((*cp)->major == major && (*cp)->baseminor >= baseminor)) + ((*cp)->major == major && + (((*cp)->baseminor >= baseminor) || + ((*cp)->baseminor + (*cp)->minorct > baseminor)))) break; - if (*cp && (*cp)->major == major && - (*cp)->baseminor < baseminor + minorct) { - ret = -EBUSY; - goto out; + + /* Check for overlapping minor ranges. */ + if (*cp && (*cp)->major == major) { + int old_min = (*cp)->baseminor; + int old_max = (*cp)->baseminor + (*cp)->minorct - 1; + int new_min = baseminor; + int new_max = baseminor + minorct - 1; + + /* New driver overlaps from the left. */ + if (new_max >= old_min && new_max <= old_max) { + ret = -EBUSY; + goto out; + } + + /* New driver overlaps from the right. */ + if (new_min <= old_max && new_min >= old_min) { + ret = -EBUSY; + goto out; + } } + cd->next = *cp; *cp = cd; mutex_unlock(&chrdevs_lock); @@ -148,6 +183,15 @@ __unregister_chrdev_region(unsigned major, unsigned baseminor, int minorct) return cd; } +/** + * register_chrdev_region() - register a range of device numbers + * @from: the first in the desired range of device numbers; must include + * the major number. + * @count: the number of consecutive device numbers required + * @name: the name of the device or driver. + * + * Return value is zero on success, a negative error code on failure. + */ int register_chrdev_region(dev_t from, unsigned count, const char *name) { struct char_device_struct *cd; @@ -173,6 +217,17 @@ fail: return PTR_ERR(cd); } +/** + * alloc_chrdev_region() - register a range of char device numbers + * @dev: output parameter for first assigned number + * @baseminor: first of the requested range of minor numbers + * @count: the number of minor numbers required + * @name: the name of the associated device or driver + * + * Allocates a range of char device numbers. The major number will be + * chosen dynamically, and returned (along with the first minor number) + * in @dev. Returns zero or a negative error code. + */ int alloc_chrdev_region(dev_t *dev, unsigned baseminor, unsigned count, const char *name) { @@ -184,6 +239,28 @@ int alloc_chrdev_region(dev_t *dev, unsigned baseminor, unsigned count, return 0; } +/** + * register_chrdev() - Register a major number for character devices. + * @major: major device number or 0 for dynamic allocation + * @name: name of this range of devices + * @fops: file operations associated with this devices + * + * If @major == 0 this functions will dynamically allocate a major and return + * its number. + * + * If @major > 0 this function will attempt to reserve a device with the given + * major number and will return zero on success. + * + * Returns a -ve errno on failure. + * + * The name of this device has nothing to do with the name of the device in + * /dev. It only helps to keep track of the different owners of devices. If + * your module name has only one type of devices it's ok to use e.g. the name + * of the module here. + * + * This function registers a range of 256 minor numbers. The first minor number + * is 0. + */ int register_chrdev(unsigned int major, const char *name, const struct file_operations *fops) { @@ -220,6 +297,15 @@ out2: return err; } +/** + * unregister_chrdev_region() - return a range of device numbers + * @from: the first in the range of numbers to unregister + * @count: the number of device numbers to unregister + * + * This function will unregister a range of @count device numbers, + * starting with @from. The caller should normally be the one who + * allocated those numbers in the first place... + */ void unregister_chrdev_region(dev_t from, unsigned count) { dev_t to = from + count; @@ -357,6 +443,16 @@ static int exact_lock(dev_t dev, void *data) return cdev_get(p) ? 0 : -1; } +/** + * cdev_add() - add a char device to the system + * @p: the cdev structure for the device + * @dev: the first device number for which this device is responsible + * @count: the number of consecutive minor numbers corresponding to this + * device + * + * cdev_add() adds the device represented by @p to the system, making it + * live immediately. A negative error code is returned on failure. + */ int cdev_add(struct cdev *p, dev_t dev, unsigned count) { p->dev = dev; @@ -369,6 +465,13 @@ static void cdev_unmap(dev_t dev, unsigned count) kobj_unmap(cdev_map, dev, count); } +/** + * cdev_del() - remove a cdev from the system + * @p: the cdev structure to be removed + * + * cdev_del() removes @p from the system, possibly freeing the structure + * itself. + */ void cdev_del(struct cdev *p) { cdev_unmap(p->dev, p->count); @@ -397,6 +500,11 @@ static struct kobj_type ktype_cdev_dynamic = { .release = cdev_dynamic_release, }; +/** + * cdev_alloc() - allocate a cdev structure + * + * Allocates and returns a cdev structure, or NULL on failure. + */ struct cdev *cdev_alloc(void) { struct cdev *p = kzalloc(sizeof(struct cdev), GFP_KERNEL); @@ -408,6 +516,14 @@ struct cdev *cdev_alloc(void) return p; } +/** + * cdev_init() - initialize a cdev structure + * @cdev: the structure to initialize + * @fops: the file_operations for this device + * + * Initializes @cdev, remembering @fops, making it ready to add to the + * system with cdev_add(). + */ void cdev_init(struct cdev *cdev, const struct file_operations *fops) { memset(cdev, 0, sizeof *cdev); @@ -441,3 +557,4 @@ EXPORT_SYMBOL(cdev_del); EXPORT_SYMBOL(cdev_add); EXPORT_SYMBOL(register_chrdev); EXPORT_SYMBOL(unregister_chrdev); +EXPORT_SYMBOL(directly_mappable_cdev_bdi); diff --git a/fs/cifs/CHANGES b/fs/cifs/CHANGES index 7271bb0257f..1eb9a2ec0a3 100644 --- a/fs/cifs/CHANGES +++ b/fs/cifs/CHANGES @@ -1,9 +1,39 @@ +Version 1.46 +------------ +Support deep tree mounts. Better support OS/2, Win9x (DOS) time stamps. + +Version 1.45 +------------ +Do not time out lockw calls when using posix extensions. Do not +time out requests if server still responding reasonably fast +on requests on other threads. Improve POSIX locking emulation, +(lock cancel now works, and unlock of merged range works even +to Windows servers now). Fix oops on mount to lanman servers +(win9x, os/2 etc.) when null password. Do not send listxattr +(SMB to query all EAs) if nouser_xattr specified. Fix SE Linux +problem (instantiate inodes/dentries in right order for readdir). + +Version 1.44 +------------ +Rewritten sessionsetup support, including support for legacy SMB +session setup needed for OS/2 and older servers such as Windows 95 and 98. +Fix oops on ls to OS/2 servers. Add support for level 1 FindFirst +so we can do search (ls etc.) to OS/2. Do not send NTCreateX +or recent levels of FindFirst unless server says it supports NT SMBs +(instead use legacy equivalents from LANMAN dialect). Fix to allow +NTLMv2 authentication support (now can use stronger password hashing +on mount if corresponding /proc/fs/cifs/SecurityFlags is set (0x4004). +Allow override of global cifs security flags on mount via "sec=" option(s). + Version 1.43 ------------ POSIX locking to servers which support CIFS POSIX Extensions (disabled by default controlled by proc/fs/cifs/Experimental). Handle conversion of long share names (especially Asian languages) -to Unicode during mount. +to Unicode during mount. Fix memory leak in sess struct on reconnect. +Fix rare oops after acpi suspend. Fix O_TRUNC opens to overwrite on +cifs open which helps rare case when setpathinfo fails or server does +not support it. Version 1.42 ------------ diff --git a/fs/cifs/Makefile b/fs/cifs/Makefile index 58c77254a23..a26f26ed5a1 100644 --- a/fs/cifs/Makefile +++ b/fs/cifs/Makefile @@ -3,4 +3,4 @@ # obj-$(CONFIG_CIFS) += cifs.o -cifs-objs := cifsfs.o cifssmb.o cifs_debug.o connect.o dir.o file.o inode.o link.o misc.o netmisc.o smbdes.o smbencrypt.o transport.o asn1.o md4.o md5.o cifs_unicode.o nterr.o xattr.o cifsencrypt.o fcntl.o readdir.o ioctl.o ntlmssp.o +cifs-objs := cifsfs.o cifssmb.o cifs_debug.o connect.o dir.o file.o inode.o link.o misc.o netmisc.o smbdes.o smbencrypt.o transport.o asn1.o md4.o md5.o cifs_unicode.o nterr.o xattr.o cifsencrypt.o fcntl.o readdir.o ioctl.o sess.o diff --git a/fs/cifs/README b/fs/cifs/README index 0355003f4f0..5f0e1bd64fe 100644 --- a/fs/cifs/README +++ b/fs/cifs/README @@ -408,7 +408,7 @@ A partial list of the supported mount options follows: user_xattr Allow getting and setting user xattrs as OS/2 EAs (extended attributes) to the server (default) e.g. via setfattr and getfattr utilities. - nouser_xattr Do not allow getfattr/setfattr to get/set xattrs + nouser_xattr Do not allow getfattr/setfattr to get/set/list xattrs mapchars Translate six of the seven reserved characters (not backslash) *?<>|: to the remap range (above 0xF000), which also @@ -443,7 +443,10 @@ A partial list of the supported mount options follows: SFU does). In the future the bottom 9 bits of the mode mode also will be emulated using queries of the security descriptor (ACL). -sec Security mode. Allowed values are: + sign Must use packet signing (helps avoid unwanted data modification + by intermediate systems in the route). Note that signing + does not work with lanman or plaintext authentication. + sec Security mode. Allowed values are: none attempt to connection as a null user (no name) krb5 Use Kerberos version 5 authentication krb5i Use Kerberos authentication and packet signing @@ -453,6 +456,8 @@ sec Security mode. Allowed values are: server requires signing also can be the default) ntlmv2 Use NTLMv2 password hashing ntlmv2i Use NTLMv2 password hashing with packet signing + lanman (if configured in kernel config) use older + lanman hash The mount.cifs mount helper also accepts a few mount options before -o including: @@ -485,14 +490,34 @@ PacketSigningEnabled If set to one, cifs packet signing is enabled it. If set to two, cifs packet signing is required even if the server considers packet signing optional. (default 1) +SecurityFlags Flags which control security negotiation and + also packet signing. Authentication (may/must) + flags (e.g. for NTLM and/or NTLMv2) may be combined with + the signing flags. Specifying two different password + hashing mechanisms (as "must use") on the other hand + does not make much sense. Default flags are + 0x07007 + (NTLM, NTLMv2 and packet signing allowed). Maximum + allowable flags if you want to allow mounts to servers + using weaker password hashes is 0x37037 (lanman, + plaintext, ntlm, ntlmv2, signing allowed): + + may use packet signing 0x00001 + must use packet signing 0x01001 + may use NTLM (most common password hash) 0x00002 + must use NTLM 0x02002 + may use NTLMv2 0x00004 + must use NTLMv2 0x04004 + may use Kerberos security (not implemented yet) 0x00008 + must use Kerberos (not implemented yet) 0x08008 + may use lanman (weak) password hash 0x00010 + must use lanman password hash 0x10010 + may use plaintext passwords 0x00020 + must use plaintext passwords 0x20020 + (reserved for future packet encryption) 0x00040 + cifsFYI If set to one, additional debug information is logged to the system error log. (default 0) -ExtendedSecurity If set to one, SPNEGO session establishment - is allowed which enables more advanced - secure CIFS session establishment (default 0) -NTLMV2Enabled If set to one, more secure password hashes - are used when the server supports them and - when kerberos is not negotiated (default 0) traceSMB If set to one, debug information is logged to the system error log with the start of smb requests and responses (default 0) diff --git a/fs/cifs/asn1.c b/fs/cifs/asn1.c index 086ae8f4a20..2e75883b7f5 100644 --- a/fs/cifs/asn1.c +++ b/fs/cifs/asn1.c @@ -17,7 +17,6 @@ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ -#include <linux/config.h> #include <linux/module.h> #include <linux/types.h> #include <linux/kernel.h> @@ -467,7 +466,7 @@ decode_negTokenInit(unsigned char *security_blob, int length, asn1_open(&ctx, security_blob, length); if (asn1_header_decode(&ctx, &end, &cls, &con, &tag) == 0) { - cFYI(1, ("Error decoding negTokenInit header ")); + cFYI(1, ("Error decoding negTokenInit header")); return 0; } else if ((cls != ASN1_APL) || (con != ASN1_CON) || (tag != ASN1_EOC)) { @@ -495,7 +494,7 @@ decode_negTokenInit(unsigned char *security_blob, int length, } if (asn1_header_decode(&ctx, &end, &cls, &con, &tag) == 0) { - cFYI(1, ("Error decoding negTokenInit ")); + cFYI(1, ("Error decoding negTokenInit")); return 0; } else if ((cls != ASN1_CTX) || (con != ASN1_CON) || (tag != ASN1_EOC)) { @@ -505,7 +504,7 @@ decode_negTokenInit(unsigned char *security_blob, int length, } if (asn1_header_decode(&ctx, &end, &cls, &con, &tag) == 0) { - cFYI(1, ("Error decoding negTokenInit ")); + cFYI(1, ("Error decoding negTokenInit")); return 0; } else if ((cls != ASN1_UNI) || (con != ASN1_CON) || (tag != ASN1_SEQ)) { @@ -515,7 +514,7 @@ decode_negTokenInit(unsigned char *security_blob, int length, } if (asn1_header_decode(&ctx, &end, &cls, &con, &tag) == 0) { - cFYI(1, ("Error decoding 2nd part of negTokenInit ")); + cFYI(1, ("Error decoding 2nd part of negTokenInit")); return 0; } else if ((cls != ASN1_CTX) || (con != ASN1_CON) || (tag != ASN1_EOC)) { @@ -527,7 +526,7 @@ decode_negTokenInit(unsigned char *security_blob, int length, if (asn1_header_decode (&ctx, &sequence_end, &cls, &con, &tag) == 0) { - cFYI(1, ("Error decoding 2nd part of negTokenInit ")); + cFYI(1, ("Error decoding 2nd part of negTokenInit")); return 0; } else if ((cls != ASN1_UNI) || (con != ASN1_CON) || (tag != ASN1_SEQ)) { diff --git a/fs/cifs/cifs_debug.c b/fs/cifs/cifs_debug.c index f4124a32bef..96abeb73897 100644 --- a/fs/cifs/cifs_debug.c +++ b/fs/cifs/cifs_debug.c @@ -39,7 +39,7 @@ cifs_dump_mem(char *label, void *data, int length) char *charptr = data; char buf[10], line[80]; - printk(KERN_DEBUG "%s: dump of %d bytes of data at 0x%p\n\n", + printk(KERN_DEBUG "%s: dump of %d bytes of data at 0x%p\n", label, length, data); for (i = 0; i < length; i += 16) { line[0] = 0; @@ -57,6 +57,57 @@ cifs_dump_mem(char *label, void *data, int length) } } +#ifdef CONFIG_CIFS_DEBUG2 +void cifs_dump_detail(struct smb_hdr * smb) +{ + cERROR(1,("Cmd: %d Err: 0x%x Flags: 0x%x Flgs2: 0x%x Mid: %d Pid: %d", + smb->Command, smb->Status.CifsError, + smb->Flags, smb->Flags2, smb->Mid, smb->Pid)); + cERROR(1,("smb buf %p len %d", smb, smbCalcSize_LE(smb))); +} + + +void cifs_dump_mids(struct TCP_Server_Info * server) +{ + struct list_head *tmp; + struct mid_q_entry * mid_entry; + + if(server == NULL) + return; + + cERROR(1,("Dump pending requests:")); + spin_lock(&GlobalMid_Lock); + list_for_each(tmp, &server->pending_mid_q) { + mid_entry = list_entry(tmp, struct mid_q_entry, qhead); + if(mid_entry) { + cERROR(1,("State: %d Cmd: %d Pid: %d Tsk: %p Mid %d", + mid_entry->midState, + (int)mid_entry->command, + mid_entry->pid, + mid_entry->tsk, + mid_entry->mid)); +#ifdef CONFIG_CIFS_STATS2 + cERROR(1,("IsLarge: %d buf: %p time rcv: %ld now: %ld", + mid_entry->largeBuf, + mid_entry->resp_buf, + mid_entry->when_received, + jiffies)); +#endif /* STATS2 */ + cERROR(1,("IsMult: %d IsEnd: %d", mid_entry->multiRsp, + mid_entry->multiEnd)); + if(mid_entry->resp_buf) { + cifs_dump_detail(mid_entry->resp_buf); + cifs_dump_mem("existing buf: ", + mid_entry->resp_buf, + 62 /* fixme */); + } + + } + } + spin_unlock(&GlobalMid_Lock); +} +#endif /* CONFIG_CIFS_DEBUG2 */ + #ifdef CONFIG_PROC_FS static int cifs_debug_data_read(char *buf, char **beginBuffer, off_t offset, @@ -73,7 +124,6 @@ cifs_debug_data_read(char *buf, char **beginBuffer, off_t offset, *beginBuffer = buf + offset; - length = sprintf(buf, "Display Internal CIFS Data Structures for Debugging\n" @@ -395,12 +445,12 @@ static read_proc_t traceSMB_read; static write_proc_t traceSMB_write; static read_proc_t multiuser_mount_read; static write_proc_t multiuser_mount_write; -static read_proc_t extended_security_read; -static write_proc_t extended_security_write; -static read_proc_t ntlmv2_enabled_read; +static read_proc_t security_flags_read; +static write_proc_t security_flags_write; +/* static read_proc_t ntlmv2_enabled_read; static write_proc_t ntlmv2_enabled_write; static read_proc_t packet_signing_enabled_read; -static write_proc_t packet_signing_enabled_write; +static write_proc_t packet_signing_enabled_write;*/ static read_proc_t experimEnabled_read; static write_proc_t experimEnabled_write; static read_proc_t linuxExtensionsEnabled_read; @@ -458,10 +508,10 @@ cifs_proc_init(void) pde->write_proc = multiuser_mount_write; pde = - create_proc_read_entry("ExtendedSecurity", 0, proc_fs_cifs, - extended_security_read, NULL); + create_proc_read_entry("SecurityFlags", 0, proc_fs_cifs, + security_flags_read, NULL); if (pde) - pde->write_proc = extended_security_write; + pde->write_proc = security_flags_write; pde = create_proc_read_entry("LookupCacheEnabled", 0, proc_fs_cifs, @@ -469,7 +519,7 @@ cifs_proc_init(void) if (pde) pde->write_proc = lookupFlag_write; - pde = +/* pde = create_proc_read_entry("NTLMV2Enabled", 0, proc_fs_cifs, ntlmv2_enabled_read, NULL); if (pde) @@ -479,7 +529,7 @@ cifs_proc_init(void) create_proc_read_entry("PacketSigningEnabled", 0, proc_fs_cifs, packet_signing_enabled_read, NULL); if (pde) - pde->write_proc = packet_signing_enabled_write; + pde->write_proc = packet_signing_enabled_write;*/ } void @@ -496,9 +546,9 @@ cifs_proc_clean(void) #endif remove_proc_entry("MultiuserMount", proc_fs_cifs); remove_proc_entry("OplockEnabled", proc_fs_cifs); - remove_proc_entry("NTLMV2Enabled",proc_fs_cifs); - remove_proc_entry("ExtendedSecurity",proc_fs_cifs); - remove_proc_entry("PacketSigningEnabled",proc_fs_cifs); +/* remove_proc_entry("NTLMV2Enabled",proc_fs_cifs); */ + remove_proc_entry("SecurityFlags",proc_fs_cifs); +/* remove_proc_entry("PacketSigningEnabled",proc_fs_cifs); */ remove_proc_entry("LinuxExtensionsEnabled",proc_fs_cifs); remove_proc_entry("Experimental",proc_fs_cifs); remove_proc_entry("LookupCacheEnabled",proc_fs_cifs); @@ -782,12 +832,12 @@ multiuser_mount_write(struct file *file, const char __user *buffer, } static int -extended_security_read(char *page, char **start, off_t off, +security_flags_read(char *page, char **start, off_t off, int count, int *eof, void *data) { int len; - len = sprintf(page, "%d\n", extended_security); + len = sprintf(page, "0x%x\n", extended_security); len -= off; *start = page + off; @@ -803,24 +853,52 @@ extended_security_read(char *page, char **start, off_t off, return len; } static int -extended_security_write(struct file *file, const char __user *buffer, +security_flags_write(struct file *file, const char __user *buffer, unsigned long count, void *data) { + unsigned int flags; + char flags_string[12]; char c; - int rc; - rc = get_user(c, buffer); - if (rc) - return rc; - if (c == '0' || c == 'n' || c == 'N') - extended_security = 0; - else if (c == '1' || c == 'y' || c == 'Y') - extended_security = 1; + if((count < 1) || (count > 11)) + return -EINVAL; + + memset(flags_string, 0, 12); + + if(copy_from_user(flags_string, buffer, count)) + return -EFAULT; + + if(count < 3) { + /* single char or single char followed by null */ + c = flags_string[0]; + if (c == '0' || c == 'n' || c == 'N') + extended_security = CIFSSEC_DEF; /* default */ + else if (c == '1' || c == 'y' || c == 'Y') + extended_security = CIFSSEC_MAX; + return count; + } + /* else we have a number */ + + flags = simple_strtoul(flags_string, NULL, 0); + + cFYI(1,("sec flags 0x%x", flags)); + + if(flags <= 0) { + cERROR(1,("invalid security flags %s",flags_string)); + return -EINVAL; + } + if(flags & ~CIFSSEC_MASK) { + cERROR(1,("attempt to set unsupported security flags 0x%x", + flags & ~CIFSSEC_MASK)); + return -EINVAL; + } + /* flags look ok - update the global security flags for cifs module */ + extended_security = flags; return count; } -static int +/* static int ntlmv2_enabled_read(char *page, char **start, off_t off, int count, int *eof, void *data) { @@ -855,6 +933,8 @@ ntlmv2_enabled_write(struct file *file, const char __user *buffer, ntlmv2_support = 0; else if (c == '1' || c == 'y' || c == 'Y') ntlmv2_support = 1; + else if (c == '2') + ntlmv2_support = 2; return count; } @@ -898,7 +978,7 @@ packet_signing_enabled_write(struct file *file, const char __user *buffer, sign_CIFS_PDUs = 2; return count; -} +} */ #endif diff --git a/fs/cifs/cifs_debug.h b/fs/cifs/cifs_debug.h index 4304d9dcfb6..c26cd0d2c6d 100644 --- a/fs/cifs/cifs_debug.h +++ b/fs/cifs/cifs_debug.h @@ -24,6 +24,10 @@ #define _H_CIFS_DEBUG void cifs_dump_mem(char *label, void *data, int length); +#ifdef CONFIG_CIFS_DEBUG2 +void cifs_dump_detail(struct smb_hdr *); +void cifs_dump_mids(struct TCP_Server_Info *); +#endif extern int traceSMB; /* flag which enables the function below */ void dump_smb(struct smb_hdr *, int); #define CIFS_INFO 0x01 diff --git a/fs/cifs/cifs_fs_sb.h b/fs/cifs/cifs_fs_sb.h index ad58eb0c4d6..fd1e52ebcee 100644 --- a/fs/cifs/cifs_fs_sb.h +++ b/fs/cifs/cifs_fs_sb.h @@ -40,5 +40,7 @@ struct cifs_sb_info { mode_t mnt_file_mode; mode_t mnt_dir_mode; int mnt_cifs_flags; + int prepathlen; + char * prepath; }; #endif /* _CIFS_FS_SB_H */ diff --git a/fs/cifs/cifs_unicode.c b/fs/cifs/cifs_unicode.c index d2b12825594..d2a8b2941fc 100644 --- a/fs/cifs/cifs_unicode.c +++ b/fs/cifs/cifs_unicode.c @@ -22,6 +22,7 @@ #include "cifs_unicode.h" #include "cifs_uniupr.h" #include "cifspdu.h" +#include "cifsglob.h" #include "cifs_debug.h" /* diff --git a/fs/cifs/cifsencrypt.c b/fs/cifs/cifsencrypt.c index e7d63737e65..4bc250b2d9f 100644 --- a/fs/cifs/cifsencrypt.c +++ b/fs/cifs/cifsencrypt.c @@ -26,6 +26,8 @@ #include "md5.h" #include "cifs_unicode.h" #include "cifsproto.h" +#include <linux/ctype.h> +#include <linux/random.h> /* Calculate and return the CIFS signature based on the mac key and the smb pdu */ /* the 16 byte signature must be allocated by the caller */ @@ -35,6 +37,8 @@ extern void mdfour(unsigned char *out, unsigned char *in, int n); extern void E_md4hash(const unsigned char *passwd, unsigned char *p16); +extern void SMBencrypt(unsigned char *passwd, unsigned char *c8, + unsigned char *p24); static int cifs_calculate_signature(const struct smb_hdr * cifs_pdu, const char * key, char * signature) @@ -45,7 +49,7 @@ static int cifs_calculate_signature(const struct smb_hdr * cifs_pdu, return -EINVAL; MD5Init(&context); - MD5Update(&context,key,CIFS_SESSION_KEY_SIZE+16); + MD5Update(&context,key,CIFS_SESS_KEY_SIZE+16); MD5Update(&context,cifs_pdu->Protocol,cifs_pdu->smb_buf_length); MD5Final(signature,&context); return 0; @@ -90,7 +94,7 @@ static int cifs_calc_signature2(const struct kvec * iov, int n_vec, return -EINVAL; MD5Init(&context); - MD5Update(&context,key,CIFS_SESSION_KEY_SIZE+16); + MD5Update(&context,key,CIFS_SESS_KEY_SIZE+16); for(i=0;i<n_vec;i++) { if(iov[i].iov_base == NULL) { cERROR(1,("null iovec entry")); @@ -204,11 +208,12 @@ int cifs_calculate_mac_key(char * key, const char * rn, const char * password) E_md4hash(password, temp_key); mdfour(key,temp_key,16); - memcpy(key+16,rn, CIFS_SESSION_KEY_SIZE); + memcpy(key+16,rn, CIFS_SESS_KEY_SIZE); return 0; } -int CalcNTLMv2_partial_mac_key(struct cifsSesInfo * ses, struct nls_table * nls_info) +int CalcNTLMv2_partial_mac_key(struct cifsSesInfo * ses, + const struct nls_table * nls_info) { char temp_hash[16]; struct HMACMD5Context ctx; @@ -225,6 +230,8 @@ int CalcNTLMv2_partial_mac_key(struct cifsSesInfo * ses, struct nls_table * nls_ user_name_len = strlen(ses->userName); if(user_name_len > MAX_USERNAME_SIZE) return -EINVAL; + if(ses->domainName == NULL) + return -EINVAL; /* BB should we use CIFS_LINUX_DOM */ dom_name_len = strlen(ses->domainName); if(dom_name_len > MAX_USERNAME_SIZE) return -EINVAL; @@ -259,16 +266,132 @@ int CalcNTLMv2_partial_mac_key(struct cifsSesInfo * ses, struct nls_table * nls_ kfree(unicode_buf); return 0; } -void CalcNTLMv2_response(const struct cifsSesInfo * ses,char * v2_session_response) + +#ifdef CONFIG_CIFS_WEAK_PW_HASH +void calc_lanman_hash(struct cifsSesInfo * ses, char * lnm_session_key) +{ + int i; + char password_with_pad[CIFS_ENCPWD_SIZE]; + + if(ses->server == NULL) + return; + + memset(password_with_pad, 0, CIFS_ENCPWD_SIZE); + if(ses->password) + strncpy(password_with_pad, ses->password, CIFS_ENCPWD_SIZE); + + if((ses->server->secMode & SECMODE_PW_ENCRYPT) == 0) + if(extended_security & CIFSSEC_MAY_PLNTXT) { + memcpy(lnm_session_key, password_with_pad, CIFS_ENCPWD_SIZE); + return; + } + + /* calculate old style session key */ + /* calling toupper is less broken than repeatedly + calling nls_toupper would be since that will never + work for UTF8, but neither handles multibyte code pages + but the only alternative would be converting to UCS-16 (Unicode) + (using a routine something like UniStrupr) then + uppercasing and then converting back from Unicode - which + would only worth doing it if we knew it were utf8. Basically + utf8 and other multibyte codepages each need their own strupper + function since a byte at a time will ont work. */ + + for(i = 0; i < CIFS_ENCPWD_SIZE; i++) { + password_with_pad[i] = toupper(password_with_pad[i]); + } + + SMBencrypt(password_with_pad, ses->server->cryptKey, lnm_session_key); + /* clear password before we return/free memory */ + memset(password_with_pad, 0, CIFS_ENCPWD_SIZE); +} +#endif /* CIFS_WEAK_PW_HASH */ + +static int calc_ntlmv2_hash(struct cifsSesInfo *ses, + const struct nls_table * nls_cp) +{ + int rc = 0; + int len; + char nt_hash[16]; + struct HMACMD5Context * pctxt; + wchar_t * user; + wchar_t * domain; + + pctxt = kmalloc(sizeof(struct HMACMD5Context), GFP_KERNEL); + + if(pctxt == NULL) + return -ENOMEM; + + /* calculate md4 hash of password */ + E_md4hash(ses->password, nt_hash); + + /* convert Domainname to unicode and uppercase */ + hmac_md5_init_limK_to_64(nt_hash, 16, pctxt); + + /* convert ses->userName to unicode and uppercase */ + len = strlen(ses->userName); + user = kmalloc(2 + (len * 2), GFP_KERNEL); + if(user == NULL) + goto calc_exit_2; + len = cifs_strtoUCS(user, ses->userName, len, nls_cp); + UniStrupr(user); + hmac_md5_update((char *)user, 2*len, pctxt); + + /* convert ses->domainName to unicode and uppercase */ + if(ses->domainName) { + len = strlen(ses->domainName); + + domain = kmalloc(2 + (len * 2), GFP_KERNEL); + if(domain == NULL) + goto calc_exit_1; + len = cifs_strtoUCS(domain, ses->domainName, len, nls_cp); + UniStrupr(domain); + + hmac_md5_update((char *)domain, 2*len, pctxt); + + kfree(domain); + } +calc_exit_1: + kfree(user); +calc_exit_2: + /* BB FIXME what about bytes 24 through 40 of the signing key? + compare with the NTLM example */ + hmac_md5_final(ses->server->mac_signing_key, pctxt); + + return rc; +} + +void setup_ntlmv2_rsp(struct cifsSesInfo * ses, char * resp_buf, + const struct nls_table * nls_cp) +{ + int rc; + struct ntlmv2_resp * buf = (struct ntlmv2_resp *)resp_buf; + + buf->blob_signature = cpu_to_le32(0x00000101); + buf->reserved = 0; + buf->time = cpu_to_le64(cifs_UnixTimeToNT(CURRENT_TIME)); + get_random_bytes(&buf->client_chal, sizeof(buf->client_chal)); + buf->reserved2 = 0; + buf->names[0].type = 0; + buf->names[0].length = 0; + + /* calculate buf->ntlmv2_hash */ + rc = calc_ntlmv2_hash(ses, nls_cp); + if(rc) + cERROR(1,("could not get v2 hash rc %d",rc)); + CalcNTLMv2_response(ses, resp_buf); +} + +void CalcNTLMv2_response(const struct cifsSesInfo * ses, char * v2_session_response) { struct HMACMD5Context context; + /* rest of v2 struct already generated */ memcpy(v2_session_response + 8, ses->server->cryptKey,8); - /* gen_blob(v2_session_response + 16); */ hmac_md5_init_limK_to_64(ses->server->mac_signing_key, 16, &context); - hmac_md5_update(ses->server->cryptKey,8,&context); -/* hmac_md5_update(v2_session_response+16)client thing,8,&context); */ /* BB fix */ + hmac_md5_update(v2_session_response+8, + sizeof(struct ntlmv2_resp) - 8, &context); hmac_md5_final(v2_session_response,&context); - cifs_dump_mem("v2_sess_rsp: ", v2_session_response, 32); /* BB removeme BB */ +/* cifs_dump_mem("v2_sess_rsp: ", v2_session_response, 32); */ } diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c index 7520f468715..22bcf4d7e7a 100644 --- a/fs/cifs/cifsfs.c +++ b/fs/cifs/cifsfs.c @@ -56,8 +56,8 @@ unsigned int experimEnabled = 0; unsigned int linuxExtEnabled = 1; unsigned int lookupCacheEnabled = 1; unsigned int multiuser_mount = 0; -unsigned int extended_security = 0; -unsigned int ntlmv2_support = 0; +unsigned int extended_security = CIFSSEC_DEF; +/* unsigned int ntlmv2_support = 0; */ unsigned int sign_CIFS_PDUs = 1; extern struct task_struct * oplockThread; /* remove sparse warning */ struct task_struct * oplockThread = NULL; @@ -189,7 +189,6 @@ cifs_statfs(struct dentry *dentry, struct kstatfs *buf) buf->f_files = 0; /* undefined */ buf->f_ffree = 0; /* unlimited */ -#ifdef CONFIG_CIFS_EXPERIMENTAL /* BB we could add a second check for a QFS Unix capability bit */ /* BB FIXME check CIFS_POSIX_EXTENSIONS Unix cap first FIXME BB */ if ((pTcon->ses->capabilities & CAP_UNIX) && (CIFS_POSIX_EXTENSIONS & @@ -199,7 +198,6 @@ cifs_statfs(struct dentry *dentry, struct kstatfs *buf) /* Only need to call the old QFSInfo if failed on newer one */ if(rc) -#endif /* CIFS_EXPERIMENTAL */ rc = CIFSSMBQFSInfo(xid, pTcon, buf); /* Old Windows servers do not support level 103, retry with level @@ -255,7 +253,6 @@ cifs_alloc_inode(struct super_block *sb) file data or metadata */ cifs_inode->clientCanCacheRead = FALSE; cifs_inode->clientCanCacheAll = FALSE; - cifs_inode->vfs_inode.i_blksize = CIFS_MAX_MSGSIZE; cifs_inode->vfs_inode.i_blkbits = 14; /* 2**14 = CIFS_MAX_MSGSIZE */ cifs_inode->vfs_inode.i_flags = S_NOATIME | S_NOCMTIME; INIT_LIST_HEAD(&cifs_inode->openFileList); @@ -402,13 +399,14 @@ static struct quotactl_ops cifs_quotactl_ops = { }; #endif -#ifdef CONFIG_CIFS_EXPERIMENTAL -static void cifs_umount_begin(struct super_block * sblock) +static void cifs_umount_begin(struct vfsmount * vfsmnt, int flags) { struct cifs_sb_info *cifs_sb; struct cifsTconInfo * tcon; - cifs_sb = CIFS_SB(sblock); + if (!(flags & MNT_FORCE)) + return; + cifs_sb = CIFS_SB(vfsmnt->mnt_sb); if(cifs_sb == NULL) return; @@ -420,7 +418,7 @@ static void cifs_umount_begin(struct super_block * sblock) tcon->tidStatus = CifsExiting; up(&tcon->tconSem); - /* cancel_brl_requests(tcon); */ + /* cancel_brl_requests(tcon); */ /* BB mark all brl mids as exiting */ /* cancel_notify_requests(tcon); */ if(tcon->ses && tcon->ses->server) { @@ -436,7 +434,6 @@ static void cifs_umount_begin(struct super_block * sblock) return; } -#endif static int cifs_remount(struct super_block *sb, int *flags, char *data) { @@ -455,9 +452,7 @@ struct super_operations cifs_super_ops = { unless later we add lazy close of inodes or unless the kernel forgets to call us with the same number of releases (closes) as opens */ .show_options = cifs_show_options, -#ifdef CONFIG_CIFS_EXPERIMENTAL .umount_begin = cifs_umount_begin, -#endif .remount_fs = cifs_remount, }; @@ -703,8 +698,7 @@ cifs_init_inodecache(void) static void cifs_destroy_inodecache(void) { - if (kmem_cache_destroy(cifs_inode_cachep)) - printk(KERN_WARNING "cifs_inode_cache: error freeing\n"); + kmem_cache_destroy(cifs_inode_cachep); } static int @@ -782,13 +776,9 @@ static void cifs_destroy_request_bufs(void) { mempool_destroy(cifs_req_poolp); - if (kmem_cache_destroy(cifs_req_cachep)) - printk(KERN_WARNING - "cifs_destroy_request_cache: error not all structures were freed\n"); + kmem_cache_destroy(cifs_req_cachep); mempool_destroy(cifs_sm_req_poolp); - if (kmem_cache_destroy(cifs_sm_req_cachep)) - printk(KERN_WARNING - "cifs_destroy_request_cache: cifs_small_rq free error\n"); + kmem_cache_destroy(cifs_sm_req_cachep); } static int @@ -823,13 +813,8 @@ static void cifs_destroy_mids(void) { mempool_destroy(cifs_mid_poolp); - if (kmem_cache_destroy(cifs_mid_cachep)) - printk(KERN_WARNING - "cifs_destroy_mids: error not all structures were freed\n"); - - if (kmem_cache_destroy(cifs_oplock_cachep)) - printk(KERN_WARNING - "error not all oplock structures were freed\n"); + kmem_cache_destroy(cifs_mid_cachep); + kmem_cache_destroy(cifs_oplock_cachep); } static int cifs_oplock_thread(void * dummyarg) @@ -906,7 +891,7 @@ static int cifs_dnotify_thread(void * dummyarg) struct cifsSesInfo *ses; do { - if(try_to_freeze()) + if (try_to_freeze()) continue; set_current_state(TASK_INTERRUPTIBLE); schedule_timeout(15*HZ); diff --git a/fs/cifs/cifsfs.h b/fs/cifs/cifsfs.h index d56c0577c71..bea875d9a46 100644 --- a/fs/cifs/cifsfs.h +++ b/fs/cifs/cifsfs.h @@ -32,7 +32,8 @@ #define TRUE 1 #endif -extern struct address_space_operations cifs_addr_ops; +extern const struct address_space_operations cifs_addr_ops; +extern const struct address_space_operations cifs_addr_ops_smallbuf; /* Functions related to super block operations */ extern struct super_operations cifs_super_ops; @@ -99,5 +100,5 @@ extern ssize_t cifs_getxattr(struct dentry *, const char *, void *, size_t); extern ssize_t cifs_listxattr(struct dentry *, char *, size_t); extern int cifs_ioctl (struct inode * inode, struct file * filep, unsigned int command, unsigned long arg); -#define CIFS_VERSION "1.43" +#define CIFS_VERSION "1.46" #endif /* _CIFSFS_H */ diff --git a/fs/cifs/cifsglob.h b/fs/cifs/cifsglob.h index 006eb33bff5..b24006c47df 100644 --- a/fs/cifs/cifsglob.h +++ b/fs/cifs/cifsglob.h @@ -3,6 +3,7 @@ * * Copyright (C) International Business Machines Corp., 2002,2006 * Author(s): Steve French (sfrench@us.ibm.com) + * Jeremy Allison (jra@samba.org) * * 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 @@ -88,7 +89,8 @@ enum statusEnum { }; enum securityEnum { - NTLM = 0, /* Legacy NTLM012 auth with NTLM hash */ + LANMAN = 0, /* Legacy LANMAN auth */ + NTLM, /* Legacy NTLM012 auth with NTLM hash */ NTLMv2, /* Legacy NTLM auth with NTLMv2 hash */ RawNTLMSSP, /* NTLMSSP without SPNEGO */ NTLMSSP, /* NTLMSSP via SPNEGO */ @@ -157,7 +159,8 @@ struct TCP_Server_Info { /* 16th byte of RFC1001 workstation name is always null */ char workstation_RFC1001_name[SERVER_NAME_LEN_WITH_NULL]; __u32 sequence_number; /* needed for CIFS PDU signature */ - char mac_signing_key[CIFS_SESSION_KEY_SIZE + 16]; + char mac_signing_key[CIFS_SESS_KEY_SIZE + 16]; + unsigned long lstrp; /* when we got last response from this server */ }; /* @@ -179,10 +182,13 @@ struct cifsUidInfo { struct cifsSesInfo { struct list_head cifsSessionList; struct semaphore sesSem; +#if 0 struct cifsUidInfo *uidInfo; /* pointer to user info */ +#endif struct TCP_Server_Info *server; /* pointer to server info */ atomic_t inUse; /* # of mounts (tree connections) on this ses */ enum statusEnum status; + unsigned overrideSecFlg; /* if non-zero override global sec flags */ __u16 ipc_tid; /* special tid for connection to IPC share */ __u16 flags; char *serverOS; /* name of operating system underlying server */ @@ -194,7 +200,7 @@ struct cifsSesInfo { char serverName[SERVER_NAME_LEN_WITH_NULL * 2]; /* BB make bigger for TCP names - will ipv6 and sctp addresses fit? */ char userName[MAX_USERNAME_SIZE + 1]; - char domainName[MAX_USERNAME_SIZE + 1]; + char * domainName; char * password; }; /* session flags */ @@ -209,12 +215,12 @@ struct cifsTconInfo { struct list_head openFileList; struct semaphore tconSem; struct cifsSesInfo *ses; /* pointer to session associated with */ - char treeName[MAX_TREE_SIZE + 1]; /* UNC name of resource (in ASCII not UTF) */ + char treeName[MAX_TREE_SIZE + 1]; /* UNC name of resource in ASCII */ char *nativeFileSystem; __u16 tid; /* The 2 byte tree id */ __u16 Flags; /* optional support bits */ enum statusEnum tidStatus; - atomic_t useCount; /* how many mounts (explicit or implicit) to this share */ + atomic_t useCount; /* how many explicit/implicit mounts to share */ #ifdef CONFIG_CIFS_STATS atomic_t num_smbs_sent; atomic_t num_writes; @@ -254,7 +260,7 @@ struct cifsTconInfo { spinlock_t stat_lock; #endif /* CONFIG_CIFS_STATS */ FILE_SYSTEM_DEVICE_INFO fsDevInfo; - FILE_SYSTEM_ATTRIBUTE_INFO fsAttrInfo; /* ok if file system name truncated */ + FILE_SYSTEM_ATTRIBUTE_INFO fsAttrInfo; /* ok if fs name truncated */ FILE_SYSTEM_UNIX_INFO fsUnixInfo; unsigned retry:1; unsigned nocase:1; @@ -262,14 +268,14 @@ struct cifsTconInfo { }; /* - * This info hangs off the cifsFileInfo structure. This is used to track - * byte stream locks on the file + * This info hangs off the cifsFileInfo structure, pointed to by llist. + * This is used to track byte stream locks on the file */ struct cifsLockInfo { - struct cifsLockInfo *next; - int start; - int length; - int type; + struct list_head llist; /* pointer to next cifsLockInfo */ + __u64 offset; + __u64 length; + __u8 type; }; /* @@ -300,12 +306,13 @@ struct cifsFileInfo { /* lock scope id (0 if none) */ struct file * pfile; /* needed for writepage */ struct inode * pInode; /* needed for oplock break */ + struct semaphore lock_sem; + struct list_head llist; /* list of byte range locks we have. */ unsigned closePend:1; /* file is marked to close */ unsigned invalidHandle:1; /* file closed via session abend */ atomic_t wrtPending; /* handle in use - defer close */ struct semaphore fh_sem; /* prevents reopen race after dead ses*/ char * search_resume_name; /* BB removeme BB */ - unsigned int resume_name_length; /* BB removeme - field renamed and moved BB */ struct cifs_search_info srch_inf; }; @@ -391,9 +398,9 @@ struct mid_q_entry { struct smb_hdr *resp_buf; /* response buffer */ int midState; /* wish this were enum but can not pass to wait_event */ __u8 command; /* smb command code */ - unsigned multiPart:1; /* multiple responses to one SMB request */ unsigned largeBuf:1; /* if valid response, is pointer to large buf */ - unsigned multiResp:1; /* multiple trans2 responses for one request */ + unsigned multiRsp:1; /* multiple trans2 responses for one request */ + unsigned multiEnd:1; /* both received */ }; struct oplock_q_entry { @@ -430,15 +437,35 @@ struct dir_notify_req { #define CIFS_LARGE_BUFFER 2 #define CIFS_IOVEC 4 /* array of response buffers */ -/* Type of session setup needed */ -#define CIFS_PLAINTEXT 0 -#define CIFS_LANMAN 1 -#define CIFS_NTLM 2 -#define CIFS_NTLMSSP_NEG 3 -#define CIFS_NTLMSSP_AUTH 4 -#define CIFS_SPNEGO_INIT 5 -#define CIFS_SPNEGO_TARG 6 - +/* Security Flags: indicate type of session setup needed */ +#define CIFSSEC_MAY_SIGN 0x00001 +#define CIFSSEC_MAY_NTLM 0x00002 +#define CIFSSEC_MAY_NTLMV2 0x00004 +#define CIFSSEC_MAY_KRB5 0x00008 +#ifdef CONFIG_CIFS_WEAK_PW_HASH +#define CIFSSEC_MAY_LANMAN 0x00010 +#define CIFSSEC_MAY_PLNTXT 0x00020 +#endif /* weak passwords */ +#define CIFSSEC_MAY_SEAL 0x00040 /* not supported yet */ + +#define CIFSSEC_MUST_SIGN 0x01001 +/* note that only one of the following can be set so the +result of setting MUST flags more than once will be to +require use of the stronger protocol */ +#define CIFSSEC_MUST_NTLM 0x02002 +#define CIFSSEC_MUST_NTLMV2 0x04004 +#define CIFSSEC_MUST_KRB5 0x08008 +#ifdef CONFIG_CIFS_WEAK_PW_HASH +#define CIFSSEC_MUST_LANMAN 0x10010 +#define CIFSSEC_MUST_PLNTXT 0x20020 +#define CIFSSEC_MASK 0x37037 /* current flags supported if weak */ +#else +#define CIFSSEC_MASK 0x07007 /* flags supported if no weak config */ +#endif /* WEAK_PW_HASH */ +#define CIFSSEC_MUST_SEAL 0x40040 /* not supported yet */ + +#define CIFSSEC_DEF CIFSSEC_MAY_SIGN | CIFSSEC_MAY_NTLM | CIFSSEC_MAY_NTLMV2 +#define CIFSSEC_MAX CIFSSEC_MUST_SIGN | CIFSSEC_MUST_NTLMV2 /* ***************************************************************** * All constants go here @@ -500,16 +527,16 @@ GLOBAL_EXTERN rwlock_t GlobalSMBSeslock; /* protects list inserts on 3 above */ GLOBAL_EXTERN struct list_head GlobalOplock_Q; GLOBAL_EXTERN struct list_head GlobalDnotifyReqList; /* Outstanding dir notify requests */ -GLOBAL_EXTERN struct list_head GlobalDnotifyRsp_Q; /* Dir notify response queue */ +GLOBAL_EXTERN struct list_head GlobalDnotifyRsp_Q;/* DirNotify response queue */ /* * Global transaction id (XID) information */ GLOBAL_EXTERN unsigned int GlobalCurrentXid; /* protected by GlobalMid_Sem */ -GLOBAL_EXTERN unsigned int GlobalTotalActiveXid; /* prot by GlobalMid_Sem */ +GLOBAL_EXTERN unsigned int GlobalTotalActiveXid; /* prot by GlobalMid_Sem */ GLOBAL_EXTERN unsigned int GlobalMaxActiveXid; /* prot by GlobalMid_Sem */ -GLOBAL_EXTERN spinlock_t GlobalMid_Lock; /* protects above and list operations */ - /* on midQ entries */ +GLOBAL_EXTERN spinlock_t GlobalMid_Lock; /* protects above & list operations */ + /* on midQ entries */ GLOBAL_EXTERN char Local_System_Name[15]; /* @@ -531,7 +558,7 @@ GLOBAL_EXTERN atomic_t smBufAllocCount; GLOBAL_EXTERN atomic_t midCount; /* Misc globals */ -GLOBAL_EXTERN unsigned int multiuser_mount; /* if enabled allows new sessions +GLOBAL_EXTERN unsigned int multiuser_mount; /* if enabled allows new sessions to be established on existing mount if we have the uid/password or Kerberos credential or equivalent for current user */ @@ -540,8 +567,8 @@ GLOBAL_EXTERN unsigned int experimEnabled; GLOBAL_EXTERN unsigned int lookupCacheEnabled; GLOBAL_EXTERN unsigned int extended_security; /* if on, session setup sent with more secure ntlmssp2 challenge/resp */ -GLOBAL_EXTERN unsigned int ntlmv2_support; /* better optional password hash */ GLOBAL_EXTERN unsigned int sign_CIFS_PDUs; /* enable smb packet signing */ +GLOBAL_EXTERN unsigned int secFlags; GLOBAL_EXTERN unsigned int linuxExtEnabled;/*enable Linux/Unix CIFS extensions*/ GLOBAL_EXTERN unsigned int CIFSMaxBufSize; /* max size not including hdr */ GLOBAL_EXTERN unsigned int cifs_min_rcv; /* min size of big ntwrk buf pool */ diff --git a/fs/cifs/cifspdu.h b/fs/cifs/cifspdu.h index b2233ac05bd..81df2bf8e75 100644 --- a/fs/cifs/cifspdu.h +++ b/fs/cifs/cifspdu.h @@ -16,7 +16,7 @@ * * 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 + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ #ifndef _CIFSPDU_H @@ -24,8 +24,14 @@ #include <net/sock.h> +#ifdef CONFIG_CIFS_WEAK_PW_HASH +#define LANMAN_PROT 0 +#define CIFS_PROT 1 +#else #define CIFS_PROT 0 -#define BAD_PROT CIFS_PROT+1 +#endif +#define POSIX_PROT CIFS_PROT+1 +#define BAD_PROT 0xFFFF /* SMB command codes */ /* Some commands have minimal (wct=0,bcc=0), or uninteresting, responses @@ -110,7 +116,7 @@ /* * Size of the session key (crypto key encrypted with the password */ -#define CIFS_SESSION_KEY_SIZE (24) +#define CIFS_SESS_KEY_SIZE (24) /* * Maximum user name length @@ -400,6 +406,29 @@ typedef struct negotiate_req { unsigned char DialectsArray[1]; } __attribute__((packed)) NEGOTIATE_REQ; +/* Dialect index is 13 for LANMAN */ + +typedef struct lanman_neg_rsp { + struct smb_hdr hdr; /* wct = 13 */ + __le16 DialectIndex; + __le16 SecurityMode; + __le16 MaxBufSize; + __le16 MaxMpxCount; + __le16 MaxNumberVcs; + __le16 RawMode; + __le32 SessionKey; + __le32 ServerTime; + __le16 ServerTimeZone; + __le16 EncryptionKeyLength; + __le16 Reserved; + __u16 ByteCount; + unsigned char EncryptionKey[1]; +} __attribute__((packed)) LANMAN_NEG_RSP; + +#define READ_RAW_ENABLE 1 +#define WRITE_RAW_ENABLE 2 +#define RAW_ENABLE (READ_RAW_ENABLE | WRITE_RAW_ENABLE) + typedef struct negotiate_rsp { struct smb_hdr hdr; /* wct = 17 */ __le16 DialectIndex; @@ -509,7 +538,7 @@ typedef union smb_com_session_setup_andx { /* unsigned char * NativeOS; */ /* unsigned char * NativeLanMan; */ /* unsigned char * PrimaryDomain; */ - } __attribute__((packed)) resp; /* NTLM response format (with or without extended security */ + } __attribute__((packed)) resp; /* NTLM response with or without extended sec*/ struct { /* request format */ struct smb_hdr hdr; /* wct = 10 */ @@ -520,8 +549,8 @@ typedef union smb_com_session_setup_andx { __le16 MaxMpxCount; __le16 VcNumber; __u32 SessionKey; - __le16 PassswordLength; - __u32 Reserved; + __le16 PasswordLength; + __u32 Reserved; /* encrypt key len and offset */ __le16 ByteCount; unsigned char AccountPassword[1]; /* followed by */ /* STRING AccountName */ @@ -543,6 +572,26 @@ typedef union smb_com_session_setup_andx { } __attribute__((packed)) old_resp; /* pre-NTLM (LANMAN2.1) response */ } __attribute__((packed)) SESSION_SETUP_ANDX; +/* format of NLTMv2 Response ie "case sensitive password" hash when NTLMv2 */ + +struct ntlmssp2_name { + __le16 type; + __le16 length; +/* char name[length]; */ +} __attribute__((packed)); + +struct ntlmv2_resp { + char ntlmv2_hash[CIFS_ENCPWD_SIZE]; + __le32 blob_signature; + __u32 reserved; + __le64 time; + __u64 client_chal; /* random */ + __u32 reserved2; + struct ntlmssp2_name names[1]; + /* array of name entries could follow ending in minimum 4 byte struct */ +} __attribute__((packed)); + + #define CIFS_NETWORK_OPSYS "CIFS VFS Client for Linux" /* Capabilities bits (for NTLM SessSetup request) */ @@ -573,7 +622,9 @@ typedef struct smb_com_tconx_req { } __attribute__((packed)) TCONX_REQ; typedef struct smb_com_tconx_rsp { - struct smb_hdr hdr; /* wct = 3 *//* note that Win2000 has sent wct=7 in some cases on responses. Four unspecified words followed OptionalSupport */ + struct smb_hdr hdr; /* wct = 3 note that Win2000 has sent wct = 7 + in some cases on responses. Four unspecified + words followed OptionalSupport */ __u8 AndXCommand; __u8 AndXReserved; __le16 AndXOffset; @@ -1293,6 +1344,7 @@ struct smb_t2_rsp { #define SMB_QUERY_ATTR_FLAGS 0x206 /* append,immutable etc. */ #define SMB_QUERY_POSIX_PERMISSION 0x207 #define SMB_QUERY_POSIX_LOCK 0x208 +/* #define SMB_POSIX_OPEN 0x209 */ #define SMB_QUERY_FILE_INTERNAL_INFO 0x3ee #define SMB_QUERY_FILE_ACCESS_INFO 0x3f0 #define SMB_QUERY_FILE_NAME_INFO2 0x3f1 /* 0x30 bytes */ @@ -1312,6 +1364,7 @@ struct smb_t2_rsp { #define SMB_SET_XATTR 0x205 #define SMB_SET_ATTR_FLAGS 0x206 /* append, immutable etc. */ #define SMB_SET_POSIX_LOCK 0x208 +#define SMB_POSIX_OPEN 0x209 #define SMB_SET_FILE_BASIC_INFO2 0x3ec #define SMB_SET_FILE_RENAME_INFORMATION 0x3f2 /* BB check if qpathinfo level too */ #define SMB_FILE_ALL_INFO2 0x3fa @@ -1323,6 +1376,9 @@ struct smb_t2_rsp { #define SMB_FILE_MAXIMUM_INFO 0x40d /* Find File infolevels */ +#define SMB_FIND_FILE_INFO_STANDARD 0x001 +#define SMB_FIND_FILE_QUERY_EA_SIZE 0x002 +#define SMB_FIND_FILE_QUERY_EAS_FROM_LIST 0x003 #define SMB_FIND_FILE_DIRECTORY_INFO 0x101 #define SMB_FIND_FILE_FULL_DIRECTORY_INFO 0x102 #define SMB_FIND_FILE_NAMES_INFO 0x103 @@ -1844,13 +1900,13 @@ typedef struct { typedef struct { __le32 DeviceType; __le32 DeviceCharacteristics; -} __attribute__((packed)) FILE_SYSTEM_DEVICE_INFO; /* device info, level 0x104 */ +} __attribute__((packed)) FILE_SYSTEM_DEVICE_INFO; /* device info level 0x104 */ typedef struct { __le32 Attributes; __le32 MaxPathNameComponentLength; __le32 FileSystemNameLen; - char FileSystemName[52]; /* do not really need to save this - so potentially get only subset of name */ + char FileSystemName[52]; /* do not have to save this - get subset? */ } __attribute__((packed)) FILE_SYSTEM_ATTRIBUTE_INFO; /******************************************************************************/ @@ -1947,7 +2003,8 @@ typedef struct { struct file_allocation_info { __le64 AllocationSize; /* Note old Samba srvr rounds this up too much */ -} __attribute__((packed)); /* size used on disk, level 0x103 for set, 0x105 for query */ +} __attribute__((packed)); /* size used on disk, for level 0x103 for set, + 0x105 for query */ struct file_end_of_file_info { __le64 FileSize; /* offset to end of file */ @@ -2054,7 +2111,7 @@ typedef struct { __le32 ExtFileAttributes; __le32 FileNameLength; char FileName[1]; -} __attribute__((packed)) FILE_DIRECTORY_INFO; /* level 0x101 FF response data area */ +} __attribute__((packed)) FILE_DIRECTORY_INFO; /* level 0x101 FF resp data */ typedef struct { __le32 NextEntryOffset; @@ -2069,7 +2126,7 @@ typedef struct { __le32 FileNameLength; __le32 EaSize; /* length of the xattrs */ char FileName[1]; -} __attribute__((packed)) FILE_FULL_DIRECTORY_INFO; /* level 0x102 FF response data area */ +} __attribute__((packed)) FILE_FULL_DIRECTORY_INFO; /* level 0x102 rsp data */ typedef struct { __le32 NextEntryOffset; @@ -2086,7 +2143,7 @@ typedef struct { __le32 Reserved; __u64 UniqueId; /* inode num - le since Samba puts ino in low 32 bit*/ char FileName[1]; -} __attribute__((packed)) SEARCH_ID_FULL_DIR_INFO; /* level 0x105 FF response data area */ +} __attribute__((packed)) SEARCH_ID_FULL_DIR_INFO; /* level 0x105 FF rsp data */ typedef struct { __le32 NextEntryOffset; @@ -2104,7 +2161,22 @@ typedef struct { __u8 Reserved; __u8 ShortName[12]; char FileName[1]; -} __attribute__((packed)) FILE_BOTH_DIRECTORY_INFO; /* level 0x104 FF response data area */ +} __attribute__((packed)) FILE_BOTH_DIRECTORY_INFO; /* level 0x104 FFrsp data */ + +typedef struct { + __u32 ResumeKey; + __le16 CreationDate; /* SMB Date */ + __le16 CreationTime; /* SMB Time */ + __le16 LastAccessDate; + __le16 LastAccessTime; + __le16 LastWriteDate; + __le16 LastWriteTime; + __le32 DataSize; /* File Size (EOF) */ + __le32 AllocationSize; + __le16 Attributes; /* verify not u32 */ + __u8 FileNameLength; + char FileName[1]; +} __attribute__((packed)) FIND_FILE_STANDARD_INFO; /* level 0x1 FF resp data */ struct win_dev { diff --git a/fs/cifs/cifsproto.h b/fs/cifs/cifsproto.h index 310ea2f0e0b..b35c55c3c8b 100644 --- a/fs/cifs/cifsproto.h +++ b/fs/cifs/cifsproto.h @@ -50,6 +50,10 @@ extern int SendReceive(const unsigned int /* xid */ , struct cifsSesInfo *, extern int SendReceive2(const unsigned int /* xid */ , struct cifsSesInfo *, struct kvec *, int /* nvec to send */, int * /* type of buf returned */ , const int long_op); +extern int SendReceiveBlockingLock(const unsigned int /* xid */ , struct cifsTconInfo *, + struct smb_hdr * /* input */ , + struct smb_hdr * /* out */ , + int * /* bytes returned */); extern int checkSMBhdr(struct smb_hdr *smb, __u16 mid); extern int checkSMB(struct smb_hdr *smb, __u16 mid, int length); extern int is_valid_oplock_break(struct smb_hdr *smb, struct TCP_Server_Info *); @@ -64,14 +68,12 @@ extern int map_smb_to_linux_error(struct smb_hdr *smb); extern void header_assemble(struct smb_hdr *, char /* command */ , const struct cifsTconInfo *, int /* length of fixed section (word count) in two byte units */); -#ifdef CONFIG_CIFS_EXPERIMENTAL extern int small_smb_init_no_tc(const int smb_cmd, const int wct, struct cifsSesInfo *ses, void ** request_buf); extern int CIFS_SessSetup(unsigned int xid, struct cifsSesInfo *ses, - const int stage, int * pNTLMv2_flg, + const int stage, const struct nls_table *nls_cp); -#endif extern __u16 GetNextMid(struct TCP_Server_Info *server); extern struct oplock_q_entry * AllocOplockQEntry(struct inode *, u16, struct cifsTconInfo *); @@ -285,8 +287,14 @@ extern int cifs_sign_smb2(struct kvec *iov, int n_vec, struct TCP_Server_Info *, extern int cifs_verify_signature(struct smb_hdr *, const char * mac_key, __u32 expected_sequence_number); extern int cifs_calculate_mac_key(char * key,const char * rn,const char * pass); -extern int CalcNTLMv2_partial_mac_key(struct cifsSesInfo *, struct nls_table *); -extern void CalcNTLMv2_response(const struct cifsSesInfo *,char * ); +extern int CalcNTLMv2_partial_mac_key(struct cifsSesInfo *, + const struct nls_table *); +extern void CalcNTLMv2_response(const struct cifsSesInfo *, char * ); +extern void setup_ntlmv2_rsp(struct cifsSesInfo *, char *, + const struct nls_table *); +#ifdef CONFIG_CIFS_WEAK_PW_HASH +extern void calc_lanman_hash(struct cifsSesInfo * ses, char * lnm_session_key); +#endif /* CIFS_WEAK_PW_HASH */ extern int CIFSSMBCopy(int xid, struct cifsTconInfo *source_tcon, const char *fromName, diff --git a/fs/cifs/cifssmb.c b/fs/cifs/cifssmb.c index 925881e00ff..075d8fb3d37 100644 --- a/fs/cifs/cifssmb.c +++ b/fs/cifs/cifssmb.c @@ -44,8 +44,11 @@ static struct { int index; char *name; } protocols[] = { +#ifdef CONFIG_CIFS_WEAK_PW_HASH + {LANMAN_PROT, "\2LM1.2X002"}, +#endif /* weak password hashing for legacy clients */ {CIFS_PROT, "\2NT LM 0.12"}, - {CIFS_PROT, "\2POSIX 2"}, + {POSIX_PROT, "\2POSIX 2"}, {BAD_PROT, "\2"} }; #else @@ -53,11 +56,29 @@ static struct { int index; char *name; } protocols[] = { +#ifdef CONFIG_CIFS_WEAK_PW_HASH + {LANMAN_PROT, "\2LM1.2X002"}, +#endif /* weak password hashing for legacy clients */ {CIFS_PROT, "\2NT LM 0.12"}, {BAD_PROT, "\2"} }; #endif +/* define the number of elements in the cifs dialect array */ +#ifdef CONFIG_CIFS_POSIX +#ifdef CONFIG_CIFS_WEAK_PW_HASH +#define CIFS_NUM_PROT 3 +#else +#define CIFS_NUM_PROT 2 +#endif /* CIFS_WEAK_PW_HASH */ +#else /* not posix */ +#ifdef CONFIG_CIFS_WEAK_PW_HASH +#define CIFS_NUM_PROT 2 +#else +#define CIFS_NUM_PROT 1 +#endif /* CONFIG_CIFS_WEAK_PW_HASH */ +#endif /* CIFS_POSIX */ + /* Mark as invalid, all open files on tree connections since they were closed when session to server was lost */ @@ -188,7 +209,6 @@ small_smb_init(int smb_command, int wct, struct cifsTconInfo *tcon, return rc; } -#ifdef CONFIG_CIFS_EXPERIMENTAL int small_smb_init_no_tc(const int smb_command, const int wct, struct cifsSesInfo *ses, void **request_buf) @@ -214,7 +234,6 @@ small_smb_init_no_tc(const int smb_command, const int wct, return rc; } -#endif /* CONFIG_CIFS_EXPERIMENTAL */ /* If the return code is zero, this function must fill in request_buf pointer */ static int @@ -322,7 +341,8 @@ smb_init(int smb_command, int wct, struct cifsTconInfo *tcon, /* potential retries of smb operations it turns out we can determine */ /* from the mid flags when the request buffer can be resent without */ /* having to use a second distinct buffer for the response */ - *response_buf = *request_buf; + if(response_buf) + *response_buf = *request_buf; header_assemble((struct smb_hdr *) *request_buf, smb_command, tcon, wct /*wct */ ); @@ -373,8 +393,10 @@ CIFSSMBNegotiate(unsigned int xid, struct cifsSesInfo *ses) NEGOTIATE_RSP *pSMBr; int rc = 0; int bytes_returned; + int i; struct TCP_Server_Info * server; u16 count; + unsigned int secFlags; if(ses->server) server = ses->server; @@ -386,101 +408,200 @@ CIFSSMBNegotiate(unsigned int xid, struct cifsSesInfo *ses) (void **) &pSMB, (void **) &pSMBr); if (rc) return rc; + + /* if any of auth flags (ie not sign or seal) are overriden use them */ + if(ses->overrideSecFlg & (~(CIFSSEC_MUST_SIGN | CIFSSEC_MUST_SEAL))) + secFlags = ses->overrideSecFlg; + else /* if override flags set only sign/seal OR them with global auth */ + secFlags = extended_security | ses->overrideSecFlg; + + cFYI(1,("secFlags 0x%x",secFlags)); + pSMB->hdr.Mid = GetNextMid(server); pSMB->hdr.Flags2 |= SMBFLG2_UNICODE; - if (extended_security) + if((secFlags & CIFSSEC_MUST_KRB5) == CIFSSEC_MUST_KRB5) pSMB->hdr.Flags2 |= SMBFLG2_EXT_SEC; - - count = strlen(protocols[0].name) + 1; - strncpy(pSMB->DialectsArray, protocols[0].name, 30); - /* null guaranteed to be at end of source and target buffers anyway */ - + + count = 0; + for(i=0;i<CIFS_NUM_PROT;i++) { + strncpy(pSMB->DialectsArray+count, protocols[i].name, 16); + count += strlen(protocols[i].name) + 1; + /* null at end of source and target buffers anyway */ + } pSMB->hdr.smb_buf_length += count; pSMB->ByteCount = cpu_to_le16(count); rc = SendReceive(xid, ses, (struct smb_hdr *) pSMB, (struct smb_hdr *) pSMBr, &bytes_returned, 0); - if (rc == 0) { - server->secMode = pSMBr->SecurityMode; - if((server->secMode & SECMODE_USER) == 0) - cFYI(1,("share mode security")); - server->secType = NTLM; /* BB override default for - NTLMv2 or kerberos v5 */ - /* one byte - no need to convert this or EncryptionKeyLen - from little endian */ - server->maxReq = le16_to_cpu(pSMBr->MaxMpxCount); - /* probably no need to store and check maxvcs */ - server->maxBuf = - min(le32_to_cpu(pSMBr->MaxBufferSize), + if (rc != 0) + goto neg_err_exit; + + cFYI(1,("Dialect: %d", pSMBr->DialectIndex)); + /* Check wct = 1 error case */ + if((pSMBr->hdr.WordCount < 13) || (pSMBr->DialectIndex == BAD_PROT)) { + /* core returns wct = 1, but we do not ask for core - otherwise + small wct just comes when dialect index is -1 indicating we + could not negotiate a common dialect */ + rc = -EOPNOTSUPP; + goto neg_err_exit; +#ifdef CONFIG_CIFS_WEAK_PW_HASH + } else if((pSMBr->hdr.WordCount == 13) + && (pSMBr->DialectIndex == LANMAN_PROT)) { + struct lanman_neg_rsp * rsp = (struct lanman_neg_rsp *)pSMBr; + + if((secFlags & CIFSSEC_MAY_LANMAN) || + (secFlags & CIFSSEC_MAY_PLNTXT)) + server->secType = LANMAN; + else { + cERROR(1, ("mount failed weak security disabled" + " in /proc/fs/cifs/SecurityFlags")); + rc = -EOPNOTSUPP; + goto neg_err_exit; + } + server->secMode = (__u8)le16_to_cpu(rsp->SecurityMode); + server->maxReq = le16_to_cpu(rsp->MaxMpxCount); + server->maxBuf = min((__u32)le16_to_cpu(rsp->MaxBufSize), + (__u32)CIFSMaxBufSize + MAX_CIFS_HDR_SIZE); + GETU32(server->sessid) = le32_to_cpu(rsp->SessionKey); + /* even though we do not use raw we might as well set this + accurately, in case we ever find a need for it */ + if((le16_to_cpu(rsp->RawMode) & RAW_ENABLE) == RAW_ENABLE) { + server->maxRw = 0xFF00; + server->capabilities = CAP_MPX_MODE | CAP_RAW_MODE; + } else { + server->maxRw = 0;/* we do not need to use raw anyway */ + server->capabilities = CAP_MPX_MODE; + } + server->timeZone = le16_to_cpu(rsp->ServerTimeZone); + + /* BB get server time for time conversions and add + code to use it and timezone since this is not UTC */ + + if (rsp->EncryptionKeyLength == cpu_to_le16(CIFS_CRYPTO_KEY_SIZE)) { + memcpy(server->cryptKey, rsp->EncryptionKey, + CIFS_CRYPTO_KEY_SIZE); + } else if (server->secMode & SECMODE_PW_ENCRYPT) { + rc = -EIO; /* need cryptkey unless plain text */ + goto neg_err_exit; + } + + cFYI(1,("LANMAN negotiated")); + /* we will not end up setting signing flags - as no signing + was in LANMAN and server did not return the flags on */ + goto signing_check; +#else /* weak security disabled */ + } else if(pSMBr->hdr.WordCount == 13) { + cERROR(1,("mount failed, cifs module not built " + "with CIFS_WEAK_PW_HASH support")); + rc = -EOPNOTSUPP; +#endif /* WEAK_PW_HASH */ + goto neg_err_exit; + } else if(pSMBr->hdr.WordCount != 17) { + /* unknown wct */ + rc = -EOPNOTSUPP; + goto neg_err_exit; + } + /* else wct == 17 NTLM */ + server->secMode = pSMBr->SecurityMode; + if((server->secMode & SECMODE_USER) == 0) + cFYI(1,("share mode security")); + + if((server->secMode & SECMODE_PW_ENCRYPT) == 0) +#ifdef CONFIG_CIFS_WEAK_PW_HASH + if ((secFlags & CIFSSEC_MAY_PLNTXT) == 0) +#endif /* CIFS_WEAK_PW_HASH */ + cERROR(1,("Server requests plain text password" + " but client support disabled")); + + if((secFlags & CIFSSEC_MUST_NTLMV2) == CIFSSEC_MUST_NTLMV2) + server->secType = NTLMv2; + else if(secFlags & CIFSSEC_MAY_NTLM) + server->secType = NTLM; + else if(secFlags & CIFSSEC_MAY_NTLMV2) + server->secType = NTLMv2; + /* else krb5 ... any others ... */ + + /* one byte, so no need to convert this or EncryptionKeyLen from + little endian */ + server->maxReq = le16_to_cpu(pSMBr->MaxMpxCount); + /* probably no need to store and check maxvcs */ + server->maxBuf = min(le32_to_cpu(pSMBr->MaxBufferSize), (__u32) CIFSMaxBufSize + MAX_CIFS_HDR_SIZE); - server->maxRw = le32_to_cpu(pSMBr->MaxRawSize); - cFYI(0, ("Max buf = %d", ses->server->maxBuf)); - GETU32(ses->server->sessid) = le32_to_cpu(pSMBr->SessionKey); - server->capabilities = le32_to_cpu(pSMBr->Capabilities); - server->timeZone = le16_to_cpu(pSMBr->ServerTimeZone); - /* BB with UTC do we ever need to be using srvr timezone? */ - if (pSMBr->EncryptionKeyLength == CIFS_CRYPTO_KEY_SIZE) { - memcpy(server->cryptKey, pSMBr->u.EncryptionKey, - CIFS_CRYPTO_KEY_SIZE); - } else if ((pSMBr->hdr.Flags2 & SMBFLG2_EXT_SEC) - && (pSMBr->EncryptionKeyLength == 0)) { - /* decode security blob */ - } else - rc = -EIO; + server->maxRw = le32_to_cpu(pSMBr->MaxRawSize); + cFYI(0, ("Max buf = %d", ses->server->maxBuf)); + GETU32(ses->server->sessid) = le32_to_cpu(pSMBr->SessionKey); + server->capabilities = le32_to_cpu(pSMBr->Capabilities); + server->timeZone = le16_to_cpu(pSMBr->ServerTimeZone); + if (pSMBr->EncryptionKeyLength == CIFS_CRYPTO_KEY_SIZE) { + memcpy(server->cryptKey, pSMBr->u.EncryptionKey, + CIFS_CRYPTO_KEY_SIZE); + } else if ((pSMBr->hdr.Flags2 & SMBFLG2_EXT_SEC) + && (pSMBr->EncryptionKeyLength == 0)) { + /* decode security blob */ + } else if (server->secMode & SECMODE_PW_ENCRYPT) { + rc = -EIO; /* no crypt key only if plain text pwd */ + goto neg_err_exit; + } - /* BB might be helpful to save off the domain of server here */ + /* BB might be helpful to save off the domain of server here */ - if ((pSMBr->hdr.Flags2 & SMBFLG2_EXT_SEC) && - (server->capabilities & CAP_EXTENDED_SECURITY)) { - count = pSMBr->ByteCount; - if (count < 16) - rc = -EIO; - else if (count == 16) { - server->secType = RawNTLMSSP; - if (server->socketUseCount.counter > 1) { - if (memcmp - (server->server_GUID, - pSMBr->u.extended_response. - GUID, 16) != 0) { - cFYI(1, ("server UID changed")); - memcpy(server-> - server_GUID, - pSMBr->u. - extended_response. - GUID, 16); - } - } else + if ((pSMBr->hdr.Flags2 & SMBFLG2_EXT_SEC) && + (server->capabilities & CAP_EXTENDED_SECURITY)) { + count = pSMBr->ByteCount; + if (count < 16) + rc = -EIO; + else if (count == 16) { + server->secType = RawNTLMSSP; + if (server->socketUseCount.counter > 1) { + if (memcmp(server->server_GUID, + pSMBr->u.extended_response. + GUID, 16) != 0) { + cFYI(1, ("server UID changed")); memcpy(server->server_GUID, - pSMBr->u.extended_response. - GUID, 16); - } else { - rc = decode_negTokenInit(pSMBr->u. - extended_response. - SecurityBlob, - count - 16, - &server->secType); - if(rc == 1) { - /* BB Need to fill struct for sessetup here */ - rc = -EOPNOTSUPP; - } else { - rc = -EINVAL; + pSMBr->u.extended_response.GUID, + 16); } + } else + memcpy(server->server_GUID, + pSMBr->u.extended_response.GUID, 16); + } else { + rc = decode_negTokenInit(pSMBr->u.extended_response. + SecurityBlob, + count - 16, + &server->secType); + if(rc == 1) { + /* BB Need to fill struct for sessetup here */ + rc = -EOPNOTSUPP; + } else { + rc = -EINVAL; } - } else - server->capabilities &= ~CAP_EXTENDED_SECURITY; - if(sign_CIFS_PDUs == FALSE) { - if(server->secMode & SECMODE_SIGN_REQUIRED) - cERROR(1, - ("Server requires /proc/fs/cifs/PacketSigningEnabled")); - server->secMode &= ~(SECMODE_SIGN_ENABLED | SECMODE_SIGN_REQUIRED); - } else if(sign_CIFS_PDUs == 1) { - if((server->secMode & SECMODE_SIGN_REQUIRED) == 0) - server->secMode &= ~(SECMODE_SIGN_ENABLED | SECMODE_SIGN_REQUIRED); } - + } else + server->capabilities &= ~CAP_EXTENDED_SECURITY; + +#ifdef CONFIG_CIFS_WEAK_PW_HASH +signing_check: +#endif + if(sign_CIFS_PDUs == FALSE) { + if(server->secMode & SECMODE_SIGN_REQUIRED) + cERROR(1,("Server requires " + "/proc/fs/cifs/PacketSigningEnabled to be on")); + server->secMode &= + ~(SECMODE_SIGN_ENABLED | SECMODE_SIGN_REQUIRED); + } else if(sign_CIFS_PDUs == 1) { + if((server->secMode & SECMODE_SIGN_REQUIRED) == 0) + server->secMode &= + ~(SECMODE_SIGN_ENABLED | SECMODE_SIGN_REQUIRED); + } else if(sign_CIFS_PDUs == 2) { + if((server->secMode & + (SECMODE_SIGN_ENABLED | SECMODE_SIGN_REQUIRED)) == 0) { + cERROR(1,("signing required but server lacks support")); + } } - +neg_err_exit: cifs_buf_release(pSMB); + + cFYI(1,("negprot rc %d",rc)); return rc; } @@ -1339,8 +1460,13 @@ CIFSSMBLock(const int xid, struct cifsTconInfo *tcon, pSMB->hdr.smb_buf_length += count; pSMB->ByteCount = cpu_to_le16(count); - rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, + if (waitFlag) { + rc = SendReceiveBlockingLock(xid, tcon, (struct smb_hdr *) pSMB, + (struct smb_hdr *) pSMBr, &bytes_returned); + } else { + rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, (struct smb_hdr *) pSMBr, &bytes_returned, timeout); + } cifs_stats_inc(&tcon->num_locks); if (rc) { cFYI(1, ("Send error in Lock = %d", rc)); @@ -1363,6 +1489,7 @@ CIFSSMBPosixLock(const int xid, struct cifsTconInfo *tcon, char *data_offset; struct cifs_posix_lock *parm_data; int rc = 0; + int timeout = 0; int bytes_returned = 0; __u16 params, param_offset, offset, byte_count, count; @@ -1382,7 +1509,6 @@ CIFSSMBPosixLock(const int xid, struct cifsTconInfo *tcon, pSMB->MaxSetupCount = 0; pSMB->Reserved = 0; pSMB->Flags = 0; - pSMB->Timeout = 0; pSMB->Reserved2 = 0; param_offset = offsetof(struct smb_com_transaction2_sfi_req, Fid) - 4; offset = param_offset + params; @@ -1408,8 +1534,13 @@ CIFSSMBPosixLock(const int xid, struct cifsTconInfo *tcon, (((char *) &pSMB->hdr.Protocol) + offset); parm_data->lock_type = cpu_to_le16(lock_type); - if(waitFlag) + if(waitFlag) { + timeout = 3; /* blocking operation, no timeout */ parm_data->lock_flags = cpu_to_le16(1); + pSMB->Timeout = cpu_to_le32(-1); + } else + pSMB->Timeout = 0; + parm_data->pid = cpu_to_le32(current->tgid); parm_data->start = cpu_to_le64(pLockData->fl_start); parm_data->length = cpu_to_le64(len); /* normalize negative numbers */ @@ -1420,8 +1551,14 @@ CIFSSMBPosixLock(const int xid, struct cifsTconInfo *tcon, pSMB->Reserved4 = 0; pSMB->hdr.smb_buf_length += byte_count; pSMB->ByteCount = cpu_to_le16(byte_count); - rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, - (struct smb_hdr *) pSMBr, &bytes_returned, 0); + if (waitFlag) { + rc = SendReceiveBlockingLock(xid, tcon, (struct smb_hdr *) pSMB, + (struct smb_hdr *) pSMBr, &bytes_returned); + } else { + rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, + (struct smb_hdr *) pSMBr, &bytes_returned, timeout); + } + if (rc) { cFYI(1, ("Send error in Posix Lock = %d", rc)); } else if (get_flag) { @@ -2239,7 +2376,7 @@ CIFSSMBQueryReparseLinkInfo(const int xid, struct cifsTconInfo *tcon, } symlinkinfo[buflen] = 0; /* just in case so the caller does not go off the end of the buffer */ - cFYI(1,("readlink result - %s ",symlinkinfo)); + cFYI(1,("readlink result - %s",symlinkinfo)); } } qreparse_out: diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index bae1479318d..0e9ba0b9d71 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -49,8 +49,6 @@ static DECLARE_COMPLETION(cifsd_complete); -extern void SMBencrypt(unsigned char *passwd, unsigned char *c8, - unsigned char *p24); extern void SMBNTencrypt(unsigned char *passwd, unsigned char *c8, unsigned char *p24); @@ -70,6 +68,7 @@ struct smb_vol { gid_t linux_gid; mode_t file_mode; mode_t dir_mode; + unsigned secFlg; unsigned rw:1; unsigned retry:1; unsigned intr:1; @@ -83,18 +82,14 @@ struct smb_vol { unsigned remap:1; /* set to remap seven reserved chars in filenames */ unsigned posix_paths:1; /* unset to not ask for posix pathnames. */ unsigned sfu_emul:1; - unsigned krb5:1; - unsigned ntlm:1; - unsigned ntlmv2:1; unsigned nullauth:1; /* attempt to authenticate with null user */ - unsigned sign:1; - unsigned seal:1; /* encrypt */ unsigned nocase; /* request case insensitive filenames */ unsigned nobrl; /* disable sending byte range locks to srv */ unsigned int rsize; unsigned int wsize; unsigned int sockopt; unsigned short int port; + char * prepath; }; static int ipv4_connect(struct sockaddr_in *psin_server, @@ -188,6 +183,7 @@ cifs_reconnect(struct TCP_Server_Info *server) while ((server->tcpStatus != CifsExiting) && (server->tcpStatus != CifsGood)) { + try_to_freeze(); if(server->protocolType == IPV6) { rc = ipv6_connect(&server->addr.sockAddr6,&server->ssocket); } else { @@ -369,21 +365,21 @@ cifs_demultiplex_thread(struct TCP_Server_Info *server) continue; if (bigbuf == NULL) { bigbuf = cifs_buf_get(); - if(bigbuf == NULL) { - cERROR(1,("No memory for large SMB response")); + if (!bigbuf) { + cERROR(1, ("No memory for large SMB response")); msleep(3000); /* retry will check if exiting */ continue; } - } else if(isLargeBuf) { - /* we are reusing a dirtry large buf, clear its start */ + } else if (isLargeBuf) { + /* we are reusing a dirty large buf, clear its start */ memset(bigbuf, 0, sizeof (struct smb_hdr)); } if (smallbuf == NULL) { smallbuf = cifs_small_buf_get(); - if(smallbuf == NULL) { - cERROR(1,("No memory for SMB response")); + if (!smallbuf) { + cERROR(1, ("No memory for SMB response")); msleep(1000); /* retry will check if exiting */ continue; @@ -403,12 +399,12 @@ cifs_demultiplex_thread(struct TCP_Server_Info *server) kernel_recvmsg(csocket, &smb_msg, &iov, 1, 4, 0 /* BB see socket.h flags */); - if(server->tcpStatus == CifsExiting) { + if (server->tcpStatus == CifsExiting) { break; } else if (server->tcpStatus == CifsNeedReconnect) { - cFYI(1,("Reconnect after server stopped responding")); + cFYI(1, ("Reconnect after server stopped responding")); cifs_reconnect(server); - cFYI(1,("call to reconnect done")); + cFYI(1, ("call to reconnect done")); csocket = server->ssocket; continue; } else if ((length == -ERESTARTSYS) || (length == -EAGAIN)) { @@ -417,15 +413,15 @@ cifs_demultiplex_thread(struct TCP_Server_Info *server) tcpStatus CifsNeedReconnect if server hung */ continue; } else if (length <= 0) { - if(server->tcpStatus == CifsNew) { - cFYI(1,("tcp session abend after SMBnegprot")); + if (server->tcpStatus == CifsNew) { + cFYI(1, ("tcp session abend after SMBnegprot")); /* some servers kill the TCP session rather than returning an SMB negprot error, in which case reconnecting here is not going to help, and so simply return error to mount */ break; } - if(length == -EINTR) { + if (!try_to_freeze() && (length == -EINTR)) { cFYI(1,("cifsd thread killed")); break; } @@ -585,9 +581,11 @@ cifs_demultiplex_thread(struct TCP_Server_Info *server) /* merge response - fix up 1st*/ if(coalesce_t2(smb_buffer, mid_entry->resp_buf)) { + mid_entry->multiRsp = 1; break; } else { /* all parts received */ + mid_entry->multiEnd = 1; goto multi_t2_fnd; } } else { @@ -616,6 +614,10 @@ multi_t2_fnd: #ifdef CONFIG_CIFS_STATS2 mid_entry->when_received = jiffies; #endif + /* so we do not time out requests to server + which is still responding (since server could + be busy but not dead) */ + server->lstrp = jiffies; break; } } @@ -632,9 +634,14 @@ multi_t2_fnd: wake_up_process(task_to_wake); } else if ((is_valid_oplock_break(smb_buffer, server) == FALSE) && (isMultiRsp == FALSE)) { - cERROR(1, ("No task to wake, unknown frame rcvd!")); + cERROR(1, ("No task to wake, unknown frame rcvd! NumMids %d", midCount.counter)); cifs_dump_mem("Received Data is: ",(char *)smb_buffer, sizeof(struct smb_hdr)); +#ifdef CONFIG_CIFS_DEBUG2 + cifs_dump_detail(smb_buffer); + cifs_dump_mids(server); +#endif /* CIFS_DEBUG2 */ + } } /* end while !EXITING */ @@ -784,7 +791,6 @@ cifs_parse_mount_options(char *options, const char *devname,struct smb_vol *vol) /* vol->retry default is 0 (i.e. "soft" limited retry not hard retry) */ vol->rw = TRUE; - vol->ntlm = TRUE; /* default is always to request posix paths. */ vol->posix_paths = 1; @@ -915,30 +921,35 @@ cifs_parse_mount_options(char *options, const char *devname,struct smb_vol *vol) cERROR(1,("no security value specified")); continue; } else if (strnicmp(value, "krb5i", 5) == 0) { - vol->sign = 1; - vol->krb5 = 1; + vol->secFlg |= CIFSSEC_MAY_KRB5 | + CIFSSEC_MUST_SIGN; } else if (strnicmp(value, "krb5p", 5) == 0) { - /* vol->seal = 1; - vol->krb5 = 1; */ + /* vol->secFlg |= CIFSSEC_MUST_SEAL | + CIFSSEC_MAY_KRB5; */ cERROR(1,("Krb5 cifs privacy not supported")); return 1; } else if (strnicmp(value, "krb5", 4) == 0) { - vol->krb5 = 1; + vol->secFlg |= CIFSSEC_MAY_KRB5; } else if (strnicmp(value, "ntlmv2i", 7) == 0) { - vol->ntlmv2 = 1; - vol->sign = 1; + vol->secFlg |= CIFSSEC_MAY_NTLMV2 | + CIFSSEC_MUST_SIGN; } else if (strnicmp(value, "ntlmv2", 6) == 0) { - vol->ntlmv2 = 1; + vol->secFlg |= CIFSSEC_MAY_NTLMV2; } else if (strnicmp(value, "ntlmi", 5) == 0) { - vol->ntlm = 1; - vol->sign = 1; + vol->secFlg |= CIFSSEC_MAY_NTLM | + CIFSSEC_MUST_SIGN; } else if (strnicmp(value, "ntlm", 4) == 0) { /* ntlm is default so can be turned off too */ - vol->ntlm = 1; + vol->secFlg |= CIFSSEC_MAY_NTLM; } else if (strnicmp(value, "nontlm", 6) == 0) { - vol->ntlm = 0; + /* BB is there a better way to do this? */ + vol->secFlg |= CIFSSEC_MAY_NTLMV2; +#ifdef CONFIG_CIFS_WEAK_PW_HASH + } else if (strnicmp(value, "lanman", 6) == 0) { + vol->secFlg |= CIFSSEC_MAY_LANMAN; +#endif } else if (strnicmp(value, "none", 4) == 0) { - vol->nullauth = 1; + vol->nullauth = 1; } else { cERROR(1,("bad security option: %s", value)); return 1; @@ -976,13 +987,35 @@ cifs_parse_mount_options(char *options, const char *devname,struct smb_vol *vol) } /* BB are there cases in which a comma can be valid in a domain name and need special handling? */ - if (strnlen(value, 65) < 65) { + if (strnlen(value, 256) < 256) { vol->domainname = value; cFYI(1, ("Domain name set")); } else { printk(KERN_WARNING "CIFS: domain name too long\n"); return 1; } + } else if (strnicmp(data, "prefixpath", 10) == 0) { + if (!value || !*value) { + printk(KERN_WARNING + "CIFS: invalid path prefix\n"); + return 1; /* needs_arg; */ + } + if ((temp_len = strnlen(value, 1024)) < 1024) { + if(value[0] != '/') + temp_len++; /* missing leading slash */ + vol->prepath = kmalloc(temp_len+1,GFP_KERNEL); + if(vol->prepath == NULL) + return 1; + if(value[0] != '/') { + vol->prepath[0] = '/'; + strcpy(vol->prepath+1,value); + } else + strcpy(vol->prepath,value); + cFYI(1,("prefix path %s",vol->prepath)); + } else { + printk(KERN_WARNING "CIFS: prefix too long\n"); + return 1; + } } else if (strnicmp(data, "iocharset", 9) == 0) { if (!value || !*value) { printk(KERN_WARNING "CIFS: invalid iocharset specified\n"); @@ -1168,6 +1201,10 @@ cifs_parse_mount_options(char *options, const char *devname,struct smb_vol *vol) vol->no_psx_acl = 0; } else if (strnicmp(data, "noacl",5) == 0) { vol->no_psx_acl = 1; + } else if (strnicmp(data, "sign",4) == 0) { + vol->secFlg |= CIFSSEC_MUST_SIGN; +/* } else if (strnicmp(data, "seal",4) == 0) { + vol->secFlg |= CIFSSEC_MUST_SEAL; */ } else if (strnicmp(data, "direct",6) == 0) { vol->direct_io = 1; } else if (strnicmp(data, "forcedirectio",13) == 0) { @@ -1257,33 +1294,35 @@ find_unc(__be32 new_target_ip_addr, char *uncName, char *userName) read_lock(&GlobalSMBSeslock); list_for_each(tmp, &GlobalTreeConnectionList) { - cFYI(1, ("Next tcon - ")); + cFYI(1, ("Next tcon")); tcon = list_entry(tmp, struct cifsTconInfo, cifsConnectionList); if (tcon->ses) { if (tcon->ses->server) { cFYI(1, - (" old ip addr: %x == new ip %x ?", + ("old ip addr: %x == new ip %x ?", tcon->ses->server->addr.sockAddr.sin_addr. s_addr, new_target_ip_addr)); if (tcon->ses->server->addr.sockAddr.sin_addr. s_addr == new_target_ip_addr) { - /* BB lock tcon and server and tcp session and increment use count here? */ + /* BB lock tcon, server and tcp session and increment use count here? */ /* found a match on the TCP session */ /* BB check if reconnection needed */ - cFYI(1,("Matched ip, old UNC: %s == new: %s ?", + cFYI(1,("IP match, old UNC: %s new: %s", tcon->treeName, uncName)); if (strncmp (tcon->treeName, uncName, MAX_TREE_SIZE) == 0) { cFYI(1, - ("Matched UNC, old user: %s == new: %s ?", + ("and old usr: %s new: %s", tcon->treeName, uncName)); if (strncmp (tcon->ses->userName, userName, MAX_USERNAME_SIZE) == 0) { read_unlock(&GlobalSMBSeslock); - return tcon;/* also matched user (smb session)*/ + /* matched smb session + (user name */ + return tcon; } } } @@ -1589,6 +1628,7 @@ cifs_mount(struct super_block *sb, struct cifs_sb_info *cifs_sb, if (cifs_parse_mount_options(mount_data, devname, &volume_info)) { kfree(volume_info.UNC); kfree(volume_info.password); + kfree(volume_info.prepath); FreeXid(xid); return -EINVAL; } @@ -1603,6 +1643,7 @@ cifs_mount(struct super_block *sb, struct cifs_sb_info *cifs_sb, locations such as env variables and files on disk */ kfree(volume_info.UNC); kfree(volume_info.password); + kfree(volume_info.prepath); FreeXid(xid); return -EINVAL; } @@ -1623,6 +1664,7 @@ cifs_mount(struct super_block *sb, struct cifs_sb_info *cifs_sb, /* we failed translating address */ kfree(volume_info.UNC); kfree(volume_info.password); + kfree(volume_info.prepath); FreeXid(xid); return -EINVAL; } @@ -1635,6 +1677,7 @@ cifs_mount(struct super_block *sb, struct cifs_sb_info *cifs_sb, cERROR(1,("Connecting to DFS root not implemented yet")); kfree(volume_info.UNC); kfree(volume_info.password); + kfree(volume_info.prepath); FreeXid(xid); return -EINVAL; } else /* which servers DFS root would we conect to */ { @@ -1642,6 +1685,7 @@ cifs_mount(struct super_block *sb, struct cifs_sb_info *cifs_sb, ("CIFS mount error: No UNC path (e.g. -o unc=//192.168.1.100/public) specified")); kfree(volume_info.UNC); kfree(volume_info.password); + kfree(volume_info.prepath); FreeXid(xid); return -EINVAL; } @@ -1656,6 +1700,7 @@ cifs_mount(struct super_block *sb, struct cifs_sb_info *cifs_sb, cERROR(1,("CIFS mount error: iocharset %s not found",volume_info.iocharset)); kfree(volume_info.UNC); kfree(volume_info.password); + kfree(volume_info.prepath); FreeXid(xid); return -ELIBACC; } @@ -1672,6 +1717,7 @@ cifs_mount(struct super_block *sb, struct cifs_sb_info *cifs_sb, else { kfree(volume_info.UNC); kfree(volume_info.password); + kfree(volume_info.prepath); FreeXid(xid); return -EINVAL; } @@ -1694,6 +1740,7 @@ cifs_mount(struct super_block *sb, struct cifs_sb_info *cifs_sb, sock_release(csocket); kfree(volume_info.UNC); kfree(volume_info.password); + kfree(volume_info.prepath); FreeXid(xid); return rc; } @@ -1704,6 +1751,7 @@ cifs_mount(struct super_block *sb, struct cifs_sb_info *cifs_sb, sock_release(csocket); kfree(volume_info.UNC); kfree(volume_info.password); + kfree(volume_info.prepath); FreeXid(xid); return rc; } else { @@ -1728,6 +1776,7 @@ cifs_mount(struct super_block *sb, struct cifs_sb_info *cifs_sb, sock_release(csocket); kfree(volume_info.UNC); kfree(volume_info.password); + kfree(volume_info.prepath); FreeXid(xid); return rc; } @@ -1762,11 +1811,18 @@ cifs_mount(struct super_block *sb, struct cifs_sb_info *cifs_sb, if (volume_info.username) strncpy(pSesInfo->userName, volume_info.username,MAX_USERNAME_SIZE); - if (volume_info.domainname) - strncpy(pSesInfo->domainName, - volume_info.domainname,MAX_USERNAME_SIZE); + if (volume_info.domainname) { + int len = strlen(volume_info.domainname); + pSesInfo->domainName = + kmalloc(len + 1, GFP_KERNEL); + if(pSesInfo->domainName) + strcpy(pSesInfo->domainName, + volume_info.domainname); + } pSesInfo->linux_uid = volume_info.linux_uid; + pSesInfo->overrideSecFlg = volume_info.secFlg; down(&pSesInfo->sesSem); + /* BB FIXME need to pass vol->secFlgs BB */ rc = cifs_setup_session(xid,pSesInfo, cifs_sb->local_nls); up(&pSesInfo->sesSem); if(!rc) @@ -1808,6 +1864,14 @@ cifs_mount(struct super_block *sb, struct cifs_sb_info *cifs_sb, /* Windows ME may prefer this */ cFYI(1,("readsize set to minimum 2048")); } + /* calculate prepath */ + cifs_sb->prepath = volume_info.prepath; + if(cifs_sb->prepath) { + cifs_sb->prepathlen = strlen(cifs_sb->prepath); + cifs_sb->prepath[0] = CIFS_DIR_SEP(cifs_sb); + volume_info.prepath = NULL; + } else + cifs_sb->prepathlen = 0; cifs_sb->mnt_uid = volume_info.linux_uid; cifs_sb->mnt_gid = volume_info.linux_gid; cifs_sb->mnt_file_mode = volume_info.file_mode; @@ -1953,7 +2017,18 @@ cifs_mount(struct super_block *sb, struct cifs_sb_info *cifs_sb, } cFYI(1,("Negotiate caps 0x%x",(int)cap)); - +#ifdef CONFIG_CIFS_DEBUG2 + if(cap & CIFS_UNIX_FCNTL_CAP) + cFYI(1,("FCNTL cap")); + if(cap & CIFS_UNIX_EXTATTR_CAP) + cFYI(1,("EXTATTR cap")); + if(cap & CIFS_UNIX_POSIX_PATHNAMES_CAP) + cFYI(1,("POSIX path cap")); + if(cap & CIFS_UNIX_XATTR_CAP) + cFYI(1,("XATTR cap")); + if(cap & CIFS_UNIX_POSIX_ACL_CAP) + cFYI(1,("POSIX ACL cap")); +#endif /* CIFS_DEBUG2 */ if (CIFSSMBSetFSUnixInfo(xid, tcon, cap)) { cFYI(1,("setting capabilities failed")); } @@ -1974,13 +2049,14 @@ cifs_mount(struct super_block *sb, struct cifs_sb_info *cifs_sb, the password ptr is put in the new session structure (in which case the password will be freed at unmount time) */ kfree(volume_info.UNC); + kfree(volume_info.prepath); FreeXid(xid); return rc; } static int CIFSSessSetup(unsigned int xid, struct cifsSesInfo *ses, - char session_key[CIFS_SESSION_KEY_SIZE], + char session_key[CIFS_SESS_KEY_SIZE], const struct nls_table *nls_codepage) { struct smb_hdr *smb_buffer; @@ -2038,15 +2114,15 @@ CIFSSessSetup(unsigned int xid, struct cifsSesInfo *ses, pSMB->req_no_secext.Capabilities = cpu_to_le32(capabilities); pSMB->req_no_secext.CaseInsensitivePasswordLength = - cpu_to_le16(CIFS_SESSION_KEY_SIZE); + cpu_to_le16(CIFS_SESS_KEY_SIZE); pSMB->req_no_secext.CaseSensitivePasswordLength = - cpu_to_le16(CIFS_SESSION_KEY_SIZE); + cpu_to_le16(CIFS_SESS_KEY_SIZE); bcc_ptr = pByteArea(smb_buffer); - memcpy(bcc_ptr, (char *) session_key, CIFS_SESSION_KEY_SIZE); - bcc_ptr += CIFS_SESSION_KEY_SIZE; - memcpy(bcc_ptr, (char *) session_key, CIFS_SESSION_KEY_SIZE); - bcc_ptr += CIFS_SESSION_KEY_SIZE; + memcpy(bcc_ptr, (char *) session_key, CIFS_SESS_KEY_SIZE); + bcc_ptr += CIFS_SESS_KEY_SIZE; + memcpy(bcc_ptr, (char *) session_key, CIFS_SESS_KEY_SIZE); + bcc_ptr += CIFS_SESS_KEY_SIZE; if (ses->capabilities & CAP_UNICODE) { if ((long) bcc_ptr % 2) { /* must be word aligned for Unicode */ @@ -2054,7 +2130,7 @@ CIFSSessSetup(unsigned int xid, struct cifsSesInfo *ses, bcc_ptr++; } if(user == NULL) - bytes_returned = 0; /* skill null user */ + bytes_returned = 0; /* skip null user */ else bytes_returned = cifs_strtoUCS((__le16 *) bcc_ptr, user, 100, @@ -2162,8 +2238,7 @@ CIFSSessSetup(unsigned int xid, struct cifsSesInfo *ses, if (remaining_words > 0) { len = UniStrnlen((wchar_t *)bcc_ptr, remaining_words-1); - if(ses->serverNOS) - kfree(ses->serverNOS); + kfree(ses->serverNOS); ses->serverNOS = kzalloc(2 * (len + 1),GFP_KERNEL); if(ses->serverNOS == NULL) goto sesssetup_nomem; @@ -2203,12 +2278,10 @@ CIFSSessSetup(unsigned int xid, struct cifsSesInfo *ses, /* if these kcallocs fail not much we can do, but better to not fail the sesssetup itself */ - if(ses->serverDomain) - kfree(ses->serverDomain); + kfree(ses->serverDomain); ses->serverDomain = kzalloc(2, GFP_KERNEL); - if(ses->serverNOS) - kfree(ses->serverNOS); + kfree(ses->serverNOS); ses->serverNOS = kzalloc(2, GFP_KERNEL); } @@ -2217,8 +2290,7 @@ CIFSSessSetup(unsigned int xid, struct cifsSesInfo *ses, if (((long) bcc_ptr + len) - (long) pByteArea(smb_buffer_response) <= BCC(smb_buffer_response)) { - if(ses->serverOS) - kfree(ses->serverOS); + kfree(ses->serverOS); ses->serverOS = kzalloc(len + 1,GFP_KERNEL); if(ses->serverOS == NULL) goto sesssetup_nomem; @@ -2229,8 +2301,7 @@ CIFSSessSetup(unsigned int xid, struct cifsSesInfo *ses, bcc_ptr++; len = strnlen(bcc_ptr, 1024); - if(ses->serverNOS) - kfree(ses->serverNOS); + kfree(ses->serverNOS); ses->serverNOS = kzalloc(len + 1,GFP_KERNEL); if(ses->serverNOS == NULL) goto sesssetup_nomem; @@ -2274,292 +2345,6 @@ sesssetup_nomem: /* do not return an error on nomem for the info strings, } static int -CIFSSpnegoSessSetup(unsigned int xid, struct cifsSesInfo *ses, - char *SecurityBlob,int SecurityBlobLength, - const struct nls_table *nls_codepage) -{ - struct smb_hdr *smb_buffer; - struct smb_hdr *smb_buffer_response; - SESSION_SETUP_ANDX *pSMB; - SESSION_SETUP_ANDX *pSMBr; - char *bcc_ptr; - char *user; - char *domain; - int rc = 0; - int remaining_words = 0; - int bytes_returned = 0; - int len; - __u32 capabilities; - __u16 count; - - cFYI(1, ("In spnego sesssetup ")); - if(ses == NULL) - return -EINVAL; - user = ses->userName; - domain = ses->domainName; - - smb_buffer = cifs_buf_get(); - if (smb_buffer == NULL) { - return -ENOMEM; - } - smb_buffer_response = smb_buffer; - pSMBr = pSMB = (SESSION_SETUP_ANDX *) smb_buffer; - - /* send SMBsessionSetup here */ - header_assemble(smb_buffer, SMB_COM_SESSION_SETUP_ANDX, - NULL /* no tCon exists yet */ , 12 /* wct */ ); - - smb_buffer->Mid = GetNextMid(ses->server); - pSMB->req.hdr.Flags2 |= SMBFLG2_EXT_SEC; - pSMB->req.AndXCommand = 0xFF; - if(ses->server->maxBuf > 64*1024) - ses->server->maxBuf = (64*1023); - pSMB->req.MaxBufferSize = cpu_to_le16(ses->server->maxBuf); - pSMB->req.MaxMpxCount = cpu_to_le16(ses->server->maxReq); - - if(ses->server->secMode & (SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED)) - smb_buffer->Flags2 |= SMBFLG2_SECURITY_SIGNATURE; - - capabilities = CAP_LARGE_FILES | CAP_NT_SMBS | CAP_LEVEL_II_OPLOCKS | - CAP_EXTENDED_SECURITY; - if (ses->capabilities & CAP_UNICODE) { - smb_buffer->Flags2 |= SMBFLG2_UNICODE; - capabilities |= CAP_UNICODE; - } - if (ses->capabilities & CAP_STATUS32) { - smb_buffer->Flags2 |= SMBFLG2_ERR_STATUS; - capabilities |= CAP_STATUS32; - } - if (ses->capabilities & CAP_DFS) { - smb_buffer->Flags2 |= SMBFLG2_DFS; - capabilities |= CAP_DFS; - } - pSMB->req.Capabilities = cpu_to_le32(capabilities); - - pSMB->req.SecurityBlobLength = cpu_to_le16(SecurityBlobLength); - bcc_ptr = pByteArea(smb_buffer); - memcpy(bcc_ptr, SecurityBlob, SecurityBlobLength); - bcc_ptr += SecurityBlobLength; - - if (ses->capabilities & CAP_UNICODE) { - if ((long) bcc_ptr % 2) { /* must be word aligned for Unicode strings */ - *bcc_ptr = 0; - bcc_ptr++; - } - bytes_returned = - cifs_strtoUCS((__le16 *) bcc_ptr, user, 100, nls_codepage); - bcc_ptr += 2 * bytes_returned; /* convert num of 16 bit words to bytes */ - bcc_ptr += 2; /* trailing null */ - if (domain == NULL) - bytes_returned = - cifs_strtoUCS((__le16 *) bcc_ptr, - "CIFS_LINUX_DOM", 32, nls_codepage); - else - bytes_returned = - cifs_strtoUCS((__le16 *) bcc_ptr, domain, 64, - nls_codepage); - bcc_ptr += 2 * bytes_returned; - bcc_ptr += 2; - bytes_returned = - cifs_strtoUCS((__le16 *) bcc_ptr, "Linux version ", - 32, nls_codepage); - bcc_ptr += 2 * bytes_returned; - bytes_returned = - cifs_strtoUCS((__le16 *) bcc_ptr, system_utsname.release, 32, - nls_codepage); - bcc_ptr += 2 * bytes_returned; - bcc_ptr += 2; - bytes_returned = - cifs_strtoUCS((__le16 *) bcc_ptr, CIFS_NETWORK_OPSYS, - 64, nls_codepage); - bcc_ptr += 2 * bytes_returned; - bcc_ptr += 2; - } else { - strncpy(bcc_ptr, user, 200); - bcc_ptr += strnlen(user, 200); - *bcc_ptr = 0; - bcc_ptr++; - if (domain == NULL) { - strcpy(bcc_ptr, "CIFS_LINUX_DOM"); - bcc_ptr += strlen("CIFS_LINUX_DOM") + 1; - } else { - strncpy(bcc_ptr, domain, 64); - bcc_ptr += strnlen(domain, 64); - *bcc_ptr = 0; - bcc_ptr++; - } - strcpy(bcc_ptr, "Linux version "); - bcc_ptr += strlen("Linux version "); - strcpy(bcc_ptr, system_utsname.release); - bcc_ptr += strlen(system_utsname.release) + 1; - strcpy(bcc_ptr, CIFS_NETWORK_OPSYS); - bcc_ptr += strlen(CIFS_NETWORK_OPSYS) + 1; - } - count = (long) bcc_ptr - (long) pByteArea(smb_buffer); - smb_buffer->smb_buf_length += count; - pSMB->req.ByteCount = cpu_to_le16(count); - - rc = SendReceive(xid, ses, smb_buffer, smb_buffer_response, - &bytes_returned, 1); - if (rc) { -/* rc = map_smb_to_linux_error(smb_buffer_response); *//* done in SendReceive now */ - } else if ((smb_buffer_response->WordCount == 3) - || (smb_buffer_response->WordCount == 4)) { - __u16 action = le16_to_cpu(pSMBr->resp.Action); - __u16 blob_len = - le16_to_cpu(pSMBr->resp.SecurityBlobLength); - if (action & GUEST_LOGIN) - cFYI(1, (" Guest login")); /* BB do we want to set anything in SesInfo struct ? */ - if (ses) { - ses->Suid = smb_buffer_response->Uid; /* UID left in wire format (le) */ - cFYI(1, ("UID = %d ", ses->Suid)); - bcc_ptr = pByteArea(smb_buffer_response); /* response can have either 3 or 4 word count - Samba sends 3 */ - - /* BB Fix below to make endian neutral !! */ - - if ((pSMBr->resp.hdr.WordCount == 3) - || ((pSMBr->resp.hdr.WordCount == 4) - && (blob_len < - pSMBr->resp.ByteCount))) { - if (pSMBr->resp.hdr.WordCount == 4) { - bcc_ptr += - blob_len; - cFYI(1, - ("Security Blob Length %d ", - blob_len)); - } - - if (smb_buffer->Flags2 & SMBFLG2_UNICODE) { - if ((long) (bcc_ptr) % 2) { - remaining_words = - (BCC(smb_buffer_response) - - 1) / 2; - bcc_ptr++; /* Unicode strings must be word aligned */ - } else { - remaining_words = - BCC - (smb_buffer_response) / 2; - } - len = - UniStrnlen((wchar_t *) bcc_ptr, - remaining_words - 1); -/* We look for obvious messed up bcc or strings in response so we do not go off - the end since (at least) WIN2K and Windows XP have a major bug in not null - terminating last Unicode string in response */ - if(ses->serverOS) - kfree(ses->serverOS); - ses->serverOS = - kzalloc(2 * (len + 1), GFP_KERNEL); - cifs_strfromUCS_le(ses->serverOS, - (__le16 *) - bcc_ptr, len, - nls_codepage); - bcc_ptr += 2 * (len + 1); - remaining_words -= len + 1; - ses->serverOS[2 * len] = 0; - ses->serverOS[1 + (2 * len)] = 0; - if (remaining_words > 0) { - len = UniStrnlen((wchar_t *)bcc_ptr, - remaining_words - - 1); - if(ses->serverNOS) - kfree(ses->serverNOS); - ses->serverNOS = - kzalloc(2 * (len + 1), - GFP_KERNEL); - cifs_strfromUCS_le(ses->serverNOS, - (__le16 *)bcc_ptr, - len, - nls_codepage); - bcc_ptr += 2 * (len + 1); - ses->serverNOS[2 * len] = 0; - ses->serverNOS[1 + (2 * len)] = 0; - remaining_words -= len + 1; - if (remaining_words > 0) { - len = UniStrnlen((wchar_t *) bcc_ptr, remaining_words); - /* last string not null terminated (e.g.Windows XP/2000) */ - if(ses->serverDomain) - kfree(ses->serverDomain); - ses->serverDomain = kzalloc(2*(len+1),GFP_KERNEL); - cifs_strfromUCS_le(ses->serverDomain, - (__le16 *)bcc_ptr, - len, nls_codepage); - bcc_ptr += 2*(len+1); - ses->serverDomain[2*len] = 0; - ses->serverDomain[1+(2*len)] = 0; - } /* else no more room so create dummy domain string */ - else { - if(ses->serverDomain) - kfree(ses->serverDomain); - ses->serverDomain = - kzalloc(2,GFP_KERNEL); - } - } else {/* no room use dummy domain&NOS */ - if(ses->serverDomain) - kfree(ses->serverDomain); - ses->serverDomain = kzalloc(2, GFP_KERNEL); - if(ses->serverNOS) - kfree(ses->serverNOS); - ses->serverNOS = kzalloc(2, GFP_KERNEL); - } - } else { /* ASCII */ - - len = strnlen(bcc_ptr, 1024); - if (((long) bcc_ptr + len) - (long) - pByteArea(smb_buffer_response) - <= BCC(smb_buffer_response)) { - if(ses->serverOS) - kfree(ses->serverOS); - ses->serverOS = kzalloc(len + 1, GFP_KERNEL); - strncpy(ses->serverOS, bcc_ptr, len); - - bcc_ptr += len; - bcc_ptr[0] = 0; /* null terminate the string */ - bcc_ptr++; - - len = strnlen(bcc_ptr, 1024); - if(ses->serverNOS) - kfree(ses->serverNOS); - ses->serverNOS = kzalloc(len + 1,GFP_KERNEL); - strncpy(ses->serverNOS, bcc_ptr, len); - bcc_ptr += len; - bcc_ptr[0] = 0; - bcc_ptr++; - - len = strnlen(bcc_ptr, 1024); - if(ses->serverDomain) - kfree(ses->serverDomain); - ses->serverDomain = kzalloc(len + 1, GFP_KERNEL); - strncpy(ses->serverDomain, bcc_ptr, len); - bcc_ptr += len; - bcc_ptr[0] = 0; - bcc_ptr++; - } else - cFYI(1, - ("Variable field of length %d extends beyond end of smb ", - len)); - } - } else { - cERROR(1, - (" Security Blob Length extends beyond end of SMB")); - } - } else { - cERROR(1, ("No session structure passed in.")); - } - } else { - cERROR(1, - (" Invalid Word count %d: ", - smb_buffer_response->WordCount)); - rc = -EIO; - } - - if (smb_buffer) - cifs_buf_release(smb_buffer); - - return rc; -} - -static int CIFSNTLMSSPNegotiateSessSetup(unsigned int xid, struct cifsSesInfo *ses, int * pNTLMv2_flag, const struct nls_table *nls_codepage) @@ -2635,8 +2420,8 @@ CIFSNTLMSSPNegotiateSessSetup(unsigned int xid, /* NTLMSSP_NEGOTIATE_ALWAYS_SIGN | */ NTLMSSP_NEGOTIATE_128; if(sign_CIFS_PDUs) negotiate_flags |= NTLMSSP_NEGOTIATE_SIGN; - if(ntlmv2_support) - negotiate_flags |= NTLMSSP_NEGOTIATE_NTLMV2; +/* if(ntlmv2_support) + negotiate_flags |= NTLMSSP_NEGOTIATE_NTLMV2;*/ /* setup pointers to domain name and workstation name */ bcc_ptr += SecurityBlobLength; @@ -2783,8 +2568,7 @@ CIFSNTLMSSPNegotiateSessSetup(unsigned int xid, bcc_ptr, remaining_words - 1); - if(ses->serverNOS) - kfree(ses->serverNOS); + kfree(ses->serverNOS); ses->serverNOS = kzalloc(2 * (len + 1), GFP_KERNEL); @@ -2802,8 +2586,7 @@ CIFSNTLMSSPNegotiateSessSetup(unsigned int xid, if (remaining_words > 0) { len = UniStrnlen((wchar_t *) bcc_ptr, remaining_words); /* last string is not always null terminated (for e.g. for Windows XP & 2000) */ - if(ses->serverDomain) - kfree(ses->serverDomain); + kfree(ses->serverDomain); ses->serverDomain = kzalloc(2 * (len + @@ -2822,19 +2605,16 @@ CIFSNTLMSSPNegotiateSessSetup(unsigned int xid, = 0; } /* else no more room so create dummy domain string */ else { - if(ses->serverDomain) - kfree(ses->serverDomain); + kfree(ses->serverDomain); ses->serverDomain = kzalloc(2, GFP_KERNEL); } } else { /* no room so create dummy domain and NOS string */ - if(ses->serverDomain); - kfree(ses->serverDomain); + kfree(ses->serverDomain); ses->serverDomain = kzalloc(2, GFP_KERNEL); - if(ses->serverNOS) - kfree(ses->serverNOS); + kfree(ses->serverNOS); ses->serverNOS = kzalloc(2, GFP_KERNEL); } @@ -2856,8 +2636,7 @@ CIFSNTLMSSPNegotiateSessSetup(unsigned int xid, bcc_ptr++; len = strnlen(bcc_ptr, 1024); - if(ses->serverNOS) - kfree(ses->serverNOS); + kfree(ses->serverNOS); ses->serverNOS = kzalloc(len + 1, GFP_KERNEL); @@ -2867,8 +2646,7 @@ CIFSNTLMSSPNegotiateSessSetup(unsigned int xid, bcc_ptr++; len = strnlen(bcc_ptr, 1024); - if(ses->serverDomain) - kfree(ses->serverDomain); + kfree(ses->serverDomain); ses->serverDomain = kzalloc(len + 1, GFP_KERNEL); @@ -2994,14 +2772,14 @@ CIFSNTLMSSPAuthSessSetup(unsigned int xid, struct cifsSesInfo *ses, SecurityBlob->LmChallengeResponse.Buffer = 0; SecurityBlob->NtChallengeResponse.Length = - cpu_to_le16(CIFS_SESSION_KEY_SIZE); + cpu_to_le16(CIFS_SESS_KEY_SIZE); SecurityBlob->NtChallengeResponse.MaximumLength = - cpu_to_le16(CIFS_SESSION_KEY_SIZE); - memcpy(bcc_ptr, ntlm_session_key, CIFS_SESSION_KEY_SIZE); + cpu_to_le16(CIFS_SESS_KEY_SIZE); + memcpy(bcc_ptr, ntlm_session_key, CIFS_SESS_KEY_SIZE); SecurityBlob->NtChallengeResponse.Buffer = cpu_to_le32(SecurityBlobLength); - SecurityBlobLength += CIFS_SESSION_KEY_SIZE; - bcc_ptr += CIFS_SESSION_KEY_SIZE; + SecurityBlobLength += CIFS_SESS_KEY_SIZE; + bcc_ptr += CIFS_SESS_KEY_SIZE; if (ses->capabilities & CAP_UNICODE) { if (domain == NULL) { @@ -3190,8 +2968,7 @@ CIFSNTLMSSPAuthSessSetup(unsigned int xid, struct cifsSesInfo *ses, bcc_ptr, remaining_words - 1); - if(ses->serverNOS) - kfree(ses->serverNOS); + kfree(ses->serverNOS); ses->serverNOS = kzalloc(2 * (len + 1), GFP_KERNEL); @@ -3244,8 +3021,7 @@ CIFSNTLMSSPAuthSessSetup(unsigned int xid, struct cifsSesInfo *ses, if(ses->serverDomain) kfree(ses->serverDomain); ses->serverDomain = kzalloc(2, GFP_KERNEL); - if(ses->serverNOS) - kfree(ses->serverNOS); + kfree(ses->serverNOS); ses->serverNOS = kzalloc(2, GFP_KERNEL); } } else { /* ASCII */ @@ -3263,8 +3039,7 @@ CIFSNTLMSSPAuthSessSetup(unsigned int xid, struct cifsSesInfo *ses, bcc_ptr++; len = strnlen(bcc_ptr, 1024); - if(ses->serverNOS) - kfree(ses->serverNOS); + kfree(ses->serverNOS); ses->serverNOS = kzalloc(len+1,GFP_KERNEL); strncpy(ses->serverNOS, bcc_ptr, len); bcc_ptr += len; @@ -3340,22 +3115,33 @@ CIFSTCon(unsigned int xid, struct cifsSesInfo *ses, bcc_ptr = &pSMB->Password[0]; if((ses->server->secMode) & SECMODE_USER) { pSMB->PasswordLength = cpu_to_le16(1); /* minimum */ + *bcc_ptr = 0; /* password is null byte */ bcc_ptr++; /* skip password */ + /* already aligned so no need to do it below */ } else { - pSMB->PasswordLength = cpu_to_le16(CIFS_SESSION_KEY_SIZE); + pSMB->PasswordLength = cpu_to_le16(CIFS_SESS_KEY_SIZE); /* BB FIXME add code to fail this if NTLMv2 or Kerberos specified as required (when that support is added to the vfs in the future) as only NTLM or the much - weaker LANMAN (which we do not send) is accepted + weaker LANMAN (which we do not send by default) is accepted by Samba (not sure whether other servers allow NTLMv2 password here) */ +#ifdef CONFIG_CIFS_WEAK_PW_HASH + if((extended_security & CIFSSEC_MAY_LANMAN) && + (ses->server->secType == LANMAN)) + calc_lanman_hash(ses, bcc_ptr); + else +#endif /* CIFS_WEAK_PW_HASH */ SMBNTencrypt(ses->password, ses->server->cryptKey, bcc_ptr); - bcc_ptr += CIFS_SESSION_KEY_SIZE; - *bcc_ptr = 0; - bcc_ptr++; /* align */ + bcc_ptr += CIFS_SESS_KEY_SIZE; + if(ses->capabilities & CAP_UNICODE) { + /* must align unicode strings */ + *bcc_ptr = 0; /* null byte password */ + bcc_ptr++; + } } if(ses->server->secMode & @@ -3429,7 +3215,10 @@ CIFSTCon(unsigned int xid, struct cifsSesInfo *ses, } /* else do not bother copying these informational fields */ } - tcon->Flags = le16_to_cpu(pSMBr->OptionalSupport); + if(smb_buffer_response->WordCount == 3) + tcon->Flags = le16_to_cpu(pSMBr->OptionalSupport); + else + tcon->Flags = 0; cFYI(1, ("Tcon flags: 0x%x ", tcon->Flags)); } else if ((rc == 0) && tcon == NULL) { /* all we need to save for IPC$ connection */ @@ -3448,6 +3237,7 @@ cifs_umount(struct super_block *sb, struct cifs_sb_info *cifs_sb) int xid; struct cifsSesInfo *ses = NULL; struct task_struct *cifsd_task; + char * tmp; xid = GetXid(); @@ -3481,6 +3271,10 @@ cifs_umount(struct super_block *sb, struct cifs_sb_info *cifs_sb) } cifs_sb->tcon = NULL; + tmp = cifs_sb->prepath; + cifs_sb->prepathlen = 0; + cifs_sb->prepath = NULL; + kfree(tmp); if (ses) schedule_timeout_interruptible(msecs_to_jiffies(500)); if (ses) @@ -3494,7 +3288,7 @@ int cifs_setup_session(unsigned int xid, struct cifsSesInfo *pSesInfo, struct nls_table * nls_info) { int rc = 0; - char ntlm_session_key[CIFS_SESSION_KEY_SIZE]; + char ntlm_session_key[CIFS_SESS_KEY_SIZE]; int ntlmv2_flag = FALSE; int first_time = 0; @@ -3526,20 +3320,13 @@ int cifs_setup_session(unsigned int xid, struct cifsSesInfo *pSesInfo, pSesInfo->server->secMode, pSesInfo->server->capabilities, pSesInfo->server->timeZone)); -#ifdef CONFIG_CIFS_EXPERIMENTAL - if(experimEnabled > 1) - rc = CIFS_SessSetup(xid, pSesInfo, CIFS_NTLM /* type */, - &ntlmv2_flag, nls_info); - else -#endif - if (extended_security + if(experimEnabled < 2) + rc = CIFS_SessSetup(xid, pSesInfo, + first_time, nls_info); + else if (extended_security && (pSesInfo->capabilities & CAP_EXTENDED_SECURITY) && (pSesInfo->server->secType == NTLMSSP)) { - cFYI(1, ("New style sesssetup")); - rc = CIFSSpnegoSessSetup(xid, pSesInfo, - NULL /* security blob */, - 0 /* blob length */, - nls_info); + rc = -EOPNOTSUPP; } else if (extended_security && (pSesInfo->capabilities & CAP_EXTENDED_SECURITY) && (pSesInfo->server->secType == RawNTLMSSP)) { diff --git a/fs/cifs/dir.c b/fs/cifs/dir.c index 82315edc77d..66b825ade3e 100644 --- a/fs/cifs/dir.c +++ b/fs/cifs/dir.c @@ -46,7 +46,8 @@ char * build_path_from_dentry(struct dentry *direntry) { struct dentry *temp; - int namelen = 0; + int namelen; + int pplen; char *full_path; char dirsep; @@ -56,7 +57,9 @@ build_path_from_dentry(struct dentry *direntry) when the server crashed */ dirsep = CIFS_DIR_SEP(CIFS_SB(direntry->d_sb)); + pplen = CIFS_SB(direntry->d_sb)->prepathlen; cifs_bp_rename_retry: + namelen = pplen; for (temp = direntry; !IS_ROOT(temp);) { namelen += (1 + temp->d_name.len); temp = temp->d_parent; @@ -70,7 +73,6 @@ cifs_bp_rename_retry: if(full_path == NULL) return full_path; full_path[namelen] = 0; /* trailing null */ - for (temp = direntry; !IS_ROOT(temp);) { namelen -= 1 + temp->d_name.len; if (namelen < 0) { @@ -79,7 +81,7 @@ cifs_bp_rename_retry: full_path[namelen] = dirsep; strncpy(full_path + namelen + 1, temp->d_name.name, temp->d_name.len); - cFYI(0, (" name: %s ", full_path + namelen)); + cFYI(0, ("name: %s", full_path + namelen)); } temp = temp->d_parent; if(temp == NULL) { @@ -88,18 +90,23 @@ cifs_bp_rename_retry: return NULL; } } - if (namelen != 0) { + if (namelen != pplen) { cERROR(1, - ("We did not end path lookup where we expected namelen is %d", + ("did not end path lookup where expected namelen is %d", namelen)); - /* presumably this is only possible if we were racing with a rename + /* presumably this is only possible if racing with a rename of one of the parent directories (we can not lock the dentries above us to prevent this, but retrying should be harmless) */ kfree(full_path); - namelen = 0; goto cifs_bp_rename_retry; } - + /* DIR_SEP already set for byte 0 / vs \ but not for + subsequent slashes in prepath which currently must + be entered the right way - not sure if there is an alternative + since the '\' is a valid posix character so we can not switch + those safely to '/' if any are found in the middle of the prepath */ + /* BB test paths to Windows with '/' in the midst of prepath */ + strncpy(full_path,CIFS_SB(direntry->d_sb)->prepath,pplen); return full_path; } @@ -113,7 +120,7 @@ cifs_bp_rename_retry: full_path[namelen+2] = 0; BB remove above eight lines BB */ -/* Inode operations in similar order to how they appear in the Linux file fs.h */ +/* Inode operations in similar order to how they appear in Linux file fs.h */ int cifs_create(struct inode *inode, struct dentry *direntry, int mode, @@ -178,11 +185,14 @@ cifs_create(struct inode *inode, struct dentry *direntry, int mode, FreeXid(xid); return -ENOMEM; } - - rc = CIFSSMBOpen(xid, pTcon, full_path, disposition, + if (cifs_sb->tcon->ses->capabilities & CAP_NT_SMBS) + rc = CIFSSMBOpen(xid, pTcon, full_path, disposition, desiredAccess, CREATE_NOT_DIR, &fileHandle, &oplock, buf, cifs_sb->local_nls, cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR); + else + rc = -EIO; /* no NT SMB support fall into legacy open below */ + if(rc == -EIO) { /* old server, retry the open legacy style */ rc = SMBLegacyOpen(xid, pTcon, full_path, disposition, @@ -191,7 +201,7 @@ cifs_create(struct inode *inode, struct dentry *direntry, int mode, cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR); } if (rc) { - cFYI(1, ("cifs_create returned 0x%x ", rc)); + cFYI(1, ("cifs_create returned 0x%x", rc)); } else { /* If Open reported that we actually created a file then we now have to set the mode if possible */ @@ -264,6 +274,10 @@ cifs_create(struct inode *inode, struct dentry *direntry, int mode, pCifsFile->invalidHandle = FALSE; pCifsFile->closePend = FALSE; init_MUTEX(&pCifsFile->fh_sem); + init_MUTEX(&pCifsFile->lock_sem); + INIT_LIST_HEAD(&pCifsFile->llist); + atomic_set(&pCifsFile->wrtPending,0); + /* set the following in open now pCifsFile->pfile = file; */ write_lock(&GlobalSMBSeslock); @@ -369,6 +383,10 @@ int cifs_mknod(struct inode *inode, struct dentry *direntry, int mode, cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR); + /* BB FIXME - add handling for backlevel servers + which need legacy open and check for all + calls to SMBOpen for fallback to + SMBLeagcyOpen */ if(!rc) { /* BB Do not bother to decode buf since no local inode yet to put timestamps in, diff --git a/fs/cifs/fcntl.c b/fs/cifs/fcntl.c index 633a9381132..d91a3d44e9e 100644 --- a/fs/cifs/fcntl.c +++ b/fs/cifs/fcntl.c @@ -91,14 +91,14 @@ int cifs_dir_notify(struct file * file, unsigned long arg) if(full_path == NULL) { rc = -ENOMEM; } else { - cERROR(1,("cifs dir notify on file %s with arg 0x%lx",full_path,arg)); /* BB removeme BB */ + cFYI(1,("dir notify on file %s Arg 0x%lx",full_path,arg)); rc = CIFSSMBOpen(xid, pTcon, full_path, FILE_OPEN, GENERIC_READ | SYNCHRONIZE, 0 /* create options */, &netfid, &oplock,NULL, cifs_sb->local_nls, cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR); /* BB fixme - add this handle to a notify handle list */ if(rc) { - cERROR(1,("Could not open directory for notify")); /* BB remove BB */ + cFYI(1,("Could not open directory for notify")); } else { filter = convert_to_cifs_notify_flags(arg); if(filter != 0) { diff --git a/fs/cifs/file.c b/fs/cifs/file.c index b4a18c1cab0..ddb012a6802 100644 --- a/fs/cifs/file.c +++ b/fs/cifs/file.c @@ -5,6 +5,7 @@ * * Copyright (C) International Business Machines Corp., 2002,2003 * Author(s): Steve French (sfrench@us.ibm.com) + * Jeremy Allison (jra@samba.org) * * 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 @@ -47,6 +48,8 @@ static inline struct cifsFileInfo *cifs_init_private( private_data->netfid = netfid; private_data->pid = current->tgid; init_MUTEX(&private_data->fh_sem); + init_MUTEX(&private_data->lock_sem); + INIT_LIST_HEAD(&private_data->llist); private_data->pfile = file; /* needed for writepage */ private_data->pInode = inode; private_data->invalidHandle = FALSE; @@ -110,7 +113,6 @@ static inline int cifs_open_inode_helper(struct inode *inode, struct file *file, &pCifsInode->openFileList); } write_unlock(&GlobalSMBSeslock); - write_unlock(&file->f_owner.lock); if (pCifsInode->clientCanCacheRead) { /* we have the inode open somewhere else no need to discard cache data */ @@ -201,7 +203,7 @@ int cifs_open(struct inode *inode, struct file *file) } else { if (file->f_flags & O_EXCL) cERROR(1, ("could not find file instance for " - "new file %p ", file)); + "new file %p", file)); } } @@ -260,10 +262,15 @@ int cifs_open(struct inode *inode, struct file *file) rc = -ENOMEM; goto out; } - rc = CIFSSMBOpen(xid, pTcon, full_path, disposition, desiredAccess, - CREATE_NOT_DIR, &netfid, &oplock, buf, + + if (cifs_sb->tcon->ses->capabilities & CAP_NT_SMBS) + rc = CIFSSMBOpen(xid, pTcon, full_path, disposition, + desiredAccess, CREATE_NOT_DIR, &netfid, &oplock, buf, cifs_sb->local_nls, cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR); + else + rc = -EIO; /* no NT SMB support fall into legacy open below */ + if (rc == -EIO) { /* Old server, try legacy style OpenX */ rc = SMBLegacyOpen(xid, pTcon, full_path, disposition, @@ -272,7 +279,7 @@ int cifs_open(struct inode *inode, struct file *file) & CIFS_MOUNT_MAP_SPECIAL_CHR); } if (rc) { - cFYI(1, ("cifs_open returned 0x%x ", rc)); + cFYI(1, ("cifs_open returned 0x%x", rc)); goto out; } file->private_data = @@ -282,7 +289,6 @@ int cifs_open(struct inode *inode, struct file *file) goto out; } pCifsFile = cifs_init_private(file->private_data, inode, file, netfid); - write_lock(&file->f_owner.lock); write_lock(&GlobalSMBSeslock); list_add(&pCifsFile->tlist, &pTcon->openFileList); @@ -293,7 +299,6 @@ int cifs_open(struct inode *inode, struct file *file) &oplock, buf, full_path, xid); } else { write_unlock(&GlobalSMBSeslock); - write_unlock(&file->f_owner.lock); } if (oplock & CIFS_CREATE_ACTION) { @@ -322,7 +327,7 @@ out: return rc; } -/* Try to reaquire byte range locks that were released when session */ +/* Try to reacquire byte range locks that were released when session */ /* to server was lost */ static int cifs_relock_file(struct cifsFileInfo *cifsFile) { @@ -409,8 +414,8 @@ static int cifs_reopen_file(struct inode *inode, struct file *file, CIFS_MOUNT_MAP_SPECIAL_CHR); if (rc) { up(&pCifsFile->fh_sem); - cFYI(1, ("cifs_open returned 0x%x ", rc)); - cFYI(1, ("oplock: %d ", oplock)); + cFYI(1, ("cifs_open returned 0x%x", rc)); + cFYI(1, ("oplock: %d", oplock)); } else { pCifsFile->netfid = netfid; pCifsFile->invalidHandle = FALSE; @@ -471,8 +476,9 @@ int cifs_close(struct inode *inode, struct file *file) cifs_sb = CIFS_SB(inode->i_sb); pTcon = cifs_sb->tcon; if (pSMBFile) { + struct cifsLockInfo *li, *tmp; + pSMBFile->closePend = TRUE; - write_lock(&file->f_owner.lock); if (pTcon) { /* no sense reconnecting to close a file that is already closed */ @@ -487,23 +493,28 @@ int cifs_close(struct inode *inode, struct file *file) the struct would be in each open file, but this should give enough time to clear the socket */ - write_unlock(&file->f_owner.lock); cERROR(1,("close with pending writes")); msleep(timeout); - write_lock(&file->f_owner.lock); timeout *= 4; } - write_unlock(&file->f_owner.lock); rc = CIFSSMBClose(xid, pTcon, pSMBFile->netfid); - write_lock(&file->f_owner.lock); } } + + /* Delete any outstanding lock records. + We'll lose them when the file is closed anyway. */ + down(&pSMBFile->lock_sem); + list_for_each_entry_safe(li, tmp, &pSMBFile->llist, llist) { + list_del(&li->llist); + kfree(li); + } + up(&pSMBFile->lock_sem); + write_lock(&GlobalSMBSeslock); list_del(&pSMBFile->flist); list_del(&pSMBFile->tlist); write_unlock(&GlobalSMBSeslock); - write_unlock(&file->f_owner.lock); kfree(pSMBFile->search_resume_name); kfree(file->private_data); file->private_data = NULL; @@ -531,7 +542,7 @@ int cifs_closedir(struct inode *inode, struct file *file) (struct cifsFileInfo *)file->private_data; char *ptmp; - cFYI(1, ("Closedir inode = 0x%p with ", inode)); + cFYI(1, ("Closedir inode = 0x%p", inode)); xid = GetXid(); @@ -574,6 +585,21 @@ int cifs_closedir(struct inode *inode, struct file *file) return rc; } +static int store_file_lock(struct cifsFileInfo *fid, __u64 len, + __u64 offset, __u8 lockType) +{ + struct cifsLockInfo *li = kmalloc(sizeof(struct cifsLockInfo), GFP_KERNEL); + if (li == NULL) + return -ENOMEM; + li->offset = offset; + li->length = len; + li->type = lockType; + down(&fid->lock_sem); + list_add(&li->llist, &fid->llist); + up(&fid->lock_sem); + return 0; +} + int cifs_lock(struct file *file, int cmd, struct file_lock *pfLock) { int rc, xid; @@ -585,6 +611,7 @@ int cifs_lock(struct file *file, int cmd, struct file_lock *pfLock) struct cifsTconInfo *pTcon; __u16 netfid; __u8 lockType = LOCKING_ANDX_LARGE_FILES; + int posix_locking; length = 1 + pfLock->fl_end - pfLock->fl_start; rc = -EACCES; @@ -605,7 +632,7 @@ int cifs_lock(struct file *file, int cmd, struct file_lock *pfLock) } if (pfLock->fl_flags & FL_ACCESS) cFYI(1, ("Process suspended by mandatory locking - " - "not implemented yet ")); + "not implemented yet")); if (pfLock->fl_flags & FL_LEASE) cFYI(1, ("Lease on file - not implemented yet")); if (pfLock->fl_flags & @@ -643,15 +670,14 @@ int cifs_lock(struct file *file, int cmd, struct file_lock *pfLock) } netfid = ((struct cifsFileInfo *)file->private_data)->netfid; + posix_locking = (cifs_sb->tcon->ses->capabilities & CAP_UNIX) && + (CIFS_UNIX_FCNTL_CAP & le64_to_cpu(cifs_sb->tcon->fsUnixInfo.Capability)); /* BB add code here to normalize offset and length to account for negative length which we can not accept over the wire */ if (IS_GETLK(cmd)) { - if(experimEnabled && - (cifs_sb->tcon->ses->capabilities & CAP_UNIX) && - (CIFS_UNIX_FCNTL_CAP & - le64_to_cpu(cifs_sb->tcon->fsUnixInfo.Capability))) { + if(posix_locking) { int posix_lock_type; if(lockType & LOCKING_ANDX_SHARED_LOCK) posix_lock_type = CIFS_RDLCK; @@ -687,10 +713,15 @@ int cifs_lock(struct file *file, int cmd, struct file_lock *pfLock) FreeXid(xid); return rc; } - if (experimEnabled && - (cifs_sb->tcon->ses->capabilities & CAP_UNIX) && - (CIFS_UNIX_FCNTL_CAP & - le64_to_cpu(cifs_sb->tcon->fsUnixInfo.Capability))) { + + if (!numLock && !numUnlock) { + /* if no lock or unlock then nothing + to do since we do not know what it is */ + FreeXid(xid); + return -EOPNOTSUPP; + } + + if (posix_locking) { int posix_lock_type; if(lockType & LOCKING_ANDX_SHARED_LOCK) posix_lock_type = CIFS_RDLCK; @@ -699,18 +730,47 @@ int cifs_lock(struct file *file, int cmd, struct file_lock *pfLock) if(numUnlock == 1) posix_lock_type = CIFS_UNLCK; - else if(numLock == 0) { - /* if no lock or unlock then nothing - to do since we do not know what it is */ - FreeXid(xid); - return -EOPNOTSUPP; - } + rc = CIFSSMBPosixLock(xid, pTcon, netfid, 0 /* set */, length, pfLock, posix_lock_type, wait_flag); - } else - rc = CIFSSMBLock(xid, pTcon, netfid, length, pfLock->fl_start, - numUnlock, numLock, lockType, wait_flag); + } else { + struct cifsFileInfo *fid = (struct cifsFileInfo *)file->private_data; + + if (numLock) { + rc = CIFSSMBLock(xid, pTcon, netfid, length, pfLock->fl_start, + 0, numLock, lockType, wait_flag); + + if (rc == 0) { + /* For Windows locks we must store them. */ + rc = store_file_lock(fid, length, + pfLock->fl_start, lockType); + } + } else if (numUnlock) { + /* For each stored lock that this unlock overlaps + completely, unlock it. */ + int stored_rc = 0; + struct cifsLockInfo *li, *tmp; + + rc = 0; + down(&fid->lock_sem); + list_for_each_entry_safe(li, tmp, &fid->llist, llist) { + if (pfLock->fl_start <= li->offset && + length >= li->length) { + stored_rc = CIFSSMBLock(xid, pTcon, netfid, + li->length, li->offset, + 1, 0, li->type, FALSE); + if (stored_rc) + rc = stored_rc; + + list_del(&li->llist); + kfree(li); + } + } + up(&fid->lock_sem); + } + } + if (pfLock->fl_flags & FL_POSIX) posix_lock_file_wait(file, pfLock); FreeXid(xid); @@ -1375,7 +1435,7 @@ int cifs_fsync(struct file *file, struct dentry *dentry, int datasync) xid = GetXid(); - cFYI(1, ("Sync file - name: %s datasync: 0x%x ", + cFYI(1, ("Sync file - name: %s datasync: 0x%x", dentry->d_name.name, datasync)); rc = filemap_fdatawrite(inode->i_mapping); @@ -1404,7 +1464,7 @@ int cifs_fsync(struct file *file, struct dentry *dentry, int datasync) /* fill in rpages then result = cifs_pagein_inode(inode, index, rpages); */ /* BB finish */ -/* cFYI(1, ("rpages is %d for sync page of Index %ld ", rpages, index)); +/* cFYI(1, ("rpages is %d for sync page of Index %ld", rpages, index)); #if 0 if (rc < 0) @@ -1836,7 +1896,7 @@ static int cifs_readpage_worker(struct file *file, struct page *page, if (rc < 0) goto io_error; else - cFYI(1, ("Bytes read %d ",rc)); + cFYI(1, ("Bytes read %d",rc)); file->f_dentry->d_inode->i_atime = current_fs_time(file->f_dentry->d_inode->i_sb); @@ -1946,7 +2006,7 @@ static int cifs_prepare_write(struct file *file, struct page *page, return 0; } -struct address_space_operations cifs_addr_ops = { +const struct address_space_operations cifs_addr_ops = { .readpage = cifs_readpage, .readpages = cifs_readpages, .writepage = cifs_writepage, @@ -1957,3 +2017,19 @@ struct address_space_operations cifs_addr_ops = { /* .sync_page = cifs_sync_page, */ /* .direct_IO = */ }; + +/* + * cifs_readpages requires the server to support a buffer large enough to + * contain the header plus one complete page of data. Otherwise, we need + * to leave cifs_readpages out of the address space operations. + */ +const struct address_space_operations cifs_addr_ops_smallbuf = { + .readpage = cifs_readpage, + .writepage = cifs_writepage, + .writepages = cifs_writepages, + .prepare_write = cifs_prepare_write, + .commit_write = cifs_commit_write, + .set_page_dirty = __set_page_dirty_nobuffers, + /* .sync_page = cifs_sync_page, */ + /* .direct_IO = */ +}; diff --git a/fs/cifs/inode.c b/fs/cifs/inode.c index 4093764ef46..b88147c1dc2 100644 --- a/fs/cifs/inode.c +++ b/fs/cifs/inode.c @@ -41,7 +41,7 @@ int cifs_get_inode_info_unix(struct inode **pinode, char *tmp_path; pTcon = cifs_sb->tcon; - cFYI(1, ("Getting info on %s ", search_path)); + cFYI(1, ("Getting info on %s", search_path)); /* could have done a find first instead but this returns more info */ rc = CIFSSMBUnixQPathInfo(xid, pTcon, search_path, &findData, cifs_sb->local_nls, cifs_sb->mnt_cifs_flags & @@ -97,9 +97,9 @@ int cifs_get_inode_info_unix(struct inode **pinode, inode = *pinode; cifsInfo = CIFS_I(inode); - cFYI(1, ("Old time %ld ", cifsInfo->time)); + cFYI(1, ("Old time %ld", cifsInfo->time)); cifsInfo->time = jiffies; - cFYI(1, ("New time %ld ", cifsInfo->time)); + cFYI(1, ("New time %ld", cifsInfo->time)); /* this is ok to set on every inode revalidate */ atomic_set(&cifsInfo->inUse,1); @@ -180,11 +180,12 @@ int cifs_get_inode_info_unix(struct inode **pinode, else /* not direct, send byte range locks */ inode->i_fop = &cifs_file_ops; - inode->i_data.a_ops = &cifs_addr_ops; /* check if server can support readpages */ if(pTcon->ses->server->maxBuf < - 4096 + MAX_CIFS_HDR_SIZE) - inode->i_data.a_ops->readpages = NULL; + PAGE_CACHE_SIZE + MAX_CIFS_HDR_SIZE) + inode->i_data.a_ops = &cifs_addr_ops_smallbuf; + else + inode->i_data.a_ops = &cifs_addr_ops; } else if (S_ISDIR(inode->i_mode)) { cFYI(1, ("Directory inode")); inode->i_op = &cifs_dir_inode_ops; @@ -421,23 +422,23 @@ int cifs_get_inode_info(struct inode **pinode, inode = *pinode; cifsInfo = CIFS_I(inode); cifsInfo->cifsAttrs = attr; - cFYI(1, ("Old time %ld ", cifsInfo->time)); + cFYI(1, ("Old time %ld", cifsInfo->time)); cifsInfo->time = jiffies; - cFYI(1, ("New time %ld ", cifsInfo->time)); + cFYI(1, ("New time %ld", cifsInfo->time)); /* blksize needs to be multiple of two. So safer to default to blksize and blkbits set in superblock so 2**blkbits and blksize will match rather than setting to: (pTcon->ses->server->maxBuf - MAX_CIFS_HDR_SIZE) & 0xFFFFFE00;*/ - /* Linux can not store file creation time unfortunately so we ignore it */ + /* Linux can not store file creation time so ignore it */ inode->i_atime = cifs_NTtimeToUnix(le64_to_cpu(pfindData->LastAccessTime)); inode->i_mtime = cifs_NTtimeToUnix(le64_to_cpu(pfindData->LastWriteTime)); inode->i_ctime = cifs_NTtimeToUnix(le64_to_cpu(pfindData->ChangeTime)); - cFYI(0, ("Attributes came in as 0x%x ", attr)); + cFYI(0, ("Attributes came in as 0x%x", attr)); /* set default mode. will override for dirs below */ if (atomic_read(&cifsInfo->inUse) == 0) @@ -519,10 +520,11 @@ int cifs_get_inode_info(struct inode **pinode, else /* not direct, send byte range locks */ inode->i_fop = &cifs_file_ops; - inode->i_data.a_ops = &cifs_addr_ops; if(pTcon->ses->server->maxBuf < - 4096 + MAX_CIFS_HDR_SIZE) - inode->i_data.a_ops->readpages = NULL; + PAGE_CACHE_SIZE + MAX_CIFS_HDR_SIZE) + inode->i_data.a_ops = &cifs_addr_ops_smallbuf; + else + inode->i_data.a_ops = &cifs_addr_ops; } else if (S_ISDIR(inode->i_mode)) { cFYI(1, ("Directory inode")); inode->i_op = &cifs_dir_inode_ops; @@ -731,7 +733,7 @@ int cifs_mkdir(struct inode *inode, struct dentry *direntry, int mode) rc = CIFSSMBMkDir(xid, pTcon, full_path, cifs_sb->local_nls, cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR); if (rc) { - cFYI(1, ("cifs_mkdir returned 0x%x ", rc)); + cFYI(1, ("cifs_mkdir returned 0x%x", rc)); d_drop(direntry); } else { inode->i_nlink++; @@ -798,7 +800,7 @@ int cifs_rmdir(struct inode *inode, struct dentry *direntry) char *full_path = NULL; struct cifsInodeInfo *cifsInode; - cFYI(1, ("cifs_rmdir, inode = 0x%p with ", inode)); + cFYI(1, ("cifs_rmdir, inode = 0x%p", inode)); xid = GetXid(); @@ -1121,7 +1123,7 @@ int cifs_setattr(struct dentry *direntry, struct iattr *attrs) xid = GetXid(); - cFYI(1, ("In cifs_setattr, name = %s attrs->iavalid 0x%x ", + cFYI(1, ("setattr on file %s attrs->iavalid 0x%x", direntry->d_name.name, attrs->ia_valid)); cifs_sb = CIFS_SB(direntry->d_inode->i_sb); @@ -1157,6 +1159,7 @@ int cifs_setattr(struct dentry *direntry, struct iattr *attrs) when the local oplock break takes longer to flush writebehind data than the SMB timeout for the SetPathInfo request would allow */ + open_file = find_writable_file(cifsInode); if (open_file) { __u16 nfid = open_file->netfid; @@ -1289,7 +1292,7 @@ int cifs_setattr(struct dentry *direntry, struct iattr *attrs) it may be useful to Windows - but we do not want to set ctime unless some other timestamp is changing */ - cFYI(1, ("CIFS - CTIME changed ")); + cFYI(1, ("CIFS - CTIME changed")); time_buf.ChangeTime = cpu_to_le64(cifs_UnixTimeToNT(attrs->ia_ctime)); } else @@ -1356,7 +1359,7 @@ cifs_setattr_exit: void cifs_delete_inode(struct inode *inode) { - cFYI(1, ("In cifs_delete_inode, inode = 0x%p ", inode)); + cFYI(1, ("In cifs_delete_inode, inode = 0x%p", inode)); /* may have to add back in if and when safe distributed caching of directories added e.g. via FindNotify */ } diff --git a/fs/cifs/link.c b/fs/cifs/link.c index 2ec99f83314..a57f5d6e621 100644 --- a/fs/cifs/link.c +++ b/fs/cifs/link.c @@ -167,7 +167,7 @@ cifs_symlink(struct inode *inode, struct dentry *direntry, const char *symname) return -ENOMEM; } - cFYI(1, ("Full path: %s ", full_path)); + cFYI(1, ("Full path: %s", full_path)); cFYI(1, ("symname is %s", symname)); /* BB what if DFS and this volume is on different share? BB */ @@ -186,8 +186,7 @@ cifs_symlink(struct inode *inode, struct dentry *direntry, const char *symname) inode->i_sb,xid); if (rc != 0) { - cFYI(1, - ("Create symlink worked but get_inode_info failed with rc = %d ", + cFYI(1, ("Create symlink ok, getinodeinfo fail rc = %d", rc)); } else { if (pTcon->nocase) @@ -289,7 +288,7 @@ cifs_readlink(struct dentry *direntry, char __user *pBuffer, int buflen) else { cFYI(1,("num referral: %d",num_referrals)); if(referrals) { - cFYI(1,("referral string: %s ",referrals)); + cFYI(1,("referral string: %s",referrals)); strncpy(tmpbuffer, referrals, len-1); } } diff --git a/fs/cifs/misc.c b/fs/cifs/misc.c index fafd056426e..22c937e5884 100644 --- a/fs/cifs/misc.c +++ b/fs/cifs/misc.c @@ -101,6 +101,7 @@ sesInfoFree(struct cifsSesInfo *buf_to_free) kfree(buf_to_free->serverDomain); kfree(buf_to_free->serverNOS); kfree(buf_to_free->password); + kfree(buf_to_free->domainName); kfree(buf_to_free); } @@ -499,11 +500,12 @@ is_valid_oplock_break(struct smb_hdr *buf, struct TCP_Server_Info *srv) if(pSMBr->ByteCount > sizeof(struct file_notify_information)) { data_offset = le32_to_cpu(pSMBr->DataOffset); - pnotify = (struct file_notify_information *)((char *)&pSMBr->hdr.Protocol - + data_offset); - cFYI(1,("dnotify on %s with action: 0x%x",pnotify->FileName, + pnotify = (struct file_notify_information *) + ((char *)&pSMBr->hdr.Protocol + data_offset); + cFYI(1,("dnotify on %s Action: 0x%x",pnotify->FileName, pnotify->Action)); /* BB removeme BB */ - /* cifs_dump_mem("Received notify Data is: ",buf,sizeof(struct smb_hdr)+60); */ + /* cifs_dump_mem("Rcvd notify Data: ",buf, + sizeof(struct smb_hdr)+60); */ return TRUE; } if(pSMBr->hdr.Status.CifsError) { diff --git a/fs/cifs/netmisc.c b/fs/cifs/netmisc.c index 5de74d216fd..ce87550e918 100644 --- a/fs/cifs/netmisc.c +++ b/fs/cifs/netmisc.c @@ -72,6 +72,7 @@ static const struct smb_to_posix_error mapping_table_ERRDOS[] = { {ERRinvlevel,-EOPNOTSUPP}, {ERRdirnotempty, -ENOTEMPTY}, {ERRnotlocked, -ENOLCK}, + {ERRcancelviolation, -ENOLCK}, {ERRalreadyexists, -EEXIST}, {ERRmoredata, -EOVERFLOW}, {ERReasnotsupported,-EOPNOTSUPP}, @@ -84,11 +85,11 @@ static const struct smb_to_posix_error mapping_table_ERRDOS[] = { static const struct smb_to_posix_error mapping_table_ERRSRV[] = { {ERRerror, -EIO}, - {ERRbadpw, -EPERM}, + {ERRbadpw, -EACCES}, /* was EPERM */ {ERRbadtype, -EREMOTE}, {ERRaccess, -EACCES}, {ERRinvtid, -ENXIO}, - {ERRinvnetname, -ENODEV}, + {ERRinvnetname, -ENXIO}, {ERRinvdevice, -ENXIO}, {ERRqfull, -ENOSPC}, {ERRqtoobig, -ENOSPC}, diff --git a/fs/cifs/ntlmssp.c b/fs/cifs/ntlmssp.c deleted file mode 100644 index 115359cc7a3..00000000000 --- a/fs/cifs/ntlmssp.c +++ /dev/null @@ -1,143 +0,0 @@ -/* - * fs/cifs/ntlmssp.h - * - * Copyright (c) International Business Machines Corp., 2006 - * Author(s): Steve French (sfrench@us.ibm.com) - * - * 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 "cifspdu.h" -#include "cifsglob.h" -#include "cifsproto.h" -#include "cifs_unicode.h" -#include "cifs_debug.h" -#include "ntlmssp.h" -#include "nterr.h" - -#ifdef CONFIG_CIFS_EXPERIMENTAL -static __u32 cifs_ssetup_hdr(struct cifsSesInfo *ses, SESSION_SETUP_ANDX *pSMB) -{ - __u32 capabilities = 0; - - /* init fields common to all four types of SessSetup */ - /* note that header is initialized to zero in header_assemble */ - pSMB->req.AndXCommand = 0xFF; - pSMB->req.MaxBufferSize = cpu_to_le16(ses->server->maxBuf); - pSMB->req.MaxMpxCount = cpu_to_le16(ses->server->maxReq); - - /* Now no need to set SMBFLG_CASELESS or obsolete CANONICAL PATH */ - - /* BB verify whether signing required on neg or just on auth frame - (and NTLM case) */ - - capabilities = CAP_LARGE_FILES | CAP_NT_SMBS | CAP_LEVEL_II_OPLOCKS | - CAP_LARGE_WRITE_X | CAP_LARGE_READ_X; - - if(ses->server->secMode & (SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED)) - pSMB->req.hdr.Flags2 |= SMBFLG2_SECURITY_SIGNATURE; - - if (ses->capabilities & CAP_UNICODE) { - pSMB->req.hdr.Flags2 |= SMBFLG2_UNICODE; - capabilities |= CAP_UNICODE; - } - if (ses->capabilities & CAP_STATUS32) { - pSMB->req.hdr.Flags2 |= SMBFLG2_ERR_STATUS; - capabilities |= CAP_STATUS32; - } - if (ses->capabilities & CAP_DFS) { - pSMB->req.hdr.Flags2 |= SMBFLG2_DFS; - capabilities |= CAP_DFS; - } - - /* BB check whether to init vcnum BB */ - return capabilities; -} -int -CIFS_SessSetup(unsigned int xid, struct cifsSesInfo *ses, const int type, - int * pNTLMv2_flg, const struct nls_table *nls_cp) -{ - int rc = 0; - int wct; - struct smb_hdr *smb_buffer; - char *bcc_ptr; - SESSION_SETUP_ANDX *pSMB; - __u32 capabilities; - - if(ses == NULL) - return -EINVAL; - - cFYI(1,("SStp type: %d",type)); - if(type < CIFS_NTLM) { -#ifndef CONFIG_CIFS_WEAK_PW_HASH - /* LANMAN and plaintext are less secure and off by default. - So we make this explicitly be turned on in kconfig (in the - build) and turned on at runtime (changed from the default) - in proc/fs/cifs or via mount parm. Unfortunately this is - needed for old Win (e.g. Win95), some obscure NAS and OS/2 */ - return -EOPNOTSUPP; -#endif - wct = 10; /* lanman 2 style sessionsetup */ - } else if(type < CIFS_NTLMSSP_NEG) - wct = 13; /* old style NTLM sessionsetup */ - else /* same size for negotiate or auth, NTLMSSP or extended security */ - wct = 12; - - rc = small_smb_init_no_tc(SMB_COM_SESSION_SETUP_ANDX, wct, ses, - (void **)&smb_buffer); - if(rc) - return rc; - - pSMB = (SESSION_SETUP_ANDX *)smb_buffer; - - capabilities = cifs_ssetup_hdr(ses, pSMB); - bcc_ptr = pByteArea(smb_buffer); - if(type > CIFS_NTLM) { - pSMB->req.hdr.Flags2 |= SMBFLG2_EXT_SEC; - capabilities |= CAP_EXTENDED_SECURITY; - pSMB->req.Capabilities = cpu_to_le32(capabilities); - /* BB set password lengths */ - } else if(type < CIFS_NTLM) /* lanman */ { - /* no capabilities flags in old lanman negotiation */ - /* pSMB->old_req.PasswordLength = */ /* BB fixme BB */ - } else /* type CIFS_NTLM */ { - pSMB->req_no_secext.Capabilities = cpu_to_le32(capabilities); - pSMB->req_no_secext.CaseInsensitivePasswordLength = - cpu_to_le16(CIFS_SESSION_KEY_SIZE); - pSMB->req_no_secext.CaseSensitivePasswordLength = - cpu_to_le16(CIFS_SESSION_KEY_SIZE); - } - - - /* copy session key */ - - /* if Unicode, align strings to two byte boundary */ - - /* copy user name */ /* BB Do we need to special case null user name? */ - - /* copy domain name */ - - /* copy Linux version */ - - /* copy network operating system name */ - - /* update bcc and smb buffer length */ - -/* rc = SendReceive2(xid, ses, iov, num_iovecs, &resp_buf_type, 0); */ - /* SMB request buf freed in SendReceive2 */ - - return rc; -} -#endif /* CONFIG_CIFS_EXPERIMENTAL */ diff --git a/fs/cifs/readdir.c b/fs/cifs/readdir.c index b689c503512..b27b34537bf 100644 --- a/fs/cifs/readdir.c +++ b/fs/cifs/readdir.c @@ -21,6 +21,7 @@ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ #include <linux/fs.h> +#include <linux/pagemap.h> #include <linux/stat.h> #include <linux/smp_lock.h> #include "cifspdu.h" @@ -31,8 +32,8 @@ #include "cifs_fs_sb.h" #include "cifsfs.h" -/* BB fixme - add debug wrappers around this function to disable it fixme BB */ -/* static void dump_cifs_file_struct(struct file *file, char *label) +#ifdef CONFIG_CIFS_DEBUG2 +static void dump_cifs_file_struct(struct file *file, char *label) { struct cifsFileInfo * cf; @@ -53,7 +54,8 @@ } } -} */ +} +#endif /* DEBUG2 */ /* Returns one if new inode created (which therefore needs to be hashed) */ /* Might check in the future if inode number changed so we can rehash inode */ @@ -80,7 +82,6 @@ static int construct_dentry(struct qstr *qstring, struct file *file, if(*ptmp_inode == NULL) return rc; rc = 1; - d_instantiate(tmp_dentry, *ptmp_inode); } } else { tmp_dentry = d_alloc(file->f_dentry, qstring); @@ -97,9 +98,7 @@ static int construct_dentry(struct qstr *qstring, struct file *file, tmp_dentry->d_op = &cifs_dentry_ops; if(*ptmp_inode == NULL) return rc; - rc = 1; - d_instantiate(tmp_dentry, *ptmp_inode); - d_rehash(tmp_dentry); + rc = 2; } tmp_dentry->d_time = jiffies; @@ -107,32 +106,52 @@ static int construct_dentry(struct qstr *qstring, struct file *file, return rc; } -static void fill_in_inode(struct inode *tmp_inode, - FILE_DIRECTORY_INFO *pfindData, int *pobject_type, int isNewInode) +static void fill_in_inode(struct inode *tmp_inode, int new_buf_type, + char * buf, int *pobject_type, int isNewInode) { loff_t local_size; struct timespec local_mtime; struct cifsInodeInfo *cifsInfo = CIFS_I(tmp_inode); struct cifs_sb_info *cifs_sb = CIFS_SB(tmp_inode->i_sb); - __u32 attr = le32_to_cpu(pfindData->ExtFileAttributes); - __u64 allocation_size = le64_to_cpu(pfindData->AllocationSize); - __u64 end_of_file = le64_to_cpu(pfindData->EndOfFile); - - cifsInfo->cifsAttrs = attr; - cifsInfo->time = jiffies; + __u32 attr; + __u64 allocation_size; + __u64 end_of_file; /* save mtime and size */ local_mtime = tmp_inode->i_mtime; local_size = tmp_inode->i_size; + if(new_buf_type) { + FILE_DIRECTORY_INFO *pfindData = (FILE_DIRECTORY_INFO *)buf; + + attr = le32_to_cpu(pfindData->ExtFileAttributes); + allocation_size = le64_to_cpu(pfindData->AllocationSize); + end_of_file = le64_to_cpu(pfindData->EndOfFile); + tmp_inode->i_atime = + cifs_NTtimeToUnix(le64_to_cpu(pfindData->LastAccessTime)); + tmp_inode->i_mtime = + cifs_NTtimeToUnix(le64_to_cpu(pfindData->LastWriteTime)); + tmp_inode->i_ctime = + cifs_NTtimeToUnix(le64_to_cpu(pfindData->ChangeTime)); + } else { /* legacy, OS2 and DOS style */ + FIND_FILE_STANDARD_INFO * pfindData = + (FIND_FILE_STANDARD_INFO *)buf; + + attr = le16_to_cpu(pfindData->Attributes); + allocation_size = le32_to_cpu(pfindData->AllocationSize); + end_of_file = le32_to_cpu(pfindData->DataSize); + tmp_inode->i_atime = CURRENT_TIME; + /* tmp_inode->i_mtime = BB FIXME - add dos time handling + tmp_inode->i_ctime = 0; BB FIXME */ + + } + /* Linux can not store file creation time unfortunately so ignore it */ - tmp_inode->i_atime = - cifs_NTtimeToUnix(le64_to_cpu(pfindData->LastAccessTime)); - tmp_inode->i_mtime = - cifs_NTtimeToUnix(le64_to_cpu(pfindData->LastWriteTime)); - tmp_inode->i_ctime = - cifs_NTtimeToUnix(le64_to_cpu(pfindData->ChangeTime)); + + cifsInfo->cifsAttrs = attr; + cifsInfo->time = jiffies; + /* treat dos attribute of read-only as read-only mode bit e.g. 555? */ /* 2767 perms - indicate mandatory locking */ /* BB fill in uid and gid here? with help from winbind? @@ -197,10 +216,9 @@ static void fill_in_inode(struct inode *tmp_inode, if (allocation_size < end_of_file) cFYI(1, ("May be sparse file, allocation less than file size")); - cFYI(1, ("File Size %ld and blocks %llu and blocksize %ld", + cFYI(1, ("File Size %ld and blocks %llu", (unsigned long)tmp_inode->i_size, - (unsigned long long)tmp_inode->i_blocks, - tmp_inode->i_blksize)); + (unsigned long long)tmp_inode->i_blocks)); if (S_ISREG(tmp_inode->i_mode)) { cFYI(1, ("File inode")); tmp_inode->i_op = &cifs_file_inode_ops; @@ -215,11 +233,13 @@ static void fill_in_inode(struct inode *tmp_inode, else tmp_inode->i_fop = &cifs_file_ops; - tmp_inode->i_data.a_ops = &cifs_addr_ops; if((cifs_sb->tcon) && (cifs_sb->tcon->ses) && (cifs_sb->tcon->ses->server->maxBuf < - 4096 + MAX_CIFS_HDR_SIZE)) - tmp_inode->i_data.a_ops->readpages = NULL; + PAGE_CACHE_SIZE + MAX_CIFS_HDR_SIZE)) + tmp_inode->i_data.a_ops = &cifs_addr_ops_smallbuf; + else + tmp_inode->i_data.a_ops = &cifs_addr_ops; + if(isNewInode) return; /* No sense invalidating pages for new inode since have not started caching readahead file @@ -338,11 +358,12 @@ static void unix_fill_in_inode(struct inode *tmp_inode, else tmp_inode->i_fop = &cifs_file_ops; - tmp_inode->i_data.a_ops = &cifs_addr_ops; if((cifs_sb->tcon) && (cifs_sb->tcon->ses) && (cifs_sb->tcon->ses->server->maxBuf < - 4096 + MAX_CIFS_HDR_SIZE)) - tmp_inode->i_data.a_ops->readpages = NULL; + PAGE_CACHE_SIZE + MAX_CIFS_HDR_SIZE)) + tmp_inode->i_data.a_ops = &cifs_addr_ops_smallbuf; + else + tmp_inode->i_data.a_ops = &cifs_addr_ops; if(isNewInode) return; /* No sense invalidating pages for new inode since we @@ -415,7 +436,10 @@ static int initiate_cifs_search(const int xid, struct file *file) ffirst_retry: /* test for Unix extensions */ if (pTcon->ses->capabilities & CAP_UNIX) { - cifsFile->srch_inf.info_level = SMB_FIND_FILE_UNIX; + cifsFile->srch_inf.info_level = SMB_FIND_FILE_UNIX; + } else if ((pTcon->ses->capabilities & + (CAP_NT_SMBS | CAP_NT_FIND)) == 0) { + cifsFile->srch_inf.info_level = SMB_FIND_FILE_INFO_STANDARD; } else if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_SERVER_INUM) { cifsFile->srch_inf.info_level = SMB_FIND_FILE_ID_FULL_DIR_INFO; } else /* not srvinos - BB fixme add check for backlevel? */ { @@ -451,12 +475,19 @@ static int cifs_unicode_bytelen(char *str) return len << 1; } -static char *nxt_dir_entry(char *old_entry, char *end_of_smb) +static char *nxt_dir_entry(char *old_entry, char *end_of_smb, int level) { char * new_entry; FILE_DIRECTORY_INFO * pDirInfo = (FILE_DIRECTORY_INFO *)old_entry; - new_entry = old_entry + le32_to_cpu(pDirInfo->NextEntryOffset); + if(level == SMB_FIND_FILE_INFO_STANDARD) { + FIND_FILE_STANDARD_INFO * pfData; + pfData = (FIND_FILE_STANDARD_INFO *)pDirInfo; + + new_entry = old_entry + sizeof(FIND_FILE_STANDARD_INFO) + + pfData->FileNameLength; + } else + new_entry = old_entry + le32_to_cpu(pDirInfo->NextEntryOffset); cFYI(1,("new entry %p old entry %p",new_entry,old_entry)); /* validate that new_entry is not past end of SMB */ if(new_entry >= end_of_smb) { @@ -464,7 +495,10 @@ static char *nxt_dir_entry(char *old_entry, char *end_of_smb) ("search entry %p began after end of SMB %p old entry %p", new_entry, end_of_smb, old_entry)); return NULL; - } else if (new_entry + sizeof(FILE_DIRECTORY_INFO) > end_of_smb) { + } else if(((level == SMB_FIND_FILE_INFO_STANDARD) && + (new_entry + sizeof(FIND_FILE_STANDARD_INFO) > end_of_smb)) || + ((level != SMB_FIND_FILE_INFO_STANDARD) && + (new_entry + sizeof(FILE_DIRECTORY_INFO) > end_of_smb))) { cERROR(1,("search entry %p extends after end of SMB %p", new_entry, end_of_smb)); return NULL; @@ -482,7 +516,7 @@ static int cifs_entry_is_dot(char *current_entry, struct cifsFileInfo *cfile) char * filename = NULL; int len = 0; - if(cfile->srch_inf.info_level == 0x202) { + if(cfile->srch_inf.info_level == SMB_FIND_FILE_UNIX) { FILE_UNIX_INFO * pFindData = (FILE_UNIX_INFO *)current_entry; filename = &pFindData->FileName[0]; if(cfile->srch_inf.unicode) { @@ -491,26 +525,34 @@ static int cifs_entry_is_dot(char *current_entry, struct cifsFileInfo *cfile) /* BB should we make this strnlen of PATH_MAX? */ len = strnlen(filename, 5); } - } else if(cfile->srch_inf.info_level == 0x101) { + } else if(cfile->srch_inf.info_level == SMB_FIND_FILE_DIRECTORY_INFO) { FILE_DIRECTORY_INFO * pFindData = (FILE_DIRECTORY_INFO *)current_entry; filename = &pFindData->FileName[0]; len = le32_to_cpu(pFindData->FileNameLength); - } else if(cfile->srch_inf.info_level == 0x102) { + } else if(cfile->srch_inf.info_level == + SMB_FIND_FILE_FULL_DIRECTORY_INFO) { FILE_FULL_DIRECTORY_INFO * pFindData = (FILE_FULL_DIRECTORY_INFO *)current_entry; filename = &pFindData->FileName[0]; len = le32_to_cpu(pFindData->FileNameLength); - } else if(cfile->srch_inf.info_level == 0x105) { + } else if(cfile->srch_inf.info_level == + SMB_FIND_FILE_ID_FULL_DIR_INFO) { SEARCH_ID_FULL_DIR_INFO * pFindData = (SEARCH_ID_FULL_DIR_INFO *)current_entry; filename = &pFindData->FileName[0]; len = le32_to_cpu(pFindData->FileNameLength); - } else if(cfile->srch_inf.info_level == 0x104) { + } else if(cfile->srch_inf.info_level == + SMB_FIND_FILE_BOTH_DIRECTORY_INFO) { FILE_BOTH_DIRECTORY_INFO * pFindData = (FILE_BOTH_DIRECTORY_INFO *)current_entry; filename = &pFindData->FileName[0]; len = le32_to_cpu(pFindData->FileNameLength); + } else if(cfile->srch_inf.info_level == SMB_FIND_FILE_INFO_STANDARD) { + FIND_FILE_STANDARD_INFO * pFindData = + (FIND_FILE_STANDARD_INFO *)current_entry; + filename = &pFindData->FileName[0]; + len = pFindData->FileNameLength; } else { cFYI(1,("Unknown findfirst level %d",cfile->srch_inf.info_level)); } @@ -597,7 +639,9 @@ static int find_cifs_entry(const int xid, struct cifsTconInfo *pTcon, . and .. for the root of a drive and for those we need to start two entries earlier */ -/* dump_cifs_file_struct(file, "In fce ");*/ +#ifdef CONFIG_CIFS_DEBUG2 + dump_cifs_file_struct(file, "In fce "); +#endif if(((index_to_find < cifsFile->srch_inf.index_of_last_entry) && is_dir_changed(file)) || (index_to_find < first_entry_in_buffer)) { @@ -644,10 +688,12 @@ static int find_cifs_entry(const int xid, struct cifsTconInfo *pTcon, first_entry_in_buffer = cifsFile->srch_inf.index_of_last_entry - cifsFile->srch_inf.entries_in_buffer; pos_in_buf = index_to_find - first_entry_in_buffer; - cFYI(1,("found entry - pos_in_buf %d",pos_in_buf)); + cFYI(1,("found entry - pos_in_buf %d",pos_in_buf)); + for(i=0;(i<(pos_in_buf)) && (current_entry != NULL);i++) { /* go entry by entry figuring out which is first */ - current_entry = nxt_dir_entry(current_entry,end_of_smb); + current_entry = nxt_dir_entry(current_entry,end_of_smb, + cifsFile->srch_inf.info_level); } if((current_entry == NULL) && (i < pos_in_buf)) { /* BB fixme - check if we should flag this error */ @@ -674,7 +720,7 @@ static int find_cifs_entry(const int xid, struct cifsTconInfo *pTcon, /* inode num, inode type and filename returned */ static int cifs_get_name_from_search_buf(struct qstr *pqst, char *current_entry, __u16 level, unsigned int unicode, - struct cifs_sb_info * cifs_sb, ino_t *pinum) + struct cifs_sb_info * cifs_sb, int max_len, ino_t *pinum) { int rc = 0; unsigned int len = 0; @@ -718,10 +764,22 @@ static int cifs_get_name_from_search_buf(struct qstr *pqst, (FILE_BOTH_DIRECTORY_INFO *)current_entry; filename = &pFindData->FileName[0]; len = le32_to_cpu(pFindData->FileNameLength); + } else if(level == SMB_FIND_FILE_INFO_STANDARD) { + FIND_FILE_STANDARD_INFO * pFindData = + (FIND_FILE_STANDARD_INFO *)current_entry; + filename = &pFindData->FileName[0]; + /* one byte length, no name conversion */ + len = (unsigned int)pFindData->FileNameLength; } else { cFYI(1,("Unknown findfirst level %d",level)); return -EINVAL; } + + if(len > max_len) { + cERROR(1,("bad search response length %d past smb end", len)); + return -EINVAL; + } + if(unicode) { /* BB fixme - test with long names */ /* Note converted filename can be longer than in unicode */ @@ -741,7 +799,7 @@ static int cifs_get_name_from_search_buf(struct qstr *pqst, } static int cifs_filldir(char *pfindEntry, struct file *file, - filldir_t filldir, void *direntry, char *scratch_buf) + filldir_t filldir, void *direntry, char *scratch_buf, int max_len) { int rc = 0; struct qstr qstring; @@ -777,6 +835,7 @@ static int cifs_filldir(char *pfindEntry, struct file *file, rc = cifs_get_name_from_search_buf(&qstring,pfindEntry, pCifsF->srch_inf.info_level, pCifsF->srch_inf.unicode,cifs_sb, + max_len, &inum /* returned */); if(rc) @@ -798,14 +857,23 @@ static int cifs_filldir(char *pfindEntry, struct file *file, /* we pass in rc below, indicating whether it is a new inode, so we can figure out whether to invalidate the inode cached data if the file has changed */ - if(pCifsF->srch_inf.info_level == SMB_FIND_FILE_UNIX) { + if(pCifsF->srch_inf.info_level == SMB_FIND_FILE_UNIX) unix_fill_in_inode(tmp_inode, - (FILE_UNIX_INFO *)pfindEntry,&obj_type, rc); - } else { - fill_in_inode(tmp_inode, - (FILE_DIRECTORY_INFO *)pfindEntry,&obj_type, rc); + (FILE_UNIX_INFO *)pfindEntry, + &obj_type, rc); + else if(pCifsF->srch_inf.info_level == SMB_FIND_FILE_INFO_STANDARD) + fill_in_inode(tmp_inode, 0 /* old level 1 buffer type */, + pfindEntry, &obj_type, rc); + else + fill_in_inode(tmp_inode, 1 /* NT */, pfindEntry, &obj_type, rc); + + if(rc) /* new inode - needs to be tied to dentry */ { + d_instantiate(tmp_dentry, tmp_inode); + if(rc == 2) + d_rehash(tmp_dentry); } + rc = filldir(direntry,qstring.name,qstring.len,file->f_pos, tmp_inode->i_ino,obj_type); if(rc) { @@ -864,6 +932,12 @@ static int cifs_save_resume_key(const char *current_entry, filename = &pFindData->FileName[0]; len = le32_to_cpu(pFindData->FileNameLength); cifsFile->srch_inf.resume_key = pFindData->FileIndex; + } else if(level == SMB_FIND_FILE_INFO_STANDARD) { + FIND_FILE_STANDARD_INFO * pFindData = + (FIND_FILE_STANDARD_INFO *)current_entry; + filename = &pFindData->FileName[0]; + /* one byte length, no name conversion */ + len = (unsigned int)pFindData->FileNameLength; } else { cFYI(1,("Unknown findfirst level %d",level)); return -EINVAL; @@ -884,6 +958,7 @@ int cifs_readdir(struct file *file, void *direntry, filldir_t filldir) int num_to_fill = 0; char * tmp_buf = NULL; char * end_of_smb; + int max_len; xid = GetXid(); @@ -909,7 +984,7 @@ int cifs_readdir(struct file *file, void *direntry, filldir_t filldir) case 1: if (filldir(direntry, "..", 2, file->f_pos, file->f_dentry->d_parent->d_inode->i_ino, DT_DIR) < 0) { - cERROR(1, ("Filldir for parent dir failed ")); + cERROR(1, ("Filldir for parent dir failed")); rc = -ENOMEM; break; } @@ -959,10 +1034,11 @@ int cifs_readdir(struct file *file, void *direntry, filldir_t filldir) goto rddir2_exit; } cFYI(1,("loop through %d times filling dir for net buf %p", - num_to_fill,cifsFile->srch_inf.ntwrk_buf_start)); - end_of_smb = cifsFile->srch_inf.ntwrk_buf_start + - smbCalcSize((struct smb_hdr *) - cifsFile->srch_inf.ntwrk_buf_start); + num_to_fill,cifsFile->srch_inf.ntwrk_buf_start)); + max_len = smbCalcSize((struct smb_hdr *) + cifsFile->srch_inf.ntwrk_buf_start); + end_of_smb = cifsFile->srch_inf.ntwrk_buf_start + max_len; + /* To be safe - for UCS to UTF-8 with strings loaded with the rare long characters alloc more to account for such multibyte target UTF-8 characters. cifs_unicode.c, @@ -977,17 +1053,19 @@ int cifs_readdir(struct file *file, void *direntry, filldir_t filldir) } /* if buggy server returns . and .. late do we want to check for that here? */ - rc = cifs_filldir(current_entry, file, - filldir, direntry,tmp_buf); + rc = cifs_filldir(current_entry, file, + filldir, direntry, tmp_buf, max_len); file->f_pos++; - if(file->f_pos == cifsFile->srch_inf.index_of_last_entry) { + if(file->f_pos == + cifsFile->srch_inf.index_of_last_entry) { cFYI(1,("last entry in buf at pos %lld %s", - file->f_pos,tmp_buf)); /* BB removeme BB */ + file->f_pos,tmp_buf)); cifs_save_resume_key(current_entry,cifsFile); break; } else - current_entry = nxt_dir_entry(current_entry, - end_of_smb); + current_entry = + nxt_dir_entry(current_entry, end_of_smb, + cifsFile->srch_inf.info_level); } kfree(tmp_buf); break; diff --git a/fs/cifs/sess.c b/fs/cifs/sess.c new file mode 100644 index 00000000000..d1705ab8136 --- /dev/null +++ b/fs/cifs/sess.c @@ -0,0 +1,538 @@ +/* + * fs/cifs/sess.c + * + * SMB/CIFS session setup handling routines + * + * Copyright (c) International Business Machines Corp., 2006 + * Author(s): Steve French (sfrench@us.ibm.com) + * + * 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 "cifspdu.h" +#include "cifsglob.h" +#include "cifsproto.h" +#include "cifs_unicode.h" +#include "cifs_debug.h" +#include "ntlmssp.h" +#include "nterr.h" +#include <linux/utsname.h> + +extern void SMBNTencrypt(unsigned char *passwd, unsigned char *c8, + unsigned char *p24); + +static __u32 cifs_ssetup_hdr(struct cifsSesInfo *ses, SESSION_SETUP_ANDX *pSMB) +{ + __u32 capabilities = 0; + + /* init fields common to all four types of SessSetup */ + /* note that header is initialized to zero in header_assemble */ + pSMB->req.AndXCommand = 0xFF; + pSMB->req.MaxBufferSize = cpu_to_le16(ses->server->maxBuf); + pSMB->req.MaxMpxCount = cpu_to_le16(ses->server->maxReq); + + /* Now no need to set SMBFLG_CASELESS or obsolete CANONICAL PATH */ + + /* BB verify whether signing required on neg or just on auth frame + (and NTLM case) */ + + capabilities = CAP_LARGE_FILES | CAP_NT_SMBS | CAP_LEVEL_II_OPLOCKS | + CAP_LARGE_WRITE_X | CAP_LARGE_READ_X; + + if(ses->server->secMode & (SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED)) + pSMB->req.hdr.Flags2 |= SMBFLG2_SECURITY_SIGNATURE; + + if (ses->capabilities & CAP_UNICODE) { + pSMB->req.hdr.Flags2 |= SMBFLG2_UNICODE; + capabilities |= CAP_UNICODE; + } + if (ses->capabilities & CAP_STATUS32) { + pSMB->req.hdr.Flags2 |= SMBFLG2_ERR_STATUS; + capabilities |= CAP_STATUS32; + } + if (ses->capabilities & CAP_DFS) { + pSMB->req.hdr.Flags2 |= SMBFLG2_DFS; + capabilities |= CAP_DFS; + } + if (ses->capabilities & CAP_UNIX) { + capabilities |= CAP_UNIX; + } + + /* BB check whether to init vcnum BB */ + return capabilities; +} + +static void unicode_ssetup_strings(char ** pbcc_area, struct cifsSesInfo *ses, + const struct nls_table * nls_cp) +{ + char * bcc_ptr = *pbcc_area; + int bytes_ret = 0; + + /* BB FIXME add check that strings total less + than 335 or will need to send them as arrays */ + + /* unicode strings, must be word aligned before the call */ +/* if ((long) bcc_ptr % 2) { + *bcc_ptr = 0; + bcc_ptr++; + } */ + /* copy user */ + if(ses->userName == NULL) { + /* BB what about null user mounts - check that we do this BB */ + } else { /* 300 should be long enough for any conceivable user name */ + bytes_ret = cifs_strtoUCS((__le16 *) bcc_ptr, ses->userName, + 300, nls_cp); + } + bcc_ptr += 2 * bytes_ret; + bcc_ptr += 2; /* account for null termination */ + /* copy domain */ + if(ses->domainName == NULL) + bytes_ret = cifs_strtoUCS((__le16 *) bcc_ptr, + "CIFS_LINUX_DOM", 32, nls_cp); + else + bytes_ret = cifs_strtoUCS((__le16 *) bcc_ptr, ses->domainName, + 256, nls_cp); + bcc_ptr += 2 * bytes_ret; + bcc_ptr += 2; /* account for null terminator */ + + /* Copy OS version */ + bytes_ret = cifs_strtoUCS((__le16 *)bcc_ptr, "Linux version ", 32, + nls_cp); + bcc_ptr += 2 * bytes_ret; + bytes_ret = cifs_strtoUCS((__le16 *) bcc_ptr, system_utsname.release, + 32, nls_cp); + bcc_ptr += 2 * bytes_ret; + bcc_ptr += 2; /* trailing null */ + + bytes_ret = cifs_strtoUCS((__le16 *) bcc_ptr, CIFS_NETWORK_OPSYS, + 32, nls_cp); + bcc_ptr += 2 * bytes_ret; + bcc_ptr += 2; /* trailing null */ + + *pbcc_area = bcc_ptr; +} + +static void ascii_ssetup_strings(char ** pbcc_area, struct cifsSesInfo *ses, + const struct nls_table * nls_cp) +{ + char * bcc_ptr = *pbcc_area; + + /* copy user */ + /* BB what about null user mounts - check that we do this BB */ + /* copy user */ + if(ses->userName == NULL) { + /* BB what about null user mounts - check that we do this BB */ + } else { /* 300 should be long enough for any conceivable user name */ + strncpy(bcc_ptr, ses->userName, 300); + } + /* BB improve check for overflow */ + bcc_ptr += strnlen(ses->userName, 300); + *bcc_ptr = 0; + bcc_ptr++; /* account for null termination */ + + /* copy domain */ + + if(ses->domainName == NULL) { + strcpy(bcc_ptr, "CIFS_LINUX_DOM"); + bcc_ptr += 14; /* strlen(CIFS_LINUX_DOM) */ + } else { + strncpy(bcc_ptr, ses->domainName, 256); + bcc_ptr += strnlen(ses->domainName, 256); + } + *bcc_ptr = 0; + bcc_ptr++; + + /* BB check for overflow here */ + + strcpy(bcc_ptr, "Linux version "); + bcc_ptr += strlen("Linux version "); + strcpy(bcc_ptr, system_utsname.release); + bcc_ptr += strlen(system_utsname.release) + 1; + + strcpy(bcc_ptr, CIFS_NETWORK_OPSYS); + bcc_ptr += strlen(CIFS_NETWORK_OPSYS) + 1; + + *pbcc_area = bcc_ptr; +} + +static int decode_unicode_ssetup(char ** pbcc_area, int bleft, struct cifsSesInfo *ses, + const struct nls_table * nls_cp) +{ + int rc = 0; + int words_left, len; + char * data = *pbcc_area; + + + + cFYI(1,("bleft %d",bleft)); + + + /* word align, if bytes remaining is not even */ + if(bleft % 2) { + bleft--; + data++; + } + words_left = bleft / 2; + + /* save off server operating system */ + len = UniStrnlen((wchar_t *) data, words_left); + +/* We look for obvious messed up bcc or strings in response so we do not go off + the end since (at least) WIN2K and Windows XP have a major bug in not null + terminating last Unicode string in response */ + if(len >= words_left) + return rc; + + if(ses->serverOS) + kfree(ses->serverOS); + /* UTF-8 string will not grow more than four times as big as UCS-16 */ + ses->serverOS = kzalloc(4 * len, GFP_KERNEL); + if(ses->serverOS != NULL) { + cifs_strfromUCS_le(ses->serverOS, (__le16 *)data, len, + nls_cp); + } + data += 2 * (len + 1); + words_left -= len + 1; + + /* save off server network operating system */ + len = UniStrnlen((wchar_t *) data, words_left); + + if(len >= words_left) + return rc; + + if(ses->serverNOS) + kfree(ses->serverNOS); + ses->serverNOS = kzalloc(4 * len, GFP_KERNEL); /* BB this is wrong length FIXME BB */ + if(ses->serverNOS != NULL) { + cifs_strfromUCS_le(ses->serverNOS, (__le16 *)data, len, + nls_cp); + if(strncmp(ses->serverNOS, "NT LAN Manager 4",16) == 0) { + cFYI(1,("NT4 server")); + ses->flags |= CIFS_SES_NT4; + } + } + data += 2 * (len + 1); + words_left -= len + 1; + + /* save off server domain */ + len = UniStrnlen((wchar_t *) data, words_left); + + if(len > words_left) + return rc; + + if(ses->serverDomain) + kfree(ses->serverDomain); + ses->serverDomain = kzalloc(2 * (len + 1), GFP_KERNEL); /* BB FIXME wrong length */ + if(ses->serverDomain != NULL) { + cifs_strfromUCS_le(ses->serverDomain, (__le16 *)data, len, + nls_cp); + ses->serverDomain[2*len] = 0; + ses->serverDomain[(2*len) + 1] = 0; + } + data += 2 * (len + 1); + words_left -= len + 1; + + cFYI(1,("words left: %d",words_left)); + + return rc; +} + +static int decode_ascii_ssetup(char ** pbcc_area, int bleft, struct cifsSesInfo *ses, + const struct nls_table * nls_cp) +{ + int rc = 0; + int len; + char * bcc_ptr = *pbcc_area; + + cFYI(1,("decode sessetup ascii. bleft %d", bleft)); + + len = strnlen(bcc_ptr, bleft); + if(len >= bleft) + return rc; + + if(ses->serverOS) + kfree(ses->serverOS); + + ses->serverOS = kzalloc(len + 1, GFP_KERNEL); + if(ses->serverOS) + strncpy(ses->serverOS, bcc_ptr, len); + + bcc_ptr += len + 1; + bleft -= len + 1; + + len = strnlen(bcc_ptr, bleft); + if(len >= bleft) + return rc; + + if(ses->serverNOS) + kfree(ses->serverNOS); + + ses->serverNOS = kzalloc(len + 1, GFP_KERNEL); + if(ses->serverNOS) + strncpy(ses->serverNOS, bcc_ptr, len); + + bcc_ptr += len + 1; + bleft -= len + 1; + + len = strnlen(bcc_ptr, bleft); + if(len > bleft) + return rc; + + if(ses->serverDomain) + kfree(ses->serverDomain); + + ses->serverDomain = kzalloc(len + 1, GFP_KERNEL); + if(ses->serverOS) + strncpy(ses->serverOS, bcc_ptr, len); + + bcc_ptr += len + 1; + bleft -= len + 1; + + cFYI(1,("ascii: bytes left %d",bleft)); + + return rc; +} + +int +CIFS_SessSetup(unsigned int xid, struct cifsSesInfo *ses, int first_time, + const struct nls_table *nls_cp) +{ + int rc = 0; + int wct; + struct smb_hdr *smb_buf; + char *bcc_ptr; + char *str_area; + SESSION_SETUP_ANDX *pSMB; + __u32 capabilities; + int count; + int resp_buf_type = 0; + struct kvec iov[2]; + enum securityEnum type; + __u16 action; + int bytes_remaining; + + if(ses == NULL) + return -EINVAL; + + type = ses->server->secType; + + cFYI(1,("sess setup type %d",type)); + if(type == LANMAN) { +#ifndef CONFIG_CIFS_WEAK_PW_HASH + /* LANMAN and plaintext are less secure and off by default. + So we make this explicitly be turned on in kconfig (in the + build) and turned on at runtime (changed from the default) + in proc/fs/cifs or via mount parm. Unfortunately this is + needed for old Win (e.g. Win95), some obscure NAS and OS/2 */ + return -EOPNOTSUPP; +#endif + wct = 10; /* lanman 2 style sessionsetup */ + } else if((type == NTLM) || (type == NTLMv2)) { + /* For NTLMv2 failures eventually may need to retry NTLM */ + wct = 13; /* old style NTLM sessionsetup */ + } else /* same size for negotiate or auth, NTLMSSP or extended security */ + wct = 12; + + rc = small_smb_init_no_tc(SMB_COM_SESSION_SETUP_ANDX, wct, ses, + (void **)&smb_buf); + if(rc) + return rc; + + pSMB = (SESSION_SETUP_ANDX *)smb_buf; + + capabilities = cifs_ssetup_hdr(ses, pSMB); + + /* we will send the SMB in two pieces, + a fixed length beginning part, and a + second part which will include the strings + and rest of bcc area, in order to avoid having + to do a large buffer 17K allocation */ + iov[0].iov_base = (char *)pSMB; + iov[0].iov_len = smb_buf->smb_buf_length + 4; + + /* 2000 big enough to fit max user, domain, NOS name etc. */ + str_area = kmalloc(2000, GFP_KERNEL); + bcc_ptr = str_area; + + if(type == LANMAN) { +#ifdef CONFIG_CIFS_WEAK_PW_HASH + char lnm_session_key[CIFS_SESS_KEY_SIZE]; + + /* no capabilities flags in old lanman negotiation */ + + pSMB->old_req.PasswordLength = cpu_to_le16(CIFS_SESS_KEY_SIZE); + /* BB calculate hash with password */ + /* and copy into bcc */ + + calc_lanman_hash(ses, lnm_session_key); + +/* #ifdef CONFIG_CIFS_DEBUG2 + cifs_dump_mem("cryptkey: ",ses->server->cryptKey, + CIFS_SESS_KEY_SIZE); +#endif */ + memcpy(bcc_ptr, (char *)lnm_session_key, CIFS_SESS_KEY_SIZE); + bcc_ptr += CIFS_SESS_KEY_SIZE; + + /* can not sign if LANMAN negotiated so no need + to calculate signing key? but what if server + changed to do higher than lanman dialect and + we reconnected would we ever calc signing_key? */ + + cFYI(1,("Negotiating LANMAN setting up strings")); + /* Unicode not allowed for LANMAN dialects */ + ascii_ssetup_strings(&bcc_ptr, ses, nls_cp); +#endif + } else if (type == NTLM) { + char ntlm_session_key[CIFS_SESS_KEY_SIZE]; + + pSMB->req_no_secext.Capabilities = cpu_to_le32(capabilities); + pSMB->req_no_secext.CaseInsensitivePasswordLength = + cpu_to_le16(CIFS_SESS_KEY_SIZE); + pSMB->req_no_secext.CaseSensitivePasswordLength = + cpu_to_le16(CIFS_SESS_KEY_SIZE); + + /* calculate session key */ + SMBNTencrypt(ses->password, ses->server->cryptKey, + ntlm_session_key); + + if(first_time) /* should this be moved into common code + with similar ntlmv2 path? */ + cifs_calculate_mac_key(ses->server->mac_signing_key, + ntlm_session_key, ses->password); + /* copy session key */ + + memcpy(bcc_ptr, (char *)ntlm_session_key,CIFS_SESS_KEY_SIZE); + bcc_ptr += CIFS_SESS_KEY_SIZE; + memcpy(bcc_ptr, (char *)ntlm_session_key,CIFS_SESS_KEY_SIZE); + bcc_ptr += CIFS_SESS_KEY_SIZE; + if(ses->capabilities & CAP_UNICODE) { + /* unicode strings must be word aligned */ + if (iov[0].iov_len % 2) { + *bcc_ptr = 0; + bcc_ptr++; + } + unicode_ssetup_strings(&bcc_ptr, ses, nls_cp); + } else + ascii_ssetup_strings(&bcc_ptr, ses, nls_cp); + } else if (type == NTLMv2) { + char * v2_sess_key = + kmalloc(sizeof(struct ntlmv2_resp), GFP_KERNEL); + + /* BB FIXME change all users of v2_sess_key to + struct ntlmv2_resp */ + + if(v2_sess_key == NULL) { + cifs_small_buf_release(smb_buf); + return -ENOMEM; + } + + pSMB->req_no_secext.Capabilities = cpu_to_le32(capabilities); + + /* LM2 password would be here if we supported it */ + pSMB->req_no_secext.CaseInsensitivePasswordLength = 0; + /* cpu_to_le16(LM2_SESS_KEY_SIZE); */ + + pSMB->req_no_secext.CaseSensitivePasswordLength = + cpu_to_le16(sizeof(struct ntlmv2_resp)); + + /* calculate session key */ + setup_ntlmv2_rsp(ses, v2_sess_key, nls_cp); + if(first_time) /* should this be moved into common code + with similar ntlmv2 path? */ + /* cifs_calculate_ntlmv2_mac_key(ses->server->mac_signing_key, + response BB FIXME, v2_sess_key); */ + + /* copy session key */ + + /* memcpy(bcc_ptr, (char *)ntlm_session_key,LM2_SESS_KEY_SIZE); + bcc_ptr += LM2_SESS_KEY_SIZE; */ + memcpy(bcc_ptr, (char *)v2_sess_key, sizeof(struct ntlmv2_resp)); + bcc_ptr += sizeof(struct ntlmv2_resp); + kfree(v2_sess_key); + if(ses->capabilities & CAP_UNICODE) { + if(iov[0].iov_len % 2) { + *bcc_ptr = 0; + } bcc_ptr++; + unicode_ssetup_strings(&bcc_ptr, ses, nls_cp); + } else + ascii_ssetup_strings(&bcc_ptr, ses, nls_cp); + } else /* NTLMSSP or SPNEGO */ { + pSMB->req.hdr.Flags2 |= SMBFLG2_EXT_SEC; + capabilities |= CAP_EXTENDED_SECURITY; + pSMB->req.Capabilities = cpu_to_le32(capabilities); + /* BB set password lengths */ + } + + count = (long) bcc_ptr - (long) str_area; + smb_buf->smb_buf_length += count; + + BCC_LE(smb_buf) = cpu_to_le16(count); + + iov[1].iov_base = str_area; + iov[1].iov_len = count; + rc = SendReceive2(xid, ses, iov, 2 /* num_iovecs */, &resp_buf_type, 0); + /* SMB request buf freed in SendReceive2 */ + + cFYI(1,("ssetup rc from sendrecv2 is %d",rc)); + if(rc) + goto ssetup_exit; + + pSMB = (SESSION_SETUP_ANDX *)iov[0].iov_base; + smb_buf = (struct smb_hdr *)iov[0].iov_base; + + if((smb_buf->WordCount != 3) && (smb_buf->WordCount != 4)) { + rc = -EIO; + cERROR(1,("bad word count %d", smb_buf->WordCount)); + goto ssetup_exit; + } + action = le16_to_cpu(pSMB->resp.Action); + if (action & GUEST_LOGIN) + cFYI(1, ("Guest login")); /* BB mark SesInfo struct? */ + ses->Suid = smb_buf->Uid; /* UID left in wire format (le) */ + cFYI(1, ("UID = %d ", ses->Suid)); + /* response can have either 3 or 4 word count - Samba sends 3 */ + /* and lanman response is 3 */ + bytes_remaining = BCC(smb_buf); + bcc_ptr = pByteArea(smb_buf); + + if(smb_buf->WordCount == 4) { + __u16 blob_len; + blob_len = le16_to_cpu(pSMB->resp.SecurityBlobLength); + bcc_ptr += blob_len; + if(blob_len > bytes_remaining) { + cERROR(1,("bad security blob length %d", blob_len)); + rc = -EINVAL; + goto ssetup_exit; + } + bytes_remaining -= blob_len; + } + + /* BB check if Unicode and decode strings */ + if(smb_buf->Flags2 & SMBFLG2_UNICODE) + rc = decode_unicode_ssetup(&bcc_ptr, bytes_remaining, + ses, nls_cp); + else + rc = decode_ascii_ssetup(&bcc_ptr, bytes_remaining, ses,nls_cp); + +ssetup_exit: + kfree(str_area); + if(resp_buf_type == CIFS_SMALL_BUFFER) { + cFYI(1,("ssetup freeing small buf %p", iov[0].iov_base)); + cifs_small_buf_release(iov[0].iov_base); + } else if(resp_buf_type == CIFS_LARGE_BUFFER) + cifs_buf_release(iov[0].iov_base); + + return rc; +} diff --git a/fs/cifs/smbencrypt.c b/fs/cifs/smbencrypt.c index 6103bcdfb16..f518c5e4503 100644 --- a/fs/cifs/smbencrypt.c +++ b/fs/cifs/smbencrypt.c @@ -30,6 +30,7 @@ #include <linux/random.h> #include "cifs_unicode.h" #include "cifspdu.h" +#include "cifsglob.h" #include "md5.h" #include "cifs_debug.h" #include "cifsencrypt.h" diff --git a/fs/cifs/smberr.h b/fs/cifs/smberr.h index cd41c67ff8d..212c3c29640 100644 --- a/fs/cifs/smberr.h +++ b/fs/cifs/smberr.h @@ -95,6 +95,7 @@ #define ERRinvlevel 124 #define ERRdirnotempty 145 #define ERRnotlocked 158 +#define ERRcancelviolation 173 #define ERRalreadyexists 183 #define ERRbadpipe 230 #define ERRpipebusy 231 diff --git a/fs/cifs/transport.c b/fs/cifs/transport.c index 3da80409466..48d47b46b1f 100644 --- a/fs/cifs/transport.c +++ b/fs/cifs/transport.c @@ -3,7 +3,8 @@ * * Copyright (C) International Business Machines Corp., 2002,2005 * Author(s): Steve French (sfrench@us.ibm.com) - * + * Jeremy Allison (jra@samba.org) 2006. + * * 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 @@ -36,7 +37,7 @@ extern mempool_t *cifs_mid_poolp; extern kmem_cache_t *cifs_oplock_cachep; static struct mid_q_entry * -AllocMidQEntry(struct smb_hdr *smb_buffer, struct cifsSesInfo *ses) +AllocMidQEntry(const struct smb_hdr *smb_buffer, struct cifsSesInfo *ses) { struct mid_q_entry *temp; @@ -203,6 +204,10 @@ smb_send(struct socket *ssocket, struct smb_hdr *smb_buffer, rc = 0; } + /* Don't want to modify the buffer as a + side effect of this call. */ + smb_buffer->smb_buf_length = smb_buf_length; + return rc; } @@ -217,6 +222,7 @@ smb_send2(struct socket *ssocket, struct kvec *iov, int n_vec, unsigned int len = iov[0].iov_len; unsigned int total_len; int first_vec = 0; + unsigned int smb_buf_length = smb_buffer->smb_buf_length; if(ssocket == NULL) return -ENOTSOCK; /* BB eventually add reconnect code here */ @@ -293,36 +299,15 @@ smb_send2(struct socket *ssocket, struct kvec *iov, int n_vec, } else rc = 0; + /* Don't want to modify the buffer as a + side effect of this call. */ + smb_buffer->smb_buf_length = smb_buf_length; + return rc; } -int -SendReceive2(const unsigned int xid, struct cifsSesInfo *ses, - struct kvec *iov, int n_vec, int * pRespBufType /* ret */, - const int long_op) +static int wait_for_free_request(struct cifsSesInfo *ses, const int long_op) { - int rc = 0; - unsigned int receive_len; - unsigned long timeout; - struct mid_q_entry *midQ; - struct smb_hdr *in_buf = iov[0].iov_base; - - *pRespBufType = CIFS_NO_BUFFER; /* no response buf yet */ - - if ((ses == NULL) || (ses->server == NULL)) { - cifs_small_buf_release(in_buf); - cERROR(1,("Null session")); - return -EIO; - } - - if(ses->server->tcpStatus == CifsExiting) { - cifs_small_buf_release(in_buf); - return -ENOENT; - } - - /* Ensure that we do not send more than 50 overlapping requests - to the same server. We may make this configurable later or - use ses->maxReq */ if(long_op == -1) { /* oplock breaks must not be held up */ atomic_inc(&ses->server->inFlight); @@ -345,53 +330,140 @@ SendReceive2(const unsigned int xid, struct cifsSesInfo *ses, } else { if(ses->server->tcpStatus == CifsExiting) { spin_unlock(&GlobalMid_Lock); - cifs_small_buf_release(in_buf); return -ENOENT; } - /* can not count locking commands against total since - they are allowed to block on server */ + /* can not count locking commands against total since + they are allowed to block on server */ - if(long_op < 3) { /* update # of requests on the wire to server */ + if (long_op < 3) atomic_inc(&ses->server->inFlight); - } spin_unlock(&GlobalMid_Lock); break; } } } - /* make sure that we sign in the same order that we send on this socket - and avoid races inside tcp sendmsg code that could cause corruption - of smb data */ - - down(&ses->server->tcpSem); + return 0; +} +static int allocate_mid(struct cifsSesInfo *ses, struct smb_hdr *in_buf, + struct mid_q_entry **ppmidQ) +{ if (ses->server->tcpStatus == CifsExiting) { - rc = -ENOENT; - goto out_unlock2; + return -ENOENT; } else if (ses->server->tcpStatus == CifsNeedReconnect) { cFYI(1,("tcp session dead - return to caller to retry")); - rc = -EAGAIN; - goto out_unlock2; + return -EAGAIN; } else if (ses->status != CifsGood) { /* check if SMB session is bad because we are setting it up */ if((in_buf->Command != SMB_COM_SESSION_SETUP_ANDX) && (in_buf->Command != SMB_COM_NEGOTIATE)) { - rc = -EAGAIN; - goto out_unlock2; + return -EAGAIN; } /* else ok - we are setting up session */ } - midQ = AllocMidQEntry(in_buf, ses); - if (midQ == NULL) { + *ppmidQ = AllocMidQEntry(in_buf, ses); + if (*ppmidQ == NULL) { + return -ENOMEM; + } + return 0; +} + +static int wait_for_response(struct cifsSesInfo *ses, + struct mid_q_entry *midQ, + unsigned long timeout, + unsigned long time_to_wait) +{ + unsigned long curr_timeout; + + for (;;) { + curr_timeout = timeout + jiffies; + wait_event(ses->server->response_q, + (!(midQ->midState == MID_REQUEST_SUBMITTED)) || + time_after(jiffies, curr_timeout) || + ((ses->server->tcpStatus != CifsGood) && + (ses->server->tcpStatus != CifsNew))); + + if (time_after(jiffies, curr_timeout) && + (midQ->midState == MID_REQUEST_SUBMITTED) && + ((ses->server->tcpStatus == CifsGood) || + (ses->server->tcpStatus == CifsNew))) { + + unsigned long lrt; + + /* We timed out. Is the server still + sending replies ? */ + spin_lock(&GlobalMid_Lock); + lrt = ses->server->lstrp; + spin_unlock(&GlobalMid_Lock); + + /* Calculate time_to_wait past last receive time. + Although we prefer not to time out if the + server is still responding - we will time + out if the server takes more than 15 (or 45 + or 180) seconds to respond to this request + and has not responded to any request from + other threads on the client within 10 seconds */ + lrt += time_to_wait; + if (time_after(jiffies, lrt)) { + /* No replies for time_to_wait. */ + cERROR(1,("server not responding")); + return -1; + } + } else { + return 0; + } + } +} + +int +SendReceive2(const unsigned int xid, struct cifsSesInfo *ses, + struct kvec *iov, int n_vec, int * pRespBufType /* ret */, + const int long_op) +{ + int rc = 0; + unsigned int receive_len; + unsigned long timeout; + struct mid_q_entry *midQ; + struct smb_hdr *in_buf = iov[0].iov_base; + + *pRespBufType = CIFS_NO_BUFFER; /* no response buf yet */ + + if ((ses == NULL) || (ses->server == NULL)) { + cifs_small_buf_release(in_buf); + cERROR(1,("Null session")); + return -EIO; + } + + if(ses->server->tcpStatus == CifsExiting) { + cifs_small_buf_release(in_buf); + return -ENOENT; + } + + /* Ensure that we do not send more than 50 overlapping requests + to the same server. We may make this configurable later or + use ses->maxReq */ + + rc = wait_for_free_request(ses, long_op); + if (rc) { + cifs_small_buf_release(in_buf); + return rc; + } + + /* make sure that we sign in the same order that we send on this socket + and avoid races inside tcp sendmsg code that could cause corruption + of smb data */ + + down(&ses->server->tcpSem); + + rc = allocate_mid(ses, in_buf, &midQ); + if (rc) { up(&ses->server->tcpSem); cifs_small_buf_release(in_buf); - /* If not lock req, update # of requests on wire to server */ - if(long_op < 3) { - atomic_dec(&ses->server->inFlight); - wake_up(&ses->server->request_q); - } - return -ENOMEM; + /* Update # of requests on wire to server */ + atomic_dec(&ses->server->inFlight); + wake_up(&ses->server->request_q); + return rc; } rc = cifs_sign_smb2(iov, n_vec, ses->server, &midQ->sequence_number); @@ -406,32 +478,23 @@ SendReceive2(const unsigned int xid, struct cifsSesInfo *ses, atomic_dec(&ses->server->inSend); midQ->when_sent = jiffies; #endif - if(rc < 0) { - DeleteMidQEntry(midQ); - up(&ses->server->tcpSem); - cifs_small_buf_release(in_buf); - /* If not lock req, update # of requests on wire to server */ - if(long_op < 3) { - atomic_dec(&ses->server->inFlight); - wake_up(&ses->server->request_q); - } - return rc; - } else { - up(&ses->server->tcpSem); - cifs_small_buf_release(in_buf); - } + + up(&ses->server->tcpSem); + cifs_small_buf_release(in_buf); + + if(rc < 0) + goto out; if (long_op == -1) - goto cifs_no_response_exit2; + goto out; else if (long_op == 2) /* writes past end of file can take loong time */ timeout = 180 * HZ; else if (long_op == 1) timeout = 45 * HZ; /* should be greater than servers oplock break timeout (about 43 seconds) */ - else if (long_op > 2) { - timeout = MAX_SCHEDULE_TIMEOUT; - } else + else timeout = 15 * HZ; + /* wait for 15 seconds or until woken up due to response arriving or due to last connection to this server being unmounted */ if (signal_pending(current)) { @@ -441,19 +504,7 @@ SendReceive2(const unsigned int xid, struct cifsSesInfo *ses, } /* No user interrupts in wait - wreaks havoc with performance */ - if(timeout != MAX_SCHEDULE_TIMEOUT) { - timeout += jiffies; - wait_event(ses->server->response_q, - (!(midQ->midState & MID_REQUEST_SUBMITTED)) || - time_after(jiffies, timeout) || - ((ses->server->tcpStatus != CifsGood) && - (ses->server->tcpStatus != CifsNew))); - } else { - wait_event(ses->server->response_q, - (!(midQ->midState & MID_REQUEST_SUBMITTED)) || - ((ses->server->tcpStatus != CifsGood) && - (ses->server->tcpStatus != CifsNew))); - } + wait_for_response(ses, midQ, timeout, 10 * HZ); spin_lock(&GlobalMid_Lock); if (midQ->resp_buf) { @@ -481,11 +532,9 @@ SendReceive2(const unsigned int xid, struct cifsSesInfo *ses, } spin_unlock(&GlobalMid_Lock); DeleteMidQEntry(midQ); - /* If not lock req, update # of requests on wire to server */ - if(long_op < 3) { - atomic_dec(&ses->server->inFlight); - wake_up(&ses->server->request_q); - } + /* Update # of requests on wire to server */ + atomic_dec(&ses->server->inFlight); + wake_up(&ses->server->request_q); return rc; } @@ -536,24 +585,12 @@ SendReceive2(const unsigned int xid, struct cifsSesInfo *ses, cFYI(1,("Bad MID state?")); } } -cifs_no_response_exit2: - DeleteMidQEntry(midQ); - - if(long_op < 3) { - atomic_dec(&ses->server->inFlight); - wake_up(&ses->server->request_q); - } - return rc; +out: -out_unlock2: - up(&ses->server->tcpSem); - cifs_small_buf_release(in_buf); - /* If not lock req, update # of requests on wire to server */ - if(long_op < 3) { - atomic_dec(&ses->server->inFlight); - wake_up(&ses->server->request_q); - } + DeleteMidQEntry(midQ); + atomic_dec(&ses->server->inFlight); + wake_up(&ses->server->request_q); return rc; } @@ -583,86 +620,34 @@ SendReceive(const unsigned int xid, struct cifsSesInfo *ses, /* Ensure that we do not send more than 50 overlapping requests to the same server. We may make this configurable later or use ses->maxReq */ - if(long_op == -1) { - /* oplock breaks must not be held up */ - atomic_inc(&ses->server->inFlight); - } else { - spin_lock(&GlobalMid_Lock); - while(1) { - if(atomic_read(&ses->server->inFlight) >= - cifs_max_pending){ - spin_unlock(&GlobalMid_Lock); -#ifdef CONFIG_CIFS_STATS2 - atomic_inc(&ses->server->num_waiters); -#endif - wait_event(ses->server->request_q, - atomic_read(&ses->server->inFlight) - < cifs_max_pending); -#ifdef CONFIG_CIFS_STATS2 - atomic_dec(&ses->server->num_waiters); -#endif - spin_lock(&GlobalMid_Lock); - } else { - if(ses->server->tcpStatus == CifsExiting) { - spin_unlock(&GlobalMid_Lock); - return -ENOENT; - } - /* can not count locking commands against total since - they are allowed to block on server */ - - if(long_op < 3) { - /* update # of requests on the wire to server */ - atomic_inc(&ses->server->inFlight); - } - spin_unlock(&GlobalMid_Lock); - break; - } - } - } + rc = wait_for_free_request(ses, long_op); + if (rc) + return rc; + /* make sure that we sign in the same order that we send on this socket and avoid races inside tcp sendmsg code that could cause corruption of smb data */ down(&ses->server->tcpSem); - if (ses->server->tcpStatus == CifsExiting) { - rc = -ENOENT; - goto out_unlock; - } else if (ses->server->tcpStatus == CifsNeedReconnect) { - cFYI(1,("tcp session dead - return to caller to retry")); - rc = -EAGAIN; - goto out_unlock; - } else if (ses->status != CifsGood) { - /* check if SMB session is bad because we are setting it up */ - if((in_buf->Command != SMB_COM_SESSION_SETUP_ANDX) && - (in_buf->Command != SMB_COM_NEGOTIATE)) { - rc = -EAGAIN; - goto out_unlock; - } /* else ok - we are setting up session */ - } - midQ = AllocMidQEntry(in_buf, ses); - if (midQ == NULL) { + rc = allocate_mid(ses, in_buf, &midQ); + if (rc) { up(&ses->server->tcpSem); - /* If not lock req, update # of requests on wire to server */ - if(long_op < 3) { - atomic_dec(&ses->server->inFlight); - wake_up(&ses->server->request_q); - } - return -ENOMEM; + /* Update # of requests on wire to server */ + atomic_dec(&ses->server->inFlight); + wake_up(&ses->server->request_q); + return rc; } if (in_buf->smb_buf_length > CIFSMaxBufSize + MAX_CIFS_HDR_SIZE - 4) { - up(&ses->server->tcpSem); - cERROR(1, - ("Illegal length, greater than maximum frame, %d ", + cERROR(1, ("Illegal length, greater than maximum frame, %d", in_buf->smb_buf_length)); DeleteMidQEntry(midQ); - /* If not lock req, update # of requests on wire to server */ - if(long_op < 3) { - atomic_dec(&ses->server->inFlight); - wake_up(&ses->server->request_q); - } + up(&ses->server->tcpSem); + /* Update # of requests on wire to server */ + atomic_dec(&ses->server->inFlight); + wake_up(&ses->server->request_q); return -EIO; } @@ -678,27 +663,19 @@ SendReceive(const unsigned int xid, struct cifsSesInfo *ses, atomic_dec(&ses->server->inSend); midQ->when_sent = jiffies; #endif - if(rc < 0) { - DeleteMidQEntry(midQ); - up(&ses->server->tcpSem); - /* If not lock req, update # of requests on wire to server */ - if(long_op < 3) { - atomic_dec(&ses->server->inFlight); - wake_up(&ses->server->request_q); - } - return rc; - } else - up(&ses->server->tcpSem); + up(&ses->server->tcpSem); + + if(rc < 0) + goto out; + if (long_op == -1) - goto cifs_no_response_exit; + goto out; else if (long_op == 2) /* writes past end of file can take loong time */ timeout = 180 * HZ; else if (long_op == 1) timeout = 45 * HZ; /* should be greater than servers oplock break timeout (about 43 seconds) */ - else if (long_op > 2) { - timeout = MAX_SCHEDULE_TIMEOUT; - } else + else timeout = 15 * HZ; /* wait for 15 seconds or until woken up due to response arriving or due to last connection to this server being unmounted */ @@ -709,19 +686,7 @@ SendReceive(const unsigned int xid, struct cifsSesInfo *ses, } /* No user interrupts in wait - wreaks havoc with performance */ - if(timeout != MAX_SCHEDULE_TIMEOUT) { - timeout += jiffies; - wait_event(ses->server->response_q, - (!(midQ->midState & MID_REQUEST_SUBMITTED)) || - time_after(jiffies, timeout) || - ((ses->server->tcpStatus != CifsGood) && - (ses->server->tcpStatus != CifsNew))); - } else { - wait_event(ses->server->response_q, - (!(midQ->midState & MID_REQUEST_SUBMITTED)) || - ((ses->server->tcpStatus != CifsGood) && - (ses->server->tcpStatus != CifsNew))); - } + wait_for_response(ses, midQ, timeout, 10 * HZ); spin_lock(&GlobalMid_Lock); if (midQ->resp_buf) { @@ -749,11 +714,9 @@ SendReceive(const unsigned int xid, struct cifsSesInfo *ses, } spin_unlock(&GlobalMid_Lock); DeleteMidQEntry(midQ); - /* If not lock req, update # of requests on wire to server */ - if(long_op < 3) { - atomic_dec(&ses->server->inFlight); - wake_up(&ses->server->request_q); - } + /* Update # of requests on wire to server */ + atomic_dec(&ses->server->inFlight); + wake_up(&ses->server->request_q); return rc; } @@ -800,23 +763,253 @@ SendReceive(const unsigned int xid, struct cifsSesInfo *ses, cERROR(1,("Bad MID state?")); } } -cifs_no_response_exit: + +out: + DeleteMidQEntry(midQ); + atomic_dec(&ses->server->inFlight); + wake_up(&ses->server->request_q); - if(long_op < 3) { - atomic_dec(&ses->server->inFlight); - wake_up(&ses->server->request_q); - } + return rc; +} + +/* Send an NT_CANCEL SMB to cause the POSIX blocking lock to return. */ + +static int +send_nt_cancel(struct cifsTconInfo *tcon, struct smb_hdr *in_buf, + struct mid_q_entry *midQ) +{ + int rc = 0; + struct cifsSesInfo *ses = tcon->ses; + __u16 mid = in_buf->Mid; + header_assemble(in_buf, SMB_COM_NT_CANCEL, tcon, 0); + in_buf->Mid = mid; + down(&ses->server->tcpSem); + rc = cifs_sign_smb(in_buf, ses->server, &midQ->sequence_number); + if (rc) { + up(&ses->server->tcpSem); + return rc; + } + rc = smb_send(ses->server->ssocket, in_buf, in_buf->smb_buf_length, + (struct sockaddr *) &(ses->server->addr.sockAddr)); + up(&ses->server->tcpSem); return rc; +} + +/* We send a LOCKINGX_CANCEL_LOCK to cause the Windows + blocking lock to return. */ + +static int +send_lock_cancel(const unsigned int xid, struct cifsTconInfo *tcon, + struct smb_hdr *in_buf, + struct smb_hdr *out_buf) +{ + int bytes_returned; + struct cifsSesInfo *ses = tcon->ses; + LOCK_REQ *pSMB = (LOCK_REQ *)in_buf; + + /* We just modify the current in_buf to change + the type of lock from LOCKING_ANDX_SHARED_LOCK + or LOCKING_ANDX_EXCLUSIVE_LOCK to + LOCKING_ANDX_CANCEL_LOCK. */ + + pSMB->LockType = LOCKING_ANDX_CANCEL_LOCK|LOCKING_ANDX_LARGE_FILES; + pSMB->Timeout = 0; + pSMB->hdr.Mid = GetNextMid(ses->server); + + return SendReceive(xid, ses, in_buf, out_buf, + &bytes_returned, 0); +} + +int +SendReceiveBlockingLock(const unsigned int xid, struct cifsTconInfo *tcon, + struct smb_hdr *in_buf, struct smb_hdr *out_buf, + int *pbytes_returned) +{ + int rc = 0; + int rstart = 0; + unsigned int receive_len; + struct mid_q_entry *midQ; + struct cifsSesInfo *ses; + + if (tcon == NULL || tcon->ses == NULL) { + cERROR(1,("Null smb session")); + return -EIO; + } + ses = tcon->ses; + + if(ses->server == NULL) { + cERROR(1,("Null tcp session")); + return -EIO; + } + + if(ses->server->tcpStatus == CifsExiting) + return -ENOENT; + + /* Ensure that we do not send more than 50 overlapping requests + to the same server. We may make this configurable later or + use ses->maxReq */ -out_unlock: + rc = wait_for_free_request(ses, 3); + if (rc) + return rc; + + /* make sure that we sign in the same order that we send on this socket + and avoid races inside tcp sendmsg code that could cause corruption + of smb data */ + + down(&ses->server->tcpSem); + + rc = allocate_mid(ses, in_buf, &midQ); + if (rc) { + up(&ses->server->tcpSem); + return rc; + } + + if (in_buf->smb_buf_length > CIFSMaxBufSize + MAX_CIFS_HDR_SIZE - 4) { + up(&ses->server->tcpSem); + cERROR(1, ("Illegal length, greater than maximum frame, %d", + in_buf->smb_buf_length)); + DeleteMidQEntry(midQ); + return -EIO; + } + + rc = cifs_sign_smb(in_buf, ses->server, &midQ->sequence_number); + + midQ->midState = MID_REQUEST_SUBMITTED; +#ifdef CONFIG_CIFS_STATS2 + atomic_inc(&ses->server->inSend); +#endif + rc = smb_send(ses->server->ssocket, in_buf, in_buf->smb_buf_length, + (struct sockaddr *) &(ses->server->addr.sockAddr)); +#ifdef CONFIG_CIFS_STATS2 + atomic_dec(&ses->server->inSend); + midQ->when_sent = jiffies; +#endif up(&ses->server->tcpSem); - /* If not lock req, update # of requests on wire to server */ - if(long_op < 3) { - atomic_dec(&ses->server->inFlight); - wake_up(&ses->server->request_q); + + if(rc < 0) { + DeleteMidQEntry(midQ); + return rc; + } + + /* Wait for a reply - allow signals to interrupt. */ + rc = wait_event_interruptible(ses->server->response_q, + (!(midQ->midState == MID_REQUEST_SUBMITTED)) || + ((ses->server->tcpStatus != CifsGood) && + (ses->server->tcpStatus != CifsNew))); + + /* Were we interrupted by a signal ? */ + if ((rc == -ERESTARTSYS) && + (midQ->midState == MID_REQUEST_SUBMITTED) && + ((ses->server->tcpStatus == CifsGood) || + (ses->server->tcpStatus == CifsNew))) { + + if (in_buf->Command == SMB_COM_TRANSACTION2) { + /* POSIX lock. We send a NT_CANCEL SMB to cause the + blocking lock to return. */ + + rc = send_nt_cancel(tcon, in_buf, midQ); + if (rc) { + DeleteMidQEntry(midQ); + return rc; + } + } else { + /* Windows lock. We send a LOCKINGX_CANCEL_LOCK + to cause the blocking lock to return. */ + + rc = send_lock_cancel(xid, tcon, in_buf, out_buf); + + /* If we get -ENOLCK back the lock may have + already been removed. Don't exit in this case. */ + if (rc && rc != -ENOLCK) { + DeleteMidQEntry(midQ); + return rc; + } + } + + /* Wait 5 seconds for the response. */ + if (wait_for_response(ses, midQ, 5 * HZ, 5 * HZ)==0) { + /* We got the response - restart system call. */ + rstart = 1; + } + } + + spin_lock(&GlobalMid_Lock); + if (midQ->resp_buf) { + spin_unlock(&GlobalMid_Lock); + receive_len = midQ->resp_buf->smb_buf_length; + } else { + cERROR(1,("No response for cmd %d mid %d", + midQ->command, midQ->mid)); + if(midQ->midState == MID_REQUEST_SUBMITTED) { + if(ses->server->tcpStatus == CifsExiting) + rc = -EHOSTDOWN; + else { + ses->server->tcpStatus = CifsNeedReconnect; + midQ->midState = MID_RETRY_NEEDED; + } + } + + if (rc != -EHOSTDOWN) { + if(midQ->midState == MID_RETRY_NEEDED) { + rc = -EAGAIN; + cFYI(1,("marking request for retry")); + } else { + rc = -EIO; + } + } + spin_unlock(&GlobalMid_Lock); + DeleteMidQEntry(midQ); + return rc; } + + if (receive_len > CIFSMaxBufSize + MAX_CIFS_HDR_SIZE) { + cERROR(1, ("Frame too large received. Length: %d Xid: %d", + receive_len, xid)); + rc = -EIO; + } else { /* rcvd frame is ok */ + + if (midQ->resp_buf && out_buf + && (midQ->midState == MID_RESPONSE_RECEIVED)) { + out_buf->smb_buf_length = receive_len; + memcpy((char *)out_buf + 4, + (char *)midQ->resp_buf + 4, + receive_len); + + dump_smb(out_buf, 92); + /* convert the length into a more usable form */ + if((receive_len > 24) && + (ses->server->secMode & (SECMODE_SIGN_REQUIRED | + SECMODE_SIGN_ENABLED))) { + rc = cifs_verify_signature(out_buf, + ses->server->mac_signing_key, + midQ->sequence_number+1); + if(rc) { + cERROR(1,("Unexpected SMB signature")); + /* BB FIXME add code to kill session */ + } + } + + *pbytes_returned = out_buf->smb_buf_length; + + /* BB special case reconnect tid and uid here? */ + rc = map_smb_to_linux_error(out_buf); + /* convert ByteCount if necessary */ + if (receive_len >= + sizeof (struct smb_hdr) - + 4 /* do not count RFC1001 header */ + + (2 * out_buf->WordCount) + 2 /* bcc */ ) + BCC(out_buf) = le16_to_cpu(BCC_LE(out_buf)); + } else { + rc = -EIO; + cERROR(1,("Bad MID state?")); + } + } + DeleteMidQEntry(midQ); + if (rstart && rc == -EACCES) + return -ERESTARTSYS; return rc; } diff --git a/fs/cifs/xattr.c b/fs/cifs/xattr.c index 7754d641775..18fcec190f8 100644 --- a/fs/cifs/xattr.c +++ b/fs/cifs/xattr.c @@ -269,7 +269,7 @@ ssize_t cifs_getxattr(struct dentry * direntry, const char * ea_name, rc = CIFSSMBGetCIFSACL(xid, pTcon, fid, ea_value, buf_size, ACL_TYPE_ACCESS); - CIFSSMBClose(xid, pTcon, fid) + CIFSSMBClose(xid, pTcon, fid); } } */ /* BB enable after fixing up return data */ @@ -330,11 +330,15 @@ ssize_t cifs_listxattr(struct dentry * direntry, char * data, size_t buf_size) sb = direntry->d_inode->i_sb; if(sb == NULL) return -EIO; - xid = GetXid(); cifs_sb = CIFS_SB(sb); pTcon = cifs_sb->tcon; + if(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NO_XATTR) + return -EOPNOTSUPP; + + xid = GetXid(); + full_path = build_path_from_dentry(direntry); if(full_path == NULL) { FreeXid(xid); diff --git a/fs/coda/coda_linux.c b/fs/coda/coda_linux.c index 5597080cb81..95a54253c04 100644 --- a/fs/coda/coda_linux.c +++ b/fs/coda/coda_linux.c @@ -110,8 +110,6 @@ void coda_vattr_to_iattr(struct inode *inode, struct coda_vattr *attr) inode->i_nlink = attr->va_nlink; if (attr->va_size != -1) inode->i_size = attr->va_size; - if (attr->va_blocksize != -1) - inode->i_blksize = attr->va_blocksize; if (attr->va_size != -1) inode->i_blocks = (attr->va_size + 511) >> 9; if (attr->va_atime.tv_sec != -1) diff --git a/fs/coda/dir.c b/fs/coda/dir.c index 71f2ea632e5..8651ea6a23b 100644 --- a/fs/coda/dir.c +++ b/fs/coda/dir.c @@ -513,7 +513,7 @@ static int coda_venus_readdir(struct file *filp, filldir_t filldir, ino_t ino; int ret, i; - vdir = (struct venus_dirent *)kmalloc(sizeof(*vdir), GFP_KERNEL); + vdir = kmalloc(sizeof(*vdir), GFP_KERNEL); if (!vdir) return -ENOMEM; i = filp->f_pos; diff --git a/fs/coda/file.c b/fs/coda/file.c index cc66c681bd1..dbfbcfa5b3c 100644 --- a/fs/coda/file.c +++ b/fs/coda/file.c @@ -136,10 +136,8 @@ int coda_open(struct inode *coda_inode, struct file *coda_file) coda_vfs_stat.open++; cfi = kmalloc(sizeof(struct coda_file_info), GFP_KERNEL); - if (!cfi) { - unlock_kernel(); + if (!cfi) return -ENOMEM; - } lock_kernel(); diff --git a/fs/coda/inode.c b/fs/coda/inode.c index 87f1dc8aa24..88d12332116 100644 --- a/fs/coda/inode.c +++ b/fs/coda/inode.c @@ -80,8 +80,7 @@ int coda_init_inodecache(void) void coda_destroy_inodecache(void) { - if (kmem_cache_destroy(coda_inode_cachep)) - printk(KERN_INFO "coda_inode_cache: not all structures were freed\n"); + kmem_cache_destroy(coda_inode_cachep); } static int coda_remount(struct super_block *sb, int *flags, char *data) diff --git a/fs/coda/psdev.c b/fs/coda/psdev.c index 6c6771db36d..803aacf0d49 100644 --- a/fs/coda/psdev.c +++ b/fs/coda/psdev.c @@ -28,7 +28,6 @@ #include <linux/delay.h> #include <linux/skbuff.h> #include <linux/proc_fs.h> -#include <linux/devfs_fs_kernel.h> #include <linux/vmalloc.h> #include <linux/fs.h> #include <linux/file.h> @@ -259,7 +258,7 @@ static ssize_t coda_psdev_read(struct file * file, char __user * buf, /* If request was not a signal, enqueue and don't free */ if (!(req->uc_flags & REQ_ASYNC)) { req->uc_flags |= REQ_READ; - list_add(&(req->uc_chain), vcp->vc_processing.prev); + list_add_tail(&(req->uc_chain), &vcp->vc_processing); goto out; } @@ -365,22 +364,12 @@ static int init_coda_psdev(void) err = PTR_ERR(coda_psdev_class); goto out_chrdev; } - devfs_mk_dir ("coda"); - for (i = 0; i < MAX_CODADEVS; i++) { + for (i = 0; i < MAX_CODADEVS; i++) class_device_create(coda_psdev_class, NULL, MKDEV(CODA_PSDEV_MAJOR,i), NULL, "cfs%d", i); - err = devfs_mk_cdev(MKDEV(CODA_PSDEV_MAJOR, i), - S_IFCHR|S_IRUSR|S_IWUSR, "coda/%d", i); - if (err) - goto out_class; - } coda_sysctl_init(); goto out; -out_class: - for (i = 0; i < MAX_CODADEVS; i++) - class_device_destroy(coda_psdev_class, MKDEV(CODA_PSDEV_MAJOR, i)); - class_destroy(coda_psdev_class); out_chrdev: unregister_chrdev(CODA_PSDEV_MAJOR, "coda"); out: @@ -419,12 +408,9 @@ static int __init init_coda(void) } return 0; out: - for (i = 0; i < MAX_CODADEVS; i++) { + for (i = 0; i < MAX_CODADEVS; i++) class_device_destroy(coda_psdev_class, MKDEV(CODA_PSDEV_MAJOR, i)); - devfs_remove("coda/%d", i); - } class_destroy(coda_psdev_class); - devfs_remove("coda"); unregister_chrdev(CODA_PSDEV_MAJOR, "coda"); coda_sysctl_clean(); out1: @@ -441,12 +427,9 @@ static void __exit exit_coda(void) if ( err != 0 ) { printk("coda: failed to unregister filesystem\n"); } - for (i = 0; i < MAX_CODADEVS; i++) { + for (i = 0; i < MAX_CODADEVS; i++) class_device_destroy(coda_psdev_class, MKDEV(CODA_PSDEV_MAJOR, i)); - devfs_remove("coda/%d", i); - } class_destroy(coda_psdev_class); - devfs_remove("coda"); unregister_chrdev(CODA_PSDEV_MAJOR, "coda"); coda_sysctl_clean(); coda_destroy_inodecache(); diff --git a/fs/coda/symlink.c b/fs/coda/symlink.c index b35e5bbd9c9..76e00a65a75 100644 --- a/fs/coda/symlink.c +++ b/fs/coda/symlink.c @@ -50,6 +50,6 @@ fail: return error; } -struct address_space_operations coda_symlink_aops = { +const struct address_space_operations coda_symlink_aops = { .readpage = coda_symlink_filler, }; diff --git a/fs/coda/sysctl.c b/fs/coda/sysctl.c index f0b10757288..1c82e9a7d7c 100644 --- a/fs/coda/sysctl.c +++ b/fs/coda/sysctl.c @@ -11,7 +11,6 @@ * */ -#include <linux/config.h> #include <linux/time.h> #include <linux/mm.h> #include <linux/sysctl.h> diff --git a/fs/coda/upcall.c b/fs/coda/upcall.c index b040eba13a7..a5b5e631ba6 100644 --- a/fs/coda/upcall.c +++ b/fs/coda/upcall.c @@ -725,7 +725,7 @@ static int coda_upcall(struct coda_sb_info *sbi, ((union inputArgs *)buffer)->ih.unique = req->uc_unique; /* Append msg to pending queue and poke Venus. */ - list_add(&(req->uc_chain), vcommp->vc_pending.prev); + list_add_tail(&(req->uc_chain), &vcommp->vc_pending); wake_up_interruptible(&vcommp->vc_waitq); /* We can be interrupted while we wait for Venus to process diff --git a/fs/compat.c b/fs/compat.c index 7e7e5bc4f3c..ce982f6e8c8 100644 --- a/fs/compat.c +++ b/fs/compat.c @@ -55,6 +55,20 @@ extern void sigset_from_compat(sigset_t *set, compat_sigset_t *compat); +int compat_log = 1; + +int compat_printk(const char *fmt, ...) +{ + va_list ap; + int ret; + if (!compat_log) + return 0; + va_start(ap, fmt); + ret = vprintk(fmt, ap); + va_end(ap); + return ret; +} + /* * Not all architectures have sys_utime, so implement this in terms * of sys_utimes. @@ -359,7 +373,7 @@ static void compat_ioctl_error(struct file *filp, unsigned int fd, sprintf(buf,"'%c'", (cmd>>24) & 0x3f); if (!isprint(buf[1])) sprintf(buf, "%02x", buf[1]); - printk("ioctl32(%s:%d): Unknown cmd fd(%d) " + compat_printk("ioctl32(%s:%d): Unknown cmd fd(%d) " "cmd(%08x){%s} arg(%08x) on %s\n", current->comm, current->pid, (int)fd, (unsigned int)cmd, buf, @@ -1841,7 +1855,7 @@ asmlinkage long compat_sys_pselect7(int n, compat_ulong_t __user *inp, } while (!ret && !timeout && tsp && (ts.tv_sec || ts.tv_nsec)); - if (tsp && !(current->personality & STICKY_TIMEOUTS)) { + if (ret == 0 && tsp && !(current->personality & STICKY_TIMEOUTS)) { struct compat_timespec rts; rts.tv_sec = timeout / HZ; @@ -1852,7 +1866,8 @@ asmlinkage long compat_sys_pselect7(int n, compat_ulong_t __user *inp, } if (compat_timespec_compare(&rts, &ts) >= 0) rts = ts; - copy_to_user(tsp, &rts, sizeof(rts)); + if (copy_to_user(tsp, &rts, sizeof(rts))) + ret = -EFAULT; } if (ret == -ERESTARTNOHAND) { diff --git a/fs/compat_ioctl.c b/fs/compat_ioctl.c index d2c38875ab2..4063a939697 100644 --- a/fs/compat_ioctl.c +++ b/fs/compat_ioctl.c @@ -10,7 +10,6 @@ * ioctls. */ -#include <linux/config.h> #include <linux/types.h> #include <linux/compat.h> #include <linux/kernel.h> @@ -44,7 +43,6 @@ #include <linux/loop.h> #include <linux/auto_fs.h> #include <linux/auto_fs4.h> -#include <linux/devfs_fs.h> #include <linux/tty.h> #include <linux/vt_kern.h> #include <linux/fb.h> @@ -80,6 +78,7 @@ #include <net/bluetooth/rfcomm.h> #include <linux/capi.h> +#include <linux/gigaset_dev.h> #include <scsi/scsi.h> #include <scsi/scsi_ioctl.h> @@ -205,38 +204,6 @@ static int do_ext3_ioctl(unsigned int fd, unsigned int cmd, unsigned long arg) return sys_ioctl(fd, cmd, (unsigned long)compat_ptr(arg)); } -struct compat_dmx_event { - dmx_event_t event; - compat_time_t timeStamp; - union - { - dmx_scrambling_status_t scrambling; - } u; -}; - -static int do_dmx_get_event(unsigned int fd, unsigned int cmd, unsigned long arg) -{ - struct dmx_event kevent; - mm_segment_t old_fs = get_fs(); - int err; - - set_fs(KERNEL_DS); - err = sys_ioctl(fd, cmd, (unsigned long) &kevent); - set_fs(old_fs); - - if (!err) { - struct compat_dmx_event __user *up = compat_ptr(arg); - - err = put_user(kevent.event, &up->event); - err |= put_user(kevent.timeStamp, &up->timeStamp); - err |= put_user(kevent.u.scrambling, &up->u.scrambling); - if (err) - err = -EFAULT; - } - - return err; -} - struct compat_video_event { int32_t type; compat_time_t timestamp; @@ -2964,7 +2931,6 @@ HANDLE_IOCTL(NCP_IOC_SETPRIVATEDATA_32, do_ncp_setprivatedata) #endif /* dvb */ -HANDLE_IOCTL(DMX_GET_EVENT, do_dmx_get_event) HANDLE_IOCTL(VIDEO_GET_EVENT, do_video_get_event) HANDLE_IOCTL(VIDEO_STILLPICTURE, do_video_stillpicture) HANDLE_IOCTL(VIDEO_SET_SPU_PALETTE, do_video_set_spu_palette) diff --git a/fs/configfs/dir.c b/fs/configfs/dir.c index 5f952187fc5..816e8ef6456 100644 --- a/fs/configfs/dir.c +++ b/fs/configfs/dir.c @@ -86,6 +86,32 @@ static struct configfs_dirent *configfs_new_dirent(struct configfs_dirent * pare return sd; } +/* + * + * Return -EEXIST if there is already a configfs element with the same + * name for the same parent. + * + * called with parent inode's i_mutex held + */ +int configfs_dirent_exists(struct configfs_dirent *parent_sd, + const unsigned char *new) +{ + struct configfs_dirent * sd; + + list_for_each_entry(sd, &parent_sd->s_children, s_sibling) { + if (sd->s_element) { + const unsigned char *existing = configfs_get_name(sd); + if (strcmp(existing, new)) + continue; + else + return -EEXIST; + } + } + + return 0; +} + + int configfs_make_dirent(struct configfs_dirent * parent_sd, struct dentry * dentry, void * element, umode_t mode, int type) @@ -136,8 +162,10 @@ static int create_dir(struct config_item * k, struct dentry * p, int error; umode_t mode = S_IFDIR| S_IRWXU | S_IRUGO | S_IXUGO; - error = configfs_make_dirent(p->d_fsdata, d, k, mode, - CONFIGFS_DIR); + error = configfs_dirent_exists(p->d_fsdata, d->d_name.name); + if (!error) + error = configfs_make_dirent(p->d_fsdata, d, k, mode, + CONFIGFS_DIR); if (!error) { error = configfs_create(d, mode, init_dir); if (!error) { @@ -211,7 +239,7 @@ static void remove_dir(struct dentry * d) struct configfs_dirent * sd; sd = d->d_fsdata; - list_del_init(&sd->s_sibling); + list_del_init(&sd->s_sibling); configfs_put(sd); if (d->d_inode) simple_rmdir(parent->d_inode,d); @@ -330,7 +358,7 @@ static int configfs_detach_prep(struct dentry *dentry) ret = configfs_detach_prep(sd->s_dentry); if (!ret) - continue; + continue; } else ret = -ENOTEMPTY; @@ -931,7 +959,7 @@ int configfs_rename_dir(struct config_item * item, const char *new_name) new_dentry = lookup_one_len(new_name, parent, strlen(new_name)); if (!IS_ERR(new_dentry)) { - if (!new_dentry->d_inode) { + if (!new_dentry->d_inode) { error = config_item_set_name(item, "%s", new_name); if (!error) { d_add(new_dentry, NULL); @@ -1009,8 +1037,7 @@ static int configfs_readdir(struct file * filp, void * dirent, filldir_t filldir /* fallthrough */ default: if (filp->f_pos == 2) { - list_del(q); - list_add(q, &parent_sd->s_children); + list_move(q, &parent_sd->s_children); } for (p=q->next; p!= &parent_sd->s_children; p=p->next) { struct configfs_dirent *next; @@ -1033,8 +1060,7 @@ static int configfs_readdir(struct file * filp, void * dirent, filldir_t filldir dt_type(next)) < 0) return 0; - list_del(q); - list_add(q, p); + list_move(q, p); p = q; filp->f_pos++; } diff --git a/fs/configfs/file.c b/fs/configfs/file.c index f499803743e..85105e50f7d 100644 --- a/fs/configfs/file.c +++ b/fs/configfs/file.c @@ -274,9 +274,8 @@ static int check_perm(struct inode * inode, struct file * file) /* No error? Great, allocate a buffer for the file, and store it * it in file->private_data for easy access. */ - buffer = kmalloc(sizeof(struct configfs_buffer),GFP_KERNEL); + buffer = kzalloc(sizeof(struct configfs_buffer),GFP_KERNEL); if (buffer) { - memset(buffer,0,sizeof(struct configfs_buffer)); init_MUTEX(&buffer->sem); buffer->needs_read_fill = 1; buffer->ops = ops; diff --git a/fs/configfs/inode.c b/fs/configfs/inode.c index c153bd9534c..fb18917954a 100644 --- a/fs/configfs/inode.c +++ b/fs/configfs/inode.c @@ -38,7 +38,7 @@ extern struct super_block * configfs_sb; -static struct address_space_operations configfs_aops = { +static const struct address_space_operations configfs_aops = { .readpage = simple_readpage, .prepare_write = simple_prepare_write, .commit_write = simple_commit_write @@ -76,11 +76,10 @@ int configfs_setattr(struct dentry * dentry, struct iattr * iattr) if (!sd_iattr) { /* setting attributes for the first time, allocate now */ - sd_iattr = kmalloc(sizeof(struct iattr), GFP_KERNEL); + sd_iattr = kzalloc(sizeof(struct iattr), GFP_KERNEL); if (!sd_iattr) return -ENOMEM; /* assign default attributes */ - memset(sd_iattr, 0, sizeof(struct iattr)); sd_iattr->ia_mode = sd->s_mode; sd_iattr->ia_uid = 0; sd_iattr->ia_gid = 0; @@ -136,7 +135,6 @@ struct inode * configfs_new_inode(mode_t mode, struct configfs_dirent * sd) { struct inode * inode = new_inode(configfs_sb); if (inode) { - inode->i_blksize = PAGE_CACHE_SIZE; inode->i_blocks = 0; inode->i_mapping->a_ops = &configfs_aops; inode->i_mapping->backing_dev_info = &configfs_backing_dev_info; diff --git a/fs/configfs/mount.c b/fs/configfs/mount.c index 94dab7bdd85..3e5fe843e1d 100644 --- a/fs/configfs/mount.c +++ b/fs/configfs/mount.c @@ -118,7 +118,7 @@ static struct file_system_type configfs_fs_type = { int configfs_pin_fs(void) { - return simple_pin_fs("configfs", &configfs_mount, + return simple_pin_fs(&configfs_fs_type, &configfs_mount, &configfs_mnt_count); } diff --git a/fs/configfs/symlink.c b/fs/configfs/symlink.c index e5512e295cf..fb65e0800a8 100644 --- a/fs/configfs/symlink.c +++ b/fs/configfs/symlink.c @@ -66,7 +66,7 @@ static void fill_item_path(struct config_item * item, char * buffer, int length) } static int create_link(struct config_item *parent_item, - struct config_item *item, + struct config_item *item, struct dentry *dentry) { struct configfs_dirent *target_sd = item->ci_dentry->d_fsdata; diff --git a/fs/cramfs/inode.c b/fs/cramfs/inode.c index c45d7386080..a624c3ec818 100644 --- a/fs/cramfs/inode.c +++ b/fs/cramfs/inode.c @@ -30,7 +30,7 @@ static struct super_operations cramfs_ops; static struct inode_operations cramfs_dir_inode_operations; static const struct file_operations cramfs_directory_operations; -static struct address_space_operations cramfs_aops; +static const struct address_space_operations cramfs_aops; static DEFINE_MUTEX(read_mutex); @@ -73,7 +73,6 @@ static int cramfs_iget5_set(struct inode *inode, void *opaque) inode->i_uid = cramfs_inode->uid; inode->i_size = cramfs_inode->size; inode->i_blocks = (cramfs_inode->size - 1) / 512 + 1; - inode->i_blksize = PAGE_CACHE_SIZE; inode->i_gid = cramfs_inode->gid; /* Struct copy intentional */ inode->i_mtime = inode->i_atime = inode->i_ctime = zerotime; @@ -242,11 +241,10 @@ static int cramfs_fill_super(struct super_block *sb, void *data, int silent) sb->s_flags |= MS_RDONLY; - sbi = kmalloc(sizeof(struct cramfs_sb_info), GFP_KERNEL); + sbi = kzalloc(sizeof(struct cramfs_sb_info), GFP_KERNEL); if (!sbi) return -ENOMEM; sb->s_fs_info = sbi; - memset(sbi, 0, sizeof(struct cramfs_sb_info)); /* Invalidate the read buffers on mount: think disk change.. */ mutex_lock(&read_mutex); @@ -501,7 +499,7 @@ static int cramfs_readpage(struct file *file, struct page * page) return 0; } -static struct address_space_operations cramfs_aops = { +static const struct address_space_operations cramfs_aops = { .readpage = cramfs_readpage }; @@ -545,8 +543,15 @@ static struct file_system_type cramfs_fs_type = { static int __init init_cramfs_fs(void) { - cramfs_uncompress_init(); - return register_filesystem(&cramfs_fs_type); + int rv; + + rv = cramfs_uncompress_init(); + if (rv < 0) + return rv; + rv = register_filesystem(&cramfs_fs_type); + if (rv < 0) + cramfs_uncompress_exit(); + return rv; } static void __exit exit_cramfs_fs(void) diff --git a/fs/cramfs/uncompress.c b/fs/cramfs/uncompress.c index 8def89f2c43..fc3ccb74626 100644 --- a/fs/cramfs/uncompress.c +++ b/fs/cramfs/uncompress.c @@ -68,11 +68,10 @@ int cramfs_uncompress_init(void) return 0; } -int cramfs_uncompress_exit(void) +void cramfs_uncompress_exit(void) { if (!--initialized) { zlib_inflateEnd(&stream); vfree(stream.workspace); } - return 0; } diff --git a/fs/dcache.c b/fs/dcache.c index 313b54b2b8f..17b392a2049 100644 --- a/fs/dcache.c +++ b/fs/dcache.c @@ -14,7 +14,6 @@ * the dcache entry is deleted or garbage collected. */ -#include <linux/config.h> #include <linux/syscalls.h> #include <linux/string.h> #include <linux/mm.h> @@ -39,7 +38,7 @@ int sysctl_vfs_cache_pressure __read_mostly = 100; EXPORT_SYMBOL_GPL(sysctl_vfs_cache_pressure); __cacheline_aligned_in_smp DEFINE_SPINLOCK(dcache_lock); -static seqlock_t rename_lock __cacheline_aligned_in_smp = SEQLOCK_UNLOCKED; +static __cacheline_aligned_in_smp DEFINE_SEQLOCK(rename_lock); EXPORT_SYMBOL(dcache_lock); @@ -406,7 +405,7 @@ static void prune_dcache(int count, struct super_block *sb) cond_resched_lock(&dcache_lock); tmp = dentry_unused.prev; - if (unlikely(sb)) { + if (sb) { /* Try to find a dentry for this sb, but don't try * too hard, if they aren't near the tail they will * be moved down again soon @@ -522,8 +521,7 @@ void shrink_dcache_sb(struct super_block * sb) dentry = list_entry(tmp, struct dentry, d_lru); if (dentry->d_sb != sb) continue; - list_del(tmp); - list_add(tmp, &dentry_unused); + list_move(tmp, &dentry_unused); } /* @@ -638,7 +636,7 @@ resume: * of the unused list for prune_dcache */ if (!atomic_read(&dentry->d_count)) { - list_add(&dentry->d_lru, dentry_unused.prev); + list_add_tail(&dentry->d_lru, &dentry_unused); dentry_stat.nr_unused++; found++; } @@ -830,17 +828,19 @@ void d_instantiate(struct dentry *entry, struct inode * inode) * (or otherwise set) by the caller to indicate that it is now * in use by the dcache. */ -struct dentry *d_instantiate_unique(struct dentry *entry, struct inode *inode) +static struct dentry *__d_instantiate_unique(struct dentry *entry, + struct inode *inode) { struct dentry *alias; int len = entry->d_name.len; const char *name = entry->d_name.name; unsigned int hash = entry->d_name.hash; - BUG_ON(!list_empty(&entry->d_alias)); - spin_lock(&dcache_lock); - if (!inode) - goto do_negative; + if (!inode) { + entry->d_inode = NULL; + return NULL; + } + list_for_each_entry(alias, &inode->i_dentry, d_alias) { struct qstr *qstr = &alias->d_name; @@ -853,19 +853,35 @@ struct dentry *d_instantiate_unique(struct dentry *entry, struct inode *inode) if (memcmp(qstr->name, name, len)) continue; dget_locked(alias); - spin_unlock(&dcache_lock); - BUG_ON(!d_unhashed(alias)); - iput(inode); return alias; } + list_add(&entry->d_alias, &inode->i_dentry); -do_negative: entry->d_inode = inode; fsnotify_d_instantiate(entry, inode); - spin_unlock(&dcache_lock); - security_d_instantiate(entry, inode); return NULL; } + +struct dentry *d_instantiate_unique(struct dentry *entry, struct inode *inode) +{ + struct dentry *result; + + BUG_ON(!list_empty(&entry->d_alias)); + + spin_lock(&dcache_lock); + result = __d_instantiate_unique(entry, inode); + spin_unlock(&dcache_lock); + + if (!result) { + security_d_instantiate(entry, inode); + return NULL; + } + + BUG_ON(!d_unhashed(result)); + iput(inode); + return result; +} + EXPORT_SYMBOL(d_instantiate_unique); /** @@ -1237,6 +1253,11 @@ static void __d_rehash(struct dentry * entry, struct hlist_head *list) hlist_add_head_rcu(&entry->d_hash, list); } +static void _d_rehash(struct dentry * entry) +{ + __d_rehash(entry, d_hash(entry->d_parent, entry->d_name.hash)); +} + /** * d_rehash - add an entry back to the hash * @entry: dentry to add to the hash @@ -1246,11 +1267,9 @@ static void __d_rehash(struct dentry * entry, struct hlist_head *list) void d_rehash(struct dentry * entry) { - struct hlist_head *list = d_hash(entry->d_parent, entry->d_name.hash); - spin_lock(&dcache_lock); spin_lock(&entry->d_lock); - __d_rehash(entry, list); + _d_rehash(entry); spin_unlock(&entry->d_lock); spin_unlock(&dcache_lock); } @@ -1341,10 +1360,10 @@ void d_move(struct dentry * dentry, struct dentry * target) */ if (target < dentry) { spin_lock(&target->d_lock); - spin_lock(&dentry->d_lock); + spin_lock_nested(&dentry->d_lock, DENTRY_D_LOCK_NESTED); } else { spin_lock(&dentry->d_lock); - spin_lock(&target->d_lock); + spin_lock_nested(&target->d_lock, DENTRY_D_LOCK_NESTED); } /* Move the dentry to the target hash queue, if on different bucket */ @@ -1388,6 +1407,120 @@ already_unhashed: spin_unlock(&dcache_lock); } +/* + * Prepare an anonymous dentry for life in the superblock's dentry tree as a + * named dentry in place of the dentry to be replaced. + */ +static void __d_materialise_dentry(struct dentry *dentry, struct dentry *anon) +{ + struct dentry *dparent, *aparent; + + switch_names(dentry, anon); + do_switch(dentry->d_name.len, anon->d_name.len); + do_switch(dentry->d_name.hash, anon->d_name.hash); + + dparent = dentry->d_parent; + aparent = anon->d_parent; + + dentry->d_parent = (aparent == anon) ? dentry : aparent; + list_del(&dentry->d_u.d_child); + if (!IS_ROOT(dentry)) + list_add(&dentry->d_u.d_child, &dentry->d_parent->d_subdirs); + else + INIT_LIST_HEAD(&dentry->d_u.d_child); + + anon->d_parent = (dparent == dentry) ? anon : dparent; + list_del(&anon->d_u.d_child); + if (!IS_ROOT(anon)) + list_add(&anon->d_u.d_child, &anon->d_parent->d_subdirs); + else + INIT_LIST_HEAD(&anon->d_u.d_child); + + anon->d_flags &= ~DCACHE_DISCONNECTED; +} + +/** + * d_materialise_unique - introduce an inode into the tree + * @dentry: candidate dentry + * @inode: inode to bind to the dentry, to which aliases may be attached + * + * Introduces an dentry into the tree, substituting an extant disconnected + * root directory alias in its place if there is one + */ +struct dentry *d_materialise_unique(struct dentry *dentry, struct inode *inode) +{ + struct dentry *alias, *actual; + + BUG_ON(!d_unhashed(dentry)); + + spin_lock(&dcache_lock); + + if (!inode) { + actual = dentry; + dentry->d_inode = NULL; + goto found_lock; + } + + /* See if a disconnected directory already exists as an anonymous root + * that we should splice into the tree instead */ + if (S_ISDIR(inode->i_mode) && (alias = __d_find_alias(inode, 1))) { + spin_lock(&alias->d_lock); + + /* Is this a mountpoint that we could splice into our tree? */ + if (IS_ROOT(alias)) + goto connect_mountpoint; + + if (alias->d_name.len == dentry->d_name.len && + alias->d_parent == dentry->d_parent && + memcmp(alias->d_name.name, + dentry->d_name.name, + dentry->d_name.len) == 0) + goto replace_with_alias; + + spin_unlock(&alias->d_lock); + + /* Doh! Seem to be aliasing directories for some reason... */ + dput(alias); + } + + /* Add a unique reference */ + actual = __d_instantiate_unique(dentry, inode); + if (!actual) + actual = dentry; + else if (unlikely(!d_unhashed(actual))) + goto shouldnt_be_hashed; + +found_lock: + spin_lock(&actual->d_lock); +found: + _d_rehash(actual); + spin_unlock(&actual->d_lock); + spin_unlock(&dcache_lock); + + if (actual == dentry) { + security_d_instantiate(dentry, inode); + return NULL; + } + + iput(inode); + return actual; + + /* Convert the anonymous/root alias into an ordinary dentry */ +connect_mountpoint: + __d_materialise_dentry(dentry, alias); + + /* Replace the candidate dentry with the alias in the tree */ +replace_with_alias: + __d_drop(alias); + actual = alias; + goto found; + +shouldnt_be_hashed: + spin_unlock(&dcache_lock); + BUG(); + goto shouldnt_be_hashed; +} + /** * d_path - return the path of a dentry * @dentry: dentry to report @@ -1786,6 +1919,7 @@ EXPORT_SYMBOL(d_instantiate); EXPORT_SYMBOL(d_invalidate); EXPORT_SYMBOL(d_lookup); EXPORT_SYMBOL(d_move); +EXPORT_SYMBOL_GPL(d_materialise_unique); EXPORT_SYMBOL(d_path); EXPORT_SYMBOL(d_prune_aliases); EXPORT_SYMBOL(d_rehash); diff --git a/fs/dcookies.c b/fs/dcookies.c index 8749339bf4f..0c4b0674854 100644 --- a/fs/dcookies.c +++ b/fs/dcookies.c @@ -12,7 +12,6 @@ * to the pair and can be looked up from userspace. */ -#include <linux/config.h> #include <linux/syscalls.h> #include <linux/module.h> #include <linux/slab.h> diff --git a/fs/debugfs/file.c b/fs/debugfs/file.c index 66a505422e5..bf3901ab174 100644 --- a/fs/debugfs/file.c +++ b/fs/debugfs/file.c @@ -13,7 +13,6 @@ * */ -#include <linux/config.h> #include <linux/module.h> #include <linux/fs.h> #include <linux/pagemap.h> @@ -33,8 +32,8 @@ static ssize_t default_write_file(struct file *file, const char __user *buf, static int default_open(struct inode *inode, struct file *file) { - if (inode->u.generic_ip) - file->private_data = inode->u.generic_ip; + if (inode->i_private) + file->private_data = inode->i_private; return 0; } @@ -56,12 +55,11 @@ static u64 debugfs_u8_get(void *data) DEFINE_SIMPLE_ATTRIBUTE(fops_u8, debugfs_u8_get, debugfs_u8_set, "%llu\n"); /** - * debugfs_create_u8 - create a file in the debugfs filesystem that is used to read and write an unsigned 8 bit value. - * + * debugfs_create_u8 - create a debugfs file that is used to read and write an unsigned 8-bit value * @name: a pointer to a string containing the name of the file to create. * @mode: the permission that the file should have * @parent: a pointer to the parent dentry for this file. This should be a - * directory dentry if set. If this paramater is NULL, then the + * directory dentry if set. If this parameter is %NULL, then the * file will be created in the root of the debugfs filesystem. * @value: a pointer to the variable that the file should read to and write * from. @@ -73,11 +71,11 @@ DEFINE_SIMPLE_ATTRIBUTE(fops_u8, debugfs_u8_get, debugfs_u8_set, "%llu\n"); * This function will return a pointer to a dentry if it succeeds. This * pointer must be passed to the debugfs_remove() function when the file is * to be removed (no automatic cleanup happens if your module is unloaded, - * you are responsible here.) If an error occurs, NULL will be returned. + * you are responsible here.) If an error occurs, %NULL will be returned. * - * If debugfs is not enabled in the kernel, the value -ENODEV will be + * If debugfs is not enabled in the kernel, the value -%ENODEV will be * returned. It is not wise to check for this value, but rather, check for - * NULL or !NULL instead as to eliminate the need for #ifdef in the calling + * %NULL or !%NULL instead as to eliminate the need for #ifdef in the calling * code. */ struct dentry *debugfs_create_u8(const char *name, mode_t mode, @@ -98,12 +96,11 @@ static u64 debugfs_u16_get(void *data) DEFINE_SIMPLE_ATTRIBUTE(fops_u16, debugfs_u16_get, debugfs_u16_set, "%llu\n"); /** - * debugfs_create_u16 - create a file in the debugfs filesystem that is used to read and write an unsigned 16 bit value. - * + * debugfs_create_u16 - create a debugfs file that is used to read and write an unsigned 16-bit value * @name: a pointer to a string containing the name of the file to create. * @mode: the permission that the file should have * @parent: a pointer to the parent dentry for this file. This should be a - * directory dentry if set. If this paramater is NULL, then the + * directory dentry if set. If this parameter is %NULL, then the * file will be created in the root of the debugfs filesystem. * @value: a pointer to the variable that the file should read to and write * from. @@ -115,11 +112,11 @@ DEFINE_SIMPLE_ATTRIBUTE(fops_u16, debugfs_u16_get, debugfs_u16_set, "%llu\n"); * This function will return a pointer to a dentry if it succeeds. This * pointer must be passed to the debugfs_remove() function when the file is * to be removed (no automatic cleanup happens if your module is unloaded, - * you are responsible here.) If an error occurs, NULL will be returned. + * you are responsible here.) If an error occurs, %NULL will be returned. * - * If debugfs is not enabled in the kernel, the value -ENODEV will be + * If debugfs is not enabled in the kernel, the value -%ENODEV will be * returned. It is not wise to check for this value, but rather, check for - * NULL or !NULL instead as to eliminate the need for #ifdef in the calling + * %NULL or !%NULL instead as to eliminate the need for #ifdef in the calling * code. */ struct dentry *debugfs_create_u16(const char *name, mode_t mode, @@ -140,12 +137,11 @@ static u64 debugfs_u32_get(void *data) DEFINE_SIMPLE_ATTRIBUTE(fops_u32, debugfs_u32_get, debugfs_u32_set, "%llu\n"); /** - * debugfs_create_u32 - create a file in the debugfs filesystem that is used to read and write an unsigned 32 bit value. - * + * debugfs_create_u32 - create a debugfs file that is used to read and write an unsigned 32-bit value * @name: a pointer to a string containing the name of the file to create. * @mode: the permission that the file should have * @parent: a pointer to the parent dentry for this file. This should be a - * directory dentry if set. If this paramater is NULL, then the + * directory dentry if set. If this parameter is %NULL, then the * file will be created in the root of the debugfs filesystem. * @value: a pointer to the variable that the file should read to and write * from. @@ -157,11 +153,11 @@ DEFINE_SIMPLE_ATTRIBUTE(fops_u32, debugfs_u32_get, debugfs_u32_set, "%llu\n"); * This function will return a pointer to a dentry if it succeeds. This * pointer must be passed to the debugfs_remove() function when the file is * to be removed (no automatic cleanup happens if your module is unloaded, - * you are responsible here.) If an error occurs, NULL will be returned. + * you are responsible here.) If an error occurs, %NULL will be returned. * - * If debugfs is not enabled in the kernel, the value -ENODEV will be + * If debugfs is not enabled in the kernel, the value -%ENODEV will be * returned. It is not wise to check for this value, but rather, check for - * NULL or !NULL instead as to eliminate the need for #ifdef in the calling + * %NULL or !%NULL instead as to eliminate the need for #ifdef in the calling * code. */ struct dentry *debugfs_create_u32(const char *name, mode_t mode, @@ -220,12 +216,11 @@ static const struct file_operations fops_bool = { }; /** - * debugfs_create_bool - create a file in the debugfs filesystem that is used to read and write a boolean value. - * + * debugfs_create_bool - create a debugfs file that is used to read and write a boolean value * @name: a pointer to a string containing the name of the file to create. * @mode: the permission that the file should have * @parent: a pointer to the parent dentry for this file. This should be a - * directory dentry if set. If this paramater is NULL, then the + * directory dentry if set. If this parameter is %NULL, then the * file will be created in the root of the debugfs filesystem. * @value: a pointer to the variable that the file should read to and write * from. @@ -237,11 +232,11 @@ static const struct file_operations fops_bool = { * This function will return a pointer to a dentry if it succeeds. This * pointer must be passed to the debugfs_remove() function when the file is * to be removed (no automatic cleanup happens if your module is unloaded, - * you are responsible here.) If an error occurs, NULL will be returned. + * you are responsible here.) If an error occurs, %NULL will be returned. * - * If debugfs is not enabled in the kernel, the value -ENODEV will be + * If debugfs is not enabled in the kernel, the value -%ENODEV will be * returned. It is not wise to check for this value, but rather, check for - * NULL or !NULL instead as to eliminate the need for #ifdef in the calling + * %NULL or !%NULL instead as to eliminate the need for #ifdef in the calling * code. */ struct dentry *debugfs_create_bool(const char *name, mode_t mode, @@ -265,13 +260,11 @@ static struct file_operations fops_blob = { }; /** - * debugfs_create_blob - create a file in the debugfs filesystem that is - * used to read and write a binary blob. - * + * debugfs_create_blob - create a debugfs file that is used to read and write a binary blob * @name: a pointer to a string containing the name of the file to create. * @mode: the permission that the file should have * @parent: a pointer to the parent dentry for this file. This should be a - * directory dentry if set. If this paramater is NULL, then the + * directory dentry if set. If this parameter is %NULL, then the * file will be created in the root of the debugfs filesystem. * @blob: a pointer to a struct debugfs_blob_wrapper which contains a pointer * to the blob data and the size of the data. @@ -283,11 +276,11 @@ static struct file_operations fops_blob = { * This function will return a pointer to a dentry if it succeeds. This * pointer must be passed to the debugfs_remove() function when the file is * to be removed (no automatic cleanup happens if your module is unloaded, - * you are responsible here.) If an error occurs, NULL will be returned. + * you are responsible here.) If an error occurs, %NULL will be returned. * - * If debugfs is not enabled in the kernel, the value -ENODEV will be + * If debugfs is not enabled in the kernel, the value -%ENODEV will be * returned. It is not wise to check for this value, but rather, check for - * NULL or !NULL instead as to eliminate the need for #ifdef in the calling + * %NULL or !%NULL instead as to eliminate the need for #ifdef in the calling * code. */ struct dentry *debugfs_create_blob(const char *name, mode_t mode, diff --git a/fs/debugfs/inode.c b/fs/debugfs/inode.c index 440128ebef3..269e649e6dc 100644 --- a/fs/debugfs/inode.c +++ b/fs/debugfs/inode.c @@ -16,7 +16,6 @@ /* uncomment to get debug messages from the debug filesystem, ah the irony. */ /* #define DEBUG */ -#include <linux/config.h> #include <linux/module.h> #include <linux/fs.h> #include <linux/mount.h> @@ -41,7 +40,6 @@ static struct inode *debugfs_get_inode(struct super_block *sb, int mode, dev_t d inode->i_mode = mode; inode->i_uid = 0; inode->i_gid = 0; - inode->i_blksize = PAGE_CACHE_SIZE; inode->i_blocks = 0; inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME; switch (mode & S_IFMT) { @@ -163,14 +161,13 @@ static int debugfs_create_by_name(const char *name, mode_t mode, /** * debugfs_create_file - create a file in the debugfs filesystem - * * @name: a pointer to a string containing the name of the file to create. * @mode: the permission that the file should have * @parent: a pointer to the parent dentry for this file. This should be a * directory dentry if set. If this paramater is NULL, then the * file will be created in the root of the debugfs filesystem. * @data: a pointer to something that the caller will want to get to later - * on. The inode.u.generic_ip pointer will point to this value on + * on. The inode.i_private pointer will point to this value on * the open() call. * @fops: a pointer to a struct file_operations that should be used for * this file. @@ -183,11 +180,11 @@ static int debugfs_create_by_name(const char *name, mode_t mode, * This function will return a pointer to a dentry if it succeeds. This * pointer must be passed to the debugfs_remove() function when the file is * to be removed (no automatic cleanup happens if your module is unloaded, - * you are responsible here.) If an error occurs, NULL will be returned. + * you are responsible here.) If an error occurs, %NULL will be returned. * - * If debugfs is not enabled in the kernel, the value -ENODEV will be + * If debugfs is not enabled in the kernel, the value -%ENODEV will be * returned. It is not wise to check for this value, but rather, check for - * NULL or !NULL instead as to eliminate the need for #ifdef in the calling + * %NULL or !%NULL instead as to eliminate the need for #ifdef in the calling * code. */ struct dentry *debugfs_create_file(const char *name, mode_t mode, @@ -199,7 +196,7 @@ struct dentry *debugfs_create_file(const char *name, mode_t mode, pr_debug("debugfs: creating file '%s'\n",name); - error = simple_pin_fs("debugfs", &debugfs_mount, &debugfs_mount_count); + error = simple_pin_fs(&debug_fs_type, &debugfs_mount, &debugfs_mount_count); if (error) goto exit; @@ -211,7 +208,7 @@ struct dentry *debugfs_create_file(const char *name, mode_t mode, if (dentry->d_inode) { if (data) - dentry->d_inode->u.generic_ip = data; + dentry->d_inode->i_private = data; if (fops) dentry->d_inode->i_fop = fops; } @@ -222,7 +219,6 @@ EXPORT_SYMBOL_GPL(debugfs_create_file); /** * debugfs_create_dir - create a directory in the debugfs filesystem - * * @name: a pointer to a string containing the name of the directory to * create. * @parent: a pointer to the parent dentry for this file. This should be a @@ -234,11 +230,11 @@ EXPORT_SYMBOL_GPL(debugfs_create_file); * This function will return a pointer to a dentry if it succeeds. This * pointer must be passed to the debugfs_remove() function when the file is * to be removed (no automatic cleanup happens if your module is unloaded, - * you are responsible here.) If an error occurs, NULL will be returned. + * you are responsible here.) If an error occurs, %NULL will be returned. * - * If debugfs is not enabled in the kernel, the value -ENODEV will be + * If debugfs is not enabled in the kernel, the value -%ENODEV will be * returned. It is not wise to check for this value, but rather, check for - * NULL or !NULL instead as to eliminate the need for #ifdef in the calling + * %NULL or !%NULL instead as to eliminate the need for #ifdef in the calling * code. */ struct dentry *debugfs_create_dir(const char *name, struct dentry *parent) @@ -251,7 +247,6 @@ EXPORT_SYMBOL_GPL(debugfs_create_dir); /** * debugfs_remove - removes a file or directory from the debugfs filesystem - * * @dentry: a pointer to a the dentry of the file or directory to be * removed. * diff --git a/fs/devfs/Makefile b/fs/devfs/Makefile deleted file mode 100644 index 6dd8d1245e2..00000000000 --- a/fs/devfs/Makefile +++ /dev/null @@ -1,8 +0,0 @@ -# -# Makefile for the linux devfs-filesystem routines. -# - -obj-$(CONFIG_DEVFS_FS) += devfs.o - -devfs-objs := base.o util.o - diff --git a/fs/devfs/base.c b/fs/devfs/base.c deleted file mode 100644 index 51a97f13274..00000000000 --- a/fs/devfs/base.c +++ /dev/null @@ -1,2836 +0,0 @@ -/* devfs (Device FileSystem) driver. - - Copyright (C) 1998-2002 Richard Gooch - - This library is free software; you can redistribute it and/or - modify it under the terms of the GNU Library General Public - License as published by the Free Software Foundation; either - version 2 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 - Library General Public License for more details. - - You should have received a copy of the GNU Library General Public - License along with this library; if not, write to the Free - Software Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. - - Richard Gooch may be reached by email at rgooch@atnf.csiro.au - The postal address is: - Richard Gooch, c/o ATNF, P. O. Box 76, Epping, N.S.W., 2121, Australia. - - ChangeLog - - 19980110 Richard Gooch <rgooch@atnf.csiro.au> - Original version. - v0.1 - 19980111 Richard Gooch <rgooch@atnf.csiro.au> - Created per-fs inode table rather than using inode->u.generic_ip - v0.2 - 19980111 Richard Gooch <rgooch@atnf.csiro.au> - Created .epoch inode which has a ctime of 0. - Fixed loss of named pipes when dentries lost. - Fixed loss of inode data when devfs_register() follows mknod(). - v0.3 - 19980111 Richard Gooch <rgooch@atnf.csiro.au> - Fix for when compiling with CONFIG_KERNELD. - 19980112 Richard Gooch <rgooch@atnf.csiro.au> - Fix for readdir() which sometimes didn't show entries. - Added <<tolerant>> option to <devfs_register>. - v0.4 - 19980113 Richard Gooch <rgooch@atnf.csiro.au> - Created <devfs_fill_file> function. - v0.5 - 19980115 Richard Gooch <rgooch@atnf.csiro.au> - Added subdirectory support. Major restructuring. - 19980116 Richard Gooch <rgooch@atnf.csiro.au> - Fixed <find_by_dev> to not search major=0,minor=0. - Added symlink support. - v0.6 - 19980120 Richard Gooch <rgooch@atnf.csiro.au> - Created <devfs_mk_dir> function and support directory unregister - 19980120 Richard Gooch <rgooch@atnf.csiro.au> - Auto-ownership uses real uid/gid rather than effective uid/gid. - v0.7 - 19980121 Richard Gooch <rgooch@atnf.csiro.au> - Supported creation of sockets. - v0.8 - 19980122 Richard Gooch <rgooch@atnf.csiro.au> - Added DEVFS_FL_HIDE_UNREG flag. - Interface change to <devfs_mk_symlink>. - Created <devfs_symlink> to support symlink(2). - v0.9 - 19980123 Richard Gooch <rgooch@atnf.csiro.au> - Added check to <devfs_fill_file> to check inode is in devfs. - Added optional traversal of symlinks. - v0.10 - 19980124 Richard Gooch <rgooch@atnf.csiro.au> - Created <devfs_get_flags> and <devfs_set_flags>. - v0.11 - 19980125 C. Scott Ananian <cananian@alumni.princeton.edu> - Created <devfs_find_handle>. - 19980125 Richard Gooch <rgooch@atnf.csiro.au> - Allow removal of symlinks. - v0.12 - 19980125 Richard Gooch <rgooch@atnf.csiro.au> - Created <devfs_set_symlink_destination>. - 19980126 Richard Gooch <rgooch@atnf.csiro.au> - Moved DEVFS_SUPER_MAGIC into header file. - Added DEVFS_FL_HIDE flag. - Created <devfs_get_maj_min>. - Created <devfs_get_handle_from_inode>. - Fixed minor bug in <find_by_dev>. - 19980127 Richard Gooch <rgooch@atnf.csiro.au> - Changed interface to <find_by_dev>, <find_entry>, - <devfs_unregister>, <devfs_fill_file> and <devfs_find_handle>. - Fixed inode times when symlink created with symlink(2). - v0.13 - 19980129 C. Scott Ananian <cananian@alumni.princeton.edu> - Exported <devfs_set_symlink_destination>, <devfs_get_maj_min> - and <devfs_get_handle_from_inode>. - 19980129 Richard Gooch <rgooch@atnf.csiro.au> - Created <devfs_unlink> to support unlink(2). - v0.14 - 19980129 Richard Gooch <rgooch@atnf.csiro.au> - Fixed kerneld support for entries in devfs subdirectories. - 19980130 Richard Gooch <rgooch@atnf.csiro.au> - Bugfixes in <call_kerneld>. - v0.15 - 19980207 Richard Gooch <rgooch@atnf.csiro.au> - Call kerneld when looking up unregistered entries. - v0.16 - 19980326 Richard Gooch <rgooch@atnf.csiro.au> - Modified interface to <devfs_find_handle> for symlink traversal. - v0.17 - 19980331 Richard Gooch <rgooch@atnf.csiro.au> - Fixed persistence bug with device numbers for manually created - device files. - Fixed problem with recreating symlinks with different content. - v0.18 - 19980401 Richard Gooch <rgooch@atnf.csiro.au> - Changed to CONFIG_KMOD. - Hide entries which are manually unlinked. - Always invalidate devfs dentry cache when registering entries. - Created <devfs_rmdir> to support rmdir(2). - Ensure directories created by <devfs_mk_dir> are visible. - v0.19 - 19980402 Richard Gooch <rgooch@atnf.csiro.au> - Invalidate devfs dentry cache when making directories. - Invalidate devfs dentry cache when removing entries. - Fixed persistence bug with fifos. - v0.20 - 19980421 Richard Gooch <rgooch@atnf.csiro.au> - Print process command when debugging kerneld/kmod. - Added debugging for register/unregister/change operations. - 19980422 Richard Gooch <rgooch@atnf.csiro.au> - Added "devfs=" boot options. - v0.21 - 19980426 Richard Gooch <rgooch@atnf.csiro.au> - No longer lock/unlock superblock in <devfs_put_super>. - Drop negative dentries when they are released. - Manage dcache more efficiently. - v0.22 - 19980427 Richard Gooch <rgooch@atnf.csiro.au> - Added DEVFS_FL_AUTO_DEVNUM flag. - v0.23 - 19980430 Richard Gooch <rgooch@atnf.csiro.au> - No longer set unnecessary methods. - v0.24 - 19980504 Richard Gooch <rgooch@atnf.csiro.au> - Added PID display to <call_kerneld> debugging message. - Added "after" debugging message to <call_kerneld>. - 19980519 Richard Gooch <rgooch@atnf.csiro.au> - Added "diread" and "diwrite" boot options. - 19980520 Richard Gooch <rgooch@atnf.csiro.au> - Fixed persistence problem with permissions. - v0.25 - 19980602 Richard Gooch <rgooch@atnf.csiro.au> - Support legacy device nodes. - Fixed bug where recreated inodes were hidden. - v0.26 - 19980602 Richard Gooch <rgooch@atnf.csiro.au> - Improved debugging in <get_vfs_inode>. - 19980607 Richard Gooch <rgooch@atnf.csiro.au> - No longer free old dentries in <devfs_mk_dir>. - Free all dentries for a given entry when deleting inodes. - v0.27 - 19980627 Richard Gooch <rgooch@atnf.csiro.au> - Limit auto-device numbering to majors 128 to 239. - v0.28 - 19980629 Richard Gooch <rgooch@atnf.csiro.au> - Fixed inode times persistence problem. - v0.29 - 19980704 Richard Gooch <rgooch@atnf.csiro.au> - Fixed spelling in <devfs_readlink> debug. - Fixed bug in <devfs_setup> parsing "dilookup". - v0.30 - 19980705 Richard Gooch <rgooch@atnf.csiro.au> - Fixed devfs inode leak when manually recreating inodes. - Fixed permission persistence problem when recreating inodes. - v0.31 - 19980727 Richard Gooch <rgooch@atnf.csiro.au> - Removed harmless "unused variable" compiler warning. - Fixed modes for manually recreated device nodes. - v0.32 - 19980728 Richard Gooch <rgooch@atnf.csiro.au> - Added NULL devfs inode warning in <devfs_read_inode>. - Force all inode nlink values to 1. - v0.33 - 19980730 Richard Gooch <rgooch@atnf.csiro.au> - Added "dimknod" boot option. - Set inode nlink to 0 when freeing dentries. - Fixed modes for manually recreated symlinks. - v0.34 - 19980802 Richard Gooch <rgooch@atnf.csiro.au> - Fixed bugs in recreated directories and symlinks. - v0.35 - 19980806 Richard Gooch <rgooch@atnf.csiro.au> - Fixed bugs in recreated device nodes. - 19980807 Richard Gooch <rgooch@atnf.csiro.au> - Fixed bug in currently unused <devfs_get_handle_from_inode>. - Defined new <devfs_handle_t> type. - Improved debugging when getting entries. - Fixed bug where directories could be emptied. - v0.36 - 19980809 Richard Gooch <rgooch@atnf.csiro.au> - Replaced dummy .epoch inode with .devfsd character device. - 19980810 Richard Gooch <rgooch@atnf.csiro.au> - Implemented devfsd protocol revision 0. - v0.37 - 19980819 Richard Gooch <rgooch@atnf.csiro.au> - Added soothing message to warning in <devfs_d_iput>. - v0.38 - 19980829 Richard Gooch <rgooch@atnf.csiro.au> - Use GCC extensions for structure initialisations. - Implemented async open notification. - Incremented devfsd protocol revision to 1. - v0.39 - 19980908 Richard Gooch <rgooch@atnf.csiro.au> - Moved async open notification to end of <devfs_open>. - v0.40 - 19980910 Richard Gooch <rgooch@atnf.csiro.au> - Prepended "/dev/" to module load request. - Renamed <call_kerneld> to <call_kmod>. - v0.41 - 19980910 Richard Gooch <rgooch@atnf.csiro.au> - Fixed typo "AYSNC" -> "ASYNC". - v0.42 - 19980910 Richard Gooch <rgooch@atnf.csiro.au> - Added open flag for files. - v0.43 - 19980927 Richard Gooch <rgooch@atnf.csiro.au> - Set i_blocks=0 and i_blksize=1024 in <devfs_read_inode>. - v0.44 - 19981005 Richard Gooch <rgooch@atnf.csiro.au> - Added test for empty <<name>> in <devfs_find_handle>. - Renamed <generate_path> to <devfs_generate_path> and published. - v0.45 - 19981006 Richard Gooch <rgooch@atnf.csiro.au> - Created <devfs_get_fops>. - v0.46 - 19981007 Richard Gooch <rgooch@atnf.csiro.au> - Limit auto-device numbering to majors 144 to 239. - v0.47 - 19981010 Richard Gooch <rgooch@atnf.csiro.au> - Updated <devfs_follow_link> for VFS change in 2.1.125. - v0.48 - 19981022 Richard Gooch <rgooch@atnf.csiro.au> - Created DEVFS_ FL_COMPAT flag. - v0.49 - 19981023 Richard Gooch <rgooch@atnf.csiro.au> - Created "nocompat" boot option. - v0.50 - 19981025 Richard Gooch <rgooch@atnf.csiro.au> - Replaced "mount" boot option with "nomount". - v0.51 - 19981110 Richard Gooch <rgooch@atnf.csiro.au> - Created "only" boot option. - v0.52 - 19981112 Richard Gooch <rgooch@atnf.csiro.au> - Added DEVFS_FL_REMOVABLE flag. - v0.53 - 19981114 Richard Gooch <rgooch@atnf.csiro.au> - Only call <scan_dir_for_removable> on first call to - <devfs_readdir>. - v0.54 - 19981205 Richard Gooch <rgooch@atnf.csiro.au> - Updated <devfs_rmdir> for VFS change in 2.1.131. - v0.55 - 19981218 Richard Gooch <rgooch@atnf.csiro.au> - Created <devfs_mk_compat>. - 19981220 Richard Gooch <rgooch@atnf.csiro.au> - Check for partitions on removable media in <devfs_lookup>. - v0.56 - 19990118 Richard Gooch <rgooch@atnf.csiro.au> - Added support for registering regular files. - Created <devfs_set_file_size>. - Update devfs inodes from entries if not changed through FS. - v0.57 - 19990124 Richard Gooch <rgooch@atnf.csiro.au> - Fixed <devfs_fill_file> to only initialise temporary inodes. - Trap for NULL fops in <devfs_register>. - Return -ENODEV in <devfs_fill_file> for non-driver inodes. - v0.58 - 19990126 Richard Gooch <rgooch@atnf.csiro.au> - Switched from PATH_MAX to DEVFS_PATHLEN. - v0.59 - 19990127 Richard Gooch <rgooch@atnf.csiro.au> - Created "nottycompat" boot option. - v0.60 - 19990318 Richard Gooch <rgooch@atnf.csiro.au> - Fixed <devfsd_read> to not overrun event buffer. - v0.61 - 19990329 Richard Gooch <rgooch@atnf.csiro.au> - Created <devfs_auto_unregister>. - v0.62 - 19990330 Richard Gooch <rgooch@atnf.csiro.au> - Don't return unregistred entries in <devfs_find_handle>. - Panic in <devfs_unregister> if entry unregistered. - 19990401 Richard Gooch <rgooch@atnf.csiro.au> - Don't panic in <devfs_auto_unregister> for duplicates. - v0.63 - 19990402 Richard Gooch <rgooch@atnf.csiro.au> - Don't unregister already unregistered entries in <unregister>. - v0.64 - 19990510 Richard Gooch <rgooch@atnf.csiro.au> - Disable warning messages when unable to read partition table for - removable media. - v0.65 - 19990512 Richard Gooch <rgooch@atnf.csiro.au> - Updated <devfs_lookup> for VFS change in 2.3.1-pre1. - Created "oops-on-panic" boot option. - Improved debugging in <devfs_register> and <devfs_unregister>. - v0.66 - 19990519 Richard Gooch <rgooch@atnf.csiro.au> - Added documentation for some functions. - 19990525 Richard Gooch <rgooch@atnf.csiro.au> - Removed "oops-on-panic" boot option: now always Oops. - v0.67 - 19990531 Richard Gooch <rgooch@atnf.csiro.au> - Improved debugging in <devfs_register>. - v0.68 - 19990604 Richard Gooch <rgooch@atnf.csiro.au> - Added "diunlink" and "nokmod" boot options. - Removed superfluous warning message in <devfs_d_iput>. - v0.69 - 19990611 Richard Gooch <rgooch@atnf.csiro.au> - Took account of change to <d_alloc_root>. - v0.70 - 19990614 Richard Gooch <rgooch@atnf.csiro.au> - Created separate event queue for each mounted devfs. - Removed <devfs_invalidate_dcache>. - Created new ioctl()s. - Incremented devfsd protocol revision to 3. - Fixed bug when re-creating directories: contents were lost. - Block access to inodes until devfsd updates permissions. - 19990615 Richard Gooch <rgooch@atnf.csiro.au> - Support 2.2.x kernels. - v0.71 - 19990623 Richard Gooch <rgooch@atnf.csiro.au> - Switched to sending process uid/gid to devfsd. - Renamed <call_kmod> to <try_modload>. - Added DEVFSD_NOTIFY_LOOKUP event. - 19990624 Richard Gooch <rgooch@atnf.csiro.au> - Added DEVFSD_NOTIFY_CHANGE event. - Incremented devfsd protocol revision to 4. - v0.72 - 19990713 Richard Gooch <rgooch@atnf.csiro.au> - Return EISDIR rather than EINVAL for read(2) on directories. - v0.73 - 19990809 Richard Gooch <rgooch@atnf.csiro.au> - Changed <devfs_setup> to new __init scheme. - v0.74 - 19990901 Richard Gooch <rgooch@atnf.csiro.au> - Changed remaining function declarations to new __init scheme. - v0.75 - 19991013 Richard Gooch <rgooch@atnf.csiro.au> - Created <devfs_get_info>, <devfs_set_info>, - <devfs_get_first_child> and <devfs_get_next_sibling>. - Added <<dir>> parameter to <devfs_register>, <devfs_mk_compat>, - <devfs_mk_dir> and <devfs_find_handle>. - Work sponsored by SGI. - v0.76 - 19991017 Richard Gooch <rgooch@atnf.csiro.au> - Allow multiple unregistrations. - Work sponsored by SGI. - v0.77 - 19991026 Richard Gooch <rgooch@atnf.csiro.au> - Added major and minor number to devfsd protocol. - Incremented devfsd protocol revision to 5. - Work sponsored by SGI. - v0.78 - 19991030 Richard Gooch <rgooch@atnf.csiro.au> - Support info pointer for all devfs entry types. - Added <<info>> parameter to <devfs_mk_dir> and - <devfs_mk_symlink>. - Work sponsored by SGI. - v0.79 - 19991031 Richard Gooch <rgooch@atnf.csiro.au> - Support "../" when searching devfs namespace. - Work sponsored by SGI. - v0.80 - 19991101 Richard Gooch <rgooch@atnf.csiro.au> - Created <devfs_get_unregister_slave>. - Work sponsored by SGI. - v0.81 - 19991103 Richard Gooch <rgooch@atnf.csiro.au> - Exported <devfs_get_parent>. - Work sponsored by SGI. - v0.82 - 19991104 Richard Gooch <rgooch@atnf.csiro.au> - Removed unused <devfs_set_symlink_destination>. - 19991105 Richard Gooch <rgooch@atnf.csiro.au> - Do not hide entries from devfsd or children. - Removed DEVFS_ FL_TTY_COMPAT flag. - Removed "nottycompat" boot option. - Removed <devfs_mk_compat>. - Work sponsored by SGI. - v0.83 - 19991107 Richard Gooch <rgooch@atnf.csiro.au> - Added DEVFS_FL_WAIT flag. - Work sponsored by SGI. - v0.84 - 19991107 Richard Gooch <rgooch@atnf.csiro.au> - Support new "disc" naming scheme in <get_removable_partition>. - Allow NULL fops in <devfs_register>. - Work sponsored by SGI. - v0.85 - 19991110 Richard Gooch <rgooch@atnf.csiro.au> - Fall back to major table if NULL fops given to <devfs_register>. - Work sponsored by SGI. - v0.86 - 19991204 Richard Gooch <rgooch@atnf.csiro.au> - Support fifos when unregistering. - Work sponsored by SGI. - v0.87 - 19991209 Richard Gooch <rgooch@atnf.csiro.au> - Removed obsolete DEVFS_ FL_COMPAT and DEVFS_ FL_TOLERANT flags. - Work sponsored by SGI. - v0.88 - 19991214 Richard Gooch <rgooch@atnf.csiro.au> - Removed kmod support. - Work sponsored by SGI. - v0.89 - 19991216 Richard Gooch <rgooch@atnf.csiro.au> - Improved debugging in <get_vfs_inode>. - Ensure dentries created by devfsd will be cleaned up. - Work sponsored by SGI. - v0.90 - 19991223 Richard Gooch <rgooch@atnf.csiro.au> - Created <devfs_get_name>. - Work sponsored by SGI. - v0.91 - 20000203 Richard Gooch <rgooch@atnf.csiro.au> - Ported to kernel 2.3.42. - Removed <devfs_fill_file>. - Work sponsored by SGI. - v0.92 - 20000306 Richard Gooch <rgooch@atnf.csiro.au> - Added DEVFS_ FL_NO_PERSISTENCE flag. - Removed unnecessary call to <update_devfs_inode_from_entry> in - <devfs_readdir>. - Work sponsored by SGI. - v0.93 - 20000413 Richard Gooch <rgooch@atnf.csiro.au> - Set inode->i_size to correct size for symlinks. - 20000414 Richard Gooch <rgooch@atnf.csiro.au> - Only give lookup() method to directories to comply with new VFS - assumptions. - Work sponsored by SGI. - 20000415 Richard Gooch <rgooch@atnf.csiro.au> - Remove unnecessary tests in symlink methods. - Don't kill existing block ops in <devfs_read_inode>. - Work sponsored by SGI. - v0.94 - 20000424 Richard Gooch <rgooch@atnf.csiro.au> - Don't create missing directories in <devfs_find_handle>. - Work sponsored by SGI. - v0.95 - 20000430 Richard Gooch <rgooch@atnf.csiro.au> - Added CONFIG_DEVFS_MOUNT. - Work sponsored by SGI. - v0.96 - 20000608 Richard Gooch <rgooch@atnf.csiro.au> - Disabled multi-mount capability (use VFS bindings instead). - Work sponsored by SGI. - v0.97 - 20000610 Richard Gooch <rgooch@atnf.csiro.au> - Switched to FS_SINGLE to disable multi-mounts. - 20000612 Richard Gooch <rgooch@atnf.csiro.au> - Removed module support. - Removed multi-mount code. - Removed compatibility macros: VFS has changed too much. - Work sponsored by SGI. - v0.98 - 20000614 Richard Gooch <rgooch@atnf.csiro.au> - Merged devfs inode into devfs entry. - Work sponsored by SGI. - v0.99 - 20000619 Richard Gooch <rgooch@atnf.csiro.au> - Removed dead code in <devfs_register> which used to call - <free_dentries>. - Work sponsored by SGI. - v0.100 - 20000621 Richard Gooch <rgooch@atnf.csiro.au> - Changed interface to <devfs_register>. - Work sponsored by SGI. - v0.101 - 20000622 Richard Gooch <rgooch@atnf.csiro.au> - Simplified interface to <devfs_mk_symlink> and <devfs_mk_dir>. - Simplified interface to <devfs_find_handle>. - Work sponsored by SGI. - v0.102 - 20010519 Richard Gooch <rgooch@atnf.csiro.au> - Ensure <devfs_generate_path> terminates string for root entry. - Exported <devfs_get_name> to modules. - 20010520 Richard Gooch <rgooch@atnf.csiro.au> - Make <devfs_mk_symlink> send events to devfsd. - Cleaned up option processing in <devfs_setup>. - 20010521 Richard Gooch <rgooch@atnf.csiro.au> - Fixed bugs in handling symlinks: could leak or cause Oops. - 20010522 Richard Gooch <rgooch@atnf.csiro.au> - Cleaned up directory handling by separating fops. - v0.103 - 20010601 Richard Gooch <rgooch@atnf.csiro.au> - Fixed handling of inverted options in <devfs_setup>. - v0.104 - 20010604 Richard Gooch <rgooch@atnf.csiro.au> - Adjusted <try_modload> to account for <devfs_generate_path> fix. - v0.105 - 20010617 Richard Gooch <rgooch@atnf.csiro.au> - Answered question posed by Al Viro and removed his comments. - Moved setting of registered flag after other fields are changed. - Fixed race between <devfsd_close> and <devfsd_notify_one>. - Global VFS changes added bogus BKL to <devfsd_close>: removed. - Widened locking in <devfs_readlink> and <devfs_follow_link>. - Replaced <devfsd_read> stack usage with <devfsd_ioctl> kmalloc. - Simplified locking in <devfsd_ioctl> and fixed memory leak. - v0.106 - 20010709 Richard Gooch <rgooch@atnf.csiro.au> - Removed broken devnum allocation and use <devfs_alloc_devnum>. - Fixed old devnum leak by calling new <devfs_dealloc_devnum>. - v0.107 - 20010712 Richard Gooch <rgooch@atnf.csiro.au> - Fixed bug in <devfs_setup> which could hang boot process. - v0.108 - 20010730 Richard Gooch <rgooch@atnf.csiro.au> - Added DEVFSD_NOTIFY_DELETE event. - 20010801 Richard Gooch <rgooch@atnf.csiro.au> - Removed #include <asm/segment.h>. - v0.109 - 20010807 Richard Gooch <rgooch@atnf.csiro.au> - Fixed inode table races by removing it and using - inode->u.generic_ip instead. - Moved <devfs_read_inode> into <get_vfs_inode>. - Moved <devfs_write_inode> into <devfs_notify_change>. - v0.110 - 20010808 Richard Gooch <rgooch@atnf.csiro.au> - Fixed race in <devfs_do_symlink> for uni-processor. - v0.111 - 20010818 Richard Gooch <rgooch@atnf.csiro.au> - Removed remnant of multi-mount support in <devfs_mknod>. - Removed unused DEVFS_FL_SHOW_UNREG flag. - v0.112 - 20010820 Richard Gooch <rgooch@atnf.csiro.au> - Removed nlink field from struct devfs_inode. - v0.113 - 20010823 Richard Gooch <rgooch@atnf.csiro.au> - Replaced BKL with global rwsem to protect symlink data (quick - and dirty hack). - v0.114 - 20010827 Richard Gooch <rgooch@atnf.csiro.au> - Replaced global rwsem for symlink with per-link refcount. - v0.115 - 20010919 Richard Gooch <rgooch@atnf.csiro.au> - Set inode->i_mapping->a_ops for block nodes in <get_vfs_inode>. - v0.116 - 20011008 Richard Gooch <rgooch@atnf.csiro.au> - Fixed overrun in <devfs_link> by removing function (not needed). - 20011009 Richard Gooch <rgooch@atnf.csiro.au> - Fixed buffer underrun in <try_modload>. - 20011029 Richard Gooch <rgooch@atnf.csiro.au> - Fixed race in <devfsd_ioctl> when setting event mask. - 20011114 Richard Gooch <rgooch@atnf.csiro.au> - First release of new locking code. - v1.0 - 20011117 Richard Gooch <rgooch@atnf.csiro.au> - Discard temporary buffer, now use "%s" for dentry names. - 20011118 Richard Gooch <rgooch@atnf.csiro.au> - Don't generate path in <try_modload>: use fake entry instead. - Use "existing" directory in <_devfs_make_parent_for_leaf>. - 20011122 Richard Gooch <rgooch@atnf.csiro.au> - Use slab cache rather than fixed buffer for devfsd events. - v1.1 - 20011125 Richard Gooch <rgooch@atnf.csiro.au> - Send DEVFSD_NOTIFY_REGISTERED events in <devfs_mk_dir>. - 20011127 Richard Gooch <rgooch@atnf.csiro.au> - Fixed locking bug in <devfs_d_revalidate_wait> due to typo. - Do not send CREATE, CHANGE, ASYNC_OPEN or DELETE events from - devfsd or children. - v1.2 - 20011202 Richard Gooch <rgooch@atnf.csiro.au> - Fixed bug in <devfsd_read>: was dereferencing freed pointer. - v1.3 - 20011203 Richard Gooch <rgooch@atnf.csiro.au> - Fixed bug in <devfsd_close>: was dereferencing freed pointer. - Added process group check for devfsd privileges. - v1.4 - 20011204 Richard Gooch <rgooch@atnf.csiro.au> - Use SLAB_ATOMIC in <devfsd_notify_de> from <devfs_d_delete>. - v1.5 - 20011211 Richard Gooch <rgooch@atnf.csiro.au> - Return old entry in <devfs_mk_dir> for 2.4.x kernels. - 20011212 Richard Gooch <rgooch@atnf.csiro.au> - Increment refcount on module in <check_disc_changed>. - 20011215 Richard Gooch <rgooch@atnf.csiro.au> - Created <devfs_get_handle> and exported <devfs_put>. - Increment refcount on module in <devfs_get_ops>. - Created <devfs_put_ops>. - v1.6 - 20011216 Richard Gooch <rgooch@atnf.csiro.au> - Added poisoning to <devfs_put>. - Improved debugging messages. - v1.7 - 20011221 Richard Gooch <rgooch@atnf.csiro.au> - Corrected (made useful) debugging message in <unregister>. - Moved <kmem_cache_create> in <mount_devfs_fs> to <init_devfs_fs> - 20011224 Richard Gooch <rgooch@atnf.csiro.au> - Added magic number to guard against scribbling drivers. - 20011226 Richard Gooch <rgooch@atnf.csiro.au> - Only return old entry in <devfs_mk_dir> if a directory. - Defined macros for error and debug messages. - v1.8 - 20020113 Richard Gooch <rgooch@atnf.csiro.au> - Fixed (rare, old) race in <devfs_lookup>. - v1.9 - 20020120 Richard Gooch <rgooch@atnf.csiro.au> - Fixed deadlock bug in <devfs_d_revalidate_wait>. - Tag VFS deletable in <devfs_mk_symlink> if handle ignored. - v1.10 - 20020129 Richard Gooch <rgooch@atnf.csiro.au> - Added KERN_* to remaining messages. - Cleaned up declaration of <stat_read>. - v1.11 - 20020219 Richard Gooch <rgooch@atnf.csiro.au> - Changed <devfs_rmdir> to allow later additions if not yet empty. - v1.12 - 20020406 Richard Gooch <rgooch@atnf.csiro.au> - Removed silently introduced calls to lock_kernel() and - unlock_kernel() due to recent VFS locking changes. BKL isn't - required in devfs. - v1.13 - 20020428 Richard Gooch <rgooch@atnf.csiro.au> - Removed 2.4.x compatibility code. - v1.14 - 20020510 Richard Gooch <rgooch@atnf.csiro.au> - Added BKL to <devfs_open> because drivers still need it. - v1.15 - 20020512 Richard Gooch <rgooch@atnf.csiro.au> - Protected <scan_dir_for_removable> and <get_removable_partition> - from changing directory contents. - v1.16 - 20020514 Richard Gooch <rgooch@atnf.csiro.au> - Minor cleanup of <scan_dir_for_removable>. - v1.17 - 20020721 Richard Gooch <rgooch@atnf.csiro.au> - Switched to ISO C structure field initialisers. - Switch to set_current_state() and move before add_wait_queue(). - 20020722 Richard Gooch <rgooch@atnf.csiro.au> - Fixed devfs entry leak in <devfs_readdir> when *readdir fails. - v1.18 - 20020725 Richard Gooch <rgooch@atnf.csiro.au> - Created <devfs_find_and_unregister>. - v1.19 - 20020728 Richard Gooch <rgooch@atnf.csiro.au> - Removed deprecated <devfs_find_handle>. - v1.20 - 20020820 Richard Gooch <rgooch@atnf.csiro.au> - Fixed module unload race in <devfs_open>. - v1.21 - 20021013 Richard Gooch <rgooch@atnf.csiro.au> - Removed DEVFS_ FL_AUTO_OWNER. - Switched lingering structure field initialiser to ISO C. - Added locking when updating FCB flags. - v1.22 -*/ -#include <linux/types.h> -#include <linux/errno.h> -#include <linux/time.h> -#include <linux/tty.h> -#include <linux/timer.h> -#include <linux/config.h> -#include <linux/kernel.h> -#include <linux/wait.h> -#include <linux/string.h> -#include <linux/slab.h> -#include <linux/ioport.h> -#include <linux/delay.h> -#include <linux/ctype.h> -#include <linux/mm.h> -#include <linux/module.h> -#include <linux/init.h> -#include <linux/devfs_fs.h> -#include <linux/devfs_fs_kernel.h> -#include <linux/smp_lock.h> -#include <linux/smp.h> -#include <linux/rwsem.h> -#include <linux/sched.h> -#include <linux/namei.h> -#include <linux/bitops.h> - -#include <asm/uaccess.h> -#include <asm/io.h> -#include <asm/processor.h> -#include <asm/system.h> -#include <asm/pgtable.h> -#include <asm/atomic.h> - -#define DEVFS_VERSION "2004-01-31" - -#define DEVFS_NAME "devfs" - -#define FIRST_INODE 1 - -#define STRING_LENGTH 256 -#define FAKE_BLOCK_SIZE 1024 -#define POISON_PTR ( *(void **) poison_array ) -#define MAGIC_VALUE 0x327db823 - -#ifndef TRUE -# define TRUE 1 -# define FALSE 0 -#endif - -#define MODE_DIR (S_IFDIR | S_IWUSR | S_IRUGO | S_IXUGO) - -#define DEBUG_NONE 0x0000000 -#define DEBUG_MODULE_LOAD 0x0000001 -#define DEBUG_REGISTER 0x0000002 -#define DEBUG_UNREGISTER 0x0000004 -#define DEBUG_FREE 0x0000008 -#define DEBUG_SET_FLAGS 0x0000010 -#define DEBUG_S_READ 0x0000100 /* Break */ -#define DEBUG_I_LOOKUP 0x0001000 /* Break */ -#define DEBUG_I_CREATE 0x0002000 -#define DEBUG_I_GET 0x0004000 -#define DEBUG_I_CHANGE 0x0008000 -#define DEBUG_I_UNLINK 0x0010000 -#define DEBUG_I_RLINK 0x0020000 -#define DEBUG_I_FLINK 0x0040000 -#define DEBUG_I_MKNOD 0x0080000 -#define DEBUG_F_READDIR 0x0100000 /* Break */ -#define DEBUG_D_DELETE 0x1000000 /* Break */ -#define DEBUG_D_RELEASE 0x2000000 -#define DEBUG_D_IPUT 0x4000000 -#define DEBUG_ALL 0xfffffff -#define DEBUG_DISABLED DEBUG_NONE - -#define OPTION_NONE 0x00 -#define OPTION_MOUNT 0x01 - -#define PRINTK(format, args...) \ - {printk (KERN_ERR "%s" format, __FUNCTION__ , ## args);} - -#define OOPS(format, args...) \ - {printk (KERN_CRIT "%s" format, __FUNCTION__ , ## args); \ - printk ("Forcing Oops\n"); \ - BUG();} - -#ifdef CONFIG_DEVFS_DEBUG -# define VERIFY_ENTRY(de) \ - {if ((de) && (de)->magic_number != MAGIC_VALUE) \ - OOPS ("(%p): bad magic value: %x\n", (de), (de)->magic_number);} -# define WRITE_ENTRY_MAGIC(de,magic) (de)->magic_number = (magic) -# define DPRINTK(flag, format, args...) \ - {if (devfs_debug & flag) \ - printk (KERN_INFO "%s" format, __FUNCTION__ , ## args);} -#else -# define VERIFY_ENTRY(de) -# define WRITE_ENTRY_MAGIC(de,magic) -# define DPRINTK(flag, format, args...) -#endif - -typedef struct devfs_entry *devfs_handle_t; - -struct directory_type { - rwlock_t lock; /* Lock for searching(R)/updating(W) */ - struct devfs_entry *first; - struct devfs_entry *last; - unsigned char no_more_additions:1; -}; - -struct symlink_type { - unsigned int length; /* Not including the NULL-termimator */ - char *linkname; /* This is NULL-terminated */ -}; - -struct devfs_inode { /* This structure is for "persistent" inode storage */ - struct dentry *dentry; - struct timespec atime; - struct timespec mtime; - struct timespec ctime; - unsigned int ino; /* Inode number as seen in the VFS */ - uid_t uid; - gid_t gid; -}; - -struct devfs_entry { -#ifdef CONFIG_DEVFS_DEBUG - unsigned int magic_number; -#endif - void *info; - atomic_t refcount; /* When this drops to zero, it's unused */ - union { - struct directory_type dir; - dev_t dev; - struct symlink_type symlink; - const char *name; /* Only used for (mode == 0) */ - } u; - struct devfs_entry *prev; /* Previous entry in the parent directory */ - struct devfs_entry *next; /* Next entry in the parent directory */ - struct devfs_entry *parent; /* The parent directory */ - struct devfs_inode inode; - umode_t mode; - unsigned short namelen; /* I think 64k+ filenames are a way off... */ - unsigned char vfs:1; /* Whether the VFS may delete the entry */ - char name[1]; /* This is just a dummy: the allocated array - is bigger. This is NULL-terminated */ -}; - -/* The root of the device tree */ -static struct devfs_entry *root_entry; - -struct devfsd_buf_entry { - struct devfs_entry *de; /* The name is generated with this */ - unsigned short type; /* The type of event */ - umode_t mode; - uid_t uid; - gid_t gid; - struct devfsd_buf_entry *next; -}; - -struct fs_info { /* This structure is for the mounted devfs */ - struct super_block *sb; - spinlock_t devfsd_buffer_lock; /* Lock when inserting/deleting events */ - struct devfsd_buf_entry *devfsd_first_event; - struct devfsd_buf_entry *devfsd_last_event; - volatile int devfsd_sleeping; - volatile struct task_struct *devfsd_task; - volatile pid_t devfsd_pgrp; - volatile struct file *devfsd_file; - struct devfsd_notify_struct *devfsd_info; - volatile unsigned long devfsd_event_mask; - atomic_t devfsd_overrun_count; - wait_queue_head_t devfsd_wait_queue; /* Wake devfsd on input */ - wait_queue_head_t revalidate_wait_queue; /* Wake when devfsd sleeps */ -}; - -static struct fs_info fs_info = {.devfsd_buffer_lock = SPIN_LOCK_UNLOCKED }; -static kmem_cache_t *devfsd_buf_cache; -#ifdef CONFIG_DEVFS_DEBUG -static unsigned int devfs_debug_init __initdata = DEBUG_NONE; -static unsigned int devfs_debug = DEBUG_NONE; -static DEFINE_SPINLOCK(stat_lock); -static unsigned int stat_num_entries; -static unsigned int stat_num_bytes; -#endif -static unsigned char poison_array[8] = - { 0x5a, 0x5a, 0x5a, 0x5a, 0x5a, 0x5a, 0x5a, 0x5a }; - -#ifdef CONFIG_DEVFS_MOUNT -static unsigned int boot_options = OPTION_MOUNT; -#else -static unsigned int boot_options = OPTION_NONE; -#endif - -/* Forward function declarations */ -static devfs_handle_t _devfs_walk_path(struct devfs_entry *dir, - const char *name, int namelen, - int traverse_symlink); -static ssize_t devfsd_read(struct file *file, char __user *buf, size_t len, - loff_t * ppos); -static int devfsd_ioctl(struct inode *inode, struct file *file, - unsigned int cmd, unsigned long arg); -static int devfsd_close(struct inode *inode, struct file *file); -#ifdef CONFIG_DEVFS_DEBUG -static ssize_t stat_read(struct file *file, char __user *buf, size_t len, - loff_t * ppos); -static const struct file_operations stat_fops = { - .open = nonseekable_open, - .read = stat_read, -}; -#endif - -/* Devfs daemon file operations */ -static const struct file_operations devfsd_fops = { - .open = nonseekable_open, - .read = devfsd_read, - .ioctl = devfsd_ioctl, - .release = devfsd_close, -}; - -/* Support functions follow */ - -/** - * devfs_get - Get a reference to a devfs entry. - * @de: The devfs entry. - */ - -static struct devfs_entry *devfs_get(struct devfs_entry *de) -{ - VERIFY_ENTRY(de); - if (de) - atomic_inc(&de->refcount); - return de; -} /* End Function devfs_get */ - -/** - * devfs_put - Put (release) a reference to a devfs entry. - * @de: The handle to the devfs entry. - */ - -static void devfs_put(devfs_handle_t de) -{ - if (!de) - return; - VERIFY_ENTRY(de); - if (de->info == POISON_PTR) - OOPS("(%p): poisoned pointer\n", de); - if (!atomic_dec_and_test(&de->refcount)) - return; - if (de == root_entry) - OOPS("(%p): root entry being freed\n", de); - DPRINTK(DEBUG_FREE, "(%s): de: %p, parent: %p \"%s\"\n", - de->name, de, de->parent, - de->parent ? de->parent->name : "no parent"); - if (S_ISLNK(de->mode)) - kfree(de->u.symlink.linkname); - WRITE_ENTRY_MAGIC(de, 0); -#ifdef CONFIG_DEVFS_DEBUG - spin_lock(&stat_lock); - --stat_num_entries; - stat_num_bytes -= sizeof *de + de->namelen; - if (S_ISLNK(de->mode)) - stat_num_bytes -= de->u.symlink.length + 1; - spin_unlock(&stat_lock); -#endif - de->info = POISON_PTR; - kfree(de); -} /* End Function devfs_put */ - -/** - * _devfs_search_dir - Search for a devfs entry in a directory. - * @dir: The directory to search. - * @name: The name of the entry to search for. - * @namelen: The number of characters in @name. - * - * Search for a devfs entry in a directory and returns a pointer to the entry - * on success, else %NULL. The directory must be locked already. - * An implicit devfs_get() is performed on the returned entry. - */ - -static struct devfs_entry *_devfs_search_dir(struct devfs_entry *dir, - const char *name, - unsigned int namelen) -{ - struct devfs_entry *curr; - - if (!S_ISDIR(dir->mode)) { - PRINTK("(%s): not a directory\n", dir->name); - return NULL; - } - for (curr = dir->u.dir.first; curr != NULL; curr = curr->next) { - if (curr->namelen != namelen) - continue; - if (memcmp(curr->name, name, namelen) == 0) - break; - /* Not found: try the next one */ - } - return devfs_get(curr); -} /* End Function _devfs_search_dir */ - -/** - * _devfs_alloc_entry - Allocate a devfs entry. - * @name: the name of the entry - * @namelen: the number of characters in @name - * @mode: the mode for the entry - * - * Allocate a devfs entry and returns a pointer to the entry on success, else - * %NULL. - */ - -static struct devfs_entry *_devfs_alloc_entry(const char *name, - unsigned int namelen, - umode_t mode) -{ - struct devfs_entry *new; - static unsigned long inode_counter = FIRST_INODE; - static DEFINE_SPINLOCK(counter_lock); - - if (name && (namelen < 1)) - namelen = strlen(name); - if ((new = kmalloc(sizeof *new + namelen, GFP_KERNEL)) == NULL) - return NULL; - memset(new, 0, sizeof *new + namelen); /* Will set '\0' on name */ - new->mode = mode; - if (S_ISDIR(mode)) - rwlock_init(&new->u.dir.lock); - atomic_set(&new->refcount, 1); - spin_lock(&counter_lock); - new->inode.ino = inode_counter++; - spin_unlock(&counter_lock); - if (name) - memcpy(new->name, name, namelen); - new->namelen = namelen; - WRITE_ENTRY_MAGIC(new, MAGIC_VALUE); -#ifdef CONFIG_DEVFS_DEBUG - spin_lock(&stat_lock); - ++stat_num_entries; - stat_num_bytes += sizeof *new + namelen; - spin_unlock(&stat_lock); -#endif - return new; -} /* End Function _devfs_alloc_entry */ - -/** - * _devfs_append_entry - Append a devfs entry to a directory's child list. - * @dir: The directory to add to. - * @de: The devfs entry to append. - * @old_de: If an existing entry exists, it will be written here. This may - * be %NULL. An implicit devfs_get() is performed on this entry. - * - * Append a devfs entry to a directory's list of children, checking first to - * see if an entry of the same name exists. The directory will be locked. - * The value 0 is returned on success, else a negative error code. - * On failure, an implicit devfs_put() is performed on %de. - */ - -static int _devfs_append_entry(devfs_handle_t dir, devfs_handle_t de, - devfs_handle_t * old_de) -{ - int retval; - - if (old_de) - *old_de = NULL; - if (!S_ISDIR(dir->mode)) { - PRINTK("(%s): dir: \"%s\" is not a directory\n", de->name, - dir->name); - devfs_put(de); - return -ENOTDIR; - } - write_lock(&dir->u.dir.lock); - if (dir->u.dir.no_more_additions) - retval = -ENOENT; - else { - struct devfs_entry *old; - - old = _devfs_search_dir(dir, de->name, de->namelen); - if (old_de) - *old_de = old; - else - devfs_put(old); - if (old == NULL) { - de->parent = dir; - de->prev = dir->u.dir.last; - /* Append to the directory's list of children */ - if (dir->u.dir.first == NULL) - dir->u.dir.first = de; - else - dir->u.dir.last->next = de; - dir->u.dir.last = de; - retval = 0; - } else - retval = -EEXIST; - } - write_unlock(&dir->u.dir.lock); - if (retval) - devfs_put(de); - return retval; -} /* End Function _devfs_append_entry */ - -/** - * _devfs_get_root_entry - Get the root devfs entry. - * - * Returns the root devfs entry on success, else %NULL. - * - * TODO it must be called asynchronously due to the fact - * that devfs is initialized relatively late. Proper way - * is to remove module_init from init_devfs_fs and manually - * call it early enough during system init - */ - -static struct devfs_entry *_devfs_get_root_entry(void) -{ - struct devfs_entry *new; - static DEFINE_SPINLOCK(root_lock); - - if (root_entry) - return root_entry; - - new = _devfs_alloc_entry(NULL, 0, MODE_DIR); - if (new == NULL) - return NULL; - - spin_lock(&root_lock); - if (root_entry) { - spin_unlock(&root_lock); - devfs_put(new); - return root_entry; - } - root_entry = new; - spin_unlock(&root_lock); - - return root_entry; -} /* End Function _devfs_get_root_entry */ - -/** - * _devfs_descend - Descend down a tree using the next component name. - * @dir: The directory to search. - * @name: The component name to search for. - * @namelen: The length of %name. - * @next_pos: The position of the next '/' or '\0' is written here. - * - * Descend into a directory, searching for a component. This function forms - * the core of a tree-walking algorithm. The directory will be locked. - * The devfs entry corresponding to the component is returned. If there is - * no matching entry, %NULL is returned. - * An implicit devfs_get() is performed on the returned entry. - */ - -static struct devfs_entry *_devfs_descend(struct devfs_entry *dir, - const char *name, int namelen, - int *next_pos) -{ - const char *stop, *ptr; - struct devfs_entry *entry; - - if ((namelen >= 3) && (strncmp(name, "../", 3) == 0)) { /* Special-case going to parent directory */ - *next_pos = 3; - return devfs_get(dir->parent); - } - stop = name + namelen; - /* Search for a possible '/' */ - for (ptr = name; (ptr < stop) && (*ptr != '/'); ++ptr) ; - *next_pos = ptr - name; - read_lock(&dir->u.dir.lock); - entry = _devfs_search_dir(dir, name, *next_pos); - read_unlock(&dir->u.dir.lock); - return entry; -} /* End Function _devfs_descend */ - -static devfs_handle_t _devfs_make_parent_for_leaf(struct devfs_entry *dir, - const char *name, - int namelen, int *leaf_pos) -{ - int next_pos = 0; - - if (dir == NULL) - dir = _devfs_get_root_entry(); - if (dir == NULL) - return NULL; - devfs_get(dir); - /* Search for possible trailing component and ignore it */ - for (--namelen; (namelen > 0) && (name[namelen] != '/'); --namelen) ; - *leaf_pos = (name[namelen] == '/') ? (namelen + 1) : 0; - for (; namelen > 0; name += next_pos, namelen -= next_pos) { - struct devfs_entry *de, *old = NULL; - - if ((de = - _devfs_descend(dir, name, namelen, &next_pos)) == NULL) { - de = _devfs_alloc_entry(name, next_pos, MODE_DIR); - devfs_get(de); - if (!de || _devfs_append_entry(dir, de, &old)) { - devfs_put(de); - if (!old || !S_ISDIR(old->mode)) { - devfs_put(old); - devfs_put(dir); - return NULL; - } - de = old; /* Use the existing directory */ - } - } - if (de == dir->parent) { - devfs_put(dir); - devfs_put(de); - return NULL; - } - devfs_put(dir); - dir = de; - if (name[next_pos] == '/') - ++next_pos; - } - return dir; -} /* End Function _devfs_make_parent_for_leaf */ - -static devfs_handle_t _devfs_prepare_leaf(devfs_handle_t * dir, - const char *name, umode_t mode) -{ - int namelen, leaf_pos; - struct devfs_entry *de; - - namelen = strlen(name); - if ((*dir = _devfs_make_parent_for_leaf(*dir, name, namelen, - &leaf_pos)) == NULL) { - PRINTK("(%s): could not create parent path\n", name); - return NULL; - } - if ((de = _devfs_alloc_entry(name + leaf_pos, namelen - leaf_pos, mode)) - == NULL) { - PRINTK("(%s): could not allocate entry\n", name); - devfs_put(*dir); - return NULL; - } - return de; -} /* End Function _devfs_prepare_leaf */ - -static devfs_handle_t _devfs_walk_path(struct devfs_entry *dir, - const char *name, int namelen, - int traverse_symlink) -{ - int next_pos = 0; - - if (dir == NULL) - dir = _devfs_get_root_entry(); - if (dir == NULL) - return NULL; - devfs_get(dir); - for (; namelen > 0; name += next_pos, namelen -= next_pos) { - struct devfs_entry *de, *link; - - if (!S_ISDIR(dir->mode)) { - devfs_put(dir); - return NULL; - } - - if ((de = - _devfs_descend(dir, name, namelen, &next_pos)) == NULL) { - devfs_put(dir); - return NULL; - } - if (S_ISLNK(de->mode) && traverse_symlink) { /* Need to follow the link: this is a stack chomper */ - /* FIXME what if it puts outside of mounted tree? */ - link = _devfs_walk_path(dir, de->u.symlink.linkname, - de->u.symlink.length, TRUE); - devfs_put(de); - if (!link) { - devfs_put(dir); - return NULL; - } - de = link; - } - devfs_put(dir); - dir = de; - if (name[next_pos] == '/') - ++next_pos; - } - return dir; -} /* End Function _devfs_walk_path */ - -/** - * _devfs_find_entry - Find a devfs entry. - * @dir: The handle to the parent devfs directory entry. If this is %NULL the - * name is relative to the root of the devfs. - * @name: The name of the entry. This may be %NULL. - * @traverse_symlink: If %TRUE then symbolic links are traversed. - * - * Returns the devfs_entry pointer on success, else %NULL. An implicit - * devfs_get() is performed. - */ - -static struct devfs_entry *_devfs_find_entry(devfs_handle_t dir, - const char *name, - int traverse_symlink) -{ - unsigned int namelen = strlen(name); - - if (name[0] == '/') { - /* Skip leading pathname component */ - if (namelen < 2) { - PRINTK("(%s): too short\n", name); - return NULL; - } - for (++name, --namelen; (*name != '/') && (namelen > 0); - ++name, --namelen) ; - if (namelen < 2) { - PRINTK("(%s): too short\n", name); - return NULL; - } - ++name; - --namelen; - } - return _devfs_walk_path(dir, name, namelen, traverse_symlink); -} /* End Function _devfs_find_entry */ - -static struct devfs_entry *get_devfs_entry_from_vfs_inode(struct inode *inode) -{ - if (inode == NULL) - return NULL; - VERIFY_ENTRY((struct devfs_entry *)inode->u.generic_ip); - return inode->u.generic_ip; -} /* End Function get_devfs_entry_from_vfs_inode */ - -/** - * free_dentry - Free the dentry for a device entry and invalidate inode. - * @de: The entry. - * - * This must only be called after the entry has been unhooked from its - * parent directory. - */ - -static void free_dentry(struct devfs_entry *de) -{ - struct dentry *dentry = de->inode.dentry; - - if (!dentry) - return; - spin_lock(&dcache_lock); - dget_locked(dentry); - spin_unlock(&dcache_lock); - /* Forcefully remove the inode */ - if (dentry->d_inode != NULL) - dentry->d_inode->i_nlink = 0; - d_drop(dentry); - dput(dentry); -} /* End Function free_dentry */ - -/** - * is_devfsd_or_child - Test if the current process is devfsd or one of its children. - * @fs_info: The filesystem information. - * - * Returns %TRUE if devfsd or child, else %FALSE. - */ - -static int is_devfsd_or_child(struct fs_info *fs_info) -{ - struct task_struct *p = current; - - if (p == fs_info->devfsd_task) - return (TRUE); - if (process_group(p) == fs_info->devfsd_pgrp) - return (TRUE); - read_lock(&tasklist_lock); - for (; p != &init_task; p = p->real_parent) { - if (p == fs_info->devfsd_task) { - read_unlock(&tasklist_lock); - return (TRUE); - } - } - read_unlock(&tasklist_lock); - return (FALSE); -} /* End Function is_devfsd_or_child */ - -/** - * devfsd_queue_empty - Test if devfsd has work pending in its event queue. - * @fs_info: The filesystem information. - * - * Returns %TRUE if the queue is empty, else %FALSE. - */ - -static inline int devfsd_queue_empty(struct fs_info *fs_info) -{ - return (fs_info->devfsd_last_event) ? FALSE : TRUE; -} /* End Function devfsd_queue_empty */ - -/** - * wait_for_devfsd_finished - Wait for devfsd to finish processing its event queue. - * @fs_info: The filesystem information. - * - * Returns %TRUE if no more waiting will be required, else %FALSE. - */ - -static int wait_for_devfsd_finished(struct fs_info *fs_info) -{ - DECLARE_WAITQUEUE(wait, current); - - if (fs_info->devfsd_task == NULL) - return (TRUE); - if (devfsd_queue_empty(fs_info) && fs_info->devfsd_sleeping) - return TRUE; - if (is_devfsd_or_child(fs_info)) - return (FALSE); - set_current_state(TASK_UNINTERRUPTIBLE); - add_wait_queue(&fs_info->revalidate_wait_queue, &wait); - if (!devfsd_queue_empty(fs_info) || !fs_info->devfsd_sleeping) - if (fs_info->devfsd_task) - schedule(); - remove_wait_queue(&fs_info->revalidate_wait_queue, &wait); - __set_current_state(TASK_RUNNING); - return (TRUE); -} /* End Function wait_for_devfsd_finished */ - -/** - * devfsd_notify_de - Notify the devfsd daemon of a change. - * @de: The devfs entry that has changed. This and all parent entries will - * have their reference counts incremented if the event was queued. - * @type: The type of change. - * @mode: The mode of the entry. - * @uid: The user ID. - * @gid: The group ID. - * @fs_info: The filesystem info. - * - * Returns %TRUE if an event was queued and devfsd woken up, else %FALSE. - */ - -static int devfsd_notify_de(struct devfs_entry *de, - unsigned short type, umode_t mode, - uid_t uid, gid_t gid, struct fs_info *fs_info) -{ - struct devfsd_buf_entry *entry; - struct devfs_entry *curr; - - if (!(fs_info->devfsd_event_mask & (1 << type))) - return (FALSE); - if ((entry = kmem_cache_alloc(devfsd_buf_cache, SLAB_KERNEL)) == NULL) { - atomic_inc(&fs_info->devfsd_overrun_count); - return (FALSE); - } - for (curr = de; curr != NULL; curr = curr->parent) - devfs_get(curr); - entry->de = de; - entry->type = type; - entry->mode = mode; - entry->uid = uid; - entry->gid = gid; - entry->next = NULL; - spin_lock(&fs_info->devfsd_buffer_lock); - if (!fs_info->devfsd_first_event) - fs_info->devfsd_first_event = entry; - if (fs_info->devfsd_last_event) - fs_info->devfsd_last_event->next = entry; - fs_info->devfsd_last_event = entry; - spin_unlock(&fs_info->devfsd_buffer_lock); - wake_up_interruptible(&fs_info->devfsd_wait_queue); - return (TRUE); -} /* End Function devfsd_notify_de */ - -/** - * devfsd_notify - Notify the devfsd daemon of a change. - * @de: The devfs entry that has changed. - * @type: The type of change event. - * @wait: If TRUE, the function waits for the daemon to finish processing - * the event. - */ - -static void devfsd_notify(struct devfs_entry *de, unsigned short type) -{ - devfsd_notify_de(de, type, de->mode, current->euid, - current->egid, &fs_info); -} - -static int devfs_mk_dev(dev_t dev, umode_t mode, const char *fmt, va_list args) -{ - struct devfs_entry *dir = NULL, *de; - char buf[64]; - int error, n; - - n = vsnprintf(buf, sizeof(buf), fmt, args); - if (n >= sizeof(buf) || !buf[0]) { - printk(KERN_WARNING "%s: invalid format string %s\n", - __FUNCTION__, fmt); - return -EINVAL; - } - - de = _devfs_prepare_leaf(&dir, buf, mode); - if (!de) { - printk(KERN_WARNING "%s: could not prepare leaf for %s\n", - __FUNCTION__, buf); - return -ENOMEM; /* could be more accurate... */ - } - - de->u.dev = dev; - - error = _devfs_append_entry(dir, de, NULL); - if (error) { - printk(KERN_WARNING "%s: could not append to parent for %s\n", - __FUNCTION__, buf); - goto out; - } - - devfsd_notify(de, DEVFSD_NOTIFY_REGISTERED); - out: - devfs_put(dir); - return error; -} - -int devfs_mk_bdev(dev_t dev, umode_t mode, const char *fmt, ...) -{ - va_list args; - - if (!S_ISBLK(mode)) { - printk(KERN_WARNING "%s: invalide mode (%u) for %s\n", - __FUNCTION__, mode, fmt); - return -EINVAL; - } - - va_start(args, fmt); - return devfs_mk_dev(dev, mode, fmt, args); -} - -EXPORT_SYMBOL(devfs_mk_bdev); - -int devfs_mk_cdev(dev_t dev, umode_t mode, const char *fmt, ...) -{ - va_list args; - - if (!S_ISCHR(mode)) { - printk(KERN_WARNING "%s: invalide mode (%u) for %s\n", - __FUNCTION__, mode, fmt); - return -EINVAL; - } - - va_start(args, fmt); - return devfs_mk_dev(dev, mode, fmt, args); -} - -EXPORT_SYMBOL(devfs_mk_cdev); - -/** - * _devfs_unhook - Unhook a device entry from its parents list - * @de: The entry to unhook. - * - * Returns %TRUE if the entry was unhooked, else %FALSE if it was - * previously unhooked. - * The caller must have a write lock on the parent directory. - */ - -static int _devfs_unhook(struct devfs_entry *de) -{ - struct devfs_entry *parent; - - if (!de || (de->prev == de)) - return FALSE; - parent = de->parent; - if (de->prev == NULL) - parent->u.dir.first = de->next; - else - de->prev->next = de->next; - if (de->next == NULL) - parent->u.dir.last = de->prev; - else - de->next->prev = de->prev; - de->prev = de; /* Indicate we're unhooked */ - de->next = NULL; /* Force early termination for <devfs_readdir> */ - return TRUE; -} /* End Function _devfs_unhook */ - -/** - * _devfs_unregister - Unregister a device entry from its parent. - * @dir: The parent directory. - * @de: The entry to unregister. - * - * The caller must have a write lock on the parent directory, which is - * unlocked by this function. - */ - -static void _devfs_unregister(struct devfs_entry *dir, struct devfs_entry *de) -{ - int unhooked = _devfs_unhook(de); - - write_unlock(&dir->u.dir.lock); - if (!unhooked) - return; - devfs_get(dir); - devfsd_notify(de, DEVFSD_NOTIFY_UNREGISTERED); - free_dentry(de); - devfs_put(dir); - if (!S_ISDIR(de->mode)) - return; - while (TRUE) { /* Recursively unregister: this is a stack chomper */ - struct devfs_entry *child; - - write_lock(&de->u.dir.lock); - de->u.dir.no_more_additions = TRUE; - child = de->u.dir.first; - VERIFY_ENTRY(child); - _devfs_unregister(de, child); - if (!child) - break; - DPRINTK(DEBUG_UNREGISTER, "(%s): child: %p refcount: %d\n", - child->name, child, atomic_read(&child->refcount)); - devfs_put(child); - } -} /* End Function _devfs_unregister */ - -static int devfs_do_symlink(devfs_handle_t dir, const char *name, - const char *link, devfs_handle_t * handle) -{ - int err; - unsigned int linklength; - char *newlink; - struct devfs_entry *de; - - if (handle != NULL) - *handle = NULL; - if (name == NULL) { - PRINTK("(): NULL name pointer\n"); - return -EINVAL; - } - if (link == NULL) { - PRINTK("(%s): NULL link pointer\n", name); - return -EINVAL; - } - linklength = strlen(link); - if ((newlink = kmalloc(linklength + 1, GFP_KERNEL)) == NULL) - return -ENOMEM; - memcpy(newlink, link, linklength); - newlink[linklength] = '\0'; - if ((de = _devfs_prepare_leaf(&dir, name, S_IFLNK | S_IRUGO | S_IXUGO)) - == NULL) { - PRINTK("(%s): could not prepare leaf\n", name); - kfree(newlink); - return -ENOTDIR; - } - de->info = NULL; - de->u.symlink.linkname = newlink; - de->u.symlink.length = linklength; - if ((err = _devfs_append_entry(dir, de, NULL)) != 0) { - PRINTK("(%s): could not append to parent, err: %d\n", name, - err); - devfs_put(dir); - return err; - } - devfs_put(dir); -#ifdef CONFIG_DEVFS_DEBUG - spin_lock(&stat_lock); - stat_num_bytes += linklength + 1; - spin_unlock(&stat_lock); -#endif - if (handle != NULL) - *handle = de; - return 0; -} /* End Function devfs_do_symlink */ - -/** - * devfs_mk_symlink Create a symbolic link in the devfs namespace. - * @from: The name of the entry. - * @to: Name of the destination - * - * Returns 0 on success, else a negative error code is returned. - */ - -int devfs_mk_symlink(const char *from, const char *to) -{ - devfs_handle_t de; - int err; - - err = devfs_do_symlink(NULL, from, to, &de); - if (!err) { - de->vfs = TRUE; - devfsd_notify(de, DEVFSD_NOTIFY_REGISTERED); - } - - return err; -} - -/** - * devfs_mk_dir - Create a directory in the devfs namespace. - * new name is relative to the root of the devfs. - * @fmt: The name of the entry. - * - * Use of this function is optional. The devfs_register() function - * will automatically create intermediate directories as needed. This function - * is provided for efficiency reasons, as it provides a handle to a directory. - * On failure %NULL is returned. - */ - -int devfs_mk_dir(const char *fmt, ...) -{ - struct devfs_entry *dir = NULL, *de = NULL, *old; - char buf[64]; - va_list args; - int error, n; - - va_start(args, fmt); - n = vsnprintf(buf, 64, fmt, args); - if (n >= 64 || !buf[0]) { - printk(KERN_WARNING "%s: invalid argument.", __FUNCTION__); - return -EINVAL; - } - - de = _devfs_prepare_leaf(&dir, buf, MODE_DIR); - if (!de) { - PRINTK("(%s): could not prepare leaf\n", buf); - return -EINVAL; - } - - error = _devfs_append_entry(dir, de, &old); - if (error == -EEXIST && S_ISDIR(old->mode)) { - /* - * devfs_mk_dir() of an already-existing directory will - * return success. - */ - error = 0; - goto out_put; - } else if (error) { - PRINTK("(%s): could not append to dir: %p \"%s\"\n", - buf, dir, dir->name); - devfs_put(old); - goto out_put; - } - - devfsd_notify(de, DEVFSD_NOTIFY_REGISTERED); - - out_put: - devfs_put(dir); - return error; -} - -void devfs_remove(const char *fmt, ...) -{ - char buf[64]; - va_list args; - int n; - - va_start(args, fmt); - n = vsnprintf(buf, sizeof(buf), fmt, args); - if (n < sizeof(buf) && buf[0]) { - devfs_handle_t de = _devfs_find_entry(NULL, buf, 0); - - if (!de) { - printk(KERN_ERR "%s: %s not found, cannot remove\n", - __FUNCTION__, buf); - dump_stack(); - return; - } - - write_lock(&de->parent->u.dir.lock); - _devfs_unregister(de->parent, de); - devfs_put(de); - devfs_put(de); - } -} - -/** - * devfs_generate_path - Generate a pathname for an entry, relative to the devfs root. - * @de: The devfs entry. - * @path: The buffer to write the pathname to. The pathname and '\0' - * terminator will be written at the end of the buffer. - * @buflen: The length of the buffer. - * - * Returns the offset in the buffer where the pathname starts on success, - * else a negative error code. - */ - -static int devfs_generate_path(devfs_handle_t de, char *path, int buflen) -{ - int pos; -#define NAMEOF(de) ( (de)->mode ? (de)->name : (de)->u.name ) - - if (de == NULL) - return -EINVAL; - VERIFY_ENTRY(de); - if (de->namelen >= buflen) - return -ENAMETOOLONG; /* Must be first */ - path[buflen - 1] = '\0'; - if (de->parent == NULL) - return buflen - 1; /* Don't prepend root */ - pos = buflen - de->namelen - 1; - memcpy(path + pos, NAMEOF(de), de->namelen); - for (de = de->parent; de->parent != NULL; de = de->parent) { - if (pos - de->namelen - 1 < 0) - return -ENAMETOOLONG; - path[--pos] = '/'; - pos -= de->namelen; - memcpy(path + pos, NAMEOF(de), de->namelen); - } - return pos; -} /* End Function devfs_generate_path */ - -/** - * devfs_setup - Process kernel boot options. - * @str: The boot options after the "devfs=". - */ - -static int __init devfs_setup(char *str) -{ - static struct { - char *name; - unsigned int mask; - unsigned int *opt; - } devfs_options_tab[] __initdata = { -#ifdef CONFIG_DEVFS_DEBUG - { - "dall", DEBUG_ALL, &devfs_debug_init}, { - "dmod", DEBUG_MODULE_LOAD, &devfs_debug_init}, { - "dreg", DEBUG_REGISTER, &devfs_debug_init}, { - "dunreg", DEBUG_UNREGISTER, &devfs_debug_init}, { - "dfree", DEBUG_FREE, &devfs_debug_init}, { - "diget", DEBUG_I_GET, &devfs_debug_init}, { - "dchange", DEBUG_SET_FLAGS, &devfs_debug_init}, { - "dsread", DEBUG_S_READ, &devfs_debug_init}, { - "dichange", DEBUG_I_CHANGE, &devfs_debug_init}, { - "dimknod", DEBUG_I_MKNOD, &devfs_debug_init}, { - "dilookup", DEBUG_I_LOOKUP, &devfs_debug_init}, { - "diunlink", DEBUG_I_UNLINK, &devfs_debug_init}, -#endif /* CONFIG_DEVFS_DEBUG */ - { - "mount", OPTION_MOUNT, &boot_options}, { - NULL, 0, NULL} - }; - - while ((*str != '\0') && !isspace(*str)) { - int i, found = 0, invert = 0; - - if (strncmp(str, "no", 2) == 0) { - invert = 1; - str += 2; - } - for (i = 0; devfs_options_tab[i].name != NULL; i++) { - int len = strlen(devfs_options_tab[i].name); - - if (strncmp(str, devfs_options_tab[i].name, len) == 0) { - if (invert) - *devfs_options_tab[i].opt &= - ~devfs_options_tab[i].mask; - else - *devfs_options_tab[i].opt |= - devfs_options_tab[i].mask; - str += len; - found = 1; - break; - } - } - if (!found) - return 0; /* No match */ - if (*str != ',') - return 0; /* No more options */ - ++str; - } - return 1; -} /* End Function devfs_setup */ - -__setup("devfs=", devfs_setup); - -EXPORT_SYMBOL(devfs_mk_dir); -EXPORT_SYMBOL(devfs_remove); - -/** - * try_modload - Notify devfsd of an inode lookup by a non-devfsd process. - * @parent: The parent devfs entry. - * @fs_info: The filesystem info. - * @name: The device name. - * @namelen: The number of characters in @name. - * @buf: A working area that will be used. This must not go out of scope - * until devfsd is idle again. - * - * Returns 0 on success (event was queued), else a negative error code. - */ - -static int try_modload(struct devfs_entry *parent, struct fs_info *fs_info, - const char *name, unsigned namelen, - struct devfs_entry *buf) -{ - if (!(fs_info->devfsd_event_mask & (1 << DEVFSD_NOTIFY_LOOKUP))) - return -ENOENT; - if (is_devfsd_or_child(fs_info)) - return -ENOENT; - memset(buf, 0, sizeof *buf); - atomic_set(&buf->refcount, 1); - buf->parent = parent; - buf->namelen = namelen; - buf->u.name = name; - WRITE_ENTRY_MAGIC(buf, MAGIC_VALUE); - if (!devfsd_notify_de(buf, DEVFSD_NOTIFY_LOOKUP, 0, - current->euid, current->egid, fs_info)) - return -ENOENT; - /* Possible success: event has been queued */ - return 0; -} /* End Function try_modload */ - -/* Superblock operations follow */ - -static struct inode_operations devfs_iops; -static struct inode_operations devfs_dir_iops; -static const struct file_operations devfs_fops; -static const struct file_operations devfs_dir_fops; -static struct inode_operations devfs_symlink_iops; - -static int devfs_notify_change(struct dentry *dentry, struct iattr *iattr) -{ - int retval; - struct devfs_entry *de; - struct inode *inode = dentry->d_inode; - struct fs_info *fs_info = inode->i_sb->s_fs_info; - - de = get_devfs_entry_from_vfs_inode(inode); - if (de == NULL) - return -ENODEV; - retval = inode_change_ok(inode, iattr); - if (retval != 0) - return retval; - retval = inode_setattr(inode, iattr); - if (retval != 0) - return retval; - DPRINTK(DEBUG_I_CHANGE, "(%d): VFS inode: %p devfs_entry: %p\n", - (int)inode->i_ino, inode, de); - DPRINTK(DEBUG_I_CHANGE, "(): mode: 0%o uid: %d gid: %d\n", - (int)inode->i_mode, (int)inode->i_uid, (int)inode->i_gid); - /* Inode is not on hash chains, thus must save permissions here rather - than in a write_inode() method */ - de->mode = inode->i_mode; - de->inode.uid = inode->i_uid; - de->inode.gid = inode->i_gid; - de->inode.atime = inode->i_atime; - de->inode.mtime = inode->i_mtime; - de->inode.ctime = inode->i_ctime; - if ((iattr->ia_valid & (ATTR_MODE | ATTR_UID | ATTR_GID)) && - !is_devfsd_or_child(fs_info)) - devfsd_notify_de(de, DEVFSD_NOTIFY_CHANGE, inode->i_mode, - inode->i_uid, inode->i_gid, fs_info); - return 0; -} /* End Function devfs_notify_change */ - -static struct super_operations devfs_sops = { - .drop_inode = generic_delete_inode, - .statfs = simple_statfs, -}; - -/** - * _devfs_get_vfs_inode - Get a VFS inode. - * @sb: The super block. - * @de: The devfs inode. - * @dentry: The dentry to register with the devfs inode. - * - * Returns the inode on success, else %NULL. An implicit devfs_get() is - * performed if the inode is created. - */ - -static struct inode *_devfs_get_vfs_inode(struct super_block *sb, - struct devfs_entry *de, - struct dentry *dentry) -{ - struct inode *inode; - - if (de->prev == de) - return NULL; /* Quick check to see if unhooked */ - if ((inode = new_inode(sb)) == NULL) { - PRINTK("(%s): new_inode() failed, de: %p\n", de->name, de); - return NULL; - } - if (de->parent) { - read_lock(&de->parent->u.dir.lock); - if (de->prev != de) - de->inode.dentry = dentry; /* Not unhooked */ - read_unlock(&de->parent->u.dir.lock); - } else - de->inode.dentry = dentry; /* Root: no locking needed */ - if (de->inode.dentry != dentry) { /* Must have been unhooked */ - iput(inode); - return NULL; - } - /* FIXME where is devfs_put? */ - inode->u.generic_ip = devfs_get(de); - inode->i_ino = de->inode.ino; - DPRINTK(DEBUG_I_GET, "(%d): VFS inode: %p devfs_entry: %p\n", - (int)inode->i_ino, inode, de); - inode->i_blocks = 0; - inode->i_blksize = FAKE_BLOCK_SIZE; - inode->i_op = &devfs_iops; - inode->i_mode = de->mode; - if (S_ISDIR(de->mode)) { - inode->i_op = &devfs_dir_iops; - inode->i_fop = &devfs_dir_fops; - } else if (S_ISLNK(de->mode)) { - inode->i_op = &devfs_symlink_iops; - inode->i_size = de->u.symlink.length; - } else if (S_ISCHR(de->mode) || S_ISBLK(de->mode)) { - init_special_inode(inode, de->mode, de->u.dev); - } else if (S_ISFIFO(de->mode) || S_ISSOCK(de->mode)) { - init_special_inode(inode, de->mode, 0); - } else { - PRINTK("(%s): unknown mode %o de: %p\n", - de->name, de->mode, de); - iput(inode); - devfs_put(de); - return NULL; - } - - inode->i_uid = de->inode.uid; - inode->i_gid = de->inode.gid; - inode->i_atime = de->inode.atime; - inode->i_mtime = de->inode.mtime; - inode->i_ctime = de->inode.ctime; - DPRINTK(DEBUG_I_GET, "(): mode: 0%o uid: %d gid: %d\n", - (int)inode->i_mode, (int)inode->i_uid, (int)inode->i_gid); - return inode; -} /* End Function _devfs_get_vfs_inode */ - -/* File operations for device entries follow */ - -static int devfs_readdir(struct file *file, void *dirent, filldir_t filldir) -{ - int err, count; - int stored = 0; - struct fs_info *fs_info; - struct devfs_entry *parent, *de, *next = NULL; - struct inode *inode = file->f_dentry->d_inode; - - fs_info = inode->i_sb->s_fs_info; - parent = get_devfs_entry_from_vfs_inode(file->f_dentry->d_inode); - if ((long)file->f_pos < 0) - return -EINVAL; - DPRINTK(DEBUG_F_READDIR, "(%s): fs_info: %p pos: %ld\n", - parent->name, fs_info, (long)file->f_pos); - switch ((long)file->f_pos) { - case 0: - err = (*filldir) (dirent, "..", 2, file->f_pos, - parent_ino(file->f_dentry), DT_DIR); - if (err == -EINVAL) - break; - if (err < 0) - return err; - file->f_pos++; - ++stored; - /* Fall through */ - case 1: - err = - (*filldir) (dirent, ".", 1, file->f_pos, inode->i_ino, - DT_DIR); - if (err == -EINVAL) - break; - if (err < 0) - return err; - file->f_pos++; - ++stored; - /* Fall through */ - default: - /* Skip entries */ - count = file->f_pos - 2; - read_lock(&parent->u.dir.lock); - for (de = parent->u.dir.first; de && (count > 0); de = de->next) - --count; - devfs_get(de); - read_unlock(&parent->u.dir.lock); - /* Now add all remaining entries */ - while (de) { - err = (*filldir) (dirent, de->name, de->namelen, - file->f_pos, de->inode.ino, - de->mode >> 12); - if (err < 0) - devfs_put(de); - else { - file->f_pos++; - ++stored; - } - if (err == -EINVAL) - break; - if (err < 0) - return err; - read_lock(&parent->u.dir.lock); - next = devfs_get(de->next); - read_unlock(&parent->u.dir.lock); - devfs_put(de); - de = next; - } - break; - } - return stored; -} /* End Function devfs_readdir */ - -/* Open devfs specific special files */ -static int devfs_open(struct inode *inode, struct file *file) -{ - int err; - int minor = MINOR(inode->i_rdev); - struct file_operations *old_fops, *new_fops; - - switch (minor) { - case 0: /* /dev/.devfsd */ - new_fops = fops_get(&devfsd_fops); - break; -#ifdef CONFIG_DEVFS_DEBUG - case 1: /* /dev/.stat */ - new_fops = fops_get(&stat_fops); - break; -#endif - default: - return -ENODEV; - } - - if (new_fops == NULL) - return -ENODEV; - old_fops = file->f_op; - file->f_op = new_fops; - err = new_fops->open ? new_fops->open(inode, file) : 0; - if (err) { - file->f_op = old_fops; - fops_put(new_fops); - } else - fops_put(old_fops); - return err; -} /* End Function devfs_open */ - -static const struct file_operations devfs_fops = { - .open = devfs_open, -}; - -static const struct file_operations devfs_dir_fops = { - .read = generic_read_dir, - .readdir = devfs_readdir, -}; - -/* Dentry operations for device entries follow */ - -/** - * devfs_d_release - Callback for when a dentry is freed. - * @dentry: The dentry. - */ - -static void devfs_d_release(struct dentry *dentry) -{ - DPRINTK(DEBUG_D_RELEASE, "(%p): inode: %p\n", dentry, dentry->d_inode); -} /* End Function devfs_d_release */ - -/** - * devfs_d_iput - Callback for when a dentry loses its inode. - * @dentry: The dentry. - * @inode: The inode. - */ - -static void devfs_d_iput(struct dentry *dentry, struct inode *inode) -{ - struct devfs_entry *de; - - de = get_devfs_entry_from_vfs_inode(inode); - DPRINTK(DEBUG_D_IPUT, - "(%s): dentry: %p inode: %p de: %p de->dentry: %p\n", de->name, - dentry, inode, de, de->inode.dentry); - if (de->inode.dentry && (de->inode.dentry != dentry)) - OOPS("(%s): de: %p dentry: %p de->dentry: %p\n", - de->name, de, dentry, de->inode.dentry); - de->inode.dentry = NULL; - iput(inode); - devfs_put(de); -} /* End Function devfs_d_iput */ - -static int devfs_d_delete(struct dentry *dentry); - -static struct dentry_operations devfs_dops = { - .d_delete = devfs_d_delete, - .d_release = devfs_d_release, - .d_iput = devfs_d_iput, -}; - -static int devfs_d_revalidate_wait(struct dentry *dentry, struct nameidata *); - -static struct dentry_operations devfs_wait_dops = { - .d_delete = devfs_d_delete, - .d_release = devfs_d_release, - .d_iput = devfs_d_iput, - .d_revalidate = devfs_d_revalidate_wait, -}; - -/** - * devfs_d_delete - Callback for when all files for a dentry are closed. - * @dentry: The dentry. - */ - -static int devfs_d_delete(struct dentry *dentry) -{ - struct inode *inode = dentry->d_inode; - - if (dentry->d_op == &devfs_wait_dops) - dentry->d_op = &devfs_dops; - /* Unhash dentry if negative (has no inode) */ - if (inode == NULL) { - DPRINTK(DEBUG_D_DELETE, "(%p): dropping negative dentry\n", - dentry); - return 1; - } - return 0; -} /* End Function devfs_d_delete */ - -struct devfs_lookup_struct { - devfs_handle_t de; - wait_queue_head_t wait_queue; -}; - -/* XXX: this doesn't handle the case where we got a negative dentry - but a devfs entry has been registered in the meanwhile */ -static int devfs_d_revalidate_wait(struct dentry *dentry, struct nameidata *nd) -{ - struct inode *dir = dentry->d_parent->d_inode; - struct fs_info *fs_info = dir->i_sb->s_fs_info; - devfs_handle_t parent = get_devfs_entry_from_vfs_inode(dir); - struct devfs_lookup_struct *lookup_info = dentry->d_fsdata; - DECLARE_WAITQUEUE(wait, current); - int need_lock; - - /* - * FIXME HACK - * - * make sure that - * d_instantiate always runs under lock - * we release i_mutex lock before going to sleep - * - * unfortunately sometimes d_revalidate is called with - * and sometimes without i_mutex lock held. The following checks - * attempt to deduce when we need to add (and drop resp.) lock - * here. This relies on current (2.6.2) calling coventions: - * - * lookup_hash is always run under i_mutex and is passing NULL - * as nd - * - * open(...,O_CREATE,...) calls _lookup_hash under i_mutex - * and sets flags to LOOKUP_OPEN|LOOKUP_CREATE - * - * all other invocations of ->d_revalidate seem to happen - * outside of i_mutex - */ - need_lock = nd && - (!(nd->flags & LOOKUP_CREATE) || (nd->flags & LOOKUP_PARENT)); - - if (need_lock) - mutex_lock(&dir->i_mutex); - - if (is_devfsd_or_child(fs_info)) { - devfs_handle_t de = lookup_info->de; - struct inode *inode; - - DPRINTK(DEBUG_I_LOOKUP, - "(%s): dentry: %p inode: %p de: %p by: \"%s\"\n", - dentry->d_name.name, dentry, dentry->d_inode, de, - current->comm); - if (dentry->d_inode) - goto out; - if (de == NULL) { - read_lock(&parent->u.dir.lock); - de = _devfs_search_dir(parent, dentry->d_name.name, - dentry->d_name.len); - read_unlock(&parent->u.dir.lock); - if (de == NULL) - goto out; - lookup_info->de = de; - } - /* Create an inode, now that the driver information is available */ - inode = _devfs_get_vfs_inode(dir->i_sb, de, dentry); - if (!inode) - goto out; - DPRINTK(DEBUG_I_LOOKUP, - "(%s): new VFS inode(%u): %p de: %p by: \"%s\"\n", - de->name, de->inode.ino, inode, de, current->comm); - d_instantiate(dentry, inode); - goto out; - } - if (lookup_info == NULL) - goto out; /* Early termination */ - read_lock(&parent->u.dir.lock); - if (dentry->d_fsdata) { - set_current_state(TASK_UNINTERRUPTIBLE); - add_wait_queue(&lookup_info->wait_queue, &wait); - read_unlock(&parent->u.dir.lock); - /* at this point it is always (hopefully) locked */ - mutex_unlock(&dir->i_mutex); - schedule(); - mutex_lock(&dir->i_mutex); - /* - * This does not need nor should remove wait from wait_queue. - * Wait queue head is never reused - nothing is ever added to it - * after all waiters have been waked up and head itself disappears - * very soon after it. Moreover it is local variable on stack that - * is likely to have already disappeared so any reference to it - * at this point is buggy. - */ - - } else - read_unlock(&parent->u.dir.lock); - - out: - if (need_lock) - mutex_unlock(&dir->i_mutex); - return 1; -} /* End Function devfs_d_revalidate_wait */ - -/* Inode operations for device entries follow */ - -static struct dentry *devfs_lookup(struct inode *dir, struct dentry *dentry, - struct nameidata *nd) -{ - struct devfs_entry tmp; /* Must stay in scope until devfsd idle again */ - struct devfs_lookup_struct lookup_info; - struct fs_info *fs_info = dir->i_sb->s_fs_info; - struct devfs_entry *parent, *de; - struct inode *inode; - struct dentry *retval = NULL; - - /* Set up the dentry operations before anything else, to ensure cleaning - up on any error */ - dentry->d_op = &devfs_dops; - /* First try to get the devfs entry for this directory */ - parent = get_devfs_entry_from_vfs_inode(dir); - DPRINTK(DEBUG_I_LOOKUP, "(%s): dentry: %p parent: %p by: \"%s\"\n", - dentry->d_name.name, dentry, parent, current->comm); - if (parent == NULL) - return ERR_PTR(-ENOENT); - read_lock(&parent->u.dir.lock); - de = _devfs_search_dir(parent, dentry->d_name.name, dentry->d_name.len); - read_unlock(&parent->u.dir.lock); - lookup_info.de = de; - init_waitqueue_head(&lookup_info.wait_queue); - dentry->d_fsdata = &lookup_info; - if (de == NULL) { /* Try with devfsd. For any kind of failure, leave a negative dentry - so someone else can deal with it (in the case where the sysadmin - does a mknod()). It's important to do this before hashing the - dentry, so that the devfsd queue is filled before revalidates - can start */ - if (try_modload(parent, fs_info, dentry->d_name.name, dentry->d_name.len, &tmp) < 0) { /* Lookup event was not queued to devfsd */ - d_add(dentry, NULL); - return NULL; - } - } - dentry->d_op = &devfs_wait_dops; - d_add(dentry, NULL); /* Open the floodgates */ - /* Unlock directory semaphore, which will release any waiters. They - will get the hashed dentry, and may be forced to wait for - revalidation */ - mutex_unlock(&dir->i_mutex); - wait_for_devfsd_finished(fs_info); /* If I'm not devfsd, must wait */ - mutex_lock(&dir->i_mutex); /* Grab it again because them's the rules */ - de = lookup_info.de; - /* If someone else has been so kind as to make the inode, we go home - early */ - if (dentry->d_inode) - goto out; - if (de == NULL) { - read_lock(&parent->u.dir.lock); - de = _devfs_search_dir(parent, dentry->d_name.name, - dentry->d_name.len); - read_unlock(&parent->u.dir.lock); - if (de == NULL) - goto out; - /* OK, there's an entry now, but no VFS inode yet */ - } - /* Create an inode, now that the driver information is available */ - inode = _devfs_get_vfs_inode(dir->i_sb, de, dentry); - if (!inode) { - retval = ERR_PTR(-ENOMEM); - goto out; - } - DPRINTK(DEBUG_I_LOOKUP, - "(%s): new VFS inode(%u): %p de: %p by: \"%s\"\n", de->name, - de->inode.ino, inode, de, current->comm); - d_instantiate(dentry, inode); - out: - write_lock(&parent->u.dir.lock); - dentry->d_op = &devfs_dops; - dentry->d_fsdata = NULL; - wake_up(&lookup_info.wait_queue); - write_unlock(&parent->u.dir.lock); - devfs_put(de); - return retval; -} /* End Function devfs_lookup */ - -static int devfs_unlink(struct inode *dir, struct dentry *dentry) -{ - int unhooked; - struct devfs_entry *de; - struct inode *inode = dentry->d_inode; - struct fs_info *fs_info = dir->i_sb->s_fs_info; - - de = get_devfs_entry_from_vfs_inode(inode); - DPRINTK(DEBUG_I_UNLINK, "(%s): de: %p\n", dentry->d_name.name, de); - if (de == NULL) - return -ENOENT; - if (!de->vfs) - return -EPERM; - write_lock(&de->parent->u.dir.lock); - unhooked = _devfs_unhook(de); - write_unlock(&de->parent->u.dir.lock); - if (!unhooked) - return -ENOENT; - if (!is_devfsd_or_child(fs_info)) - devfsd_notify_de(de, DEVFSD_NOTIFY_DELETE, inode->i_mode, - inode->i_uid, inode->i_gid, fs_info); - free_dentry(de); - devfs_put(de); - return 0; -} /* End Function devfs_unlink */ - -static int devfs_symlink(struct inode *dir, struct dentry *dentry, - const char *symname) -{ - int err; - struct fs_info *fs_info = dir->i_sb->s_fs_info; - struct devfs_entry *parent, *de; - struct inode *inode; - - /* First try to get the devfs entry for this directory */ - parent = get_devfs_entry_from_vfs_inode(dir); - if (parent == NULL) - return -ENOENT; - err = devfs_do_symlink(parent, dentry->d_name.name, symname, &de); - DPRINTK(DEBUG_DISABLED, "(%s): errcode from <devfs_do_symlink>: %d\n", - dentry->d_name.name, err); - if (err < 0) - return err; - de->vfs = TRUE; - de->inode.uid = current->euid; - de->inode.gid = current->egid; - de->inode.atime = CURRENT_TIME; - de->inode.mtime = CURRENT_TIME; - de->inode.ctime = CURRENT_TIME; - if ((inode = _devfs_get_vfs_inode(dir->i_sb, de, dentry)) == NULL) - return -ENOMEM; - DPRINTK(DEBUG_DISABLED, "(%s): new VFS inode(%u): %p dentry: %p\n", - dentry->d_name.name, de->inode.ino, inode, dentry); - d_instantiate(dentry, inode); - if (!is_devfsd_or_child(fs_info)) - devfsd_notify_de(de, DEVFSD_NOTIFY_CREATE, inode->i_mode, - inode->i_uid, inode->i_gid, fs_info); - return 0; -} /* End Function devfs_symlink */ - -static int devfs_mkdir(struct inode *dir, struct dentry *dentry, int mode) -{ - int err; - struct fs_info *fs_info = dir->i_sb->s_fs_info; - struct devfs_entry *parent, *de; - struct inode *inode; - - mode = (mode & ~S_IFMT) | S_IFDIR; /* VFS doesn't pass S_IFMT part */ - parent = get_devfs_entry_from_vfs_inode(dir); - if (parent == NULL) - return -ENOENT; - de = _devfs_alloc_entry(dentry->d_name.name, dentry->d_name.len, mode); - if (!de) - return -ENOMEM; - de->vfs = TRUE; - if ((err = _devfs_append_entry(parent, de, NULL)) != 0) - return err; - de->inode.uid = current->euid; - de->inode.gid = current->egid; - de->inode.atime = CURRENT_TIME; - de->inode.mtime = CURRENT_TIME; - de->inode.ctime = CURRENT_TIME; - if ((inode = _devfs_get_vfs_inode(dir->i_sb, de, dentry)) == NULL) - return -ENOMEM; - DPRINTK(DEBUG_DISABLED, "(%s): new VFS inode(%u): %p dentry: %p\n", - dentry->d_name.name, de->inode.ino, inode, dentry); - d_instantiate(dentry, inode); - if (!is_devfsd_or_child(fs_info)) - devfsd_notify_de(de, DEVFSD_NOTIFY_CREATE, inode->i_mode, - inode->i_uid, inode->i_gid, fs_info); - return 0; -} /* End Function devfs_mkdir */ - -static int devfs_rmdir(struct inode *dir, struct dentry *dentry) -{ - int err = 0; - struct devfs_entry *de; - struct fs_info *fs_info = dir->i_sb->s_fs_info; - struct inode *inode = dentry->d_inode; - - if (dir->i_sb->s_fs_info != inode->i_sb->s_fs_info) - return -EINVAL; - de = get_devfs_entry_from_vfs_inode(inode); - if (de == NULL) - return -ENOENT; - if (!S_ISDIR(de->mode)) - return -ENOTDIR; - if (!de->vfs) - return -EPERM; - /* First ensure the directory is empty and will stay that way */ - write_lock(&de->u.dir.lock); - if (de->u.dir.first) - err = -ENOTEMPTY; - else - de->u.dir.no_more_additions = TRUE; - write_unlock(&de->u.dir.lock); - if (err) - return err; - /* Now unhook the directory from its parent */ - write_lock(&de->parent->u.dir.lock); - if (!_devfs_unhook(de)) - err = -ENOENT; - write_unlock(&de->parent->u.dir.lock); - if (err) - return err; - if (!is_devfsd_or_child(fs_info)) - devfsd_notify_de(de, DEVFSD_NOTIFY_DELETE, inode->i_mode, - inode->i_uid, inode->i_gid, fs_info); - free_dentry(de); - devfs_put(de); - return 0; -} /* End Function devfs_rmdir */ - -static int devfs_mknod(struct inode *dir, struct dentry *dentry, int mode, - dev_t rdev) -{ - int err; - struct fs_info *fs_info = dir->i_sb->s_fs_info; - struct devfs_entry *parent, *de; - struct inode *inode; - - DPRINTK(DEBUG_I_MKNOD, "(%s): mode: 0%o dev: %u:%u\n", - dentry->d_name.name, mode, MAJOR(rdev), MINOR(rdev)); - parent = get_devfs_entry_from_vfs_inode(dir); - if (parent == NULL) - return -ENOENT; - de = _devfs_alloc_entry(dentry->d_name.name, dentry->d_name.len, mode); - if (!de) - return -ENOMEM; - de->vfs = TRUE; - if (S_ISCHR(mode) || S_ISBLK(mode)) - de->u.dev = rdev; - if ((err = _devfs_append_entry(parent, de, NULL)) != 0) - return err; - de->inode.uid = current->euid; - de->inode.gid = current->egid; - de->inode.atime = CURRENT_TIME; - de->inode.mtime = CURRENT_TIME; - de->inode.ctime = CURRENT_TIME; - if ((inode = _devfs_get_vfs_inode(dir->i_sb, de, dentry)) == NULL) - return -ENOMEM; - DPRINTK(DEBUG_I_MKNOD, ": new VFS inode(%u): %p dentry: %p\n", - de->inode.ino, inode, dentry); - d_instantiate(dentry, inode); - if (!is_devfsd_or_child(fs_info)) - devfsd_notify_de(de, DEVFSD_NOTIFY_CREATE, inode->i_mode, - inode->i_uid, inode->i_gid, fs_info); - return 0; -} /* End Function devfs_mknod */ - -static void *devfs_follow_link(struct dentry *dentry, struct nameidata *nd) -{ - struct devfs_entry *p = get_devfs_entry_from_vfs_inode(dentry->d_inode); - nd_set_link(nd, p ? p->u.symlink.linkname : ERR_PTR(-ENODEV)); - return NULL; -} /* End Function devfs_follow_link */ - -static struct inode_operations devfs_iops = { - .setattr = devfs_notify_change, -}; - -static struct inode_operations devfs_dir_iops = { - .lookup = devfs_lookup, - .unlink = devfs_unlink, - .symlink = devfs_symlink, - .mkdir = devfs_mkdir, - .rmdir = devfs_rmdir, - .mknod = devfs_mknod, - .setattr = devfs_notify_change, -}; - -static struct inode_operations devfs_symlink_iops = { - .readlink = generic_readlink, - .follow_link = devfs_follow_link, - .setattr = devfs_notify_change, -}; - -static int devfs_fill_super(struct super_block *sb, void *data, int silent) -{ - struct inode *root_inode = NULL; - - if (_devfs_get_root_entry() == NULL) - goto out_no_root; - atomic_set(&fs_info.devfsd_overrun_count, 0); - init_waitqueue_head(&fs_info.devfsd_wait_queue); - init_waitqueue_head(&fs_info.revalidate_wait_queue); - fs_info.sb = sb; - sb->s_fs_info = &fs_info; - sb->s_blocksize = 1024; - sb->s_blocksize_bits = 10; - sb->s_magic = DEVFS_SUPER_MAGIC; - sb->s_op = &devfs_sops; - sb->s_time_gran = 1; - if ((root_inode = _devfs_get_vfs_inode(sb, root_entry, NULL)) == NULL) - goto out_no_root; - sb->s_root = d_alloc_root(root_inode); - if (!sb->s_root) - goto out_no_root; - DPRINTK(DEBUG_S_READ, "(): made devfs ptr: %p\n", sb->s_fs_info); - return 0; - - out_no_root: - PRINTK("(): get root inode failed\n"); - if (root_inode) - iput(root_inode); - return -EINVAL; -} /* End Function devfs_fill_super */ - -static int devfs_get_sb(struct file_system_type *fs_type, - int flags, const char *dev_name, - void *data, struct vfsmount *mnt) -{ - return get_sb_single(fs_type, flags, data, devfs_fill_super, mnt); -} - -static struct file_system_type devfs_fs_type = { - .name = DEVFS_NAME, - .get_sb = devfs_get_sb, - .kill_sb = kill_anon_super, -}; - -/* File operations for devfsd follow */ - -static ssize_t devfsd_read(struct file *file, char __user *buf, size_t len, - loff_t * ppos) -{ - int done = FALSE; - int ival; - loff_t pos, devname_offset, tlen, rpos; - devfs_handle_t de; - struct devfsd_buf_entry *entry; - struct fs_info *fs_info = file->f_dentry->d_inode->i_sb->s_fs_info; - struct devfsd_notify_struct *info = fs_info->devfsd_info; - DECLARE_WAITQUEUE(wait, current); - - /* Verify the task has grabbed the queue */ - if (fs_info->devfsd_task != current) - return -EPERM; - info->major = 0; - info->minor = 0; - /* Block for a new entry */ - set_current_state(TASK_INTERRUPTIBLE); - add_wait_queue(&fs_info->devfsd_wait_queue, &wait); - while (devfsd_queue_empty(fs_info)) { - fs_info->devfsd_sleeping = TRUE; - wake_up(&fs_info->revalidate_wait_queue); - schedule(); - fs_info->devfsd_sleeping = FALSE; - if (signal_pending(current)) { - remove_wait_queue(&fs_info->devfsd_wait_queue, &wait); - __set_current_state(TASK_RUNNING); - return -EINTR; - } - set_current_state(TASK_INTERRUPTIBLE); - } - remove_wait_queue(&fs_info->devfsd_wait_queue, &wait); - __set_current_state(TASK_RUNNING); - /* Now play with the data */ - ival = atomic_read(&fs_info->devfsd_overrun_count); - info->overrun_count = ival; - entry = fs_info->devfsd_first_event; - info->type = entry->type; - info->mode = entry->mode; - info->uid = entry->uid; - info->gid = entry->gid; - de = entry->de; - if (S_ISCHR(de->mode) || S_ISBLK(de->mode)) { - info->major = MAJOR(de->u.dev); - info->minor = MINOR(de->u.dev); - } - pos = devfs_generate_path(de, info->devname, DEVFS_PATHLEN); - if (pos < 0) - return pos; - info->namelen = DEVFS_PATHLEN - pos - 1; - if (info->mode == 0) - info->mode = de->mode; - devname_offset = info->devname - (char *)info; - rpos = *ppos; - if (rpos < devname_offset) { - /* Copy parts of the header */ - tlen = devname_offset - rpos; - if (tlen > len) - tlen = len; - if (copy_to_user(buf, (char *)info + rpos, tlen)) { - return -EFAULT; - } - rpos += tlen; - buf += tlen; - len -= tlen; - } - if ((rpos >= devname_offset) && (len > 0)) { - /* Copy the name */ - tlen = info->namelen + 1; - if (tlen > len) - tlen = len; - else - done = TRUE; - if (copy_to_user - (buf, info->devname + pos + rpos - devname_offset, tlen)) { - return -EFAULT; - } - rpos += tlen; - } - tlen = rpos - *ppos; - if (done) { - devfs_handle_t parent; - - spin_lock(&fs_info->devfsd_buffer_lock); - fs_info->devfsd_first_event = entry->next; - if (entry->next == NULL) - fs_info->devfsd_last_event = NULL; - spin_unlock(&fs_info->devfsd_buffer_lock); - for (; de != NULL; de = parent) { - parent = de->parent; - devfs_put(de); - } - kmem_cache_free(devfsd_buf_cache, entry); - if (ival > 0) - atomic_sub(ival, &fs_info->devfsd_overrun_count); - *ppos = 0; - } else - *ppos = rpos; - return tlen; -} /* End Function devfsd_read */ - -static int devfsd_ioctl(struct inode *inode, struct file *file, - unsigned int cmd, unsigned long arg) -{ - int ival; - struct fs_info *fs_info = inode->i_sb->s_fs_info; - - switch (cmd) { - case DEVFSDIOC_GET_PROTO_REV: - ival = DEVFSD_PROTOCOL_REVISION_KERNEL; - if (copy_to_user((void __user *)arg, &ival, sizeof ival)) - return -EFAULT; - break; - case DEVFSDIOC_SET_EVENT_MASK: - /* Ensure only one reader has access to the queue. This scheme will - work even if the global kernel lock were to be removed, because it - doesn't matter who gets in first, as long as only one gets it */ - if (fs_info->devfsd_task == NULL) { - static DEFINE_SPINLOCK(lock); - - if (!spin_trylock(&lock)) - return -EBUSY; - if (fs_info->devfsd_task != NULL) { /* We lost the race... */ - spin_unlock(&lock); - return -EBUSY; - } - fs_info->devfsd_task = current; - spin_unlock(&lock); - fs_info->devfsd_pgrp = - (process_group(current) == - current->pid) ? process_group(current) : 0; - fs_info->devfsd_file = file; - fs_info->devfsd_info = - kmalloc(sizeof *fs_info->devfsd_info, GFP_KERNEL); - if (!fs_info->devfsd_info) { - devfsd_close(inode, file); - return -ENOMEM; - } - } else if (fs_info->devfsd_task != current) - return -EBUSY; - fs_info->devfsd_event_mask = arg; /* Let the masses come forth */ - break; - case DEVFSDIOC_RELEASE_EVENT_QUEUE: - if (fs_info->devfsd_file != file) - return -EPERM; - return devfsd_close(inode, file); - /*break; */ -#ifdef CONFIG_DEVFS_DEBUG - case DEVFSDIOC_SET_DEBUG_MASK: - if (copy_from_user(&ival, (void __user *)arg, sizeof ival)) - return -EFAULT; - devfs_debug = ival; - break; -#endif - default: - return -ENOIOCTLCMD; - } - return 0; -} /* End Function devfsd_ioctl */ - -static int devfsd_close(struct inode *inode, struct file *file) -{ - struct devfsd_buf_entry *entry, *next; - struct fs_info *fs_info = inode->i_sb->s_fs_info; - - if (fs_info->devfsd_file != file) - return 0; - fs_info->devfsd_event_mask = 0; - fs_info->devfsd_file = NULL; - spin_lock(&fs_info->devfsd_buffer_lock); - entry = fs_info->devfsd_first_event; - fs_info->devfsd_first_event = NULL; - fs_info->devfsd_last_event = NULL; - kfree(fs_info->devfsd_info); - fs_info->devfsd_info = NULL; - spin_unlock(&fs_info->devfsd_buffer_lock); - fs_info->devfsd_pgrp = 0; - fs_info->devfsd_task = NULL; - wake_up(&fs_info->revalidate_wait_queue); - for (; entry; entry = next) { - next = entry->next; - kmem_cache_free(devfsd_buf_cache, entry); - } - return 0; -} /* End Function devfsd_close */ - -#ifdef CONFIG_DEVFS_DEBUG -static ssize_t stat_read(struct file *file, char __user *buf, size_t len, - loff_t * ppos) -{ - ssize_t num; - char txt[80]; - - num = sprintf(txt, "Number of entries: %u number of bytes: %u\n", - stat_num_entries, stat_num_bytes) + 1; - if (*ppos >= num) - return 0; - if (*ppos + len > num) - len = num - *ppos; - if (copy_to_user(buf, txt + *ppos, len)) - return -EFAULT; - *ppos += len; - return len; -} /* End Function stat_read */ -#endif - -static int __init init_devfs_fs(void) -{ - int err; - int major; - struct devfs_entry *devfsd; -#ifdef CONFIG_DEVFS_DEBUG - struct devfs_entry *stat; -#endif - - if (_devfs_get_root_entry() == NULL) - return -ENOMEM; - - printk(KERN_INFO "%s: %s Richard Gooch (rgooch@atnf.csiro.au)\n", - DEVFS_NAME, DEVFS_VERSION); - devfsd_buf_cache = kmem_cache_create("devfsd_event", - sizeof(struct devfsd_buf_entry), - 0, 0, NULL, NULL); - if (!devfsd_buf_cache) - OOPS("(): unable to allocate event slab\n"); -#ifdef CONFIG_DEVFS_DEBUG - devfs_debug = devfs_debug_init; - printk(KERN_INFO "%s: devfs_debug: 0x%0x\n", DEVFS_NAME, devfs_debug); -#endif - printk(KERN_INFO "%s: boot_options: 0x%0x\n", DEVFS_NAME, boot_options); - - /* register special device for devfsd communication */ - major = register_chrdev(0, "devfs", &devfs_fops); - if (major < 0) - return major; - - /* And create the entry for ".devfsd" */ - devfsd = _devfs_alloc_entry(".devfsd", 0, S_IFCHR | S_IRUSR | S_IWUSR); - if (devfsd == NULL) - return -ENOMEM; - devfsd->u.dev = MKDEV(major, 0); - _devfs_append_entry(root_entry, devfsd, NULL); - -#ifdef CONFIG_DEVFS_DEBUG - stat = _devfs_alloc_entry(".stat", 0, S_IFCHR | S_IRUGO); - if (stat == NULL) - return -ENOMEM; - stat->u.dev = MKDEV(major, 1); - _devfs_append_entry(root_entry, stat, NULL); -#endif - - err = register_filesystem(&devfs_fs_type); - return err; -} /* End Function init_devfs_fs */ - -void __init mount_devfs_fs(void) -{ - int err; - - if (!(boot_options & OPTION_MOUNT)) - return; - err = do_mount("none", "/dev", "devfs", 0, NULL); - if (err == 0) - printk(KERN_INFO "Mounted devfs on /dev\n"); - else - PRINTK("(): unable to mount devfs, err: %d\n", err); -} /* End Function mount_devfs_fs */ - -module_init(init_devfs_fs) diff --git a/fs/devfs/util.c b/fs/devfs/util.c deleted file mode 100644 index db06d388c9a..00000000000 --- a/fs/devfs/util.c +++ /dev/null @@ -1,97 +0,0 @@ -/* devfs (Device FileSystem) utilities. - - Copyright (C) 1999-2002 Richard Gooch - - This library is free software; you can redistribute it and/or - modify it under the terms of the GNU Library General Public - License as published by the Free Software Foundation; either - version 2 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 - Library General Public License for more details. - - You should have received a copy of the GNU Library General Public - License along with this library; if not, write to the Free - Software Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. - - Richard Gooch may be reached by email at rgooch@atnf.csiro.au - The postal address is: - Richard Gooch, c/o ATNF, P. O. Box 76, Epping, N.S.W., 2121, Australia. - - ChangeLog - - 19991031 Richard Gooch <rgooch@atnf.csiro.au> - Created. - 19991103 Richard Gooch <rgooch@atnf.csiro.au> - Created <_devfs_convert_name> and supported SCSI and IDE CD-ROMs - 20000203 Richard Gooch <rgooch@atnf.csiro.au> - Changed operations pointer type to void *. - 20000621 Richard Gooch <rgooch@atnf.csiro.au> - Changed interface to <devfs_register_series>. - 20000622 Richard Gooch <rgooch@atnf.csiro.au> - Took account of interface change to <devfs_mk_symlink>. - Took account of interface change to <devfs_mk_dir>. - 20010519 Richard Gooch <rgooch@atnf.csiro.au> - Documentation cleanup. - 20010709 Richard Gooch <rgooch@atnf.csiro.au> - Created <devfs_*alloc_major> and <devfs_*alloc_devnum>. - 20010710 Richard Gooch <rgooch@atnf.csiro.au> - Created <devfs_*alloc_unique_number>. - 20010730 Richard Gooch <rgooch@atnf.csiro.au> - Documentation typo fix. - 20010806 Richard Gooch <rgooch@atnf.csiro.au> - Made <block_semaphore> and <char_semaphore> private. - 20010813 Richard Gooch <rgooch@atnf.csiro.au> - Fixed bug in <devfs_alloc_unique_number>: limited to 128 numbers - 20010818 Richard Gooch <rgooch@atnf.csiro.au> - Updated major masks up to Linus' "no new majors" proclamation. - Block: were 126 now 122 free, char: were 26 now 19 free. - 20020324 Richard Gooch <rgooch@atnf.csiro.au> - Fixed bug in <devfs_alloc_unique_number>: was clearing beyond - bitfield. - 20020326 Richard Gooch <rgooch@atnf.csiro.au> - Fixed bitfield data type for <devfs_*alloc_devnum>. - Made major bitfield type and initialiser 64 bit safe. - 20020413 Richard Gooch <rgooch@atnf.csiro.au> - Fixed shift warning on 64 bit machines. - 20020428 Richard Gooch <rgooch@atnf.csiro.au> - Copied and used macro for error messages from fs/devfs/base.c - 20021013 Richard Gooch <rgooch@atnf.csiro.au> - Documentation fix. - 20030101 Adam J. Richter <adam@yggdrasil.com> - Eliminate DEVFS_SPECIAL_{CHR,BLK}. Use mode_t instead. - 20030106 Christoph Hellwig <hch@infradead.org> - Rewrite devfs_{,de}alloc_devnum to look like C code. -*/ -#include <linux/module.h> -#include <linux/init.h> -#include <linux/devfs_fs_kernel.h> -#include <linux/slab.h> -#include <linux/vmalloc.h> -#include <linux/genhd.h> -#include <linux/bitops.h> - -int devfs_register_tape(const char *name) -{ - char tname[32], dest[64]; - static unsigned int tape_counter; - unsigned int n = tape_counter++; - - sprintf(dest, "../%s", name); - sprintf(tname, "tapes/tape%u", n); - devfs_mk_symlink(tname, dest); - - return n; -} - -EXPORT_SYMBOL(devfs_register_tape); - -void devfs_unregister_tape(int num) -{ - if (num >= 0) - devfs_remove("tapes/tape%u", num); -} - -EXPORT_SYMBOL(devfs_unregister_tape); diff --git a/fs/devpts/inode.c b/fs/devpts/inode.c index f7aef5bb584..5f7b5a6025b 100644 --- a/fs/devpts/inode.c +++ b/fs/devpts/inode.c @@ -113,7 +113,6 @@ devpts_fill_super(struct super_block *s, void *data, int silent) inode->i_ino = 1; inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME; inode->i_blocks = 0; - inode->i_blksize = 1024; inode->i_uid = inode->i_gid = 0; inode->i_mode = S_IFDIR | S_IRUGO | S_IXUGO | S_IWUSR; inode->i_op = &simple_dir_inode_operations; @@ -172,12 +171,11 @@ int devpts_pty_new(struct tty_struct *tty) return -ENOMEM; inode->i_ino = number+2; - inode->i_blksize = 1024; inode->i_uid = config.setuid ? config.uid : current->fsuid; inode->i_gid = config.setgid ? config.gid : current->fsgid; inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME; init_special_inode(inode, S_IFCHR|config.mode, device); - inode->u.generic_ip = tty; + inode->i_private = tty; dentry = get_node(number); if (!IS_ERR(dentry) && !dentry->d_inode) @@ -196,7 +194,7 @@ struct tty_struct *devpts_get_tty(int number) tty = NULL; if (!IS_ERR(dentry)) { if (dentry->d_inode) - tty = dentry->d_inode->u.generic_ip; + tty = dentry->d_inode->i_private; dput(dentry); } diff --git a/fs/direct-io.c b/fs/direct-io.c index 538fb0418fb..5981e17f46f 100644 --- a/fs/direct-io.c +++ b/fs/direct-io.c @@ -220,7 +220,8 @@ static void dio_complete(struct dio *dio, loff_t offset, ssize_t bytes) if (dio->end_io && dio->result) dio->end_io(dio->iocb, offset, bytes, dio->map_bh.b_private); if (dio->lock_type == DIO_LOCKING) - up_read(&dio->inode->i_alloc_sem); + /* lockdep: non-owner release */ + up_read_non_owner(&dio->inode->i_alloc_sem); } /* @@ -1261,7 +1262,8 @@ __blockdev_direct_IO(int rw, struct kiocb *iocb, struct inode *inode, } if (dio_lock_type == DIO_LOCKING) - down_read(&inode->i_alloc_sem); + /* lockdep: not the owner will release it */ + down_read_non_owner(&inode->i_alloc_sem); } /* diff --git a/fs/dquot.c b/fs/dquot.c index 81d87a413c6..9af789567e5 100644 --- a/fs/dquot.c +++ b/fs/dquot.c @@ -250,7 +250,7 @@ static inline struct dquot *find_dquot(unsigned int hashent, struct super_block /* Add a dquot to the tail of the free list */ static inline void put_dquot_last(struct dquot *dquot) { - list_add(&dquot->dq_free, free_dquots.prev); + list_add_tail(&dquot->dq_free, &free_dquots); dqstats.free_dquots++; } @@ -266,7 +266,7 @@ static inline void put_inuse(struct dquot *dquot) { /* We add to the back of inuse list so we don't have to restart * when traversing this list and we block */ - list_add(&dquot->dq_inuse, inuse_list.prev); + list_add_tail(&dquot->dq_inuse, &inuse_list); dqstats.allocated_dquots++; } @@ -834,6 +834,9 @@ static void print_warning(struct dquot *dquot, const char warntype) if (!need_print_warning(dquot) || (flag && test_and_set_bit(flag, &dquot->dq_flags))) return; + mutex_lock(&tty_mutex); + if (!current->signal->tty) + goto out_lock; tty_write_message(current->signal->tty, dquot->dq_sb->s_id); if (warntype == ISOFTWARN || warntype == BSOFTWARN) tty_write_message(current->signal->tty, ": warning, "); @@ -861,6 +864,8 @@ static void print_warning(struct dquot *dquot, const char warntype) break; } tty_write_message(current->signal->tty, msg); +out_lock: + mutex_unlock(&tty_mutex); } static inline void flush_warnings(struct dquot **dquots, char *warntype) diff --git a/fs/efs/inode.c b/fs/efs/inode.c index 180607f9314..174696f9bf1 100644 --- a/fs/efs/inode.c +++ b/fs/efs/inode.c @@ -21,7 +21,7 @@ static sector_t _efs_bmap(struct address_space *mapping, sector_t block) { return generic_block_bmap(mapping,block,efs_get_block); } -static struct address_space_operations efs_aops = { +static const struct address_space_operations efs_aops = { .readpage = efs_readpage, .sync_page = block_sync_page, .bmap = _efs_bmap diff --git a/fs/efs/super.c b/fs/efs/super.c index 8ac2462ae5d..b3f50651eb6 100644 --- a/fs/efs/super.c +++ b/fs/efs/super.c @@ -90,8 +90,7 @@ static int init_inodecache(void) static void destroy_inodecache(void) { - if (kmem_cache_destroy(efs_inode_cachep)) - printk(KERN_INFO "efs_inode_cache: not all structures were freed\n"); + kmem_cache_destroy(efs_inode_cachep); } static void efs_put_super(struct super_block *s) @@ -248,11 +247,10 @@ static int efs_fill_super(struct super_block *s, void *d, int silent) struct buffer_head *bh; struct inode *root; - sb = kmalloc(sizeof(struct efs_sb_info), GFP_KERNEL); + sb = kzalloc(sizeof(struct efs_sb_info), GFP_KERNEL); if (!sb) return -ENOMEM; s->s_fs_info = sb; - memset(sb, 0, sizeof(struct efs_sb_info)); s->s_magic = EFS_SUPER_MAGIC; if (!sb_set_blocksize(s, EFS_BLOCKSIZE)) { diff --git a/fs/efs/symlink.c b/fs/efs/symlink.c index 3d9a350e3e7..1d30d2ff440 100644 --- a/fs/efs/symlink.c +++ b/fs/efs/symlink.c @@ -22,7 +22,7 @@ static int efs_symlink_readpage(struct file *file, struct page *page) err = -ENAMETOOLONG; if (size > 2 * EFS_BLOCKSIZE) - goto fail; + goto fail_notlocked; lock_kernel(); /* read first 512 bytes of link target */ @@ -47,12 +47,13 @@ static int efs_symlink_readpage(struct file *file, struct page *page) return 0; fail: unlock_kernel(); +fail_notlocked: SetPageError(page); kunmap(page); unlock_page(page); return err; } -struct address_space_operations efs_symlink_aops = { +const struct address_space_operations efs_symlink_aops = { .readpage = efs_symlink_readpage }; diff --git a/fs/eventpoll.c b/fs/eventpoll.c index 08e7e6a555c..8d544334bcd 100644 --- a/fs/eventpoll.c +++ b/fs/eventpoll.c @@ -1,6 +1,6 @@ /* * fs/eventpoll.c ( Efficent event polling implementation ) - * Copyright (C) 2001,...,2003 Davide Libenzi + * Copyright (C) 2001,...,2006 Davide Libenzi * * 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 @@ -120,7 +120,7 @@ struct epoll_filefd { */ struct wake_task_node { struct list_head llink; - task_t *task; + struct task_struct *task; wait_queue_head_t *wq; }; @@ -413,7 +413,7 @@ static void ep_poll_safewake(struct poll_safewake *psw, wait_queue_head_t *wq) { int wake_nests = 0; unsigned long flags; - task_t *this_task = current; + struct task_struct *this_task = current; struct list_head *lsthead = &psw->wake_task_list, *lnk; struct wake_task_node *tncur; struct wake_task_node tnode; @@ -1004,7 +1004,7 @@ static int ep_insert(struct eventpoll *ep, struct epoll_event *event, /* Notify waiting tasks that events are available */ if (waitqueue_active(&ep->wq)) - wake_up(&ep->wq); + __wake_up_locked(&ep->wq, TASK_UNINTERRUPTIBLE | TASK_INTERRUPTIBLE); if (waitqueue_active(&ep->poll_wait)) pwake++; } @@ -1083,7 +1083,8 @@ static int ep_modify(struct eventpoll *ep, struct epitem *epi, struct epoll_even /* Notify waiting tasks that events are available */ if (waitqueue_active(&ep->wq)) - wake_up(&ep->wq); + __wake_up_locked(&ep->wq, TASK_UNINTERRUPTIBLE | + TASK_INTERRUPTIBLE); if (waitqueue_active(&ep->poll_wait)) pwake++; } @@ -1167,7 +1168,7 @@ static int ep_unlink(struct eventpoll *ep, struct epitem *epi) eexit_1: DNPRINTK(3, (KERN_INFO "[%p] eventpoll: ep_unlink(%p, %p) = %d\n", - current, ep, epi->file, error)); + current, ep, epi->ffd.file, error)); return error; } @@ -1235,7 +1236,7 @@ static int ep_poll_callback(wait_queue_t *wait, unsigned mode, int sync, void *k struct eventpoll *ep = epi->ep; DNPRINTK(3, (KERN_INFO "[%p] eventpoll: poll_callback(%p) epi=%p ep=%p\n", - current, epi->file, epi, ep)); + current, epi->ffd.file, epi, ep)); write_lock_irqsave(&ep->lock, flags); @@ -1260,7 +1261,8 @@ is_linked: * wait list. */ if (waitqueue_active(&ep->wq)) - wake_up(&ep->wq); + __wake_up_locked(&ep->wq, TASK_UNINTERRUPTIBLE | + TASK_INTERRUPTIBLE); if (waitqueue_active(&ep->poll_wait)) pwake++; @@ -1444,7 +1446,8 @@ static void ep_reinject_items(struct eventpoll *ep, struct list_head *txlist) * wait list. */ if (waitqueue_active(&ep->wq)) - wake_up(&ep->wq); + __wake_up_locked(&ep->wq, TASK_UNINTERRUPTIBLE | + TASK_INTERRUPTIBLE); if (waitqueue_active(&ep->poll_wait)) pwake++; } @@ -1516,7 +1519,7 @@ retry: * ep_poll_callback() when events will become available. */ init_waitqueue_entry(&wait, current); - add_wait_queue(&ep->wq, &wait); + __add_wait_queue(&ep->wq, &wait); for (;;) { /* @@ -1536,7 +1539,7 @@ retry: jtimeout = schedule_timeout(jtimeout); write_lock_irqsave(&ep->lock, flags); } - remove_wait_queue(&ep->wq, &wait); + __remove_wait_queue(&ep->wq, &wait); set_current_state(TASK_RUNNING); } @@ -1587,7 +1590,6 @@ static struct inode *ep_eventpoll_inode(void) inode->i_uid = current->fsuid; inode->i_gid = current->fsgid; inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME; - inode->i_blksize = PAGE_SIZE; return inode; eexit_1: diff --git a/fs/exec.c b/fs/exec.c index 0b88bf64614..a8efe35176b 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -22,7 +22,6 @@ * formats. */ -#include <linux/config.h> #include <linux/slab.h> #include <linux/file.h> #include <linux/mman.h> @@ -487,8 +486,6 @@ struct file *open_exec(const char *name) if (!(nd.mnt->mnt_flags & MNT_NOEXEC) && S_ISREG(inode->i_mode)) { int err = vfs_permission(&nd, MAY_EXEC); - if (!err && !(inode->i_mode & 0111)) - err = -EACCES; file = ERR_PTR(err); if (!err) { file = nameidata_to_filp(&nd, O_RDONLY); @@ -598,7 +595,7 @@ static int de_thread(struct task_struct *tsk) if (!newsighand) return -ENOMEM; - if (thread_group_empty(current)) + if (thread_group_empty(tsk)) goto no_thread_group; /* @@ -623,17 +620,17 @@ static int de_thread(struct task_struct *tsk) * Reparenting needs write_lock on tasklist_lock, * so it is safe to do it under read_lock. */ - if (unlikely(current->group_leader == child_reaper)) - child_reaper = current; + if (unlikely(tsk->group_leader == child_reaper)) + child_reaper = tsk; - zap_other_threads(current); + zap_other_threads(tsk); read_unlock(&tasklist_lock); /* * Account for the thread group leader hanging around: */ count = 1; - if (!thread_group_leader(current)) { + if (!thread_group_leader(tsk)) { count = 2; /* * The SIGALRM timer survives the exec, but needs to point @@ -642,14 +639,14 @@ static int de_thread(struct task_struct *tsk) * synchronize with any firing (by calling del_timer_sync) * before we can safely let the old group leader die. */ - sig->tsk = current; + sig->tsk = tsk; spin_unlock_irq(lock); if (hrtimer_cancel(&sig->real_timer)) hrtimer_restart(&sig->real_timer); spin_lock_irq(lock); } while (atomic_read(&sig->count) > count) { - sig->group_exit_task = current; + sig->group_exit_task = tsk; sig->notify_count = count; __set_current_state(TASK_UNINTERRUPTIBLE); spin_unlock_irq(lock); @@ -665,15 +662,13 @@ static int de_thread(struct task_struct *tsk) * do is to wait for the thread group leader to become inactive, * and to assume its PID: */ - if (!thread_group_leader(current)) { - struct dentry *proc_dentry1, *proc_dentry2; - + if (!thread_group_leader(tsk)) { /* * Wait for the thread group leader to be a zombie. * It should already be zombie at this point, most * of the time. */ - leader = current->group_leader; + leader = tsk->group_leader; while (leader->exit_state != EXIT_ZOMBIE) yield(); @@ -687,16 +682,12 @@ static int de_thread(struct task_struct *tsk) * When we take on its identity by switching to its PID, we * also take its birthdate (always earlier than our own). */ - current->start_time = leader->start_time; + tsk->start_time = leader->start_time; - spin_lock(&leader->proc_lock); - spin_lock(¤t->proc_lock); - proc_dentry1 = proc_pid_unhash(current); - proc_dentry2 = proc_pid_unhash(leader); write_lock_irq(&tasklist_lock); - BUG_ON(leader->tgid != current->tgid); - BUG_ON(current->pid == current->tgid); + BUG_ON(leader->tgid != tsk->tgid); + BUG_ON(tsk->pid == tsk->tgid); /* * An exec() starts a new thread group with the * TGID of the previous thread group. Rehash the @@ -705,34 +696,26 @@ static int de_thread(struct task_struct *tsk) */ /* Become a process group leader with the old leader's pid. - * Note: The old leader also uses thispid until release_task + * The old leader becomes a thread of the this thread group. + * Note: The old leader also uses this pid until release_task * is called. Odd but simple and correct. */ - detach_pid(current, PIDTYPE_PID); - current->pid = leader->pid; - attach_pid(current, PIDTYPE_PID, current->pid); - attach_pid(current, PIDTYPE_PGID, current->signal->pgrp); - attach_pid(current, PIDTYPE_SID, current->signal->session); - list_add_tail_rcu(¤t->tasks, &init_task.tasks); - - current->group_leader = current; - leader->group_leader = current; + detach_pid(tsk, PIDTYPE_PID); + tsk->pid = leader->pid; + attach_pid(tsk, PIDTYPE_PID, tsk->pid); + transfer_pid(leader, tsk, PIDTYPE_PGID); + transfer_pid(leader, tsk, PIDTYPE_SID); + list_replace_rcu(&leader->tasks, &tsk->tasks); - /* Reduce leader to a thread */ - detach_pid(leader, PIDTYPE_PGID); - detach_pid(leader, PIDTYPE_SID); - list_del_init(&leader->tasks); + tsk->group_leader = tsk; + leader->group_leader = tsk; - current->exit_signal = SIGCHLD; + tsk->exit_signal = SIGCHLD; BUG_ON(leader->exit_state != EXIT_ZOMBIE); leader->exit_state = EXIT_DEAD; write_unlock_irq(&tasklist_lock); - spin_unlock(&leader->proc_lock); - spin_unlock(¤t->proc_lock); - proc_pid_flush(proc_dentry1); - proc_pid_flush(proc_dentry2); } /* @@ -765,9 +748,9 @@ no_thread_group: write_lock_irq(&tasklist_lock); spin_lock(&oldsighand->siglock); - spin_lock(&newsighand->siglock); + spin_lock_nested(&newsighand->siglock, SINGLE_DEPTH_NESTING); - rcu_assign_pointer(current->sighand, newsighand); + rcu_assign_pointer(tsk->sighand, newsighand); recalc_sigpending(); spin_unlock(&newsighand->siglock); @@ -778,7 +761,7 @@ no_thread_group: kmem_cache_free(sighand_cachep, oldsighand); } - BUG_ON(!thread_group_leader(current)); + BUG_ON(!thread_group_leader(tsk)); return 0; } @@ -915,8 +898,7 @@ int flush_old_exec(struct linux_binprm * bprm) return 0; mmap_failed: - put_files_struct(current->files); - current->files = files; + reset_files_struct(current, files); out: return retval; } @@ -934,12 +916,6 @@ int prepare_binprm(struct linux_binprm *bprm) int retval; mode = inode->i_mode; - /* - * Check execute perms again - if the caller has CAP_DAC_OVERRIDE, - * generic_permission lets a non-executable through - */ - if (!(mode & 0111)) /* with at least _one_ execute bit set */ - return -EACCES; if (bprm->file->f_op == NULL) return -EACCES; @@ -1379,67 +1355,102 @@ static void format_corename(char *corename, const char *pattern, long signr) *out_ptr = 0; } -static void zap_threads (struct mm_struct *mm) +static void zap_process(struct task_struct *start) { - struct task_struct *g, *p; - struct task_struct *tsk = current; - struct completion *vfork_done = tsk->vfork_done; - int traced = 0; + struct task_struct *t; - /* - * Make sure nobody is waiting for us to release the VM, - * otherwise we can deadlock when we wait on each other - */ - if (vfork_done) { - tsk->vfork_done = NULL; - complete(vfork_done); - } + start->signal->flags = SIGNAL_GROUP_EXIT; + start->signal->group_stop_count = 0; - read_lock(&tasklist_lock); - do_each_thread(g,p) - if (mm == p->mm && p != tsk) { - force_sig_specific(SIGKILL, p); - mm->core_waiters++; - if (unlikely(p->ptrace) && - unlikely(p->parent->mm == mm)) - traced = 1; + t = start; + do { + if (t != current && t->mm) { + t->mm->core_waiters++; + sigaddset(&t->pending.signal, SIGKILL); + signal_wake_up(t, 1); } - while_each_thread(g,p); + } while ((t = next_thread(t)) != start); +} - read_unlock(&tasklist_lock); +static inline int zap_threads(struct task_struct *tsk, struct mm_struct *mm, + int exit_code) +{ + struct task_struct *g, *p; + unsigned long flags; + int err = -EAGAIN; + + spin_lock_irq(&tsk->sighand->siglock); + if (!(tsk->signal->flags & SIGNAL_GROUP_EXIT)) { + tsk->signal->group_exit_code = exit_code; + zap_process(tsk); + err = 0; + } + spin_unlock_irq(&tsk->sighand->siglock); + if (err) + return err; - if (unlikely(traced)) { - /* - * We are zapping a thread and the thread it ptraces. - * If the tracee went into a ptrace stop for exit tracing, - * we could deadlock since the tracer is waiting for this - * coredump to finish. Detach them so they can both die. - */ - write_lock_irq(&tasklist_lock); - do_each_thread(g,p) { - if (mm == p->mm && p != tsk && - p->ptrace && p->parent->mm == mm) { - __ptrace_detach(p, 0); + if (atomic_read(&mm->mm_users) == mm->core_waiters + 1) + goto done; + + rcu_read_lock(); + for_each_process(g) { + if (g == tsk->group_leader) + continue; + + p = g; + do { + if (p->mm) { + if (p->mm == mm) { + /* + * p->sighand can't disappear, but + * may be changed by de_thread() + */ + lock_task_sighand(p, &flags); + zap_process(p); + unlock_task_sighand(p, &flags); + } + break; } - } while_each_thread(g,p); - write_unlock_irq(&tasklist_lock); + } while ((p = next_thread(p)) != g); } + rcu_read_unlock(); +done: + return mm->core_waiters; } -static void coredump_wait(struct mm_struct *mm) +static int coredump_wait(int exit_code) { - DECLARE_COMPLETION(startup_done); + struct task_struct *tsk = current; + struct mm_struct *mm = tsk->mm; + struct completion startup_done; + struct completion *vfork_done; int core_waiters; + init_completion(&mm->core_done); + init_completion(&startup_done); mm->core_startup_done = &startup_done; - zap_threads(mm); - core_waiters = mm->core_waiters; + core_waiters = zap_threads(tsk, mm, exit_code); up_write(&mm->mmap_sem); + if (unlikely(core_waiters < 0)) + goto fail; + + /* + * Make sure nobody is waiting for us to release the VM, + * otherwise we can deadlock when we wait on each other + */ + vfork_done = tsk->vfork_done; + if (vfork_done) { + tsk->vfork_done = NULL; + complete(vfork_done); + } + if (core_waiters) wait_for_completion(&startup_done); +fail: BUG_ON(mm->core_waiters); + return core_waiters; } int do_coredump(long signr, int exit_code, struct pt_regs * regs) @@ -1473,22 +1484,9 @@ int do_coredump(long signr, int exit_code, struct pt_regs * regs) } mm->dumpable = 0; - retval = -EAGAIN; - spin_lock_irq(¤t->sighand->siglock); - if (!(current->signal->flags & SIGNAL_GROUP_EXIT)) { - current->signal->flags = SIGNAL_GROUP_EXIT; - current->signal->group_exit_code = exit_code; - current->signal->group_stop_count = 0; - retval = 0; - } - spin_unlock_irq(¤t->sighand->siglock); - if (retval) { - up_write(&mm->mmap_sem); + retval = coredump_wait(exit_code); + if (retval < 0) goto fail; - } - - init_completion(&mm->core_done); - coredump_wait(mm); /* * Clear any false indication of pending signals that might diff --git a/fs/ext2/Makefile b/fs/ext2/Makefile index c5d02da73bc..e0b2b43c1fd 100644 --- a/fs/ext2/Makefile +++ b/fs/ext2/Makefile @@ -4,7 +4,7 @@ obj-$(CONFIG_EXT2_FS) += ext2.o -ext2-y := balloc.o bitmap.o dir.o file.o fsync.o ialloc.o inode.o \ +ext2-y := balloc.o dir.o file.o fsync.o ialloc.o inode.o \ ioctl.o namei.o super.o symlink.o ext2-$(CONFIG_EXT2_FS_XATTR) += xattr.o xattr_user.o xattr_trusted.o diff --git a/fs/ext2/acl.c b/fs/ext2/acl.c index da52b4a5db6..7c420b800c3 100644 --- a/fs/ext2/acl.c +++ b/fs/ext2/acl.c @@ -89,8 +89,8 @@ ext2_acl_to_disk(const struct posix_acl *acl, size_t *size) size_t n; *size = ext2_acl_size(acl->a_count); - ext_acl = (ext2_acl_header *)kmalloc(sizeof(ext2_acl_header) + - acl->a_count * sizeof(ext2_acl_entry), GFP_KERNEL); + ext_acl = kmalloc(sizeof(ext2_acl_header) + acl->a_count * + sizeof(ext2_acl_entry), GFP_KERNEL); if (!ext_acl) return ERR_PTR(-ENOMEM); ext_acl->a_version = cpu_to_le32(EXT2_ACL_VERSION); diff --git a/fs/ext2/balloc.c b/fs/ext2/balloc.c index 2c00953d4b0..b1981d0e95a 100644 --- a/fs/ext2/balloc.c +++ b/fs/ext2/balloc.c @@ -11,7 +11,6 @@ * David S. Miller (davem@caip.rutgers.edu), 1995 */ -#include <linux/config.h> #include "ext2.h" #include <linux/quotaops.h> #include <linux/sched.h> @@ -521,6 +520,25 @@ io_error: goto out_release; } +#ifdef EXT2FS_DEBUG + +static int nibblemap[] = {4, 3, 3, 2, 3, 2, 2, 1, 3, 2, 2, 1, 2, 1, 1, 0}; + +unsigned long ext2_count_free (struct buffer_head * map, unsigned int numchars) +{ + unsigned int i; + unsigned long sum = 0; + + if (!map) + return (0); + for (i = 0; i < numchars; i++) + sum += nibblemap[map->b_data[i] & 0xf] + + nibblemap[(map->b_data[i] >> 4) & 0xf]; + return (sum); +} + +#endif /* EXT2FS_DEBUG */ + unsigned long ext2_count_free_blocks (struct super_block * sb) { struct ext2_group_desc * desc; @@ -530,7 +548,6 @@ unsigned long ext2_count_free_blocks (struct super_block * sb) unsigned long bitmap_count, x; struct ext2_super_block *es; - lock_super (sb); es = EXT2_SB(sb)->s_es; desc_count = 0; bitmap_count = 0; @@ -554,7 +571,6 @@ unsigned long ext2_count_free_blocks (struct super_block * sb) printk("ext2_count_free_blocks: stored = %lu, computed = %lu, %lu\n", (long)le32_to_cpu(es->s_free_blocks_count), desc_count, bitmap_count); - unlock_super (sb); return bitmap_count; #else for (i = 0; i < EXT2_SB(sb)->s_groups_count; i++) { diff --git a/fs/ext2/bitmap.c b/fs/ext2/bitmap.c deleted file mode 100644 index e9983a0dd39..00000000000 --- a/fs/ext2/bitmap.c +++ /dev/null @@ -1,32 +0,0 @@ -/* - * linux/fs/ext2/bitmap.c - * - * Copyright (C) 1992, 1993, 1994, 1995 - * Remy Card (card@masi.ibp.fr) - * Laboratoire MASI - Institut Blaise Pascal - * Universite Pierre et Marie Curie (Paris VI) - */ - -#ifdef EXT2FS_DEBUG - -#include <linux/buffer_head.h> - -#include "ext2.h" - -static int nibblemap[] = {4, 3, 3, 2, 3, 2, 2, 1, 3, 2, 2, 1, 2, 1, 1, 0}; - -unsigned long ext2_count_free (struct buffer_head * map, unsigned int numchars) -{ - unsigned int i; - unsigned long sum = 0; - - if (!map) - return (0); - for (i = 0; i < numchars; i++) - sum += nibblemap[map->b_data[i] & 0xf] + - nibblemap[(map->b_data[i] >> 4) & 0xf]; - return (sum); -} - -#endif /* EXT2FS_DEBUG */ - diff --git a/fs/ext2/dir.c b/fs/ext2/dir.c index 3c1c9aaaca6..92ea8265d7d 100644 --- a/fs/ext2/dir.c +++ b/fs/ext2/dir.c @@ -399,8 +399,7 @@ ino_t ext2_inode_by_name(struct inode * dir, struct dentry *dentry) de = ext2_find_entry (dir, dentry, &page); if (de) { res = le32_to_cpu(de->inode); - kunmap(page); - page_cache_release(page); + ext2_put_page(page); } return res; } diff --git a/fs/ext2/ext2.h b/fs/ext2/ext2.h index 9f74a62be55..e65a019fc7a 100644 --- a/fs/ext2/ext2.h +++ b/fs/ext2/ext2.h @@ -162,9 +162,9 @@ extern const struct file_operations ext2_file_operations; extern const struct file_operations ext2_xip_file_operations; /* inode.c */ -extern struct address_space_operations ext2_aops; -extern struct address_space_operations ext2_aops_xip; -extern struct address_space_operations ext2_nobh_aops; +extern const struct address_space_operations ext2_aops; +extern const struct address_space_operations ext2_aops_xip; +extern const struct address_space_operations ext2_nobh_aops; /* namei.c */ extern struct inode_operations ext2_dir_inode_operations; diff --git a/fs/ext2/fsync.c b/fs/ext2/fsync.c index c9c2e5ffa48..7806b9e8155 100644 --- a/fs/ext2/fsync.c +++ b/fs/ext2/fsync.c @@ -24,7 +24,7 @@ #include "ext2.h" #include <linux/smp_lock.h> -#include <linux/buffer_head.h> /* for fsync_inode_buffers() */ +#include <linux/buffer_head.h> /* for sync_mapping_buffers() */ /* diff --git a/fs/ext2/ialloc.c b/fs/ext2/ialloc.c index e52765219e1..2cb545bf0f3 100644 --- a/fs/ext2/ialloc.c +++ b/fs/ext2/ialloc.c @@ -12,7 +12,6 @@ * David S. Miller (davem@caip.rutgers.edu), 1995 */ -#include <linux/config.h> #include <linux/quotaops.h> #include <linux/sched.h> #include <linux/backing-dev.h> @@ -575,7 +574,6 @@ got: inode->i_mode = mode; inode->i_ino = ino; - inode->i_blksize = PAGE_SIZE; /* This is the optimal IO size (for stat), not the fs block size */ inode->i_blocks = 0; inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME_SEC; memset(ei->i_data, 0, sizeof(ei->i_data)); @@ -649,7 +647,6 @@ unsigned long ext2_count_free_inodes (struct super_block * sb) unsigned long bitmap_count = 0; struct buffer_head *bitmap_bh = NULL; - lock_super (sb); es = EXT2_SB(sb)->s_es; for (i = 0; i < EXT2_SB(sb)->s_groups_count; i++) { unsigned x; @@ -672,7 +669,6 @@ unsigned long ext2_count_free_inodes (struct super_block * sb) printk("ext2_count_free_inodes: stored = %lu, computed = %lu, %lu\n", percpu_counter_read(&EXT2_SB(sb)->s_freeinodes_counter), desc_count, bitmap_count); - unlock_super(sb); return desc_count; #else for (i = 0; i < EXT2_SB(sb)->s_groups_count; i++) { diff --git a/fs/ext2/inode.c b/fs/ext2/inode.c index 04af9c45dce..dd4e14c221e 100644 --- a/fs/ext2/inode.c +++ b/fs/ext2/inode.c @@ -684,7 +684,7 @@ ext2_writepages(struct address_space *mapping, struct writeback_control *wbc) return mpage_writepages(mapping, wbc, ext2_get_block); } -struct address_space_operations ext2_aops = { +const struct address_space_operations ext2_aops = { .readpage = ext2_readpage, .readpages = ext2_readpages, .writepage = ext2_writepage, @@ -697,12 +697,12 @@ struct address_space_operations ext2_aops = { .migratepage = buffer_migrate_page, }; -struct address_space_operations ext2_aops_xip = { +const struct address_space_operations ext2_aops_xip = { .bmap = ext2_bmap, .get_xip_page = ext2_get_xip_page, }; -struct address_space_operations ext2_nobh_aops = { +const struct address_space_operations ext2_nobh_aops = { .readpage = ext2_readpage, .readpages = ext2_readpages, .writepage = ext2_nobh_writepage, @@ -1094,7 +1094,6 @@ void ext2_read_inode (struct inode * inode) brelse (bh); goto bad_inode; } - inode->i_blksize = PAGE_SIZE; /* This is the optimal IO size (for stat), not the fs block size */ inode->i_blocks = le32_to_cpu(raw_inode->i_blocks); ei->i_flags = le32_to_cpu(raw_inode->i_flags); ei->i_faddr = le32_to_cpu(raw_inode->i_faddr); diff --git a/fs/ext2/super.c b/fs/ext2/super.c index ee4ba759581..513cd421ac0 100644 --- a/fs/ext2/super.c +++ b/fs/ext2/super.c @@ -16,7 +16,6 @@ * David S. Miller (davem@caip.rutgers.edu), 1995 */ -#include <linux/config.h> #include <linux/module.h> #include <linux/string.h> #include <linux/fs.h> @@ -185,8 +184,7 @@ static int init_inodecache(void) static void destroy_inodecache(void) { - if (kmem_cache_destroy(ext2_inode_cachep)) - printk(KERN_INFO "ext2_inode_cache: not all structures were freed\n"); + kmem_cache_destroy(ext2_inode_cachep); } static void ext2_clear_inode(struct inode *inode) @@ -252,6 +250,44 @@ static struct super_operations ext2_sops = { #endif }; +static struct dentry *ext2_get_dentry(struct super_block *sb, void *vobjp) +{ + __u32 *objp = vobjp; + unsigned long ino = objp[0]; + __u32 generation = objp[1]; + struct inode *inode; + struct dentry *result; + + if (ino < EXT2_FIRST_INO(sb) && ino != EXT2_ROOT_INO) + return ERR_PTR(-ESTALE); + if (ino > le32_to_cpu(EXT2_SB(sb)->s_es->s_inodes_count)) + return ERR_PTR(-ESTALE); + + /* iget isn't really right if the inode is currently unallocated!! + * ext2_read_inode currently does appropriate checks, but + * it might be "neater" to call ext2_get_inode first and check + * if the inode is valid..... + */ + inode = iget(sb, ino); + if (inode == NULL) + return ERR_PTR(-ENOMEM); + if (is_bad_inode(inode) || + (generation && inode->i_generation != generation)) { + /* we didn't find the right inode.. */ + iput(inode); + return ERR_PTR(-ESTALE); + } + /* now to find a dentry. + * If possible, get a well-connected one + */ + result = d_alloc_anon(inode); + if (!result) { + iput(inode); + return ERR_PTR(-ENOMEM); + } + return result; +} + /* Yes, most of these are left as NULL!! * A NULL value implies the default, which works with ext2-like file * systems, but can be improved upon. @@ -259,6 +295,7 @@ static struct super_operations ext2_sops = { */ static struct export_operations ext2_export_ops = { .get_parent = ext2_get_parent, + .get_dentry = ext2_get_dentry, }; static unsigned long get_sb_block(void **data) @@ -506,17 +543,24 @@ static int ext2_check_descriptors (struct super_block * sb) int i; int desc_block = 0; struct ext2_sb_info *sbi = EXT2_SB(sb); - unsigned long block = le32_to_cpu(sbi->s_es->s_first_data_block); + unsigned long first_block = le32_to_cpu(sbi->s_es->s_first_data_block); + unsigned long last_block; struct ext2_group_desc * gdp = NULL; ext2_debug ("Checking group descriptors"); for (i = 0; i < sbi->s_groups_count; i++) { + if (i == sbi->s_groups_count - 1) + last_block = le32_to_cpu(sbi->s_es->s_blocks_count) - 1; + else + last_block = first_block + + (EXT2_BLOCKS_PER_GROUP(sb) - 1); + if ((i % EXT2_DESC_PER_BLOCK(sb)) == 0) gdp = (struct ext2_group_desc *) sbi->s_group_desc[desc_block++]->b_data; - if (le32_to_cpu(gdp->bg_block_bitmap) < block || - le32_to_cpu(gdp->bg_block_bitmap) >= block + EXT2_BLOCKS_PER_GROUP(sb)) + if (le32_to_cpu(gdp->bg_block_bitmap) < first_block || + le32_to_cpu(gdp->bg_block_bitmap) > last_block) { ext2_error (sb, "ext2_check_descriptors", "Block bitmap for group %d" @@ -524,8 +568,8 @@ static int ext2_check_descriptors (struct super_block * sb) i, (unsigned long) le32_to_cpu(gdp->bg_block_bitmap)); return 0; } - if (le32_to_cpu(gdp->bg_inode_bitmap) < block || - le32_to_cpu(gdp->bg_inode_bitmap) >= block + EXT2_BLOCKS_PER_GROUP(sb)) + if (le32_to_cpu(gdp->bg_inode_bitmap) < first_block || + le32_to_cpu(gdp->bg_inode_bitmap) > last_block) { ext2_error (sb, "ext2_check_descriptors", "Inode bitmap for group %d" @@ -533,9 +577,9 @@ static int ext2_check_descriptors (struct super_block * sb) i, (unsigned long) le32_to_cpu(gdp->bg_inode_bitmap)); return 0; } - if (le32_to_cpu(gdp->bg_inode_table) < block || - le32_to_cpu(gdp->bg_inode_table) + sbi->s_itb_per_group >= - block + EXT2_BLOCKS_PER_GROUP(sb)) + if (le32_to_cpu(gdp->bg_inode_table) < first_block || + le32_to_cpu(gdp->bg_inode_table) + sbi->s_itb_per_group > + last_block) { ext2_error (sb, "ext2_check_descriptors", "Inode table for group %d" @@ -543,7 +587,7 @@ static int ext2_check_descriptors (struct super_block * sb) i, (unsigned long) le32_to_cpu(gdp->bg_inode_table)); return 0; } - block += EXT2_BLOCKS_PER_GROUP(sb); + first_block += EXT2_BLOCKS_PER_GROUP(sb); gdp++; } return 1; @@ -610,11 +654,10 @@ static int ext2_fill_super(struct super_block *sb, void *data, int silent) int i, j; __le32 features; - sbi = kmalloc(sizeof(*sbi), GFP_KERNEL); + sbi = kzalloc(sizeof(*sbi), GFP_KERNEL); if (!sbi) return -ENOMEM; sb->s_fs_info = sbi; - memset(sbi, 0, sizeof(*sbi)); /* * See what the current blocksize for the device is, and @@ -776,7 +819,7 @@ static int ext2_fill_super(struct super_block *sb, void *data, int silent) if (EXT2_INODE_SIZE(sb) == 0) goto cantfind_ext2; sbi->s_inodes_per_block = sb->s_blocksize / EXT2_INODE_SIZE(sb); - if (sbi->s_inodes_per_block == 0) + if (sbi->s_inodes_per_block == 0 || sbi->s_inodes_per_group == 0) goto cantfind_ext2; sbi->s_itb_per_group = sbi->s_inodes_per_group / sbi->s_inodes_per_block; @@ -823,10 +866,9 @@ static int ext2_fill_super(struct super_block *sb, void *data, int silent) if (EXT2_BLOCKS_PER_GROUP(sb) == 0) goto cantfind_ext2; - sbi->s_groups_count = (le32_to_cpu(es->s_blocks_count) - - le32_to_cpu(es->s_first_data_block) + - EXT2_BLOCKS_PER_GROUP(sb) - 1) / - EXT2_BLOCKS_PER_GROUP(sb); + sbi->s_groups_count = ((le32_to_cpu(es->s_blocks_count) - + le32_to_cpu(es->s_first_data_block) - 1) + / EXT2_BLOCKS_PER_GROUP(sb)) + 1; db_count = (sbi->s_groups_count + EXT2_DESC_PER_BLOCK(sb) - 1) / EXT2_DESC_PER_BLOCK(sb); sbi->s_group_desc = kmalloc (db_count * sizeof (struct buffer_head *), GFP_KERNEL); @@ -854,7 +896,6 @@ static int ext2_fill_super(struct super_block *sb, void *data, int silent) } if (!ext2_check_descriptors (sb)) { printk ("EXT2-fs: group descriptors corrupted!\n"); - db_count = i; goto failed_mount2; } sbi->s_gdb_count = db_count; @@ -1157,7 +1198,7 @@ static ssize_t ext2_quota_write(struct super_block *sb, int type, struct buffer_head tmp_bh; struct buffer_head *bh; - mutex_lock(&inode->i_mutex); + mutex_lock_nested(&inode->i_mutex, I_MUTEX_QUOTA); while (towrite > 0) { tocopy = sb->s_blocksize - offset < towrite ? sb->s_blocksize - offset : towrite; diff --git a/fs/ext2/xattr.c b/fs/ext2/xattr.c index 86ae8e93adb..af52a7f8b29 100644 --- a/fs/ext2/xattr.c +++ b/fs/ext2/xattr.c @@ -521,11 +521,10 @@ bad_block: ext2_error(sb, "ext2_xattr_set", } } else { /* Allocate a buffer where we construct the new block. */ - header = kmalloc(sb->s_blocksize, GFP_KERNEL); + header = kzalloc(sb->s_blocksize, GFP_KERNEL); error = -ENOMEM; if (header == NULL) goto cleanup; - memset(header, 0, sb->s_blocksize); end = (char *)header + sb->s_blocksize; header->h_magic = cpu_to_le32(EXT2_XATTR_MAGIC); header->h_blocks = header->h_refcount = cpu_to_le32(1); diff --git a/fs/ext2/xattr.h b/fs/ext2/xattr.h index 67cfeb66e89..bf8175b2ced 100644 --- a/fs/ext2/xattr.h +++ b/fs/ext2/xattr.h @@ -6,7 +6,6 @@ (C) 2001 Andreas Gruenbacher, <a.gruenbacher@computer.org> */ -#include <linux/config.h> #include <linux/init.h> #include <linux/xattr.h> diff --git a/fs/ext3/acl.c b/fs/ext3/acl.c index 0d21d558b87..1e5038d9a01 100644 --- a/fs/ext3/acl.c +++ b/fs/ext3/acl.c @@ -90,8 +90,8 @@ ext3_acl_to_disk(const struct posix_acl *acl, size_t *size) size_t n; *size = ext3_acl_size(acl->a_count); - ext_acl = (ext3_acl_header *)kmalloc(sizeof(ext3_acl_header) + - acl->a_count * sizeof(ext3_acl_entry), GFP_KERNEL); + ext_acl = kmalloc(sizeof(ext3_acl_header) + acl->a_count * + sizeof(ext3_acl_entry), GFP_KERNEL); if (!ext_acl) return ERR_PTR(-ENOMEM); ext_acl->a_version = cpu_to_le32(EXT3_ACL_VERSION); @@ -258,7 +258,7 @@ ext3_set_acl(handle_t *handle, struct inode *inode, int type, default: return -EINVAL; } - if (acl) { + if (acl) { value = ext3_acl_to_disk(acl, &size); if (IS_ERR(value)) return (int)PTR_ERR(value); diff --git a/fs/ext3/acl.h b/fs/ext3/acl.h index 92d50b53a93..0d1e6279cbf 100644 --- a/fs/ext3/acl.h +++ b/fs/ext3/acl.h @@ -62,9 +62,6 @@ extern int ext3_permission (struct inode *, int, struct nameidata *); extern int ext3_acl_chmod (struct inode *); extern int ext3_init_acl (handle_t *, struct inode *, struct inode *); -extern int init_ext3_acl(void); -extern void exit_ext3_acl(void); - #else /* CONFIG_EXT3_FS_POSIX_ACL */ #include <linux/sched.h> #define ext3_permission NULL diff --git a/fs/ext3/balloc.c b/fs/ext3/balloc.c index 77927d6938f..b41a7d7e20f 100644 --- a/fs/ext3/balloc.c +++ b/fs/ext3/balloc.c @@ -11,7 +11,6 @@ * David S. Miller (davem@caip.rutgers.edu), 1995 */ -#include <linux/config.h> #include <linux/time.h> #include <linux/capability.h> #include <linux/fs.h> @@ -39,6 +38,13 @@ #define in_range(b, first, len) ((b) >= (first) && (b) <= (first) + (len) - 1) +/** + * ext3_get_group_desc() -- load group descriptor from disk + * @sb: super block + * @block_group: given block group + * @bh: pointer to the buffer head to store the block + * group descriptor + */ struct ext3_group_desc * ext3_get_group_desc(struct super_block * sb, unsigned int block_group, struct buffer_head ** bh) @@ -74,8 +80,12 @@ struct ext3_group_desc * ext3_get_group_desc(struct super_block * sb, return desc + offset; } -/* - * Read the bitmap for a given block_group, reading into the specified +/** + * read_block_bitmap() + * @sb: super block + * @block_group: given block group + * + * Read the bitmap for a given block_group, reading into the specified * slot in the superblock's bitmap cache. * * Return buffer_head on success or NULL in case of failure. @@ -104,15 +114,22 @@ error_out: * Operations include: * dump, find, add, remove, is_empty, find_next_reservable_window, etc. * - * We use sorted double linked list for the per-filesystem reservation - * window list. (like in vm_region). + * We use a red-black tree to represent per-filesystem reservation + * windows. * - * Initially, we keep those small operations in the abstract functions, - * so later if we need a better searching tree than double linked-list, - * we could easily switch to that without changing too much - * code. */ -#if 0 + +/** + * __rsv_window_dump() -- Dump the filesystem block allocation reservation map + * @rb_root: root of per-filesystem reservation rb tree + * @verbose: verbose mode + * @fn: function which wishes to dump the reservation map + * + * If verbose is turned on, it will print the whole block reservation + * windows(start, end). Otherwise, it will only print out the "bad" windows, + * those windows that overlap with their immediate neighbors. + */ +#if 1 static void __rsv_window_dump(struct rb_root *root, int verbose, const char *fn) { @@ -130,7 +147,7 @@ restart: rsv = list_entry(n, struct ext3_reserve_window_node, rsv_node); if (verbose) printk("reservation window 0x%p " - "start: %d, end: %d\n", + "start: %lu, end: %lu\n", rsv, rsv->rsv_start, rsv->rsv_end); if (rsv->rsv_start && rsv->rsv_start >= rsv->rsv_end) { printk("Bad reservation %p (start >= end)\n", @@ -162,32 +179,51 @@ restart: #define rsv_window_dump(root, verbose) do {} while (0) #endif +/** + * goal_in_my_reservation() + * @rsv: inode's reservation window + * @grp_goal: given goal block relative to the allocation block group + * @group: the current allocation block group + * @sb: filesystem super block + * + * Test if the given goal block (group relative) is within the file's + * own block reservation window range. + * + * If the reservation window is outside the goal allocation group, return 0; + * grp_goal (given goal block) could be -1, which means no specific + * goal block. In this case, always return 1. + * If the goal block is within the reservation window, return 1; + * otherwise, return 0; + */ static int -goal_in_my_reservation(struct ext3_reserve_window *rsv, int goal, +goal_in_my_reservation(struct ext3_reserve_window *rsv, ext3_grpblk_t grp_goal, unsigned int group, struct super_block * sb) { - unsigned long group_first_block, group_last_block; + ext3_fsblk_t group_first_block, group_last_block; - group_first_block = le32_to_cpu(EXT3_SB(sb)->s_es->s_first_data_block) + - group * EXT3_BLOCKS_PER_GROUP(sb); - group_last_block = group_first_block + EXT3_BLOCKS_PER_GROUP(sb) - 1; + group_first_block = ext3_group_first_block_no(sb, group); + group_last_block = group_first_block + (EXT3_BLOCKS_PER_GROUP(sb) - 1); if ((rsv->_rsv_start > group_last_block) || (rsv->_rsv_end < group_first_block)) return 0; - if ((goal >= 0) && ((goal + group_first_block < rsv->_rsv_start) - || (goal + group_first_block > rsv->_rsv_end))) + if ((grp_goal >= 0) && ((grp_goal + group_first_block < rsv->_rsv_start) + || (grp_goal + group_first_block > rsv->_rsv_end))) return 0; return 1; } -/* +/** + * search_reserve_window() + * @rb_root: root of reservation tree + * @goal: target allocation block + * * Find the reserved window which includes the goal, or the previous one * if the goal is not in any window. * Returns NULL if there are no windows or if all windows start after the goal. */ static struct ext3_reserve_window_node * -search_reserve_window(struct rb_root *root, unsigned long goal) +search_reserve_window(struct rb_root *root, ext3_fsblk_t goal) { struct rb_node *n = root->rb_node; struct ext3_reserve_window_node *rsv; @@ -218,12 +254,19 @@ search_reserve_window(struct rb_root *root, unsigned long goal) return rsv; } +/** + * ext3_rsv_window_add() -- Insert a window to the block reservation rb tree. + * @sb: super block + * @rsv: reservation window to add + * + * Must be called with rsv_lock hold. + */ void ext3_rsv_window_add(struct super_block *sb, struct ext3_reserve_window_node *rsv) { struct rb_root *root = &EXT3_SB(sb)->s_rsv_window_root; struct rb_node *node = &rsv->rsv_node; - unsigned int start = rsv->rsv_start; + ext3_fsblk_t start = rsv->rsv_start; struct rb_node ** p = &root->rb_node; struct rb_node * parent = NULL; @@ -238,14 +281,25 @@ void ext3_rsv_window_add(struct super_block *sb, p = &(*p)->rb_left; else if (start > this->rsv_end) p = &(*p)->rb_right; - else + else { + rsv_window_dump(root, 1); BUG(); + } } rb_link_node(node, parent, p); rb_insert_color(node, root); } +/** + * ext3_rsv_window_remove() -- unlink a window from the reservation rb tree + * @sb: super block + * @rsv: reservation window to remove + * + * Mark the block reservation window as not allocated, and unlink it + * from the filesystem reservation window rb tree. Must be called with + * rsv_lock hold. + */ static void rsv_window_remove(struct super_block *sb, struct ext3_reserve_window_node *rsv) { @@ -255,11 +309,39 @@ static void rsv_window_remove(struct super_block *sb, rb_erase(&rsv->rsv_node, &EXT3_SB(sb)->s_rsv_window_root); } +/* + * rsv_is_empty() -- Check if the reservation window is allocated. + * @rsv: given reservation window to check + * + * returns 1 if the end block is EXT3_RESERVE_WINDOW_NOT_ALLOCATED. + */ static inline int rsv_is_empty(struct ext3_reserve_window *rsv) { /* a valid reservation end block could not be 0 */ - return (rsv->_rsv_end == EXT3_RESERVE_WINDOW_NOT_ALLOCATED); + return rsv->_rsv_end == EXT3_RESERVE_WINDOW_NOT_ALLOCATED; } + +/** + * ext3_init_block_alloc_info() + * @inode: file inode structure + * + * Allocate and initialize the reservation window structure, and + * link the window to the ext3 inode structure at last + * + * The reservation window structure is only dynamically allocated + * and linked to ext3 inode the first time the open file + * needs a new block. So, before every ext3_new_block(s) call, for + * regular files, we should check whether the reservation window + * structure exists or not. In the latter case, this function is called. + * Fail to do so will result in block reservation being turned off for that + * open file. + * + * This function is called from ext3_get_blocks_handle(), also called + * when setting the reservation window size through ioctl before the file + * is open for write (needs block allocation). + * + * Needs truncate_mutex protection prior to call this function. + */ void ext3_init_block_alloc_info(struct inode *inode) { struct ext3_inode_info *ei = EXT3_I(inode); @@ -273,7 +355,7 @@ void ext3_init_block_alloc_info(struct inode *inode) rsv->rsv_start = EXT3_RESERVE_WINDOW_NOT_ALLOCATED; rsv->rsv_end = EXT3_RESERVE_WINDOW_NOT_ALLOCATED; - /* + /* * if filesystem is mounted with NORESERVATION, the goal * reservation window size is set to zero to indicate * block reservation is off @@ -289,6 +371,19 @@ void ext3_init_block_alloc_info(struct inode *inode) ei->i_block_alloc_info = block_i; } +/** + * ext3_discard_reservation() + * @inode: inode + * + * Discard(free) block reservation window on last file close, or truncate + * or at last iput(). + * + * It is being called in three cases: + * ext3_release_file(): last writer close the file + * ext3_clear_inode(): last iput(), when nobody link to this file. + * ext3_truncate(): when the block indirect map is about to change. + * + */ void ext3_discard_reservation(struct inode *inode) { struct ext3_inode_info *ei = EXT3_I(inode); @@ -308,22 +403,29 @@ void ext3_discard_reservation(struct inode *inode) } } -/* Free given blocks, update quota and i_blocks field */ +/** + * ext3_free_blocks_sb() -- Free given blocks and update quota + * @handle: handle to this transaction + * @sb: super block + * @block: start physcial block to free + * @count: number of blocks to free + * @pdquot_freed_blocks: pointer to quota + */ void ext3_free_blocks_sb(handle_t *handle, struct super_block *sb, - unsigned long block, unsigned long count, - int *pdquot_freed_blocks) + ext3_fsblk_t block, unsigned long count, + unsigned long *pdquot_freed_blocks) { struct buffer_head *bitmap_bh = NULL; struct buffer_head *gd_bh; unsigned long block_group; - unsigned long bit; + ext3_grpblk_t bit; unsigned long i; unsigned long overflow; struct ext3_group_desc * desc; struct ext3_super_block * es; struct ext3_sb_info *sbi; int err = 0, ret; - unsigned group_freed; + ext3_grpblk_t group_freed; *pdquot_freed_blocks = 0; sbi = EXT3_SB(sb); @@ -333,7 +435,7 @@ void ext3_free_blocks_sb(handle_t *handle, struct super_block *sb, block + count > le32_to_cpu(es->s_blocks_count)) { ext3_error (sb, "ext3_free_blocks", "Freeing blocks not in datazone - " - "block = %lu, count = %lu", block, count); + "block = "E3FSBLK", count = %lu", block, count); goto error_return; } @@ -369,7 +471,7 @@ do_more: sbi->s_itb_per_group)) ext3_error (sb, "ext3_free_blocks", "Freeing blocks in system zones - " - "Block = %lu, count = %lu", + "Block = "E3FSBLK", count = %lu", block, count); /* @@ -421,8 +523,8 @@ do_more: } /* @@@ This prevents newly-allocated data from being * freed and then reallocated within the same - * transaction. - * + * transaction. + * * Ideally we would want to allow that to happen, but to * do so requires making journal_forget() capable of * revoking the queued write of a data block, which @@ -435,7 +537,7 @@ do_more: * safe not to set the allocation bit in the committed * bitmap, because we know that there is no outstanding * activity on the buffer any more and so it is safe to - * reallocate it. + * reallocate it. */ BUFFER_TRACE(bitmap_bh, "set in b_committed_data"); J_ASSERT_BH(bitmap_bh, @@ -453,7 +555,8 @@ do_more: bit + i, bitmap_bh->b_data)) { jbd_unlock_bh_state(bitmap_bh); ext3_error(sb, __FUNCTION__, - "bit already cleared for block %lu", block + i); + "bit already cleared for block "E3FSBLK, + block + i); jbd_lock_bh_state(bitmap_bh); BUFFER_TRACE(bitmap_bh, "bit already cleared"); } else { @@ -491,12 +594,18 @@ error_return: return; } -/* Free given blocks, update quota and i_blocks field */ +/** + * ext3_free_blocks() -- Free given blocks and update quota + * @handle: handle for this transaction + * @inode: inode + * @block: start physical block to free + * @count: number of blocks to count + */ void ext3_free_blocks(handle_t *handle, struct inode *inode, - unsigned long block, unsigned long count) + ext3_fsblk_t block, unsigned long count) { struct super_block * sb; - int dquot_freed_blocks; + unsigned long dquot_freed_blocks; sb = inode->i_sb; if (!sb) { @@ -509,7 +618,11 @@ void ext3_free_blocks(handle_t *handle, struct inode *inode, return; } -/* +/** + * ext3_test_allocatable() + * @nr: given allocation block group + * @bh: bufferhead contains the bitmap of the given block group + * * For ext3 allocations, we must not reuse any blocks which are * allocated in the bitmap buffer's "last committed data" copy. This * prevents deletes from freeing up the page for reuse until we have @@ -519,13 +632,13 @@ void ext3_free_blocks(handle_t *handle, struct inode *inode, * data would allow the old block to be overwritten before the * transaction committed (because we force data to disk before commit). * This would lead to corruption if we crashed between overwriting the - * data and committing the delete. + * data and committing the delete. * * @@@ We may want to make this allocation behaviour conditional on * data-writes at some point, and disable it for metadata allocations or * sync-data inodes. */ -static int ext3_test_allocatable(int nr, struct buffer_head *bh) +static int ext3_test_allocatable(ext3_grpblk_t nr, struct buffer_head *bh) { int ret; struct journal_head *jh = bh2jh(bh); @@ -542,18 +655,23 @@ static int ext3_test_allocatable(int nr, struct buffer_head *bh) return ret; } -static int -bitmap_search_next_usable_block(int start, struct buffer_head *bh, - int maxblocks) +/** + * bitmap_search_next_usable_block() + * @start: the starting block (group relative) of the search + * @bh: bufferhead contains the block group bitmap + * @maxblocks: the ending block (group relative) of the reservation + * + * The bitmap search --- search forward alternately through the actual + * bitmap on disk and the last-committed copy in journal, until we find a + * bit free in both bitmaps. + */ +static ext3_grpblk_t +bitmap_search_next_usable_block(ext3_grpblk_t start, struct buffer_head *bh, + ext3_grpblk_t maxblocks) { - int next; + ext3_grpblk_t next; struct journal_head *jh = bh2jh(bh); - /* - * The bitmap search --- search forward alternately through the actual - * bitmap and the last-committed copy until we find a bit free in - * both - */ while (start < maxblocks) { next = ext3_find_next_zero_bit(bh->b_data, maxblocks, start); if (next >= maxblocks) @@ -563,35 +681,42 @@ bitmap_search_next_usable_block(int start, struct buffer_head *bh, jbd_lock_bh_state(bh); if (jh->b_committed_data) start = ext3_find_next_zero_bit(jh->b_committed_data, - maxblocks, next); + maxblocks, next); jbd_unlock_bh_state(bh); } return -1; } -/* - * Find an allocatable block in a bitmap. We honour both the bitmap and +/** + * find_next_usable_block() + * @start: the starting block (group relative) to find next + * allocatable block in bitmap. + * @bh: bufferhead contains the block group bitmap + * @maxblocks: the ending block (group relative) for the search + * + * Find an allocatable block in a bitmap. We honor both the bitmap and * its last-committed copy (if that exists), and perform the "most * appropriate allocation" algorithm of looking for a free block near * the initial goal; then for a free byte somewhere in the bitmap; then * for any free bit in the bitmap. */ -static int -find_next_usable_block(int start, struct buffer_head *bh, int maxblocks) +static ext3_grpblk_t +find_next_usable_block(ext3_grpblk_t start, struct buffer_head *bh, + ext3_grpblk_t maxblocks) { - int here, next; + ext3_grpblk_t here, next; char *p, *r; if (start > 0) { /* - * The goal was occupied; search forward for a free + * The goal was occupied; search forward for a free * block within the next XX blocks. * * end_goal is more or less random, but it has to be * less than EXT3_BLOCKS_PER_GROUP. Aligning up to the * next 64-bit boundary is simple.. */ - int end_goal = (start + 63) & ~63; + ext3_grpblk_t end_goal = (start + 63) & ~63; if (end_goal > maxblocks) end_goal = maxblocks; here = ext3_find_next_zero_bit(bh->b_data, end_goal, start); @@ -620,7 +745,11 @@ find_next_usable_block(int start, struct buffer_head *bh, int maxblocks) return here; } -/* +/** + * claim_block() + * @block: the free block (group relative) to allocate + * @bh: the bufferhead containts the block group bitmap + * * We think we can allocate this block in this bitmap. Try to set the bit. * If that succeeds then check that nobody has allocated and then freed the * block since we saw that is was not marked in b_committed_data. If it _was_ @@ -628,7 +757,7 @@ find_next_usable_block(int start, struct buffer_head *bh, int maxblocks) * zero (failure). */ static inline int -claim_block(spinlock_t *lock, int block, struct buffer_head *bh) +claim_block(spinlock_t *lock, ext3_grpblk_t block, struct buffer_head *bh) { struct journal_head *jh = bh2jh(bh); int ret; @@ -646,24 +775,42 @@ claim_block(spinlock_t *lock, int block, struct buffer_head *bh) return ret; } -/* +/** + * ext3_try_to_allocate() + * @sb: superblock + * @handle: handle to this transaction + * @group: given allocation block group + * @bitmap_bh: bufferhead holds the block bitmap + * @grp_goal: given target block within the group + * @count: target number of blocks to allocate + * @my_rsv: reservation window + * + * Attempt to allocate blocks within a give range. Set the range of allocation + * first, then find the first free bit(s) from the bitmap (within the range), + * and at last, allocate the blocks by claiming the found free bit as allocated. + * + * To set the range of this allocation: + * if there is a reservation window, only try to allocate block(s) from the + * file's own reservation window; + * Otherwise, the allocation range starts from the give goal block, ends at + * the block group's last block. + * * If we failed to allocate the desired block then we may end up crossing to a * new bitmap. In that case we must release write access to the old one via * ext3_journal_release_buffer(), else we'll run out of credits. */ -static int +static ext3_grpblk_t ext3_try_to_allocate(struct super_block *sb, handle_t *handle, int group, - struct buffer_head *bitmap_bh, int goal, + struct buffer_head *bitmap_bh, ext3_grpblk_t grp_goal, unsigned long *count, struct ext3_reserve_window *my_rsv) { - int group_first_block, start, end; + ext3_fsblk_t group_first_block; + ext3_grpblk_t start, end; unsigned long num = 0; /* we do allocation within the reservation window if we have a window */ if (my_rsv) { - group_first_block = - le32_to_cpu(EXT3_SB(sb)->s_es->s_first_data_block) + - group * EXT3_BLOCKS_PER_GROUP(sb); + group_first_block = ext3_group_first_block_no(sb, group); if (my_rsv->_rsv_start >= group_first_block) start = my_rsv->_rsv_start - group_first_block; else @@ -673,13 +820,13 @@ ext3_try_to_allocate(struct super_block *sb, handle_t *handle, int group, if (end > EXT3_BLOCKS_PER_GROUP(sb)) /* reservation window crosses group boundary */ end = EXT3_BLOCKS_PER_GROUP(sb); - if ((start <= goal) && (goal < end)) - start = goal; + if ((start <= grp_goal) && (grp_goal < end)) + start = grp_goal; else - goal = -1; + grp_goal = -1; } else { - if (goal > 0) - start = goal; + if (grp_goal > 0) + start = grp_goal; else start = 0; end = EXT3_BLOCKS_PER_GROUP(sb); @@ -688,55 +835,57 @@ ext3_try_to_allocate(struct super_block *sb, handle_t *handle, int group, BUG_ON(start > EXT3_BLOCKS_PER_GROUP(sb)); repeat: - if (goal < 0 || !ext3_test_allocatable(goal, bitmap_bh)) { - goal = find_next_usable_block(start, bitmap_bh, end); - if (goal < 0) + if (grp_goal < 0 || !ext3_test_allocatable(grp_goal, bitmap_bh)) { + grp_goal = find_next_usable_block(start, bitmap_bh, end); + if (grp_goal < 0) goto fail_access; if (!my_rsv) { int i; - for (i = 0; i < 7 && goal > start && - ext3_test_allocatable(goal - 1, + for (i = 0; i < 7 && grp_goal > start && + ext3_test_allocatable(grp_goal - 1, bitmap_bh); - i++, goal--) + i++, grp_goal--) ; } } - start = goal; + start = grp_goal; - if (!claim_block(sb_bgl_lock(EXT3_SB(sb), group), goal, bitmap_bh)) { + if (!claim_block(sb_bgl_lock(EXT3_SB(sb), group), + grp_goal, bitmap_bh)) { /* * The block was allocated by another thread, or it was * allocated and then freed by another thread */ start++; - goal++; + grp_goal++; if (start >= end) goto fail_access; goto repeat; } num++; - goal++; - while (num < *count && goal < end - && ext3_test_allocatable(goal, bitmap_bh) - && claim_block(sb_bgl_lock(EXT3_SB(sb), group), goal, bitmap_bh)) { + grp_goal++; + while (num < *count && grp_goal < end + && ext3_test_allocatable(grp_goal, bitmap_bh) + && claim_block(sb_bgl_lock(EXT3_SB(sb), group), + grp_goal, bitmap_bh)) { num++; - goal++; + grp_goal++; } *count = num; - return goal - num; + return grp_goal - num; fail_access: *count = num; return -1; } /** - * find_next_reservable_window(): + * find_next_reservable_window(): * find a reservable space within the given range. * It does not allocate the reservation window for now: * alloc_new_reservation() will do the work later. * - * @search_head: the head of the searching list; + * @search_head: the head of the searching list; * This is not necessarily the list head of the whole filesystem * * We have both head and start_block to assist the search @@ -744,12 +893,12 @@ fail_access: * but we will shift to the place where start_block is, * then start from there, when looking for a reservable space. * - * @size: the target new reservation window size + * @size: the target new reservation window size * - * @group_first_block: the first block we consider to start + * @group_first_block: the first block we consider to start * the real search from * - * @last_block: + * @last_block: * the maximum block number that our goal reservable space * could start from. This is normally the last block in this * group. The search will end when we found the start of next @@ -757,21 +906,22 @@ fail_access: * This could handle the cross boundary reservation window * request. * - * basically we search from the given range, rather than the whole - * reservation double linked list, (start_block, last_block) - * to find a free region that is of my size and has not - * been reserved. + * basically we search from the given range, rather than the whole + * reservation double linked list, (start_block, last_block) + * to find a free region that is of my size and has not + * been reserved. * */ static int find_next_reservable_window( struct ext3_reserve_window_node *search_head, struct ext3_reserve_window_node *my_rsv, - struct super_block * sb, int start_block, - int last_block) + struct super_block * sb, + ext3_fsblk_t start_block, + ext3_fsblk_t last_block) { struct rb_node *next; struct ext3_reserve_window_node *rsv, *prev; - int cur; + ext3_fsblk_t cur; int size = my_rsv->rsv_goal_size; /* TODO: make the start of the reservation window byte-aligned */ @@ -812,7 +962,7 @@ static int find_next_reservable_window( /* * Found a reserveable space big enough. We could * have a reservation across the group boundary here - */ + */ break; } } @@ -848,7 +998,7 @@ static int find_next_reservable_window( } /** - * alloc_new_reservation()--allocate a new reservation window + * alloc_new_reservation()--allocate a new reservation window * * To make a new reservation, we search part of the filesystem * reservation list (the list that inside the group). We try to @@ -873,10 +1023,10 @@ static int find_next_reservable_window( * * @rsv: the reservation * - * @goal: The goal (group-relative). It is where the search for a + * @grp_goal: The goal (group-relative). It is where the search for a * free reservable space should start from. - * if we have a goal(goal >0 ), then start from there, - * no goal(goal = -1), we start from the first block + * if we have a grp_goal(grp_goal >0 ), then start from there, + * no grp_goal(grp_goal = -1), we start from the first block * of the group. * * @sb: the super block @@ -885,25 +1035,24 @@ static int find_next_reservable_window( * */ static int alloc_new_reservation(struct ext3_reserve_window_node *my_rsv, - int goal, struct super_block *sb, + ext3_grpblk_t grp_goal, struct super_block *sb, unsigned int group, struct buffer_head *bitmap_bh) { struct ext3_reserve_window_node *search_head; - int group_first_block, group_end_block, start_block; - int first_free_block; + ext3_fsblk_t group_first_block, group_end_block, start_block; + ext3_grpblk_t first_free_block; struct rb_root *fs_rsv_root = &EXT3_SB(sb)->s_rsv_window_root; unsigned long size; int ret; spinlock_t *rsv_lock = &EXT3_SB(sb)->s_rsv_window_lock; - group_first_block = le32_to_cpu(EXT3_SB(sb)->s_es->s_first_data_block) + - group * EXT3_BLOCKS_PER_GROUP(sb); - group_end_block = group_first_block + EXT3_BLOCKS_PER_GROUP(sb) - 1; + group_first_block = ext3_group_first_block_no(sb, group); + group_end_block = group_first_block + (EXT3_BLOCKS_PER_GROUP(sb) - 1); - if (goal < 0) + if (grp_goal < 0) start_block = group_first_block; else - start_block = goal + group_first_block; + start_block = grp_goal + group_first_block; size = my_rsv->rsv_goal_size; @@ -930,9 +1079,10 @@ static int alloc_new_reservation(struct ext3_reserve_window_node *my_rsv, if ((my_rsv->rsv_alloc_hit > (my_rsv->rsv_end - my_rsv->rsv_start + 1) / 2)) { /* - * if we previously allocation hit ration is greater than half - * we double the size of reservation window next time - * otherwise keep the same + * if the previously allocation hit ratio is + * greater than 1/2, then we double the size of + * the reservation window the next time, + * otherwise we keep the same size window */ size = size * 2; if (size > EXT3_MAX_RESERVE_BLOCKS) @@ -1011,6 +1161,23 @@ retry: goto retry; } +/** + * try_to_extend_reservation() + * @my_rsv: given reservation window + * @sb: super block + * @size: the delta to extend + * + * Attempt to expand the reservation window large enough to have + * required number of free blocks + * + * Since ext3_try_to_allocate() will always allocate blocks within + * the reservation window range, if the window size is too small, + * multiple blocks allocation has to stop at the end of the reservation + * window. To make this more efficient, given the total number of + * blocks needed and the current size of the window, we try to + * expand the reservation window size if necessary on a best-effort + * basis before ext3_new_blocks() tries to allocate blocks, + */ static void try_to_extend_reservation(struct ext3_reserve_window_node *my_rsv, struct super_block *sb, int size) { @@ -1036,7 +1203,17 @@ static void try_to_extend_reservation(struct ext3_reserve_window_node *my_rsv, spin_unlock(rsv_lock); } -/* +/** + * ext3_try_to_allocate_with_rsv() + * @sb: superblock + * @handle: handle to this transaction + * @group: given allocation block group + * @bitmap_bh: bufferhead holds the block bitmap + * @grp_goal: given target block within the group + * @count: target number of blocks to allocate + * @my_rsv: reservation window + * @errp: pointer to store the error code + * * This is the main function used to allocate a new block and its reservation * window. * @@ -1052,19 +1229,18 @@ static void try_to_extend_reservation(struct ext3_reserve_window_node *my_rsv, * reservation), and there are lots of free blocks, but they are all * being reserved. * - * We use a sorted double linked list for the per-filesystem reservation list. - * The insert, remove and find a free space(non-reserved) operations for the - * sorted double linked list should be fast. + * We use a red-black tree for the per-filesystem reservation list. * */ -static int +static ext3_grpblk_t ext3_try_to_allocate_with_rsv(struct super_block *sb, handle_t *handle, unsigned int group, struct buffer_head *bitmap_bh, - int goal, struct ext3_reserve_window_node * my_rsv, + ext3_grpblk_t grp_goal, + struct ext3_reserve_window_node * my_rsv, unsigned long *count, int *errp) { - unsigned long group_first_block; - int ret = 0; + ext3_fsblk_t group_first_block, group_last_block; + ext3_grpblk_t ret = 0; int fatal; unsigned long num = *count; @@ -1090,17 +1266,17 @@ ext3_try_to_allocate_with_rsv(struct super_block *sb, handle_t *handle, */ if (my_rsv == NULL ) { ret = ext3_try_to_allocate(sb, handle, group, bitmap_bh, - goal, count, NULL); + grp_goal, count, NULL); goto out; } /* - * goal is a group relative block number (if there is a goal) - * 0 < goal < EXT3_BLOCKS_PER_GROUP(sb) + * grp_goal is a group relative block number (if there is a goal) + * 0 < grp_goal < EXT3_BLOCKS_PER_GROUP(sb) * first block is a filesystem wide block number * first block is the block number of the first block in this group */ - group_first_block = le32_to_cpu(EXT3_SB(sb)->s_es->s_first_data_block) + - group * EXT3_BLOCKS_PER_GROUP(sb); + group_first_block = ext3_group_first_block_no(sb, group); + group_last_block = group_first_block + (EXT3_BLOCKS_PER_GROUP(sb) - 1); /* * Basically we will allocate a new block from inode's reservation @@ -1119,25 +1295,30 @@ ext3_try_to_allocate_with_rsv(struct super_block *sb, handle_t *handle, */ while (1) { if (rsv_is_empty(&my_rsv->rsv_window) || (ret < 0) || - !goal_in_my_reservation(&my_rsv->rsv_window, goal, group, sb)) { + !goal_in_my_reservation(&my_rsv->rsv_window, + grp_goal, group, sb)) { if (my_rsv->rsv_goal_size < *count) my_rsv->rsv_goal_size = *count; - ret = alloc_new_reservation(my_rsv, goal, sb, + ret = alloc_new_reservation(my_rsv, grp_goal, sb, group, bitmap_bh); if (ret < 0) break; /* failed */ - if (!goal_in_my_reservation(&my_rsv->rsv_window, goal, group, sb)) - goal = -1; - } else if (goal > 0 && (my_rsv->rsv_end-goal+1) < *count) + if (!goal_in_my_reservation(&my_rsv->rsv_window, + grp_goal, group, sb)) + grp_goal = -1; + } else if (grp_goal > 0 && + (my_rsv->rsv_end-grp_goal+1) < *count) try_to_extend_reservation(my_rsv, sb, - *count-my_rsv->rsv_end + goal - 1); + *count-my_rsv->rsv_end + grp_goal - 1); - if ((my_rsv->rsv_start >= group_first_block + EXT3_BLOCKS_PER_GROUP(sb)) - || (my_rsv->rsv_end < group_first_block)) + if ((my_rsv->rsv_start > group_last_block) || + (my_rsv->rsv_end < group_first_block)) { + rsv_window_dump(&EXT3_SB(sb)->s_rsv_window_root, 1); BUG(); - ret = ext3_try_to_allocate(sb, handle, group, bitmap_bh, goal, - &num, &my_rsv->rsv_window); + } + ret = ext3_try_to_allocate(sb, handle, group, bitmap_bh, + grp_goal, &num, &my_rsv->rsv_window); if (ret >= 0) { my_rsv->rsv_alloc_hit += num; *count = num; @@ -1162,9 +1343,15 @@ out: return ret; } +/** + * ext3_has_free_blocks() + * @sbi: in-core super block structure. + * + * Check if filesystem has at least 1 free block available for allocation. + */ static int ext3_has_free_blocks(struct ext3_sb_info *sbi) { - int free_blocks, root_blocks; + ext3_fsblk_t free_blocks, root_blocks; free_blocks = percpu_counter_read_positive(&sbi->s_freeblocks_counter); root_blocks = le32_to_cpu(sbi->s_es->s_r_blocks_count); @@ -1176,11 +1363,17 @@ static int ext3_has_free_blocks(struct ext3_sb_info *sbi) return 1; } -/* +/** + * ext3_should_retry_alloc() + * @sb: super block + * @retries number of attemps has been made + * * ext3_should_retry_alloc() is called when ENOSPC is returned, and if * it is profitable to retry the operation, this function will wait * for the current or commiting transaction to complete, and then * return TRUE. + * + * if the total number of retries exceed three times, return FALSE. */ int ext3_should_retry_alloc(struct super_block *sb, int *retries) { @@ -1192,27 +1385,34 @@ int ext3_should_retry_alloc(struct super_block *sb, int *retries) return journal_force_commit_nested(EXT3_SB(sb)->s_journal); } -/* - * ext3_new_block uses a goal block to assist allocation. If the goal is - * free, or there is a free block within 32 blocks of the goal, that block - * is allocated. Otherwise a forward search is made for a free block; within - * each block group the search first looks for an entire free byte in the block - * bitmap, and then for any free bit if that fails. - * This function also updates quota and i_blocks field. +/** + * ext3_new_blocks() -- core block(s) allocation function + * @handle: handle to this transaction + * @inode: file inode + * @goal: given target block(filesystem wide) + * @count: target number of blocks to allocate + * @errp: error code + * + * ext3_new_blocks uses a goal block to assist allocation. It tries to + * allocate block(s) from the block group contains the goal block first. If that + * fails, it will try to allocate block(s) from other block groups without + * any specific goal block. + * */ -int ext3_new_blocks(handle_t *handle, struct inode *inode, - unsigned long goal, unsigned long *count, int *errp) +ext3_fsblk_t ext3_new_blocks(handle_t *handle, struct inode *inode, + ext3_fsblk_t goal, unsigned long *count, int *errp) { struct buffer_head *bitmap_bh = NULL; struct buffer_head *gdp_bh; int group_no; int goal_group; - int ret_block; + ext3_grpblk_t grp_target_blk; /* blockgroup relative goal block */ + ext3_grpblk_t grp_alloc_blk; /* blockgroup-relative allocated block*/ + ext3_fsblk_t ret_block; /* filesyetem-wide allocated block */ int bgi; /* blockgroup iteration index */ - int target_block; int fatal = 0, err; int performed_allocation = 0; - int free_blocks; + ext3_grpblk_t free_blocks; /* number of free blocks in a group */ struct super_block *sb; struct ext3_group_desc *gdp; struct ext3_super_block *es; @@ -1269,12 +1469,12 @@ int ext3_new_blocks(handle_t *handle, struct inode *inode, goal = le32_to_cpu(es->s_first_data_block); group_no = (goal - le32_to_cpu(es->s_first_data_block)) / EXT3_BLOCKS_PER_GROUP(sb); + goal_group = group_no; +retry_alloc: gdp = ext3_get_group_desc(sb, group_no, &gdp_bh); if (!gdp) goto io_error; - goal_group = group_no; -retry: free_blocks = le16_to_cpu(gdp->bg_free_blocks_count); /* * if there is not enough free blocks to make a new resevation @@ -1285,16 +1485,17 @@ retry: my_rsv = NULL; if (free_blocks > 0) { - ret_block = ((goal - le32_to_cpu(es->s_first_data_block)) % + grp_target_blk = ((goal - le32_to_cpu(es->s_first_data_block)) % EXT3_BLOCKS_PER_GROUP(sb)); bitmap_bh = read_block_bitmap(sb, group_no); if (!bitmap_bh) goto io_error; - ret_block = ext3_try_to_allocate_with_rsv(sb, handle, group_no, - bitmap_bh, ret_block, my_rsv, &num, &fatal); + grp_alloc_blk = ext3_try_to_allocate_with_rsv(sb, handle, + group_no, bitmap_bh, grp_target_blk, + my_rsv, &num, &fatal); if (fatal) goto out; - if (ret_block >= 0) + if (grp_alloc_blk >= 0) goto allocated; } @@ -1302,7 +1503,7 @@ retry: smp_rmb(); /* - * Now search the rest of the groups. We assume that + * Now search the rest of the groups. We assume that * i and gdp correctly point to the last group visited. */ for (bgi = 0; bgi < ngroups; bgi++) { @@ -1327,11 +1528,15 @@ retry: bitmap_bh = read_block_bitmap(sb, group_no); if (!bitmap_bh) goto io_error; - ret_block = ext3_try_to_allocate_with_rsv(sb, handle, group_no, - bitmap_bh, -1, my_rsv, &num, &fatal); + /* + * try to allocate block(s) from this group, without a goal(-1). + */ + grp_alloc_blk = ext3_try_to_allocate_with_rsv(sb, handle, + group_no, bitmap_bh, -1, my_rsv, + &num, &fatal); if (fatal) goto out; - if (ret_block >= 0) + if (grp_alloc_blk >= 0) goto allocated; } /* @@ -1344,7 +1549,7 @@ retry: if (my_rsv) { my_rsv = NULL; group_no = goal_group; - goto retry; + goto retry_alloc; } /* No space left on the device */ *errp = -ENOSPC; @@ -1360,18 +1565,18 @@ allocated: if (fatal) goto out; - target_block = ret_block + group_no * EXT3_BLOCKS_PER_GROUP(sb) - + le32_to_cpu(es->s_first_data_block); + ret_block = grp_alloc_blk + ext3_group_first_block_no(sb, group_no); - if (in_range(le32_to_cpu(gdp->bg_block_bitmap), target_block, num) || - in_range(le32_to_cpu(gdp->bg_inode_bitmap), target_block, num) || - in_range(target_block, le32_to_cpu(gdp->bg_inode_table), + if (in_range(le32_to_cpu(gdp->bg_block_bitmap), ret_block, num) || + in_range(le32_to_cpu(gdp->bg_inode_bitmap), ret_block, num) || + in_range(ret_block, le32_to_cpu(gdp->bg_inode_table), EXT3_SB(sb)->s_itb_per_group) || - in_range(target_block + num - 1, le32_to_cpu(gdp->bg_inode_table), + in_range(ret_block + num - 1, le32_to_cpu(gdp->bg_inode_table), EXT3_SB(sb)->s_itb_per_group)) ext3_error(sb, "ext3_new_block", "Allocating block in system zone - " - "blocks from %u, length %lu", target_block, num); + "blocks from "E3FSBLK", length %lu", + ret_block, num); performed_allocation = 1; @@ -1380,7 +1585,7 @@ allocated: struct buffer_head *debug_bh; /* Record bitmap buffer state in the newly allocated block */ - debug_bh = sb_find_get_block(sb, target_block); + debug_bh = sb_find_get_block(sb, ret_block); if (debug_bh) { BUFFER_TRACE(debug_bh, "state when allocated"); BUFFER_TRACE2(debug_bh, bitmap_bh, "bitmap state"); @@ -1393,24 +1598,21 @@ allocated: int i; for (i = 0; i < num; i++) { - if (ext3_test_bit(ret_block, + if (ext3_test_bit(grp_alloc_blk+i, bh2jh(bitmap_bh)->b_committed_data)) { printk("%s: block was unexpectedly set in " "b_committed_data\n", __FUNCTION__); } } } - ext3_debug("found bit %d\n", ret_block); + ext3_debug("found bit %d\n", grp_alloc_blk); spin_unlock(sb_bgl_lock(sbi, group_no)); jbd_unlock_bh_state(bitmap_bh); #endif - /* ret_block was blockgroup-relative. Now it becomes fs-relative */ - ret_block = target_block; - if (ret_block + num - 1 >= le32_to_cpu(es->s_blocks_count)) { ext3_error(sb, "ext3_new_block", - "block(%d) >= blocks count(%d) - " + "block("E3FSBLK") >= blocks count(%d) - " "block_group = %d, es == %p ", ret_block, le32_to_cpu(es->s_blocks_count), group_no, es); goto out; @@ -1421,12 +1623,12 @@ allocated: * list of some description. We don't know in advance whether * the caller wants to use it as metadata or data. */ - ext3_debug("allocating block %d. Goal hits %d of %d.\n", + ext3_debug("allocating block %lu. Goal hits %d of %d.\n", ret_block, goal_hits, goal_attempts); spin_lock(sb_bgl_lock(sbi, group_no)); gdp->bg_free_blocks_count = - cpu_to_le16(le16_to_cpu(gdp->bg_free_blocks_count) - num); + cpu_to_le16(le16_to_cpu(gdp->bg_free_blocks_count)-num); spin_unlock(sb_bgl_lock(sbi, group_no)); percpu_counter_mod(&sbi->s_freeblocks_counter, -num); @@ -1461,23 +1663,30 @@ out: return 0; } -int ext3_new_block(handle_t *handle, struct inode *inode, - unsigned long goal, int *errp) +ext3_fsblk_t ext3_new_block(handle_t *handle, struct inode *inode, + ext3_fsblk_t goal, int *errp) { unsigned long count = 1; return ext3_new_blocks(handle, inode, goal, &count, errp); } -unsigned long ext3_count_free_blocks(struct super_block *sb) +/** + * ext3_count_free_blocks() -- count filesystem free blocks + * @sb: superblock + * + * Adds up the number of free blocks from each block group. + */ +ext3_fsblk_t ext3_count_free_blocks(struct super_block *sb) { - unsigned long desc_count; + ext3_fsblk_t desc_count; struct ext3_group_desc *gdp; int i; unsigned long ngroups = EXT3_SB(sb)->s_groups_count; #ifdef EXT3FS_DEBUG struct ext3_super_block *es; - unsigned long bitmap_count, x; + ext3_fsblk_t bitmap_count; + unsigned long x; struct buffer_head *bitmap_bh = NULL; es = EXT3_SB(sb)->s_es; @@ -1502,8 +1711,10 @@ unsigned long ext3_count_free_blocks(struct super_block *sb) bitmap_count += x; } brelse(bitmap_bh); - printk("ext3_count_free_blocks: stored = %u, computed = %lu, %lu\n", - le32_to_cpu(es->s_free_blocks_count), desc_count, bitmap_count); + printk("ext3_count_free_blocks: stored = "E3FSBLK + ", computed = "E3FSBLK", "E3FSBLK"\n", + le32_to_cpu(es->s_free_blocks_count), + desc_count, bitmap_count); return bitmap_count; #else desc_count = 0; @@ -1520,7 +1731,7 @@ unsigned long ext3_count_free_blocks(struct super_block *sb) } static inline int -block_in_use(unsigned long block, struct super_block *sb, unsigned char *map) +block_in_use(ext3_fsblk_t block, struct super_block *sb, unsigned char *map) { return ext3_test_bit ((block - le32_to_cpu(EXT3_SB(sb)->s_es->s_first_data_block)) % diff --git a/fs/ext3/bitmap.c b/fs/ext3/bitmap.c index ce4f82b9e52..b9176eed98d 100644 --- a/fs/ext3/bitmap.c +++ b/fs/ext3/bitmap.c @@ -20,7 +20,7 @@ unsigned long ext3_count_free (struct buffer_head * map, unsigned int numchars) unsigned int i; unsigned long sum = 0; - if (!map) + if (!map) return (0); for (i = 0; i < numchars; i++) sum += nibblemap[map->b_data[i] & 0xf] + diff --git a/fs/ext3/dir.c b/fs/ext3/dir.c index fbb0d4ed07d..429acbb4e06 100644 --- a/fs/ext3/dir.c +++ b/fs/ext3/dir.c @@ -59,7 +59,7 @@ static unsigned char get_dtype(struct super_block *sb, int filetype) return (ext3_filetype_table[filetype]); } - + int ext3_check_dir_entry (const char * function, struct inode * dir, struct ext3_dir_entry_2 * de, @@ -67,7 +67,7 @@ int ext3_check_dir_entry (const char * function, struct inode * dir, unsigned long offset) { const char * error_msg = NULL; - const int rlen = le16_to_cpu(de->rec_len); + const int rlen = le16_to_cpu(de->rec_len); if (rlen < EXT3_DIR_REC_LEN(1)) error_msg = "rec_len is smaller than minimal"; @@ -162,7 +162,7 @@ revalidate: * to make sure. */ if (filp->f_version != inode->i_version) { for (i = 0; i < sb->s_blocksize && i < offset; ) { - de = (struct ext3_dir_entry_2 *) + de = (struct ext3_dir_entry_2 *) (bh->b_data + i); /* It's too expensive to do a full * dirent test each time round this @@ -181,7 +181,7 @@ revalidate: filp->f_version = inode->i_version; } - while (!error && filp->f_pos < inode->i_size + while (!error && filp->f_pos < inode->i_size && offset < sb->s_blocksize) { de = (struct ext3_dir_entry_2 *) (bh->b_data + offset); if (!ext3_check_dir_entry ("ext3_readdir", inode, de, @@ -229,7 +229,7 @@ out: /* * These functions convert from the major/minor hash to an f_pos * value. - * + * * Currently we only use major hash numer. This is unfortunate, but * on 32-bit machines, the same VFS interface is used for lseek and * llseek, so if we use the 64 bit offset, then the 32-bit versions of @@ -250,7 +250,7 @@ out: struct fname { __u32 hash; __u32 minor_hash; - struct rb_node rb_hash; + struct rb_node rb_hash; struct fname *next; __u32 inode; __u8 name_len; @@ -343,10 +343,9 @@ int ext3_htree_store_dirent(struct file *dir_file, __u32 hash, /* Create and allocate the fname structure */ len = sizeof(struct fname) + dirent->name_len + 1; - new_fn = kmalloc(len, GFP_KERNEL); + new_fn = kzalloc(len, GFP_KERNEL); if (!new_fn) return -ENOMEM; - memset(new_fn, 0, len); new_fn->hash = hash; new_fn->minor_hash = minor_hash; new_fn->inode = le32_to_cpu(dirent->inode); @@ -410,7 +409,7 @@ static int call_filldir(struct file * filp, void * dirent, curr_pos = hash2pos(fname->hash, fname->minor_hash); while (fname) { error = filldir(dirent, fname->name, - fname->name_len, curr_pos, + fname->name_len, curr_pos, fname->inode, get_dtype(sb, fname->file_type)); if (error) { @@ -465,7 +464,7 @@ static int ext3_dx_readdir(struct file * filp, /* * Fill the rbtree if we have no more entries, * or the inode has changed since we last read in the - * cached entries. + * cached entries. */ if ((!info->curr_node) || (filp->f_version != inode->i_version)) { diff --git a/fs/ext3/file.c b/fs/ext3/file.c index 1efefb630ea..994efd189f4 100644 --- a/fs/ext3/file.c +++ b/fs/ext3/file.c @@ -100,7 +100,7 @@ ext3_file_write(struct kiocb *iocb, const char __user *buf, size_t count, loff_t force_commit: err = ext3_force_commit(inode->i_sb); - if (err) + if (err) return err; return ret; } diff --git a/fs/ext3/fsync.c b/fs/ext3/fsync.c index 49382a208e0..dd1fd3c0fc0 100644 --- a/fs/ext3/fsync.c +++ b/fs/ext3/fsync.c @@ -8,14 +8,14 @@ * Universite Pierre et Marie Curie (Paris VI) * from * linux/fs/minix/truncate.c Copyright (C) 1991, 1992 Linus Torvalds - * + * * ext3fs fsync primitive * * Big-endian to little-endian byte-swapping/bitmaps by * David S. Miller (davem@caip.rutgers.edu), 1995 - * + * * Removed unnecessary code duplication for little endian machines - * and excessive __inline__s. + * and excessive __inline__s. * Andi Kleen, 1997 * * Major simplications and cleanup - we only need to do the metadata, because diff --git a/fs/ext3/hash.c b/fs/ext3/hash.c index 5a2d1235ead..deeb27b5ba8 100644 --- a/fs/ext3/hash.c +++ b/fs/ext3/hash.c @@ -4,7 +4,7 @@ * Copyright (C) 2002 by Theodore Ts'o * * This file is released under the GPL v2. - * + * * This file may be redistributed under the terms of the GNU Public * License. */ @@ -80,11 +80,11 @@ static void str2hashbuf(const char *msg, int len, __u32 *buf, int num) * Returns the hash of a filename. If len is 0 and name is NULL, then * this function can be used to test whether or not a hash version is * supported. - * + * * The seed is an 4 longword (32 bits) "secret" which can be used to * uniquify a hash. If the seed is all zero's, then some default seed * may be used. - * + * * A particular hash version specifies whether or not the seed is * represented, and whether or not the returned hash is 32 bits or 64 * bits. 32 bit hashes will return 0 for the minor hash. @@ -95,7 +95,7 @@ int ext3fs_dirhash(const char *name, int len, struct dx_hash_info *hinfo) __u32 minor_hash = 0; const char *p; int i; - __u32 in[8], buf[4]; + __u32 in[8], buf[4]; /* Initialize the default seed for the hash checksum functions */ buf[0] = 0x67452301; diff --git a/fs/ext3/ialloc.c b/fs/ext3/ialloc.c index dc826464f31..e45dbd65173 100644 --- a/fs/ext3/ialloc.c +++ b/fs/ext3/ialloc.c @@ -202,7 +202,7 @@ error_return: static int find_group_dir(struct super_block *sb, struct inode *parent) { int ngroups = EXT3_SB(sb)->s_groups_count; - int freei, avefreei; + unsigned int freei, avefreei; struct ext3_group_desc *desc, *best_desc = NULL; struct buffer_head *bh; int group, best_group = -1; @@ -216,7 +216,7 @@ static int find_group_dir(struct super_block *sb, struct inode *parent) continue; if (le16_to_cpu(desc->bg_free_inodes_count) < avefreei) continue; - if (!best_desc || + if (!best_desc || (le16_to_cpu(desc->bg_free_blocks_count) > le16_to_cpu(best_desc->bg_free_blocks_count))) { best_group = group; @@ -226,30 +226,30 @@ static int find_group_dir(struct super_block *sb, struct inode *parent) return best_group; } -/* - * Orlov's allocator for directories. - * +/* + * Orlov's allocator for directories. + * * We always try to spread first-level directories. * - * If there are blockgroups with both free inodes and free blocks counts - * not worse than average we return one with smallest directory count. - * Otherwise we simply return a random group. - * - * For the rest rules look so: - * - * It's OK to put directory into a group unless - * it has too many directories already (max_dirs) or - * it has too few free inodes left (min_inodes) or - * it has too few free blocks left (min_blocks) or - * it's already running too large debt (max_debt). - * Parent's group is prefered, if it doesn't satisfy these - * conditions we search cyclically through the rest. If none - * of the groups look good we just look for a group with more - * free inodes than average (starting at parent's group). - * - * Debt is incremented each time we allocate a directory and decremented - * when we allocate an inode, within 0--255. - */ + * If there are blockgroups with both free inodes and free blocks counts + * not worse than average we return one with smallest directory count. + * Otherwise we simply return a random group. + * + * For the rest rules look so: + * + * It's OK to put directory into a group unless + * it has too many directories already (max_dirs) or + * it has too few free inodes left (min_inodes) or + * it has too few free blocks left (min_blocks) or + * it's already running too large debt (max_debt). + * Parent's group is prefered, if it doesn't satisfy these + * conditions we search cyclically through the rest. If none + * of the groups look good we just look for a group with more + * free inodes than average (starting at parent's group). + * + * Debt is incremented each time we allocate a directory and decremented + * when we allocate an inode, within 0--255. + */ #define INODE_COST 64 #define BLOCK_COST 256 @@ -261,10 +261,12 @@ static int find_group_orlov(struct super_block *sb, struct inode *parent) struct ext3_super_block *es = sbi->s_es; int ngroups = sbi->s_groups_count; int inodes_per_group = EXT3_INODES_PER_GROUP(sb); - int freei, avefreei; - int freeb, avefreeb; - int blocks_per_dir, ndirs; - int max_debt, max_dirs, min_blocks, min_inodes; + unsigned int freei, avefreei; + ext3_fsblk_t freeb, avefreeb; + ext3_fsblk_t blocks_per_dir; + unsigned int ndirs; + int max_debt, max_dirs, min_inodes; + ext3_grpblk_t min_blocks; int group = -1, i; struct ext3_group_desc *desc; struct buffer_head *bh; @@ -307,7 +309,7 @@ static int find_group_orlov(struct super_block *sb, struct inode *parent) min_inodes = avefreei - inodes_per_group / 4; min_blocks = avefreeb - EXT3_BLOCKS_PER_GROUP(sb) / 4; - max_debt = EXT3_BLOCKS_PER_GROUP(sb) / max(blocks_per_dir, BLOCK_COST); + max_debt = EXT3_BLOCKS_PER_GROUP(sb) / max(blocks_per_dir, (ext3_fsblk_t)BLOCK_COST); if (max_debt * INODE_COST > inodes_per_group) max_debt = inodes_per_group / INODE_COST; if (max_debt > 255) @@ -452,7 +454,7 @@ struct inode *ext3_new_inode(handle_t *handle, struct inode * dir, int mode) group = find_group_dir(sb, dir); else group = find_group_orlov(sb, dir); - } else + } else group = find_group_other(sb, dir); err = -ENOSPC; @@ -557,7 +559,6 @@ got: inode->i_ino = ino; /* This is the optimal IO size (for stat), not the fs block size */ - inode->i_blksize = PAGE_SIZE; inode->i_blocks = 0; inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME_SEC; diff --git a/fs/ext3/inode.c b/fs/ext3/inode.c index 2edd7eec88f..dcf4f1dd108 100644 --- a/fs/ext3/inode.c +++ b/fs/ext3/inode.c @@ -13,11 +13,11 @@ * Copyright (C) 1991, 1992 Linus Torvalds * * Goal-directed block allocation by Stephen Tweedie - * (sct@redhat.com), 1993, 1998 + * (sct@redhat.com), 1993, 1998 * Big-endian to little-endian byte-swapping/bitmaps by * David S. Miller (davem@caip.rutgers.edu), 1995 * 64-bit file support on 64-bit platforms by Jakub Jelinek - * (jj@sunsite.ms.mff.cuni.cz) + * (jj@sunsite.ms.mff.cuni.cz) * * Assorted race fixes, rewrite of ext3_get_block() by Al Viro, 2000 */ @@ -55,14 +55,14 @@ static int ext3_inode_is_fast_symlink(struct inode *inode) /* * The ext3 forget function must perform a revoke if we are freeing data * which has been journaled. Metadata (eg. indirect blocks) must be - * revoked in all cases. + * revoked in all cases. * * "bh" may be NULL: a metadata block may have been freed from memory * but there may still be a record of it in the journal, and that record * still needs to be revoked. */ int ext3_forget(handle_t *handle, int is_metadata, struct inode *inode, - struct buffer_head *bh, int blocknr) + struct buffer_head *bh, ext3_fsblk_t blocknr) { int err; @@ -105,7 +105,7 @@ int ext3_forget(handle_t *handle, int is_metadata, struct inode *inode, * Work out how many blocks we need to proceed with the next chunk of a * truncate transaction. */ -static unsigned long blocks_for_truncate(struct inode *inode) +static unsigned long blocks_for_truncate(struct inode *inode) { unsigned long needed; @@ -122,13 +122,13 @@ static unsigned long blocks_for_truncate(struct inode *inode) /* But we need to bound the transaction so we don't overflow the * journal. */ - if (needed > EXT3_MAX_TRANS_DATA) + if (needed > EXT3_MAX_TRANS_DATA) needed = EXT3_MAX_TRANS_DATA; return EXT3_DATA_TRANS_BLOCKS(inode->i_sb) + needed; } -/* +/* * Truncate transactions can be complex and absolutely huge. So we need to * be able to restart the transaction at a conventient checkpoint to make * sure we don't overflow the journal. @@ -136,9 +136,9 @@ static unsigned long blocks_for_truncate(struct inode *inode) * start_transaction gets us a new handle for a truncate transaction, * and extend_transaction tries to extend the existing one a bit. If * extend fails, we need to propagate the failure up and restart the - * transaction in the top-level truncate loop. --sct + * transaction in the top-level truncate loop. --sct */ -static handle_t *start_transaction(struct inode *inode) +static handle_t *start_transaction(struct inode *inode) { handle_t *result; @@ -215,12 +215,12 @@ void ext3_delete_inode (struct inode * inode) ext3_orphan_del(handle, inode); EXT3_I(inode)->i_dtime = get_seconds(); - /* + /* * One subtle ordering requirement: if anything has gone wrong * (transaction abort, IO errors, whatever), then we can still * do these next steps (the fs will already have been marked as * having errors), but we can't free the inode if the mark_dirty - * fails. + * fails. */ if (ext3_mark_inode_dirty(handle, inode)) /* If that failed, just do the required in-core inode clear. */ @@ -398,7 +398,7 @@ no_block: * + if there is a block to the left of our position - allocate near it. * + if pointer will live in indirect block - allocate near that block. * + if pointer will live in inode - allocate in the same - * cylinder group. + * cylinder group. * * In the latter case we colour the starting block by the callers PID to * prevent it from clashing with concurrent allocations for a different inode @@ -407,13 +407,13 @@ no_block: * * Caller must make sure that @ind is valid and will stay that way. */ -static unsigned long ext3_find_near(struct inode *inode, Indirect *ind) +static ext3_fsblk_t ext3_find_near(struct inode *inode, Indirect *ind) { struct ext3_inode_info *ei = EXT3_I(inode); __le32 *start = ind->bh ? (__le32*) ind->bh->b_data : ei->i_data; __le32 *p; - unsigned long bg_start; - unsigned long colour; + ext3_fsblk_t bg_start; + ext3_grpblk_t colour; /* Try to find previous block */ for (p = ind->p - 1; p >= start; p--) { @@ -429,8 +429,7 @@ static unsigned long ext3_find_near(struct inode *inode, Indirect *ind) * It is going to be referred to from the inode itself? OK, just put it * into the same cylinder group then. */ - bg_start = (ei->i_block_group * EXT3_BLOCKS_PER_GROUP(inode->i_sb)) + - le32_to_cpu(EXT3_SB(inode->i_sb)->s_es->s_first_data_block); + bg_start = ext3_group_first_block_no(inode->i_sb, ei->i_block_group); colour = (current->pid % 16) * (EXT3_BLOCKS_PER_GROUP(inode->i_sb) / 16); return bg_start + colour; @@ -448,7 +447,7 @@ static unsigned long ext3_find_near(struct inode *inode, Indirect *ind) * stores it in *@goal and returns zero. */ -static unsigned long ext3_find_goal(struct inode *inode, long block, +static ext3_fsblk_t ext3_find_goal(struct inode *inode, long block, Indirect chain[4], Indirect *partial) { struct ext3_block_alloc_info *block_i; @@ -471,7 +470,7 @@ static unsigned long ext3_find_goal(struct inode *inode, long block, * ext3_blks_to_allocate: Look up the block map and count the number * of direct blocks need to be allocated for the given branch. * - * @branch: chain of indirect blocks + * @branch: chain of indirect blocks * @k: number of blocks need for indirect blocks * @blks: number of data blocks to be mapped. * @blocks_to_boundary: the offset in the indirect block @@ -516,13 +515,13 @@ static int ext3_blks_to_allocate(Indirect *branch, int k, unsigned long blks, * direct blocks */ static int ext3_alloc_blocks(handle_t *handle, struct inode *inode, - unsigned long goal, int indirect_blks, int blks, - unsigned long long new_blocks[4], int *err) + ext3_fsblk_t goal, int indirect_blks, int blks, + ext3_fsblk_t new_blocks[4], int *err) { int target, i; unsigned long count = 0; int index = 0; - unsigned long current_block = 0; + ext3_fsblk_t current_block = 0; int ret = 0; /* @@ -592,7 +591,7 @@ failed_out: * as described above and return 0. */ static int ext3_alloc_branch(handle_t *handle, struct inode *inode, - int indirect_blks, int *blks, unsigned long goal, + int indirect_blks, int *blks, ext3_fsblk_t goal, int *offsets, Indirect *branch) { int blocksize = inode->i_sb->s_blocksize; @@ -600,8 +599,8 @@ static int ext3_alloc_branch(handle_t *handle, struct inode *inode, int err = 0; struct buffer_head *bh; int num; - unsigned long long new_blocks[4]; - unsigned long long current_block; + ext3_fsblk_t new_blocks[4]; + ext3_fsblk_t current_block; num = ext3_alloc_blocks(handle, inode, goal, indirect_blks, *blks, new_blocks, &err); @@ -688,7 +687,7 @@ static int ext3_splice_branch(handle_t *handle, struct inode *inode, int i; int err = 0; struct ext3_block_alloc_info *block_i; - unsigned long current_block; + ext3_fsblk_t current_block; block_i = EXT3_I(inode)->i_block_alloc_info; /* @@ -745,7 +744,7 @@ static int ext3_splice_branch(handle_t *handle, struct inode *inode, jbd_debug(5, "splicing indirect only\n"); BUFFER_TRACE(where->bh, "call ext3_journal_dirty_metadata"); err = ext3_journal_dirty_metadata(handle, where->bh); - if (err) + if (err) goto err_out; } else { /* @@ -795,13 +794,13 @@ int ext3_get_blocks_handle(handle_t *handle, struct inode *inode, int offsets[4]; Indirect chain[4]; Indirect *partial; - unsigned long goal; + ext3_fsblk_t goal; int indirect_blks; int blocks_to_boundary = 0; int depth; struct ext3_inode_info *ei = EXT3_I(inode); int count = 0; - unsigned long first_block = 0; + ext3_fsblk_t first_block = 0; J_ASSERT(handle != NULL || create == 0); @@ -819,7 +818,7 @@ int ext3_get_blocks_handle(handle_t *handle, struct inode *inode, count++; /*map more blocks*/ while (count < maxblocks && count <= blocks_to_boundary) { - unsigned long blk; + ext3_fsblk_t blk; if (!verify_chain(chain, partial)) { /* @@ -926,7 +925,7 @@ int ext3_get_blocks_handle(handle_t *handle, struct inode *inode, set_buffer_new(bh_result); got_it: map_bh(bh_result, inode->i_sb, le32_to_cpu(chain[depth-1].key)); - if (blocks_to_boundary == 0) + if (count > blocks_to_boundary) set_buffer_boundary(bh_result); err = count; /* Clean up and exit */ @@ -1010,11 +1009,14 @@ struct buffer_head *ext3_getblk(handle_t *handle, struct inode *inode, buffer_trace_init(&dummy.b_history); err = ext3_get_blocks_handle(handle, inode, block, 1, &dummy, create, 1); - if (err == 1) { + /* + * ext3_get_blocks_handle() returns number of blocks + * mapped. 0 in case of a HOLE. + */ + if (err > 0) { + if (err > 1) + WARN_ON(1); err = 0; - } else if (err >= 0) { - WARN_ON(1); - err = -EIO; } *errp = err; if (!err && buffer_mapped(&dummy)) { @@ -1096,7 +1098,7 @@ static int walk_page_buffers( handle_t *handle, for ( bh = head, block_start = 0; ret == 0 && (bh != head || !block_start); - block_start = block_end, bh = next) + block_start = block_end, bh = next) { next = bh->b_this_page; block_end = block_start + blocksize; @@ -1135,7 +1137,7 @@ static int walk_page_buffers( handle_t *handle, * So what we do is to rely on the fact that journal_stop/journal_start * will _not_ run commit under these circumstances because handle->h_ref * is elevated. We'll still have enough credits for the tiny quotafile - * write. + * write. */ static int do_journal_get_write_access(handle_t *handle, struct buffer_head *bh) @@ -1159,7 +1161,7 @@ retry: ret = PTR_ERR(handle); goto out; } - if (test_opt(inode->i_sb, NOBH)) + if (test_opt(inode->i_sb, NOBH) && ext3_should_writeback_data(inode)) ret = nobh_prepare_write(page, from, to, ext3_get_block); else ret = block_prepare_write(page, from, to, ext3_get_block); @@ -1245,7 +1247,7 @@ static int ext3_writeback_commit_write(struct file *file, struct page *page, if (new_i_size > EXT3_I(inode)->i_disksize) EXT3_I(inode)->i_disksize = new_i_size; - if (test_opt(inode->i_sb, NOBH)) + if (test_opt(inode->i_sb, NOBH) && ext3_should_writeback_data(inode)) ret = nobh_commit_write(file, page, from, to); else ret = generic_commit_write(file, page, from, to); @@ -1280,7 +1282,7 @@ static int ext3_journalled_commit_write(struct file *file, if (inode->i_size > EXT3_I(inode)->i_disksize) { EXT3_I(inode)->i_disksize = inode->i_size; ret2 = ext3_mark_inode_dirty(handle, inode); - if (!ret) + if (!ret) ret = ret2; } ret2 = ext3_journal_stop(handle); @@ -1289,7 +1291,7 @@ static int ext3_journalled_commit_write(struct file *file, return ret; } -/* +/* * bmap() is special. It gets used by applications such as lilo and by * the swapper to find the on-disk block of a specific piece of data. * @@ -1298,10 +1300,10 @@ static int ext3_journalled_commit_write(struct file *file, * filesystem and enables swap, then they may get a nasty shock when the * data getting swapped to that swapfile suddenly gets overwritten by * the original zero's written out previously to the journal and - * awaiting writeback in the kernel's buffer cache. + * awaiting writeback in the kernel's buffer cache. * * So, if we see any bmap calls here on a modified, data-journaled file, - * take extra steps to flush any blocks which might be in the cache. + * take extra steps to flush any blocks which might be in the cache. */ static sector_t ext3_bmap(struct address_space *mapping, sector_t block) { @@ -1310,16 +1312,16 @@ static sector_t ext3_bmap(struct address_space *mapping, sector_t block) int err; if (EXT3_I(inode)->i_state & EXT3_STATE_JDATA) { - /* + /* * This is a REALLY heavyweight approach, but the use of * bmap on dirty files is expected to be extremely rare: * only if we run lilo or swapon on a freshly made file - * do we expect this to happen. + * do we expect this to happen. * * (bmap requires CAP_SYS_RAWIO so this does not * represent an unprivileged user DOS attack --- we'd be * in trouble if mortal users could trigger this path at - * will.) + * will.) * * NB. EXT3_STATE_JDATA is not set on files other than * regular files. If somebody wants to bmap a directory @@ -1455,7 +1457,7 @@ static int ext3_ordered_writepage(struct page *page, */ /* - * And attach them to the current transaction. But only if + * And attach them to the current transaction. But only if * block_write_full_page() succeeded. Otherwise they are unmapped, * and generally junk. */ @@ -1495,7 +1497,7 @@ static int ext3_writeback_writepage(struct page *page, goto out_fail; } - if (test_opt(inode->i_sb, NOBH)) + 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); @@ -1642,7 +1644,7 @@ static ssize_t ext3_direct_IO(int rw, struct kiocb *iocb, } } - ret = blockdev_direct_IO(rw, iocb, inode, inode->i_sb->s_bdev, iov, + ret = blockdev_direct_IO(rw, iocb, inode, inode->i_sb->s_bdev, iov, offset, nr_segs, ext3_get_block, NULL); @@ -1699,7 +1701,7 @@ static int ext3_journalled_set_page_dirty(struct page *page) return __set_page_dirty_nobuffers(page); } -static struct address_space_operations ext3_ordered_aops = { +static const struct address_space_operations ext3_ordered_aops = { .readpage = ext3_readpage, .readpages = ext3_readpages, .writepage = ext3_ordered_writepage, @@ -1713,7 +1715,7 @@ static struct address_space_operations ext3_ordered_aops = { .migratepage = buffer_migrate_page, }; -static struct address_space_operations ext3_writeback_aops = { +static const struct address_space_operations ext3_writeback_aops = { .readpage = ext3_readpage, .readpages = ext3_readpages, .writepage = ext3_writeback_writepage, @@ -1727,7 +1729,7 @@ static struct address_space_operations ext3_writeback_aops = { .migratepage = buffer_migrate_page, }; -static struct address_space_operations ext3_journalled_aops = { +static const struct address_space_operations ext3_journalled_aops = { .readpage = ext3_readpage, .readpages = ext3_readpages, .writepage = ext3_journalled_writepage, @@ -1759,7 +1761,7 @@ void ext3_set_aops(struct inode *inode) static int ext3_block_truncate_page(handle_t *handle, struct page *page, struct address_space *mapping, loff_t from) { - unsigned long index = from >> PAGE_CACHE_SHIFT; + ext3_fsblk_t index = from >> PAGE_CACHE_SHIFT; unsigned offset = from & (PAGE_CACHE_SIZE-1); unsigned blocksize, iblock, length, pos; struct inode *inode = mapping->host; @@ -1960,7 +1962,7 @@ no_top: * than `count' because there can be holes in there. */ static void ext3_clear_blocks(handle_t *handle, struct inode *inode, - struct buffer_head *bh, unsigned long block_to_free, + struct buffer_head *bh, ext3_fsblk_t block_to_free, unsigned long count, __le32 *first, __le32 *last) { __le32 *p; @@ -2022,12 +2024,12 @@ static void ext3_free_data(handle_t *handle, struct inode *inode, struct buffer_head *this_bh, __le32 *first, __le32 *last) { - unsigned long block_to_free = 0; /* Starting block # of a run */ - unsigned long count = 0; /* Number of blocks in the run */ + ext3_fsblk_t block_to_free = 0; /* Starting block # of a run */ + unsigned long count = 0; /* Number of blocks in the run */ __le32 *block_to_free_p = NULL; /* Pointer into inode/ind corresponding to block_to_free */ - unsigned long nr; /* Current block # */ + ext3_fsblk_t nr; /* Current block # */ __le32 *p; /* Pointer into inode/ind for current block */ int err; @@ -2052,7 +2054,7 @@ static void ext3_free_data(handle_t *handle, struct inode *inode, } else if (nr == block_to_free + count) { count++; } else { - ext3_clear_blocks(handle, inode, this_bh, + ext3_clear_blocks(handle, inode, this_bh, block_to_free, count, block_to_free_p, p); block_to_free = nr; @@ -2089,7 +2091,7 @@ static void ext3_free_branches(handle_t *handle, struct inode *inode, struct buffer_head *parent_bh, __le32 *first, __le32 *last, int depth) { - unsigned long nr; + ext3_fsblk_t nr; __le32 *p; if (is_handle_aborted(handle)) @@ -2113,7 +2115,7 @@ static void ext3_free_branches(handle_t *handle, struct inode *inode, */ if (!bh) { ext3_error(inode->i_sb, "ext3_free_branches", - "Read failure, inode=%ld, block=%ld", + "Read failure, inode=%lu, block="E3FSBLK, inode->i_ino, nr); continue; } @@ -2182,7 +2184,7 @@ static void ext3_free_branches(handle_t *handle, struct inode *inode, *p = 0; BUFFER_TRACE(parent_bh, "call ext3_journal_dirty_metadata"); - ext3_journal_dirty_metadata(handle, + ext3_journal_dirty_metadata(handle, parent_bh); } } @@ -2394,22 +2396,24 @@ out_stop: ext3_journal_stop(handle); } -static unsigned long ext3_get_inode_block(struct super_block *sb, +static ext3_fsblk_t ext3_get_inode_block(struct super_block *sb, unsigned long ino, struct ext3_iloc *iloc) { unsigned long desc, group_desc, block_group; - unsigned long offset, block; + unsigned long offset; + ext3_fsblk_t block; struct buffer_head *bh; struct ext3_group_desc * gdp; - - if ((ino != EXT3_ROOT_INO && ino != EXT3_JOURNAL_INO && - ino != EXT3_RESIZE_INO && ino < EXT3_FIRST_INO(sb)) || - ino > le32_to_cpu(EXT3_SB(sb)->s_es->s_inodes_count)) { - ext3_error(sb, "ext3_get_inode_block", - "bad inode number: %lu", ino); + if (!ext3_valid_inum(sb, ino)) { + /* + * This error is already checked for in namei.c unless we are + * looking at an NFS filehandle, in which case no error + * report is needed + */ return 0; } + block_group = (ino - 1) / EXT3_INODES_PER_GROUP(sb); if (block_group >= EXT3_SB(sb)->s_groups_count) { ext3_error(sb,"ext3_get_inode_block","group >= groups count"); @@ -2448,7 +2452,7 @@ static unsigned long ext3_get_inode_block(struct super_block *sb, static int __ext3_get_inode_loc(struct inode *inode, struct ext3_iloc *iloc, int in_mem) { - unsigned long block; + ext3_fsblk_t block; struct buffer_head *bh; block = ext3_get_inode_block(inode->i_sb, inode->i_ino, iloc); @@ -2459,7 +2463,8 @@ static int __ext3_get_inode_loc(struct inode *inode, if (!bh) { ext3_error (inode->i_sb, "ext3_get_inode_loc", "unable to read inode block - " - "inode=%lu, block=%lu", inode->i_ino, block); + "inode=%lu, block="E3FSBLK, + inode->i_ino, block); return -EIO; } if (!buffer_uptodate(bh)) { @@ -2540,7 +2545,7 @@ make_io: if (!buffer_uptodate(bh)) { ext3_error(inode->i_sb, "ext3_get_inode_loc", "unable to read inode block - " - "inode=%lu, block=%lu", + "inode=%lu, block="E3FSBLK, inode->i_ino, block); brelse(bh); return -EIO; @@ -2627,9 +2632,6 @@ void ext3_read_inode(struct inode * inode) * recovery code: that's fine, we're about to complete * the process of deleting those. */ } - inode->i_blksize = PAGE_SIZE; /* This is the optimal IO size - * (for stat), not the fs block - * size */ inode->i_blocks = le32_to_cpu(raw_inode->i_blocks); ei->i_flags = le32_to_cpu(raw_inode->i_flags); #ifdef EXT3_FRAGMENTS @@ -2699,7 +2701,7 @@ void ext3_read_inode(struct inode * inode) if (raw_inode->i_block[0]) init_special_inode(inode, inode->i_mode, old_decode_dev(le32_to_cpu(raw_inode->i_block[0]))); - else + else init_special_inode(inode, inode->i_mode, new_decode_dev(le32_to_cpu(raw_inode->i_block[1]))); } @@ -2719,8 +2721,8 @@ bad_inode: * * The caller must have write access to iloc->bh. */ -static int ext3_do_update_inode(handle_t *handle, - struct inode *inode, +static int ext3_do_update_inode(handle_t *handle, + struct inode *inode, struct ext3_iloc *iloc) { struct ext3_inode *raw_inode = ext3_raw_inode(iloc); @@ -2895,7 +2897,7 @@ int ext3_write_inode(struct inode *inode, int wait) * commit will leave the blocks being flushed in an unused state on * disk. (On recovery, the inode will get truncated and the blocks will * be freed, so we have a strong guarantee that no future commit will - * leave these blocks visible to the user.) + * leave these blocks visible to the user.) * * Called with inode->sem down. */ @@ -3038,13 +3040,13 @@ int ext3_mark_iloc_dirty(handle_t *handle, return err; } -/* +/* * On success, We end up with an outstanding reference count against - * iloc->bh. This _must_ be cleaned up later. + * iloc->bh. This _must_ be cleaned up later. */ int -ext3_reserve_inode_write(handle_t *handle, struct inode *inode, +ext3_reserve_inode_write(handle_t *handle, struct inode *inode, struct ext3_iloc *iloc) { int err = 0; @@ -3134,7 +3136,7 @@ out: } #if 0 -/* +/* * Bind an inode's backing buffer_head into this transaction, to prevent * it from being flushed to disk early. Unlike * ext3_reserve_inode_write, this leaves behind no bh reference and @@ -3152,7 +3154,7 @@ static int ext3_pin_inode(handle_t *handle, struct inode *inode) BUFFER_TRACE(iloc.bh, "get_write_access"); err = journal_get_write_access(handle, iloc.bh); if (!err) - err = ext3_journal_dirty_metadata(handle, + err = ext3_journal_dirty_metadata(handle, iloc.bh); brelse(iloc.bh); } diff --git a/fs/ext3/ioctl.c b/fs/ext3/ioctl.c index 8c22aa9a7fb..3a6b012d120 100644 --- a/fs/ext3/ioctl.c +++ b/fs/ext3/ioctl.c @@ -204,7 +204,7 @@ flags_err: return 0; } case EXT3_IOC_GROUP_EXTEND: { - unsigned long n_blocks_count; + ext3_fsblk_t n_blocks_count; struct super_block *sb = inode->i_sb; int err; diff --git a/fs/ext3/namei.c b/fs/ext3/namei.c index b8f5cd1e540..85d132c37ee 100644 --- a/fs/ext3/namei.c +++ b/fs/ext3/namei.c @@ -15,13 +15,13 @@ * Big-endian to little-endian byte-swapping/bitmaps by * David S. Miller (davem@caip.rutgers.edu), 1995 * Directory entry file type support and forward compatibility hooks - * for B-tree directories by Theodore Ts'o (tytso@mit.edu), 1998 + * for B-tree directories by Theodore Ts'o (tytso@mit.edu), 1998 * Hash Tree Directory indexing (c) - * Daniel Phillips, 2001 + * Daniel Phillips, 2001 * Hash Tree Directory indexing porting - * Christopher Li, 2002 + * Christopher Li, 2002 * Hash Tree Directory indexing cleanup - * Theodore Ts'o, 2002 + * Theodore Ts'o, 2002 */ #include <linux/fs.h> @@ -76,7 +76,7 @@ static struct buffer_head *ext3_append(handle_t *handle, #ifdef DX_DEBUG #define dxtrace(command) command #else -#define dxtrace(command) +#define dxtrace(command) #endif struct fake_dirent @@ -169,7 +169,7 @@ static struct ext3_dir_entry_2* dx_pack_dirents (char *base, int size); static void dx_insert_block (struct dx_frame *frame, u32 hash, u32 block); static int ext3_htree_next_block(struct inode *dir, __u32 hash, struct dx_frame *frame, - struct dx_frame *frames, + struct dx_frame *frames, __u32 *start_hash); static struct buffer_head * ext3_dx_find_entry(struct dentry *dentry, struct ext3_dir_entry_2 **res_dir, int *err); @@ -250,7 +250,7 @@ static void dx_show_index (char * label, struct dx_entry *entries) } struct stats -{ +{ unsigned names; unsigned space; unsigned bcount; @@ -278,7 +278,7 @@ static struct stats dx_show_leaf(struct dx_hash_info *hinfo, struct ext3_dir_ent ((char *) de - base)); } space += EXT3_DIR_REC_LEN(de->name_len); - names++; + names++; } de = (struct ext3_dir_entry_2 *) ((char *) de + le16_to_cpu(de->rec_len)); } @@ -464,7 +464,7 @@ static void dx_release (struct dx_frame *frames) */ static int ext3_htree_next_block(struct inode *dir, __u32 hash, struct dx_frame *frame, - struct dx_frame *frames, + struct dx_frame *frames, __u32 *start_hash) { struct dx_frame *p; @@ -632,7 +632,7 @@ int ext3_htree_fill_tree(struct file *dir_file, __u32 start_hash, } count += ret; hashval = ~0; - ret = ext3_htree_next_block(dir, HASH_NB_ALWAYS, + ret = ext3_htree_next_block(dir, HASH_NB_ALWAYS, frame, frames, &hashval); *next_hash = hashval; if (ret < 0) { @@ -649,7 +649,7 @@ int ext3_htree_fill_tree(struct file *dir_file, __u32 start_hash, break; } dx_release(frames); - dxtrace(printk("Fill tree: returned %d entries, next hash: %x\n", + dxtrace(printk("Fill tree: returned %d entries, next hash: %x\n", count, *next_hash)); return count; errout: @@ -1000,7 +1000,12 @@ static struct dentry *ext3_lookup(struct inode * dir, struct dentry *dentry, str if (bh) { unsigned long ino = le32_to_cpu(de->inode); brelse (bh); - inode = iget(dir->i_sb, ino); + if (!ext3_valid_inum(dir->i_sb, ino)) { + ext3_error(dir->i_sb, "ext3_lookup", + "bad inode number: %lu", ino); + inode = NULL; + } else + inode = iget(dir->i_sb, ino); if (!inode) return ERR_PTR(-EACCES); @@ -1028,7 +1033,13 @@ struct dentry *ext3_get_parent(struct dentry *child) return ERR_PTR(-ENOENT); ino = le32_to_cpu(de->inode); brelse(bh); - inode = iget(child->d_inode->i_sb, ino); + + if (!ext3_valid_inum(child->d_inode->i_sb, ino)) { + ext3_error(child->d_inode->i_sb, "ext3_get_parent", + "bad inode number: %lu", ino); + inode = NULL; + } else + inode = iget(child->d_inode->i_sb, ino); if (!inode) return ERR_PTR(-EACCES); @@ -1039,7 +1050,7 @@ struct dentry *ext3_get_parent(struct dentry *child) parent = ERR_PTR(-ENOMEM); } return parent; -} +} #define S_SHIFT 12 static unsigned char ext3_type_by_mode[S_IFMT >> S_SHIFT] = { @@ -1187,7 +1198,7 @@ errout: * add_dirent_to_buf will attempt search the directory block for * space. It will return -ENOSPC if no space is available, and -EIO * and -EEXIST if directory entry already exists. - * + * * NOTE! bh is NOT released in the case where ENOSPC is returned. In * all other cases bh is released. */ @@ -1379,7 +1390,6 @@ static int ext3_add_entry (handle_t *handle, struct dentry *dentry, int dx_fallback=0; #endif unsigned blocksize; - unsigned nlen, rlen; u32 block, blocks; sb = dir->i_sb; @@ -1417,8 +1427,7 @@ static int ext3_add_entry (handle_t *handle, struct dentry *dentry, return retval; de = (struct ext3_dir_entry_2 *) bh->b_data; de->inode = 0; - de->rec_len = cpu_to_le16(rlen = blocksize); - nlen = 0; + de->rec_len = cpu_to_le16(blocksize); return add_dirent_to_buf(handle, dentry, inode, de, bh); } @@ -1563,7 +1572,7 @@ cleanup: * ext3_delete_entry deletes a directory entry by merging it with the * previous entry */ -static int ext3_delete_entry (handle_t *handle, +static int ext3_delete_entry (handle_t *handle, struct inode * dir, struct ext3_dir_entry_2 * de_del, struct buffer_head * bh) @@ -1634,12 +1643,12 @@ static int ext3_add_nondir(handle_t *handle, * is so far negative - it has no inode. * * If the create succeeds, we fill in the inode information - * with d_instantiate(). + * with d_instantiate(). */ static int ext3_create (struct inode * dir, struct dentry * dentry, int mode, struct nameidata *nd) { - handle_t *handle; + handle_t *handle; struct inode * inode; int err, retries = 0; @@ -1679,7 +1688,7 @@ static int ext3_mknod (struct inode * dir, struct dentry *dentry, retry: handle = ext3_journal_start(dir, EXT3_DATA_TRANS_BLOCKS(dir->i_sb) + - EXT3_INDEX_EXTRA_TRANS_BLOCKS + 3 + + EXT3_INDEX_EXTRA_TRANS_BLOCKS + 3 + 2*EXT3_QUOTA_INIT_BLOCKS(dir->i_sb)); if (IS_ERR(handle)) return PTR_ERR(handle); @@ -1804,10 +1813,10 @@ static int empty_dir (struct inode * inode) de1 = (struct ext3_dir_entry_2 *) ((char *) de + le16_to_cpu(de->rec_len)); if (le32_to_cpu(de->inode) != inode->i_ino || - !le32_to_cpu(de1->inode) || + !le32_to_cpu(de1->inode) || strcmp (".", de->name) || strcmp ("..", de1->name)) { - ext3_warning (inode->i_sb, "empty_dir", + ext3_warning (inode->i_sb, "empty_dir", "bad directory (dir #%lu) - no `.' or `..'", inode->i_ino); brelse (bh); @@ -1874,7 +1883,7 @@ int ext3_orphan_add(handle_t *handle, struct inode *inode) * being truncated, or files being unlinked. */ /* @@@ FIXME: Observation from aviro: - * I think I can trigger J_ASSERT in ext3_orphan_add(). We block + * I think I can trigger J_ASSERT in ext3_orphan_add(). We block * here (on lock_super()), so race with ext3_link() which might bump * ->i_nlink. For, say it, character device. Not a regular file, * not a directory, not a symlink and ->i_nlink > 0. @@ -1910,8 +1919,8 @@ int ext3_orphan_add(handle_t *handle, struct inode *inode) if (!err) list_add(&EXT3_I(inode)->i_orphan, &EXT3_SB(sb)->s_orphan); - jbd_debug(4, "superblock will point to %ld\n", inode->i_ino); - jbd_debug(4, "orphan inode %ld will point to %d\n", + jbd_debug(4, "superblock will point to %lu\n", inode->i_ino); + jbd_debug(4, "orphan inode %lu will point to %d\n", inode->i_ino, NEXT_ORPHAN(inode)); out_unlock: unlock_super(sb); @@ -2120,7 +2129,7 @@ static int ext3_symlink (struct inode * dir, retry: handle = ext3_journal_start(dir, EXT3_DATA_TRANS_BLOCKS(dir->i_sb) + - EXT3_INDEX_EXTRA_TRANS_BLOCKS + 5 + + EXT3_INDEX_EXTRA_TRANS_BLOCKS + 5 + 2*EXT3_QUOTA_INIT_BLOCKS(dir->i_sb)); if (IS_ERR(handle)) return PTR_ERR(handle); @@ -2218,7 +2227,7 @@ static int ext3_rename (struct inode * old_dir, struct dentry *old_dentry, DQUOT_INIT(new_dentry->d_inode); handle = ext3_journal_start(old_dir, 2 * EXT3_DATA_TRANS_BLOCKS(old_dir->i_sb) + - EXT3_INDEX_EXTRA_TRANS_BLOCKS + 2); + EXT3_INDEX_EXTRA_TRANS_BLOCKS + 2); if (IS_ERR(handle)) return PTR_ERR(handle); @@ -2384,4 +2393,4 @@ struct inode_operations ext3_special_inode_operations = { .removexattr = generic_removexattr, #endif .permission = ext3_permission, -}; +}; diff --git a/fs/ext3/resize.c b/fs/ext3/resize.c index 34b39e9a1e5..b73cba12f79 100644 --- a/fs/ext3/resize.c +++ b/fs/ext3/resize.c @@ -8,7 +8,6 @@ * This could probably be made into a module, because it is not often in use. */ -#include <linux/config.h> #define EXT3FS_DEBUG @@ -28,16 +27,16 @@ static int verify_group_input(struct super_block *sb, { struct ext3_sb_info *sbi = EXT3_SB(sb); struct ext3_super_block *es = sbi->s_es; - unsigned start = le32_to_cpu(es->s_blocks_count); - unsigned end = start + input->blocks_count; + ext3_fsblk_t start = le32_to_cpu(es->s_blocks_count); + ext3_fsblk_t end = start + input->blocks_count; unsigned group = input->group; - unsigned itend = input->inode_table + sbi->s_itb_per_group; + ext3_fsblk_t itend = input->inode_table + sbi->s_itb_per_group; unsigned overhead = ext3_bg_has_super(sb, group) ? (1 + ext3_bg_num_gdb(sb, group) + le16_to_cpu(es->s_reserved_gdt_blocks)) : 0; - unsigned metaend = start + overhead; + ext3_fsblk_t metaend = start + overhead; struct buffer_head *bh = NULL; - int free_blocks_count; + ext3_grpblk_t free_blocks_count; int err = -EINVAL; input->free_blocks_count = free_blocks_count = @@ -64,7 +63,8 @@ static int verify_group_input(struct super_block *sb, ext3_warning(sb, __FUNCTION__, "Bad blocks count %u", input->blocks_count); else if (!(bh = sb_bread(sb, end - 1))) - ext3_warning(sb, __FUNCTION__, "Cannot read last block (%u)", + ext3_warning(sb, __FUNCTION__, + "Cannot read last block ("E3FSBLK")", end - 1); else if (outside(input->block_bitmap, start, end)) ext3_warning(sb, __FUNCTION__, @@ -77,7 +77,7 @@ static int verify_group_input(struct super_block *sb, else if (outside(input->inode_table, start, end) || outside(itend - 1, start, end)) ext3_warning(sb, __FUNCTION__, - "Inode table not in group (blocks %u-%u)", + "Inode table not in group (blocks %u-"E3FSBLK")", input->inode_table, itend - 1); else if (input->inode_bitmap == input->block_bitmap) ext3_warning(sb, __FUNCTION__, @@ -85,24 +85,27 @@ static int verify_group_input(struct super_block *sb, input->block_bitmap); else if (inside(input->block_bitmap, input->inode_table, itend)) ext3_warning(sb, __FUNCTION__, - "Block bitmap (%u) in inode table (%u-%u)", + "Block bitmap (%u) in inode table (%u-"E3FSBLK")", input->block_bitmap, input->inode_table, itend-1); else if (inside(input->inode_bitmap, input->inode_table, itend)) ext3_warning(sb, __FUNCTION__, - "Inode bitmap (%u) in inode table (%u-%u)", + "Inode bitmap (%u) in inode table (%u-"E3FSBLK")", input->inode_bitmap, input->inode_table, itend-1); else if (inside(input->block_bitmap, start, metaend)) ext3_warning(sb, __FUNCTION__, - "Block bitmap (%u) in GDT table (%u-%u)", + "Block bitmap (%u) in GDT table" + " ("E3FSBLK"-"E3FSBLK")", input->block_bitmap, start, metaend - 1); else if (inside(input->inode_bitmap, start, metaend)) ext3_warning(sb, __FUNCTION__, - "Inode bitmap (%u) in GDT table (%u-%u)", + "Inode bitmap (%u) in GDT table" + " ("E3FSBLK"-"E3FSBLK")", input->inode_bitmap, start, metaend - 1); else if (inside(input->inode_table, start, metaend) || inside(itend - 1, start, metaend)) ext3_warning(sb, __FUNCTION__, - "Inode table (%u-%u) overlaps GDT table (%u-%u)", + "Inode table (%u-"E3FSBLK") overlaps" + "GDT table ("E3FSBLK"-"E3FSBLK")", input->inode_table, itend - 1, start, metaend - 1); else err = 0; @@ -112,7 +115,7 @@ static int verify_group_input(struct super_block *sb, } static struct buffer_head *bclean(handle_t *handle, struct super_block *sb, - unsigned long blk) + ext3_fsblk_t blk) { struct buffer_head *bh; int err; @@ -163,15 +166,14 @@ static int setup_new_group_blocks(struct super_block *sb, struct ext3_new_group_data *input) { struct ext3_sb_info *sbi = EXT3_SB(sb); - unsigned long start = input->group * sbi->s_blocks_per_group + - le32_to_cpu(sbi->s_es->s_first_data_block); + ext3_fsblk_t start = ext3_group_first_block_no(sb, input->group); int reserved_gdb = ext3_bg_has_super(sb, input->group) ? le16_to_cpu(sbi->s_es->s_reserved_gdt_blocks) : 0; unsigned long gdblocks = ext3_bg_num_gdb(sb, input->group); struct buffer_head *bh; handle_t *handle; - unsigned long block; - int bit; + ext3_fsblk_t block; + ext3_grpblk_t bit; int i; int err = 0, err2; @@ -328,19 +330,20 @@ static unsigned ext3_list_backups(struct super_block *sb, unsigned *three, static int verify_reserved_gdb(struct super_block *sb, struct buffer_head *primary) { - const unsigned long blk = primary->b_blocknr; + const ext3_fsblk_t blk = primary->b_blocknr; const unsigned long end = EXT3_SB(sb)->s_groups_count; unsigned three = 1; unsigned five = 5; unsigned seven = 7; unsigned grp; - __u32 *p = (__u32 *)primary->b_data; + __le32 *p = (__le32 *)primary->b_data; int gdbackups = 0; while ((grp = ext3_list_backups(sb, &three, &five, &seven)) < end) { if (le32_to_cpu(*p++) != grp * EXT3_BLOCKS_PER_GROUP(sb) + blk){ ext3_warning(sb, __FUNCTION__, - "reserved GDT %ld missing grp %d (%ld)", + "reserved GDT "E3FSBLK + " missing grp %d ("E3FSBLK")", blk, grp, grp * EXT3_BLOCKS_PER_GROUP(sb) + blk); return -EINVAL; @@ -372,12 +375,12 @@ static int add_new_gdb(handle_t *handle, struct inode *inode, struct super_block *sb = inode->i_sb; struct ext3_super_block *es = EXT3_SB(sb)->s_es; unsigned long gdb_num = input->group / EXT3_DESC_PER_BLOCK(sb); - unsigned long gdblock = EXT3_SB(sb)->s_sbh->b_blocknr + 1 + gdb_num; + ext3_fsblk_t gdblock = EXT3_SB(sb)->s_sbh->b_blocknr + 1 + gdb_num; struct buffer_head **o_group_desc, **n_group_desc; struct buffer_head *dind; int gdbackups; struct ext3_iloc iloc; - __u32 *data; + __le32 *data; int err; if (test_opt(sb, DEBUG)) @@ -414,10 +417,10 @@ static int add_new_gdb(handle_t *handle, struct inode *inode, goto exit_bh; } - data = (__u32 *)dind->b_data; + data = (__le32 *)dind->b_data; if (le32_to_cpu(data[gdb_num % EXT3_ADDR_PER_BLOCK(sb)]) != gdblock) { ext3_warning(sb, __FUNCTION__, - "new group %u GDT block %lu not reserved", + "new group %u GDT block "E3FSBLK" not reserved", input->group, gdblock); err = -EINVAL; goto exit_dind; @@ -436,8 +439,8 @@ static int add_new_gdb(handle_t *handle, struct inode *inode, if ((err = ext3_reserve_inode_write(handle, inode, &iloc))) goto exit_dindj; - n_group_desc = (struct buffer_head **)kmalloc((gdb_num + 1) * - sizeof(struct buffer_head *), GFP_KERNEL); + n_group_desc = kmalloc((gdb_num + 1) * sizeof(struct buffer_head *), + GFP_KERNEL); if (!n_group_desc) { err = -ENOMEM; ext3_warning (sb, __FUNCTION__, @@ -515,8 +518,8 @@ static int reserve_backup_gdb(handle_t *handle, struct inode *inode, struct buffer_head **primary; struct buffer_head *dind; struct ext3_iloc iloc; - unsigned long blk; - __u32 *data, *end; + ext3_fsblk_t blk; + __le32 *data, *end; int gdbackups = 0; int res, i; int err; @@ -533,15 +536,17 @@ static int reserve_backup_gdb(handle_t *handle, struct inode *inode, } blk = EXT3_SB(sb)->s_sbh->b_blocknr + 1 + EXT3_SB(sb)->s_gdb_count; - data = (__u32 *)dind->b_data + EXT3_SB(sb)->s_gdb_count; - end = (__u32 *)dind->b_data + EXT3_ADDR_PER_BLOCK(sb); + data = (__le32 *)dind->b_data + EXT3_SB(sb)->s_gdb_count; + end = (__le32 *)dind->b_data + EXT3_ADDR_PER_BLOCK(sb); /* Get each reserved primary GDT block and verify it holds backups */ for (res = 0; res < reserved_gdb; res++, blk++) { if (le32_to_cpu(*data) != blk) { ext3_warning(sb, __FUNCTION__, - "reserved block %lu not at offset %ld", - blk, (long)(data - (__u32 *)dind->b_data)); + "reserved block "E3FSBLK + " not at offset %ld", + blk, + (long)(data - (__le32 *)dind->b_data)); err = -EINVAL; goto exit_bh; } @@ -556,7 +561,7 @@ static int reserve_backup_gdb(handle_t *handle, struct inode *inode, goto exit_bh; } if (++data >= end) - data = (__u32 *)dind->b_data; + data = (__le32 *)dind->b_data; } for (i = 0; i < reserved_gdb; i++) { @@ -580,7 +585,7 @@ static int reserve_backup_gdb(handle_t *handle, struct inode *inode, blk = input->group * EXT3_BLOCKS_PER_GROUP(sb); for (i = 0; i < reserved_gdb; i++) { int err2; - data = (__u32 *)primary[i]->b_data; + data = (__le32 *)primary[i]->b_data; /* printk("reserving backup %lu[%u] = %lu\n", primary[i]->b_blocknr, gdbackups, blk + primary[i]->b_blocknr); */ @@ -685,7 +690,7 @@ exit_err: "can't update backup for group %d (err %d), " "forcing fsck on next reboot", group, err); sbi->s_mount_state &= ~EXT3_VALID_FS; - sbi->s_es->s_state &= ~cpu_to_le16(EXT3_VALID_FS); + sbi->s_es->s_state &= cpu_to_le16(~EXT3_VALID_FS); mark_buffer_dirty(sbi->s_sbh); } } @@ -726,6 +731,18 @@ int ext3_group_add(struct super_block *sb, struct ext3_new_group_data *input) return -EPERM; } + if (le32_to_cpu(es->s_blocks_count) + input->blocks_count < + le32_to_cpu(es->s_blocks_count)) { + ext3_warning(sb, __FUNCTION__, "blocks_count overflow\n"); + return -EINVAL; + } + + if (le32_to_cpu(es->s_inodes_count) + EXT3_INODES_PER_GROUP(sb) < + le32_to_cpu(es->s_inodes_count)) { + ext3_warning(sb, __FUNCTION__, "inodes_count overflow\n"); + return -EINVAL; + } + if (reserved_gdb || gdb_off == 0) { if (!EXT3_HAS_COMPAT_FEATURE(sb, EXT3_FEATURE_COMPAT_RESIZE_INODE)){ @@ -902,15 +919,16 @@ exit_put: * GDT blocks are reserved to grow to the desired size. */ int ext3_group_extend(struct super_block *sb, struct ext3_super_block *es, - unsigned long n_blocks_count) + ext3_fsblk_t n_blocks_count) { - unsigned long o_blocks_count; + ext3_fsblk_t o_blocks_count; unsigned long o_groups_count; - unsigned long last; - int add; + ext3_grpblk_t last; + ext3_grpblk_t add; struct buffer_head * bh; handle_t *handle; - int err, freed_blocks; + int err; + unsigned long freed_blocks; /* We don't need to worry about locking wrt other resizers just * yet: we're going to revalidate es->s_blocks_count after @@ -919,12 +937,22 @@ int ext3_group_extend(struct super_block *sb, struct ext3_super_block *es, o_groups_count = EXT3_SB(sb)->s_groups_count; if (test_opt(sb, DEBUG)) - printk(KERN_DEBUG "EXT3-fs: extending last group from %lu to %lu blocks\n", + printk(KERN_DEBUG "EXT3-fs: extending last group from "E3FSBLK" uto "E3FSBLK" blocks\n", o_blocks_count, n_blocks_count); if (n_blocks_count == 0 || n_blocks_count == o_blocks_count) return 0; + if (n_blocks_count > (sector_t)(~0ULL) >> (sb->s_blocksize_bits - 9)) { + printk(KERN_ERR "EXT3-fs: filesystem on %s:" + " too large to resize to %lu blocks safely\n", + sb->s_id, n_blocks_count); + if (sizeof(sector_t) < 8) + ext3_warning(sb, __FUNCTION__, + "CONFIG_LBD not enabled\n"); + return -EINVAL; + } + if (n_blocks_count < o_blocks_count) { ext3_warning(sb, __FUNCTION__, "can't shrink FS - resize aborted"); @@ -943,12 +971,18 @@ int ext3_group_extend(struct super_block *sb, struct ext3_super_block *es, add = EXT3_BLOCKS_PER_GROUP(sb) - last; + if (o_blocks_count + add < o_blocks_count) { + ext3_warning(sb, __FUNCTION__, "blocks_count overflow"); + return -EINVAL; + } + if (o_blocks_count + add > n_blocks_count) add = n_blocks_count - o_blocks_count; if (o_blocks_count + add < n_blocks_count) ext3_warning(sb, __FUNCTION__, - "will only finish group (%lu blocks, %u new)", + "will only finish group ("E3FSBLK + " blocks, %u new)", o_blocks_count + add, add); /* See if the device is actually as big as what was requested */ @@ -991,10 +1025,10 @@ int ext3_group_extend(struct super_block *sb, struct ext3_super_block *es, ext3_journal_dirty_metadata(handle, EXT3_SB(sb)->s_sbh); sb->s_dirt = 1; unlock_super(sb); - ext3_debug("freeing blocks %ld through %ld\n", o_blocks_count, + ext3_debug("freeing blocks %lu through "E3FSBLK"\n", o_blocks_count, o_blocks_count + add); ext3_free_blocks_sb(handle, sb, o_blocks_count, add, &freed_blocks); - ext3_debug("freed blocks %ld through %ld\n", o_blocks_count, + ext3_debug("freed blocks "E3FSBLK" through "E3FSBLK"\n", o_blocks_count, o_blocks_count + add); if ((err = ext3_journal_stop(handle))) goto exit_put; diff --git a/fs/ext3/super.c b/fs/ext3/super.c index a60cc6ec130..8bfd56ef18c 100644 --- a/fs/ext3/super.c +++ b/fs/ext3/super.c @@ -16,7 +16,6 @@ * David S. Miller (davem@caip.rutgers.edu), 1995 */ -#include <linux/config.h> #include <linux/module.h> #include <linux/string.h> #include <linux/fs.h> @@ -46,7 +45,7 @@ static int ext3_load_journal(struct super_block *, struct ext3_super_block *, unsigned long journal_devnum); static int ext3_create_journal(struct super_block *, struct ext3_super_block *, - int); + unsigned int); static void ext3_commit_super (struct super_block * sb, struct ext3_super_block * es, int sync); @@ -63,13 +62,13 @@ static void ext3_unlockfs(struct super_block *sb); static void ext3_write_super (struct super_block * sb); static void ext3_write_super_lockfs(struct super_block *sb); -/* +/* * Wrappers for journal_start/end. * * The only special thing we need to do here is to make sure that all * journal_end calls result in the superblock being marked dirty, so * that sync() will call the filesystem's write_super callback if - * appropriate. + * appropriate. */ handle_t *ext3_journal_start_sb(struct super_block *sb, int nblocks) { @@ -91,11 +90,11 @@ handle_t *ext3_journal_start_sb(struct super_block *sb, int nblocks) return journal_start(journal, nblocks); } -/* +/* * The only special thing we need to do here is to make sure that all * journal_stop calls result in the superblock being marked dirty, so * that sync() will call the filesystem's write_super callback if - * appropriate. + * appropriate. */ int __ext3_journal_stop(const char *where, handle_t *handle) { @@ -160,20 +159,21 @@ static void ext3_handle_error(struct super_block *sb) if (sb->s_flags & MS_RDONLY) return; - if (test_opt (sb, ERRORS_RO)) { - printk (KERN_CRIT "Remounting filesystem read-only\n"); - sb->s_flags |= MS_RDONLY; - } else { + if (!test_opt (sb, ERRORS_CONT)) { journal_t *journal = EXT3_SB(sb)->s_journal; EXT3_SB(sb)->s_mount_opt |= EXT3_MOUNT_ABORT; if (journal) journal_abort(journal, -EIO); } + if (test_opt (sb, ERRORS_RO)) { + printk (KERN_CRIT "Remounting filesystem read-only\n"); + sb->s_flags |= MS_RDONLY; + } + ext3_commit_super(sb, es, 1); if (test_opt(sb, ERRORS_PANIC)) panic("EXT3-fs (device %s): panic forced after error\n", sb->s_id); - ext3_commit_super(sb, es, 1); } void ext3_error (struct super_block * sb, const char * function, @@ -370,16 +370,16 @@ static void dump_orphan_list(struct super_block *sb, struct ext3_sb_info *sbi) { struct list_head *l; - printk(KERN_ERR "sb orphan head is %d\n", + printk(KERN_ERR "sb orphan head is %d\n", le32_to_cpu(sbi->s_es->s_last_orphan)); printk(KERN_ERR "sb_info orphan list:\n"); list_for_each(l, &sbi->s_orphan) { struct inode *inode = orphan_list_entry(l); printk(KERN_ERR " " - "inode %s:%ld at %p: mode %o, nlink %d, next %d\n", + "inode %s:%lu at %p: mode %o, nlink %d, next %d\n", inode->i_sb->s_id, inode->i_ino, inode, - inode->i_mode, inode->i_nlink, + inode->i_mode, inode->i_nlink, NEXT_ORPHAN(inode)); } } @@ -476,7 +476,7 @@ static void init_once(void * foo, kmem_cache_t * cachep, unsigned long flags) inode_init_once(&ei->vfs_inode); } } - + static int init_inodecache(void) { ext3_inode_cachep = kmem_cache_create("ext3_inode_cache", @@ -491,8 +491,7 @@ static int init_inodecache(void) static void destroy_inodecache(void) { - if (kmem_cache_destroy(ext3_inode_cachep)) - printk(KERN_INFO "ext3_inode_cache: not all structures were freed\n"); + kmem_cache_destroy(ext3_inode_cachep); } static void ext3_clear_inode(struct inode *inode) @@ -555,6 +554,47 @@ static int ext3_show_options(struct seq_file *seq, struct vfsmount *vfs) return 0; } + +static struct dentry *ext3_get_dentry(struct super_block *sb, void *vobjp) +{ + __u32 *objp = vobjp; + unsigned long ino = objp[0]; + __u32 generation = objp[1]; + struct inode *inode; + struct dentry *result; + + if (ino < EXT3_FIRST_INO(sb) && ino != EXT3_ROOT_INO) + return ERR_PTR(-ESTALE); + if (ino > le32_to_cpu(EXT3_SB(sb)->s_es->s_inodes_count)) + return ERR_PTR(-ESTALE); + + /* iget isn't really right if the inode is currently unallocated!! + * + * ext3_read_inode will return a bad_inode if the inode had been + * deleted, so we should be safe. + * + * Currently we don't know the generation for parent directory, so + * a generation of 0 means "accept any" + */ + inode = iget(sb, ino); + if (inode == NULL) + return ERR_PTR(-ENOMEM); + if (is_bad_inode(inode) || + (generation && inode->i_generation != generation)) { + iput(inode); + return ERR_PTR(-ESTALE); + } + /* now to find a dentry. + * If possible, get a well-connected one + */ + result = d_alloc_anon(inode); + if (!result) { + iput(inode); + return ERR_PTR(-ENOMEM); + } + return result; +} + #ifdef CONFIG_QUOTA #define QTYPE2NAME(t) ((t)==USRQUOTA?"user":"group") #define QTYPE2MOPT(on, t) ((t)==USRQUOTA?((on)##USRJQUOTA):((on)##GRPJQUOTA)) @@ -623,6 +663,7 @@ static struct super_operations ext3_sops = { static struct export_operations ext3_export_ops = { .get_parent = ext3_get_parent, + .get_dentry = ext3_get_dentry, }; enum { @@ -630,7 +671,7 @@ enum { Opt_resgid, Opt_resuid, Opt_sb, Opt_err_cont, Opt_err_panic, Opt_err_ro, Opt_nouid32, Opt_nocheck, Opt_debug, Opt_oldalloc, Opt_orlov, Opt_user_xattr, Opt_nouser_xattr, Opt_acl, Opt_noacl, - Opt_reservation, Opt_noreservation, Opt_noload, Opt_nobh, + Opt_reservation, Opt_noreservation, Opt_noload, Opt_nobh, Opt_bh, Opt_commit, Opt_journal_update, Opt_journal_inum, Opt_journal_dev, Opt_abort, Opt_data_journal, Opt_data_ordered, Opt_data_writeback, Opt_usrjquota, Opt_grpjquota, Opt_offusrjquota, Opt_offgrpjquota, @@ -666,6 +707,7 @@ static match_table_t tokens = { {Opt_noreservation, "noreservation"}, {Opt_noload, "noload"}, {Opt_nobh, "nobh"}, + {Opt_bh, "bh"}, {Opt_commit, "commit=%u"}, {Opt_journal_update, "journal=update"}, {Opt_journal_inum, "journal=%u"}, @@ -689,14 +731,15 @@ static match_table_t tokens = { {Opt_resize, "resize"}, }; -static unsigned long get_sb_block(void **data) +static ext3_fsblk_t get_sb_block(void **data) { - unsigned long sb_block; - char *options = (char *) *data; + ext3_fsblk_t sb_block; + char *options = (char *) *data; if (!options || strncmp(options, "sb=", 3) != 0) return 1; /* Default location */ options += 3; + /*todo: use simple_strtoll with >32bit ext3 */ sb_block = simple_strtoul(options, &options, 0); if (*options && *options != ',') { printk("EXT3-fs: Invalid sb specification: %s\n", @@ -710,8 +753,8 @@ static unsigned long get_sb_block(void **data) } static int parse_options (char *options, struct super_block *sb, - unsigned long *inum, unsigned long *journal_devnum, - unsigned long *n_blocks_count, int is_remount) + unsigned int *inum, unsigned long *journal_devnum, + ext3_fsblk_t *n_blocks_count, int is_remount) { struct ext3_sb_info *sbi = EXT3_SB(sb); char * p; @@ -1013,6 +1056,9 @@ clear_qf_name: case Opt_nobh: set_opt(sbi->s_mount_opt, NOBH); break; + case Opt_bh: + clear_opt(sbi->s_mount_opt, NOBH); + break; default: printk (KERN_ERR "EXT3-fs: Unrecognized mount option \"%s\" " @@ -1128,7 +1174,8 @@ static int ext3_setup_super(struct super_block *sb, struct ext3_super_block *es, static int ext3_check_descriptors (struct super_block * sb) { struct ext3_sb_info *sbi = EXT3_SB(sb); - unsigned long block = le32_to_cpu(sbi->s_es->s_first_data_block); + ext3_fsblk_t first_block = le32_to_cpu(sbi->s_es->s_first_data_block); + ext3_fsblk_t last_block; struct ext3_group_desc * gdp = NULL; int desc_block = 0; int i; @@ -1137,12 +1184,17 @@ static int ext3_check_descriptors (struct super_block * sb) for (i = 0; i < sbi->s_groups_count; i++) { + if (i == sbi->s_groups_count - 1) + last_block = le32_to_cpu(sbi->s_es->s_blocks_count) - 1; + else + last_block = first_block + + (EXT3_BLOCKS_PER_GROUP(sb) - 1); + if ((i % EXT3_DESC_PER_BLOCK(sb)) == 0) gdp = (struct ext3_group_desc *) sbi->s_group_desc[desc_block++]->b_data; - if (le32_to_cpu(gdp->bg_block_bitmap) < block || - le32_to_cpu(gdp->bg_block_bitmap) >= - block + EXT3_BLOCKS_PER_GROUP(sb)) + if (le32_to_cpu(gdp->bg_block_bitmap) < first_block || + le32_to_cpu(gdp->bg_block_bitmap) > last_block) { ext3_error (sb, "ext3_check_descriptors", "Block bitmap for group %d" @@ -1151,9 +1203,8 @@ static int ext3_check_descriptors (struct super_block * sb) le32_to_cpu(gdp->bg_block_bitmap)); return 0; } - if (le32_to_cpu(gdp->bg_inode_bitmap) < block || - le32_to_cpu(gdp->bg_inode_bitmap) >= - block + EXT3_BLOCKS_PER_GROUP(sb)) + if (le32_to_cpu(gdp->bg_inode_bitmap) < first_block || + le32_to_cpu(gdp->bg_inode_bitmap) > last_block) { ext3_error (sb, "ext3_check_descriptors", "Inode bitmap for group %d" @@ -1162,9 +1213,9 @@ static int ext3_check_descriptors (struct super_block * sb) le32_to_cpu(gdp->bg_inode_bitmap)); return 0; } - if (le32_to_cpu(gdp->bg_inode_table) < block || - le32_to_cpu(gdp->bg_inode_table) + sbi->s_itb_per_group >= - block + EXT3_BLOCKS_PER_GROUP(sb)) + if (le32_to_cpu(gdp->bg_inode_table) < first_block || + le32_to_cpu(gdp->bg_inode_table) + sbi->s_itb_per_group > + last_block) { ext3_error (sb, "ext3_check_descriptors", "Inode table for group %d" @@ -1173,7 +1224,7 @@ static int ext3_check_descriptors (struct super_block * sb) le32_to_cpu(gdp->bg_inode_table)); return 0; } - block += EXT3_BLOCKS_PER_GROUP(sb); + first_block += EXT3_BLOCKS_PER_GROUP(sb); gdp++; } @@ -1255,17 +1306,17 @@ static void ext3_orphan_cleanup (struct super_block * sb, DQUOT_INIT(inode); if (inode->i_nlink) { printk(KERN_DEBUG - "%s: truncating inode %ld to %Ld bytes\n", + "%s: truncating inode %lu to %Ld bytes\n", __FUNCTION__, inode->i_ino, inode->i_size); - jbd_debug(2, "truncating inode %ld to %Ld bytes\n", + jbd_debug(2, "truncating inode %lu to %Ld bytes\n", inode->i_ino, inode->i_size); ext3_truncate(inode); nr_truncates++; } else { printk(KERN_DEBUG - "%s: deleting unreferenced inode %ld\n", + "%s: deleting unreferenced inode %lu\n", __FUNCTION__, inode->i_ino); - jbd_debug(2, "deleting unreferenced inode %ld\n", + jbd_debug(2, "deleting unreferenced inode %lu\n", inode->i_ino); nr_orphans++; } @@ -1315,15 +1366,14 @@ static loff_t ext3_max_size(int bits) return res; } -static unsigned long descriptor_loc(struct super_block *sb, - unsigned long logic_sb_block, +static ext3_fsblk_t descriptor_loc(struct super_block *sb, + ext3_fsblk_t logic_sb_block, int nr) { struct ext3_sb_info *sbi = EXT3_SB(sb); - unsigned long bg, first_data_block, first_meta_bg; + unsigned long bg, first_meta_bg; int has_super = 0; - first_data_block = le32_to_cpu(sbi->s_es->s_first_data_block); first_meta_bg = le32_to_cpu(sbi->s_es->s_first_meta_bg); if (!EXT3_HAS_INCOMPAT_FEATURE(sb, EXT3_FEATURE_INCOMPAT_META_BG) || @@ -1332,7 +1382,7 @@ static unsigned long descriptor_loc(struct super_block *sb, bg = sbi->s_desc_per_block * nr; if (ext3_bg_has_super(sb, bg)) has_super = 1; - return (first_data_block + has_super + (bg * sbi->s_blocks_per_group)); + return (has_super + ext3_group_first_block_no(sb, bg)); } @@ -1341,11 +1391,11 @@ static int ext3_fill_super (struct super_block *sb, void *data, int silent) struct buffer_head * bh; struct ext3_super_block *es = NULL; struct ext3_sb_info *sbi; - unsigned long block; - unsigned long sb_block = get_sb_block(&data); - unsigned long logic_sb_block; + ext3_fsblk_t block; + ext3_fsblk_t sb_block = get_sb_block(&data); + ext3_fsblk_t logic_sb_block; unsigned long offset = 0; - unsigned long journal_inum = 0; + unsigned int journal_inum = 0; unsigned long journal_devnum = 0; unsigned long def_mount_opts; struct inode *root; @@ -1356,11 +1406,10 @@ static int ext3_fill_super (struct super_block *sb, void *data, int silent) int needs_recovery; __le32 features; - sbi = kmalloc(sizeof(*sbi), GFP_KERNEL); + sbi = kzalloc(sizeof(*sbi), GFP_KERNEL); if (!sbi) return -ENOMEM; sb->s_fs_info = sbi; - memset(sbi, 0, sizeof(*sbi)); sbi->s_mount_opt = 0; sbi->s_resuid = EXT3_DEF_RESUID; sbi->s_resgid = EXT3_DEF_RESGID; @@ -1438,7 +1487,7 @@ static int ext3_fill_super (struct super_block *sb, void *data, int silent) (EXT3_HAS_COMPAT_FEATURE(sb, ~0U) || EXT3_HAS_RO_COMPAT_FEATURE(sb, ~0U) || EXT3_HAS_INCOMPAT_FEATURE(sb, ~0U))) - printk(KERN_WARNING + printk(KERN_WARNING "EXT3-fs warning: feature flags set on rev 0 fs, " "running e2fsck is recommended\n"); /* @@ -1464,7 +1513,7 @@ static int ext3_fill_super (struct super_block *sb, void *data, int silent) if (blocksize < EXT3_MIN_BLOCK_SIZE || blocksize > EXT3_MAX_BLOCK_SIZE) { - printk(KERN_ERR + printk(KERN_ERR "EXT3-fs: Unsupported filesystem blocksize %d on %s.\n", blocksize, sb->s_id); goto failed_mount; @@ -1488,14 +1537,14 @@ static int ext3_fill_super (struct super_block *sb, void *data, int silent) offset = (sb_block * EXT3_MIN_BLOCK_SIZE) % blocksize; bh = sb_bread(sb, logic_sb_block); if (!bh) { - printk(KERN_ERR + printk(KERN_ERR "EXT3-fs: Can't read superblock on 2nd try.\n"); goto failed_mount; } es = (struct ext3_super_block *)(((char *)bh->b_data) + offset); sbi->s_es = es; if (es->s_magic != cpu_to_le16(EXT3_SUPER_MAGIC)) { - printk (KERN_ERR + printk (KERN_ERR "EXT3-fs: Magic mismatch, very weird !\n"); goto failed_mount; } @@ -1565,12 +1614,21 @@ static int ext3_fill_super (struct super_block *sb, void *data, int silent) goto failed_mount; } + if (le32_to_cpu(es->s_blocks_count) > + (sector_t)(~0ULL) >> (sb->s_blocksize_bits - 9)) { + printk(KERN_ERR "EXT3-fs: filesystem on %s:" + " too large to mount safely\n", sb->s_id); + if (sizeof(sector_t) < 8) + printk(KERN_WARNING "EXT3-fs: CONFIG_LBD not " + "enabled\n"); + goto failed_mount; + } + if (EXT3_BLOCKS_PER_GROUP(sb) == 0) goto cantfind_ext3; - sbi->s_groups_count = (le32_to_cpu(es->s_blocks_count) - - le32_to_cpu(es->s_first_data_block) + - EXT3_BLOCKS_PER_GROUP(sb) - 1) / - EXT3_BLOCKS_PER_GROUP(sb); + sbi->s_groups_count = ((le32_to_cpu(es->s_blocks_count) - + le32_to_cpu(es->s_first_data_block) - 1) + / EXT3_BLOCKS_PER_GROUP(sb)) + 1; db_count = (sbi->s_groups_count + EXT3_DESC_PER_BLOCK(sb) - 1) / EXT3_DESC_PER_BLOCK(sb); sbi->s_group_desc = kmalloc(db_count * sizeof (struct buffer_head *), @@ -1593,7 +1651,7 @@ static int ext3_fill_super (struct super_block *sb, void *data, int silent) } } if (!ext3_check_descriptors (sb)) { - printk (KERN_ERR "EXT3-fs: group descriptors corrupted !\n"); + printk(KERN_ERR "EXT3-fs: group descriptors corrupted!\n"); goto failed_mount2; } sbi->s_gdb_count = db_count; @@ -1765,7 +1823,7 @@ out_fail: /* * Setup any per-fs journal parameters now. We'll do this both on * initial mount, once the journal has been initialised but before we've - * done any recovery; and again on any subsequent remount. + * done any recovery; and again on any subsequent remount. */ static void ext3_init_journal_params(struct super_block *sb, journal_t *journal) { @@ -1785,7 +1843,8 @@ static void ext3_init_journal_params(struct super_block *sb, journal_t *journal) spin_unlock(&journal->j_state_lock); } -static journal_t *ext3_get_journal(struct super_block *sb, int journal_inum) +static journal_t *ext3_get_journal(struct super_block *sb, + unsigned int journal_inum) { struct inode *journal_inode; journal_t *journal; @@ -1830,10 +1889,10 @@ static journal_t *ext3_get_dev_journal(struct super_block *sb, { struct buffer_head * bh; journal_t *journal; - int start; - int len; + ext3_fsblk_t start; + ext3_fsblk_t len; int hblock, blocksize; - unsigned long sb_block; + ext3_fsblk_t sb_block; unsigned long offset; struct ext3_super_block * es; struct block_device *bdev; @@ -1920,7 +1979,7 @@ static int ext3_load_journal(struct super_block *sb, unsigned long journal_devnum) { journal_t *journal; - int journal_inum = le32_to_cpu(es->s_journal_inum); + unsigned int journal_inum = le32_to_cpu(es->s_journal_inum); dev_t journal_dev; int err = 0; int really_read_only; @@ -2006,7 +2065,7 @@ static int ext3_load_journal(struct super_block *sb, static int ext3_create_journal(struct super_block * sb, struct ext3_super_block * es, - int journal_inum) + unsigned int journal_inum) { journal_t *journal; @@ -2019,7 +2078,7 @@ static int ext3_create_journal(struct super_block * sb, if (!(journal = ext3_get_journal(sb, journal_inum))) return -EINVAL; - printk(KERN_INFO "EXT3-fs: creating new journal on inode %d\n", + printk(KERN_INFO "EXT3-fs: creating new journal on inode %u\n", journal_inum); if (journal_create(journal)) { @@ -2206,7 +2265,7 @@ static int ext3_remount (struct super_block * sb, int * flags, char * data) { struct ext3_super_block * es; struct ext3_sb_info *sbi = EXT3_SB(sb); - unsigned long n_blocks_count = 0; + ext3_fsblk_t n_blocks_count = 0; unsigned long old_sb_flags; struct ext3_mount_options old_opts; int err; @@ -2287,10 +2346,8 @@ static int ext3_remount (struct super_block * sb, int * flags, char * data) */ ext3_clear_journal_err(sb, es); sbi->s_mount_state = le16_to_cpu(es->s_state); - if ((ret = ext3_group_extend(sb, es, n_blocks_count))) { - err = ret; + if ((err = ext3_group_extend(sb, es, n_blocks_count))) goto restore_opts; - } if (!ext3_setup_super (sb, es, 0)) sb->s_flags &= ~MS_RDONLY; } @@ -2326,7 +2383,7 @@ static int ext3_statfs (struct dentry * dentry, struct kstatfs * buf) struct super_block *sb = dentry->d_sb; struct ext3_sb_info *sbi = EXT3_SB(sb); struct ext3_super_block *es = sbi->s_es; - unsigned long overhead; + ext3_fsblk_t overhead; int i; if (test_opt (sb, MINIX_DF)) @@ -2601,7 +2658,7 @@ static ssize_t ext3_quota_write(struct super_block *sb, int type, struct buffer_head *bh; handle_t *handle = journal_current_handle(); - mutex_lock(&inode->i_mutex); + mutex_lock_nested(&inode->i_mutex, I_MUTEX_QUOTA); while (towrite > 0) { tocopy = sb->s_blocksize - offset < towrite ? sb->s_blocksize - offset : towrite; @@ -2679,7 +2736,7 @@ static int __init init_ext3_fs(void) out: destroy_inodecache(); out1: - exit_ext3_xattr(); + exit_ext3_xattr(); return err; } diff --git a/fs/ext3/xattr.c b/fs/ext3/xattr.c index e8d60bf6b7d..f86f2482f01 100644 --- a/fs/ext3/xattr.c +++ b/fs/ext3/xattr.c @@ -75,7 +75,7 @@ #ifdef EXT3_XATTR_DEBUG # define ea_idebug(inode, f...) do { \ - printk(KERN_DEBUG "inode %s:%ld: ", \ + printk(KERN_DEBUG "inode %s:%lu: ", \ inode->i_sb->s_id, inode->i_ino); \ printk(f); \ printk("\n"); \ @@ -225,7 +225,7 @@ ext3_xattr_block_get(struct inode *inode, int name_index, const char *name, error = -ENODATA; if (!EXT3_I(inode)->i_file_acl) goto cleanup; - ea_idebug(inode, "reading block %d", EXT3_I(inode)->i_file_acl); + ea_idebug(inode, "reading block %u", EXT3_I(inode)->i_file_acl); bh = sb_bread(inode->i_sb, EXT3_I(inode)->i_file_acl); if (!bh) goto cleanup; @@ -233,7 +233,7 @@ ext3_xattr_block_get(struct inode *inode, int name_index, const char *name, atomic_read(&(bh->b_count)), le32_to_cpu(BHDR(bh)->h_refcount)); if (ext3_xattr_check_block(bh)) { bad_block: ext3_error(inode->i_sb, __FUNCTION__, - "inode %ld: bad block %d", inode->i_ino, + "inode %lu: bad block "E3FSBLK, inode->i_ino, EXT3_I(inode)->i_file_acl); error = -EIO; goto cleanup; @@ -366,7 +366,7 @@ ext3_xattr_block_list(struct inode *inode, char *buffer, size_t buffer_size) error = 0; if (!EXT3_I(inode)->i_file_acl) goto cleanup; - ea_idebug(inode, "reading block %d", EXT3_I(inode)->i_file_acl); + ea_idebug(inode, "reading block %u", EXT3_I(inode)->i_file_acl); bh = sb_bread(inode->i_sb, EXT3_I(inode)->i_file_acl); error = -EIO; if (!bh) @@ -375,7 +375,7 @@ ext3_xattr_block_list(struct inode *inode, char *buffer, size_t buffer_size) atomic_read(&(bh->b_count)), le32_to_cpu(BHDR(bh)->h_refcount)); if (ext3_xattr_check_block(bh)) { ext3_error(inode->i_sb, __FUNCTION__, - "inode %ld: bad block %d", inode->i_ino, + "inode %lu: bad block "E3FSBLK, inode->i_ino, EXT3_I(inode)->i_file_acl); error = -EIO; goto cleanup; @@ -647,7 +647,7 @@ ext3_xattr_block_find(struct inode *inode, struct ext3_xattr_info *i, le32_to_cpu(BHDR(bs->bh)->h_refcount)); if (ext3_xattr_check_block(bs->bh)) { ext3_error(sb, __FUNCTION__, - "inode %ld: bad block %d", inode->i_ino, + "inode %lu: bad block "E3FSBLK, inode->i_ino, EXT3_I(inode)->i_file_acl); error = -EIO; goto cleanup; @@ -792,11 +792,12 @@ inserted: get_bh(new_bh); } else { /* We need to allocate a new block */ - int goal = le32_to_cpu( + ext3_fsblk_t goal = le32_to_cpu( EXT3_SB(sb)->s_es->s_first_data_block) + - EXT3_I(inode)->i_block_group * + (ext3_fsblk_t)EXT3_I(inode)->i_block_group * EXT3_BLOCKS_PER_GROUP(sb); - int block = ext3_new_block(handle, inode, goal, &error); + ext3_fsblk_t block = ext3_new_block(handle, inode, + goal, &error); if (error) goto cleanup; ea_idebug(inode, "creating block %d", block); @@ -847,7 +848,7 @@ cleanup_dquot: bad_block: ext3_error(inode->i_sb, __FUNCTION__, - "inode %ld: bad block %d", inode->i_ino, + "inode %lu: bad block "E3FSBLK, inode->i_ino, EXT3_I(inode)->i_file_acl); goto cleanup; @@ -1076,14 +1077,14 @@ ext3_xattr_delete_inode(handle_t *handle, struct inode *inode) bh = sb_bread(inode->i_sb, EXT3_I(inode)->i_file_acl); if (!bh) { ext3_error(inode->i_sb, __FUNCTION__, - "inode %ld: block %d read error", inode->i_ino, + "inode %lu: block "E3FSBLK" read error", inode->i_ino, EXT3_I(inode)->i_file_acl); goto cleanup; } if (BHDR(bh)->h_magic != cpu_to_le32(EXT3_XATTR_MAGIC) || BHDR(bh)->h_blocks != cpu_to_le32(1)) { ext3_error(inode->i_sb, __FUNCTION__, - "inode %ld: bad block %d", inode->i_ino, + "inode %lu: bad block "E3FSBLK, inode->i_ino, EXT3_I(inode)->i_file_acl); goto cleanup; } @@ -1210,11 +1211,11 @@ again: bh = sb_bread(inode->i_sb, ce->e_block); if (!bh) { ext3_error(inode->i_sb, __FUNCTION__, - "inode %ld: block %ld read error", + "inode %lu: block %lu read error", inode->i_ino, (unsigned long) ce->e_block); } else if (le32_to_cpu(BHDR(bh)->h_refcount) >= EXT3_XATTR_REFCOUNT_MAX) { - ea_idebug(inode, "block %ld refcount %d>=%d", + ea_idebug(inode, "block %lu refcount %d>=%d", (unsigned long) ce->e_block, le32_to_cpu(BHDR(bh)->h_refcount), EXT3_XATTR_REFCOUNT_MAX); diff --git a/fs/ext3/xattr.h b/fs/ext3/xattr.h index 2ceae38f3d4..6b1ae1c6182 100644 --- a/fs/ext3/xattr.h +++ b/fs/ext3/xattr.h @@ -6,7 +6,6 @@ (C) 2001 Andreas Gruenbacher, <a.gruenbacher@computer.org> */ -#include <linux/config.h> #include <linux/xattr.h> /* Magic value in attribute blocks */ diff --git a/fs/fat/cache.c b/fs/fat/cache.c index 97b967b84fc..82cc4f59e3b 100644 --- a/fs/fat/cache.c +++ b/fs/fat/cache.c @@ -58,8 +58,7 @@ int __init fat_cache_init(void) void fat_cache_destroy(void) { - if (kmem_cache_destroy(fat_cache_cachep)) - printk(KERN_INFO "fat_cache: not all structures were freed\n"); + kmem_cache_destroy(fat_cache_cachep); } static inline struct fat_cache *fat_cache_alloc(struct inode *inode) diff --git a/fs/fat/file.c b/fs/fat/file.c index 1ee25232e6a..d50fc47169c 100644 --- a/fs/fat/file.c +++ b/fs/fat/file.c @@ -13,6 +13,7 @@ #include <linux/smp_lock.h> #include <linux/buffer_head.h> #include <linux/writeback.h> +#include <linux/blkdev.h> int fat_generic_ioctl(struct inode *inode, struct file *filp, unsigned int cmd, unsigned long arg) @@ -112,6 +113,16 @@ int fat_generic_ioctl(struct inode *inode, struct file *filp, } } +static int fat_file_release(struct inode *inode, struct file *filp) +{ + if ((filp->f_mode & FMODE_WRITE) && + MSDOS_SB(inode->i_sb)->options.flush) { + fat_flush_inodes(inode->i_sb, inode, NULL); + blk_congestion_wait(WRITE, HZ/10); + } + return 0; +} + const struct file_operations fat_file_operations = { .llseek = generic_file_llseek, .read = do_sync_read, @@ -121,6 +132,7 @@ const struct file_operations fat_file_operations = { .aio_read = generic_file_aio_read, .aio_write = generic_file_aio_write, .mmap = generic_file_mmap, + .release = fat_file_release, .ioctl = fat_generic_ioctl, .fsync = file_fsync, .sendfile = generic_file_sendfile, @@ -289,6 +301,7 @@ void fat_truncate(struct inode *inode) lock_kernel(); fat_free(inode, nr_clusters); unlock_kernel(); + fat_flush_inodes(inode->i_sb, inode, NULL); } struct inode_operations fat_file_inode_operations = { diff --git a/fs/fat/inode.c b/fs/fat/inode.c index 7c35d582ec1..045738032a8 100644 --- a/fs/fat/inode.c +++ b/fs/fat/inode.c @@ -24,6 +24,7 @@ #include <linux/vfs.h> #include <linux/parser.h> #include <linux/uio.h> +#include <linux/writeback.h> #include <asm/unaligned.h> #ifndef CONFIG_FAT_DEFAULT_IOCHARSET @@ -50,14 +51,14 @@ static int fat_add_cluster(struct inode *inode) return err; } -static int __fat_get_blocks(struct inode *inode, sector_t iblock, - unsigned long *max_blocks, - struct buffer_head *bh_result, int create) +static inline int __fat_get_block(struct inode *inode, sector_t iblock, + unsigned long *max_blocks, + struct buffer_head *bh_result, int create) { struct super_block *sb = inode->i_sb; struct msdos_sb_info *sbi = MSDOS_SB(sb); - sector_t phys; unsigned long mapped_blocks; + sector_t phys; int err, offset; err = fat_bmap(inode, iblock, &phys, &mapped_blocks); @@ -73,7 +74,7 @@ static int __fat_get_blocks(struct inode *inode, sector_t iblock, if (iblock != MSDOS_I(inode)->mmu_private >> sb->s_blocksize_bits) { fat_fs_panic(sb, "corrupted file size (i_pos %lld, %lld)", - MSDOS_I(inode)->i_pos, MSDOS_I(inode)->mmu_private); + MSDOS_I(inode)->i_pos, MSDOS_I(inode)->mmu_private); return -EIO; } @@ -93,34 +94,29 @@ static int __fat_get_blocks(struct inode *inode, sector_t iblock, err = fat_bmap(inode, iblock, &phys, &mapped_blocks); if (err) return err; + BUG_ON(!phys); BUG_ON(*max_blocks != mapped_blocks); set_buffer_new(bh_result); map_bh(bh_result, sb, phys); + return 0; } -static int fat_get_blocks(struct inode *inode, sector_t iblock, - struct buffer_head *bh_result, int create) +static int fat_get_block(struct inode *inode, sector_t iblock, + struct buffer_head *bh_result, int create) { struct super_block *sb = inode->i_sb; - int err; unsigned long max_blocks = bh_result->b_size >> inode->i_blkbits; + int err; - err = __fat_get_blocks(inode, iblock, &max_blocks, bh_result, create); + err = __fat_get_block(inode, iblock, &max_blocks, bh_result, create); if (err) return err; bh_result->b_size = max_blocks << sb->s_blocksize_bits; return 0; } -static int fat_get_block(struct inode *inode, sector_t iblock, - struct buffer_head *bh_result, int create) -{ - unsigned long max_blocks = 1; - return __fat_get_blocks(inode, iblock, &max_blocks, bh_result, create); -} - static int fat_writepage(struct page *page, struct writeback_control *wbc) { return block_write_full_page(page, fat_get_block, wbc); @@ -188,7 +184,7 @@ static ssize_t fat_direct_IO(int rw, struct kiocb *iocb, * condition of fat_get_block() and ->truncate(). */ return blockdev_direct_IO(rw, iocb, inode, inode->i_sb->s_bdev, iov, - offset, nr_segs, fat_get_blocks, NULL); + offset, nr_segs, fat_get_block, NULL); } static sector_t _fat_bmap(struct address_space *mapping, sector_t block) @@ -196,7 +192,7 @@ static sector_t _fat_bmap(struct address_space *mapping, sector_t block) return generic_block_bmap(mapping, block, fat_get_block); } -static struct address_space_operations fat_aops = { +static const struct address_space_operations fat_aops = { .readpage = fat_readpage, .readpages = fat_readpages, .writepage = fat_writepage, @@ -375,8 +371,6 @@ static int fat_fill_inode(struct inode *inode, struct msdos_dir_entry *de) inode->i_flags |= S_IMMUTABLE; } MSDOS_I(inode)->i_attrs = de->attr & ATTR_UNUSED; - /* this is as close to the truth as we can get ... */ - inode->i_blksize = sbi->cluster_size; inode->i_blocks = ((inode->i_size + (sbi->cluster_size - 1)) & ~((loff_t)sbi->cluster_size - 1)) >> 9; inode->i_mtime.tv_sec = @@ -528,8 +522,7 @@ static int __init fat_init_inodecache(void) static void __exit fat_destroy_inodecache(void) { - if (kmem_cache_destroy(fat_inode_cachep)) - printk(KERN_INFO "fat_inode_cache: not all structures were freed\n"); + kmem_cache_destroy(fat_inode_cachep); } static int fat_remount(struct super_block *sb, int *flags, char *data) @@ -861,7 +854,7 @@ enum { Opt_charset, Opt_shortname_lower, Opt_shortname_win95, Opt_shortname_winnt, Opt_shortname_mixed, Opt_utf8_no, Opt_utf8_yes, Opt_uni_xl_no, Opt_uni_xl_yes, Opt_nonumtail_no, Opt_nonumtail_yes, - Opt_obsolate, Opt_err, + Opt_obsolate, Opt_flush, Opt_err, }; static match_table_t fat_tokens = { @@ -893,7 +886,8 @@ static match_table_t fat_tokens = { {Opt_obsolate, "cvf_format=%20s"}, {Opt_obsolate, "cvf_options=%100s"}, {Opt_obsolate, "posix"}, - {Opt_err, NULL} + {Opt_flush, "flush"}, + {Opt_err, NULL}, }; static match_table_t msdos_tokens = { {Opt_nodots, "nodots"}, @@ -1034,6 +1028,9 @@ static int parse_options(char *options, int is_vfat, int silent, int *debug, return 0; opts->codepage = option; break; + case Opt_flush: + opts->flush = 1; + break; /* msdos specific */ case Opt_dots: @@ -1137,7 +1134,6 @@ static int fat_read_root(struct inode *inode) MSDOS_I(inode)->i_start = 0; inode->i_size = sbi->dir_entries * sizeof(struct msdos_dir_entry); } - inode->i_blksize = sbi->cluster_size; inode->i_blocks = ((inode->i_size + (sbi->cluster_size - 1)) & ~((loff_t)sbi->cluster_size - 1)) >> 9; MSDOS_I(inode)->i_logstart = 0; @@ -1168,11 +1164,10 @@ int fat_fill_super(struct super_block *sb, void *data, int silent, long error; char buf[50]; - sbi = kmalloc(sizeof(struct msdos_sb_info), GFP_KERNEL); + sbi = kzalloc(sizeof(struct msdos_sb_info), GFP_KERNEL); if (!sbi) return -ENOMEM; sb->s_fs_info = sbi; - memset(sbi, 0, sizeof(struct msdos_sb_info)); sb->s_flags |= MS_NODIRATIME; sb->s_magic = MSDOS_SUPER_MAGIC; @@ -1435,6 +1430,56 @@ out_fail: EXPORT_SYMBOL_GPL(fat_fill_super); +/* + * helper function for fat_flush_inodes. This writes both the inode + * and the file data blocks, waiting for in flight data blocks before + * the start of the call. It does not wait for any io started + * during the call + */ +static int writeback_inode(struct inode *inode) +{ + + int ret; + struct address_space *mapping = inode->i_mapping; + struct writeback_control wbc = { + .sync_mode = WB_SYNC_NONE, + .nr_to_write = 0, + }; + /* if we used WB_SYNC_ALL, sync_inode waits for the io for the + * inode to finish. So WB_SYNC_NONE is sent down to sync_inode + * and filemap_fdatawrite is used for the data blocks + */ + ret = sync_inode(inode, &wbc); + if (!ret) + ret = filemap_fdatawrite(mapping); + return ret; +} + +/* + * write data and metadata corresponding to i1 and i2. The io is + * started but we do not wait for any of it to finish. + * + * filemap_flush is used for the block device, so if there is a dirty + * page for a block already in flight, we will not wait and start the + * io over again + */ +int fat_flush_inodes(struct super_block *sb, struct inode *i1, struct inode *i2) +{ + int ret = 0; + if (!MSDOS_SB(sb)->options.flush) + return 0; + if (i1) + ret = writeback_inode(i1); + if (!ret && i2) + ret = writeback_inode(i2); + if (!ret && sb) { + struct address_space *mapping = sb->s_bdev->bd_inode->i_mapping; + ret = filemap_flush(mapping); + } + return ret; +} +EXPORT_SYMBOL_GPL(fat_flush_inodes); + static int __init init_fat_fs(void) { int err; diff --git a/fs/file.c b/fs/file.c index 55f4e702256..8e81775c5dc 100644 --- a/fs/file.c +++ b/fs/file.c @@ -240,13 +240,9 @@ static struct fdtable *alloc_fdtable(int nr) if (!fdt) goto out; - nfds = 8 * L1_CACHE_BYTES; - /* Expand to the max in easy steps */ - while (nfds <= nr) { - nfds = nfds * 2; - if (nfds > NR_OPEN) - nfds = NR_OPEN; - } + nfds = max_t(int, 8 * L1_CACHE_BYTES, roundup_pow_of_two(nr + 1)); + if (nfds > NR_OPEN) + nfds = NR_OPEN; new_openset = alloc_fdset(nfds); new_execset = alloc_fdset(nfds); @@ -277,86 +273,78 @@ static struct fdtable *alloc_fdtable(int nr) } while (nfds <= nr); new_fds = alloc_fd_array(nfds); if (!new_fds) - goto out; + goto out2; fdt->fd = new_fds; fdt->max_fds = nfds; fdt->free_files = NULL; return fdt; +out2: + nfds = fdt->max_fdset; out: - if (new_openset) - free_fdset(new_openset, nfds); - if (new_execset) - free_fdset(new_execset, nfds); + free_fdset(new_openset, nfds); + free_fdset(new_execset, nfds); kfree(fdt); return NULL; } /* - * Expands the file descriptor table - it will allocate a new fdtable and - * both fd array and fdset. It is expected to be called with the - * files_lock held. + * Expand the file descriptor table. + * This function will allocate a new fdtable and both fd array and fdset, of + * the given size. + * Return <0 error code on error; 1 on successful completion. + * The files->file_lock should be held on entry, and will be held on exit. */ static int expand_fdtable(struct files_struct *files, int nr) __releases(files->file_lock) __acquires(files->file_lock) { - int error = 0; - struct fdtable *fdt; - struct fdtable *nfdt = NULL; + struct fdtable *new_fdt, *cur_fdt; spin_unlock(&files->file_lock); - nfdt = alloc_fdtable(nr); - if (!nfdt) { - error = -ENOMEM; - spin_lock(&files->file_lock); - goto out; - } - + new_fdt = alloc_fdtable(nr); spin_lock(&files->file_lock); - fdt = files_fdtable(files); + if (!new_fdt) + return -ENOMEM; /* - * Check again since another task may have expanded the - * fd table while we dropped the lock + * Check again since another task may have expanded the fd table while + * we dropped the lock */ - if (nr >= fdt->max_fds || nr >= fdt->max_fdset) { - copy_fdtable(nfdt, fdt); + cur_fdt = files_fdtable(files); + if (nr >= cur_fdt->max_fds || nr >= cur_fdt->max_fdset) { + /* Continue as planned */ + copy_fdtable(new_fdt, cur_fdt); + rcu_assign_pointer(files->fdt, new_fdt); + free_fdtable(cur_fdt); } else { - /* Somebody expanded while we dropped file_lock */ - spin_unlock(&files->file_lock); - __free_fdtable(nfdt); - spin_lock(&files->file_lock); - goto out; + /* Somebody else expanded, so undo our attempt */ + __free_fdtable(new_fdt); } - rcu_assign_pointer(files->fdt, nfdt); - free_fdtable(fdt); -out: - return error; + return 1; } /* * Expand files. - * Return <0 on error; 0 nothing done; 1 files expanded, we may have blocked. - * Should be called with the files->file_lock spinlock held for write. + * This function will expand the file structures, if the requested size exceeds + * the current capacity and there is room for expansion. + * Return <0 error code on error; 0 when nothing done; 1 when files were + * expanded and execution may have blocked. + * The files->file_lock should be held on entry, and will be held on exit. */ int expand_files(struct files_struct *files, int nr) { - int err, expand = 0; struct fdtable *fdt; fdt = files_fdtable(files); - if (nr >= fdt->max_fdset || nr >= fdt->max_fds) { - if (fdt->max_fdset >= NR_OPEN || - fdt->max_fds >= NR_OPEN || nr >= NR_OPEN) { - err = -EMFILE; - goto out; - } - expand = 1; - if ((err = expand_fdtable(files, nr))) - goto out; - } - err = expand; -out: - return err; + /* Do we need to expand? */ + if (nr < fdt->max_fdset && nr < fdt->max_fds) + return 0; + /* Can we expand? */ + if (fdt->max_fdset >= NR_OPEN || fdt->max_fds >= NR_OPEN || + nr >= NR_OPEN) + return -EMFILE; + + /* All good, so we try */ + return expand_fdtable(files, nr); } static void __devinit fdtable_defer_list_init(int cpu) diff --git a/fs/file_table.c b/fs/file_table.c index 506d5307108..bc35a40417d 100644 --- a/fs/file_table.c +++ b/fs/file_table.c @@ -5,7 +5,6 @@ * Copyright (C) 1997 David S. Miller (davem@caip.rutgers.edu) */ -#include <linux/config.h> #include <linux/string.h> #include <linux/slab.h> #include <linux/file.h> @@ -170,7 +169,7 @@ void fastcall __fput(struct file *file) if (file->f_op && file->f_op->release) file->f_op->release(inode, file); security_file_free(file); - if (unlikely(inode->i_cdev != NULL)) + if (unlikely(S_ISCHR(inode->i_mode) && inode->i_cdev != NULL)) cdev_put(inode->i_cdev); fops_put(file->f_op); if (file->f_mode & FMODE_WRITE) diff --git a/fs/filesystems.c b/fs/filesystems.c index 9f1072836c8..e3fa77c6ed5 100644 --- a/fs/filesystems.c +++ b/fs/filesystems.c @@ -69,8 +69,6 @@ int register_filesystem(struct file_system_type * fs) int res = 0; struct file_system_type ** p; - if (!fs) - return -EINVAL; if (fs->next) return -EBUSY; INIT_LIST_HEAD(&fs->fs_supers); diff --git a/fs/freevxfs/vxfs.h b/fs/freevxfs/vxfs.h index 583bd78086d..c8a92652612 100644 --- a/fs/freevxfs/vxfs.h +++ b/fs/freevxfs/vxfs.h @@ -159,11 +159,11 @@ struct vxfs_sb { * In core superblock filesystem private data for VxFS. */ struct vxfs_sb_info { - struct vxfs_sb *vsi_raw; /* raw (on disk) supeblock */ + struct vxfs_sb *vsi_raw; /* raw (on disk) superblock */ struct buffer_head *vsi_bp; /* buffer for raw superblock*/ struct inode *vsi_fship; /* fileset header inode */ struct inode *vsi_ilist; /* inode list inode */ - struct inode *vsi_stilist; /* structual inode list inode */ + struct inode *vsi_stilist; /* structural inode list inode */ u_long vsi_iext; /* initial inode list */ ino_t vsi_fshino; /* fileset header inode */ daddr_t vsi_oltext; /* OLT extent */ @@ -252,7 +252,7 @@ enum { * Get filesystem private data from VFS inode. */ #define VXFS_INO(ip) \ - ((struct vxfs_inode_info *)(ip)->u.generic_ip) + ((struct vxfs_inode_info *)(ip)->i_private) /* * Get filesystem private data from VFS superblock. diff --git a/fs/freevxfs/vxfs_fshead.c b/fs/freevxfs/vxfs_fshead.c index 6dee109aeea..78948b4b189 100644 --- a/fs/freevxfs/vxfs_fshead.c +++ b/fs/freevxfs/vxfs_fshead.c @@ -112,7 +112,7 @@ vxfs_read_fshead(struct super_block *sbp) vip = vxfs_blkiget(sbp, infp->vsi_iext, infp->vsi_fshino); if (!vip) { - printk(KERN_ERR "vxfs: unabled to read fsh inode\n"); + printk(KERN_ERR "vxfs: unable to read fsh inode\n"); return -EINVAL; } if (!VXFS_ISFSH(vip)) { @@ -129,13 +129,13 @@ vxfs_read_fshead(struct super_block *sbp) infp->vsi_fship = vxfs_get_fake_inode(sbp, vip); if (!infp->vsi_fship) { - printk(KERN_ERR "vxfs: unabled to get fsh inode\n"); + printk(KERN_ERR "vxfs: unable to get fsh inode\n"); goto out_free_fship; } sfp = vxfs_getfsh(infp->vsi_fship, 0); if (!sfp) { - printk(KERN_ERR "vxfs: unabled to get structural fsh\n"); + printk(KERN_ERR "vxfs: unable to get structural fsh\n"); goto out_iput_fship; } @@ -145,7 +145,7 @@ vxfs_read_fshead(struct super_block *sbp) pfp = vxfs_getfsh(infp->vsi_fship, 1); if (!pfp) { - printk(KERN_ERR "vxfs: unabled to get primary fsh\n"); + printk(KERN_ERR "vxfs: unable to get primary fsh\n"); goto out_free_sfp; } @@ -159,7 +159,7 @@ vxfs_read_fshead(struct super_block *sbp) infp->vsi_stilist = vxfs_get_fake_inode(sbp, tip); if (!infp->vsi_stilist) { - printk(KERN_ERR "vxfs: unabled to get structual list inode\n"); + printk(KERN_ERR "vxfs: unable to get structural list inode\n"); kfree(tip); goto out_free_pfp; } @@ -174,7 +174,7 @@ vxfs_read_fshead(struct super_block *sbp) goto out_iput_stilist; infp->vsi_ilist = vxfs_get_fake_inode(sbp, tip); if (!infp->vsi_ilist) { - printk(KERN_ERR "vxfs: unabled to get inode list inode\n"); + printk(KERN_ERR "vxfs: unable to get inode list inode\n"); kfree(tip); goto out_iput_stilist; } diff --git a/fs/freevxfs/vxfs_immed.c b/fs/freevxfs/vxfs_immed.c index 6f5df1700e9..4e25f3fbed8 100644 --- a/fs/freevxfs/vxfs_immed.c +++ b/fs/freevxfs/vxfs_immed.c @@ -56,7 +56,7 @@ struct inode_operations vxfs_immed_symlink_iops = { /* * Adress space operations for immed files and directories. */ -struct address_space_operations vxfs_immed_aops = { +const struct address_space_operations vxfs_immed_aops = { .readpage = vxfs_immed_readpage, }; diff --git a/fs/freevxfs/vxfs_inode.c b/fs/freevxfs/vxfs_inode.c index f544aae9169..4786d51ad3b 100644 --- a/fs/freevxfs/vxfs_inode.c +++ b/fs/freevxfs/vxfs_inode.c @@ -41,8 +41,8 @@ #include "vxfs_extern.h" -extern struct address_space_operations vxfs_aops; -extern struct address_space_operations vxfs_immed_aops; +extern const struct address_space_operations vxfs_aops; +extern const struct address_space_operations vxfs_immed_aops; extern struct inode_operations vxfs_immed_symlink_iops; @@ -239,11 +239,10 @@ vxfs_iinit(struct inode *ip, struct vxfs_inode_info *vip) ip->i_ctime.tv_nsec = 0; ip->i_mtime.tv_nsec = 0; - ip->i_blksize = PAGE_SIZE; ip->i_blocks = vip->vii_blocks; ip->i_generation = vip->vii_gen; - ip->u.generic_ip = (void *)vip; + ip->i_private = vip; } @@ -295,7 +294,7 @@ vxfs_read_inode(struct inode *ip) { struct super_block *sbp = ip->i_sb; struct vxfs_inode_info *vip; - struct address_space_operations *aops; + const struct address_space_operations *aops; ino_t ino = ip->i_ino; if (!(vip = __vxfs_iget(ino, VXFS_SBI(sbp)->vsi_ilist))) @@ -338,5 +337,5 @@ vxfs_read_inode(struct inode *ip) void vxfs_clear_inode(struct inode *ip) { - kmem_cache_free(vxfs_inode_cachep, ip->u.generic_ip); + kmem_cache_free(vxfs_inode_cachep, ip->i_private); } diff --git a/fs/freevxfs/vxfs_lookup.c b/fs/freevxfs/vxfs_lookup.c index 29cce456c7c..43886fa00a2 100644 --- a/fs/freevxfs/vxfs_lookup.c +++ b/fs/freevxfs/vxfs_lookup.c @@ -246,6 +246,8 @@ vxfs_readdir(struct file *fp, void *retp, filldir_t filler) u_long page, npages, block, pblocks, nblocks, offset; loff_t pos; + lock_kernel(); + switch ((long)fp->f_pos) { case 0: if (filler(retp, ".", 1, fp->f_pos, ip->i_ino, DT_DIR) < 0) diff --git a/fs/freevxfs/vxfs_subr.c b/fs/freevxfs/vxfs_subr.c index c1be118fc06..decac62efe5 100644 --- a/fs/freevxfs/vxfs_subr.c +++ b/fs/freevxfs/vxfs_subr.c @@ -42,7 +42,7 @@ static int vxfs_readpage(struct file *, struct page *); static sector_t vxfs_bmap(struct address_space *, sector_t); -struct address_space_operations vxfs_aops = { +const struct address_space_operations vxfs_aops = { .readpage = vxfs_readpage, .bmap = vxfs_bmap, .sync_page = block_sync_page, diff --git a/fs/freevxfs/vxfs_super.c b/fs/freevxfs/vxfs_super.c index b74b791fc23..ac28b0835ff 100644 --- a/fs/freevxfs/vxfs_super.c +++ b/fs/freevxfs/vxfs_super.c @@ -260,12 +260,17 @@ static struct file_system_type vxfs_fs_type = { static int __init vxfs_init(void) { + int rv; + vxfs_inode_cachep = kmem_cache_create("vxfs_inode", sizeof(struct vxfs_inode_info), 0, SLAB_RECLAIM_ACCOUNT|SLAB_MEM_SPREAD, NULL, NULL); - if (vxfs_inode_cachep) - return register_filesystem(&vxfs_fs_type); - return -ENOMEM; + if (!vxfs_inode_cachep) + return -ENOMEM; + rv = register_filesystem(&vxfs_fs_type); + if (rv < 0) + kmem_cache_destroy(vxfs_inode_cachep); + return rv; } static void __exit diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c index 031b27a4bc9..892643dc9af 100644 --- a/fs/fs-writeback.c +++ b/fs/fs-writeback.c @@ -464,8 +464,8 @@ void sync_inodes_sb(struct super_block *sb, int wait) .range_start = 0, .range_end = LLONG_MAX, }; - unsigned long nr_dirty = read_page_state(nr_dirty); - unsigned long nr_unstable = read_page_state(nr_unstable); + unsigned long nr_dirty = global_page_state(NR_FILE_DIRTY); + unsigned long nr_unstable = global_page_state(NR_UNSTABLE_NFS); wbc.nr_to_write = nr_dirty + nr_unstable + (inodes_stat.nr_inodes - inodes_stat.nr_unused) + diff --git a/fs/fuse/Makefile b/fs/fuse/Makefile index c3e1f760cac..72437065f6a 100644 --- a/fs/fuse/Makefile +++ b/fs/fuse/Makefile @@ -4,4 +4,4 @@ obj-$(CONFIG_FUSE_FS) += fuse.o -fuse-objs := dev.o dir.o file.o inode.o +fuse-objs := dev.o dir.o file.o inode.o control.o diff --git a/fs/fuse/control.c b/fs/fuse/control.c new file mode 100644 index 00000000000..79ec1f23d4d --- /dev/null +++ b/fs/fuse/control.c @@ -0,0 +1,218 @@ +/* + FUSE: Filesystem in Userspace + Copyright (C) 2001-2006 Miklos Szeredi <miklos@szeredi.hu> + + This program can be distributed under the terms of the GNU GPL. + See the file COPYING. +*/ + +#include "fuse_i.h" + +#include <linux/init.h> +#include <linux/module.h> + +#define FUSE_CTL_SUPER_MAGIC 0x65735543 + +/* + * This is non-NULL when the single instance of the control filesystem + * exists. Protected by fuse_mutex + */ +static struct super_block *fuse_control_sb; + +static struct fuse_conn *fuse_ctl_file_conn_get(struct file *file) +{ + struct fuse_conn *fc; + mutex_lock(&fuse_mutex); + fc = file->f_dentry->d_inode->i_private; + if (fc) + fc = fuse_conn_get(fc); + mutex_unlock(&fuse_mutex); + return fc; +} + +static ssize_t fuse_conn_abort_write(struct file *file, const char __user *buf, + size_t count, loff_t *ppos) +{ + struct fuse_conn *fc = fuse_ctl_file_conn_get(file); + if (fc) { + fuse_abort_conn(fc); + fuse_conn_put(fc); + } + return count; +} + +static ssize_t fuse_conn_waiting_read(struct file *file, char __user *buf, + size_t len, loff_t *ppos) +{ + char tmp[32]; + size_t size; + + if (!*ppos) { + struct fuse_conn *fc = fuse_ctl_file_conn_get(file); + if (!fc) + return 0; + + file->private_data=(void *)(long)atomic_read(&fc->num_waiting); + fuse_conn_put(fc); + } + size = sprintf(tmp, "%ld\n", (long)file->private_data); + return simple_read_from_buffer(buf, len, ppos, tmp, size); +} + +static const struct file_operations fuse_ctl_abort_ops = { + .open = nonseekable_open, + .write = fuse_conn_abort_write, +}; + +static const struct file_operations fuse_ctl_waiting_ops = { + .open = nonseekable_open, + .read = fuse_conn_waiting_read, +}; + +static struct dentry *fuse_ctl_add_dentry(struct dentry *parent, + struct fuse_conn *fc, + const char *name, + int mode, int nlink, + struct inode_operations *iop, + const struct file_operations *fop) +{ + struct dentry *dentry; + struct inode *inode; + + BUG_ON(fc->ctl_ndents >= FUSE_CTL_NUM_DENTRIES); + dentry = d_alloc_name(parent, name); + if (!dentry) + return NULL; + + fc->ctl_dentry[fc->ctl_ndents++] = dentry; + inode = new_inode(fuse_control_sb); + if (!inode) + return NULL; + + inode->i_mode = mode; + inode->i_uid = fc->user_id; + inode->i_gid = fc->group_id; + inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME; + /* setting ->i_op to NULL is not allowed */ + if (iop) + inode->i_op = iop; + inode->i_fop = fop; + inode->i_nlink = nlink; + inode->i_private = fc; + d_add(dentry, inode); + return dentry; +} + +/* + * Add a connection to the control filesystem (if it exists). Caller + * must hold fuse_mutex + */ +int fuse_ctl_add_conn(struct fuse_conn *fc) +{ + struct dentry *parent; + char name[32]; + + if (!fuse_control_sb) + return 0; + + parent = fuse_control_sb->s_root; + parent->d_inode->i_nlink++; + sprintf(name, "%llu", (unsigned long long) fc->id); + parent = fuse_ctl_add_dentry(parent, fc, name, S_IFDIR | 0500, 2, + &simple_dir_inode_operations, + &simple_dir_operations); + if (!parent) + goto err; + + if (!fuse_ctl_add_dentry(parent, fc, "waiting", S_IFREG | 0400, 1, + NULL, &fuse_ctl_waiting_ops) || + !fuse_ctl_add_dentry(parent, fc, "abort", S_IFREG | 0200, 1, + NULL, &fuse_ctl_abort_ops)) + goto err; + + return 0; + + err: + fuse_ctl_remove_conn(fc); + return -ENOMEM; +} + +/* + * Remove a connection from the control filesystem (if it exists). + * Caller must hold fuse_mutex + */ +void fuse_ctl_remove_conn(struct fuse_conn *fc) +{ + int i; + + if (!fuse_control_sb) + return; + + for (i = fc->ctl_ndents - 1; i >= 0; i--) { + struct dentry *dentry = fc->ctl_dentry[i]; + dentry->d_inode->i_private = NULL; + d_drop(dentry); + dput(dentry); + } + fuse_control_sb->s_root->d_inode->i_nlink--; +} + +static int fuse_ctl_fill_super(struct super_block *sb, void *data, int silent) +{ + struct tree_descr empty_descr = {""}; + struct fuse_conn *fc; + int err; + + err = simple_fill_super(sb, FUSE_CTL_SUPER_MAGIC, &empty_descr); + if (err) + return err; + + mutex_lock(&fuse_mutex); + BUG_ON(fuse_control_sb); + fuse_control_sb = sb; + list_for_each_entry(fc, &fuse_conn_list, entry) { + err = fuse_ctl_add_conn(fc); + if (err) { + fuse_control_sb = NULL; + mutex_unlock(&fuse_mutex); + return err; + } + } + mutex_unlock(&fuse_mutex); + + return 0; +} + +static int fuse_ctl_get_sb(struct file_system_type *fs_type, int flags, + const char *dev_name, void *raw_data, + struct vfsmount *mnt) +{ + return get_sb_single(fs_type, flags, raw_data, + fuse_ctl_fill_super, mnt); +} + +static void fuse_ctl_kill_sb(struct super_block *sb) +{ + mutex_lock(&fuse_mutex); + fuse_control_sb = NULL; + mutex_unlock(&fuse_mutex); + + kill_litter_super(sb); +} + +static struct file_system_type fuse_ctl_fs_type = { + .owner = THIS_MODULE, + .name = "fusectl", + .get_sb = fuse_ctl_get_sb, + .kill_sb = fuse_ctl_kill_sb, +}; + +int __init fuse_ctl_init(void) +{ + return register_filesystem(&fuse_ctl_fs_type); +} + +void fuse_ctl_cleanup(void) +{ + unregister_filesystem(&fuse_ctl_fs_type); +} diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index 104a62dadb9..4fc557c40cc 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -34,6 +34,7 @@ static void fuse_request_init(struct fuse_req *req) { memset(req, 0, sizeof(*req)); INIT_LIST_HEAD(&req->list); + INIT_LIST_HEAD(&req->intr_entry); init_waitqueue_head(&req->waitq); atomic_set(&req->count, 1); } @@ -64,18 +65,6 @@ static void restore_sigs(sigset_t *oldset) sigprocmask(SIG_SETMASK, oldset, NULL); } -/* - * Reset request, so that it can be reused - * - * The caller must be _very_ careful to make sure, that it is holding - * the only reference to req - */ -void fuse_reset_request(struct fuse_req *req) -{ - BUG_ON(atomic_read(&req->count) != 1); - fuse_request_init(req); -} - static void __fuse_get_request(struct fuse_req *req) { atomic_inc(&req->count); @@ -88,6 +77,13 @@ static void __fuse_put_request(struct fuse_req *req) atomic_dec(&req->count); } +static void fuse_req_init_context(struct fuse_req *req) +{ + req->in.h.uid = current->fsuid; + req->in.h.gid = current->fsgid; + req->in.h.pid = current->pid; +} + struct fuse_req *fuse_get_req(struct fuse_conn *fc) { struct fuse_req *req; @@ -103,14 +99,16 @@ struct fuse_req *fuse_get_req(struct fuse_conn *fc) if (intr) goto out; + err = -ENOTCONN; + if (!fc->connected) + goto out; + req = fuse_request_alloc(); err = -ENOMEM; if (!req) goto out; - req->in.h.uid = current->fsuid; - req->in.h.gid = current->fsgid; - req->in.h.pid = current->pid; + fuse_req_init_context(req); req->waiting = 1; return req; @@ -119,142 +117,184 @@ struct fuse_req *fuse_get_req(struct fuse_conn *fc) return ERR_PTR(err); } -void fuse_put_request(struct fuse_conn *fc, struct fuse_req *req) +/* + * Return request in fuse_file->reserved_req. However that may + * currently be in use. If that is the case, wait for it to become + * available. + */ +static struct fuse_req *get_reserved_req(struct fuse_conn *fc, + struct file *file) { - if (atomic_dec_and_test(&req->count)) { - if (req->waiting) - atomic_dec(&fc->num_waiting); - fuse_request_free(req); - } + struct fuse_req *req = NULL; + struct fuse_file *ff = file->private_data; + + do { + wait_event(fc->blocked_waitq, ff->reserved_req); + spin_lock(&fc->lock); + if (ff->reserved_req) { + req = ff->reserved_req; + ff->reserved_req = NULL; + get_file(file); + req->stolen_file = file; + } + spin_unlock(&fc->lock); + } while (!req); + + return req; } /* - * Called with sbput_sem held for read (request_end) or write - * (fuse_put_super). By the time fuse_put_super() is finished, all - * inodes belonging to background requests must be released, so the - * iputs have to be done within the locked region. + * Put stolen request back into fuse_file->reserved_req */ -void fuse_release_background(struct fuse_conn *fc, struct fuse_req *req) +static void put_reserved_req(struct fuse_conn *fc, struct fuse_req *req) { - iput(req->inode); - iput(req->inode2); + struct file *file = req->stolen_file; + struct fuse_file *ff = file->private_data; + spin_lock(&fc->lock); - list_del(&req->bg_entry); - if (fc->num_background == FUSE_MAX_BACKGROUND) { - fc->blocked = 0; - wake_up_all(&fc->blocked_waitq); - } - fc->num_background--; + fuse_request_init(req); + BUG_ON(ff->reserved_req); + ff->reserved_req = req; + wake_up(&fc->blocked_waitq); spin_unlock(&fc->lock); + fput(file); } /* - * This function is called when a request is finished. Either a reply - * has arrived or it was interrupted (and not yet sent) or some error - * occurred during communication with userspace, or the device file - * was closed. In case of a background request the reference to the - * stored objects are released. The requester thread is woken up (if - * still waiting), the 'end' callback is called if given, else the - * reference to the request is released + * Gets a requests for a file operation, always succeeds * - * Releasing extra reference for foreground requests must be done - * within the same locked region as setting state to finished. This - * is because fuse_reset_request() may be called after request is - * finished and it must be the sole possessor. If request is - * interrupted and put in the background, it will return with an error - * and hence never be reset and reused. + * This is used for sending the FLUSH request, which must get to + * userspace, due to POSIX locks which may need to be unlocked. * - * Called with fc->lock, unlocks it + * If allocation fails due to OOM, use the reserved request in + * fuse_file. + * + * This is very unlikely to deadlock accidentally, since the + * filesystem should not have it's own file open. If deadlock is + * intentional, it can still be broken by "aborting" the filesystem. */ -static void request_end(struct fuse_conn *fc, struct fuse_req *req) +struct fuse_req *fuse_get_req_nofail(struct fuse_conn *fc, struct file *file) { - list_del(&req->list); - req->state = FUSE_REQ_FINISHED; - if (!req->background) { - spin_unlock(&fc->lock); - wake_up(&req->waitq); - fuse_put_request(fc, req); - } else { - void (*end) (struct fuse_conn *, struct fuse_req *) = req->end; - req->end = NULL; - spin_unlock(&fc->lock); - down_read(&fc->sbput_sem); - if (fc->mounted) - fuse_release_background(fc, req); - up_read(&fc->sbput_sem); + struct fuse_req *req; - /* fput must go outside sbput_sem, otherwise it can deadlock */ - if (req->file) - fput(req->file); + atomic_inc(&fc->num_waiting); + wait_event(fc->blocked_waitq, !fc->blocked); + req = fuse_request_alloc(); + if (!req) + req = get_reserved_req(fc, file); - if (end) - end(fc, req); + fuse_req_init_context(req); + req->waiting = 1; + return req; +} + +void fuse_put_request(struct fuse_conn *fc, struct fuse_req *req) +{ + if (atomic_dec_and_test(&req->count)) { + if (req->waiting) + atomic_dec(&fc->num_waiting); + + if (req->stolen_file) + put_reserved_req(fc, req); else - fuse_put_request(fc, req); + fuse_request_free(req); } } /* - * Unfortunately request interruption not just solves the deadlock - * problem, it causes problems too. These stem from the fact, that an - * interrupted request is continued to be processed in userspace, - * while all the locks and object references (inode and file) held - * during the operation are released. - * - * To release the locks is exactly why there's a need to interrupt the - * request, so there's not a lot that can be done about this, except - * introduce additional locking in userspace. - * - * More important is to keep inode and file references until userspace - * has replied, otherwise FORGET and RELEASE could be sent while the - * inode/file is still used by the filesystem. - * - * For this reason the concept of "background" request is introduced. - * An interrupted request is backgrounded if it has been already sent - * to userspace. Backgrounding involves getting an extra reference to - * inode(s) or file used in the request, and adding the request to - * fc->background list. When a reply is received for a background - * request, the object references are released, and the request is - * removed from the list. If the filesystem is unmounted while there - * are still background requests, the list is walked and references - * are released as if a reply was received. + * This function is called when a request is finished. Either a reply + * has arrived or it was aborted (and not yet sent) or some error + * occurred during communication with userspace, or the device file + * was closed. The requester thread is woken up (if still waiting), + * the 'end' callback is called if given, else the reference to the + * request is released * - * There's one more use for a background request. The RELEASE message is - * always sent as background, since it doesn't return an error or - * data. + * Called with fc->lock, unlocks it */ -static void background_request(struct fuse_conn *fc, struct fuse_req *req) -{ - req->background = 1; - list_add(&req->bg_entry, &fc->background); - fc->num_background++; - if (fc->num_background == FUSE_MAX_BACKGROUND) - fc->blocked = 1; - if (req->inode) - req->inode = igrab(req->inode); - if (req->inode2) - req->inode2 = igrab(req->inode2); +static void request_end(struct fuse_conn *fc, struct fuse_req *req) + __releases(fc->lock) +{ + void (*end) (struct fuse_conn *, struct fuse_req *) = req->end; + req->end = NULL; + list_del(&req->list); + list_del(&req->intr_entry); + req->state = FUSE_REQ_FINISHED; + if (req->background) { + if (fc->num_background == FUSE_MAX_BACKGROUND) { + fc->blocked = 0; + wake_up_all(&fc->blocked_waitq); + } + fc->num_background--; + } + spin_unlock(&fc->lock); + dput(req->dentry); + mntput(req->vfsmount); if (req->file) - get_file(req->file); + fput(req->file); + wake_up(&req->waitq); + if (end) + end(fc, req); + else + fuse_put_request(fc, req); } -/* Called with fc->lock held. Releases, and then reacquires it. */ -static void request_wait_answer(struct fuse_conn *fc, struct fuse_req *req) +static void wait_answer_interruptible(struct fuse_conn *fc, + struct fuse_req *req) { - sigset_t oldset; + if (signal_pending(current)) + return; spin_unlock(&fc->lock); - block_sigs(&oldset); wait_event_interruptible(req->waitq, req->state == FUSE_REQ_FINISHED); - restore_sigs(&oldset); spin_lock(&fc->lock); - if (req->state == FUSE_REQ_FINISHED && !req->interrupted) - return; +} + +static void queue_interrupt(struct fuse_conn *fc, struct fuse_req *req) +{ + list_add_tail(&req->intr_entry, &fc->interrupts); + wake_up(&fc->waitq); + kill_fasync(&fc->fasync, SIGIO, POLL_IN); +} + +/* Called with fc->lock held. Releases, and then reacquires it. */ +static void request_wait_answer(struct fuse_conn *fc, struct fuse_req *req) +{ + if (!fc->no_interrupt) { + /* Any signal may interrupt this */ + wait_answer_interruptible(fc, req); + + if (req->aborted) + goto aborted; + if (req->state == FUSE_REQ_FINISHED) + return; - if (!req->interrupted) { - req->out.h.error = -EINTR; req->interrupted = 1; + if (req->state == FUSE_REQ_SENT) + queue_interrupt(fc, req); + } + + if (req->force) { + spin_unlock(&fc->lock); + wait_event(req->waitq, req->state == FUSE_REQ_FINISHED); + spin_lock(&fc->lock); + } else { + sigset_t oldset; + + /* Only fatal signals may interrupt this */ + block_sigs(&oldset); + wait_answer_interruptible(fc, req); + restore_sigs(&oldset); } + + if (req->aborted) + goto aborted; + if (req->state == FUSE_REQ_FINISHED) + return; + + req->out.h.error = -EINTR; + req->aborted = 1; + + aborted: if (req->locked) { /* This is uninterruptible sleep, because data is being copied to/from the buffers of req. During @@ -268,8 +308,11 @@ static void request_wait_answer(struct fuse_conn *fc, struct fuse_req *req) if (req->state == FUSE_REQ_PENDING) { list_del(&req->list); __fuse_put_request(req); - } else if (req->state == FUSE_REQ_SENT) - background_request(fc, req); + } else if (req->state == FUSE_REQ_SENT) { + spin_unlock(&fc->lock); + wait_event(req->waitq, req->state == FUSE_REQ_FINISHED); + spin_lock(&fc->lock); + } } static unsigned len_args(unsigned numargs, struct fuse_arg *args) @@ -283,13 +326,19 @@ static unsigned len_args(unsigned numargs, struct fuse_arg *args) return nbytes; } +static u64 fuse_get_unique(struct fuse_conn *fc) + { + fc->reqctr++; + /* zero is special */ + if (fc->reqctr == 0) + fc->reqctr = 1; + + return fc->reqctr; +} + static void queue_request(struct fuse_conn *fc, struct fuse_req *req) { - fc->reqctr++; - /* zero is special */ - if (fc->reqctr == 0) - fc->reqctr = 1; - req->in.h.unique = fc->reqctr; + 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); @@ -302,9 +351,6 @@ static void queue_request(struct fuse_conn *fc, struct fuse_req *req) kill_fasync(&fc->fasync, SIGIO, POLL_IN); } -/* - * This can only be interrupted by a SIGKILL - */ void request_send(struct fuse_conn *fc, struct fuse_req *req) { req->isreply = 1; @@ -327,8 +373,12 @@ void request_send(struct fuse_conn *fc, struct fuse_req *req) static void request_send_nowait(struct fuse_conn *fc, struct fuse_req *req) { spin_lock(&fc->lock); - background_request(fc, req); if (fc->connected) { + req->background = 1; + fc->num_background++; + if (fc->num_background == FUSE_MAX_BACKGROUND) + fc->blocked = 1; + queue_request(fc, req); spin_unlock(&fc->lock); } else { @@ -352,14 +402,14 @@ void request_send_background(struct fuse_conn *fc, struct fuse_req *req) /* * Lock the request. Up to the next unlock_request() there mustn't be * anything that could cause a page-fault. If the request was already - * interrupted bail out. + * aborted bail out. */ static int lock_request(struct fuse_conn *fc, struct fuse_req *req) { int err = 0; if (req) { spin_lock(&fc->lock); - if (req->interrupted) + if (req->aborted) err = -ENOENT; else req->locked = 1; @@ -369,7 +419,7 @@ static int lock_request(struct fuse_conn *fc, struct fuse_req *req) } /* - * Unlock request. If it was interrupted during being locked, the + * Unlock request. If it was aborted during being locked, the * requester thread is currently waiting for it to be unlocked, so * wake it up. */ @@ -378,7 +428,7 @@ static void unlock_request(struct fuse_conn *fc, struct fuse_req *req) if (req) { spin_lock(&fc->lock); req->locked = 0; - if (req->interrupted) + if (req->aborted) wake_up(&req->waitq); spin_unlock(&fc->lock); } @@ -557,13 +607,18 @@ static int fuse_copy_args(struct fuse_copy_state *cs, unsigned numargs, return err; } +static int request_pending(struct fuse_conn *fc) +{ + return !list_empty(&fc->pending) || !list_empty(&fc->interrupts); +} + /* Wait until a request is available on the pending list */ static void request_wait(struct fuse_conn *fc) { DECLARE_WAITQUEUE(wait, current); add_wait_queue_exclusive(&fc->waitq, &wait); - while (fc->connected && list_empty(&fc->pending)) { + while (fc->connected && !request_pending(fc)) { set_current_state(TASK_INTERRUPTIBLE); if (signal_pending(current)) break; @@ -577,11 +632,51 @@ static void request_wait(struct fuse_conn *fc) } /* + * Transfer an interrupt request to userspace + * + * Unlike other requests this is assembled on demand, without a need + * to allocate a separate fuse_req structure. + * + * Called with fc->lock held, releases it + */ +static int fuse_read_interrupt(struct fuse_conn *fc, struct fuse_req *req, + const struct iovec *iov, unsigned long nr_segs) + __releases(fc->lock) +{ + struct fuse_copy_state cs; + struct fuse_in_header ih; + struct fuse_interrupt_in arg; + unsigned reqsize = sizeof(ih) + sizeof(arg); + int err; + + list_del_init(&req->intr_entry); + req->intr_unique = fuse_get_unique(fc); + memset(&ih, 0, sizeof(ih)); + memset(&arg, 0, sizeof(arg)); + ih.len = reqsize; + ih.opcode = FUSE_INTERRUPT; + ih.unique = req->intr_unique; + arg.unique = req->in.h.unique; + + spin_unlock(&fc->lock); + if (iov_length(iov, nr_segs) < reqsize) + return -EINVAL; + + fuse_copy_init(&cs, fc, 1, NULL, iov, nr_segs); + err = fuse_copy_one(&cs, &ih, sizeof(ih)); + if (!err) + err = fuse_copy_one(&cs, &arg, sizeof(arg)); + fuse_copy_finish(&cs); + + return err ? err : reqsize; +} + +/* * Read a single request into the userspace filesystem's buffer. This * function waits until a request is available, then removes it from * the pending list and copies request data to userspace buffer. If - * no reply is needed (FORGET) or request has been interrupted or - * there was an error during the copying then it's finished by calling + * no reply is needed (FORGET) or request has been aborted or there + * was an error during the copying then it's finished by calling * request_end(). Otherwise add it to the processing list, and set * the 'sent' flag. */ @@ -601,7 +696,7 @@ static ssize_t fuse_dev_readv(struct file *file, const struct iovec *iov, spin_lock(&fc->lock); err = -EAGAIN; if ((file->f_flags & O_NONBLOCK) && fc->connected && - list_empty(&fc->pending)) + !request_pending(fc)) goto err_unlock; request_wait(fc); @@ -609,9 +704,15 @@ static ssize_t fuse_dev_readv(struct file *file, const struct iovec *iov, if (!fc->connected) goto err_unlock; err = -ERESTARTSYS; - if (list_empty(&fc->pending)) + if (!request_pending(fc)) goto err_unlock; + if (!list_empty(&fc->interrupts)) { + req = list_entry(fc->interrupts.next, struct fuse_req, + intr_entry); + return fuse_read_interrupt(fc, req, iov, nr_segs); + } + req = list_entry(fc->pending.next, struct fuse_req, list); req->state = FUSE_REQ_READING; list_move(&req->list, &fc->io); @@ -636,10 +737,10 @@ static ssize_t fuse_dev_readv(struct file *file, const struct iovec *iov, fuse_copy_finish(&cs); spin_lock(&fc->lock); req->locked = 0; - if (!err && req->interrupted) + if (!err && req->aborted) err = -ENOENT; if (err) { - if (!req->interrupted) + if (!req->aborted) req->out.h.error = -EIO; request_end(fc, req); return err; @@ -649,6 +750,8 @@ static ssize_t fuse_dev_readv(struct file *file, const struct iovec *iov, else { req->state = FUSE_REQ_SENT; list_move_tail(&req->list, &fc->processing); + if (req->interrupted) + queue_interrupt(fc, req); spin_unlock(&fc->lock); } return reqsize; @@ -675,7 +778,7 @@ static struct fuse_req *request_find(struct fuse_conn *fc, u64 unique) list_for_each(entry, &fc->processing) { struct fuse_req *req; req = list_entry(entry, struct fuse_req, list); - if (req->in.h.unique == unique) + if (req->in.h.unique == unique || req->intr_unique == unique) return req; } return NULL; @@ -741,17 +844,33 @@ static ssize_t fuse_dev_writev(struct file *file, const struct iovec *iov, goto err_unlock; req = request_find(fc, oh.unique); - err = -EINVAL; if (!req) goto err_unlock; - if (req->interrupted) { + if (req->aborted) { spin_unlock(&fc->lock); fuse_copy_finish(&cs); spin_lock(&fc->lock); request_end(fc, req); return -ENOENT; } + /* Is it an interrupt reply? */ + if (req->intr_unique == oh.unique) { + err = -EINVAL; + if (nbytes != sizeof(struct fuse_out_header)) + goto err_unlock; + + if (oh.error == -ENOSYS) + fc->no_interrupt = 1; + else if (oh.error == -EAGAIN) + queue_interrupt(fc, req); + + spin_unlock(&fc->lock); + fuse_copy_finish(&cs); + return nbytes; + } + + req->state = FUSE_REQ_WRITING; list_move(&req->list, &fc->io); req->out.h = oh; req->locked = 1; @@ -764,9 +883,9 @@ static ssize_t fuse_dev_writev(struct file *file, const struct iovec *iov, spin_lock(&fc->lock); req->locked = 0; if (!err) { - if (req->interrupted) + if (req->aborted) err = -ENOENT; - } else if (!req->interrupted) + } else if (!req->aborted) req->out.h.error = -EIO; request_end(fc, req); @@ -800,7 +919,7 @@ static unsigned fuse_dev_poll(struct file *file, poll_table *wait) spin_lock(&fc->lock); if (!fc->connected) mask = POLLERR; - else if (!list_empty(&fc->pending)) + else if (request_pending(fc)) mask |= POLLIN | POLLRDNORM; spin_unlock(&fc->lock); @@ -826,7 +945,7 @@ static void end_requests(struct fuse_conn *fc, struct list_head *head) /* * Abort requests under I/O * - * The requests are set to interrupted and finished, and the request + * The requests are set to aborted and finished, and the request * waiter is woken up. This will make request_wait_answer() wait * until the request is unlocked and then return. * @@ -841,7 +960,7 @@ static void end_io_requests(struct fuse_conn *fc) list_entry(fc->io.next, struct fuse_req, list); void (*end) (struct fuse_conn *, struct fuse_req *) = req->end; - req->interrupted = 1; + req->aborted = 1; req->out.h.error = -ECONNABORTED; req->state = FUSE_REQ_FINISHED; list_del_init(&req->list); @@ -874,19 +993,20 @@ static void end_io_requests(struct fuse_conn *fc) * onto the pending list is prevented by req->connected being false. * * Progression of requests under I/O to the processing list is - * prevented by the req->interrupted flag being true for these - * requests. For this reason requests on the io list must be aborted - * first. + * prevented by the req->aborted flag being true for these requests. + * For this reason requests on the io list must be aborted first. */ void fuse_abort_conn(struct fuse_conn *fc) { spin_lock(&fc->lock); if (fc->connected) { fc->connected = 0; + fc->blocked = 0; end_io_requests(fc); end_requests(fc, &fc->pending); end_requests(fc, &fc->processing); wake_up_all(&fc->waitq); + wake_up_all(&fc->blocked_waitq); kill_fasync(&fc->fasync, SIGIO, POLL_IN); } spin_unlock(&fc->lock); @@ -902,7 +1022,7 @@ static int fuse_dev_release(struct inode *inode, struct file *file) end_requests(fc, &fc->processing); spin_unlock(&fc->lock); fasync_helper(-1, file, 0, &fc->fasync); - kobject_put(&fc->kobj); + fuse_conn_put(fc); } return 0; diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c index 8d7546e832e..f85b2a282f1 100644 --- a/fs/fuse/dir.c +++ b/fs/fuse/dir.c @@ -1,6 +1,6 @@ /* FUSE: Filesystem in Userspace - Copyright (C) 2001-2005 Miklos Szeredi <miklos@szeredi.hu> + Copyright (C) 2001-2006 Miklos Szeredi <miklos@szeredi.hu> This program can be distributed under the terms of the GNU GPL. See the file COPYING. @@ -14,6 +14,33 @@ #include <linux/sched.h> #include <linux/namei.h> +#if BITS_PER_LONG >= 64 +static inline void fuse_dentry_settime(struct dentry *entry, u64 time) +{ + entry->d_time = time; +} + +static inline u64 fuse_dentry_time(struct dentry *entry) +{ + return entry->d_time; +} +#else +/* + * On 32 bit archs store the high 32 bits of time in d_fsdata + */ +static void fuse_dentry_settime(struct dentry *entry, u64 time) +{ + entry->d_time = time; + entry->d_fsdata = (void *) (unsigned long) (time >> 32); +} + +static u64 fuse_dentry_time(struct dentry *entry) +{ + return (u64) entry->d_time + + ((u64) (unsigned long) entry->d_fsdata << 32); +} +#endif + /* * FUSE caches dentries and attributes with separate timeout. The * time in jiffies until the dentry/attributes are valid is stored in @@ -23,10 +50,13 @@ /* * Calculate the time in jiffies until a dentry/attributes are valid */ -static unsigned long time_to_jiffies(unsigned long sec, unsigned long nsec) +static u64 time_to_jiffies(unsigned long sec, unsigned long nsec) { - struct timespec ts = {sec, nsec}; - return jiffies + timespec_to_jiffies(&ts); + if (sec || nsec) { + struct timespec ts = {sec, nsec}; + return get_jiffies_64() + timespec_to_jiffies(&ts); + } else + return 0; } /* @@ -35,7 +65,8 @@ static unsigned long time_to_jiffies(unsigned long sec, unsigned long nsec) */ static void fuse_change_timeout(struct dentry *entry, struct fuse_entry_out *o) { - entry->d_time = time_to_jiffies(o->entry_valid, o->entry_valid_nsec); + fuse_dentry_settime(entry, + time_to_jiffies(o->entry_valid, o->entry_valid_nsec)); if (entry->d_inode) get_fuse_inode(entry->d_inode)->i_time = time_to_jiffies(o->attr_valid, o->attr_valid_nsec); @@ -47,7 +78,7 @@ static void fuse_change_timeout(struct dentry *entry, struct fuse_entry_out *o) */ void fuse_invalidate_attr(struct inode *inode) { - get_fuse_inode(inode)->i_time = jiffies - 1; + get_fuse_inode(inode)->i_time = 0; } /* @@ -60,7 +91,7 @@ void fuse_invalidate_attr(struct inode *inode) */ static void fuse_invalidate_entry_cache(struct dentry *entry) { - entry->d_time = jiffies - 1; + fuse_dentry_settime(entry, 0); } /* @@ -79,7 +110,6 @@ static void fuse_lookup_init(struct fuse_req *req, struct inode *dir, { req->in.h.opcode = FUSE_LOOKUP; req->in.h.nodeid = get_node_id(dir); - req->inode = dir; req->in.numargs = 1; req->in.args[0].size = entry->d_name.len + 1; req->in.args[0].value = entry->d_name.name; @@ -103,7 +133,7 @@ static int fuse_dentry_revalidate(struct dentry *entry, struct nameidata *nd) if (inode && is_bad_inode(inode)) return 0; - else if (time_after(jiffies, entry->d_time)) { + else if (fuse_dentry_time(entry) < get_jiffies_64()) { int err; struct fuse_entry_out outarg; struct fuse_conn *fc; @@ -225,6 +255,20 @@ static struct dentry *fuse_lookup(struct inode *dir, struct dentry *entry, } /* + * Synchronous release for the case when something goes wrong in CREATE_OPEN + */ +static void fuse_sync_release(struct fuse_conn *fc, struct fuse_file *ff, + u64 nodeid, int flags) +{ + struct fuse_req *req; + + req = fuse_release_fill(ff, nodeid, flags, FUSE_RELEASE); + req->force = 1; + request_send(fc, req); + fuse_put_request(fc, req); +} + +/* * Atomic create+open operation * * If the filesystem doesn't support this, then fall back to separate @@ -237,6 +281,7 @@ static int fuse_create_open(struct inode *dir, struct dentry *entry, int mode, struct inode *inode; struct fuse_conn *fc = get_fuse_conn(dir); struct fuse_req *req; + struct fuse_req *forget_req; struct fuse_open_in inarg; struct fuse_open_out outopen; struct fuse_entry_out outentry; @@ -247,9 +292,14 @@ static int fuse_create_open(struct inode *dir, struct dentry *entry, int mode, if (fc->no_create) return -ENOSYS; + forget_req = fuse_get_req(fc); + if (IS_ERR(forget_req)) + return PTR_ERR(forget_req); + req = fuse_get_req(fc); + err = PTR_ERR(req); if (IS_ERR(req)) - return PTR_ERR(req); + goto out_put_forget_req; err = -ENOMEM; ff = fuse_file_alloc(); @@ -262,7 +312,6 @@ static int fuse_create_open(struct inode *dir, struct dentry *entry, int mode, inarg.mode = mode; req->in.h.opcode = FUSE_CREATE; req->in.h.nodeid = get_node_id(dir); - req->inode = dir; req->in.numargs = 2; req->in.args[0].size = sizeof(inarg); req->in.args[0].value = &inarg; @@ -285,25 +334,23 @@ static int fuse_create_open(struct inode *dir, struct dentry *entry, int mode, if (!S_ISREG(outentry.attr.mode) || invalid_nodeid(outentry.nodeid)) goto out_free_ff; + fuse_put_request(fc, req); inode = fuse_iget(dir->i_sb, outentry.nodeid, outentry.generation, &outentry.attr); - err = -ENOMEM; if (!inode) { flags &= ~(O_CREAT | O_EXCL | O_TRUNC); ff->fh = outopen.fh; - /* Special release, with inode = NULL, this will - trigger a 'forget' request when the release is - complete */ - fuse_send_release(fc, ff, outentry.nodeid, NULL, flags, 0); - goto out_put_request; + fuse_sync_release(fc, ff, outentry.nodeid, flags); + fuse_send_forget(fc, forget_req, outentry.nodeid, 1); + return -ENOMEM; } - fuse_put_request(fc, req); + fuse_put_request(fc, forget_req); d_instantiate(entry, inode); fuse_change_timeout(entry, &outentry); file = lookup_instantiate_filp(nd, entry, generic_file_open); if (IS_ERR(file)) { ff->fh = outopen.fh; - fuse_send_release(fc, ff, outentry.nodeid, inode, flags, 0); + fuse_sync_release(fc, ff, outentry.nodeid, flags); return PTR_ERR(file); } fuse_finish_open(inode, file, ff, &outopen); @@ -313,6 +360,8 @@ static int fuse_create_open(struct inode *dir, struct dentry *entry, int mode, fuse_file_free(ff); out_put_request: fuse_put_request(fc, req); + out_put_forget_req: + fuse_put_request(fc, forget_req); return err; } @@ -328,7 +377,6 @@ static int create_new_entry(struct fuse_conn *fc, struct fuse_req *req, int err; req->in.h.nodeid = get_node_id(dir); - req->inode = dir; req->out.numargs = 1; req->out.args[0].size = sizeof(outarg); req->out.args[0].value = &outarg; @@ -448,7 +496,6 @@ static int fuse_unlink(struct inode *dir, struct dentry *entry) req->in.h.opcode = FUSE_UNLINK; req->in.h.nodeid = get_node_id(dir); - req->inode = dir; req->in.numargs = 1; req->in.args[0].size = entry->d_name.len + 1; req->in.args[0].value = entry->d_name.name; @@ -480,7 +527,6 @@ static int fuse_rmdir(struct inode *dir, struct dentry *entry) req->in.h.opcode = FUSE_RMDIR; req->in.h.nodeid = get_node_id(dir); - req->inode = dir; req->in.numargs = 1; req->in.args[0].size = entry->d_name.len + 1; req->in.args[0].value = entry->d_name.name; @@ -510,8 +556,6 @@ static int fuse_rename(struct inode *olddir, struct dentry *oldent, inarg.newdir = get_node_id(newdir); req->in.h.opcode = FUSE_RENAME; req->in.h.nodeid = get_node_id(olddir); - req->inode = olddir; - req->inode2 = newdir; req->in.numargs = 3; req->in.args[0].size = sizeof(inarg); req->in.args[0].value = &inarg; @@ -558,7 +602,6 @@ static int fuse_link(struct dentry *entry, struct inode *newdir, memset(&inarg, 0, sizeof(inarg)); inarg.oldnodeid = get_node_id(inode); req->in.h.opcode = FUSE_LINK; - req->inode2 = inode; req->in.numargs = 2; req->in.args[0].size = sizeof(inarg); req->in.args[0].value = &inarg; @@ -587,7 +630,6 @@ int fuse_do_getattr(struct inode *inode) req->in.h.opcode = FUSE_GETATTR; req->in.h.nodeid = get_node_id(inode); - req->inode = inode; req->out.numargs = 1; req->out.args[0].size = sizeof(arg); req->out.args[0].value = &arg; @@ -655,7 +697,7 @@ static int fuse_revalidate(struct dentry *entry) if (!fuse_allow_task(fc, current)) return -EACCES; if (get_node_id(inode) != FUSE_ROOT_ID && - time_before_eq(jiffies, fi->i_time)) + fi->i_time >= get_jiffies_64()) return 0; return fuse_do_getattr(inode); @@ -679,7 +721,6 @@ static int fuse_access(struct inode *inode, int mask) inarg.mask = mask; req->in.h.opcode = FUSE_ACCESS; req->in.h.nodeid = get_node_id(inode); - req->inode = inode; req->in.numargs = 1; req->in.args[0].size = sizeof(inarg); req->in.args[0].value = &inarg; @@ -735,7 +776,7 @@ static int fuse_permission(struct inode *inode, int mask, struct nameidata *nd) if ((mask & MAY_EXEC) && !S_ISDIR(mode) && !(mode & S_IXUGO)) return -EACCES; - if (nd && (nd->flags & LOOKUP_ACCESS)) + if (nd && (nd->flags & (LOOKUP_ACCESS | LOOKUP_CHDIR))) return fuse_access(inode, mask); return 0; } @@ -820,7 +861,6 @@ static char *read_link(struct dentry *dentry) } req->in.h.opcode = FUSE_READLINK; req->in.h.nodeid = get_node_id(inode); - req->inode = inode; req->out.argvar = 1; req->out.numargs = 1; req->out.args[0].size = PAGE_SIZE - 1; @@ -939,7 +979,6 @@ static int fuse_setattr(struct dentry *entry, struct iattr *attr) iattr_to_fattr(attr, &inarg); req->in.h.opcode = FUSE_SETATTR; req->in.h.nodeid = get_node_id(inode); - req->inode = inode; req->in.numargs = 1; req->in.args[0].size = sizeof(inarg); req->in.args[0].value = &inarg; @@ -1002,7 +1041,6 @@ static int fuse_setxattr(struct dentry *entry, const char *name, inarg.flags = flags; req->in.h.opcode = FUSE_SETXATTR; req->in.h.nodeid = get_node_id(inode); - req->inode = inode; req->in.numargs = 3; req->in.args[0].size = sizeof(inarg); req->in.args[0].value = &inarg; @@ -1041,7 +1079,6 @@ static ssize_t fuse_getxattr(struct dentry *entry, const char *name, inarg.size = size; req->in.h.opcode = FUSE_GETXATTR; req->in.h.nodeid = get_node_id(inode); - req->inode = inode; req->in.numargs = 2; req->in.args[0].size = sizeof(inarg); req->in.args[0].value = &inarg; @@ -1091,7 +1128,6 @@ static ssize_t fuse_listxattr(struct dentry *entry, char *list, size_t size) inarg.size = size; req->in.h.opcode = FUSE_LISTXATTR; req->in.h.nodeid = get_node_id(inode); - req->inode = inode; req->in.numargs = 1; req->in.args[0].size = sizeof(inarg); req->in.args[0].value = &inarg; @@ -1135,7 +1171,6 @@ static int fuse_removexattr(struct dentry *entry, const char *name) req->in.h.opcode = FUSE_REMOVEXATTR; req->in.h.nodeid = get_node_id(inode); - req->inode = inode; req->in.numargs = 1; req->in.args[0].size = strlen(name) + 1; req->in.args[0].value = name; diff --git a/fs/fuse/file.c b/fs/fuse/file.c index 087f3b734f4..5c4fcd1dbf5 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -30,7 +30,6 @@ static int fuse_send_open(struct inode *inode, struct file *file, int isdir, inarg.flags = file->f_flags & ~(O_CREAT | O_EXCL | O_NOCTTY | O_TRUNC); req->in.h.opcode = isdir ? FUSE_OPENDIR : FUSE_OPEN; req->in.h.nodeid = get_node_id(inode); - req->inode = inode; req->in.numargs = 1; req->in.args[0].size = sizeof(inarg); req->in.args[0].value = &inarg; @@ -49,8 +48,8 @@ struct fuse_file *fuse_file_alloc(void) struct fuse_file *ff; ff = kmalloc(sizeof(struct fuse_file), GFP_KERNEL); if (ff) { - ff->release_req = fuse_request_alloc(); - if (!ff->release_req) { + ff->reserved_req = fuse_request_alloc(); + if (!ff->reserved_req) { kfree(ff); ff = NULL; } @@ -60,7 +59,7 @@ struct fuse_file *fuse_file_alloc(void) void fuse_file_free(struct fuse_file *ff) { - fuse_request_free(ff->release_req); + fuse_request_free(ff->reserved_req); kfree(ff); } @@ -113,37 +112,22 @@ int fuse_open_common(struct inode *inode, struct file *file, int isdir) return err; } -/* Special case for failed iget in CREATE */ -static void fuse_release_end(struct fuse_conn *fc, struct fuse_req *req) +struct fuse_req *fuse_release_fill(struct fuse_file *ff, u64 nodeid, int flags, + int opcode) { - /* If called from end_io_requests(), req has more than one - reference and fuse_reset_request() cannot work */ - if (fc->connected) { - u64 nodeid = req->in.h.nodeid; - fuse_reset_request(req); - fuse_send_forget(fc, req, nodeid, 1); - } else - fuse_put_request(fc, req); -} - -void fuse_send_release(struct fuse_conn *fc, struct fuse_file *ff, - u64 nodeid, struct inode *inode, int flags, int isdir) -{ - struct fuse_req * req = ff->release_req; + struct fuse_req *req = ff->reserved_req; struct fuse_release_in *inarg = &req->misc.release_in; inarg->fh = ff->fh; inarg->flags = flags; - req->in.h.opcode = isdir ? FUSE_RELEASEDIR : FUSE_RELEASE; + req->in.h.opcode = opcode; req->in.h.nodeid = nodeid; - req->inode = inode; req->in.numargs = 1; req->in.args[0].size = sizeof(struct fuse_release_in); req->in.args[0].value = inarg; - request_send_background(fc, req); - if (!inode) - req->end = fuse_release_end; kfree(ff); + + return req; } int fuse_release_common(struct inode *inode, struct file *file, int isdir) @@ -151,8 +135,15 @@ int fuse_release_common(struct inode *inode, struct file *file, int isdir) struct fuse_file *ff = file->private_data; if (ff) { struct fuse_conn *fc = get_fuse_conn(inode); - u64 nodeid = get_node_id(inode); - fuse_send_release(fc, ff, nodeid, inode, file->f_flags, isdir); + struct fuse_req *req; + + req = fuse_release_fill(ff, get_node_id(inode), file->f_flags, + isdir ? FUSE_RELEASEDIR : FUSE_RELEASE); + + /* Hold vfsmount and dentry until release is finished */ + req->vfsmount = mntget(file->f_vfsmnt); + req->dentry = dget(file->f_dentry); + request_send_background(fc, req); } /* Return value is ignored by VFS */ @@ -169,6 +160,28 @@ static int fuse_release(struct inode *inode, struct file *file) return fuse_release_common(inode, file, 0); } +/* + * Scramble the ID space with XTEA, so that the value of the files_struct + * pointer is not exposed to userspace. + */ +static u64 fuse_lock_owner_id(struct fuse_conn *fc, fl_owner_t id) +{ + u32 *k = fc->scramble_key; + u64 v = (unsigned long) id; + u32 v0 = v; + u32 v1 = v >> 32; + u32 sum = 0; + int i; + + for (i = 0; i < 32; i++) { + v0 += ((v1 << 4 ^ v1 >> 5) + v1) ^ (sum + k[sum & 3]); + sum += 0x9E3779B9; + v1 += ((v0 << 4 ^ v0 >> 5) + v0) ^ (sum + k[sum>>11 & 3]); + } + + return (u64) v0 + ((u64) v1 << 32); +} + static int fuse_flush(struct file *file, fl_owner_t id) { struct inode *inode = file->f_dentry->d_inode; @@ -184,19 +197,16 @@ static int fuse_flush(struct file *file, fl_owner_t id) if (fc->no_flush) return 0; - req = fuse_get_req(fc); - if (IS_ERR(req)) - return PTR_ERR(req); - + req = fuse_get_req_nofail(fc, file); memset(&inarg, 0, sizeof(inarg)); inarg.fh = ff->fh; + inarg.lock_owner = fuse_lock_owner_id(fc, id); req->in.h.opcode = FUSE_FLUSH; req->in.h.nodeid = get_node_id(inode); - req->inode = inode; - req->file = file; req->in.numargs = 1; req->in.args[0].size = sizeof(inarg); req->in.args[0].value = &inarg; + req->force = 1; request_send(fc, req); err = req->out.h.error; fuse_put_request(fc, req); @@ -232,8 +242,6 @@ int fuse_fsync_common(struct file *file, struct dentry *de, int datasync, inarg.fsync_flags = datasync ? 1 : 0; req->in.h.opcode = isdir ? FUSE_FSYNCDIR : FUSE_FSYNC; req->in.h.nodeid = get_node_id(inode); - req->inode = inode; - req->file = file; req->in.numargs = 1; req->in.args[0].size = sizeof(inarg); req->in.args[0].value = &inarg; @@ -266,8 +274,6 @@ void fuse_read_fill(struct fuse_req *req, struct file *file, inarg->size = count; req->in.h.opcode = opcode; req->in.h.nodeid = get_node_id(inode); - req->inode = inode; - req->file = file; req->in.numargs = 1; req->in.args[0].size = sizeof(struct fuse_read_in); req->in.args[0].value = inarg; @@ -342,6 +348,8 @@ static void fuse_send_readpages(struct fuse_req *req, struct file *file, req->out.page_zeroing = 1; fuse_read_fill(req, file, inode, pos, count, FUSE_READ); if (fc->async_read) { + get_file(file); + req->file = file; req->end = fuse_readpages_end; request_send_background(fc, req); } else { @@ -387,14 +395,16 @@ static int fuse_readpages(struct file *file, struct address_space *mapping, struct fuse_readpages_data data; int err; + err = -EIO; if (is_bad_inode(inode)) - return -EIO; + goto clean_pages_up; data.file = file; data.inode = inode; data.req = fuse_get_req(fc); + err = PTR_ERR(data.req); if (IS_ERR(data.req)) - return PTR_ERR(data.req); + goto clean_pages_up; err = read_cache_pages(mapping, pages, fuse_readpages_fill, &data); if (!err) { @@ -404,6 +414,10 @@ static int fuse_readpages(struct file *file, struct address_space *mapping, fuse_put_request(fc, data.req); } return err; + +clean_pages_up: + put_pages_list(pages); + return err; } static size_t fuse_send_write(struct fuse_req *req, struct file *file, @@ -420,8 +434,6 @@ static size_t fuse_send_write(struct fuse_req *req, struct file *file, inarg.size = count; req->in.h.opcode = FUSE_WRITE; req->in.h.nodeid = get_node_id(inode); - req->inode = inode; - req->file = file; req->in.argpages = 1; req->in.numargs = 2; req->in.args[0].size = sizeof(struct fuse_write_in); @@ -619,6 +631,126 @@ static int fuse_set_page_dirty(struct page *page) return 0; } +static int convert_fuse_file_lock(const struct fuse_file_lock *ffl, + struct file_lock *fl) +{ + switch (ffl->type) { + case F_UNLCK: + break; + + case F_RDLCK: + case F_WRLCK: + if (ffl->start > OFFSET_MAX || ffl->end > OFFSET_MAX || + ffl->end < ffl->start) + return -EIO; + + fl->fl_start = ffl->start; + fl->fl_end = ffl->end; + fl->fl_pid = ffl->pid; + break; + + default: + return -EIO; + } + fl->fl_type = ffl->type; + return 0; +} + +static void fuse_lk_fill(struct fuse_req *req, struct file *file, + const struct file_lock *fl, int opcode, pid_t pid) +{ + struct inode *inode = file->f_dentry->d_inode; + struct fuse_conn *fc = get_fuse_conn(inode); + struct fuse_file *ff = file->private_data; + struct fuse_lk_in *arg = &req->misc.lk_in; + + arg->fh = ff->fh; + arg->owner = fuse_lock_owner_id(fc, fl->fl_owner); + arg->lk.start = fl->fl_start; + arg->lk.end = fl->fl_end; + arg->lk.type = fl->fl_type; + arg->lk.pid = pid; + req->in.h.opcode = opcode; + req->in.h.nodeid = get_node_id(inode); + req->in.numargs = 1; + req->in.args[0].size = sizeof(*arg); + req->in.args[0].value = arg; +} + +static int fuse_getlk(struct file *file, struct file_lock *fl) +{ + struct inode *inode = file->f_dentry->d_inode; + struct fuse_conn *fc = get_fuse_conn(inode); + struct fuse_req *req; + struct fuse_lk_out outarg; + int err; + + req = fuse_get_req(fc); + if (IS_ERR(req)) + return PTR_ERR(req); + + fuse_lk_fill(req, file, fl, FUSE_GETLK, 0); + req->out.numargs = 1; + req->out.args[0].size = sizeof(outarg); + req->out.args[0].value = &outarg; + request_send(fc, req); + err = req->out.h.error; + fuse_put_request(fc, req); + if (!err) + err = convert_fuse_file_lock(&outarg.lk, fl); + + return err; +} + +static int fuse_setlk(struct file *file, struct file_lock *fl) +{ + struct inode *inode = file->f_dentry->d_inode; + struct fuse_conn *fc = get_fuse_conn(inode); + struct fuse_req *req; + int opcode = (fl->fl_flags & FL_SLEEP) ? FUSE_SETLKW : FUSE_SETLK; + pid_t pid = fl->fl_type != F_UNLCK ? current->tgid : 0; + int err; + + /* Unlock on close is handled by the flush method */ + if (fl->fl_flags & FL_CLOSE) + return 0; + + req = fuse_get_req(fc); + if (IS_ERR(req)) + return PTR_ERR(req); + + fuse_lk_fill(req, file, fl, opcode, pid); + request_send(fc, req); + err = req->out.h.error; + /* locking is restartable */ + if (err == -EINTR) + err = -ERESTARTSYS; + fuse_put_request(fc, req); + return err; +} + +static int fuse_file_lock(struct file *file, int cmd, struct file_lock *fl) +{ + struct inode *inode = file->f_dentry->d_inode; + struct fuse_conn *fc = get_fuse_conn(inode); + int err; + + if (cmd == F_GETLK) { + if (fc->no_lock) { + if (!posix_test_lock(file, fl, fl)) + fl->fl_type = F_UNLCK; + err = 0; + } else + err = fuse_getlk(file, fl); + } else { + if (fc->no_lock) + err = posix_lock_file_wait(file, fl); + else + err = fuse_setlk(file, fl); + } + return err; +} + static const struct file_operations fuse_file_operations = { .llseek = generic_file_llseek, .read = generic_file_read, @@ -628,6 +760,7 @@ static const struct file_operations fuse_file_operations = { .flush = fuse_flush, .release = fuse_release, .fsync = fuse_fsync, + .lock = fuse_file_lock, .sendfile = generic_file_sendfile, }; @@ -639,10 +772,11 @@ static const struct file_operations fuse_direct_io_file_operations = { .flush = fuse_flush, .release = fuse_release, .fsync = fuse_fsync, + .lock = fuse_file_lock, /* no mmap and sendfile */ }; -static struct address_space_operations fuse_file_aops = { +static const struct address_space_operations fuse_file_aops = { .readpage = fuse_readpage, .prepare_write = fuse_prepare_write, .commit_write = fuse_commit_write, diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h index 0474202cb5d..69c7750d55b 100644 --- a/fs/fuse/fuse_i.h +++ b/fs/fuse/fuse_i.h @@ -8,12 +8,13 @@ #include <linux/fuse.h> #include <linux/fs.h> +#include <linux/mount.h> #include <linux/wait.h> #include <linux/list.h> #include <linux/spinlock.h> #include <linux/mm.h> #include <linux/backing-dev.h> -#include <asm/semaphore.h> +#include <linux/mutex.h> /** Max number of pages that can be used in a single read request */ #define FUSE_MAX_PAGES_PER_REQ 32 @@ -24,6 +25,9 @@ /** It could be as large as PATH_MAX, but would that have any uses? */ #define FUSE_NAME_MAX 1024 +/** Number of dentries for each connection in the control filesystem */ +#define FUSE_CTL_NUM_DENTRIES 3 + /** If the FUSE_DEFAULT_PERMISSIONS flag is given, the filesystem module will check permissions based on the file mode. Otherwise no permission checking is done in the kernel */ @@ -33,6 +37,11 @@ doing the mount will be allowed to access the filesystem */ #define FUSE_ALLOW_OTHER (1 << 1) +/** List of active connections */ +extern struct list_head fuse_conn_list; + +/** Global mutex protecting fuse_conn_list and the control filesystem */ +extern struct mutex fuse_mutex; /** FUSE inode */ struct fuse_inode { @@ -50,13 +59,13 @@ struct fuse_inode { struct fuse_req *forget_req; /** Time in jiffies until the file attributes are valid */ - unsigned long i_time; + u64 i_time; }; /** FUSE specific file data */ struct fuse_file { /** Request reserved for flush and release */ - struct fuse_req *release_req; + struct fuse_req *reserved_req; /** File handle used by userspace */ u64 fh; @@ -122,6 +131,7 @@ enum fuse_req_state { FUSE_REQ_PENDING, FUSE_REQ_READING, FUSE_REQ_SENT, + FUSE_REQ_WRITING, FUSE_REQ_FINISHED }; @@ -135,12 +145,15 @@ struct fuse_req { fuse_conn */ struct list_head list; - /** Entry on the background list */ - struct list_head bg_entry; + /** Entry on the interrupts list */ + struct list_head intr_entry; /** refcount */ atomic_t count; + /** Unique ID for the interrupt request */ + u64 intr_unique; + /* * The following bitfields are either set once before the * request is queued or setting/clearing them is protected by @@ -150,12 +163,18 @@ struct fuse_req { /** True if the request has reply */ unsigned isreply:1; - /** The request was interrupted */ - unsigned interrupted:1; + /** Force sending of the request even if interrupted */ + unsigned force:1; + + /** The request was aborted */ + unsigned aborted:1; /** Request is sent in the background */ unsigned background:1; + /** The request has been interrupted */ + unsigned interrupted:1; + /** Data is being copied to/from the request */ unsigned locked:1; @@ -181,6 +200,7 @@ struct fuse_req { struct fuse_init_in init_in; struct fuse_init_out init_out; struct fuse_read_in read_in; + struct fuse_lk_in lk_in; } misc; /** page vector */ @@ -192,17 +212,20 @@ struct fuse_req { /** offset of data on first page */ unsigned page_offset; - /** Inode used in the request */ - struct inode *inode; - - /** Second inode used in the request (or NULL) */ - struct inode *inode2; - /** File used in the request (or NULL) */ struct file *file; + /** vfsmount used in release */ + struct vfsmount *vfsmount; + + /** dentry used in release */ + struct dentry *dentry; + /** Request completion callback */ void (*end)(struct fuse_conn *, struct fuse_req *); + + /** Request is stolen from fuse_file->reserved_req */ + struct file *stolen_file; }; /** @@ -216,6 +239,9 @@ struct fuse_conn { /** Lock protecting accessess to members of this structure */ spinlock_t lock; + /** Refcount */ + atomic_t count; + /** The user id for this mount */ uid_t user_id; @@ -243,13 +269,12 @@ struct fuse_conn { /** The list of requests under I/O */ struct list_head io; - /** Requests put in the background (RELEASE or any other - interrupted request) */ - struct list_head background; - /** Number of requests currently in the background */ unsigned num_background; + /** Pending interrupts */ + struct list_head interrupts; + /** Flag indicating if connection is blocked. This will be the case before the INIT reply is received, and if there are too many outstading backgrounds requests */ @@ -258,15 +283,9 @@ struct fuse_conn { /** waitq for blocked connection */ wait_queue_head_t blocked_waitq; - /** RW semaphore for exclusion with fuse_put_super() */ - struct rw_semaphore sbput_sem; - /** The next unique request id */ u64 reqctr; - /** Mount is active */ - unsigned mounted; - /** Connection established, cleared on umount, connection abort and device release */ unsigned connected; @@ -305,12 +324,18 @@ struct fuse_conn { /** Is removexattr not implemented by fs? */ unsigned no_removexattr : 1; + /** Are file locking primitives not implemented by fs? */ + unsigned no_lock : 1; + /** Is access not implemented by fs? */ unsigned no_access : 1; /** Is create not implemented by fs? */ unsigned no_create : 1; + /** Is interrupt not implemented by fs? */ + unsigned no_interrupt : 1; + /** The number of requests waiting for completion */ atomic_t num_waiting; @@ -320,11 +345,23 @@ struct fuse_conn { /** Backing dev info */ struct backing_dev_info bdi; - /** kobject */ - struct kobject kobj; + /** Entry on the fuse_conn_list */ + struct list_head entry; + + /** Unique ID */ + u64 id; + + /** Dentries in the control filesystem */ + struct dentry *ctl_dentry[FUSE_CTL_NUM_DENTRIES]; + + /** number of dentries used in the above array */ + int ctl_ndents; /** O_ASYNC requests */ struct fasync_struct *fasync; + + /** Key for lock owner ID scrambling */ + u32 scramble_key[4]; }; static inline struct fuse_conn *get_fuse_conn_super(struct super_block *sb) @@ -337,11 +374,6 @@ static inline struct fuse_conn *get_fuse_conn(struct inode *inode) return get_fuse_conn_super(inode->i_sb); } -static inline struct fuse_conn *get_fuse_conn_kobj(struct kobject *obj) -{ - return container_of(obj, struct fuse_conn, kobj); -} - static inline struct fuse_inode *get_fuse_inode(struct inode *inode) { return container_of(inode, struct fuse_inode, inode); @@ -383,12 +415,9 @@ void fuse_file_free(struct fuse_file *ff); void fuse_finish_open(struct inode *inode, struct file *file, struct fuse_file *ff, struct fuse_open_out *outarg); -/** - * Send a RELEASE request - */ -void fuse_send_release(struct fuse_conn *fc, struct fuse_file *ff, - u64 nodeid, struct inode *inode, int flags, int isdir); - +/** */ +struct fuse_req *fuse_release_fill(struct fuse_file *ff, u64 nodeid, int flags, + int opcode); /** * Send RELEASE or RELEASEDIR request */ @@ -435,6 +464,9 @@ int fuse_dev_init(void); */ void fuse_dev_cleanup(void); +int fuse_ctl_init(void); +void fuse_ctl_cleanup(void); + /** * Allocate a request */ @@ -446,14 +478,14 @@ struct fuse_req *fuse_request_alloc(void); void fuse_request_free(struct fuse_req *req); /** - * Reinitialize a request, the preallocated flag is left unmodified + * Get a request, may fail with -ENOMEM */ -void fuse_reset_request(struct fuse_req *req); +struct fuse_req *fuse_get_req(struct fuse_conn *fc); /** - * Reserve a preallocated request + * Gets a requests for a file operation, always succeeds */ -struct fuse_req *fuse_get_req(struct fuse_conn *fc); +struct fuse_req *fuse_get_req_nofail(struct fuse_conn *fc, struct file *file); /** * Decrement reference count of a request. If count goes to zero free @@ -476,11 +508,6 @@ void request_send_noreply(struct fuse_conn *fc, struct fuse_req *req); */ void request_send_background(struct fuse_conn *fc, struct fuse_req *req); -/** - * Release inodes and file associated with background request - */ -void fuse_release_background(struct fuse_conn *fc, struct fuse_req *req); - /* Abort all requests */ void fuse_abort_conn(struct fuse_conn *fc); @@ -493,3 +520,23 @@ int fuse_do_getattr(struct inode *inode); * Invalidate inode attributes */ void fuse_invalidate_attr(struct inode *inode); + +/** + * Acquire reference to fuse_conn + */ +struct fuse_conn *fuse_conn_get(struct fuse_conn *fc); + +/** + * Release reference to fuse_conn + */ +void fuse_conn_put(struct fuse_conn *fc); + +/** + * Add connection to control filesystem + */ +int fuse_ctl_add_conn(struct fuse_conn *fc); + +/** + * Remove connection from control filesystem + */ +void fuse_ctl_remove_conn(struct fuse_conn *fc); diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c index a13c0f52905..7d0a9aee01f 100644 --- a/fs/fuse/inode.c +++ b/fs/fuse/inode.c @@ -11,25 +11,20 @@ #include <linux/pagemap.h> #include <linux/slab.h> #include <linux/file.h> -#include <linux/mount.h> #include <linux/seq_file.h> #include <linux/init.h> #include <linux/module.h> #include <linux/parser.h> #include <linux/statfs.h> +#include <linux/random.h> MODULE_AUTHOR("Miklos Szeredi <miklos@szeredi.hu>"); MODULE_DESCRIPTION("Filesystem in Userspace"); MODULE_LICENSE("GPL"); static kmem_cache_t *fuse_inode_cachep; -static struct subsystem connections_subsys; - -struct fuse_conn_attr { - struct attribute attr; - ssize_t (*show)(struct fuse_conn *, char *); - ssize_t (*store)(struct fuse_conn *, const char *, size_t); -}; +struct list_head fuse_conn_list; +DEFINE_MUTEX(fuse_mutex); #define FUSE_SUPER_MAGIC 0x65735546 @@ -56,7 +51,7 @@ static struct inode *fuse_alloc_inode(struct super_block *sb) return NULL; fi = get_fuse_inode(inode); - fi->i_time = jiffies - 1; + fi->i_time = 0; fi->nodeid = 0; fi->nlookup = 0; fi->forget_req = fuse_request_alloc(); @@ -104,6 +99,14 @@ static void fuse_clear_inode(struct inode *inode) } } +static int fuse_remount_fs(struct super_block *sb, int *flags, char *data) +{ + if (*flags & MS_MANDLOCK) + return -EINVAL; + + return 0; +} + void fuse_change_attributes(struct inode *inode, struct fuse_attr *attr) { if (S_ISREG(inode->i_mode) && i_size_read(inode) != attr->size) @@ -115,7 +118,6 @@ void fuse_change_attributes(struct inode *inode, struct fuse_attr *attr) inode->i_uid = attr->uid; inode->i_gid = attr->gid; i_size_write(inode, attr->size); - inode->i_blksize = PAGE_CACHE_SIZE; inode->i_blocks = attr->blocks; inode->i_atime.tv_sec = attr->atime; inode->i_atime.tv_nsec = attr->atimensec; @@ -195,31 +197,29 @@ struct inode *fuse_iget(struct super_block *sb, unsigned long nodeid, return inode; } -static void fuse_umount_begin(struct super_block *sb) +static void fuse_umount_begin(struct vfsmount *vfsmnt, int flags) { - fuse_abort_conn(get_fuse_conn_super(sb)); + if (flags & MNT_FORCE) + fuse_abort_conn(get_fuse_conn_super(vfsmnt->mnt_sb)); } static void fuse_put_super(struct super_block *sb) { struct fuse_conn *fc = get_fuse_conn_super(sb); - down_write(&fc->sbput_sem); - while (!list_empty(&fc->background)) - fuse_release_background(fc, - list_entry(fc->background.next, - struct fuse_req, bg_entry)); - spin_lock(&fc->lock); - fc->mounted = 0; fc->connected = 0; + fc->blocked = 0; spin_unlock(&fc->lock); - up_write(&fc->sbput_sem); /* Flush all readers on this fs */ kill_fasync(&fc->fasync, SIGIO, POLL_IN); wake_up_all(&fc->waitq); - kobject_del(&fc->kobj); - kobject_put(&fc->kobj); + wake_up_all(&fc->blocked_waitq); + mutex_lock(&fuse_mutex); + list_del(&fc->entry); + fuse_ctl_remove_conn(fc); + mutex_unlock(&fuse_mutex); + fuse_conn_put(fc); } static void convert_fuse_statfs(struct kstatfs *stbuf, struct fuse_kstatfs *attr) @@ -251,6 +251,7 @@ static int fuse_statfs(struct dentry *dentry, struct kstatfs *buf) memset(&outarg, 0, sizeof(outarg)); req->in.numargs = 0; req->in.h.opcode = FUSE_STATFS; + req->in.h.nodeid = get_node_id(dentry->d_inode); req->out.numargs = 1; req->out.args[0].size = fc->minor < 4 ? FUSE_COMPAT_STATFS_SIZE : sizeof(outarg); @@ -369,11 +370,6 @@ static int fuse_show_options(struct seq_file *m, struct vfsmount *mnt) return 0; } -static void fuse_conn_release(struct kobject *kobj) -{ - kfree(get_fuse_conn_kobj(kobj)); -} - static struct fuse_conn *new_conn(void) { struct fuse_conn *fc; @@ -381,24 +377,35 @@ static struct fuse_conn *new_conn(void) fc = kzalloc(sizeof(*fc), GFP_KERNEL); if (fc) { spin_lock_init(&fc->lock); + atomic_set(&fc->count, 1); init_waitqueue_head(&fc->waitq); init_waitqueue_head(&fc->blocked_waitq); INIT_LIST_HEAD(&fc->pending); INIT_LIST_HEAD(&fc->processing); INIT_LIST_HEAD(&fc->io); - INIT_LIST_HEAD(&fc->background); - init_rwsem(&fc->sbput_sem); - kobj_set_kset_s(fc, connections_subsys); - kobject_init(&fc->kobj); + INIT_LIST_HEAD(&fc->interrupts); atomic_set(&fc->num_waiting, 0); fc->bdi.ra_pages = (VM_MAX_READAHEAD * 1024) / PAGE_CACHE_SIZE; fc->bdi.unplug_io_fn = default_unplug_io_fn; fc->reqctr = 0; fc->blocked = 1; + get_random_bytes(&fc->scramble_key, sizeof(fc->scramble_key)); } return fc; } +void fuse_conn_put(struct fuse_conn *fc) +{ + if (atomic_dec_and_test(&fc->count)) + kfree(fc); +} + +struct fuse_conn *fuse_conn_get(struct fuse_conn *fc) +{ + atomic_inc(&fc->count); + return fc; +} + static struct inode *get_root_inode(struct super_block *sb, unsigned mode) { struct fuse_attr attr; @@ -414,6 +421,7 @@ static struct super_operations fuse_super_operations = { .destroy_inode = fuse_destroy_inode, .read_inode = fuse_read_inode, .clear_inode = fuse_clear_inode, + .remount_fs = fuse_remount_fs, .put_super = fuse_put_super, .umount_begin = fuse_umount_begin, .statfs = fuse_statfs, @@ -433,8 +441,12 @@ static void process_init_reply(struct fuse_conn *fc, struct fuse_req *req) ra_pages = arg->max_readahead / PAGE_CACHE_SIZE; if (arg->flags & FUSE_ASYNC_READ) fc->async_read = 1; - } else + if (!(arg->flags & FUSE_POSIX_LOCKS)) + fc->no_lock = 1; + } else { ra_pages = fc->max_read / PAGE_CACHE_SIZE; + fc->no_lock = 1; + } fc->bdi.ra_pages = min(fc->bdi.ra_pages, ra_pages); fc->minor = arg->minor; @@ -452,7 +464,7 @@ static void fuse_send_init(struct fuse_conn *fc, struct fuse_req *req) arg->major = FUSE_KERNEL_VERSION; arg->minor = FUSE_KERNEL_MINOR_VERSION; arg->max_readahead = fc->bdi.ra_pages * PAGE_CACHE_SIZE; - arg->flags |= FUSE_ASYNC_READ; + arg->flags |= FUSE_ASYNC_READ | FUSE_POSIX_LOCKS; req->in.h.opcode = FUSE_INIT; req->in.numargs = 1; req->in.args[0].size = sizeof(*arg); @@ -468,10 +480,9 @@ static void fuse_send_init(struct fuse_conn *fc, struct fuse_req *req) request_send_background(fc, req); } -static unsigned long long conn_id(void) +static u64 conn_id(void) { - /* BKL is held for ->get_sb() */ - static unsigned long long ctr = 1; + static u64 ctr = 1; return ctr++; } @@ -485,6 +496,9 @@ static int fuse_fill_super(struct super_block *sb, void *data, int silent) struct fuse_req *init_req; int err; + if (sb->s_flags & MS_MANDLOCK) + return -EINVAL; + if (!parse_fuse_opt((char *) data, &d)) return -EINVAL; @@ -528,25 +542,21 @@ static int fuse_fill_super(struct super_block *sb, void *data, int silent) if (!init_req) goto err_put_root; - err = kobject_set_name(&fc->kobj, "%llu", conn_id()); - if (err) - goto err_free_req; - - err = kobject_add(&fc->kobj); - if (err) - goto err_free_req; - - /* Setting file->private_data can't race with other mount() - instances, since BKL is held for ->get_sb() */ + mutex_lock(&fuse_mutex); err = -EINVAL; if (file->private_data) - goto err_kobject_del; + goto err_unlock; + fc->id = conn_id(); + err = fuse_ctl_add_conn(fc); + if (err) + goto err_unlock; + + list_add_tail(&fc->entry, &fuse_conn_list); sb->s_root = root_dentry; - fc->mounted = 1; fc->connected = 1; - kobject_get(&fc->kobj); - file->private_data = fc; + file->private_data = fuse_conn_get(fc); + mutex_unlock(&fuse_mutex); /* * atomic_dec_and_test() in fput() provides the necessary * memory barrier for file->private_data to be visible on all @@ -558,15 +568,14 @@ static int fuse_fill_super(struct super_block *sb, void *data, int silent) return 0; - err_kobject_del: - kobject_del(&fc->kobj); - err_free_req: + err_unlock: + mutex_unlock(&fuse_mutex); fuse_request_free(init_req); err_put_root: dput(root_dentry); err: fput(file); - kobject_put(&fc->kobj); + fuse_conn_put(fc); return err; } @@ -584,68 +593,8 @@ static struct file_system_type fuse_fs_type = { .kill_sb = kill_anon_super, }; -static ssize_t fuse_conn_waiting_show(struct fuse_conn *fc, char *page) -{ - return sprintf(page, "%i\n", atomic_read(&fc->num_waiting)); -} - -static ssize_t fuse_conn_abort_store(struct fuse_conn *fc, const char *page, - size_t count) -{ - fuse_abort_conn(fc); - return count; -} - -static struct fuse_conn_attr fuse_conn_waiting = - __ATTR(waiting, 0400, fuse_conn_waiting_show, NULL); -static struct fuse_conn_attr fuse_conn_abort = - __ATTR(abort, 0600, NULL, fuse_conn_abort_store); - -static struct attribute *fuse_conn_attrs[] = { - &fuse_conn_waiting.attr, - &fuse_conn_abort.attr, - NULL, -}; - -static ssize_t fuse_conn_attr_show(struct kobject *kobj, - struct attribute *attr, - char *page) -{ - struct fuse_conn_attr *fca = - container_of(attr, struct fuse_conn_attr, attr); - - if (fca->show) - return fca->show(get_fuse_conn_kobj(kobj), page); - else - return -EACCES; -} - -static ssize_t fuse_conn_attr_store(struct kobject *kobj, - struct attribute *attr, - const char *page, size_t count) -{ - struct fuse_conn_attr *fca = - container_of(attr, struct fuse_conn_attr, attr); - - if (fca->store) - return fca->store(get_fuse_conn_kobj(kobj), page, count); - else - return -EACCES; -} - -static struct sysfs_ops fuse_conn_sysfs_ops = { - .show = &fuse_conn_attr_show, - .store = &fuse_conn_attr_store, -}; - -static struct kobj_type ktype_fuse_conn = { - .release = fuse_conn_release, - .sysfs_ops = &fuse_conn_sysfs_ops, - .default_attrs = fuse_conn_attrs, -}; - static decl_subsys(fuse, NULL, NULL); -static decl_subsys(connections, &ktype_fuse_conn, NULL); +static decl_subsys(connections, NULL, NULL); static void fuse_inode_init_once(void *foo, kmem_cache_t *cachep, unsigned long flags) @@ -719,6 +668,7 @@ static int __init fuse_init(void) printk("fuse init (API version %i.%i)\n", FUSE_KERNEL_VERSION, FUSE_KERNEL_MINOR_VERSION); + INIT_LIST_HEAD(&fuse_conn_list); res = fuse_fs_init(); if (res) goto err; @@ -731,8 +681,14 @@ static int __init fuse_init(void) if (res) goto err_dev_cleanup; + res = fuse_ctl_init(); + if (res) + goto err_sysfs_cleanup; + return 0; + err_sysfs_cleanup: + fuse_sysfs_cleanup(); err_dev_cleanup: fuse_dev_cleanup(); err_fs_cleanup: @@ -745,6 +701,7 @@ static void __exit fuse_exit(void) { printk(KERN_DEBUG "fuse exit\n"); + fuse_ctl_cleanup(); fuse_sysfs_cleanup(); fuse_fs_cleanup(); fuse_dev_cleanup(); diff --git a/fs/generic_acl.c b/fs/generic_acl.c new file mode 100644 index 00000000000..9ccb7894717 --- /dev/null +++ b/fs/generic_acl.c @@ -0,0 +1,197 @@ +/* + * fs/generic_acl.c + * + * (C) 2005 Andreas Gruenbacher <agruen@suse.de> + * + * This file is released under the GPL. + */ + +#include <linux/sched.h> +#include <linux/fs.h> +#include <linux/generic_acl.h> + +/** + * generic_acl_list - Generic xattr_handler->list() operation + * @ops: Filesystem specific getacl and setacl callbacks + */ +size_t +generic_acl_list(struct inode *inode, struct generic_acl_operations *ops, + int type, char *list, size_t list_size) +{ + struct posix_acl *acl; + const char *name; + size_t size; + + acl = ops->getacl(inode, type); + if (!acl) + return 0; + posix_acl_release(acl); + + switch(type) { + case ACL_TYPE_ACCESS: + name = POSIX_ACL_XATTR_ACCESS; + break; + + case ACL_TYPE_DEFAULT: + name = POSIX_ACL_XATTR_DEFAULT; + break; + + default: + return 0; + } + size = strlen(name) + 1; + if (list && size <= list_size) + memcpy(list, name, size); + return size; +} + +/** + * generic_acl_get - Generic xattr_handler->get() operation + * @ops: Filesystem specific getacl and setacl callbacks + */ +int +generic_acl_get(struct inode *inode, struct generic_acl_operations *ops, + int type, void *buffer, size_t size) +{ + struct posix_acl *acl; + int error; + + acl = ops->getacl(inode, type); + if (!acl) + return -ENODATA; + error = posix_acl_to_xattr(acl, buffer, size); + posix_acl_release(acl); + + return error; +} + +/** + * generic_acl_set - Generic xattr_handler->set() operation + * @ops: Filesystem specific getacl and setacl callbacks + */ +int +generic_acl_set(struct inode *inode, struct generic_acl_operations *ops, + int type, const void *value, size_t size) +{ + struct posix_acl *acl = NULL; + int error; + + if (S_ISLNK(inode->i_mode)) + return -EOPNOTSUPP; + if (current->fsuid != inode->i_uid && !capable(CAP_FOWNER)) + return -EPERM; + if (value) { + acl = posix_acl_from_xattr(value, size); + if (IS_ERR(acl)) + return PTR_ERR(acl); + } + if (acl) { + mode_t mode; + + error = posix_acl_valid(acl); + if (error) + goto failed; + switch(type) { + case ACL_TYPE_ACCESS: + mode = inode->i_mode; + error = posix_acl_equiv_mode(acl, &mode); + if (error < 0) + goto failed; + inode->i_mode = mode; + if (error == 0) { + posix_acl_release(acl); + acl = NULL; + } + break; + + case ACL_TYPE_DEFAULT: + if (!S_ISDIR(inode->i_mode)) { + error = -EINVAL; + goto failed; + } + break; + } + } + ops->setacl(inode, type, acl); + error = 0; +failed: + posix_acl_release(acl); + return error; +} + +/** + * generic_acl_init - Take care of acl inheritance at @inode create time + * @ops: Filesystem specific getacl and setacl callbacks + * + * Files created inside a directory with a default ACL inherit the + * directory's default ACL. + */ +int +generic_acl_init(struct inode *inode, struct inode *dir, + struct generic_acl_operations *ops) +{ + struct posix_acl *acl = NULL; + mode_t mode = inode->i_mode; + int error; + + inode->i_mode = mode & ~current->fs->umask; + if (!S_ISLNK(inode->i_mode)) + acl = ops->getacl(dir, ACL_TYPE_DEFAULT); + if (acl) { + struct posix_acl *clone; + + if (S_ISDIR(inode->i_mode)) { + clone = posix_acl_clone(acl, GFP_KERNEL); + error = -ENOMEM; + if (!clone) + goto cleanup; + ops->setacl(inode, ACL_TYPE_DEFAULT, clone); + posix_acl_release(clone); + } + clone = posix_acl_clone(acl, GFP_KERNEL); + error = -ENOMEM; + if (!clone) + goto cleanup; + error = posix_acl_create_masq(clone, &mode); + if (error >= 0) { + inode->i_mode = mode; + if (error > 0) + ops->setacl(inode, ACL_TYPE_ACCESS, clone); + } + posix_acl_release(clone); + } + error = 0; + +cleanup: + posix_acl_release(acl); + return error; +} + +/** + * generic_acl_chmod - change the access acl of @inode upon chmod() + * @ops: FIlesystem specific getacl and setacl callbacks + * + * A chmod also changes the permissions of the owner, group/mask, and + * other ACL entries. + */ +int +generic_acl_chmod(struct inode *inode, struct generic_acl_operations *ops) +{ + struct posix_acl *acl, *clone; + int error = 0; + + if (S_ISLNK(inode->i_mode)) + return -EOPNOTSUPP; + acl = ops->getacl(inode, ACL_TYPE_ACCESS); + if (acl) { + clone = posix_acl_clone(acl, GFP_KERNEL); + posix_acl_release(acl); + if (!clone) + return -ENOMEM; + error = posix_acl_chmod_masq(clone, inode->i_mode); + if (!error) + ops->setacl(inode, ACL_TYPE_ACCESS, clone); + posix_acl_release(clone); + } + return error; +} diff --git a/fs/hfs/bnode.c b/fs/hfs/bnode.c index 13231dd5ce6..0d200068d0a 100644 --- a/fs/hfs/bnode.c +++ b/fs/hfs/bnode.c @@ -249,10 +249,9 @@ static struct hfs_bnode *__hfs_bnode_create(struct hfs_btree *tree, u32 cnid) sb = tree->inode->i_sb; size = sizeof(struct hfs_bnode) + tree->pages_per_bnode * sizeof(struct page *); - node = kmalloc(size, GFP_KERNEL); + node = kzalloc(size, GFP_KERNEL); if (!node) return NULL; - memset(node, 0, size); node->tree = tree; node->this = cnid; set_bit(HFS_BNODE_NEW, &node->flags); diff --git a/fs/hfs/btree.c b/fs/hfs/btree.c index 40035799431..5fd0ed71f92 100644 --- a/fs/hfs/btree.c +++ b/fs/hfs/btree.c @@ -21,10 +21,9 @@ struct hfs_btree *hfs_btree_open(struct super_block *sb, u32 id, btree_keycmp ke struct page *page; unsigned int size; - tree = kmalloc(sizeof(*tree), GFP_KERNEL); + tree = kzalloc(sizeof(*tree), GFP_KERNEL); if (!tree) return NULL; - memset(tree, 0, sizeof(*tree)); init_MUTEX(&tree->tree_lock); spin_lock_init(&tree->hash_lock); diff --git a/fs/hfs/hfs_fs.h b/fs/hfs/hfs_fs.h index 3ed8663a8db..735332dfd1b 100644 --- a/fs/hfs/hfs_fs.h +++ b/fs/hfs/hfs_fs.h @@ -182,8 +182,8 @@ extern void hfs_file_truncate(struct inode *); extern int hfs_get_block(struct inode *, sector_t, struct buffer_head *, int); /* inode.c */ -extern struct address_space_operations hfs_aops; -extern struct address_space_operations hfs_btree_aops; +extern const struct address_space_operations hfs_aops; +extern const struct address_space_operations hfs_btree_aops; extern struct inode *hfs_new_inode(struct inode *, struct qstr *, int); extern void hfs_inode_write_fork(struct inode *, struct hfs_extent *, __be32 *, __be32 *); diff --git a/fs/hfs/inode.c b/fs/hfs/inode.c index 2d4ced22201..d05641c35fc 100644 --- a/fs/hfs/inode.c +++ b/fs/hfs/inode.c @@ -114,7 +114,7 @@ static int hfs_writepages(struct address_space *mapping, return mpage_writepages(mapping, wbc, hfs_get_block); } -struct address_space_operations hfs_btree_aops = { +const struct address_space_operations hfs_btree_aops = { .readpage = hfs_readpage, .writepage = hfs_writepage, .sync_page = block_sync_page, @@ -124,7 +124,7 @@ struct address_space_operations hfs_btree_aops = { .releasepage = hfs_releasepage, }; -struct address_space_operations hfs_aops = { +const struct address_space_operations hfs_aops = { .readpage = hfs_readpage, .writepage = hfs_writepage, .sync_page = block_sync_page, @@ -154,7 +154,6 @@ struct inode *hfs_new_inode(struct inode *dir, struct qstr *name, int mode) inode->i_gid = current->fsgid; inode->i_nlink = 1; inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME_SEC; - inode->i_blksize = HFS_SB(sb)->alloc_blksz; HFS_I(inode)->flags = 0; HFS_I(inode)->rsrc_inode = NULL; HFS_I(inode)->fs_blocks = 0; @@ -284,7 +283,6 @@ static int hfs_read_inode(struct inode *inode, void *data) inode->i_uid = hsb->s_uid; inode->i_gid = hsb->s_gid; inode->i_nlink = 1; - inode->i_blksize = HFS_SB(inode->i_sb)->alloc_blksz; if (idata->key) HFS_I(inode)->cat_key = *idata->key; diff --git a/fs/hfs/super.c b/fs/hfs/super.c index d9227bf14e8..d43b4fcc8ad 100644 --- a/fs/hfs/super.c +++ b/fs/hfs/super.c @@ -12,7 +12,6 @@ * Based on the minix file system code, (C) 1991, 1992 by Linus Torvalds */ -#include <linux/config.h> #include <linux/module.h> #include <linux/blkdev.h> #include <linux/mount.h> @@ -357,11 +356,10 @@ static int hfs_fill_super(struct super_block *sb, void *data, int silent) struct inode *root_inode; int res; - sbi = kmalloc(sizeof(struct hfs_sb_info), GFP_KERNEL); + sbi = kzalloc(sizeof(struct hfs_sb_info), GFP_KERNEL); if (!sbi) return -ENOMEM; sb->s_fs_info = sbi; - memset(sbi, 0, sizeof(struct hfs_sb_info)); INIT_HLIST_HEAD(&sbi->rsrc_inodes); res = -EINVAL; @@ -456,8 +454,7 @@ static int __init init_hfs_fs(void) static void __exit exit_hfs_fs(void) { unregister_filesystem(&hfs_fs_type); - if (kmem_cache_destroy(hfs_inode_cachep)) - printk(KERN_ERR "hfs_inode_cache: not all structures were freed\n"); + kmem_cache_destroy(hfs_inode_cachep); } module_init(init_hfs_fs) diff --git a/fs/hfsplus/bnode.c b/fs/hfsplus/bnode.c index 77bf434da67..29da6574ba7 100644 --- a/fs/hfsplus/bnode.c +++ b/fs/hfsplus/bnode.c @@ -409,10 +409,9 @@ static struct hfs_bnode *__hfs_bnode_create(struct hfs_btree *tree, u32 cnid) sb = tree->inode->i_sb; size = sizeof(struct hfs_bnode) + tree->pages_per_bnode * sizeof(struct page *); - node = kmalloc(size, GFP_KERNEL); + node = kzalloc(size, GFP_KERNEL); if (!node) return NULL; - memset(node, 0, size); node->tree = tree; node->this = cnid; set_bit(HFS_BNODE_NEW, &node->flags); diff --git a/fs/hfsplus/btree.c b/fs/hfsplus/btree.c index cfc852fdd1b..a9b9e872e29 100644 --- a/fs/hfsplus/btree.c +++ b/fs/hfsplus/btree.c @@ -24,10 +24,9 @@ struct hfs_btree *hfs_btree_open(struct super_block *sb, u32 id) struct page *page; unsigned int size; - tree = kmalloc(sizeof(*tree), GFP_KERNEL); + tree = kzalloc(sizeof(*tree), GFP_KERNEL); if (!tree) return NULL; - memset(tree, 0, sizeof(*tree)); init_MUTEX(&tree->tree_lock); spin_lock_init(&tree->hash_lock); diff --git a/fs/hfsplus/hfsplus_fs.h b/fs/hfsplus/hfsplus_fs.h index 7ae393637a0..8a1ca5ef7ad 100644 --- a/fs/hfsplus/hfsplus_fs.h +++ b/fs/hfsplus/hfsplus_fs.h @@ -323,8 +323,8 @@ int hfsplus_file_extend(struct inode *); void hfsplus_file_truncate(struct inode *); /* inode.c */ -extern struct address_space_operations hfsplus_aops; -extern struct address_space_operations hfsplus_btree_aops; +extern const struct address_space_operations hfsplus_aops; +extern const struct address_space_operations hfsplus_btree_aops; void hfsplus_inode_read_fork(struct inode *, struct hfsplus_fork_raw *); void hfsplus_inode_write_fork(struct inode *, struct hfsplus_fork_raw *); diff --git a/fs/hfsplus/inode.c b/fs/hfsplus/inode.c index acf66dba3e0..0eb1a609266 100644 --- a/fs/hfsplus/inode.c +++ b/fs/hfsplus/inode.c @@ -109,7 +109,7 @@ static int hfsplus_writepages(struct address_space *mapping, return mpage_writepages(mapping, wbc, hfsplus_get_block); } -struct address_space_operations hfsplus_btree_aops = { +const struct address_space_operations hfsplus_btree_aops = { .readpage = hfsplus_readpage, .writepage = hfsplus_writepage, .sync_page = block_sync_page, @@ -119,7 +119,7 @@ struct address_space_operations hfsplus_btree_aops = { .releasepage = hfsplus_releasepage, }; -struct address_space_operations hfsplus_aops = { +const struct address_space_operations hfsplus_aops = { .readpage = hfsplus_readpage, .writepage = hfsplus_writepage, .sync_page = block_sync_page, @@ -304,7 +304,6 @@ struct inode *hfsplus_new_inode(struct super_block *sb, int mode) inode->i_gid = current->fsgid; inode->i_nlink = 1; inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME_SEC; - inode->i_blksize = HFSPLUS_SB(sb).alloc_blksz; INIT_LIST_HEAD(&HFSPLUS_I(inode).open_dir_list); init_MUTEX(&HFSPLUS_I(inode).extents_lock); atomic_set(&HFSPLUS_I(inode).opencnt, 0); @@ -407,7 +406,6 @@ int hfsplus_cat_read_inode(struct inode *inode, struct hfs_find_data *fd) type = hfs_bnode_read_u16(fd->bnode, fd->entryoffset); HFSPLUS_I(inode).dev = 0; - inode->i_blksize = HFSPLUS_SB(inode->i_sb).alloc_blksz; if (type == HFSPLUS_FOLDER) { struct hfsplus_cat_folder *folder = &entry.folder; diff --git a/fs/hfsplus/super.c b/fs/hfsplus/super.c index 0a92fa2336a..194eede52fa 100644 --- a/fs/hfsplus/super.c +++ b/fs/hfsplus/super.c @@ -7,7 +7,6 @@ * */ -#include <linux/config.h> #include <linux/module.h> #include <linux/init.h> #include <linux/pagemap.h> @@ -494,8 +493,7 @@ static int __init init_hfsplus_fs(void) static void __exit exit_hfsplus_fs(void) { unregister_filesystem(&hfsplus_fs_type); - if (kmem_cache_destroy(hfsplus_inode_cachep)) - printk(KERN_ERR "hfsplus_inode_cache: not all structures were freed\n"); + kmem_cache_destroy(hfsplus_inode_cachep); } module_init(init_hfsplus_fs) diff --git a/fs/hostfs/hostfs_kern.c b/fs/hostfs/hostfs_kern.c index 8e0d37743e7..322e876c35e 100644 --- a/fs/hostfs/hostfs_kern.c +++ b/fs/hostfs/hostfs_kern.c @@ -54,7 +54,7 @@ static int append = 0; static struct inode_operations hostfs_iops; static struct inode_operations hostfs_dir_iops; -static struct address_space_operations hostfs_link_aops; +static const struct address_space_operations hostfs_link_aops; #ifndef MODULE static int __init hostfs_args(char *options, int *add) @@ -156,7 +156,6 @@ static int read_name(struct inode *ino, char *name) ino->i_mode = i_mode; ino->i_nlink = i_nlink; ino->i_size = i_size; - ino->i_blksize = i_blksize; ino->i_blocks = i_blocks; return(0); } @@ -518,7 +517,7 @@ int hostfs_commit_write(struct file *file, struct page *page, unsigned from, return(err); } -static struct address_space_operations hostfs_aops = { +static const struct address_space_operations hostfs_aops = { .writepage = hostfs_writepage, .readpage = hostfs_readpage, .set_page_dirty = __set_page_dirty_nobuffers, @@ -935,7 +934,7 @@ int hostfs_link_readpage(struct file *file, struct page *page) return(err); } -static struct address_space_operations hostfs_link_aops = { +static const struct address_space_operations hostfs_link_aops = { .readpage = hostfs_link_readpage, }; diff --git a/fs/hpfs/buffer.c b/fs/hpfs/buffer.c index 2807aa833e6..b52b7381d10 100644 --- a/fs/hpfs/buffer.c +++ b/fs/hpfs/buffer.c @@ -76,7 +76,7 @@ void *hpfs_map_4sectors(struct super_block *s, unsigned secno, struct quad_buffe return NULL; } - qbh->data = data = (char *)kmalloc(2048, GFP_NOFS); + qbh->data = data = kmalloc(2048, GFP_NOFS); if (!data) { printk("HPFS: hpfs_map_4sectors: out of memory\n"); goto bail; diff --git a/fs/hpfs/file.c b/fs/hpfs/file.c index d3b9fffe45a..d9eb19b7b8a 100644 --- a/fs/hpfs/file.c +++ b/fs/hpfs/file.c @@ -99,7 +99,7 @@ static sector_t _hpfs_bmap(struct address_space *mapping, sector_t block) { return generic_block_bmap(mapping,block,hpfs_get_block); } -struct address_space_operations hpfs_aops = { +const struct address_space_operations hpfs_aops = { .readpage = hpfs_readpage, .writepage = hpfs_writepage, .sync_page = block_sync_page, diff --git a/fs/hpfs/hpfs_fn.h b/fs/hpfs/hpfs_fn.h index 29b7a3e5517..32ab51e42b9 100644 --- a/fs/hpfs/hpfs_fn.h +++ b/fs/hpfs/hpfs_fn.h @@ -12,7 +12,6 @@ #include <linux/mutex.h> #include <linux/pagemap.h> #include <linux/buffer_head.h> -#include <linux/hpfs_fs.h> #include <linux/slab.h> #include <linux/smp_lock.h> @@ -268,7 +267,7 @@ void hpfs_set_ea(struct inode *, struct fnode *, char *, char *, int); int hpfs_file_fsync(struct file *, struct dentry *, int); extern const struct file_operations hpfs_file_ops; extern struct inode_operations hpfs_file_iops; -extern struct address_space_operations hpfs_aops; +extern const struct address_space_operations hpfs_aops; /* inode.c */ @@ -304,7 +303,7 @@ void hpfs_decide_conv(struct inode *, unsigned char *, unsigned); /* namei.c */ extern struct inode_operations hpfs_dir_iops; -extern struct address_space_operations hpfs_symlink_aops; +extern const struct address_space_operations hpfs_symlink_aops; static inline struct hpfs_inode_info *hpfs_i(struct inode *inode) { diff --git a/fs/hpfs/inode.c b/fs/hpfs/inode.c index 56f2c338c4d..bcf6ee36e06 100644 --- a/fs/hpfs/inode.c +++ b/fs/hpfs/inode.c @@ -17,7 +17,6 @@ void hpfs_init_inode(struct inode *i) i->i_gid = hpfs_sb(sb)->sb_gid; i->i_mode = hpfs_sb(sb)->sb_mode; hpfs_inode->i_conv = hpfs_sb(sb)->sb_conv; - i->i_blksize = 512; i->i_size = -1; i->i_blocks = -1; diff --git a/fs/hpfs/namei.c b/fs/hpfs/namei.c index a03abb12c61..59e7dc182a0 100644 --- a/fs/hpfs/namei.c +++ b/fs/hpfs/namei.c @@ -538,7 +538,7 @@ fail: return err; } -struct address_space_operations hpfs_symlink_aops = { +const struct address_space_operations hpfs_symlink_aops = { .readpage = hpfs_symlink_readpage }; diff --git a/fs/hpfs/super.c b/fs/hpfs/super.c index f798480a363..450b5e0b478 100644 --- a/fs/hpfs/super.c +++ b/fs/hpfs/super.c @@ -11,6 +11,7 @@ #include <linux/parser.h> #include <linux/init.h> #include <linux/statfs.h> +#include <linux/magic.h> /* Mark the filesystem dirty, so that chkdsk checks it when os/2 booted */ @@ -202,8 +203,7 @@ static int init_inodecache(void) static void destroy_inodecache(void) { - if (kmem_cache_destroy(hpfs_inode_cachep)) - printk(KERN_INFO "hpfs_inode_cache: not all structures were freed\n"); + kmem_cache_destroy(hpfs_inode_cachep); } /* @@ -461,11 +461,10 @@ static int hpfs_fill_super(struct super_block *s, void *options, int silent) int o; - sbi = kmalloc(sizeof(*sbi), GFP_KERNEL); + sbi = kzalloc(sizeof(*sbi), GFP_KERNEL); if (!sbi) return -ENOMEM; s->s_fs_info = sbi; - memset(sbi, 0, sizeof(*sbi)); sbi->sb_bmp_dir = NULL; sbi->sb_cp_table = NULL; diff --git a/fs/hppfs/hppfs_kern.c b/fs/hppfs/hppfs_kern.c index 3a9bdf58166..dcb6d2e988b 100644 --- a/fs/hppfs/hppfs_kern.c +++ b/fs/hppfs/hppfs_kern.c @@ -152,7 +152,6 @@ static void hppfs_read_inode(struct inode *ino) ino->i_mode = proc_ino->i_mode; ino->i_nlink = proc_ino->i_nlink; ino->i_size = proc_ino->i_size; - ino->i_blksize = proc_ino->i_blksize; ino->i_blocks = proc_ino->i_blocks; } diff --git a/fs/hugetlbfs/inode.c b/fs/hugetlbfs/inode.c index e6410d8edd0..f5b8f329aca 100644 --- a/fs/hugetlbfs/inode.c +++ b/fs/hugetlbfs/inode.c @@ -34,7 +34,7 @@ #define HUGETLBFS_MAGIC 0x958458f6 static struct super_operations hugetlbfs_ops; -static struct address_space_operations hugetlbfs_aops; +static const struct address_space_operations hugetlbfs_aops; const struct file_operations hugetlbfs_file_operations; static struct inode_operations hugetlbfs_dir_inode_operations; static struct inode_operations hugetlbfs_inode_operations; @@ -83,8 +83,6 @@ static int hugetlbfs_file_mmap(struct file *file, struct vm_area_struct *vma) ret = -ENOMEM; len = vma_len + ((loff_t)vma->vm_pgoff << PAGE_SHIFT); - if (!(vma->vm_flags & VM_WRITE) && len > inode->i_size) - goto out; if (vma->vm_flags & VM_MAYSHARE && hugetlb_reserve_pages(inode, vma->vm_pgoff >> (HPAGE_SHIFT-PAGE_SHIFT), @@ -93,7 +91,7 @@ static int hugetlbfs_file_mmap(struct file *file, struct vm_area_struct *vma) ret = 0; hugetlb_prefault_arch_hook(vma->vm_mm); - if (inode->i_size < len) + if (vma->vm_flags & VM_WRITE && inode->i_size < len) inode->i_size = len; out: mutex_unlock(&inode->i_mutex); @@ -231,7 +229,7 @@ static void hugetlbfs_delete_inode(struct inode *inode) clear_inode(inode); } -static void hugetlbfs_forget_inode(struct inode *inode) +static void hugetlbfs_forget_inode(struct inode *inode) __releases(inode_lock) { struct super_block *sb = inode->i_sb; @@ -359,7 +357,6 @@ static struct inode *hugetlbfs_get_inode(struct super_block *sb, uid_t uid, inode->i_mode = mode; inode->i_uid = uid; inode->i_gid = gid; - inode->i_blksize = HPAGE_SIZE; inode->i_blocks = 0; inode->i_mapping->a_ops = &hugetlbfs_aops; inode->i_mapping->backing_dev_info =&hugetlbfs_backing_dev_info; @@ -547,7 +544,7 @@ static void hugetlbfs_destroy_inode(struct inode *inode) kmem_cache_free(hugetlbfs_inode_cachep, HUGETLBFS_I(inode)); } -static struct address_space_operations hugetlbfs_aops = { +static const struct address_space_operations hugetlbfs_aops = { .readpage = hugetlbfs_readpage, .prepare_write = hugetlbfs_prepare_write, .commit_write = hugetlbfs_commit_write, diff --git a/fs/inode.c b/fs/inode.c index 3a2446a27d2..abf77471e6c 100644 --- a/fs/inode.c +++ b/fs/inode.c @@ -4,7 +4,6 @@ * (C) 1997 Linus Torvalds */ -#include <linux/config.h> #include <linux/fs.h> #include <linux/mm.h> #include <linux/dcache.h> @@ -102,7 +101,7 @@ static kmem_cache_t * inode_cachep __read_mostly; static struct inode *alloc_inode(struct super_block *sb) { - static struct address_space_operations empty_aops; + static const struct address_space_operations empty_aops; static struct inode_operations empty_iops; static const struct file_operations empty_fops; struct inode *inode; @@ -134,7 +133,6 @@ static struct inode *alloc_inode(struct super_block *sb) inode->i_bdev = NULL; inode->i_cdev = NULL; inode->i_rdev = 0; - inode->i_security = NULL; inode->dirtied_when = 0; if (security_inode_alloc(inode)) { if (inode->i_sb->s_op->destroy_inode) @@ -164,7 +162,7 @@ static struct inode *alloc_inode(struct super_block *sb) bdi = sb->s_bdev->bd_inode->i_mapping->backing_dev_info; mapping->backing_dev_info = bdi; } - memset(&inode->u, 0, sizeof(inode->u)); + inode->i_private = 0; inode->i_mapping = mapping; } return inode; @@ -255,9 +253,9 @@ void clear_inode(struct inode *inode) DQUOT_DROP(inode); if (inode->i_sb && inode->i_sb->s_op->clear_inode) inode->i_sb->s_op->clear_inode(inode); - if (inode->i_bdev) + if (S_ISBLK(inode->i_mode) && inode->i_bdev) bd_forget(inode); - if (inode->i_cdev) + if (S_ISCHR(inode->i_mode) && inode->i_cdev) cd_forget(inode); inode->i_state = I_CLEAR; } @@ -452,15 +450,14 @@ static void prune_icache(int nr_to_scan) nr_pruned++; } inodes_stat.nr_unused -= nr_pruned; + if (current_is_kswapd()) + __count_vm_events(KSWAPD_INODESTEAL, reap); + else + __count_vm_events(PGINODESTEAL, reap); spin_unlock(&inode_lock); dispose_list(&freeable); mutex_unlock(&iprune_mutex); - - if (current_is_kswapd()) - mod_page_state(kswapd_inodesteal, reap); - else - mod_page_state(pginodesteal, reap); } /* diff --git a/fs/inotify_user.c b/fs/inotify_user.c index f2386442ade..017cb0f134d 100644 --- a/fs/inotify_user.c +++ b/fs/inotify_user.c @@ -187,7 +187,7 @@ static struct inotify_kernel_event * kernel_event(s32 wd, u32 mask, u32 cookie, { struct inotify_kernel_event *kevent; - kevent = kmem_cache_alloc(event_cachep, GFP_KERNEL); + kevent = kmem_cache_alloc(event_cachep, GFP_NOFS); if (unlikely(!kevent)) return NULL; diff --git a/fs/ioctl.c b/fs/ioctl.c index f8aeec3ca10..4b7660b09ac 100644 --- a/fs/ioctl.c +++ b/fs/ioctl.c @@ -4,7 +4,6 @@ * Copyright (C) 1991, 1992 Linus Torvalds */ -#include <linux/config.h> #include <linux/syscalls.h> #include <linux/mm.h> #include <linux/smp_lock.h> diff --git a/fs/ioprio.c b/fs/ioprio.c index 7fa76ed53c1..78b1deae3fa 100644 --- a/fs/ioprio.c +++ b/fs/ioprio.c @@ -44,6 +44,9 @@ static int set_task_ioprio(struct task_struct *task, int ioprio) task->ioprio = ioprio; ioc = task->io_context; + /* see wmb() in current_io_context() */ + smp_read_barrier_depends(); + if (ioc && ioc->set_ioprio) ioc->set_ioprio(ioc, ioprio); @@ -111,9 +114,9 @@ asmlinkage long sys_ioprio_set(int which, int who, int ioprio) continue; ret = set_task_ioprio(p, ioprio); if (ret) - break; + goto free_uid; } while_each_thread(g, p); - +free_uid: if (who) free_uid(user); break; @@ -125,11 +128,47 @@ asmlinkage long sys_ioprio_set(int which, int who, int ioprio) return ret; } +static int get_task_ioprio(struct task_struct *p) +{ + int ret; + + ret = security_task_getioprio(p); + if (ret) + goto out; + ret = p->ioprio; +out: + return ret; +} + +int ioprio_best(unsigned short aprio, unsigned short bprio) +{ + unsigned short aclass = IOPRIO_PRIO_CLASS(aprio); + unsigned short bclass = IOPRIO_PRIO_CLASS(bprio); + + if (!ioprio_valid(aprio)) + return bprio; + if (!ioprio_valid(bprio)) + return aprio; + + if (aclass == IOPRIO_CLASS_NONE) + aclass = IOPRIO_CLASS_BE; + if (bclass == IOPRIO_CLASS_NONE) + bclass = IOPRIO_CLASS_BE; + + if (aclass == bclass) + return min(aprio, bprio); + if (aclass > bclass) + return bprio; + else + return aprio; +} + asmlinkage long sys_ioprio_get(int which, int who) { struct task_struct *g, *p; struct user_struct *user; int ret = -ESRCH; + int tmpio; read_lock_irq(&tasklist_lock); switch (which) { @@ -139,16 +178,19 @@ asmlinkage long sys_ioprio_get(int which, int who) else p = find_task_by_pid(who); if (p) - ret = p->ioprio; + ret = get_task_ioprio(p); break; case IOPRIO_WHO_PGRP: if (!who) who = process_group(current); do_each_task_pid(who, PIDTYPE_PGID, p) { + tmpio = get_task_ioprio(p); + if (tmpio < 0) + continue; if (ret == -ESRCH) - ret = p->ioprio; + ret = tmpio; else - ret = ioprio_best(ret, p->ioprio); + ret = ioprio_best(ret, tmpio); } while_each_task_pid(who, PIDTYPE_PGID, p); break; case IOPRIO_WHO_USER: @@ -163,10 +205,13 @@ asmlinkage long sys_ioprio_get(int which, int who) do_each_thread(g, p) { if (p->uid != user->uid) continue; + tmpio = get_task_ioprio(p); + if (tmpio < 0) + continue; if (ret == -ESRCH) - ret = p->ioprio; + ret = tmpio; else - ret = ioprio_best(ret, p->ioprio); + ret = ioprio_best(ret, tmpio); } while_each_thread(g, p); if (who) diff --git a/fs/isofs/compress.c b/fs/isofs/compress.c index 4917315db73..731816332b1 100644 --- a/fs/isofs/compress.c +++ b/fs/isofs/compress.c @@ -16,7 +16,6 @@ * Transparent decompression of files on an iso9660 filesystem */ -#include <linux/config.h> #include <linux/module.h> #include <linux/init.h> @@ -312,7 +311,7 @@ eio: return err; } -struct address_space_operations zisofs_aops = { +const struct address_space_operations zisofs_aops = { .readpage = zisofs_readpage, /* No sync_page operation supported? */ /* No bmap operation supported */ diff --git a/fs/isofs/dir.c b/fs/isofs/dir.c index 5440ea292c6..27e276987fd 100644 --- a/fs/isofs/dir.c +++ b/fs/isofs/dir.c @@ -10,7 +10,6 @@ * * isofs directory handling functions */ -#include <linux/config.h> #include <linux/smp_lock.h> #include "isofs.h" diff --git a/fs/isofs/inode.c b/fs/isofs/inode.c index 3f9c8ba1fa1..c34b862cdbf 100644 --- a/fs/isofs/inode.c +++ b/fs/isofs/inode.c @@ -11,7 +11,6 @@ * 2004 Paul Serice - NFS Export Operations */ -#include <linux/config.h> #include <linux/init.h> #include <linux/module.h> @@ -97,9 +96,7 @@ static int init_inodecache(void) static void destroy_inodecache(void) { - if (kmem_cache_destroy(isofs_inode_cachep)) - printk(KERN_INFO "iso_inode_cache: not all structures were " - "freed\n"); + kmem_cache_destroy(isofs_inode_cachep); } static int isofs_remount(struct super_block *sb, int *flags, char *data) @@ -558,11 +555,10 @@ static int isofs_fill_super(struct super_block *s, void *data, int silent) struct iso9660_options opt; struct isofs_sb_info * sbi; - sbi = kmalloc(sizeof(*sbi), GFP_KERNEL); + sbi = kzalloc(sizeof(*sbi), GFP_KERNEL); if (!sbi) return -ENOMEM; s->s_fs_info = sbi; - memset(sbi, 0, sizeof(*sbi)); if (!parse_options((char *)data, &opt)) goto out_freesbi; @@ -964,30 +960,30 @@ int isofs_get_blocks(struct inode *inode, sector_t iblock_s, goto abort; } - if (nextblk) { - while (b_off >= (offset + sect_size)) { - struct inode *ninode; - - offset += sect_size; - if (nextblk == 0) - goto abort; - ninode = isofs_iget(inode->i_sb, nextblk, nextoff); - if (!ninode) - goto abort; - firstext = ISOFS_I(ninode)->i_first_extent; - sect_size = ISOFS_I(ninode)->i_section_size >> ISOFS_BUFFER_BITS(ninode); - nextblk = ISOFS_I(ninode)->i_next_section_block; - nextoff = ISOFS_I(ninode)->i_next_section_offset; - iput(ninode); - - if (++section > 100) { - printk("isofs_get_blocks: More than 100 file sections ?!?, aborting...\n"); - printk("isofs_get_blocks: block=%ld firstext=%u sect_size=%u " - "nextblk=%lu nextoff=%lu\n", - iblock, firstext, (unsigned) sect_size, - nextblk, nextoff); - goto abort; - } + /* On the last section, nextblk == 0, section size is likely to + * exceed sect_size by a partial block, and access beyond the + * end of the file will reach beyond the section size, too. + */ + while (nextblk && (b_off >= (offset + sect_size))) { + struct inode *ninode; + + offset += sect_size; + ninode = isofs_iget(inode->i_sb, nextblk, nextoff); + if (!ninode) + goto abort; + firstext = ISOFS_I(ninode)->i_first_extent; + sect_size = ISOFS_I(ninode)->i_section_size >> ISOFS_BUFFER_BITS(ninode); + nextblk = ISOFS_I(ninode)->i_next_section_block; + nextoff = ISOFS_I(ninode)->i_next_section_offset; + iput(ninode); + + if (++section > 100) { + printk("isofs_get_blocks: More than 100 file sections ?!?, aborting...\n"); + printk("isofs_get_blocks: block=%ld firstext=%u sect_size=%u " + "nextblk=%lu nextoff=%lu\n", + iblock, firstext, (unsigned) sect_size, + nextblk, nextoff); + goto abort; } } @@ -1054,7 +1050,7 @@ static sector_t _isofs_bmap(struct address_space *mapping, sector_t block) return generic_block_bmap(mapping,block,isofs_get_block); } -static struct address_space_operations isofs_aops = { +static const struct address_space_operations isofs_aops = { .readpage = isofs_readpage, .sync_page = block_sync_page, .bmap = _isofs_bmap @@ -1239,7 +1235,7 @@ static void isofs_read_inode(struct inode *inode) } inode->i_uid = sbi->s_uid; inode->i_gid = sbi->s_gid; - inode->i_blocks = inode->i_blksize = 0; + inode->i_blocks = 0; ei->i_format_parm[0] = 0; ei->i_format_parm[1] = 0; @@ -1295,7 +1291,6 @@ static void isofs_read_inode(struct inode *inode) isonum_711 (de->ext_attr_length)); /* Set the number of blocks for stat() - should be done before RR */ - inode->i_blksize = PAGE_CACHE_SIZE; /* For stat() only */ inode->i_blocks = (inode->i_size + 511) >> 9; /* diff --git a/fs/isofs/isofs.h b/fs/isofs/isofs.h index b87ba066f5e..e6308c8b573 100644 --- a/fs/isofs/isofs.h +++ b/fs/isofs/isofs.h @@ -176,5 +176,5 @@ isofs_normalize_block_and_offset(struct iso_directory_record* de, extern struct inode_operations isofs_dir_inode_operations; extern const struct file_operations isofs_dir_operations; -extern struct address_space_operations isofs_symlink_aops; +extern const struct address_space_operations isofs_symlink_aops; extern struct export_operations isofs_export_ops; diff --git a/fs/isofs/rock.c b/fs/isofs/rock.c index 4326cb47f8f..f3a1db3098d 100644 --- a/fs/isofs/rock.c +++ b/fs/isofs/rock.c @@ -754,6 +754,6 @@ error: return -EIO; } -struct address_space_operations isofs_symlink_aops = { +const struct address_space_operations isofs_symlink_aops = { .readpage = rock_ridge_symlink_readpage }; diff --git a/fs/isofs/zisofs.h b/fs/isofs/zisofs.h index d78485d101c..27379570915 100644 --- a/fs/isofs/zisofs.h +++ b/fs/isofs/zisofs.h @@ -15,7 +15,7 @@ */ #ifdef CONFIG_ZISOFS -extern struct address_space_operations zisofs_aops; +extern const struct address_space_operations zisofs_aops; extern int __init zisofs_init(void); extern void zisofs_cleanup(void); #endif diff --git a/fs/jbd/checkpoint.c b/fs/jbd/checkpoint.c index 47678a26c13..0208cc7ac5d 100644 --- a/fs/jbd/checkpoint.c +++ b/fs/jbd/checkpoint.c @@ -1,6 +1,6 @@ /* * linux/fs/checkpoint.c - * + * * Written by Stephen C. Tweedie <sct@redhat.com>, 1999 * * Copyright 1999 Red Hat Software --- All Rights Reserved @@ -9,8 +9,8 @@ * the terms of the GNU General Public License, version 2, or at your * option, any later version, incorporated herein by reference. * - * Checkpoint routines for the generic filesystem journaling code. - * Part of the ext2fs journaling system. + * Checkpoint routines for the generic filesystem journaling code. + * Part of the ext2fs journaling system. * * Checkpointing is the process of ensuring that a section of the log is * committed fully to disk, so that that portion of the log can be @@ -145,6 +145,7 @@ void __log_wait_for_space(journal_t *journal) * jbd_unlock_bh_state(). */ static void jbd_sync_bh(journal_t *journal, struct buffer_head *bh) + __releases(journal->j_list_lock) { get_bh(bh); spin_unlock(&journal->j_list_lock); @@ -225,7 +226,7 @@ __flush_batch(journal_t *journal, struct buffer_head **bhs, int *batch_count) * Try to flush one buffer from the checkpoint list to disk. * * Return 1 if something happened which requires us to abort the current - * scan of the checkpoint list. + * scan of the checkpoint list. * * Called with j_list_lock held and drops it if 1 is returned * Called under jbd_lock_bh_state(jh2bh(jh)), and drops it @@ -269,7 +270,7 @@ static int __process_buffer(journal_t *journal, struct journal_head *jh, * possibly block, while still holding the journal lock. * We cannot afford to let the transaction logic start * messing around with this buffer before we write it to - * disk, as that would break recoverability. + * disk, as that would break recoverability. */ BUFFER_TRACE(bh, "queue"); get_bh(bh); @@ -292,7 +293,7 @@ static int __process_buffer(journal_t *journal, struct journal_head *jh, * Perform an actual checkpoint. We take the first transaction on the * list of transactions to be checkpointed and send all its buffers * to disk. We submit larger chunks of data at once. - * + * * The journal should be locked before calling this function. */ int log_do_checkpoint(journal_t *journal) @@ -303,10 +304,10 @@ int log_do_checkpoint(journal_t *journal) jbd_debug(1, "Start checkpoint\n"); - /* + /* * First thing: if there are any transactions in the log which * don't need checkpointing, just eliminate them from the - * journal straight away. + * journal straight away. */ result = cleanup_journal_tail(journal); jbd_debug(1, "cleanup_journal_tail returned %d\n", result); @@ -384,9 +385,9 @@ out: * we have already got rid of any since the last update of the log tail * in the journal superblock. If so, we can instantly roll the * superblock forward to remove those transactions from the log. - * + * * Return <0 on error, 0 on success, 1 if there was nothing to clean up. - * + * * Called with the journal lock held. * * This is the only part of the journaling code which really needs to be @@ -403,8 +404,8 @@ int cleanup_journal_tail(journal_t *journal) unsigned long blocknr, freed; /* OK, work out the oldest transaction remaining in the log, and - * the log block it starts at. - * + * the log block it starts at. + * * If the log is now empty, we need to work out which is the * next transaction ID we will write, and where it will * start. */ @@ -479,7 +480,7 @@ static int journal_clean_one_cp_list(struct journal_head *jh, int *released) if (!jh) return 0; - last_jh = jh->b_cpprev; + last_jh = jh->b_cpprev; do { jh = next_jh; next_jh = jh->b_cpnext; @@ -557,7 +558,7 @@ out: return ret; } -/* +/* * journal_remove_checkpoint: called after a buffer has been committed * to disk (either by being write-back flushed to disk, or being * committed to the log). @@ -635,7 +636,7 @@ out: * Called with the journal locked. * Called with j_list_lock held. */ -void __journal_insert_checkpoint(struct journal_head *jh, +void __journal_insert_checkpoint(struct journal_head *jh, transaction_t *transaction) { JBUFFER_TRACE(jh, "entry"); @@ -657,7 +658,7 @@ void __journal_insert_checkpoint(struct journal_head *jh, /* * We've finished with this transaction structure: adios... - * + * * The transaction must have no links except for the checkpoint by this * point. * diff --git a/fs/jbd/commit.c b/fs/jbd/commit.c index 0971814c38b..32a8caf0c41 100644 --- a/fs/jbd/commit.c +++ b/fs/jbd/commit.c @@ -160,6 +160,117 @@ static int journal_write_commit_record(journal_t *journal, return (ret == -EIO); } +static void journal_do_submit_data(struct buffer_head **wbuf, int bufs) +{ + int i; + + for (i = 0; i < bufs; i++) { + wbuf[i]->b_end_io = end_buffer_write_sync; + /* We use-up our safety reference in submit_bh() */ + submit_bh(WRITE, wbuf[i]); + } +} + +/* + * Submit all the data buffers to disk + */ +static void journal_submit_data_buffers(journal_t *journal, + transaction_t *commit_transaction) +{ + struct journal_head *jh; + struct buffer_head *bh; + int locked; + int bufs = 0; + struct buffer_head **wbuf = journal->j_wbuf; + + /* + * Whenever we unlock the journal and sleep, things can get added + * onto ->t_sync_datalist, so we have to keep looping back to + * write_out_data until we *know* that the list is empty. + * + * Cleanup any flushed data buffers from the data list. Even in + * abort mode, we want to flush this out as soon as possible. + */ +write_out_data: + cond_resched(); + spin_lock(&journal->j_list_lock); + + while (commit_transaction->t_sync_datalist) { + jh = commit_transaction->t_sync_datalist; + bh = jh2bh(jh); + locked = 0; + + /* Get reference just to make sure buffer does not disappear + * when we are forced to drop various locks */ + get_bh(bh); + /* If the buffer is dirty, we need to submit IO and hence + * we need the buffer lock. We try to lock the buffer without + * blocking. If we fail, we need to drop j_list_lock and do + * blocking lock_buffer(). + */ + if (buffer_dirty(bh)) { + if (test_set_buffer_locked(bh)) { + BUFFER_TRACE(bh, "needs blocking lock"); + spin_unlock(&journal->j_list_lock); + /* Write out all data to prevent deadlocks */ + journal_do_submit_data(wbuf, bufs); + bufs = 0; + lock_buffer(bh); + spin_lock(&journal->j_list_lock); + } + locked = 1; + } + /* We have to get bh_state lock. Again out of order, sigh. */ + if (!inverted_lock(journal, bh)) { + jbd_lock_bh_state(bh); + spin_lock(&journal->j_list_lock); + } + /* Someone already cleaned up the buffer? */ + if (!buffer_jbd(bh) + || jh->b_transaction != commit_transaction + || jh->b_jlist != BJ_SyncData) { + jbd_unlock_bh_state(bh); + if (locked) + unlock_buffer(bh); + BUFFER_TRACE(bh, "already cleaned up"); + put_bh(bh); + continue; + } + if (locked && test_clear_buffer_dirty(bh)) { + BUFFER_TRACE(bh, "needs writeout, adding to array"); + wbuf[bufs++] = bh; + __journal_file_buffer(jh, commit_transaction, + BJ_Locked); + jbd_unlock_bh_state(bh); + if (bufs == journal->j_wbufsize) { + spin_unlock(&journal->j_list_lock); + journal_do_submit_data(wbuf, bufs); + bufs = 0; + goto write_out_data; + } + } + else { + BUFFER_TRACE(bh, "writeout complete: unfile"); + __journal_unfile_buffer(jh); + jbd_unlock_bh_state(bh); + if (locked) + unlock_buffer(bh); + journal_remove_journal_head(bh); + /* Once for our safety reference, once for + * journal_remove_journal_head() */ + put_bh(bh); + put_bh(bh); + } + + if (lock_need_resched(&journal->j_list_lock)) { + spin_unlock(&journal->j_list_lock); + goto write_out_data; + } + } + spin_unlock(&journal->j_list_lock); + journal_do_submit_data(wbuf, bufs); +} + /* * journal_commit_transaction * @@ -261,7 +372,7 @@ void journal_commit_transaction(journal_t *journal) struct buffer_head *bh = jh2bh(jh); jbd_lock_bh_state(bh); - kfree(jh->b_committed_data); + jbd_slab_free(jh->b_committed_data, bh->b_size); jh->b_committed_data = NULL; jbd_unlock_bh_state(bh); } @@ -313,80 +424,13 @@ void journal_commit_transaction(journal_t *journal) * Now start flushing things to disk, in the order they appear * on the transaction lists. Data blocks go first. */ - err = 0; - /* - * Whenever we unlock the journal and sleep, things can get added - * onto ->t_sync_datalist, so we have to keep looping back to - * write_out_data until we *know* that the list is empty. - */ - bufs = 0; - /* - * Cleanup any flushed data buffers from the data list. Even in - * abort mode, we want to flush this out as soon as possible. - */ -write_out_data: - cond_resched(); - spin_lock(&journal->j_list_lock); - - while (commit_transaction->t_sync_datalist) { - struct buffer_head *bh; - - jh = commit_transaction->t_sync_datalist; - commit_transaction->t_sync_datalist = jh->b_tnext; - bh = jh2bh(jh); - if (buffer_locked(bh)) { - BUFFER_TRACE(bh, "locked"); - if (!inverted_lock(journal, bh)) - goto write_out_data; - __journal_temp_unlink_buffer(jh); - __journal_file_buffer(jh, commit_transaction, - BJ_Locked); - jbd_unlock_bh_state(bh); - if (lock_need_resched(&journal->j_list_lock)) { - spin_unlock(&journal->j_list_lock); - goto write_out_data; - } - } else { - if (buffer_dirty(bh)) { - BUFFER_TRACE(bh, "start journal writeout"); - get_bh(bh); - wbuf[bufs++] = bh; - if (bufs == journal->j_wbufsize) { - jbd_debug(2, "submit %d writes\n", - bufs); - spin_unlock(&journal->j_list_lock); - ll_rw_block(SWRITE, bufs, wbuf); - journal_brelse_array(wbuf, bufs); - bufs = 0; - goto write_out_data; - } - } else { - BUFFER_TRACE(bh, "writeout complete: unfile"); - if (!inverted_lock(journal, bh)) - goto write_out_data; - __journal_unfile_buffer(jh); - jbd_unlock_bh_state(bh); - journal_remove_journal_head(bh); - put_bh(bh); - if (lock_need_resched(&journal->j_list_lock)) { - spin_unlock(&journal->j_list_lock); - goto write_out_data; - } - } - } - } - - if (bufs) { - spin_unlock(&journal->j_list_lock); - ll_rw_block(SWRITE, bufs, wbuf); - journal_brelse_array(wbuf, bufs); - spin_lock(&journal->j_list_lock); - } + journal_submit_data_buffers(journal, commit_transaction); /* * Wait for all previously submitted IO to complete. */ + spin_lock(&journal->j_list_lock); while (commit_transaction->t_locked_list) { struct buffer_head *bh; @@ -745,14 +789,14 @@ restart_loop: * Otherwise, we can just throw away the frozen data now. */ if (jh->b_committed_data) { - kfree(jh->b_committed_data); + jbd_slab_free(jh->b_committed_data, bh->b_size); jh->b_committed_data = NULL; if (jh->b_frozen_data) { jh->b_committed_data = jh->b_frozen_data; jh->b_frozen_data = NULL; } } else if (jh->b_frozen_data) { - kfree(jh->b_frozen_data); + jbd_slab_free(jh->b_frozen_data, bh->b_size); jh->b_frozen_data = NULL; } diff --git a/fs/jbd/journal.c b/fs/jbd/journal.c index 7f96b5cb678..7af6099c911 100644 --- a/fs/jbd/journal.c +++ b/fs/jbd/journal.c @@ -34,6 +34,7 @@ #include <linux/suspend.h> #include <linux/pagemap.h> #include <linux/kthread.h> +#include <linux/poison.h> #include <linux/proc_fs.h> #include <asm/uaccess.h> @@ -83,6 +84,7 @@ EXPORT_SYMBOL(journal_force_commit); static int journal_convert_superblock_v1(journal_t *, journal_superblock_t *); static void __journal_abort_soft (journal_t *journal, int errno); +static int journal_create_jbd_slab(size_t slab_size); /* * Helper function used to manage commit timeouts @@ -179,7 +181,7 @@ loop: transaction->t_expires)) should_sleep = 0; if (journal->j_flags & JFS_UNMOUNT) - should_sleep = 0; + should_sleep = 0; if (should_sleep) { spin_unlock(&journal->j_state_lock); schedule(); @@ -269,7 +271,7 @@ static void journal_kill_thread(journal_t *journal) int journal_write_metadata_buffer(transaction_t *transaction, struct journal_head *jh_in, struct journal_head **jh_out, - int blocknr) + unsigned long blocknr) { int need_copy_out = 0; int done_copy_out = 0; @@ -327,10 +329,10 @@ repeat: char *tmp; jbd_unlock_bh_state(bh_in); - tmp = jbd_rep_kmalloc(bh_in->b_size, GFP_NOFS); + tmp = jbd_slab_alloc(bh_in->b_size, GFP_NOFS); jbd_lock_bh_state(bh_in); if (jh_in->b_frozen_data) { - kfree(tmp); + jbd_slab_free(tmp, bh_in->b_size); goto repeat; } @@ -576,7 +578,7 @@ int journal_next_log_block(journal_t *journal, unsigned long *retp) * this is a no-op. If needed, we can use j_blk_offset - everything is * ready. */ -int journal_bmap(journal_t *journal, unsigned long blocknr, +int journal_bmap(journal_t *journal, unsigned long blocknr, unsigned long *retp) { int err = 0; @@ -694,13 +696,13 @@ fail: * @bdev: Block device on which to create the journal * @fs_dev: Device which hold journalled filesystem for this journal. * @start: Block nr Start of journal. - * @len: Lenght of the journal in blocks. + * @len: Length of the journal in blocks. * @blocksize: blocksize of journalling device * @returns: a newly created journal_t * - * + * * journal_init_dev creates a journal which maps a fixed contiguous * range of blocks on an arbitrary block device. - * + * */ journal_t * journal_init_dev(struct block_device *bdev, struct block_device *fs_dev, @@ -713,18 +715,8 @@ journal_t * journal_init_dev(struct block_device *bdev, if (!journal) return NULL; - journal->j_dev = bdev; - journal->j_fs_dev = fs_dev; - journal->j_blk_offset = start; - journal->j_maxlen = len; - journal->j_blocksize = blocksize; - - bh = __getblk(journal->j_dev, start, journal->j_blocksize); - J_ASSERT(bh != NULL); - journal->j_sb_buffer = bh; - journal->j_superblock = (journal_superblock_t *)bh->b_data; - /* journal descriptor can store up to n blocks -bzzz */ + journal->j_blocksize = blocksize; n = journal->j_blocksize / sizeof(journal_block_tag_t); journal->j_wbufsize = n; journal->j_wbuf = kmalloc(n * sizeof(struct buffer_head*), GFP_KERNEL); @@ -734,14 +726,23 @@ journal_t * journal_init_dev(struct block_device *bdev, kfree(journal); journal = NULL; } + journal->j_dev = bdev; + journal->j_fs_dev = fs_dev; + journal->j_blk_offset = start; + journal->j_maxlen = len; + + bh = __getblk(journal->j_dev, start, journal->j_blocksize); + J_ASSERT(bh != NULL); + journal->j_sb_buffer = bh; + journal->j_superblock = (journal_superblock_t *)bh->b_data; return journal; } - -/** + +/** * journal_t * journal_init_inode () - creates a journal which maps to a inode. * @inode: An inode to create the journal in - * + * * journal_init_inode creates a journal which maps an on-disk inode as * the journal. The inode must exist already, must support bmap() and * must have all data blocks preallocated. @@ -761,7 +762,7 @@ journal_t * journal_init_inode (struct inode *inode) journal->j_inode = inode; jbd_debug(1, "journal %p: inode %s/%ld, size %Ld, bits %d, blksize %ld\n", - journal, inode->i_sb->s_id, inode->i_ino, + journal, inode->i_sb->s_id, inode->i_ino, (long long) inode->i_size, inode->i_sb->s_blocksize_bits, inode->i_sb->s_blocksize); @@ -796,10 +797,10 @@ journal_t * journal_init_inode (struct inode *inode) return journal; } -/* +/* * If the journal init or create aborts, we need to mark the journal * superblock as being NULL to prevent the journal destroy from writing - * back a bogus superblock. + * back a bogus superblock. */ static void journal_fail_superblock (journal_t *journal) { @@ -818,7 +819,7 @@ static void journal_fail_superblock (journal_t *journal) static int journal_reset(journal_t *journal) { journal_superblock_t *sb = journal->j_superblock; - unsigned int first, last; + unsigned long first, last; first = be32_to_cpu(sb->s_first); last = be32_to_cpu(sb->s_maxlen); @@ -842,13 +843,13 @@ static int journal_reset(journal_t *journal) return 0; } -/** +/** * int journal_create() - Initialise the new journal file * @journal: Journal to create. This structure must have been initialised - * + * * Given a journal_t structure which tells us which disk blocks we can * use, create a new journal superblock and initialise all of the - * journal fields from scratch. + * journal fields from scratch. **/ int journal_create(journal_t *journal) { @@ -913,7 +914,7 @@ int journal_create(journal_t *journal) return journal_reset(journal); } -/** +/** * void journal_update_superblock() - Update journal sb on disk. * @journal: The journal to update. * @wait: Set to '0' if you don't want to wait for IO completion. @@ -937,7 +938,7 @@ void journal_update_superblock(journal_t *journal, int wait) journal->j_transaction_sequence) { jbd_debug(1,"JBD: Skipping superblock update on recovered sb " "(start %ld, seq %d, errno %d)\n", - journal->j_tail, journal->j_tail_sequence, + journal->j_tail, journal->j_tail_sequence, journal->j_errno); goto out; } @@ -1060,7 +1061,7 @@ static int load_superblock(journal_t *journal) /** * int journal_load() - Read journal from disk. * @journal: Journal to act on. - * + * * Given a journal_t structure which tells us which disk blocks contain * a journal, read the journal from disk to initialise the in-memory * structures. @@ -1068,17 +1069,17 @@ static int load_superblock(journal_t *journal) int journal_load(journal_t *journal) { int err; + journal_superblock_t *sb; err = load_superblock(journal); if (err) return err; + sb = journal->j_superblock; /* If this is a V2 superblock, then we have to check the * features flags on it. */ if (journal->j_format_version >= 2) { - journal_superblock_t *sb = journal->j_superblock; - if ((sb->s_feature_ro_compat & ~cpu_to_be32(JFS_KNOWN_ROCOMPAT_FEATURES)) || (sb->s_feature_incompat & @@ -1089,6 +1090,13 @@ int journal_load(journal_t *journal) } } + /* + * Create a slab for this blocksize + */ + err = journal_create_jbd_slab(be32_to_cpu(sb->s_blocksize)); + if (err) + return err; + /* Let the recovery code check whether it needs to recover any * data from the journal. */ if (journal_recover(journal)) @@ -1163,9 +1171,9 @@ void journal_destroy(journal_t *journal) * @compat: bitmask of compatible features * @ro: bitmask of features that force read-only mount * @incompat: bitmask of incompatible features - * + * * Check whether the journal uses all of a given set of - * features. Return true (non-zero) if it does. + * features. Return true (non-zero) if it does. **/ int journal_check_used_features (journal_t *journal, unsigned long compat, @@ -1194,7 +1202,7 @@ int journal_check_used_features (journal_t *journal, unsigned long compat, * @compat: bitmask of compatible features * @ro: bitmask of features that force read-only mount * @incompat: bitmask of incompatible features - * + * * Check whether the journaling code supports the use of * all of a given set of features on this journal. Return true * (non-zero) if it can. */ @@ -1232,7 +1240,7 @@ int journal_check_available_features (journal_t *journal, unsigned long compat, * @incompat: bitmask of incompatible features * * Mark a given journal feature as present on the - * superblock. Returns true if the requested features could be set. + * superblock. Returns true if the requested features could be set. * */ @@ -1318,7 +1326,7 @@ static int journal_convert_superblock_v1(journal_t *journal, /** * int journal_flush () - Flush journal * @journal: Journal to act on. - * + * * Flush all data for a given journal to disk and empty the journal. * Filesystems can use this when remounting readonly to ensure that * recovery does not need to happen on remount. @@ -1385,7 +1393,7 @@ int journal_flush(journal_t *journal) * int journal_wipe() - Wipe journal contents * @journal: Journal to act on. * @write: flag (see below) - * + * * Wipe out all of the contents of a journal, safely. This will produce * a warning if the journal contains any valid recovery information. * Must be called between journal_init_*() and journal_load(). @@ -1440,7 +1448,7 @@ static const char *journal_dev_name(journal_t *journal, char *buffer) /* * Journal abort has very specific semantics, which we describe - * for journal abort. + * for journal abort. * * Two internal function, which provide abort to te jbd layer * itself are here. @@ -1495,7 +1503,7 @@ static void __journal_abort_soft (journal_t *journal, int errno) * Perform a complete, immediate shutdown of the ENTIRE * journal (not of a single transaction). This operation cannot be * undone without closing and reopening the journal. - * + * * The journal_abort function is intended to support higher level error * recovery mechanisms such as the ext2/ext3 remount-readonly error * mode. @@ -1529,7 +1537,7 @@ static void __journal_abort_soft (journal_t *journal, int errno) * supply an errno; a null errno implies that absolutely no further * writes are done to the journal (unless there are any already in * progress). - * + * */ void journal_abort(journal_t *journal, int errno) @@ -1537,7 +1545,7 @@ void journal_abort(journal_t *journal, int errno) __journal_abort_soft(journal, errno); } -/** +/** * int journal_errno () - returns the journal's error state. * @journal: journal to examine. * @@ -1561,7 +1569,7 @@ int journal_errno(journal_t *journal) return err; } -/** +/** * int journal_clear_err () - clears the journal's error state * @journal: journal to act on. * @@ -1581,7 +1589,7 @@ int journal_clear_err(journal_t *journal) return err; } -/** +/** * void journal_ack_err() - Ack journal err. * @journal: journal to act on. * @@ -1603,7 +1611,7 @@ int journal_blocks_per_page(struct inode *inode) /* * Simple support for retrying memory allocations. Introduced to help to - * debug different VM deadlock avoidance strategies. + * debug different VM deadlock avoidance strategies. */ void * __jbd_kmalloc (const char *where, size_t size, gfp_t flags, int retry) { @@ -1611,6 +1619,77 @@ void * __jbd_kmalloc (const char *where, size_t size, gfp_t flags, int retry) } /* + * jbd slab management: create 1k, 2k, 4k, 8k slabs as needed + * and allocate frozen and commit buffers from these slabs. + * + * Reason for doing this is to avoid, SLAB_DEBUG - since it could + * cause bh to cross page boundary. + */ + +#define JBD_MAX_SLABS 5 +#define JBD_SLAB_INDEX(size) (size >> 11) + +static kmem_cache_t *jbd_slab[JBD_MAX_SLABS]; +static const char *jbd_slab_names[JBD_MAX_SLABS] = { + "jbd_1k", "jbd_2k", "jbd_4k", NULL, "jbd_8k" +}; + +static void journal_destroy_jbd_slabs(void) +{ + int i; + + for (i = 0; i < JBD_MAX_SLABS; i++) { + if (jbd_slab[i]) + kmem_cache_destroy(jbd_slab[i]); + jbd_slab[i] = NULL; + } +} + +static int journal_create_jbd_slab(size_t slab_size) +{ + int i = JBD_SLAB_INDEX(slab_size); + + BUG_ON(i >= JBD_MAX_SLABS); + + /* + * Check if we already have a slab created for this size + */ + if (jbd_slab[i]) + return 0; + + /* + * Create a slab and force alignment to be same as slabsize - + * this will make sure that allocations won't cross the page + * boundary. + */ + jbd_slab[i] = kmem_cache_create(jbd_slab_names[i], + slab_size, slab_size, 0, NULL, NULL); + if (!jbd_slab[i]) { + printk(KERN_EMERG "JBD: no memory for jbd_slab cache\n"); + return -ENOMEM; + } + return 0; +} + +void * jbd_slab_alloc(size_t size, gfp_t flags) +{ + int idx; + + idx = JBD_SLAB_INDEX(size); + BUG_ON(jbd_slab[idx] == NULL); + return kmem_cache_alloc(jbd_slab[idx], flags | __GFP_NOFAIL); +} + +void jbd_slab_free(void *ptr, size_t size) +{ + int idx; + + idx = JBD_SLAB_INDEX(size); + BUG_ON(jbd_slab[idx] == NULL); + kmem_cache_free(jbd_slab[idx], ptr); +} + +/* * Journal_head storage management */ static kmem_cache_t *journal_head_cache; @@ -1675,7 +1754,7 @@ static void journal_free_journal_head(struct journal_head *jh) { #ifdef CONFIG_JBD_DEBUG atomic_dec(&nr_journal_heads); - memset(jh, 0x5b, sizeof(*jh)); + memset(jh, JBD_POISON_FREE, sizeof(*jh)); #endif kmem_cache_free(journal_head_cache, jh); } @@ -1798,13 +1877,13 @@ static void __journal_remove_journal_head(struct buffer_head *bh) printk(KERN_WARNING "%s: freeing " "b_frozen_data\n", __FUNCTION__); - kfree(jh->b_frozen_data); + jbd_slab_free(jh->b_frozen_data, bh->b_size); } if (jh->b_committed_data) { printk(KERN_WARNING "%s: freeing " "b_committed_data\n", __FUNCTION__); - kfree(jh->b_committed_data); + jbd_slab_free(jh->b_committed_data, bh->b_size); } bh->b_private = NULL; jh->b_bh = NULL; /* debug, really */ @@ -1960,19 +2039,14 @@ static void journal_destroy_caches(void) journal_destroy_revoke_caches(); journal_destroy_journal_head_cache(); journal_destroy_handle_cache(); + journal_destroy_jbd_slabs(); } static int __init journal_init(void) { int ret; -/* Static check for data structure consistency. There's no code - * invoked --- we'll just get a linker failure if things aren't right. - */ - extern void journal_bad_superblock_size(void); - if (sizeof(struct journal_superblock_s) != 1024) - journal_bad_superblock_size(); - + BUILD_BUG_ON(sizeof(struct journal_superblock_s) != 1024); ret = journal_init_caches(); if (ret != 0) diff --git a/fs/jbd/recovery.c b/fs/jbd/recovery.c index 80d7f53fd0a..11563fe2a52 100644 --- a/fs/jbd/recovery.c +++ b/fs/jbd/recovery.c @@ -1,6 +1,6 @@ /* * linux/fs/recovery.c - * + * * Written by Stephen C. Tweedie <sct@redhat.com>, 1999 * * Copyright 1999-2000 Red Hat Software --- All Rights Reserved @@ -10,7 +10,7 @@ * option, any later version, incorporated herein by reference. * * Journal recovery routines for the generic filesystem journaling code; - * part of the ext2fs journaling system. + * part of the ext2fs journaling system. */ #ifndef __KERNEL__ @@ -25,9 +25,9 @@ /* * Maintain information about the progress of the recovery job, so that - * the different passes can carry information between them. + * the different passes can carry information between them. */ -struct recovery_info +struct recovery_info { tid_t start_transaction; tid_t end_transaction; @@ -46,7 +46,7 @@ static int scan_revoke_records(journal_t *, struct buffer_head *, #ifdef __KERNEL__ /* Release readahead buffers after use */ -void journal_brelse_array(struct buffer_head *b[], int n) +static void journal_brelse_array(struct buffer_head *b[], int n) { while (--n >= 0) brelse (b[n]); @@ -116,7 +116,7 @@ static int do_readahead(journal_t *journal, unsigned int start) err = 0; failed: - if (nbufs) + if (nbufs) journal_brelse_array(bufs, nbufs); return err; } @@ -128,7 +128,7 @@ failed: * Read a block from the journal */ -static int jread(struct buffer_head **bhp, journal_t *journal, +static int jread(struct buffer_head **bhp, journal_t *journal, unsigned int offset) { int err; @@ -212,14 +212,14 @@ do { \ /** * journal_recover - recovers a on-disk journal * @journal: the journal to recover - * + * * The primary function for recovering the log contents when mounting a - * journaled device. + * journaled device. * * Recovery is done in three passes. In the first pass, we look for the * end of the log. In the second, we assemble the list of revoke * blocks. In the third and final pass, we replay any un-revoked blocks - * in the log. + * in the log. */ int journal_recover(journal_t *journal) { @@ -231,10 +231,10 @@ int journal_recover(journal_t *journal) memset(&info, 0, sizeof(info)); sb = journal->j_superblock; - /* + /* * The journal superblock's s_start field (the current log head) * is always zero if, and only if, the journal was cleanly - * unmounted. + * unmounted. */ if (!sb->s_start) { @@ -253,7 +253,7 @@ int journal_recover(journal_t *journal) jbd_debug(0, "JBD: recovery, exit status %d, " "recovered transactions %u to %u\n", err, info.start_transaction, info.end_transaction); - jbd_debug(0, "JBD: Replayed %d and revoked %d/%d blocks\n", + jbd_debug(0, "JBD: Replayed %d and revoked %d/%d blocks\n", info.nr_replays, info.nr_revoke_hits, info.nr_revokes); /* Restart the log at the next transaction ID, thus invalidating @@ -268,15 +268,15 @@ int journal_recover(journal_t *journal) /** * journal_skip_recovery - Start journal and wipe exiting records * @journal: journal to startup - * + * * Locate any valid recovery information from the journal and set up the * journal structures in memory to ignore it (presumably because the - * caller has evidence that it is out of date). + * caller has evidence that it is out of date). * This function does'nt appear to be exorted.. * * We perform one pass over the journal to allow us to tell the user how * much recovery information is being erased, and to let us initialise - * the journal transaction sequence numbers to the next unused ID. + * the journal transaction sequence numbers to the next unused ID. */ int journal_skip_recovery(journal_t *journal) { @@ -297,7 +297,7 @@ int journal_skip_recovery(journal_t *journal) #ifdef CONFIG_JBD_DEBUG int dropped = info.end_transaction - be32_to_cpu(sb->s_sequence); #endif - jbd_debug(0, + jbd_debug(0, "JBD: ignoring %d transaction%s from the journal.\n", dropped, (dropped == 1) ? "" : "s"); journal->j_transaction_sequence = ++info.end_transaction; @@ -314,7 +314,7 @@ static int do_one_pass(journal_t *journal, unsigned long next_log_block; int err, success = 0; journal_superblock_t * sb; - journal_header_t * tmp; + journal_header_t * tmp; struct buffer_head * bh; unsigned int sequence; int blocktype; @@ -324,10 +324,10 @@ static int do_one_pass(journal_t *journal, 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 - * block offsets): query the superblock. + * block offsets): query the superblock. */ sb = journal->j_superblock; @@ -344,7 +344,7 @@ static int do_one_pass(journal_t *journal, * Now we walk through the log, transaction by transaction, * making sure that each transaction has a commit block in the * expected place. Each complete transaction gets replayed back - * into the main filesystem. + * into the main filesystem. */ while (1) { @@ -379,8 +379,8 @@ static int do_one_pass(journal_t *journal, next_log_block++; wrap(journal, next_log_block); - /* What kind of buffer is it? - * + /* What kind of buffer is it? + * * If it is a descriptor block, check that it has the * expected sequence number. Otherwise, we're all done * here. */ @@ -394,7 +394,7 @@ static int do_one_pass(journal_t *journal, blocktype = be32_to_cpu(tmp->h_blocktype); sequence = be32_to_cpu(tmp->h_sequence); - jbd_debug(3, "Found magic %d, sequence %d\n", + jbd_debug(3, "Found magic %d, sequence %d\n", blocktype, sequence); if (sequence != next_commit_ID) { @@ -438,7 +438,7 @@ static int do_one_pass(journal_t *journal, /* Recover what we can, but * report failure at the end. */ success = err; - printk (KERN_ERR + printk (KERN_ERR "JBD: IO error %d recovering " "block %ld in log\n", err, io_block); @@ -452,7 +452,7 @@ static int do_one_pass(journal_t *journal, * revoked, then we're all done * here. */ if (journal_test_revoke - (journal, blocknr, + (journal, blocknr, next_commit_ID)) { brelse(obh); ++info->nr_revoke_hits; @@ -465,7 +465,7 @@ static int do_one_pass(journal_t *journal, blocknr, journal->j_blocksize); if (nbh == NULL) { - printk(KERN_ERR + printk(KERN_ERR "JBD: Out of memory " "during recovery.\n"); err = -ENOMEM; @@ -531,12 +531,13 @@ static int do_one_pass(journal_t *journal, default: jbd_debug(3, "Unrecognised magic %d, end of scan.\n", blocktype); + brelse(bh); goto done; } } done: - /* + /* * We broke out of the log scan loop: either we came to the * known end of the log or we found an unexpected block in the * log. If the latter happened, then we know that the "current" @@ -566,7 +567,7 @@ static int do_one_pass(journal_t *journal, /* Scan a revoke record, marking all blocks mentioned as revoked. */ -static int scan_revoke_records(journal_t *journal, struct buffer_head *bh, +static int scan_revoke_records(journal_t *journal, struct buffer_head *bh, tid_t sequence, struct recovery_info *info) { journal_revoke_header_t *header; diff --git a/fs/jbd/revoke.c b/fs/jbd/revoke.c index a5614418346..c532429d8d9 100644 --- a/fs/jbd/revoke.c +++ b/fs/jbd/revoke.c @@ -1,6 +1,6 @@ /* * linux/fs/revoke.c - * + * * Written by Stephen C. Tweedie <sct@redhat.com>, 2000 * * Copyright 2000 Red Hat corp --- All Rights Reserved @@ -15,10 +15,10 @@ * Revoke is the mechanism used to prevent old log records for deleted * metadata from being replayed on top of newer data using the same * blocks. The revoke mechanism is used in two separate places: - * + * * + Commit: during commit we write the entire list of the current * transaction's revoked blocks to the journal - * + * * + Recovery: during recovery we record the transaction ID of all * revoked blocks. If there are multiple revoke records in the log * for a single block, only the last one counts, and if there is a log @@ -29,7 +29,7 @@ * single transaction: * * Block is revoked and then journaled: - * The desired end result is the journaling of the new block, so we + * The desired end result is the journaling of the new block, so we * cancel the revoke before the transaction commits. * * Block is journaled and then revoked: @@ -41,7 +41,7 @@ * transaction must have happened after the block was journaled and so * the revoke must take precedence. * - * Block is revoked and then written as data: + * Block is revoked and then written as data: * The data write is allowed to succeed, but the revoke is _not_ * cancelled. We still need to prevent old log records from * overwriting the new data. We don't even need to clear the revoke @@ -54,7 +54,7 @@ * buffer has not been revoked, and cancel_revoke * need do nothing. * RevokeValid set, Revoked set: - * buffer has been revoked. + * buffer has been revoked. */ #ifndef __KERNEL__ @@ -77,7 +77,7 @@ static kmem_cache_t *revoke_table_cache; journal replay, this involves recording the transaction ID of the last transaction to revoke this block. */ -struct jbd_revoke_record_s +struct jbd_revoke_record_s { struct list_head hash; tid_t sequence; /* Used for recovery only */ @@ -90,8 +90,8 @@ struct jbd_revoke_table_s { /* It is conceivable that we might want a larger hash table * for recovery. Must be a power of two. */ - int hash_size; - int hash_shift; + int hash_size; + int hash_shift; struct list_head *hash_table; }; @@ -301,22 +301,22 @@ void journal_destroy_revoke(journal_t *journal) #ifdef __KERNEL__ -/* +/* * journal_revoke: revoke a given buffer_head from the journal. This * prevents the block from being replayed during recovery if we take a * crash after this current transaction commits. Any subsequent * metadata writes of the buffer in this transaction cancel the - * revoke. + * revoke. * * Note that this call may block --- it is up to the caller to make * sure that there are no further calls to journal_write_metadata * before the revoke is complete. In ext3, this implies calling the * revoke before clearing the block bitmap when we are deleting - * metadata. + * metadata. * * Revoke performs a journal_forget on any buffer_head passed in as a * parameter, but does _not_ forget the buffer_head if the bh was only - * found implicitly. + * found implicitly. * * bh_in may not be a journalled buffer - it may have come off * the hash tables without an attached journal_head. @@ -325,7 +325,7 @@ void journal_destroy_revoke(journal_t *journal) * by one. */ -int journal_revoke(handle_t *handle, unsigned long blocknr, +int journal_revoke(handle_t *handle, unsigned long blocknr, struct buffer_head *bh_in) { struct buffer_head *bh = NULL; @@ -487,7 +487,7 @@ void journal_switch_revoke_table(journal_t *journal) else journal->j_revoke = journal->j_revoke_table[0]; - for (i = 0; i < journal->j_revoke->hash_size; i++) + for (i = 0; i < journal->j_revoke->hash_size; i++) INIT_LIST_HEAD(&journal->j_revoke->hash_table[i]); } @@ -498,7 +498,7 @@ void journal_switch_revoke_table(journal_t *journal) * Called with the journal lock held. */ -void journal_write_revoke_records(journal_t *journal, +void journal_write_revoke_records(journal_t *journal, transaction_t *transaction) { struct journal_head *descriptor; @@ -507,7 +507,7 @@ void journal_write_revoke_records(journal_t *journal, struct list_head *hash_list; int i, offset, count; - descriptor = NULL; + descriptor = NULL; offset = 0; count = 0; @@ -519,10 +519,10 @@ void journal_write_revoke_records(journal_t *journal, hash_list = &revoke->hash_table[i]; while (!list_empty(hash_list)) { - record = (struct jbd_revoke_record_s *) + record = (struct jbd_revoke_record_s *) hash_list->next; write_one_revoke_record(journal, transaction, - &descriptor, &offset, + &descriptor, &offset, record); count++; list_del(&record->hash); @@ -534,14 +534,14 @@ void journal_write_revoke_records(journal_t *journal, jbd_debug(1, "Wrote %d revoke records\n", count); } -/* +/* * Write out one revoke record. We need to create a new descriptor - * block if the old one is full or if we have not already created one. + * block if the old one is full or if we have not already created one. */ -static void write_one_revoke_record(journal_t *journal, +static void write_one_revoke_record(journal_t *journal, transaction_t *transaction, - struct journal_head **descriptorp, + struct journal_head **descriptorp, int *offsetp, struct jbd_revoke_record_s *record) { @@ -584,21 +584,21 @@ static void write_one_revoke_record(journal_t *journal, *descriptorp = descriptor; } - * ((__be32 *)(&jh2bh(descriptor)->b_data[offset])) = + * ((__be32 *)(&jh2bh(descriptor)->b_data[offset])) = cpu_to_be32(record->blocknr); offset += 4; *offsetp = offset; } -/* +/* * Flush a revoke descriptor out to the journal. If we are aborting, * this is a noop; otherwise we are generating a buffer which needs to * be waited for during commit, so it has to go onto the appropriate * journal buffer list. */ -static void flush_descriptor(journal_t *journal, - struct journal_head *descriptor, +static void flush_descriptor(journal_t *journal, + struct journal_head *descriptor, int offset) { journal_revoke_header_t *header; @@ -618,7 +618,7 @@ static void flush_descriptor(journal_t *journal, } #endif -/* +/* * Revoke support for recovery. * * Recovery needs to be able to: @@ -629,7 +629,7 @@ static void flush_descriptor(journal_t *journal, * check whether a given block in a given transaction should be replayed * (ie. has not been revoked by a revoke record in that or a subsequent * transaction) - * + * * empty the revoke table after recovery. */ @@ -637,11 +637,11 @@ static void flush_descriptor(journal_t *journal, * First, setting revoke records. We create a new revoke record for * every block ever revoked in the log as we scan it for recovery, and * we update the existing records if we find multiple revokes for a - * single block. + * single block. */ -int journal_set_revoke(journal_t *journal, - unsigned long blocknr, +int journal_set_revoke(journal_t *journal, + unsigned long blocknr, tid_t sequence) { struct jbd_revoke_record_s *record; @@ -653,18 +653,18 @@ int journal_set_revoke(journal_t *journal, if (tid_gt(sequence, record->sequence)) record->sequence = sequence; return 0; - } + } return insert_revoke_hash(journal, blocknr, sequence); } -/* +/* * Test revoke records. For a given block referenced in the log, has * that block been revoked? A revoke record with a given transaction * sequence number revokes all blocks in that transaction and earlier * ones, but later transactions still need replayed. */ -int journal_test_revoke(journal_t *journal, +int journal_test_revoke(journal_t *journal, unsigned long blocknr, tid_t sequence) { diff --git a/fs/jbd/transaction.c b/fs/jbd/transaction.c index 508b2ea91f4..e1b3c8af4d1 100644 --- a/fs/jbd/transaction.c +++ b/fs/jbd/transaction.c @@ -1,6 +1,6 @@ /* * linux/fs/transaction.c - * + * * Written by Stephen C. Tweedie <sct@redhat.com>, 1998 * * Copyright 1998 Red Hat corp --- All Rights Reserved @@ -10,7 +10,7 @@ * option, any later version, incorporated herein by reference. * * Generic filesystem transaction handling code; part of the ext2fs - * journaling system. + * journaling system. * * This file manages transactions (compound commits managed by the * journaling code) and handles (individual atomic operations by the @@ -74,7 +74,7 @@ get_transaction(journal_t *journal, transaction_t *transaction) * start_this_handle: Given a handle, deal with any locking or stalling * needed to make sure that there is enough journal space for the handle * to begin. Attach the handle to a transaction and set up the - * transaction's buffer credits. + * transaction's buffer credits. */ static int start_this_handle(journal_t *journal, handle_t *handle) @@ -117,7 +117,7 @@ repeat_locked: if (is_journal_aborted(journal) || (journal->j_errno != 0 && !(journal->j_flags & JFS_ACK_ERR))) { spin_unlock(&journal->j_state_lock); - ret = -EROFS; + ret = -EROFS; goto out; } @@ -182,7 +182,7 @@ repeat_locked: goto repeat; } - /* + /* * The commit code assumes that it can get enough log space * without forcing a checkpoint. This is *critical* for * correctness: a checkpoint of a buffer which is also @@ -191,7 +191,7 @@ repeat_locked: * * We must therefore ensure the necessary space in the journal * *before* starting to dirty potentially checkpointed buffers - * in the new transaction. + * in the new transaction. * * The worst part is, any transaction currently committing can * reduce the free space arbitrarily. Be careful to account for @@ -246,13 +246,13 @@ static handle_t *new_handle(int nblocks) } /** - * handle_t *journal_start() - Obtain a new handle. + * handle_t *journal_start() - Obtain a new handle. * @journal: Journal to start transaction on. * @nblocks: number of block buffer we might modify * * We make sure that the transaction can guarantee at least nblocks of * modified buffers in the log. We block until the log can guarantee - * that much space. + * that much space. * * This function is visible to journal users (like ext3fs), so is not * called with the journal already locked. @@ -292,11 +292,11 @@ handle_t *journal_start(journal_t *journal, int nblocks) * int journal_extend() - extend buffer credits. * @handle: handle to 'extend' * @nblocks: nr blocks to try to extend by. - * + * * Some transactions, such as large extends and truncates, can be done * atomically all at once or in several stages. The operation requests * a credit for a number of buffer modications in advance, but can - * extend its credit if it needs more. + * extend its credit if it needs more. * * journal_extend tries to give the running handle more buffer credits. * It does not guarantee that allocation - this is a best-effort only. @@ -363,7 +363,7 @@ out: * int journal_restart() - restart a handle . * @handle: handle to restart * @nblocks: nr credits requested - * + * * Restart a handle for a multi-transaction filesystem * operation. * @@ -462,7 +462,7 @@ void journal_lock_updates(journal_t *journal) /** * void journal_unlock_updates (journal_t* journal) - release barrier * @journal: Journal to release the barrier on. - * + * * Release a transaction barrier obtained with journal_lock_updates(). * * Should be called without the journal lock held. @@ -547,8 +547,8 @@ repeat: jbd_lock_bh_state(bh); /* We now hold the buffer lock so it is safe to query the buffer - * state. Is the buffer dirty? - * + * state. Is the buffer dirty? + * * If so, there are two possibilities. The buffer may be * non-journaled, and undergoing a quite legitimate writeback. * Otherwise, it is journaled, and we don't expect dirty buffers @@ -566,7 +566,7 @@ repeat: */ if (jh->b_transaction) { J_ASSERT_JH(jh, - jh->b_transaction == transaction || + jh->b_transaction == transaction || jh->b_transaction == journal->j_committing_transaction); if (jh->b_next_transaction) @@ -580,7 +580,7 @@ repeat: */ JBUFFER_TRACE(jh, "Unexpected dirty buffer"); jbd_unexpected_dirty_buffer(jh); - } + } unlock_buffer(bh); @@ -653,7 +653,7 @@ repeat: * buffer had better remain locked during the kmalloc, * but that should be true --- we hold the journal lock * still and the buffer is already on the BUF_JOURNAL - * list so won't be flushed. + * list so won't be flushed. * * Subtle point, though: if this is a get_undo_access, * then we will be relying on the frozen_data to contain @@ -666,8 +666,9 @@ repeat: if (!frozen_buffer) { JBUFFER_TRACE(jh, "allocate memory for buffer"); jbd_unlock_bh_state(bh); - frozen_buffer = jbd_kmalloc(jh2bh(jh)->b_size, - GFP_NOFS); + frozen_buffer = + jbd_slab_alloc(jh2bh(jh)->b_size, + GFP_NOFS); if (!frozen_buffer) { printk(KERN_EMERG "%s: OOM for frozen_buffer\n", @@ -726,7 +727,7 @@ done: out: if (unlikely(frozen_buffer)) /* It's usually NULL */ - kfree(frozen_buffer); + jbd_slab_free(frozen_buffer, bh->b_size); JBUFFER_TRACE(jh, "exit"); return error; @@ -764,8 +765,8 @@ int journal_get_write_access(handle_t *handle, struct buffer_head *bh) * manually rather than reading off disk), then we need to keep the * buffer_head locked until it has been completely filled with new * data. In this case, we should be able to make the assertion that - * the bh is not already part of an existing transaction. - * + * the bh is not already part of an existing transaction. + * * The buffer should already be locked by the caller by this point. * There is no lock ranking violation: it was a newly created, * unlocked buffer beforehand. */ @@ -777,7 +778,7 @@ int journal_get_write_access(handle_t *handle, struct buffer_head *bh) * * Call this if you create a new bh. */ -int journal_get_create_access(handle_t *handle, struct buffer_head *bh) +int journal_get_create_access(handle_t *handle, struct buffer_head *bh) { transaction_t *transaction = handle->h_transaction; journal_t *journal = transaction->t_journal; @@ -846,13 +847,13 @@ out: * do not reuse freed space until the deallocation has been committed, * since if we overwrote that space we would make the delete * un-rewindable in case of a crash. - * + * * To deal with that, journal_get_undo_access requests write access to a * buffer for parts of non-rewindable operations such as delete * operations on the bitmaps. The journaling code must keep a copy of * the buffer's contents prior to the undo_access call until such time * as we know that the buffer has definitely been committed to disk. - * + * * We never need to know which transaction the committed data is part * of, buffers touched here are guaranteed to be dirtied later and so * will be committed to a new transaction in due course, at which point @@ -879,7 +880,7 @@ int journal_get_undo_access(handle_t *handle, struct buffer_head *bh) repeat: if (!jh->b_committed_data) { - committed_data = jbd_kmalloc(jh2bh(jh)->b_size, GFP_NOFS); + committed_data = jbd_slab_alloc(jh2bh(jh)->b_size, GFP_NOFS); if (!committed_data) { printk(KERN_EMERG "%s: No memory for committed data\n", __FUNCTION__); @@ -906,17 +907,17 @@ repeat: out: journal_put_journal_head(jh); if (unlikely(committed_data)) - kfree(committed_data); + jbd_slab_free(committed_data, bh->b_size); return err; } -/** +/** * int journal_dirty_data() - mark a buffer as containing dirty data which * needs to be flushed before we can commit the - * current transaction. + * current transaction. * @handle: transaction * @bh: bufferhead to mark - * + * * The buffer is placed on the transaction's data list and is marked as * belonging to the transaction. * @@ -945,15 +946,15 @@ int journal_dirty_data(handle_t *handle, struct buffer_head *bh) /* * What if the buffer is already part of a running transaction? - * + * * There are two cases: * 1) It is part of the current running transaction. Refile it, * just in case we have allocated it as metadata, deallocated - * it, then reallocated it as data. + * it, then reallocated it as data. * 2) It is part of the previous, still-committing transaction. * If all we want to do is to guarantee that the buffer will be * written to disk before this new transaction commits, then - * being sure that the *previous* transaction has this same + * being sure that the *previous* transaction has this same * property is sufficient for us! Just leave it on its old * transaction. * @@ -1075,18 +1076,18 @@ no_journal: return 0; } -/** +/** * int journal_dirty_metadata() - mark a buffer as containing dirty metadata * @handle: transaction to add buffer to. - * @bh: buffer to mark - * + * @bh: buffer to mark + * * mark dirty metadata which needs to be journaled as part of the current * transaction. * * The buffer is placed on the transaction's metadata list and is marked - * as belonging to the transaction. + * as belonging to the transaction. * - * Returns error number or 0 on success. + * Returns error number or 0 on success. * * Special care needs to be taken if the buffer already belongs to the * current committing transaction (in which case we should have frozen @@ -1134,11 +1135,11 @@ int journal_dirty_metadata(handle_t *handle, struct buffer_head *bh) set_buffer_jbddirty(bh); - /* + /* * Metadata already on the current transaction list doesn't * need to be filed. Metadata on another transaction's list must * be committing, and will be refiled once the commit completes: - * leave it alone for now. + * leave it alone for now. */ if (jh->b_transaction != transaction) { JBUFFER_TRACE(jh, "already on other transaction"); @@ -1164,7 +1165,7 @@ out: return 0; } -/* +/* * journal_release_buffer: undo a get_write_access without any buffer * updates, if the update decided in the end that it didn't need access. * @@ -1175,20 +1176,20 @@ journal_release_buffer(handle_t *handle, struct buffer_head *bh) BUFFER_TRACE(bh, "entry"); } -/** +/** * void journal_forget() - bforget() for potentially-journaled buffers. * @handle: transaction handle * @bh: bh to 'forget' * * We can only do the bforget if there are no commits pending against the * buffer. If the buffer is dirty in the current running transaction we - * can safely unlink it. + * can safely unlink it. * * bh may not be a journalled buffer at all - it may be a non-JBD * buffer which came off the hashtable. Check for this. * * Decrements bh->b_count by one. - * + * * Allow this call even if the handle has aborted --- it may be part of * the caller's cleanup after an abort. */ @@ -1236,7 +1237,7 @@ int journal_forget (handle_t *handle, struct buffer_head *bh) drop_reserve = 1; - /* + /* * We are no longer going to journal this buffer. * However, the commit of this transaction is still * important to the buffer: the delete that we are now @@ -1245,7 +1246,7 @@ int journal_forget (handle_t *handle, struct buffer_head *bh) * * So, if we have a checkpoint on the buffer, we should * now refile the buffer on our BJ_Forget list so that - * we know to remove the checkpoint after we commit. + * we know to remove the checkpoint after we commit. */ if (jh->b_cp_transaction) { @@ -1263,7 +1264,7 @@ int journal_forget (handle_t *handle, struct buffer_head *bh) } } } else if (jh->b_transaction) { - J_ASSERT_JH(jh, (jh->b_transaction == + J_ASSERT_JH(jh, (jh->b_transaction == journal->j_committing_transaction)); /* However, if the buffer is still owned by a prior * (committing) transaction, we can't drop it yet... */ @@ -1293,7 +1294,7 @@ drop: /** * int journal_stop() - complete a transaction * @handle: tranaction to complete. - * + * * All done for a particular handle. * * There is not much action needed here. We just return any remaining @@ -1302,7 +1303,7 @@ drop: * filesystem is marked for synchronous update. * * journal_stop itself will not usually return an error, but it may - * do so in unusual circumstances. In particular, expect it to + * do so in unusual circumstances. In particular, expect it to * return -EIO if a journal_abort has been executed since the * transaction began. */ @@ -1372,7 +1373,7 @@ int journal_stop(handle_t *handle) if (handle->h_sync || transaction->t_outstanding_credits > journal->j_max_transaction_buffers || - time_after_eq(jiffies, transaction->t_expires)) { + 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. */ @@ -1387,7 +1388,7 @@ int journal_stop(handle_t *handle) /* * Special case: JFS_SYNC synchronous updates require us - * to wait for the commit to complete. + * to wait for the commit to complete. */ if (handle->h_sync && !(current->flags & PF_MEMALLOC)) err = log_wait_commit(journal, tid); @@ -1438,7 +1439,7 @@ int journal_force_commit(journal_t *journal) * jbd_lock_bh_state(jh2bh(jh)) is held. */ -static inline void +static inline void __blist_add_buffer(struct journal_head **list, struct journal_head *jh) { if (!*list) { @@ -1453,7 +1454,7 @@ __blist_add_buffer(struct journal_head **list, struct journal_head *jh) } } -/* +/* * Remove a buffer from a transaction list, given the transaction's list * head pointer. * @@ -1474,7 +1475,7 @@ __blist_del_buffer(struct journal_head **list, struct journal_head *jh) jh->b_tnext->b_tprev = jh->b_tprev; } -/* +/* * Remove a buffer from the appropriate transaction list. * * Note that this function can *change* the value of @@ -1594,17 +1595,17 @@ out: } -/** +/** * int journal_try_to_free_buffers() - try to free page buffers. * @journal: journal for operation * @page: to try and free * @unused_gfp_mask: unused * - * + * * For all the buffers on this page, * if they are fully written out ordered data, move them onto BUF_CLEAN * so try_to_free_buffers() can reap them. - * + * * This function returns non-zero if we wish try_to_free_buffers() * to be called. We do this if the page is releasable by try_to_free_buffers(). * We also do it if the page has locked or dirty buffers and the caller wants @@ -1628,7 +1629,7 @@ out: * cannot happen because we never reallocate freed data as metadata * while the data is part of a transaction. Yes? */ -int journal_try_to_free_buffers(journal_t *journal, +int journal_try_to_free_buffers(journal_t *journal, struct page *page, gfp_t unused_gfp_mask) { struct buffer_head *head; @@ -1696,7 +1697,7 @@ static int __dispose_buffer(struct journal_head *jh, transaction_t *transaction) } /* - * journal_invalidatepage + * journal_invalidatepage * * This code is tricky. It has a number of cases to deal with. * @@ -1704,15 +1705,15 @@ static int __dispose_buffer(struct journal_head *jh, transaction_t *transaction) * * i_size must be updated on disk before we start calling invalidatepage on the * data. - * + * * This is done in ext3 by defining an ext3_setattr method which * updates i_size before truncate gets going. By maintaining this * invariant, we can be sure that it is safe to throw away any buffers * attached to the current transaction: once the transaction commits, * we know that the data will not be needed. - * + * * Note however that we can *not* throw away data belonging to the - * previous, committing transaction! + * previous, committing transaction! * * Any disk blocks which *are* part of the previous, committing * transaction (and which therefore cannot be discarded immediately) are @@ -1731,7 +1732,7 @@ static int __dispose_buffer(struct journal_head *jh, transaction_t *transaction) * don't make guarantees about the order in which data hits disk --- in * particular we don't guarantee that new dirty data is flushed before * transaction commit --- so it is always safe just to discard data - * immediately in that mode. --sct + * immediately in that mode. --sct */ /* @@ -1875,9 +1876,9 @@ zap_buffer_unlocked: return may_free; } -/** +/** * void journal_invalidatepage() - * @journal: journal to use for flush... + * @journal: journal to use for flush... * @page: page to flush * @offset: length of page to invalidate. * @@ -1885,7 +1886,7 @@ zap_buffer_unlocked: * */ void journal_invalidatepage(journal_t *journal, - struct page *page, + struct page *page, unsigned long offset) { struct buffer_head *head, *bh, *next; @@ -1907,7 +1908,7 @@ void journal_invalidatepage(journal_t *journal, next = bh->b_this_page; if (offset <= curr_off) { - /* This block is wholly outside the truncation point */ + /* This block is wholly outside the truncation point */ lock_buffer(bh); may_free &= journal_unmap_buffer(journal, bh); unlock_buffer(bh); @@ -1923,8 +1924,8 @@ void journal_invalidatepage(journal_t *journal, } } -/* - * File a buffer on the given transaction list. +/* + * File a buffer on the given transaction list. */ void __journal_file_buffer(struct journal_head *jh, transaction_t *transaction, int jlist) @@ -1947,7 +1948,7 @@ void __journal_file_buffer(struct journal_head *jh, * with __jbd_unexpected_dirty_buffer()'s handling of dirty * state. */ - if (jlist == BJ_Metadata || jlist == BJ_Reserved || + if (jlist == BJ_Metadata || jlist == BJ_Reserved || jlist == BJ_Shadow || jlist == BJ_Forget) { if (test_clear_buffer_dirty(bh) || test_clear_buffer_jbddirty(bh)) @@ -2007,7 +2008,7 @@ void journal_file_buffer(struct journal_head *jh, jbd_unlock_bh_state(jh2bh(jh)); } -/* +/* * Remove a buffer from its current buffer list in preparation for * dropping it from its current transaction entirely. If the buffer has * already started to be used by a subsequent transaction, refile the @@ -2059,7 +2060,7 @@ void __journal_refile_buffer(struct journal_head *jh) * to the caller to remove the journal_head if necessary. For the * unlocked journal_refile_buffer call, the caller isn't going to be * doing anything else to the buffer so we need to do the cleanup - * ourselves to avoid a jh leak. + * ourselves to avoid a jh leak. * * *** The journal_head may be freed by this call! *** */ diff --git a/fs/jffs/inode-v23.c b/fs/jffs/inode-v23.c index 9e46ea6da75..f5cf9c93e24 100644 --- a/fs/jffs/inode-v23.c +++ b/fs/jffs/inode-v23.c @@ -59,7 +59,7 @@ static const struct file_operations jffs_file_operations; static struct inode_operations jffs_file_inode_operations; static const struct file_operations jffs_dir_operations; static struct inode_operations jffs_dir_inode_operations; -static struct address_space_operations jffs_address_operations; +static const struct address_space_operations jffs_address_operations; kmem_cache_t *node_cache = NULL; kmem_cache_t *fm_cache = NULL; @@ -364,12 +364,11 @@ jffs_new_inode(const struct inode * dir, struct jffs_raw_inode *raw_inode, inode->i_ctime.tv_nsec = 0; inode->i_mtime.tv_nsec = 0; inode->i_atime.tv_nsec = 0; - inode->i_blksize = PAGE_SIZE; inode->i_blocks = (inode->i_size + 511) >> 9; f = jffs_find_file(c, raw_inode->ino); - inode->u.generic_ip = (void *)f; + inode->i_private = (void *)f; insert_inode_hash(inode); return inode; @@ -442,7 +441,7 @@ jffs_rename(struct inode *old_dir, struct dentry *old_dentry, }); result = -ENOTDIR; - if (!(old_dir_f = (struct jffs_file *)old_dir->u.generic_ip)) { + if (!(old_dir_f = old_dir->i_private)) { D(printk("jffs_rename(): Old dir invalid.\n")); goto jffs_rename_end; } @@ -456,7 +455,7 @@ jffs_rename(struct inode *old_dir, struct dentry *old_dentry, /* Find the new directory. */ result = -ENOTDIR; - if (!(new_dir_f = (struct jffs_file *)new_dir->u.generic_ip)) { + if (!(new_dir_f = new_dir->i_private)) { D(printk("jffs_rename(): New dir invalid.\n")); goto jffs_rename_end; } @@ -593,7 +592,7 @@ jffs_readdir(struct file *filp, void *dirent, filldir_t filldir) } else { ddino = ((struct jffs_file *) - inode->u.generic_ip)->pino; + inode->i_private)->pino; } D3(printk("jffs_readdir(): \"..\" %u\n", ddino)); if (filldir(dirent, "..", 2, filp->f_pos, ddino, DT_DIR) < 0) { @@ -604,7 +603,7 @@ jffs_readdir(struct file *filp, void *dirent, filldir_t filldir) } filp->f_pos++; } - f = ((struct jffs_file *)inode->u.generic_ip)->children; + f = ((struct jffs_file *)inode->i_private)->children; j = 2; while(f && (f->deleted || j++ < filp->f_pos )) { @@ -652,7 +651,7 @@ jffs_lookup(struct inode *dir, struct dentry *dentry, struct nameidata *nd) lock_kernel(); D3({ - char *s = (char *)kmalloc(len + 1, GFP_KERNEL); + char *s = kmalloc(len + 1, GFP_KERNEL); memcpy(s, name, len); s[len] = '\0'; printk("jffs_lookup(): dir: 0x%p, name: \"%s\"\n", dir, s); @@ -668,7 +667,7 @@ jffs_lookup(struct inode *dir, struct dentry *dentry, struct nameidata *nd) } r = -EACCES; - if (!(d = (struct jffs_file *)dir->u.generic_ip)) { + if (!(d = (struct jffs_file *)dir->i_private)) { D(printk("jffs_lookup(): No such inode! (%lu)\n", dir->i_ino)); goto jffs_lookup_end; @@ -739,7 +738,7 @@ jffs_do_readpage_nolock(struct file *file, struct page *page) unsigned long read_len; int result; struct inode *inode = (struct inode*)page->mapping->host; - struct jffs_file *f = (struct jffs_file *)inode->u.generic_ip; + struct jffs_file *f = (struct jffs_file *)inode->i_private; struct jffs_control *c = (struct jffs_control *)inode->i_sb->s_fs_info; int r; loff_t offset; @@ -828,7 +827,7 @@ jffs_mkdir(struct inode *dir, struct dentry *dentry, int mode) }); lock_kernel(); - dir_f = (struct jffs_file *)dir->u.generic_ip; + dir_f = dir->i_private; ASSERT(if (!dir_f) { printk(KERN_ERR "jffs_mkdir(): No reference to a " @@ -972,7 +971,7 @@ jffs_remove(struct inode *dir, struct dentry *dentry, int type) kfree(_name); }); - dir_f = (struct jffs_file *) dir->u.generic_ip; + dir_f = dir->i_private; c = dir_f->c; result = -ENOENT; @@ -1082,7 +1081,7 @@ jffs_mknod(struct inode *dir, struct dentry *dentry, int mode, dev_t rdev) if (!old_valid_dev(rdev)) return -EINVAL; lock_kernel(); - dir_f = (struct jffs_file *)dir->u.generic_ip; + dir_f = dir->i_private; c = dir_f->c; D3(printk (KERN_NOTICE "mknod(): down biglock\n")); @@ -1173,8 +1172,8 @@ jffs_symlink(struct inode *dir, struct dentry *dentry, const char *symname) lock_kernel(); D1({ int len = dentry->d_name.len; - char *_name = (char *)kmalloc(len + 1, GFP_KERNEL); - char *_symname = (char *)kmalloc(symname_len + 1, GFP_KERNEL); + char *_name = kmalloc(len + 1, GFP_KERNEL); + char *_symname = kmalloc(symname_len + 1, GFP_KERNEL); memcpy(_name, dentry->d_name.name, len); _name[len] = '\0'; memcpy(_symname, symname, symname_len); @@ -1186,7 +1185,7 @@ jffs_symlink(struct inode *dir, struct dentry *dentry, const char *symname) kfree(_symname); }); - dir_f = (struct jffs_file *)dir->u.generic_ip; + dir_f = dir->i_private; ASSERT(if (!dir_f) { printk(KERN_ERR "jffs_symlink(): No reference to a " "jffs_file struct in inode.\n"); @@ -1282,14 +1281,14 @@ jffs_create(struct inode *dir, struct dentry *dentry, int mode, lock_kernel(); D1({ int len = dentry->d_name.len; - char *s = (char *)kmalloc(len + 1, GFP_KERNEL); + char *s = kmalloc(len + 1, GFP_KERNEL); memcpy(s, dentry->d_name.name, len); s[len] = '\0'; printk("jffs_create(): dir: 0x%p, name: \"%s\"\n", dir, s); kfree(s); }); - dir_f = (struct jffs_file *)dir->u.generic_ip; + dir_f = dir->i_private; ASSERT(if (!dir_f) { printk(KERN_ERR "jffs_create(): No reference to a " "jffs_file struct in inode.\n"); @@ -1403,9 +1402,9 @@ jffs_file_write(struct file *filp, const char *buf, size_t count, goto out_isem; } - if (!(f = (struct jffs_file *)inode->u.generic_ip)) { - D(printk("jffs_file_write(): inode->u.generic_ip = 0x%p\n", - inode->u.generic_ip)); + if (!(f = inode->i_private)) { + D(printk("jffs_file_write(): inode->i_private = 0x%p\n", + inode->i_private)); goto out_isem; } @@ -1614,7 +1613,7 @@ jffs_ioctl(struct inode *inode, struct file *filp, unsigned int cmd, } /* jffs_ioctl() */ -static struct address_space_operations jffs_address_operations = { +static const struct address_space_operations jffs_address_operations = { .readpage = jffs_readpage, .prepare_write = jffs_prepare_write, .commit_write = jffs_commit_write, @@ -1693,7 +1692,7 @@ jffs_read_inode(struct inode *inode) mutex_unlock(&c->fmc->biglock); return; } - inode->u.generic_ip = (void *)f; + inode->i_private = f; inode->i_mode = f->mode; inode->i_nlink = f->nlink; inode->i_uid = f->uid; @@ -1706,7 +1705,6 @@ jffs_read_inode(struct inode *inode) inode->i_mtime.tv_nsec = inode->i_ctime.tv_nsec = 0; - inode->i_blksize = PAGE_SIZE; inode->i_blocks = (inode->i_size + 511) >> 9; if (S_ISREG(inode->i_mode)) { inode->i_op = &jffs_file_inode_operations; @@ -1748,7 +1746,7 @@ jffs_delete_inode(struct inode *inode) lock_kernel(); inode->i_size = 0; inode->i_blocks = 0; - inode->u.generic_ip = NULL; + inode->i_private = NULL; clear_inode(inode); if (inode->i_nlink == 0) { c = (struct jffs_control *) inode->i_sb->s_fs_info; diff --git a/fs/jffs/intrep.c b/fs/jffs/intrep.c index 5371a403130..4a543e11497 100644 --- a/fs/jffs/intrep.c +++ b/fs/jffs/intrep.c @@ -55,7 +55,6 @@ * */ -#include <linux/config.h> #include <linux/types.h> #include <linux/slab.h> #include <linux/jffs.h> @@ -489,13 +488,11 @@ jffs_create_file(struct jffs_control *c, { struct jffs_file *f; - if (!(f = (struct jffs_file *)kmalloc(sizeof(struct jffs_file), - GFP_KERNEL))) { + if (!(f = kzalloc(sizeof(*f), GFP_KERNEL))) { D(printk("jffs_create_file(): Failed!\n")); return NULL; } no_jffs_file++; - memset(f, 0, sizeof(struct jffs_file)); f->ino = raw_inode->ino; f->pino = raw_inode->pino; f->nlink = raw_inode->nlink; @@ -517,7 +514,7 @@ jffs_create_control(struct super_block *sb) D2(printk("jffs_create_control()\n")); - if (!(c = (struct jffs_control *)kmalloc(s, GFP_KERNEL))) { + if (!(c = kmalloc(s, GFP_KERNEL))) { goto fail_control; } DJM(no_jffs_control++); @@ -525,7 +522,7 @@ jffs_create_control(struct super_block *sb) c->gc_task = NULL; c->hash_len = JFFS_HASH_SIZE; s = sizeof(struct list_head) * c->hash_len; - if (!(c->hash = (struct list_head *)kmalloc(s, GFP_KERNEL))) { + if (!(c->hash = kmalloc(s, GFP_KERNEL))) { goto fail_hash; } DJM(no_hash++); @@ -594,8 +591,7 @@ jffs_add_virtual_root(struct jffs_control *c) D2(printk("jffs_add_virtual_root(): " "Creating a virtual root directory.\n")); - if (!(root = (struct jffs_file *)kmalloc(sizeof(struct jffs_file), - GFP_KERNEL))) { + if (!(root = kmalloc(sizeof(struct jffs_file), GFP_KERNEL))) { return -ENOMEM; } no_jffs_file++; diff --git a/fs/jffs/jffs_fm.c b/fs/jffs/jffs_fm.c index 7d8ca1aeace..29b68d939bd 100644 --- a/fs/jffs/jffs_fm.c +++ b/fs/jffs/jffs_fm.c @@ -94,8 +94,7 @@ jffs_build_begin(struct jffs_control *c, int unit) struct mtd_info *mtd; D3(printk("jffs_build_begin()\n")); - fmc = (struct jffs_fmcontrol *)kmalloc(sizeof(struct jffs_fmcontrol), - GFP_KERNEL); + fmc = kmalloc(sizeof(*fmc), GFP_KERNEL); if (!fmc) { D(printk("jffs_build_begin(): Allocation of " "struct jffs_fmcontrol failed!\n")); @@ -486,8 +485,7 @@ jffs_add_node(struct jffs_node *node) D3(printk("jffs_add_node(): ino = %u\n", node->ino)); - ref = (struct jffs_node_ref *)kmalloc(sizeof(struct jffs_node_ref), - GFP_KERNEL); + ref = kmalloc(sizeof(*ref), GFP_KERNEL); if (!ref) return -ENOMEM; diff --git a/fs/jffs/jffs_fm.h b/fs/jffs/jffs_fm.h index c794d923df2..9ee6ad29eff 100644 --- a/fs/jffs/jffs_fm.h +++ b/fs/jffs/jffs_fm.h @@ -20,7 +20,6 @@ #ifndef __LINUX_JFFS_FM_H__ #define __LINUX_JFFS_FM_H__ -#include <linux/config.h> #include <linux/types.h> #include <linux/jffs.h> #include <linux/mtd/mtd.h> diff --git a/fs/jffs2/acl.c b/fs/jffs2/acl.c index 320dd48b834..0ae3cd10702 100644 --- a/fs/jffs2/acl.c +++ b/fs/jffs2/acl.c @@ -267,6 +267,8 @@ static int jffs2_set_acl(struct inode *inode, int type, struct posix_acl *acl) } rc = do_jffs2_setxattr(inode, xprefix, "", value, size, 0); + if (!value && rc == -ENODATA) + rc = 0; if (value) kfree(value); if (!rc) { @@ -343,10 +345,8 @@ int jffs2_init_acl(struct inode *inode, struct inode *dir) return rc; } -void jffs2_clear_acl(struct inode *inode) +void jffs2_clear_acl(struct jffs2_inode_info *f) { - struct jffs2_inode_info *f = JFFS2_INODE_INFO(inode); - if (f->i_acl_access && f->i_acl_access != JFFS2_ACL_NOT_CACHED) { posix_acl_release(f->i_acl_access); f->i_acl_access = JFFS2_ACL_NOT_CACHED; diff --git a/fs/jffs2/acl.h b/fs/jffs2/acl.h index 8893bd1a6ba..fa327dbd317 100644 --- a/fs/jffs2/acl.h +++ b/fs/jffs2/acl.h @@ -30,7 +30,7 @@ struct jffs2_acl_header { extern int jffs2_permission(struct inode *, int, struct nameidata *); extern int jffs2_acl_chmod(struct inode *); extern int jffs2_init_acl(struct inode *, struct inode *); -extern void jffs2_clear_acl(struct inode *); +extern void jffs2_clear_acl(struct jffs2_inode_info *); extern struct xattr_handler jffs2_acl_access_xattr_handler; extern struct xattr_handler jffs2_acl_default_xattr_handler; @@ -40,6 +40,6 @@ extern struct xattr_handler jffs2_acl_default_xattr_handler; #define jffs2_permission NULL #define jffs2_acl_chmod(inode) (0) #define jffs2_init_acl(inode,dir) (0) -#define jffs2_clear_acl(inode) +#define jffs2_clear_acl(f) #endif /* CONFIG_JFFS2_FS_POSIX_ACL */ diff --git a/fs/jffs2/compr_zlib.c b/fs/jffs2/compr_zlib.c index 5c63e0cdcf4..3681d0728ac 100644 --- a/fs/jffs2/compr_zlib.c +++ b/fs/jffs2/compr_zlib.c @@ -15,7 +15,6 @@ #error "The userspace support got too messy and was removed. Update your mkfs.jffs2" #endif -#include <linux/config.h> #include <linux/kernel.h> #include <linux/sched.h> #include <linux/slab.h> diff --git a/fs/jffs2/debug.h b/fs/jffs2/debug.h index 5fa494a792b..3daf3bca037 100644 --- a/fs/jffs2/debug.h +++ b/fs/jffs2/debug.h @@ -13,7 +13,6 @@ #ifndef _JFFS2_DEBUG_H_ #define _JFFS2_DEBUG_H_ -#include <linux/config.h> #ifndef CONFIG_JFFS2_FS_DEBUG #define CONFIG_JFFS2_FS_DEBUG 0 diff --git a/fs/jffs2/erase.c b/fs/jffs2/erase.c index 1862e8bc101..ad0121088dd 100644 --- a/fs/jffs2/erase.c +++ b/fs/jffs2/erase.c @@ -53,8 +53,7 @@ static void jffs2_erase_block(struct jffs2_sb_info *c, if (!instr) { printk(KERN_WARNING "kmalloc for struct erase_info in jffs2_erase_block failed. Refiling block for later\n"); spin_lock(&c->erase_completion_lock); - list_del(&jeb->list); - list_add(&jeb->list, &c->erase_pending_list); + list_move(&jeb->list, &c->erase_pending_list); c->erasing_size -= c->sector_size; c->dirty_size += c->sector_size; jeb->dirty_size = c->sector_size; @@ -86,8 +85,7 @@ static void jffs2_erase_block(struct jffs2_sb_info *c, /* Erase failed immediately. Refile it on the list */ D1(printk(KERN_DEBUG "Erase at 0x%08x failed: %d. Refiling on erase_pending_list\n", jeb->offset, ret)); spin_lock(&c->erase_completion_lock); - list_del(&jeb->list); - list_add(&jeb->list, &c->erase_pending_list); + list_move(&jeb->list, &c->erase_pending_list); c->erasing_size -= c->sector_size; c->dirty_size += c->sector_size; jeb->dirty_size = c->sector_size; @@ -161,8 +159,7 @@ static void jffs2_erase_succeeded(struct jffs2_sb_info *c, struct jffs2_eraseblo { D1(printk(KERN_DEBUG "Erase completed successfully at 0x%08x\n", jeb->offset)); spin_lock(&c->erase_completion_lock); - list_del(&jeb->list); - list_add_tail(&jeb->list, &c->erase_complete_list); + list_move_tail(&jeb->list, &c->erase_complete_list); spin_unlock(&c->erase_completion_lock); /* Ensure that kupdated calls us again to mark them clean */ jffs2_erase_pending_trigger(c); @@ -178,8 +175,7 @@ static void jffs2_erase_failed(struct jffs2_sb_info *c, struct jffs2_eraseblock if (!jffs2_write_nand_badblock(c, jeb, bad_offset)) { /* We'd like to give this block another try. */ spin_lock(&c->erase_completion_lock); - list_del(&jeb->list); - list_add(&jeb->list, &c->erase_pending_list); + list_move(&jeb->list, &c->erase_pending_list); c->erasing_size -= c->sector_size; c->dirty_size += c->sector_size; jeb->dirty_size = c->sector_size; @@ -191,8 +187,7 @@ static void jffs2_erase_failed(struct jffs2_sb_info *c, struct jffs2_eraseblock spin_lock(&c->erase_completion_lock); c->erasing_size -= c->sector_size; c->bad_size += c->sector_size; - list_del(&jeb->list); - list_add(&jeb->list, &c->bad_list); + list_move(&jeb->list, &c->bad_list); c->nr_erasing_blocks--; spin_unlock(&c->erase_completion_lock); wake_up(&c->erase_wait); @@ -230,7 +225,6 @@ static inline void jffs2_remove_node_refs_from_ino_list(struct jffs2_sb_info *c, at the end of the linked list. Stash it and continue from the beginning of the list */ ic = (struct jffs2_inode_cache *)(*prev); - BUG_ON(ic->class != RAWNODE_CLASS_INODE_CACHE); prev = &ic->nodes; continue; } @@ -254,7 +248,8 @@ static inline void jffs2_remove_node_refs_from_ino_list(struct jffs2_sb_info *c, /* PARANOIA */ if (!ic) { - printk(KERN_WARNING "inode_cache not found in remove_node_refs()!!\n"); + JFFS2_WARNING("inode_cache/xattr_datum/xattr_ref" + " not found in remove_node_refs()!!\n"); return; } @@ -279,8 +274,19 @@ static inline void jffs2_remove_node_refs_from_ino_list(struct jffs2_sb_info *c, printk("\n"); }); - if (ic->nodes == (void *)ic && ic->nlink == 0) - jffs2_del_ino_cache(c, ic); + switch (ic->class) { +#ifdef CONFIG_JFFS2_FS_XATTR + case RAWNODE_CLASS_XATTR_DATUM: + jffs2_release_xattr_datum(c, (struct jffs2_xattr_datum *)ic); + break; + case RAWNODE_CLASS_XATTR_REF: + jffs2_release_xattr_ref(c, (struct jffs2_xattr_ref *)ic); + break; +#endif + default: + if (ic->nodes == (void *)ic && ic->nlink == 0) + jffs2_del_ino_cache(c, ic); + } } void jffs2_free_jeb_node_refs(struct jffs2_sb_info *c, struct jffs2_eraseblock *jeb) diff --git a/fs/jffs2/file.c b/fs/jffs2/file.c index bb8844f40e4..3ed6e3e120b 100644 --- a/fs/jffs2/file.c +++ b/fs/jffs2/file.c @@ -62,7 +62,7 @@ struct inode_operations jffs2_file_inode_operations = .removexattr = jffs2_removexattr }; -struct address_space_operations jffs2_file_address_operations = +const struct address_space_operations jffs2_file_address_operations = { .readpage = jffs2_readpage, .prepare_write =jffs2_prepare_write, diff --git a/fs/jffs2/fs.c b/fs/jffs2/fs.c index 2900ec3ec3a..72d9909d95f 100644 --- a/fs/jffs2/fs.c +++ b/fs/jffs2/fs.c @@ -12,7 +12,6 @@ */ #include <linux/capability.h> -#include <linux/config.h> #include <linux/kernel.h> #include <linux/sched.h> #include <linux/fs.h> @@ -227,8 +226,6 @@ void jffs2_clear_inode (struct inode *inode) struct jffs2_inode_info *f = JFFS2_INODE_INFO(inode); D1(printk(KERN_DEBUG "jffs2_clear_inode(): ino #%lu mode %o\n", inode->i_ino, inode->i_mode)); - - jffs2_xattr_delete_inode(c, f->inocache); jffs2_do_clear_inode(c, f); } @@ -266,7 +263,6 @@ void jffs2_read_inode (struct inode *inode) inode->i_nlink = f->inocache->nlink; - inode->i_blksize = PAGE_SIZE; inode->i_blocks = (inode->i_size + 511) >> 9; switch (inode->i_mode & S_IFMT) { @@ -452,7 +448,6 @@ struct inode *jffs2_new_inode (struct inode *dir_i, int mode, struct jffs2_raw_i inode->i_atime = inode->i_ctime = inode->i_mtime = CURRENT_TIME_SEC; ri->atime = ri->mtime = ri->ctime = cpu_to_je32(I_SEC(inode->i_mtime)); - inode->i_blksize = PAGE_SIZE; inode->i_blocks = 0; inode->i_size = 0; diff --git a/fs/jffs2/gc.c b/fs/jffs2/gc.c index 477c526d638..daff3341ff9 100644 --- a/fs/jffs2/gc.c +++ b/fs/jffs2/gc.c @@ -165,6 +165,7 @@ int jffs2_garbage_collect_pass(struct jffs2_sb_info *c) D1(printk(KERN_DEBUG "Skipping check of ino #%d with nlink zero\n", ic->ino)); spin_unlock(&c->inocache_lock); + jffs2_xattr_delete_inode(c, ic); continue; } switch(ic->state) { @@ -275,13 +276,12 @@ int jffs2_garbage_collect_pass(struct jffs2_sb_info *c) * We can decide whether this node is inode or xattr by ic->class. */ if (ic->class == RAWNODE_CLASS_XATTR_DATUM || ic->class == RAWNODE_CLASS_XATTR_REF) { - BUG_ON(raw->next_in_ino != (void *)ic); spin_unlock(&c->erase_completion_lock); if (ic->class == RAWNODE_CLASS_XATTR_DATUM) { - ret = jffs2_garbage_collect_xattr_datum(c, (struct jffs2_xattr_datum *)ic); + ret = jffs2_garbage_collect_xattr_datum(c, (struct jffs2_xattr_datum *)ic, raw); } else { - ret = jffs2_garbage_collect_xattr_ref(c, (struct jffs2_xattr_ref *)ic); + ret = jffs2_garbage_collect_xattr_ref(c, (struct jffs2_xattr_ref *)ic, raw); } goto release_sem; } diff --git a/fs/jffs2/jffs2_fs_i.h b/fs/jffs2/jffs2_fs_i.h index 2e0cc8e00b8..3a566077ac9 100644 --- a/fs/jffs2/jffs2_fs_i.h +++ b/fs/jffs2/jffs2_fs_i.h @@ -41,11 +41,7 @@ struct jffs2_inode_info { uint16_t flags; uint8_t usercompr; -#if !defined (__ECOS) -#if LINUX_VERSION_CODE > KERNEL_VERSION(2,5,2) struct inode vfs_inode; -#endif -#endif #ifdef CONFIG_JFFS2_FS_POSIX_ACL struct posix_acl *i_acl_access; struct posix_acl *i_acl_default; diff --git a/fs/jffs2/jffs2_fs_sb.h b/fs/jffs2/jffs2_fs_sb.h index 935fec1b120..b98594992ee 100644 --- a/fs/jffs2/jffs2_fs_sb.h +++ b/fs/jffs2/jffs2_fs_sb.h @@ -119,8 +119,11 @@ struct jffs2_sb_info { #ifdef CONFIG_JFFS2_FS_XATTR #define XATTRINDEX_HASHSIZE (57) uint32_t highest_xid; + uint32_t highest_xseqno; struct list_head xattrindex[XATTRINDEX_HASHSIZE]; struct list_head xattr_unchecked; + struct list_head xattr_dead_list; + struct jffs2_xattr_ref *xref_dead_list; struct jffs2_xattr_ref *xref_temp; struct rw_semaphore xattr_sem; uint32_t xdatum_mem_usage; diff --git a/fs/jffs2/malloc.c b/fs/jffs2/malloc.c index 4889d0700c0..33f29100501 100644 --- a/fs/jffs2/malloc.c +++ b/fs/jffs2/malloc.c @@ -190,7 +190,7 @@ void jffs2_free_tmp_dnode_info(struct jffs2_tmp_dnode_info *x) kmem_cache_free(tmp_dnode_info_slab, x); } -struct jffs2_raw_node_ref *jffs2_alloc_refblock(void) +static struct jffs2_raw_node_ref *jffs2_alloc_refblock(void) { struct jffs2_raw_node_ref *ret; @@ -291,6 +291,7 @@ struct jffs2_xattr_datum *jffs2_alloc_xattr_datum(void) memset(xd, 0, sizeof(struct jffs2_xattr_datum)); xd->class = RAWNODE_CLASS_XATTR_DATUM; + xd->node = (void *)xd; INIT_LIST_HEAD(&xd->xindex); return xd; } @@ -309,6 +310,7 @@ struct jffs2_xattr_ref *jffs2_alloc_xattr_ref(void) memset(ref, 0, sizeof(struct jffs2_xattr_ref)); ref->class = RAWNODE_CLASS_XATTR_REF; + ref->node = (void *)ref; return ref; } diff --git a/fs/jffs2/nodelist.c b/fs/jffs2/nodelist.c index 927dfe42ba7..5a6b4d64206 100644 --- a/fs/jffs2/nodelist.c +++ b/fs/jffs2/nodelist.c @@ -21,6 +21,9 @@ #include <linux/pagemap.h> #include "nodelist.h" +static void jffs2_obsolete_node_frag(struct jffs2_sb_info *c, + struct jffs2_node_frag *this); + void jffs2_add_fd_to_list(struct jffs2_sb_info *c, struct jffs2_full_dirent *new, struct jffs2_full_dirent **list) { struct jffs2_full_dirent **prev = list; @@ -87,7 +90,8 @@ void jffs2_truncate_fragtree(struct jffs2_sb_info *c, struct rb_root *list, uint } } -void jffs2_obsolete_node_frag(struct jffs2_sb_info *c, struct jffs2_node_frag *this) +static void jffs2_obsolete_node_frag(struct jffs2_sb_info *c, + struct jffs2_node_frag *this) { if (this->node) { this->node->frags--; @@ -906,6 +910,9 @@ void jffs2_del_ino_cache(struct jffs2_sb_info *c, struct jffs2_inode_cache *old) { struct jffs2_inode_cache **prev; +#ifdef CONFIG_JFFS2_FS_XATTR + BUG_ON(old->xref); +#endif dbg_inocache("del %p (ino #%u)\n", old, old->ino); spin_lock(&c->inocache_lock); diff --git a/fs/jffs2/nodelist.h b/fs/jffs2/nodelist.h index b16c60bbcf6..0ddfd70307f 100644 --- a/fs/jffs2/nodelist.h +++ b/fs/jffs2/nodelist.h @@ -14,7 +14,6 @@ #ifndef __JFFS2_NODELIST_H__ #define __JFFS2_NODELIST_H__ -#include <linux/config.h> #include <linux/fs.h> #include <linux/types.h> #include <linux/jffs2.h> @@ -335,7 +334,6 @@ void jffs2_kill_fragtree(struct rb_root *root, struct jffs2_sb_info *c_delete); struct rb_node *rb_next(struct rb_node *); struct rb_node *rb_prev(struct rb_node *); void rb_replace_node(struct rb_node *victim, struct rb_node *new, struct rb_root *root); -void jffs2_obsolete_node_frag(struct jffs2_sb_info *c, struct jffs2_node_frag *this); int jffs2_add_full_dnode_to_inode(struct jffs2_sb_info *c, struct jffs2_inode_info *f, struct jffs2_full_dnode *fn); void jffs2_truncate_fragtree (struct jffs2_sb_info *c, struct rb_root *list, uint32_t size); int jffs2_add_older_frag_to_fragtree(struct jffs2_sb_info *c, struct jffs2_inode_info *f, struct jffs2_tmp_dnode_info *tn); @@ -427,8 +425,6 @@ char *jffs2_getlink(struct jffs2_sb_info *c, struct jffs2_inode_info *f); /* scan.c */ int jffs2_scan_medium(struct jffs2_sb_info *c); void jffs2_rotate_lists(struct jffs2_sb_info *c); -int jffs2_fill_scan_buf(struct jffs2_sb_info *c, void *buf, - uint32_t ofs, uint32_t len); struct jffs2_inode_cache *jffs2_scan_make_ino_cache(struct jffs2_sb_info *c, uint32_t ino); int jffs2_scan_classify_jeb(struct jffs2_sb_info *c, struct jffs2_eraseblock *jeb); int jffs2_scan_dirty_space(struct jffs2_sb_info *c, struct jffs2_eraseblock *jeb, uint32_t size); diff --git a/fs/jffs2/nodemgmt.c b/fs/jffs2/nodemgmt.c index 8bedfd2ff68..d88376992ed 100644 --- a/fs/jffs2/nodemgmt.c +++ b/fs/jffs2/nodemgmt.c @@ -211,8 +211,7 @@ static int jffs2_find_nextblock(struct jffs2_sb_info *c) struct jffs2_eraseblock *ejeb; ejeb = list_entry(c->erasable_list.next, struct jffs2_eraseblock, list); - list_del(&ejeb->list); - list_add_tail(&ejeb->list, &c->erase_pending_list); + list_move_tail(&ejeb->list, &c->erase_pending_list); c->nr_erasing_blocks++; jffs2_erase_pending_trigger(c); D1(printk(KERN_DEBUG "jffs2_find_nextblock: Triggering erase of erasable block at 0x%08x\n", @@ -684,19 +683,26 @@ void jffs2_mark_node_obsolete(struct jffs2_sb_info *c, struct jffs2_raw_node_ref spin_lock(&c->erase_completion_lock); ic = jffs2_raw_ref_to_ic(ref); - /* It seems we should never call jffs2_mark_node_obsolete() for - XATTR nodes.... yet. Make sure we notice if/when we change - that :) */ - BUG_ON(ic->class != RAWNODE_CLASS_INODE_CACHE); for (p = &ic->nodes; (*p) != ref; p = &((*p)->next_in_ino)) ; *p = ref->next_in_ino; ref->next_in_ino = NULL; - if (ic->nodes == (void *)ic && ic->nlink == 0) - jffs2_del_ino_cache(c, ic); - + switch (ic->class) { +#ifdef CONFIG_JFFS2_FS_XATTR + case RAWNODE_CLASS_XATTR_DATUM: + jffs2_release_xattr_datum(c, (struct jffs2_xattr_datum *)ic); + break; + case RAWNODE_CLASS_XATTR_REF: + jffs2_release_xattr_ref(c, (struct jffs2_xattr_ref *)ic); + break; +#endif + default: + if (ic->nodes == (void *)ic && ic->nlink == 0) + jffs2_del_ino_cache(c, ic); + break; + } spin_unlock(&c->erase_completion_lock); } diff --git a/fs/jffs2/os-linux.h b/fs/jffs2/os-linux.h index 6b522356540..9f41fc01a37 100644 --- a/fs/jffs2/os-linux.h +++ b/fs/jffs2/os-linux.h @@ -158,7 +158,7 @@ extern struct inode_operations jffs2_dir_inode_operations; /* file.c */ extern const struct file_operations jffs2_file_operations; extern struct inode_operations jffs2_file_inode_operations; -extern struct address_space_operations jffs2_file_address_operations; +extern const struct address_space_operations jffs2_file_address_operations; int jffs2_fsync(struct file *, struct dentry *, int); int jffs2_do_readpage_unlock (struct inode *inode, struct page *pg); diff --git a/fs/jffs2/readinode.c b/fs/jffs2/readinode.c index 5fec012b02e..266423b2709 100644 --- a/fs/jffs2/readinode.c +++ b/fs/jffs2/readinode.c @@ -968,6 +968,8 @@ void jffs2_do_clear_inode(struct jffs2_sb_info *c, struct jffs2_inode_info *f) struct jffs2_full_dirent *fd, *fds; int deleted; + jffs2_clear_acl(f); + jffs2_xattr_delete_inode(c, f->inocache); down(&f->sem); deleted = f->inocache && !f->inocache->nlink; diff --git a/fs/jffs2/scan.c b/fs/jffs2/scan.c index 61618080b86..e2413466ddd 100644 --- a/fs/jffs2/scan.c +++ b/fs/jffs2/scan.c @@ -274,8 +274,8 @@ int jffs2_scan_medium(struct jffs2_sb_info *c) return ret; } -int jffs2_fill_scan_buf (struct jffs2_sb_info *c, void *buf, - uint32_t ofs, uint32_t len) +static int jffs2_fill_scan_buf(struct jffs2_sb_info *c, void *buf, + uint32_t ofs, uint32_t len) { int ret; size_t retlen; @@ -317,20 +317,23 @@ static int jffs2_scan_xattr_node(struct jffs2_sb_info *c, struct jffs2_erasebloc struct jffs2_summary *s) { struct jffs2_xattr_datum *xd; - uint32_t totlen, crc; + uint32_t xid, version, totlen, crc; int err; crc = crc32(0, rx, sizeof(struct jffs2_raw_xattr) - 4); if (crc != je32_to_cpu(rx->node_crc)) { - if (je32_to_cpu(rx->node_crc) != 0xffffffff) - JFFS2_WARNING("node CRC failed at %#08x, read=%#08x, calc=%#08x\n", - ofs, je32_to_cpu(rx->node_crc), crc); + JFFS2_WARNING("node CRC failed at %#08x, read=%#08x, calc=%#08x\n", + ofs, je32_to_cpu(rx->node_crc), crc); if ((err = jffs2_scan_dirty_space(c, jeb, je32_to_cpu(rx->totlen)))) return err; return 0; } - totlen = PAD(sizeof(*rx) + rx->name_len + 1 + je16_to_cpu(rx->value_len)); + xid = je32_to_cpu(rx->xid); + version = je32_to_cpu(rx->version); + + totlen = PAD(sizeof(struct jffs2_raw_xattr) + + rx->name_len + 1 + je16_to_cpu(rx->value_len)); if (totlen != je32_to_cpu(rx->totlen)) { JFFS2_WARNING("node length mismatch at %#08x, read=%u, calc=%u\n", ofs, je32_to_cpu(rx->totlen), totlen); @@ -339,22 +342,24 @@ static int jffs2_scan_xattr_node(struct jffs2_sb_info *c, struct jffs2_erasebloc return 0; } - xd = jffs2_setup_xattr_datum(c, je32_to_cpu(rx->xid), je32_to_cpu(rx->version)); - if (IS_ERR(xd)) { - if (PTR_ERR(xd) == -EEXIST) { - if ((err = jffs2_scan_dirty_space(c, jeb, PAD(je32_to_cpu(rx->totlen))))) - return err; - return 0; - } + xd = jffs2_setup_xattr_datum(c, xid, version); + if (IS_ERR(xd)) return PTR_ERR(xd); - } - xd->xprefix = rx->xprefix; - xd->name_len = rx->name_len; - xd->value_len = je16_to_cpu(rx->value_len); - xd->data_crc = je32_to_cpu(rx->data_crc); - xd->node = jffs2_link_node_ref(c, jeb, ofs | REF_PRISTINE, totlen, NULL); - /* FIXME */ xd->node->next_in_ino = (void *)xd; + if (xd->version > version) { + struct jffs2_raw_node_ref *raw + = jffs2_link_node_ref(c, jeb, ofs | REF_PRISTINE, totlen, NULL); + raw->next_in_ino = xd->node->next_in_ino; + xd->node->next_in_ino = raw; + } else { + xd->version = version; + xd->xprefix = rx->xprefix; + xd->name_len = rx->name_len; + xd->value_len = je16_to_cpu(rx->value_len); + xd->data_crc = je32_to_cpu(rx->data_crc); + + jffs2_link_node_ref(c, jeb, ofs | REF_PRISTINE, totlen, (void *)xd); + } if (jffs2_sum_active()) jffs2_sum_add_xattr_mem(s, rx, ofs - jeb->offset); @@ -373,9 +378,8 @@ static int jffs2_scan_xref_node(struct jffs2_sb_info *c, struct jffs2_eraseblock crc = crc32(0, rr, sizeof(*rr) - 4); if (crc != je32_to_cpu(rr->node_crc)) { - if (je32_to_cpu(rr->node_crc) != 0xffffffff) - JFFS2_WARNING("node CRC failed at %#08x, read=%#08x, calc=%#08x\n", - ofs, je32_to_cpu(rr->node_crc), crc); + JFFS2_WARNING("node CRC failed at %#08x, read=%#08x, calc=%#08x\n", + ofs, je32_to_cpu(rr->node_crc), crc); if ((err = jffs2_scan_dirty_space(c, jeb, PAD(je32_to_cpu(rr->totlen))))) return err; return 0; @@ -395,6 +399,7 @@ static int jffs2_scan_xref_node(struct jffs2_sb_info *c, struct jffs2_eraseblock return -ENOMEM; /* BEFORE jffs2_build_xattr_subsystem() called, + * and AFTER xattr_ref is marked as a dead xref, * ref->xid is used to store 32bit xid, xd is not used * ref->ino is used to store 32bit inode-number, ic is not used * Thoes variables are declared as union, thus using those @@ -404,11 +409,13 @@ static int jffs2_scan_xref_node(struct jffs2_sb_info *c, struct jffs2_eraseblock */ ref->ino = je32_to_cpu(rr->ino); ref->xid = je32_to_cpu(rr->xid); + ref->xseqno = je32_to_cpu(rr->xseqno); + if (ref->xseqno > c->highest_xseqno) + c->highest_xseqno = (ref->xseqno & ~XREF_DELETE_MARKER); ref->next = c->xref_temp; c->xref_temp = ref; - ref->node = jffs2_link_node_ref(c, jeb, ofs | REF_PRISTINE, PAD(je32_to_cpu(rr->totlen)), NULL); - /* FIXME */ ref->node->next_in_ino = (void *)ref; + jffs2_link_node_ref(c, jeb, ofs | REF_PRISTINE, PAD(je32_to_cpu(rr->totlen)), (void *)ref); if (jffs2_sum_active()) jffs2_sum_add_xref_mem(s, rr, ofs - jeb->offset); diff --git a/fs/jffs2/summary.c b/fs/jffs2/summary.c index 0b02fc79e4d..e52cef526d9 100644 --- a/fs/jffs2/summary.c +++ b/fs/jffs2/summary.c @@ -5,7 +5,7 @@ * Zoltan Sogor <weth@inf.u-szeged.hu>, * Patrik Kluba <pajko@halom.u-szeged.hu>, * University of Szeged, Hungary - * 2005 KaiGai Kohei <kaigai@ak.jp.nec.com> + * 2006 KaiGai Kohei <kaigai@ak.jp.nec.com> * * For licensing information, see the file 'LICENCE' in this directory. * @@ -43,7 +43,7 @@ int jffs2_sum_init(struct jffs2_sb_info *c) return -ENOMEM; } - dbg_summary("returned succesfully\n"); + dbg_summary("returned successfully\n"); return 0; } @@ -252,6 +252,11 @@ int jffs2_sum_add_kvec(struct jffs2_sb_info *c, const struct kvec *invecs, union jffs2_node_union *node; struct jffs2_eraseblock *jeb; + if (c->summary->sum_size == JFFS2_SUMMARY_NOSUM_SIZE) { + dbg_summary("Summary is disabled for this jeb! Skipping summary info!\n"); + return 0; + } + node = invecs[0].iov_base; jeb = &c->blocks[ofs / c->sector_size]; ofs -= jeb->offset; @@ -310,8 +315,6 @@ int jffs2_sum_add_kvec(struct jffs2_sb_info *c, const struct kvec *invecs, #ifdef CONFIG_JFFS2_FS_XATTR case JFFS2_NODETYPE_XATTR: { struct jffs2_sum_xattr_mem *temp; - if (je32_to_cpu(node->x.version) == 0xffffffff) - return 0; temp = kmalloc(sizeof(struct jffs2_sum_xattr_mem), GFP_KERNEL); if (!temp) goto no_mem; @@ -327,10 +330,6 @@ int jffs2_sum_add_kvec(struct jffs2_sb_info *c, const struct kvec *invecs, } case JFFS2_NODETYPE_XREF: { struct jffs2_sum_xref_mem *temp; - - if (je32_to_cpu(node->r.ino) == 0xffffffff - && je32_to_cpu(node->r.xid) == 0xffffffff) - return 0; temp = kmalloc(sizeof(struct jffs2_sum_xref_mem), GFP_KERNEL); if (!temp) goto no_mem; @@ -483,22 +482,20 @@ static int jffs2_sum_process_sum_data(struct jffs2_sb_info *c, struct jffs2_eras xd = jffs2_setup_xattr_datum(c, je32_to_cpu(spx->xid), je32_to_cpu(spx->version)); - if (IS_ERR(xd)) { - if (PTR_ERR(xd) == -EEXIST) { - /* a newer version of xd exists */ - if ((err = jffs2_scan_dirty_space(c, jeb, je32_to_cpu(spx->totlen)))) - return err; - sp += JFFS2_SUMMARY_XATTR_SIZE; - break; - } - JFFS2_NOTICE("allocation of xattr_datum failed\n"); + if (IS_ERR(xd)) return PTR_ERR(xd); + if (xd->version > je32_to_cpu(spx->version)) { + /* node is not the newest one */ + struct jffs2_raw_node_ref *raw + = sum_link_node_ref(c, jeb, je32_to_cpu(spx->offset) | REF_UNCHECKED, + PAD(je32_to_cpu(spx->totlen)), NULL); + raw->next_in_ino = xd->node->next_in_ino; + xd->node->next_in_ino = raw; + } else { + xd->version = je32_to_cpu(spx->version); + sum_link_node_ref(c, jeb, je32_to_cpu(spx->offset) | REF_UNCHECKED, + PAD(je32_to_cpu(spx->totlen)), (void *)xd); } - - xd->node = sum_link_node_ref(c, jeb, je32_to_cpu(spx->offset) | REF_UNCHECKED, - PAD(je32_to_cpu(spx->totlen)), NULL); - /* FIXME */ xd->node->next_in_ino = (void *)xd; - *pseudo_random += je32_to_cpu(spx->xid); sp += JFFS2_SUMMARY_XATTR_SIZE; @@ -519,14 +516,11 @@ static int jffs2_sum_process_sum_data(struct jffs2_sb_info *c, struct jffs2_eras JFFS2_NOTICE("allocation of xattr_datum failed\n"); return -ENOMEM; } - ref->ino = 0xfffffffe; - ref->xid = 0xfffffffd; ref->next = c->xref_temp; c->xref_temp = ref; - ref->node = sum_link_node_ref(c, jeb, je32_to_cpu(spr->offset) | REF_UNCHECKED, - PAD(sizeof(struct jffs2_raw_xref)), NULL); - /* FIXME */ ref->node->next_in_ino = (void *)ref; + sum_link_node_ref(c, jeb, je32_to_cpu(spr->offset) | REF_UNCHECKED, + PAD(sizeof(struct jffs2_raw_xref)), (void *)ref); *pseudo_random += ref->node->flash_offset; sp += JFFS2_SUMMARY_XREF_SIZE; diff --git a/fs/jffs2/super.c b/fs/jffs2/super.c index 2378a662c25..6de374513c0 100644 --- a/fs/jffs2/super.c +++ b/fs/jffs2/super.c @@ -11,7 +11,6 @@ * */ -#include <linux/config.h> #include <linux/kernel.h> #include <linux/module.h> #include <linux/slab.h> @@ -120,10 +119,9 @@ static int jffs2_get_sb_mtd(struct file_system_type *fs_type, struct jffs2_sb_info *c; int ret; - c = kmalloc(sizeof(*c), GFP_KERNEL); + c = kzalloc(sizeof(*c), GFP_KERNEL); if (!c) return -ENOMEM; - memset(c, 0, sizeof(*c)); c->mtd = mtd; sb = sget(fs_type, jffs2_sb_compare, jffs2_sb_set, c); diff --git a/fs/jffs2/wbuf.c b/fs/jffs2/wbuf.c index a7f153f79ec..b9b700730df 100644 --- a/fs/jffs2/wbuf.c +++ b/fs/jffs2/wbuf.c @@ -495,8 +495,7 @@ static void jffs2_wbuf_recover(struct jffs2_sb_info *c) /* Fix up the original jeb now it's on the bad_list */ if (first_raw == jeb->first_node) { D1(printk(KERN_DEBUG "Failing block at %08x is now empty. Moving to erase_pending_list\n", jeb->offset)); - list_del(&jeb->list); - list_add(&jeb->list, &c->erase_pending_list); + list_move(&jeb->list, &c->erase_pending_list); c->nr_erasing_blocks++; jffs2_erase_pending_trigger(c); } diff --git a/fs/jffs2/xattr.c b/fs/jffs2/xattr.c index 2d82e250be3..4da09ce1d1f 100644 --- a/fs/jffs2/xattr.c +++ b/fs/jffs2/xattr.c @@ -23,18 +23,15 @@ * xattr_datum_hashkey(xprefix, xname, xvalue, xsize) * is used to calcurate xdatum hashkey. The reminder of hashkey into XATTRINDEX_HASHSIZE is * the index of the xattr name/value pair cache (c->xattrindex). + * is_xattr_datum_unchecked(c, xd) + * returns 1, if xdatum contains any unchecked raw nodes. if all raw nodes are not + * unchecked, it returns 0. * unload_xattr_datum(c, xd) * is used to release xattr name/value pair and detach from c->xattrindex. * reclaim_xattr_datum(c) * is used to reclaim xattr name/value pairs on the xattr name/value pair cache when * memory usage by cache is over c->xdatum_mem_threshold. Currentry, this threshold * is hard coded as 32KiB. - * delete_xattr_datum_node(c, xd) - * is used to delete a jffs2 node is dominated by xdatum. When EBS(Erase Block Summary) is - * enabled, it overwrites the obsolete node by myself. - * delete_xattr_datum(c, xd) - * is used to delete jffs2_xattr_datum object. It must be called with 0-value of reference - * counter. (It means how many jffs2_xattr_ref object refers this xdatum.) * do_verify_xattr_datum(c, xd) * is used to load the xdatum informations without name/value pair from the medium. * It's necessary once, because those informations are not collected during mounting @@ -53,8 +50,11 @@ * is used to write xdatum to medium. xd->version will be incremented. * create_xattr_datum(c, xprefix, xname, xvalue, xsize) * is used to create new xdatum and write to medium. + * unrefer_xattr_datum(c, xd) + * is used to delete a xdatum. When nobody refers this xdatum, JFFS2_XFLAGS_DEAD + * is set on xd->flags and chained xattr_dead_list or release it immediately. + * In the first case, the garbage collector release it later. * -------------------------------------------------- */ - static uint32_t xattr_datum_hashkey(int xprefix, const char *xname, const char *xvalue, int xsize) { int name_len = strlen(xname); @@ -62,6 +62,22 @@ static uint32_t xattr_datum_hashkey(int xprefix, const char *xname, const char * return crc32(xprefix, xname, name_len) ^ crc32(xprefix, xvalue, xsize); } +static int is_xattr_datum_unchecked(struct jffs2_sb_info *c, struct jffs2_xattr_datum *xd) +{ + struct jffs2_raw_node_ref *raw; + int rc = 0; + + spin_lock(&c->erase_completion_lock); + for (raw=xd->node; raw != (void *)xd; raw=raw->next_in_ino) { + if (ref_flags(raw) == REF_UNCHECKED) { + rc = 1; + break; + } + } + spin_unlock(&c->erase_completion_lock); + return rc; +} + static void unload_xattr_datum(struct jffs2_sb_info *c, struct jffs2_xattr_datum *xd) { /* must be called under down_write(xattr_sem) */ @@ -107,77 +123,33 @@ static void reclaim_xattr_datum(struct jffs2_sb_info *c) before, c->xdatum_mem_usage, before - c->xdatum_mem_usage); } -static void delete_xattr_datum_node(struct jffs2_sb_info *c, struct jffs2_xattr_datum *xd) -{ - /* must be called under down_write(xattr_sem) */ - struct jffs2_raw_xattr rx; - size_t length; - int rc; - - if (!xd->node) { - JFFS2_WARNING("xdatum (xid=%u) is removed twice.\n", xd->xid); - return; - } - if (jffs2_sum_active()) { - memset(&rx, 0xff, sizeof(struct jffs2_raw_xattr)); - rc = jffs2_flash_read(c, ref_offset(xd->node), - sizeof(struct jffs2_unknown_node), - &length, (char *)&rx); - if (rc || length != sizeof(struct jffs2_unknown_node)) { - JFFS2_ERROR("jffs2_flash_read()=%d, req=%zu, read=%zu at %#08x\n", - rc, sizeof(struct jffs2_unknown_node), - length, ref_offset(xd->node)); - } - rc = jffs2_flash_write(c, ref_offset(xd->node), sizeof(rx), - &length, (char *)&rx); - if (rc || length != sizeof(struct jffs2_raw_xattr)) { - JFFS2_ERROR("jffs2_flash_write()=%d, req=%zu, wrote=%zu ar %#08x\n", - rc, sizeof(rx), length, ref_offset(xd->node)); - } - } - spin_lock(&c->erase_completion_lock); - xd->node->next_in_ino = NULL; - spin_unlock(&c->erase_completion_lock); - jffs2_mark_node_obsolete(c, xd->node); - xd->node = NULL; -} - -static void delete_xattr_datum(struct jffs2_sb_info *c, struct jffs2_xattr_datum *xd) -{ - /* must be called under down_write(xattr_sem) */ - BUG_ON(xd->refcnt); - - unload_xattr_datum(c, xd); - if (xd->node) { - delete_xattr_datum_node(c, xd); - xd->node = NULL; - } - jffs2_free_xattr_datum(xd); -} - static int do_verify_xattr_datum(struct jffs2_sb_info *c, struct jffs2_xattr_datum *xd) { /* must be called under down_write(xattr_sem) */ struct jffs2_eraseblock *jeb; + struct jffs2_raw_node_ref *raw; struct jffs2_raw_xattr rx; size_t readlen; - uint32_t crc, totlen; + uint32_t crc, offset, totlen; int rc; - BUG_ON(!xd->node); - BUG_ON(ref_flags(xd->node) != REF_UNCHECKED); + spin_lock(&c->erase_completion_lock); + offset = ref_offset(xd->node); + if (ref_flags(xd->node) == REF_PRISTINE) + goto complete; + spin_unlock(&c->erase_completion_lock); - rc = jffs2_flash_read(c, ref_offset(xd->node), sizeof(rx), &readlen, (char *)&rx); + rc = jffs2_flash_read(c, offset, sizeof(rx), &readlen, (char *)&rx); if (rc || readlen != sizeof(rx)) { JFFS2_WARNING("jffs2_flash_read()=%d, req=%zu, read=%zu at %#08x\n", - rc, sizeof(rx), readlen, ref_offset(xd->node)); + rc, sizeof(rx), readlen, offset); return rc ? rc : -EIO; } crc = crc32(0, &rx, sizeof(rx) - 4); if (crc != je32_to_cpu(rx.node_crc)) { - if (je32_to_cpu(rx.node_crc) != 0xffffffff) - JFFS2_ERROR("node CRC failed at %#08x, read=%#08x, calc=%#08x\n", - ref_offset(xd->node), je32_to_cpu(rx.hdr_crc), crc); + JFFS2_ERROR("node CRC failed at %#08x, read=%#08x, calc=%#08x\n", + offset, je32_to_cpu(rx.hdr_crc), crc); + xd->flags |= JFFS2_XFLAGS_INVALID; return EIO; } totlen = PAD(sizeof(rx) + rx.name_len + 1 + je16_to_cpu(rx.value_len)); @@ -188,11 +160,12 @@ static int do_verify_xattr_datum(struct jffs2_sb_info *c, struct jffs2_xattr_dat || je32_to_cpu(rx.version) != xd->version) { JFFS2_ERROR("inconsistent xdatum at %#08x, magic=%#04x/%#04x, " "nodetype=%#04x/%#04x, totlen=%u/%u, xid=%u/%u, version=%u/%u\n", - ref_offset(xd->node), je16_to_cpu(rx.magic), JFFS2_MAGIC_BITMASK, + offset, je16_to_cpu(rx.magic), JFFS2_MAGIC_BITMASK, je16_to_cpu(rx.nodetype), JFFS2_NODETYPE_XATTR, je32_to_cpu(rx.totlen), totlen, je32_to_cpu(rx.xid), xd->xid, je32_to_cpu(rx.version), xd->version); + xd->flags |= JFFS2_XFLAGS_INVALID; return EIO; } xd->xprefix = rx.xprefix; @@ -200,14 +173,17 @@ static int do_verify_xattr_datum(struct jffs2_sb_info *c, struct jffs2_xattr_dat xd->value_len = je16_to_cpu(rx.value_len); xd->data_crc = je32_to_cpu(rx.data_crc); - /* This JFFS2_NODETYPE_XATTR node is checked */ - jeb = &c->blocks[ref_offset(xd->node) / c->sector_size]; - totlen = PAD(je32_to_cpu(rx.totlen)); - spin_lock(&c->erase_completion_lock); - c->unchecked_size -= totlen; c->used_size += totlen; - jeb->unchecked_size -= totlen; jeb->used_size += totlen; - xd->node->flash_offset = ref_offset(xd->node) | REF_PRISTINE; + complete: + for (raw=xd->node; raw != (void *)xd; raw=raw->next_in_ino) { + jeb = &c->blocks[ref_offset(raw) / c->sector_size]; + totlen = PAD(ref_totlen(c, jeb, raw)); + if (ref_flags(raw) == REF_UNCHECKED) { + c->unchecked_size -= totlen; c->used_size += totlen; + jeb->unchecked_size -= totlen; jeb->used_size += totlen; + } + raw->flash_offset = ref_offset(raw) | ((xd->node==raw) ? REF_PRISTINE : REF_NORMAL); + } spin_unlock(&c->erase_completion_lock); /* unchecked xdatum is chained with c->xattr_unchecked */ @@ -227,7 +203,6 @@ static int do_load_xattr_datum(struct jffs2_sb_info *c, struct jffs2_xattr_datum uint32_t crc, length; int i, ret, retry = 0; - BUG_ON(!xd->node); BUG_ON(ref_flags(xd->node) != REF_PRISTINE); BUG_ON(!list_empty(&xd->xindex)); retry: @@ -253,6 +228,7 @@ static int do_load_xattr_datum(struct jffs2_sb_info *c, struct jffs2_xattr_datum " at %#08x, read: 0x%08x calculated: 0x%08x\n", ref_offset(xd->node), xd->data_crc, crc); kfree(data); + xd->flags |= JFFS2_XFLAGS_INVALID; return EIO; } @@ -286,16 +262,14 @@ static int load_xattr_datum(struct jffs2_sb_info *c, struct jffs2_xattr_datum *x * rc > 0 : Unrecoverable error, this node should be deleted. */ int rc = 0; - BUG_ON(xd->xname); - if (!xd->node) + + BUG_ON(xd->flags & JFFS2_XFLAGS_DEAD); + if (xd->xname) + return 0; + if (xd->flags & JFFS2_XFLAGS_INVALID) return EIO; - if (unlikely(ref_flags(xd->node) != REF_PRISTINE)) { + if (unlikely(is_xattr_datum_unchecked(c, xd))) rc = do_verify_xattr_datum(c, xd); - if (rc > 0) { - list_del_init(&xd->xindex); - delete_xattr_datum_node(c, xd); - } - } if (!rc) rc = do_load_xattr_datum(c, xd); return rc; @@ -304,7 +278,6 @@ static int load_xattr_datum(struct jffs2_sb_info *c, struct jffs2_xattr_datum *x static int save_xattr_datum(struct jffs2_sb_info *c, struct jffs2_xattr_datum *xd) { /* must be called under down_write(xattr_sem) */ - struct jffs2_raw_node_ref *raw; struct jffs2_raw_xattr rx; struct kvec vecs[2]; size_t length; @@ -312,14 +285,16 @@ static int save_xattr_datum(struct jffs2_sb_info *c, struct jffs2_xattr_datum *x uint32_t phys_ofs = write_ofs(c); BUG_ON(!xd->xname); + BUG_ON(xd->flags & (JFFS2_XFLAGS_DEAD|JFFS2_XFLAGS_INVALID)); vecs[0].iov_base = ℞ - vecs[0].iov_len = PAD(sizeof(rx)); + vecs[0].iov_len = sizeof(rx); vecs[1].iov_base = xd->xname; vecs[1].iov_len = xd->name_len + 1 + xd->value_len; totlen = vecs[0].iov_len + vecs[1].iov_len; /* Setup raw-xattr */ + memset(&rx, 0, sizeof(rx)); rx.magic = cpu_to_je16(JFFS2_MAGIC_BITMASK); rx.nodetype = cpu_to_je16(JFFS2_NODETYPE_XATTR); rx.totlen = cpu_to_je32(PAD(totlen)); @@ -343,14 +318,8 @@ static int save_xattr_datum(struct jffs2_sb_info *c, struct jffs2_xattr_datum *x return rc; } - /* success */ - raw = jffs2_add_physical_node_ref(c, phys_ofs | REF_PRISTINE, PAD(totlen), NULL); - /* FIXME */ raw->next_in_ino = (void *)xd; - - if (xd->node) - delete_xattr_datum_node(c, xd); - xd->node = raw; + jffs2_add_physical_node_ref(c, phys_ofs | REF_PRISTINE, PAD(totlen), (void *)xd); dbg_xattr("success on saving xdatum (xid=%u, version=%u, xprefix=%u, xname='%s')\n", xd->xid, xd->version, xd->xprefix, xd->xname); @@ -377,7 +346,7 @@ static struct jffs2_xattr_datum *create_xattr_datum(struct jffs2_sb_info *c, && xd->value_len==xsize && !strcmp(xd->xname, xname) && !memcmp(xd->xvalue, xvalue, xsize)) { - xd->refcnt++; + atomic_inc(&xd->refcnt); return xd; } } @@ -397,7 +366,7 @@ static struct jffs2_xattr_datum *create_xattr_datum(struct jffs2_sb_info *c, strcpy(data, xname); memcpy(data + name_len + 1, xvalue, xsize); - xd->refcnt = 1; + atomic_set(&xd->refcnt, 1); xd->xid = ++c->highest_xid; xd->flags |= JFFS2_XFLAGS_HOT; xd->xprefix = xprefix; @@ -426,20 +395,38 @@ static struct jffs2_xattr_datum *create_xattr_datum(struct jffs2_sb_info *c, return xd; } +static void unrefer_xattr_datum(struct jffs2_sb_info *c, struct jffs2_xattr_datum *xd) +{ + /* must be called under down_write(xattr_sem) */ + if (atomic_dec_and_lock(&xd->refcnt, &c->erase_completion_lock)) { + uint32_t xid = xd->xid, version = xd->version; + + unload_xattr_datum(c, xd); + xd->flags |= JFFS2_XFLAGS_DEAD; + if (xd->node == (void *)xd) { + BUG_ON(!(xd->flags & JFFS2_XFLAGS_INVALID)); + jffs2_free_xattr_datum(xd); + } else { + list_add(&xd->xindex, &c->xattr_dead_list); + } + spin_unlock(&c->erase_completion_lock); + + dbg_xattr("xdatum(xid=%u, version=%u) was removed.\n", xid, version); + } +} + /* -------- xref related functions ------------------ * verify_xattr_ref(c, ref) * is used to load xref information from medium. Because summary data does not * contain xid/ino, it's necessary to verify once while mounting process. - * delete_xattr_ref_node(c, ref) - * is used to delete a jffs2 node is dominated by xref. When EBS is enabled, - * it overwrites the obsolete node by myself. - * delete_xattr_ref(c, ref) - * is used to delete jffs2_xattr_ref object. If the reference counter of xdatum - * is refered by this xref become 0, delete_xattr_datum() is called later. * save_xattr_ref(c, ref) - * is used to write xref to medium. + * is used to write xref to medium. If delete marker is marked, it write + * a delete marker of xref into medium. * create_xattr_ref(c, ic, xd) * is used to create a new xref and write to medium. + * delete_xattr_ref(c, ref) + * is used to delete jffs2_xattr_ref. It marks xref XREF_DELETE_MARKER, + * and allows GC to reclaim those physical nodes. * jffs2_xattr_delete_inode(c, ic) * is called to remove xrefs related to obsolete inode when inode is unlinked. * jffs2_xattr_free_inode(c, ic) @@ -450,25 +437,29 @@ static struct jffs2_xattr_datum *create_xattr_datum(struct jffs2_sb_info *c, static int verify_xattr_ref(struct jffs2_sb_info *c, struct jffs2_xattr_ref *ref) { struct jffs2_eraseblock *jeb; + struct jffs2_raw_node_ref *raw; struct jffs2_raw_xref rr; size_t readlen; - uint32_t crc, totlen; + uint32_t crc, offset, totlen; int rc; - BUG_ON(ref_flags(ref->node) != REF_UNCHECKED); + spin_lock(&c->erase_completion_lock); + if (ref_flags(ref->node) != REF_UNCHECKED) + goto complete; + offset = ref_offset(ref->node); + spin_unlock(&c->erase_completion_lock); - rc = jffs2_flash_read(c, ref_offset(ref->node), sizeof(rr), &readlen, (char *)&rr); + rc = jffs2_flash_read(c, offset, sizeof(rr), &readlen, (char *)&rr); if (rc || sizeof(rr) != readlen) { JFFS2_WARNING("jffs2_flash_read()=%d, req=%zu, read=%zu, at %#08x\n", - rc, sizeof(rr), readlen, ref_offset(ref->node)); + rc, sizeof(rr), readlen, offset); return rc ? rc : -EIO; } /* obsolete node */ crc = crc32(0, &rr, sizeof(rr) - 4); if (crc != je32_to_cpu(rr.node_crc)) { - if (je32_to_cpu(rr.node_crc) != 0xffffffff) - JFFS2_ERROR("node CRC failed at %#08x, read=%#08x, calc=%#08x\n", - ref_offset(ref->node), je32_to_cpu(rr.node_crc), crc); + JFFS2_ERROR("node CRC failed at %#08x, read=%#08x, calc=%#08x\n", + offset, je32_to_cpu(rr.node_crc), crc); return EIO; } if (je16_to_cpu(rr.magic) != JFFS2_MAGIC_BITMASK @@ -476,22 +467,28 @@ static int verify_xattr_ref(struct jffs2_sb_info *c, struct jffs2_xattr_ref *ref || je32_to_cpu(rr.totlen) != PAD(sizeof(rr))) { JFFS2_ERROR("inconsistent xref at %#08x, magic=%#04x/%#04x, " "nodetype=%#04x/%#04x, totlen=%u/%zu\n", - ref_offset(ref->node), je16_to_cpu(rr.magic), JFFS2_MAGIC_BITMASK, + offset, je16_to_cpu(rr.magic), JFFS2_MAGIC_BITMASK, je16_to_cpu(rr.nodetype), JFFS2_NODETYPE_XREF, je32_to_cpu(rr.totlen), PAD(sizeof(rr))); return EIO; } ref->ino = je32_to_cpu(rr.ino); ref->xid = je32_to_cpu(rr.xid); - - /* fixup superblock/eraseblock info */ - jeb = &c->blocks[ref_offset(ref->node) / c->sector_size]; - totlen = PAD(sizeof(rr)); + ref->xseqno = je32_to_cpu(rr.xseqno); + if (ref->xseqno > c->highest_xseqno) + c->highest_xseqno = (ref->xseqno & ~XREF_DELETE_MARKER); spin_lock(&c->erase_completion_lock); - c->unchecked_size -= totlen; c->used_size += totlen; - jeb->unchecked_size -= totlen; jeb->used_size += totlen; - ref->node->flash_offset = ref_offset(ref->node) | REF_PRISTINE; + complete: + for (raw=ref->node; raw != (void *)ref; raw=raw->next_in_ino) { + jeb = &c->blocks[ref_offset(raw) / c->sector_size]; + totlen = PAD(ref_totlen(c, jeb, raw)); + if (ref_flags(raw) == REF_UNCHECKED) { + c->unchecked_size -= totlen; c->used_size += totlen; + jeb->unchecked_size -= totlen; jeb->used_size += totlen; + } + raw->flash_offset = ref_offset(raw) | ((ref->node==raw) ? REF_PRISTINE : REF_NORMAL); + } spin_unlock(&c->erase_completion_lock); dbg_xattr("success on verifying xref (ino=%u, xid=%u) at %#08x\n", @@ -499,58 +496,12 @@ static int verify_xattr_ref(struct jffs2_sb_info *c, struct jffs2_xattr_ref *ref return 0; } -static void delete_xattr_ref_node(struct jffs2_sb_info *c, struct jffs2_xattr_ref *ref) -{ - struct jffs2_raw_xref rr; - size_t length; - int rc; - - if (jffs2_sum_active()) { - memset(&rr, 0xff, sizeof(rr)); - rc = jffs2_flash_read(c, ref_offset(ref->node), - sizeof(struct jffs2_unknown_node), - &length, (char *)&rr); - if (rc || length != sizeof(struct jffs2_unknown_node)) { - JFFS2_ERROR("jffs2_flash_read()=%d, req=%zu, read=%zu at %#08x\n", - rc, sizeof(struct jffs2_unknown_node), - length, ref_offset(ref->node)); - } - rc = jffs2_flash_write(c, ref_offset(ref->node), sizeof(rr), - &length, (char *)&rr); - if (rc || length != sizeof(struct jffs2_raw_xref)) { - JFFS2_ERROR("jffs2_flash_write()=%d, req=%zu, wrote=%zu at %#08x\n", - rc, sizeof(rr), length, ref_offset(ref->node)); - } - } - spin_lock(&c->erase_completion_lock); - ref->node->next_in_ino = NULL; - spin_unlock(&c->erase_completion_lock); - jffs2_mark_node_obsolete(c, ref->node); - ref->node = NULL; -} - -static void delete_xattr_ref(struct jffs2_sb_info *c, struct jffs2_xattr_ref *ref) -{ - /* must be called under down_write(xattr_sem) */ - struct jffs2_xattr_datum *xd; - - BUG_ON(!ref->node); - delete_xattr_ref_node(c, ref); - - xd = ref->xd; - xd->refcnt--; - if (!xd->refcnt) - delete_xattr_datum(c, xd); - jffs2_free_xattr_ref(ref); -} - static int save_xattr_ref(struct jffs2_sb_info *c, struct jffs2_xattr_ref *ref) { /* must be called under down_write(xattr_sem) */ - struct jffs2_raw_node_ref *raw; struct jffs2_raw_xref rr; size_t length; - uint32_t phys_ofs = write_ofs(c); + uint32_t xseqno, phys_ofs = write_ofs(c); int ret; rr.magic = cpu_to_je16(JFFS2_MAGIC_BITMASK); @@ -558,8 +509,16 @@ static int save_xattr_ref(struct jffs2_sb_info *c, struct jffs2_xattr_ref *ref) rr.totlen = cpu_to_je32(PAD(sizeof(rr))); rr.hdr_crc = cpu_to_je32(crc32(0, &rr, sizeof(struct jffs2_unknown_node) - 4)); - rr.ino = cpu_to_je32(ref->ic->ino); - rr.xid = cpu_to_je32(ref->xd->xid); + xseqno = (c->highest_xseqno += 2); + if (is_xattr_ref_dead(ref)) { + xseqno |= XREF_DELETE_MARKER; + rr.ino = cpu_to_je32(ref->ino); + rr.xid = cpu_to_je32(ref->xid); + } else { + rr.ino = cpu_to_je32(ref->ic->ino); + rr.xid = cpu_to_je32(ref->xd->xid); + } + rr.xseqno = cpu_to_je32(xseqno); rr.node_crc = cpu_to_je32(crc32(0, &rr, sizeof(rr) - 4)); ret = jffs2_flash_write(c, phys_ofs, sizeof(rr), &length, (char *)&rr); @@ -572,12 +531,9 @@ static int save_xattr_ref(struct jffs2_sb_info *c, struct jffs2_xattr_ref *ref) return ret; } - - raw = jffs2_add_physical_node_ref(c, phys_ofs | REF_PRISTINE, PAD(sizeof(rr)), NULL); - /* FIXME */ raw->next_in_ino = (void *)ref; - if (ref->node) - delete_xattr_ref_node(c, ref); - ref->node = raw; + /* success */ + ref->xseqno = xseqno; + jffs2_add_physical_node_ref(c, phys_ofs | REF_PRISTINE, PAD(sizeof(rr)), (void *)ref); dbg_xattr("success on saving xref (ino=%u, xid=%u)\n", ref->ic->ino, ref->xd->xid); @@ -610,6 +566,26 @@ static struct jffs2_xattr_ref *create_xattr_ref(struct jffs2_sb_info *c, struct return ref; /* success */ } +static void delete_xattr_ref(struct jffs2_sb_info *c, struct jffs2_xattr_ref *ref) +{ + /* must be called under down_write(xattr_sem) */ + struct jffs2_xattr_datum *xd; + + xd = ref->xd; + ref->xseqno |= XREF_DELETE_MARKER; + ref->ino = ref->ic->ino; + ref->xid = ref->xd->xid; + spin_lock(&c->erase_completion_lock); + ref->next = c->xref_dead_list; + c->xref_dead_list = ref; + spin_unlock(&c->erase_completion_lock); + + dbg_xattr("xref(ino=%u, xid=%u, xseqno=%u) was removed.\n", + ref->ino, ref->xid, ref->xseqno); + + unrefer_xattr_datum(c, xd); +} + void jffs2_xattr_delete_inode(struct jffs2_sb_info *c, struct jffs2_inode_cache *ic) { /* It's called from jffs2_clear_inode() on inode removing. @@ -638,8 +614,7 @@ void jffs2_xattr_free_inode(struct jffs2_sb_info *c, struct jffs2_inode_cache *i for (ref = ic->xref; ref; ref = _ref) { _ref = ref->next; xd = ref->xd; - xd->refcnt--; - if (!xd->refcnt) { + if (atomic_dec_and_test(&xd->refcnt)) { unload_xattr_datum(c, xd); jffs2_free_xattr_datum(xd); } @@ -655,7 +630,7 @@ static int check_xattr_ref_inode(struct jffs2_sb_info *c, struct jffs2_inode_cac * duplicate name/value pairs. If duplicate name/value pair would be found, * one will be removed. */ - struct jffs2_xattr_ref *ref, *cmp, **pref; + struct jffs2_xattr_ref *ref, *cmp, **pref, **pcmp; int rc = 0; if (likely(ic->flags & INO_FLAGS_XATTR_CHECKED)) @@ -673,13 +648,13 @@ static int check_xattr_ref_inode(struct jffs2_sb_info *c, struct jffs2_inode_cac } else if (unlikely(rc < 0)) goto out; } - for (cmp=ref->next, pref=&ref->next; cmp; pref=&cmp->next, cmp=cmp->next) { + for (cmp=ref->next, pcmp=&ref->next; cmp; pcmp=&cmp->next, cmp=cmp->next) { if (!cmp->xd->xname) { ref->xd->flags |= JFFS2_XFLAGS_BIND; rc = load_xattr_datum(c, cmp->xd); ref->xd->flags &= ~JFFS2_XFLAGS_BIND; if (unlikely(rc > 0)) { - *pref = cmp->next; + *pcmp = cmp->next; delete_xattr_ref(c, cmp); goto retry; } else if (unlikely(rc < 0)) @@ -687,8 +662,13 @@ static int check_xattr_ref_inode(struct jffs2_sb_info *c, struct jffs2_inode_cac } if (ref->xd->xprefix == cmp->xd->xprefix && !strcmp(ref->xd->xname, cmp->xd->xname)) { - *pref = cmp->next; - delete_xattr_ref(c, cmp); + if (ref->xseqno > cmp->xseqno) { + *pcmp = cmp->next; + delete_xattr_ref(c, cmp); + } else { + *pref = ref->next; + delete_xattr_ref(c, ref); + } goto retry; } } @@ -719,9 +699,13 @@ void jffs2_init_xattr_subsystem(struct jffs2_sb_info *c) for (i=0; i < XATTRINDEX_HASHSIZE; i++) INIT_LIST_HEAD(&c->xattrindex[i]); INIT_LIST_HEAD(&c->xattr_unchecked); + INIT_LIST_HEAD(&c->xattr_dead_list); + c->xref_dead_list = NULL; c->xref_temp = NULL; init_rwsem(&c->xattr_sem); + c->highest_xid = 0; + c->highest_xseqno = 0; c->xdatum_mem_usage = 0; c->xdatum_mem_threshold = 32 * 1024; /* Default 32KB */ } @@ -751,7 +735,11 @@ void jffs2_clear_xattr_subsystem(struct jffs2_sb_info *c) _ref = ref->next; jffs2_free_xattr_ref(ref); } - c->xref_temp = NULL; + + for (ref=c->xref_dead_list; ref; ref = _ref) { + _ref = ref->next; + jffs2_free_xattr_ref(ref); + } for (i=0; i < XATTRINDEX_HASHSIZE; i++) { list_for_each_entry_safe(xd, _xd, &c->xattrindex[i], xindex) { @@ -761,100 +749,143 @@ void jffs2_clear_xattr_subsystem(struct jffs2_sb_info *c) jffs2_free_xattr_datum(xd); } } + + list_for_each_entry_safe(xd, _xd, &c->xattr_dead_list, xindex) { + list_del(&xd->xindex); + jffs2_free_xattr_datum(xd); + } } +#define XREF_TMPHASH_SIZE (128) void jffs2_build_xattr_subsystem(struct jffs2_sb_info *c) { struct jffs2_xattr_ref *ref, *_ref; + struct jffs2_xattr_ref *xref_tmphash[XREF_TMPHASH_SIZE]; struct jffs2_xattr_datum *xd, *_xd; struct jffs2_inode_cache *ic; - int i, xdatum_count =0, xdatum_unchecked_count = 0, xref_count = 0; + struct jffs2_raw_node_ref *raw; + int i, xdatum_count = 0, xdatum_unchecked_count = 0, xref_count = 0; + int xdatum_orphan_count = 0, xref_orphan_count = 0, xref_dead_count = 0; BUG_ON(!(c->flags & JFFS2_SB_FLAG_BUILDING)); - /* Phase.1 */ + /* Phase.1 : Merge same xref */ + for (i=0; i < XREF_TMPHASH_SIZE; i++) + xref_tmphash[i] = NULL; for (ref=c->xref_temp; ref; ref=_ref) { + struct jffs2_xattr_ref *tmp; + _ref = ref->next; - /* checking REF_UNCHECKED nodes */ if (ref_flags(ref->node) != REF_PRISTINE) { if (verify_xattr_ref(c, ref)) { - delete_xattr_ref_node(c, ref); + BUG_ON(ref->node->next_in_ino != (void *)ref); + ref->node->next_in_ino = NULL; + jffs2_mark_node_obsolete(c, ref->node); jffs2_free_xattr_ref(ref); continue; } } - /* At this point, ref->xid and ref->ino contain XID and inode number. - ref->xd and ref->ic are not valid yet. */ - xd = jffs2_find_xattr_datum(c, ref->xid); - ic = jffs2_get_ino_cache(c, ref->ino); - if (!xd || !ic) { - if (ref_flags(ref->node) != REF_UNCHECKED) - JFFS2_WARNING("xref(ino=%u, xid=%u) is orphan. \n", - ref->ino, ref->xid); - delete_xattr_ref_node(c, ref); + + i = (ref->ino ^ ref->xid) % XREF_TMPHASH_SIZE; + for (tmp=xref_tmphash[i]; tmp; tmp=tmp->next) { + if (tmp->ino == ref->ino && tmp->xid == ref->xid) + break; + } + if (tmp) { + raw = ref->node; + if (ref->xseqno > tmp->xseqno) { + tmp->xseqno = ref->xseqno; + raw->next_in_ino = tmp->node; + tmp->node = raw; + } else { + raw->next_in_ino = tmp->node->next_in_ino; + tmp->node->next_in_ino = raw; + } jffs2_free_xattr_ref(ref); continue; + } else { + ref->next = xref_tmphash[i]; + xref_tmphash[i] = ref; } - ref->xd = xd; - ref->ic = ic; - xd->refcnt++; - ref->next = ic->xref; - ic->xref = ref; - xref_count++; } c->xref_temp = NULL; - /* After this, ref->xid/ino are NEVER used. */ - /* Phase.2 */ + /* Phase.2 : Bind xref with inode_cache and xattr_datum */ + for (i=0; i < XREF_TMPHASH_SIZE; i++) { + for (ref=xref_tmphash[i]; ref; ref=_ref) { + xref_count++; + _ref = ref->next; + if (is_xattr_ref_dead(ref)) { + ref->next = c->xref_dead_list; + c->xref_dead_list = ref; + xref_dead_count++; + continue; + } + /* At this point, ref->xid and ref->ino contain XID and inode number. + ref->xd and ref->ic are not valid yet. */ + xd = jffs2_find_xattr_datum(c, ref->xid); + ic = jffs2_get_ino_cache(c, ref->ino); + if (!xd || !ic) { + dbg_xattr("xref(ino=%u, xid=%u, xseqno=%u) is orphan.\n", + ref->ino, ref->xid, ref->xseqno); + ref->xseqno |= XREF_DELETE_MARKER; + ref->next = c->xref_dead_list; + c->xref_dead_list = ref; + xref_orphan_count++; + continue; + } + ref->xd = xd; + ref->ic = ic; + atomic_inc(&xd->refcnt); + ref->next = ic->xref; + ic->xref = ref; + } + } + + /* Phase.3 : Link unchecked xdatum to xattr_unchecked list */ for (i=0; i < XATTRINDEX_HASHSIZE; i++) { list_for_each_entry_safe(xd, _xd, &c->xattrindex[i], xindex) { + xdatum_count++; list_del_init(&xd->xindex); - if (!xd->refcnt) { - if (ref_flags(xd->node) != REF_UNCHECKED) - JFFS2_WARNING("orphan xdatum(xid=%u, version=%u) at %#08x\n", - xd->xid, xd->version, ref_offset(xd->node)); - delete_xattr_datum(c, xd); + if (!atomic_read(&xd->refcnt)) { + dbg_xattr("xdatum(xid=%u, version=%u) is orphan.\n", + xd->xid, xd->version); + xd->flags |= JFFS2_XFLAGS_DEAD; + list_add(&xd->xindex, &c->xattr_unchecked); + xdatum_orphan_count++; continue; } - if (ref_flags(xd->node) != REF_PRISTINE) { - dbg_xattr("unchecked xdatum(xid=%u) at %#08x\n", - xd->xid, ref_offset(xd->node)); + if (is_xattr_datum_unchecked(c, xd)) { + dbg_xattr("unchecked xdatum(xid=%u, version=%u)\n", + xd->xid, xd->version); list_add(&xd->xindex, &c->xattr_unchecked); xdatum_unchecked_count++; } - xdatum_count++; } } /* build complete */ - JFFS2_NOTICE("complete building xattr subsystem, %u of xdatum (%u unchecked) and " - "%u of xref found.\n", xdatum_count, xdatum_unchecked_count, xref_count); + JFFS2_NOTICE("complete building xattr subsystem, %u of xdatum" + " (%u unchecked, %u orphan) and " + "%u of xref (%u dead, %u orphan) found.\n", + xdatum_count, xdatum_unchecked_count, xdatum_orphan_count, + xref_count, xref_dead_count, xref_orphan_count); } struct jffs2_xattr_datum *jffs2_setup_xattr_datum(struct jffs2_sb_info *c, uint32_t xid, uint32_t version) { - struct jffs2_xattr_datum *xd, *_xd; + struct jffs2_xattr_datum *xd; - _xd = jffs2_find_xattr_datum(c, xid); - if (_xd) { - dbg_xattr("duplicate xdatum (xid=%u, version=%u/%u) at %#08x\n", - xid, version, _xd->version, ref_offset(_xd->node)); - if (version < _xd->version) - return ERR_PTR(-EEXIST); - } - xd = jffs2_alloc_xattr_datum(); - if (!xd) - return ERR_PTR(-ENOMEM); - xd->xid = xid; - xd->version = version; - if (xd->xid > c->highest_xid) - c->highest_xid = xd->xid; - list_add_tail(&xd->xindex, &c->xattrindex[xid % XATTRINDEX_HASHSIZE]); - - if (_xd) { - list_del_init(&_xd->xindex); - delete_xattr_datum_node(c, _xd); - jffs2_free_xattr_datum(_xd); + xd = jffs2_find_xattr_datum(c, xid); + if (!xd) { + xd = jffs2_alloc_xattr_datum(); + if (!xd) + return ERR_PTR(-ENOMEM); + xd->xid = xid; + xd->version = version; + if (xd->xid > c->highest_xid) + c->highest_xid = xd->xid; + list_add_tail(&xd->xindex, &c->xattrindex[xid % XATTRINDEX_HASHSIZE]); } return xd; } @@ -1080,9 +1111,22 @@ int do_jffs2_setxattr(struct inode *inode, int xprefix, const char *xname, goto out; } if (!buffer) { - *pref = ref->next; - delete_xattr_ref(c, ref); - rc = 0; + ref->ino = ic->ino; + ref->xid = xd->xid; + ref->xseqno |= XREF_DELETE_MARKER; + rc = save_xattr_ref(c, ref); + if (!rc) { + *pref = ref->next; + spin_lock(&c->erase_completion_lock); + ref->next = c->xref_dead_list; + c->xref_dead_list = ref; + spin_unlock(&c->erase_completion_lock); + unrefer_xattr_datum(c, xd); + } else { + ref->ic = ic; + ref->xd = xd; + ref->xseqno &= ~XREF_DELETE_MARKER; + } goto out; } goto found; @@ -1094,7 +1138,7 @@ int do_jffs2_setxattr(struct inode *inode, int xprefix, const char *xname, goto out; } if (!buffer) { - rc = -EINVAL; + rc = -ENODATA; goto out; } found: @@ -1110,16 +1154,13 @@ int do_jffs2_setxattr(struct inode *inode, int xprefix, const char *xname, request = PAD(sizeof(struct jffs2_raw_xref)); rc = jffs2_reserve_space(c, request, &length, ALLOC_NORMAL, JFFS2_SUMMARY_XREF_SIZE); + down_write(&c->xattr_sem); if (rc) { JFFS2_WARNING("jffs2_reserve_space()=%d, request=%u\n", rc, request); - down_write(&c->xattr_sem); - xd->refcnt--; - if (!xd->refcnt) - delete_xattr_datum(c, xd); + unrefer_xattr_datum(c, xd); up_write(&c->xattr_sem); return rc; } - down_write(&c->xattr_sem); if (ref) *pref = ref->next; newref = create_xattr_ref(c, ic, xd); @@ -1129,9 +1170,7 @@ int do_jffs2_setxattr(struct inode *inode, int xprefix, const char *xname, ic->xref = ref; } rc = PTR_ERR(newref); - xd->refcnt--; - if (!xd->refcnt) - delete_xattr_datum(c, xd); + unrefer_xattr_datum(c, xd); } else if (ref) { delete_xattr_ref(c, ref); } @@ -1142,39 +1181,40 @@ int do_jffs2_setxattr(struct inode *inode, int xprefix, const char *xname, } /* -------- garbage collector functions ------------- - * jffs2_garbage_collect_xattr_datum(c, xd) + * jffs2_garbage_collect_xattr_datum(c, xd, raw) * is used to move xdatum into new node. - * jffs2_garbage_collect_xattr_ref(c, ref) + * jffs2_garbage_collect_xattr_ref(c, ref, raw) * is used to move xref into new node. * jffs2_verify_xattr(c) * is used to call do_verify_xattr_datum() before garbage collecting. + * jffs2_release_xattr_datum(c, xd) + * is used to release an in-memory object of xdatum. + * jffs2_release_xattr_ref(c, ref) + * is used to release an in-memory object of xref. * -------------------------------------------------- */ -int jffs2_garbage_collect_xattr_datum(struct jffs2_sb_info *c, struct jffs2_xattr_datum *xd) +int jffs2_garbage_collect_xattr_datum(struct jffs2_sb_info *c, struct jffs2_xattr_datum *xd, + struct jffs2_raw_node_ref *raw) { uint32_t totlen, length, old_ofs; - int rc = -EINVAL; + int rc = 0; down_write(&c->xattr_sem); - BUG_ON(!xd->node); - - old_ofs = ref_offset(xd->node); - totlen = ref_totlen(c, c->gcblock, xd->node); - if (totlen < sizeof(struct jffs2_raw_xattr)) + if (xd->node != raw) + goto out; + if (xd->flags & (JFFS2_XFLAGS_DEAD|JFFS2_XFLAGS_INVALID)) goto out; - if (!xd->xname) { - rc = load_xattr_datum(c, xd); - if (unlikely(rc > 0)) { - delete_xattr_datum_node(c, xd); - rc = 0; - goto out; - } else if (unlikely(rc < 0)) - goto out; + rc = load_xattr_datum(c, xd); + if (unlikely(rc)) { + rc = (rc > 0) ? 0 : rc; + goto out; } + old_ofs = ref_offset(xd->node); + totlen = PAD(sizeof(struct jffs2_raw_xattr) + + xd->name_len + 1 + xd->value_len); rc = jffs2_reserve_space_gc(c, totlen, &length, JFFS2_SUMMARY_XATTR_SIZE); - if (rc || length < totlen) { - JFFS2_WARNING("jffs2_reserve_space()=%d, request=%u\n", rc, totlen); - rc = rc ? rc : -EBADFD; + if (rc) { + JFFS2_WARNING("jffs2_reserve_space_gc()=%d, request=%u\n", rc, totlen); goto out; } rc = save_xattr_datum(c, xd); @@ -1182,27 +1222,32 @@ int jffs2_garbage_collect_xattr_datum(struct jffs2_sb_info *c, struct jffs2_xatt dbg_xattr("xdatum (xid=%u, version=%u) GC'ed from %#08x to %08x\n", xd->xid, xd->version, old_ofs, ref_offset(xd->node)); out: + if (!rc) + jffs2_mark_node_obsolete(c, raw); up_write(&c->xattr_sem); return rc; } - -int jffs2_garbage_collect_xattr_ref(struct jffs2_sb_info *c, struct jffs2_xattr_ref *ref) +int jffs2_garbage_collect_xattr_ref(struct jffs2_sb_info *c, struct jffs2_xattr_ref *ref, + struct jffs2_raw_node_ref *raw) { uint32_t totlen, length, old_ofs; - int rc = -EINVAL; + int rc = 0; down_write(&c->xattr_sem); BUG_ON(!ref->node); + if (ref->node != raw) + goto out; + if (is_xattr_ref_dead(ref) && (raw->next_in_ino == (void *)ref)) + goto out; + old_ofs = ref_offset(ref->node); totlen = ref_totlen(c, c->gcblock, ref->node); - if (totlen != sizeof(struct jffs2_raw_xref)) - goto out; rc = jffs2_reserve_space_gc(c, totlen, &length, JFFS2_SUMMARY_XREF_SIZE); - if (rc || length < totlen) { - JFFS2_WARNING("%s: jffs2_reserve_space() = %d, request = %u\n", + if (rc) { + JFFS2_WARNING("%s: jffs2_reserve_space_gc() = %d, request = %u\n", __FUNCTION__, rc, totlen); rc = rc ? rc : -EBADFD; goto out; @@ -1212,6 +1257,8 @@ int jffs2_garbage_collect_xattr_ref(struct jffs2_sb_info *c, struct jffs2_xattr_ dbg_xattr("xref (ino=%u, xid=%u) GC'ed from %#08x to %08x\n", ref->ic->ino, ref->xd->xid, old_ofs, ref_offset(ref->node)); out: + if (!rc) + jffs2_mark_node_obsolete(c, raw); up_write(&c->xattr_sem); return rc; } @@ -1219,20 +1266,59 @@ int jffs2_garbage_collect_xattr_ref(struct jffs2_sb_info *c, struct jffs2_xattr_ int jffs2_verify_xattr(struct jffs2_sb_info *c) { struct jffs2_xattr_datum *xd, *_xd; + struct jffs2_eraseblock *jeb; + struct jffs2_raw_node_ref *raw; + uint32_t totlen; int rc; down_write(&c->xattr_sem); list_for_each_entry_safe(xd, _xd, &c->xattr_unchecked, xindex) { rc = do_verify_xattr_datum(c, xd); - if (rc == 0) { - list_del_init(&xd->xindex); - break; - } else if (rc > 0) { - list_del_init(&xd->xindex); - delete_xattr_datum_node(c, xd); + if (rc < 0) + continue; + list_del_init(&xd->xindex); + spin_lock(&c->erase_completion_lock); + for (raw=xd->node; raw != (void *)xd; raw=raw->next_in_ino) { + if (ref_flags(raw) != REF_UNCHECKED) + continue; + jeb = &c->blocks[ref_offset(raw) / c->sector_size]; + totlen = PAD(ref_totlen(c, jeb, raw)); + c->unchecked_size -= totlen; c->used_size += totlen; + jeb->unchecked_size -= totlen; jeb->used_size += totlen; + raw->flash_offset = ref_offset(raw) + | ((xd->node == (void *)raw) ? REF_PRISTINE : REF_NORMAL); } + if (xd->flags & JFFS2_XFLAGS_DEAD) + list_add(&xd->xindex, &c->xattr_dead_list); + spin_unlock(&c->erase_completion_lock); } up_write(&c->xattr_sem); - return list_empty(&c->xattr_unchecked) ? 1 : 0; } + +void jffs2_release_xattr_datum(struct jffs2_sb_info *c, struct jffs2_xattr_datum *xd) +{ + /* must be called under spin_lock(&c->erase_completion_lock) */ + if (atomic_read(&xd->refcnt) || xd->node != (void *)xd) + return; + + list_del(&xd->xindex); + jffs2_free_xattr_datum(xd); +} + +void jffs2_release_xattr_ref(struct jffs2_sb_info *c, struct jffs2_xattr_ref *ref) +{ + /* must be called under spin_lock(&c->erase_completion_lock) */ + struct jffs2_xattr_ref *tmp, **ptmp; + + if (ref->node != (void *)ref) + return; + + for (tmp=c->xref_dead_list, ptmp=&c->xref_dead_list; tmp; ptmp=&tmp->next, tmp=tmp->next) { + if (ref == tmp) { + *ptmp = tmp->next; + break; + } + } + jffs2_free_xattr_ref(ref); +} diff --git a/fs/jffs2/xattr.h b/fs/jffs2/xattr.h index 2c199856c58..06a5c69dcf8 100644 --- a/fs/jffs2/xattr.h +++ b/fs/jffs2/xattr.h @@ -16,6 +16,8 @@ #define JFFS2_XFLAGS_HOT (0x01) /* This datum is HOT */ #define JFFS2_XFLAGS_BIND (0x02) /* This datum is not reclaimed */ +#define JFFS2_XFLAGS_DEAD (0x40) /* This datum is already dead */ +#define JFFS2_XFLAGS_INVALID (0x80) /* This datum contains crc error */ struct jffs2_xattr_datum { @@ -23,10 +25,10 @@ struct jffs2_xattr_datum struct jffs2_raw_node_ref *node; uint8_t class; uint8_t flags; - uint16_t xprefix; /* see JFFS2_XATTR_PREFIX_* */ + uint16_t xprefix; /* see JFFS2_XATTR_PREFIX_* */ struct list_head xindex; /* chained from c->xattrindex[n] */ - uint32_t refcnt; /* # of xattr_ref refers this */ + atomic_t refcnt; /* # of xattr_ref refers this */ uint32_t xid; uint32_t version; @@ -47,6 +49,7 @@ struct jffs2_xattr_ref uint8_t flags; /* Currently unused */ u16 unused; + uint32_t xseqno; union { struct jffs2_inode_cache *ic; /* reference to jffs2_inode_cache */ uint32_t ino; /* only used in scanning/building */ @@ -58,6 +61,12 @@ struct jffs2_xattr_ref struct jffs2_xattr_ref *next; /* chained from ic->xref_list */ }; +#define XREF_DELETE_MARKER (0x00000001) +static inline int is_xattr_ref_dead(struct jffs2_xattr_ref *ref) +{ + return ((ref->xseqno & XREF_DELETE_MARKER) != 0); +} + #ifdef CONFIG_JFFS2_FS_XATTR extern void jffs2_init_xattr_subsystem(struct jffs2_sb_info *c); @@ -70,9 +79,13 @@ extern struct jffs2_xattr_datum *jffs2_setup_xattr_datum(struct jffs2_sb_info *c extern void jffs2_xattr_delete_inode(struct jffs2_sb_info *c, struct jffs2_inode_cache *ic); extern void jffs2_xattr_free_inode(struct jffs2_sb_info *c, struct jffs2_inode_cache *ic); -extern int jffs2_garbage_collect_xattr_datum(struct jffs2_sb_info *c, struct jffs2_xattr_datum *xd); -extern int jffs2_garbage_collect_xattr_ref(struct jffs2_sb_info *c, struct jffs2_xattr_ref *ref); +extern int jffs2_garbage_collect_xattr_datum(struct jffs2_sb_info *c, struct jffs2_xattr_datum *xd, + struct jffs2_raw_node_ref *raw); +extern int jffs2_garbage_collect_xattr_ref(struct jffs2_sb_info *c, struct jffs2_xattr_ref *ref, + struct jffs2_raw_node_ref *raw); extern int jffs2_verify_xattr(struct jffs2_sb_info *c); +extern void jffs2_release_xattr_datum(struct jffs2_sb_info *c, struct jffs2_xattr_datum *xd); +extern void jffs2_release_xattr_ref(struct jffs2_sb_info *c, struct jffs2_xattr_ref *ref); extern int do_jffs2_getxattr(struct inode *inode, int xprefix, const char *xname, char *buffer, size_t size); diff --git a/fs/jfs/inode.c b/fs/jfs/inode.c index 04eb78f1252..a223cf4faa9 100644 --- a/fs/jfs/inode.c +++ b/fs/jfs/inode.c @@ -168,16 +168,15 @@ void jfs_dirty_inode(struct inode *inode) set_cflag(COMMIT_Dirty, inode); } -static int -jfs_get_blocks(struct inode *ip, sector_t lblock, unsigned long max_blocks, - struct buffer_head *bh_result, int create) +int jfs_get_block(struct inode *ip, sector_t lblock, + struct buffer_head *bh_result, int create) { s64 lblock64 = lblock; int rc = 0; xad_t xad; s64 xaddr; int xflag; - s32 xlen = max_blocks; + s32 xlen = bh_result->b_size >> ip->i_blkbits; /* * Take appropriate lock on inode @@ -188,7 +187,7 @@ jfs_get_blocks(struct inode *ip, sector_t lblock, unsigned long max_blocks, IREAD_LOCK(ip); if (((lblock64 << ip->i_sb->s_blocksize_bits) < ip->i_size) && - (!xtLookup(ip, lblock64, max_blocks, &xflag, &xaddr, &xlen, 0)) && + (!xtLookup(ip, lblock64, xlen, &xflag, &xaddr, &xlen, 0)) && xaddr) { if (xflag & XAD_NOTRECORDED) { if (!create) @@ -255,13 +254,6 @@ jfs_get_blocks(struct inode *ip, sector_t lblock, unsigned long max_blocks, return rc; } -static int jfs_get_block(struct inode *ip, sector_t lblock, - struct buffer_head *bh_result, int create) -{ - return jfs_get_blocks(ip, lblock, bh_result->b_size >> ip->i_blkbits, - bh_result, create); -} - static int jfs_writepage(struct page *page, struct writeback_control *wbc) { return nobh_writepage(page, jfs_get_block, wbc); @@ -305,7 +297,7 @@ static ssize_t jfs_direct_IO(int rw, struct kiocb *iocb, offset, nr_segs, jfs_get_block, NULL); } -struct address_space_operations jfs_aops = { +const struct address_space_operations jfs_aops = { .readpage = jfs_readpage, .readpages = jfs_readpages, .writepage = jfs_writepage, diff --git a/fs/jfs/jfs_extent.c b/fs/jfs/jfs_extent.c index 5549378358b..4c74f0944f7 100644 --- a/fs/jfs/jfs_extent.c +++ b/fs/jfs/jfs_extent.c @@ -126,7 +126,7 @@ extAlloc(struct inode *ip, s64 xlen, s64 pno, xad_t * xp, boolean_t abnr) /* allocate the disk blocks for the extent. initially, extBalloc() * will try to allocate disk blocks for the requested size (xlen). - * if this fails (xlen contigious free blocks not avaliable), it'll + * if this fails (xlen contiguous free blocks not avaliable), it'll * try to allocate a smaller number of blocks (producing a smaller * extent), with this smaller number of blocks consisting of the * requested number of blocks rounded down to the next smaller @@ -468,7 +468,7 @@ int extRecord(struct inode *ip, xad_t * xp) int extFill(struct inode *ip, xad_t * xp) { int rc, nbperpage = JFS_SBI(ip->i_sb)->nbperpage; - s64 blkno = offsetXAD(xp) >> ip->i_blksize; + s64 blkno = offsetXAD(xp) >> ip->i_blkbits; // assert(ISSPARSE(ip)); @@ -493,7 +493,7 @@ int extFill(struct inode *ip, xad_t * xp) * * initially, we will try to allocate disk blocks for the * requested size (nblocks). if this fails (nblocks - * contigious free blocks not avaliable), we'll try to allocate + * contiguous free blocks not avaliable), we'll try to allocate * a smaller number of blocks (producing a smaller extent), with * this smaller number of blocks consisting of the requested * number of blocks rounded down to the next smaller power of 2 @@ -529,7 +529,7 @@ extBalloc(struct inode *ip, s64 hint, s64 * nblocks, s64 * blkno) /* get the number of blocks to initially attempt to allocate. * we'll first try the number of blocks requested unless this - * number is greater than the maximum number of contigious free + * number is greater than the maximum number of contiguous free * blocks in the map. in that case, we'll start off with the * maximum free. */ @@ -586,7 +586,7 @@ extBalloc(struct inode *ip, s64 hint, s64 * nblocks, s64 * blkno) * in place. if this fails, we'll try to move the extent * to a new set of blocks. if moving the extent, we initially * will try to allocate disk blocks for the requested size - * (nnew). if this fails (nnew contigious free blocks not + * (nnew). if this fails (new contiguous free blocks not * avaliable), we'll try to allocate a smaller number of * blocks (producing a smaller extent), with this smaller * number of blocks consisting of the requested number of diff --git a/fs/jfs/jfs_imap.c b/fs/jfs/jfs_imap.c index ccbe60aff83..369d7f39c04 100644 --- a/fs/jfs/jfs_imap.c +++ b/fs/jfs/jfs_imap.c @@ -3115,7 +3115,6 @@ static int copy_from_dinode(struct dinode * dip, struct inode *ip) ip->i_mtime.tv_nsec = le32_to_cpu(dip->di_mtime.tv_nsec); ip->i_ctime.tv_sec = le32_to_cpu(dip->di_ctime.tv_sec); ip->i_ctime.tv_nsec = le32_to_cpu(dip->di_ctime.tv_nsec); - ip->i_blksize = ip->i_sb->s_blocksize; ip->i_blocks = LBLK2PBLK(ip->i_sb, le64_to_cpu(dip->di_nblocks)); ip->i_generation = le32_to_cpu(dip->di_gen); diff --git a/fs/jfs/jfs_inode.c b/fs/jfs/jfs_inode.c index 495df402916..bffaca9ae3a 100644 --- a/fs/jfs/jfs_inode.c +++ b/fs/jfs/jfs_inode.c @@ -115,7 +115,6 @@ struct inode *ialloc(struct inode *parent, umode_t mode) } jfs_inode->mode2 |= mode; - inode->i_blksize = sb->s_blocksize; inode->i_blocks = 0; inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME; jfs_inode->otime = inode->i_ctime.tv_sec; diff --git a/fs/jfs/jfs_inode.h b/fs/jfs/jfs_inode.h index c3007267446..1fc48df670c 100644 --- a/fs/jfs/jfs_inode.h +++ b/fs/jfs/jfs_inode.h @@ -32,8 +32,9 @@ extern void jfs_truncate_nolock(struct inode *, loff_t); extern void jfs_free_zero_link(struct inode *); extern struct dentry *jfs_get_parent(struct dentry *dentry); extern void jfs_set_inode_flags(struct inode *); +extern int jfs_get_block(struct inode *, sector_t, struct buffer_head *, int); -extern struct address_space_operations jfs_aops; +extern const struct address_space_operations jfs_aops; extern struct inode_operations jfs_dir_inode_operations; extern const struct file_operations jfs_dir_operations; extern struct inode_operations jfs_file_inode_operations; diff --git a/fs/jfs/jfs_metapage.c b/fs/jfs/jfs_metapage.c index 7f6e8803970..f5afc129d6b 100644 --- a/fs/jfs/jfs_metapage.c +++ b/fs/jfs/jfs_metapage.c @@ -257,7 +257,7 @@ static sector_t metapage_get_blocks(struct inode *inode, sector_t lblock, int rc = 0; int xflag; s64 xaddr; - sector_t file_blocks = (inode->i_size + inode->i_blksize - 1) >> + sector_t file_blocks = (inode->i_size + inode->i_sb->s_blocksize - 1) >> inode->i_blkbits; if (lblock >= file_blocks) @@ -577,7 +577,7 @@ static void metapage_invalidatepage(struct page *page, unsigned long offset) metapage_releasepage(page, 0); } -struct address_space_operations jfs_metapage_aops = { +const struct address_space_operations jfs_metapage_aops = { .readpage = metapage_readpage, .writepage = metapage_writepage, .sync_page = block_sync_page, diff --git a/fs/jfs/jfs_metapage.h b/fs/jfs/jfs_metapage.h index f0b7d3282b0..d17a3290f5a 100644 --- a/fs/jfs/jfs_metapage.h +++ b/fs/jfs/jfs_metapage.h @@ -139,7 +139,7 @@ static inline void metapage_homeok(struct metapage *mp) put_metapage(mp); } -extern struct address_space_operations jfs_metapage_aops; +extern const struct address_space_operations jfs_metapage_aops; /* * This routines invalidate all pages for an extent. diff --git a/fs/jfs/jfs_txnmgr.c b/fs/jfs/jfs_txnmgr.c index ac3d66948e8..3856efc399c 100644 --- a/fs/jfs/jfs_txnmgr.c +++ b/fs/jfs/jfs_txnmgr.c @@ -282,7 +282,7 @@ int txInit(void) TxLockVHWM = (nTxLock * 8) / 10; size = sizeof(struct tblock) * nTxBlock; - TxBlock = (struct tblock *) vmalloc(size); + TxBlock = vmalloc(size); if (TxBlock == NULL) return -ENOMEM; @@ -307,7 +307,7 @@ int txInit(void) * tlock id = 0 is reserved. */ size = sizeof(struct tlock) * nTxLock; - TxLock = (struct tlock *) vmalloc(size); + TxLock = vmalloc(size); if (TxLock == NULL) { vfree(TxBlock); return -ENOMEM; @@ -842,7 +842,7 @@ struct tlock *txLock(tid_t tid, struct inode *ip, struct metapage * mp, TXN_UNLOCK(); release_metapage(mp); TXN_LOCK(); - xtid = tlck->tid; /* reaquire after dropping TXN_LOCK */ + xtid = tlck->tid; /* reacquire after dropping TXN_LOCK */ jfs_info("txLock: in waitLock, tid = %d, xtid = %d, lid = %d", tid, xtid, lid); @@ -2944,7 +2944,7 @@ int jfs_sync(void *arg) * Inode is being freed */ list_del_init(&jfs_ip->anon_inode_list); - } else if (! !mutex_trylock(&jfs_ip->commit_mutex)) { + } else if (mutex_trylock(&jfs_ip->commit_mutex)) { /* * inode will be removed from anonymous list * when it is committed diff --git a/fs/jfs/namei.c b/fs/jfs/namei.c index 09ea03f6227..295268ad231 100644 --- a/fs/jfs/namei.c +++ b/fs/jfs/namei.c @@ -165,8 +165,8 @@ static int jfs_create(struct inode *dip, struct dentry *dentry, int mode, out3: txEnd(tid); - mutex_unlock(&JFS_IP(dip)->commit_mutex); mutex_unlock(&JFS_IP(ip)->commit_mutex); + mutex_unlock(&JFS_IP(dip)->commit_mutex); if (rc) { free_ea_wmap(ip); ip->i_nlink = 0; @@ -300,8 +300,8 @@ static int jfs_mkdir(struct inode *dip, struct dentry *dentry, int mode) out3: txEnd(tid); - mutex_unlock(&JFS_IP(dip)->commit_mutex); mutex_unlock(&JFS_IP(ip)->commit_mutex); + mutex_unlock(&JFS_IP(dip)->commit_mutex); if (rc) { free_ea_wmap(ip); ip->i_nlink = 0; @@ -384,8 +384,8 @@ static int jfs_rmdir(struct inode *dip, struct dentry *dentry) if (rc == -EIO) txAbort(tid, 1); txEnd(tid); - mutex_unlock(&JFS_IP(dip)->commit_mutex); mutex_unlock(&JFS_IP(ip)->commit_mutex); + mutex_unlock(&JFS_IP(dip)->commit_mutex); goto out2; } @@ -422,8 +422,8 @@ static int jfs_rmdir(struct inode *dip, struct dentry *dentry) txEnd(tid); - mutex_unlock(&JFS_IP(dip)->commit_mutex); mutex_unlock(&JFS_IP(ip)->commit_mutex); + mutex_unlock(&JFS_IP(dip)->commit_mutex); /* * Truncating the directory index table is not guaranteed. It @@ -503,8 +503,8 @@ static int jfs_unlink(struct inode *dip, struct dentry *dentry) if (rc == -EIO) txAbort(tid, 1); /* Marks FS Dirty */ txEnd(tid); - mutex_unlock(&JFS_IP(dip)->commit_mutex); mutex_unlock(&JFS_IP(ip)->commit_mutex); + mutex_unlock(&JFS_IP(dip)->commit_mutex); IWRITE_UNLOCK(ip); goto out1; } @@ -527,8 +527,8 @@ static int jfs_unlink(struct inode *dip, struct dentry *dentry) if ((new_size = commitZeroLink(tid, ip)) < 0) { txAbort(tid, 1); /* Marks FS Dirty */ txEnd(tid); - mutex_unlock(&JFS_IP(dip)->commit_mutex); mutex_unlock(&JFS_IP(ip)->commit_mutex); + mutex_unlock(&JFS_IP(dip)->commit_mutex); IWRITE_UNLOCK(ip); rc = new_size; goto out1; @@ -556,9 +556,8 @@ static int jfs_unlink(struct inode *dip, struct dentry *dentry) txEnd(tid); - mutex_unlock(&JFS_IP(dip)->commit_mutex); mutex_unlock(&JFS_IP(ip)->commit_mutex); - + mutex_unlock(&JFS_IP(dip)->commit_mutex); while (new_size && (rc == 0)) { tid = txBegin(dip->i_sb, 0); @@ -847,8 +846,8 @@ static int jfs_link(struct dentry *old_dentry, out: txEnd(tid); - mutex_unlock(&JFS_IP(dir)->commit_mutex); mutex_unlock(&JFS_IP(ip)->commit_mutex); + mutex_unlock(&JFS_IP(dir)->commit_mutex); jfs_info("jfs_link: rc:%d", rc); return rc; @@ -1037,8 +1036,8 @@ static int jfs_symlink(struct inode *dip, struct dentry *dentry, out3: txEnd(tid); - mutex_unlock(&JFS_IP(dip)->commit_mutex); mutex_unlock(&JFS_IP(ip)->commit_mutex); + mutex_unlock(&JFS_IP(dip)->commit_mutex); if (rc) { free_ea_wmap(ip); ip->i_nlink = 0; @@ -1160,10 +1159,11 @@ static int jfs_rename(struct inode *old_dir, struct dentry *old_dentry, if (S_ISDIR(new_ip->i_mode)) { new_ip->i_nlink--; if (new_ip->i_nlink) { - mutex_unlock(&JFS_IP(new_dir)->commit_mutex); - mutex_unlock(&JFS_IP(old_ip)->commit_mutex); + mutex_unlock(&JFS_IP(new_ip)->commit_mutex); if (old_dir != new_dir) mutex_unlock(&JFS_IP(old_dir)->commit_mutex); + mutex_unlock(&JFS_IP(old_ip)->commit_mutex); + mutex_unlock(&JFS_IP(new_dir)->commit_mutex); if (!S_ISDIR(old_ip->i_mode) && new_ip) IWRITE_UNLOCK(new_ip); jfs_error(new_ip->i_sb, @@ -1281,13 +1281,12 @@ static int jfs_rename(struct inode *old_dir, struct dentry *old_dentry, out4: txEnd(tid); - - mutex_unlock(&JFS_IP(new_dir)->commit_mutex); - mutex_unlock(&JFS_IP(old_ip)->commit_mutex); - if (old_dir != new_dir) - mutex_unlock(&JFS_IP(old_dir)->commit_mutex); if (new_ip) mutex_unlock(&JFS_IP(new_ip)->commit_mutex); + if (old_dir != new_dir) + mutex_unlock(&JFS_IP(old_dir)->commit_mutex); + mutex_unlock(&JFS_IP(old_ip)->commit_mutex); + mutex_unlock(&JFS_IP(new_dir)->commit_mutex); while (new_size && (rc == 0)) { tid = txBegin(new_ip->i_sb, 0); diff --git a/fs/jfs/super.c b/fs/jfs/super.c index 73d2aba084c..143bcd1d5ea 100644 --- a/fs/jfs/super.c +++ b/fs/jfs/super.c @@ -18,7 +18,6 @@ */ #include <linux/fs.h> -#include <linux/config.h> #include <linux/module.h> #include <linux/parser.h> #include <linux/completion.h> @@ -27,6 +26,7 @@ #include <linux/moduleparam.h> #include <linux/kthread.h> #include <linux/posix_acl.h> +#include <linux/buffer_head.h> #include <asm/uaccess.h> #include <linux/seq_file.h> @@ -299,7 +299,7 @@ static int parse_options(char *options, struct super_block *sb, s64 *newLVSize, break; } -#if defined(CONFIG_QUOTA) +#ifdef CONFIG_QUOTA case Opt_quota: case Opt_usrquota: *flag |= JFS_USRQUOTA; @@ -598,7 +598,7 @@ static int jfs_show_options(struct seq_file *seq, struct vfsmount *vfs) if (sbi->flag & JFS_NOINTEGRITY) seq_puts(seq, ",nointegrity"); -#if defined(CONFIG_QUOTA) +#ifdef CONFIG_QUOTA if (sbi->flag & JFS_USRQUOTA) seq_puts(seq, ",usrquota"); @@ -609,6 +609,113 @@ static int jfs_show_options(struct seq_file *seq, struct vfsmount *vfs) return 0; } +#ifdef CONFIG_QUOTA + +/* 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) + * we don't have to be afraid of races */ +static ssize_t jfs_quota_read(struct super_block *sb, int type, char *data, + size_t len, loff_t off) +{ + struct inode *inode = sb_dqopt(sb)->files[type]; + sector_t blk = off >> sb->s_blocksize_bits; + int err = 0; + int offset = off & (sb->s_blocksize - 1); + int tocopy; + size_t toread; + struct buffer_head tmp_bh; + struct buffer_head *bh; + loff_t i_size = i_size_read(inode); + + if (off > i_size) + return 0; + if (off+len > i_size) + len = i_size-off; + toread = len; + while (toread > 0) { + tocopy = sb->s_blocksize - offset < toread ? + sb->s_blocksize - offset : toread; + + tmp_bh.b_state = 0; + tmp_bh.b_size = 1 << inode->i_blkbits; + err = jfs_get_block(inode, blk, &tmp_bh, 0); + if (err) + return err; + if (!buffer_mapped(&tmp_bh)) /* A hole? */ + memset(data, 0, tocopy); + else { + bh = sb_bread(sb, tmp_bh.b_blocknr); + if (!bh) + return -EIO; + memcpy(data, bh->b_data+offset, tocopy); + brelse(bh); + } + offset = 0; + toread -= tocopy; + data += tocopy; + blk++; + } + return len; +} + +/* Write to quotafile */ +static ssize_t jfs_quota_write(struct super_block *sb, int type, + const char *data, size_t len, loff_t off) +{ + struct inode *inode = sb_dqopt(sb)->files[type]; + sector_t blk = off >> sb->s_blocksize_bits; + int err = 0; + int offset = off & (sb->s_blocksize - 1); + int tocopy; + size_t towrite = len; + struct buffer_head tmp_bh; + struct buffer_head *bh; + + mutex_lock(&inode->i_mutex); + while (towrite > 0) { + tocopy = sb->s_blocksize - offset < towrite ? + sb->s_blocksize - offset : towrite; + + tmp_bh.b_state = 0; + tmp_bh.b_size = 1 << inode->i_blkbits; + err = jfs_get_block(inode, blk, &tmp_bh, 1); + if (err) + goto out; + if (offset || tocopy != sb->s_blocksize) + bh = sb_bread(sb, tmp_bh.b_blocknr); + else + bh = sb_getblk(sb, tmp_bh.b_blocknr); + if (!bh) { + err = -EIO; + goto out; + } + lock_buffer(bh); + memcpy(bh->b_data+offset, data, tocopy); + flush_dcache_page(bh->b_page); + set_buffer_uptodate(bh); + mark_buffer_dirty(bh); + unlock_buffer(bh); + brelse(bh); + offset = 0; + towrite -= tocopy; + data += tocopy; + blk++; + } +out: + if (len == towrite) + return err; + if (inode->i_size < off+len-towrite) + i_size_write(inode, off+len-towrite); + inode->i_version++; + inode->i_mtime = inode->i_ctime = CURRENT_TIME; + mark_inode_dirty(inode); + mutex_unlock(&inode->i_mutex); + return len - towrite; +} + +#endif + static struct super_operations jfs_super_operations = { .alloc_inode = jfs_alloc_inode, .destroy_inode = jfs_destroy_inode, @@ -622,7 +729,11 @@ static struct super_operations jfs_super_operations = { .unlockfs = jfs_unlockfs, .statfs = jfs_statfs, .remount_fs = jfs_remount, - .show_options = jfs_show_options + .show_options = jfs_show_options, +#ifdef CONFIG_QUOTA + .quota_read = jfs_quota_read, + .quota_write = jfs_quota_write, +#endif }; static struct export_operations jfs_export_operations = { diff --git a/fs/libfs.c b/fs/libfs.c index 1b115638178..3793aaa1457 100644 --- a/fs/libfs.c +++ b/fs/libfs.c @@ -149,10 +149,9 @@ int dcache_readdir(struct file * filp, void * dirent, filldir_t filldir) /* fallthrough */ default: spin_lock(&dcache_lock); - if (filp->f_pos == 2) { - list_del(q); - list_add(q, &dentry->d_subdirs); - } + if (filp->f_pos == 2) + list_move(q, &dentry->d_subdirs); + for (p=q->next; p != &dentry->d_subdirs; p=p->next) { struct dentry *next; next = list_entry(p, struct dentry, d_u.d_child); @@ -164,8 +163,7 @@ int dcache_readdir(struct file * filp, void * dirent, filldir_t filldir) return 0; spin_lock(&dcache_lock); /* next is still alive */ - list_del(q); - list_add(q, p); + list_move(q, p); p = q; filp->f_pos++; } @@ -319,17 +317,9 @@ int simple_rename(struct inode *old_dir, struct dentry *old_dentry, int simple_readpage(struct file *file, struct page *page) { - void *kaddr; - - if (PageUptodate(page)) - goto out; - - kaddr = kmap_atomic(page, KM_USER0); - memset(kaddr, 0, PAGE_CACHE_SIZE); - kunmap_atomic(kaddr, KM_USER0); + clear_highpage(page); flush_dcache_page(page); SetPageUptodate(page); -out: unlock_page(page); return 0; } @@ -385,7 +375,6 @@ int simple_fill_super(struct super_block *s, int magic, struct tree_descr *files return -ENOMEM; inode->i_mode = S_IFDIR | 0755; inode->i_uid = inode->i_gid = 0; - inode->i_blksize = PAGE_CACHE_SIZE; inode->i_blocks = 0; inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME; inode->i_op = &simple_dir_inode_operations; @@ -407,7 +396,6 @@ int simple_fill_super(struct super_block *s, int magic, struct tree_descr *files goto out; inode->i_mode = S_IFREG | files->mode; inode->i_uid = inode->i_gid = 0; - inode->i_blksize = PAGE_CACHE_SIZE; inode->i_blocks = 0; inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME; inode->i_fop = files->ops; @@ -424,13 +412,13 @@ out: static DEFINE_SPINLOCK(pin_fs_lock); -int simple_pin_fs(char *name, struct vfsmount **mount, int *count) +int simple_pin_fs(struct file_system_type *type, struct vfsmount **mount, int *count) { struct vfsmount *mnt = NULL; spin_lock(&pin_fs_lock); if (unlikely(!*mount)) { spin_unlock(&pin_fs_lock); - mnt = do_kern_mount(name, 0, name, NULL); + mnt = vfs_kern_mount(type, 0, type->name, NULL); if (IS_ERR(mnt)) return PTR_ERR(mnt); spin_lock(&pin_fs_lock); @@ -549,7 +537,7 @@ int simple_attr_open(struct inode *inode, struct file *file, attr->get = get; attr->set = set; - attr->data = inode->u.generic_ip; + attr->data = inode->i_private; attr->fmt = fmt; mutex_init(&attr->mutex); diff --git a/fs/lockd/clntlock.c b/fs/lockd/clntlock.c index bce74446870..f95cc3f3c42 100644 --- a/fs/lockd/clntlock.c +++ b/fs/lockd/clntlock.c @@ -147,11 +147,10 @@ u32 nlmclnt_grant(const struct sockaddr_in *addr, const struct nlm_lock *lock) * Someone has sent us an SM_NOTIFY. Ensure we bind to the new port number, * that we mark locks for reclaiming, and that we bump the pseudo NSM state. */ -static inline -void nlmclnt_prepare_reclaim(struct nlm_host *host, u32 newstate) +static void nlmclnt_prepare_reclaim(struct nlm_host *host) { + down_write(&host->h_rwsem); host->h_monitored = 0; - host->h_nsmstate = newstate; host->h_state++; host->h_nextrebind = 0; nlm_rebind_host(host); @@ -161,7 +160,14 @@ void nlmclnt_prepare_reclaim(struct nlm_host *host, u32 newstate) */ list_splice_init(&host->h_granted, &host->h_reclaim); - dprintk("NLM: reclaiming locks for host %s", host->h_name); + dprintk("NLM: reclaiming locks for host %s\n", host->h_name); +} + +static void nlmclnt_finish_reclaim(struct nlm_host *host) +{ + host->h_reclaiming = 0; + up_write(&host->h_rwsem); + dprintk("NLM: done reclaiming locks for host %s", host->h_name); } /* @@ -171,12 +177,10 @@ void nlmclnt_prepare_reclaim(struct nlm_host *host, u32 newstate) void nlmclnt_recovery(struct nlm_host *host, u32 newstate) { - if (host->h_reclaiming++) { - if (host->h_nsmstate == newstate) - return; - nlmclnt_prepare_reclaim(host, newstate); - } else { - nlmclnt_prepare_reclaim(host, newstate); + if (host->h_nsmstate == newstate) + return; + host->h_nsmstate = newstate; + if (!host->h_reclaiming++) { nlm_get_host(host); __module_get(THIS_MODULE); if (kernel_thread(reclaimer, host, CLONE_KERNEL) < 0) @@ -190,6 +194,7 @@ reclaimer(void *ptr) struct nlm_host *host = (struct nlm_host *) ptr; struct nlm_wait *block; struct file_lock *fl, *next; + u32 nsmstate; daemonize("%s-reclaim", host->h_name); allow_signal(SIGKILL); @@ -199,19 +204,25 @@ reclaimer(void *ptr) lock_kernel(); lockd_up(); + nlmclnt_prepare_reclaim(host); /* First, reclaim all locks that have been marked. */ restart: + nsmstate = host->h_nsmstate; list_for_each_entry_safe(fl, next, &host->h_reclaim, fl_u.nfs_fl.list) { list_del_init(&fl->fl_u.nfs_fl.list); if (signalled()) continue; - if (nlmclnt_reclaim(host, fl) == 0) - list_add_tail(&fl->fl_u.nfs_fl.list, &host->h_granted); - goto restart; + if (nlmclnt_reclaim(host, fl) != 0) + continue; + list_add_tail(&fl->fl_u.nfs_fl.list, &host->h_granted); + if (host->h_nsmstate != nsmstate) { + /* Argh! The server rebooted again! */ + list_splice_init(&host->h_granted, &host->h_reclaim); + goto restart; + } } - - host->h_reclaiming = 0; + nlmclnt_finish_reclaim(host); /* Now, wake up all processes that sleep on a blocked lock */ list_for_each_entry(block, &nlm_blocked, b_list) { diff --git a/fs/lockd/clntproc.c b/fs/lockd/clntproc.c index f96e38155b5..271e2165fff 100644 --- a/fs/lockd/clntproc.c +++ b/fs/lockd/clntproc.c @@ -6,7 +6,6 @@ * Copyright (C) 1996, Olaf Kirch <okir@monad.swb.de> */ -#include <linux/config.h> #include <linux/module.h> #include <linux/types.h> #include <linux/errno.h> @@ -101,7 +100,7 @@ static struct nlm_lockowner *nlm_find_lockowner(struct nlm_host *host, fl_owner_ res = __nlm_find_lockowner(host, owner); if (res == NULL) { spin_unlock(&host->h_lock); - new = (struct nlm_lockowner *)kmalloc(sizeof(*new), GFP_KERNEL); + new = kmalloc(sizeof(*new), GFP_KERNEL); spin_lock(&host->h_lock); res = __nlm_find_lockowner(host, owner); if (res == NULL && new != NULL) { @@ -152,11 +151,13 @@ static void nlmclnt_release_lockargs(struct nlm_rqst *req) int nlmclnt_proc(struct inode *inode, int cmd, struct file_lock *fl) { + struct rpc_clnt *client = NFS_CLIENT(inode); + struct sockaddr_in addr; struct nlm_host *host; struct nlm_rqst *call; sigset_t oldset; unsigned long flags; - int status, proto, vers; + int status, vers; vers = (NFS_PROTO(inode)->version == 3) ? 4 : 1; if (NFS_PROTO(inode)->version > 3) { @@ -164,10 +165,8 @@ nlmclnt_proc(struct inode *inode, int cmd, struct file_lock *fl) return -ENOLCK; } - /* Retrieve transport protocol from NFS client */ - proto = NFS_CLIENT(inode)->cl_xprt->prot; - - host = nlmclnt_lookup_host(NFS_ADDR(inode), proto, vers); + rpc_peeraddr(client, (struct sockaddr *) &addr, sizeof(addr)); + host = nlmclnt_lookup_host(&addr, client->cl_xprt->prot, vers); if (host == NULL) return -ENOLCK; @@ -455,7 +454,7 @@ static void nlmclnt_locks_init_private(struct file_lock *fl, struct nlm_host *ho fl->fl_ops = &nlmclnt_lock_ops; } -static void do_vfs_lock(struct file_lock *fl) +static int do_vfs_lock(struct file_lock *fl) { int res = 0; switch (fl->fl_flags & (FL_POSIX|FL_FLOCK)) { @@ -468,9 +467,7 @@ static void do_vfs_lock(struct file_lock *fl) default: BUG(); } - if (res < 0) - printk(KERN_WARNING "%s: VFS is out of sync with lock manager!\n", - __FUNCTION__); + return res; } /* @@ -499,6 +496,7 @@ nlmclnt_lock(struct nlm_rqst *req, struct file_lock *fl) struct nlm_host *host = req->a_host; struct nlm_res *resp = &req->a_res; struct nlm_wait *block = NULL; + unsigned char fl_flags = fl->fl_flags; int status = -ENOLCK; if (!host->h_monitored && nsm_monitor(host) < 0) { @@ -506,9 +504,16 @@ nlmclnt_lock(struct nlm_rqst *req, struct file_lock *fl) host->h_name); goto out; } + fl->fl_flags |= FL_ACCESS; + status = do_vfs_lock(fl); + if (status < 0) + goto out; block = nlmclnt_prepare_block(host, fl); +again: for(;;) { + /* Reboot protection */ + fl->fl_u.nfs_fl.state = host->h_state; status = nlmclnt_call(req, NLMPROC_LOCK); if (status < 0) goto out_unblock; @@ -531,10 +536,17 @@ nlmclnt_lock(struct nlm_rqst *req, struct file_lock *fl) } if (resp->status == NLM_LCK_GRANTED) { - fl->fl_u.nfs_fl.state = host->h_state; - fl->fl_flags |= FL_SLEEP; + down_read(&host->h_rwsem); + /* Check whether or not the server has rebooted */ + if (fl->fl_u.nfs_fl.state != host->h_state) { + up_read(&host->h_rwsem); + goto again; + } /* Ensure the resulting lock will get added to granted list */ - do_vfs_lock(fl); + fl->fl_flags = fl_flags | FL_SLEEP; + if (do_vfs_lock(fl) < 0) + printk(KERN_WARNING "%s: VFS is out of sync with lock manager!\n", __FUNCTION__); + up_read(&host->h_rwsem); } status = nlm_stat_to_errno(resp->status); out_unblock: @@ -544,6 +556,7 @@ out_unblock: nlmclnt_cancel(host, req->a_args.block, fl); out: nlm_release_call(req); + fl->fl_flags = fl_flags; return status; } @@ -596,15 +609,22 @@ nlmclnt_reclaim(struct nlm_host *host, struct file_lock *fl) static int nlmclnt_unlock(struct nlm_rqst *req, struct file_lock *fl) { + struct nlm_host *host = req->a_host; struct nlm_res *resp = &req->a_res; - int status; + int status = 0; /* * Note: the server is supposed to either grant us the unlock * request, or to deny it with NLM_LCK_DENIED_GRACE_PERIOD. In either * case, we want to unlock. */ - do_vfs_lock(fl); + fl->fl_flags |= FL_EXISTS; + down_read(&host->h_rwsem); + if (do_vfs_lock(fl) == -ENOENT) { + up_read(&host->h_rwsem); + goto out; + } + up_read(&host->h_rwsem); if (req->a_flags & RPC_TASK_ASYNC) return nlm_async_call(req, NLMPROC_UNLOCK, &nlmclnt_unlock_ops); @@ -613,7 +633,6 @@ nlmclnt_unlock(struct nlm_rqst *req, struct file_lock *fl) if (status < 0) goto out; - status = 0; if (resp->status == NLM_LCK_GRANTED) goto out; diff --git a/fs/lockd/host.c b/fs/lockd/host.c index 729ac427d35..a0d0b58ce7a 100644 --- a/fs/lockd/host.c +++ b/fs/lockd/host.c @@ -26,7 +26,6 @@ #define NLM_HOST_REBIND (60 * HZ) #define NLM_HOST_EXPIRE ((nrhosts > NLM_HOST_MAX)? 300 * HZ : 120 * HZ) #define NLM_HOST_COLLECT ((nrhosts > NLM_HOST_MAX)? 120 * HZ : 60 * HZ) -#define NLM_HOST_ADDR(sv) (&(sv)->s_nlmclnt->cl_xprt->addr) static struct nlm_host * nlm_hosts[NLM_HOST_NRHASH]; static unsigned long next_gc; @@ -100,9 +99,9 @@ nlm_lookup_host(int server, struct sockaddr_in *sin, /* Ooops, no host found, create it */ dprintk("lockd: creating host entry\n"); - if (!(host = (struct nlm_host *) kmalloc(sizeof(*host), GFP_KERNEL))) + host = kzalloc(sizeof(*host), GFP_KERNEL); + if (!host) goto nohost; - memset(host, 0, sizeof(*host)); addr = sin->sin_addr.s_addr; sprintf(host->h_name, "%u.%u.%u.%u", NIPQUAD(addr)); @@ -112,11 +111,12 @@ nlm_lookup_host(int server, struct sockaddr_in *sin, host->h_version = version; host->h_proto = proto; host->h_rpcclnt = NULL; - init_MUTEX(&host->h_sema); + mutex_init(&host->h_mutex); host->h_nextrebind = jiffies + NLM_HOST_REBIND; host->h_expires = jiffies + NLM_HOST_EXPIRE; atomic_set(&host->h_count, 1); init_waitqueue_head(&host->h_gracewait); + init_rwsem(&host->h_rwsem); host->h_state = 0; /* pseudo NSM state */ host->h_nsmstate = 0; /* real NSM state */ host->h_server = server; @@ -166,19 +166,17 @@ struct rpc_clnt * nlm_bind_host(struct nlm_host *host) { struct rpc_clnt *clnt; - struct rpc_xprt *xprt; dprintk("lockd: nlm_bind_host(%08x)\n", (unsigned)ntohl(host->h_addr.sin_addr.s_addr)); /* Lock host handle */ - down(&host->h_sema); + mutex_lock(&host->h_mutex); /* If we've already created an RPC client, check whether * RPC rebind is required */ if ((clnt = host->h_rpcclnt) != NULL) { - xprt = clnt->cl_xprt; if (time_after_eq(jiffies, host->h_nextrebind)) { rpc_force_rebind(clnt); host->h_nextrebind = jiffies + NLM_HOST_REBIND; @@ -186,31 +184,37 @@ nlm_bind_host(struct nlm_host *host) host->h_nextrebind - jiffies); } } else { - xprt = xprt_create_proto(host->h_proto, &host->h_addr, NULL); - if (IS_ERR(xprt)) - goto forgetit; - - xprt_set_timeout(&xprt->timeout, 5, nlmsvc_timeout); - xprt->resvport = 1; /* NLM requires a reserved port */ - - /* Existing NLM servers accept AUTH_UNIX only */ - clnt = rpc_new_client(xprt, host->h_name, &nlm_program, - host->h_version, RPC_AUTH_UNIX); - if (IS_ERR(clnt)) - goto forgetit; - clnt->cl_autobind = 1; /* turn on pmap queries */ - clnt->cl_softrtry = 1; /* All queries are soft */ - - host->h_rpcclnt = clnt; + unsigned long increment = nlmsvc_timeout * HZ; + struct rpc_timeout timeparms = { + .to_initval = increment, + .to_increment = increment, + .to_maxval = increment * 6UL, + .to_retries = 5U, + }; + struct rpc_create_args args = { + .protocol = host->h_proto, + .address = (struct sockaddr *)&host->h_addr, + .addrsize = sizeof(host->h_addr), + .timeout = &timeparms, + .servername = host->h_name, + .program = &nlm_program, + .version = host->h_version, + .authflavor = RPC_AUTH_UNIX, + .flags = (RPC_CLNT_CREATE_HARDRTRY | + RPC_CLNT_CREATE_AUTOBIND), + }; + + clnt = rpc_create(&args); + if (!IS_ERR(clnt)) + host->h_rpcclnt = clnt; + else { + printk("lockd: couldn't create RPC handle for %s\n", host->h_name); + clnt = NULL; + } } - up(&host->h_sema); + mutex_unlock(&host->h_mutex); return clnt; - -forgetit: - printk("lockd: couldn't create RPC handle for %s\n", host->h_name); - up(&host->h_sema); - return NULL; } /* diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c index 3fc683f46b3..5954dcb497e 100644 --- a/fs/lockd/mon.c +++ b/fs/lockd/mon.c @@ -109,30 +109,23 @@ nsm_unmonitor(struct nlm_host *host) static struct rpc_clnt * nsm_create(void) { - struct rpc_xprt *xprt; - struct rpc_clnt *clnt; - struct sockaddr_in sin; - - sin.sin_family = AF_INET; - sin.sin_addr.s_addr = htonl(INADDR_LOOPBACK); - sin.sin_port = 0; - - xprt = xprt_create_proto(IPPROTO_UDP, &sin, NULL); - if (IS_ERR(xprt)) - return (struct rpc_clnt *)xprt; - xprt->resvport = 1; /* NSM requires a reserved port */ - - clnt = rpc_create_client(xprt, "localhost", - &nsm_program, SM_VERSION, - RPC_AUTH_NULL); - if (IS_ERR(clnt)) - goto out_err; - clnt->cl_softrtry = 1; - clnt->cl_oneshot = 1; - return clnt; - -out_err: - return clnt; + struct sockaddr_in sin = { + .sin_family = AF_INET, + .sin_addr.s_addr = htonl(INADDR_LOOPBACK), + .sin_port = 0, + }; + struct rpc_create_args args = { + .protocol = IPPROTO_UDP, + .address = (struct sockaddr *)&sin, + .addrsize = sizeof(sin), + .servername = "localhost", + .program = &nsm_program, + .version = SM_VERSION, + .authflavor = RPC_AUTH_NULL, + .flags = (RPC_CLNT_CREATE_ONESHOT), + }; + + return rpc_create(&args); } /* diff --git a/fs/lockd/svc.c b/fs/lockd/svc.c index fd56c8872f3..9a991b52c64 100644 --- a/fs/lockd/svc.c +++ b/fs/lockd/svc.c @@ -12,7 +12,6 @@ * Copyright (C) 1995, 1996 Olaf Kirch <okir@monad.swb.de> */ -#include <linux/config.h> #include <linux/module.h> #include <linux/init.h> #include <linux/sysctl.h> diff --git a/fs/lockd/svclock.c b/fs/lockd/svclock.c index 3ef739120df..c9d419703cf 100644 --- a/fs/lockd/svclock.c +++ b/fs/lockd/svclock.c @@ -20,7 +20,6 @@ * Copyright (C) 1996, Olaf Kirch <okir@monad.swb.de> */ -#include <linux/config.h> #include <linux/types.h> #include <linux/errno.h> #include <linux/kernel.h> @@ -639,9 +638,6 @@ static void nlmsvc_grant_callback(struct rpc_task *task, void *data) if (task->tk_status < 0) { /* RPC error: Re-insert for retransmission */ timeout = 10 * HZ; - } else if (block->b_done) { - /* Block already removed, kill it for real */ - timeout = 0; } else { /* Call was successful, now wait for client callback */ timeout = 60 * HZ; @@ -710,13 +706,10 @@ nlmsvc_retry_blocked(void) break; if (time_after(block->b_when,jiffies)) break; - dprintk("nlmsvc_retry_blocked(%p, when=%ld, done=%d)\n", - block, block->b_when, block->b_done); + dprintk("nlmsvc_retry_blocked(%p, when=%ld)\n", + block, block->b_when); kref_get(&block->b_count); - if (block->b_done) - nlmsvc_unlink_block(block); - else - nlmsvc_grant_blocked(block); + nlmsvc_grant_blocked(block); nlmsvc_release_block(block); } diff --git a/fs/lockd/svcproc.c b/fs/lockd/svcproc.c index d210cf304e9..dbb66a3b5cd 100644 --- a/fs/lockd/svcproc.c +++ b/fs/lockd/svcproc.c @@ -7,7 +7,6 @@ * Copyright (C) 1996, Olaf Kirch <okir@monad.swb.de> */ -#include <linux/config.h> #include <linux/types.h> #include <linux/time.h> #include <linux/slab.h> diff --git a/fs/lockd/svcsubs.c b/fs/lockd/svcsubs.c index a570e5c8a93..a92dd98f840 100644 --- a/fs/lockd/svcsubs.c +++ b/fs/lockd/svcsubs.c @@ -6,7 +6,6 @@ * Copyright (C) 1996, Olaf Kirch <okir@monad.swb.de> */ -#include <linux/config.h> #include <linux/types.h> #include <linux/string.h> #include <linux/time.h> @@ -101,11 +100,10 @@ nlm_lookup_file(struct svc_rqst *rqstp, struct nlm_file **result, nlm_debug_print_fh("creating file for", f); nfserr = nlm_lck_denied_nolocks; - file = (struct nlm_file *) kmalloc(sizeof(*file), GFP_KERNEL); + file = kzalloc(sizeof(*file), GFP_KERNEL); if (!file) goto out_unlock; - memset(file, 0, sizeof(*file)); memcpy(&file->f_handle, f, sizeof(struct nfs_fh)); file->f_hash = hash; init_MUTEX(&file->f_sema); @@ -238,19 +236,22 @@ static int nlm_traverse_files(struct nlm_host *host, int action) { struct nlm_file *file, **fp; - int i; + int i, ret = 0; mutex_lock(&nlm_file_mutex); for (i = 0; i < FILE_NRHASH; i++) { fp = nlm_files + i; while ((file = *fp) != NULL) { + file->f_count++; + mutex_unlock(&nlm_file_mutex); + /* Traverse locks, blocks and shares of this file * and update file->f_locks count */ - if (nlm_inspect_file(host, file, action)) { - mutex_unlock(&nlm_file_mutex); - return 1; - } + if (nlm_inspect_file(host, file, action)) + ret = 1; + mutex_lock(&nlm_file_mutex); + file->f_count--; /* No more references to this file. Let go of it. */ if (!file->f_blocks && !file->f_locks && !file->f_shares && !file->f_count) { @@ -263,7 +264,7 @@ nlm_traverse_files(struct nlm_host *host, int action) } } mutex_unlock(&nlm_file_mutex); - return 0; + return ret; } /* diff --git a/fs/lockd/xdr.c b/fs/lockd/xdr.c index f22a3764461..033ea4ac2c3 100644 --- a/fs/lockd/xdr.c +++ b/fs/lockd/xdr.c @@ -6,7 +6,6 @@ * Copyright (C) 1995, 1996 Olaf Kirch <okir@monad.swb.de> */ -#include <linux/config.h> #include <linux/types.h> #include <linux/sched.h> #include <linux/utsname.h> diff --git a/fs/locks.c b/fs/locks.c index 1ad29c9b625..d7c53392cac 100644 --- a/fs/locks.c +++ b/fs/locks.c @@ -725,6 +725,10 @@ next_task: /* Try to create a FLOCK lock on filp. We always insert new FLOCK locks * at the head of the list, but that's secret knowledge known only to * flock_lock_file and posix_lock_file. + * + * Note that if called with an FL_EXISTS argument, the caller may determine + * whether or not a lock was successfully freed by testing the return + * value for -ENOENT. */ static int flock_lock_file(struct file *filp, struct file_lock *request) { @@ -735,6 +739,8 @@ static int flock_lock_file(struct file *filp, struct file_lock *request) int found = 0; lock_kernel(); + if (request->fl_flags & FL_ACCESS) + goto find_conflict; for_each_lock(inode, before) { struct file_lock *fl = *before; if (IS_POSIX(fl)) @@ -750,8 +756,11 @@ static int flock_lock_file(struct file *filp, struct file_lock *request) break; } - if (request->fl_type == F_UNLCK) + if (request->fl_type == F_UNLCK) { + if ((request->fl_flags & FL_EXISTS) && !found) + error = -ENOENT; goto out; + } error = -ENOMEM; new_fl = locks_alloc_lock(); @@ -764,6 +773,7 @@ static int flock_lock_file(struct file *filp, struct file_lock *request) if (found) cond_resched(); +find_conflict: for_each_lock(inode, before) { struct file_lock *fl = *before; if (IS_POSIX(fl)) @@ -777,6 +787,8 @@ static int flock_lock_file(struct file *filp, struct file_lock *request) locks_insert_block(fl, request); goto out; } + if (request->fl_flags & FL_ACCESS) + goto out; locks_copy_lock(new_fl, request); locks_insert_lock(&inode->i_flock, new_fl); new_fl = NULL; @@ -948,8 +960,11 @@ static int __posix_lock_file_conf(struct inode *inode, struct file_lock *request error = 0; if (!added) { - if (request->fl_type == F_UNLCK) + if (request->fl_type == F_UNLCK) { + if (request->fl_flags & FL_EXISTS) + error = -ENOENT; goto out; + } if (!new_fl) { error = -ENOLCK; @@ -996,6 +1011,10 @@ static int __posix_lock_file_conf(struct inode *inode, struct file_lock *request * Add a POSIX style lock to a file. * We merge adjacent & overlapping locks whenever possible. * POSIX locks are sorted by owner task, then by starting address + * + * Note that if called with an FL_EXISTS argument, the caller may determine + * whether or not a lock was successfully freed by testing the return + * value for -ENOENT. */ int posix_lock_file(struct file *filp, struct file_lock *fl) { @@ -1402,8 +1421,9 @@ static int __setlease(struct file *filp, long arg, struct file_lock **flp) if (!leases_enable) goto out; - error = lease_alloc(filp, arg, &fl); - if (error) + error = -ENOMEM; + fl = locks_alloc_lock(); + if (fl == NULL) goto out; locks_copy_lock(fl, lease); @@ -1411,6 +1431,7 @@ static int __setlease(struct file *filp, long arg, struct file_lock **flp) locks_insert_lock(before, fl); *flp = fl; + error = 0; out: return error; } diff --git a/fs/mbcache.c b/fs/mbcache.c index e4fde1ab22c..0ff71256e65 100644 --- a/fs/mbcache.c +++ b/fs/mbcache.c @@ -160,6 +160,7 @@ __mb_cache_entry_forget(struct mb_cache_entry *ce, gfp_t gfp_mask) static void __mb_cache_entry_release_unlock(struct mb_cache_entry *ce) + __releases(mb_cache_spinlock) { /* Wake up all processes queuing for this cache entry. */ if (ce->e_queued) diff --git a/fs/minix/bitmap.c b/fs/minix/bitmap.c index 4a6abc49418..df6b1075b54 100644 --- a/fs/minix/bitmap.c +++ b/fs/minix/bitmap.c @@ -254,7 +254,7 @@ struct inode * minix_new_inode(const struct inode * dir, int * error) inode->i_gid = (dir->i_mode & S_ISGID) ? dir->i_gid : current->fsgid; inode->i_ino = j; inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME_SEC; - inode->i_blocks = inode->i_blksize = 0; + inode->i_blocks = 0; memset(&minix_i(inode)->u, 0, sizeof(minix_i(inode)->u)); insert_inode_hash(inode); mark_inode_dirty(inode); diff --git a/fs/minix/inode.c b/fs/minix/inode.c index a6fb509b734..c11a4b9fb86 100644 --- a/fs/minix/inode.c +++ b/fs/minix/inode.c @@ -90,8 +90,7 @@ static int init_inodecache(void) static void destroy_inodecache(void) { - if (kmem_cache_destroy(minix_inode_cachep)) - printk(KERN_INFO "minix_inode_cache: not all structures were freed\n"); + kmem_cache_destroy(minix_inode_cachep); } static struct super_operations minix_sops = { @@ -145,11 +144,10 @@ static int minix_fill_super(struct super_block *s, void *data, int silent) struct inode *root_inode; struct minix_sb_info *sbi; - sbi = kmalloc(sizeof(struct minix_sb_info), GFP_KERNEL); + sbi = kzalloc(sizeof(struct minix_sb_info), GFP_KERNEL); if (!sbi) return -ENOMEM; s->s_fs_info = sbi; - memset(sbi, 0, sizeof(struct minix_sb_info)); /* N.B. These should be compile-time tests. Unfortunately that is impossible. */ @@ -204,11 +202,12 @@ static int minix_fill_super(struct super_block *s, void *data, int silent) /* * Allocate the buffer map to keep the superblock small. */ + if (sbi->s_imap_blocks == 0 || sbi->s_zmap_blocks == 0) + goto out_illegal_sb; i = (sbi->s_imap_blocks + sbi->s_zmap_blocks) * sizeof(bh); - map = kmalloc(i, GFP_KERNEL); + map = kzalloc(i, GFP_KERNEL); if (!map) goto out_no_map; - memset(map, 0, i); sbi->s_imap = &map[0]; sbi->s_zmap = &map[sbi->s_imap_blocks]; @@ -263,7 +262,7 @@ out_no_root: out_no_bitmap: printk("MINIX-fs: bad superblock or unable to read bitmaps\n"); - out_freemap: +out_freemap: for (i = 0; i < sbi->s_imap_blocks; i++) brelse(sbi->s_imap[i]); for (i = 0; i < sbi->s_zmap_blocks; i++) @@ -276,11 +275,16 @@ out_no_map: printk("MINIX-fs: can't allocate map\n"); goto out_release; +out_illegal_sb: + if (!silent) + printk("MINIX-fs: bad superblock\n"); + goto out_release; + out_no_fs: if (!silent) printk("VFS: Can't find a Minix or Minix V2 filesystem " "on device %s\n", s->s_id); - out_release: +out_release: brelse(bh); goto out; @@ -290,7 +294,7 @@ out_bad_hblock: out_bad_sb: printk("MINIX-fs: unable to read superblock\n"); - out: +out: s->s_fs_info = NULL; kfree(sbi); return -EINVAL; @@ -335,7 +339,7 @@ static sector_t minix_bmap(struct address_space *mapping, sector_t block) { return generic_block_bmap(mapping,block,minix_get_block); } -static struct address_space_operations minix_aops = { +static const struct address_space_operations minix_aops = { .readpage = minix_readpage, .writepage = minix_writepage, .sync_page = block_sync_page, @@ -392,7 +396,7 @@ static void V1_minix_read_inode(struct inode * inode) inode->i_mtime.tv_nsec = 0; inode->i_atime.tv_nsec = 0; inode->i_ctime.tv_nsec = 0; - inode->i_blocks = inode->i_blksize = 0; + inode->i_blocks = 0; for (i = 0; i < 9; i++) minix_inode->u.i1_data[i] = raw_inode->i_zone[i]; minix_set_inode(inode, old_decode_dev(raw_inode->i_zone[0])); @@ -425,7 +429,7 @@ static void V2_minix_read_inode(struct inode * inode) inode->i_mtime.tv_nsec = 0; inode->i_atime.tv_nsec = 0; inode->i_ctime.tv_nsec = 0; - inode->i_blocks = inode->i_blksize = 0; + inode->i_blocks = 0; for (i = 0; i < 10; i++) minix_inode->u.i2_data[i] = raw_inode->i_zone[i]; minix_set_inode(inode, old_decode_dev(raw_inode->i_zone[0])); diff --git a/fs/msdos/namei.c b/fs/msdos/namei.c index 9e44158a754..d220165d491 100644 --- a/fs/msdos/namei.c +++ b/fs/msdos/namei.c @@ -280,7 +280,7 @@ static int msdos_create(struct inode *dir, struct dentry *dentry, int mode, struct nameidata *nd) { struct super_block *sb = dir->i_sb; - struct inode *inode; + struct inode *inode = NULL; struct fat_slot_info sinfo; struct timespec ts; unsigned char msdos_name[MSDOS_NAME]; @@ -316,6 +316,8 @@ static int msdos_create(struct inode *dir, struct dentry *dentry, int mode, d_instantiate(dentry, inode); out: unlock_kernel(); + if (!err) + err = fat_flush_inodes(sb, dir, inode); return err; } @@ -348,6 +350,8 @@ static int msdos_rmdir(struct inode *dir, struct dentry *dentry) fat_detach(inode); out: unlock_kernel(); + if (!err) + err = fat_flush_inodes(inode->i_sb, dir, inode); return err; } @@ -401,6 +405,7 @@ static int msdos_mkdir(struct inode *dir, struct dentry *dentry, int mode) d_instantiate(dentry, inode); unlock_kernel(); + fat_flush_inodes(sb, dir, inode); return 0; out_free: @@ -430,6 +435,8 @@ static int msdos_unlink(struct inode *dir, struct dentry *dentry) fat_detach(inode); out: unlock_kernel(); + if (!err) + err = fat_flush_inodes(inode->i_sb, dir, inode); return err; } @@ -635,6 +642,8 @@ static int msdos_rename(struct inode *old_dir, struct dentry *old_dentry, new_dir, new_msdos_name, new_dentry, is_hid); out: unlock_kernel(); + if (!err) + err = fat_flush_inodes(old_dir->i_sb, old_dir, new_dir); return err; } diff --git a/fs/namei.c b/fs/namei.c index bb4a3e40e43..2892e68d3a8 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -159,7 +159,7 @@ char * getname(const char __user * filename) #ifdef CONFIG_AUDITSYSCALL void putname(const char *name) { - if (unlikely(current->audit_context)) + if (unlikely(!audit_dummy_context())) audit_putname(name); else __putname(name); @@ -227,10 +227,10 @@ int generic_permission(struct inode *inode, int mask, int permission(struct inode *inode, int mask, struct nameidata *nd) { + umode_t mode = inode->i_mode; int retval, submask; if (mask & MAY_WRITE) { - umode_t mode = inode->i_mode; /* * Nobody gets write access to a read-only fs. @@ -247,6 +247,13 @@ int permission(struct inode *inode, int mask, struct nameidata *nd) } + /* + * MAY_EXEC on regular files requires special handling: We override + * filesystem execute permissions if the mode bits aren't set. + */ + if ((mask & MAY_EXEC) && S_ISREG(mode) && !(mode & S_IXUGO)) + return -EACCES; + /* Ordinary permission routines do not understand MAY_APPEND. */ submask = mask & ~MAY_APPEND; if (inode->i_op && inode->i_op->permission) @@ -365,6 +372,30 @@ void release_open_intent(struct nameidata *nd) fput(nd->intent.open.file); } +static inline struct dentry * +do_revalidate(struct dentry *dentry, struct nameidata *nd) +{ + int status = dentry->d_op->d_revalidate(dentry, nd); + if (unlikely(status <= 0)) { + /* + * The dentry failed validation. + * If d_revalidate returned 0 attempt to invalidate + * the dentry otherwise d_revalidate is asking us + * to return a fail status. + */ + if (!status) { + if (!d_invalidate(dentry)) { + dput(dentry); + dentry = NULL; + } + } else { + dput(dentry); + dentry = ERR_PTR(status); + } + } + return dentry; +} + /* * Internal lookup() using the new generic dcache. * SMP-safe @@ -379,12 +410,9 @@ static struct dentry * cached_lookup(struct dentry * parent, struct qstr * name, if (!dentry) dentry = d_lookup(parent, name); - if (dentry && dentry->d_op && dentry->d_op->d_revalidate) { - if (!dentry->d_op->d_revalidate(dentry, nd) && !d_invalidate(dentry)) { - dput(dentry); - dentry = NULL; - } - } + if (dentry && dentry->d_op && dentry->d_op->d_revalidate) + dentry = do_revalidate(dentry, nd); + return dentry; } @@ -477,10 +505,9 @@ static struct dentry * real_lookup(struct dentry * parent, struct qstr * name, s */ mutex_unlock(&dir->i_mutex); if (result->d_op && result->d_op->d_revalidate) { - if (!result->d_op->d_revalidate(result, nd) && !d_invalidate(result)) { - dput(result); + result = do_revalidate(result, nd); + if (!result) result = ERR_PTR(-ENOENT); - } } return result; } @@ -491,18 +518,20 @@ static int __emul_lookup_dentry(const char *, struct nameidata *); static __always_inline int walk_init_root(const char *name, struct nameidata *nd) { - read_lock(¤t->fs->lock); - if (current->fs->altroot && !(nd->flags & LOOKUP_NOALT)) { - nd->mnt = mntget(current->fs->altrootmnt); - nd->dentry = dget(current->fs->altroot); - read_unlock(¤t->fs->lock); + struct fs_struct *fs = current->fs; + + read_lock(&fs->lock); + if (fs->altroot && !(nd->flags & LOOKUP_NOALT)) { + nd->mnt = mntget(fs->altrootmnt); + nd->dentry = dget(fs->altroot); + read_unlock(&fs->lock); if (__emul_lookup_dentry(name,nd)) return 0; - read_lock(¤t->fs->lock); + read_lock(&fs->lock); } - nd->mnt = mntget(current->fs->rootmnt); - nd->dentry = dget(current->fs->root); - read_unlock(¤t->fs->lock); + nd->mnt = mntget(fs->rootmnt); + nd->dentry = dget(fs->root); + read_unlock(&fs->lock); return 1; } @@ -697,17 +726,19 @@ int follow_down(struct vfsmount **mnt, struct dentry **dentry) static __always_inline void follow_dotdot(struct nameidata *nd) { + struct fs_struct *fs = current->fs; + while(1) { struct vfsmount *parent; struct dentry *old = nd->dentry; - read_lock(¤t->fs->lock); - if (nd->dentry == current->fs->root && - nd->mnt == current->fs->rootmnt) { - read_unlock(¤t->fs->lock); + read_lock(&fs->lock); + if (nd->dentry == fs->root && + nd->mnt == fs->rootmnt) { + read_unlock(&fs->lock); break; } - read_unlock(¤t->fs->lock); + read_unlock(&fs->lock); spin_lock(&dcache_lock); if (nd->dentry != nd->mnt->mnt_root) { nd->dentry = dget(nd->dentry->d_parent); @@ -760,12 +791,12 @@ need_lookup: goto done; need_revalidate: - if (dentry->d_op->d_revalidate(dentry, nd)) - goto done; - if (d_invalidate(dentry)) - goto done; - dput(dentry); - goto need_lookup; + dentry = do_revalidate(dentry, nd); + if (!dentry) + goto need_lookup; + if (IS_ERR(dentry)) + goto fail; + goto done; fail: return PTR_ERR(dentry); @@ -1015,15 +1046,17 @@ static int __emul_lookup_dentry(const char *name, struct nameidata *nd) struct vfsmount *old_mnt = nd->mnt; struct qstr last = nd->last; int last_type = nd->last_type; + struct fs_struct *fs = current->fs; + /* - * NAME was not found in alternate root or it's a directory. Try to find - * it in the normal root: + * NAME was not found in alternate root or it's a directory. + * Try to find it in the normal root: */ nd->last_type = LAST_ROOT; - read_lock(¤t->fs->lock); - nd->mnt = mntget(current->fs->rootmnt); - nd->dentry = dget(current->fs->root); - read_unlock(¤t->fs->lock); + read_lock(&fs->lock); + nd->mnt = mntget(fs->rootmnt); + nd->dentry = dget(fs->root); + read_unlock(&fs->lock); if (path_walk(name, nd) == 0) { if (nd->dentry->d_inode) { dput(old_dentry); @@ -1047,6 +1080,7 @@ void set_fs_altroot(void) struct vfsmount *mnt = NULL, *oldmnt; struct dentry *dentry = NULL, *olddentry; int err; + struct fs_struct *fs = current->fs; if (!emul) goto set_it; @@ -1056,12 +1090,12 @@ void set_fs_altroot(void) dentry = nd.dentry; } set_it: - write_lock(¤t->fs->lock); - oldmnt = current->fs->altrootmnt; - olddentry = current->fs->altroot; - current->fs->altrootmnt = mnt; - current->fs->altroot = dentry; - write_unlock(¤t->fs->lock); + write_lock(&fs->lock); + oldmnt = fs->altrootmnt; + olddentry = fs->altroot; + fs->altrootmnt = mnt; + fs->altroot = dentry; + write_unlock(&fs->lock); if (olddentry) { dput(olddentry); mntput(oldmnt); @@ -1075,29 +1109,30 @@ static int fastcall do_path_lookup(int dfd, const char *name, int retval = 0; int fput_needed; struct file *file; + struct fs_struct *fs = current->fs; nd->last_type = LAST_ROOT; /* if there are only slashes... */ nd->flags = flags; nd->depth = 0; if (*name=='/') { - read_lock(¤t->fs->lock); - if (current->fs->altroot && !(nd->flags & LOOKUP_NOALT)) { - nd->mnt = mntget(current->fs->altrootmnt); - nd->dentry = dget(current->fs->altroot); - read_unlock(¤t->fs->lock); + read_lock(&fs->lock); + if (fs->altroot && !(nd->flags & LOOKUP_NOALT)) { + nd->mnt = mntget(fs->altrootmnt); + nd->dentry = dget(fs->altroot); + read_unlock(&fs->lock); if (__emul_lookup_dentry(name,nd)) goto out; /* found in altroot */ - read_lock(¤t->fs->lock); + read_lock(&fs->lock); } - nd->mnt = mntget(current->fs->rootmnt); - nd->dentry = dget(current->fs->root); - read_unlock(¤t->fs->lock); + nd->mnt = mntget(fs->rootmnt); + nd->dentry = dget(fs->root); + read_unlock(&fs->lock); } else if (dfd == AT_FDCWD) { - read_lock(¤t->fs->lock); - nd->mnt = mntget(current->fs->pwdmnt); - nd->dentry = dget(current->fs->pwd); - read_unlock(¤t->fs->lock); + read_lock(&fs->lock); + nd->mnt = mntget(fs->pwdmnt); + nd->dentry = dget(fs->pwd); + read_unlock(&fs->lock); } else { struct dentry *dentry; @@ -1125,7 +1160,7 @@ static int fastcall do_path_lookup(int dfd, const char *name, retval = link_path_walk(name, nd); out: if (likely(retval == 0)) { - if (unlikely(current->audit_context && nd && nd->dentry && + if (unlikely(!audit_dummy_context() && nd && nd->dentry && nd->dentry->d_inode)) audit_inode(name, nd->dentry->d_inode); } @@ -1357,7 +1392,7 @@ static int may_delete(struct inode *dir,struct dentry *victim,int isdir) return -ENOENT; BUG_ON(victim->d_parent->d_inode != dir); - audit_inode_child(victim->d_name.name, victim->d_inode, dir->i_ino); + audit_inode_child(victim->d_name.name, victim->d_inode, dir); error = permission(dir,MAY_WRITE | MAY_EXEC, NULL); if (error) @@ -1423,7 +1458,7 @@ struct dentry *lock_rename(struct dentry *p1, struct dentry *p2) struct dentry *p; if (p1 == p2) { - mutex_lock(&p1->d_inode->i_mutex); + mutex_lock_nested(&p1->d_inode->i_mutex, I_MUTEX_PARENT); return NULL; } @@ -1431,22 +1466,22 @@ struct dentry *lock_rename(struct dentry *p1, struct dentry *p2) for (p = p1; p->d_parent != p; p = p->d_parent) { if (p->d_parent == p2) { - mutex_lock(&p2->d_inode->i_mutex); - mutex_lock(&p1->d_inode->i_mutex); + mutex_lock_nested(&p2->d_inode->i_mutex, I_MUTEX_PARENT); + mutex_lock_nested(&p1->d_inode->i_mutex, I_MUTEX_CHILD); return p; } } for (p = p2; p->d_parent != p; p = p->d_parent) { if (p->d_parent == p1) { - mutex_lock(&p1->d_inode->i_mutex); - mutex_lock(&p2->d_inode->i_mutex); + mutex_lock_nested(&p1->d_inode->i_mutex, I_MUTEX_PARENT); + mutex_lock_nested(&p2->d_inode->i_mutex, I_MUTEX_CHILD); return p; } } - mutex_lock(&p1->d_inode->i_mutex); - mutex_lock(&p2->d_inode->i_mutex); + mutex_lock_nested(&p1->d_inode->i_mutex, I_MUTEX_PARENT); + mutex_lock_nested(&p2->d_inode->i_mutex, I_MUTEX_CHILD); return NULL; } @@ -1659,6 +1694,7 @@ do_last: * It already exists. */ mutex_unlock(&dir->d_inode->i_mutex); + audit_inode_update(path.dentry->d_inode); error = -EEXIST; if (flag & O_EXCL) @@ -1669,6 +1705,7 @@ do_last: if (flag & O_NOFOLLOW) goto exit_dput; } + error = -ENOENT; if (!path.dentry->d_inode) goto exit_dput; @@ -1712,8 +1749,14 @@ do_link: if (error) goto exit_dput; error = __do_follow_link(&path, nd); - if (error) + if (error) { + /* Does someone understand code flow here? Or it is only + * me so stupid? Anathema to whoever designed this non-sense + * with "intent.open". + */ + release_open_intent(nd); return error; + } nd->flags &= ~LOOKUP_PARENT; if (nd->last_type == LAST_BIND) goto ok; @@ -1751,7 +1794,7 @@ struct dentry *lookup_create(struct nameidata *nd, int is_dir) { struct dentry *dentry = ERR_PTR(-EEXIST); - mutex_lock(&nd->dentry->d_inode->i_mutex); + mutex_lock_nested(&nd->dentry->d_inode->i_mutex, I_MUTEX_PARENT); /* * Yucky last component or no last component at all? * (foo/., foo/.., /////) @@ -1759,6 +1802,8 @@ struct dentry *lookup_create(struct nameidata *nd, int is_dir) if (nd->last_type != LAST_NORM) goto fail; nd->flags &= ~LOOKUP_PARENT; + nd->flags |= LOOKUP_CREATE; + nd->intent.open.flags = O_EXCL; /* * Do the final lookup. @@ -2008,7 +2053,7 @@ static long do_rmdir(int dfd, const char __user *pathname) error = -EBUSY; goto exit1; } - mutex_lock(&nd.dentry->d_inode->i_mutex); + mutex_lock_nested(&nd.dentry->d_inode->i_mutex, I_MUTEX_PARENT); dentry = lookup_hash(&nd); error = PTR_ERR(dentry); if (!IS_ERR(dentry)) { @@ -2082,7 +2127,7 @@ static long do_unlinkat(int dfd, const char __user *pathname) error = -EISDIR; if (nd.last_type != LAST_NORM) goto exit1; - mutex_lock(&nd.dentry->d_inode->i_mutex); + mutex_lock_nested(&nd.dentry->d_inode->i_mutex, I_MUTEX_PARENT); dentry = lookup_hash(&nd); error = PTR_ERR(dentry); if (!IS_ERR(dentry)) { @@ -2243,14 +2288,16 @@ asmlinkage long sys_linkat(int olddfd, const char __user *oldname, int error; char * to; - if (flags != 0) + if ((flags & ~AT_SYMLINK_FOLLOW) != 0) return -EINVAL; to = getname(newname); if (IS_ERR(to)) return PTR_ERR(to); - error = __user_walk_fd(olddfd, oldname, 0, &old_nd); + error = __user_walk_fd(olddfd, oldname, + flags & AT_SYMLINK_FOLLOW ? LOOKUP_FOLLOW : 0, + &old_nd); if (error) goto exit; error = do_path_lookup(newdfd, to, LOOKUP_PARENT, &nd); @@ -2351,7 +2398,8 @@ static int vfs_rename_dir(struct inode *old_dir, struct dentry *old_dentry, dput(new_dentry); } if (!error) - d_move(old_dentry,new_dentry); + if (!(old_dir->i_sb->s_type->fs_flags & FS_RENAME_DOES_D_MOVE)) + d_move(old_dentry,new_dentry); return error; } @@ -2374,8 +2422,7 @@ static int vfs_rename_other(struct inode *old_dir, struct dentry *old_dentry, else error = old_dir->i_op->rename(old_dir, old_dentry, new_dir, new_dentry); if (!error) { - /* The following d_move() should become unconditional */ - if (!(old_dir->i_sb->s_type->fs_flags & FS_ODD_RENAME)) + if (!(old_dir->i_sb->s_type->fs_flags & FS_RENAME_DOES_D_MOVE)) d_move(old_dentry, new_dentry); } if (target) diff --git a/fs/namespace.c b/fs/namespace.c index c13072a5f1e..6ede3a539ed 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -8,16 +8,17 @@ * Heavily rewritten. */ -#include <linux/config.h> #include <linux/syscalls.h> #include <linux/slab.h> #include <linux/sched.h> #include <linux/smp_lock.h> #include <linux/init.h> +#include <linux/kernel.h> #include <linux/quotaops.h> #include <linux/acct.h> #include <linux/capability.h> #include <linux/module.h> +#include <linux/sysfs.h> #include <linux/seq_file.h> #include <linux/namespace.h> #include <linux/namei.h> @@ -29,15 +30,6 @@ extern int __init init_rootfs(void); -#ifdef CONFIG_SYSFS -extern int __init sysfs_init(void); -#else -static inline int sysfs_init(void) -{ - return 0; -} -#endif - /* spinlock for vfsmount related operations, inplace of dcache_lock */ __cacheline_aligned_in_smp DEFINE_SPINLOCK(vfsmount_lock); @@ -526,10 +518,8 @@ void umount_tree(struct vfsmount *mnt, int propagate, struct list_head *kill) { struct vfsmount *p; - for (p = mnt; p; p = next_mnt(p, mnt)) { - list_del(&p->mnt_hash); - list_add(&p->mnt_hash, kill); - } + for (p = mnt; p; p = next_mnt(p, mnt)) + list_move(&p->mnt_hash, kill); if (propagate) propagate_umount(kill); @@ -585,8 +575,8 @@ static int do_umount(struct vfsmount *mnt, int flags) */ lock_kernel(); - if ((flags & MNT_FORCE) && sb->s_op->umount_begin) - sb->s_op->umount_begin(sb); + if (sb->s_op->umount_begin) + sb->s_op->umount_begin(mnt, flags); unlock_kernel(); /* @@ -1172,13 +1162,46 @@ static void expire_mount(struct vfsmount *mnt, struct list_head *mounts, } /* + * go through the vfsmounts we've just consigned to the graveyard to + * - check that they're still dead + * - delete the vfsmount from the appropriate namespace under lock + * - dispose of the corpse + */ +static void expire_mount_list(struct list_head *graveyard, struct list_head *mounts) +{ + struct namespace *namespace; + struct vfsmount *mnt; + + while (!list_empty(graveyard)) { + LIST_HEAD(umounts); + mnt = list_entry(graveyard->next, struct vfsmount, mnt_expire); + list_del_init(&mnt->mnt_expire); + + /* don't do anything if the namespace is dead - all the + * vfsmounts from it are going away anyway */ + namespace = mnt->mnt_namespace; + if (!namespace || !namespace->root) + continue; + get_namespace(namespace); + + spin_unlock(&vfsmount_lock); + down_write(&namespace_sem); + expire_mount(mnt, mounts, &umounts); + up_write(&namespace_sem); + release_mounts(&umounts); + mntput(mnt); + put_namespace(namespace); + spin_lock(&vfsmount_lock); + } +} + +/* * process a list of expirable mountpoints with the intent of discarding any * mountpoints that aren't in use and haven't been touched since last we came * here */ void mark_mounts_for_expiry(struct list_head *mounts) { - struct namespace *namespace; struct vfsmount *mnt, *next; LIST_HEAD(graveyard); @@ -1202,38 +1225,79 @@ void mark_mounts_for_expiry(struct list_head *mounts) list_move(&mnt->mnt_expire, &graveyard); } - /* - * go through the vfsmounts we've just consigned to the graveyard to - * - check that they're still dead - * - delete the vfsmount from the appropriate namespace under lock - * - dispose of the corpse - */ - while (!list_empty(&graveyard)) { - LIST_HEAD(umounts); - mnt = list_entry(graveyard.next, struct vfsmount, mnt_expire); - list_del_init(&mnt->mnt_expire); + expire_mount_list(&graveyard, mounts); - /* don't do anything if the namespace is dead - all the - * vfsmounts from it are going away anyway */ - namespace = mnt->mnt_namespace; - if (!namespace || !namespace->root) + spin_unlock(&vfsmount_lock); +} + +EXPORT_SYMBOL_GPL(mark_mounts_for_expiry); + +/* + * Ripoff of 'select_parent()' + * + * search the list of submounts for a given mountpoint, and move any + * shrinkable submounts to the 'graveyard' list. + */ +static int select_submounts(struct vfsmount *parent, struct list_head *graveyard) +{ + struct vfsmount *this_parent = parent; + struct list_head *next; + int found = 0; + +repeat: + next = this_parent->mnt_mounts.next; +resume: + while (next != &this_parent->mnt_mounts) { + struct list_head *tmp = next; + struct vfsmount *mnt = list_entry(tmp, struct vfsmount, mnt_child); + + next = tmp->next; + if (!(mnt->mnt_flags & MNT_SHRINKABLE)) continue; - get_namespace(namespace); + /* + * Descend a level if the d_mounts list is non-empty. + */ + if (!list_empty(&mnt->mnt_mounts)) { + this_parent = mnt; + goto repeat; + } - spin_unlock(&vfsmount_lock); - down_write(&namespace_sem); - expire_mount(mnt, mounts, &umounts); - up_write(&namespace_sem); - release_mounts(&umounts); - mntput(mnt); - put_namespace(namespace); - spin_lock(&vfsmount_lock); + if (!propagate_mount_busy(mnt, 1)) { + mntget(mnt); + list_move_tail(&mnt->mnt_expire, graveyard); + found++; + } + } + /* + * All done at this level ... ascend and resume the search + */ + if (this_parent != parent) { + next = this_parent->mnt_child.next; + this_parent = this_parent->mnt_parent; + goto resume; } + return found; +} + +/* + * process a list of expirable mountpoints with the intent of discarding any + * submounts of a specific parent mountpoint + */ +void shrink_submounts(struct vfsmount *mountpoint, struct list_head *mounts) +{ + LIST_HEAD(graveyard); + int found; + + spin_lock(&vfsmount_lock); + + /* extract submounts of 'mountpoint' from the expiration list */ + while ((found = select_submounts(mountpoint, &graveyard)) != 0) + expire_mount_list(&graveyard, mounts); spin_unlock(&vfsmount_lock); } -EXPORT_SYMBOL_GPL(mark_mounts_for_expiry); +EXPORT_SYMBOL_GPL(shrink_submounts); /* * Some copy_from_user() implementations do not return the exact number of @@ -1750,6 +1814,7 @@ void __init mnt_init(unsigned long mempages) struct list_head *d; unsigned int nr_hash; int i; + int err; init_rwsem(&namespace_sem); @@ -1790,8 +1855,14 @@ void __init mnt_init(unsigned long mempages) d++; i--; } while (i); - sysfs_init(); - subsystem_register(&fs_subsys); + err = sysfs_init(); + if (err) + printk(KERN_WARNING "%s: sysfs_init error: %d\n", + __FUNCTION__, err); + err = subsystem_register(&fs_subsys); + if (err) + printk(KERN_WARNING "%s: subsystem_register error: %d\n", + __FUNCTION__, err); init_rootfs(); init_mount_tree(); } diff --git a/fs/ncpfs/dir.c b/fs/ncpfs/dir.c index f0860c602d8..b4ee89250e9 100644 --- a/fs/ncpfs/dir.c +++ b/fs/ncpfs/dir.c @@ -10,7 +10,6 @@ * */ -#include <linux/config.h> #include <linux/time.h> #include <linux/errno.h> diff --git a/fs/ncpfs/inode.c b/fs/ncpfs/inode.c index 90d2ea28f33..42e3bef270c 100644 --- a/fs/ncpfs/inode.c +++ b/fs/ncpfs/inode.c @@ -9,7 +9,6 @@ * */ -#include <linux/config.h> #include <linux/module.h> #include <asm/system.h> @@ -82,8 +81,7 @@ static int init_inodecache(void) static void destroy_inodecache(void) { - if (kmem_cache_destroy(ncp_inode_cachep)) - printk(KERN_INFO "ncp_inode_cache: not all structures were freed\n"); + kmem_cache_destroy(ncp_inode_cachep); } static int ncp_remount(struct super_block *sb, int *flags, char* data) @@ -105,7 +103,7 @@ static struct super_operations ncp_sops = extern struct dentry_operations ncp_root_dentry_operations; #if defined(CONFIG_NCPFS_EXTRAS) || defined(CONFIG_NCPFS_NFS_NS) -extern struct address_space_operations ncp_symlink_aops; +extern const struct address_space_operations ncp_symlink_aops; extern int ncp_symlink(struct inode*, struct dentry*, const char*); #endif @@ -225,7 +223,6 @@ static void ncp_set_attr(struct inode *inode, struct ncp_entry_info *nwinfo) inode->i_nlink = 1; inode->i_uid = server->m.uid; inode->i_gid = server->m.gid; - inode->i_blksize = NCP_BLOCK_SIZE; ncp_update_dates(inode, &nwinfo->i); ncp_update_inode(inode, nwinfo); @@ -412,11 +409,10 @@ static int ncp_fill_super(struct super_block *sb, void *raw_data, int silent) #endif struct ncp_entry_info finfo; - server = kmalloc(sizeof(struct ncp_server), GFP_KERNEL); + server = kzalloc(sizeof(struct ncp_server), GFP_KERNEL); if (!server) return -ENOMEM; sb->s_fs_info = server; - memset(server, 0, sizeof(struct ncp_server)); error = -EFAULT; if (raw_data == NULL) diff --git a/fs/ncpfs/ioctl.c b/fs/ncpfs/ioctl.c index eb3813ad136..42039fe0653 100644 --- a/fs/ncpfs/ioctl.c +++ b/fs/ncpfs/ioctl.c @@ -7,7 +7,6 @@ * */ -#include <linux/config.h> #include <asm/uaccess.h> #include <linux/capability.h> diff --git a/fs/ncpfs/mmap.c b/fs/ncpfs/mmap.c index 52d60c3d899..e7d5a3097fe 100644 --- a/fs/ncpfs/mmap.c +++ b/fs/ncpfs/mmap.c @@ -93,7 +93,7 @@ static struct page* ncp_file_mmap_nopage(struct vm_area_struct *area, */ if (type) *type = VM_FAULT_MAJOR; - inc_page_state(pgmajfault); + count_vm_event(PGMAJFAULT); return page; } diff --git a/fs/ncpfs/ncplib_kernel.c b/fs/ncpfs/ncplib_kernel.c index d9ebf6439f5..551e0bac7aa 100644 --- a/fs/ncpfs/ncplib_kernel.c +++ b/fs/ncpfs/ncplib_kernel.c @@ -10,7 +10,6 @@ */ -#include <linux/config.h> #include "ncplib_kernel.h" diff --git a/fs/ncpfs/ncplib_kernel.h b/fs/ncpfs/ncplib_kernel.h index 799e5c2bec5..2441d1ab57d 100644 --- a/fs/ncpfs/ncplib_kernel.h +++ b/fs/ncpfs/ncplib_kernel.h @@ -12,7 +12,6 @@ #ifndef _NCPLIB_H #define _NCPLIB_H -#include <linux/config.h> #include <linux/fs.h> #include <linux/types.h> diff --git a/fs/ncpfs/ncpsign_kernel.c b/fs/ncpfs/ncpsign_kernel.c index a6ec90cd889..749a18d3359 100644 --- a/fs/ncpfs/ncpsign_kernel.c +++ b/fs/ncpfs/ncpsign_kernel.c @@ -5,7 +5,6 @@ * */ -#include <linux/config.h> #ifdef CONFIG_NCPFS_PACKET_SIGNING diff --git a/fs/ncpfs/sock.c b/fs/ncpfs/sock.c index 8783eb7ec64..11c2b252ebe 100644 --- a/fs/ncpfs/sock.c +++ b/fs/ncpfs/sock.c @@ -8,7 +8,6 @@ * */ -#include <linux/config.h> #include <linux/time.h> #include <linux/errno.h> diff --git a/fs/ncpfs/symlink.c b/fs/ncpfs/symlink.c index e935f1b34bc..e3d26c1bd10 100644 --- a/fs/ncpfs/symlink.c +++ b/fs/ncpfs/symlink.c @@ -20,7 +20,6 @@ * */ -#include <linux/config.h> #include <asm/uaccess.h> @@ -49,7 +48,7 @@ static int ncp_symlink_readpage(struct file *file, struct page *page) char *buf = kmap(page); error = -ENOMEM; - rawlink=(char *)kmalloc(NCP_MAX_SYMLINK_SIZE, GFP_KERNEL); + rawlink = kmalloc(NCP_MAX_SYMLINK_SIZE, GFP_KERNEL); if (!rawlink) goto fail; @@ -99,7 +98,7 @@ fail: /* * symlinks can't do much... */ -struct address_space_operations ncp_symlink_aops = { +const struct address_space_operations ncp_symlink_aops = { .readpage = ncp_symlink_readpage, }; @@ -127,7 +126,7 @@ int ncp_symlink(struct inode *dir, struct dentry *dentry, const char *symname) { /* EPERM is returned by VFS if symlink procedure does not exist */ return -EPERM; - rawlink=(char *)kmalloc(NCP_MAX_SYMLINK_SIZE, GFP_KERNEL); + rawlink = kmalloc(NCP_MAX_SYMLINK_SIZE, GFP_KERNEL); if (!rawlink) return -ENOMEM; diff --git a/fs/nfs/Makefile b/fs/nfs/Makefile index ec61fd56a1a..f4580b44eef 100644 --- a/fs/nfs/Makefile +++ b/fs/nfs/Makefile @@ -4,14 +4,16 @@ obj-$(CONFIG_NFS_FS) += nfs.o -nfs-y := dir.o file.o inode.o nfs2xdr.o pagelist.o \ - proc.o read.o symlink.o unlink.o write.o +nfs-y := client.o dir.o file.o getroot.o inode.o super.o nfs2xdr.o \ + pagelist.o proc.o read.o symlink.o unlink.o \ + write.o namespace.o nfs-$(CONFIG_ROOT_NFS) += nfsroot.o mount_clnt.o nfs-$(CONFIG_NFS_V3) += nfs3proc.o nfs3xdr.o nfs-$(CONFIG_NFS_V3_ACL) += nfs3acl.o nfs-$(CONFIG_NFS_V4) += nfs4proc.o nfs4xdr.o nfs4state.o nfs4renewd.o \ delegation.o idmap.o \ - callback.o callback_xdr.o callback_proc.o + callback.o callback_xdr.o callback_proc.o \ + nfs4namespace.o nfs-$(CONFIG_NFS_DIRECTIO) += direct.o nfs-$(CONFIG_SYSCTL) += sysctl.o nfs-objs := $(nfs-y) diff --git a/fs/nfs/callback.c b/fs/nfs/callback.c index 90c95adc8c1..a3ee11364db 100644 --- a/fs/nfs/callback.c +++ b/fs/nfs/callback.c @@ -6,7 +6,6 @@ * NFSv4 callback handling */ -#include <linux/config.h> #include <linux/completion.h> #include <linux/ip.h> #include <linux/module.h> @@ -20,6 +19,7 @@ #include "nfs4_fs.h" #include "callback.h" +#include "internal.h" #define NFSDBG_FACILITY NFSDBG_CALLBACK @@ -37,6 +37,21 @@ static struct svc_program nfs4_callback_program; unsigned int nfs_callback_set_tcpport; unsigned short nfs_callback_tcpport; +static const int nfs_set_port_min = 0; +static const int nfs_set_port_max = 65535; + +static int param_set_port(const char *val, struct kernel_param *kp) +{ + char *endp; + int num = simple_strtol(val, &endp, 0); + if (endp == val || *endp || num < nfs_set_port_min || num > nfs_set_port_max) + return -EINVAL; + *((int *)kp->arg) = num; + return 0; +} + +module_param_call(callback_tcpport, param_set_port, param_get_int, + &nfs_callback_set_tcpport, 0644); /* * This is the callback kernel thread. @@ -135,10 +150,8 @@ out_err: /* * Kill the server process if it is not already up. */ -int nfs_callback_down(void) +void nfs_callback_down(void) { - int ret = 0; - lock_kernel(); mutex_lock(&nfs_callback_mutex); nfs_callback_info.users--; @@ -150,20 +163,19 @@ int nfs_callback_down(void) } while (wait_for_completion_timeout(&nfs_callback_info.stopped, 5*HZ) == 0); mutex_unlock(&nfs_callback_mutex); unlock_kernel(); - return ret; } static int nfs_callback_authenticate(struct svc_rqst *rqstp) { - struct in_addr *addr = &rqstp->rq_addr.sin_addr; - struct nfs4_client *clp; + struct sockaddr_in *addr = &rqstp->rq_addr; + struct nfs_client *clp; /* Don't talk to strangers */ - clp = nfs4_find_client(addr); + clp = nfs_find_client(addr, 4); if (clp == NULL) return SVC_DROP; - dprintk("%s: %u.%u.%u.%u NFSv4 callback!\n", __FUNCTION__, NIPQUAD(addr)); - nfs4_put_client(clp); + dprintk("%s: %u.%u.%u.%u NFSv4 callback!\n", __FUNCTION__, NIPQUAD(addr->sin_addr)); + nfs_put_client(clp); switch (rqstp->rq_authop->flavour) { case RPC_AUTH_NULL: if (rqstp->rq_proc != CB_NULL) @@ -182,8 +194,6 @@ static int nfs_callback_authenticate(struct svc_rqst *rqstp) /* * Define NFS4 callback program */ -extern struct svc_version nfs4_callback_version1; - static struct svc_version *nfs4_callback_version[] = { [1] = &nfs4_callback_version1, }; diff --git a/fs/nfs/callback.h b/fs/nfs/callback.h index b252e7fe53a..5676163d26e 100644 --- a/fs/nfs/callback.h +++ b/fs/nfs/callback.h @@ -62,8 +62,13 @@ struct cb_recallargs { extern unsigned nfs4_callback_getattr(struct cb_getattrargs *args, struct cb_getattrres *res); extern unsigned nfs4_callback_recall(struct cb_recallargs *args, void *dummy); +#ifdef CONFIG_NFS_V4 extern int nfs_callback_up(void); -extern int nfs_callback_down(void); +extern void nfs_callback_down(void); +#else +#define nfs_callback_up() (0) +#define nfs_callback_down() do {} while(0) +#endif extern unsigned int nfs_callback_set_tcpport; extern unsigned short nfs_callback_tcpport; diff --git a/fs/nfs/callback_proc.c b/fs/nfs/callback_proc.c index 462cfceb50c..97cf8f71451 100644 --- a/fs/nfs/callback_proc.c +++ b/fs/nfs/callback_proc.c @@ -5,25 +5,25 @@ * * NFSv4 callback procedures */ -#include <linux/config.h> #include <linux/nfs4.h> #include <linux/nfs_fs.h> #include "nfs4_fs.h" #include "callback.h" #include "delegation.h" +#include "internal.h" #define NFSDBG_FACILITY NFSDBG_CALLBACK unsigned nfs4_callback_getattr(struct cb_getattrargs *args, struct cb_getattrres *res) { - struct nfs4_client *clp; + struct nfs_client *clp; struct nfs_delegation *delegation; struct nfs_inode *nfsi; struct inode *inode; res->bitmap[0] = res->bitmap[1] = 0; res->status = htonl(NFS4ERR_BADHANDLE); - clp = nfs4_find_client(&args->addr->sin_addr); + clp = nfs_find_client(args->addr, 4); if (clp == NULL) goto out; inode = nfs_delegation_find_inode(clp, &args->fh); @@ -49,7 +49,7 @@ out_iput: up_read(&nfsi->rwsem); iput(inode); out_putclient: - nfs4_put_client(clp); + nfs_put_client(clp); out: dprintk("%s: exit with status = %d\n", __FUNCTION__, ntohl(res->status)); return res->status; @@ -57,12 +57,12 @@ out: unsigned nfs4_callback_recall(struct cb_recallargs *args, void *dummy) { - struct nfs4_client *clp; + struct nfs_client *clp; struct inode *inode; unsigned res; res = htonl(NFS4ERR_BADHANDLE); - clp = nfs4_find_client(&args->addr->sin_addr); + clp = nfs_find_client(args->addr, 4); if (clp == NULL) goto out; inode = nfs_delegation_find_inode(clp, &args->fh); @@ -81,7 +81,7 @@ unsigned nfs4_callback_recall(struct cb_recallargs *args, void *dummy) } iput(inode); out_putclient: - nfs4_put_client(clp); + nfs_put_client(clp); out: dprintk("%s: exit with status = %d\n", __FUNCTION__, ntohl(res)); return res; diff --git a/fs/nfs/callback_xdr.c b/fs/nfs/callback_xdr.c index 05c38cf40b6..29f93219205 100644 --- a/fs/nfs/callback_xdr.c +++ b/fs/nfs/callback_xdr.c @@ -5,7 +5,6 @@ * * NFSv4 callback encode/decode procedures */ -#include <linux/config.h> #include <linux/kernel.h> #include <linux/sunrpc/svc.h> #include <linux/nfs4.h> @@ -202,7 +201,7 @@ static unsigned decode_recall_args(struct svc_rqst *rqstp, struct xdr_stream *xd status = decode_fh(xdr, &args->fh); out: dprintk("%s: exit with status = %d\n", __FUNCTION__, status); - return 0; + return status; } static unsigned encode_string(struct xdr_stream *xdr, unsigned int len, const char *str) diff --git a/fs/nfs/client.c b/fs/nfs/client.c new file mode 100644 index 00000000000..ec1938d4b81 --- /dev/null +++ b/fs/nfs/client.c @@ -0,0 +1,1448 @@ +/* client.c: NFS client sharing and management code + * + * Copyright (C) 2006 Red Hat, Inc. All Rights Reserved. + * Written by David Howells (dhowells@redhat.com) + * + * 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. + */ + + +#include <linux/config.h> +#include <linux/module.h> +#include <linux/init.h> + +#include <linux/time.h> +#include <linux/kernel.h> +#include <linux/mm.h> +#include <linux/string.h> +#include <linux/stat.h> +#include <linux/errno.h> +#include <linux/unistd.h> +#include <linux/sunrpc/clnt.h> +#include <linux/sunrpc/stats.h> +#include <linux/sunrpc/metrics.h> +#include <linux/nfs_fs.h> +#include <linux/nfs_mount.h> +#include <linux/nfs4_mount.h> +#include <linux/lockd/bind.h> +#include <linux/smp_lock.h> +#include <linux/seq_file.h> +#include <linux/mount.h> +#include <linux/nfs_idmap.h> +#include <linux/vfs.h> +#include <linux/inet.h> +#include <linux/nfs_xdr.h> + +#include <asm/system.h> + +#include "nfs4_fs.h" +#include "callback.h" +#include "delegation.h" +#include "iostat.h" +#include "internal.h" + +#define NFSDBG_FACILITY NFSDBG_CLIENT + +static DEFINE_SPINLOCK(nfs_client_lock); +static LIST_HEAD(nfs_client_list); +static LIST_HEAD(nfs_volume_list); +static DECLARE_WAIT_QUEUE_HEAD(nfs_client_active_wq); + +/* + * RPC cruft for NFS + */ +static struct rpc_version *nfs_version[5] = { + [2] = &nfs_version2, +#ifdef CONFIG_NFS_V3 + [3] = &nfs_version3, +#endif +#ifdef CONFIG_NFS_V4 + [4] = &nfs_version4, +#endif +}; + +struct rpc_program nfs_program = { + .name = "nfs", + .number = NFS_PROGRAM, + .nrvers = ARRAY_SIZE(nfs_version), + .version = nfs_version, + .stats = &nfs_rpcstat, + .pipe_dir_name = "/nfs", +}; + +struct rpc_stat nfs_rpcstat = { + .program = &nfs_program +}; + + +#ifdef CONFIG_NFS_V3_ACL +static struct rpc_stat nfsacl_rpcstat = { &nfsacl_program }; +static struct rpc_version * nfsacl_version[] = { + [3] = &nfsacl_version3, +}; + +struct rpc_program nfsacl_program = { + .name = "nfsacl", + .number = NFS_ACL_PROGRAM, + .nrvers = ARRAY_SIZE(nfsacl_version), + .version = nfsacl_version, + .stats = &nfsacl_rpcstat, +}; +#endif /* CONFIG_NFS_V3_ACL */ + +/* + * Allocate a shared client record + * + * Since these are allocated/deallocated very rarely, we don't + * bother putting them in a slab cache... + */ +static struct nfs_client *nfs_alloc_client(const char *hostname, + const struct sockaddr_in *addr, + int nfsversion) +{ + struct nfs_client *clp; + int error; + + if ((clp = kzalloc(sizeof(*clp), GFP_KERNEL)) == NULL) + goto error_0; + + error = rpciod_up(); + if (error < 0) { + dprintk("%s: couldn't start rpciod! Error = %d\n", + __FUNCTION__, error); + goto error_1; + } + __set_bit(NFS_CS_RPCIOD, &clp->cl_res_state); + + if (nfsversion == 4) { + if (nfs_callback_up() < 0) + goto error_2; + __set_bit(NFS_CS_CALLBACK, &clp->cl_res_state); + } + + atomic_set(&clp->cl_count, 1); + clp->cl_cons_state = NFS_CS_INITING; + + clp->cl_nfsversion = nfsversion; + memcpy(&clp->cl_addr, addr, sizeof(clp->cl_addr)); + + if (hostname) { + clp->cl_hostname = kstrdup(hostname, GFP_KERNEL); + if (!clp->cl_hostname) + goto error_3; + } + + INIT_LIST_HEAD(&clp->cl_superblocks); + clp->cl_rpcclient = ERR_PTR(-EINVAL); + +#ifdef CONFIG_NFS_V4 + init_rwsem(&clp->cl_sem); + INIT_LIST_HEAD(&clp->cl_delegations); + INIT_LIST_HEAD(&clp->cl_state_owners); + INIT_LIST_HEAD(&clp->cl_unused); + spin_lock_init(&clp->cl_lock); + INIT_WORK(&clp->cl_renewd, nfs4_renew_state, clp); + rpc_init_wait_queue(&clp->cl_rpcwaitq, "NFS client"); + clp->cl_boot_time = CURRENT_TIME; + clp->cl_state = 1 << NFS4CLNT_LEASE_EXPIRED; +#endif + + return clp; + +error_3: + if (__test_and_clear_bit(NFS_CS_CALLBACK, &clp->cl_res_state)) + nfs_callback_down(); +error_2: + rpciod_down(); + __clear_bit(NFS_CS_RPCIOD, &clp->cl_res_state); +error_1: + kfree(clp); +error_0: + return NULL; +} + +static void nfs4_shutdown_client(struct nfs_client *clp) +{ +#ifdef CONFIG_NFS_V4 + if (__test_and_clear_bit(NFS_CS_RENEWD, &clp->cl_res_state)) + nfs4_kill_renewd(clp); + while (!list_empty(&clp->cl_unused)) { + struct nfs4_state_owner *sp; + + sp = list_entry(clp->cl_unused.next, + struct nfs4_state_owner, + so_list); + list_del(&sp->so_list); + kfree(sp); + } + BUG_ON(!list_empty(&clp->cl_state_owners)); + if (__test_and_clear_bit(NFS_CS_IDMAP, &clp->cl_res_state)) + nfs_idmap_delete(clp); +#endif +} + +/* + * Destroy a shared client record + */ +static void nfs_free_client(struct nfs_client *clp) +{ + dprintk("--> nfs_free_client(%d)\n", clp->cl_nfsversion); + + nfs4_shutdown_client(clp); + + /* -EIO all pending I/O */ + if (!IS_ERR(clp->cl_rpcclient)) + rpc_shutdown_client(clp->cl_rpcclient); + + if (__test_and_clear_bit(NFS_CS_CALLBACK, &clp->cl_res_state)) + nfs_callback_down(); + + if (__test_and_clear_bit(NFS_CS_RPCIOD, &clp->cl_res_state)) + rpciod_down(); + + kfree(clp->cl_hostname); + kfree(clp); + + dprintk("<-- nfs_free_client()\n"); +} + +/* + * Release a reference to a shared client record + */ +void nfs_put_client(struct nfs_client *clp) +{ + if (!clp) + return; + + dprintk("--> nfs_put_client({%d})\n", atomic_read(&clp->cl_count)); + + if (atomic_dec_and_lock(&clp->cl_count, &nfs_client_lock)) { + list_del(&clp->cl_share_link); + spin_unlock(&nfs_client_lock); + + BUG_ON(!list_empty(&clp->cl_superblocks)); + + nfs_free_client(clp); + } +} + +/* + * Find a client by address + * - caller must hold nfs_client_lock + */ +static struct nfs_client *__nfs_find_client(const struct sockaddr_in *addr, int nfsversion) +{ + struct nfs_client *clp; + + list_for_each_entry(clp, &nfs_client_list, cl_share_link) { + /* Different NFS versions cannot share the same nfs_client */ + if (clp->cl_nfsversion != nfsversion) + continue; + + if (memcmp(&clp->cl_addr.sin_addr, &addr->sin_addr, + sizeof(clp->cl_addr.sin_addr)) != 0) + continue; + + if (clp->cl_addr.sin_port == addr->sin_port) + goto found; + } + + return NULL; + +found: + atomic_inc(&clp->cl_count); + return clp; +} + +/* + * Find a client by IP address and protocol version + * - returns NULL if no such client + */ +struct nfs_client *nfs_find_client(const struct sockaddr_in *addr, int nfsversion) +{ + struct nfs_client *clp; + + spin_lock(&nfs_client_lock); + clp = __nfs_find_client(addr, nfsversion); + spin_unlock(&nfs_client_lock); + + BUG_ON(clp && clp->cl_cons_state == 0); + + return clp; +} + +/* + * Look up a client by IP address and protocol version + * - creates a new record if one doesn't yet exist + */ +static struct nfs_client *nfs_get_client(const char *hostname, + const struct sockaddr_in *addr, + int nfsversion) +{ + struct nfs_client *clp, *new = NULL; + int error; + + dprintk("--> nfs_get_client(%s,"NIPQUAD_FMT":%d,%d)\n", + hostname ?: "", NIPQUAD(addr->sin_addr), + addr->sin_port, nfsversion); + + /* see if the client already exists */ + do { + spin_lock(&nfs_client_lock); + + clp = __nfs_find_client(addr, nfsversion); + if (clp) + goto found_client; + if (new) + goto install_client; + + spin_unlock(&nfs_client_lock); + + new = nfs_alloc_client(hostname, addr, nfsversion); + } while (new); + + return ERR_PTR(-ENOMEM); + + /* install a new client and return with it unready */ +install_client: + clp = new; + list_add(&clp->cl_share_link, &nfs_client_list); + spin_unlock(&nfs_client_lock); + dprintk("--> nfs_get_client() = %p [new]\n", clp); + return clp; + + /* found an existing client + * - make sure it's ready before returning + */ +found_client: + spin_unlock(&nfs_client_lock); + + if (new) + nfs_free_client(new); + + if (clp->cl_cons_state == NFS_CS_INITING) { + DECLARE_WAITQUEUE(myself, current); + + add_wait_queue(&nfs_client_active_wq, &myself); + + for (;;) { + set_current_state(TASK_INTERRUPTIBLE); + if (signal_pending(current) || + clp->cl_cons_state > NFS_CS_READY) + break; + schedule(); + } + + remove_wait_queue(&nfs_client_active_wq, &myself); + + if (signal_pending(current)) { + nfs_put_client(clp); + return ERR_PTR(-ERESTARTSYS); + } + } + + if (clp->cl_cons_state < NFS_CS_READY) { + error = clp->cl_cons_state; + nfs_put_client(clp); + return ERR_PTR(error); + } + + BUG_ON(clp->cl_cons_state != NFS_CS_READY); + + dprintk("--> nfs_get_client() = %p [share]\n", clp); + return clp; +} + +/* + * Mark a server as ready or failed + */ +static void nfs_mark_client_ready(struct nfs_client *clp, int state) +{ + clp->cl_cons_state = state; + wake_up_all(&nfs_client_active_wq); +} + +/* + * Initialise the timeout values for a connection + */ +static void nfs_init_timeout_values(struct rpc_timeout *to, int proto, + unsigned int timeo, unsigned int retrans) +{ + to->to_initval = timeo * HZ / 10; + to->to_retries = retrans; + if (!to->to_retries) + to->to_retries = 2; + + switch (proto) { + case IPPROTO_TCP: + if (!to->to_initval) + to->to_initval = 60 * HZ; + if (to->to_initval > NFS_MAX_TCP_TIMEOUT) + to->to_initval = NFS_MAX_TCP_TIMEOUT; + to->to_increment = to->to_initval; + to->to_maxval = to->to_initval + (to->to_increment * to->to_retries); + to->to_exponential = 0; + break; + case IPPROTO_UDP: + default: + if (!to->to_initval) + to->to_initval = 11 * HZ / 10; + if (to->to_initval > NFS_MAX_UDP_TIMEOUT) + to->to_initval = NFS_MAX_UDP_TIMEOUT; + to->to_maxval = NFS_MAX_UDP_TIMEOUT; + to->to_exponential = 1; + break; + } +} + +/* + * Create an RPC client handle + */ +static int nfs_create_rpc_client(struct nfs_client *clp, int proto, + unsigned int timeo, + unsigned int retrans, + rpc_authflavor_t flavor) +{ + struct rpc_timeout timeparms; + struct rpc_clnt *clnt = NULL; + struct rpc_create_args args = { + .protocol = proto, + .address = (struct sockaddr *)&clp->cl_addr, + .addrsize = sizeof(clp->cl_addr), + .timeout = &timeparms, + .servername = clp->cl_hostname, + .program = &nfs_program, + .version = clp->rpc_ops->version, + .authflavor = flavor, + }; + + if (!IS_ERR(clp->cl_rpcclient)) + return 0; + + nfs_init_timeout_values(&timeparms, proto, timeo, retrans); + clp->retrans_timeo = timeparms.to_initval; + clp->retrans_count = timeparms.to_retries; + + clnt = rpc_create(&args); + if (IS_ERR(clnt)) { + dprintk("%s: cannot create RPC client. Error = %ld\n", + __FUNCTION__, PTR_ERR(clnt)); + return PTR_ERR(clnt); + } + + clp->cl_rpcclient = clnt; + return 0; +} + +/* + * Version 2 or 3 client destruction + */ +static void nfs_destroy_server(struct nfs_server *server) +{ + if (!IS_ERR(server->client_acl)) + rpc_shutdown_client(server->client_acl); + + if (!(server->flags & NFS_MOUNT_NONLM)) + lockd_down(); /* release rpc.lockd */ +} + +/* + * Version 2 or 3 lockd setup + */ +static int nfs_start_lockd(struct nfs_server *server) +{ + int error = 0; + + if (server->nfs_client->cl_nfsversion > 3) + goto out; + if (server->flags & NFS_MOUNT_NONLM) + goto out; + error = lockd_up(); + if (error < 0) + server->flags |= NFS_MOUNT_NONLM; + else + server->destroy = nfs_destroy_server; +out: + return error; +} + +/* + * Initialise an NFSv3 ACL client connection + */ +#ifdef CONFIG_NFS_V3_ACL +static void nfs_init_server_aclclient(struct nfs_server *server) +{ + if (server->nfs_client->cl_nfsversion != 3) + goto out_noacl; + if (server->flags & NFS_MOUNT_NOACL) + goto out_noacl; + + server->client_acl = rpc_bind_new_program(server->client, &nfsacl_program, 3); + if (IS_ERR(server->client_acl)) + goto out_noacl; + + /* No errors! Assume that Sun nfsacls are supported */ + server->caps |= NFS_CAP_ACLS; + return; + +out_noacl: + server->caps &= ~NFS_CAP_ACLS; +} +#else +static inline void nfs_init_server_aclclient(struct nfs_server *server) +{ + server->flags &= ~NFS_MOUNT_NOACL; + server->caps &= ~NFS_CAP_ACLS; +} +#endif + +/* + * Create a general RPC client + */ +static int nfs_init_server_rpcclient(struct nfs_server *server, rpc_authflavor_t pseudoflavour) +{ + struct nfs_client *clp = server->nfs_client; + + server->client = rpc_clone_client(clp->cl_rpcclient); + if (IS_ERR(server->client)) { + dprintk("%s: couldn't create rpc_client!\n", __FUNCTION__); + return PTR_ERR(server->client); + } + + if (pseudoflavour != clp->cl_rpcclient->cl_auth->au_flavor) { + struct rpc_auth *auth; + + auth = rpcauth_create(pseudoflavour, server->client); + if (IS_ERR(auth)) { + dprintk("%s: couldn't create credcache!\n", __FUNCTION__); + return PTR_ERR(auth); + } + } + server->client->cl_softrtry = 0; + if (server->flags & NFS_MOUNT_SOFT) + server->client->cl_softrtry = 1; + + server->client->cl_intr = 0; + if (server->flags & NFS4_MOUNT_INTR) + server->client->cl_intr = 1; + + return 0; +} + +/* + * Initialise an NFS2 or NFS3 client + */ +static int nfs_init_client(struct nfs_client *clp, const struct nfs_mount_data *data) +{ + int proto = (data->flags & NFS_MOUNT_TCP) ? IPPROTO_TCP : IPPROTO_UDP; + int error; + + if (clp->cl_cons_state == NFS_CS_READY) { + /* the client is already initialised */ + dprintk("<-- nfs_init_client() = 0 [already %p]\n", clp); + return 0; + } + + /* Check NFS protocol revision and initialize RPC op vector */ + clp->rpc_ops = &nfs_v2_clientops; +#ifdef CONFIG_NFS_V3 + if (clp->cl_nfsversion == 3) + clp->rpc_ops = &nfs_v3_clientops; +#endif + /* + * Create a client RPC handle for doing FSSTAT with UNIX auth only + * - RFC 2623, sec 2.3.2 + */ + error = nfs_create_rpc_client(clp, proto, data->timeo, data->retrans, + RPC_AUTH_UNIX); + if (error < 0) + goto error; + nfs_mark_client_ready(clp, NFS_CS_READY); + return 0; + +error: + nfs_mark_client_ready(clp, error); + dprintk("<-- nfs_init_client() = xerror %d\n", error); + return error; +} + +/* + * Create a version 2 or 3 client + */ +static int nfs_init_server(struct nfs_server *server, const struct nfs_mount_data *data) +{ + struct nfs_client *clp; + int error, nfsvers = 2; + + dprintk("--> nfs_init_server()\n"); + +#ifdef CONFIG_NFS_V3 + if (data->flags & NFS_MOUNT_VER3) + nfsvers = 3; +#endif + + /* Allocate or find a client reference we can use */ + clp = nfs_get_client(data->hostname, &data->addr, nfsvers); + if (IS_ERR(clp)) { + dprintk("<-- nfs_init_server() = error %ld\n", PTR_ERR(clp)); + return PTR_ERR(clp); + } + + error = nfs_init_client(clp, data); + if (error < 0) + goto error; + + server->nfs_client = clp; + + /* Initialise the client representation from the mount data */ + server->flags = data->flags & NFS_MOUNT_FLAGMASK; + + if (data->rsize) + server->rsize = nfs_block_size(data->rsize, NULL); + if (data->wsize) + server->wsize = nfs_block_size(data->wsize, NULL); + + server->acregmin = data->acregmin * HZ; + server->acregmax = data->acregmax * HZ; + server->acdirmin = data->acdirmin * HZ; + server->acdirmax = data->acdirmax * HZ; + + /* Start lockd here, before we might error out */ + error = nfs_start_lockd(server); + if (error < 0) + goto error; + + error = nfs_init_server_rpcclient(server, data->pseudoflavor); + if (error < 0) + goto error; + + server->namelen = data->namlen; + /* Create a client RPC handle for the NFSv3 ACL management interface */ + nfs_init_server_aclclient(server); + if (clp->cl_nfsversion == 3) { + if (server->namelen == 0 || server->namelen > NFS3_MAXNAMLEN) + server->namelen = NFS3_MAXNAMLEN; + server->caps |= NFS_CAP_READDIRPLUS; + } else { + if (server->namelen == 0 || server->namelen > NFS2_MAXNAMLEN) + server->namelen = NFS2_MAXNAMLEN; + } + + dprintk("<-- nfs_init_server() = 0 [new %p]\n", clp); + return 0; + +error: + server->nfs_client = NULL; + nfs_put_client(clp); + dprintk("<-- nfs_init_server() = xerror %d\n", error); + return error; +} + +/* + * Load up the server record from information gained in an fsinfo record + */ +static void nfs_server_set_fsinfo(struct nfs_server *server, struct nfs_fsinfo *fsinfo) +{ + unsigned long max_rpc_payload; + + /* Work out a lot of parameters */ + if (server->rsize == 0) + server->rsize = nfs_block_size(fsinfo->rtpref, NULL); + if (server->wsize == 0) + server->wsize = nfs_block_size(fsinfo->wtpref, NULL); + + if (fsinfo->rtmax >= 512 && server->rsize > fsinfo->rtmax) + server->rsize = nfs_block_size(fsinfo->rtmax, NULL); + if (fsinfo->wtmax >= 512 && server->wsize > fsinfo->wtmax) + server->wsize = nfs_block_size(fsinfo->wtmax, NULL); + + max_rpc_payload = nfs_block_size(rpc_max_payload(server->client), NULL); + if (server->rsize > max_rpc_payload) + server->rsize = max_rpc_payload; + if (server->rsize > NFS_MAX_FILE_IO_SIZE) + server->rsize = NFS_MAX_FILE_IO_SIZE; + server->rpages = (server->rsize + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT; + server->backing_dev_info.ra_pages = server->rpages * NFS_MAX_READAHEAD; + + if (server->wsize > max_rpc_payload) + server->wsize = max_rpc_payload; + if (server->wsize > NFS_MAX_FILE_IO_SIZE) + server->wsize = NFS_MAX_FILE_IO_SIZE; + server->wpages = (server->wsize + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT; + server->wtmult = nfs_block_bits(fsinfo->wtmult, NULL); + + server->dtsize = nfs_block_size(fsinfo->dtpref, NULL); + if (server->dtsize > PAGE_CACHE_SIZE) + server->dtsize = PAGE_CACHE_SIZE; + if (server->dtsize > server->rsize) + server->dtsize = server->rsize; + + if (server->flags & NFS_MOUNT_NOAC) { + server->acregmin = server->acregmax = 0; + server->acdirmin = server->acdirmax = 0; + } + + server->maxfilesize = fsinfo->maxfilesize; + + /* We're airborne Set socket buffersize */ + rpc_setbufsize(server->client, server->wsize + 100, server->rsize + 100); +} + +/* + * Probe filesystem information, including the FSID on v2/v3 + */ +static int nfs_probe_fsinfo(struct nfs_server *server, struct nfs_fh *mntfh, struct nfs_fattr *fattr) +{ + struct nfs_fsinfo fsinfo; + struct nfs_client *clp = server->nfs_client; + int error; + + dprintk("--> nfs_probe_fsinfo()\n"); + + if (clp->rpc_ops->set_capabilities != NULL) { + error = clp->rpc_ops->set_capabilities(server, mntfh); + if (error < 0) + goto out_error; + } + + fsinfo.fattr = fattr; + nfs_fattr_init(fattr); + error = clp->rpc_ops->fsinfo(server, mntfh, &fsinfo); + if (error < 0) + goto out_error; + + nfs_server_set_fsinfo(server, &fsinfo); + + /* Get some general file system info */ + if (server->namelen == 0) { + struct nfs_pathconf pathinfo; + + pathinfo.fattr = fattr; + nfs_fattr_init(fattr); + + if (clp->rpc_ops->pathconf(server, mntfh, &pathinfo) >= 0) + server->namelen = pathinfo.max_namelen; + } + + dprintk("<-- nfs_probe_fsinfo() = 0\n"); + return 0; + +out_error: + dprintk("nfs_probe_fsinfo: error = %d\n", -error); + return error; +} + +/* + * Copy useful information when duplicating a server record + */ +static void nfs_server_copy_userdata(struct nfs_server *target, struct nfs_server *source) +{ + target->flags = source->flags; + target->acregmin = source->acregmin; + target->acregmax = source->acregmax; + target->acdirmin = source->acdirmin; + target->acdirmax = source->acdirmax; + target->caps = source->caps; +} + +/* + * Allocate and initialise a server record + */ +static struct nfs_server *nfs_alloc_server(void) +{ + struct nfs_server *server; + + server = kzalloc(sizeof(struct nfs_server), GFP_KERNEL); + if (!server) + return NULL; + + server->client = server->client_acl = ERR_PTR(-EINVAL); + + /* Zero out the NFS state stuff */ + INIT_LIST_HEAD(&server->client_link); + INIT_LIST_HEAD(&server->master_link); + + server->io_stats = nfs_alloc_iostats(); + if (!server->io_stats) { + kfree(server); + return NULL; + } + + return server; +} + +/* + * Free up a server record + */ +void nfs_free_server(struct nfs_server *server) +{ + dprintk("--> nfs_free_server()\n"); + + spin_lock(&nfs_client_lock); + list_del(&server->client_link); + list_del(&server->master_link); + spin_unlock(&nfs_client_lock); + + if (server->destroy != NULL) + server->destroy(server); + if (!IS_ERR(server->client)) + rpc_shutdown_client(server->client); + + nfs_put_client(server->nfs_client); + + nfs_free_iostats(server->io_stats); + kfree(server); + nfs_release_automount_timer(); + dprintk("<-- nfs_free_server()\n"); +} + +/* + * Create a version 2 or 3 volume record + * - keyed on server and FSID + */ +struct nfs_server *nfs_create_server(const struct nfs_mount_data *data, + struct nfs_fh *mntfh) +{ + struct nfs_server *server; + struct nfs_fattr fattr; + int error; + + server = nfs_alloc_server(); + if (!server) + return ERR_PTR(-ENOMEM); + + /* Get a client representation */ + error = nfs_init_server(server, data); + if (error < 0) + goto error; + + BUG_ON(!server->nfs_client); + BUG_ON(!server->nfs_client->rpc_ops); + BUG_ON(!server->nfs_client->rpc_ops->file_inode_ops); + + /* Probe the root fh to retrieve its FSID */ + error = nfs_probe_fsinfo(server, mntfh, &fattr); + if (error < 0) + goto error; + if (!(fattr.valid & NFS_ATTR_FATTR)) { + error = server->nfs_client->rpc_ops->getattr(server, mntfh, &fattr); + if (error < 0) { + dprintk("nfs_create_server: getattr error = %d\n", -error); + goto error; + } + } + memcpy(&server->fsid, &fattr.fsid, sizeof(server->fsid)); + + dprintk("Server FSID: %llx:%llx\n", + (unsigned long long) server->fsid.major, + (unsigned long long) server->fsid.minor); + + BUG_ON(!server->nfs_client); + BUG_ON(!server->nfs_client->rpc_ops); + BUG_ON(!server->nfs_client->rpc_ops->file_inode_ops); + + spin_lock(&nfs_client_lock); + list_add_tail(&server->client_link, &server->nfs_client->cl_superblocks); + list_add_tail(&server->master_link, &nfs_volume_list); + spin_unlock(&nfs_client_lock); + + server->mount_time = jiffies; + return server; + +error: + nfs_free_server(server); + return ERR_PTR(error); +} + +#ifdef CONFIG_NFS_V4 +/* + * Initialise an NFS4 client record + */ +static int nfs4_init_client(struct nfs_client *clp, + int proto, int timeo, int retrans, + rpc_authflavor_t authflavour) +{ + int error; + + if (clp->cl_cons_state == NFS_CS_READY) { + /* the client is initialised already */ + dprintk("<-- nfs4_init_client() = 0 [already %p]\n", clp); + return 0; + } + + /* Check NFS protocol revision and initialize RPC op vector */ + clp->rpc_ops = &nfs_v4_clientops; + + error = nfs_create_rpc_client(clp, proto, timeo, retrans, authflavour); + if (error < 0) + goto error; + + error = nfs_idmap_new(clp); + if (error < 0) { + dprintk("%s: failed to create idmapper. Error = %d\n", + __FUNCTION__, error); + goto error; + } + __set_bit(NFS_CS_IDMAP, &clp->cl_res_state); + + nfs_mark_client_ready(clp, NFS_CS_READY); + return 0; + +error: + nfs_mark_client_ready(clp, error); + dprintk("<-- nfs4_init_client() = xerror %d\n", error); + return error; +} + +/* + * Set up an NFS4 client + */ +static int nfs4_set_client(struct nfs_server *server, + const char *hostname, const struct sockaddr_in *addr, + rpc_authflavor_t authflavour, + int proto, int timeo, int retrans) +{ + struct nfs_client *clp; + int error; + + dprintk("--> nfs4_set_client()\n"); + + /* Allocate or find a client reference we can use */ + clp = nfs_get_client(hostname, addr, 4); + if (IS_ERR(clp)) { + error = PTR_ERR(clp); + goto error; + } + error = nfs4_init_client(clp, proto, timeo, retrans, authflavour); + if (error < 0) + goto error_put; + + server->nfs_client = clp; + dprintk("<-- nfs4_set_client() = 0 [new %p]\n", clp); + return 0; + +error_put: + nfs_put_client(clp); +error: + dprintk("<-- nfs4_set_client() = xerror %d\n", error); + return error; +} + +/* + * Create a version 4 volume record + */ +static int nfs4_init_server(struct nfs_server *server, + const struct nfs4_mount_data *data, rpc_authflavor_t authflavour) +{ + int error; + + dprintk("--> nfs4_init_server()\n"); + + /* Initialise the client representation from the mount data */ + server->flags = data->flags & NFS_MOUNT_FLAGMASK; + server->caps |= NFS_CAP_ATOMIC_OPEN; + + if (data->rsize) + server->rsize = nfs_block_size(data->rsize, NULL); + if (data->wsize) + server->wsize = nfs_block_size(data->wsize, NULL); + + server->acregmin = data->acregmin * HZ; + server->acregmax = data->acregmax * HZ; + server->acdirmin = data->acdirmin * HZ; + server->acdirmax = data->acdirmax * HZ; + + error = nfs_init_server_rpcclient(server, authflavour); + + /* Done */ + dprintk("<-- nfs4_init_server() = %d\n", error); + return error; +} + +/* + * Create a version 4 volume record + * - keyed on server and FSID + */ +struct nfs_server *nfs4_create_server(const struct nfs4_mount_data *data, + const char *hostname, + const struct sockaddr_in *addr, + const char *mntpath, + const char *ip_addr, + rpc_authflavor_t authflavour, + struct nfs_fh *mntfh) +{ + struct nfs_fattr fattr; + struct nfs_server *server; + int error; + + dprintk("--> nfs4_create_server()\n"); + + server = nfs_alloc_server(); + if (!server) + return ERR_PTR(-ENOMEM); + + /* Get a client record */ + error = nfs4_set_client(server, hostname, addr, authflavour, + data->proto, data->timeo, data->retrans); + if (error < 0) + goto error; + + /* set up the general RPC client */ + error = nfs4_init_server(server, data, authflavour); + if (error < 0) + goto error; + + BUG_ON(!server->nfs_client); + BUG_ON(!server->nfs_client->rpc_ops); + BUG_ON(!server->nfs_client->rpc_ops->file_inode_ops); + + /* Probe the root fh to retrieve its FSID */ + error = nfs4_path_walk(server, mntfh, mntpath); + if (error < 0) + goto error; + + dprintk("Server FSID: %llx:%llx\n", + (unsigned long long) server->fsid.major, + (unsigned long long) server->fsid.minor); + dprintk("Mount FH: %d\n", mntfh->size); + + error = nfs_probe_fsinfo(server, mntfh, &fattr); + if (error < 0) + goto error; + + BUG_ON(!server->nfs_client); + BUG_ON(!server->nfs_client->rpc_ops); + BUG_ON(!server->nfs_client->rpc_ops->file_inode_ops); + + spin_lock(&nfs_client_lock); + list_add_tail(&server->client_link, &server->nfs_client->cl_superblocks); + list_add_tail(&server->master_link, &nfs_volume_list); + spin_unlock(&nfs_client_lock); + + server->mount_time = jiffies; + dprintk("<-- nfs4_create_server() = %p\n", server); + return server; + +error: + nfs_free_server(server); + dprintk("<-- nfs4_create_server() = error %d\n", error); + return ERR_PTR(error); +} + +/* + * Create an NFS4 referral server record + */ +struct nfs_server *nfs4_create_referral_server(struct nfs_clone_mount *data, + struct nfs_fh *fh) +{ + struct nfs_client *parent_client; + struct nfs_server *server, *parent_server; + struct nfs_fattr fattr; + int error; + + dprintk("--> nfs4_create_referral_server()\n"); + + server = nfs_alloc_server(); + if (!server) + return ERR_PTR(-ENOMEM); + + parent_server = NFS_SB(data->sb); + parent_client = parent_server->nfs_client; + + /* Get a client representation. + * Note: NFSv4 always uses TCP, */ + error = nfs4_set_client(server, data->hostname, data->addr, + data->authflavor, + parent_server->client->cl_xprt->prot, + parent_client->retrans_timeo, + parent_client->retrans_count); + if (error < 0) + goto error; + + /* Initialise the client representation from the parent server */ + nfs_server_copy_userdata(server, parent_server); + server->caps |= NFS_CAP_ATOMIC_OPEN; + + error = nfs_init_server_rpcclient(server, data->authflavor); + if (error < 0) + goto error; + + BUG_ON(!server->nfs_client); + BUG_ON(!server->nfs_client->rpc_ops); + BUG_ON(!server->nfs_client->rpc_ops->file_inode_ops); + + /* probe the filesystem info for this server filesystem */ + error = nfs_probe_fsinfo(server, fh, &fattr); + if (error < 0) + goto error; + + dprintk("Referral FSID: %llx:%llx\n", + (unsigned long long) server->fsid.major, + (unsigned long long) server->fsid.minor); + + spin_lock(&nfs_client_lock); + list_add_tail(&server->client_link, &server->nfs_client->cl_superblocks); + list_add_tail(&server->master_link, &nfs_volume_list); + spin_unlock(&nfs_client_lock); + + server->mount_time = jiffies; + + dprintk("<-- nfs_create_referral_server() = %p\n", server); + return server; + +error: + nfs_free_server(server); + dprintk("<-- nfs4_create_referral_server() = error %d\n", error); + return ERR_PTR(error); +} + +#endif /* CONFIG_NFS_V4 */ + +/* + * Clone an NFS2, NFS3 or NFS4 server record + */ +struct nfs_server *nfs_clone_server(struct nfs_server *source, + struct nfs_fh *fh, + struct nfs_fattr *fattr) +{ + struct nfs_server *server; + struct nfs_fattr fattr_fsinfo; + int error; + + dprintk("--> nfs_clone_server(,%llx:%llx,)\n", + (unsigned long long) fattr->fsid.major, + (unsigned long long) fattr->fsid.minor); + + server = nfs_alloc_server(); + if (!server) + return ERR_PTR(-ENOMEM); + + /* Copy data from the source */ + server->nfs_client = source->nfs_client; + atomic_inc(&server->nfs_client->cl_count); + nfs_server_copy_userdata(server, source); + + server->fsid = fattr->fsid; + + error = nfs_init_server_rpcclient(server, source->client->cl_auth->au_flavor); + if (error < 0) + goto out_free_server; + if (!IS_ERR(source->client_acl)) + nfs_init_server_aclclient(server); + + /* probe the filesystem info for this server filesystem */ + error = nfs_probe_fsinfo(server, fh, &fattr_fsinfo); + if (error < 0) + goto out_free_server; + + dprintk("Cloned FSID: %llx:%llx\n", + (unsigned long long) server->fsid.major, + (unsigned long long) server->fsid.minor); + + error = nfs_start_lockd(server); + if (error < 0) + goto out_free_server; + + spin_lock(&nfs_client_lock); + list_add_tail(&server->client_link, &server->nfs_client->cl_superblocks); + list_add_tail(&server->master_link, &nfs_volume_list); + spin_unlock(&nfs_client_lock); + + server->mount_time = jiffies; + + dprintk("<-- nfs_clone_server() = %p\n", server); + return server; + +out_free_server: + nfs_free_server(server); + dprintk("<-- nfs_clone_server() = error %d\n", error); + return ERR_PTR(error); +} + +#ifdef CONFIG_PROC_FS +static struct proc_dir_entry *proc_fs_nfs; + +static int nfs_server_list_open(struct inode *inode, struct file *file); +static void *nfs_server_list_start(struct seq_file *p, loff_t *pos); +static void *nfs_server_list_next(struct seq_file *p, void *v, loff_t *pos); +static void nfs_server_list_stop(struct seq_file *p, void *v); +static int nfs_server_list_show(struct seq_file *m, void *v); + +static struct seq_operations nfs_server_list_ops = { + .start = nfs_server_list_start, + .next = nfs_server_list_next, + .stop = nfs_server_list_stop, + .show = nfs_server_list_show, +}; + +static struct file_operations nfs_server_list_fops = { + .open = nfs_server_list_open, + .read = seq_read, + .llseek = seq_lseek, + .release = seq_release, +}; + +static int nfs_volume_list_open(struct inode *inode, struct file *file); +static void *nfs_volume_list_start(struct seq_file *p, loff_t *pos); +static void *nfs_volume_list_next(struct seq_file *p, void *v, loff_t *pos); +static void nfs_volume_list_stop(struct seq_file *p, void *v); +static int nfs_volume_list_show(struct seq_file *m, void *v); + +static struct seq_operations nfs_volume_list_ops = { + .start = nfs_volume_list_start, + .next = nfs_volume_list_next, + .stop = nfs_volume_list_stop, + .show = nfs_volume_list_show, +}; + +static struct file_operations nfs_volume_list_fops = { + .open = nfs_volume_list_open, + .read = seq_read, + .llseek = seq_lseek, + .release = seq_release, +}; + +/* + * open "/proc/fs/nfsfs/servers" which provides a summary of servers with which + * we're dealing + */ +static int nfs_server_list_open(struct inode *inode, struct file *file) +{ + struct seq_file *m; + int ret; + + ret = seq_open(file, &nfs_server_list_ops); + if (ret < 0) + return ret; + + m = file->private_data; + m->private = PDE(inode)->data; + + return 0; +} + +/* + * set up the iterator to start reading from the server list and return the first item + */ +static void *nfs_server_list_start(struct seq_file *m, loff_t *_pos) +{ + struct list_head *_p; + loff_t pos = *_pos; + + /* lock the list against modification */ + spin_lock(&nfs_client_lock); + + /* allow for the header line */ + if (!pos) + return SEQ_START_TOKEN; + pos--; + + /* find the n'th element in the list */ + list_for_each(_p, &nfs_client_list) + if (!pos--) + break; + + return _p != &nfs_client_list ? _p : NULL; +} + +/* + * move to next server + */ +static void *nfs_server_list_next(struct seq_file *p, void *v, loff_t *pos) +{ + struct list_head *_p; + + (*pos)++; + + _p = v; + _p = (v == SEQ_START_TOKEN) ? nfs_client_list.next : _p->next; + + return _p != &nfs_client_list ? _p : NULL; +} + +/* + * clean up after reading from the transports list + */ +static void nfs_server_list_stop(struct seq_file *p, void *v) +{ + spin_unlock(&nfs_client_lock); +} + +/* + * display a header line followed by a load of call lines + */ +static int nfs_server_list_show(struct seq_file *m, void *v) +{ + struct nfs_client *clp; + + /* display header on line 1 */ + if (v == SEQ_START_TOKEN) { + seq_puts(m, "NV SERVER PORT USE HOSTNAME\n"); + return 0; + } + + /* display one transport per line on subsequent lines */ + clp = list_entry(v, struct nfs_client, cl_share_link); + + seq_printf(m, "v%d %02x%02x%02x%02x %4hx %3d %s\n", + clp->cl_nfsversion, + NIPQUAD(clp->cl_addr.sin_addr), + ntohs(clp->cl_addr.sin_port), + atomic_read(&clp->cl_count), + clp->cl_hostname); + + return 0; +} + +/* + * open "/proc/fs/nfsfs/volumes" which provides a summary of extant volumes + */ +static int nfs_volume_list_open(struct inode *inode, struct file *file) +{ + struct seq_file *m; + int ret; + + ret = seq_open(file, &nfs_volume_list_ops); + if (ret < 0) + return ret; + + m = file->private_data; + m->private = PDE(inode)->data; + + return 0; +} + +/* + * set up the iterator to start reading from the volume list and return the first item + */ +static void *nfs_volume_list_start(struct seq_file *m, loff_t *_pos) +{ + struct list_head *_p; + loff_t pos = *_pos; + + /* lock the list against modification */ + spin_lock(&nfs_client_lock); + + /* allow for the header line */ + if (!pos) + return SEQ_START_TOKEN; + pos--; + + /* find the n'th element in the list */ + list_for_each(_p, &nfs_volume_list) + if (!pos--) + break; + + return _p != &nfs_volume_list ? _p : NULL; +} + +/* + * move to next volume + */ +static void *nfs_volume_list_next(struct seq_file *p, void *v, loff_t *pos) +{ + struct list_head *_p; + + (*pos)++; + + _p = v; + _p = (v == SEQ_START_TOKEN) ? nfs_volume_list.next : _p->next; + + return _p != &nfs_volume_list ? _p : NULL; +} + +/* + * clean up after reading from the transports list + */ +static void nfs_volume_list_stop(struct seq_file *p, void *v) +{ + spin_unlock(&nfs_client_lock); +} + +/* + * display a header line followed by a load of call lines + */ +static int nfs_volume_list_show(struct seq_file *m, void *v) +{ + struct nfs_server *server; + struct nfs_client *clp; + char dev[8], fsid[17]; + + /* display header on line 1 */ + if (v == SEQ_START_TOKEN) { + seq_puts(m, "NV SERVER PORT DEV FSID\n"); + return 0; + } + /* display one transport per line on subsequent lines */ + server = list_entry(v, struct nfs_server, master_link); + clp = server->nfs_client; + + snprintf(dev, 8, "%u:%u", + MAJOR(server->s_dev), MINOR(server->s_dev)); + + snprintf(fsid, 17, "%llx:%llx", + (unsigned long long) server->fsid.major, + (unsigned long long) server->fsid.minor); + + seq_printf(m, "v%d %02x%02x%02x%02x %4hx %-7s %-17s\n", + clp->cl_nfsversion, + NIPQUAD(clp->cl_addr.sin_addr), + ntohs(clp->cl_addr.sin_port), + dev, + fsid); + + return 0; +} + +/* + * initialise the /proc/fs/nfsfs/ directory + */ +int __init nfs_fs_proc_init(void) +{ + struct proc_dir_entry *p; + + proc_fs_nfs = proc_mkdir("nfsfs", proc_root_fs); + if (!proc_fs_nfs) + goto error_0; + + proc_fs_nfs->owner = THIS_MODULE; + + /* a file of servers with which we're dealing */ + p = create_proc_entry("servers", S_IFREG|S_IRUGO, proc_fs_nfs); + if (!p) + goto error_1; + + p->proc_fops = &nfs_server_list_fops; + p->owner = THIS_MODULE; + + /* a file of volumes that we have mounted */ + p = create_proc_entry("volumes", S_IFREG|S_IRUGO, proc_fs_nfs); + if (!p) + goto error_2; + + p->proc_fops = &nfs_volume_list_fops; + p->owner = THIS_MODULE; + return 0; + +error_2: + remove_proc_entry("servers", proc_fs_nfs); +error_1: + remove_proc_entry("nfsfs", proc_root_fs); +error_0: + return -ENOMEM; +} + +/* + * clean up the /proc/fs/nfsfs/ directory + */ +void nfs_fs_proc_exit(void) +{ + remove_proc_entry("volumes", proc_fs_nfs); + remove_proc_entry("servers", proc_fs_nfs); + remove_proc_entry("nfsfs", proc_root_fs); +} + +#endif /* CONFIG_PROC_FS */ diff --git a/fs/nfs/delegation.c b/fs/nfs/delegation.c index d3be923d4e4..841c99a9b11 100644 --- a/fs/nfs/delegation.c +++ b/fs/nfs/delegation.c @@ -6,7 +6,6 @@ * NFS file delegation management * */ -#include <linux/config.h> #include <linux/completion.h> #include <linux/kthread.h> #include <linux/module.h> @@ -19,11 +18,7 @@ #include "nfs4_fs.h" #include "delegation.h" - -static struct nfs_delegation *nfs_alloc_delegation(void) -{ - return (struct nfs_delegation *)kmalloc(sizeof(struct nfs_delegation), GFP_KERNEL); -} +#include "internal.h" static void nfs_free_delegation(struct nfs_delegation *delegation) { @@ -53,7 +48,7 @@ static int nfs_delegation_claim_locks(struct nfs_open_context *ctx, struct nfs4_ case -NFS4ERR_EXPIRED: /* kill_proc(fl->fl_pid, SIGLOST, 1); */ case -NFS4ERR_STALE_CLIENTID: - nfs4_schedule_state_recovery(NFS_SERVER(inode)->nfs4_state); + nfs4_schedule_state_recovery(NFS_SERVER(inode)->nfs_client); goto out_err; } } @@ -115,7 +110,7 @@ void nfs_inode_reclaim_delegation(struct inode *inode, struct rpc_cred *cred, st */ int nfs_inode_set_delegation(struct inode *inode, struct rpc_cred *cred, struct nfs_openres *res) { - struct nfs4_client *clp = NFS_SERVER(inode)->nfs4_state; + struct nfs_client *clp = NFS_SERVER(inode)->nfs_client; struct nfs_inode *nfsi = NFS_I(inode); struct nfs_delegation *delegation; int status = 0; @@ -124,7 +119,7 @@ int nfs_inode_set_delegation(struct inode *inode, struct rpc_cred *cred, struct if ((nfsi->cache_validity & (NFS_INO_REVAL_PAGECACHE|NFS_INO_INVALID_ATTR))) __nfs_revalidate_inode(NFS_SERVER(inode), inode); - delegation = nfs_alloc_delegation(); + delegation = kmalloc(sizeof(*delegation), GFP_KERNEL); if (delegation == NULL) return -ENOMEM; memcpy(delegation->stateid.data, res->delegation.data, @@ -146,7 +141,7 @@ int nfs_inode_set_delegation(struct inode *inode, struct rpc_cred *cred, struct sizeof(delegation->stateid)) != 0 || delegation->type != nfsi->delegation->type) { printk("%s: server %u.%u.%u.%u, handed out a duplicate delegation!\n", - __FUNCTION__, NIPQUAD(clp->cl_addr)); + __FUNCTION__, NIPQUAD(clp->cl_addr.sin_addr)); status = -EIO; } } @@ -177,7 +172,7 @@ static void nfs_msync_inode(struct inode *inode) */ int __nfs_inode_return_delegation(struct inode *inode) { - struct nfs4_client *clp = NFS_SERVER(inode)->nfs4_state; + struct nfs_client *clp = NFS_SERVER(inode)->nfs_client; struct nfs_inode *nfsi = NFS_I(inode); struct nfs_delegation *delegation; int res = 0; @@ -209,7 +204,7 @@ int __nfs_inode_return_delegation(struct inode *inode) */ void nfs_return_all_delegations(struct super_block *sb) { - struct nfs4_client *clp = NFS_SB(sb)->nfs4_state; + struct nfs_client *clp = NFS_SB(sb)->nfs_client; struct nfs_delegation *delegation; struct inode *inode; @@ -233,7 +228,7 @@ restart: int nfs_do_expire_all_delegations(void *ptr) { - struct nfs4_client *clp = ptr; + struct nfs_client *clp = ptr; struct nfs_delegation *delegation; struct inode *inode; @@ -255,11 +250,11 @@ restart: } out: spin_unlock(&clp->cl_lock); - nfs4_put_client(clp); + nfs_put_client(clp); module_put_and_exit(0); } -void nfs_expire_all_delegations(struct nfs4_client *clp) +void nfs_expire_all_delegations(struct nfs_client *clp) { struct task_struct *task; @@ -267,17 +262,17 @@ void nfs_expire_all_delegations(struct nfs4_client *clp) atomic_inc(&clp->cl_count); task = kthread_run(nfs_do_expire_all_delegations, clp, "%u.%u.%u.%u-delegreturn", - NIPQUAD(clp->cl_addr)); + NIPQUAD(clp->cl_addr.sin_addr)); if (!IS_ERR(task)) return; - nfs4_put_client(clp); + nfs_put_client(clp); module_put(THIS_MODULE); } /* * Return all delegations following an NFS4ERR_CB_PATH_DOWN error. */ -void nfs_handle_cb_pathdown(struct nfs4_client *clp) +void nfs_handle_cb_pathdown(struct nfs_client *clp) { struct nfs_delegation *delegation; struct inode *inode; @@ -300,7 +295,7 @@ restart: struct recall_threadargs { struct inode *inode; - struct nfs4_client *clp; + struct nfs_client *clp; const nfs4_stateid *stateid; struct completion started; @@ -311,7 +306,7 @@ static int recall_thread(void *data) { struct recall_threadargs *args = (struct recall_threadargs *)data; struct inode *inode = igrab(args->inode); - struct nfs4_client *clp = NFS_SERVER(inode)->nfs4_state; + struct nfs_client *clp = NFS_SERVER(inode)->nfs_client; struct nfs_inode *nfsi = NFS_I(inode); struct nfs_delegation *delegation; @@ -372,7 +367,7 @@ out_module_put: /* * Retrieve the inode associated with a delegation */ -struct inode *nfs_delegation_find_inode(struct nfs4_client *clp, const struct nfs_fh *fhandle) +struct inode *nfs_delegation_find_inode(struct nfs_client *clp, const struct nfs_fh *fhandle) { struct nfs_delegation *delegation; struct inode *res = NULL; @@ -390,7 +385,7 @@ struct inode *nfs_delegation_find_inode(struct nfs4_client *clp, const struct nf /* * Mark all delegations as needing to be reclaimed */ -void nfs_delegation_mark_reclaim(struct nfs4_client *clp) +void nfs_delegation_mark_reclaim(struct nfs_client *clp) { struct nfs_delegation *delegation; spin_lock(&clp->cl_lock); @@ -402,7 +397,7 @@ void nfs_delegation_mark_reclaim(struct nfs4_client *clp) /* * Reap all unclaimed delegations after reboot recovery is done */ -void nfs_delegation_reap_unclaimed(struct nfs4_client *clp) +void nfs_delegation_reap_unclaimed(struct nfs_client *clp) { struct nfs_delegation *delegation, *n; LIST_HEAD(head); @@ -424,7 +419,7 @@ void nfs_delegation_reap_unclaimed(struct nfs4_client *clp) int nfs4_copy_delegation_stateid(nfs4_stateid *dst, struct inode *inode) { - struct nfs4_client *clp = NFS_SERVER(inode)->nfs4_state; + struct nfs_client *clp = NFS_SERVER(inode)->nfs_client; struct nfs_inode *nfsi = NFS_I(inode); struct nfs_delegation *delegation; int res = 0; diff --git a/fs/nfs/delegation.h b/fs/nfs/delegation.h index 3858694652f..2cfd4b24c7f 100644 --- a/fs/nfs/delegation.h +++ b/fs/nfs/delegation.h @@ -29,13 +29,13 @@ void nfs_inode_reclaim_delegation(struct inode *inode, struct rpc_cred *cred, st int __nfs_inode_return_delegation(struct inode *inode); int nfs_async_inode_return_delegation(struct inode *inode, const nfs4_stateid *stateid); -struct inode *nfs_delegation_find_inode(struct nfs4_client *clp, const struct nfs_fh *fhandle); +struct inode *nfs_delegation_find_inode(struct nfs_client *clp, const struct nfs_fh *fhandle); void nfs_return_all_delegations(struct super_block *sb); -void nfs_expire_all_delegations(struct nfs4_client *clp); -void nfs_handle_cb_pathdown(struct nfs4_client *clp); +void nfs_expire_all_delegations(struct nfs_client *clp); +void nfs_handle_cb_pathdown(struct nfs_client *clp); -void nfs_delegation_mark_reclaim(struct nfs4_client *clp); -void nfs_delegation_reap_unclaimed(struct nfs4_client *clp); +void nfs_delegation_mark_reclaim(struct nfs_client *clp); +void nfs_delegation_reap_unclaimed(struct nfs_client *clp); /* NFSv4 delegation-related procedures */ int nfs4_proc_delegreturn(struct inode *inode, struct rpc_cred *cred, const nfs4_stateid *stateid); diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c index cae74dd4c7f..7432f1a43f3 100644 --- a/fs/nfs/dir.c +++ b/fs/nfs/dir.c @@ -30,7 +30,9 @@ #include <linux/nfs_mount.h> #include <linux/pagemap.h> #include <linux/smp_lock.h> +#include <linux/pagevec.h> #include <linux/namei.h> +#include <linux/mount.h> #include "nfs4_fs.h" #include "delegation.h" @@ -528,7 +530,7 @@ static int nfs_readdir(struct file *filp, void *dirent, filldir_t filldir) lock_kernel(); - res = nfs_revalidate_inode(NFS_SERVER(inode), inode); + res = nfs_revalidate_mapping(inode, filp->f_mapping); if (res < 0) { unlock_kernel(); return res; @@ -690,7 +692,9 @@ int nfs_lookup_verify_inode(struct inode *inode, struct nameidata *nd) goto out_force; /* This is an open(2) */ if (nfs_lookup_check_intent(nd, LOOKUP_OPEN) != 0 && - !(server->flags & NFS_MOUNT_NOCTO)) + !(server->flags & NFS_MOUNT_NOCTO) && + (S_ISREG(inode->i_mode) || + S_ISDIR(inode->i_mode))) goto out_force; } return nfs_revalidate_inode(server, inode); @@ -868,6 +872,17 @@ int nfs_is_exclusive_create(struct inode *dir, struct nameidata *nd) return (nd->intent.open.flags & O_EXCL) != 0; } +static inline int nfs_reval_fsid(struct vfsmount *mnt, struct inode *dir, + struct nfs_fh *fh, struct nfs_fattr *fattr) +{ + struct nfs_server *server = NFS_SERVER(dir); + + if (!nfs_fsid_equal(&server->fsid, &fattr->fsid)) + /* Revalidate fsid on root dir */ + return __nfs_revalidate_inode(server, mnt->mnt_root->d_inode); + return 0; +} + static struct dentry *nfs_lookup(struct inode *dir, struct dentry * dentry, struct nameidata *nd) { struct dentry *res; @@ -889,9 +904,15 @@ static struct dentry *nfs_lookup(struct inode *dir, struct dentry * dentry, stru lock_kernel(); - /* If we're doing an exclusive create, optimize away the lookup */ - if (nfs_is_exclusive_create(dir, nd)) - goto no_entry; + /* + * If we're doing an exclusive create, optimize away the lookup + * but don't hash the dentry. + */ + if (nfs_is_exclusive_create(dir, nd)) { + d_instantiate(dentry, NULL); + res = NULL; + goto out_unlock; + } error = NFS_PROTO(dir)->lookup(dir, &dentry->d_name, &fhandle, &fattr); if (error == -ENOENT) @@ -900,12 +921,18 @@ static struct dentry *nfs_lookup(struct inode *dir, struct dentry * dentry, stru res = ERR_PTR(error); goto out_unlock; } + error = nfs_reval_fsid(nd->mnt, dir, &fhandle, &fattr); + if (error < 0) { + res = ERR_PTR(error); + goto out_unlock; + } inode = nfs_fhget(dentry->d_sb, &fhandle, &fattr); res = (struct dentry *)inode; if (IS_ERR(res)) goto out_unlock; + no_entry: - res = d_add_unique(dentry, inode); + res = d_materialise_unique(dentry, inode); if (res != NULL) dentry = res; nfs_renew_times(dentry); @@ -1099,11 +1126,13 @@ static struct dentry *nfs_readdir_lookup(nfs_readdir_descriptor_t *desc) dput(dentry); return NULL; } - alias = d_add_unique(dentry, inode); + + alias = d_materialise_unique(dentry, inode); if (alias != NULL) { dput(dentry); dentry = alias; } + nfs_renew_times(dentry); nfs_set_verifier(dentry, nfs_save_change_attribute(dir)); return dentry; @@ -1125,23 +1154,22 @@ int nfs_instantiate(struct dentry *dentry, struct nfs_fh *fhandle, struct inode *dir = dentry->d_parent->d_inode; error = NFS_PROTO(dir)->lookup(dir, &dentry->d_name, fhandle, fattr); if (error) - goto out_err; + return error; } if (!(fattr->valid & NFS_ATTR_FATTR)) { struct nfs_server *server = NFS_SB(dentry->d_sb); - error = server->rpc_ops->getattr(server, fhandle, fattr); + error = server->nfs_client->rpc_ops->getattr(server, fhandle, fattr); if (error < 0) - goto out_err; + return error; } inode = nfs_fhget(dentry->d_sb, fhandle, fattr); error = PTR_ERR(inode); if (IS_ERR(inode)) - goto out_err; + return error; d_instantiate(dentry, inode); + if (d_unhashed(dentry)) + d_rehash(dentry); return 0; -out_err: - d_drop(dentry); - return error; } /* @@ -1422,48 +1450,82 @@ static int nfs_unlink(struct inode *dir, struct dentry *dentry) return error; } -static int -nfs_symlink(struct inode *dir, struct dentry *dentry, const char *symname) +/* + * To create a symbolic link, most file systems instantiate a new inode, + * add a page to it containing the path, then write it out to the disk + * using prepare_write/commit_write. + * + * Unfortunately the NFS client can't create the in-core inode first + * because it needs a file handle to create an in-core inode (see + * fs/nfs/inode.c:nfs_fhget). We only have a file handle *after* the + * symlink request has completed on the server. + * + * So instead we allocate a raw page, copy the symname into it, then do + * the SYMLINK request with the page as the buffer. If it succeeds, we + * now have a new file handle and can instantiate an in-core NFS inode + * and move the raw page into its mapping. + */ +static int nfs_symlink(struct inode *dir, struct dentry *dentry, const char *symname) { + struct pagevec lru_pvec; + struct page *page; + char *kaddr; struct iattr attr; - struct nfs_fattr sym_attr; - struct nfs_fh sym_fh; - struct qstr qsymname; + unsigned int pathlen = strlen(symname); int error; dfprintk(VFS, "NFS: symlink(%s/%ld, %s, %s)\n", dir->i_sb->s_id, dir->i_ino, dentry->d_name.name, symname); -#ifdef NFS_PARANOIA -if (dentry->d_inode) -printk("nfs_proc_symlink: %s/%s not negative!\n", -dentry->d_parent->d_name.name, dentry->d_name.name); -#endif - /* - * Fill in the sattr for the call. - * Note: SunOS 4.1.2 crashes if the mode isn't initialized! - */ - attr.ia_valid = ATTR_MODE; - attr.ia_mode = S_IFLNK | S_IRWXUGO; + if (pathlen > PAGE_SIZE) + return -ENAMETOOLONG; - qsymname.name = symname; - qsymname.len = strlen(symname); + attr.ia_mode = S_IFLNK | S_IRWXUGO; + attr.ia_valid = ATTR_MODE; lock_kernel(); + + page = alloc_page(GFP_KERNEL); + if (!page) { + unlock_kernel(); + return -ENOMEM; + } + + kaddr = kmap_atomic(page, KM_USER0); + memcpy(kaddr, symname, pathlen); + if (pathlen < PAGE_SIZE) + memset(kaddr + pathlen, 0, PAGE_SIZE - pathlen); + kunmap_atomic(kaddr, KM_USER0); + nfs_begin_data_update(dir); - error = NFS_PROTO(dir)->symlink(dir, &dentry->d_name, &qsymname, - &attr, &sym_fh, &sym_attr); + error = NFS_PROTO(dir)->symlink(dir, dentry, page, pathlen, &attr); nfs_end_data_update(dir); - if (!error) { - error = nfs_instantiate(dentry, &sym_fh, &sym_attr); - } else { - if (error == -EEXIST) - printk("nfs_proc_symlink: %s/%s already exists??\n", - dentry->d_parent->d_name.name, dentry->d_name.name); + if (error != 0) { + dfprintk(VFS, "NFS: symlink(%s/%ld, %s, %s) error %d\n", + dir->i_sb->s_id, dir->i_ino, + dentry->d_name.name, symname, error); d_drop(dentry); + __free_page(page); + unlock_kernel(); + return error; } + + /* + * No big deal if we can't add this page to the page cache here. + * READLINK will get the missing page from the server if needed. + */ + pagevec_init(&lru_pvec, 0); + if (!add_to_page_cache(page, dentry->d_inode->i_mapping, 0, + GFP_KERNEL)) { + if (!pagevec_add(&lru_pvec, page)) + __pagevec_lru_add(&lru_pvec); + SetPageUptodate(page); + unlock_page(page); + } else + __free_page(page); + unlock_kernel(); - return error; + return 0; } static int @@ -1607,8 +1669,7 @@ out: if (rehash) d_rehash(rehash); if (!error) { - if (!S_ISDIR(old_inode->i_mode)) - d_move(old_dentry, new_dentry); + d_move(old_dentry, new_dentry); nfs_renew_times(new_dentry); nfs_set_verifier(new_dentry, nfs_save_change_attribute(new_dir)); } @@ -1620,35 +1681,211 @@ out: return error; } +static DEFINE_SPINLOCK(nfs_access_lru_lock); +static LIST_HEAD(nfs_access_lru_list); +static atomic_long_t nfs_access_nr_entries; + +static void nfs_access_free_entry(struct nfs_access_entry *entry) +{ + put_rpccred(entry->cred); + kfree(entry); + smp_mb__before_atomic_dec(); + atomic_long_dec(&nfs_access_nr_entries); + smp_mb__after_atomic_dec(); +} + +int nfs_access_cache_shrinker(int nr_to_scan, gfp_t gfp_mask) +{ + LIST_HEAD(head); + struct nfs_inode *nfsi; + struct nfs_access_entry *cache; + + spin_lock(&nfs_access_lru_lock); +restart: + list_for_each_entry(nfsi, &nfs_access_lru_list, access_cache_inode_lru) { + struct inode *inode; + + if (nr_to_scan-- == 0) + break; + inode = igrab(&nfsi->vfs_inode); + if (inode == NULL) + continue; + spin_lock(&inode->i_lock); + if (list_empty(&nfsi->access_cache_entry_lru)) + goto remove_lru_entry; + cache = list_entry(nfsi->access_cache_entry_lru.next, + struct nfs_access_entry, lru); + list_move(&cache->lru, &head); + rb_erase(&cache->rb_node, &nfsi->access_cache); + if (!list_empty(&nfsi->access_cache_entry_lru)) + list_move_tail(&nfsi->access_cache_inode_lru, + &nfs_access_lru_list); + else { +remove_lru_entry: + list_del_init(&nfsi->access_cache_inode_lru); + clear_bit(NFS_INO_ACL_LRU_SET, &nfsi->flags); + } + spin_unlock(&inode->i_lock); + iput(inode); + goto restart; + } + spin_unlock(&nfs_access_lru_lock); + while (!list_empty(&head)) { + cache = list_entry(head.next, struct nfs_access_entry, lru); + list_del(&cache->lru); + nfs_access_free_entry(cache); + } + return (atomic_long_read(&nfs_access_nr_entries) / 100) * sysctl_vfs_cache_pressure; +} + +static void __nfs_access_zap_cache(struct inode *inode) +{ + struct nfs_inode *nfsi = NFS_I(inode); + struct rb_root *root_node = &nfsi->access_cache; + struct rb_node *n, *dispose = NULL; + struct nfs_access_entry *entry; + + /* Unhook entries from the cache */ + while ((n = rb_first(root_node)) != NULL) { + entry = rb_entry(n, struct nfs_access_entry, rb_node); + rb_erase(n, root_node); + list_del(&entry->lru); + n->rb_left = dispose; + dispose = n; + } + nfsi->cache_validity &= ~NFS_INO_INVALID_ACCESS; + spin_unlock(&inode->i_lock); + + /* Now kill them all! */ + while (dispose != NULL) { + n = dispose; + dispose = n->rb_left; + nfs_access_free_entry(rb_entry(n, struct nfs_access_entry, rb_node)); + } +} + +void nfs_access_zap_cache(struct inode *inode) +{ + /* Remove from global LRU init */ + if (test_and_clear_bit(NFS_INO_ACL_LRU_SET, &NFS_FLAGS(inode))) { + spin_lock(&nfs_access_lru_lock); + list_del_init(&NFS_I(inode)->access_cache_inode_lru); + spin_unlock(&nfs_access_lru_lock); + } + + spin_lock(&inode->i_lock); + /* This will release the spinlock */ + __nfs_access_zap_cache(inode); +} + +static struct nfs_access_entry *nfs_access_search_rbtree(struct inode *inode, struct rpc_cred *cred) +{ + struct rb_node *n = NFS_I(inode)->access_cache.rb_node; + struct nfs_access_entry *entry; + + while (n != NULL) { + entry = rb_entry(n, struct nfs_access_entry, rb_node); + + if (cred < entry->cred) + n = n->rb_left; + else if (cred > entry->cred) + n = n->rb_right; + else + return entry; + } + return NULL; +} + int nfs_access_get_cached(struct inode *inode, struct rpc_cred *cred, struct nfs_access_entry *res) { struct nfs_inode *nfsi = NFS_I(inode); - struct nfs_access_entry *cache = &nfsi->cache_access; + struct nfs_access_entry *cache; + int err = -ENOENT; - if (cache->cred != cred - || time_after(jiffies, cache->jiffies + NFS_ATTRTIMEO(inode)) - || (nfsi->cache_validity & NFS_INO_INVALID_ACCESS)) - return -ENOENT; - memcpy(res, cache, sizeof(*res)); - return 0; + spin_lock(&inode->i_lock); + if (nfsi->cache_validity & NFS_INO_INVALID_ACCESS) + goto out_zap; + cache = nfs_access_search_rbtree(inode, cred); + if (cache == NULL) + goto out; + if (time_after(jiffies, cache->jiffies + NFS_ATTRTIMEO(inode))) + goto out_stale; + res->jiffies = cache->jiffies; + res->cred = cache->cred; + res->mask = cache->mask; + list_move_tail(&cache->lru, &nfsi->access_cache_entry_lru); + err = 0; +out: + spin_unlock(&inode->i_lock); + return err; +out_stale: + rb_erase(&cache->rb_node, &nfsi->access_cache); + list_del(&cache->lru); + spin_unlock(&inode->i_lock); + nfs_access_free_entry(cache); + return -ENOENT; +out_zap: + /* This will release the spinlock */ + __nfs_access_zap_cache(inode); + return -ENOENT; } -void nfs_access_add_cache(struct inode *inode, struct nfs_access_entry *set) +static void nfs_access_add_rbtree(struct inode *inode, struct nfs_access_entry *set) { struct nfs_inode *nfsi = NFS_I(inode); - struct nfs_access_entry *cache = &nfsi->cache_access; + struct rb_root *root_node = &nfsi->access_cache; + struct rb_node **p = &root_node->rb_node; + struct rb_node *parent = NULL; + struct nfs_access_entry *entry; - if (cache->cred != set->cred) { - if (cache->cred) - put_rpccred(cache->cred); - cache->cred = get_rpccred(set->cred); - } - /* FIXME: replace current access_cache BKL reliance with inode->i_lock */ spin_lock(&inode->i_lock); - nfsi->cache_validity &= ~NFS_INO_INVALID_ACCESS; + while (*p != NULL) { + parent = *p; + entry = rb_entry(parent, struct nfs_access_entry, rb_node); + + if (set->cred < entry->cred) + p = &parent->rb_left; + else if (set->cred > entry->cred) + p = &parent->rb_right; + else + goto found; + } + rb_link_node(&set->rb_node, parent, p); + rb_insert_color(&set->rb_node, root_node); + list_add_tail(&set->lru, &nfsi->access_cache_entry_lru); spin_unlock(&inode->i_lock); + return; +found: + rb_replace_node(parent, &set->rb_node, root_node); + list_add_tail(&set->lru, &nfsi->access_cache_entry_lru); + list_del(&entry->lru); + spin_unlock(&inode->i_lock); + nfs_access_free_entry(entry); +} + +void nfs_access_add_cache(struct inode *inode, struct nfs_access_entry *set) +{ + struct nfs_access_entry *cache = kmalloc(sizeof(*cache), GFP_KERNEL); + if (cache == NULL) + return; + RB_CLEAR_NODE(&cache->rb_node); cache->jiffies = set->jiffies; + cache->cred = get_rpccred(set->cred); cache->mask = set->mask; + + nfs_access_add_rbtree(inode, cache); + + /* Update accounting */ + smp_mb__before_atomic_inc(); + atomic_long_inc(&nfs_access_nr_entries); + smp_mb__after_atomic_inc(); + + /* Add inode to global LRU list */ + if (!test_and_set_bit(NFS_INO_ACL_LRU_SET, &NFS_FLAGS(inode))) { + spin_lock(&nfs_access_lru_lock); + list_add_tail(&NFS_I(inode)->access_cache_inode_lru, &nfs_access_lru_list); + spin_unlock(&nfs_access_lru_lock); + } } static int nfs_do_access(struct inode *inode, struct rpc_cred *cred, int mask) diff --git a/fs/nfs/direct.c b/fs/nfs/direct.c index 3c72b0c0728..377839bed17 100644 --- a/fs/nfs/direct.c +++ b/fs/nfs/direct.c @@ -38,7 +38,6 @@ * */ -#include <linux/config.h> #include <linux/errno.h> #include <linux/sched.h> #include <linux/kernel.h> @@ -68,25 +67,19 @@ struct nfs_direct_req { struct kref kref; /* release manager */ /* I/O parameters */ - struct list_head list, /* nfs_read/write_data structs */ - rewrite_list; /* saved nfs_write_data structs */ struct nfs_open_context *ctx; /* file open context info */ struct kiocb * iocb; /* controlling i/o request */ struct inode * inode; /* target file of i/o */ - unsigned long user_addr; /* location of user's buffer */ - size_t user_count; /* total bytes to move */ - loff_t pos; /* starting offset in file */ - struct page ** pages; /* pages in our buffer */ - unsigned int npages; /* count of pages */ /* completion state */ + atomic_t io_count; /* i/os we're waiting for */ spinlock_t lock; /* protect completion state */ - int outstanding; /* i/os we're waiting for */ ssize_t count, /* bytes actually processed */ error; /* any reported error */ struct completion completion; /* wait for i/o completion */ /* commit state */ + struct list_head rewrite_list; /* saved nfs_write_data structs */ struct nfs_write_data * commit_data; /* special write_data for commits */ int flags; #define NFS_ODIRECT_DO_COMMIT (1) /* an unstable reply was received */ @@ -94,8 +87,18 @@ struct nfs_direct_req { struct nfs_writeverf verf; /* unstable write verifier */ }; -static void nfs_direct_write_schedule(struct nfs_direct_req *dreq, int sync); static void nfs_direct_write_complete(struct nfs_direct_req *dreq, struct inode *inode); +static const struct rpc_call_ops nfs_write_direct_ops; + +static inline void get_dreq(struct nfs_direct_req *dreq) +{ + atomic_inc(&dreq->io_count); +} + +static inline int put_dreq(struct nfs_direct_req *dreq) +{ + return atomic_dec_and_test(&dreq->io_count); +} /** * nfs_direct_IO - NFS address space operation for direct I/O @@ -119,50 +122,21 @@ ssize_t nfs_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov, loff_ return -EINVAL; } -static void nfs_free_user_pages(struct page **pages, int npages, int do_dirty) +static void nfs_direct_dirty_pages(struct page **pages, int npages) { int i; for (i = 0; i < npages; i++) { struct page *page = pages[i]; - if (do_dirty && !PageCompound(page)) + if (!PageCompound(page)) set_page_dirty_lock(page); - page_cache_release(page); } - kfree(pages); } -static inline int nfs_get_user_pages(int rw, unsigned long user_addr, size_t size, struct page ***pages) +static void nfs_direct_release_pages(struct page **pages, int npages) { - int result = -ENOMEM; - unsigned long page_count; - size_t array_size; - - page_count = (user_addr + size + PAGE_SIZE - 1) >> PAGE_SHIFT; - page_count -= user_addr >> PAGE_SHIFT; - - array_size = (page_count * sizeof(struct page *)); - *pages = kmalloc(array_size, GFP_KERNEL); - if (*pages) { - down_read(¤t->mm->mmap_sem); - result = get_user_pages(current, current->mm, user_addr, - page_count, (rw == READ), 0, - *pages, NULL); - up_read(¤t->mm->mmap_sem); - if (result != page_count) { - /* - * If we got fewer pages than expected from - * get_user_pages(), the user buffer runs off the - * end of a mapping; return EFAULT. - */ - if (result >= 0) { - nfs_free_user_pages(*pages, result, 0); - result = -EFAULT; - } else - kfree(*pages); - *pages = NULL; - } - } - return result; + int i; + for (i = 0; i < npages; i++) + page_cache_release(pages[i]); } static inline struct nfs_direct_req *nfs_direct_req_alloc(void) @@ -174,13 +148,13 @@ static inline struct nfs_direct_req *nfs_direct_req_alloc(void) return NULL; kref_init(&dreq->kref); + kref_get(&dreq->kref); init_completion(&dreq->completion); - INIT_LIST_HEAD(&dreq->list); INIT_LIST_HEAD(&dreq->rewrite_list); dreq->iocb = NULL; dreq->ctx = NULL; spin_lock_init(&dreq->lock); - dreq->outstanding = 0; + atomic_set(&dreq->io_count, 0); dreq->count = 0; dreq->error = 0; dreq->flags = 0; @@ -221,18 +195,11 @@ out: } /* - * We must hold a reference to all the pages in this direct read request - * until the RPCs complete. This could be long *after* we are woken up in - * nfs_direct_wait (for instance, if someone hits ^C on a slow server). - * - * In addition, synchronous I/O uses a stack-allocated iocb. Thus we - * can't trust the iocb is still valid here if this is a synchronous - * request. If the waiter is woken prematurely, the iocb is long gone. + * Synchronous I/O uses a stack-allocated iocb. Thus we can't trust + * the iocb is still valid here if this is a synchronous request. */ static void nfs_direct_complete(struct nfs_direct_req *dreq) { - nfs_free_user_pages(dreq->pages, dreq->npages, 1); - if (dreq->iocb) { long res = (long) dreq->error; if (!res) @@ -245,48 +212,10 @@ static void nfs_direct_complete(struct nfs_direct_req *dreq) } /* - * Note we also set the number of requests we have in the dreq when we are - * done. This prevents races with I/O completion so we will always wait - * until all requests have been dispatched and completed. + * We must hold a reference to all the pages in this direct read request + * until the RPCs complete. This could be long *after* we are woken up in + * nfs_direct_wait (for instance, if someone hits ^C on a slow server). */ -static struct nfs_direct_req *nfs_direct_read_alloc(size_t nbytes, size_t rsize) -{ - struct list_head *list; - struct nfs_direct_req *dreq; - unsigned int rpages = (rsize + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT; - - dreq = nfs_direct_req_alloc(); - if (!dreq) - return NULL; - - list = &dreq->list; - for(;;) { - struct nfs_read_data *data = nfs_readdata_alloc(rpages); - - if (unlikely(!data)) { - while (!list_empty(list)) { - data = list_entry(list->next, - struct nfs_read_data, pages); - list_del(&data->pages); - nfs_readdata_free(data); - } - kref_put(&dreq->kref, nfs_direct_req_release); - return NULL; - } - - INIT_LIST_HEAD(&data->pages); - list_add(&data->pages, list); - - data->req = (struct nfs_page *) dreq; - dreq->outstanding++; - if (nbytes <= rsize) - break; - nbytes -= rsize; - } - kref_get(&dreq->kref); - return dreq; -} - static void nfs_direct_read_result(struct rpc_task *task, void *calldata) { struct nfs_read_data *data = calldata; @@ -295,6 +224,9 @@ static void nfs_direct_read_result(struct rpc_task *task, void *calldata) if (nfs_readpage_result(task, data) != 0) return; + nfs_direct_dirty_pages(data->pagevec, data->npages); + nfs_direct_release_pages(data->pagevec, data->npages); + spin_lock(&dreq->lock); if (likely(task->tk_status >= 0)) @@ -302,13 +234,10 @@ static void nfs_direct_read_result(struct rpc_task *task, void *calldata) else dreq->error = task->tk_status; - if (--dreq->outstanding) { - spin_unlock(&dreq->lock); - return; - } - spin_unlock(&dreq->lock); - nfs_direct_complete(dreq); + + if (put_dreq(dreq)) + nfs_direct_complete(dreq); } static const struct rpc_call_ops nfs_read_direct_ops = { @@ -317,41 +246,56 @@ static const struct rpc_call_ops nfs_read_direct_ops = { }; /* - * For each nfs_read_data struct that was allocated on the list, dispatch - * an NFS READ operation + * For each rsize'd chunk of the user's buffer, dispatch an NFS READ + * operation. If nfs_readdata_alloc() or get_user_pages() fails, + * bail and stop sending more reads. Read length accounting is + * handled automatically by nfs_direct_read_result(). Otherwise, if + * no requests have been sent, just return an error. */ -static void nfs_direct_read_schedule(struct nfs_direct_req *dreq) +static ssize_t nfs_direct_read_schedule(struct nfs_direct_req *dreq, unsigned long user_addr, size_t count, loff_t pos) { struct nfs_open_context *ctx = dreq->ctx; struct inode *inode = ctx->dentry->d_inode; - struct list_head *list = &dreq->list; - struct page **pages = dreq->pages; - size_t count = dreq->user_count; - loff_t pos = dreq->pos; size_t rsize = NFS_SERVER(inode)->rsize; - unsigned int curpage, pgbase; + unsigned int pgbase; + int result; + ssize_t started = 0; + + get_dreq(dreq); - curpage = 0; - pgbase = dreq->user_addr & ~PAGE_MASK; do { struct nfs_read_data *data; size_t bytes; - bytes = rsize; - if (count < rsize) - bytes = count; + pgbase = user_addr & ~PAGE_MASK; + bytes = min(rsize,count); - BUG_ON(list_empty(list)); - data = list_entry(list->next, struct nfs_read_data, pages); - list_del_init(&data->pages); + result = -ENOMEM; + data = nfs_readdata_alloc(pgbase + bytes); + if (unlikely(!data)) + break; + + down_read(¤t->mm->mmap_sem); + result = get_user_pages(current, current->mm, user_addr, + data->npages, 1, 0, data->pagevec, NULL); + up_read(¤t->mm->mmap_sem); + if (unlikely(result < data->npages)) { + if (result > 0) + nfs_direct_release_pages(data->pagevec, result); + nfs_readdata_release(data); + break; + } + + get_dreq(dreq); + data->req = (struct nfs_page *) dreq; data->inode = inode; data->cred = ctx->cred; data->args.fh = NFS_FH(inode); data->args.context = ctx; data->args.offset = pos; data->args.pgbase = pgbase; - data->args.pages = &pages[curpage]; + data->args.pages = data->pagevec; data->args.count = bytes; data->res.fattr = &data->fattr; data->res.eof = 0; @@ -374,33 +318,37 @@ static void nfs_direct_read_schedule(struct nfs_direct_req *dreq) bytes, (unsigned long long)data->args.offset); + started += bytes; + user_addr += bytes; pos += bytes; + /* FIXME: Remove this unnecessary math from final patch */ pgbase += bytes; - curpage += pgbase >> PAGE_SHIFT; pgbase &= ~PAGE_MASK; + BUG_ON(pgbase != (user_addr & ~PAGE_MASK)); count -= bytes; } while (count != 0); - BUG_ON(!list_empty(list)); + + if (put_dreq(dreq)) + nfs_direct_complete(dreq); + + if (started) + return 0; + return result < 0 ? (ssize_t) result : -EFAULT; } -static ssize_t nfs_direct_read(struct kiocb *iocb, unsigned long user_addr, size_t count, loff_t pos, struct page **pages, unsigned int nr_pages) +static ssize_t nfs_direct_read(struct kiocb *iocb, unsigned long user_addr, size_t count, loff_t pos) { - ssize_t result; + ssize_t result = 0; sigset_t oldset; struct inode *inode = iocb->ki_filp->f_mapping->host; struct rpc_clnt *clnt = NFS_CLIENT(inode); struct nfs_direct_req *dreq; - dreq = nfs_direct_read_alloc(count, NFS_SERVER(inode)->rsize); + dreq = nfs_direct_req_alloc(); if (!dreq) return -ENOMEM; - dreq->user_addr = user_addr; - dreq->user_count = count; - dreq->pos = pos; - dreq->pages = pages; - dreq->npages = nr_pages; dreq->inode = inode; dreq->ctx = get_nfs_open_context((struct nfs_open_context *)iocb->ki_filp->private_data); if (!is_sync_kiocb(iocb)) @@ -408,8 +356,9 @@ static ssize_t nfs_direct_read(struct kiocb *iocb, unsigned long user_addr, size nfs_add_stats(inode, NFSIOS_DIRECTREADBYTES, count); rpc_clnt_sigmask(clnt, &oldset); - nfs_direct_read_schedule(dreq); - result = nfs_direct_wait(dreq); + result = nfs_direct_read_schedule(dreq, user_addr, count, pos); + if (!result) + result = nfs_direct_wait(dreq); rpc_clnt_sigunmask(clnt, &oldset); return result; @@ -417,10 +366,10 @@ static ssize_t nfs_direct_read(struct kiocb *iocb, unsigned long user_addr, size static void nfs_direct_free_writedata(struct nfs_direct_req *dreq) { - list_splice_init(&dreq->rewrite_list, &dreq->list); - while (!list_empty(&dreq->list)) { - struct nfs_write_data *data = list_entry(dreq->list.next, struct nfs_write_data, pages); + while (!list_empty(&dreq->rewrite_list)) { + struct nfs_write_data *data = list_entry(dreq->rewrite_list.next, struct nfs_write_data, pages); list_del(&data->pages); + nfs_direct_release_pages(data->pagevec, data->npages); nfs_writedata_release(data); } } @@ -428,14 +377,51 @@ static void nfs_direct_free_writedata(struct nfs_direct_req *dreq) #if defined(CONFIG_NFS_V3) || defined(CONFIG_NFS_V4) static void nfs_direct_write_reschedule(struct nfs_direct_req *dreq) { - struct list_head *pos; + struct inode *inode = dreq->inode; + struct list_head *p; + struct nfs_write_data *data; - list_splice_init(&dreq->rewrite_list, &dreq->list); - list_for_each(pos, &dreq->list) - dreq->outstanding++; dreq->count = 0; + get_dreq(dreq); + + list_for_each(p, &dreq->rewrite_list) { + data = list_entry(p, struct nfs_write_data, pages); + + get_dreq(dreq); - nfs_direct_write_schedule(dreq, FLUSH_STABLE); + /* + * Reset data->res. + */ + nfs_fattr_init(&data->fattr); + data->res.count = data->args.count; + memset(&data->verf, 0, sizeof(data->verf)); + + /* + * Reuse data->task; data->args should not have changed + * since the original request was sent. + */ + rpc_init_task(&data->task, NFS_CLIENT(inode), RPC_TASK_ASYNC, + &nfs_write_direct_ops, data); + NFS_PROTO(inode)->write_setup(data, FLUSH_STABLE); + + data->task.tk_priority = RPC_PRIORITY_NORMAL; + data->task.tk_cookie = (unsigned long) inode; + + /* + * We're called via an RPC callback, so BKL is already held. + */ + rpc_execute(&data->task); + + dprintk("NFS: %5u rescheduled direct write call (req %s/%Ld, %u bytes @ offset %Lu)\n", + data->task.tk_pid, + inode->i_sb->s_id, + (long long)NFS_FILEID(inode), + data->args.count, + (unsigned long long)data->args.offset); + } + + if (put_dreq(dreq)) + nfs_direct_write_complete(dreq, inode); } static void nfs_direct_commit_result(struct rpc_task *task, void *calldata) @@ -472,8 +458,8 @@ static void nfs_direct_commit_schedule(struct nfs_direct_req *dreq) data->cred = dreq->ctx->cred; data->args.fh = NFS_FH(data->inode); - data->args.offset = dreq->pos; - data->args.count = dreq->user_count; + data->args.offset = 0; + data->args.count = 0; data->res.count = 0; data->res.fattr = &data->fattr; data->res.verf = &data->verf; @@ -517,7 +503,7 @@ static void nfs_direct_write_complete(struct nfs_direct_req *dreq, struct inode static void nfs_alloc_commit_data(struct nfs_direct_req *dreq) { - dreq->commit_data = nfs_commit_alloc(0); + dreq->commit_data = nfs_commit_alloc(); if (dreq->commit_data != NULL) dreq->commit_data->req = (struct nfs_page *) dreq; } @@ -535,47 +521,6 @@ static void nfs_direct_write_complete(struct nfs_direct_req *dreq, struct inode } #endif -static struct nfs_direct_req *nfs_direct_write_alloc(size_t nbytes, size_t wsize) -{ - struct list_head *list; - struct nfs_direct_req *dreq; - unsigned int wpages = (wsize + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT; - - dreq = nfs_direct_req_alloc(); - if (!dreq) - return NULL; - - list = &dreq->list; - for(;;) { - struct nfs_write_data *data = nfs_writedata_alloc(wpages); - - if (unlikely(!data)) { - while (!list_empty(list)) { - data = list_entry(list->next, - struct nfs_write_data, pages); - list_del(&data->pages); - nfs_writedata_free(data); - } - kref_put(&dreq->kref, nfs_direct_req_release); - return NULL; - } - - INIT_LIST_HEAD(&data->pages); - list_add(&data->pages, list); - - data->req = (struct nfs_page *) dreq; - dreq->outstanding++; - if (nbytes <= wsize) - break; - nbytes -= wsize; - } - - nfs_alloc_commit_data(dreq); - - kref_get(&dreq->kref); - return dreq; -} - static void nfs_direct_write_result(struct rpc_task *task, void *calldata) { struct nfs_write_data *data = calldata; @@ -605,8 +550,6 @@ static void nfs_direct_write_result(struct rpc_task *task, void *calldata) } } } - /* In case we have to resend */ - data->args.stable = NFS_FILE_SYNC; spin_unlock(&dreq->lock); } @@ -620,14 +563,8 @@ static void nfs_direct_write_release(void *calldata) struct nfs_write_data *data = calldata; struct nfs_direct_req *dreq = (struct nfs_direct_req *) data->req; - spin_lock(&dreq->lock); - if (--dreq->outstanding) { - spin_unlock(&dreq->lock); - return; - } - spin_unlock(&dreq->lock); - - nfs_direct_write_complete(dreq, data->inode); + if (put_dreq(dreq)) + nfs_direct_write_complete(dreq, data->inode); } static const struct rpc_call_ops nfs_write_direct_ops = { @@ -636,41 +573,58 @@ static const struct rpc_call_ops nfs_write_direct_ops = { }; /* - * For each nfs_write_data struct that was allocated on the list, dispatch - * an NFS WRITE operation + * For each wsize'd chunk of the user's buffer, dispatch an NFS WRITE + * operation. If nfs_writedata_alloc() or get_user_pages() fails, + * bail and stop sending more writes. Write length accounting is + * handled automatically by nfs_direct_write_result(). Otherwise, if + * no requests have been sent, just return an error. */ -static void nfs_direct_write_schedule(struct nfs_direct_req *dreq, int sync) +static ssize_t nfs_direct_write_schedule(struct nfs_direct_req *dreq, unsigned long user_addr, size_t count, loff_t pos, int sync) { struct nfs_open_context *ctx = dreq->ctx; struct inode *inode = ctx->dentry->d_inode; - struct list_head *list = &dreq->list; - struct page **pages = dreq->pages; - size_t count = dreq->user_count; - loff_t pos = dreq->pos; size_t wsize = NFS_SERVER(inode)->wsize; - unsigned int curpage, pgbase; + unsigned int pgbase; + int result; + ssize_t started = 0; + + get_dreq(dreq); - curpage = 0; - pgbase = dreq->user_addr & ~PAGE_MASK; do { struct nfs_write_data *data; size_t bytes; - bytes = wsize; - if (count < wsize) - bytes = count; + pgbase = user_addr & ~PAGE_MASK; + bytes = min(wsize,count); + + result = -ENOMEM; + data = nfs_writedata_alloc(pgbase + bytes); + if (unlikely(!data)) + break; + + down_read(¤t->mm->mmap_sem); + result = get_user_pages(current, current->mm, user_addr, + data->npages, 0, 0, data->pagevec, NULL); + up_read(¤t->mm->mmap_sem); + if (unlikely(result < data->npages)) { + if (result > 0) + nfs_direct_release_pages(data->pagevec, result); + nfs_writedata_release(data); + break; + } + + get_dreq(dreq); - BUG_ON(list_empty(list)); - data = list_entry(list->next, struct nfs_write_data, pages); list_move_tail(&data->pages, &dreq->rewrite_list); + data->req = (struct nfs_page *) dreq; data->inode = inode; data->cred = ctx->cred; data->args.fh = NFS_FH(inode); data->args.context = ctx; data->args.offset = pos; data->args.pgbase = pgbase; - data->args.pages = &pages[curpage]; + data->args.pages = data->pagevec; data->args.count = bytes; data->res.fattr = &data->fattr; data->res.count = bytes; @@ -694,19 +648,29 @@ static void nfs_direct_write_schedule(struct nfs_direct_req *dreq, int sync) bytes, (unsigned long long)data->args.offset); + started += bytes; + user_addr += bytes; pos += bytes; + + /* FIXME: Remove this useless math from the final patch */ pgbase += bytes; - curpage += pgbase >> PAGE_SHIFT; pgbase &= ~PAGE_MASK; + BUG_ON(pgbase != (user_addr & ~PAGE_MASK)); count -= bytes; } while (count != 0); - BUG_ON(!list_empty(list)); + + if (put_dreq(dreq)) + nfs_direct_write_complete(dreq, inode); + + if (started) + return 0; + return result < 0 ? (ssize_t) result : -EFAULT; } -static ssize_t nfs_direct_write(struct kiocb *iocb, unsigned long user_addr, size_t count, loff_t pos, struct page **pages, int nr_pages) +static ssize_t nfs_direct_write(struct kiocb *iocb, unsigned long user_addr, size_t count, loff_t pos) { - ssize_t result; + ssize_t result = 0; sigset_t oldset; struct inode *inode = iocb->ki_filp->f_mapping->host; struct rpc_clnt *clnt = NFS_CLIENT(inode); @@ -714,17 +678,14 @@ static ssize_t nfs_direct_write(struct kiocb *iocb, unsigned long user_addr, siz size_t wsize = NFS_SERVER(inode)->wsize; int sync = 0; - dreq = nfs_direct_write_alloc(count, wsize); + dreq = nfs_direct_req_alloc(); if (!dreq) return -ENOMEM; + nfs_alloc_commit_data(dreq); + if (dreq->commit_data == NULL || count < wsize) sync = FLUSH_STABLE; - dreq->user_addr = user_addr; - dreq->user_count = count; - dreq->pos = pos; - dreq->pages = pages; - dreq->npages = nr_pages; dreq->inode = inode; dreq->ctx = get_nfs_open_context((struct nfs_open_context *)iocb->ki_filp->private_data); if (!is_sync_kiocb(iocb)) @@ -735,8 +696,9 @@ static ssize_t nfs_direct_write(struct kiocb *iocb, unsigned long user_addr, siz nfs_begin_data_update(inode); rpc_clnt_sigmask(clnt, &oldset); - nfs_direct_write_schedule(dreq, sync); - result = nfs_direct_wait(dreq); + result = nfs_direct_write_schedule(dreq, user_addr, count, pos, sync); + if (!result) + result = nfs_direct_wait(dreq); rpc_clnt_sigunmask(clnt, &oldset); return result; @@ -766,8 +728,6 @@ static ssize_t nfs_direct_write(struct kiocb *iocb, unsigned long user_addr, siz ssize_t nfs_file_direct_read(struct kiocb *iocb, char __user *buf, size_t count, loff_t pos) { ssize_t retval = -EINVAL; - int page_count; - struct page **pages; struct file *file = iocb->ki_filp; struct address_space *mapping = file->f_mapping; @@ -789,14 +749,7 @@ ssize_t nfs_file_direct_read(struct kiocb *iocb, char __user *buf, size_t count, if (retval) goto out; - retval = nfs_get_user_pages(READ, (unsigned long) buf, - count, &pages); - if (retval < 0) - goto out; - page_count = retval; - - retval = nfs_direct_read(iocb, (unsigned long) buf, count, pos, - pages, page_count); + retval = nfs_direct_read(iocb, (unsigned long) buf, count, pos); if (retval > 0) iocb->ki_pos = pos + retval; @@ -832,8 +785,6 @@ out: ssize_t nfs_file_direct_write(struct kiocb *iocb, const char __user *buf, size_t count, loff_t pos) { ssize_t retval; - int page_count; - struct page **pages; struct file *file = iocb->ki_filp; struct address_space *mapping = file->f_mapping; @@ -861,14 +812,7 @@ ssize_t nfs_file_direct_write(struct kiocb *iocb, const char __user *buf, size_t if (retval) goto out; - retval = nfs_get_user_pages(WRITE, (unsigned long) buf, - count, &pages); - if (retval < 0) - goto out; - page_count = retval; - - retval = nfs_direct_write(iocb, (unsigned long) buf, count, - pos, pages, page_count); + retval = nfs_direct_write(iocb, (unsigned long) buf, count, pos); /* * XXX: nfs_end_data_update() already ensures this file's @@ -892,7 +836,7 @@ out: * nfs_init_directcache - create a slab cache for nfs_direct_req structures * */ -int nfs_init_directcache(void) +int __init nfs_init_directcache(void) { nfs_direct_cachep = kmem_cache_create("nfs_direct_cache", sizeof(struct nfs_direct_req), @@ -906,11 +850,10 @@ int nfs_init_directcache(void) } /** - * nfs_init_directcache - destroy the slab cache for nfs_direct_req structures + * nfs_destroy_directcache - destroy the slab cache for nfs_direct_req structures * */ void nfs_destroy_directcache(void) { - if (kmem_cache_destroy(nfs_direct_cachep)) - printk(KERN_INFO "nfs_direct_cache: not all structures were freed\n"); + kmem_cache_destroy(nfs_direct_cachep); } diff --git a/fs/nfs/file.c b/fs/nfs/file.c index fa05c027ea1..be997d64912 100644 --- a/fs/nfs/file.c +++ b/fs/nfs/file.c @@ -111,7 +111,7 @@ nfs_file_open(struct inode *inode, struct file *filp) nfs_inc_stats(inode, NFSIOS_VFSOPEN); lock_kernel(); - res = NFS_SERVER(inode)->rpc_ops->file_open(inode, filp); + res = NFS_PROTO(inode)->file_open(inode, filp); unlock_kernel(); return res; } @@ -127,23 +127,6 @@ nfs_file_release(struct inode *inode, struct file *filp) } /** - * nfs_revalidate_file - Revalidate the page cache & related metadata - * @inode - pointer to inode struct - * @file - pointer to file - */ -static int nfs_revalidate_file(struct inode *inode, struct file *filp) -{ - struct nfs_inode *nfsi = NFS_I(inode); - int retval = 0; - - if ((nfsi->cache_validity & (NFS_INO_REVAL_PAGECACHE|NFS_INO_INVALID_ATTR)) - || nfs_attribute_timeout(inode)) - retval = __nfs_revalidate_inode(NFS_SERVER(inode), inode); - nfs_revalidate_mapping(inode, filp->f_mapping); - return 0; -} - -/** * nfs_revalidate_size - Revalidate the file size * @inode - pointer to inode struct * @file - pointer to struct file @@ -174,7 +157,7 @@ force_reval: static loff_t nfs_file_llseek(struct file *filp, loff_t offset, int origin) { /* origin == SEEK_END => we must revalidate the cached file length */ - if (origin == 2) { + if (origin == SEEK_END) { struct inode *inode = filp->f_mapping->host; int retval = nfs_revalidate_file_size(inode, filp); if (retval < 0) @@ -228,7 +211,7 @@ nfs_file_read(struct kiocb *iocb, char __user * buf, size_t count, loff_t pos) dentry->d_parent->d_name.name, dentry->d_name.name, (unsigned long) count, (unsigned long) pos); - result = nfs_revalidate_file(inode, iocb->ki_filp); + result = nfs_revalidate_mapping(inode, iocb->ki_filp->f_mapping); nfs_add_stats(inode, NFSIOS_NORMALREADBYTES, count); if (!result) result = generic_file_aio_read(iocb, buf, count, pos); @@ -247,7 +230,7 @@ nfs_file_sendfile(struct file *filp, loff_t *ppos, size_t count, dentry->d_parent->d_name.name, dentry->d_name.name, (unsigned long) count, (unsigned long long) *ppos); - res = nfs_revalidate_file(inode, filp); + res = nfs_revalidate_mapping(inode, filp->f_mapping); if (!res) res = generic_file_sendfile(filp, ppos, count, actor, target); return res; @@ -263,7 +246,7 @@ nfs_file_mmap(struct file * file, struct vm_area_struct * vma) dfprintk(VFS, "nfs: mmap(%s/%s)\n", dentry->d_parent->d_name.name, dentry->d_name.name); - status = nfs_revalidate_file(inode, file); + status = nfs_revalidate_mapping(inode, file->f_mapping); if (!status) status = generic_file_mmap(file, vma); return status; @@ -320,15 +303,25 @@ static int nfs_commit_write(struct file *file, struct page *page, unsigned offse static void nfs_invalidate_page(struct page *page, unsigned long offset) { - /* FIXME: we really should cancel any unstarted writes on this page */ + struct inode *inode = page->mapping->host; + + /* Cancel any unstarted writes on this page */ + if (offset == 0) + nfs_sync_inode_wait(inode, page->index, 1, FLUSH_INVALIDATE); } static int nfs_release_page(struct page *page, gfp_t gfp) { - return !nfs_wb_page(page->mapping->host, page); + if (gfp & __GFP_FS) + return !nfs_wb_page(page->mapping->host, page); + else + /* + * Avoid deadlock on nfs_wait_on_request(). + */ + return 0; } -struct address_space_operations nfs_file_aops = { +const struct address_space_operations nfs_file_aops = { .readpage = nfs_readpage, .readpages = nfs_readpages, .set_page_dirty = __set_page_dirty_nobuffers, @@ -373,7 +366,6 @@ nfs_file_write(struct kiocb *iocb, const char __user *buf, size_t count, loff_t if (result) goto out; } - nfs_revalidate_mapping(inode, iocb->ki_filp->f_mapping); result = count; if (!count) diff --git a/fs/nfs/getroot.c b/fs/nfs/getroot.c new file mode 100644 index 00000000000..76b08ae9ed8 --- /dev/null +++ b/fs/nfs/getroot.c @@ -0,0 +1,311 @@ +/* getroot.c: get the root dentry for an NFS mount + * + * Copyright (C) 2006 Red Hat, Inc. All Rights Reserved. + * Written by David Howells (dhowells@redhat.com) + * + * 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. + */ + +#include <linux/config.h> +#include <linux/module.h> +#include <linux/init.h> + +#include <linux/time.h> +#include <linux/kernel.h> +#include <linux/mm.h> +#include <linux/string.h> +#include <linux/stat.h> +#include <linux/errno.h> +#include <linux/unistd.h> +#include <linux/sunrpc/clnt.h> +#include <linux/sunrpc/stats.h> +#include <linux/nfs_fs.h> +#include <linux/nfs_mount.h> +#include <linux/nfs4_mount.h> +#include <linux/lockd/bind.h> +#include <linux/smp_lock.h> +#include <linux/seq_file.h> +#include <linux/mount.h> +#include <linux/nfs_idmap.h> +#include <linux/vfs.h> +#include <linux/namei.h> +#include <linux/namespace.h> +#include <linux/security.h> + +#include <asm/system.h> +#include <asm/uaccess.h> + +#include "nfs4_fs.h" +#include "delegation.h" +#include "internal.h" + +#define NFSDBG_FACILITY NFSDBG_CLIENT +#define NFS_PARANOIA 1 + +/* + * get an NFS2/NFS3 root dentry from the root filehandle + */ +struct dentry *nfs_get_root(struct super_block *sb, struct nfs_fh *mntfh) +{ + struct nfs_server *server = NFS_SB(sb); + struct nfs_fsinfo fsinfo; + struct nfs_fattr fattr; + struct dentry *mntroot; + struct inode *inode; + int error; + + /* create a dummy root dentry with dummy inode for this superblock */ + if (!sb->s_root) { + struct nfs_fh dummyfh; + struct dentry *root; + struct inode *iroot; + + memset(&dummyfh, 0, sizeof(dummyfh)); + memset(&fattr, 0, sizeof(fattr)); + nfs_fattr_init(&fattr); + fattr.valid = NFS_ATTR_FATTR; + fattr.type = NFDIR; + fattr.mode = S_IFDIR | S_IRUSR | S_IWUSR; + fattr.nlink = 2; + + iroot = nfs_fhget(sb, &dummyfh, &fattr); + if (IS_ERR(iroot)) + return ERR_PTR(PTR_ERR(iroot)); + + root = d_alloc_root(iroot); + if (!root) { + iput(iroot); + return ERR_PTR(-ENOMEM); + } + + sb->s_root = root; + } + + /* get the actual root for this mount */ + fsinfo.fattr = &fattr; + + error = server->nfs_client->rpc_ops->getroot(server, mntfh, &fsinfo); + if (error < 0) { + dprintk("nfs_get_root: getattr error = %d\n", -error); + return ERR_PTR(error); + } + + inode = nfs_fhget(sb, mntfh, fsinfo.fattr); + if (IS_ERR(inode)) { + dprintk("nfs_get_root: get root inode failed\n"); + return ERR_PTR(PTR_ERR(inode)); + } + + /* root dentries normally start off anonymous and get spliced in later + * if the dentry tree reaches them; however if the dentry already + * exists, we'll pick it up at this point and use it as the root + */ + mntroot = d_alloc_anon(inode); + if (!mntroot) { + iput(inode); + dprintk("nfs_get_root: get root dentry failed\n"); + return ERR_PTR(-ENOMEM); + } + + security_d_instantiate(mntroot, inode); + + if (!mntroot->d_op) + mntroot->d_op = server->nfs_client->rpc_ops->dentry_ops; + + return mntroot; +} + +#ifdef CONFIG_NFS_V4 + +/* + * Do a simple pathwalk from the root FH of the server to the nominated target + * of the mountpoint + * - give error on symlinks + * - give error on ".." occurring in the path + * - follow traversals + */ +int nfs4_path_walk(struct nfs_server *server, + struct nfs_fh *mntfh, + const char *path) +{ + struct nfs_fsinfo fsinfo; + struct nfs_fattr fattr; + struct nfs_fh lastfh; + struct qstr name; + int ret; + //int referral_count = 0; + + dprintk("--> nfs4_path_walk(,,%s)\n", path); + + fsinfo.fattr = &fattr; + nfs_fattr_init(&fattr); + + if (*path++ != '/') { + dprintk("nfs4_get_root: Path does not begin with a slash\n"); + return -EINVAL; + } + + /* Start by getting the root filehandle from the server */ + ret = server->nfs_client->rpc_ops->getroot(server, mntfh, &fsinfo); + if (ret < 0) { + dprintk("nfs4_get_root: getroot error = %d\n", -ret); + return ret; + } + + if (fattr.type != NFDIR) { + printk(KERN_ERR "nfs4_get_root:" + " getroot encountered non-directory\n"); + return -ENOTDIR; + } + + if (fattr.valid & NFS_ATTR_FATTR_V4_REFERRAL) { + printk(KERN_ERR "nfs4_get_root:" + " getroot obtained referral\n"); + return -EREMOTE; + } + +next_component: + dprintk("Next: %s\n", path); + + /* extract the next bit of the path */ + if (!*path) + goto path_walk_complete; + + name.name = path; + while (*path && *path != '/') + path++; + name.len = path - (const char *) name.name; + +eat_dot_dir: + while (*path == '/') + path++; + + if (path[0] == '.' && (path[1] == '/' || !path[1])) { + path += 2; + goto eat_dot_dir; + } + + if (path[0] == '.' && path[1] == '.' && (path[2] == '/' || !path[2]) + ) { + printk(KERN_ERR "nfs4_get_root:" + " Mount path contains reference to \"..\"\n"); + return -EINVAL; + } + + /* lookup the next FH in the sequence */ + memcpy(&lastfh, mntfh, sizeof(lastfh)); + + dprintk("LookupFH: %*.*s [%s]\n", name.len, name.len, name.name, path); + + ret = server->nfs_client->rpc_ops->lookupfh(server, &lastfh, &name, + mntfh, &fattr); + if (ret < 0) { + dprintk("nfs4_get_root: getroot error = %d\n", -ret); + return ret; + } + + if (fattr.type != NFDIR) { + printk(KERN_ERR "nfs4_get_root:" + " lookupfh encountered non-directory\n"); + return -ENOTDIR; + } + + if (fattr.valid & NFS_ATTR_FATTR_V4_REFERRAL) { + printk(KERN_ERR "nfs4_get_root:" + " lookupfh obtained referral\n"); + return -EREMOTE; + } + + goto next_component; + +path_walk_complete: + memcpy(&server->fsid, &fattr.fsid, sizeof(server->fsid)); + dprintk("<-- nfs4_path_walk() = 0\n"); + return 0; +} + +/* + * get an NFS4 root dentry from the root filehandle + */ +struct dentry *nfs4_get_root(struct super_block *sb, struct nfs_fh *mntfh) +{ + struct nfs_server *server = NFS_SB(sb); + struct nfs_fattr fattr; + struct dentry *mntroot; + struct inode *inode; + int error; + + dprintk("--> nfs4_get_root()\n"); + + /* create a dummy root dentry with dummy inode for this superblock */ + if (!sb->s_root) { + struct nfs_fh dummyfh; + struct dentry *root; + struct inode *iroot; + + memset(&dummyfh, 0, sizeof(dummyfh)); + memset(&fattr, 0, sizeof(fattr)); + nfs_fattr_init(&fattr); + fattr.valid = NFS_ATTR_FATTR; + fattr.type = NFDIR; + fattr.mode = S_IFDIR | S_IRUSR | S_IWUSR; + fattr.nlink = 2; + + iroot = nfs_fhget(sb, &dummyfh, &fattr); + if (IS_ERR(iroot)) + return ERR_PTR(PTR_ERR(iroot)); + + root = d_alloc_root(iroot); + if (!root) { + iput(iroot); + return ERR_PTR(-ENOMEM); + } + + sb->s_root = root; + } + + /* get the info about the server and filesystem */ + error = nfs4_server_capabilities(server, mntfh); + if (error < 0) { + dprintk("nfs_get_root: getcaps error = %d\n", + -error); + return ERR_PTR(error); + } + + /* get the actual root for this mount */ + error = server->nfs_client->rpc_ops->getattr(server, mntfh, &fattr); + if (error < 0) { + dprintk("nfs_get_root: getattr error = %d\n", -error); + return ERR_PTR(error); + } + + inode = nfs_fhget(sb, mntfh, &fattr); + if (IS_ERR(inode)) { + dprintk("nfs_get_root: get root inode failed\n"); + return ERR_PTR(PTR_ERR(inode)); + } + + /* root dentries normally start off anonymous and get spliced in later + * if the dentry tree reaches them; however if the dentry already + * exists, we'll pick it up at this point and use it as the root + */ + mntroot = d_alloc_anon(inode); + if (!mntroot) { + iput(inode); + dprintk("nfs_get_root: get root dentry failed\n"); + return ERR_PTR(-ENOMEM); + } + + security_d_instantiate(mntroot, inode); + + if (!mntroot->d_op) + mntroot->d_op = server->nfs_client->rpc_ops->dentry_ops; + + dprintk("<-- nfs4_get_root()\n"); + return mntroot; +} + +#endif /* CONFIG_NFS_V4 */ diff --git a/fs/nfs/idmap.c b/fs/nfs/idmap.c index 3fab5b0cfc5..82ad7110a1c 100644 --- a/fs/nfs/idmap.c +++ b/fs/nfs/idmap.c @@ -47,7 +47,6 @@ #include <linux/workqueue.h> #include <linux/sunrpc/rpc_pipe_fs.h> -#include <linux/nfs_fs_sb.h> #include <linux/nfs_fs.h> #include <linux/nfs_idmap.h> @@ -58,6 +57,20 @@ /* Default cache timeout is 10 minutes */ unsigned int nfs_idmap_cache_timeout = 600 * HZ; +static int param_set_idmap_timeout(const char *val, struct kernel_param *kp) +{ + char *endp; + int num = simple_strtol(val, &endp, 0); + int jif = num * HZ; + if (endp == val || *endp || num < 0 || jif < num) + return -EINVAL; + *((int *)kp->arg) = jif; + return 0; +} + +module_param_call(idmap_cache_timeout, param_set_idmap_timeout, param_get_int, + &nfs_idmap_cache_timeout, 0644); + struct idmap_hashent { unsigned long ih_expires; __u32 ih_id; @@ -71,7 +84,6 @@ struct idmap_hashtable { }; struct idmap { - char idmap_path[48]; struct dentry *idmap_dentry; wait_queue_head_t idmap_wq; struct idmap_msg idmap_im; @@ -95,24 +107,23 @@ static struct rpc_pipe_ops idmap_upcall_ops = { .destroy_msg = idmap_pipe_destroy_msg, }; -void -nfs_idmap_new(struct nfs4_client *clp) +int +nfs_idmap_new(struct nfs_client *clp) { struct idmap *idmap; + int error; - if (clp->cl_idmap != NULL) - return; - if ((idmap = kzalloc(sizeof(*idmap), GFP_KERNEL)) == NULL) - return; + BUG_ON(clp->cl_idmap != NULL); - snprintf(idmap->idmap_path, sizeof(idmap->idmap_path), - "%s/idmap", clp->cl_rpcclient->cl_pathname); + if ((idmap = kzalloc(sizeof(*idmap), GFP_KERNEL)) == NULL) + return -ENOMEM; - idmap->idmap_dentry = rpc_mkpipe(idmap->idmap_path, + idmap->idmap_dentry = rpc_mkpipe(clp->cl_rpcclient->cl_dentry, "idmap", idmap, &idmap_upcall_ops, 0); if (IS_ERR(idmap->idmap_dentry)) { + error = PTR_ERR(idmap->idmap_dentry); kfree(idmap); - return; + return error; } mutex_init(&idmap->idmap_lock); @@ -122,18 +133,17 @@ nfs_idmap_new(struct nfs4_client *clp) idmap->idmap_group_hash.h_type = IDMAP_TYPE_GROUP; clp->cl_idmap = idmap; + return 0; } void -nfs_idmap_delete(struct nfs4_client *clp) +nfs_idmap_delete(struct nfs_client *clp) { struct idmap *idmap = clp->cl_idmap; if (!idmap) return; - dput(idmap->idmap_dentry); - idmap->idmap_dentry = NULL; - rpc_unlink(idmap->idmap_path); + rpc_unlink(idmap->idmap_dentry); clp->cl_idmap = NULL; kfree(idmap); } @@ -480,27 +490,27 @@ static unsigned int fnvhash32(const void *buf, size_t buflen) return (hash); } -int nfs_map_name_to_uid(struct nfs4_client *clp, const char *name, size_t namelen, __u32 *uid) +int nfs_map_name_to_uid(struct nfs_client *clp, const char *name, size_t namelen, __u32 *uid) { struct idmap *idmap = clp->cl_idmap; return nfs_idmap_id(idmap, &idmap->idmap_user_hash, name, namelen, uid); } -int nfs_map_group_to_gid(struct nfs4_client *clp, const char *name, size_t namelen, __u32 *uid) +int nfs_map_group_to_gid(struct nfs_client *clp, const char *name, size_t namelen, __u32 *uid) { struct idmap *idmap = clp->cl_idmap; return nfs_idmap_id(idmap, &idmap->idmap_group_hash, name, namelen, uid); } -int nfs_map_uid_to_name(struct nfs4_client *clp, __u32 uid, char *buf) +int nfs_map_uid_to_name(struct nfs_client *clp, __u32 uid, char *buf) { struct idmap *idmap = clp->cl_idmap; return nfs_idmap_name(idmap, &idmap->idmap_user_hash, uid, buf); } -int nfs_map_gid_to_group(struct nfs4_client *clp, __u32 uid, char *buf) +int nfs_map_gid_to_group(struct nfs_client *clp, __u32 uid, char *buf) { struct idmap *idmap = clp->cl_idmap; diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c index 937fbfc381b..bc9376ca86c 100644 --- a/fs/nfs/inode.c +++ b/fs/nfs/inode.c @@ -13,7 +13,6 @@ * */ -#include <linux/config.h> #include <linux/module.h> #include <linux/init.h> @@ -36,6 +35,8 @@ #include <linux/mount.h> #include <linux/nfs_idmap.h> #include <linux/vfs.h> +#include <linux/inet.h> +#include <linux/nfs_xdr.h> #include <asm/system.h> #include <asm/uaccess.h> @@ -44,89 +45,17 @@ #include "callback.h" #include "delegation.h" #include "iostat.h" +#include "internal.h" #define NFSDBG_FACILITY NFSDBG_VFS #define NFS_PARANOIA 1 -/* Maximum number of readahead requests - * FIXME: this should really be a sysctl so that users may tune it to suit - * their needs. People that do NFS over a slow network, might for - * instance want to reduce it to something closer to 1 for improved - * interactive response. - */ -#define NFS_MAX_READAHEAD (RPC_DEF_SLOT_TABLE - 1) - static void nfs_invalidate_inode(struct inode *); static int nfs_update_inode(struct inode *, struct nfs_fattr *); -static struct inode *nfs_alloc_inode(struct super_block *sb); -static void nfs_destroy_inode(struct inode *); -static int nfs_write_inode(struct inode *,int); -static void nfs_delete_inode(struct inode *); -static void nfs_clear_inode(struct inode *); -static void nfs_umount_begin(struct super_block *); -static int nfs_statfs(struct dentry *, struct kstatfs *); -static int nfs_show_options(struct seq_file *, struct vfsmount *); -static int nfs_show_stats(struct seq_file *, struct vfsmount *); static void nfs_zap_acl_cache(struct inode *); -static struct rpc_program nfs_program; - -static struct super_operations nfs_sops = { - .alloc_inode = nfs_alloc_inode, - .destroy_inode = nfs_destroy_inode, - .write_inode = nfs_write_inode, - .delete_inode = nfs_delete_inode, - .statfs = nfs_statfs, - .clear_inode = nfs_clear_inode, - .umount_begin = nfs_umount_begin, - .show_options = nfs_show_options, - .show_stats = nfs_show_stats, -}; - -/* - * RPC cruft for NFS - */ -static struct rpc_stat nfs_rpcstat = { - .program = &nfs_program -}; -static struct rpc_version * nfs_version[] = { - NULL, - NULL, - &nfs_version2, -#if defined(CONFIG_NFS_V3) - &nfs_version3, -#elif defined(CONFIG_NFS_V4) - NULL, -#endif -#if defined(CONFIG_NFS_V4) - &nfs_version4, -#endif -}; - -static struct rpc_program nfs_program = { - .name = "nfs", - .number = NFS_PROGRAM, - .nrvers = ARRAY_SIZE(nfs_version), - .version = nfs_version, - .stats = &nfs_rpcstat, - .pipe_dir_name = "/nfs", -}; - -#ifdef CONFIG_NFS_V3_ACL -static struct rpc_stat nfsacl_rpcstat = { &nfsacl_program }; -static struct rpc_version * nfsacl_version[] = { - [3] = &nfsacl_version3, -}; - -struct rpc_program nfsacl_program = { - .name = "nfsacl", - .number = NFS_ACL_PROGRAM, - .nrvers = ARRAY_SIZE(nfsacl_version), - .version = nfsacl_version, - .stats = &nfsacl_rpcstat, -}; -#endif /* CONFIG_NFS_V3_ACL */ +static kmem_cache_t * nfs_inode_cachep; static inline unsigned long nfs_fattr_to_ino_t(struct nfs_fattr *fattr) @@ -134,8 +63,7 @@ nfs_fattr_to_ino_t(struct nfs_fattr *fattr) return nfs_fileid_to_ino_t(fattr->fileid); } -static int -nfs_write_inode(struct inode *inode, int sync) +int nfs_write_inode(struct inode *inode, int sync) { int flags = sync ? FLUSH_SYNC : 0; int ret; @@ -146,586 +74,16 @@ nfs_write_inode(struct inode *inode, int sync) return 0; } -static void -nfs_delete_inode(struct inode * inode) +void nfs_clear_inode(struct inode *inode) { - dprintk("NFS: delete_inode(%s/%ld)\n", inode->i_sb->s_id, inode->i_ino); - - truncate_inode_pages(&inode->i_data, 0); - - nfs_wb_all(inode); /* * The following should never happen... */ - if (nfs_have_writebacks(inode)) { - printk(KERN_ERR "nfs_delete_inode: inode %ld has pending RPC requests\n", inode->i_ino); - } - - clear_inode(inode); -} - -static void -nfs_clear_inode(struct inode *inode) -{ - struct nfs_inode *nfsi = NFS_I(inode); - struct rpc_cred *cred; - - nfs_wb_all(inode); - BUG_ON (!list_empty(&nfsi->open_files)); + BUG_ON(nfs_have_writebacks(inode)); + BUG_ON(!list_empty(&NFS_I(inode)->open_files)); + BUG_ON(atomic_read(&NFS_I(inode)->data_updates) != 0); nfs_zap_acl_cache(inode); - cred = nfsi->cache_access.cred; - if (cred) - put_rpccred(cred); - BUG_ON(atomic_read(&nfsi->data_updates) != 0); -} - -void -nfs_umount_begin(struct super_block *sb) -{ - struct rpc_clnt *rpc = NFS_SB(sb)->client; - - /* -EIO all pending I/O */ - if (!IS_ERR(rpc)) - rpc_killall_tasks(rpc); - rpc = NFS_SB(sb)->client_acl; - if (!IS_ERR(rpc)) - rpc_killall_tasks(rpc); -} - - -static inline unsigned long -nfs_block_bits(unsigned long bsize, unsigned char *nrbitsp) -{ - /* make sure blocksize is a power of two */ - if ((bsize & (bsize - 1)) || nrbitsp) { - unsigned char nrbits; - - for (nrbits = 31; nrbits && !(bsize & (1 << nrbits)); nrbits--) - ; - bsize = 1 << nrbits; - if (nrbitsp) - *nrbitsp = nrbits; - } - - return bsize; -} - -/* - * Calculate the number of 512byte blocks used. - */ -static inline unsigned long -nfs_calc_block_size(u64 tsize) -{ - loff_t used = (tsize + 511) >> 9; - return (used > ULONG_MAX) ? ULONG_MAX : used; -} - -/* - * Compute and set NFS server blocksize - */ -static inline unsigned long -nfs_block_size(unsigned long bsize, unsigned char *nrbitsp) -{ - if (bsize < NFS_MIN_FILE_IO_SIZE) - bsize = NFS_DEF_FILE_IO_SIZE; - else if (bsize >= NFS_MAX_FILE_IO_SIZE) - bsize = NFS_MAX_FILE_IO_SIZE; - - return nfs_block_bits(bsize, nrbitsp); -} - -/* - * Obtain the root inode of the file system. - */ -static struct inode * -nfs_get_root(struct super_block *sb, struct nfs_fh *rootfh, struct nfs_fsinfo *fsinfo) -{ - struct nfs_server *server = NFS_SB(sb); - int error; - - error = server->rpc_ops->getroot(server, rootfh, fsinfo); - if (error < 0) { - dprintk("nfs_get_root: getattr error = %d\n", -error); - return ERR_PTR(error); - } - - return nfs_fhget(sb, rootfh, fsinfo->fattr); -} - -/* - * Do NFS version-independent mount processing, and sanity checking - */ -static int -nfs_sb_init(struct super_block *sb, rpc_authflavor_t authflavor) -{ - struct nfs_server *server; - struct inode *root_inode; - struct nfs_fattr fattr; - struct nfs_fsinfo fsinfo = { - .fattr = &fattr, - }; - struct nfs_pathconf pathinfo = { - .fattr = &fattr, - }; - int no_root_error = 0; - unsigned long max_rpc_payload; - - /* We probably want something more informative here */ - snprintf(sb->s_id, sizeof(sb->s_id), "%x:%x", MAJOR(sb->s_dev), MINOR(sb->s_dev)); - - server = NFS_SB(sb); - - sb->s_magic = NFS_SUPER_MAGIC; - - server->io_stats = nfs_alloc_iostats(); - if (server->io_stats == NULL) - return -ENOMEM; - - root_inode = nfs_get_root(sb, &server->fh, &fsinfo); - /* Did getting the root inode fail? */ - if (IS_ERR(root_inode)) { - no_root_error = PTR_ERR(root_inode); - goto out_no_root; - } - sb->s_root = d_alloc_root(root_inode); - if (!sb->s_root) { - no_root_error = -ENOMEM; - goto out_no_root; - } - sb->s_root->d_op = server->rpc_ops->dentry_ops; - - /* mount time stamp, in seconds */ - server->mount_time = jiffies; - - /* Get some general file system info */ - if (server->namelen == 0 && - server->rpc_ops->pathconf(server, &server->fh, &pathinfo) >= 0) - server->namelen = pathinfo.max_namelen; - /* Work out a lot of parameters */ - if (server->rsize == 0) - server->rsize = nfs_block_size(fsinfo.rtpref, NULL); - if (server->wsize == 0) - server->wsize = nfs_block_size(fsinfo.wtpref, NULL); - - if (fsinfo.rtmax >= 512 && server->rsize > fsinfo.rtmax) - server->rsize = nfs_block_size(fsinfo.rtmax, NULL); - if (fsinfo.wtmax >= 512 && server->wsize > fsinfo.wtmax) - server->wsize = nfs_block_size(fsinfo.wtmax, NULL); - - max_rpc_payload = nfs_block_size(rpc_max_payload(server->client), NULL); - if (server->rsize > max_rpc_payload) - server->rsize = max_rpc_payload; - if (server->rsize > NFS_MAX_FILE_IO_SIZE) - server->rsize = NFS_MAX_FILE_IO_SIZE; - server->rpages = (server->rsize + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT; - - if (server->wsize > max_rpc_payload) - server->wsize = max_rpc_payload; - if (server->wsize > NFS_MAX_FILE_IO_SIZE) - server->wsize = NFS_MAX_FILE_IO_SIZE; - server->wpages = (server->wsize + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT; - - if (sb->s_blocksize == 0) - sb->s_blocksize = nfs_block_bits(server->wsize, - &sb->s_blocksize_bits); - server->wtmult = nfs_block_bits(fsinfo.wtmult, NULL); - - server->dtsize = nfs_block_size(fsinfo.dtpref, NULL); - if (server->dtsize > PAGE_CACHE_SIZE) - server->dtsize = PAGE_CACHE_SIZE; - if (server->dtsize > server->rsize) - server->dtsize = server->rsize; - - if (server->flags & NFS_MOUNT_NOAC) { - server->acregmin = server->acregmax = 0; - server->acdirmin = server->acdirmax = 0; - sb->s_flags |= MS_SYNCHRONOUS; - } - server->backing_dev_info.ra_pages = server->rpages * NFS_MAX_READAHEAD; - - sb->s_maxbytes = fsinfo.maxfilesize; - if (sb->s_maxbytes > MAX_LFS_FILESIZE) - sb->s_maxbytes = MAX_LFS_FILESIZE; - - server->client->cl_intr = (server->flags & NFS_MOUNT_INTR) ? 1 : 0; - server->client->cl_softrtry = (server->flags & NFS_MOUNT_SOFT) ? 1 : 0; - - /* We're airborne Set socket buffersize */ - rpc_setbufsize(server->client, server->wsize + 100, server->rsize + 100); - return 0; - /* Yargs. It didn't work out. */ -out_no_root: - dprintk("nfs_sb_init: get root inode failed: errno %d\n", -no_root_error); - if (!IS_ERR(root_inode)) - iput(root_inode); - return no_root_error; -} - -static void nfs_init_timeout_values(struct rpc_timeout *to, int proto, unsigned int timeo, unsigned int retrans) -{ - to->to_initval = timeo * HZ / 10; - to->to_retries = retrans; - if (!to->to_retries) - to->to_retries = 2; - - switch (proto) { - case IPPROTO_TCP: - if (!to->to_initval) - to->to_initval = 60 * HZ; - if (to->to_initval > NFS_MAX_TCP_TIMEOUT) - to->to_initval = NFS_MAX_TCP_TIMEOUT; - to->to_increment = to->to_initval; - to->to_maxval = to->to_initval + (to->to_increment * to->to_retries); - to->to_exponential = 0; - break; - case IPPROTO_UDP: - default: - if (!to->to_initval) - to->to_initval = 11 * HZ / 10; - if (to->to_initval > NFS_MAX_UDP_TIMEOUT) - to->to_initval = NFS_MAX_UDP_TIMEOUT; - to->to_maxval = NFS_MAX_UDP_TIMEOUT; - to->to_exponential = 1; - break; - } -} - -/* - * Create an RPC client handle. - */ -static struct rpc_clnt * -nfs_create_client(struct nfs_server *server, const struct nfs_mount_data *data) -{ - struct rpc_timeout timeparms; - struct rpc_xprt *xprt = NULL; - struct rpc_clnt *clnt = NULL; - int proto = (data->flags & NFS_MOUNT_TCP) ? IPPROTO_TCP : IPPROTO_UDP; - - nfs_init_timeout_values(&timeparms, proto, data->timeo, data->retrans); - - server->retrans_timeo = timeparms.to_initval; - server->retrans_count = timeparms.to_retries; - - /* create transport and client */ - xprt = xprt_create_proto(proto, &server->addr, &timeparms); - if (IS_ERR(xprt)) { - dprintk("%s: cannot create RPC transport. Error = %ld\n", - __FUNCTION__, PTR_ERR(xprt)); - return (struct rpc_clnt *)xprt; - } - clnt = rpc_create_client(xprt, server->hostname, &nfs_program, - server->rpc_ops->version, data->pseudoflavor); - if (IS_ERR(clnt)) { - dprintk("%s: cannot create RPC client. Error = %ld\n", - __FUNCTION__, PTR_ERR(xprt)); - goto out_fail; - } - - clnt->cl_intr = 1; - clnt->cl_softrtry = 1; - - return clnt; - -out_fail: - return clnt; -} - -/* - * The way this works is that the mount process passes a structure - * in the data argument which contains the server's IP address - * and the root file handle obtained from the server's mount - * daemon. We stash these away in the private superblock fields. - */ -static int -nfs_fill_super(struct super_block *sb, struct nfs_mount_data *data, int silent) -{ - struct nfs_server *server; - rpc_authflavor_t authflavor; - - server = NFS_SB(sb); - sb->s_blocksize_bits = 0; - sb->s_blocksize = 0; - if (data->bsize) - sb->s_blocksize = nfs_block_size(data->bsize, &sb->s_blocksize_bits); - if (data->rsize) - server->rsize = nfs_block_size(data->rsize, NULL); - if (data->wsize) - server->wsize = nfs_block_size(data->wsize, NULL); - server->flags = data->flags & NFS_MOUNT_FLAGMASK; - - server->acregmin = data->acregmin*HZ; - server->acregmax = data->acregmax*HZ; - server->acdirmin = data->acdirmin*HZ; - server->acdirmax = data->acdirmax*HZ; - - /* Start lockd here, before we might error out */ - if (!(server->flags & NFS_MOUNT_NONLM)) - lockd_up(); - - server->namelen = data->namlen; - server->hostname = kmalloc(strlen(data->hostname) + 1, GFP_KERNEL); - if (!server->hostname) - return -ENOMEM; - strcpy(server->hostname, data->hostname); - - /* Check NFS protocol revision and initialize RPC op vector - * and file handle pool. */ -#ifdef CONFIG_NFS_V3 - if (server->flags & NFS_MOUNT_VER3) { - server->rpc_ops = &nfs_v3_clientops; - server->caps |= NFS_CAP_READDIRPLUS; - } else { - server->rpc_ops = &nfs_v2_clientops; - } -#else - server->rpc_ops = &nfs_v2_clientops; -#endif - - /* Fill in pseudoflavor for mount version < 5 */ - if (!(data->flags & NFS_MOUNT_SECFLAVOUR)) - data->pseudoflavor = RPC_AUTH_UNIX; - authflavor = data->pseudoflavor; /* save for sb_init() */ - /* XXX maybe we want to add a server->pseudoflavor field */ - - /* Create RPC client handles */ - server->client = nfs_create_client(server, data); - if (IS_ERR(server->client)) - return PTR_ERR(server->client); - /* RFC 2623, sec 2.3.2 */ - if (authflavor != RPC_AUTH_UNIX) { - struct rpc_auth *auth; - - server->client_sys = rpc_clone_client(server->client); - if (IS_ERR(server->client_sys)) - return PTR_ERR(server->client_sys); - auth = rpcauth_create(RPC_AUTH_UNIX, server->client_sys); - if (IS_ERR(auth)) - return PTR_ERR(auth); - } else { - atomic_inc(&server->client->cl_count); - server->client_sys = server->client; - } - if (server->flags & NFS_MOUNT_VER3) { -#ifdef CONFIG_NFS_V3_ACL - if (!(server->flags & NFS_MOUNT_NOACL)) { - server->client_acl = rpc_bind_new_program(server->client, &nfsacl_program, 3); - /* No errors! Assume that Sun nfsacls are supported */ - if (!IS_ERR(server->client_acl)) - server->caps |= NFS_CAP_ACLS; - } -#else - server->flags &= ~NFS_MOUNT_NOACL; -#endif /* CONFIG_NFS_V3_ACL */ - /* - * The VFS shouldn't apply the umask to mode bits. We will - * do so ourselves when necessary. - */ - sb->s_flags |= MS_POSIXACL; - if (server->namelen == 0 || server->namelen > NFS3_MAXNAMLEN) - server->namelen = NFS3_MAXNAMLEN; - sb->s_time_gran = 1; - } else { - if (server->namelen == 0 || server->namelen > NFS2_MAXNAMLEN) - server->namelen = NFS2_MAXNAMLEN; - } - - sb->s_op = &nfs_sops; - return nfs_sb_init(sb, authflavor); -} - -static int -nfs_statfs(struct dentry *dentry, struct kstatfs *buf) -{ - struct super_block *sb = dentry->d_sb; - struct nfs_server *server = NFS_SB(sb); - unsigned char blockbits; - unsigned long blockres; - struct nfs_fh *rootfh = NFS_FH(sb->s_root->d_inode); - struct nfs_fattr fattr; - struct nfs_fsstat res = { - .fattr = &fattr, - }; - int error; - - lock_kernel(); - - error = server->rpc_ops->statfs(server, rootfh, &res); - buf->f_type = NFS_SUPER_MAGIC; - if (error < 0) - goto out_err; - - /* - * Current versions of glibc do not correctly handle the - * case where f_frsize != f_bsize. Eventually we want to - * report the value of wtmult in this field. - */ - buf->f_frsize = sb->s_blocksize; - - /* - * On most *nix systems, f_blocks, f_bfree, and f_bavail - * are reported in units of f_frsize. Linux hasn't had - * an f_frsize field in its statfs struct until recently, - * thus historically Linux's sys_statfs reports these - * fields in units of f_bsize. - */ - buf->f_bsize = sb->s_blocksize; - blockbits = sb->s_blocksize_bits; - blockres = (1 << blockbits) - 1; - buf->f_blocks = (res.tbytes + blockres) >> blockbits; - buf->f_bfree = (res.fbytes + blockres) >> blockbits; - buf->f_bavail = (res.abytes + blockres) >> blockbits; - - buf->f_files = res.tfiles; - buf->f_ffree = res.afiles; - - buf->f_namelen = server->namelen; - out: - unlock_kernel(); - return 0; - - out_err: - dprintk("%s: statfs error = %d\n", __FUNCTION__, -error); - buf->f_bsize = buf->f_blocks = buf->f_bfree = buf->f_bavail = -1; - goto out; - -} - -static void nfs_show_mount_options(struct seq_file *m, struct nfs_server *nfss, int showdefaults) -{ - static struct proc_nfs_info { - int flag; - char *str; - char *nostr; - } nfs_info[] = { - { NFS_MOUNT_SOFT, ",soft", ",hard" }, - { NFS_MOUNT_INTR, ",intr", "" }, - { NFS_MOUNT_NOCTO, ",nocto", "" }, - { NFS_MOUNT_NOAC, ",noac", "" }, - { NFS_MOUNT_NONLM, ",nolock", "" }, - { NFS_MOUNT_NOACL, ",noacl", "" }, - { 0, NULL, NULL } - }; - struct proc_nfs_info *nfs_infop; - char buf[12]; - char *proto; - - seq_printf(m, ",vers=%d", nfss->rpc_ops->version); - seq_printf(m, ",rsize=%d", nfss->rsize); - seq_printf(m, ",wsize=%d", nfss->wsize); - if (nfss->acregmin != 3*HZ || showdefaults) - seq_printf(m, ",acregmin=%d", nfss->acregmin/HZ); - if (nfss->acregmax != 60*HZ || showdefaults) - seq_printf(m, ",acregmax=%d", nfss->acregmax/HZ); - if (nfss->acdirmin != 30*HZ || showdefaults) - seq_printf(m, ",acdirmin=%d", nfss->acdirmin/HZ); - if (nfss->acdirmax != 60*HZ || showdefaults) - seq_printf(m, ",acdirmax=%d", nfss->acdirmax/HZ); - for (nfs_infop = nfs_info; nfs_infop->flag; nfs_infop++) { - if (nfss->flags & nfs_infop->flag) - seq_puts(m, nfs_infop->str); - else - seq_puts(m, nfs_infop->nostr); - } - switch (nfss->client->cl_xprt->prot) { - case IPPROTO_TCP: - proto = "tcp"; - break; - case IPPROTO_UDP: - proto = "udp"; - break; - default: - snprintf(buf, sizeof(buf), "%u", nfss->client->cl_xprt->prot); - proto = buf; - } - seq_printf(m, ",proto=%s", proto); - seq_printf(m, ",timeo=%lu", 10U * nfss->retrans_timeo / HZ); - seq_printf(m, ",retrans=%u", nfss->retrans_count); -} - -static int nfs_show_options(struct seq_file *m, struct vfsmount *mnt) -{ - struct nfs_server *nfss = NFS_SB(mnt->mnt_sb); - - nfs_show_mount_options(m, nfss, 0); - - seq_puts(m, ",addr="); - seq_escape(m, nfss->hostname, " \t\n\\"); - - return 0; -} - -static int nfs_show_stats(struct seq_file *m, struct vfsmount *mnt) -{ - int i, cpu; - struct nfs_server *nfss = NFS_SB(mnt->mnt_sb); - struct rpc_auth *auth = nfss->client->cl_auth; - struct nfs_iostats totals = { }; - - seq_printf(m, "statvers=%s", NFS_IOSTAT_VERS); - - /* - * Display all mount option settings - */ - seq_printf(m, "\n\topts:\t"); - seq_puts(m, mnt->mnt_sb->s_flags & MS_RDONLY ? "ro" : "rw"); - seq_puts(m, mnt->mnt_sb->s_flags & MS_SYNCHRONOUS ? ",sync" : ""); - seq_puts(m, mnt->mnt_sb->s_flags & MS_NOATIME ? ",noatime" : ""); - seq_puts(m, mnt->mnt_sb->s_flags & MS_NODIRATIME ? ",nodiratime" : ""); - nfs_show_mount_options(m, nfss, 1); - - seq_printf(m, "\n\tage:\t%lu", (jiffies - nfss->mount_time) / HZ); - - seq_printf(m, "\n\tcaps:\t"); - seq_printf(m, "caps=0x%x", nfss->caps); - seq_printf(m, ",wtmult=%d", nfss->wtmult); - seq_printf(m, ",dtsize=%d", nfss->dtsize); - seq_printf(m, ",bsize=%d", nfss->bsize); - seq_printf(m, ",namelen=%d", nfss->namelen); - -#ifdef CONFIG_NFS_V4 - if (nfss->rpc_ops->version == 4) { - seq_printf(m, "\n\tnfsv4:\t"); - seq_printf(m, "bm0=0x%x", nfss->attr_bitmask[0]); - seq_printf(m, ",bm1=0x%x", nfss->attr_bitmask[1]); - seq_printf(m, ",acl=0x%x", nfss->acl_bitmask); - } -#endif - - /* - * Display security flavor in effect for this mount - */ - seq_printf(m, "\n\tsec:\tflavor=%d", auth->au_ops->au_flavor); - if (auth->au_flavor) - seq_printf(m, ",pseudoflavor=%d", auth->au_flavor); - - /* - * Display superblock I/O counters - */ - for_each_possible_cpu(cpu) { - struct nfs_iostats *stats; - - preempt_disable(); - stats = per_cpu_ptr(nfss->io_stats, cpu); - - for (i = 0; i < __NFSIOS_COUNTSMAX; i++) - totals.events[i] += stats->events[i]; - for (i = 0; i < __NFSIOS_BYTESMAX; i++) - totals.bytes[i] += stats->bytes[i]; - - preempt_enable(); - } - - seq_printf(m, "\n\tevents:\t"); - for (i = 0; i < __NFSIOS_COUNTSMAX; i++) - seq_printf(m, "%lu ", totals.events[i]); - seq_printf(m, "\n\tbytes:\t"); - for (i = 0; i < __NFSIOS_BYTESMAX; i++) - seq_printf(m, "%Lu ", totals.bytes[i]); - seq_printf(m, "\n"); - - rpc_print_iostats(m, nfss->client); - - return 0; + nfs_access_zap_cache(inode); } /** @@ -879,17 +237,25 @@ nfs_fhget(struct super_block *sb, struct nfs_fh *fh, struct nfs_fattr *fattr) /* Why so? Because we want revalidate for devices/FIFOs, and * that's precisely what we have in nfs_file_inode_operations. */ - inode->i_op = NFS_SB(sb)->rpc_ops->file_inode_ops; + inode->i_op = NFS_SB(sb)->nfs_client->rpc_ops->file_inode_ops; if (S_ISREG(inode->i_mode)) { inode->i_fop = &nfs_file_operations; inode->i_data.a_ops = &nfs_file_aops; inode->i_data.backing_dev_info = &NFS_SB(sb)->backing_dev_info; } else if (S_ISDIR(inode->i_mode)) { - inode->i_op = NFS_SB(sb)->rpc_ops->dir_inode_ops; + inode->i_op = NFS_SB(sb)->nfs_client->rpc_ops->dir_inode_ops; inode->i_fop = &nfs_dir_operations; if (nfs_server_capable(inode, NFS_CAP_READDIRPLUS) && fattr->size <= NFS_LIMIT_READDIRPLUS) set_bit(NFS_INO_ADVISE_RDPLUS, &NFS_FLAGS(inode)); + /* Deal with crossing mountpoints */ + if (!nfs_fsid_equal(&NFS_SB(sb)->fsid, &fattr->fsid)) { + if (fattr->valid & NFS_ATTR_FATTR_V4_REFERRAL) + inode->i_op = &nfs_referral_inode_operations; + else + inode->i_op = &nfs_mountpoint_inode_operations; + inode->i_fop = NULL; + } } else if (S_ISLNK(inode->i_mode)) inode->i_op = &nfs_symlink_inode_operations; else @@ -911,15 +277,13 @@ nfs_fhget(struct super_block *sb, struct nfs_fh *fh, struct nfs_fattr *fattr) * report the blocks in 512byte units */ inode->i_blocks = nfs_calc_block_size(fattr->du.nfs3.used); - inode->i_blksize = inode->i_sb->s_blocksize; } else { inode->i_blocks = fattr->du.nfs2.blocks; - inode->i_blksize = fattr->du.nfs2.blocksize; } nfsi->attrtimeo = NFS_MINATTRTIMEO(inode); nfsi->attrtimeo_timestamp = jiffies; memset(nfsi->cookieverf, 0, sizeof(nfsi->cookieverf)); - nfsi->cache_access.cred = NULL; + nfsi->access_cache = RB_ROOT; unlock_new_inode(inode); } else @@ -1077,7 +441,7 @@ static struct nfs_open_context *alloc_nfs_open_context(struct vfsmount *mnt, str { struct nfs_open_context *ctx; - ctx = (struct nfs_open_context *)kmalloc(sizeof(*ctx), GFP_KERNEL); + ctx = kmalloc(sizeof(*ctx), GFP_KERNEL); if (ctx != NULL) { atomic_set(&ctx->count, 1); ctx->dentry = dget(dentry); @@ -1208,6 +572,7 @@ __nfs_revalidate_inode(struct nfs_server *server, struct inode *inode) dfprintk(PAGECACHE, "NFS: revalidating (%s/%Ld)\n", inode->i_sb->s_id, (long long)NFS_FILEID(inode)); + nfs_inc_stats(inode, NFSIOS_INODEREVALIDATE); lock_kernel(); if (!inode || is_bad_inode(inode)) goto out_nowait; @@ -1221,7 +586,7 @@ __nfs_revalidate_inode(struct nfs_server *server, struct inode *inode) status = -ESTALE; /* Do we trust the cached ESTALE? */ if (NFS_ATTRTIMEO(inode) != 0) { - if (nfsi->cache_validity & (NFS_INO_INVALID_ATTR|NFS_INO_INVALID_DATA|NFS_INO_INVALID_ATIME)) { + if (nfsi->cache_validity & (NFS_INO_INVALID_ATTR|NFS_INO_INVALID_ATIME)) { /* no */ } else goto out; @@ -1252,8 +617,6 @@ __nfs_revalidate_inode(struct nfs_server *server, struct inode *inode) } spin_unlock(&inode->i_lock); - nfs_revalidate_mapping(inode, inode->i_mapping); - if (nfsi->cache_validity & NFS_INO_INVALID_ACL) nfs_zap_acl_cache(inode); @@ -1287,8 +650,7 @@ int nfs_attribute_timeout(struct inode *inode) */ int nfs_revalidate_inode(struct nfs_server *server, struct inode *inode) { - nfs_inc_stats(inode, NFSIOS_INODEREVALIDATE); - if (!(NFS_I(inode)->cache_validity & (NFS_INO_INVALID_ATTR|NFS_INO_INVALID_DATA)) + if (!(NFS_I(inode)->cache_validity & NFS_INO_INVALID_ATTR) && !nfs_attribute_timeout(inode)) return NFS_STALE(inode) ? -ESTALE : 0; return __nfs_revalidate_inode(server, inode); @@ -1299,9 +661,16 @@ int nfs_revalidate_inode(struct nfs_server *server, struct inode *inode) * @inode - pointer to host inode * @mapping - pointer to mapping */ -void nfs_revalidate_mapping(struct inode *inode, struct address_space *mapping) +int nfs_revalidate_mapping(struct inode *inode, struct address_space *mapping) { struct nfs_inode *nfsi = NFS_I(inode); + int ret = 0; + + if (NFS_STALE(inode)) + ret = -ESTALE; + if ((nfsi->cache_validity & NFS_INO_REVAL_PAGECACHE) + || nfs_attribute_timeout(inode)) + ret = __nfs_revalidate_inode(NFS_SERVER(inode), inode); if (nfsi->cache_validity & NFS_INO_INVALID_DATA) { nfs_inc_stats(inode, NFSIOS_DATAINVALIDATE); @@ -1322,6 +691,7 @@ void nfs_revalidate_mapping(struct inode *inode, struct address_space *mapping) inode->i_sb->s_id, (long long)NFS_FILEID(inode)); } + return ret; } /** @@ -1345,13 +715,11 @@ void nfs_end_data_update(struct inode *inode) { struct nfs_inode *nfsi = NFS_I(inode); - if (!nfs_have_delegation(inode, FMODE_READ)) { - /* Directories and symlinks: invalidate page cache */ - if (S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode)) { - spin_lock(&inode->i_lock); - nfsi->cache_validity |= NFS_INO_INVALID_DATA; - spin_unlock(&inode->i_lock); - } + /* Directories: invalidate page cache */ + if (S_ISDIR(inode->i_mode)) { + spin_lock(&inode->i_lock); + nfsi->cache_validity |= NFS_INO_INVALID_DATA; + spin_unlock(&inode->i_lock); } nfsi->cache_change_attribute = jiffies; atomic_dec(&nfsi->data_updates); @@ -1361,12 +729,6 @@ static void nfs_wcc_update_inode(struct inode *inode, struct nfs_fattr *fattr) { struct nfs_inode *nfsi = NFS_I(inode); - if ((fattr->valid & NFS_ATTR_PRE_CHANGE) != 0 - && nfsi->change_attr == fattr->pre_change_attr) { - nfsi->change_attr = fattr->change_attr; - nfsi->cache_change_attribute = jiffies; - } - /* If we have atomic WCC data, we may update some attributes */ if ((fattr->valid & NFS_ATTR_WCC) != 0) { if (timespec_equal(&inode->i_ctime, &fattr->pre_ctime)) { @@ -1400,9 +762,6 @@ static int nfs_check_inode_attributes(struct inode *inode, struct nfs_fattr *fat int data_unstable; - if ((fattr->valid & NFS_ATTR_FATTR) == 0) - return 0; - /* Has the inode gone and changed behind our back? */ if (nfsi->fileid != fattr->fileid || (inode->i_mode & S_IFMT) != (fattr->mode & S_IFMT)) { @@ -1415,20 +774,13 @@ static int nfs_check_inode_attributes(struct inode *inode, struct nfs_fattr *fat /* Do atomic weak cache consistency updates */ nfs_wcc_update_inode(inode, fattr); - if ((fattr->valid & NFS_ATTR_FATTR_V4) != 0) { - if (nfsi->change_attr == fattr->change_attr) - goto out; - nfsi->cache_validity |= NFS_INO_INVALID_ATTR; - if (!data_unstable) - nfsi->cache_validity |= NFS_INO_REVAL_PAGECACHE; - } + if ((fattr->valid & NFS_ATTR_FATTR_V4) != 0 && + nfsi->change_attr != fattr->change_attr) + nfsi->cache_validity |= NFS_INO_INVALID_ATTR|NFS_INO_REVAL_PAGECACHE; /* Verify a few of the more important attributes */ - if (!timespec_equal(&inode->i_mtime, &fattr->mtime)) { - nfsi->cache_validity |= NFS_INO_INVALID_ATTR; - if (!data_unstable) - nfsi->cache_validity |= NFS_INO_REVAL_PAGECACHE; - } + if (!timespec_equal(&inode->i_mtime, &fattr->mtime)) + nfsi->cache_validity |= NFS_INO_INVALID_ATTR|NFS_INO_REVAL_PAGECACHE; cur_size = i_size_read(inode); new_isize = nfs_size_to_loff_t(fattr->size); @@ -1445,7 +797,6 @@ static int nfs_check_inode_attributes(struct inode *inode, struct nfs_fattr *fat if (inode->i_nlink != fattr->nlink) nfsi->cache_validity |= NFS_INO_INVALID_ATTR; -out: if (!timespec_equal(&inode->i_atime, &fattr->atime)) nfsi->cache_validity |= NFS_INO_INVALID_ATIME; @@ -1471,7 +822,6 @@ int nfs_refresh_inode(struct inode *inode, struct nfs_fattr *fattr) if ((fattr->valid & NFS_ATTR_FATTR) == 0) return 0; spin_lock(&inode->i_lock); - nfsi->cache_validity &= ~NFS_INO_REVAL_PAGECACHE; if (time_after(fattr->time_start, nfsi->last_updated)) status = nfs_update_inode(inode, fattr); else @@ -1488,6 +838,12 @@ int nfs_refresh_inode(struct inode *inode, struct nfs_fattr *fattr) * * After an operation that has changed the inode metadata, mark the * attribute cache as being invalid, then try to update it. + * + * NB: if the server didn't return any post op attributes, this + * function will force the retrieval of attributes before the next + * NFS request. Thus it should be used only for operations that + * are expected to change one or more attributes, to avoid + * unnecessary NFS requests and trips through nfs_update_inode(). */ int nfs_post_op_update_inode(struct inode *inode, struct nfs_fattr *fattr) { @@ -1496,7 +852,7 @@ int nfs_post_op_update_inode(struct inode *inode, struct nfs_fattr *fattr) spin_lock(&inode->i_lock); if (unlikely((fattr->valid & NFS_ATTR_FATTR) == 0)) { - nfsi->cache_validity |= NFS_INO_INVALID_ATTR | NFS_INO_INVALID_ACCESS; + nfsi->cache_validity |= NFS_INO_INVALID_ACCESS|NFS_INO_INVALID_ATTR|NFS_INO_REVAL_PAGECACHE; goto out; } status = nfs_update_inode(inode, fattr); @@ -1519,6 +875,7 @@ out: */ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr) { + struct nfs_server *server; struct nfs_inode *nfsi = NFS_I(inode); loff_t cur_isize, new_isize; unsigned int invalid = 0; @@ -1528,9 +885,6 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr) __FUNCTION__, inode->i_sb->s_id, inode->i_ino, atomic_read(&inode->i_count), fattr->valid); - if ((fattr->valid & NFS_ATTR_FATTR) == 0) - return 0; - if (nfsi->fileid != fattr->fileid) goto out_fileid; @@ -1540,6 +894,12 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr) if ((inode->i_mode & S_IFMT) != (fattr->mode & S_IFMT)) goto out_changed; + server = NFS_SERVER(inode); + /* Update the fsid if and only if this is the root directory */ + if (inode == inode->i_sb->s_root->d_inode + && !nfs_fsid_equal(&server->fsid, &fattr->fsid)) + server->fsid = fattr->fsid; + /* * Update the read time so we don't revalidate too often. */ @@ -1549,7 +909,7 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr) /* Are we racing with known updates of the metadata on the server? */ data_stable = nfs_verify_change_attribute(inode, fattr->time_start); if (data_stable) - nfsi->cache_validity &= ~(NFS_INO_INVALID_ATTR|NFS_INO_INVALID_ATIME); + nfsi->cache_validity &= ~(NFS_INO_INVALID_ATTR|NFS_INO_REVAL_PAGECACHE|NFS_INO_INVALID_ATIME); /* Do atomic weak cache consistency updates */ nfs_wcc_update_inode(inode, fattr); @@ -1607,21 +967,17 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr) * report the blocks in 512byte units */ inode->i_blocks = nfs_calc_block_size(fattr->du.nfs3.used); - inode->i_blksize = inode->i_sb->s_blocksize; } else { inode->i_blocks = fattr->du.nfs2.blocks; - inode->i_blksize = fattr->du.nfs2.blocksize; } - if ((fattr->valid & NFS_ATTR_FATTR_V4)) { - if (nfsi->change_attr != fattr->change_attr) { - dprintk("NFS: change_attr change on server for file %s/%ld\n", - inode->i_sb->s_id, inode->i_ino); - nfsi->change_attr = fattr->change_attr; - invalid |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_DATA|NFS_INO_INVALID_ACCESS|NFS_INO_INVALID_ACL; - nfsi->cache_change_attribute = jiffies; - } else - invalid &= ~(NFS_INO_INVALID_ATTR|NFS_INO_INVALID_DATA); + if ((fattr->valid & NFS_ATTR_FATTR_V4) != 0 && + nfsi->change_attr != fattr->change_attr) { + dprintk("NFS: change_attr change on server for file %s/%ld\n", + inode->i_sb->s_id, inode->i_ino); + nfsi->change_attr = fattr->change_attr; + invalid |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_DATA|NFS_INO_INVALID_ACCESS|NFS_INO_INVALID_ACL; + nfsi->cache_change_attribute = jiffies; } /* Update attrtimeo value if we're out of the unstable period */ @@ -1664,207 +1020,20 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr) out_fileid: printk(KERN_ERR "NFS: server %s error: fileid changed\n" "fsid %s: expected fileid 0x%Lx, got 0x%Lx\n", - NFS_SERVER(inode)->hostname, inode->i_sb->s_id, + NFS_SERVER(inode)->nfs_client->cl_hostname, inode->i_sb->s_id, (long long)nfsi->fileid, (long long)fattr->fileid); goto out_err; } -/* - * File system information - */ - -static int nfs_set_super(struct super_block *s, void *data) -{ - s->s_fs_info = data; - return set_anon_super(s, data); -} - -static int nfs_compare_super(struct super_block *sb, void *data) -{ - struct nfs_server *server = data; - struct nfs_server *old = NFS_SB(sb); - - if (old->addr.sin_addr.s_addr != server->addr.sin_addr.s_addr) - return 0; - if (old->addr.sin_port != server->addr.sin_port) - return 0; - return !nfs_compare_fh(&old->fh, &server->fh); -} - -static int nfs_get_sb(struct file_system_type *fs_type, - int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt) -{ - int error; - struct nfs_server *server = NULL; - struct super_block *s; - struct nfs_fh *root; - struct nfs_mount_data *data = raw_data; - - error = -EINVAL; - if (data == NULL) { - dprintk("%s: missing data argument\n", __FUNCTION__); - goto out_err_noserver; - } - if (data->version <= 0 || data->version > NFS_MOUNT_VERSION) { - dprintk("%s: bad mount version\n", __FUNCTION__); - goto out_err_noserver; - } - switch (data->version) { - case 1: - data->namlen = 0; - case 2: - data->bsize = 0; - case 3: - if (data->flags & NFS_MOUNT_VER3) { - dprintk("%s: mount structure version %d does not support NFSv3\n", - __FUNCTION__, - data->version); - goto out_err_noserver; - } - data->root.size = NFS2_FHSIZE; - memcpy(data->root.data, data->old_root.data, NFS2_FHSIZE); - case 4: - if (data->flags & NFS_MOUNT_SECFLAVOUR) { - dprintk("%s: mount structure version %d does not support strong security\n", - __FUNCTION__, - data->version); - goto out_err_noserver; - } - case 5: - memset(data->context, 0, sizeof(data->context)); - } -#ifndef CONFIG_NFS_V3 - /* If NFSv3 is not compiled in, return -EPROTONOSUPPORT */ - error = -EPROTONOSUPPORT; - if (data->flags & NFS_MOUNT_VER3) { - dprintk("%s: NFSv3 not compiled into kernel\n", __FUNCTION__); - goto out_err_noserver; - } -#endif /* CONFIG_NFS_V3 */ - - error = -ENOMEM; - server = kzalloc(sizeof(struct nfs_server), GFP_KERNEL); - if (!server) - goto out_err_noserver; - /* Zero out the NFS state stuff */ - init_nfsv4_state(server); - server->client = server->client_sys = server->client_acl = ERR_PTR(-EINVAL); - - root = &server->fh; - if (data->flags & NFS_MOUNT_VER3) - root->size = data->root.size; - else - root->size = NFS2_FHSIZE; - error = -EINVAL; - if (root->size > sizeof(root->data)) { - dprintk("%s: invalid root filehandle\n", __FUNCTION__); - goto out_err; - } - memcpy(root->data, data->root.data, root->size); - - /* We now require that the mount process passes the remote address */ - memcpy(&server->addr, &data->addr, sizeof(server->addr)); - if (server->addr.sin_addr.s_addr == INADDR_ANY) { - dprintk("%s: mount program didn't pass remote address!\n", - __FUNCTION__); - goto out_err; - } - - /* Fire up rpciod if not yet running */ - error = rpciod_up(); - if (error < 0) { - dprintk("%s: couldn't start rpciod! Error = %d\n", - __FUNCTION__, error); - goto out_err; - } - - s = sget(fs_type, nfs_compare_super, nfs_set_super, server); - if (IS_ERR(s)) { - error = PTR_ERR(s); - goto out_err_rpciod; - } - - if (s->s_root) - goto out_rpciod_down; - - s->s_flags = flags; - - error = nfs_fill_super(s, data, flags & MS_SILENT ? 1 : 0); - if (error) { - up_write(&s->s_umount); - deactivate_super(s); - return error; - } - s->s_flags |= MS_ACTIVE; - return simple_set_mnt(mnt, s); - -out_rpciod_down: - rpciod_down(); - kfree(server); - return simple_set_mnt(mnt, s); - -out_err_rpciod: - rpciod_down(); -out_err: - kfree(server); -out_err_noserver: - return error; -} - -static void nfs_kill_super(struct super_block *s) -{ - struct nfs_server *server = NFS_SB(s); - - kill_anon_super(s); - - if (!IS_ERR(server->client)) - rpc_shutdown_client(server->client); - if (!IS_ERR(server->client_sys)) - rpc_shutdown_client(server->client_sys); - if (!IS_ERR(server->client_acl)) - rpc_shutdown_client(server->client_acl); - - if (!(server->flags & NFS_MOUNT_NONLM)) - lockd_down(); /* release rpc.lockd */ - - rpciod_down(); /* release rpciod */ - - nfs_free_iostats(server->io_stats); - kfree(server->hostname); - kfree(server); -} - -static struct file_system_type nfs_fs_type = { - .owner = THIS_MODULE, - .name = "nfs", - .get_sb = nfs_get_sb, - .kill_sb = nfs_kill_super, - .fs_flags = FS_ODD_RENAME|FS_REVAL_DOT|FS_BINARY_MOUNTDATA, -}; #ifdef CONFIG_NFS_V4 -static void nfs4_clear_inode(struct inode *); - - -static struct super_operations nfs4_sops = { - .alloc_inode = nfs_alloc_inode, - .destroy_inode = nfs_destroy_inode, - .write_inode = nfs_write_inode, - .delete_inode = nfs_delete_inode, - .statfs = nfs_statfs, - .clear_inode = nfs4_clear_inode, - .umount_begin = nfs_umount_begin, - .show_options = nfs_show_options, - .show_stats = nfs_show_stats, -}; - /* * Clean out any remaining NFSv4 state that might be left over due * to open() calls that passed nfs_atomic_lookup, but failed to call * nfs_open(). */ -static void nfs4_clear_inode(struct inode *inode) +void nfs4_clear_inode(struct inode *inode) { struct nfs_inode *nfsi = NFS_I(inode); @@ -1888,365 +1057,9 @@ static void nfs4_clear_inode(struct inode *inode) nfs4_close_state(state, state->state); } } - - -static int nfs4_fill_super(struct super_block *sb, struct nfs4_mount_data *data, int silent) -{ - struct nfs_server *server; - struct nfs4_client *clp = NULL; - struct rpc_xprt *xprt = NULL; - struct rpc_clnt *clnt = NULL; - struct rpc_timeout timeparms; - rpc_authflavor_t authflavour; - int err = -EIO; - - sb->s_blocksize_bits = 0; - sb->s_blocksize = 0; - server = NFS_SB(sb); - if (data->rsize != 0) - server->rsize = nfs_block_size(data->rsize, NULL); - if (data->wsize != 0) - server->wsize = nfs_block_size(data->wsize, NULL); - server->flags = data->flags & NFS_MOUNT_FLAGMASK; - server->caps = NFS_CAP_ATOMIC_OPEN; - - server->acregmin = data->acregmin*HZ; - server->acregmax = data->acregmax*HZ; - server->acdirmin = data->acdirmin*HZ; - server->acdirmax = data->acdirmax*HZ; - - server->rpc_ops = &nfs_v4_clientops; - - nfs_init_timeout_values(&timeparms, data->proto, data->timeo, data->retrans); - - server->retrans_timeo = timeparms.to_initval; - server->retrans_count = timeparms.to_retries; - - clp = nfs4_get_client(&server->addr.sin_addr); - if (!clp) { - dprintk("%s: failed to create NFS4 client.\n", __FUNCTION__); - return -EIO; - } - - /* Now create transport and client */ - authflavour = RPC_AUTH_UNIX; - if (data->auth_flavourlen != 0) { - if (data->auth_flavourlen != 1) { - dprintk("%s: Invalid number of RPC auth flavours %d.\n", - __FUNCTION__, data->auth_flavourlen); - err = -EINVAL; - goto out_fail; - } - if (copy_from_user(&authflavour, data->auth_flavours, sizeof(authflavour))) { - err = -EFAULT; - goto out_fail; - } - } - - down_write(&clp->cl_sem); - if (IS_ERR(clp->cl_rpcclient)) { - xprt = xprt_create_proto(data->proto, &server->addr, &timeparms); - if (IS_ERR(xprt)) { - up_write(&clp->cl_sem); - err = PTR_ERR(xprt); - dprintk("%s: cannot create RPC transport. Error = %d\n", - __FUNCTION__, err); - goto out_fail; - } - clnt = rpc_create_client(xprt, server->hostname, &nfs_program, - server->rpc_ops->version, authflavour); - if (IS_ERR(clnt)) { - up_write(&clp->cl_sem); - err = PTR_ERR(clnt); - dprintk("%s: cannot create RPC client. Error = %d\n", - __FUNCTION__, err); - goto out_fail; - } - clnt->cl_intr = 1; - clnt->cl_softrtry = 1; - clp->cl_rpcclient = clnt; - memcpy(clp->cl_ipaddr, server->ip_addr, sizeof(clp->cl_ipaddr)); - nfs_idmap_new(clp); - } - list_add_tail(&server->nfs4_siblings, &clp->cl_superblocks); - clnt = rpc_clone_client(clp->cl_rpcclient); - if (!IS_ERR(clnt)) - server->nfs4_state = clp; - up_write(&clp->cl_sem); - clp = NULL; - - if (IS_ERR(clnt)) { - err = PTR_ERR(clnt); - dprintk("%s: cannot create RPC client. Error = %d\n", - __FUNCTION__, err); - return err; - } - - server->client = clnt; - - if (server->nfs4_state->cl_idmap == NULL) { - dprintk("%s: failed to create idmapper.\n", __FUNCTION__); - return -ENOMEM; - } - - if (clnt->cl_auth->au_flavor != authflavour) { - struct rpc_auth *auth; - - auth = rpcauth_create(authflavour, clnt); - if (IS_ERR(auth)) { - dprintk("%s: couldn't create credcache!\n", __FUNCTION__); - return PTR_ERR(auth); - } - } - - sb->s_time_gran = 1; - - sb->s_op = &nfs4_sops; - err = nfs_sb_init(sb, authflavour); - if (err == 0) - return 0; -out_fail: - if (clp) - nfs4_put_client(clp); - return err; -} - -static int nfs4_compare_super(struct super_block *sb, void *data) -{ - struct nfs_server *server = data; - struct nfs_server *old = NFS_SB(sb); - - if (strcmp(server->hostname, old->hostname) != 0) - return 0; - if (strcmp(server->mnt_path, old->mnt_path) != 0) - return 0; - return 1; -} - -static void * -nfs_copy_user_string(char *dst, struct nfs_string *src, int maxlen) -{ - void *p = NULL; - - if (!src->len) - return ERR_PTR(-EINVAL); - if (src->len < maxlen) - maxlen = src->len; - if (dst == NULL) { - p = dst = kmalloc(maxlen + 1, GFP_KERNEL); - if (p == NULL) - return ERR_PTR(-ENOMEM); - } - if (copy_from_user(dst, src->data, maxlen)) { - kfree(p); - return ERR_PTR(-EFAULT); - } - dst[maxlen] = '\0'; - return dst; -} - -static int nfs4_get_sb(struct file_system_type *fs_type, - int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt) -{ - int error; - struct nfs_server *server; - struct super_block *s; - struct nfs4_mount_data *data = raw_data; - void *p; - - if (data == NULL) { - dprintk("%s: missing data argument\n", __FUNCTION__); - return -EINVAL; - } - if (data->version <= 0 || data->version > NFS4_MOUNT_VERSION) { - dprintk("%s: bad mount version\n", __FUNCTION__); - return -EINVAL; - } - - server = kzalloc(sizeof(struct nfs_server), GFP_KERNEL); - if (!server) - return -ENOMEM; - /* Zero out the NFS state stuff */ - init_nfsv4_state(server); - server->client = server->client_sys = server->client_acl = ERR_PTR(-EINVAL); - - p = nfs_copy_user_string(NULL, &data->hostname, 256); - if (IS_ERR(p)) - goto out_err; - server->hostname = p; - - p = nfs_copy_user_string(NULL, &data->mnt_path, 1024); - if (IS_ERR(p)) - goto out_err; - server->mnt_path = p; - - p = nfs_copy_user_string(server->ip_addr, &data->client_addr, - sizeof(server->ip_addr) - 1); - if (IS_ERR(p)) - goto out_err; - - /* We now require that the mount process passes the remote address */ - if (data->host_addrlen != sizeof(server->addr)) { - error = -EINVAL; - goto out_free; - } - if (copy_from_user(&server->addr, data->host_addr, sizeof(server->addr))) { - error = -EFAULT; - goto out_free; - } - if (server->addr.sin_family != AF_INET || - server->addr.sin_addr.s_addr == INADDR_ANY) { - dprintk("%s: mount program didn't pass remote IP address!\n", - __FUNCTION__); - error = -EINVAL; - goto out_free; - } - - /* Fire up rpciod if not yet running */ - error = rpciod_up(); - if (error < 0) { - dprintk("%s: couldn't start rpciod! Error = %d\n", - __FUNCTION__, error); - goto out_free; - } - - s = sget(fs_type, nfs4_compare_super, nfs_set_super, server); - if (IS_ERR(s)) { - error = PTR_ERR(s); - goto out_free; - } - - if (s->s_root) { - kfree(server->mnt_path); - kfree(server->hostname); - kfree(server); - return simple_set_mnt(mnt, s); - } - - s->s_flags = flags; - - error = nfs4_fill_super(s, data, flags & MS_SILENT ? 1 : 0); - if (error) { - up_write(&s->s_umount); - deactivate_super(s); - return error; - } - s->s_flags |= MS_ACTIVE; - return simple_set_mnt(mnt, s); -out_err: - error = PTR_ERR(p); -out_free: - kfree(server->mnt_path); - kfree(server->hostname); - kfree(server); - return error; -} - -static void nfs4_kill_super(struct super_block *sb) -{ - struct nfs_server *server = NFS_SB(sb); - - nfs_return_all_delegations(sb); - kill_anon_super(sb); - - nfs4_renewd_prepare_shutdown(server); - - if (server->client != NULL && !IS_ERR(server->client)) - rpc_shutdown_client(server->client); - - destroy_nfsv4_state(server); - - rpciod_down(); - - nfs_free_iostats(server->io_stats); - kfree(server->hostname); - kfree(server); -} - -static struct file_system_type nfs4_fs_type = { - .owner = THIS_MODULE, - .name = "nfs4", - .get_sb = nfs4_get_sb, - .kill_sb = nfs4_kill_super, - .fs_flags = FS_ODD_RENAME|FS_REVAL_DOT|FS_BINARY_MOUNTDATA, -}; - -static const int nfs_set_port_min = 0; -static const int nfs_set_port_max = 65535; -static int param_set_port(const char *val, struct kernel_param *kp) -{ - char *endp; - int num = simple_strtol(val, &endp, 0); - if (endp == val || *endp || num < nfs_set_port_min || num > nfs_set_port_max) - return -EINVAL; - *((int *)kp->arg) = num; - return 0; -} - -module_param_call(callback_tcpport, param_set_port, param_get_int, - &nfs_callback_set_tcpport, 0644); - -static int param_set_idmap_timeout(const char *val, struct kernel_param *kp) -{ - char *endp; - int num = simple_strtol(val, &endp, 0); - int jif = num * HZ; - if (endp == val || *endp || num < 0 || jif < num) - return -EINVAL; - *((int *)kp->arg) = jif; - return 0; -} - -module_param_call(idmap_cache_timeout, param_set_idmap_timeout, param_get_int, - &nfs_idmap_cache_timeout, 0644); - -#define nfs4_init_once(nfsi) \ - do { \ - INIT_LIST_HEAD(&(nfsi)->open_states); \ - nfsi->delegation = NULL; \ - nfsi->delegation_state = 0; \ - init_rwsem(&nfsi->rwsem); \ - } while(0) - -static inline int register_nfs4fs(void) -{ - int ret; - - ret = nfs_register_sysctl(); - if (ret != 0) - return ret; - ret = register_filesystem(&nfs4_fs_type); - if (ret != 0) - nfs_unregister_sysctl(); - return ret; -} - -static inline void unregister_nfs4fs(void) -{ - unregister_filesystem(&nfs4_fs_type); - nfs_unregister_sysctl(); -} -#else -#define nfs4_init_once(nfsi) \ - do { } while (0) -#define register_nfs4fs() (0) -#define unregister_nfs4fs() -#endif - -extern int nfs_init_nfspagecache(void); -extern void nfs_destroy_nfspagecache(void); -extern int nfs_init_readpagecache(void); -extern void nfs_destroy_readpagecache(void); -extern int nfs_init_writepagecache(void); -extern void nfs_destroy_writepagecache(void); -#ifdef CONFIG_NFS_DIRECTIO -extern int nfs_init_directcache(void); -extern void nfs_destroy_directcache(void); #endif -static kmem_cache_t * nfs_inode_cachep; - -static struct inode *nfs_alloc_inode(struct super_block *sb) +struct inode *nfs_alloc_inode(struct super_block *sb) { struct nfs_inode *nfsi; nfsi = (struct nfs_inode *)kmem_cache_alloc(nfs_inode_cachep, SLAB_KERNEL); @@ -2265,11 +1078,21 @@ static struct inode *nfs_alloc_inode(struct super_block *sb) return &nfsi->vfs_inode; } -static void nfs_destroy_inode(struct inode *inode) +void nfs_destroy_inode(struct inode *inode) { kmem_cache_free(nfs_inode_cachep, NFS_I(inode)); } +static inline void nfs4_init_once(struct nfs_inode *nfsi) +{ +#ifdef CONFIG_NFS_V4 + INIT_LIST_HEAD(&nfsi->open_states); + nfsi->delegation = NULL; + nfsi->delegation_state = 0; + init_rwsem(&nfsi->rwsem); +#endif +} + static void init_once(void * foo, kmem_cache_t * cachep, unsigned long flags) { struct nfs_inode *nfsi = (struct nfs_inode *) foo; @@ -2281,6 +1104,8 @@ static void init_once(void * foo, kmem_cache_t * cachep, unsigned long flags) INIT_LIST_HEAD(&nfsi->dirty); INIT_LIST_HEAD(&nfsi->commit); INIT_LIST_HEAD(&nfsi->open_files); + INIT_LIST_HEAD(&nfsi->access_cache_entry_lru); + INIT_LIST_HEAD(&nfsi->access_cache_inode_lru); INIT_RADIX_TREE(&nfsi->nfs_page_tree, GFP_ATOMIC); atomic_set(&nfsi->data_updates, 0); nfsi->ndirty = 0; @@ -2290,7 +1115,7 @@ static void init_once(void * foo, kmem_cache_t * cachep, unsigned long flags) } } -static int nfs_init_inodecache(void) +static int __init nfs_init_inodecache(void) { nfs_inode_cachep = kmem_cache_create("nfs_inode_cache", sizeof(struct nfs_inode), @@ -2305,8 +1130,7 @@ static int nfs_init_inodecache(void) static void nfs_destroy_inodecache(void) { - if (kmem_cache_destroy(nfs_inode_cachep)) - printk(KERN_INFO "nfs_inode_cache: not all structures were freed\n"); + kmem_cache_destroy(nfs_inode_cachep); } /* @@ -2316,6 +1140,10 @@ static int __init init_nfs_fs(void) { int err; + err = nfs_fs_proc_init(); + if (err) + goto out5; + err = nfs_init_nfspagecache(); if (err) goto out4; @@ -2332,29 +1160,22 @@ static int __init init_nfs_fs(void) if (err) goto out1; -#ifdef CONFIG_NFS_DIRECTIO err = nfs_init_directcache(); if (err) goto out0; -#endif #ifdef CONFIG_PROC_FS rpc_proc_register(&nfs_rpcstat); #endif - err = register_filesystem(&nfs_fs_type); - if (err) - goto out; - if ((err = register_nfs4fs()) != 0) + if ((err = register_nfs_fs()) != 0) goto out; return 0; out: #ifdef CONFIG_PROC_FS rpc_proc_unregister("nfs"); #endif -#ifdef CONFIG_NFS_DIRECTIO nfs_destroy_directcache(); out0: -#endif nfs_destroy_writepagecache(); out1: nfs_destroy_readpagecache(); @@ -2363,14 +1184,14 @@ out2: out3: nfs_destroy_nfspagecache(); out4: + nfs_fs_proc_exit(); +out5: return err; } static void __exit exit_nfs_fs(void) { -#ifdef CONFIG_NFS_DIRECTIO nfs_destroy_directcache(); -#endif nfs_destroy_writepagecache(); nfs_destroy_readpagecache(); nfs_destroy_inodecache(); @@ -2378,8 +1199,8 @@ static void __exit exit_nfs_fs(void) #ifdef CONFIG_PROC_FS rpc_proc_unregister("nfs"); #endif - unregister_filesystem(&nfs_fs_type); - unregister_nfs4fs(); + unregister_nfs_fs(); + nfs_fs_proc_exit(); } /* Not quite true; I just maintain it */ diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h new file mode 100644 index 00000000000..bea0b016bd7 --- /dev/null +++ b/fs/nfs/internal.h @@ -0,0 +1,219 @@ +/* + * NFS internal definitions + */ + +#include <linux/mount.h> + +struct nfs_string; +struct nfs_mount_data; +struct nfs4_mount_data; + +/* Maximum number of readahead requests + * FIXME: this should really be a sysctl so that users may tune it to suit + * their needs. People that do NFS over a slow network, might for + * instance want to reduce it to something closer to 1 for improved + * interactive response. + */ +#define NFS_MAX_READAHEAD (RPC_DEF_SLOT_TABLE - 1) + +struct nfs_clone_mount { + const struct super_block *sb; + const struct dentry *dentry; + struct nfs_fh *fh; + struct nfs_fattr *fattr; + char *hostname; + char *mnt_path; + struct sockaddr_in *addr; + rpc_authflavor_t authflavor; +}; + +/* client.c */ +extern struct rpc_program nfs_program; + +extern void nfs_put_client(struct nfs_client *); +extern struct nfs_client *nfs_find_client(const struct sockaddr_in *, int); +extern struct nfs_server *nfs_create_server(const struct nfs_mount_data *, + struct nfs_fh *); +extern struct nfs_server *nfs4_create_server(const struct nfs4_mount_data *, + const char *, + const struct sockaddr_in *, + const char *, + const char *, + rpc_authflavor_t, + struct nfs_fh *); +extern struct nfs_server *nfs4_create_referral_server(struct nfs_clone_mount *, + struct nfs_fh *); +extern void nfs_free_server(struct nfs_server *server); +extern struct nfs_server *nfs_clone_server(struct nfs_server *, + struct nfs_fh *, + struct nfs_fattr *); +#ifdef CONFIG_PROC_FS +extern int __init nfs_fs_proc_init(void); +extern void nfs_fs_proc_exit(void); +#else +static inline int nfs_fs_proc_init(void) +{ + return 0; +} +static inline void nfs_fs_proc_exit(void) +{ +} +#endif + +/* nfs4namespace.c */ +#ifdef CONFIG_NFS_V4 +extern struct vfsmount *nfs_do_refmount(const struct vfsmount *mnt_parent, struct dentry *dentry); +#else +static inline +struct vfsmount *nfs_do_refmount(const struct vfsmount *mnt_parent, struct dentry *dentry) +{ + return ERR_PTR(-ENOENT); +} +#endif + +/* callback_xdr.c */ +extern struct svc_version nfs4_callback_version1; + +/* pagelist.c */ +extern int __init nfs_init_nfspagecache(void); +extern void nfs_destroy_nfspagecache(void); +extern int __init nfs_init_readpagecache(void); +extern void nfs_destroy_readpagecache(void); +extern int __init nfs_init_writepagecache(void); +extern void nfs_destroy_writepagecache(void); + +#ifdef CONFIG_NFS_DIRECTIO +extern int __init nfs_init_directcache(void); +extern void nfs_destroy_directcache(void); +#else +#define nfs_init_directcache() (0) +#define nfs_destroy_directcache() do {} while(0) +#endif + +/* nfs2xdr.c */ +extern int nfs_stat_to_errno(int); +extern struct rpc_procinfo nfs_procedures[]; +extern u32 * nfs_decode_dirent(u32 *, struct nfs_entry *, int); + +/* nfs3xdr.c */ +extern struct rpc_procinfo nfs3_procedures[]; +extern u32 *nfs3_decode_dirent(u32 *, struct nfs_entry *, int); + +/* nfs4xdr.c */ +#ifdef CONFIG_NFS_V4 +extern u32 *nfs4_decode_dirent(u32 *p, struct nfs_entry *entry, int plus); +#endif + +/* nfs4proc.c */ +#ifdef CONFIG_NFS_V4 +extern struct rpc_procinfo nfs4_procedures[]; + +extern int nfs4_proc_fs_locations(struct inode *dir, struct dentry *dentry, + struct nfs4_fs_locations *fs_locations, + struct page *page); +#endif + +/* dir.c */ +extern int nfs_access_cache_shrinker(int nr_to_scan, gfp_t gfp_mask); + +/* inode.c */ +extern struct inode *nfs_alloc_inode(struct super_block *sb); +extern void nfs_destroy_inode(struct inode *); +extern int nfs_write_inode(struct inode *,int); +extern void nfs_clear_inode(struct inode *); +#ifdef CONFIG_NFS_V4 +extern void nfs4_clear_inode(struct inode *); +#endif + +/* super.c */ +extern struct file_system_type nfs_xdev_fs_type; +#ifdef CONFIG_NFS_V4 +extern struct file_system_type nfs4_xdev_fs_type; +extern struct file_system_type nfs4_referral_fs_type; +#endif + +extern struct rpc_stat nfs_rpcstat; + +extern int __init register_nfs_fs(void); +extern void __exit unregister_nfs_fs(void); + +/* namespace.c */ +extern char *nfs_path(const char *base, + const struct dentry *droot, + const struct dentry *dentry, + char *buffer, ssize_t buflen); + +/* getroot.c */ +extern struct dentry *nfs_get_root(struct super_block *, struct nfs_fh *); +#ifdef CONFIG_NFS_V4 +extern struct dentry *nfs4_get_root(struct super_block *, struct nfs_fh *); + +extern int nfs4_path_walk(struct nfs_server *server, + struct nfs_fh *mntfh, + const char *path); +#endif + +/* + * Determine the device name as a string + */ +static inline char *nfs_devname(const struct vfsmount *mnt_parent, + const struct dentry *dentry, + char *buffer, ssize_t buflen) +{ + return nfs_path(mnt_parent->mnt_devname, mnt_parent->mnt_root, + dentry, buffer, buflen); +} + +/* + * Determine the actual block size (and log2 thereof) + */ +static inline +unsigned long nfs_block_bits(unsigned long bsize, unsigned char *nrbitsp) +{ + /* make sure blocksize is a power of two */ + if ((bsize & (bsize - 1)) || nrbitsp) { + unsigned char nrbits; + + for (nrbits = 31; nrbits && !(bsize & (1 << nrbits)); nrbits--) + ; + bsize = 1 << nrbits; + if (nrbitsp) + *nrbitsp = nrbits; + } + + return bsize; +} + +/* + * Calculate the number of 512byte blocks used. + */ +static inline unsigned long nfs_calc_block_size(u64 tsize) +{ + loff_t used = (tsize + 511) >> 9; + return (used > ULONG_MAX) ? ULONG_MAX : used; +} + +/* + * Compute and set NFS server blocksize + */ +static inline +unsigned long nfs_block_size(unsigned long bsize, unsigned char *nrbitsp) +{ + if (bsize < NFS_MIN_FILE_IO_SIZE) + bsize = NFS_DEF_FILE_IO_SIZE; + else if (bsize >= NFS_MAX_FILE_IO_SIZE) + bsize = NFS_MAX_FILE_IO_SIZE; + + return nfs_block_bits(bsize, nrbitsp); +} + +/* + * Determine the maximum file size for a superblock + */ +static inline +void nfs_super_set_maxbytes(struct super_block *sb, __u64 maxfilesize) +{ + sb->s_maxbytes = (loff_t)maxfilesize; + if (sb->s_maxbytes > MAX_LFS_FILESIZE || sb->s_maxbytes <= 0) + sb->s_maxbytes = MAX_LFS_FILESIZE; +} diff --git a/fs/nfs/mount_clnt.c b/fs/nfs/mount_clnt.c index 445abb4d421..d507b021207 100644 --- a/fs/nfs/mount_clnt.c +++ b/fs/nfs/mount_clnt.c @@ -14,7 +14,6 @@ #include <linux/net.h> #include <linux/in.h> #include <linux/sunrpc/clnt.h> -#include <linux/sunrpc/xprt.h> #include <linux/sunrpc/sched.h> #include <linux/nfs_fs.h> @@ -77,22 +76,19 @@ static struct rpc_clnt * mnt_create(char *hostname, struct sockaddr_in *srvaddr, int version, int protocol) { - struct rpc_xprt *xprt; - struct rpc_clnt *clnt; - - xprt = xprt_create_proto(protocol, srvaddr, NULL); - if (IS_ERR(xprt)) - return (struct rpc_clnt *)xprt; - - clnt = rpc_create_client(xprt, hostname, - &mnt_program, version, - RPC_AUTH_UNIX); - if (!IS_ERR(clnt)) { - clnt->cl_softrtry = 1; - clnt->cl_oneshot = 1; - clnt->cl_intr = 1; - } - return clnt; + struct rpc_create_args args = { + .protocol = protocol, + .address = (struct sockaddr *)srvaddr, + .addrsize = sizeof(*srvaddr), + .servername = hostname, + .program = &mnt_program, + .version = version, + .authflavor = RPC_AUTH_UNIX, + .flags = (RPC_CLNT_CREATE_ONESHOT | + RPC_CLNT_CREATE_INTR), + }; + + return rpc_create(&args); } /* diff --git a/fs/nfs/namespace.c b/fs/nfs/namespace.c new file mode 100644 index 00000000000..60408646176 --- /dev/null +++ b/fs/nfs/namespace.c @@ -0,0 +1,251 @@ +/* + * linux/fs/nfs/namespace.c + * + * Copyright (C) 2005 Trond Myklebust <Trond.Myklebust@netapp.com> + * - Modified by David Howells <dhowells@redhat.com> + * + * NFS namespace + */ + +#include <linux/config.h> + +#include <linux/dcache.h> +#include <linux/mount.h> +#include <linux/namei.h> +#include <linux/nfs_fs.h> +#include <linux/string.h> +#include <linux/sunrpc/clnt.h> +#include <linux/vfs.h> +#include "internal.h" + +#define NFSDBG_FACILITY NFSDBG_VFS + +static void nfs_expire_automounts(void *list); + +LIST_HEAD(nfs_automount_list); +static DECLARE_WORK(nfs_automount_task, nfs_expire_automounts, &nfs_automount_list); +int nfs_mountpoint_expiry_timeout = 500 * HZ; + +static struct vfsmount *nfs_do_submount(const struct vfsmount *mnt_parent, + const struct dentry *dentry, + struct nfs_fh *fh, + struct nfs_fattr *fattr); + +/* + * nfs_path - reconstruct the path given an arbitrary dentry + * @base - arbitrary string to prepend to the path + * @droot - pointer to root dentry for mountpoint + * @dentry - pointer to dentry + * @buffer - result buffer + * @buflen - length of buffer + * + * Helper function for constructing the path from the + * root dentry to an arbitrary hashed dentry. + * + * This is mainly for use in figuring out the path on the + * server side when automounting on top of an existing partition. + */ +char *nfs_path(const char *base, + const struct dentry *droot, + const struct dentry *dentry, + char *buffer, ssize_t buflen) +{ + char *end = buffer+buflen; + int namelen; + + *--end = '\0'; + buflen--; + spin_lock(&dcache_lock); + while (!IS_ROOT(dentry) && dentry != droot) { + namelen = dentry->d_name.len; + buflen -= namelen + 1; + if (buflen < 0) + goto Elong_unlock; + end -= namelen; + memcpy(end, dentry->d_name.name, namelen); + *--end = '/'; + dentry = dentry->d_parent; + } + spin_unlock(&dcache_lock); + namelen = strlen(base); + /* Strip off excess slashes in base string */ + while (namelen > 0 && base[namelen - 1] == '/') + namelen--; + buflen -= namelen; + if (buflen < 0) + goto Elong; + end -= namelen; + memcpy(end, base, namelen); + return end; +Elong_unlock: + spin_unlock(&dcache_lock); +Elong: + return ERR_PTR(-ENAMETOOLONG); +} + +/* + * nfs_follow_mountpoint - handle crossing a mountpoint on the server + * @dentry - dentry of mountpoint + * @nd - nameidata info + * + * When we encounter a mountpoint on the server, we want to set up + * a mountpoint on the client too, to prevent inode numbers from + * colliding, and to allow "df" to work properly. + * On NFSv4, we also want to allow for the fact that different + * filesystems may be migrated to different servers in a failover + * situation, and that different filesystems may want to use + * different security flavours. + */ +static void * nfs_follow_mountpoint(struct dentry *dentry, struct nameidata *nd) +{ + struct vfsmount *mnt; + struct nfs_server *server = NFS_SERVER(dentry->d_inode); + struct dentry *parent; + struct nfs_fh fh; + struct nfs_fattr fattr; + int err; + + dprintk("--> nfs_follow_mountpoint()\n"); + + BUG_ON(IS_ROOT(dentry)); + dprintk("%s: enter\n", __FUNCTION__); + dput(nd->dentry); + nd->dentry = dget(dentry); + + /* Look it up again */ + parent = dget_parent(nd->dentry); + err = server->nfs_client->rpc_ops->lookup(parent->d_inode, + &nd->dentry->d_name, + &fh, &fattr); + dput(parent); + if (err != 0) + goto out_err; + + if (fattr.valid & NFS_ATTR_FATTR_V4_REFERRAL) + mnt = nfs_do_refmount(nd->mnt, nd->dentry); + else + mnt = nfs_do_submount(nd->mnt, nd->dentry, &fh, &fattr); + err = PTR_ERR(mnt); + if (IS_ERR(mnt)) + goto out_err; + + mntget(mnt); + err = do_add_mount(mnt, nd, nd->mnt->mnt_flags|MNT_SHRINKABLE, &nfs_automount_list); + if (err < 0) { + mntput(mnt); + if (err == -EBUSY) + goto out_follow; + goto out_err; + } + mntput(nd->mnt); + dput(nd->dentry); + nd->mnt = mnt; + nd->dentry = dget(mnt->mnt_root); + schedule_delayed_work(&nfs_automount_task, nfs_mountpoint_expiry_timeout); +out: + dprintk("%s: done, returned %d\n", __FUNCTION__, err); + + dprintk("<-- nfs_follow_mountpoint() = %d\n", err); + return ERR_PTR(err); +out_err: + path_release(nd); + goto out; +out_follow: + while(d_mountpoint(nd->dentry) && follow_down(&nd->mnt, &nd->dentry)) + ; + err = 0; + goto out; +} + +struct inode_operations nfs_mountpoint_inode_operations = { + .follow_link = nfs_follow_mountpoint, + .getattr = nfs_getattr, +}; + +struct inode_operations nfs_referral_inode_operations = { + .follow_link = nfs_follow_mountpoint, +}; + +static void nfs_expire_automounts(void *data) +{ + struct list_head *list = (struct list_head *)data; + + mark_mounts_for_expiry(list); + if (!list_empty(list)) + schedule_delayed_work(&nfs_automount_task, nfs_mountpoint_expiry_timeout); +} + +void nfs_release_automount_timer(void) +{ + if (list_empty(&nfs_automount_list)) { + cancel_delayed_work(&nfs_automount_task); + flush_scheduled_work(); + } +} + +/* + * Clone a mountpoint of the appropriate type + */ +static struct vfsmount *nfs_do_clone_mount(struct nfs_server *server, + const char *devname, + struct nfs_clone_mount *mountdata) +{ +#ifdef CONFIG_NFS_V4 + struct vfsmount *mnt = NULL; + switch (server->nfs_client->cl_nfsversion) { + case 2: + case 3: + mnt = vfs_kern_mount(&nfs_xdev_fs_type, 0, devname, mountdata); + break; + case 4: + mnt = vfs_kern_mount(&nfs4_xdev_fs_type, 0, devname, mountdata); + } + return mnt; +#else + return vfs_kern_mount(&nfs_xdev_fs_type, 0, devname, mountdata); +#endif +} + +/** + * nfs_do_submount - set up mountpoint when crossing a filesystem boundary + * @mnt_parent - mountpoint of parent directory + * @dentry - parent directory + * @fh - filehandle for new root dentry + * @fattr - attributes for new root inode + * + */ +static struct vfsmount *nfs_do_submount(const struct vfsmount *mnt_parent, + const struct dentry *dentry, + struct nfs_fh *fh, + struct nfs_fattr *fattr) +{ + struct nfs_clone_mount mountdata = { + .sb = mnt_parent->mnt_sb, + .dentry = dentry, + .fh = fh, + .fattr = fattr, + }; + struct vfsmount *mnt = ERR_PTR(-ENOMEM); + char *page = (char *) __get_free_page(GFP_USER); + char *devname; + + dprintk("--> nfs_do_submount()\n"); + + dprintk("%s: submounting on %s/%s\n", __FUNCTION__, + dentry->d_parent->d_name.name, + dentry->d_name.name); + if (page == NULL) + goto out; + devname = nfs_devname(mnt_parent, dentry, page, PAGE_SIZE); + mnt = (struct vfsmount *)devname; + if (IS_ERR(devname)) + goto free_page; + mnt = nfs_do_clone_mount(NFS_SB(mnt_parent->mnt_sb), devname, &mountdata); +free_page: + free_page((unsigned long)page); +out: + dprintk("%s: done\n", __FUNCTION__); + + dprintk("<-- nfs_do_submount() = %p\n", mnt); + return mnt; +} diff --git a/fs/nfs/nfs2xdr.c b/fs/nfs/nfs2xdr.c index f0015fa876e..b49501fc0a7 100644 --- a/fs/nfs/nfs2xdr.c +++ b/fs/nfs/nfs2xdr.c @@ -23,12 +23,11 @@ #include <linux/nfs.h> #include <linux/nfs2.h> #include <linux/nfs_fs.h> +#include "internal.h" #define NFSDBG_FACILITY NFSDBG_XDR /* #define NFS_PARANOIA 1 */ -extern int nfs_stat_to_errno(int stat); - /* Mapping from NFS error code to "errno" error code. */ #define errno_NFSERR_IO EIO @@ -52,7 +51,7 @@ extern int nfs_stat_to_errno(int stat); #define NFS_createargs_sz (NFS_diropargs_sz+NFS_sattr_sz) #define NFS_renameargs_sz (NFS_diropargs_sz+NFS_diropargs_sz) #define NFS_linkargs_sz (NFS_fhandle_sz+NFS_diropargs_sz) -#define NFS_symlinkargs_sz (NFS_diropargs_sz+NFS_path_sz+NFS_sattr_sz) +#define NFS_symlinkargs_sz (NFS_diropargs_sz+1+NFS_sattr_sz) #define NFS_readdirargs_sz (NFS_fhandle_sz+2) #define NFS_attrstat_sz (1+NFS_fattr_sz) @@ -131,7 +130,8 @@ xdr_decode_fattr(u32 *p, struct nfs_fattr *fattr) fattr->du.nfs2.blocksize = ntohl(*p++); rdev = ntohl(*p++); fattr->du.nfs2.blocks = ntohl(*p++); - fattr->fsid_u.nfs3 = ntohl(*p++); + fattr->fsid.major = ntohl(*p++); + fattr->fsid.minor = 0; fattr->fileid = ntohl(*p++); p = xdr_decode_time(p, &fattr->atime); p = xdr_decode_time(p, &fattr->mtime); @@ -351,11 +351,26 @@ nfs_xdr_linkargs(struct rpc_rqst *req, u32 *p, struct nfs_linkargs *args) static int nfs_xdr_symlinkargs(struct rpc_rqst *req, u32 *p, struct nfs_symlinkargs *args) { + struct xdr_buf *sndbuf = &req->rq_snd_buf; + size_t pad; + p = xdr_encode_fhandle(p, args->fromfh); p = xdr_encode_array(p, args->fromname, args->fromlen); - p = xdr_encode_array(p, args->topath, args->tolen); + *p++ = htonl(args->pathlen); + sndbuf->len = xdr_adjust_iovec(sndbuf->head, p); + + xdr_encode_pages(sndbuf, args->pages, 0, args->pathlen); + + /* + * xdr_encode_pages may have added a few bytes to ensure the + * pathname ends on a 4-byte boundary. Start encoding the + * attributes after the pad bytes. + */ + pad = sndbuf->tail->iov_len; + if (pad > 0) + p++; p = xdr_encode_sattr(p, args->sattr); - req->rq_slen = xdr_adjust_iovec(req->rq_svec, p); + sndbuf->len += xdr_adjust_iovec(sndbuf->tail, p) - pad; return 0; } diff --git a/fs/nfs/nfs3acl.c b/fs/nfs/nfs3acl.c index 33287879bd2..7322da4d205 100644 --- a/fs/nfs/nfs3acl.c +++ b/fs/nfs/nfs3acl.c @@ -172,8 +172,10 @@ static void nfs3_cache_acls(struct inode *inode, struct posix_acl *acl, inode->i_ino, acl, dfacl); spin_lock(&inode->i_lock); __nfs3_forget_cached_acls(NFS_I(inode)); - nfsi->acl_access = posix_acl_dup(acl); - nfsi->acl_default = posix_acl_dup(dfacl); + if (!IS_ERR(acl)) + nfsi->acl_access = posix_acl_dup(acl); + if (!IS_ERR(dfacl)) + nfsi->acl_default = posix_acl_dup(dfacl); spin_unlock(&inode->i_lock); } @@ -254,7 +256,9 @@ struct posix_acl *nfs3_proc_getacl(struct inode *inode, int type) res.acl_access = NULL; } } - nfs3_cache_acls(inode, res.acl_access, res.acl_default); + nfs3_cache_acls(inode, + (res.mask & NFS_ACL) ? res.acl_access : ERR_PTR(-EINVAL), + (res.mask & NFS_DFACL) ? res.acl_default : ERR_PTR(-EINVAL)); switch(type) { case ACL_TYPE_ACCESS: @@ -329,6 +333,7 @@ static int nfs3_proc_setacls(struct inode *inode, struct posix_acl *acl, switch (status) { case 0: status = nfs_refresh_inode(inode, &fattr); + nfs3_cache_acls(inode, acl, dfacl); break; case -EPFNOSUPPORT: case -EPROTONOSUPPORT: diff --git a/fs/nfs/nfs3proc.c b/fs/nfs/nfs3proc.c index cf186f0d2b3..3b234d4601e 100644 --- a/fs/nfs/nfs3proc.c +++ b/fs/nfs/nfs3proc.c @@ -20,11 +20,10 @@ #include <linux/nfs_mount.h> #include "iostat.h" +#include "internal.h" #define NFSDBG_FACILITY NFSDBG_PROC -extern struct rpc_procinfo nfs3_procedures[]; - /* A wrapper to handle the EJUKEBOX error message */ static int nfs3_rpc_wrapper(struct rpc_clnt *clnt, struct rpc_message *msg, int flags) @@ -82,7 +81,7 @@ do_proc_get_root(struct rpc_clnt *client, struct nfs_fh *fhandle, } /* - * Bare-bones access to getattr: this is for nfs_read_super. + * Bare-bones access to getattr: this is for nfs_get_root/nfs_get_sb */ static int nfs3_proc_get_root(struct nfs_server *server, struct nfs_fh *fhandle, @@ -91,8 +90,8 @@ nfs3_proc_get_root(struct nfs_server *server, struct nfs_fh *fhandle, int status; status = do_proc_get_root(server->client, fhandle, info); - if (status && server->client_sys != server->client) - status = do_proc_get_root(server->client_sys, fhandle, info); + if (status && server->nfs_client->cl_rpcclient != server->client) + status = do_proc_get_root(server->nfs_client->cl_rpcclient, fhandle, info); return status; } @@ -450,7 +449,7 @@ nfs3_proc_unlink_setup(struct rpc_message *msg, struct dentry *dir, struct qstr struct nfs_fattr res; } *ptr; - ptr = (struct unlinkxdr *)kmalloc(sizeof(*ptr), GFP_KERNEL); + ptr = kmalloc(sizeof(*ptr), GFP_KERNEL); if (!ptr) return -ENOMEM; ptr->arg.fh = NFS_FH(dir->d_inode); @@ -545,23 +544,23 @@ nfs3_proc_link(struct inode *inode, struct inode *dir, struct qstr *name) } static int -nfs3_proc_symlink(struct inode *dir, struct qstr *name, struct qstr *path, - struct iattr *sattr, struct nfs_fh *fhandle, - struct nfs_fattr *fattr) +nfs3_proc_symlink(struct inode *dir, struct dentry *dentry, struct page *page, + unsigned int len, struct iattr *sattr) { - struct nfs_fattr dir_attr; + struct nfs_fh fhandle; + struct nfs_fattr fattr, dir_attr; struct nfs3_symlinkargs arg = { .fromfh = NFS_FH(dir), - .fromname = name->name, - .fromlen = name->len, - .topath = path->name, - .tolen = path->len, + .fromname = dentry->d_name.name, + .fromlen = dentry->d_name.len, + .pages = &page, + .pathlen = len, .sattr = sattr }; struct nfs3_diropres res = { .dir_attr = &dir_attr, - .fh = fhandle, - .fattr = fattr + .fh = &fhandle, + .fattr = &fattr }; struct rpc_message msg = { .rpc_proc = &nfs3_procedures[NFS3PROC_SYMLINK], @@ -570,13 +569,19 @@ nfs3_proc_symlink(struct inode *dir, struct qstr *name, struct qstr *path, }; int status; - if (path->len > NFS3_MAXPATHLEN) + if (len > NFS3_MAXPATHLEN) return -ENAMETOOLONG; - dprintk("NFS call symlink %s -> %s\n", name->name, path->name); + + dprintk("NFS call symlink %s\n", dentry->d_name.name); + nfs_fattr_init(&dir_attr); - nfs_fattr_init(fattr); + nfs_fattr_init(&fattr); status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0); nfs_post_op_update_inode(dir, &dir_attr); + if (status != 0) + goto out; + status = nfs_instantiate(dentry, &fhandle, &fattr); +out: dprintk("NFS reply symlink: %d\n", status); return status; } @@ -786,7 +791,7 @@ nfs3_proc_fsinfo(struct nfs_server *server, struct nfs_fh *fhandle, dprintk("NFS call fsinfo\n"); nfs_fattr_init(info->fattr); - status = rpc_call_sync(server->client_sys, &msg, 0); + status = rpc_call_sync(server->nfs_client->cl_rpcclient, &msg, 0); dprintk("NFS reply fsinfo: %d\n", status); return status; } @@ -809,8 +814,6 @@ nfs3_proc_pathconf(struct nfs_server *server, struct nfs_fh *fhandle, return status; } -extern u32 *nfs3_decode_dirent(u32 *, struct nfs_entry *, int); - static int nfs3_read_done(struct rpc_task *task, struct nfs_read_data *data) { if (nfs3_async_handle_jukebox(task, data->inode)) @@ -889,7 +892,7 @@ nfs3_proc_lock(struct file *filp, int cmd, struct file_lock *fl) return nlmclnt_proc(filp->f_dentry->d_inode, cmd, fl); } -struct nfs_rpc_ops nfs_v3_clientops = { +const struct nfs_rpc_ops nfs_v3_clientops = { .version = 3, /* protocol version */ .dentry_ops = &nfs_dentry_operations, .dir_inode_ops = &nfs3_dir_inode_operations, diff --git a/fs/nfs/nfs3xdr.c b/fs/nfs/nfs3xdr.c index ec233619687..16556fa4eff 100644 --- a/fs/nfs/nfs3xdr.c +++ b/fs/nfs/nfs3xdr.c @@ -22,14 +22,13 @@ #include <linux/nfs3.h> #include <linux/nfs_fs.h> #include <linux/nfsacl.h> +#include "internal.h" #define NFSDBG_FACILITY NFSDBG_XDR /* Mapping from NFS error code to "errno" error code. */ #define errno_NFSERR_IO EIO -extern int nfs_stat_to_errno(int); - /* * Declare the space requirements for NFS arguments and replies as * number of 32bit-words @@ -57,7 +56,7 @@ extern int nfs_stat_to_errno(int); #define NFS3_writeargs_sz (NFS3_fh_sz+5) #define NFS3_createargs_sz (NFS3_diropargs_sz+NFS3_sattr_sz) #define NFS3_mkdirargs_sz (NFS3_diropargs_sz+NFS3_sattr_sz) -#define NFS3_symlinkargs_sz (NFS3_diropargs_sz+NFS3_path_sz+NFS3_sattr_sz) +#define NFS3_symlinkargs_sz (NFS3_diropargs_sz+1+NFS3_sattr_sz) #define NFS3_mknodargs_sz (NFS3_diropargs_sz+2+NFS3_sattr_sz) #define NFS3_renameargs_sz (NFS3_diropargs_sz+NFS3_diropargs_sz) #define NFS3_linkargs_sz (NFS3_fh_sz+NFS3_diropargs_sz) @@ -166,7 +165,8 @@ xdr_decode_fattr(u32 *p, struct nfs_fattr *fattr) if (MAJOR(fattr->rdev) != major || MINOR(fattr->rdev) != minor) fattr->rdev = 0; - p = xdr_decode_hyper(p, &fattr->fsid_u.nfs3); + p = xdr_decode_hyper(p, &fattr->fsid.major); + fattr->fsid.minor = 0; p = xdr_decode_hyper(p, &fattr->fileid); p = xdr_decode_time3(p, &fattr->atime); p = xdr_decode_time3(p, &fattr->mtime); @@ -398,8 +398,11 @@ nfs3_xdr_symlinkargs(struct rpc_rqst *req, u32 *p, struct nfs3_symlinkargs *args p = xdr_encode_fhandle(p, args->fromfh); p = xdr_encode_array(p, args->fromname, args->fromlen); p = xdr_encode_sattr(p, args->sattr); - p = xdr_encode_array(p, args->topath, args->tolen); + *p++ = htonl(args->pathlen); req->rq_slen = xdr_adjust_iovec(req->rq_svec, p); + + /* Copy the page */ + xdr_encode_pages(&req->rq_snd_buf, args->pages, 0, args->pathlen); return 0; } diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h index 0f5e4e7cdde..61095fe4b5c 100644 --- a/fs/nfs/nfs4_fs.h +++ b/fs/nfs/nfs4_fs.h @@ -43,55 +43,6 @@ enum nfs4_client_state { }; /* - * The nfs4_client identifies our client state to the server. - */ -struct nfs4_client { - struct list_head cl_servers; /* Global list of servers */ - struct in_addr cl_addr; /* Server identifier */ - u64 cl_clientid; /* constant */ - nfs4_verifier cl_confirm; - unsigned long cl_state; - - u32 cl_lockowner_id; - - /* - * The following rwsem ensures exclusive access to the server - * while we recover the state following a lease expiration. - */ - struct rw_semaphore cl_sem; - - struct list_head cl_delegations; - struct list_head cl_state_owners; - struct list_head cl_unused; - int cl_nunused; - spinlock_t cl_lock; - atomic_t cl_count; - - struct rpc_clnt * cl_rpcclient; - - struct list_head cl_superblocks; /* List of nfs_server structs */ - - unsigned long cl_lease_time; - unsigned long cl_last_renewal; - struct work_struct cl_renewd; - struct work_struct cl_recoverd; - - struct rpc_wait_queue cl_rpcwaitq; - - /* used for the setclientid verifier */ - struct timespec cl_boot_time; - - /* idmapper */ - struct idmap * cl_idmap; - - /* Our own IP address, as a null-terminated string. - * This is used to generate the clientid, and the callback address. - */ - char cl_ipaddr[16]; - unsigned char cl_id_uniquifier; -}; - -/* * struct rpc_sequence ensures that RPC calls are sent in the exact * order that they appear on the list. */ @@ -127,7 +78,7 @@ static inline void nfs_confirm_seqid(struct nfs_seqid_counter *seqid, int status struct nfs4_state_owner { spinlock_t so_lock; struct list_head so_list; /* per-clientid list of state_owners */ - struct nfs4_client *so_client; + struct nfs_client *so_client; u32 so_id; /* 32-bit identifier, unique */ atomic_t so_count; @@ -210,13 +161,16 @@ extern ssize_t nfs4_listxattr(struct dentry *, char *, size_t); /* nfs4proc.c */ extern int nfs4_map_errors(int err); -extern int nfs4_proc_setclientid(struct nfs4_client *, u32, unsigned short, struct rpc_cred *); -extern int nfs4_proc_setclientid_confirm(struct nfs4_client *, struct rpc_cred *); -extern int nfs4_proc_async_renew(struct nfs4_client *, struct rpc_cred *); -extern int nfs4_proc_renew(struct nfs4_client *, struct rpc_cred *); +extern int nfs4_proc_setclientid(struct nfs_client *, u32, unsigned short, struct rpc_cred *); +extern int nfs4_proc_setclientid_confirm(struct nfs_client *, struct rpc_cred *); +extern int nfs4_proc_async_renew(struct nfs_client *, struct rpc_cred *); +extern int nfs4_proc_renew(struct nfs_client *, struct rpc_cred *); extern int nfs4_do_close(struct inode *inode, struct nfs4_state *state); extern struct dentry *nfs4_atomic_open(struct inode *, struct dentry *, struct nameidata *); extern int nfs4_open_revalidate(struct inode *, struct dentry *, int, struct nameidata *); +extern int nfs4_server_capabilities(struct nfs_server *server, struct nfs_fh *fhandle); +extern int nfs4_proc_fs_locations(struct inode *dir, struct dentry *dentry, + struct nfs4_fs_locations *fs_locations, struct page *page); extern struct nfs4_state_recovery_ops nfs4_reboot_recovery_ops; extern struct nfs4_state_recovery_ops nfs4_network_partition_recovery_ops; @@ -225,21 +179,17 @@ extern const u32 nfs4_fattr_bitmap[2]; extern const u32 nfs4_statfs_bitmap[2]; extern const u32 nfs4_pathconf_bitmap[2]; extern const u32 nfs4_fsinfo_bitmap[2]; +extern const u32 nfs4_fs_locations_bitmap[2]; /* nfs4renewd.c */ -extern void nfs4_schedule_state_renewal(struct nfs4_client *); +extern void nfs4_schedule_state_renewal(struct nfs_client *); extern void nfs4_renewd_prepare_shutdown(struct nfs_server *); -extern void nfs4_kill_renewd(struct nfs4_client *); +extern void nfs4_kill_renewd(struct nfs_client *); extern void nfs4_renew_state(void *); /* nfs4state.c */ -extern void init_nfsv4_state(struct nfs_server *); -extern void destroy_nfsv4_state(struct nfs_server *); -extern struct nfs4_client *nfs4_get_client(struct in_addr *); -extern void nfs4_put_client(struct nfs4_client *clp); -extern struct nfs4_client *nfs4_find_client(struct in_addr *); -struct rpc_cred *nfs4_get_renew_cred(struct nfs4_client *clp); -extern u32 nfs4_alloc_lockowner_id(struct nfs4_client *); +struct rpc_cred *nfs4_get_renew_cred(struct nfs_client *clp); +extern u32 nfs4_alloc_lockowner_id(struct nfs_client *); extern struct nfs4_state_owner * nfs4_get_state_owner(struct nfs_server *, struct rpc_cred *); extern void nfs4_put_state_owner(struct nfs4_state_owner *); @@ -248,7 +198,7 @@ extern struct nfs4_state * nfs4_get_open_state(struct inode *, struct nfs4_state extern void nfs4_put_open_state(struct nfs4_state *); extern void nfs4_close_state(struct nfs4_state *, mode_t); extern void nfs4_state_set_mode_locked(struct nfs4_state *, mode_t); -extern void nfs4_schedule_state_recovery(struct nfs4_client *); +extern void nfs4_schedule_state_recovery(struct nfs_client *); 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); @@ -272,10 +222,6 @@ extern struct svc_version nfs4_callback_version1; #else -#define init_nfsv4_state(server) do { } while (0) -#define destroy_nfsv4_state(server) do { } while (0) -#define nfs4_put_state_owner(inode, owner) do { } while (0) -#define nfs4_put_open_state(state) do { } while (0) #define nfs4_close_state(a, b) do { } while (0) #endif /* CONFIG_NFS_V4 */ diff --git a/fs/nfs/nfs4namespace.c b/fs/nfs/nfs4namespace.c new file mode 100644 index 00000000000..24e47f3bbd1 --- /dev/null +++ b/fs/nfs/nfs4namespace.c @@ -0,0 +1,261 @@ +/* + * linux/fs/nfs/nfs4namespace.c + * + * Copyright (C) 2005 Trond Myklebust <Trond.Myklebust@netapp.com> + * - Modified by David Howells <dhowells@redhat.com> + * + * NFSv4 namespace + */ + +#include <linux/config.h> + +#include <linux/dcache.h> +#include <linux/mount.h> +#include <linux/namei.h> +#include <linux/nfs_fs.h> +#include <linux/string.h> +#include <linux/sunrpc/clnt.h> +#include <linux/vfs.h> +#include <linux/inet.h> +#include "internal.h" + +#define NFSDBG_FACILITY NFSDBG_VFS + +/* + * Check if fs_root is valid + */ +static inline char *nfs4_pathname_string(const struct nfs4_pathname *pathname, + char *buffer, ssize_t buflen) +{ + char *end = buffer + buflen; + int n; + + *--end = '\0'; + buflen--; + + n = pathname->ncomponents; + while (--n >= 0) { + const struct nfs4_string *component = &pathname->components[n]; + buflen -= component->len + 1; + if (buflen < 0) + goto Elong; + end -= component->len; + memcpy(end, component->data, component->len); + *--end = '/'; + } + return end; +Elong: + return ERR_PTR(-ENAMETOOLONG); +} + +/* + * Determine the mount path as a string + */ +static char *nfs4_path(const struct vfsmount *mnt_parent, + const struct dentry *dentry, + char *buffer, ssize_t buflen) +{ + const char *srvpath; + + srvpath = strchr(mnt_parent->mnt_devname, ':'); + if (srvpath) + srvpath++; + else + srvpath = mnt_parent->mnt_devname; + + return nfs_path(srvpath, mnt_parent->mnt_root, dentry, buffer, buflen); +} + +/* + * Check that fs_locations::fs_root [RFC3530 6.3] is a prefix for what we + * believe to be the server path to this dentry + */ +static int nfs4_validate_fspath(const struct vfsmount *mnt_parent, + const struct dentry *dentry, + const struct nfs4_fs_locations *locations, + char *page, char *page2) +{ + const char *path, *fs_path; + + path = nfs4_path(mnt_parent, dentry, page, PAGE_SIZE); + if (IS_ERR(path)) + return PTR_ERR(path); + + fs_path = nfs4_pathname_string(&locations->fs_path, page2, PAGE_SIZE); + if (IS_ERR(fs_path)) + return PTR_ERR(fs_path); + + if (strncmp(path, fs_path, strlen(fs_path)) != 0) { + dprintk("%s: path %s does not begin with fsroot %s\n", + __FUNCTION__, path, fs_path); + return -ENOENT; + } + + return 0; +} + +/* + * Check if the string represents a "valid" IPv4 address + */ +static inline int valid_ipaddr4(const char *buf) +{ + int rc, count, in[4]; + + rc = sscanf(buf, "%d.%d.%d.%d", &in[0], &in[1], &in[2], &in[3]); + if (rc != 4) + return -EINVAL; + for (count = 0; count < 4; count++) { + if (in[count] > 255) + return -EINVAL; + } + return 0; +} + +/** + * nfs_follow_referral - set up mountpoint when hitting a referral on moved error + * @mnt_parent - mountpoint of parent directory + * @dentry - parent directory + * @fspath - fs path returned in fs_locations + * @mntpath - mount path to new server + * @hostname - hostname of new server + * @addr - host addr of new server + * + */ +static struct vfsmount *nfs_follow_referral(const struct vfsmount *mnt_parent, + const struct dentry *dentry, + const struct nfs4_fs_locations *locations) +{ + struct vfsmount *mnt = ERR_PTR(-ENOENT); + struct nfs_clone_mount mountdata = { + .sb = mnt_parent->mnt_sb, + .dentry = dentry, + .authflavor = NFS_SB(mnt_parent->mnt_sb)->client->cl_auth->au_flavor, + }; + char *page = NULL, *page2 = NULL; + char *devname; + int loc, s, error; + + if (locations == NULL || locations->nlocations <= 0) + goto out; + + dprintk("%s: referral at %s/%s\n", __FUNCTION__, + dentry->d_parent->d_name.name, dentry->d_name.name); + + page = (char *) __get_free_page(GFP_USER); + if (!page) + goto out; + + page2 = (char *) __get_free_page(GFP_USER); + if (!page2) + goto out; + + /* Ensure fs path is a prefix of current dentry path */ + error = nfs4_validate_fspath(mnt_parent, dentry, locations, page, page2); + if (error < 0) { + mnt = ERR_PTR(error); + goto out; + } + + devname = nfs_devname(mnt_parent, dentry, page, PAGE_SIZE); + if (IS_ERR(devname)) { + mnt = (struct vfsmount *)devname; + goto out; + } + + loc = 0; + while (loc < locations->nlocations && IS_ERR(mnt)) { + const struct nfs4_fs_location *location = &locations->locations[loc]; + char *mnt_path; + + if (location == NULL || location->nservers <= 0 || + location->rootpath.ncomponents == 0) { + loc++; + continue; + } + + mnt_path = nfs4_pathname_string(&location->rootpath, page2, PAGE_SIZE); + if (IS_ERR(mnt_path)) { + loc++; + continue; + } + mountdata.mnt_path = mnt_path; + + s = 0; + while (s < location->nservers) { + struct sockaddr_in addr = {}; + + if (location->servers[s].len <= 0 || + valid_ipaddr4(location->servers[s].data) < 0) { + s++; + continue; + } + + mountdata.hostname = location->servers[s].data; + addr.sin_addr.s_addr = in_aton(mountdata.hostname); + addr.sin_family = AF_INET; + addr.sin_port = htons(NFS_PORT); + mountdata.addr = &addr; + + mnt = vfs_kern_mount(&nfs4_referral_fs_type, 0, devname, &mountdata); + if (!IS_ERR(mnt)) { + break; + } + s++; + } + loc++; + } + +out: + free_page((unsigned long) page); + free_page((unsigned long) page2); + dprintk("%s: done\n", __FUNCTION__); + return mnt; +} + +/* + * nfs_do_refmount - handle crossing a referral on server + * @dentry - dentry of referral + * @nd - nameidata info + * + */ +struct vfsmount *nfs_do_refmount(const struct vfsmount *mnt_parent, struct dentry *dentry) +{ + struct vfsmount *mnt = ERR_PTR(-ENOMEM); + struct dentry *parent; + struct nfs4_fs_locations *fs_locations = NULL; + struct page *page; + int err; + + /* BUG_ON(IS_ROOT(dentry)); */ + dprintk("%s: enter\n", __FUNCTION__); + + page = alloc_page(GFP_KERNEL); + if (page == NULL) + goto out; + + fs_locations = kmalloc(sizeof(struct nfs4_fs_locations), GFP_KERNEL); + if (fs_locations == NULL) + goto out_free; + + /* Get locations */ + mnt = ERR_PTR(-ENOENT); + + parent = dget_parent(dentry); + dprintk("%s: getting locations for %s/%s\n", + __FUNCTION__, parent->d_name.name, dentry->d_name.name); + + err = nfs4_proc_fs_locations(parent->d_inode, dentry, fs_locations, page); + dput(parent); + if (err != 0 || + fs_locations->nlocations <= 0 || + fs_locations->fs_path.ncomponents <= 0) + goto out_free; + + mnt = nfs_follow_referral(mnt_parent, dentry, fs_locations); +out_free: + __free_page(page); + kfree(fs_locations); +out: + dprintk("%s: done\n", __FUNCTION__); + return mnt; +} diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index d86c0db7b1e..47c7e6e3910 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -55,7 +55,7 @@ #define NFSDBG_FACILITY NFSDBG_PROC -#define NFS4_POLL_RETRY_MIN (1*HZ) +#define NFS4_POLL_RETRY_MIN (HZ/10) #define NFS4_POLL_RETRY_MAX (15*HZ) struct nfs4_opendata; @@ -64,9 +64,7 @@ static int nfs4_do_fsinfo(struct nfs_server *, struct nfs_fh *, struct nfs_fsinf static int nfs4_async_handle_error(struct rpc_task *, const struct nfs_server *); static int _nfs4_proc_access(struct inode *inode, struct nfs_access_entry *entry); static int nfs4_handle_exception(const struct nfs_server *server, int errorcode, struct nfs4_exception *exception); -static int nfs4_wait_clnt_recover(struct rpc_clnt *clnt, struct nfs4_client *clp); -extern u32 *nfs4_decode_dirent(u32 *p, struct nfs_entry *entry, int plus); -extern struct rpc_procinfo nfs4_procedures[]; +static int nfs4_wait_clnt_recover(struct rpc_clnt *clnt, struct nfs_client *clp); /* Prevent leaks of NFSv4 errors into userland */ int nfs4_map_errors(int err) @@ -121,6 +119,25 @@ const u32 nfs4_fsinfo_bitmap[2] = { FATTR4_WORD0_MAXFILESIZE 0 }; +const u32 nfs4_fs_locations_bitmap[2] = { + FATTR4_WORD0_TYPE + | FATTR4_WORD0_CHANGE + | FATTR4_WORD0_SIZE + | FATTR4_WORD0_FSID + | FATTR4_WORD0_FILEID + | FATTR4_WORD0_FS_LOCATIONS, + FATTR4_WORD1_MODE + | FATTR4_WORD1_NUMLINKS + | FATTR4_WORD1_OWNER + | FATTR4_WORD1_OWNER_GROUP + | FATTR4_WORD1_RAWDEV + | FATTR4_WORD1_SPACE_USED + | FATTR4_WORD1_TIME_ACCESS + | FATTR4_WORD1_TIME_METADATA + | FATTR4_WORD1_TIME_MODIFY + | FATTR4_WORD1_MOUNTED_ON_FILEID +}; + static void nfs4_setup_readdir(u64 cookie, u32 *verifier, struct dentry *dentry, struct nfs4_readdir_arg *readdir) { @@ -178,22 +195,22 @@ static void nfs4_setup_readdir(u64 cookie, u32 *verifier, struct dentry *dentry, static void renew_lease(const struct nfs_server *server, unsigned long timestamp) { - struct nfs4_client *clp = server->nfs4_state; + 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 update_changeattr(struct inode *inode, struct nfs4_change_info *cinfo) +static void update_changeattr(struct inode *dir, struct nfs4_change_info *cinfo) { - struct nfs_inode *nfsi = NFS_I(inode); + struct nfs_inode *nfsi = NFS_I(dir); - spin_lock(&inode->i_lock); - nfsi->cache_validity |= NFS_INO_INVALID_ATTR; + spin_lock(&dir->i_lock); + nfsi->cache_validity |= NFS_INO_INVALID_ATTR|NFS_INO_REVAL_PAGECACHE|NFS_INO_INVALID_DATA; if (cinfo->before == nfsi->change_attr && cinfo->atomic) nfsi->change_attr = cinfo->after; - spin_unlock(&inode->i_lock); + spin_unlock(&dir->i_lock); } struct nfs4_opendata { @@ -235,7 +252,7 @@ static struct nfs4_opendata *nfs4_opendata_alloc(struct dentry *dentry, atomic_inc(&sp->so_count); p->o_arg.fh = NFS_FH(dir); p->o_arg.open_flags = flags, - p->o_arg.clientid = server->nfs4_state->cl_clientid; + p->o_arg.clientid = server->nfs_client->cl_clientid; p->o_arg.id = sp->so_id; p->o_arg.name = &dentry->d_name; p->o_arg.server = server; @@ -533,7 +550,7 @@ int nfs4_open_delegation_recall(struct dentry *dentry, struct nfs4_state *state) case -NFS4ERR_STALE_STATEID: case -NFS4ERR_EXPIRED: /* Don't recall a delegation if it was lost */ - nfs4_schedule_state_recovery(server->nfs4_state); + nfs4_schedule_state_recovery(server->nfs_client); return err; } err = nfs4_handle_exception(server, err, &exception); @@ -741,7 +758,7 @@ static int _nfs4_proc_open(struct nfs4_opendata *data) } nfs_confirm_seqid(&data->owner->so_seqid, 0); if (!(o_res->f_attr->valid & NFS_ATTR_FATTR)) - return server->rpc_ops->getattr(server, &o_res->fh, o_res->f_attr); + return server->nfs_client->rpc_ops->getattr(server, &o_res->fh, o_res->f_attr); return 0; } @@ -775,11 +792,18 @@ out: int nfs4_recover_expired_lease(struct nfs_server *server) { - struct nfs4_client *clp = server->nfs4_state; + struct nfs_client *clp = server->nfs_client; + int ret; - if (test_and_clear_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state)) + for (;;) { + ret = nfs4_wait_clnt_recover(server->client, clp); + if (ret != 0) + return ret; + if (!test_and_clear_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state)) + break; nfs4_schedule_state_recovery(clp); - return nfs4_wait_clnt_recover(server->client, clp); + } + return 0; } /* @@ -850,7 +874,7 @@ static int _nfs4_open_delegated(struct inode *inode, int flags, struct rpc_cred { struct nfs_delegation *delegation; struct nfs_server *server = NFS_SERVER(inode); - struct nfs4_client *clp = server->nfs4_state; + struct nfs_client *clp = server->nfs_client; struct nfs_inode *nfsi = NFS_I(inode); struct nfs4_state_owner *sp = NULL; struct nfs4_state *state = NULL; @@ -936,7 +960,7 @@ static int _nfs4_do_open(struct inode *dir, struct dentry *dentry, int flags, st struct nfs4_state_owner *sp; struct nfs4_state *state = NULL; struct nfs_server *server = NFS_SERVER(dir); - struct nfs4_client *clp = server->nfs4_state; + struct nfs_client *clp = server->nfs_client; struct nfs4_opendata *opendata; int status; @@ -953,7 +977,7 @@ static int _nfs4_do_open(struct inode *dir, struct dentry *dentry, int flags, st status = -ENOMEM; opendata = nfs4_opendata_alloc(dentry, sp, flags, sattr); if (opendata == NULL) - goto err_put_state_owner; + goto err_release_rwsem; status = _nfs4_proc_open(opendata); if (status != 0) @@ -972,11 +996,11 @@ static int _nfs4_do_open(struct inode *dir, struct dentry *dentry, int flags, st return 0; err_opendata_free: nfs4_opendata_free(opendata); +err_release_rwsem: + up_read(&clp->cl_sem); err_put_state_owner: nfs4_put_state_owner(sp); out_err: - /* Note: clp->cl_sem must be released before nfs4_put_open_state()! */ - up_read(&clp->cl_sem); *res = NULL; return status; } @@ -1116,7 +1140,7 @@ static void nfs4_close_done(struct rpc_task *task, void *data) break; case -NFS4ERR_STALE_STATEID: case -NFS4ERR_EXPIRED: - nfs4_schedule_state_recovery(server->nfs4_state); + nfs4_schedule_state_recovery(server->nfs_client); break; default: if (nfs4_async_handle_error(task, server) == -EAGAIN) { @@ -1251,7 +1275,7 @@ nfs4_atomic_open(struct inode *dir, struct dentry *dentry, struct nameidata *nd) BUG_ON(nd->intent.open.flags & O_CREAT); } - cred = rpcauth_lookupcred(NFS_SERVER(dir)->client->cl_auth, 0); + cred = rpcauth_lookupcred(NFS_CLIENT(dir)->cl_auth, 0); if (IS_ERR(cred)) return (struct dentry *)cred; state = nfs4_do_open(dir, dentry, nd->intent.open.flags, &attr, cred); @@ -1274,7 +1298,7 @@ nfs4_open_revalidate(struct inode *dir, struct dentry *dentry, int openflags, st struct rpc_cred *cred; struct nfs4_state *state; - cred = rpcauth_lookupcred(NFS_SERVER(dir)->client->cl_auth, 0); + cred = rpcauth_lookupcred(NFS_CLIENT(dir)->cl_auth, 0); if (IS_ERR(cred)) return PTR_ERR(cred); state = nfs4_open_delegated(dentry->d_inode, openflags, cred); @@ -1331,7 +1355,7 @@ static int _nfs4_server_capabilities(struct nfs_server *server, struct nfs_fh *f return status; } -static int nfs4_server_capabilities(struct nfs_server *server, struct nfs_fh *fhandle) +int nfs4_server_capabilities(struct nfs_server *server, struct nfs_fh *fhandle) { struct nfs4_exception exception = { }; int err; @@ -1376,73 +1400,66 @@ static int nfs4_lookup_root(struct nfs_server *server, struct nfs_fh *fhandle, return err; } +/* + * get the file handle for the "/" directory on the server + */ static int nfs4_proc_get_root(struct nfs_server *server, struct nfs_fh *fhandle, - struct nfs_fsinfo *info) + struct nfs_fsinfo *info) { - struct nfs_fattr * fattr = info->fattr; - unsigned char * p; - struct qstr q; - struct nfs4_lookup_arg args = { - .dir_fh = fhandle, - .name = &q, - .bitmask = nfs4_fattr_bitmap, - }; - struct nfs4_lookup_res res = { - .server = server, - .fattr = fattr, - .fh = fhandle, - }; - struct rpc_message msg = { - .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_LOOKUP], - .rpc_argp = &args, - .rpc_resp = &res, - }; int status; - /* - * Now we do a separate LOOKUP for each component of the mount path. - * The LOOKUPs are done separately so that we can conveniently - * catch an ERR_WRONGSEC if it occurs along the way... - */ status = nfs4_lookup_root(server, fhandle, info); - if (status) - goto out; - - p = server->mnt_path; - for (;;) { - struct nfs4_exception exception = { }; - - while (*p == '/') - p++; - if (!*p) - break; - q.name = p; - while (*p && (*p != '/')) - p++; - q.len = p - q.name; - - do { - nfs_fattr_init(fattr); - status = nfs4_handle_exception(server, - rpc_call_sync(server->client, &msg, 0), - &exception); - } while (exception.retry); - if (status == 0) - continue; - if (status == -ENOENT) { - printk(KERN_NOTICE "NFS: mount path %s does not exist!\n", server->mnt_path); - printk(KERN_NOTICE "NFS: suggestion: try mounting '/' instead.\n"); - } - break; - } if (status == 0) status = nfs4_server_capabilities(server, fhandle); if (status == 0) status = nfs4_do_fsinfo(server, fhandle, info); -out: return nfs4_map_errors(status); } +/* + * Get locations and (maybe) other attributes of a referral. + * Note that we'll actually follow the referral later when + * we detect fsid mismatch in inode revalidation + */ +static int nfs4_get_referral(struct inode *dir, struct qstr *name, struct nfs_fattr *fattr, struct nfs_fh *fhandle) +{ + int status = -ENOMEM; + struct page *page = NULL; + struct nfs4_fs_locations *locations = NULL; + struct dentry dentry = {}; + + page = alloc_page(GFP_KERNEL); + if (page == NULL) + goto out; + locations = kmalloc(sizeof(struct nfs4_fs_locations), GFP_KERNEL); + if (locations == NULL) + goto out; + + dentry.d_name.name = name->name; + dentry.d_name.len = name->len; + status = nfs4_proc_fs_locations(dir, &dentry, locations, page); + if (status != 0) + goto out; + /* Make sure server returned a different fsid for the referral */ + if (nfs_fsid_equal(&NFS_SERVER(dir)->fsid, &locations->fattr.fsid)) { + dprintk("%s: server did not return a different fsid for a referral at %s\n", __FUNCTION__, name->name); + status = -EIO; + goto out; + } + + memcpy(fattr, &locations->fattr, sizeof(struct nfs_fattr)); + fattr->valid |= NFS_ATTR_FATTR_V4_REFERRAL; + if (!fattr->mode) + fattr->mode = S_IFDIR; + memset(fhandle, 0, sizeof(struct nfs_fh)); +out: + if (page) + __free_page(page); + if (locations) + kfree(locations); + return status; +} + static int _nfs4_proc_getattr(struct nfs_server *server, struct nfs_fh *fhandle, struct nfs_fattr *fattr) { struct nfs4_getattr_arg args = { @@ -1504,7 +1521,7 @@ nfs4_proc_setattr(struct dentry *dentry, struct nfs_fattr *fattr, nfs_fattr_init(fattr); - cred = rpcauth_lookupcred(NFS_SERVER(inode)->client->cl_auth, 0); + cred = rpcauth_lookupcred(NFS_CLIENT(inode)->cl_auth, 0); if (IS_ERR(cred)) return PTR_ERR(cred); @@ -1522,6 +1539,52 @@ nfs4_proc_setattr(struct dentry *dentry, struct nfs_fattr *fattr, return status; } +static int _nfs4_proc_lookupfh(struct nfs_server *server, struct nfs_fh *dirfh, + struct qstr *name, struct nfs_fh *fhandle, + struct nfs_fattr *fattr) +{ + int status; + struct nfs4_lookup_arg args = { + .bitmask = server->attr_bitmask, + .dir_fh = dirfh, + .name = name, + }; + struct nfs4_lookup_res res = { + .server = server, + .fattr = fattr, + .fh = fhandle, + }; + struct rpc_message msg = { + .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_LOOKUP], + .rpc_argp = &args, + .rpc_resp = &res, + }; + + nfs_fattr_init(fattr); + + dprintk("NFS call lookupfh %s\n", name->name); + status = rpc_call_sync(server->client, &msg, 0); + dprintk("NFS reply lookupfh: %d\n", status); + if (status == -NFS4ERR_MOVED) + status = -EREMOTE; + return status; +} + +static int nfs4_proc_lookupfh(struct nfs_server *server, struct nfs_fh *dirfh, + struct qstr *name, struct nfs_fh *fhandle, + struct nfs_fattr *fattr) +{ + struct nfs4_exception exception = { }; + int err; + do { + err = nfs4_handle_exception(server, + _nfs4_proc_lookupfh(server, dirfh, name, + fhandle, fattr), + &exception); + } while (exception.retry); + return err; +} + static int _nfs4_proc_lookup(struct inode *dir, struct qstr *name, struct nfs_fh *fhandle, struct nfs_fattr *fattr) { @@ -1547,6 +1610,8 @@ static int _nfs4_proc_lookup(struct inode *dir, struct qstr *name, dprintk("NFS call lookup %s\n", name->name); status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0); + if (status == -NFS4ERR_MOVED) + status = nfs4_get_referral(dir, name, fattr, fhandle); dprintk("NFS reply lookup: %d\n", status); return status; } @@ -1818,7 +1883,7 @@ nfs4_proc_create(struct inode *dir, struct dentry *dentry, struct iattr *sattr, struct rpc_cred *cred; int status = 0; - cred = rpcauth_lookupcred(NFS_SERVER(dir)->client->cl_auth, 0); + cred = rpcauth_lookupcred(NFS_CLIENT(dir)->cl_auth, 0); if (IS_ERR(cred)) { status = PTR_ERR(cred); goto out; @@ -2008,7 +2073,7 @@ static int _nfs4_proc_link(struct inode *inode, struct inode *dir, struct qstr * if (!status) { update_changeattr(dir, &res.cinfo); nfs_post_op_update_inode(dir, res.dir_attr); - nfs_refresh_inode(inode, res.fattr); + nfs_post_op_update_inode(inode, res.fattr); } return status; @@ -2026,24 +2091,24 @@ static int nfs4_proc_link(struct inode *inode, struct inode *dir, struct qstr *n return err; } -static int _nfs4_proc_symlink(struct inode *dir, struct qstr *name, - struct qstr *path, struct iattr *sattr, struct nfs_fh *fhandle, - struct nfs_fattr *fattr) +static int _nfs4_proc_symlink(struct inode *dir, struct dentry *dentry, + struct page *page, unsigned int len, struct iattr *sattr) { struct nfs_server *server = NFS_SERVER(dir); - struct nfs_fattr dir_fattr; + struct nfs_fh fhandle; + struct nfs_fattr fattr, dir_fattr; struct nfs4_create_arg arg = { .dir_fh = NFS_FH(dir), .server = server, - .name = name, + .name = &dentry->d_name, .attrs = sattr, .ftype = NF4LNK, .bitmask = server->attr_bitmask, }; struct nfs4_create_res res = { .server = server, - .fh = fhandle, - .fattr = fattr, + .fh = &fhandle, + .fattr = &fattr, .dir_fattr = &dir_fattr, }; struct rpc_message msg = { @@ -2053,29 +2118,32 @@ static int _nfs4_proc_symlink(struct inode *dir, struct qstr *name, }; int status; - if (path->len > NFS4_MAXPATHLEN) + if (len > NFS4_MAXPATHLEN) return -ENAMETOOLONG; - arg.u.symlink = path; - nfs_fattr_init(fattr); + + arg.u.symlink.pages = &page; + arg.u.symlink.len = len; + nfs_fattr_init(&fattr); nfs_fattr_init(&dir_fattr); status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0); - if (!status) + if (!status) { update_changeattr(dir, &res.dir_cinfo); - nfs_post_op_update_inode(dir, res.dir_fattr); + nfs_post_op_update_inode(dir, res.dir_fattr); + status = nfs_instantiate(dentry, &fhandle, &fattr); + } return status; } -static int nfs4_proc_symlink(struct inode *dir, struct qstr *name, - struct qstr *path, struct iattr *sattr, struct nfs_fh *fhandle, - struct nfs_fattr *fattr) +static int nfs4_proc_symlink(struct inode *dir, struct dentry *dentry, + struct page *page, unsigned int len, struct iattr *sattr) { struct nfs4_exception exception = { }; int err; do { err = nfs4_handle_exception(NFS_SERVER(dir), - _nfs4_proc_symlink(dir, name, path, sattr, - fhandle, fattr), + _nfs4_proc_symlink(dir, dentry, page, + len, sattr), &exception); } while (exception.retry); return err; @@ -2458,7 +2526,7 @@ static void nfs4_proc_commit_setup(struct nfs_write_data *data, int how) */ static void nfs4_renew_done(struct rpc_task *task, void *data) { - struct nfs4_client *clp = (struct nfs4_client *)task->tk_msg.rpc_argp; + struct nfs_client *clp = (struct nfs_client *)task->tk_msg.rpc_argp; unsigned long timestamp = (unsigned long)data; if (task->tk_status < 0) { @@ -2480,7 +2548,7 @@ static const struct rpc_call_ops nfs4_renew_ops = { .rpc_call_done = nfs4_renew_done, }; -int nfs4_proc_async_renew(struct nfs4_client *clp, struct rpc_cred *cred) +int nfs4_proc_async_renew(struct nfs_client *clp, struct rpc_cred *cred) { struct rpc_message msg = { .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_RENEW], @@ -2492,7 +2560,7 @@ int nfs4_proc_async_renew(struct nfs4_client *clp, struct rpc_cred *cred) &nfs4_renew_ops, (void *)jiffies); } -int nfs4_proc_renew(struct nfs4_client *clp, struct rpc_cred *cred) +int nfs4_proc_renew(struct nfs_client *clp, struct rpc_cred *cred) { struct rpc_message msg = { .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_RENEW], @@ -2605,7 +2673,7 @@ out: nfs4_set_cached_acl(inode, acl); } -static inline ssize_t nfs4_get_acl_uncached(struct inode *inode, void *buf, size_t buflen) +static ssize_t __nfs4_get_acl_uncached(struct inode *inode, void *buf, size_t buflen) { struct page *pages[NFS4ACL_MAXPAGES]; struct nfs_getaclargs args = { @@ -2658,6 +2726,19 @@ out_free: return ret; } +static ssize_t nfs4_get_acl_uncached(struct inode *inode, void *buf, size_t buflen) +{ + struct nfs4_exception exception = { }; + ssize_t ret; + do { + ret = __nfs4_get_acl_uncached(inode, buf, buflen); + if (ret >= 0) + break; + ret = nfs4_handle_exception(NFS_SERVER(inode), ret, &exception); + } while (exception.retry); + return ret; +} + static ssize_t nfs4_proc_get_acl(struct inode *inode, void *buf, size_t buflen) { struct nfs_server *server = NFS_SERVER(inode); @@ -2674,7 +2755,7 @@ static ssize_t nfs4_proc_get_acl(struct inode *inode, void *buf, size_t buflen) return nfs4_get_acl_uncached(inode, buf, buflen); } -static int nfs4_proc_set_acl(struct inode *inode, const void *buf, size_t buflen) +static int __nfs4_proc_set_acl(struct inode *inode, const void *buf, size_t buflen) { struct nfs_server *server = NFS_SERVER(inode); struct page *pages[NFS4ACL_MAXPAGES]; @@ -2694,16 +2775,28 @@ static int nfs4_proc_set_acl(struct inode *inode, const void *buf, size_t buflen return -EOPNOTSUPP; nfs_inode_return_delegation(inode); buf_to_pages(buf, buflen, arg.acl_pages, &arg.acl_pgbase); - ret = rpc_call_sync(NFS_SERVER(inode)->client, &msg, 0); + ret = rpc_call_sync(NFS_CLIENT(inode), &msg, 0); if (ret == 0) nfs4_write_cached_acl(inode, buf, buflen); return ret; } +static int nfs4_proc_set_acl(struct inode *inode, const void *buf, size_t buflen) +{ + struct nfs4_exception exception = { }; + int err; + do { + err = nfs4_handle_exception(NFS_SERVER(inode), + __nfs4_proc_set_acl(inode, buf, buflen), + &exception); + } while (exception.retry); + return err; +} + static int nfs4_async_handle_error(struct rpc_task *task, const struct nfs_server *server) { - struct nfs4_client *clp = server->nfs4_state; + struct nfs_client *clp = server->nfs_client; if (!clp || task->tk_status >= 0) return 0; @@ -2740,7 +2833,7 @@ static int nfs4_wait_bit_interruptible(void *word) return 0; } -static int nfs4_wait_clnt_recover(struct rpc_clnt *clnt, struct nfs4_client *clp) +static int nfs4_wait_clnt_recover(struct rpc_clnt *clnt, struct nfs_client *clp) { sigset_t oldset; int res; @@ -2783,7 +2876,7 @@ static int nfs4_delay(struct rpc_clnt *clnt, long *timeout) */ int nfs4_handle_exception(const struct nfs_server *server, int errorcode, struct nfs4_exception *exception) { - struct nfs4_client *clp = server->nfs4_state; + struct nfs_client *clp = server->nfs_client; int ret = errorcode; exception->retry = 0; @@ -2798,6 +2891,7 @@ int nfs4_handle_exception(const struct nfs_server *server, int errorcode, struct if (ret == 0) exception->retry = 1; break; + case -NFS4ERR_FILE_OPEN: case -NFS4ERR_GRACE: case -NFS4ERR_DELAY: ret = nfs4_delay(server->client, &exception->timeout); @@ -2810,7 +2904,7 @@ int nfs4_handle_exception(const struct nfs_server *server, int errorcode, struct return nfs4_map_errors(ret); } -int nfs4_proc_setclientid(struct nfs4_client *clp, u32 program, unsigned short port, struct rpc_cred *cred) +int nfs4_proc_setclientid(struct nfs_client *clp, u32 program, unsigned short port, struct rpc_cred *cred) { nfs4_verifier sc_verifier; struct nfs4_setclientid setclientid = { @@ -2834,7 +2928,7 @@ int nfs4_proc_setclientid(struct nfs4_client *clp, u32 program, unsigned short p for(;;) { setclientid.sc_name_len = scnprintf(setclientid.sc_name, sizeof(setclientid.sc_name), "%s/%u.%u.%u.%u %s %u", - clp->cl_ipaddr, NIPQUAD(clp->cl_addr.s_addr), + clp->cl_ipaddr, NIPQUAD(clp->cl_addr.sin_addr), cred->cr_ops->cr_name, clp->cl_id_uniquifier); setclientid.sc_netid_len = scnprintf(setclientid.sc_netid, @@ -2857,7 +2951,7 @@ int nfs4_proc_setclientid(struct nfs4_client *clp, u32 program, unsigned short p return status; } -static int _nfs4_proc_setclientid_confirm(struct nfs4_client *clp, struct rpc_cred *cred) +static int _nfs4_proc_setclientid_confirm(struct nfs_client *clp, struct rpc_cred *cred) { struct nfs_fsinfo fsinfo; struct rpc_message msg = { @@ -2881,7 +2975,7 @@ static int _nfs4_proc_setclientid_confirm(struct nfs4_client *clp, struct rpc_cr return status; } -int nfs4_proc_setclientid_confirm(struct nfs4_client *clp, struct rpc_cred *cred) +int nfs4_proc_setclientid_confirm(struct nfs_client *clp, struct rpc_cred *cred) { long timeout; int err; @@ -2989,7 +3083,7 @@ int nfs4_proc_delegreturn(struct inode *inode, struct rpc_cred *cred, const nfs4 switch (err) { case -NFS4ERR_STALE_STATEID: case -NFS4ERR_EXPIRED: - nfs4_schedule_state_recovery(server->nfs4_state); + nfs4_schedule_state_recovery(server->nfs_client); case 0: return 0; } @@ -3018,7 +3112,7 @@ static int _nfs4_proc_getlk(struct nfs4_state *state, int cmd, struct file_lock { struct inode *inode = state->inode; struct nfs_server *server = NFS_SERVER(inode); - struct nfs4_client *clp = server->nfs4_state; + struct nfs_client *clp = server->nfs_client; struct nfs_lockt_args arg = { .fh = NFS_FH(inode), .fl = request, @@ -3081,9 +3175,6 @@ static int do_vfs_lock(struct file *file, struct file_lock *fl) default: BUG(); } - if (res < 0) - printk(KERN_WARNING "%s: VFS is out of sync with lock manager!\n", - __FUNCTION__); return res; } @@ -3146,7 +3237,7 @@ static void nfs4_locku_done(struct rpc_task *task, void *data) break; case -NFS4ERR_STALE_STATEID: case -NFS4ERR_EXPIRED: - nfs4_schedule_state_recovery(calldata->server->nfs4_state); + nfs4_schedule_state_recovery(calldata->server->nfs_client); break; default: if (nfs4_async_handle_error(task, calldata->server) == -EAGAIN) { @@ -3195,8 +3286,6 @@ static struct rpc_task *nfs4_do_unlck(struct file_lock *fl, return ERR_PTR(-ENOMEM); } - /* Unlock _before_ we do the RPC call */ - do_vfs_lock(fl->fl_file, fl); return rpc_run_task(NFS_CLIENT(lsp->ls_state->inode), RPC_TASK_ASYNC, &nfs4_locku_ops, data); } @@ -3207,30 +3296,28 @@ static int nfs4_proc_unlck(struct nfs4_state *state, int cmd, struct file_lock * struct rpc_task *task; int status = 0; - /* Is this a delegated lock? */ - if (test_bit(NFS_DELEGATED_STATE, &state->flags)) - goto out_unlock; - /* Is this open_owner holding any locks on the server? */ - if (test_bit(LK_STATE_IN_USE, &state->flags) == 0) - goto out_unlock; - status = nfs4_set_lock_state(state, request); + /* Unlock _before_ we do the RPC call */ + request->fl_flags |= FL_EXISTS; + if (do_vfs_lock(request->fl_file, request) == -ENOENT) + goto out; if (status != 0) - goto out_unlock; + goto out; + /* Is this a delegated lock? */ + if (test_bit(NFS_DELEGATED_STATE, &state->flags)) + goto out; lsp = request->fl_u.nfs4_fl.owner; - status = -ENOMEM; seqid = nfs_alloc_seqid(&lsp->ls_seqid); + status = -ENOMEM; if (seqid == NULL) - goto out_unlock; + goto out; task = nfs4_do_unlck(request, request->fl_file->private_data, lsp, seqid); status = PTR_ERR(task); if (IS_ERR(task)) - goto out_unlock; + goto out; status = nfs4_wait_for_completion_rpc_task(task); rpc_release_task(task); - return status; -out_unlock: - do_vfs_lock(request->fl_file, request); +out: return status; } @@ -3262,7 +3349,7 @@ static struct nfs4_lockdata *nfs4_alloc_lockdata(struct file_lock *fl, if (p->arg.lock_seqid == NULL) goto out_free; p->arg.lock_stateid = &lsp->ls_stateid; - p->arg.lock_owner.clientid = server->nfs4_state->cl_clientid; + p->arg.lock_owner.clientid = server->nfs_client->cl_clientid; p->arg.lock_owner.id = lsp->ls_id; p->lsp = lsp; atomic_inc(&lsp->ls_count); @@ -3398,10 +3485,10 @@ static int nfs4_lock_reclaim(struct nfs4_state *state, struct file_lock *request struct nfs4_exception exception = { }; int err; - /* Cache the lock if possible... */ - if (test_bit(NFS_DELEGATED_STATE, &state->flags)) - return 0; do { + /* Cache the lock if possible... */ + if (test_bit(NFS_DELEGATED_STATE, &state->flags) != 0) + return 0; err = _nfs4_do_setlk(state, F_SETLK, request, 1); if (err != -NFS4ERR_DELAY) break; @@ -3420,6 +3507,8 @@ static int nfs4_lock_expired(struct nfs4_state *state, struct file_lock *request if (err != 0) return err; do { + if (test_bit(NFS_DELEGATED_STATE, &state->flags) != 0) + return 0; err = _nfs4_do_setlk(state, F_SETLK, request, 0); if (err != -NFS4ERR_DELAY) break; @@ -3430,30 +3519,43 @@ static int nfs4_lock_expired(struct nfs4_state *state, struct file_lock *request static int _nfs4_proc_setlk(struct nfs4_state *state, int cmd, struct file_lock *request) { - struct nfs4_client *clp = state->owner->so_client; + struct nfs_client *clp = state->owner->so_client; + unsigned char fl_flags = request->fl_flags; int status; /* Is this a delegated open? */ - if (NFS_I(state->inode)->delegation_state != 0) { - /* Yes: cache locks! */ - status = do_vfs_lock(request->fl_file, request); - /* ...but avoid races with delegation recall... */ - if (status < 0 || test_bit(NFS_DELEGATED_STATE, &state->flags)) - return status; - } - down_read(&clp->cl_sem); status = nfs4_set_lock_state(state, request); if (status != 0) goto out; + request->fl_flags |= FL_ACCESS; + status = do_vfs_lock(request->fl_file, request); + if (status < 0) + goto out; + down_read(&clp->cl_sem); + if (test_bit(NFS_DELEGATED_STATE, &state->flags)) { + struct nfs_inode *nfsi = NFS_I(state->inode); + /* Yes: cache locks! */ + down_read(&nfsi->rwsem); + /* ...but avoid races with delegation recall... */ + if (test_bit(NFS_DELEGATED_STATE, &state->flags)) { + request->fl_flags = fl_flags & ~FL_SLEEP; + status = do_vfs_lock(request->fl_file, request); + up_read(&nfsi->rwsem); + goto out_unlock; + } + up_read(&nfsi->rwsem); + } status = _nfs4_do_setlk(state, cmd, request, 0); if (status != 0) - goto out; + goto out_unlock; /* Note: we always want to sleep here! */ - request->fl_flags |= FL_SLEEP; + request->fl_flags = fl_flags | FL_SLEEP; if (do_vfs_lock(request->fl_file, request) < 0) printk(KERN_WARNING "%s: VFS is out of sync with lock manager!\n", __FUNCTION__); -out: +out_unlock: up_read(&clp->cl_sem); +out: + request->fl_flags = fl_flags; return status; } @@ -3570,6 +3672,36 @@ ssize_t nfs4_listxattr(struct dentry *dentry, char *buf, size_t buflen) return len; } +int nfs4_proc_fs_locations(struct inode *dir, struct dentry *dentry, + struct nfs4_fs_locations *fs_locations, struct page *page) +{ + struct nfs_server *server = NFS_SERVER(dir); + u32 bitmask[2] = { + [0] = FATTR4_WORD0_FSID | FATTR4_WORD0_FS_LOCATIONS, + [1] = FATTR4_WORD1_MOUNTED_ON_FILEID, + }; + struct nfs4_fs_locations_arg args = { + .dir_fh = NFS_FH(dir), + .name = &dentry->d_name, + .page = page, + .bitmask = bitmask, + }; + struct rpc_message msg = { + .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_FS_LOCATIONS], + .rpc_argp = &args, + .rpc_resp = fs_locations, + }; + int status; + + dprintk("%s: start\n", __FUNCTION__); + fs_locations->fattr.valid = 0; + fs_locations->server = server; + fs_locations->nlocations = 0; + status = rpc_call_sync(server->client, &msg, 0); + dprintk("%s: returned status = %d\n", __FUNCTION__, status); + return status; +} + struct nfs4_state_recovery_ops nfs4_reboot_recovery_ops = { .recover_open = nfs4_open_reclaim, .recover_lock = nfs4_lock_reclaim, @@ -3589,7 +3721,7 @@ static struct inode_operations nfs4_file_inode_operations = { .listxattr = nfs4_listxattr, }; -struct nfs_rpc_ops nfs_v4_clientops = { +const struct nfs_rpc_ops nfs_v4_clientops = { .version = 4, /* protocol version */ .dentry_ops = &nfs4_dentry_operations, .dir_inode_ops = &nfs4_dir_inode_operations, @@ -3597,6 +3729,7 @@ struct nfs_rpc_ops nfs_v4_clientops = { .getroot = nfs4_proc_get_root, .getattr = nfs4_proc_getattr, .setattr = nfs4_proc_setattr, + .lookupfh = nfs4_proc_lookupfh, .lookup = nfs4_proc_lookup, .access = nfs4_proc_access, .readlink = nfs4_proc_readlink, @@ -3617,6 +3750,7 @@ struct nfs_rpc_ops nfs_v4_clientops = { .statfs = nfs4_proc_statfs, .fsinfo = nfs4_proc_fsinfo, .pathconf = nfs4_proc_pathconf, + .set_capabilities = nfs4_server_capabilities, .decode_dirent = nfs4_decode_dirent, .read_setup = nfs4_proc_read_setup, .read_done = nfs4_read_done, diff --git a/fs/nfs/nfs4renewd.c b/fs/nfs/nfs4renewd.c index 5d764d8e6d8..7b6df1852e7 100644 --- a/fs/nfs/nfs4renewd.c +++ b/fs/nfs/nfs4renewd.c @@ -61,7 +61,7 @@ void nfs4_renew_state(void *data) { - struct nfs4_client *clp = (struct nfs4_client *)data; + struct nfs_client *clp = (struct nfs_client *)data; struct rpc_cred *cred; long lease, timeout; unsigned long last, now; @@ -108,7 +108,7 @@ out: /* Must be called with clp->cl_sem locked for writes */ void -nfs4_schedule_state_renewal(struct nfs4_client *clp) +nfs4_schedule_state_renewal(struct nfs_client *clp) { long timeout; @@ -121,32 +121,20 @@ nfs4_schedule_state_renewal(struct nfs4_client *clp) __FUNCTION__, (timeout + HZ - 1) / HZ); cancel_delayed_work(&clp->cl_renewd); schedule_delayed_work(&clp->cl_renewd, timeout); + set_bit(NFS_CS_RENEWD, &clp->cl_res_state); spin_unlock(&clp->cl_lock); } void nfs4_renewd_prepare_shutdown(struct nfs_server *server) { - struct nfs4_client *clp = server->nfs4_state; - - if (!clp) - return; flush_scheduled_work(); - down_write(&clp->cl_sem); - if (!list_empty(&server->nfs4_siblings)) - list_del_init(&server->nfs4_siblings); - up_write(&clp->cl_sem); } -/* Must be called with clp->cl_sem locked for writes */ void -nfs4_kill_renewd(struct nfs4_client *clp) +nfs4_kill_renewd(struct nfs_client *clp) { down_read(&clp->cl_sem); - if (!list_empty(&clp->cl_superblocks)) { - up_read(&clp->cl_sem); - return; - } cancel_delayed_work(&clp->cl_renewd); up_read(&clp->cl_sem); flush_scheduled_work(); diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c index 96e5b82c153..5fffbdfa971 100644 --- a/fs/nfs/nfs4state.c +++ b/fs/nfs/nfs4state.c @@ -38,7 +38,6 @@ * subsequent patch. */ -#include <linux/config.h> #include <linux/slab.h> #include <linux/smp_lock.h> #include <linux/nfs_fs.h> @@ -51,149 +50,15 @@ #include "nfs4_fs.h" #include "callback.h" #include "delegation.h" +#include "internal.h" #define OPENOWNER_POOL_SIZE 8 const nfs4_stateid zero_stateid; -static DEFINE_SPINLOCK(state_spinlock); static LIST_HEAD(nfs4_clientid_list); -void -init_nfsv4_state(struct nfs_server *server) -{ - server->nfs4_state = NULL; - INIT_LIST_HEAD(&server->nfs4_siblings); -} - -void -destroy_nfsv4_state(struct nfs_server *server) -{ - kfree(server->mnt_path); - server->mnt_path = NULL; - if (server->nfs4_state) { - nfs4_put_client(server->nfs4_state); - server->nfs4_state = NULL; - } -} - -/* - * nfs4_get_client(): returns an empty client structure - * nfs4_put_client(): drops reference to client structure - * - * Since these are allocated/deallocated very rarely, we don't - * bother putting them in a slab cache... - */ -static struct nfs4_client * -nfs4_alloc_client(struct in_addr *addr) -{ - struct nfs4_client *clp; - - if (nfs_callback_up() < 0) - return NULL; - if ((clp = kzalloc(sizeof(*clp), GFP_KERNEL)) == NULL) { - nfs_callback_down(); - return NULL; - } - memcpy(&clp->cl_addr, addr, sizeof(clp->cl_addr)); - init_rwsem(&clp->cl_sem); - INIT_LIST_HEAD(&clp->cl_delegations); - INIT_LIST_HEAD(&clp->cl_state_owners); - INIT_LIST_HEAD(&clp->cl_unused); - spin_lock_init(&clp->cl_lock); - atomic_set(&clp->cl_count, 1); - INIT_WORK(&clp->cl_renewd, nfs4_renew_state, clp); - INIT_LIST_HEAD(&clp->cl_superblocks); - rpc_init_wait_queue(&clp->cl_rpcwaitq, "NFS4 client"); - clp->cl_rpcclient = ERR_PTR(-EINVAL); - clp->cl_boot_time = CURRENT_TIME; - clp->cl_state = 1 << NFS4CLNT_LEASE_EXPIRED; - return clp; -} - -static void -nfs4_free_client(struct nfs4_client *clp) -{ - struct nfs4_state_owner *sp; - - while (!list_empty(&clp->cl_unused)) { - sp = list_entry(clp->cl_unused.next, - struct nfs4_state_owner, - so_list); - list_del(&sp->so_list); - kfree(sp); - } - BUG_ON(!list_empty(&clp->cl_state_owners)); - nfs_idmap_delete(clp); - if (!IS_ERR(clp->cl_rpcclient)) - rpc_shutdown_client(clp->cl_rpcclient); - kfree(clp); - nfs_callback_down(); -} - -static struct nfs4_client *__nfs4_find_client(struct in_addr *addr) -{ - struct nfs4_client *clp; - list_for_each_entry(clp, &nfs4_clientid_list, cl_servers) { - if (memcmp(&clp->cl_addr, addr, sizeof(clp->cl_addr)) == 0) { - atomic_inc(&clp->cl_count); - return clp; - } - } - return NULL; -} - -struct nfs4_client *nfs4_find_client(struct in_addr *addr) -{ - struct nfs4_client *clp; - spin_lock(&state_spinlock); - clp = __nfs4_find_client(addr); - spin_unlock(&state_spinlock); - return clp; -} - -struct nfs4_client * -nfs4_get_client(struct in_addr *addr) -{ - struct nfs4_client *clp, *new = NULL; - - spin_lock(&state_spinlock); - for (;;) { - clp = __nfs4_find_client(addr); - if (clp != NULL) - break; - clp = new; - if (clp != NULL) { - list_add(&clp->cl_servers, &nfs4_clientid_list); - new = NULL; - break; - } - spin_unlock(&state_spinlock); - new = nfs4_alloc_client(addr); - spin_lock(&state_spinlock); - if (new == NULL) - break; - } - spin_unlock(&state_spinlock); - if (new) - nfs4_free_client(new); - return clp; -} - -void -nfs4_put_client(struct nfs4_client *clp) -{ - if (!atomic_dec_and_lock(&clp->cl_count, &state_spinlock)) - return; - list_del(&clp->cl_servers); - spin_unlock(&state_spinlock); - BUG_ON(!list_empty(&clp->cl_superblocks)); - rpc_wake_up(&clp->cl_rpcwaitq); - nfs4_kill_renewd(clp); - nfs4_free_client(clp); -} - -static int nfs4_init_client(struct nfs4_client *clp, struct rpc_cred *cred) +static int nfs4_init_client(struct nfs_client *clp, struct rpc_cred *cred) { int status = nfs4_proc_setclientid(clp, NFS4_CALLBACK, nfs_callback_tcpport, cred); @@ -205,13 +70,13 @@ static int nfs4_init_client(struct nfs4_client *clp, struct rpc_cred *cred) } u32 -nfs4_alloc_lockowner_id(struct nfs4_client *clp) +nfs4_alloc_lockowner_id(struct nfs_client *clp) { return clp->cl_lockowner_id ++; } static struct nfs4_state_owner * -nfs4_client_grab_unused(struct nfs4_client *clp, struct rpc_cred *cred) +nfs4_client_grab_unused(struct nfs_client *clp, struct rpc_cred *cred) { struct nfs4_state_owner *sp = NULL; @@ -225,7 +90,7 @@ nfs4_client_grab_unused(struct nfs4_client *clp, struct rpc_cred *cred) return sp; } -struct rpc_cred *nfs4_get_renew_cred(struct nfs4_client *clp) +struct rpc_cred *nfs4_get_renew_cred(struct nfs_client *clp) { struct nfs4_state_owner *sp; struct rpc_cred *cred = NULL; @@ -239,7 +104,7 @@ struct rpc_cred *nfs4_get_renew_cred(struct nfs4_client *clp) return cred; } -struct rpc_cred *nfs4_get_setclientid_cred(struct nfs4_client *clp) +struct rpc_cred *nfs4_get_setclientid_cred(struct nfs_client *clp) { struct nfs4_state_owner *sp; @@ -252,7 +117,7 @@ struct rpc_cred *nfs4_get_setclientid_cred(struct nfs4_client *clp) } static struct nfs4_state_owner * -nfs4_find_state_owner(struct nfs4_client *clp, struct rpc_cred *cred) +nfs4_find_state_owner(struct nfs_client *clp, struct rpc_cred *cred) { struct nfs4_state_owner *sp, *res = NULL; @@ -295,7 +160,7 @@ nfs4_alloc_state_owner(void) void nfs4_drop_state_owner(struct nfs4_state_owner *sp) { - struct nfs4_client *clp = sp->so_client; + struct nfs_client *clp = sp->so_client; spin_lock(&clp->cl_lock); list_del_init(&sp->so_list); spin_unlock(&clp->cl_lock); @@ -307,7 +172,7 @@ nfs4_drop_state_owner(struct nfs4_state_owner *sp) */ struct nfs4_state_owner *nfs4_get_state_owner(struct nfs_server *server, struct rpc_cred *cred) { - struct nfs4_client *clp = server->nfs4_state; + struct nfs_client *clp = server->nfs_client; struct nfs4_state_owner *sp, *new; get_rpccred(cred); @@ -338,7 +203,7 @@ struct nfs4_state_owner *nfs4_get_state_owner(struct nfs_server *server, struct */ void nfs4_put_state_owner(struct nfs4_state_owner *sp) { - struct nfs4_client *clp = sp->so_client; + struct nfs_client *clp = sp->so_client; struct rpc_cred *cred = sp->so_cred; if (!atomic_dec_and_lock(&sp->so_count, &clp->cl_lock)) @@ -541,7 +406,7 @@ __nfs4_find_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) { struct nfs4_lock_state *lsp; - struct nfs4_client *clp = state->owner->so_client; + struct nfs_client *clp = state->owner->so_client; lsp = kzalloc(sizeof(*lsp), GFP_KERNEL); if (lsp == NULL) @@ -753,7 +618,7 @@ out: static int reclaimer(void *); -static inline void nfs4_clear_recover_bit(struct nfs4_client *clp) +static inline void nfs4_clear_recover_bit(struct nfs_client *clp) { smp_mb__before_clear_bit(); clear_bit(NFS4CLNT_STATE_RECOVER, &clp->cl_state); @@ -765,25 +630,25 @@ static inline void nfs4_clear_recover_bit(struct nfs4_client *clp) /* * State recovery routine */ -static void nfs4_recover_state(struct nfs4_client *clp) +static void nfs4_recover_state(struct nfs_client *clp) { struct task_struct *task; __module_get(THIS_MODULE); atomic_inc(&clp->cl_count); task = kthread_run(reclaimer, clp, "%u.%u.%u.%u-reclaim", - NIPQUAD(clp->cl_addr)); + NIPQUAD(clp->cl_addr.sin_addr)); if (!IS_ERR(task)) return; nfs4_clear_recover_bit(clp); - nfs4_put_client(clp); + nfs_put_client(clp); module_put(THIS_MODULE); } /* * Schedule a state recovery attempt */ -void nfs4_schedule_state_recovery(struct nfs4_client *clp) +void nfs4_schedule_state_recovery(struct nfs_client *clp) { if (!clp) return; @@ -880,7 +745,7 @@ out_err: return status; } -static void nfs4_state_mark_reclaim(struct nfs4_client *clp) +static void nfs4_state_mark_reclaim(struct nfs_client *clp) { struct nfs4_state_owner *sp; struct nfs4_state *state; @@ -904,7 +769,7 @@ static void nfs4_state_mark_reclaim(struct nfs4_client *clp) static int reclaimer(void *ptr) { - struct nfs4_client *clp = ptr; + struct nfs_client *clp = ptr; struct nfs4_state_owner *sp; struct nfs4_state_recovery_ops *ops; struct rpc_cred *cred; @@ -971,12 +836,12 @@ out: if (status == -NFS4ERR_CB_PATH_DOWN) nfs_handle_cb_pathdown(clp); nfs4_clear_recover_bit(clp); - nfs4_put_client(clp); + nfs_put_client(clp); module_put_and_exit(0); return 0; out_error: printk(KERN_WARNING "Error: state recovery failed on NFSv4 server %u.%u.%u.%u with error %d\n", - NIPQUAD(clp->cl_addr.s_addr), -status); + NIPQUAD(clp->cl_addr.sin_addr), -status); set_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state); goto out; } diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index 7c5d70efe72..3dd413f52da 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -58,7 +58,7 @@ /* Mapping from NFS error code to "errno" error code. */ #define errno_NFSERR_IO EIO -static int nfs_stat_to_errno(int); +static int nfs4_stat_to_errno(int); /* NFSv4 COMPOUND tags are only wanted for debugging purposes */ #ifdef DEBUG @@ -128,7 +128,7 @@ static int nfs_stat_to_errno(int); #define decode_link_maxsz (op_decode_hdr_maxsz + 5) #define encode_symlink_maxsz (op_encode_hdr_maxsz + \ 1 + nfs4_name_maxsz + \ - nfs4_path_maxsz + \ + 1 + \ nfs4_fattr_maxsz) #define decode_symlink_maxsz (op_decode_hdr_maxsz + 8) #define encode_create_maxsz (op_encode_hdr_maxsz + \ @@ -411,6 +411,15 @@ static int nfs_stat_to_errno(int); #define NFS4_dec_setacl_sz (compound_decode_hdr_maxsz + \ decode_putfh_maxsz + \ op_decode_hdr_maxsz + nfs4_fattr_bitmap_maxsz) +#define NFS4_enc_fs_locations_sz \ + (compound_encode_hdr_maxsz + \ + encode_putfh_maxsz + \ + encode_getattr_maxsz) +#define NFS4_dec_fs_locations_sz \ + (compound_decode_hdr_maxsz + \ + decode_putfh_maxsz + \ + op_decode_hdr_maxsz + \ + nfs4_fattr_bitmap_maxsz) static struct { unsigned int mode; @@ -520,7 +529,7 @@ static int encode_attrs(struct xdr_stream *xdr, const struct iattr *iap, const s if (iap->ia_valid & ATTR_MODE) len += 4; if (iap->ia_valid & ATTR_UID) { - owner_namelen = nfs_map_uid_to_name(server->nfs4_state, iap->ia_uid, owner_name); + owner_namelen = nfs_map_uid_to_name(server->nfs_client, iap->ia_uid, owner_name); if (owner_namelen < 0) { printk(KERN_WARNING "nfs: couldn't resolve uid %d to string\n", iap->ia_uid); @@ -532,7 +541,7 @@ static int encode_attrs(struct xdr_stream *xdr, const struct iattr *iap, const s len += 4 + (XDR_QUADLEN(owner_namelen) << 2); } if (iap->ia_valid & ATTR_GID) { - owner_grouplen = nfs_map_gid_to_group(server->nfs4_state, iap->ia_gid, owner_group); + owner_grouplen = nfs_map_gid_to_group(server->nfs_client, iap->ia_gid, owner_group); if (owner_grouplen < 0) { printk(KERN_WARNING "nfs4: couldn't resolve gid %d to string\n", iap->ia_gid); @@ -664,9 +673,9 @@ static int encode_create(struct xdr_stream *xdr, const struct nfs4_create_arg *c switch (create->ftype) { case NF4LNK: - RESERVE_SPACE(4 + create->u.symlink->len); - WRITE32(create->u.symlink->len); - WRITEMEM(create->u.symlink->name, create->u.symlink->len); + RESERVE_SPACE(4); + WRITE32(create->u.symlink.len); + xdr_write_pages(xdr, create->u.symlink.pages, 0, create->u.symlink.len); break; case NF4BLK: case NF4CHR: @@ -722,6 +731,13 @@ static int encode_fsinfo(struct xdr_stream *xdr, const u32* bitmask) bitmask[1] & nfs4_fsinfo_bitmap[1]); } +static int encode_fs_locations(struct xdr_stream *xdr, const u32* bitmask) +{ + return encode_getattr_two(xdr, + bitmask[0] & nfs4_fs_locations_bitmap[0], + bitmask[1] & nfs4_fs_locations_bitmap[1]); +} + static int encode_getfh(struct xdr_stream *xdr) { uint32_t *p; @@ -1144,7 +1160,7 @@ static int encode_rename(struct xdr_stream *xdr, const struct qstr *oldname, con return 0; } -static int encode_renew(struct xdr_stream *xdr, const struct nfs4_client *client_stateid) +static int encode_renew(struct xdr_stream *xdr, const struct nfs_client *client_stateid) { uint32_t *p; @@ -1230,7 +1246,7 @@ static int encode_setclientid(struct xdr_stream *xdr, const struct nfs4_setclien return 0; } -static int encode_setclientid_confirm(struct xdr_stream *xdr, const struct nfs4_client *client_state) +static int encode_setclientid_confirm(struct xdr_stream *xdr, const struct nfs_client *client_state) { uint32_t *p; @@ -1929,7 +1945,7 @@ static int nfs4_xdr_enc_server_caps(struct rpc_rqst *req, uint32_t *p, const str /* * a RENEW request */ -static int nfs4_xdr_enc_renew(struct rpc_rqst *req, uint32_t *p, struct nfs4_client *clp) +static int nfs4_xdr_enc_renew(struct rpc_rqst *req, uint32_t *p, struct nfs_client *clp) { struct xdr_stream xdr; struct compound_hdr hdr = { @@ -1959,7 +1975,7 @@ static int nfs4_xdr_enc_setclientid(struct rpc_rqst *req, uint32_t *p, struct nf /* * a SETCLIENTID_CONFIRM request */ -static int nfs4_xdr_enc_setclientid_confirm(struct rpc_rqst *req, uint32_t *p, struct nfs4_client *clp) +static int nfs4_xdr_enc_setclientid_confirm(struct rpc_rqst *req, uint32_t *p, struct nfs_client *clp) { struct xdr_stream xdr; struct compound_hdr hdr = { @@ -2003,6 +2019,38 @@ out: } /* + * Encode FS_LOCATIONS request + */ +static int nfs4_xdr_enc_fs_locations(struct rpc_rqst *req, uint32_t *p, struct nfs4_fs_locations_arg *args) +{ + struct xdr_stream xdr; + struct compound_hdr hdr = { + .nops = 3, + }; + struct rpc_auth *auth = req->rq_task->tk_auth; + int replen; + int status; + + xdr_init_encode(&xdr, &req->rq_snd_buf, p); + encode_compound_hdr(&xdr, &hdr); + if ((status = encode_putfh(&xdr, args->dir_fh)) != 0) + goto out; + if ((status = encode_lookup(&xdr, args->name)) != 0) + goto out; + if ((status = encode_fs_locations(&xdr, args->bitmask)) != 0) + goto out; + /* set up reply + * toplevel_status + OP_PUTFH + status + * + OP_LOOKUP + status + OP_GETATTR + status = 7 + */ + replen = (RPC_REPHDRSIZE + auth->au_rslack + 7) << 2; + xdr_inline_pages(&req->rq_rcv_buf, replen, &args->page, + 0, PAGE_SIZE); +out: + return status; +} + +/* * START OF "GENERIC" DECODE ROUTINES. * These may look a little ugly since they are imported from a "generic" * set of XDR encode/decode routines which are intended to be shared by @@ -2036,7 +2084,7 @@ out: } \ } while (0) -static int decode_opaque_inline(struct xdr_stream *xdr, uint32_t *len, char **string) +static int decode_opaque_inline(struct xdr_stream *xdr, unsigned int *len, char **string) { uint32_t *p; @@ -2079,15 +2127,15 @@ static int decode_op_hdr(struct xdr_stream *xdr, enum nfs_opnum4 expected) } READ32(nfserr); if (nfserr != NFS_OK) - return -nfs_stat_to_errno(nfserr); + return -nfs4_stat_to_errno(nfserr); return 0; } /* Dummy routine */ -static int decode_ace(struct xdr_stream *xdr, void *ace, struct nfs4_client *clp) +static int decode_ace(struct xdr_stream *xdr, void *ace, struct nfs_client *clp) { uint32_t *p; - uint32_t strlen; + unsigned int strlen; char *str; READ_BUF(12); @@ -2217,7 +2265,7 @@ static int decode_attr_symlink_support(struct xdr_stream *xdr, uint32_t *bitmap, return 0; } -static int decode_attr_fsid(struct xdr_stream *xdr, uint32_t *bitmap, struct nfs4_fsid *fsid) +static int decode_attr_fsid(struct xdr_stream *xdr, uint32_t *bitmap, struct nfs_fsid *fsid) { uint32_t *p; @@ -2285,6 +2333,22 @@ static int decode_attr_fileid(struct xdr_stream *xdr, uint32_t *bitmap, uint64_t return 0; } +static int decode_attr_mounted_on_fileid(struct xdr_stream *xdr, uint32_t *bitmap, uint64_t *fileid) +{ + uint32_t *p; + + *fileid = 0; + if (unlikely(bitmap[1] & (FATTR4_WORD1_MOUNTED_ON_FILEID - 1U))) + return -EIO; + if (likely(bitmap[1] & FATTR4_WORD1_MOUNTED_ON_FILEID)) { + READ_BUF(8); + READ64(*fileid); + bitmap[1] &= ~FATTR4_WORD1_MOUNTED_ON_FILEID; + } + dprintk("%s: fileid=%Lu\n", __FUNCTION__, (unsigned long long)*fileid); + return 0; +} + static int decode_attr_files_avail(struct xdr_stream *xdr, uint32_t *bitmap, uint64_t *res) { uint32_t *p; @@ -2336,6 +2400,116 @@ static int decode_attr_files_total(struct xdr_stream *xdr, uint32_t *bitmap, uin return status; } +static int decode_pathname(struct xdr_stream *xdr, struct nfs4_pathname *path) +{ + int n; + uint32_t *p; + int status = 0; + + READ_BUF(4); + READ32(n); + if (n < 0) + goto out_eio; + if (n == 0) + goto root_path; + dprintk("path "); + path->ncomponents = 0; + while (path->ncomponents < n) { + struct nfs4_string *component = &path->components[path->ncomponents]; + status = decode_opaque_inline(xdr, &component->len, &component->data); + if (unlikely(status != 0)) + goto out_eio; + if (path->ncomponents != n) + dprintk("/"); + dprintk("%s", component->data); + if (path->ncomponents < NFS4_PATHNAME_MAXCOMPONENTS) + path->ncomponents++; + else { + dprintk("cannot parse %d components in path\n", n); + goto out_eio; + } + } +out: + dprintk("\n"); + return status; +root_path: +/* a root pathname is sent as a zero component4 */ + path->ncomponents = 1; + path->components[0].len=0; + path->components[0].data=NULL; + dprintk("path /\n"); + goto out; +out_eio: + dprintk(" status %d", status); + status = -EIO; + goto out; +} + +static int decode_attr_fs_locations(struct xdr_stream *xdr, uint32_t *bitmap, struct nfs4_fs_locations *res) +{ + int n; + uint32_t *p; + int status = -EIO; + + if (unlikely(bitmap[0] & (FATTR4_WORD0_FS_LOCATIONS -1U))) + goto out; + status = 0; + if (unlikely(!(bitmap[0] & FATTR4_WORD0_FS_LOCATIONS))) + goto out; + dprintk("%s: fsroot ", __FUNCTION__); + status = decode_pathname(xdr, &res->fs_path); + if (unlikely(status != 0)) + goto out; + READ_BUF(4); + READ32(n); + if (n <= 0) + goto out_eio; + res->nlocations = 0; + while (res->nlocations < n) { + int m; + struct nfs4_fs_location *loc = &res->locations[res->nlocations]; + + READ_BUF(4); + READ32(m); + if (m <= 0) + goto out_eio; + + loc->nservers = 0; + dprintk("%s: servers ", __FUNCTION__); + while (loc->nservers < m) { + struct nfs4_string *server = &loc->servers[loc->nservers]; + status = decode_opaque_inline(xdr, &server->len, &server->data); + if (unlikely(status != 0)) + goto out_eio; + dprintk("%s ", server->data); + if (loc->nservers < NFS4_FS_LOCATION_MAXSERVERS) + loc->nservers++; + else { + int i; + dprintk("%s: using first %d of %d servers returned for location %d\n", __FUNCTION__, NFS4_FS_LOCATION_MAXSERVERS, m, res->nlocations); + for (i = loc->nservers; i < m; i++) { + int len; + char *data; + status = decode_opaque_inline(xdr, &len, &data); + if (unlikely(status != 0)) + goto out_eio; + } + } + } + status = decode_pathname(xdr, &loc->rootpath); + if (unlikely(status != 0)) + goto out_eio; + if (res->nlocations < NFS4_FS_LOCATIONS_MAXENTRIES) + res->nlocations++; + } +out: + dprintk("%s: fs_locations done, error = %d\n", __FUNCTION__, status); + return status; +out_eio: + status = -EIO; + goto out; +} + static int decode_attr_maxfilesize(struct xdr_stream *xdr, uint32_t *bitmap, uint64_t *res) { uint32_t *p; @@ -2462,7 +2636,7 @@ static int decode_attr_nlink(struct xdr_stream *xdr, uint32_t *bitmap, uint32_t return 0; } -static int decode_attr_owner(struct xdr_stream *xdr, uint32_t *bitmap, struct nfs4_client *clp, int32_t *uid) +static int decode_attr_owner(struct xdr_stream *xdr, uint32_t *bitmap, struct nfs_client *clp, int32_t *uid) { uint32_t len, *p; @@ -2486,7 +2660,7 @@ static int decode_attr_owner(struct xdr_stream *xdr, uint32_t *bitmap, struct nf return 0; } -static int decode_attr_group(struct xdr_stream *xdr, uint32_t *bitmap, struct nfs4_client *clp, int32_t *gid) +static int decode_attr_group(struct xdr_stream *xdr, uint32_t *bitmap, struct nfs_client *clp, int32_t *gid) { uint32_t len, *p; @@ -2841,6 +3015,7 @@ static int decode_getfattr(struct xdr_stream *xdr, struct nfs_fattr *fattr, cons bitmap[2] = {0}, type; int status, fmode = 0; + uint64_t fileid; if ((status = decode_op_hdr(xdr, OP_GETATTR)) != 0) goto xdr_error; @@ -2863,18 +3038,22 @@ static int decode_getfattr(struct xdr_stream *xdr, struct nfs_fattr *fattr, cons goto xdr_error; if ((status = decode_attr_size(xdr, bitmap, &fattr->size)) != 0) goto xdr_error; - if ((status = decode_attr_fsid(xdr, bitmap, &fattr->fsid_u.nfs4)) != 0) + if ((status = decode_attr_fsid(xdr, bitmap, &fattr->fsid)) != 0) goto xdr_error; if ((status = decode_attr_fileid(xdr, bitmap, &fattr->fileid)) != 0) goto xdr_error; + if ((status = decode_attr_fs_locations(xdr, bitmap, container_of(fattr, + struct nfs4_fs_locations, + fattr))) != 0) + goto xdr_error; if ((status = decode_attr_mode(xdr, bitmap, &fattr->mode)) != 0) goto xdr_error; fattr->mode |= fmode; if ((status = decode_attr_nlink(xdr, bitmap, &fattr->nlink)) != 0) goto xdr_error; - if ((status = decode_attr_owner(xdr, bitmap, server->nfs4_state, &fattr->uid)) != 0) + if ((status = decode_attr_owner(xdr, bitmap, server->nfs_client, &fattr->uid)) != 0) goto xdr_error; - if ((status = decode_attr_group(xdr, bitmap, server->nfs4_state, &fattr->gid)) != 0) + if ((status = decode_attr_group(xdr, bitmap, server->nfs_client, &fattr->gid)) != 0) goto xdr_error; if ((status = decode_attr_rdev(xdr, bitmap, &fattr->rdev)) != 0) goto xdr_error; @@ -2886,6 +3065,10 @@ static int decode_getfattr(struct xdr_stream *xdr, struct nfs_fattr *fattr, cons goto xdr_error; if ((status = decode_attr_time_modify(xdr, bitmap, &fattr->mtime)) != 0) goto xdr_error; + if ((status = decode_attr_mounted_on_fileid(xdr, bitmap, &fileid)) != 0) + goto xdr_error; + if (fattr->fileid == 0 && fileid != 0) + fattr->fileid = fileid; if ((status = verify_attr_len(xdr, savep, attrlen)) == 0) fattr->valid = NFS_ATTR_FATTR | NFS_ATTR_FATTR_V3 | NFS_ATTR_FATTR_V4; xdr_error: @@ -3071,7 +3254,7 @@ static int decode_delegation(struct xdr_stream *xdr, struct nfs_openres *res) if (decode_space_limit(xdr, &res->maxsize) < 0) return -EIO; } - return decode_ace(xdr, NULL, res->server->nfs4_state); + return decode_ace(xdr, NULL, res->server->nfs_client); } static int decode_open(struct xdr_stream *xdr, struct nfs_openres *res) @@ -3172,7 +3355,7 @@ static int decode_readdir(struct xdr_stream *xdr, struct rpc_rqst *req, struct n struct kvec *iov = rcvbuf->head; unsigned int nr, pglen = rcvbuf->page_len; uint32_t *end, *entry, *p, *kaddr; - uint32_t len, attrlen; + uint32_t len, attrlen, xlen; int hdrlen, recvd, status; status = decode_op_hdr(xdr, OP_READDIR); @@ -3194,10 +3377,10 @@ static int decode_readdir(struct xdr_stream *xdr, struct rpc_rqst *req, struct n BUG_ON(pglen + readdir->pgbase > PAGE_CACHE_SIZE); kaddr = p = (uint32_t *) kmap_atomic(page, KM_USER0); - end = (uint32_t *) ((char *)p + pglen + readdir->pgbase); + end = p + ((pglen + readdir->pgbase) >> 2); entry = p; for (nr = 0; *p++; nr++) { - if (p + 3 > end) + if (end - p < 3) goto short_pkt; dprintk("cookie = %Lu, ", *((unsigned long long *)p)); p += 2; /* cookie */ @@ -3206,18 +3389,19 @@ static int decode_readdir(struct xdr_stream *xdr, struct rpc_rqst *req, struct n printk(KERN_WARNING "NFS: giant filename in readdir (len 0x%x)\n", len); goto err_unmap; } - dprintk("filename = %*s\n", len, (char *)p); - p += XDR_QUADLEN(len); - if (p + 1 > end) + xlen = XDR_QUADLEN(len); + if (end - p < xlen + 1) goto short_pkt; + dprintk("filename = %*s\n", len, (char *)p); + p += xlen; len = ntohl(*p++); /* bitmap length */ - p += len; - if (p + 1 > end) + if (end - p < len + 1) goto short_pkt; + p += len; attrlen = XDR_QUADLEN(ntohl(*p++)); - p += attrlen; /* attributes */ - if (p + 2 > end) + if (end - p < attrlen + 2) goto short_pkt; + p += attrlen; /* attributes */ entry = p; } if (!nr && (entry[0] != 0 || entry[1] == 0)) @@ -3350,8 +3534,7 @@ static int decode_getacl(struct xdr_stream *xdr, struct rpc_rqst *req, attrlen, recvd); return -EINVAL; } - if (attrlen <= *acl_len) - xdr_read_pages(xdr, attrlen); + xdr_read_pages(xdr, attrlen); *acl_len = attrlen; } else status = -EOPNOTSUPP; @@ -3382,7 +3565,7 @@ static int decode_setattr(struct xdr_stream *xdr, struct nfs_setattrres *res) return 0; } -static int decode_setclientid(struct xdr_stream *xdr, struct nfs4_client *clp) +static int decode_setclientid(struct xdr_stream *xdr, struct nfs_client *clp) { uint32_t *p; uint32_t opnum; @@ -3415,7 +3598,7 @@ static int decode_setclientid(struct xdr_stream *xdr, struct nfs4_client *clp) READ_BUF(len); return -NFSERR_CLID_INUSE; } else - return -nfs_stat_to_errno(nfserr); + return -nfs4_stat_to_errno(nfserr); return 0; } @@ -4073,7 +4256,7 @@ static int nfs4_xdr_dec_fsinfo(struct rpc_rqst *req, uint32_t *p, struct nfs_fsi if (!status) status = decode_fsinfo(&xdr, fsinfo); if (!status) - status = -nfs_stat_to_errno(hdr.status); + status = -nfs4_stat_to_errno(hdr.status); return status; } @@ -4152,7 +4335,7 @@ static int nfs4_xdr_dec_renew(struct rpc_rqst *rqstp, uint32_t *p, void *dummy) * a SETCLIENTID request */ static int nfs4_xdr_dec_setclientid(struct rpc_rqst *req, uint32_t *p, - struct nfs4_client *clp) + struct nfs_client *clp) { struct xdr_stream xdr; struct compound_hdr hdr; @@ -4163,7 +4346,7 @@ static int nfs4_xdr_dec_setclientid(struct rpc_rqst *req, uint32_t *p, if (!status) status = decode_setclientid(&xdr, clp); if (!status) - status = -nfs_stat_to_errno(hdr.status); + status = -nfs4_stat_to_errno(hdr.status); return status; } @@ -4185,7 +4368,7 @@ static int nfs4_xdr_dec_setclientid_confirm(struct rpc_rqst *req, uint32_t *p, s if (!status) status = decode_fsinfo(&xdr, fsinfo); if (!status) - status = -nfs_stat_to_errno(hdr.status); + status = -nfs4_stat_to_errno(hdr.status); return status; } @@ -4211,6 +4394,29 @@ out: return status; } +/* + * FS_LOCATIONS request + */ +static int nfs4_xdr_dec_fs_locations(struct rpc_rqst *req, uint32_t *p, struct nfs4_fs_locations *res) +{ + struct xdr_stream xdr; + struct compound_hdr hdr; + int status; + + xdr_init_decode(&xdr, &req->rq_rcv_buf, p); + status = decode_compound_hdr(&xdr, &hdr); + if (status != 0) + goto out; + if ((status = decode_putfh(&xdr)) != 0) + goto out; + if ((status = decode_lookup(&xdr)) != 0) + goto out; + xdr_enter_page(&xdr, PAGE_SIZE); + status = decode_getfattr(&xdr, &res->fattr, res->server); +out: + return status; +} + uint32_t *nfs4_decode_dirent(uint32_t *p, struct nfs_entry *entry, int plus) { uint32_t bitmap[2] = {0}; @@ -4315,7 +4521,7 @@ static struct { * This one is used jointly by NFSv2 and NFSv3. */ static int -nfs_stat_to_errno(int stat) +nfs4_stat_to_errno(int stat) { int i; for (i = 0; nfs_errtbl[i].stat != -1; i++) { @@ -4382,6 +4588,7 @@ struct rpc_procinfo nfs4_procedures[] = { PROC(DELEGRETURN, enc_delegreturn, dec_delegreturn), PROC(GETACL, enc_getacl, dec_getacl), PROC(SETACL, enc_setacl, dec_setacl), + PROC(FS_LOCATIONS, enc_fs_locations, dec_fs_locations), }; struct rpc_version nfs_version4 = { diff --git a/fs/nfs/pagelist.c b/fs/nfs/pagelist.c index 106aca388eb..829af323f28 100644 --- a/fs/nfs/pagelist.c +++ b/fs/nfs/pagelist.c @@ -9,7 +9,6 @@ * */ -#include <linux/config.h> #include <linux/slab.h> #include <linux/file.h> #include <linux/sunrpc/clnt.h> @@ -315,6 +314,7 @@ nfs_scan_lock_dirty(struct nfs_inode *nfsi, struct list_head *dst, req->wb_index, NFS_PAGE_TAG_DIRTY); nfs_list_remove_request(req); nfs_list_add_request(req, dst); + dec_zone_page_state(req->wb_page, NR_FILE_DIRTY); res++; } } @@ -325,6 +325,7 @@ out: /** * nfs_scan_list - Scan a list for matching requests + * @nfsi: NFS inode * @head: One of the NFS inode request lists * @dst: Destination list * @idx_start: lower bound of page->index to scan @@ -336,14 +337,15 @@ out: * The requests are *not* checked to ensure that they form a contiguous set. * You must be holding the inode's req_lock when calling this function */ -int -nfs_scan_list(struct list_head *head, struct list_head *dst, - unsigned long idx_start, unsigned int npages) +int nfs_scan_list(struct nfs_inode *nfsi, struct list_head *head, + struct list_head *dst, unsigned long idx_start, + unsigned int npages) { - struct list_head *pos, *tmp; - struct nfs_page *req; - unsigned long idx_end; - int res; + struct nfs_page *pgvec[NFS_SCAN_MAXENTRIES]; + struct nfs_page *req; + unsigned long idx_end; + int found, i; + int res; res = 0; if (npages == 0) @@ -351,25 +353,32 @@ nfs_scan_list(struct list_head *head, struct list_head *dst, else idx_end = idx_start + npages - 1; - list_for_each_safe(pos, tmp, head) { - - req = nfs_list_entry(pos); - - if (req->wb_index < idx_start) - continue; - if (req->wb_index > idx_end) + for (;;) { + found = radix_tree_gang_lookup(&nfsi->nfs_page_tree, + (void **)&pgvec[0], idx_start, + NFS_SCAN_MAXENTRIES); + if (found <= 0) break; + for (i = 0; i < found; i++) { + req = pgvec[i]; + if (req->wb_index > idx_end) + goto out; + idx_start = req->wb_index + 1; + if (req->wb_list_head != head) + continue; + if (nfs_set_page_writeback_locked(req)) { + nfs_list_remove_request(req); + nfs_list_add_request(req, dst); + res++; + } + } - if (!nfs_set_page_writeback_locked(req)) - continue; - nfs_list_remove_request(req); - nfs_list_add_request(req, dst); - res++; } +out: return res; } -int nfs_init_nfspagecache(void) +int __init nfs_init_nfspagecache(void) { nfs_page_cachep = kmem_cache_create("nfs_page", sizeof(struct nfs_page), @@ -383,7 +392,6 @@ int nfs_init_nfspagecache(void) void nfs_destroy_nfspagecache(void) { - if (kmem_cache_destroy(nfs_page_cachep)) - printk(KERN_INFO "nfs_page: not all structures were freed\n"); + kmem_cache_destroy(nfs_page_cachep); } diff --git a/fs/nfs/proc.c b/fs/nfs/proc.c index 9dd85cac2df..4529cc4f3f8 100644 --- a/fs/nfs/proc.c +++ b/fs/nfs/proc.c @@ -44,11 +44,10 @@ #include <linux/nfs_page.h> #include <linux/lockd/bind.h> #include <linux/smp_lock.h> +#include "internal.h" #define NFSDBG_FACILITY NFSDBG_PROC -extern struct rpc_procinfo nfs_procedures[]; - /* * Bare-bones access to getattr: this is for nfs_read_super. */ @@ -67,14 +66,14 @@ nfs_proc_get_root(struct nfs_server *server, struct nfs_fh *fhandle, dprintk("%s: call getattr\n", __FUNCTION__); nfs_fattr_init(fattr); - status = rpc_call_sync(server->client_sys, &msg, 0); + status = rpc_call_sync(server->nfs_client->cl_rpcclient, &msg, 0); dprintk("%s: reply getattr: %d\n", __FUNCTION__, status); if (status) return status; dprintk("%s: call statfs\n", __FUNCTION__); msg.rpc_proc = &nfs_procedures[NFSPROC_STATFS]; msg.rpc_resp = &fsinfo; - status = rpc_call_sync(server->client_sys, &msg, 0); + status = rpc_call_sync(server->nfs_client->cl_rpcclient, &msg, 0); dprintk("%s: reply statfs: %d\n", __FUNCTION__, status); if (status) return status; @@ -353,7 +352,7 @@ nfs_proc_unlink_setup(struct rpc_message *msg, struct dentry *dir, struct qstr * { struct nfs_diropargs *arg; - arg = (struct nfs_diropargs *)kmalloc(sizeof(*arg), GFP_KERNEL); + arg = kmalloc(sizeof(*arg), GFP_KERNEL); if (!arg) return -ENOMEM; arg->fh = NFS_FH(dir->d_inode); @@ -426,16 +425,17 @@ nfs_proc_link(struct inode *inode, struct inode *dir, struct qstr *name) } static int -nfs_proc_symlink(struct inode *dir, struct qstr *name, struct qstr *path, - struct iattr *sattr, struct nfs_fh *fhandle, - struct nfs_fattr *fattr) +nfs_proc_symlink(struct inode *dir, struct dentry *dentry, struct page *page, + unsigned int len, struct iattr *sattr) { + struct nfs_fh fhandle; + struct nfs_fattr fattr; struct nfs_symlinkargs arg = { .fromfh = NFS_FH(dir), - .fromname = name->name, - .fromlen = name->len, - .topath = path->name, - .tolen = path->len, + .fromname = dentry->d_name.name, + .fromlen = dentry->d_name.len, + .pages = &page, + .pathlen = len, .sattr = sattr }; struct rpc_message msg = { @@ -444,13 +444,25 @@ nfs_proc_symlink(struct inode *dir, struct qstr *name, struct qstr *path, }; int status; - if (path->len > NFS2_MAXPATHLEN) + if (len > NFS2_MAXPATHLEN) return -ENAMETOOLONG; - dprintk("NFS call symlink %s -> %s\n", name->name, path->name); - nfs_fattr_init(fattr); - fhandle->size = 0; + + dprintk("NFS call symlink %s\n", dentry->d_name.name); + status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0); nfs_mark_for_revalidate(dir); + + /* + * V2 SYMLINK requests don't return any attributes. Setting the + * filehandle size to zero indicates to nfs_instantiate that it + * should fill in the data with a LOOKUP call on the wire. + */ + if (status == 0) { + nfs_fattr_init(&fattr); + fhandle.size = 0; + status = nfs_instantiate(dentry, &fhandle, &fattr); + } + dprintk("NFS reply symlink: %d\n", status); return status; } @@ -611,8 +623,6 @@ nfs_proc_pathconf(struct nfs_server *server, struct nfs_fh *fhandle, return 0; } -extern u32 * nfs_decode_dirent(u32 *, struct nfs_entry *, int); - static int nfs_read_done(struct rpc_task *task, struct nfs_read_data *data) { if (task->tk_status >= 0) { @@ -674,7 +684,7 @@ nfs_proc_lock(struct file *filp, int cmd, struct file_lock *fl) } -struct nfs_rpc_ops nfs_v2_clientops = { +const struct nfs_rpc_ops nfs_v2_clientops = { .version = 2, /* protocol version */ .dentry_ops = &nfs_dentry_operations, .dir_inode_ops = &nfs_dir_inode_operations, diff --git a/fs/nfs/read.c b/fs/nfs/read.c index 624ca7146b6..c2e49c397a2 100644 --- a/fs/nfs/read.c +++ b/fs/nfs/read.c @@ -15,7 +15,6 @@ * within the RPC code when root squashing is suspected. */ -#include <linux/config.h> #include <linux/time.h> #include <linux/kernel.h> #include <linux/errno.h> @@ -44,21 +43,20 @@ static mempool_t *nfs_rdata_mempool; #define MIN_POOL_READ (32) -struct nfs_read_data *nfs_readdata_alloc(unsigned int pagecount) +struct nfs_read_data *nfs_readdata_alloc(size_t len) { + unsigned int pagecount = (len + PAGE_SIZE - 1) >> PAGE_SHIFT; struct nfs_read_data *p = mempool_alloc(nfs_rdata_mempool, SLAB_NOFS); if (p) { memset(p, 0, sizeof(*p)); INIT_LIST_HEAD(&p->pages); - if (pagecount < NFS_PAGEVEC_SIZE) - p->pagevec = &p->page_array[0]; + p->npages = pagecount; + if (pagecount <= ARRAY_SIZE(p->page_array)) + p->pagevec = p->page_array; else { - size_t size = ++pagecount * sizeof(struct page *); - p->pagevec = kmalloc(size, GFP_NOFS); - if (p->pagevec) { - memset(p->pagevec, 0, size); - } else { + p->pagevec = kcalloc(pagecount, sizeof(struct page *), GFP_NOFS); + if (!p->pagevec) { mempool_free(p, nfs_rdata_mempool); p = NULL; } @@ -67,7 +65,7 @@ struct nfs_read_data *nfs_readdata_alloc(unsigned int pagecount) return p; } -void nfs_readdata_free(struct nfs_read_data *p) +static void nfs_readdata_free(struct nfs_read_data *p) { if (p && (p->pagevec != &p->page_array[0])) kfree(p->pagevec); @@ -104,6 +102,35 @@ int nfs_return_empty_page(struct page *page) return 0; } +static void nfs_readpage_truncate_uninitialised_page(struct nfs_read_data *data) +{ + unsigned int remainder = data->args.count - data->res.count; + unsigned int base = data->args.pgbase + data->res.count; + unsigned int pglen; + struct page **pages; + + if (data->res.eof == 0 || remainder == 0) + return; + /* + * Note: "remainder" can never be negative, since we check for + * this in the XDR code. + */ + pages = &data->args.pages[base >> PAGE_CACHE_SHIFT]; + base &= ~PAGE_CACHE_MASK; + pglen = PAGE_CACHE_SIZE - base; + for (;;) { + if (remainder <= pglen) { + memclear_highpage_flush(*pages, base, remainder); + break; + } + memclear_highpage_flush(*pages, base, pglen); + pages++; + remainder -= pglen; + pglen = PAGE_CACHE_SIZE; + base = 0; + } +} + /* * Read a page synchronously. */ @@ -115,7 +142,7 @@ static int nfs_readpage_sync(struct nfs_open_context *ctx, struct inode *inode, int result; struct nfs_read_data *rdata; - rdata = nfs_readdata_alloc(1); + rdata = nfs_readdata_alloc(count); if (!rdata) return -ENOMEM; @@ -144,7 +171,7 @@ static int nfs_readpage_sync(struct nfs_open_context *ctx, struct inode *inode, rdata->args.offset = page_offset(page) + rdata->args.pgbase; dprintk("NFS: nfs_proc_read(%s, (%s/%Ld), %Lu, %u)\n", - NFS_SERVER(inode)->hostname, + NFS_SERVER(inode)->nfs_client->cl_hostname, inode->i_sb->s_id, (long long)NFS_FILEID(inode), (unsigned long long)rdata->args.pgbase, @@ -177,11 +204,11 @@ static int nfs_readpage_sync(struct nfs_open_context *ctx, struct inode *inode, NFS_I(inode)->cache_validity |= NFS_INO_INVALID_ATIME; spin_unlock(&inode->i_lock); - if (count) - memclear_highpage_flush(page, rdata->args.pgbase, count); - SetPageUptodate(page); - if (PageError(page)) - ClearPageError(page); + if (rdata->res.eof || rdata->res.count == rdata->args.count) { + SetPageUptodate(page); + if (rdata->res.eof && count != 0) + memclear_highpage_flush(page, rdata->args.pgbase, count); + } result = 0; io_error: @@ -313,25 +340,25 @@ static int nfs_pagein_multi(struct list_head *head, struct inode *inode) struct nfs_page *req = nfs_list_entry(head->next); struct page *page = req->wb_page; struct nfs_read_data *data; - unsigned int rsize = NFS_SERVER(inode)->rsize; - unsigned int nbytes, offset; + size_t rsize = NFS_SERVER(inode)->rsize, nbytes; + unsigned int offset; int requests = 0; LIST_HEAD(list); nfs_list_remove_request(req); nbytes = req->wb_bytes; - for(;;) { - data = nfs_readdata_alloc(1); + do { + size_t len = min(nbytes,rsize); + + data = nfs_readdata_alloc(len); if (!data) goto out_bad; INIT_LIST_HEAD(&data->pages); list_add(&data->pages, &list); requests++; - if (nbytes <= rsize) - break; - nbytes -= rsize; - } + nbytes -= len; + } while(nbytes != 0); atomic_set(&req->wb_complete, requests); ClearPageError(page); @@ -379,7 +406,7 @@ static int nfs_pagein_one(struct list_head *head, struct inode *inode) if (NFS_SERVER(inode)->rsize < PAGE_CACHE_SIZE) return nfs_pagein_multi(head, inode); - data = nfs_readdata_alloc(NFS_SERVER(inode)->rpages); + data = nfs_readdata_alloc(NFS_SERVER(inode)->rsize); if (!data) goto out_bad; @@ -436,20 +463,12 @@ static void nfs_readpage_result_partial(struct rpc_task *task, void *calldata) struct nfs_page *req = data->req; struct page *page = req->wb_page; + if (likely(task->tk_status >= 0)) + nfs_readpage_truncate_uninitialised_page(data); + else + SetPageError(page); if (nfs_readpage_result(task, data) != 0) return; - if (task->tk_status >= 0) { - unsigned int request = data->args.count; - unsigned int result = data->res.count; - - if (result < request) { - memclear_highpage_flush(page, - data->args.pgbase + result, - request - result); - } - } else - SetPageError(page); - if (atomic_dec_and_test(&req->wb_complete)) { if (!PageError(page)) SetPageUptodate(page); @@ -462,6 +481,40 @@ static const struct rpc_call_ops nfs_read_partial_ops = { .rpc_release = nfs_readdata_release, }; +static void nfs_readpage_set_pages_uptodate(struct nfs_read_data *data) +{ + unsigned int count = data->res.count; + unsigned int base = data->args.pgbase; + struct page **pages; + + if (data->res.eof) + count = data->args.count; + if (unlikely(count == 0)) + return; + pages = &data->args.pages[base >> PAGE_CACHE_SHIFT]; + base &= ~PAGE_CACHE_MASK; + count += base; + for (;count >= PAGE_CACHE_SIZE; count -= PAGE_CACHE_SIZE, pages++) + SetPageUptodate(*pages); + if (count != 0) + SetPageUptodate(*pages); +} + +static void nfs_readpage_set_pages_error(struct nfs_read_data *data) +{ + unsigned int count = data->args.count; + unsigned int base = data->args.pgbase; + struct page **pages; + + pages = &data->args.pages[base >> PAGE_CACHE_SHIFT]; + base &= ~PAGE_CACHE_MASK; + count += base; + for (;count >= PAGE_CACHE_SIZE; count -= PAGE_CACHE_SIZE, pages++) + SetPageError(*pages); + if (count != 0) + SetPageError(*pages); +} + /* * This is the callback from RPC telling us whether a reply was * received or some error occurred (timeout or socket shutdown). @@ -469,27 +522,24 @@ static const struct rpc_call_ops nfs_read_partial_ops = { static void nfs_readpage_result_full(struct rpc_task *task, void *calldata) { struct nfs_read_data *data = calldata; - unsigned int count = data->res.count; + /* + * Note: nfs_readpage_result may change the values of + * data->args. In the multi-page case, we therefore need + * to ensure that we call the next nfs_readpage_set_page_uptodate() + * first in the multi-page case. + */ + if (likely(task->tk_status >= 0)) { + nfs_readpage_truncate_uninitialised_page(data); + nfs_readpage_set_pages_uptodate(data); + } else + nfs_readpage_set_pages_error(data); if (nfs_readpage_result(task, data) != 0) return; while (!list_empty(&data->pages)) { struct nfs_page *req = nfs_list_entry(data->pages.next); - struct page *page = req->wb_page; - nfs_list_remove_request(req); - if (task->tk_status >= 0) { - if (count < PAGE_CACHE_SIZE) { - if (count < req->wb_bytes) - memclear_highpage_flush(page, - req->wb_pgbase + count, - req->wb_bytes - count); - count = 0; - } else - count -= PAGE_CACHE_SIZE; - SetPageUptodate(page); - } else - SetPageError(page); + nfs_list_remove_request(req); nfs_readpage_release(req); } } @@ -518,8 +568,13 @@ int nfs_readpage_result(struct rpc_task *task, struct nfs_read_data *data) nfs_add_stats(data->inode, NFSIOS_SERVERREADBYTES, resp->count); - /* Is this a short read? */ - if (task->tk_status >= 0 && resp->count < argp->count && !resp->eof) { + if (task->tk_status < 0) { + if (task->tk_status == -ESTALE) { + set_bit(NFS_INO_STALE, &NFS_FLAGS(data->inode)); + nfs_mark_for_revalidate(data->inode); + } + } else if (resp->count < argp->count && !resp->eof) { + /* This is a short read! */ nfs_inc_stats(data->inode, NFSIOS_SHORTREAD); /* Has the server at least made some progress? */ if (resp->count != 0) { @@ -566,6 +621,10 @@ int nfs_readpage(struct file *file, struct page *page) if (error) goto out_error; + error = -ESTALE; + if (NFS_STALE(inode)) + goto out_error; + if (file == NULL) { ctx = nfs_find_open_context(inode, NULL, FMODE_READ); if (ctx == NULL) @@ -628,7 +687,7 @@ int nfs_readpages(struct file *filp, struct address_space *mapping, }; struct inode *inode = mapping->host; struct nfs_server *server = NFS_SERVER(inode); - int ret; + int ret = -ESTALE; dprintk("NFS: nfs_readpages (%s/%Ld %d)\n", inode->i_sb->s_id, @@ -636,6 +695,9 @@ int nfs_readpages(struct file *filp, struct address_space *mapping, nr_pages); nfs_inc_stats(inode, NFSIOS_VFSREADPAGES); + if (NFS_STALE(inode)) + goto out; + if (filp == NULL) { desc.ctx = nfs_find_open_context(inode, NULL, FMODE_READ); if (desc.ctx == NULL) @@ -651,10 +713,11 @@ int nfs_readpages(struct file *filp, struct address_space *mapping, ret = err; } put_nfs_open_context(desc.ctx); +out: return ret; } -int nfs_init_readpagecache(void) +int __init nfs_init_readpagecache(void) { nfs_rdata_cachep = kmem_cache_create("nfs_read_data", sizeof(struct nfs_read_data), @@ -674,6 +737,5 @@ int nfs_init_readpagecache(void) void nfs_destroy_readpagecache(void) { mempool_destroy(nfs_rdata_mempool); - if (kmem_cache_destroy(nfs_rdata_cachep)) - printk(KERN_INFO "nfs_read_data: not all structures were freed\n"); + kmem_cache_destroy(nfs_rdata_cachep); } diff --git a/fs/nfs/super.c b/fs/nfs/super.c new file mode 100644 index 00000000000..e8d40030cab --- /dev/null +++ b/fs/nfs/super.c @@ -0,0 +1,1074 @@ +/* + * linux/fs/nfs/super.c + * + * Copyright (C) 1992 Rick Sladkey + * + * nfs superblock handling functions + * + * Modularised by Alan Cox <Alan.Cox@linux.org>, while hacking some + * experimental NFS changes. Modularisation taken straight from SYS5 fs. + * + * Change to nfs_read_super() to permit NFS mounts to multi-homed hosts. + * J.S.Peatfield@damtp.cam.ac.uk + * + * Split from inode.c by David Howells <dhowells@redhat.com> + * + * - superblocks are indexed on server only - all inodes, dentries, etc. associated with a + * particular server are held in the same superblock + * - NFS superblocks can have several effective roots to the dentry tree + * - directory type roots are spliced into the tree when a path from one root reaches the root + * of another (see nfs_lookup()) + */ + +#include <linux/config.h> +#include <linux/module.h> +#include <linux/init.h> + +#include <linux/time.h> +#include <linux/kernel.h> +#include <linux/mm.h> +#include <linux/string.h> +#include <linux/stat.h> +#include <linux/errno.h> +#include <linux/unistd.h> +#include <linux/sunrpc/clnt.h> +#include <linux/sunrpc/stats.h> +#include <linux/sunrpc/metrics.h> +#include <linux/nfs_fs.h> +#include <linux/nfs_mount.h> +#include <linux/nfs4_mount.h> +#include <linux/lockd/bind.h> +#include <linux/smp_lock.h> +#include <linux/seq_file.h> +#include <linux/mount.h> +#include <linux/nfs_idmap.h> +#include <linux/vfs.h> +#include <linux/inet.h> +#include <linux/nfs_xdr.h> + +#include <asm/system.h> +#include <asm/uaccess.h> + +#include "nfs4_fs.h" +#include "callback.h" +#include "delegation.h" +#include "iostat.h" +#include "internal.h" + +#define NFSDBG_FACILITY NFSDBG_VFS + +static void nfs_umount_begin(struct vfsmount *, int); +static int nfs_statfs(struct dentry *, struct kstatfs *); +static int nfs_show_options(struct seq_file *, struct vfsmount *); +static int nfs_show_stats(struct seq_file *, struct vfsmount *); +static int nfs_get_sb(struct file_system_type *, int, const char *, void *, struct vfsmount *); +static int nfs_xdev_get_sb(struct file_system_type *fs_type, + int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt); +static void nfs_kill_super(struct super_block *); + +static struct file_system_type nfs_fs_type = { + .owner = THIS_MODULE, + .name = "nfs", + .get_sb = nfs_get_sb, + .kill_sb = nfs_kill_super, + .fs_flags = FS_RENAME_DOES_D_MOVE|FS_REVAL_DOT|FS_BINARY_MOUNTDATA, +}; + +struct file_system_type nfs_xdev_fs_type = { + .owner = THIS_MODULE, + .name = "nfs", + .get_sb = nfs_xdev_get_sb, + .kill_sb = nfs_kill_super, + .fs_flags = FS_RENAME_DOES_D_MOVE|FS_REVAL_DOT|FS_BINARY_MOUNTDATA, +}; + +static struct super_operations nfs_sops = { + .alloc_inode = nfs_alloc_inode, + .destroy_inode = nfs_destroy_inode, + .write_inode = nfs_write_inode, + .statfs = nfs_statfs, + .clear_inode = nfs_clear_inode, + .umount_begin = nfs_umount_begin, + .show_options = nfs_show_options, + .show_stats = nfs_show_stats, +}; + +#ifdef CONFIG_NFS_V4 +static int nfs4_get_sb(struct file_system_type *fs_type, + int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt); +static int nfs4_xdev_get_sb(struct file_system_type *fs_type, + int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt); +static int nfs4_referral_get_sb(struct file_system_type *fs_type, + int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt); +static void nfs4_kill_super(struct super_block *sb); + +static struct file_system_type nfs4_fs_type = { + .owner = THIS_MODULE, + .name = "nfs4", + .get_sb = nfs4_get_sb, + .kill_sb = nfs4_kill_super, + .fs_flags = FS_RENAME_DOES_D_MOVE|FS_REVAL_DOT|FS_BINARY_MOUNTDATA, +}; + +struct file_system_type nfs4_xdev_fs_type = { + .owner = THIS_MODULE, + .name = "nfs4", + .get_sb = nfs4_xdev_get_sb, + .kill_sb = nfs4_kill_super, + .fs_flags = FS_RENAME_DOES_D_MOVE|FS_REVAL_DOT|FS_BINARY_MOUNTDATA, +}; + +struct file_system_type nfs4_referral_fs_type = { + .owner = THIS_MODULE, + .name = "nfs4", + .get_sb = nfs4_referral_get_sb, + .kill_sb = nfs4_kill_super, + .fs_flags = FS_RENAME_DOES_D_MOVE|FS_REVAL_DOT|FS_BINARY_MOUNTDATA, +}; + +static struct super_operations nfs4_sops = { + .alloc_inode = nfs_alloc_inode, + .destroy_inode = nfs_destroy_inode, + .write_inode = nfs_write_inode, + .statfs = nfs_statfs, + .clear_inode = nfs4_clear_inode, + .umount_begin = nfs_umount_begin, + .show_options = nfs_show_options, + .show_stats = nfs_show_stats, +}; +#endif + +static struct shrinker *acl_shrinker; + +/* + * Register the NFS filesystems + */ +int __init register_nfs_fs(void) +{ + int ret; + + ret = register_filesystem(&nfs_fs_type); + if (ret < 0) + goto error_0; + +#ifdef CONFIG_NFS_V4 + ret = nfs_register_sysctl(); + if (ret < 0) + goto error_1; + ret = register_filesystem(&nfs4_fs_type); + if (ret < 0) + goto error_2; +#endif + acl_shrinker = set_shrinker(DEFAULT_SEEKS, nfs_access_cache_shrinker); + return 0; + +#ifdef CONFIG_NFS_V4 +error_2: + nfs_unregister_sysctl(); +error_1: + unregister_filesystem(&nfs_fs_type); +#endif +error_0: + return ret; +} + +/* + * Unregister the NFS filesystems + */ +void __exit unregister_nfs_fs(void) +{ + if (acl_shrinker != NULL) + remove_shrinker(acl_shrinker); +#ifdef CONFIG_NFS_V4 + unregister_filesystem(&nfs4_fs_type); + nfs_unregister_sysctl(); +#endif + unregister_filesystem(&nfs_fs_type); +} + +/* + * Deliver file system statistics to userspace + */ +static int nfs_statfs(struct dentry *dentry, struct kstatfs *buf) +{ + struct nfs_server *server = NFS_SB(dentry->d_sb); + unsigned char blockbits; + unsigned long blockres; + struct nfs_fh *fh = NFS_FH(dentry->d_inode); + struct nfs_fattr fattr; + struct nfs_fsstat res = { + .fattr = &fattr, + }; + int error; + + lock_kernel(); + + error = server->nfs_client->rpc_ops->statfs(server, fh, &res); + buf->f_type = NFS_SUPER_MAGIC; + if (error < 0) + goto out_err; + + /* + * Current versions of glibc do not correctly handle the + * case where f_frsize != f_bsize. Eventually we want to + * report the value of wtmult in this field. + */ + buf->f_frsize = dentry->d_sb->s_blocksize; + + /* + * On most *nix systems, f_blocks, f_bfree, and f_bavail + * are reported in units of f_frsize. Linux hasn't had + * an f_frsize field in its statfs struct until recently, + * thus historically Linux's sys_statfs reports these + * fields in units of f_bsize. + */ + buf->f_bsize = dentry->d_sb->s_blocksize; + blockbits = dentry->d_sb->s_blocksize_bits; + blockres = (1 << blockbits) - 1; + buf->f_blocks = (res.tbytes + blockres) >> blockbits; + buf->f_bfree = (res.fbytes + blockres) >> blockbits; + buf->f_bavail = (res.abytes + blockres) >> blockbits; + + buf->f_files = res.tfiles; + buf->f_ffree = res.afiles; + + buf->f_namelen = server->namelen; + out: + unlock_kernel(); + return 0; + + out_err: + dprintk("%s: statfs error = %d\n", __FUNCTION__, -error); + buf->f_bsize = buf->f_blocks = buf->f_bfree = buf->f_bavail = -1; + goto out; + +} + +/* + * Map the security flavour number to a name + */ +static const char *nfs_pseudoflavour_to_name(rpc_authflavor_t flavour) +{ + static const struct { + rpc_authflavor_t flavour; + const char *str; + } sec_flavours[] = { + { RPC_AUTH_NULL, "null" }, + { RPC_AUTH_UNIX, "sys" }, + { RPC_AUTH_GSS_KRB5, "krb5" }, + { RPC_AUTH_GSS_KRB5I, "krb5i" }, + { RPC_AUTH_GSS_KRB5P, "krb5p" }, + { RPC_AUTH_GSS_LKEY, "lkey" }, + { RPC_AUTH_GSS_LKEYI, "lkeyi" }, + { RPC_AUTH_GSS_LKEYP, "lkeyp" }, + { RPC_AUTH_GSS_SPKM, "spkm" }, + { RPC_AUTH_GSS_SPKMI, "spkmi" }, + { RPC_AUTH_GSS_SPKMP, "spkmp" }, + { -1, "unknown" } + }; + int i; + + for (i=0; sec_flavours[i].flavour != -1; i++) { + if (sec_flavours[i].flavour == flavour) + break; + } + return sec_flavours[i].str; +} + +/* + * Describe the mount options in force on this server representation + */ +static void nfs_show_mount_options(struct seq_file *m, struct nfs_server *nfss, int showdefaults) +{ + static const struct proc_nfs_info { + int flag; + const char *str; + const char *nostr; + } nfs_info[] = { + { NFS_MOUNT_SOFT, ",soft", ",hard" }, + { NFS_MOUNT_INTR, ",intr", "" }, + { NFS_MOUNT_NOCTO, ",nocto", "" }, + { NFS_MOUNT_NOAC, ",noac", "" }, + { NFS_MOUNT_NONLM, ",nolock", "" }, + { NFS_MOUNT_NOACL, ",noacl", "" }, + { 0, NULL, NULL } + }; + const struct proc_nfs_info *nfs_infop; + struct nfs_client *clp = nfss->nfs_client; + char buf[12]; + const char *proto; + + seq_printf(m, ",vers=%d", clp->rpc_ops->version); + seq_printf(m, ",rsize=%d", nfss->rsize); + seq_printf(m, ",wsize=%d", nfss->wsize); + if (nfss->acregmin != 3*HZ || showdefaults) + seq_printf(m, ",acregmin=%d", nfss->acregmin/HZ); + if (nfss->acregmax != 60*HZ || showdefaults) + seq_printf(m, ",acregmax=%d", nfss->acregmax/HZ); + if (nfss->acdirmin != 30*HZ || showdefaults) + seq_printf(m, ",acdirmin=%d", nfss->acdirmin/HZ); + if (nfss->acdirmax != 60*HZ || showdefaults) + seq_printf(m, ",acdirmax=%d", nfss->acdirmax/HZ); + for (nfs_infop = nfs_info; nfs_infop->flag; nfs_infop++) { + if (nfss->flags & nfs_infop->flag) + seq_puts(m, nfs_infop->str); + else + seq_puts(m, nfs_infop->nostr); + } + switch (nfss->client->cl_xprt->prot) { + case IPPROTO_TCP: + proto = "tcp"; + break; + case IPPROTO_UDP: + proto = "udp"; + break; + default: + snprintf(buf, sizeof(buf), "%u", nfss->client->cl_xprt->prot); + proto = buf; + } + seq_printf(m, ",proto=%s", proto); + seq_printf(m, ",timeo=%lu", 10U * clp->retrans_timeo / HZ); + seq_printf(m, ",retrans=%u", clp->retrans_count); + seq_printf(m, ",sec=%s", nfs_pseudoflavour_to_name(nfss->client->cl_auth->au_flavor)); +} + +/* + * Describe the mount options on this VFS mountpoint + */ +static int nfs_show_options(struct seq_file *m, struct vfsmount *mnt) +{ + struct nfs_server *nfss = NFS_SB(mnt->mnt_sb); + + nfs_show_mount_options(m, nfss, 0); + + seq_puts(m, ",addr="); + seq_escape(m, nfss->nfs_client->cl_hostname, " \t\n\\"); + + return 0; +} + +/* + * Present statistical information for this VFS mountpoint + */ +static int nfs_show_stats(struct seq_file *m, struct vfsmount *mnt) +{ + int i, cpu; + struct nfs_server *nfss = NFS_SB(mnt->mnt_sb); + struct rpc_auth *auth = nfss->client->cl_auth; + struct nfs_iostats totals = { }; + + seq_printf(m, "statvers=%s", NFS_IOSTAT_VERS); + + /* + * Display all mount option settings + */ + seq_printf(m, "\n\topts:\t"); + seq_puts(m, mnt->mnt_sb->s_flags & MS_RDONLY ? "ro" : "rw"); + seq_puts(m, mnt->mnt_sb->s_flags & MS_SYNCHRONOUS ? ",sync" : ""); + seq_puts(m, mnt->mnt_sb->s_flags & MS_NOATIME ? ",noatime" : ""); + seq_puts(m, mnt->mnt_sb->s_flags & MS_NODIRATIME ? ",nodiratime" : ""); + nfs_show_mount_options(m, nfss, 1); + + seq_printf(m, "\n\tage:\t%lu", (jiffies - nfss->mount_time) / HZ); + + seq_printf(m, "\n\tcaps:\t"); + seq_printf(m, "caps=0x%x", nfss->caps); + seq_printf(m, ",wtmult=%d", nfss->wtmult); + seq_printf(m, ",dtsize=%d", nfss->dtsize); + seq_printf(m, ",bsize=%d", nfss->bsize); + seq_printf(m, ",namelen=%d", nfss->namelen); + +#ifdef CONFIG_NFS_V4 + if (nfss->nfs_client->cl_nfsversion == 4) { + seq_printf(m, "\n\tnfsv4:\t"); + seq_printf(m, "bm0=0x%x", nfss->attr_bitmask[0]); + seq_printf(m, ",bm1=0x%x", nfss->attr_bitmask[1]); + seq_printf(m, ",acl=0x%x", nfss->acl_bitmask); + } +#endif + + /* + * Display security flavor in effect for this mount + */ + seq_printf(m, "\n\tsec:\tflavor=%d", auth->au_ops->au_flavor); + if (auth->au_flavor) + seq_printf(m, ",pseudoflavor=%d", auth->au_flavor); + + /* + * Display superblock I/O counters + */ + for_each_possible_cpu(cpu) { + struct nfs_iostats *stats; + + preempt_disable(); + stats = per_cpu_ptr(nfss->io_stats, cpu); + + for (i = 0; i < __NFSIOS_COUNTSMAX; i++) + totals.events[i] += stats->events[i]; + for (i = 0; i < __NFSIOS_BYTESMAX; i++) + totals.bytes[i] += stats->bytes[i]; + + preempt_enable(); + } + + seq_printf(m, "\n\tevents:\t"); + for (i = 0; i < __NFSIOS_COUNTSMAX; i++) + seq_printf(m, "%lu ", totals.events[i]); + seq_printf(m, "\n\tbytes:\t"); + for (i = 0; i < __NFSIOS_BYTESMAX; i++) + seq_printf(m, "%Lu ", totals.bytes[i]); + seq_printf(m, "\n"); + + rpc_print_iostats(m, nfss->client); + + return 0; +} + +/* + * Begin unmount by attempting to remove all automounted mountpoints we added + * in response to xdev traversals and referrals + */ +static void nfs_umount_begin(struct vfsmount *vfsmnt, int flags) +{ + shrink_submounts(vfsmnt, &nfs_automount_list); +} + +/* + * Validate the NFS2/NFS3 mount data + * - fills in the mount root filehandle + */ +static int nfs_validate_mount_data(struct nfs_mount_data *data, + struct nfs_fh *mntfh) +{ + if (data == NULL) { + dprintk("%s: missing data argument\n", __FUNCTION__); + return -EINVAL; + } + + if (data->version <= 0 || data->version > NFS_MOUNT_VERSION) { + dprintk("%s: bad mount version\n", __FUNCTION__); + return -EINVAL; + } + + switch (data->version) { + case 1: + data->namlen = 0; + case 2: + data->bsize = 0; + case 3: + if (data->flags & NFS_MOUNT_VER3) { + dprintk("%s: mount structure version %d does not support NFSv3\n", + __FUNCTION__, + data->version); + return -EINVAL; + } + data->root.size = NFS2_FHSIZE; + memcpy(data->root.data, data->old_root.data, NFS2_FHSIZE); + case 4: + if (data->flags & NFS_MOUNT_SECFLAVOUR) { + dprintk("%s: mount structure version %d does not support strong security\n", + __FUNCTION__, + data->version); + return -EINVAL; + } + case 5: + memset(data->context, 0, sizeof(data->context)); + } + + /* Set the pseudoflavor */ + if (!(data->flags & NFS_MOUNT_SECFLAVOUR)) + data->pseudoflavor = RPC_AUTH_UNIX; + +#ifndef CONFIG_NFS_V3 + /* If NFSv3 is not compiled in, return -EPROTONOSUPPORT */ + if (data->flags & NFS_MOUNT_VER3) { + dprintk("%s: NFSv3 not compiled into kernel\n", __FUNCTION__); + return -EPROTONOSUPPORT; + } +#endif /* CONFIG_NFS_V3 */ + + /* We now require that the mount process passes the remote address */ + if (data->addr.sin_addr.s_addr == INADDR_ANY) { + dprintk("%s: mount program didn't pass remote address!\n", + __FUNCTION__); + return -EINVAL; + } + + /* Prepare the root filehandle */ + if (data->flags & NFS_MOUNT_VER3) + mntfh->size = data->root.size; + else + mntfh->size = NFS2_FHSIZE; + + if (mntfh->size > sizeof(mntfh->data)) { + dprintk("%s: invalid root filehandle\n", __FUNCTION__); + return -EINVAL; + } + + memcpy(mntfh->data, data->root.data, mntfh->size); + if (mntfh->size < sizeof(mntfh->data)) + memset(mntfh->data + mntfh->size, 0, + sizeof(mntfh->data) - mntfh->size); + + return 0; +} + +/* + * Initialise the common bits of the superblock + */ +static inline void nfs_initialise_sb(struct super_block *sb) +{ + struct nfs_server *server = NFS_SB(sb); + + sb->s_magic = NFS_SUPER_MAGIC; + + /* We probably want something more informative here */ + snprintf(sb->s_id, sizeof(sb->s_id), + "%x:%x", MAJOR(sb->s_dev), MINOR(sb->s_dev)); + + if (sb->s_blocksize == 0) + sb->s_blocksize = nfs_block_bits(server->wsize, + &sb->s_blocksize_bits); + + if (server->flags & NFS_MOUNT_NOAC) + sb->s_flags |= MS_SYNCHRONOUS; + + nfs_super_set_maxbytes(sb, server->maxfilesize); +} + +/* + * Finish setting up an NFS2/3 superblock + */ +static void nfs_fill_super(struct super_block *sb, struct nfs_mount_data *data) +{ + struct nfs_server *server = NFS_SB(sb); + + sb->s_blocksize_bits = 0; + sb->s_blocksize = 0; + if (data->bsize) + sb->s_blocksize = nfs_block_size(data->bsize, &sb->s_blocksize_bits); + + if (server->flags & NFS_MOUNT_VER3) { + /* The VFS shouldn't apply the umask to mode bits. We will do + * so ourselves when necessary. + */ + sb->s_flags |= MS_POSIXACL; + sb->s_time_gran = 1; + } + + sb->s_op = &nfs_sops; + nfs_initialise_sb(sb); +} + +/* + * Finish setting up a cloned NFS2/3 superblock + */ +static void nfs_clone_super(struct super_block *sb, + const struct super_block *old_sb) +{ + struct nfs_server *server = NFS_SB(sb); + + sb->s_blocksize_bits = old_sb->s_blocksize_bits; + sb->s_blocksize = old_sb->s_blocksize; + sb->s_maxbytes = old_sb->s_maxbytes; + + if (server->flags & NFS_MOUNT_VER3) { + /* The VFS shouldn't apply the umask to mode bits. We will do + * so ourselves when necessary. + */ + sb->s_flags |= MS_POSIXACL; + sb->s_time_gran = 1; + } + + sb->s_op = old_sb->s_op; + nfs_initialise_sb(sb); +} + +static int nfs_set_super(struct super_block *s, void *_server) +{ + struct nfs_server *server = _server; + int ret; + + s->s_fs_info = server; + ret = set_anon_super(s, server); + if (ret == 0) + server->s_dev = s->s_dev; + return ret; +} + +static int nfs_compare_super(struct super_block *sb, void *data) +{ + struct nfs_server *server = data, *old = NFS_SB(sb); + + if (old->nfs_client != server->nfs_client) + return 0; + if (memcmp(&old->fsid, &server->fsid, sizeof(old->fsid)) != 0) + return 0; + return 1; +} + +static int nfs_get_sb(struct file_system_type *fs_type, + int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt) +{ + struct nfs_server *server = NULL; + struct super_block *s; + struct nfs_fh mntfh; + struct nfs_mount_data *data = raw_data; + struct dentry *mntroot; + int error; + + /* Validate the mount data */ + error = nfs_validate_mount_data(data, &mntfh); + if (error < 0) + return error; + + /* Get a volume representation */ + server = nfs_create_server(data, &mntfh); + if (IS_ERR(server)) { + error = PTR_ERR(server); + goto out_err_noserver; + } + + /* Get a superblock - note that we may end up sharing one that already exists */ + s = sget(fs_type, nfs_compare_super, nfs_set_super, server); + if (IS_ERR(s)) { + error = PTR_ERR(s); + goto out_err_nosb; + } + + if (s->s_fs_info != server) { + nfs_free_server(server); + server = NULL; + } + + if (!s->s_root) { + /* initial superblock/root creation */ + s->s_flags = flags; + nfs_fill_super(s, data); + } + + mntroot = nfs_get_root(s, &mntfh); + if (IS_ERR(mntroot)) { + error = PTR_ERR(mntroot); + goto error_splat_super; + } + + s->s_flags |= MS_ACTIVE; + mnt->mnt_sb = s; + mnt->mnt_root = mntroot; + return 0; + +out_err_nosb: + nfs_free_server(server); +out_err_noserver: + return error; + +error_splat_super: + up_write(&s->s_umount); + deactivate_super(s); + return error; +} + +/* + * Destroy an NFS2/3 superblock + */ +static void nfs_kill_super(struct super_block *s) +{ + struct nfs_server *server = NFS_SB(s); + + kill_anon_super(s); + nfs_free_server(server); +} + +/* + * Clone an NFS2/3 server record on xdev traversal (FSID-change) + */ +static int nfs_xdev_get_sb(struct file_system_type *fs_type, int flags, + const char *dev_name, void *raw_data, + struct vfsmount *mnt) +{ + struct nfs_clone_mount *data = raw_data; + struct super_block *s; + struct nfs_server *server; + struct dentry *mntroot; + int error; + + dprintk("--> nfs_xdev_get_sb()\n"); + + /* create a new volume representation */ + server = nfs_clone_server(NFS_SB(data->sb), data->fh, data->fattr); + if (IS_ERR(server)) { + error = PTR_ERR(server); + goto out_err_noserver; + } + + /* Get a superblock - note that we may end up sharing one that already exists */ + s = sget(&nfs_fs_type, nfs_compare_super, nfs_set_super, server); + if (IS_ERR(s)) { + error = PTR_ERR(s); + goto out_err_nosb; + } + + if (s->s_fs_info != server) { + nfs_free_server(server); + server = NULL; + } + + if (!s->s_root) { + /* initial superblock/root creation */ + s->s_flags = flags; + nfs_clone_super(s, data->sb); + } + + mntroot = nfs_get_root(s, data->fh); + if (IS_ERR(mntroot)) { + error = PTR_ERR(mntroot); + goto error_splat_super; + } + + s->s_flags |= MS_ACTIVE; + mnt->mnt_sb = s; + mnt->mnt_root = mntroot; + + dprintk("<-- nfs_xdev_get_sb() = 0\n"); + return 0; + +out_err_nosb: + nfs_free_server(server); +out_err_noserver: + dprintk("<-- nfs_xdev_get_sb() = %d [error]\n", error); + return error; + +error_splat_super: + up_write(&s->s_umount); + deactivate_super(s); + dprintk("<-- nfs_xdev_get_sb() = %d [splat]\n", error); + return error; +} + +#ifdef CONFIG_NFS_V4 + +/* + * Finish setting up a cloned NFS4 superblock + */ +static void nfs4_clone_super(struct super_block *sb, + const struct super_block *old_sb) +{ + sb->s_blocksize_bits = old_sb->s_blocksize_bits; + sb->s_blocksize = old_sb->s_blocksize; + sb->s_maxbytes = old_sb->s_maxbytes; + sb->s_time_gran = 1; + sb->s_op = old_sb->s_op; + nfs_initialise_sb(sb); +} + +/* + * Set up an NFS4 superblock + */ +static void nfs4_fill_super(struct super_block *sb) +{ + sb->s_time_gran = 1; + sb->s_op = &nfs4_sops; + nfs_initialise_sb(sb); +} + +static void *nfs_copy_user_string(char *dst, struct nfs_string *src, int maxlen) +{ + void *p = NULL; + + if (!src->len) + return ERR_PTR(-EINVAL); + if (src->len < maxlen) + maxlen = src->len; + if (dst == NULL) { + p = dst = kmalloc(maxlen + 1, GFP_KERNEL); + if (p == NULL) + return ERR_PTR(-ENOMEM); + } + if (copy_from_user(dst, src->data, maxlen)) { + kfree(p); + return ERR_PTR(-EFAULT); + } + dst[maxlen] = '\0'; + return dst; +} + +/* + * Get the superblock for an NFS4 mountpoint + */ +static int nfs4_get_sb(struct file_system_type *fs_type, + int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt) +{ + struct nfs4_mount_data *data = raw_data; + struct super_block *s; + struct nfs_server *server; + struct sockaddr_in addr; + rpc_authflavor_t authflavour; + struct nfs_fh mntfh; + struct dentry *mntroot; + char *mntpath = NULL, *hostname = NULL, ip_addr[16]; + void *p; + int error; + + if (data == NULL) { + dprintk("%s: missing data argument\n", __FUNCTION__); + return -EINVAL; + } + if (data->version <= 0 || data->version > NFS4_MOUNT_VERSION) { + dprintk("%s: bad mount version\n", __FUNCTION__); + return -EINVAL; + } + + /* We now require that the mount process passes the remote address */ + if (data->host_addrlen != sizeof(addr)) + return -EINVAL; + + if (copy_from_user(&addr, data->host_addr, sizeof(addr))) + return -EFAULT; + + if (addr.sin_family != AF_INET || + addr.sin_addr.s_addr == INADDR_ANY + ) { + dprintk("%s: mount program didn't pass remote IP address!\n", + __FUNCTION__); + return -EINVAL; + } + /* RFC3530: The default port for NFS is 2049 */ + if (addr.sin_port == 0) + addr.sin_port = NFS_PORT; + + /* Grab the authentication type */ + authflavour = RPC_AUTH_UNIX; + if (data->auth_flavourlen != 0) { + if (data->auth_flavourlen != 1) { + dprintk("%s: Invalid number of RPC auth flavours %d.\n", + __FUNCTION__, data->auth_flavourlen); + error = -EINVAL; + goto out_err_noserver; + } + + if (copy_from_user(&authflavour, data->auth_flavours, + sizeof(authflavour))) { + error = -EFAULT; + goto out_err_noserver; + } + } + + p = nfs_copy_user_string(NULL, &data->hostname, 256); + if (IS_ERR(p)) + goto out_err; + hostname = p; + + p = nfs_copy_user_string(NULL, &data->mnt_path, 1024); + if (IS_ERR(p)) + goto out_err; + mntpath = p; + + dprintk("MNTPATH: %s\n", mntpath); + + p = nfs_copy_user_string(ip_addr, &data->client_addr, + sizeof(ip_addr) - 1); + if (IS_ERR(p)) + goto out_err; + + /* Get a volume representation */ + server = nfs4_create_server(data, hostname, &addr, mntpath, ip_addr, + authflavour, &mntfh); + if (IS_ERR(server)) { + error = PTR_ERR(server); + goto out_err_noserver; + } + + /* Get a superblock - note that we may end up sharing one that already exists */ + s = sget(fs_type, nfs_compare_super, nfs_set_super, server); + if (IS_ERR(s)) { + error = PTR_ERR(s); + goto out_free; + } + + if (s->s_fs_info != server) { + nfs_free_server(server); + server = NULL; + } + + if (!s->s_root) { + /* initial superblock/root creation */ + s->s_flags = flags; + nfs4_fill_super(s); + } + + mntroot = nfs4_get_root(s, &mntfh); + if (IS_ERR(mntroot)) { + error = PTR_ERR(mntroot); + goto error_splat_super; + } + + s->s_flags |= MS_ACTIVE; + mnt->mnt_sb = s; + mnt->mnt_root = mntroot; + kfree(mntpath); + kfree(hostname); + return 0; + +out_err: + error = PTR_ERR(p); + goto out_err_noserver; + +out_free: + nfs_free_server(server); +out_err_noserver: + kfree(mntpath); + kfree(hostname); + return error; + +error_splat_super: + up_write(&s->s_umount); + deactivate_super(s); + goto out_err_noserver; +} + +static void nfs4_kill_super(struct super_block *sb) +{ + struct nfs_server *server = NFS_SB(sb); + + nfs_return_all_delegations(sb); + kill_anon_super(sb); + + nfs4_renewd_prepare_shutdown(server); + nfs_free_server(server); +} + +/* + * Clone an NFS4 server record on xdev traversal (FSID-change) + */ +static int nfs4_xdev_get_sb(struct file_system_type *fs_type, int flags, + const char *dev_name, void *raw_data, + struct vfsmount *mnt) +{ + struct nfs_clone_mount *data = raw_data; + struct super_block *s; + struct nfs_server *server; + struct dentry *mntroot; + int error; + + dprintk("--> nfs4_xdev_get_sb()\n"); + + /* create a new volume representation */ + server = nfs_clone_server(NFS_SB(data->sb), data->fh, data->fattr); + if (IS_ERR(server)) { + error = PTR_ERR(server); + goto out_err_noserver; + } + + /* Get a superblock - note that we may end up sharing one that already exists */ + s = sget(&nfs_fs_type, nfs_compare_super, nfs_set_super, server); + if (IS_ERR(s)) { + error = PTR_ERR(s); + goto out_err_nosb; + } + + if (s->s_fs_info != server) { + nfs_free_server(server); + server = NULL; + } + + if (!s->s_root) { + /* initial superblock/root creation */ + s->s_flags = flags; + nfs4_clone_super(s, data->sb); + } + + mntroot = nfs4_get_root(s, data->fh); + if (IS_ERR(mntroot)) { + error = PTR_ERR(mntroot); + goto error_splat_super; + } + + s->s_flags |= MS_ACTIVE; + mnt->mnt_sb = s; + mnt->mnt_root = mntroot; + + dprintk("<-- nfs4_xdev_get_sb() = 0\n"); + return 0; + +out_err_nosb: + nfs_free_server(server); +out_err_noserver: + dprintk("<-- nfs4_xdev_get_sb() = %d [error]\n", error); + return error; + +error_splat_super: + up_write(&s->s_umount); + deactivate_super(s); + dprintk("<-- nfs4_xdev_get_sb() = %d [splat]\n", error); + return error; +} + +/* + * Create an NFS4 server record on referral traversal + */ +static int nfs4_referral_get_sb(struct file_system_type *fs_type, int flags, + const char *dev_name, void *raw_data, + struct vfsmount *mnt) +{ + struct nfs_clone_mount *data = raw_data; + struct super_block *s; + struct nfs_server *server; + struct dentry *mntroot; + struct nfs_fh mntfh; + int error; + + dprintk("--> nfs4_referral_get_sb()\n"); + + /* create a new volume representation */ + server = nfs4_create_referral_server(data, &mntfh); + if (IS_ERR(server)) { + error = PTR_ERR(server); + goto out_err_noserver; + } + + /* Get a superblock - note that we may end up sharing one that already exists */ + s = sget(&nfs_fs_type, nfs_compare_super, nfs_set_super, server); + if (IS_ERR(s)) { + error = PTR_ERR(s); + goto out_err_nosb; + } + + if (s->s_fs_info != server) { + nfs_free_server(server); + server = NULL; + } + + if (!s->s_root) { + /* initial superblock/root creation */ + s->s_flags = flags; + nfs4_fill_super(s); + } + + mntroot = nfs4_get_root(s, data->fh); + if (IS_ERR(mntroot)) { + error = PTR_ERR(mntroot); + goto error_splat_super; + } + + s->s_flags |= MS_ACTIVE; + mnt->mnt_sb = s; + mnt->mnt_root = mntroot; + + dprintk("<-- nfs4_referral_get_sb() = 0\n"); + return 0; + +out_err_nosb: + nfs_free_server(server); +out_err_noserver: + dprintk("<-- nfs4_referral_get_sb() = %d [error]\n", error); + return error; + +error_splat_super: + up_write(&s->s_umount); + deactivate_super(s); + dprintk("<-- nfs4_referral_get_sb() = %d [splat]\n", error); + return error; +} + +#endif /* CONFIG_NFS_V4 */ diff --git a/fs/nfs/symlink.c b/fs/nfs/symlink.c index 18dc95b0b64..600bbe630ab 100644 --- a/fs/nfs/symlink.c +++ b/fs/nfs/symlink.c @@ -52,7 +52,7 @@ static void *nfs_follow_link(struct dentry *dentry, struct nameidata *nd) { struct inode *inode = dentry->d_inode; struct page *page; - void *err = ERR_PTR(nfs_revalidate_inode(NFS_SERVER(inode), inode)); + void *err = ERR_PTR(nfs_revalidate_mapping(inode, inode->i_mapping)); if (err) goto read_failed; page = read_cache_page(&inode->i_data, 0, @@ -75,22 +75,13 @@ read_failed: return NULL; } -static void nfs_put_link(struct dentry *dentry, struct nameidata *nd, void *cookie) -{ - if (cookie) { - struct page *page = cookie; - kunmap(page); - page_cache_release(page); - } -} - /* * symlinks can't do much... */ struct inode_operations nfs_symlink_inode_operations = { .readlink = generic_readlink, .follow_link = nfs_follow_link, - .put_link = nfs_put_link, + .put_link = page_put_link, .getattr = nfs_getattr, .setattr = nfs_setattr, }; diff --git a/fs/nfs/sysctl.c b/fs/nfs/sysctl.c index 4c486eb867c..2fe3403c240 100644 --- a/fs/nfs/sysctl.c +++ b/fs/nfs/sysctl.c @@ -3,7 +3,6 @@ * * Sysctl interface to NFS parameters */ -#include <linux/config.h> #include <linux/types.h> #include <linux/linkage.h> #include <linux/ctype.h> @@ -12,6 +11,7 @@ #include <linux/module.h> #include <linux/nfs4.h> #include <linux/nfs_idmap.h> +#include <linux/nfs_fs.h> #include "callback.h" @@ -46,6 +46,15 @@ static ctl_table nfs_cb_sysctls[] = { .strategy = &sysctl_jiffies, }, #endif + { + .ctl_name = CTL_UNNUMBERED, + .procname = "nfs_mountpoint_timeout", + .data = &nfs_mountpoint_expiry_timeout, + .maxlen = sizeof(nfs_mountpoint_expiry_timeout), + .mode = 0644, + .proc_handler = &proc_dointvec_jiffies, + .strategy = &sysctl_jiffies, + }, { .ctl_name = 0 } }; diff --git a/fs/nfs/write.c b/fs/nfs/write.c index 4cfada2cc09..b674462793d 100644 --- a/fs/nfs/write.c +++ b/fs/nfs/write.c @@ -46,7 +46,6 @@ * Copyright (C) 1996, 1997, Olaf Kirch <okir@monad.swb.de> */ -#include <linux/config.h> #include <linux/types.h> #include <linux/slab.h> #include <linux/mm.h> @@ -91,23 +90,13 @@ static mempool_t *nfs_commit_mempool; static DECLARE_WAIT_QUEUE_HEAD(nfs_write_congestion); -struct nfs_write_data *nfs_commit_alloc(unsigned int pagecount) +struct nfs_write_data *nfs_commit_alloc(void) { struct nfs_write_data *p = mempool_alloc(nfs_commit_mempool, SLAB_NOFS); if (p) { memset(p, 0, sizeof(*p)); INIT_LIST_HEAD(&p->pages); - if (pagecount < NFS_PAGEVEC_SIZE) - p->pagevec = &p->page_array[0]; - else { - size_t size = ++pagecount * sizeof(struct page *); - p->pagevec = kzalloc(size, GFP_NOFS); - if (!p->pagevec) { - mempool_free(p, nfs_commit_mempool); - p = NULL; - } - } } return p; } @@ -119,21 +108,20 @@ void nfs_commit_free(struct nfs_write_data *p) mempool_free(p, nfs_commit_mempool); } -struct nfs_write_data *nfs_writedata_alloc(unsigned int pagecount) +struct nfs_write_data *nfs_writedata_alloc(size_t len) { + unsigned int pagecount = (len + PAGE_SIZE - 1) >> PAGE_SHIFT; struct nfs_write_data *p = mempool_alloc(nfs_wdata_mempool, SLAB_NOFS); if (p) { memset(p, 0, sizeof(*p)); INIT_LIST_HEAD(&p->pages); - if (pagecount < NFS_PAGEVEC_SIZE) - p->pagevec = &p->page_array[0]; + p->npages = pagecount; + if (pagecount <= ARRAY_SIZE(p->page_array)) + p->pagevec = p->page_array; else { - size_t size = ++pagecount * sizeof(struct page *); - p->pagevec = kmalloc(size, GFP_NOFS); - if (p->pagevec) { - memset(p->pagevec, 0, size); - } else { + p->pagevec = kcalloc(pagecount, sizeof(struct page *), GFP_NOFS); + if (!p->pagevec) { mempool_free(p, nfs_wdata_mempool); p = NULL; } @@ -142,7 +130,7 @@ struct nfs_write_data *nfs_writedata_alloc(unsigned int pagecount) return p; } -void nfs_writedata_free(struct nfs_write_data *p) +static void nfs_writedata_free(struct nfs_write_data *p) { if (p && (p->pagevec != &p->page_array[0])) kfree(p->pagevec); @@ -213,7 +201,7 @@ static int nfs_writepage_sync(struct nfs_open_context *ctx, struct inode *inode, int result, written = 0; struct nfs_write_data *wdata; - wdata = nfs_writedata_alloc(1); + wdata = nfs_writedata_alloc(wsize); if (!wdata) return -ENOMEM; @@ -408,6 +396,7 @@ int nfs_writepages(struct address_space *mapping, struct writeback_control *wbc) out: clear_bit(BDI_write_congested, &bdi->state); wake_up_all(&nfs_write_congestion); + writeback_congestion_end(); return err; } @@ -501,7 +490,7 @@ nfs_mark_request_dirty(struct nfs_page *req) nfs_list_add_request(req, &nfsi->dirty); nfsi->ndirty++; spin_unlock(&nfsi->req_lock); - inc_page_state(nr_dirty); + inc_zone_page_state(req->wb_page, NR_FILE_DIRTY); mark_inode_dirty(inode); } @@ -529,7 +518,7 @@ nfs_mark_request_commit(struct nfs_page *req) nfs_list_add_request(req, &nfsi->commit); nfsi->ncommit++; spin_unlock(&nfsi->req_lock); - inc_page_state(nr_unstable); + inc_zone_page_state(req->wb_page, NR_UNSTABLE_NFS); mark_inode_dirty(inode); } #endif @@ -583,6 +572,30 @@ static int nfs_wait_on_requests(struct inode *inode, unsigned long idx_start, un return ret; } +static void nfs_cancel_dirty_list(struct list_head *head) +{ + struct nfs_page *req; + while(!list_empty(head)) { + req = nfs_list_entry(head->next); + nfs_list_remove_request(req); + nfs_inode_remove_request(req); + nfs_clear_page_writeback(req); + } +} + +static void nfs_cancel_commit_list(struct list_head *head) +{ + struct nfs_page *req; + + while(!list_empty(head)) { + req = nfs_list_entry(head->next); + nfs_list_remove_request(req); + nfs_inode_remove_request(req); + dec_zone_page_state(req->wb_page, NR_UNSTABLE_NFS); + nfs_clear_page_writeback(req); + } +} + /* * nfs_scan_dirty - Scan an inode for dirty requests * @inode: NFS inode to scan @@ -602,7 +615,6 @@ nfs_scan_dirty(struct inode *inode, struct list_head *dst, unsigned long idx_sta if (nfsi->ndirty != 0) { res = nfs_scan_lock_dirty(nfsi, dst, idx_start, npages); nfsi->ndirty -= res; - sub_page_state(nr_dirty,res); if ((nfsi->ndirty == 0) != list_empty(&nfsi->dirty)) printk(KERN_ERR "NFS: desynchronized value of nfs_i.ndirty.\n"); } @@ -627,7 +639,7 @@ nfs_scan_commit(struct inode *inode, struct list_head *dst, unsigned long idx_st int res = 0; if (nfsi->ncommit != 0) { - res = nfs_scan_list(&nfsi->commit, dst, idx_start, npages); + res = nfs_scan_list(nfsi, &nfsi->commit, dst, idx_start, npages); nfsi->ncommit -= res; if ((nfsi->ncommit == 0) != list_empty(&nfsi->commit)) printk(KERN_ERR "NFS: desynchronized value of nfs_i.ncommit.\n"); @@ -981,24 +993,24 @@ static int nfs_flush_multi(struct inode *inode, struct list_head *head, int how) struct nfs_page *req = nfs_list_entry(head->next); struct page *page = req->wb_page; struct nfs_write_data *data; - unsigned int wsize = NFS_SERVER(inode)->wsize; - unsigned int nbytes, offset; + size_t wsize = NFS_SERVER(inode)->wsize, nbytes; + unsigned int offset; int requests = 0; LIST_HEAD(list); nfs_list_remove_request(req); nbytes = req->wb_bytes; - for (;;) { - data = nfs_writedata_alloc(1); + do { + size_t len = min(nbytes, wsize); + + data = nfs_writedata_alloc(len); if (!data) goto out_bad; list_add(&data->pages, &list); requests++; - if (nbytes <= wsize) - break; - nbytes -= wsize; - } + nbytes -= len; + } while (nbytes != 0); atomic_set(&req->wb_complete, requests); ClearPageError(page); @@ -1052,7 +1064,7 @@ static int nfs_flush_one(struct inode *inode, struct list_head *head, int how) struct nfs_write_data *data; unsigned int count; - data = nfs_writedata_alloc(NFS_SERVER(inode)->wpages); + data = nfs_writedata_alloc(NFS_SERVER(inode)->wsize); if (!data) goto out_bad; @@ -1241,7 +1253,13 @@ int nfs_writeback_done(struct rpc_task *task, struct nfs_write_data *data) dprintk("NFS: %4d nfs_writeback_done (status %d)\n", task->tk_pid, task->tk_status); - /* Call the NFS version-specific code */ + /* + * ->write_done will attempt to use post-op attributes to detect + * conflicting writes by other clients. A strict interpretation + * of close-to-open would allow us to continue caching even if + * another writer had changed the file, but some applications + * depend on tighter cache coherency when writing. + */ status = NFS_PROTO(data->inode)->write_done(task, data); if (status != 0) return status; @@ -1262,7 +1280,7 @@ int nfs_writeback_done(struct rpc_task *task, struct nfs_write_data *data) if (time_before(complain, jiffies)) { dprintk("NFS: faulty NFS server %s:" " (committed = %d) != (stable = %d)\n", - NFS_SERVER(data->inode)->hostname, + NFS_SERVER(data->inode)->nfs_client->cl_hostname, resp->verf->committed, argp->stable); complain = jiffies + 300 * HZ; } @@ -1360,7 +1378,7 @@ nfs_commit_list(struct inode *inode, struct list_head *head, int how) struct nfs_write_data *data; struct nfs_page *req; - data = nfs_commit_alloc(NFS_SERVER(inode)->wpages); + data = nfs_commit_alloc(); if (!data) goto out_bad; @@ -1375,6 +1393,7 @@ nfs_commit_list(struct inode *inode, struct list_head *head, int how) req = nfs_list_entry(head->next); nfs_list_remove_request(req); nfs_mark_request_commit(req); + dec_zone_page_state(req->wb_page, NR_UNSTABLE_NFS); nfs_clear_page_writeback(req); } return -ENOMEM; @@ -1387,7 +1406,6 @@ static void nfs_commit_done(struct rpc_task *task, void *calldata) { struct nfs_write_data *data = calldata; struct nfs_page *req; - int res = 0; dprintk("NFS: %4d nfs_commit_done (status %d)\n", task->tk_pid, task->tk_status); @@ -1399,6 +1417,7 @@ static void nfs_commit_done(struct rpc_task *task, void *calldata) while (!list_empty(&data->pages)) { req = nfs_list_entry(data->pages.next); nfs_list_remove_request(req); + dec_zone_page_state(req->wb_page, NR_UNSTABLE_NFS); dprintk("NFS: commit (%s/%Ld %d@%Ld)", req->wb_context->dentry->d_inode->i_sb->s_id, @@ -1425,9 +1444,7 @@ static void nfs_commit_done(struct rpc_task *task, void *calldata) nfs_mark_request_dirty(req); next: nfs_clear_page_writeback(req); - res++; } - sub_page_state(nr_unstable,res); } static const struct rpc_call_ops nfs_commit_ops = { @@ -1495,15 +1512,25 @@ int nfs_sync_inode_wait(struct inode *inode, unsigned long idx_start, pages = nfs_scan_dirty(inode, &head, idx_start, npages); if (pages != 0) { spin_unlock(&nfsi->req_lock); - ret = nfs_flush_list(inode, &head, pages, how); + if (how & FLUSH_INVALIDATE) + nfs_cancel_dirty_list(&head); + else + ret = nfs_flush_list(inode, &head, pages, how); spin_lock(&nfsi->req_lock); continue; } if (nocommit) break; - pages = nfs_scan_commit(inode, &head, 0, 0); + pages = nfs_scan_commit(inode, &head, idx_start, npages); if (pages == 0) break; + if (how & FLUSH_INVALIDATE) { + spin_unlock(&nfsi->req_lock); + nfs_cancel_commit_list(&head); + spin_lock(&nfsi->req_lock); + continue; + } + pages += nfs_scan_commit(inode, &head, 0, 0); spin_unlock(&nfsi->req_lock); ret = nfs_commit_list(inode, &head, how); spin_lock(&nfsi->req_lock); @@ -1512,7 +1539,7 @@ int nfs_sync_inode_wait(struct inode *inode, unsigned long idx_start, return ret; } -int nfs_init_writepagecache(void) +int __init nfs_init_writepagecache(void) { nfs_wdata_cachep = kmem_cache_create("nfs_write_data", sizeof(struct nfs_write_data), @@ -1538,7 +1565,6 @@ void nfs_destroy_writepagecache(void) { mempool_destroy(nfs_commit_mempool); mempool_destroy(nfs_wdata_mempool); - if (kmem_cache_destroy(nfs_wdata_cachep)) - printk(KERN_INFO "nfs_write_data: not all structures were freed\n"); + kmem_cache_destroy(nfs_wdata_cachep); } diff --git a/fs/nfsctl.c b/fs/nfsctl.c index a5a18d4aca4..c043136a82c 100644 --- a/fs/nfsctl.c +++ b/fs/nfsctl.c @@ -4,7 +4,6 @@ * This should eventually move to userland. * */ -#include <linux/config.h> #include <linux/types.h> #include <linux/file.h> #include <linux/fs.h> diff --git a/fs/nfsd/export.c b/fs/nfsd/export.c index 3eec30000f3..01bc68c628a 100644 --- a/fs/nfsd/export.c +++ b/fs/nfsd/export.c @@ -126,7 +126,7 @@ static int expkey_parse(struct cache_detail *cd, char *mesg, int mlen) if (*ep) goto out; dprintk("found fsidtype %d\n", fsidtype); - if (fsidtype > 2) + if (key_len(fsidtype)==0) /* invalid type */ goto out; if ((len=qword_get(&mesg, buf, PAGE_SIZE)) <= 0) goto out; diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c index dbaf3f93f32..8583d99ee74 100644 --- a/fs/nfsd/nfs4callback.c +++ b/fs/nfsd/nfs4callback.c @@ -33,7 +33,6 @@ * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. */ -#include <linux/config.h> #include <linux/module.h> #include <linux/list.h> #include <linux/inet.h> @@ -376,16 +375,28 @@ nfsd4_probe_callback(struct nfs4_client *clp) { struct sockaddr_in addr; struct nfs4_callback *cb = &clp->cl_callback; - struct rpc_timeout timeparms; - struct rpc_xprt * xprt; + struct rpc_timeout timeparms = { + .to_initval = (NFSD_LEASE_TIME/4) * HZ, + .to_retries = 5, + .to_maxval = (NFSD_LEASE_TIME/2) * HZ, + .to_exponential = 1, + }; struct rpc_program * program = &cb->cb_program; - struct rpc_stat * stat = &cb->cb_stat; - struct rpc_clnt * clnt; + struct rpc_create_args args = { + .protocol = IPPROTO_TCP, + .address = (struct sockaddr *)&addr, + .addrsize = sizeof(addr), + .timeout = &timeparms, + .servername = clp->cl_name.data, + .program = program, + .version = nfs_cb_version[1]->number, + .authflavor = RPC_AUTH_UNIX, /* XXX: need AUTH_GSS... */ + .flags = (RPC_CLNT_CREATE_NOPING), + }; struct rpc_message msg = { .rpc_proc = &nfs4_cb_procedures[NFSPROC4_CLNT_CB_NULL], .rpc_argp = clp, }; - char hostname[32]; int status; if (atomic_read(&cb->cb_set)) @@ -397,51 +408,27 @@ nfsd4_probe_callback(struct nfs4_client *clp) addr.sin_port = htons(cb->cb_port); addr.sin_addr.s_addr = htonl(cb->cb_addr); - /* Initialize timeout */ - timeparms.to_initval = (NFSD_LEASE_TIME/4) * HZ; - timeparms.to_retries = 0; - timeparms.to_maxval = (NFSD_LEASE_TIME/2) * HZ; - timeparms.to_exponential = 1; - - /* Create RPC transport */ - xprt = xprt_create_proto(IPPROTO_TCP, &addr, &timeparms); - if (IS_ERR(xprt)) { - dprintk("NFSD: couldn't create callback transport!\n"); - goto out_err; - } - /* Initialize rpc_program */ program->name = "nfs4_cb"; program->number = cb->cb_prog; program->nrvers = ARRAY_SIZE(nfs_cb_version); program->version = nfs_cb_version; - program->stats = stat; + program->stats = &cb->cb_stat; /* Initialize rpc_stat */ - memset(stat, 0, sizeof(struct rpc_stat)); - stat->program = program; - - /* Create RPC client - * - * XXX AUTH_UNIX only - need AUTH_GSS.... - */ - sprintf(hostname, "%u.%u.%u.%u", NIPQUAD(addr.sin_addr.s_addr)); - clnt = rpc_new_client(xprt, hostname, program, 1, RPC_AUTH_UNIX); - if (IS_ERR(clnt)) { + memset(program->stats, 0, sizeof(cb->cb_stat)); + program->stats->program = program; + + /* Create RPC client */ + cb->cb_client = rpc_create(&args); + if (!cb->cb_client) { dprintk("NFSD: couldn't create callback client\n"); goto out_err; } - clnt->cl_intr = 0; - clnt->cl_softrtry = 1; /* Kick rpciod, put the call on the wire. */ - - if (rpciod_up() != 0) { - dprintk("nfsd: couldn't start rpciod for callbacks!\n"); + if (rpciod_up() != 0) goto out_clnt; - } - - cb->cb_client = clnt; /* the task holds a reference to the nfs4_client struct */ atomic_inc(&clp->cl_count); @@ -449,7 +436,7 @@ nfsd4_probe_callback(struct nfs4_client *clp) msg.rpc_cred = nfsd4_lookupcred(clp,0); if (IS_ERR(msg.rpc_cred)) goto out_rpciod; - status = rpc_call_async(clnt, &msg, RPC_TASK_ASYNC, &nfs4_cb_null_ops, NULL); + status = rpc_call_async(cb->cb_client, &msg, RPC_TASK_ASYNC, &nfs4_cb_null_ops, NULL); put_rpccred(msg.rpc_cred); if (status != 0) { @@ -463,7 +450,7 @@ out_rpciod: rpciod_down(); cb->cb_client = NULL; out_clnt: - rpc_shutdown_client(clnt); + rpc_shutdown_client(cb->cb_client); out_err: dprintk("NFSD: warning: no callback path to client %.*s\n", (int)clp->cl_name.len, clp->cl_name.data); diff --git a/fs/nfsd/nfs4idmap.c b/fs/nfsd/nfs4idmap.c index 4b6aa60dfce..b1902ebaab4 100644 --- a/fs/nfsd/nfs4idmap.c +++ b/fs/nfsd/nfs4idmap.c @@ -34,7 +34,6 @@ * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. */ -#include <linux/config.h> #include <linux/module.h> #include <linux/init.h> @@ -574,10 +573,9 @@ idmap_lookup(struct svc_rqst *rqstp, struct idmap_defer_req *mdr; int ret; - mdr = kmalloc(sizeof(*mdr), GFP_KERNEL); + mdr = kzalloc(sizeof(*mdr), GFP_KERNEL); if (!mdr) return -ENOMEM; - memset(mdr, 0, sizeof(*mdr)); atomic_set(&mdr->count, 1); init_waitqueue_head(&mdr->waitq); mdr->req.defer = idmap_defer; diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index b0e095ea0c0..ee4eff27aed 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -721,6 +721,12 @@ nfsd4_proc_null(struct svc_rqst *rqstp, void *argp, void *resp) return nfs_ok; } +static inline void nfsd4_increment_op_stats(u32 opnum) +{ + if (opnum >= FIRST_NFS4_OP && opnum <= LAST_NFS4_OP) + nfsdstats.nfs4_opcount[opnum]++; +} + /* * COMPOUND call. @@ -930,6 +936,8 @@ encode_op: /* XXX Ugh, we need to get rid of this kind of special case: */ if (op->opnum == OP_READ && op->u.read.rd_filp) fput(op->u.read.rd_filp); + + nfsd4_increment_op_stats(op->opnum); } out: diff --git a/fs/nfsd/nfs4recover.c b/fs/nfsd/nfs4recover.c index 06da7506363..e35d7e52fde 100644 --- a/fs/nfsd/nfs4recover.c +++ b/fs/nfsd/nfs4recover.c @@ -33,7 +33,7 @@ * */ - +#include <linux/err.h> #include <linux/sunrpc/svc.h> #include <linux/nfsd/nfsd.h> #include <linux/nfs4.h> @@ -87,34 +87,35 @@ int nfs4_make_rec_clidname(char *dname, struct xdr_netobj *clname) { struct xdr_netobj cksum; - struct crypto_tfm *tfm; + struct hash_desc desc; struct scatterlist sg[1]; int status = nfserr_resource; dprintk("NFSD: nfs4_make_rec_clidname for %.*s\n", clname->len, clname->data); - tfm = crypto_alloc_tfm("md5", CRYPTO_TFM_REQ_MAY_SLEEP); - if (tfm == NULL) - goto out; - cksum.len = crypto_tfm_alg_digestsize(tfm); + desc.flags = CRYPTO_TFM_REQ_MAY_SLEEP; + desc.tfm = crypto_alloc_hash("md5", 0, CRYPTO_ALG_ASYNC); + if (IS_ERR(desc.tfm)) + goto out_no_tfm; + cksum.len = crypto_hash_digestsize(desc.tfm); cksum.data = kmalloc(cksum.len, GFP_KERNEL); if (cksum.data == NULL) goto out; - crypto_digest_init(tfm); sg[0].page = virt_to_page(clname->data); sg[0].offset = offset_in_page(clname->data); sg[0].length = clname->len; - crypto_digest_update(tfm, sg, 1); - crypto_digest_final(tfm, cksum.data); + if (crypto_hash_digest(&desc, sg, sg->length, cksum.data)) + goto out; md5_to_hex(dname, cksum.data); kfree(cksum.data); status = nfs_ok; out: - crypto_free_tfm(tfm); + crypto_free_hash(desc.tfm); +out_no_tfm: return status; } diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 96c7578cbe1..ebcf226a9e4 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -123,7 +123,7 @@ static void release_stateid(struct nfs4_stateid *stp, int flags); */ /* recall_lock protects the del_recall_lru */ -static spinlock_t recall_lock = SPIN_LOCK_UNLOCKED; +static DEFINE_SPINLOCK(recall_lock); static struct list_head del_recall_lru; static void @@ -339,8 +339,7 @@ alloc_client(struct xdr_netobj name) { struct nfs4_client *clp; - if ((clp = kmalloc(sizeof(struct nfs4_client), GFP_KERNEL))!= NULL) { - memset(clp, 0, sizeof(*clp)); + if ((clp = kzalloc(sizeof(struct nfs4_client), GFP_KERNEL))!= NULL) { if ((clp->cl_name.data = kmalloc(name.len, GFP_KERNEL)) != NULL) { memcpy(clp->cl_name.data, name.data, name.len); clp->cl_name.len = name.len; @@ -529,8 +528,7 @@ move_to_confirmed(struct nfs4_client *clp) dprintk("NFSD: move_to_confirm nfs4_client %p\n", clp); list_del_init(&clp->cl_strhash); - list_del_init(&clp->cl_idhash); - list_add(&clp->cl_idhash, &conf_id_hashtbl[idhashval]); + list_move(&clp->cl_idhash, &conf_id_hashtbl[idhashval]); strhashval = clientstr_hashval(clp->cl_recdir); list_add(&clp->cl_strhash, &conf_str_hashtbl[strhashval]); renew_client(clp); @@ -1007,13 +1005,10 @@ alloc_init_file(struct inode *ino) static void nfsd4_free_slab(kmem_cache_t **slab) { - int status; - if (*slab == NULL) return; - status = kmem_cache_destroy(*slab); + kmem_cache_destroy(*slab); *slab = NULL; - WARN_ON(status); } static void @@ -1238,8 +1233,15 @@ find_file(struct inode *ino) return NULL; } -#define TEST_ACCESS(x) ((x > 0 || x < 4)?1:0) -#define TEST_DENY(x) ((x >= 0 || x < 5)?1:0) +static int access_valid(u32 x) +{ + return (x > 0 && x < 4); +} + +static int deny_valid(u32 x) +{ + return (x >= 0 && x < 5); +} static void set_access(unsigned int *access, unsigned long bmap) { @@ -1746,7 +1748,8 @@ nfsd4_process_open2(struct svc_rqst *rqstp, struct svc_fh *current_fh, struct nf int status; status = nfserr_inval; - if (!TEST_ACCESS(open->op_share_access) || !TEST_DENY(open->op_share_deny)) + if (!access_valid(open->op_share_access) + || !deny_valid(open->op_share_deny)) goto out; /* * Lookup file; if found, lookup stateid and check open request, @@ -1783,10 +1786,10 @@ nfsd4_process_open2(struct svc_rqst *rqstp, struct svc_fh *current_fh, struct nf } else { /* Stateid was not found, this is a new OPEN */ int flags = 0; + if (open->op_share_access & NFS4_SHARE_ACCESS_READ) + flags |= MAY_READ; if (open->op_share_access & NFS4_SHARE_ACCESS_WRITE) - flags = MAY_WRITE; - else - flags = MAY_READ; + flags |= MAY_WRITE; status = nfs4_new_open(rqstp, &stp, dp, current_fh, flags); if (status) goto out; @@ -2071,16 +2074,12 @@ nfs4_preprocess_stateid_op(struct svc_fh *current_fh, stateid_t *stateid, int fl if (!stateid->si_fileid) { /* delegation stateid */ if(!(dp = find_delegation_stateid(ino, stateid))) { dprintk("NFSD: delegation stateid not found\n"); - if (nfs4_in_grace()) - status = nfserr_grace; goto out; } stidp = &dp->dl_stateid; } else { /* open or lock stateid */ if (!(stp = find_stateid(stateid, flags))) { dprintk("NFSD: open or lock stateid not found\n"); - if (nfs4_in_grace()) - status = nfserr_grace; goto out; } if ((flags & CHECK_FH) && nfs4_check_fh(current_fh, stp)) @@ -2253,8 +2252,9 @@ nfsd4_open_confirm(struct svc_rqst *rqstp, struct svc_fh *current_fh, struct nfs (int)current_fh->fh_dentry->d_name.len, current_fh->fh_dentry->d_name.name); - if ((status = fh_verify(rqstp, current_fh, S_IFREG, 0))) - goto out; + status = fh_verify(rqstp, current_fh, S_IFREG, 0); + if (status) + return status; nfs4_lock_state(); @@ -2321,7 +2321,8 @@ nfsd4_open_downgrade(struct svc_rqst *rqstp, struct svc_fh *current_fh, struct n (int)current_fh->fh_dentry->d_name.len, current_fh->fh_dentry->d_name.name); - if (!TEST_ACCESS(od->od_share_access) || !TEST_DENY(od->od_share_deny)) + if (!access_valid(od->od_share_access) + || !deny_valid(od->od_share_deny)) return nfserr_inval; nfs4_lock_state(); diff --git a/fs/nfsd/nfscache.c b/fs/nfsd/nfscache.c index d852ebb538e..fdf7cf3dfad 100644 --- a/fs/nfsd/nfscache.c +++ b/fs/nfsd/nfscache.c @@ -103,8 +103,7 @@ nfsd_cache_shutdown(void) static void lru_put_end(struct svc_cacherep *rp) { - list_del(&rp->c_lru); - list_add_tail(&rp->c_lru, &lru_head); + list_move_tail(&rp->c_lru, &lru_head); } /* diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index a1810e6a93e..7046ac9cf97 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -6,7 +6,6 @@ * Copyright (C) 1995, 1996 Olaf Kirch <okir@monad.swb.de> */ -#include <linux/config.h> #include <linux/module.h> #include <linux/linkage.h> diff --git a/fs/nfsd/nfsfh.c b/fs/nfsd/nfsfh.c index 3f2ec2e6d06..501d8388453 100644 --- a/fs/nfsd/nfsfh.c +++ b/fs/nfsd/nfsfh.c @@ -188,11 +188,9 @@ fh_verify(struct svc_rqst *rqstp, struct svc_fh *fhp, int type, int access) } /* Set user creds for this exportpoint */ - error = nfsd_setuser(rqstp, exp); - if (error) { - error = nfserrno(error); + error = nfserrno(nfsd_setuser(rqstp, exp)); + if (error) goto out; - } /* * Look up the dentry using the NFS file handle. @@ -248,9 +246,18 @@ fh_verify(struct svc_rqst *rqstp, struct svc_fh *fhp, int type, int access) dprintk("nfsd: fh_verify - just checking\n"); dentry = fhp->fh_dentry; exp = fhp->fh_export; + /* Set user creds for this exportpoint; necessary even + * in the "just checking" case because this may be a + * filehandle that was created by fh_compose, and that + * is about to be used in another nfsv4 compound + * operation */ + error = nfserrno(nfsd_setuser(rqstp, exp)); + if (error) + goto out; } cache_get(&exp->h); + error = nfsd_mode_check(rqstp, dentry->d_inode->i_mode, type); if (error) goto out; @@ -312,8 +319,8 @@ int fh_compose(struct svc_fh *fhp, struct svc_export *exp, struct dentry *dentry, struct svc_fh *ref_fh) { /* ref_fh is a reference file handle. - * if it is non-null, then we should compose a filehandle which is - * of the same version, where possible. + * if it is non-null and for the same filesystem, then we should compose + * a filehandle which is of the same version, where possible. * Currently, that means that if ref_fh->fh_handle.fh_version == 0xca * Then create a 32byte filehandle using nfs_fhbase_old * @@ -332,7 +339,7 @@ fh_compose(struct svc_fh *fhp, struct svc_export *exp, struct dentry *dentry, st parent->d_name.name, dentry->d_name.name, (inode ? inode->i_ino : 0)); - if (ref_fh) { + if (ref_fh && ref_fh->fh_export == exp) { ref_fh_version = ref_fh->fh_handle.fh_version; if (ref_fh_version == 0xca) ref_fh_fsid_type = 0; @@ -461,7 +468,7 @@ fh_update(struct svc_fh *fhp) } else { int size; if (fhp->fh_handle.fh_fileid_type != 0) - goto out_uptodate; + goto out; datap = fhp->fh_handle.fh_auth+ fhp->fh_handle.fh_size/4 -1; size = (fhp->fh_maxsize - fhp->fh_handle.fh_size)/4; @@ -481,10 +488,6 @@ out_negative: printk(KERN_ERR "fh_update: %s/%s still negative!\n", dentry->d_parent->d_name.name, dentry->d_name.name); goto out; -out_uptodate: - printk(KERN_ERR "fh_update: %s/%s already up-to-date!\n", - dentry->d_parent->d_name.name, dentry->d_name.name); - goto out; } /* diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c index 3790727e5df..ec1decf29ba 100644 --- a/fs/nfsd/nfssvc.c +++ b/fs/nfsd/nfssvc.c @@ -8,7 +8,6 @@ * Copyright (C) 1995, 1996, 1997 Olaf Kirch <okir@monad.swb.de> */ -#include <linux/config.h> #include <linux/module.h> #include <linux/time.h> diff --git a/fs/nfsd/stats.c b/fs/nfsd/stats.c index 57265d56380..71944cddf68 100644 --- a/fs/nfsd/stats.c +++ b/fs/nfsd/stats.c @@ -72,6 +72,16 @@ static int nfsd_proc_show(struct seq_file *seq, void *v) /* show my rpc info */ svc_seq_show(seq, &nfsd_svcstats); +#ifdef CONFIG_NFSD_V4 + /* Show count for individual nfsv4 operations */ + /* Writing operation numbers 0 1 2 also for maintaining uniformity */ + seq_printf(seq,"proc4ops %u", LAST_NFS4_OP + 1); + for (i = 0; i <= LAST_NFS4_OP; i++) + seq_printf(seq, " %u", nfsdstats.nfs4_opcount[i]); + + seq_putc(seq, '\n'); +#endif + return 0; } diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c index 245eaa1fb59..c9e3b5a8fe0 100644 --- a/fs/nfsd/vfs.c +++ b/fs/nfsd/vfs.c @@ -16,7 +16,6 @@ * Zerocpy NFS support (C) 2002 Hirokazu Takahashi <taka@valinux.co.jp> */ -#include <linux/config.h> #include <linux/string.h> #include <linux/time.h> #include <linux/errno.h> @@ -673,7 +672,10 @@ nfsd_open(struct svc_rqst *rqstp, struct svc_fh *fhp, int type, goto out_nfserr; if (access & MAY_WRITE) { - flags = O_WRONLY|O_LARGEFILE; + if (access & MAY_READ) + flags = O_RDWR|O_LARGEFILE; + else + flags = O_WRONLY|O_LARGEFILE; DQUOT_INIT(inode); } @@ -834,7 +836,7 @@ nfsd_vfs_read(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file, if (ra && ra->p_set) file->f_ra = ra->p_ra; - if (file->f_op->sendfile) { + if (file->f_op->sendfile && rqstp->rq_sendfile_ok) { svc_pushback_unused_pages(rqstp); err = file->f_op->sendfile(file, &offset, *count, nfsd_read_actor, rqstp); @@ -1517,14 +1519,15 @@ nfsd_link(struct svc_rqst *rqstp, struct svc_fh *ffhp, err = nfserrno(err); } - fh_unlock(ffhp); dput(dnew); +out_unlock: + fh_unlock(ffhp); out: return err; out_nfserr: err = nfserrno(err); - goto out; + goto out_unlock; } /* @@ -1553,7 +1556,7 @@ nfsd_rename(struct svc_rqst *rqstp, struct svc_fh *ffhp, char *fname, int flen, tdir = tdentry->d_inode; err = (rqstp->rq_vers == 2) ? nfserr_acces : nfserr_xdev; - if (fdir->i_sb != tdir->i_sb) + if (ffhp->fh_export != tfhp->fh_export) goto out; err = nfserr_perm; diff --git a/fs/nls/nls_base.c b/fs/nls/nls_base.c index a912debcd20..9de6b495f11 100644 --- a/fs/nls/nls_base.c +++ b/fs/nls/nls_base.c @@ -10,7 +10,6 @@ #include <linux/module.h> #include <linux/string.h> -#include <linux/config.h> #include <linux/nls.h> #include <linux/kernel.h> #include <linux/errno.h> diff --git a/fs/ntfs/aops.c b/fs/ntfs/aops.c index 580412d330c..bc579bfdfbd 100644 --- a/fs/ntfs/aops.c +++ b/fs/ntfs/aops.c @@ -1544,7 +1544,7 @@ err_out: /** * ntfs_aops - general address space operations for inodes and attributes */ -struct address_space_operations ntfs_aops = { +const struct address_space_operations ntfs_aops = { .readpage = ntfs_readpage, /* Fill page with data. */ .sync_page = block_sync_page, /* Currently, just unplugs the disk request queue. */ @@ -1560,7 +1560,7 @@ struct address_space_operations ntfs_aops = { * ntfs_mst_aops - general address space operations for mst protecteed inodes * and attributes */ -struct address_space_operations ntfs_mst_aops = { +const struct address_space_operations ntfs_mst_aops = { .readpage = ntfs_readpage, /* Fill page with data. */ .sync_page = block_sync_page, /* Currently, just unplugs the disk request queue. */ diff --git a/fs/ntfs/dir.c b/fs/ntfs/dir.c index d1e2c6f9f05..85c36b8ca45 100644 --- a/fs/ntfs/dir.c +++ b/fs/ntfs/dir.c @@ -1149,8 +1149,7 @@ static int ntfs_readdir(struct file *filp, void *dirent, filldir_t filldir) * Allocate a buffer to store the current name being processed * converted to format determined by current NLS. */ - name = (u8*)kmalloc(NTFS_MAX_NAME_LEN * NLS_MAX_CHARSET_SIZE + 1, - GFP_NOFS); + name = kmalloc(NTFS_MAX_NAME_LEN * NLS_MAX_CHARSET_SIZE + 1, GFP_NOFS); if (unlikely(!name)) { err = -ENOMEM; goto err_out; @@ -1191,7 +1190,7 @@ static int ntfs_readdir(struct file *filp, void *dirent, filldir_t filldir) * map the mft record without deadlocking. */ rc = le32_to_cpu(ctx->attr->data.resident.value_length); - ir = (INDEX_ROOT*)kmalloc(rc, GFP_NOFS); + ir = kmalloc(rc, GFP_NOFS); if (unlikely(!ir)) { err = -ENOMEM; goto err_out; diff --git a/fs/ntfs/file.c b/fs/ntfs/file.c index 88292f9e4b9..2e42c2dcae1 100644 --- a/fs/ntfs/file.c +++ b/fs/ntfs/file.c @@ -1358,7 +1358,7 @@ err_out: goto out; } -static size_t __ntfs_copy_from_user_iovec(char *vaddr, +static size_t __ntfs_copy_from_user_iovec_inatomic(char *vaddr, const struct iovec *iov, size_t iov_ofs, size_t bytes) { size_t total = 0; @@ -1376,10 +1376,6 @@ static size_t __ntfs_copy_from_user_iovec(char *vaddr, bytes -= len; vaddr += len; if (unlikely(left)) { - /* - * Zero the rest of the target like __copy_from_user(). - */ - memset(vaddr, 0, bytes); total -= left; break; } @@ -1420,11 +1416,13 @@ static inline void ntfs_set_next_iovec(const struct iovec **iovp, * pages (out to offset + bytes), to emulate ntfs_copy_from_user()'s * single-segment behaviour. * - * We call the same helper (__ntfs_copy_from_user_iovec()) both when atomic and - * when not atomic. This is ok because __ntfs_copy_from_user_iovec() calls - * __copy_from_user_inatomic() and it is ok to call this when non-atomic. In - * fact, the only difference between __copy_from_user_inatomic() and - * __copy_from_user() is that the latter calls might_sleep(). And on many + * We call the same helper (__ntfs_copy_from_user_iovec_inatomic()) both + * when atomic and when not atomic. This is ok because + * __ntfs_copy_from_user_iovec_inatomic() calls __copy_from_user_inatomic() + * and it is ok to call this when non-atomic. + * Infact, the only difference between __copy_from_user_inatomic() and + * __copy_from_user() is that the latter calls might_sleep() and the former + * should not zero the tail of the buffer on error. And on many * architectures __copy_from_user_inatomic() is just defined to * __copy_from_user() so it makes no difference at all on those architectures. */ @@ -1441,14 +1439,18 @@ static inline size_t ntfs_copy_from_user_iovec(struct page **pages, if (len > bytes) len = bytes; kaddr = kmap_atomic(*pages, KM_USER0); - copied = __ntfs_copy_from_user_iovec(kaddr + ofs, + copied = __ntfs_copy_from_user_iovec_inatomic(kaddr + ofs, *iov, *iov_ofs, len); kunmap_atomic(kaddr, KM_USER0); if (unlikely(copied != len)) { /* Do it the slow way. */ kaddr = kmap(*pages); - copied = __ntfs_copy_from_user_iovec(kaddr + ofs, + copied = __ntfs_copy_from_user_iovec_inatomic(kaddr + ofs, *iov, *iov_ofs, len); + /* + * Zero the rest of the target like __copy_from_user(). + */ + memset(kaddr + ofs + copied, 0, len - copied); kunmap(*pages); if (unlikely(copied != len)) goto err_out; diff --git a/fs/ntfs/inode.c b/fs/ntfs/inode.c index 4c86b7e1d1e..933dbd89c2a 100644 --- a/fs/ntfs/inode.c +++ b/fs/ntfs/inode.c @@ -137,7 +137,7 @@ static int ntfs_init_locked_inode(struct inode *vi, ntfs_attr *na) BUG_ON(!na->name); i = na->name_len * sizeof(ntfschar); - ni->name = (ntfschar*)kmalloc(i + sizeof(ntfschar), GFP_ATOMIC); + ni->name = kmalloc(i + sizeof(ntfschar), GFP_ATOMIC); if (!ni->name) return -ENOMEM; memcpy(ni->name, na->name, i); @@ -367,6 +367,12 @@ static void ntfs_destroy_extent_inode(ntfs_inode *ni) kmem_cache_free(ntfs_inode_cache, ni); } +/* + * The attribute runlist lock has separate locking rules from the + * normal runlist lock, so split the two lock-classes: + */ +static struct lock_class_key attr_list_rl_lock_class; + /** * __ntfs_init_inode - initialize ntfs specific part of an inode * @sb: super block of mounted volume @@ -394,6 +400,8 @@ void __ntfs_init_inode(struct super_block *sb, ntfs_inode *ni) ni->attr_list_size = 0; ni->attr_list = NULL; ntfs_init_runlist(&ni->attr_list_rl); + lockdep_set_class(&ni->attr_list_rl.lock, + &attr_list_rl_lock_class); ni->itype.index.bmp_ino = NULL; ni->itype.index.block_size = 0; ni->itype.index.vcn_size = 0; @@ -405,6 +413,13 @@ void __ntfs_init_inode(struct super_block *sb, ntfs_inode *ni) ni->ext.base_ntfs_ino = NULL; } +/* + * Extent inodes get MFT-mapped in a nested way, while the base inode + * is still mapped. Teach this nesting to the lock validator by creating + * a separate class for nested inode's mrec_lock's: + */ +static struct lock_class_key extent_inode_mrec_lock_key; + inline ntfs_inode *ntfs_new_extent_inode(struct super_block *sb, unsigned long mft_no) { @@ -413,6 +428,7 @@ inline ntfs_inode *ntfs_new_extent_inode(struct super_block *sb, ntfs_debug("Entering."); if (likely(ni != NULL)) { __ntfs_init_inode(sb, ni); + lockdep_set_class(&ni->mrec_lock, &extent_inode_mrec_lock_key); ni->mft_no = mft_no; ni->type = AT_UNUSED; ni->name = NULL; @@ -540,8 +556,6 @@ static int ntfs_read_locked_inode(struct inode *vi) /* Setup the generic vfs inode parts now. */ - /* This is the optimal IO size (for stat), not the fs block size. */ - vi->i_blksize = PAGE_CACHE_SIZE; /* * This is for checking whether an inode has changed w.r.t. a file so * that the file can be updated if necessary (compare with f_version). @@ -1218,7 +1232,6 @@ static int ntfs_read_locked_attr_inode(struct inode *base_vi, struct inode *vi) base_ni = NTFS_I(base_vi); /* Just mirror the values from the base inode. */ - vi->i_blksize = base_vi->i_blksize; vi->i_version = base_vi->i_version; vi->i_uid = base_vi->i_uid; vi->i_gid = base_vi->i_gid; @@ -1488,7 +1501,6 @@ static int ntfs_read_locked_index_inode(struct inode *base_vi, struct inode *vi) ni = NTFS_I(vi); base_ni = NTFS_I(base_vi); /* Just mirror the values from the base inode. */ - vi->i_blksize = base_vi->i_blksize; vi->i_version = base_vi->i_version; vi->i_uid = base_vi->i_uid; vi->i_gid = base_vi->i_gid; @@ -1722,6 +1734,15 @@ err_out: return err; } +/* + * The MFT inode has special locking, so teach the lock validator + * about this by splitting off the locking rules of the MFT from + * the locking rules of other inodes. The MFT inode can never be + * accessed from the VFS side (or even internally), only by the + * map_mft functions. + */ +static struct lock_class_key mft_ni_runlist_lock_key, mft_ni_mrec_lock_key; + /** * ntfs_read_inode_mount - special read_inode for mount time use only * @vi: inode to read @@ -2148,6 +2169,14 @@ int ntfs_read_inode_mount(struct inode *vi) ntfs_attr_put_search_ctx(ctx); ntfs_debug("Done."); ntfs_free(m); + + /* + * Split the locking rules of the MFT inode from the + * locking rules of other inodes: + */ + lockdep_set_class(&ni->runlist.lock, &mft_ni_runlist_lock_key); + lockdep_set_class(&ni->mrec_lock, &mft_ni_mrec_lock_key); + return 0; em_put_err_out: diff --git a/fs/ntfs/mft.c b/fs/ntfs/mft.c index 2438c00ec0c..584260fd684 100644 --- a/fs/ntfs/mft.c +++ b/fs/ntfs/mft.c @@ -331,7 +331,7 @@ map_err_out: ntfs_inode **tmp; int new_size = (base_ni->nr_extents + 4) * sizeof(ntfs_inode *); - tmp = (ntfs_inode **)kmalloc(new_size, GFP_NOFS); + tmp = kmalloc(new_size, GFP_NOFS); if (unlikely(!tmp)) { ntfs_error(base_ni->vol->sb, "Failed to allocate " "internal buffer."); @@ -2638,11 +2638,6 @@ mft_rec_already_initialized: } vi->i_ino = bit; /* - * This is the optimal IO size (for stat), not the fs block - * size. - */ - vi->i_blksize = PAGE_CACHE_SIZE; - /* * This is for checking whether an inode has changed w.r.t. a * file so that the file can be updated if necessary (compare * with f_version). @@ -2893,7 +2888,7 @@ rollback: if (!(base_ni->nr_extents & 3)) { int new_size = (base_ni->nr_extents + 4) * sizeof(ntfs_inode*); - extent_nis = (ntfs_inode**)kmalloc(new_size, GFP_NOFS); + extent_nis = kmalloc(new_size, GFP_NOFS); if (unlikely(!extent_nis)) { ntfs_error(vol->sb, "Failed to allocate internal " "buffer during rollback.%s", es); diff --git a/fs/ntfs/ntfs.h b/fs/ntfs/ntfs.h index bf7b3d7c093..ddd3d503097 100644 --- a/fs/ntfs/ntfs.h +++ b/fs/ntfs/ntfs.h @@ -57,8 +57,8 @@ extern struct kmem_cache *ntfs_attr_ctx_cache; extern struct kmem_cache *ntfs_index_ctx_cache; /* The various operations structs defined throughout the driver files. */ -extern struct address_space_operations ntfs_aops; -extern struct address_space_operations ntfs_mst_aops; +extern const struct address_space_operations ntfs_aops; +extern const struct address_space_operations ntfs_mst_aops; extern const struct file_operations ntfs_file_ops; extern struct inode_operations ntfs_file_inode_ops; diff --git a/fs/ntfs/super.c b/fs/ntfs/super.c index 0e14acea3f8..6b2712f10dd 100644 --- a/fs/ntfs/super.c +++ b/fs/ntfs/super.c @@ -1724,6 +1724,14 @@ upcase_failed: return FALSE; } +/* + * The lcn and mft bitmap inodes are NTFS-internal inodes with + * their own special locking rules: + */ +static struct lock_class_key + lcnbmp_runlist_lock_key, lcnbmp_mrec_lock_key, + mftbmp_runlist_lock_key, mftbmp_mrec_lock_key; + /** * load_system_files - open the system files using normal functions * @vol: ntfs super block describing device whose system files to load @@ -1780,6 +1788,10 @@ static BOOL load_system_files(ntfs_volume *vol) ntfs_error(sb, "Failed to load $MFT/$BITMAP attribute."); goto iput_mirr_err_out; } + lockdep_set_class(&NTFS_I(vol->mftbmp_ino)->runlist.lock, + &mftbmp_runlist_lock_key); + lockdep_set_class(&NTFS_I(vol->mftbmp_ino)->mrec_lock, + &mftbmp_mrec_lock_key); /* Read upcase table and setup @vol->upcase and @vol->upcase_len. */ if (!load_and_init_upcase(vol)) goto iput_mftbmp_err_out; @@ -1802,6 +1814,11 @@ static BOOL load_system_files(ntfs_volume *vol) iput(vol->lcnbmp_ino); goto bitmap_failed; } + lockdep_set_class(&NTFS_I(vol->lcnbmp_ino)->runlist.lock, + &lcnbmp_runlist_lock_key); + lockdep_set_class(&NTFS_I(vol->lcnbmp_ino)->mrec_lock, + &lcnbmp_mrec_lock_key); + NInoSetSparseDisabled(NTFS_I(vol->lcnbmp_ino)); if ((vol->nr_clusters + 7) >> 3 > i_size_read(vol->lcnbmp_ino)) { iput(vol->lcnbmp_ino); @@ -2743,6 +2760,17 @@ static int ntfs_fill_super(struct super_block *sb, void *opt, const int silent) struct inode *tmp_ino; int blocksize, result; + /* + * We do a pretty difficult piece of bootstrap by reading the + * MFT (and other metadata) from disk into memory. We'll only + * release this metadata during umount, so the locking patterns + * observed during bootstrap do not count. So turn off the + * observation of locking patterns (strictly for this context + * only) while mounting NTFS. [The validator is still active + * otherwise, even for this context: it will for example record + * lock class registrations.] + */ + lockdep_off(); ntfs_debug("Entering."); #ifndef NTFS_RW sb->s_flags |= MS_RDONLY; @@ -2754,6 +2782,7 @@ static int ntfs_fill_super(struct super_block *sb, void *opt, const int silent) if (!silent) ntfs_error(sb, "Allocation of NTFS volume structure " "failed. Aborting mount..."); + lockdep_on(); return -ENOMEM; } /* Initialize ntfs_volume structure. */ @@ -2940,6 +2969,7 @@ static int ntfs_fill_super(struct super_block *sb, void *opt, const int silent) mutex_unlock(&ntfs_lock); sb->s_export_op = &ntfs_export_ops; lock_kernel(); + lockdep_on(); return 0; } ntfs_error(sb, "Failed to allocate root directory."); @@ -3059,6 +3089,7 @@ err_out_now: sb->s_fs_info = NULL; kfree(vol); ntfs_debug("Failed, returning -EINVAL."); + lockdep_on(); return -EINVAL; } @@ -3217,32 +3248,14 @@ ictx_err_out: static void __exit exit_ntfs_fs(void) { - int err = 0; - ntfs_debug("Unregistering NTFS driver."); unregister_filesystem(&ntfs_fs_type); - - if (kmem_cache_destroy(ntfs_big_inode_cache) && (err = 1)) - printk(KERN_CRIT "NTFS: Failed to destory %s.\n", - ntfs_big_inode_cache_name); - if (kmem_cache_destroy(ntfs_inode_cache) && (err = 1)) - printk(KERN_CRIT "NTFS: Failed to destory %s.\n", - ntfs_inode_cache_name); - if (kmem_cache_destroy(ntfs_name_cache) && (err = 1)) - printk(KERN_CRIT "NTFS: Failed to destory %s.\n", - ntfs_name_cache_name); - if (kmem_cache_destroy(ntfs_attr_ctx_cache) && (err = 1)) - printk(KERN_CRIT "NTFS: Failed to destory %s.\n", - ntfs_attr_ctx_cache_name); - if (kmem_cache_destroy(ntfs_index_ctx_cache) && (err = 1)) - printk(KERN_CRIT "NTFS: Failed to destory %s.\n", - ntfs_index_ctx_cache_name); - if (err) - printk(KERN_CRIT "NTFS: This causes memory to leak! There is " - "probably a BUG in the driver! Please report " - "you saw this message to " - "linux-ntfs-dev@lists.sourceforge.net\n"); + kmem_cache_destroy(ntfs_big_inode_cache); + kmem_cache_destroy(ntfs_inode_cache); + kmem_cache_destroy(ntfs_name_cache); + kmem_cache_destroy(ntfs_attr_ctx_cache); + kmem_cache_destroy(ntfs_index_ctx_cache); /* Unregister the ntfs sysctls. */ ntfs_sysctl(0); } diff --git a/fs/ntfs/sysctl.h b/fs/ntfs/sysctl.h index c8064cae8f1..beda5bf9640 100644 --- a/fs/ntfs/sysctl.h +++ b/fs/ntfs/sysctl.h @@ -24,7 +24,6 @@ #ifndef _LINUX_NTFS_SYSCTL_H #define _LINUX_NTFS_SYSCTL_H -#include <linux/config.h> #if defined(DEBUG) && defined(CONFIG_SYSCTL) diff --git a/fs/ntfs/unistr.c b/fs/ntfs/unistr.c index b123c0fa6bf..a1b572196fe 100644 --- a/fs/ntfs/unistr.c +++ b/fs/ntfs/unistr.c @@ -350,7 +350,7 @@ int ntfs_ucstonls(const ntfs_volume *vol, const ntfschar *ins, } if (!ns) { ns_len = ins_len * NLS_MAX_CHARSET_SIZE; - ns = (unsigned char*)kmalloc(ns_len + 1, GFP_NOFS); + ns = kmalloc(ns_len + 1, GFP_NOFS); if (!ns) goto mem_err_out; } @@ -365,7 +365,7 @@ retry: wc = nls->uni2char(le16_to_cpu(ins[i]), ns + o, else if (wc == -ENAMETOOLONG && ns != *outs) { unsigned char *tc; /* Grow in multiples of 64 bytes. */ - tc = (unsigned char*)kmalloc((ns_len + 64) & + tc = kmalloc((ns_len + 64) & ~63, GFP_NOFS); if (tc) { memcpy(tc, ns, ns_len); diff --git a/fs/ocfs2/Makefile b/fs/ocfs2/Makefile index 7d3be845a61..9fb8132f19b 100644 --- a/fs/ocfs2/Makefile +++ b/fs/ocfs2/Makefile @@ -16,6 +16,7 @@ ocfs2-objs := \ file.o \ heartbeat.o \ inode.o \ + ioctl.o \ journal.o \ localalloc.o \ mmap.o \ diff --git a/fs/ocfs2/alloc.c b/fs/ocfs2/alloc.c index edaab05a93e..f43bc5f18a3 100644 --- a/fs/ocfs2/alloc.c +++ b/fs/ocfs2/alloc.c @@ -1717,17 +1717,29 @@ static int ocfs2_do_truncate(struct ocfs2_super *osb, ocfs2_remove_from_cache(inode, eb_bh); - BUG_ON(eb->h_suballoc_slot); BUG_ON(el->l_recs[0].e_clusters); BUG_ON(el->l_recs[0].e_cpos); BUG_ON(el->l_recs[0].e_blkno); - status = ocfs2_free_extent_block(handle, - tc->tc_ext_alloc_inode, - tc->tc_ext_alloc_bh, - eb); - if (status < 0) { - mlog_errno(status); - goto bail; + if (eb->h_suballoc_slot == 0) { + /* + * This code only understands how to + * lock the suballocator in slot 0, + * which is fine because allocation is + * only ever done out of that + * suballocator too. A future version + * might change that however, so avoid + * a free if we don't know how to + * handle it. This way an fs incompat + * bit will not be necessary. + */ + status = ocfs2_free_extent_block(handle, + tc->tc_ext_alloc_inode, + tc->tc_ext_alloc_bh, + eb); + if (status < 0) { + mlog_errno(status); + goto bail; + } } } brelse(eb_bh); diff --git a/fs/ocfs2/aops.c b/fs/ocfs2/aops.c index 47152bf9a7f..3d7c082a8f5 100644 --- a/fs/ocfs2/aops.c +++ b/fs/ocfs2/aops.c @@ -391,31 +391,28 @@ out: static int ocfs2_commit_write(struct file *file, struct page *page, unsigned from, unsigned to) { - int ret, extending = 0, locklevel = 0; - loff_t new_i_size; + int ret; struct buffer_head *di_bh = NULL; struct inode *inode = page->mapping->host; struct ocfs2_journal_handle *handle = NULL; + struct ocfs2_dinode *di; mlog_entry("(0x%p, 0x%p, %u, %u)\n", file, page, from, to); /* NOTE: ocfs2_file_aio_write has ensured that it's safe for - * us to sample inode->i_size here without the metadata lock: + * us to continue here without rechecking the I/O against + * changed inode values. * * 1) We're currently holding the inode alloc lock, so no * nodes can change it underneath us. * * 2) We've had to take the metadata lock at least once - * already to check for extending writes, hence insuring - * that our current copy is also up to date. + * already to check for extending writes, suid removal, etc. + * The meta data update code then ensures that we don't get a + * stale inode allocation image (i_size, i_clusters, etc). */ - new_i_size = ((loff_t)page->index << PAGE_CACHE_SHIFT) + to; - if (new_i_size > i_size_read(inode)) { - extending = 1; - locklevel = 1; - } - ret = ocfs2_meta_lock_with_page(inode, NULL, &di_bh, locklevel, page); + ret = ocfs2_meta_lock_with_page(inode, NULL, &di_bh, 1, page); if (ret != 0) { mlog_errno(ret); goto out; @@ -427,23 +424,20 @@ static int ocfs2_commit_write(struct file *file, struct page *page, goto out_unlock_meta; } - if (extending) { - handle = ocfs2_start_walk_page_trans(inode, page, from, to); - if (IS_ERR(handle)) { - ret = PTR_ERR(handle); - handle = NULL; - goto out_unlock_data; - } + handle = ocfs2_start_walk_page_trans(inode, page, from, to); + if (IS_ERR(handle)) { + ret = PTR_ERR(handle); + goto out_unlock_data; + } - /* Mark our buffer early. We'd rather catch this error up here - * as opposed to after a successful commit_write which would - * require us to set back inode->i_size. */ - ret = ocfs2_journal_access(handle, inode, di_bh, - OCFS2_JOURNAL_ACCESS_WRITE); - if (ret < 0) { - mlog_errno(ret); - goto out_commit; - } + /* Mark our buffer early. We'd rather catch this error up here + * as opposed to after a successful commit_write which would + * require us to set back inode->i_size. */ + ret = ocfs2_journal_access(handle, inode, di_bh, + OCFS2_JOURNAL_ACCESS_WRITE); + if (ret < 0) { + mlog_errno(ret); + goto out_commit; } /* might update i_size */ @@ -453,37 +447,28 @@ static int ocfs2_commit_write(struct file *file, struct page *page, goto out_commit; } - if (extending) { - loff_t size = (u64) i_size_read(inode); - struct ocfs2_dinode *di = - (struct ocfs2_dinode *)di_bh->b_data; + di = (struct ocfs2_dinode *)di_bh->b_data; - /* ocfs2_mark_inode_dirty is too heavy to use here. */ - inode->i_blocks = ocfs2_align_bytes_to_sectors(size); - inode->i_ctime = inode->i_mtime = CURRENT_TIME; + /* ocfs2_mark_inode_dirty() is too heavy to use here. */ + inode->i_mtime = inode->i_ctime = CURRENT_TIME; + di->i_mtime = di->i_ctime = cpu_to_le64(inode->i_mtime.tv_sec); + di->i_mtime_nsec = di->i_ctime_nsec = cpu_to_le32(inode->i_mtime.tv_nsec); - di->i_size = cpu_to_le64(size); - di->i_ctime = di->i_mtime = - cpu_to_le64(inode->i_mtime.tv_sec); - di->i_ctime_nsec = di->i_mtime_nsec = - cpu_to_le32(inode->i_mtime.tv_nsec); + inode->i_blocks = ocfs2_align_bytes_to_sectors((u64)(i_size_read(inode))); + di->i_size = cpu_to_le64((u64)i_size_read(inode)); - ret = ocfs2_journal_dirty(handle, di_bh); - if (ret < 0) { - mlog_errno(ret); - goto out_commit; - } + ret = ocfs2_journal_dirty(handle, di_bh); + if (ret < 0) { + mlog_errno(ret); + goto out_commit; } - BUG_ON(extending && (i_size_read(inode) != new_i_size)); - out_commit: - if (handle) - ocfs2_commit_trans(handle); + ocfs2_commit_trans(handle); out_unlock_data: ocfs2_data_unlock(inode, 1); out_unlock_meta: - ocfs2_meta_unlock(inode, locklevel); + ocfs2_meta_unlock(inode, 1); out: if (di_bh) brelse(di_bh); @@ -558,16 +543,9 @@ static int ocfs2_direct_IO_get_blocks(struct inode *inode, sector_t iblock, u64 vbo_max; /* file offset, max_blocks from iblock */ u64 p_blkno; int contig_blocks; - unsigned char blocksize_bits; + unsigned char blocksize_bits = inode->i_sb->s_blocksize_bits; unsigned long max_blocks = bh_result->b_size >> inode->i_blkbits; - if (!inode || !bh_result) { - mlog(ML_ERROR, "inode or bh_result is null\n"); - return -EIO; - } - - blocksize_bits = inode->i_sb->s_blocksize_bits; - /* This function won't even be called if the request isn't all * nicely aligned and of the right size, so there's no need * for us to check any of that. */ @@ -666,7 +644,7 @@ out: return ret; } -struct address_space_operations ocfs2_aops = { +const struct address_space_operations ocfs2_aops = { .readpage = ocfs2_readpage, .writepage = ocfs2_writepage, .prepare_write = ocfs2_prepare_write, diff --git a/fs/ocfs2/buffer_head_io.c b/fs/ocfs2/buffer_head_io.c index 9a24adf9be6..c9037414f4f 100644 --- a/fs/ocfs2/buffer_head_io.c +++ b/fs/ocfs2/buffer_head_io.c @@ -100,6 +100,9 @@ int ocfs2_read_blocks(struct ocfs2_super *osb, u64 block, int nr, mlog_entry("(block=(%llu), nr=(%d), flags=%d, inode=%p)\n", (unsigned long long)block, nr, flags, inode); + BUG_ON((flags & OCFS2_BH_READAHEAD) && + (!inode || !(flags & OCFS2_BH_CACHED))); + if (osb == NULL || osb->sb == NULL || bhs == NULL) { status = -EINVAL; mlog_errno(status); @@ -140,6 +143,30 @@ int ocfs2_read_blocks(struct ocfs2_super *osb, u64 block, int nr, bh = bhs[i]; ignore_cache = 0; + /* There are three read-ahead cases here which we need to + * be concerned with. All three assume a buffer has + * previously been submitted with OCFS2_BH_READAHEAD + * and it hasn't yet completed I/O. + * + * 1) The current request is sync to disk. This rarely + * happens these days, and never when performance + * matters - the code can just wait on the buffer + * lock and re-submit. + * + * 2) The current request is cached, but not + * readahead. ocfs2_buffer_uptodate() will return + * false anyway, so we'll wind up waiting on the + * buffer lock to do I/O. We re-check the request + * with after getting the lock to avoid a re-submit. + * + * 3) The current request is readahead (and so must + * also be a caching one). We short circuit if the + * buffer is locked (under I/O) and if it's in the + * uptodate cache. The re-check from #2 catches the + * case that the previous read-ahead completes just + * before our is-it-in-flight check. + */ + if (flags & OCFS2_BH_CACHED && !ocfs2_buffer_uptodate(inode, bh)) { mlog(ML_UPTODATE, @@ -169,6 +196,14 @@ int ocfs2_read_blocks(struct ocfs2_super *osb, u64 block, int nr, continue; } + /* A read-ahead request was made - if the + * buffer is already under read-ahead from a + * previously submitted request than we are + * done here. */ + if ((flags & OCFS2_BH_READAHEAD) + && ocfs2_buffer_read_ahead(inode, bh)) + continue; + lock_buffer(bh); if (buffer_jbd(bh)) { #ifdef CATCH_BH_JBD_RACES @@ -181,13 +216,22 @@ int ocfs2_read_blocks(struct ocfs2_super *osb, u64 block, int nr, continue; #endif } + + /* Re-check ocfs2_buffer_uptodate() as a + * previously read-ahead buffer may have + * completed I/O while we were waiting for the + * buffer lock. */ + if ((flags & OCFS2_BH_CACHED) + && !(flags & OCFS2_BH_READAHEAD) + && ocfs2_buffer_uptodate(inode, bh)) { + unlock_buffer(bh); + continue; + } + clear_buffer_uptodate(bh); get_bh(bh); /* for end_buffer_read_sync() */ bh->b_end_io = end_buffer_read_sync; - if (flags & OCFS2_BH_READAHEAD) - submit_bh(READA, bh); - else - submit_bh(READ, bh); + submit_bh(READ, bh); continue; } } @@ -197,34 +241,39 @@ int ocfs2_read_blocks(struct ocfs2_super *osb, u64 block, int nr, for (i = (nr - 1); i >= 0; i--) { bh = bhs[i]; - /* We know this can't have changed as we hold the - * inode sem. Avoid doing any work on the bh if the - * journal has it. */ - if (!buffer_jbd(bh)) - wait_on_buffer(bh); - - if (!buffer_uptodate(bh)) { - /* Status won't be cleared from here on out, - * so we can safely record this and loop back - * to cleanup the other buffers. Don't need to - * remove the clustered uptodate information - * for this bh as it's not marked locally - * uptodate. */ - status = -EIO; - brelse(bh); - bhs[i] = NULL; - continue; + if (!(flags & OCFS2_BH_READAHEAD)) { + /* We know this can't have changed as we hold the + * inode sem. Avoid doing any work on the bh if the + * journal has it. */ + if (!buffer_jbd(bh)) + wait_on_buffer(bh); + + if (!buffer_uptodate(bh)) { + /* Status won't be cleared from here on out, + * so we can safely record this and loop back + * to cleanup the other buffers. Don't need to + * remove the clustered uptodate information + * for this bh as it's not marked locally + * uptodate. */ + status = -EIO; + brelse(bh); + bhs[i] = NULL; + continue; + } } + /* Always set the buffer in the cache, even if it was + * a forced read, or read-ahead which hasn't yet + * completed. */ if (inode) ocfs2_set_buffer_uptodate(inode, bh); } if (inode) mutex_unlock(&OCFS2_I(inode)->ip_io_mutex); - mlog(ML_BH_IO, "block=(%llu), nr=(%d), cached=%s\n", + mlog(ML_BH_IO, "block=(%llu), nr=(%d), cached=%s, flags=0x%x\n", (unsigned long long)block, nr, - (!(flags & OCFS2_BH_CACHED) || ignore_cache) ? "no" : "yes"); + (!(flags & OCFS2_BH_CACHED) || ignore_cache) ? "no" : "yes", flags); bail: diff --git a/fs/ocfs2/buffer_head_io.h b/fs/ocfs2/buffer_head_io.h index 6ecb90937b6..6cc20930fac 100644 --- a/fs/ocfs2/buffer_head_io.h +++ b/fs/ocfs2/buffer_head_io.h @@ -49,7 +49,7 @@ int ocfs2_read_blocks(struct ocfs2_super *osb, #define OCFS2_BH_CACHED 1 -#define OCFS2_BH_READAHEAD 8 /* use this to pass READA down to submit_bh */ +#define OCFS2_BH_READAHEAD 8 static inline int ocfs2_read_block(struct ocfs2_super * osb, u64 off, struct buffer_head **bh, int flags, diff --git a/fs/ocfs2/cluster/heartbeat.c b/fs/ocfs2/cluster/heartbeat.c index 21f38accd03..305cba3681f 100644 --- a/fs/ocfs2/cluster/heartbeat.c +++ b/fs/ocfs2/cluster/heartbeat.c @@ -54,7 +54,7 @@ static DECLARE_RWSEM(o2hb_callback_sem); * multiple hb threads are watching multiple regions. A node is live * whenever any of the threads sees activity from the node in its region. */ -static spinlock_t o2hb_live_lock = SPIN_LOCK_UNLOCKED; +static DEFINE_SPINLOCK(o2hb_live_lock); static struct list_head o2hb_live_slots[O2NM_MAX_NODES]; static unsigned long o2hb_live_node_bitmap[BITS_TO_LONGS(O2NM_MAX_NODES)]; static LIST_HEAD(o2hb_node_events); @@ -320,8 +320,12 @@ static int compute_max_sectors(struct block_device *bdev) max_pages = q->max_hw_segments; max_pages--; /* Handle I/Os that straddle a page */ - max_sectors = max_pages << (PAGE_SHIFT - 9); - + if (max_pages) { + max_sectors = max_pages << (PAGE_SHIFT - 9); + } else { + /* If BIO contains 1 or less than 1 page. */ + max_sectors = q->max_sectors; + } /* Why is fls() 1-based???? */ pow_two_sectors = 1 << (fls(max_sectors) - 1); @@ -517,6 +521,7 @@ static inline void o2hb_prepare_block(struct o2hb_region *reg, hb_block->hb_seq = cpu_to_le64(cputime); hb_block->hb_node = node_num; hb_block->hb_generation = cpu_to_le64(generation); + hb_block->hb_dead_ms = cpu_to_le32(o2hb_dead_threshold * O2HB_REGION_TIMEOUT_MS); /* This step must always happen last! */ hb_block->hb_cksum = cpu_to_le32(o2hb_compute_block_crc_le(reg, @@ -645,6 +650,8 @@ static int o2hb_check_slot(struct o2hb_region *reg, struct o2nm_node *node; struct o2hb_disk_heartbeat_block *hb_block = reg->hr_tmp_block; u64 cputime; + unsigned int dead_ms = o2hb_dead_threshold * O2HB_REGION_TIMEOUT_MS; + unsigned int slot_dead_ms; memcpy(hb_block, slot->ds_raw_block, reg->hr_block_bytes); @@ -733,6 +740,23 @@ fire_callbacks: &o2hb_live_slots[slot->ds_node_num]); slot->ds_equal_samples = 0; + + /* We want to be sure that all nodes agree on the + * number of milliseconds before a node will be + * considered dead. The self-fencing timeout is + * computed from this value, and a discrepancy might + * result in heartbeat calling a node dead when it + * hasn't self-fenced yet. */ + slot_dead_ms = le32_to_cpu(hb_block->hb_dead_ms); + if (slot_dead_ms && slot_dead_ms != dead_ms) { + /* TODO: Perhaps we can fail the region here. */ + mlog(ML_ERROR, "Node %d on device %s has a dead count " + "of %u ms, but our count is %u ms.\n" + "Please double check your configuration values " + "for 'O2CB_HEARTBEAT_THRESHOLD'\n", + slot->ds_node_num, reg->hr_dev_name, slot_dead_ms, + dead_ms); + } goto out; } diff --git a/fs/ocfs2/cluster/masklog.h b/fs/ocfs2/cluster/masklog.h index 73edad78253..a42628ba9dd 100644 --- a/fs/ocfs2/cluster/masklog.h +++ b/fs/ocfs2/cluster/masklog.h @@ -123,6 +123,17 @@ #define MLOG_MASK_PREFIX 0 #endif +/* + * When logging is disabled, force the bit test to 0 for anything other + * than errors and notices, allowing gcc to remove the code completely. + * When enabled, allow all masks. + */ +#if defined(CONFIG_OCFS2_DEBUG_MASKLOG) +#define ML_ALLOWED_BITS ~0 +#else +#define ML_ALLOWED_BITS (ML_ERROR|ML_NOTICE) +#endif + #define MLOG_MAX_BITS 64 struct mlog_bits { @@ -187,7 +198,8 @@ extern struct mlog_bits mlog_and_bits, mlog_not_bits; #define mlog(mask, fmt, args...) do { \ u64 __m = MLOG_MASK_PREFIX | (mask); \ - if (__mlog_test_u64(__m, mlog_and_bits) && \ + if ((__m & ML_ALLOWED_BITS) && \ + __mlog_test_u64(__m, mlog_and_bits) && \ !__mlog_test_u64(__m, mlog_not_bits)) { \ if (__m & ML_ERROR) \ __mlog_printk(KERN_ERR, "ERROR: "fmt , ##args); \ @@ -204,6 +216,7 @@ extern struct mlog_bits mlog_and_bits, mlog_not_bits; mlog(ML_ERROR, "status = %lld\n", (long long)_st); \ } while (0) +#if defined(CONFIG_OCFS2_DEBUG_MASKLOG) #define mlog_entry(fmt, args...) do { \ mlog(ML_ENTRY, "ENTRY:" fmt , ##args); \ } while (0) @@ -247,6 +260,13 @@ extern struct mlog_bits mlog_and_bits, mlog_not_bits; #define mlog_exit_void() do { \ mlog(ML_EXIT, "EXIT\n"); \ } while (0) +#else +#define mlog_entry(...) do { } while (0) +#define mlog_entry_void(...) do { } while (0) +#define mlog_exit(...) do { } while (0) +#define mlog_exit_ptr(...) do { } while (0) +#define mlog_exit_void(...) do { } while (0) +#endif /* defined(CONFIG_OCFS2_DEBUG_MASKLOG) */ #define mlog_bug_on_msg(cond, fmt, args...) do { \ if (cond) { \ diff --git a/fs/ocfs2/cluster/ocfs2_heartbeat.h b/fs/ocfs2/cluster/ocfs2_heartbeat.h index 94096069cb4..3f4151da970 100644 --- a/fs/ocfs2/cluster/ocfs2_heartbeat.h +++ b/fs/ocfs2/cluster/ocfs2_heartbeat.h @@ -32,6 +32,7 @@ struct o2hb_disk_heartbeat_block { __u8 hb_pad1[3]; __le32 hb_cksum; __le64 hb_generation; + __le32 hb_dead_ms; }; #endif /* _OCFS2_HEARTBEAT_H */ diff --git a/fs/ocfs2/cluster/tcp.c b/fs/ocfs2/cluster/tcp.c index 0f60cc0d398..b650efa8c8b 100644 --- a/fs/ocfs2/cluster/tcp.c +++ b/fs/ocfs2/cluster/tcp.c @@ -108,7 +108,7 @@ ##args); \ } while (0) -static rwlock_t o2net_handler_lock = RW_LOCK_UNLOCKED; +static DEFINE_RWLOCK(o2net_handler_lock); static struct rb_root o2net_handler_tree = RB_ROOT; static struct o2net_node o2net_nodes[O2NM_MAX_NODES]; @@ -396,8 +396,8 @@ static void o2net_set_nn_state(struct o2net_node *nn, } if (was_valid && !valid) { - mlog(ML_NOTICE, "no longer connected to " SC_NODEF_FMT "\n", - SC_NODEF_ARGS(old_sc)); + printk(KERN_INFO "o2net: no longer connected to " + SC_NODEF_FMT "\n", SC_NODEF_ARGS(old_sc)); o2net_complete_nodes_nsw(nn); } @@ -409,10 +409,10 @@ static void o2net_set_nn_state(struct o2net_node *nn, * the only way to start connecting again is to down * heartbeat and bring it back up. */ cancel_delayed_work(&nn->nn_connect_expired); - mlog(ML_NOTICE, "%s " SC_NODEF_FMT "\n", - o2nm_this_node() > sc->sc_node->nd_num ? - "connected to" : "accepted connection from", - SC_NODEF_ARGS(sc)); + printk(KERN_INFO "o2net: %s " SC_NODEF_FMT "\n", + o2nm_this_node() > sc->sc_node->nd_num ? + "connected to" : "accepted connection from", + SC_NODEF_ARGS(sc)); } /* trigger the connecting worker func as long as we're not valid, @@ -1280,7 +1280,7 @@ static void o2net_idle_timer(unsigned long data) do_gettimeofday(&now); - mlog(ML_NOTICE, "connection to " SC_NODEF_FMT " has been idle for 10 " + printk(KERN_INFO "o2net: connection to " SC_NODEF_FMT " has been idle for 10 " "seconds, shutting it down.\n", SC_NODEF_ARGS(sc)); mlog(ML_NOTICE, "here are some times that might help debug the " "situation: (tmr %ld.%ld now %ld.%ld dr %ld.%ld adv " diff --git a/fs/ocfs2/cluster/tcp_internal.h b/fs/ocfs2/cluster/tcp_internal.h index ff9e2e2104c..4b46aac7d24 100644 --- a/fs/ocfs2/cluster/tcp_internal.h +++ b/fs/ocfs2/cluster/tcp_internal.h @@ -44,11 +44,17 @@ * locking semantics of the file system using the protocol. It should * be somewhere else, I'm sure, but right now it isn't. * + * New in version 4: + * - Remove i_generation from lock names for better stat performance. + * + * New in version 3: + * - Replace dentry votes with a cluster lock + * * New in version 2: * - full 64 bit i_size in the metadata lock lvbs * - introduction of "rw" lock and pushing meta/data locking down */ -#define O2NET_PROTOCOL_VERSION 2ULL +#define O2NET_PROTOCOL_VERSION 4ULL struct o2net_handshake { __be64 protocol_version; __be64 connector_id; diff --git a/fs/ocfs2/dcache.c b/fs/ocfs2/dcache.c index 1a01380e387..014e73978da 100644 --- a/fs/ocfs2/dcache.c +++ b/fs/ocfs2/dcache.c @@ -35,15 +35,17 @@ #include "alloc.h" #include "dcache.h" +#include "dlmglue.h" #include "file.h" #include "inode.h" + static int ocfs2_dentry_revalidate(struct dentry *dentry, struct nameidata *nd) { struct inode *inode = dentry->d_inode; int ret = 0; /* if all else fails, just return false */ - struct ocfs2_super *osb; + struct ocfs2_super *osb = OCFS2_SB(dentry->d_sb); mlog_entry("(0x%p, '%.*s')\n", dentry, dentry->d_name.len, dentry->d_name.name); @@ -55,28 +57,31 @@ static int ocfs2_dentry_revalidate(struct dentry *dentry, goto bail; } - osb = OCFS2_SB(inode->i_sb); - BUG_ON(!osb); - if (inode != osb->root_inode) { - spin_lock(&OCFS2_I(inode)->ip_lock); - /* did we or someone else delete this inode? */ - if (OCFS2_I(inode)->ip_flags & OCFS2_INODE_DELETED) { - spin_unlock(&OCFS2_I(inode)->ip_lock); - mlog(0, "inode (%llu) deleted, returning false\n", - (unsigned long long)OCFS2_I(inode)->ip_blkno); - goto bail; - } + if (inode == osb->root_inode || is_bad_inode(inode)) + goto bail; + + spin_lock(&OCFS2_I(inode)->ip_lock); + /* did we or someone else delete this inode? */ + if (OCFS2_I(inode)->ip_flags & OCFS2_INODE_DELETED) { spin_unlock(&OCFS2_I(inode)->ip_lock); + mlog(0, "inode (%llu) deleted, returning false\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno); + goto bail; + } + spin_unlock(&OCFS2_I(inode)->ip_lock); - if (!inode->i_nlink) { - mlog(0, "Inode %llu orphaned, returning false " - "dir = %d\n", - (unsigned long long)OCFS2_I(inode)->ip_blkno, - S_ISDIR(inode->i_mode)); - goto bail; - } + /* + * We don't need a cluster lock to test this because once an + * inode nlink hits zero, it never goes back. + */ + if (inode->i_nlink == 0) { + mlog(0, "Inode %llu orphaned, returning false " + "dir = %d\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno, + S_ISDIR(inode->i_mode)); + goto bail; } ret = 1; @@ -87,6 +92,322 @@ bail: return ret; } +static int ocfs2_match_dentry(struct dentry *dentry, + u64 parent_blkno, + int skip_unhashed) +{ + struct inode *parent; + + /* + * ocfs2_lookup() does a d_splice_alias() _before_ attaching + * to the lock data, so we skip those here, otherwise + * ocfs2_dentry_attach_lock() will get its original dentry + * back. + */ + if (!dentry->d_fsdata) + return 0; + + if (!dentry->d_parent) + return 0; + + if (skip_unhashed && d_unhashed(dentry)) + return 0; + + parent = dentry->d_parent->d_inode; + /* Negative parent dentry? */ + if (!parent) + return 0; + + /* Name is in a different directory. */ + if (OCFS2_I(parent)->ip_blkno != parent_blkno) + return 0; + + return 1; +} + +/* + * Walk the inode alias list, and find a dentry which has a given + * parent. ocfs2_dentry_attach_lock() wants to find _any_ alias as it + * is looking for a dentry_lock reference. The vote thread is looking + * to unhash aliases, so we allow it to skip any that already have + * that property. + */ +struct dentry *ocfs2_find_local_alias(struct inode *inode, + u64 parent_blkno, + int skip_unhashed) +{ + struct list_head *p; + struct dentry *dentry = NULL; + + spin_lock(&dcache_lock); + + list_for_each(p, &inode->i_dentry) { + dentry = list_entry(p, struct dentry, d_alias); + + if (ocfs2_match_dentry(dentry, parent_blkno, skip_unhashed)) { + mlog(0, "dentry found: %.*s\n", + dentry->d_name.len, dentry->d_name.name); + + dget_locked(dentry); + break; + } + + dentry = NULL; + } + + spin_unlock(&dcache_lock); + + return dentry; +} + +DEFINE_SPINLOCK(dentry_attach_lock); + +/* + * Attach this dentry to a cluster lock. + * + * Dentry locks cover all links in a given directory to a particular + * inode. We do this so that ocfs2 can build a lock name which all + * nodes in the cluster can agree on at all times. Shoving full names + * in the cluster lock won't work due to size restrictions. Covering + * links inside of a directory is a good compromise because it still + * allows us to use the parent directory lock to synchronize + * operations. + * + * Call this function with the parent dir semaphore and the parent dir + * cluster lock held. + * + * The dir semaphore will protect us from having to worry about + * concurrent processes on our node trying to attach a lock at the + * same time. + * + * The dir cluster lock (held at either PR or EX mode) protects us + * from unlink and rename on other nodes. + * + * A dput() can happen asynchronously due to pruning, so we cover + * attaching and detaching the dentry lock with a + * dentry_attach_lock. + * + * A node which has done lookup on a name retains a protected read + * lock until final dput. If the user requests and unlink or rename, + * the protected read is upgraded to an exclusive lock. Other nodes + * who have seen the dentry will then be informed that they need to + * downgrade their lock, which will involve d_delete on the + * dentry. This happens in ocfs2_dentry_convert_worker(). + */ +int ocfs2_dentry_attach_lock(struct dentry *dentry, + struct inode *inode, + u64 parent_blkno) +{ + int ret; + struct dentry *alias; + struct ocfs2_dentry_lock *dl = dentry->d_fsdata; + + mlog(0, "Attach \"%.*s\", parent %llu, fsdata: %p\n", + dentry->d_name.len, dentry->d_name.name, + (unsigned long long)parent_blkno, dl); + + /* + * Negative dentry. We ignore these for now. + * + * XXX: Could we can improve ocfs2_dentry_revalidate() by + * tracking these? + */ + if (!inode) + return 0; + + if (dl) { + mlog_bug_on_msg(dl->dl_parent_blkno != parent_blkno, + " \"%.*s\": old parent: %llu, new: %llu\n", + dentry->d_name.len, dentry->d_name.name, + (unsigned long long)parent_blkno, + (unsigned long long)dl->dl_parent_blkno); + return 0; + } + + alias = ocfs2_find_local_alias(inode, parent_blkno, 0); + if (alias) { + /* + * Great, an alias exists, which means we must have a + * dentry lock already. We can just grab the lock off + * the alias and add it to the list. + * + * We're depending here on the fact that this dentry + * was found and exists in the dcache and so must have + * a reference to the dentry_lock because we can't + * race creates. Final dput() cannot happen on it + * since we have it pinned, so our reference is safe. + */ + dl = alias->d_fsdata; + mlog_bug_on_msg(!dl, "parent %llu, ino %llu\n", + (unsigned long long)parent_blkno, + (unsigned long long)OCFS2_I(inode)->ip_blkno); + + mlog_bug_on_msg(dl->dl_parent_blkno != parent_blkno, + " \"%.*s\": old parent: %llu, new: %llu\n", + dentry->d_name.len, dentry->d_name.name, + (unsigned long long)parent_blkno, + (unsigned long long)dl->dl_parent_blkno); + + mlog(0, "Found: %s\n", dl->dl_lockres.l_name); + + goto out_attach; + } + + /* + * There are no other aliases + */ + dl = kmalloc(sizeof(*dl), GFP_NOFS); + if (!dl) { + ret = -ENOMEM; + mlog_errno(ret); + return ret; + } + + dl->dl_count = 0; + /* + * Does this have to happen below, for all attaches, in case + * the struct inode gets blown away by votes? + */ + dl->dl_inode = igrab(inode); + dl->dl_parent_blkno = parent_blkno; + ocfs2_dentry_lock_res_init(dl, parent_blkno, inode); + +out_attach: + spin_lock(&dentry_attach_lock); + dentry->d_fsdata = dl; + dl->dl_count++; + spin_unlock(&dentry_attach_lock); + + /* + * This actually gets us our PRMODE level lock. From now on, + * we'll have a notification if one of these names is + * destroyed on another node. + */ + ret = ocfs2_dentry_lock(dentry, 0); + if (!ret) + ocfs2_dentry_unlock(dentry, 0); + else + mlog_errno(ret); + + dput(alias); + + return ret; +} + +/* + * ocfs2_dentry_iput() and friends. + * + * At this point, our particular dentry is detached from the inodes + * alias list, so there's no way that the locking code can find it. + * + * The interesting stuff happens when we determine that our lock needs + * to go away because this is the last subdir alias in the + * system. This function needs to handle a couple things: + * + * 1) Synchronizing lock shutdown with the downconvert threads. This + * is already handled for us via the lockres release drop function + * called in ocfs2_release_dentry_lock() + * + * 2) A race may occur when we're doing our lock shutdown and + * another process wants to create a new dentry lock. Right now we + * let them race, which means that for a very short while, this + * node might have two locks on a lock resource. This should be a + * problem though because one of them is in the process of being + * thrown out. + */ +static void ocfs2_drop_dentry_lock(struct ocfs2_super *osb, + struct ocfs2_dentry_lock *dl) +{ + ocfs2_simple_drop_lockres(osb, &dl->dl_lockres); + ocfs2_lock_res_free(&dl->dl_lockres); + iput(dl->dl_inode); + kfree(dl); +} + +void ocfs2_dentry_lock_put(struct ocfs2_super *osb, + struct ocfs2_dentry_lock *dl) +{ + int unlock = 0; + + BUG_ON(dl->dl_count == 0); + + spin_lock(&dentry_attach_lock); + dl->dl_count--; + unlock = !dl->dl_count; + spin_unlock(&dentry_attach_lock); + + if (unlock) + ocfs2_drop_dentry_lock(osb, dl); +} + +static void ocfs2_dentry_iput(struct dentry *dentry, struct inode *inode) +{ + struct ocfs2_dentry_lock *dl = dentry->d_fsdata; + + mlog_bug_on_msg(!dl && !(dentry->d_flags & DCACHE_DISCONNECTED), + "dentry: %.*s\n", dentry->d_name.len, + dentry->d_name.name); + + if (!dl) + goto out; + + mlog_bug_on_msg(dl->dl_count == 0, "dentry: %.*s, count: %u\n", + dentry->d_name.len, dentry->d_name.name, + dl->dl_count); + + ocfs2_dentry_lock_put(OCFS2_SB(dentry->d_sb), dl); + +out: + iput(inode); +} + +/* + * d_move(), but keep the locks in sync. + * + * When we are done, "dentry" will have the parent dir and name of + * "target", which will be thrown away. + * + * We manually update the lock of "dentry" if need be. + * + * "target" doesn't have it's dentry lock touched - we allow the later + * dput() to handle this for us. + * + * This is called during ocfs2_rename(), while holding parent + * directory locks. The dentries have already been deleted on other + * nodes via ocfs2_remote_dentry_delete(). + * + * Normally, the VFS handles the d_move() for the file sytem, after + * the ->rename() callback. OCFS2 wants to handle this internally, so + * the new lock can be created atomically with respect to the cluster. + */ +void ocfs2_dentry_move(struct dentry *dentry, struct dentry *target, + struct inode *old_dir, struct inode *new_dir) +{ + int ret; + struct ocfs2_super *osb = OCFS2_SB(old_dir->i_sb); + struct inode *inode = dentry->d_inode; + + /* + * Move within the same directory, so the actual lock info won't + * change. + * + * XXX: Is there any advantage to dropping the lock here? + */ + if (old_dir == new_dir) + goto out_move; + + ocfs2_dentry_lock_put(osb, dentry->d_fsdata); + + dentry->d_fsdata = NULL; + ret = ocfs2_dentry_attach_lock(dentry, inode, OCFS2_I(new_dir)->ip_blkno); + if (ret) + mlog_errno(ret); + +out_move: + d_move(dentry, target); +} + struct dentry_operations ocfs2_dentry_ops = { .d_revalidate = ocfs2_dentry_revalidate, + .d_iput = ocfs2_dentry_iput, }; diff --git a/fs/ocfs2/dcache.h b/fs/ocfs2/dcache.h index 90072771114..c091c34d988 100644 --- a/fs/ocfs2/dcache.h +++ b/fs/ocfs2/dcache.h @@ -28,4 +28,31 @@ extern struct dentry_operations ocfs2_dentry_ops; +struct ocfs2_dentry_lock { + unsigned int dl_count; + u64 dl_parent_blkno; + + /* + * The ocfs2_dentry_lock keeps an inode reference until + * dl_lockres has been destroyed. This is usually done in + * ->d_iput() anyway, so there should be minimal impact. + */ + struct inode *dl_inode; + struct ocfs2_lock_res dl_lockres; +}; + +int ocfs2_dentry_attach_lock(struct dentry *dentry, struct inode *inode, + u64 parent_blkno); + +void ocfs2_dentry_lock_put(struct ocfs2_super *osb, + struct ocfs2_dentry_lock *dl); + +struct dentry *ocfs2_find_local_alias(struct inode *inode, u64 parent_blkno, + int skip_unhashed); + +void ocfs2_dentry_move(struct dentry *dentry, struct dentry *target, + struct inode *old_dir, struct inode *new_dir); + +extern spinlock_t dentry_attach_lock; + #endif /* OCFS2_DCACHE_H */ diff --git a/fs/ocfs2/dir.c b/fs/ocfs2/dir.c index ae47f450792..04e01915b86 100644 --- a/fs/ocfs2/dir.c +++ b/fs/ocfs2/dir.c @@ -74,14 +74,14 @@ static int ocfs2_extend_dir(struct ocfs2_super *osb, int ocfs2_readdir(struct file * filp, void * dirent, filldir_t filldir) { int error = 0; - unsigned long offset, blk; - int i, num, stored; + unsigned long offset, blk, last_ra_blk = 0; + int i, stored; struct buffer_head * bh, * tmp; struct ocfs2_dir_entry * de; int err; struct inode *inode = filp->f_dentry->d_inode; struct super_block * sb = inode->i_sb; - int have_disk_lock = 0; + unsigned int ra_sectors = 16; mlog_entry("dirino=%llu\n", (unsigned long long)OCFS2_I(inode)->ip_blkno); @@ -95,9 +95,8 @@ int ocfs2_readdir(struct file * filp, void * dirent, filldir_t filldir) mlog_errno(error); /* we haven't got any yet, so propagate the error. */ stored = error; - goto bail; + goto bail_nolock; } - have_disk_lock = 1; offset = filp->f_pos & (sb->s_blocksize - 1); @@ -113,16 +112,21 @@ int ocfs2_readdir(struct file * filp, void * dirent, filldir_t filldir) continue; } - /* - * Do the readahead (8k) - */ - if (!offset) { - for (i = 16 >> (sb->s_blocksize_bits - 9), num = 0; + /* The idea here is to begin with 8k read-ahead and to stay + * 4k ahead of our current position. + * + * TODO: Use the pagecache for this. We just need to + * make sure it's cluster-safe... */ + if (!last_ra_blk + || (((last_ra_blk - blk) << 9) <= (ra_sectors / 2))) { + for (i = ra_sectors >> (sb->s_blocksize_bits - 9); i > 0; i--) { tmp = ocfs2_bread(inode, ++blk, &err, 1); if (tmp) brelse(tmp); } + last_ra_blk = blk; + ra_sectors = 8; } revalidate: @@ -194,9 +198,9 @@ revalidate: stored = 0; bail: - if (have_disk_lock) - ocfs2_meta_unlock(inode, 0); + ocfs2_meta_unlock(inode, 0); +bail_nolock: mlog_exit(stored); return stored; @@ -213,11 +217,9 @@ int ocfs2_find_files_on_disk(const char *name, struct ocfs2_dir_entry **dirent) { int status = -ENOENT; - struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); - mlog_entry("(osb=%p, parent=%llu, name='%.*s', blkno=%p, inode=%p)\n", - osb, (unsigned long long)OCFS2_I(inode)->ip_blkno, - namelen, name, blkno, inode); + mlog_entry("(name=%.*s, blkno=%p, inode=%p, dirent_bh=%p, dirent=%p)\n", + namelen, name, blkno, inode, dirent_bh, dirent); *dirent_bh = ocfs2_find_entry(name, namelen, inode, dirent); if (!*dirent_bh || !*dirent) { diff --git a/fs/ocfs2/dlm/dlmapi.h b/fs/ocfs2/dlm/dlmapi.h index 53652f51c0e..cfd5cb65cab 100644 --- a/fs/ocfs2/dlm/dlmapi.h +++ b/fs/ocfs2/dlm/dlmapi.h @@ -182,6 +182,7 @@ enum dlm_status dlmlock(struct dlm_ctxt *dlm, struct dlm_lockstatus *lksb, int flags, const char *name, + int namelen, dlm_astlockfunc_t *ast, void *data, dlm_bastlockfunc_t *bast); diff --git a/fs/ocfs2/dlm/dlmast.c b/fs/ocfs2/dlm/dlmast.c index 355593dd8ef..681046d5139 100644 --- a/fs/ocfs2/dlm/dlmast.c +++ b/fs/ocfs2/dlm/dlmast.c @@ -197,12 +197,14 @@ static void dlm_update_lvb(struct dlm_ctxt *dlm, struct dlm_lock_resource *res, lock->ml.node == dlm->node_num ? "master" : "remote"); memcpy(lksb->lvb, res->lvb, DLM_LVB_LEN); - } else if (lksb->flags & DLM_LKSB_PUT_LVB) { - mlog(0, "setting lvb from lockres for %s node\n", - lock->ml.node == dlm->node_num ? "master" : - "remote"); - memcpy(res->lvb, lksb->lvb, DLM_LVB_LEN); } + /* Do nothing for lvb put requests - they should be done in + * place when the lock is downconverted - otherwise we risk + * racing gets and puts which could result in old lvb data + * being propagated. We leave the put flag set and clear it + * here. In the future we might want to clear it at the time + * the put is actually done. + */ spin_unlock(&res->spinlock); } @@ -318,8 +320,8 @@ int dlm_proxy_ast_handler(struct o2net_msg *msg, u32 len, void *data) res = dlm_lookup_lockres(dlm, name, locklen); if (!res) { - mlog(ML_ERROR, "got %sast for unknown lockres! " - "cookie=%u:%llu, name=%.*s, namelen=%u\n", + mlog(0, "got %sast for unknown lockres! " + "cookie=%u:%llu, name=%.*s, namelen=%u\n", past->type == DLM_AST ? "" : "b", dlm_get_lock_cookie_node(cookie), dlm_get_lock_cookie_seq(cookie), @@ -365,12 +367,10 @@ int dlm_proxy_ast_handler(struct o2net_msg *msg, u32 len, void *data) goto do_ast; } - mlog(ML_ERROR, "got %sast for unknown lock! cookie=%u:%llu, " - "name=%.*s, namelen=%u\n", - past->type == DLM_AST ? "" : "b", - dlm_get_lock_cookie_node(cookie), - dlm_get_lock_cookie_seq(cookie), - locklen, name, locklen); + mlog(0, "got %sast for unknown lock! cookie=%u:%llu, " + "name=%.*s, namelen=%u\n", past->type == DLM_AST ? "" : "b", + dlm_get_lock_cookie_node(cookie), dlm_get_lock_cookie_seq(cookie), + locklen, name, locklen); ret = DLM_NORMAL; unlock_out: @@ -381,8 +381,7 @@ do_ast: ret = DLM_NORMAL; if (past->type == DLM_AST) { /* do not alter lock refcount. switching lists. */ - list_del_init(&lock->list); - list_add_tail(&lock->list, &res->granted); + list_move_tail(&lock->list, &res->granted); mlog(0, "ast: adding to granted list... type=%d, " "convert_type=%d\n", lock->ml.type, lock->ml.convert_type); if (lock->ml.convert_type != LKM_IVMODE) { @@ -463,7 +462,7 @@ int dlm_send_proxy_ast_msg(struct dlm_ctxt *dlm, struct dlm_lock_resource *res, mlog(ML_ERROR, "sent AST to node %u, it returned " "DLM_MIGRATING!\n", lock->ml.node); BUG(); - } else if (status != DLM_NORMAL) { + } else if (status != DLM_NORMAL && status != DLM_IVLOCKID) { mlog(ML_ERROR, "AST to node %u returned %d!\n", lock->ml.node, status); /* ignore it */ diff --git a/fs/ocfs2/dlm/dlmcommon.h b/fs/ocfs2/dlm/dlmcommon.h index 88cc43df18f..fa968180b07 100644 --- a/fs/ocfs2/dlm/dlmcommon.h +++ b/fs/ocfs2/dlm/dlmcommon.h @@ -37,7 +37,17 @@ #define DLM_THREAD_SHUFFLE_INTERVAL 5 // flush everything every 5 passes #define DLM_THREAD_MS 200 // flush at least every 200 ms -#define DLM_HASH_BUCKETS (PAGE_SIZE / sizeof(struct hlist_head)) +#define DLM_HASH_SIZE_DEFAULT (1 << 14) +#if DLM_HASH_SIZE_DEFAULT < PAGE_SIZE +# define DLM_HASH_PAGES 1 +#else +# define DLM_HASH_PAGES (DLM_HASH_SIZE_DEFAULT / PAGE_SIZE) +#endif +#define DLM_BUCKETS_PER_PAGE (PAGE_SIZE / sizeof(struct hlist_head)) +#define DLM_HASH_BUCKETS (DLM_HASH_PAGES * DLM_BUCKETS_PER_PAGE) + +/* Intended to make it easier for us to switch out hash functions */ +#define dlm_lockid_hash(_n, _l) full_name_hash(_n, _l) enum dlm_ast_type { DLM_AST = 0, @@ -61,7 +71,8 @@ static inline int dlm_is_recovery_lock(const char *lock_name, int name_len) return 0; } -#define DLM_RECO_STATE_ACTIVE 0x0001 +#define DLM_RECO_STATE_ACTIVE 0x0001 +#define DLM_RECO_STATE_FINALIZE 0x0002 struct dlm_recovery_ctxt { @@ -85,7 +96,7 @@ enum dlm_ctxt_state { struct dlm_ctxt { struct list_head list; - struct hlist_head *lockres_hash; + struct hlist_head **lockres_hash; struct list_head dirty_list; struct list_head purge_list; struct list_head pending_asts; @@ -120,6 +131,7 @@ struct dlm_ctxt struct o2hb_callback_func dlm_hb_down; struct task_struct *dlm_thread_task; struct task_struct *dlm_reco_thread_task; + struct workqueue_struct *dlm_worker; wait_queue_head_t dlm_thread_wq; wait_queue_head_t dlm_reco_thread_wq; wait_queue_head_t ast_wq; @@ -132,6 +144,11 @@ struct dlm_ctxt struct list_head dlm_eviction_callbacks; }; +static inline struct hlist_head *dlm_lockres_hash(struct dlm_ctxt *dlm, unsigned i) +{ + return dlm->lockres_hash[(i / DLM_BUCKETS_PER_PAGE) % DLM_HASH_PAGES] + (i % DLM_BUCKETS_PER_PAGE); +} + /* these keventd work queue items are for less-frequently * called functions that cannot be directly called from the * net message handlers for some reason, usually because @@ -216,20 +233,29 @@ struct dlm_lock_resource /* WARNING: Please see the comment in dlm_init_lockres before * adding fields here. */ struct hlist_node hash_node; + struct qstr lockname; struct kref refs; - /* please keep these next 3 in this order - * some funcs want to iterate over all lists */ + /* + * Please keep granted, converting, and blocked in this order, + * as some funcs want to iterate over all lists. + * + * All four lists are protected by the hash's reference. + */ struct list_head granted; struct list_head converting; struct list_head blocked; + struct list_head purge; + /* + * These two lists require you to hold an additional reference + * while they are on the list. + */ struct list_head dirty; struct list_head recovering; // dlm_recovery_ctxt.resources list /* unused lock resources have their last_used stamped and are * put on a list for the dlm thread to run. */ - struct list_head purge; unsigned long last_used; unsigned migration_pending:1; @@ -238,7 +264,6 @@ struct dlm_lock_resource wait_queue_head_t wq; u8 owner; //node which owns the lock resource, or unknown u16 state; - struct qstr lockname; char lvb[DLM_LVB_LEN]; }; @@ -300,6 +325,15 @@ enum dlm_lockres_list { DLM_BLOCKED_LIST }; +static inline int dlm_lvb_is_empty(char *lvb) +{ + int i; + for (i=0; i<DLM_LVB_LEN; i++) + if (lvb[i]) + return 0; + return 1; +} + static inline struct list_head * dlm_list_idx_to_ptr(struct dlm_lock_resource *res, enum dlm_lockres_list idx) { @@ -609,7 +643,8 @@ struct dlm_finalize_reco { u8 node_idx; u8 dead_node; - __be16 pad1; + u8 flags; + u8 pad1; __be32 pad2; }; @@ -676,6 +711,7 @@ void dlm_wait_for_recovery(struct dlm_ctxt *dlm); void dlm_kick_recovery_thread(struct dlm_ctxt *dlm); int dlm_is_node_dead(struct dlm_ctxt *dlm, u8 node); int dlm_wait_for_node_death(struct dlm_ctxt *dlm, u8 node, int timeout); +int dlm_wait_for_node_recovery(struct dlm_ctxt *dlm, u8 node, int timeout); void dlm_put(struct dlm_ctxt *dlm); struct dlm_ctxt *dlm_grab(struct dlm_ctxt *dlm); @@ -687,14 +723,20 @@ void dlm_lockres_calc_usage(struct dlm_ctxt *dlm, struct dlm_lock_resource *res); void dlm_purge_lockres(struct dlm_ctxt *dlm, struct dlm_lock_resource *lockres); -void dlm_lockres_get(struct dlm_lock_resource *res); +static inline void dlm_lockres_get(struct dlm_lock_resource *res) +{ + /* This is called on every lookup, so it might be worth + * inlining. */ + kref_get(&res->refs); +} void dlm_lockres_put(struct dlm_lock_resource *res); void __dlm_unhash_lockres(struct dlm_lock_resource *res); void __dlm_insert_lockres(struct dlm_ctxt *dlm, struct dlm_lock_resource *res); struct dlm_lock_resource * __dlm_lookup_lockres(struct dlm_ctxt *dlm, const char *name, - unsigned int len); + unsigned int len, + unsigned int hash); struct dlm_lock_resource * dlm_lookup_lockres(struct dlm_ctxt *dlm, const char *name, unsigned int len); @@ -705,6 +747,7 @@ void dlm_change_lockres_owner(struct dlm_ctxt *dlm, u8 owner); struct dlm_lock_resource * dlm_get_lock_resource(struct dlm_ctxt *dlm, const char *lockid, + int namelen, int flags); struct dlm_lock_resource *dlm_new_lockres(struct dlm_ctxt *dlm, const char *name, @@ -780,8 +823,6 @@ int dlm_begin_reco_handler(struct o2net_msg *msg, u32 len, void *data); int dlm_finalize_reco_handler(struct o2net_msg *msg, u32 len, void *data); int dlm_do_master_requery(struct dlm_ctxt *dlm, struct dlm_lock_resource *res, u8 nodenum, u8 *real_master); -int dlm_lockres_master_requery(struct dlm_ctxt *dlm, - struct dlm_lock_resource *res, u8 *real_master); int dlm_dispatch_assert_master(struct dlm_ctxt *dlm, @@ -819,6 +860,7 @@ void dlm_clean_master_list(struct dlm_ctxt *dlm, u8 dead_node); int dlm_lock_basts_flushed(struct dlm_ctxt *dlm, struct dlm_lock *lock); +int __dlm_lockres_unused(struct dlm_lock_resource *res); static inline const char * dlm_lock_mode_name(int mode) { diff --git a/fs/ocfs2/dlm/dlmconvert.c b/fs/ocfs2/dlm/dlmconvert.c index 8285228d9e3..c764dc8e40a 100644 --- a/fs/ocfs2/dlm/dlmconvert.c +++ b/fs/ocfs2/dlm/dlmconvert.c @@ -214,6 +214,9 @@ grant: if (lock->ml.node == dlm->node_num) mlog(0, "doing in-place convert for nonlocal lock\n"); lock->ml.type = type; + if (lock->lksb->flags & DLM_LKSB_PUT_LVB) + memcpy(res->lvb, lock->lksb->lvb, DLM_LVB_LEN); + status = DLM_NORMAL; *call_ast = 1; goto unlock_exit; @@ -231,8 +234,7 @@ switch_queues: lock->ml.convert_type = type; /* do not alter lock refcount. switching lists. */ - list_del_init(&lock->list); - list_add_tail(&lock->list, &res->converting); + list_move_tail(&lock->list, &res->converting); unlock_exit: spin_unlock(&lock->spinlock); @@ -248,8 +250,7 @@ void dlm_revert_pending_convert(struct dlm_lock_resource *res, struct dlm_lock *lock) { /* do not alter lock refcount. switching lists. */ - list_del_init(&lock->list); - list_add_tail(&lock->list, &res->granted); + list_move_tail(&lock->list, &res->granted); lock->ml.convert_type = LKM_IVMODE; lock->lksb->flags &= ~(DLM_LKSB_GET_LVB|DLM_LKSB_PUT_LVB); } @@ -294,8 +295,7 @@ enum dlm_status dlmconvert_remote(struct dlm_ctxt *dlm, res->state |= DLM_LOCK_RES_IN_PROGRESS; /* move lock to local convert queue */ /* do not alter lock refcount. switching lists. */ - list_del_init(&lock->list); - list_add_tail(&lock->list, &res->converting); + list_move_tail(&lock->list, &res->converting); lock->convert_pending = 1; lock->ml.convert_type = type; @@ -464,6 +464,12 @@ int dlm_convert_lock_handler(struct o2net_msg *msg, u32 len, void *data) } spin_lock(&res->spinlock); + status = __dlm_lockres_state_to_status(res); + if (status != DLM_NORMAL) { + spin_unlock(&res->spinlock); + dlm_error(status); + goto leave; + } list_for_each(iter, &res->granted) { lock = list_entry(iter, struct dlm_lock, list); if (lock->ml.cookie == cnv->cookie && @@ -473,6 +479,21 @@ int dlm_convert_lock_handler(struct o2net_msg *msg, u32 len, void *data) } lock = NULL; } + if (!lock) { + __dlm_print_one_lock_resource(res); + list_for_each(iter, &res->granted) { + lock = list_entry(iter, struct dlm_lock, list); + if (lock->ml.node == cnv->node_idx) { + mlog(ML_ERROR, "There is something here " + "for node %u, lock->ml.cookie=%llu, " + "cnv->cookie=%llu\n", cnv->node_idx, + (unsigned long long)lock->ml.cookie, + (unsigned long long)cnv->cookie); + break; + } + } + lock = NULL; + } spin_unlock(&res->spinlock); if (!lock) { status = DLM_IVLOCKID; diff --git a/fs/ocfs2/dlm/dlmdebug.c b/fs/ocfs2/dlm/dlmdebug.c index c7eae5d3324..3f6c8d88f7a 100644 --- a/fs/ocfs2/dlm/dlmdebug.c +++ b/fs/ocfs2/dlm/dlmdebug.c @@ -37,10 +37,8 @@ #include "dlmapi.h" #include "dlmcommon.h" -#include "dlmdebug.h" #include "dlmdomain.h" -#include "dlmdebug.h" #define MLOG_MASK_PREFIX ML_DLM #include "cluster/masklog.h" @@ -120,6 +118,7 @@ void dlm_print_one_lock(struct dlm_lock *lockid) } EXPORT_SYMBOL_GPL(dlm_print_one_lock); +#if 0 void dlm_dump_lock_resources(struct dlm_ctxt *dlm) { struct dlm_lock_resource *res; @@ -136,12 +135,13 @@ void dlm_dump_lock_resources(struct dlm_ctxt *dlm) spin_lock(&dlm->spinlock); for (i=0; i<DLM_HASH_BUCKETS; i++) { - bucket = &(dlm->lockres_hash[i]); + bucket = dlm_lockres_hash(dlm, i); hlist_for_each_entry(res, iter, bucket, hash_node) dlm_print_one_lock_resource(res); } spin_unlock(&dlm->spinlock); } +#endif /* 0 */ static const char *dlm_errnames[] = { [DLM_NORMAL] = "DLM_NORMAL", diff --git a/fs/ocfs2/dlm/dlmdebug.h b/fs/ocfs2/dlm/dlmdebug.h deleted file mode 100644 index 6858510c3cc..00000000000 --- a/fs/ocfs2/dlm/dlmdebug.h +++ /dev/null @@ -1,30 +0,0 @@ -/* -*- mode: c; c-basic-offset: 8; -*- - * vim: noexpandtab sw=8 ts=8 sts=0: - * - * dlmdebug.h - * - * Copyright (C) 2004 Oracle. All rights reserved. - * - * This program is free software; you can redistribute it and/or - * modify it under the terms of the GNU General Public - * License as published by the Free Software Foundation; either - * version 2 of the License, or (at your option) any later version. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU - * General Public License for more details. - * - * You should have received a copy of the GNU General Public - * License along with this program; if not, write to the - * Free Software Foundation, Inc., 59 Temple Place - Suite 330, - * Boston, MA 021110-1307, USA. - * - */ - -#ifndef DLMDEBUG_H -#define DLMDEBUG_H - -void dlm_dump_lock_resources(struct dlm_ctxt *dlm); - -#endif diff --git a/fs/ocfs2/dlm/dlmdomain.c b/fs/ocfs2/dlm/dlmdomain.c index 8f3a9e3106f..8d1065f8b3b 100644 --- a/fs/ocfs2/dlm/dlmdomain.c +++ b/fs/ocfs2/dlm/dlmdomain.c @@ -41,7 +41,6 @@ #include "dlmapi.h" #include "dlmcommon.h" -#include "dlmdebug.h" #include "dlmdomain.h" #include "dlmver.h" @@ -49,6 +48,33 @@ #define MLOG_MASK_PREFIX (ML_DLM|ML_DLM_DOMAIN) #include "cluster/masklog.h" +static void dlm_free_pagevec(void **vec, int pages) +{ + while (pages--) + free_page((unsigned long)vec[pages]); + kfree(vec); +} + +static void **dlm_alloc_pagevec(int pages) +{ + void **vec = kmalloc(pages * sizeof(void *), GFP_KERNEL); + int i; + + if (!vec) + return NULL; + + for (i = 0; i < pages; i++) + if (!(vec[i] = (void *)__get_free_page(GFP_KERNEL))) + goto out_free; + + mlog(0, "Allocated DLM hash pagevec; %d pages (%lu expected), %lu buckets per page\n", + pages, DLM_HASH_PAGES, (unsigned long)DLM_BUCKETS_PER_PAGE); + return vec; +out_free: + dlm_free_pagevec(vec, i); + return NULL; +} + /* * * spinlock lock ordering: if multiple locks are needed, obey this ordering: @@ -62,7 +88,7 @@ * */ -spinlock_t dlm_domain_lock = SPIN_LOCK_UNLOCKED; +DEFINE_SPINLOCK(dlm_domain_lock); LIST_HEAD(dlm_domains); static DECLARE_WAIT_QUEUE_HEAD(dlm_domain_events); @@ -90,8 +116,7 @@ void __dlm_insert_lockres(struct dlm_ctxt *dlm, assert_spin_locked(&dlm->spinlock); q = &res->lockname; - q->hash = full_name_hash(q->name, q->len); - bucket = &(dlm->lockres_hash[q->hash % DLM_HASH_BUCKETS]); + bucket = dlm_lockres_hash(dlm, q->hash); /* get a reference for our hashtable */ dlm_lockres_get(res); @@ -100,34 +125,32 @@ void __dlm_insert_lockres(struct dlm_ctxt *dlm, } struct dlm_lock_resource * __dlm_lookup_lockres(struct dlm_ctxt *dlm, - const char *name, - unsigned int len) + const char *name, + unsigned int len, + unsigned int hash) { - unsigned int hash; - struct hlist_node *iter; - struct dlm_lock_resource *tmpres=NULL; struct hlist_head *bucket; + struct hlist_node *list; mlog_entry("%.*s\n", len, name); assert_spin_locked(&dlm->spinlock); - hash = full_name_hash(name, len); - - bucket = &(dlm->lockres_hash[hash % DLM_HASH_BUCKETS]); - - /* check for pre-existing lock */ - hlist_for_each(iter, bucket) { - tmpres = hlist_entry(iter, struct dlm_lock_resource, hash_node); - if (tmpres->lockname.len == len && - memcmp(tmpres->lockname.name, name, len) == 0) { - dlm_lockres_get(tmpres); - break; - } + bucket = dlm_lockres_hash(dlm, hash); - tmpres = NULL; + hlist_for_each(list, bucket) { + struct dlm_lock_resource *res = hlist_entry(list, + struct dlm_lock_resource, hash_node); + if (res->lockname.name[0] != name[0]) + continue; + if (unlikely(res->lockname.len != len)) + continue; + if (memcmp(res->lockname.name + 1, name + 1, len - 1)) + continue; + dlm_lockres_get(res); + return res; } - return tmpres; + return NULL; } struct dlm_lock_resource * dlm_lookup_lockres(struct dlm_ctxt *dlm, @@ -135,9 +158,10 @@ struct dlm_lock_resource * dlm_lookup_lockres(struct dlm_ctxt *dlm, unsigned int len) { struct dlm_lock_resource *res; + unsigned int hash = dlm_lockid_hash(name, len); spin_lock(&dlm->spinlock); - res = __dlm_lookup_lockres(dlm, name, len); + res = __dlm_lookup_lockres(dlm, name, len, hash); spin_unlock(&dlm->spinlock); return res; } @@ -194,7 +218,7 @@ static int dlm_wait_on_domain_helper(const char *domain) static void dlm_free_ctxt_mem(struct dlm_ctxt *dlm) { if (dlm->lockres_hash) - free_page((unsigned long) dlm->lockres_hash); + dlm_free_pagevec((void **)dlm->lockres_hash, DLM_HASH_PAGES); if (dlm->name) kfree(dlm->name); @@ -278,11 +302,21 @@ int dlm_domain_fully_joined(struct dlm_ctxt *dlm) return ret; } +static void dlm_destroy_dlm_worker(struct dlm_ctxt *dlm) +{ + if (dlm->dlm_worker) { + flush_workqueue(dlm->dlm_worker); + destroy_workqueue(dlm->dlm_worker); + dlm->dlm_worker = NULL; + } +} + static void dlm_complete_dlm_shutdown(struct dlm_ctxt *dlm) { dlm_unregister_domain_handlers(dlm); dlm_complete_thread(dlm); dlm_complete_recovery_thread(dlm); + dlm_destroy_dlm_worker(dlm); /* We've left the domain. Now we can take ourselves out of the * list and allow the kref stuff to help us free the @@ -304,8 +338,8 @@ static void dlm_migrate_all_locks(struct dlm_ctxt *dlm) restart: spin_lock(&dlm->spinlock); for (i = 0; i < DLM_HASH_BUCKETS; i++) { - while (!hlist_empty(&dlm->lockres_hash[i])) { - res = hlist_entry(dlm->lockres_hash[i].first, + while (!hlist_empty(dlm_lockres_hash(dlm, i))) { + res = hlist_entry(dlm_lockres_hash(dlm, i)->first, struct dlm_lock_resource, hash_node); /* need reference when manually grabbing lockres */ dlm_lockres_get(res); @@ -374,12 +408,13 @@ static void __dlm_print_nodes(struct dlm_ctxt *dlm) assert_spin_locked(&dlm->spinlock); - mlog(ML_NOTICE, "Nodes in my domain (\"%s\"):\n", dlm->name); + printk(KERN_INFO "ocfs2_dlm: Nodes in domain (\"%s\"): ", dlm->name); while ((node = find_next_bit(dlm->domain_map, O2NM_MAX_NODES, node + 1)) < O2NM_MAX_NODES) { - mlog(ML_NOTICE, " node %d\n", node); + printk("%d ", node); } + printk("\n"); } static int dlm_exit_domain_handler(struct o2net_msg *msg, u32 len, void *data) @@ -395,7 +430,7 @@ static int dlm_exit_domain_handler(struct o2net_msg *msg, u32 len, void *data) node = exit_msg->node_idx; - mlog(0, "Node %u leaves domain %s\n", node, dlm->name); + printk(KERN_INFO "ocfs2_dlm: Node %u leaves domain %s\n", node, dlm->name); spin_lock(&dlm->spinlock); clear_bit(node, dlm->domain_map); @@ -644,6 +679,8 @@ static int dlm_assert_joined_handler(struct o2net_msg *msg, u32 len, void *data) set_bit(assert->node_idx, dlm->domain_map); __dlm_set_joining_node(dlm, DLM_LOCK_RES_OWNER_UNKNOWN); + printk(KERN_INFO "ocfs2_dlm: Node %u joins domain %s\n", + assert->node_idx, dlm->name); __dlm_print_nodes(dlm); /* notify anything attached to the heartbeat events */ @@ -1126,6 +1163,13 @@ static int dlm_join_domain(struct dlm_ctxt *dlm) goto bail; } + dlm->dlm_worker = create_singlethread_workqueue("dlm_wq"); + if (!dlm->dlm_worker) { + status = -ENOMEM; + mlog_errno(status); + goto bail; + } + do { unsigned int backoff; status = dlm_try_to_join_domain(dlm); @@ -1166,6 +1210,7 @@ bail: dlm_unregister_domain_handlers(dlm); dlm_complete_thread(dlm); dlm_complete_recovery_thread(dlm); + dlm_destroy_dlm_worker(dlm); } return status; @@ -1191,7 +1236,7 @@ static struct dlm_ctxt *dlm_alloc_ctxt(const char *domain, goto leave; } - dlm->lockres_hash = (struct hlist_head *) __get_free_page(GFP_KERNEL); + dlm->lockres_hash = (struct hlist_head **)dlm_alloc_pagevec(DLM_HASH_PAGES); if (!dlm->lockres_hash) { mlog_errno(-ENOMEM); kfree(dlm->name); @@ -1200,8 +1245,8 @@ static struct dlm_ctxt *dlm_alloc_ctxt(const char *domain, goto leave; } - for (i=0; i<DLM_HASH_BUCKETS; i++) - INIT_HLIST_HEAD(&dlm->lockres_hash[i]); + for (i = 0; i < DLM_HASH_BUCKETS; i++) + INIT_HLIST_HEAD(dlm_lockres_hash(dlm, i)); strcpy(dlm->name, domain); dlm->key = key; @@ -1231,6 +1276,7 @@ static struct dlm_ctxt *dlm_alloc_ctxt(const char *domain, dlm->dlm_thread_task = NULL; dlm->dlm_reco_thread_task = NULL; + dlm->dlm_worker = NULL; init_waitqueue_head(&dlm->dlm_thread_wq); init_waitqueue_head(&dlm->dlm_reco_thread_wq); init_waitqueue_head(&dlm->reco.event); diff --git a/fs/ocfs2/dlm/dlmfs.c b/fs/ocfs2/dlm/dlmfs.c index 7273d9fa6ba..0368c640218 100644 --- a/fs/ocfs2/dlm/dlmfs.c +++ b/fs/ocfs2/dlm/dlmfs.c @@ -116,7 +116,7 @@ static int dlmfs_file_open(struct inode *inode, * doesn't make sense for LVB writes. */ file->f_flags &= ~O_APPEND; - fp = kmalloc(sizeof(*fp), GFP_KERNEL); + fp = kmalloc(sizeof(*fp), GFP_NOFS); if (!fp) { status = -ENOMEM; goto bail; @@ -196,7 +196,7 @@ static ssize_t dlmfs_file_read(struct file *filp, else readlen = count - *ppos; - lvb_buf = kmalloc(readlen, GFP_KERNEL); + lvb_buf = kmalloc(readlen, GFP_NOFS); if (!lvb_buf) return -ENOMEM; @@ -240,7 +240,7 @@ static ssize_t dlmfs_file_write(struct file *filp, else writelen = count - *ppos; - lvb_buf = kmalloc(writelen, GFP_KERNEL); + lvb_buf = kmalloc(writelen, GFP_NOFS); if (!lvb_buf) return -ENOMEM; @@ -335,7 +335,6 @@ static struct inode *dlmfs_get_root_inode(struct super_block *sb) inode->i_mode = mode; inode->i_uid = current->fsuid; inode->i_gid = current->fsgid; - inode->i_blksize = PAGE_CACHE_SIZE; inode->i_blocks = 0; inode->i_mapping->backing_dev_info = &dlmfs_backing_dev_info; inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME; @@ -362,7 +361,6 @@ static struct inode *dlmfs_get_inode(struct inode *parent, inode->i_mode = mode; inode->i_uid = current->fsuid; inode->i_gid = current->fsgid; - inode->i_blksize = PAGE_CACHE_SIZE; inode->i_blocks = 0; inode->i_mapping->backing_dev_info = &dlmfs_backing_dev_info; inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME; @@ -629,9 +627,7 @@ static void __exit exit_dlmfs_fs(void) flush_workqueue(user_dlm_worker); destroy_workqueue(user_dlm_worker); - if (kmem_cache_destroy(dlmfs_inode_cache)) - printk(KERN_INFO "dlmfs_inode_cache: not all structures " - "were freed\n"); + kmem_cache_destroy(dlmfs_inode_cache); } MODULE_AUTHOR("Oracle"); diff --git a/fs/ocfs2/dlm/dlmlock.c b/fs/ocfs2/dlm/dlmlock.c index 6fea28318d6..42a1b91979b 100644 --- a/fs/ocfs2/dlm/dlmlock.c +++ b/fs/ocfs2/dlm/dlmlock.c @@ -53,7 +53,7 @@ #define MLOG_MASK_PREFIX ML_DLM #include "cluster/masklog.h" -static spinlock_t dlm_cookie_lock = SPIN_LOCK_UNLOCKED; +static DEFINE_SPINLOCK(dlm_cookie_lock); static u64 dlm_next_cookie = 1; static enum dlm_status dlm_send_remote_lock_request(struct dlm_ctxt *dlm, @@ -201,6 +201,7 @@ static enum dlm_status dlmlock_remote(struct dlm_ctxt *dlm, struct dlm_lock *lock, int flags) { enum dlm_status status = DLM_DENIED; + int lockres_changed = 1; mlog_entry("type=%d\n", lock->ml.type); mlog(0, "lockres %.*s, flags = 0x%x\n", res->lockname.len, @@ -226,8 +227,25 @@ static enum dlm_status dlmlock_remote(struct dlm_ctxt *dlm, res->state &= ~DLM_LOCK_RES_IN_PROGRESS; lock->lock_pending = 0; if (status != DLM_NORMAL) { - if (status != DLM_NOTQUEUED) + if (status == DLM_RECOVERING && + dlm_is_recovery_lock(res->lockname.name, + res->lockname.len)) { + /* recovery lock was mastered by dead node. + * we need to have calc_usage shoot down this + * lockres and completely remaster it. */ + mlog(0, "%s: recovery lock was owned by " + "dead node %u, remaster it now.\n", + dlm->name, res->owner); + } else if (status != DLM_NOTQUEUED) { + /* + * DO NOT call calc_usage, as this would unhash + * the remote lockres before we ever get to use + * it. treat as if we never made any change to + * the lockres. + */ + lockres_changed = 0; dlm_error(status); + } dlm_revert_pending_lock(res, lock); dlm_lock_put(lock); } else if (dlm_is_recovery_lock(res->lockname.name, @@ -239,12 +257,12 @@ static enum dlm_status dlmlock_remote(struct dlm_ctxt *dlm, mlog(0, "%s: $RECOVERY lock for this node (%u) is " "mastered by %u; got lock, manually granting (no ast)\n", dlm->name, dlm->node_num, res->owner); - list_del_init(&lock->list); - list_add_tail(&lock->list, &res->granted); + list_move_tail(&lock->list, &res->granted); } spin_unlock(&res->spinlock); - dlm_lockres_calc_usage(dlm, res); + if (lockres_changed) + dlm_lockres_calc_usage(dlm, res); wake_up(&res->wq); return status; @@ -281,6 +299,14 @@ static enum dlm_status dlm_send_remote_lock_request(struct dlm_ctxt *dlm, if (tmpret >= 0) { // successfully sent and received ret = status; // this is already a dlm_status + if (ret == DLM_REJECTED) { + mlog(ML_ERROR, "%s:%.*s: BUG. this is a stale lockres " + "no longer owned by %u. that node is coming back " + "up currently.\n", dlm->name, create.namelen, + create.name, res->owner); + dlm_print_one_lock_resource(res); + BUG(); + } } else { mlog_errno(tmpret); if (dlm_is_host_down(tmpret)) { @@ -382,13 +408,13 @@ struct dlm_lock * dlm_new_lock(int type, u8 node, u64 cookie, struct dlm_lock *lock; int kernel_allocated = 0; - lock = kcalloc(1, sizeof(*lock), GFP_KERNEL); + lock = kcalloc(1, sizeof(*lock), GFP_NOFS); if (!lock) return NULL; if (!lksb) { /* zero memory only if kernel-allocated */ - lksb = kcalloc(1, sizeof(*lksb), GFP_KERNEL); + lksb = kcalloc(1, sizeof(*lksb), GFP_NOFS); if (!lksb) { kfree(lock); return NULL; @@ -429,11 +455,16 @@ int dlm_create_lock_handler(struct o2net_msg *msg, u32 len, void *data) if (!dlm_grab(dlm)) return DLM_REJECTED; - mlog_bug_on_msg(!dlm_domain_fully_joined(dlm), - "Domain %s not fully joined!\n", dlm->name); - name = create->name; namelen = create->namelen; + status = DLM_REJECTED; + if (!dlm_domain_fully_joined(dlm)) { + mlog(ML_ERROR, "Domain %s not fully joined, but node %u is " + "sending a create_lock message for lock %.*s!\n", + dlm->name, create->node_idx, namelen, name); + dlm_error(status); + goto leave; + } status = DLM_IVBUFLEN; if (namelen > DLM_LOCKID_NAME_MAX) { @@ -509,8 +540,8 @@ static inline void dlm_get_next_cookie(u8 node_num, u64 *cookie) enum dlm_status dlmlock(struct dlm_ctxt *dlm, int mode, struct dlm_lockstatus *lksb, int flags, - const char *name, dlm_astlockfunc_t *ast, void *data, - dlm_bastlockfunc_t *bast) + const char *name, int namelen, dlm_astlockfunc_t *ast, + void *data, dlm_bastlockfunc_t *bast) { enum dlm_status status; struct dlm_lock_resource *res = NULL; @@ -540,7 +571,7 @@ enum dlm_status dlmlock(struct dlm_ctxt *dlm, int mode, recovery = (flags & LKM_RECOVERY); if (recovery && - (!dlm_is_recovery_lock(name, strlen(name)) || convert) ) { + (!dlm_is_recovery_lock(name, namelen) || convert) ) { dlm_error(status); goto error; } @@ -612,7 +643,7 @@ retry_convert: } status = DLM_IVBUFLEN; - if (strlen(name) > DLM_LOCKID_NAME_MAX || strlen(name) < 1) { + if (namelen > DLM_LOCKID_NAME_MAX || namelen < 1) { dlm_error(status); goto error; } @@ -628,7 +659,7 @@ retry_convert: dlm_wait_for_recovery(dlm); /* find or create the lock resource */ - res = dlm_get_lock_resource(dlm, name, flags); + res = dlm_get_lock_resource(dlm, name, namelen, flags); if (!res) { status = DLM_IVLOCKID; dlm_error(status); @@ -669,18 +700,22 @@ retry_lock: msleep(100); /* no waiting for dlm_reco_thread */ if (recovery) { - if (status == DLM_RECOVERING) { - mlog(0, "%s: got RECOVERING " - "for $REOCVERY lock, master " - "was %u\n", dlm->name, - res->owner); - dlm_wait_for_node_death(dlm, res->owner, - DLM_NODE_DEATH_WAIT_MAX); - } + if (status != DLM_RECOVERING) + goto retry_lock; + + mlog(0, "%s: got RECOVERING " + "for $RECOVERY lock, master " + "was %u\n", dlm->name, + res->owner); + /* wait to see the node go down, then + * drop down and allow the lockres to + * get cleaned up. need to remaster. */ + dlm_wait_for_node_death(dlm, res->owner, + DLM_NODE_DEATH_WAIT_MAX); } else { dlm_wait_for_recovery(dlm); + goto retry_lock; } - goto retry_lock; } if (status != DLM_NORMAL) { diff --git a/fs/ocfs2/dlm/dlmmaster.c b/fs/ocfs2/dlm/dlmmaster.c index 940be4c13b1..f784177b624 100644 --- a/fs/ocfs2/dlm/dlmmaster.c +++ b/fs/ocfs2/dlm/dlmmaster.c @@ -47,7 +47,6 @@ #include "dlmapi.h" #include "dlmcommon.h" -#include "dlmdebug.h" #include "dlmdomain.h" #define MLOG_MASK_PREFIX (ML_DLM|ML_DLM_MASTER) @@ -74,6 +73,7 @@ struct dlm_master_list_entry wait_queue_head_t wq; atomic_t woken; struct kref mle_refs; + int inuse; unsigned long maybe_map[BITS_TO_LONGS(O2NM_MAX_NODES)]; unsigned long vote_map[BITS_TO_LONGS(O2NM_MAX_NODES)]; unsigned long response_map[BITS_TO_LONGS(O2NM_MAX_NODES)]; @@ -127,18 +127,30 @@ static inline int dlm_mle_equal(struct dlm_ctxt *dlm, return 1; } -#if 0 -/* Code here is included but defined out as it aids debugging */ +#define dlm_print_nodemap(m) _dlm_print_nodemap(m,#m) +static void _dlm_print_nodemap(unsigned long *map, const char *mapname) +{ + int i; + printk("%s=[ ", mapname); + for (i=0; i<O2NM_MAX_NODES; i++) + if (test_bit(i, map)) + printk("%d ", i); + printk("]"); +} -void dlm_print_one_mle(struct dlm_master_list_entry *mle) +static void dlm_print_one_mle(struct dlm_master_list_entry *mle) { - int i = 0, refs; + int refs; char *type; char attached; u8 master; unsigned int namelen; const char *name; struct kref *k; + unsigned long *maybe = mle->maybe_map, + *vote = mle->vote_map, + *resp = mle->response_map, + *node = mle->node_map; k = &mle->mle_refs; if (mle->type == DLM_MLE_BLOCK) @@ -159,18 +171,29 @@ void dlm_print_one_mle(struct dlm_master_list_entry *mle) name = mle->u.res->lockname.name; } - mlog(ML_NOTICE, " #%3d: %3s %3d %3u %3u %c (%d)%.*s\n", - i, type, refs, master, mle->new_master, attached, - namelen, namelen, name); + mlog(ML_NOTICE, "%.*s: %3s refs=%3d mas=%3u new=%3u evt=%c inuse=%d ", + namelen, name, type, refs, master, mle->new_master, attached, + mle->inuse); + dlm_print_nodemap(maybe); + printk(", "); + dlm_print_nodemap(vote); + printk(", "); + dlm_print_nodemap(resp); + printk(", "); + dlm_print_nodemap(node); + printk(", "); + printk("\n"); } +#if 0 +/* Code here is included but defined out as it aids debugging */ + static void dlm_dump_mles(struct dlm_ctxt *dlm) { struct dlm_master_list_entry *mle; struct list_head *iter; mlog(ML_NOTICE, "dumping all mles for domain %s:\n", dlm->name); - mlog(ML_NOTICE, " ####: type refs owner new events? lockname nodemap votemap respmap maybemap\n"); spin_lock(&dlm->master_lock); list_for_each(iter, &dlm->master_list) { mle = list_entry(iter, struct dlm_master_list_entry, list); @@ -314,6 +337,31 @@ static inline void dlm_mle_detach_hb_events(struct dlm_ctxt *dlm, spin_unlock(&dlm->spinlock); } +static void dlm_get_mle_inuse(struct dlm_master_list_entry *mle) +{ + struct dlm_ctxt *dlm; + dlm = mle->dlm; + + assert_spin_locked(&dlm->spinlock); + assert_spin_locked(&dlm->master_lock); + mle->inuse++; + kref_get(&mle->mle_refs); +} + +static void dlm_put_mle_inuse(struct dlm_master_list_entry *mle) +{ + struct dlm_ctxt *dlm; + dlm = mle->dlm; + + spin_lock(&dlm->spinlock); + spin_lock(&dlm->master_lock); + mle->inuse--; + __dlm_put_mle(mle); + spin_unlock(&dlm->master_lock); + spin_unlock(&dlm->spinlock); + +} + /* remove from list and free */ static void __dlm_put_mle(struct dlm_master_list_entry *mle) { @@ -322,9 +370,14 @@ static void __dlm_put_mle(struct dlm_master_list_entry *mle) assert_spin_locked(&dlm->spinlock); assert_spin_locked(&dlm->master_lock); - BUG_ON(!atomic_read(&mle->mle_refs.refcount)); - - kref_put(&mle->mle_refs, dlm_mle_release); + if (!atomic_read(&mle->mle_refs.refcount)) { + /* this may or may not crash, but who cares. + * it's a BUG. */ + mlog(ML_ERROR, "bad mle: %p\n", mle); + dlm_print_one_mle(mle); + BUG(); + } else + kref_put(&mle->mle_refs, dlm_mle_release); } @@ -367,6 +420,7 @@ static void dlm_init_mle(struct dlm_master_list_entry *mle, memset(mle->response_map, 0, sizeof(mle->response_map)); mle->master = O2NM_MAX_NODES; mle->new_master = O2NM_MAX_NODES; + mle->inuse = 0; if (mle->type == DLM_MLE_MASTER) { BUG_ON(!res); @@ -564,6 +618,28 @@ static void dlm_lockres_release(struct kref *kref) mlog(0, "destroying lockres %.*s\n", res->lockname.len, res->lockname.name); + if (!hlist_unhashed(&res->hash_node) || + !list_empty(&res->granted) || + !list_empty(&res->converting) || + !list_empty(&res->blocked) || + !list_empty(&res->dirty) || + !list_empty(&res->recovering) || + !list_empty(&res->purge)) { + mlog(ML_ERROR, + "Going to BUG for resource %.*s." + " We're on a list! [%c%c%c%c%c%c%c]\n", + res->lockname.len, res->lockname.name, + !hlist_unhashed(&res->hash_node) ? 'H' : ' ', + !list_empty(&res->granted) ? 'G' : ' ', + !list_empty(&res->converting) ? 'C' : ' ', + !list_empty(&res->blocked) ? 'B' : ' ', + !list_empty(&res->dirty) ? 'D' : ' ', + !list_empty(&res->recovering) ? 'R' : ' ', + !list_empty(&res->purge) ? 'P' : ' '); + + dlm_print_one_lock_resource(res); + } + /* By the time we're ready to blow this guy away, we shouldn't * be on any lists. */ BUG_ON(!hlist_unhashed(&res->hash_node)); @@ -579,11 +655,6 @@ static void dlm_lockres_release(struct kref *kref) kfree(res); } -void dlm_lockres_get(struct dlm_lock_resource *res) -{ - kref_get(&res->refs); -} - void dlm_lockres_put(struct dlm_lock_resource *res) { kref_put(&res->refs, dlm_lockres_release); @@ -603,7 +674,7 @@ static void dlm_init_lockres(struct dlm_ctxt *dlm, memcpy(qname, name, namelen); res->lockname.len = namelen; - res->lockname.hash = full_name_hash(name, namelen); + res->lockname.hash = dlm_lockid_hash(name, namelen); init_waitqueue_head(&res->wq); spin_lock_init(&res->spinlock); @@ -637,11 +708,11 @@ struct dlm_lock_resource *dlm_new_lockres(struct dlm_ctxt *dlm, { struct dlm_lock_resource *res; - res = kmalloc(sizeof(struct dlm_lock_resource), GFP_KERNEL); + res = kmalloc(sizeof(struct dlm_lock_resource), GFP_NOFS); if (!res) return NULL; - res->lockname.name = kmalloc(namelen, GFP_KERNEL); + res->lockname.name = kmalloc(namelen, GFP_NOFS); if (!res->lockname.name) { kfree(res); return NULL; @@ -669,6 +740,7 @@ struct dlm_lock_resource *dlm_new_lockres(struct dlm_ctxt *dlm, */ struct dlm_lock_resource * dlm_get_lock_resource(struct dlm_ctxt *dlm, const char *lockid, + int namelen, int flags) { struct dlm_lock_resource *tmpres=NULL, *res=NULL; @@ -677,19 +749,19 @@ struct dlm_lock_resource * dlm_get_lock_resource(struct dlm_ctxt *dlm, int blocked = 0; int ret, nodenum; struct dlm_node_iter iter; - unsigned int namelen; + unsigned int hash; int tries = 0; int bit, wait_on_recovery = 0; BUG_ON(!lockid); - namelen = strlen(lockid); + hash = dlm_lockid_hash(lockid, namelen); mlog(0, "get lockres %s (len %d)\n", lockid, namelen); lookup: spin_lock(&dlm->spinlock); - tmpres = __dlm_lookup_lockres(dlm, lockid, namelen); + tmpres = __dlm_lookup_lockres(dlm, lockid, namelen, hash); if (tmpres) { spin_unlock(&dlm->spinlock); mlog(0, "found in hash!\n"); @@ -704,7 +776,7 @@ lookup: mlog(0, "allocating a new resource\n"); /* nothing found and we need to allocate one. */ alloc_mle = (struct dlm_master_list_entry *) - kmem_cache_alloc(dlm_mle_cache, GFP_KERNEL); + kmem_cache_alloc(dlm_mle_cache, GFP_NOFS); if (!alloc_mle) goto leave; res = dlm_new_lockres(dlm, lockid, namelen); @@ -790,10 +862,11 @@ lookup: * if so, the creator of the BLOCK may try to put the last * ref at this time in the assert master handler, so we * need an extra one to keep from a bad ptr deref. */ - dlm_get_mle(mle); + dlm_get_mle_inuse(mle); spin_unlock(&dlm->master_lock); spin_unlock(&dlm->spinlock); +redo_request: while (wait_on_recovery) { /* any cluster changes that occurred after dropping the * dlm spinlock would be detectable be a change on the mle, @@ -812,7 +885,7 @@ lookup: } dlm_kick_recovery_thread(dlm); - msleep(100); + msleep(1000); dlm_wait_for_recovery(dlm); spin_lock(&dlm->spinlock); @@ -825,13 +898,15 @@ lookup: } else wait_on_recovery = 0; spin_unlock(&dlm->spinlock); + + if (wait_on_recovery) + dlm_wait_for_node_recovery(dlm, bit, 10000); } /* must wait for lock to be mastered elsewhere */ if (blocked) goto wait; -redo_request: ret = -EINVAL; dlm_node_iter_init(mle->vote_map, &iter); while ((nodenum = dlm_node_iter_next(&iter)) >= 0) { @@ -856,6 +931,7 @@ wait: /* keep going until the response map includes all nodes */ ret = dlm_wait_for_lock_mastery(dlm, res, mle, &blocked); if (ret < 0) { + wait_on_recovery = 1; mlog(0, "%s:%.*s: node map changed, redo the " "master request now, blocked=%d\n", dlm->name, res->lockname.len, @@ -866,7 +942,7 @@ wait: dlm->name, res->lockname.len, res->lockname.name, blocked); dlm_print_one_lock_resource(res); - /* dlm_print_one_mle(mle); */ + dlm_print_one_mle(mle); tries = 0; } goto redo_request; @@ -880,7 +956,7 @@ wait: dlm_mle_detach_hb_events(dlm, mle); dlm_put_mle(mle); /* put the extra ref */ - dlm_put_mle(mle); + dlm_put_mle_inuse(mle); wake_waiters: spin_lock(&res->spinlock); @@ -921,12 +997,14 @@ recheck: spin_unlock(&res->spinlock); /* this will cause the master to re-assert across * the whole cluster, freeing up mles */ - ret = dlm_do_master_request(mle, res->owner); - if (ret < 0) { - /* give recovery a chance to run */ - mlog(ML_ERROR, "link to %u went down?: %d\n", res->owner, ret); - msleep(500); - goto recheck; + if (res->owner != dlm->node_num) { + ret = dlm_do_master_request(mle, res->owner); + if (ret < 0) { + /* give recovery a chance to run */ + mlog(ML_ERROR, "link to %u went down?: %d\n", res->owner, ret); + msleep(500); + goto recheck; + } } ret = 0; goto leave; @@ -962,6 +1040,12 @@ recheck: "rechecking now\n", dlm->name, res->lockname.len, res->lockname.name); goto recheck; + } else { + if (!voting_done) { + mlog(0, "map not changed and voting not done " + "for %s:%.*s\n", dlm->name, res->lockname.len, + res->lockname.name); + } } if (m != O2NM_MAX_NODES) { @@ -1129,18 +1213,6 @@ static int dlm_restart_lock_mastery(struct dlm_ctxt *dlm, set_bit(node, mle->vote_map); } else { mlog(ML_ERROR, "node down! %d\n", node); - - /* if the node wasn't involved in mastery skip it, - * but clear it out from the maps so that it will - * not affect mastery of this lockres */ - clear_bit(node, mle->response_map); - clear_bit(node, mle->vote_map); - if (!test_bit(node, mle->maybe_map)) - goto next; - - /* if we're already blocked on lock mastery, and the - * dead node wasn't the expected master, or there is - * another node in the maybe_map, keep waiting */ if (blocked) { int lowest = find_next_bit(mle->maybe_map, O2NM_MAX_NODES, 0); @@ -1148,54 +1220,53 @@ static int dlm_restart_lock_mastery(struct dlm_ctxt *dlm, /* act like it was never there */ clear_bit(node, mle->maybe_map); - if (node != lowest) - goto next; - - mlog(ML_ERROR, "expected master %u died while " - "this node was blocked waiting on it!\n", - node); - lowest = find_next_bit(mle->maybe_map, - O2NM_MAX_NODES, - lowest+1); - if (lowest < O2NM_MAX_NODES) { - mlog(0, "still blocked. waiting " - "on %u now\n", lowest); - goto next; + if (node == lowest) { + mlog(0, "expected master %u died" + " while this node was blocked " + "waiting on it!\n", node); + lowest = find_next_bit(mle->maybe_map, + O2NM_MAX_NODES, + lowest+1); + if (lowest < O2NM_MAX_NODES) { + mlog(0, "%s:%.*s:still " + "blocked. waiting on %u " + "now\n", dlm->name, + res->lockname.len, + res->lockname.name, + lowest); + } else { + /* mle is an MLE_BLOCK, but + * there is now nothing left to + * block on. we need to return + * all the way back out and try + * again with an MLE_MASTER. + * dlm_do_local_recovery_cleanup + * has already run, so the mle + * refcount is ok */ + mlog(0, "%s:%.*s: no " + "longer blocking. try to " + "master this here\n", + dlm->name, + res->lockname.len, + res->lockname.name); + mle->type = DLM_MLE_MASTER; + mle->u.res = res; + } } - - /* mle is an MLE_BLOCK, but there is now - * nothing left to block on. we need to return - * all the way back out and try again with - * an MLE_MASTER. dlm_do_local_recovery_cleanup - * has already run, so the mle refcount is ok */ - mlog(0, "no longer blocking. we can " - "try to master this here\n"); - mle->type = DLM_MLE_MASTER; - memset(mle->maybe_map, 0, - sizeof(mle->maybe_map)); - memset(mle->response_map, 0, - sizeof(mle->maybe_map)); - memcpy(mle->vote_map, mle->node_map, - sizeof(mle->node_map)); - mle->u.res = res; - set_bit(dlm->node_num, mle->maybe_map); - - ret = -EAGAIN; - goto next; } - clear_bit(node, mle->maybe_map); - if (node > dlm->node_num) - goto next; - - mlog(0, "dead node in map!\n"); - /* yuck. go back and re-contact all nodes - * in the vote_map, removing this node. */ - memset(mle->response_map, 0, - sizeof(mle->response_map)); + /* now blank out everything, as if we had never + * contacted anyone */ + memset(mle->maybe_map, 0, sizeof(mle->maybe_map)); + memset(mle->response_map, 0, sizeof(mle->response_map)); + /* reset the vote_map to the current node_map */ + memcpy(mle->vote_map, mle->node_map, + sizeof(mle->node_map)); + /* put myself into the maybe map */ + if (mle->type != DLM_MLE_BLOCK) + set_bit(dlm->node_num, mle->maybe_map); } ret = -EAGAIN; -next: node = dlm_bitmap_diff_iter_next(&bdi, &sc); } return ret; @@ -1316,7 +1387,7 @@ int dlm_master_request_handler(struct o2net_msg *msg, u32 len, void *data) struct dlm_master_request *request = (struct dlm_master_request *) msg->buf; struct dlm_master_list_entry *mle = NULL, *tmpmle = NULL; char *name; - unsigned int namelen; + unsigned int namelen, hash; int found, ret; int set_maybe; int dispatch_assert = 0; @@ -1331,6 +1402,7 @@ int dlm_master_request_handler(struct o2net_msg *msg, u32 len, void *data) name = request->name; namelen = request->namelen; + hash = dlm_lockid_hash(name, namelen); if (namelen > DLM_LOCKID_NAME_MAX) { response = DLM_IVBUFLEN; @@ -1339,7 +1411,7 @@ int dlm_master_request_handler(struct o2net_msg *msg, u32 len, void *data) way_up_top: spin_lock(&dlm->spinlock); - res = __dlm_lookup_lockres(dlm, name, namelen); + res = __dlm_lookup_lockres(dlm, name, namelen, hash); if (res) { spin_unlock(&dlm->spinlock); @@ -1459,21 +1531,18 @@ way_up_top: spin_unlock(&dlm->spinlock); mle = (struct dlm_master_list_entry *) - kmem_cache_alloc(dlm_mle_cache, GFP_KERNEL); + kmem_cache_alloc(dlm_mle_cache, GFP_NOFS); if (!mle) { response = DLM_MASTER_RESP_ERROR; mlog_errno(-ENOMEM); goto send_response; } - spin_lock(&dlm->spinlock); - dlm_init_mle(mle, DLM_MLE_BLOCK, dlm, NULL, - name, namelen); - spin_unlock(&dlm->spinlock); goto way_up_top; } // mlog(0, "this is second time thru, already allocated, " // "add the block.\n"); + dlm_init_mle(mle, DLM_MLE_BLOCK, dlm, NULL, name, namelen); set_bit(request->node_idx, mle->maybe_map); list_add(&mle->list, &dlm->master_list); response = DLM_MASTER_RESP_NO; @@ -1556,6 +1625,8 @@ again: dlm_node_iter_init(nodemap, &iter); while ((to = dlm_node_iter_next(&iter)) >= 0) { int r = 0; + struct dlm_master_list_entry *mle = NULL; + mlog(0, "sending assert master to %d (%.*s)\n", to, namelen, lockname); memset(&assert, 0, sizeof(assert)); @@ -1567,20 +1638,28 @@ again: tmpret = o2net_send_message(DLM_ASSERT_MASTER_MSG, dlm->key, &assert, sizeof(assert), to, &r); if (tmpret < 0) { - mlog(ML_ERROR, "assert_master returned %d!\n", tmpret); + mlog(0, "assert_master returned %d!\n", tmpret); if (!dlm_is_host_down(tmpret)) { - mlog(ML_ERROR, "unhandled error!\n"); + mlog(ML_ERROR, "unhandled error=%d!\n", tmpret); BUG(); } /* a node died. finish out the rest of the nodes. */ - mlog(ML_ERROR, "link to %d went down!\n", to); + mlog(0, "link to %d went down!\n", to); /* any nonzero status return will do */ ret = tmpret; } else if (r < 0) { /* ok, something horribly messed. kill thyself. */ mlog(ML_ERROR,"during assert master of %.*s to %u, " "got %d.\n", namelen, lockname, to, r); - dlm_dump_lock_resources(dlm); + spin_lock(&dlm->spinlock); + spin_lock(&dlm->master_lock); + if (dlm_find_mle(dlm, &mle, (char *)lockname, + namelen)) { + dlm_print_one_mle(mle); + __dlm_put_mle(mle); + } + spin_unlock(&dlm->master_lock); + spin_unlock(&dlm->spinlock); BUG(); } else if (r == EAGAIN) { mlog(0, "%.*s: node %u create mles on other " @@ -1612,7 +1691,7 @@ int dlm_assert_master_handler(struct o2net_msg *msg, u32 len, void *data) struct dlm_assert_master *assert = (struct dlm_assert_master *)msg->buf; struct dlm_lock_resource *res = NULL; char *name; - unsigned int namelen; + unsigned int namelen, hash; u32 flags; int master_request = 0; int ret = 0; @@ -1622,6 +1701,7 @@ int dlm_assert_master_handler(struct o2net_msg *msg, u32 len, void *data) name = assert->name; namelen = assert->namelen; + hash = dlm_lockid_hash(name, namelen); flags = be32_to_cpu(assert->flags); if (namelen > DLM_LOCKID_NAME_MAX) { @@ -1646,7 +1726,7 @@ int dlm_assert_master_handler(struct o2net_msg *msg, u32 len, void *data) if (bit >= O2NM_MAX_NODES) { /* not necessarily an error, though less likely. * could be master just re-asserting. */ - mlog(ML_ERROR, "no bits set in the maybe_map, but %u " + mlog(0, "no bits set in the maybe_map, but %u " "is asserting! (%.*s)\n", assert->node_idx, namelen, name); } else if (bit != assert->node_idx) { @@ -1658,19 +1738,36 @@ int dlm_assert_master_handler(struct o2net_msg *msg, u32 len, void *data) * number winning the mastery will respond * YES to mastery requests, but this node * had no way of knowing. let it pass. */ - mlog(ML_ERROR, "%u is the lowest node, " + mlog(0, "%u is the lowest node, " "%u is asserting. (%.*s) %u must " "have begun after %u won.\n", bit, assert->node_idx, namelen, name, bit, assert->node_idx); } } + if (mle->type == DLM_MLE_MIGRATION) { + if (flags & DLM_ASSERT_MASTER_MLE_CLEANUP) { + mlog(0, "%s:%.*s: got cleanup assert" + " from %u for migration\n", + dlm->name, namelen, name, + assert->node_idx); + } else if (!(flags & DLM_ASSERT_MASTER_FINISH_MIGRATION)) { + mlog(0, "%s:%.*s: got unrelated assert" + " from %u for migration, ignoring\n", + dlm->name, namelen, name, + assert->node_idx); + __dlm_put_mle(mle); + spin_unlock(&dlm->master_lock); + spin_unlock(&dlm->spinlock); + goto done; + } + } } spin_unlock(&dlm->master_lock); /* ok everything checks out with the MLE * now check to see if there is a lockres */ - res = __dlm_lookup_lockres(dlm, name, namelen); + res = __dlm_lookup_lockres(dlm, name, namelen, hash); if (res) { spin_lock(&res->spinlock); if (res->state & DLM_LOCK_RES_RECOVERING) { @@ -1679,7 +1776,8 @@ int dlm_assert_master_handler(struct o2net_msg *msg, u32 len, void *data) goto kill; } if (!mle) { - if (res->owner != assert->node_idx) { + if (res->owner != DLM_LOCK_RES_OWNER_UNKNOWN && + res->owner != assert->node_idx) { mlog(ML_ERROR, "assert_master from " "%u, but current owner is " "%u! (%.*s)\n", @@ -1732,6 +1830,7 @@ ok: if (mle) { int extra_ref = 0; int nn = -1; + int rr, err = 0; spin_lock(&mle->spinlock); if (mle->type == DLM_MLE_BLOCK || mle->type == DLM_MLE_MIGRATION) @@ -1751,27 +1850,64 @@ ok: wake_up(&mle->wq); spin_unlock(&mle->spinlock); - if (mle->type == DLM_MLE_MIGRATION && res) { - mlog(0, "finishing off migration of lockres %.*s, " - "from %u to %u\n", - res->lockname.len, res->lockname.name, - dlm->node_num, mle->new_master); + if (res) { spin_lock(&res->spinlock); - res->state &= ~DLM_LOCK_RES_MIGRATING; - dlm_change_lockres_owner(dlm, res, mle->new_master); - BUG_ON(res->state & DLM_LOCK_RES_DIRTY); + if (mle->type == DLM_MLE_MIGRATION) { + mlog(0, "finishing off migration of lockres %.*s, " + "from %u to %u\n", + res->lockname.len, res->lockname.name, + dlm->node_num, mle->new_master); + res->state &= ~DLM_LOCK_RES_MIGRATING; + dlm_change_lockres_owner(dlm, res, mle->new_master); + BUG_ON(res->state & DLM_LOCK_RES_DIRTY); + } else { + dlm_change_lockres_owner(dlm, res, mle->master); + } spin_unlock(&res->spinlock); } - /* master is known, detach if not already detached */ - dlm_mle_detach_hb_events(dlm, mle); - dlm_put_mle(mle); - + + /* master is known, detach if not already detached. + * ensures that only one assert_master call will happen + * on this mle. */ + spin_lock(&dlm->spinlock); + spin_lock(&dlm->master_lock); + + rr = atomic_read(&mle->mle_refs.refcount); + if (mle->inuse > 0) { + if (extra_ref && rr < 3) + err = 1; + else if (!extra_ref && rr < 2) + err = 1; + } else { + if (extra_ref && rr < 2) + err = 1; + else if (!extra_ref && rr < 1) + err = 1; + } + if (err) { + mlog(ML_ERROR, "%s:%.*s: got assert master from %u " + "that will mess up this node, refs=%d, extra=%d, " + "inuse=%d\n", dlm->name, namelen, name, + assert->node_idx, rr, extra_ref, mle->inuse); + dlm_print_one_mle(mle); + } + list_del_init(&mle->list); + __dlm_mle_detach_hb_events(dlm, mle); + __dlm_put_mle(mle); if (extra_ref) { /* the assert master message now balances the extra * ref given by the master / migration request message. * if this is the last put, it will be removed * from the list. */ - dlm_put_mle(mle); + __dlm_put_mle(mle); + } + spin_unlock(&dlm->master_lock); + spin_unlock(&dlm->spinlock); + } else if (res) { + if (res->owner != assert->node_idx) { + mlog(0, "assert_master from %u, but current " + "owner is %u (%.*s), no mle\n", assert->node_idx, + res->owner, namelen, name); } } @@ -1788,12 +1924,12 @@ done: kill: /* kill the caller! */ + mlog(ML_ERROR, "Bad message received from another node. Dumping state " + "and killing the other node now! This node is OK and can continue.\n"); + __dlm_print_one_lock_resource(res); spin_unlock(&res->spinlock); spin_unlock(&dlm->spinlock); dlm_lockres_put(res); - mlog(ML_ERROR, "Bad message received from another node. Dumping state " - "and killing the other node now! This node is OK and can continue.\n"); - dlm_dump_lock_resources(dlm); dlm_put(dlm); return -EINVAL; } @@ -1803,7 +1939,7 @@ int dlm_dispatch_assert_master(struct dlm_ctxt *dlm, int ignore_higher, u8 request_from, u32 flags) { struct dlm_work_item *item; - item = kcalloc(1, sizeof(*item), GFP_KERNEL); + item = kcalloc(1, sizeof(*item), GFP_NOFS); if (!item) return -ENOMEM; @@ -1825,7 +1961,7 @@ int dlm_dispatch_assert_master(struct dlm_ctxt *dlm, list_add_tail(&item->list, &dlm->work_list); spin_unlock(&dlm->work_lock); - schedule_work(&dlm->dispatched_work); + queue_work(dlm->dlm_worker, &dlm->dispatched_work); return 0; } @@ -1866,6 +2002,23 @@ static void dlm_assert_master_worker(struct dlm_work_item *item, void *data) } } + /* + * If we're migrating this lock to someone else, we are no + * longer allowed to assert out own mastery. OTOH, we need to + * prevent migration from starting while we're still asserting + * our dominance. The reserved ast delays migration. + */ + spin_lock(&res->spinlock); + if (res->state & DLM_LOCK_RES_MIGRATING) { + mlog(0, "Someone asked us to assert mastery, but we're " + "in the middle of migration. Skipping assert, " + "the new master will handle that.\n"); + spin_unlock(&res->spinlock); + goto put; + } else + __dlm_lockres_reserve_ast(res); + spin_unlock(&res->spinlock); + /* this call now finishes out the nodemap * even if one or more nodes die */ mlog(0, "worker about to master %.*s here, this=%u\n", @@ -1875,9 +2028,14 @@ static void dlm_assert_master_worker(struct dlm_work_item *item, void *data) nodemap, flags); if (ret < 0) { /* no need to restart, we are done */ - mlog_errno(ret); + if (!dlm_is_host_down(ret)) + mlog_errno(ret); } + /* Ok, we've asserted ourselves. Let's let migration start. */ + dlm_lockres_release_ast(dlm, res); + +put: dlm_lockres_put(res); mlog(0, "finished with dlm_assert_master_worker\n"); @@ -1916,6 +2074,7 @@ static int dlm_pre_master_reco_lockres(struct dlm_ctxt *dlm, BUG(); /* host is down, so answer for that node would be * DLM_LOCK_RES_OWNER_UNKNOWN. continue. */ + ret = 0; } if (master != DLM_LOCK_RES_OWNER_UNKNOWN) { @@ -2016,14 +2175,14 @@ int dlm_migrate_lockres(struct dlm_ctxt *dlm, struct dlm_lock_resource *res, */ ret = -ENOMEM; - mres = (struct dlm_migratable_lockres *) __get_free_page(GFP_KERNEL); + mres = (struct dlm_migratable_lockres *) __get_free_page(GFP_NOFS); if (!mres) { mlog_errno(ret); goto leave; } mle = (struct dlm_master_list_entry *) kmem_cache_alloc(dlm_mle_cache, - GFP_KERNEL); + GFP_NOFS); if (!mle) { mlog_errno(ret); goto leave; @@ -2117,7 +2276,7 @@ fail: * take both dlm->spinlock and dlm->master_lock */ spin_lock(&dlm->spinlock); spin_lock(&dlm->master_lock); - dlm_get_mle(mle); + dlm_get_mle_inuse(mle); spin_unlock(&dlm->master_lock); spin_unlock(&dlm->spinlock); @@ -2134,7 +2293,10 @@ fail: /* migration failed, detach and clean up mle */ dlm_mle_detach_hb_events(dlm, mle); dlm_put_mle(mle); - dlm_put_mle(mle); + dlm_put_mle_inuse(mle); + spin_lock(&res->spinlock); + res->state &= ~DLM_LOCK_RES_MIGRATING; + spin_unlock(&res->spinlock); goto leave; } @@ -2164,8 +2326,8 @@ fail: /* avoid hang during shutdown when migrating lockres * to a node which also goes down */ if (dlm_is_node_dead(dlm, target)) { - mlog(0, "%s:%.*s: expected migration target %u " - "is no longer up. restarting.\n", + mlog(0, "%s:%.*s: expected migration " + "target %u is no longer up, restarting\n", dlm->name, res->lockname.len, res->lockname.name, target); ret = -ERESTARTSYS; @@ -2175,7 +2337,10 @@ fail: /* migration failed, detach and clean up mle */ dlm_mle_detach_hb_events(dlm, mle); dlm_put_mle(mle); - dlm_put_mle(mle); + dlm_put_mle_inuse(mle); + spin_lock(&res->spinlock); + res->state &= ~DLM_LOCK_RES_MIGRATING; + spin_unlock(&res->spinlock); goto leave; } /* TODO: if node died: stop, clean up, return error */ @@ -2191,7 +2356,7 @@ fail: /* master is known, detach if not already detached */ dlm_mle_detach_hb_events(dlm, mle); - dlm_put_mle(mle); + dlm_put_mle_inuse(mle); ret = 0; dlm_lockres_calc_usage(dlm, res); @@ -2210,7 +2375,6 @@ leave: mlog(0, "returning %d\n", ret); return ret; } -EXPORT_SYMBOL_GPL(dlm_migrate_lockres); int dlm_lock_basts_flushed(struct dlm_ctxt *dlm, struct dlm_lock *lock) { @@ -2462,7 +2626,7 @@ int dlm_migrate_request_handler(struct o2net_msg *msg, u32 len, void *data) struct dlm_migrate_request *migrate = (struct dlm_migrate_request *) msg->buf; struct dlm_master_list_entry *mle = NULL, *oldmle = NULL; const char *name; - unsigned int namelen; + unsigned int namelen, hash; int ret = 0; if (!dlm_grab(dlm)) @@ -2470,10 +2634,11 @@ int dlm_migrate_request_handler(struct o2net_msg *msg, u32 len, void *data) name = migrate->name; namelen = migrate->namelen; + hash = dlm_lockid_hash(name, namelen); /* preallocate.. if this fails, abort */ mle = (struct dlm_master_list_entry *) kmem_cache_alloc(dlm_mle_cache, - GFP_KERNEL); + GFP_NOFS); if (!mle) { ret = -ENOMEM; @@ -2482,7 +2647,7 @@ int dlm_migrate_request_handler(struct o2net_msg *msg, u32 len, void *data) /* check for pre-existing lock */ spin_lock(&dlm->spinlock); - res = __dlm_lookup_lockres(dlm, name, namelen); + res = __dlm_lookup_lockres(dlm, name, namelen, hash); spin_lock(&dlm->master_lock); if (res) { @@ -2580,6 +2745,7 @@ static int dlm_add_migration_mle(struct dlm_ctxt *dlm, /* remove it from the list so that only one * mle will be found */ list_del_init(&tmp->list); + __dlm_mle_detach_hb_events(dlm, mle); } spin_unlock(&tmp->spinlock); } @@ -2601,6 +2767,7 @@ void dlm_clean_master_list(struct dlm_ctxt *dlm, u8 dead_node) struct list_head *iter, *iter2; struct dlm_master_list_entry *mle; struct dlm_lock_resource *res; + unsigned int hash; mlog_entry("dlm=%s, dead node=%u\n", dlm->name, dead_node); top: @@ -2640,7 +2807,7 @@ top: * may result in the mle being unlinked and * freed, but there may still be a process * waiting in the dlmlock path which is fine. */ - mlog(ML_ERROR, "node %u was expected master\n", + mlog(0, "node %u was expected master\n", dead_node); atomic_set(&mle->woken, 1); spin_unlock(&mle->spinlock); @@ -2673,19 +2840,21 @@ top: /* remove from the list early. NOTE: unlinking * list_head while in list_for_each_safe */ + __dlm_mle_detach_hb_events(dlm, mle); spin_lock(&mle->spinlock); list_del_init(&mle->list); atomic_set(&mle->woken, 1); spin_unlock(&mle->spinlock); wake_up(&mle->wq); - mlog(0, "node %u died during migration from " - "%u to %u!\n", dead_node, + mlog(0, "%s: node %u died during migration from " + "%u to %u!\n", dlm->name, dead_node, mle->master, mle->new_master); /* if there is a lockres associated with this * mle, find it and set its owner to UNKNOWN */ + hash = dlm_lockid_hash(mle->u.name.name, mle->u.name.len); res = __dlm_lookup_lockres(dlm, mle->u.name.name, - mle->u.name.len); + mle->u.name.len, hash); if (res) { /* unfortunately if we hit this rare case, our * lock ordering is messed. we need to drop diff --git a/fs/ocfs2/dlm/dlmrecovery.c b/fs/ocfs2/dlm/dlmrecovery.c index 805cbabac05..9d950d7cea3 100644 --- a/fs/ocfs2/dlm/dlmrecovery.c +++ b/fs/ocfs2/dlm/dlmrecovery.c @@ -95,11 +95,14 @@ static void dlm_reco_unlock_ast(void *astdata, enum dlm_status st); static void dlm_request_all_locks_worker(struct dlm_work_item *item, void *data); static void dlm_mig_lockres_worker(struct dlm_work_item *item, void *data); +static int dlm_lockres_master_requery(struct dlm_ctxt *dlm, + struct dlm_lock_resource *res, + u8 *real_master); static u64 dlm_get_next_mig_cookie(void); -static spinlock_t dlm_reco_state_lock = SPIN_LOCK_UNLOCKED; -static spinlock_t dlm_mig_cookie_lock = SPIN_LOCK_UNLOCKED; +static DEFINE_SPINLOCK(dlm_reco_state_lock); +static DEFINE_SPINLOCK(dlm_mig_cookie_lock); static u64 dlm_mig_cookie = 1; static u64 dlm_get_next_mig_cookie(void) @@ -115,12 +118,37 @@ static u64 dlm_get_next_mig_cookie(void) return c; } +static inline void dlm_set_reco_dead_node(struct dlm_ctxt *dlm, + u8 dead_node) +{ + assert_spin_locked(&dlm->spinlock); + if (dlm->reco.dead_node != dead_node) + mlog(0, "%s: changing dead_node from %u to %u\n", + dlm->name, dlm->reco.dead_node, dead_node); + dlm->reco.dead_node = dead_node; +} + +static inline void dlm_set_reco_master(struct dlm_ctxt *dlm, + u8 master) +{ + assert_spin_locked(&dlm->spinlock); + mlog(0, "%s: changing new_master from %u to %u\n", + dlm->name, dlm->reco.new_master, master); + dlm->reco.new_master = master; +} + +static inline void __dlm_reset_recovery(struct dlm_ctxt *dlm) +{ + assert_spin_locked(&dlm->spinlock); + clear_bit(dlm->reco.dead_node, dlm->recovery_map); + dlm_set_reco_dead_node(dlm, O2NM_INVALID_NODE_NUM); + dlm_set_reco_master(dlm, O2NM_INVALID_NODE_NUM); +} + static inline void dlm_reset_recovery(struct dlm_ctxt *dlm) { spin_lock(&dlm->spinlock); - clear_bit(dlm->reco.dead_node, dlm->recovery_map); - dlm->reco.dead_node = O2NM_INVALID_NODE_NUM; - dlm->reco.new_master = O2NM_INVALID_NODE_NUM; + __dlm_reset_recovery(dlm); spin_unlock(&dlm->spinlock); } @@ -132,12 +160,21 @@ void dlm_dispatch_work(void *data) struct list_head *iter, *iter2; struct dlm_work_item *item; dlm_workfunc_t *workfunc; + int tot=0; + + if (!dlm_joined(dlm)) + return; spin_lock(&dlm->work_lock); list_splice_init(&dlm->work_list, &tmp_list); spin_unlock(&dlm->work_lock); list_for_each_safe(iter, iter2, &tmp_list) { + tot++; + } + mlog(0, "%s: work thread has %d work items\n", dlm->name, tot); + + list_for_each_safe(iter, iter2, &tmp_list) { item = list_entry(iter, struct dlm_work_item, list); workfunc = item->func; list_del_init(&item->list); @@ -220,6 +257,52 @@ void dlm_complete_recovery_thread(struct dlm_ctxt *dlm) * */ +static void dlm_print_reco_node_status(struct dlm_ctxt *dlm) +{ + struct dlm_reco_node_data *ndata; + struct dlm_lock_resource *res; + + mlog(ML_NOTICE, "%s(%d): recovery info, state=%s, dead=%u, master=%u\n", + dlm->name, dlm->dlm_reco_thread_task->pid, + dlm->reco.state & DLM_RECO_STATE_ACTIVE ? "ACTIVE" : "inactive", + dlm->reco.dead_node, dlm->reco.new_master); + + list_for_each_entry(ndata, &dlm->reco.node_data, list) { + char *st = "unknown"; + switch (ndata->state) { + case DLM_RECO_NODE_DATA_INIT: + st = "init"; + break; + case DLM_RECO_NODE_DATA_REQUESTING: + st = "requesting"; + break; + case DLM_RECO_NODE_DATA_DEAD: + st = "dead"; + break; + case DLM_RECO_NODE_DATA_RECEIVING: + st = "receiving"; + break; + case DLM_RECO_NODE_DATA_REQUESTED: + st = "requested"; + break; + case DLM_RECO_NODE_DATA_DONE: + st = "done"; + break; + case DLM_RECO_NODE_DATA_FINALIZE_SENT: + st = "finalize-sent"; + break; + default: + st = "bad"; + break; + } + mlog(ML_NOTICE, "%s: reco state, node %u, state=%s\n", + dlm->name, ndata->node_num, st); + } + list_for_each_entry(res, &dlm->reco.resources, recovering) { + mlog(ML_NOTICE, "%s: lockres %.*s on recovering list\n", + dlm->name, res->lockname.len, res->lockname.name); + } +} #define DLM_RECO_THREAD_TIMEOUT_MS (5 * 1000) @@ -267,11 +350,23 @@ int dlm_is_node_dead(struct dlm_ctxt *dlm, u8 node) { int dead; spin_lock(&dlm->spinlock); - dead = test_bit(node, dlm->domain_map); + dead = !test_bit(node, dlm->domain_map); spin_unlock(&dlm->spinlock); return dead; } +/* returns true if node is no longer in the domain + * could be dead or just not joined */ +static int dlm_is_node_recovered(struct dlm_ctxt *dlm, u8 node) +{ + int recovered; + spin_lock(&dlm->spinlock); + recovered = !test_bit(node, dlm->recovery_map); + spin_unlock(&dlm->spinlock); + return recovered; +} + + int dlm_wait_for_node_death(struct dlm_ctxt *dlm, u8 node, int timeout) { if (timeout) { @@ -290,6 +385,24 @@ int dlm_wait_for_node_death(struct dlm_ctxt *dlm, u8 node, int timeout) return 0; } +int dlm_wait_for_node_recovery(struct dlm_ctxt *dlm, u8 node, int timeout) +{ + if (timeout) { + mlog(0, "%s: waiting %dms for notification of " + "recovery of node %u\n", dlm->name, timeout, node); + wait_event_timeout(dlm->dlm_reco_thread_wq, + dlm_is_node_recovered(dlm, node), + msecs_to_jiffies(timeout)); + } else { + mlog(0, "%s: waiting indefinitely for notification " + "of recovery of node %u\n", dlm->name, node); + wait_event(dlm->dlm_reco_thread_wq, + dlm_is_node_recovered(dlm, node)); + } + /* for now, return 0 */ + return 0; +} + /* callers of the top-level api calls (dlmlock/dlmunlock) should * block on the dlm->reco.event when recovery is in progress. * the dlm recovery thread will set this state when it begins @@ -308,6 +421,13 @@ static int dlm_in_recovery(struct dlm_ctxt *dlm) void dlm_wait_for_recovery(struct dlm_ctxt *dlm) { + if (dlm_in_recovery(dlm)) { + mlog(0, "%s: reco thread %d in recovery: " + "state=%d, master=%u, dead=%u\n", + dlm->name, dlm->dlm_reco_thread_task->pid, + dlm->reco.state, dlm->reco.new_master, + dlm->reco.dead_node); + } wait_event(dlm->reco.event, !dlm_in_recovery(dlm)); } @@ -341,7 +461,7 @@ static int dlm_do_recovery(struct dlm_ctxt *dlm) mlog(0, "new master %u died while recovering %u!\n", dlm->reco.new_master, dlm->reco.dead_node); /* unset the new_master, leave dead_node */ - dlm->reco.new_master = O2NM_INVALID_NODE_NUM; + dlm_set_reco_master(dlm, O2NM_INVALID_NODE_NUM); } /* select a target to recover */ @@ -350,14 +470,14 @@ static int dlm_do_recovery(struct dlm_ctxt *dlm) bit = find_next_bit (dlm->recovery_map, O2NM_MAX_NODES+1, 0); if (bit >= O2NM_MAX_NODES || bit < 0) - dlm->reco.dead_node = O2NM_INVALID_NODE_NUM; + dlm_set_reco_dead_node(dlm, O2NM_INVALID_NODE_NUM); else - dlm->reco.dead_node = bit; + dlm_set_reco_dead_node(dlm, bit); } else if (!test_bit(dlm->reco.dead_node, dlm->recovery_map)) { /* BUG? */ mlog(ML_ERROR, "dead_node %u no longer in recovery map!\n", dlm->reco.dead_node); - dlm->reco.dead_node = O2NM_INVALID_NODE_NUM; + dlm_set_reco_dead_node(dlm, O2NM_INVALID_NODE_NUM); } if (dlm->reco.dead_node == O2NM_INVALID_NODE_NUM) { @@ -366,7 +486,8 @@ static int dlm_do_recovery(struct dlm_ctxt *dlm) /* return to main thread loop and sleep. */ return 0; } - mlog(0, "recovery thread found node %u in the recovery map!\n", + mlog(0, "%s(%d):recovery thread found node %u in the recovery map!\n", + dlm->name, dlm->dlm_reco_thread_task->pid, dlm->reco.dead_node); spin_unlock(&dlm->spinlock); @@ -389,8 +510,8 @@ static int dlm_do_recovery(struct dlm_ctxt *dlm) } mlog(0, "another node will master this recovery session.\n"); } - mlog(0, "dlm=%s, new_master=%u, this node=%u, dead_node=%u\n", - dlm->name, dlm->reco.new_master, + mlog(0, "dlm=%s (%d), new_master=%u, this node=%u, dead_node=%u\n", + dlm->name, dlm->dlm_reco_thread_task->pid, dlm->reco.new_master, dlm->node_num, dlm->reco.dead_node); /* it is safe to start everything back up here @@ -402,11 +523,13 @@ static int dlm_do_recovery(struct dlm_ctxt *dlm) return 0; master_here: - mlog(0, "mastering recovery of %s:%u here(this=%u)!\n", + mlog(0, "(%d) mastering recovery of %s:%u here(this=%u)!\n", + dlm->dlm_reco_thread_task->pid, dlm->name, dlm->reco.dead_node, dlm->node_num); status = dlm_remaster_locks(dlm, dlm->reco.dead_node); if (status < 0) { + /* we should never hit this anymore */ mlog(ML_ERROR, "error %d remastering locks for node %u, " "retrying.\n", status, dlm->reco.dead_node); /* yield a bit to allow any final network messages @@ -433,9 +556,16 @@ static int dlm_remaster_locks(struct dlm_ctxt *dlm, u8 dead_node) int destroy = 0; int pass = 0; - status = dlm_init_recovery_area(dlm, dead_node); - if (status < 0) - goto leave; + do { + /* we have become recovery master. there is no escaping + * this, so just keep trying until we get it. */ + status = dlm_init_recovery_area(dlm, dead_node); + if (status < 0) { + mlog(ML_ERROR, "%s: failed to alloc recovery area, " + "retrying\n", dlm->name); + msleep(1000); + } + } while (status != 0); /* safe to access the node data list without a lock, since this * process is the only one to change the list */ @@ -452,16 +582,36 @@ static int dlm_remaster_locks(struct dlm_ctxt *dlm, u8 dead_node) continue; } - status = dlm_request_all_locks(dlm, ndata->node_num, dead_node); - if (status < 0) { - mlog_errno(status); - if (dlm_is_host_down(status)) - ndata->state = DLM_RECO_NODE_DATA_DEAD; - else { - destroy = 1; - goto leave; + do { + status = dlm_request_all_locks(dlm, ndata->node_num, + dead_node); + if (status < 0) { + mlog_errno(status); + if (dlm_is_host_down(status)) { + /* node died, ignore it for recovery */ + status = 0; + ndata->state = DLM_RECO_NODE_DATA_DEAD; + /* wait for the domain map to catch up + * with the network state. */ + wait_event_timeout(dlm->dlm_reco_thread_wq, + dlm_is_node_dead(dlm, + ndata->node_num), + msecs_to_jiffies(1000)); + mlog(0, "waited 1 sec for %u, " + "dead? %s\n", ndata->node_num, + dlm_is_node_dead(dlm, ndata->node_num) ? + "yes" : "no"); + } else { + /* -ENOMEM on the other node */ + mlog(0, "%s: node %u returned " + "%d during recovery, retrying " + "after a short wait\n", + dlm->name, ndata->node_num, + status); + msleep(100); + } } - } + } while (status != 0); switch (ndata->state) { case DLM_RECO_NODE_DATA_INIT: @@ -473,10 +623,9 @@ static int dlm_remaster_locks(struct dlm_ctxt *dlm, u8 dead_node) mlog(0, "node %u died after requesting " "recovery info for node %u\n", ndata->node_num, dead_node); - // start all over - destroy = 1; - status = -EAGAIN; - goto leave; + /* fine. don't need this node's info. + * continue without it. */ + break; case DLM_RECO_NODE_DATA_REQUESTING: ndata->state = DLM_RECO_NODE_DATA_REQUESTED; mlog(0, "now receiving recovery data from " @@ -520,35 +669,26 @@ static int dlm_remaster_locks(struct dlm_ctxt *dlm, u8 dead_node) BUG(); break; case DLM_RECO_NODE_DATA_DEAD: - mlog(ML_NOTICE, "node %u died after " + mlog(0, "node %u died after " "requesting recovery info for " "node %u\n", ndata->node_num, dead_node); - spin_unlock(&dlm_reco_state_lock); - // start all over - destroy = 1; - status = -EAGAIN; - /* instead of spinning like crazy here, - * wait for the domain map to catch up - * with the network state. otherwise this - * can be hit hundreds of times before - * the node is really seen as dead. */ - wait_event_timeout(dlm->dlm_reco_thread_wq, - dlm_is_node_dead(dlm, - ndata->node_num), - msecs_to_jiffies(1000)); - mlog(0, "waited 1 sec for %u, " - "dead? %s\n", ndata->node_num, - dlm_is_node_dead(dlm, ndata->node_num) ? - "yes" : "no"); - goto leave; + break; case DLM_RECO_NODE_DATA_RECEIVING: case DLM_RECO_NODE_DATA_REQUESTED: + mlog(0, "%s: node %u still in state %s\n", + dlm->name, ndata->node_num, + ndata->state==DLM_RECO_NODE_DATA_RECEIVING ? + "receiving" : "requested"); all_nodes_done = 0; break; case DLM_RECO_NODE_DATA_DONE: + mlog(0, "%s: node %u state is done\n", + dlm->name, ndata->node_num); break; case DLM_RECO_NODE_DATA_FINALIZE_SENT: + mlog(0, "%s: node %u state is finalize\n", + dlm->name, ndata->node_num); break; } } @@ -578,7 +718,7 @@ static int dlm_remaster_locks(struct dlm_ctxt *dlm, u8 dead_node) jiffies, dlm->reco.dead_node, dlm->node_num, dlm->reco.new_master); destroy = 1; - status = ret; + status = 0; /* rescan everything marked dirty along the way */ dlm_kick_thread(dlm, NULL); break; @@ -591,7 +731,6 @@ static int dlm_remaster_locks(struct dlm_ctxt *dlm, u8 dead_node) } -leave: if (destroy) dlm_destroy_recovery_area(dlm, dead_node); @@ -617,7 +756,7 @@ static int dlm_init_recovery_area(struct dlm_ctxt *dlm, u8 dead_node) } BUG_ON(num == dead_node); - ndata = kcalloc(1, sizeof(*ndata), GFP_KERNEL); + ndata = kcalloc(1, sizeof(*ndata), GFP_NOFS); if (!ndata) { dlm_destroy_recovery_area(dlm, dead_node); return -ENOMEM; @@ -691,16 +830,25 @@ int dlm_request_all_locks_handler(struct o2net_msg *msg, u32 len, void *data) if (!dlm_grab(dlm)) return -EINVAL; + if (lr->dead_node != dlm->reco.dead_node) { + mlog(ML_ERROR, "%s: node %u sent dead_node=%u, but local " + "dead_node is %u\n", dlm->name, lr->node_idx, + lr->dead_node, dlm->reco.dead_node); + dlm_print_reco_node_status(dlm); + /* this is a hack */ + dlm_put(dlm); + return -ENOMEM; + } BUG_ON(lr->dead_node != dlm->reco.dead_node); - item = kcalloc(1, sizeof(*item), GFP_KERNEL); + item = kcalloc(1, sizeof(*item), GFP_NOFS); if (!item) { dlm_put(dlm); return -ENOMEM; } /* this will get freed by dlm_request_all_locks_worker */ - buf = (char *) __get_free_page(GFP_KERNEL); + buf = (char *) __get_free_page(GFP_NOFS); if (!buf) { kfree(item); dlm_put(dlm); @@ -715,7 +863,7 @@ int dlm_request_all_locks_handler(struct o2net_msg *msg, u32 len, void *data) spin_lock(&dlm->work_lock); list_add_tail(&item->list, &dlm->work_list); spin_unlock(&dlm->work_lock); - schedule_work(&dlm->dispatched_work); + queue_work(dlm->dlm_worker, &dlm->dispatched_work); dlm_put(dlm); return 0; @@ -730,32 +878,34 @@ static void dlm_request_all_locks_worker(struct dlm_work_item *item, void *data) struct list_head *iter; int ret; u8 dead_node, reco_master; + int skip_all_done = 0; dlm = item->dlm; dead_node = item->u.ral.dead_node; reco_master = item->u.ral.reco_master; mres = (struct dlm_migratable_lockres *)data; + mlog(0, "%s: recovery worker started, dead=%u, master=%u\n", + dlm->name, dead_node, reco_master); + if (dead_node != dlm->reco.dead_node || reco_master != dlm->reco.new_master) { - /* show extra debug info if the recovery state is messed */ - mlog(ML_ERROR, "%s: bad reco state: reco(dead=%u, master=%u), " - "request(dead=%u, master=%u)\n", - dlm->name, dlm->reco.dead_node, dlm->reco.new_master, - dead_node, reco_master); - mlog(ML_ERROR, "%s: name=%.*s master=%u locks=%u/%u flags=%u " - "entry[0]={c=%u:%llu,l=%u,f=%u,t=%d,ct=%d,hb=%d,n=%u}\n", - dlm->name, mres->lockname_len, mres->lockname, mres->master, - mres->num_locks, mres->total_locks, mres->flags, - dlm_get_lock_cookie_node(mres->ml[0].cookie), - dlm_get_lock_cookie_seq(mres->ml[0].cookie), - mres->ml[0].list, mres->ml[0].flags, - mres->ml[0].type, mres->ml[0].convert_type, - mres->ml[0].highest_blocked, mres->ml[0].node); - BUG(); + /* worker could have been created before the recovery master + * died. if so, do not continue, but do not error. */ + if (dlm->reco.new_master == O2NM_INVALID_NODE_NUM) { + mlog(ML_NOTICE, "%s: will not send recovery state, " + "recovery master %u died, thread=(dead=%u,mas=%u)" + " current=(dead=%u,mas=%u)\n", dlm->name, + reco_master, dead_node, reco_master, + dlm->reco.dead_node, dlm->reco.new_master); + } else { + mlog(ML_NOTICE, "%s: reco state invalid: reco(dead=%u, " + "master=%u), request(dead=%u, master=%u)\n", + dlm->name, dlm->reco.dead_node, + dlm->reco.new_master, dead_node, reco_master); + } + goto leave; } - BUG_ON(dead_node != dlm->reco.dead_node); - BUG_ON(reco_master != dlm->reco.new_master); /* lock resources should have already been moved to the * dlm->reco.resources list. now move items from that list @@ -766,12 +916,20 @@ static void dlm_request_all_locks_worker(struct dlm_work_item *item, void *data) dlm_move_reco_locks_to_list(dlm, &resources, dead_node); /* now we can begin blasting lockreses without the dlm lock */ + + /* any errors returned will be due to the new_master dying, + * the dlm_reco_thread should detect this */ list_for_each(iter, &resources) { res = list_entry (iter, struct dlm_lock_resource, recovering); ret = dlm_send_one_lockres(dlm, res, mres, reco_master, DLM_MRES_RECOVERY); - if (ret < 0) - mlog_errno(ret); + if (ret < 0) { + mlog(ML_ERROR, "%s: node %u went down while sending " + "recovery state for dead node %u, ret=%d\n", dlm->name, + reco_master, dead_node, ret); + skip_all_done = 1; + break; + } } /* move the resources back to the list */ @@ -779,10 +937,15 @@ static void dlm_request_all_locks_worker(struct dlm_work_item *item, void *data) list_splice_init(&resources, &dlm->reco.resources); spin_unlock(&dlm->spinlock); - ret = dlm_send_all_done_msg(dlm, dead_node, reco_master); - if (ret < 0) - mlog_errno(ret); - + if (!skip_all_done) { + ret = dlm_send_all_done_msg(dlm, dead_node, reco_master); + if (ret < 0) { + mlog(ML_ERROR, "%s: node %u went down while sending " + "recovery all-done for dead node %u, ret=%d\n", + dlm->name, reco_master, dead_node, ret); + } + } +leave: free_page((unsigned long)data); } @@ -801,8 +964,14 @@ static int dlm_send_all_done_msg(struct dlm_ctxt *dlm, u8 dead_node, u8 send_to) ret = o2net_send_message(DLM_RECO_DATA_DONE_MSG, dlm->key, &done_msg, sizeof(done_msg), send_to, &tmpret); - /* negative status is ignored by the caller */ - if (ret >= 0) + if (ret < 0) { + if (!dlm_is_host_down(ret)) { + mlog_errno(ret); + mlog(ML_ERROR, "%s: unknown error sending data-done " + "to %u\n", dlm->name, send_to); + BUG(); + } + } else ret = tmpret; return ret; } @@ -822,7 +991,11 @@ int dlm_reco_data_done_handler(struct o2net_msg *msg, u32 len, void *data) mlog(0, "got DATA DONE: dead_node=%u, reco.dead_node=%u, " "node_idx=%u, this node=%u\n", done->dead_node, dlm->reco.dead_node, done->node_idx, dlm->node_num); - BUG_ON(done->dead_node != dlm->reco.dead_node); + + mlog_bug_on_msg((done->dead_node != dlm->reco.dead_node), + "Got DATA DONE: dead_node=%u, reco.dead_node=%u, " + "node_idx=%u, this node=%u\n", done->dead_node, + dlm->reco.dead_node, done->node_idx, dlm->node_num); spin_lock(&dlm_reco_state_lock); list_for_each(iter, &dlm->reco.node_data) { @@ -905,13 +1078,11 @@ static void dlm_move_reco_locks_to_list(struct dlm_ctxt *dlm, mlog(0, "found lockres owned by dead node while " "doing recovery for node %u. sending it.\n", dead_node); - list_del_init(&res->recovering); - list_add_tail(&res->recovering, list); + list_move_tail(&res->recovering, list); } else if (res->owner == DLM_LOCK_RES_OWNER_UNKNOWN) { mlog(0, "found UNKNOWN owner while doing recovery " "for node %u. sending it.\n", dead_node); - list_del_init(&res->recovering); - list_add_tail(&res->recovering, list); + list_move_tail(&res->recovering, list); } } spin_unlock(&dlm->spinlock); @@ -1023,8 +1194,9 @@ static int dlm_add_lock_to_array(struct dlm_lock *lock, ml->type == LKM_PRMODE) { /* if it is already set, this had better be a PR * and it has to match */ - if (mres->lvb[0] && (ml->type == LKM_EXMODE || - memcmp(mres->lvb, lock->lksb->lvb, DLM_LVB_LEN))) { + if (!dlm_lvb_is_empty(mres->lvb) && + (ml->type == LKM_EXMODE || + memcmp(mres->lvb, lock->lksb->lvb, DLM_LVB_LEN))) { mlog(ML_ERROR, "mismatched lvbs!\n"); __dlm_print_one_lock_resource(lock->lockres); BUG(); @@ -1083,22 +1255,25 @@ int dlm_send_one_lockres(struct dlm_ctxt *dlm, struct dlm_lock_resource *res, * we must send it immediately. */ ret = dlm_send_mig_lockres_msg(dlm, mres, send_to, res, total_locks); - if (ret < 0) { - // TODO - mlog(ML_ERROR, "dlm_send_mig_lockres_msg " - "returned %d, TODO\n", ret); - BUG(); - } + if (ret < 0) + goto error; } } /* flush any remaining locks */ ret = dlm_send_mig_lockres_msg(dlm, mres, send_to, res, total_locks); - if (ret < 0) { - // TODO - mlog(ML_ERROR, "dlm_send_mig_lockres_msg returned %d, " - "TODO\n", ret); + if (ret < 0) + goto error; + return ret; + +error: + mlog(ML_ERROR, "%s: dlm_send_mig_lockres_msg returned %d\n", + dlm->name, ret); + if (!dlm_is_host_down(ret)) BUG(); - } + mlog(0, "%s: node %u went down while sending %s " + "lockres %.*s\n", dlm->name, send_to, + flags & DLM_MRES_RECOVERY ? "recovery" : "migration", + res->lockname.len, res->lockname.name); return ret; } @@ -1146,8 +1321,8 @@ int dlm_mig_lockres_handler(struct o2net_msg *msg, u32 len, void *data) mlog(0, "all done flag. all lockres data received!\n"); ret = -ENOMEM; - buf = kmalloc(be16_to_cpu(msg->data_len), GFP_KERNEL); - item = kcalloc(1, sizeof(*item), GFP_KERNEL); + buf = kmalloc(be16_to_cpu(msg->data_len), GFP_NOFS); + item = kcalloc(1, sizeof(*item), GFP_NOFS); if (!buf || !item) goto leave; @@ -1238,7 +1413,7 @@ int dlm_mig_lockres_handler(struct o2net_msg *msg, u32 len, void *data) spin_lock(&dlm->work_lock); list_add_tail(&item->list, &dlm->work_list); spin_unlock(&dlm->work_lock); - schedule_work(&dlm->dispatched_work); + queue_work(dlm->dlm_worker, &dlm->dispatched_work); leave: dlm_put(dlm); @@ -1312,8 +1487,9 @@ leave: -int dlm_lockres_master_requery(struct dlm_ctxt *dlm, - struct dlm_lock_resource *res, u8 *real_master) +static int dlm_lockres_master_requery(struct dlm_ctxt *dlm, + struct dlm_lock_resource *res, + u8 *real_master) { struct dlm_node_iter iter; int nodenum; @@ -1406,6 +1582,7 @@ int dlm_master_requery_handler(struct o2net_msg *msg, u32 len, void *data) struct dlm_ctxt *dlm = data; struct dlm_master_requery *req = (struct dlm_master_requery *)msg->buf; struct dlm_lock_resource *res = NULL; + unsigned int hash; int master = DLM_LOCK_RES_OWNER_UNKNOWN; u32 flags = DLM_ASSERT_MASTER_REQUERY; @@ -1415,8 +1592,10 @@ int dlm_master_requery_handler(struct o2net_msg *msg, u32 len, void *data) return master; } + hash = dlm_lockid_hash(req->name, req->namelen); + spin_lock(&dlm->spinlock); - res = __dlm_lookup_lockres(dlm, req->name, req->namelen); + res = __dlm_lookup_lockres(dlm, req->name, req->namelen, hash); if (res) { spin_lock(&res->spinlock); master = res->owner; @@ -1483,7 +1662,7 @@ static int dlm_process_recovery_data(struct dlm_ctxt *dlm, struct dlm_lock *newlock = NULL; struct dlm_lockstatus *lksb = NULL; int ret = 0; - int i; + int i, bad; struct list_head *iter; struct dlm_lock *lock = NULL; @@ -1529,8 +1708,7 @@ static int dlm_process_recovery_data(struct dlm_ctxt *dlm, /* move the lock to its proper place */ /* do not alter lock refcount. switching lists. */ - list_del_init(&lock->list); - list_add_tail(&lock->list, queue); + list_move_tail(&lock->list, queue); spin_unlock(&res->spinlock); mlog(0, "just reordered a local lock!\n"); @@ -1553,28 +1731,48 @@ static int dlm_process_recovery_data(struct dlm_ctxt *dlm, } lksb->flags |= (ml->flags & (DLM_LKSB_PUT_LVB|DLM_LKSB_GET_LVB)); - - if (mres->lvb[0]) { + + if (ml->type == LKM_NLMODE) + goto skip_lvb; + + if (!dlm_lvb_is_empty(mres->lvb)) { if (lksb->flags & DLM_LKSB_PUT_LVB) { /* other node was trying to update * lvb when node died. recreate the * lksb with the updated lvb. */ memcpy(lksb->lvb, mres->lvb, DLM_LVB_LEN); + /* the lock resource lvb update must happen + * NOW, before the spinlock is dropped. + * we no longer wait for the AST to update + * the lvb. */ + memcpy(res->lvb, mres->lvb, DLM_LVB_LEN); } else { /* otherwise, the node is sending its * most recent valid lvb info */ BUG_ON(ml->type != LKM_EXMODE && ml->type != LKM_PRMODE); - if (res->lvb[0] && (ml->type == LKM_EXMODE || - memcmp(res->lvb, mres->lvb, DLM_LVB_LEN))) { - mlog(ML_ERROR, "received bad lvb!\n"); - __dlm_print_one_lock_resource(res); - BUG(); + if (!dlm_lvb_is_empty(res->lvb) && + (ml->type == LKM_EXMODE || + memcmp(res->lvb, mres->lvb, DLM_LVB_LEN))) { + int i; + mlog(ML_ERROR, "%s:%.*s: received bad " + "lvb! type=%d\n", dlm->name, + res->lockname.len, + res->lockname.name, ml->type); + printk("lockres lvb=["); + for (i=0; i<DLM_LVB_LEN; i++) + printk("%02x", res->lvb[i]); + printk("]\nmigrated lvb=["); + for (i=0; i<DLM_LVB_LEN; i++) + printk("%02x", mres->lvb[i]); + printk("]\n"); + dlm_print_one_lock_resource(res); + BUG(); } memcpy(res->lvb, mres->lvb, DLM_LVB_LEN); } } - +skip_lvb: /* NOTE: * wrt lock queue ordering and recovery: @@ -1592,9 +1790,33 @@ static int dlm_process_recovery_data(struct dlm_ctxt *dlm, * relative to each other, but clearly *not* * preserved relative to locks from other nodes. */ + bad = 0; spin_lock(&res->spinlock); - dlm_lock_get(newlock); - list_add_tail(&newlock->list, queue); + list_for_each_entry(lock, queue, list) { + if (lock->ml.cookie == ml->cookie) { + u64 c = lock->ml.cookie; + mlog(ML_ERROR, "%s:%.*s: %u:%llu: lock already " + "exists on this lockres!\n", dlm->name, + res->lockname.len, res->lockname.name, + dlm_get_lock_cookie_node(c), + dlm_get_lock_cookie_seq(c)); + + mlog(ML_NOTICE, "sent lock: type=%d, conv=%d, " + "node=%u, cookie=%u:%llu, queue=%d\n", + ml->type, ml->convert_type, ml->node, + dlm_get_lock_cookie_node(ml->cookie), + dlm_get_lock_cookie_seq(ml->cookie), + ml->list); + + __dlm_print_one_lock_resource(res); + bad = 1; + break; + } + } + if (!bad) { + dlm_lock_get(newlock); + list_add_tail(&newlock->list, queue); + } spin_unlock(&res->spinlock); } mlog(0, "done running all the locks\n"); @@ -1618,8 +1840,14 @@ void dlm_move_lockres_to_recovery_list(struct dlm_ctxt *dlm, struct dlm_lock *lock; res->state |= DLM_LOCK_RES_RECOVERING; - if (!list_empty(&res->recovering)) + if (!list_empty(&res->recovering)) { + mlog(0, + "Recovering res %s:%.*s, is already on recovery list!\n", + dlm->name, res->lockname.len, res->lockname.name); list_del_init(&res->recovering); + } + /* We need to hold a reference while on the recovery list */ + dlm_lockres_get(res); list_add_tail(&res->recovering, &dlm->reco.resources); /* find any pending locks and put them back on proper list */ @@ -1708,9 +1936,11 @@ static void dlm_finish_local_lockres_recovery(struct dlm_ctxt *dlm, spin_lock(&res->spinlock); dlm_change_lockres_owner(dlm, res, new_master); res->state &= ~DLM_LOCK_RES_RECOVERING; - __dlm_dirty_lockres(dlm, res); + if (!__dlm_lockres_unused(res)) + __dlm_dirty_lockres(dlm, res); spin_unlock(&res->spinlock); wake_up(&res->wq); + dlm_lockres_put(res); } } @@ -1719,7 +1949,7 @@ static void dlm_finish_local_lockres_recovery(struct dlm_ctxt *dlm, * the RECOVERING state and set the owner * if necessary */ for (i = 0; i < DLM_HASH_BUCKETS; i++) { - bucket = &(dlm->lockres_hash[i]); + bucket = dlm_lockres_hash(dlm, i); hlist_for_each_entry(res, hash_iter, bucket, hash_node) { if (res->state & DLM_LOCK_RES_RECOVERING) { if (res->owner == dead_node) { @@ -1743,11 +1973,13 @@ static void dlm_finish_local_lockres_recovery(struct dlm_ctxt *dlm, dlm->name, res->lockname.len, res->lockname.name, res->owner); list_del_init(&res->recovering); + dlm_lockres_put(res); } spin_lock(&res->spinlock); dlm_change_lockres_owner(dlm, res, new_master); res->state &= ~DLM_LOCK_RES_RECOVERING; - __dlm_dirty_lockres(dlm, res); + if (!__dlm_lockres_unused(res)) + __dlm_dirty_lockres(dlm, res); spin_unlock(&res->spinlock); wake_up(&res->wq); } @@ -1884,7 +2116,7 @@ static void dlm_do_local_recovery_cleanup(struct dlm_ctxt *dlm, u8 dead_node) * need to be fired as a result. */ for (i = 0; i < DLM_HASH_BUCKETS; i++) { - bucket = &(dlm->lockres_hash[i]); + bucket = dlm_lockres_hash(dlm, i); hlist_for_each_entry(res, iter, bucket, hash_node) { /* always prune any $RECOVERY entries for dead nodes, * otherwise hangs can occur during later recovery */ @@ -1924,6 +2156,20 @@ static void __dlm_hb_node_down(struct dlm_ctxt *dlm, int idx) { assert_spin_locked(&dlm->spinlock); + if (dlm->reco.new_master == idx) { + mlog(0, "%s: recovery master %d just died\n", + dlm->name, idx); + if (dlm->reco.state & DLM_RECO_STATE_FINALIZE) { + /* finalize1 was reached, so it is safe to clear + * the new_master and dead_node. that recovery + * is complete. */ + mlog(0, "%s: dead master %d had reached " + "finalize1 state, clearing\n", dlm->name, idx); + dlm->reco.state &= ~DLM_RECO_STATE_FINALIZE; + __dlm_reset_recovery(dlm); + } + } + /* check to see if the node is already considered dead */ if (!test_bit(idx, dlm->live_nodes_map)) { mlog(0, "for domain %s, node %d is already dead. " @@ -2039,7 +2285,8 @@ again: memset(&lksb, 0, sizeof(lksb)); ret = dlmlock(dlm, LKM_EXMODE, &lksb, LKM_NOQUEUE|LKM_RECOVERY, - DLM_RECOVERY_LOCK_NAME, dlm_reco_ast, dlm, dlm_reco_bast); + DLM_RECOVERY_LOCK_NAME, DLM_RECOVERY_LOCK_NAME_LEN, + dlm_reco_ast, dlm, dlm_reco_bast); mlog(0, "%s: dlmlock($RECOVERY) returned %d, lksb=%d\n", dlm->name, ret, lksb.status); @@ -2087,7 +2334,7 @@ again: /* set the new_master to this node */ spin_lock(&dlm->spinlock); - dlm->reco.new_master = dlm->node_num; + dlm_set_reco_master(dlm, dlm->node_num); spin_unlock(&dlm->spinlock); } @@ -2125,6 +2372,10 @@ again: mlog(0, "%s: reco master %u is ready to recover %u\n", dlm->name, dlm->reco.new_master, dlm->reco.dead_node); status = -EEXIST; + } else if (ret == DLM_RECOVERING) { + mlog(0, "dlm=%s dlmlock says master node died (this=%u)\n", + dlm->name, dlm->node_num); + goto again; } else { struct dlm_lock_resource *res; @@ -2156,7 +2407,7 @@ static int dlm_send_begin_reco_message(struct dlm_ctxt *dlm, u8 dead_node) mlog_entry("%u\n", dead_node); - mlog(0, "dead node is %u\n", dead_node); + mlog(0, "%s: dead node is %u\n", dlm->name, dead_node); spin_lock(&dlm->spinlock); dlm_node_iter_init(dlm->domain_map, &iter); @@ -2214,6 +2465,14 @@ retry: * another ENOMEM */ msleep(100); goto retry; + } else if (ret == EAGAIN) { + mlog(0, "%s: trying to start recovery of node " + "%u, but node %u is waiting for last recovery " + "to complete, backoff for a bit\n", dlm->name, + dead_node, nodenum); + /* TODO Look into replacing msleep with cond_resched() */ + msleep(100); + goto retry; } } @@ -2229,8 +2488,20 @@ int dlm_begin_reco_handler(struct o2net_msg *msg, u32 len, void *data) if (!dlm_grab(dlm)) return 0; - mlog(0, "node %u wants to recover node %u\n", - br->node_idx, br->dead_node); + spin_lock(&dlm->spinlock); + if (dlm->reco.state & DLM_RECO_STATE_FINALIZE) { + mlog(0, "%s: node %u wants to recover node %u (%u:%u) " + "but this node is in finalize state, waiting on finalize2\n", + dlm->name, br->node_idx, br->dead_node, + dlm->reco.dead_node, dlm->reco.new_master); + spin_unlock(&dlm->spinlock); + return EAGAIN; + } + spin_unlock(&dlm->spinlock); + + mlog(0, "%s: node %u wants to recover node %u (%u:%u)\n", + dlm->name, br->node_idx, br->dead_node, + dlm->reco.dead_node, dlm->reco.new_master); dlm_fire_domain_eviction_callbacks(dlm, br->dead_node); @@ -2252,8 +2523,8 @@ int dlm_begin_reco_handler(struct o2net_msg *msg, u32 len, void *data) "node %u changing it to %u\n", dlm->name, dlm->reco.dead_node, br->node_idx, br->dead_node); } - dlm->reco.new_master = br->node_idx; - dlm->reco.dead_node = br->dead_node; + dlm_set_reco_master(dlm, br->node_idx); + dlm_set_reco_dead_node(dlm, br->dead_node); if (!test_bit(br->dead_node, dlm->recovery_map)) { mlog(0, "recovery master %u sees %u as dead, but this " "node has not yet. marking %u as dead\n", @@ -2272,10 +2543,16 @@ int dlm_begin_reco_handler(struct o2net_msg *msg, u32 len, void *data) spin_unlock(&dlm->spinlock); dlm_kick_recovery_thread(dlm); + + mlog(0, "%s: recovery started by node %u, for %u (%u:%u)\n", + dlm->name, br->node_idx, br->dead_node, + dlm->reco.dead_node, dlm->reco.new_master); + dlm_put(dlm); return 0; } +#define DLM_FINALIZE_STAGE2 0x01 static int dlm_send_finalize_reco_message(struct dlm_ctxt *dlm) { int ret = 0; @@ -2283,25 +2560,31 @@ static int dlm_send_finalize_reco_message(struct dlm_ctxt *dlm) struct dlm_node_iter iter; int nodenum; int status; + int stage = 1; - mlog(0, "finishing recovery for node %s:%u\n", - dlm->name, dlm->reco.dead_node); + mlog(0, "finishing recovery for node %s:%u, " + "stage %d\n", dlm->name, dlm->reco.dead_node, stage); spin_lock(&dlm->spinlock); dlm_node_iter_init(dlm->domain_map, &iter); spin_unlock(&dlm->spinlock); +stage2: memset(&fr, 0, sizeof(fr)); fr.node_idx = dlm->node_num; fr.dead_node = dlm->reco.dead_node; + if (stage == 2) + fr.flags |= DLM_FINALIZE_STAGE2; while ((nodenum = dlm_node_iter_next(&iter)) >= 0) { if (nodenum == dlm->node_num) continue; ret = o2net_send_message(DLM_FINALIZE_RECO_MSG, dlm->key, &fr, sizeof(fr), nodenum, &status); - if (ret >= 0) { + if (ret >= 0) ret = status; + if (ret < 0) { + mlog_errno(ret); if (dlm_is_host_down(ret)) { /* this has no effect on this recovery * session, so set the status to zero to @@ -2309,13 +2592,17 @@ static int dlm_send_finalize_reco_message(struct dlm_ctxt *dlm) mlog(ML_ERROR, "node %u went down after this " "node finished recovery.\n", nodenum); ret = 0; + continue; } - } - if (ret < 0) { - mlog_errno(ret); break; } } + if (stage == 1) { + /* reset the node_iter back to the top and send finalize2 */ + iter.curnode = -1; + stage = 2; + goto stage2; + } return ret; } @@ -2324,14 +2611,19 @@ int dlm_finalize_reco_handler(struct o2net_msg *msg, u32 len, void *data) { struct dlm_ctxt *dlm = data; struct dlm_finalize_reco *fr = (struct dlm_finalize_reco *)msg->buf; + int stage = 1; /* ok to return 0, domain has gone away */ if (!dlm_grab(dlm)) return 0; - mlog(0, "node %u finalizing recovery of node %u\n", - fr->node_idx, fr->dead_node); + if (fr->flags & DLM_FINALIZE_STAGE2) + stage = 2; + mlog(0, "%s: node %u finalizing recovery stage%d of " + "node %u (%u:%u)\n", dlm->name, fr->node_idx, stage, + fr->dead_node, dlm->reco.dead_node, dlm->reco.new_master); + spin_lock(&dlm->spinlock); if (dlm->reco.new_master != fr->node_idx) { @@ -2347,13 +2639,41 @@ int dlm_finalize_reco_handler(struct o2net_msg *msg, u32 len, void *data) BUG(); } - dlm_finish_local_lockres_recovery(dlm, fr->dead_node, fr->node_idx); - - spin_unlock(&dlm->spinlock); + switch (stage) { + case 1: + dlm_finish_local_lockres_recovery(dlm, fr->dead_node, fr->node_idx); + if (dlm->reco.state & DLM_RECO_STATE_FINALIZE) { + mlog(ML_ERROR, "%s: received finalize1 from " + "new master %u for dead node %u, but " + "this node has already received it!\n", + dlm->name, fr->node_idx, fr->dead_node); + dlm_print_reco_node_status(dlm); + BUG(); + } + dlm->reco.state |= DLM_RECO_STATE_FINALIZE; + spin_unlock(&dlm->spinlock); + break; + case 2: + if (!(dlm->reco.state & DLM_RECO_STATE_FINALIZE)) { + mlog(ML_ERROR, "%s: received finalize2 from " + "new master %u for dead node %u, but " + "this node did not have finalize1!\n", + dlm->name, fr->node_idx, fr->dead_node); + dlm_print_reco_node_status(dlm); + BUG(); + } + dlm->reco.state &= ~DLM_RECO_STATE_FINALIZE; + spin_unlock(&dlm->spinlock); + dlm_reset_recovery(dlm); + dlm_kick_recovery_thread(dlm); + break; + default: + BUG(); + } - dlm_reset_recovery(dlm); + mlog(0, "%s: recovery done, reco master was %u, dead now %u, master now %u\n", + dlm->name, fr->node_idx, dlm->reco.dead_node, dlm->reco.new_master); - dlm_kick_recovery_thread(dlm); dlm_put(dlm); return 0; } diff --git a/fs/ocfs2/dlm/dlmthread.c b/fs/ocfs2/dlm/dlmthread.c index 5be9d14f12c..0c822f3ffb0 100644 --- a/fs/ocfs2/dlm/dlmthread.c +++ b/fs/ocfs2/dlm/dlmthread.c @@ -39,6 +39,7 @@ #include <linux/inet.h> #include <linux/timer.h> #include <linux/kthread.h> +#include <linux/delay.h> #include "cluster/heartbeat.h" @@ -53,6 +54,8 @@ #include "cluster/masklog.h" static int dlm_thread(void *data); +static void dlm_purge_lockres_now(struct dlm_ctxt *dlm, + struct dlm_lock_resource *lockres); static void dlm_flush_asts(struct dlm_ctxt *dlm); @@ -80,7 +83,7 @@ repeat: } -static int __dlm_lockres_unused(struct dlm_lock_resource *res) +int __dlm_lockres_unused(struct dlm_lock_resource *res) { if (list_empty(&res->granted) && list_empty(&res->converting) && @@ -103,6 +106,20 @@ void __dlm_lockres_calc_usage(struct dlm_ctxt *dlm, assert_spin_locked(&res->spinlock); if (__dlm_lockres_unused(res)){ + /* For now, just keep any resource we master */ + if (res->owner == dlm->node_num) + { + if (!list_empty(&res->purge)) { + mlog(0, "we master %s:%.*s, but it is on " + "the purge list. Removing\n", + dlm->name, res->lockname.len, + res->lockname.name); + list_del_init(&res->purge); + dlm->purge_count--; + } + return; + } + if (list_empty(&res->purge)) { mlog(0, "putting lockres %.*s from purge list\n", res->lockname.len, res->lockname.name); @@ -110,10 +127,23 @@ void __dlm_lockres_calc_usage(struct dlm_ctxt *dlm, res->last_used = jiffies; list_add_tail(&res->purge, &dlm->purge_list); dlm->purge_count++; + + /* if this node is not the owner, there is + * no way to keep track of who the owner could be. + * unhash it to avoid serious problems. */ + if (res->owner != dlm->node_num) { + mlog(0, "%s:%.*s: doing immediate " + "purge of lockres owned by %u\n", + dlm->name, res->lockname.len, + res->lockname.name, res->owner); + + dlm_purge_lockres_now(dlm, res); + } } } else if (!list_empty(&res->purge)) { - mlog(0, "removing lockres %.*s from purge list\n", - res->lockname.len, res->lockname.name); + mlog(0, "removing lockres %.*s from purge list, " + "owner=%u\n", res->lockname.len, res->lockname.name, + res->owner); list_del_init(&res->purge); dlm->purge_count--; @@ -165,6 +195,7 @@ again: } else if (ret < 0) { mlog(ML_NOTICE, "lockres %.*s: migrate failed, retrying\n", lockres->lockname.len, lockres->lockname.name); + msleep(100); goto again; } @@ -178,6 +209,24 @@ finish: __dlm_unhash_lockres(lockres); } +/* make an unused lockres go away immediately. + * as soon as the dlm spinlock is dropped, this lockres + * will not be found. kfree still happens on last put. */ +static void dlm_purge_lockres_now(struct dlm_ctxt *dlm, + struct dlm_lock_resource *lockres) +{ + assert_spin_locked(&dlm->spinlock); + assert_spin_locked(&lockres->spinlock); + + BUG_ON(!__dlm_lockres_unused(lockres)); + + if (!list_empty(&lockres->purge)) { + list_del_init(&lockres->purge); + dlm->purge_count--; + } + __dlm_unhash_lockres(lockres); +} + static void dlm_run_purge_list(struct dlm_ctxt *dlm, int purge_now) { @@ -318,8 +367,7 @@ converting: target->ml.type = target->ml.convert_type; target->ml.convert_type = LKM_IVMODE; - list_del_init(&target->list); - list_add_tail(&target->list, &res->granted); + list_move_tail(&target->list, &res->granted); BUG_ON(!target->lksb); target->lksb->status = DLM_NORMAL; @@ -380,8 +428,7 @@ blocked: target->ml.type, target->ml.node); // target->ml.type is already correct - list_del_init(&target->list); - list_add_tail(&target->list, &res->granted); + list_move_tail(&target->list, &res->granted); BUG_ON(!target->lksb); target->lksb->status = DLM_NORMAL; @@ -422,6 +469,8 @@ void __dlm_dirty_lockres(struct dlm_ctxt *dlm, struct dlm_lock_resource *res) /* don't shuffle secondary queues */ if ((res->owner == dlm->node_num) && !(res->state & DLM_LOCK_RES_DIRTY)) { + /* ref for dirty_list */ + dlm_lockres_get(res); list_add_tail(&res->dirty, &dlm->dirty_list); res->state |= DLM_LOCK_RES_DIRTY; } @@ -606,6 +655,8 @@ static int dlm_thread(void *data) list_del_init(&res->dirty); spin_unlock(&res->spinlock); spin_unlock(&dlm->spinlock); + /* Drop dirty_list ref */ + dlm_lockres_put(res); /* lockres can be re-dirtied/re-added to the * dirty_list in this gap, but that is ok */ @@ -642,8 +693,9 @@ static int dlm_thread(void *data) * spinlock and do NOT have the dlm lock. * safe to reserve/queue asts and run the lists. */ - mlog(0, "calling dlm_shuffle_lists with dlm=%p, " - "res=%p\n", dlm, res); + mlog(0, "calling dlm_shuffle_lists with dlm=%s, " + "res=%.*s\n", dlm->name, + res->lockname.len, res->lockname.name); /* called while holding lockres lock */ dlm_shuffle_lists(dlm, res); @@ -657,6 +709,8 @@ in_progress: /* if the lock was in-progress, stick * it on the back of the list */ if (delay) { + /* ref for dirty_list */ + dlm_lockres_get(res); spin_lock(&res->spinlock); list_add_tail(&res->dirty, &dlm->dirty_list); res->state |= DLM_LOCK_RES_DIRTY; @@ -677,7 +731,7 @@ in_progress: /* yield and continue right away if there is more work to do */ if (!n) { - yield(); + cond_resched(); continue; } diff --git a/fs/ocfs2/dlm/dlmunlock.c b/fs/ocfs2/dlm/dlmunlock.c index 7b1a2754267..37be4b2e0d4 100644 --- a/fs/ocfs2/dlm/dlmunlock.c +++ b/fs/ocfs2/dlm/dlmunlock.c @@ -155,7 +155,7 @@ static enum dlm_status dlmunlock_common(struct dlm_ctxt *dlm, else status = dlm_get_unlock_actions(dlm, res, lock, lksb, &actions); - if (status != DLM_NORMAL) + if (status != DLM_NORMAL && (status != DLM_CANCELGRANT || !master_node)) goto leave; /* By now this has been masked out of cancel requests. */ @@ -183,8 +183,7 @@ static enum dlm_status dlmunlock_common(struct dlm_ctxt *dlm, spin_lock(&lock->spinlock); /* if the master told us the lock was already granted, * let the ast handle all of these actions */ - if (status == DLM_NORMAL && - lksb->status == DLM_CANCELGRANT) { + if (status == DLM_CANCELGRANT) { actions &= ~(DLM_UNLOCK_REMOVE_LOCK| DLM_UNLOCK_REGRANT_LOCK| DLM_UNLOCK_CLEAR_CONVERT_TYPE); @@ -271,8 +270,7 @@ void dlm_commit_pending_unlock(struct dlm_lock_resource *res, void dlm_commit_pending_cancel(struct dlm_lock_resource *res, struct dlm_lock *lock) { - list_del_init(&lock->list); - list_add_tail(&lock->list, &res->granted); + list_move_tail(&lock->list, &res->granted); lock->ml.convert_type = LKM_IVMODE; } @@ -319,6 +317,16 @@ static enum dlm_status dlm_send_remote_unlock_request(struct dlm_ctxt *dlm, mlog_entry("%.*s\n", res->lockname.len, res->lockname.name); + if (owner == dlm->node_num) { + /* ended up trying to contact ourself. this means + * that the lockres had been remote but became local + * via a migration. just retry it, now as local */ + mlog(0, "%s:%.*s: this node became the master due to a " + "migration, re-evaluate now\n", dlm->name, + res->lockname.len, res->lockname.name); + return DLM_FORWARD; + } + memset(&unlock, 0, sizeof(unlock)); unlock.node_idx = dlm->node_num; unlock.flags = cpu_to_be32(flags); @@ -340,14 +348,9 @@ static enum dlm_status dlm_send_remote_unlock_request(struct dlm_ctxt *dlm, vec, veclen, owner, &status); if (tmpret >= 0) { // successfully sent and received - if (status == DLM_CANCELGRANT) - ret = DLM_NORMAL; - else if (status == DLM_FORWARD) { + if (status == DLM_FORWARD) mlog(0, "master was in-progress. retry\n"); - ret = DLM_FORWARD; - } else - ret = status; - lksb->status = status; + ret = status; } else { mlog_errno(tmpret); if (dlm_is_host_down(tmpret)) { @@ -363,7 +366,6 @@ static enum dlm_status dlm_send_remote_unlock_request(struct dlm_ctxt *dlm, /* something bad. this will BUG in ocfs2 */ ret = dlm_err_to_dlm_status(tmpret); } - lksb->status = ret; } return ret; @@ -474,6 +476,10 @@ int dlm_unlock_lock_handler(struct o2net_msg *msg, u32 len, void *data) /* lock was found on queue */ lksb = lock->lksb; + if (flags & (LKM_VALBLK|LKM_PUT_LVB) && + lock->ml.type != LKM_EXMODE) + flags &= ~(LKM_VALBLK|LKM_PUT_LVB); + /* unlockast only called on originating node */ if (flags & LKM_PUT_LVB) { lksb->flags |= DLM_LKSB_PUT_LVB; @@ -498,11 +504,8 @@ not_found: "cookie=%u:%llu\n", dlm_get_lock_cookie_node(unlock->cookie), dlm_get_lock_cookie_seq(unlock->cookie)); - else { - /* send the lksb->status back to the other node */ - status = lksb->status; + else dlm_lock_put(lock); - } leave: if (res) @@ -524,26 +527,22 @@ static enum dlm_status dlm_get_cancel_actions(struct dlm_ctxt *dlm, if (dlm_lock_on_list(&res->blocked, lock)) { /* cancel this outright */ - lksb->status = DLM_NORMAL; status = DLM_NORMAL; *actions = (DLM_UNLOCK_CALL_AST | DLM_UNLOCK_REMOVE_LOCK); } else if (dlm_lock_on_list(&res->converting, lock)) { /* cancel the request, put back on granted */ - lksb->status = DLM_NORMAL; status = DLM_NORMAL; *actions = (DLM_UNLOCK_CALL_AST | DLM_UNLOCK_REMOVE_LOCK | DLM_UNLOCK_REGRANT_LOCK | DLM_UNLOCK_CLEAR_CONVERT_TYPE); } else if (dlm_lock_on_list(&res->granted, lock)) { - /* too late, already granted. DLM_CANCELGRANT */ - lksb->status = DLM_CANCELGRANT; - status = DLM_NORMAL; + /* too late, already granted. */ + status = DLM_CANCELGRANT; *actions = DLM_UNLOCK_CALL_AST; } else { mlog(ML_ERROR, "lock to cancel is not on any list!\n"); - lksb->status = DLM_IVLOCKID; status = DLM_IVLOCKID; *actions = 0; } @@ -560,13 +559,11 @@ static enum dlm_status dlm_get_unlock_actions(struct dlm_ctxt *dlm, /* unlock request */ if (!dlm_lock_on_list(&res->granted, lock)) { - lksb->status = DLM_DENIED; status = DLM_DENIED; dlm_error(status); *actions = 0; } else { /* unlock granted lock */ - lksb->status = DLM_NORMAL; status = DLM_NORMAL; *actions = (DLM_UNLOCK_FREE_LOCK | DLM_UNLOCK_CALL_AST | @@ -623,6 +620,8 @@ retry: spin_lock(&res->spinlock); is_master = (res->owner == dlm->node_num); + if (flags & LKM_VALBLK && lock->ml.type != LKM_EXMODE) + flags &= ~LKM_VALBLK; spin_unlock(&res->spinlock); if (is_master) { @@ -656,7 +655,7 @@ retry: } if (call_ast) { - mlog(0, "calling unlockast(%p, %d)\n", data, lksb->status); + mlog(0, "calling unlockast(%p, %d)\n", data, status); if (is_master) { /* it is possible that there is one last bast * pending. make sure it is flushed, then @@ -668,9 +667,12 @@ retry: wait_event(dlm->ast_wq, dlm_lock_basts_flushed(dlm, lock)); } - (*unlockast)(data, lksb->status); + (*unlockast)(data, status); } + if (status == DLM_CANCELGRANT) + status = DLM_NORMAL; + if (status == DLM_NORMAL) { mlog(0, "kicking the thread\n"); dlm_kick_thread(dlm, res); diff --git a/fs/ocfs2/dlm/userdlm.c b/fs/ocfs2/dlm/userdlm.c index 74ca4e5f976..eead48bbfac 100644 --- a/fs/ocfs2/dlm/userdlm.c +++ b/fs/ocfs2/dlm/userdlm.c @@ -102,10 +102,10 @@ static inline void user_recover_from_dlm_error(struct user_lock_res *lockres) spin_unlock(&lockres->l_lock); } -#define user_log_dlm_error(_func, _stat, _lockres) do { \ - mlog(ML_ERROR, "Dlm error \"%s\" while calling %s on " \ - "resource %s: %s\n", dlm_errname(_stat), _func, \ - _lockres->l_name, dlm_errmsg(_stat)); \ +#define user_log_dlm_error(_func, _stat, _lockres) do { \ + mlog(ML_ERROR, "Dlm error \"%s\" while calling %s on " \ + "resource %.*s: %s\n", dlm_errname(_stat), _func, \ + _lockres->l_namelen, _lockres->l_name, dlm_errmsg(_stat)); \ } while (0) /* WARNING: This function lives in a world where the only three lock @@ -127,21 +127,22 @@ static void user_ast(void *opaque) struct user_lock_res *lockres = opaque; struct dlm_lockstatus *lksb; - mlog(0, "AST fired for lockres %s\n", lockres->l_name); + mlog(0, "AST fired for lockres %.*s\n", lockres->l_namelen, + lockres->l_name); spin_lock(&lockres->l_lock); lksb = &(lockres->l_lksb); if (lksb->status != DLM_NORMAL) { - mlog(ML_ERROR, "lksb status value of %u on lockres %s\n", - lksb->status, lockres->l_name); + mlog(ML_ERROR, "lksb status value of %u on lockres %.*s\n", + lksb->status, lockres->l_namelen, lockres->l_name); spin_unlock(&lockres->l_lock); return; } mlog_bug_on_msg(lockres->l_requested == LKM_IVMODE, - "Lockres %s, requested ivmode. flags 0x%x\n", - lockres->l_name, lockres->l_flags); + "Lockres %.*s, requested ivmode. flags 0x%x\n", + lockres->l_namelen, lockres->l_name, lockres->l_flags); /* we're downconverting. */ if (lockres->l_requested < lockres->l_level) { @@ -213,8 +214,8 @@ static void user_bast(void *opaque, int level) { struct user_lock_res *lockres = opaque; - mlog(0, "Blocking AST fired for lockres %s. Blocking level %d\n", - lockres->l_name, level); + mlog(0, "Blocking AST fired for lockres %.*s. Blocking level %d\n", + lockres->l_namelen, lockres->l_name, level); spin_lock(&lockres->l_lock); lockres->l_flags |= USER_LOCK_BLOCKED; @@ -231,7 +232,8 @@ static void user_unlock_ast(void *opaque, enum dlm_status status) { struct user_lock_res *lockres = opaque; - mlog(0, "UNLOCK AST called on lock %s\n", lockres->l_name); + mlog(0, "UNLOCK AST called on lock %.*s\n", lockres->l_namelen, + lockres->l_name); if (status != DLM_NORMAL && status != DLM_CANCELGRANT) mlog(ML_ERROR, "Dlm returns status %d\n", status); @@ -244,8 +246,6 @@ static void user_unlock_ast(void *opaque, enum dlm_status status) && !(lockres->l_flags & USER_LOCK_IN_CANCEL)) { lockres->l_level = LKM_IVMODE; } else if (status == DLM_CANCELGRANT) { - mlog(0, "Lock %s, cancel fails, flags 0x%x\n", - lockres->l_name, lockres->l_flags); /* We tried to cancel a convert request, but it was * already granted. Don't clear the busy flag - the * ast should've done this already. */ @@ -255,8 +255,6 @@ static void user_unlock_ast(void *opaque, enum dlm_status status) } else { BUG_ON(!(lockres->l_flags & USER_LOCK_IN_CANCEL)); /* Cancel succeeded, we want to re-queue */ - mlog(0, "Lock %s, cancel succeeds, flags 0x%x\n", - lockres->l_name, lockres->l_flags); lockres->l_requested = LKM_IVMODE; /* cancel an * upconvert * request. */ @@ -287,13 +285,14 @@ static void user_dlm_unblock_lock(void *opaque) struct user_lock_res *lockres = (struct user_lock_res *) opaque; struct dlm_ctxt *dlm = dlm_ctxt_from_user_lockres(lockres); - mlog(0, "processing lockres %s\n", lockres->l_name); + mlog(0, "processing lockres %.*s\n", lockres->l_namelen, + lockres->l_name); spin_lock(&lockres->l_lock); mlog_bug_on_msg(!(lockres->l_flags & USER_LOCK_QUEUED), - "Lockres %s, flags 0x%x\n", - lockres->l_name, lockres->l_flags); + "Lockres %.*s, flags 0x%x\n", + lockres->l_namelen, lockres->l_name, lockres->l_flags); /* notice that we don't clear USER_LOCK_BLOCKED here. If it's * set, we want user_ast clear it. */ @@ -305,22 +304,16 @@ static void user_dlm_unblock_lock(void *opaque) * flag, and finally we might get another bast which re-queues * us before our ast for the downconvert is called. */ if (!(lockres->l_flags & USER_LOCK_BLOCKED)) { - mlog(0, "Lockres %s, flags 0x%x: queued but not blocking\n", - lockres->l_name, lockres->l_flags); spin_unlock(&lockres->l_lock); goto drop_ref; } if (lockres->l_flags & USER_LOCK_IN_TEARDOWN) { - mlog(0, "lock is in teardown so we do nothing\n"); spin_unlock(&lockres->l_lock); goto drop_ref; } if (lockres->l_flags & USER_LOCK_BUSY) { - mlog(0, "Cancel lock %s, flags 0x%x\n", - lockres->l_name, lockres->l_flags); - if (lockres->l_flags & USER_LOCK_IN_CANCEL) { spin_unlock(&lockres->l_lock); goto drop_ref; @@ -372,6 +365,7 @@ static void user_dlm_unblock_lock(void *opaque) &lockres->l_lksb, LKM_CONVERT|LKM_VALBLK, lockres->l_name, + lockres->l_namelen, user_ast, lockres, user_bast); @@ -420,16 +414,16 @@ int user_dlm_cluster_lock(struct user_lock_res *lockres, if (level != LKM_EXMODE && level != LKM_PRMODE) { - mlog(ML_ERROR, "lockres %s: invalid request!\n", - lockres->l_name); + mlog(ML_ERROR, "lockres %.*s: invalid request!\n", + lockres->l_namelen, lockres->l_name); status = -EINVAL; goto bail; } - mlog(0, "lockres %s: asking for %s lock, passed flags = 0x%x\n", - lockres->l_name, - (level == LKM_EXMODE) ? "LKM_EXMODE" : "LKM_PRMODE", - lkm_flags); + mlog(0, "lockres %.*s: asking for %s lock, passed flags = 0x%x\n", + lockres->l_namelen, lockres->l_name, + (level == LKM_EXMODE) ? "LKM_EXMODE" : "LKM_PRMODE", + lkm_flags); again: if (signal_pending(current)) { @@ -474,15 +468,13 @@ again: BUG_ON(level == LKM_IVMODE); BUG_ON(level == LKM_NLMODE); - mlog(0, "lock %s, get lock from %d to level = %d\n", - lockres->l_name, lockres->l_level, level); - /* call dlm_lock to upgrade lock now */ status = dlmlock(dlm, level, &lockres->l_lksb, local_flags, lockres->l_name, + lockres->l_namelen, user_ast, lockres, user_bast); @@ -498,9 +490,6 @@ again: goto bail; } - mlog(0, "lock %s, successfull return from dlmlock\n", - lockres->l_name); - user_wait_on_busy_lock(lockres); goto again; } @@ -508,9 +497,6 @@ again: user_dlm_inc_holders(lockres, level); spin_unlock(&lockres->l_lock); - mlog(0, "lockres %s: Got %s lock!\n", lockres->l_name, - (level == LKM_EXMODE) ? "LKM_EXMODE" : "LKM_PRMODE"); - status = 0; bail: return status; @@ -538,13 +524,11 @@ void user_dlm_cluster_unlock(struct user_lock_res *lockres, { if (level != LKM_EXMODE && level != LKM_PRMODE) { - mlog(ML_ERROR, "lockres %s: invalid request!\n", lockres->l_name); + mlog(ML_ERROR, "lockres %.*s: invalid request!\n", + lockres->l_namelen, lockres->l_name); return; } - mlog(0, "lockres %s: dropping %s lock\n", lockres->l_name, - (level == LKM_EXMODE) ? "LKM_EXMODE" : "LKM_PRMODE"); - spin_lock(&lockres->l_lock); user_dlm_dec_holders(lockres, level); __user_dlm_cond_queue_lockres(lockres); @@ -602,6 +586,7 @@ void user_dlm_lock_res_init(struct user_lock_res *lockres, memcpy(lockres->l_name, dentry->d_name.name, dentry->d_name.len); + lockres->l_namelen = dentry->d_name.len; } int user_dlm_destroy_lock(struct user_lock_res *lockres) @@ -609,11 +594,10 @@ int user_dlm_destroy_lock(struct user_lock_res *lockres) int status = -EBUSY; struct dlm_ctxt *dlm = dlm_ctxt_from_user_lockres(lockres); - mlog(0, "asked to destroy %s\n", lockres->l_name); + mlog(0, "asked to destroy %.*s\n", lockres->l_namelen, lockres->l_name); spin_lock(&lockres->l_lock); if (lockres->l_flags & USER_LOCK_IN_TEARDOWN) { - mlog(0, "Lock is already torn down\n"); spin_unlock(&lockres->l_lock); return 0; } @@ -623,8 +607,6 @@ int user_dlm_destroy_lock(struct user_lock_res *lockres) while (lockres->l_flags & USER_LOCK_BUSY) { spin_unlock(&lockres->l_lock); - mlog(0, "lock %s is busy\n", lockres->l_name); - user_wait_on_busy_lock(lockres); spin_lock(&lockres->l_lock); @@ -632,14 +614,12 @@ int user_dlm_destroy_lock(struct user_lock_res *lockres) if (lockres->l_ro_holders || lockres->l_ex_holders) { spin_unlock(&lockres->l_lock); - mlog(0, "lock %s has holders\n", lockres->l_name); goto bail; } status = 0; if (!(lockres->l_flags & USER_LOCK_ATTACHED)) { spin_unlock(&lockres->l_lock); - mlog(0, "lock %s is not attached\n", lockres->l_name); goto bail; } @@ -647,7 +627,6 @@ int user_dlm_destroy_lock(struct user_lock_res *lockres) lockres->l_flags |= USER_LOCK_BUSY; spin_unlock(&lockres->l_lock); - mlog(0, "unlocking lockres %s\n", lockres->l_name); status = dlmunlock(dlm, &lockres->l_lksb, LKM_VALBLK, @@ -672,7 +651,7 @@ struct dlm_ctxt *user_dlm_register_context(struct qstr *name) u32 dlm_key; char *domain; - domain = kmalloc(name->len + 1, GFP_KERNEL); + domain = kmalloc(name->len + 1, GFP_NOFS); if (!domain) { mlog_errno(-ENOMEM); return ERR_PTR(-ENOMEM); diff --git a/fs/ocfs2/dlm/userdlm.h b/fs/ocfs2/dlm/userdlm.h index 04178bc40b7..c400e93bbf7 100644 --- a/fs/ocfs2/dlm/userdlm.h +++ b/fs/ocfs2/dlm/userdlm.h @@ -53,6 +53,7 @@ struct user_lock_res { #define USER_DLM_LOCK_ID_MAX_LEN 32 char l_name[USER_DLM_LOCK_ID_MAX_LEN]; + int l_namelen; int l_level; unsigned int l_ro_holders; unsigned int l_ex_holders; diff --git a/fs/ocfs2/dlmglue.c b/fs/ocfs2/dlmglue.c index 64cd52860c8..8801e41afe8 100644 --- a/fs/ocfs2/dlmglue.c +++ b/fs/ocfs2/dlmglue.c @@ -46,6 +46,7 @@ #include "ocfs2.h" #include "alloc.h" +#include "dcache.h" #include "dlmglue.h" #include "extent_map.h" #include "heartbeat.h" @@ -66,78 +67,161 @@ struct ocfs2_mask_waiter { unsigned long mw_goal; }; -static void ocfs2_inode_ast_func(void *opaque); -static void ocfs2_inode_bast_func(void *opaque, - int level); -static void ocfs2_super_ast_func(void *opaque); -static void ocfs2_super_bast_func(void *opaque, - int level); -static void ocfs2_rename_ast_func(void *opaque); -static void ocfs2_rename_bast_func(void *opaque, - int level); - -/* so far, all locks have gotten along with the same unlock ast */ -static void ocfs2_unlock_ast_func(void *opaque, - enum dlm_status status); -static int ocfs2_do_unblock_meta(struct inode *inode, - int *requeue); -static int ocfs2_unblock_meta(struct ocfs2_lock_res *lockres, - int *requeue); -static int ocfs2_unblock_data(struct ocfs2_lock_res *lockres, - int *requeue); -static int ocfs2_unblock_inode_lock(struct ocfs2_lock_res *lockres, - int *requeue); -static int ocfs2_unblock_osb_lock(struct ocfs2_lock_res *lockres, - int *requeue); -typedef void (ocfs2_convert_worker_t)(struct ocfs2_lock_res *, int); -static int ocfs2_generic_unblock_lock(struct ocfs2_super *osb, - struct ocfs2_lock_res *lockres, - int *requeue, - ocfs2_convert_worker_t *worker); +static struct ocfs2_super *ocfs2_get_dentry_osb(struct ocfs2_lock_res *lockres); +static struct ocfs2_super *ocfs2_get_inode_osb(struct ocfs2_lock_res *lockres); +/* + * Return value from ->downconvert_worker functions. + * + * These control the precise actions of ocfs2_unblock_lock() + * and ocfs2_process_blocked_lock() + * + */ +enum ocfs2_unblock_action { + UNBLOCK_CONTINUE = 0, /* Continue downconvert */ + UNBLOCK_CONTINUE_POST = 1, /* Continue downconvert, fire + * ->post_unlock callback */ + UNBLOCK_STOP_POST = 2, /* Do not downconvert, fire + * ->post_unlock() callback. */ +}; + +struct ocfs2_unblock_ctl { + int requeue; + enum ocfs2_unblock_action unblock_action; +}; + +static int ocfs2_check_meta_downconvert(struct ocfs2_lock_res *lockres, + int new_level); +static void ocfs2_set_meta_lvb(struct ocfs2_lock_res *lockres); + +static int ocfs2_data_convert_worker(struct ocfs2_lock_res *lockres, + int blocking); + +static int ocfs2_dentry_convert_worker(struct ocfs2_lock_res *lockres, + int blocking); + +static void ocfs2_dentry_post_unlock(struct ocfs2_super *osb, + struct ocfs2_lock_res *lockres); + +/* + * OCFS2 Lock Resource Operations + * + * These fine tune the behavior of the generic dlmglue locking infrastructure. + * + * The most basic of lock types can point ->l_priv to their respective + * struct ocfs2_super and allow the default actions to manage things. + * + * Right now, each lock type also needs to implement an init function, + * and trivial lock/unlock wrappers. ocfs2_simple_drop_lockres() + * should be called when the lock is no longer needed (i.e., object + * destruction time). + */ struct ocfs2_lock_res_ops { - void (*ast)(void *); - void (*bast)(void *, int); - void (*unlock_ast)(void *, enum dlm_status); - int (*unblock)(struct ocfs2_lock_res *, int *); + /* + * Translate an ocfs2_lock_res * into an ocfs2_super *. Define + * this callback if ->l_priv is not an ocfs2_super pointer + */ + struct ocfs2_super * (*get_osb)(struct ocfs2_lock_res *); + + /* + * Optionally called in the downconvert (or "vote") thread + * after a successful downconvert. The lockres will not be + * referenced after this callback is called, so it is safe to + * free memory, etc. + * + * The exact semantics of when this is called are controlled + * by ->downconvert_worker() + */ + void (*post_unlock)(struct ocfs2_super *, struct ocfs2_lock_res *); + + /* + * Allow a lock type to add checks to determine whether it is + * safe to downconvert a lock. Return 0 to re-queue the + * downconvert at a later time, nonzero to continue. + * + * For most locks, the default checks that there are no + * incompatible holders are sufficient. + * + * Called with the lockres spinlock held. + */ + int (*check_downconvert)(struct ocfs2_lock_res *, int); + + /* + * Allows a lock type to populate the lock value block. This + * is called on downconvert, and when we drop a lock. + * + * Locks that want to use this should set LOCK_TYPE_USES_LVB + * in the flags field. + * + * Called with the lockres spinlock held. + */ + void (*set_lvb)(struct ocfs2_lock_res *); + + /* + * Called from the downconvert thread when it is determined + * that a lock will be downconverted. This is called without + * any locks held so the function can do work that might + * schedule (syncing out data, etc). + * + * This should return any one of the ocfs2_unblock_action + * values, depending on what it wants the thread to do. + */ + int (*downconvert_worker)(struct ocfs2_lock_res *, int); + + /* + * LOCK_TYPE_* flags which describe the specific requirements + * of a lock type. Descriptions of each individual flag follow. + */ + int flags; }; +/* + * Some locks want to "refresh" potentially stale data when a + * meaningful (PRMODE or EXMODE) lock level is first obtained. If this + * flag is set, the OCFS2_LOCK_NEEDS_REFRESH flag will be set on the + * individual lockres l_flags member from the ast function. It is + * expected that the locking wrapper will clear the + * OCFS2_LOCK_NEEDS_REFRESH flag when done. + */ +#define LOCK_TYPE_REQUIRES_REFRESH 0x1 + +/* + * Indicate that a lock type makes use of the lock value block. The + * ->set_lvb lock type callback must be defined. + */ +#define LOCK_TYPE_USES_LVB 0x2 + static struct ocfs2_lock_res_ops ocfs2_inode_rw_lops = { - .ast = ocfs2_inode_ast_func, - .bast = ocfs2_inode_bast_func, - .unlock_ast = ocfs2_unlock_ast_func, - .unblock = ocfs2_unblock_inode_lock, + .get_osb = ocfs2_get_inode_osb, + .flags = 0, }; static struct ocfs2_lock_res_ops ocfs2_inode_meta_lops = { - .ast = ocfs2_inode_ast_func, - .bast = ocfs2_inode_bast_func, - .unlock_ast = ocfs2_unlock_ast_func, - .unblock = ocfs2_unblock_meta, + .get_osb = ocfs2_get_inode_osb, + .check_downconvert = ocfs2_check_meta_downconvert, + .set_lvb = ocfs2_set_meta_lvb, + .flags = LOCK_TYPE_REQUIRES_REFRESH|LOCK_TYPE_USES_LVB, }; -static void ocfs2_data_convert_worker(struct ocfs2_lock_res *lockres, - int blocking); - static struct ocfs2_lock_res_ops ocfs2_inode_data_lops = { - .ast = ocfs2_inode_ast_func, - .bast = ocfs2_inode_bast_func, - .unlock_ast = ocfs2_unlock_ast_func, - .unblock = ocfs2_unblock_data, + .get_osb = ocfs2_get_inode_osb, + .downconvert_worker = ocfs2_data_convert_worker, + .flags = 0, }; static struct ocfs2_lock_res_ops ocfs2_super_lops = { - .ast = ocfs2_super_ast_func, - .bast = ocfs2_super_bast_func, - .unlock_ast = ocfs2_unlock_ast_func, - .unblock = ocfs2_unblock_osb_lock, + .flags = LOCK_TYPE_REQUIRES_REFRESH, }; static struct ocfs2_lock_res_ops ocfs2_rename_lops = { - .ast = ocfs2_rename_ast_func, - .bast = ocfs2_rename_bast_func, - .unlock_ast = ocfs2_unlock_ast_func, - .unblock = ocfs2_unblock_osb_lock, + .flags = 0, +}; + +static struct ocfs2_lock_res_ops ocfs2_dentry_lops = { + .get_osb = ocfs2_get_dentry_osb, + .post_unlock = ocfs2_dentry_post_unlock, + .downconvert_worker = ocfs2_dentry_convert_worker, + .flags = 0, }; static inline int ocfs2_is_inode_lock(struct ocfs2_lock_res *lockres) @@ -147,29 +231,26 @@ static inline int ocfs2_is_inode_lock(struct ocfs2_lock_res *lockres) lockres->l_type == OCFS2_LOCK_TYPE_RW; } -static inline int ocfs2_is_super_lock(struct ocfs2_lock_res *lockres) +static inline struct inode *ocfs2_lock_res_inode(struct ocfs2_lock_res *lockres) { - return lockres->l_type == OCFS2_LOCK_TYPE_SUPER; -} + BUG_ON(!ocfs2_is_inode_lock(lockres)); -static inline int ocfs2_is_rename_lock(struct ocfs2_lock_res *lockres) -{ - return lockres->l_type == OCFS2_LOCK_TYPE_RENAME; + return (struct inode *) lockres->l_priv; } -static inline struct ocfs2_super *ocfs2_lock_res_super(struct ocfs2_lock_res *lockres) +static inline struct ocfs2_dentry_lock *ocfs2_lock_res_dl(struct ocfs2_lock_res *lockres) { - BUG_ON(!ocfs2_is_super_lock(lockres) - && !ocfs2_is_rename_lock(lockres)); + BUG_ON(lockres->l_type != OCFS2_LOCK_TYPE_DENTRY); - return (struct ocfs2_super *) lockres->l_priv; + return (struct ocfs2_dentry_lock *)lockres->l_priv; } -static inline struct inode *ocfs2_lock_res_inode(struct ocfs2_lock_res *lockres) +static inline struct ocfs2_super *ocfs2_get_lockres_osb(struct ocfs2_lock_res *lockres) { - BUG_ON(!ocfs2_is_inode_lock(lockres)); + if (lockres->l_ops->get_osb) + return lockres->l_ops->get_osb(lockres); - return (struct inode *) lockres->l_priv; + return (struct ocfs2_super *)lockres->l_priv; } static int ocfs2_lock_create(struct ocfs2_super *osb, @@ -200,25 +281,6 @@ static int ocfs2_meta_lock_update(struct inode *inode, struct buffer_head **bh); static void ocfs2_drop_osb_locks(struct ocfs2_super *osb); static inline int ocfs2_highest_compat_lock_level(int level); -static inline int ocfs2_can_downconvert_meta_lock(struct inode *inode, - struct ocfs2_lock_res *lockres, - int new_level); - -static char *ocfs2_lock_type_strings[] = { - [OCFS2_LOCK_TYPE_META] = "Meta", - [OCFS2_LOCK_TYPE_DATA] = "Data", - [OCFS2_LOCK_TYPE_SUPER] = "Super", - [OCFS2_LOCK_TYPE_RENAME] = "Rename", - /* Need to differntiate from [R]ename.. serializing writes is the - * important job it does, anyway. */ - [OCFS2_LOCK_TYPE_RW] = "Write/Read", -}; - -static char *ocfs2_lock_type_string(enum ocfs2_lock_type type) -{ - mlog_bug_on_msg(type >= OCFS2_NUM_LOCK_TYPES, "%d\n", type); - return ocfs2_lock_type_strings[type]; -} static void ocfs2_build_lock_name(enum ocfs2_lock_type type, u64 blkno, @@ -242,7 +304,7 @@ static void ocfs2_build_lock_name(enum ocfs2_lock_type type, mlog_exit_void(); } -static spinlock_t ocfs2_dlm_tracking_lock = SPIN_LOCK_UNLOCKED; +static DEFINE_SPINLOCK(ocfs2_dlm_tracking_lock); static void ocfs2_add_lockres_tracking(struct ocfs2_lock_res *res, struct ocfs2_dlm_debug *dlm_debug) @@ -265,13 +327,9 @@ static void ocfs2_remove_lockres_tracking(struct ocfs2_lock_res *res) static void ocfs2_lock_res_init_common(struct ocfs2_super *osb, struct ocfs2_lock_res *res, enum ocfs2_lock_type type, - u64 blkno, - u32 generation, struct ocfs2_lock_res_ops *ops, void *priv) { - ocfs2_build_lock_name(type, blkno, generation, res->l_name); - res->l_type = type; res->l_ops = ops; res->l_priv = priv; @@ -299,6 +357,7 @@ void ocfs2_lock_res_init_once(struct ocfs2_lock_res *res) void ocfs2_inode_lock_res_init(struct ocfs2_lock_res *res, enum ocfs2_lock_type type, + unsigned int generation, struct inode *inode) { struct ocfs2_lock_res_ops *ops; @@ -319,9 +378,73 @@ void ocfs2_inode_lock_res_init(struct ocfs2_lock_res *res, break; }; - ocfs2_lock_res_init_common(OCFS2_SB(inode->i_sb), res, type, - OCFS2_I(inode)->ip_blkno, - inode->i_generation, ops, inode); + ocfs2_build_lock_name(type, OCFS2_I(inode)->ip_blkno, + generation, res->l_name); + ocfs2_lock_res_init_common(OCFS2_SB(inode->i_sb), res, type, ops, inode); +} + +static struct ocfs2_super *ocfs2_get_inode_osb(struct ocfs2_lock_res *lockres) +{ + struct inode *inode = ocfs2_lock_res_inode(lockres); + + return OCFS2_SB(inode->i_sb); +} + +static __u64 ocfs2_get_dentry_lock_ino(struct ocfs2_lock_res *lockres) +{ + __be64 inode_blkno_be; + + memcpy(&inode_blkno_be, &lockres->l_name[OCFS2_DENTRY_LOCK_INO_START], + sizeof(__be64)); + + return be64_to_cpu(inode_blkno_be); +} + +static struct ocfs2_super *ocfs2_get_dentry_osb(struct ocfs2_lock_res *lockres) +{ + struct ocfs2_dentry_lock *dl = lockres->l_priv; + + return OCFS2_SB(dl->dl_inode->i_sb); +} + +void ocfs2_dentry_lock_res_init(struct ocfs2_dentry_lock *dl, + u64 parent, struct inode *inode) +{ + int len; + u64 inode_blkno = OCFS2_I(inode)->ip_blkno; + __be64 inode_blkno_be = cpu_to_be64(inode_blkno); + struct ocfs2_lock_res *lockres = &dl->dl_lockres; + + ocfs2_lock_res_init_once(lockres); + + /* + * Unfortunately, the standard lock naming scheme won't work + * here because we have two 16 byte values to use. Instead, + * we'll stuff the inode number as a binary value. We still + * want error prints to show something without garbling the + * display, so drop a null byte in there before the inode + * number. A future version of OCFS2 will likely use all + * binary lock names. The stringified names have been a + * tremendous aid in debugging, but now that the debugfs + * interface exists, we can mangle things there if need be. + * + * NOTE: We also drop the standard "pad" value (the total lock + * name size stays the same though - the last part is all + * zeros due to the memset in ocfs2_lock_res_init_once() + */ + len = snprintf(lockres->l_name, OCFS2_DENTRY_LOCK_INO_START, + "%c%016llx", + ocfs2_lock_type_char(OCFS2_LOCK_TYPE_DENTRY), + (long long)parent); + + BUG_ON(len != (OCFS2_DENTRY_LOCK_INO_START - 1)); + + memcpy(&lockres->l_name[OCFS2_DENTRY_LOCK_INO_START], &inode_blkno_be, + sizeof(__be64)); + + ocfs2_lock_res_init_common(OCFS2_SB(inode->i_sb), lockres, + OCFS2_LOCK_TYPE_DENTRY, &ocfs2_dentry_lops, + dl); } static void ocfs2_super_lock_res_init(struct ocfs2_lock_res *res, @@ -330,8 +453,9 @@ static void ocfs2_super_lock_res_init(struct ocfs2_lock_res *res, /* Superblock lockres doesn't come from a slab so we call init * once on it manually. */ ocfs2_lock_res_init_once(res); + ocfs2_build_lock_name(OCFS2_LOCK_TYPE_SUPER, OCFS2_SUPER_BLOCK_BLKNO, + 0, res->l_name); ocfs2_lock_res_init_common(osb, res, OCFS2_LOCK_TYPE_SUPER, - OCFS2_SUPER_BLOCK_BLKNO, 0, &ocfs2_super_lops, osb); } @@ -341,7 +465,8 @@ static void ocfs2_rename_lock_res_init(struct ocfs2_lock_res *res, /* Rename lockres doesn't come from a slab so we call init * once on it manually. */ ocfs2_lock_res_init_once(res); - ocfs2_lock_res_init_common(osb, res, OCFS2_LOCK_TYPE_RENAME, 0, 0, + ocfs2_build_lock_name(OCFS2_LOCK_TYPE_RENAME, 0, 0, res->l_name); + ocfs2_lock_res_init_common(osb, res, OCFS2_LOCK_TYPE_RENAME, &ocfs2_rename_lops, osb); } @@ -495,7 +620,8 @@ static inline void ocfs2_generic_handle_convert_action(struct ocfs2_lock_res *lo * information is already up to data. Convert from NL to * *anything* however should mark ourselves as needing an * update */ - if (lockres->l_level == LKM_NLMODE) + if (lockres->l_level == LKM_NLMODE && + lockres->l_ops->flags & LOCK_TYPE_REQUIRES_REFRESH) lockres_or_flags(lockres, OCFS2_LOCK_NEEDS_REFRESH); lockres->l_level = lockres->l_requested; @@ -512,7 +638,8 @@ static inline void ocfs2_generic_handle_attach_action(struct ocfs2_lock_res *loc BUG_ON(lockres->l_flags & OCFS2_LOCK_ATTACHED); if (lockres->l_requested > LKM_NLMODE && - !(lockres->l_flags & OCFS2_LOCK_LOCAL)) + !(lockres->l_flags & OCFS2_LOCK_LOCAL) && + lockres->l_ops->flags & LOCK_TYPE_REQUIRES_REFRESH) lockres_or_flags(lockres, OCFS2_LOCK_NEEDS_REFRESH); lockres->l_level = lockres->l_requested; @@ -522,68 +649,6 @@ static inline void ocfs2_generic_handle_attach_action(struct ocfs2_lock_res *loc mlog_exit_void(); } -static void ocfs2_inode_ast_func(void *opaque) -{ - struct ocfs2_lock_res *lockres = opaque; - struct inode *inode; - struct dlm_lockstatus *lksb; - unsigned long flags; - - mlog_entry_void(); - - inode = ocfs2_lock_res_inode(lockres); - - mlog(0, "AST fired for inode %llu, l_action = %u, type = %s\n", - (unsigned long long)OCFS2_I(inode)->ip_blkno, lockres->l_action, - ocfs2_lock_type_string(lockres->l_type)); - - BUG_ON(!ocfs2_is_inode_lock(lockres)); - - spin_lock_irqsave(&lockres->l_lock, flags); - - lksb = &(lockres->l_lksb); - if (lksb->status != DLM_NORMAL) { - mlog(ML_ERROR, "ocfs2_inode_ast_func: lksb status value of %u " - "on inode %llu\n", lksb->status, - (unsigned long long)OCFS2_I(inode)->ip_blkno); - spin_unlock_irqrestore(&lockres->l_lock, flags); - mlog_exit_void(); - return; - } - - switch(lockres->l_action) { - case OCFS2_AST_ATTACH: - ocfs2_generic_handle_attach_action(lockres); - lockres_clear_flags(lockres, OCFS2_LOCK_LOCAL); - break; - case OCFS2_AST_CONVERT: - ocfs2_generic_handle_convert_action(lockres); - break; - case OCFS2_AST_DOWNCONVERT: - ocfs2_generic_handle_downconvert_action(lockres); - break; - default: - mlog(ML_ERROR, "lockres %s: ast fired with invalid action: %u " - "lockres flags = 0x%lx, unlock action: %u\n", - lockres->l_name, lockres->l_action, lockres->l_flags, - lockres->l_unlock_action); - - BUG(); - } - - /* data and rw locking ignores refresh flag for now. */ - if (lockres->l_type != OCFS2_LOCK_TYPE_META) - lockres_clear_flags(lockres, OCFS2_LOCK_NEEDS_REFRESH); - - /* set it to something invalid so if we get called again we - * can catch it. */ - lockres->l_action = OCFS2_AST_INVALID; - spin_unlock_irqrestore(&lockres->l_lock, flags); - wake_up(&lockres->l_event); - - mlog_exit_void(); -} - static int ocfs2_generic_handle_bast(struct ocfs2_lock_res *lockres, int level) { @@ -610,54 +675,33 @@ static int ocfs2_generic_handle_bast(struct ocfs2_lock_res *lockres, return needs_downconvert; } -static void ocfs2_generic_bast_func(struct ocfs2_super *osb, - struct ocfs2_lock_res *lockres, - int level) +static void ocfs2_blocking_ast(void *opaque, int level) { + struct ocfs2_lock_res *lockres = opaque; + struct ocfs2_super *osb = ocfs2_get_lockres_osb(lockres); int needs_downconvert; unsigned long flags; - mlog_entry_void(); - BUG_ON(level <= LKM_NLMODE); + mlog(0, "BAST fired for lockres %s, blocking %d, level %d type %s\n", + lockres->l_name, level, lockres->l_level, + ocfs2_lock_type_string(lockres->l_type)); + spin_lock_irqsave(&lockres->l_lock, flags); needs_downconvert = ocfs2_generic_handle_bast(lockres, level); if (needs_downconvert) ocfs2_schedule_blocked_lock(osb, lockres); spin_unlock_irqrestore(&lockres->l_lock, flags); - ocfs2_kick_vote_thread(osb); - wake_up(&lockres->l_event); - mlog_exit_void(); -} - -static void ocfs2_inode_bast_func(void *opaque, int level) -{ - struct ocfs2_lock_res *lockres = opaque; - struct inode *inode; - struct ocfs2_super *osb; - - mlog_entry_void(); - - BUG_ON(!ocfs2_is_inode_lock(lockres)); - inode = ocfs2_lock_res_inode(lockres); - osb = OCFS2_SB(inode->i_sb); - - mlog(0, "BAST fired for inode %llu, blocking %d, level %d type %s\n", - (unsigned long long)OCFS2_I(inode)->ip_blkno, level, - lockres->l_level, ocfs2_lock_type_string(lockres->l_type)); - - ocfs2_generic_bast_func(osb, lockres, level); - - mlog_exit_void(); + ocfs2_kick_vote_thread(osb); } -static void ocfs2_generic_ast_func(struct ocfs2_lock_res *lockres, - int ignore_refresh) +static void ocfs2_locking_ast(void *opaque) { + struct ocfs2_lock_res *lockres = opaque; struct dlm_lockstatus *lksb = &lockres->l_lksb; unsigned long flags; @@ -673,6 +717,7 @@ static void ocfs2_generic_ast_func(struct ocfs2_lock_res *lockres, switch(lockres->l_action) { case OCFS2_AST_ATTACH: ocfs2_generic_handle_attach_action(lockres); + lockres_clear_flags(lockres, OCFS2_LOCK_LOCAL); break; case OCFS2_AST_CONVERT: ocfs2_generic_handle_convert_action(lockres); @@ -681,80 +726,19 @@ static void ocfs2_generic_ast_func(struct ocfs2_lock_res *lockres, ocfs2_generic_handle_downconvert_action(lockres); break; default: + mlog(ML_ERROR, "lockres %s: ast fired with invalid action: %u " + "lockres flags = 0x%lx, unlock action: %u\n", + lockres->l_name, lockres->l_action, lockres->l_flags, + lockres->l_unlock_action); BUG(); } - if (ignore_refresh) - lockres_clear_flags(lockres, OCFS2_LOCK_NEEDS_REFRESH); - /* set it to something invalid so if we get called again we * can catch it. */ lockres->l_action = OCFS2_AST_INVALID; - spin_unlock_irqrestore(&lockres->l_lock, flags); wake_up(&lockres->l_event); -} - -static void ocfs2_super_ast_func(void *opaque) -{ - struct ocfs2_lock_res *lockres = opaque; - - mlog_entry_void(); - mlog(0, "Superblock AST fired\n"); - - BUG_ON(!ocfs2_is_super_lock(lockres)); - ocfs2_generic_ast_func(lockres, 0); - - mlog_exit_void(); -} - -static void ocfs2_super_bast_func(void *opaque, - int level) -{ - struct ocfs2_lock_res *lockres = opaque; - struct ocfs2_super *osb; - - mlog_entry_void(); - mlog(0, "Superblock BAST fired\n"); - - BUG_ON(!ocfs2_is_super_lock(lockres)); - osb = ocfs2_lock_res_super(lockres); - ocfs2_generic_bast_func(osb, lockres, level); - - mlog_exit_void(); -} - -static void ocfs2_rename_ast_func(void *opaque) -{ - struct ocfs2_lock_res *lockres = opaque; - - mlog_entry_void(); - - mlog(0, "Rename AST fired\n"); - - BUG_ON(!ocfs2_is_rename_lock(lockres)); - - ocfs2_generic_ast_func(lockres, 1); - - mlog_exit_void(); -} - -static void ocfs2_rename_bast_func(void *opaque, - int level) -{ - struct ocfs2_lock_res *lockres = opaque; - struct ocfs2_super *osb; - - mlog_entry_void(); - - mlog(0, "Rename BAST fired\n"); - - BUG_ON(!ocfs2_is_rename_lock(lockres)); - - osb = ocfs2_lock_res_super(lockres); - ocfs2_generic_bast_func(osb, lockres, level); - - mlog_exit_void(); + spin_unlock_irqrestore(&lockres->l_lock, flags); } static inline void ocfs2_recover_from_dlm_error(struct ocfs2_lock_res *lockres, @@ -810,9 +794,10 @@ static int ocfs2_lock_create(struct ocfs2_super *osb, &lockres->l_lksb, dlm_flags, lockres->l_name, - lockres->l_ops->ast, + OCFS2_LOCK_ID_MAX_LEN - 1, + ocfs2_locking_ast, lockres, - lockres->l_ops->bast); + ocfs2_blocking_ast); if (status != DLM_NORMAL) { ocfs2_log_dlm_error("dlmlock", status, lockres); ret = -EINVAL; @@ -930,6 +915,9 @@ static int ocfs2_cluster_lock(struct ocfs2_super *osb, ocfs2_init_mask_waiter(&mw); + if (lockres->l_ops->flags & LOCK_TYPE_USES_LVB) + lkm_flags |= LKM_VALBLK; + again: wait = 0; @@ -997,11 +985,12 @@ again: status = dlmlock(osb->dlm, level, &lockres->l_lksb, - lkm_flags|LKM_CONVERT|LKM_VALBLK, + lkm_flags|LKM_CONVERT, lockres->l_name, - lockres->l_ops->ast, + OCFS2_LOCK_ID_MAX_LEN - 1, + ocfs2_locking_ast, lockres, - lockres->l_ops->bast); + ocfs2_blocking_ast); if (status != DLM_NORMAL) { if ((lkm_flags & LKM_NOQUEUE) && (status == DLM_NOTQUEUED)) @@ -1074,18 +1063,21 @@ static void ocfs2_cluster_unlock(struct ocfs2_super *osb, mlog_exit_void(); } -static int ocfs2_create_new_inode_lock(struct inode *inode, - struct ocfs2_lock_res *lockres) +int ocfs2_create_new_lock(struct ocfs2_super *osb, + struct ocfs2_lock_res *lockres, + int ex, + int local) { - struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); + int level = ex ? LKM_EXMODE : LKM_PRMODE; unsigned long flags; + int lkm_flags = local ? LKM_LOCAL : 0; spin_lock_irqsave(&lockres->l_lock, flags); BUG_ON(lockres->l_flags & OCFS2_LOCK_ATTACHED); lockres_or_flags(lockres, OCFS2_LOCK_LOCAL); spin_unlock_irqrestore(&lockres->l_lock, flags); - return ocfs2_lock_create(osb, lockres, LKM_EXMODE, LKM_LOCAL); + return ocfs2_lock_create(osb, lockres, level, lkm_flags); } /* Grants us an EX lock on the data and metadata resources, skipping @@ -1097,6 +1089,7 @@ static int ocfs2_create_new_inode_lock(struct inode *inode, int ocfs2_create_new_inode_locks(struct inode *inode) { int ret; + struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); BUG_ON(!inode); BUG_ON(!ocfs2_inode_is_new(inode)); @@ -1113,22 +1106,23 @@ int ocfs2_create_new_inode_locks(struct inode *inode) * on a resource which has an invalid one -- we'll set it * valid when we release the EX. */ - ret = ocfs2_create_new_inode_lock(inode, - &OCFS2_I(inode)->ip_rw_lockres); + ret = ocfs2_create_new_lock(osb, &OCFS2_I(inode)->ip_rw_lockres, 1, 1); if (ret) { mlog_errno(ret); goto bail; } - ret = ocfs2_create_new_inode_lock(inode, - &OCFS2_I(inode)->ip_meta_lockres); + /* + * We don't want to use LKM_LOCAL on a meta data lock as they + * don't use a generation in their lock names. + */ + ret = ocfs2_create_new_lock(osb, &OCFS2_I(inode)->ip_meta_lockres, 1, 0); if (ret) { mlog_errno(ret); goto bail; } - ret = ocfs2_create_new_inode_lock(inode, - &OCFS2_I(inode)->ip_data_lockres); + ret = ocfs2_create_new_lock(osb, &OCFS2_I(inode)->ip_data_lockres, 1, 1); if (ret) { mlog_errno(ret); goto bail; @@ -1317,7 +1311,17 @@ static void __ocfs2_stuff_meta_lvb(struct inode *inode) lvb = (struct ocfs2_meta_lvb *) lockres->l_lksb.lvb; - lvb->lvb_version = cpu_to_be32(OCFS2_LVB_VERSION); + /* + * Invalidate the LVB of a deleted inode - this way other + * nodes are forced to go to disk and discover the new inode + * status. + */ + if (oi->ip_flags & OCFS2_INODE_DELETED) { + lvb->lvb_version = 0; + goto out; + } + + lvb->lvb_version = OCFS2_LVB_VERSION; lvb->lvb_isize = cpu_to_be64(i_size_read(inode)); lvb->lvb_iclusters = cpu_to_be32(oi->ip_clusters); lvb->lvb_iuid = cpu_to_be32(inode->i_uid); @@ -1330,7 +1334,10 @@ static void __ocfs2_stuff_meta_lvb(struct inode *inode) cpu_to_be64(ocfs2_pack_timespec(&inode->i_ctime)); lvb->lvb_imtime_packed = cpu_to_be64(ocfs2_pack_timespec(&inode->i_mtime)); + lvb->lvb_iattr = cpu_to_be32(oi->ip_attr); + lvb->lvb_igeneration = cpu_to_be32(inode->i_generation); +out: mlog_meta_lvb(0, lockres); mlog_exit_void(); @@ -1360,6 +1367,9 @@ static void ocfs2_refresh_inode_from_lvb(struct inode *inode) oi->ip_clusters = be32_to_cpu(lvb->lvb_iclusters); i_size_write(inode, be64_to_cpu(lvb->lvb_isize)); + oi->ip_attr = be32_to_cpu(lvb->lvb_iattr); + ocfs2_set_inode_flags(inode); + /* fast-symlinks are a special case */ if (S_ISLNK(inode->i_mode) && !oi->ip_clusters) inode->i_blocks = 0; @@ -1382,11 +1392,13 @@ static void ocfs2_refresh_inode_from_lvb(struct inode *inode) mlog_exit_void(); } -static inline int ocfs2_meta_lvb_is_trustable(struct ocfs2_lock_res *lockres) +static inline int ocfs2_meta_lvb_is_trustable(struct inode *inode, + struct ocfs2_lock_res *lockres) { struct ocfs2_meta_lvb *lvb = (struct ocfs2_meta_lvb *) lockres->l_lksb.lvb; - if (be32_to_cpu(lvb->lvb_version) == OCFS2_LVB_VERSION) + if (lvb->lvb_version == OCFS2_LVB_VERSION + && be32_to_cpu(lvb->lvb_igeneration) == inode->i_generation) return 1; return 0; } @@ -1483,7 +1495,7 @@ static int ocfs2_meta_lock_update(struct inode *inode, * map (directories, bitmap files, etc) */ ocfs2_extent_map_trunc(inode, 0); - if (ocfs2_meta_lvb_is_trustable(lockres)) { + if (ocfs2_meta_lvb_is_trustable(inode, lockres)) { mlog(0, "Trusting LVB on inode %llu\n", (unsigned long long)oi->ip_blkno); ocfs2_refresh_inode_from_lvb(inode); @@ -1624,6 +1636,18 @@ int ocfs2_meta_lock_full(struct inode *inode, wait_event(osb->recovery_event, ocfs2_node_map_is_empty(osb, &osb->recovery_map)); + /* + * We only see this flag if we're being called from + * ocfs2_read_locked_inode(). It means we're locking an inode + * which hasn't been populated yet, so clear the refresh flag + * and let the caller handle it. + */ + if (inode->i_state & I_NEW) { + status = 0; + ocfs2_complete_lock_res_refresh(lockres, 0); + goto bail; + } + /* This is fun. The caller may want a bh back, or it may * not. ocfs2_meta_lock_update definitely wants one in, but * may or may not read one, depending on what's in the @@ -1803,6 +1827,34 @@ void ocfs2_rename_unlock(struct ocfs2_super *osb) ocfs2_cluster_unlock(osb, lockres, LKM_EXMODE); } +int ocfs2_dentry_lock(struct dentry *dentry, int ex) +{ + int ret; + int level = ex ? LKM_EXMODE : LKM_PRMODE; + struct ocfs2_dentry_lock *dl = dentry->d_fsdata; + struct ocfs2_super *osb = OCFS2_SB(dentry->d_sb); + + BUG_ON(!dl); + + if (ocfs2_is_hard_readonly(osb)) + return -EROFS; + + ret = ocfs2_cluster_lock(osb, &dl->dl_lockres, level, 0, 0); + if (ret < 0) + mlog_errno(ret); + + return ret; +} + +void ocfs2_dentry_unlock(struct dentry *dentry, int ex) +{ + int level = ex ? LKM_EXMODE : LKM_PRMODE; + struct ocfs2_dentry_lock *dl = dentry->d_fsdata; + struct ocfs2_super *osb = OCFS2_SB(dentry->d_sb); + + ocfs2_cluster_unlock(osb, &dl->dl_lockres, level); +} + /* Reference counting of the dlm debug structure. We want this because * open references on the debug inodes can live on after a mount, so * we can't rely on the ocfs2_super to always exist. */ @@ -1933,9 +1985,16 @@ static int ocfs2_dlm_seq_show(struct seq_file *m, void *v) if (!lockres) return -EINVAL; - seq_printf(m, "0x%x\t" - "%.*s\t" - "%d\t" + seq_printf(m, "0x%x\t", OCFS2_DLM_DEBUG_STR_VERSION); + + if (lockres->l_type == OCFS2_LOCK_TYPE_DENTRY) + seq_printf(m, "%.*s%08x\t", OCFS2_DENTRY_LOCK_INO_START - 1, + lockres->l_name, + (unsigned int)ocfs2_get_dentry_lock_ino(lockres)); + else + seq_printf(m, "%.*s\t", OCFS2_LOCK_ID_MAX_LEN, lockres->l_name); + + seq_printf(m, "%d\t" "0x%lx\t" "0x%x\t" "0x%x\t" @@ -1943,8 +2002,6 @@ static int ocfs2_dlm_seq_show(struct seq_file *m, void *v) "%u\t" "%d\t" "%d\t", - OCFS2_DLM_DEBUG_STR_VERSION, - OCFS2_LOCK_ID_MAX_LEN, lockres->l_name, lockres->l_level, lockres->l_flags, lockres->l_action, @@ -1995,7 +2052,7 @@ static int ocfs2_dlm_debug_open(struct inode *inode, struct file *file) mlog_errno(ret); goto out; } - osb = (struct ocfs2_super *) inode->u.generic_ip; + osb = inode->i_private; ocfs2_get_dlm_debug(osb->osb_dlm_debug); priv->p_dlm_debug = osb->osb_dlm_debug; INIT_LIST_HEAD(&priv->p_iter_res.l_debug_list); @@ -2071,8 +2128,7 @@ int ocfs2_dlm_init(struct ocfs2_super *osb) } /* launch vote thread */ - osb->vote_task = kthread_run(ocfs2_vote_thread, osb, "ocfs2vote-%d", - osb->osb_id); + osb->vote_task = kthread_run(ocfs2_vote_thread, osb, "ocfs2vote"); if (IS_ERR(osb->vote_task)) { status = PTR_ERR(osb->vote_task); osb->vote_task = NULL; @@ -2135,7 +2191,7 @@ void ocfs2_dlm_shutdown(struct ocfs2_super *osb) mlog_exit_void(); } -static void ocfs2_unlock_ast_func(void *opaque, enum dlm_status status) +static void ocfs2_unlock_ast(void *opaque, enum dlm_status status) { struct ocfs2_lock_res *lockres = opaque; unsigned long flags; @@ -2191,24 +2247,20 @@ complete_unlock: mlog_exit_void(); } -typedef void (ocfs2_pre_drop_cb_t)(struct ocfs2_lock_res *, void *); - -struct drop_lock_cb { - ocfs2_pre_drop_cb_t *drop_func; - void *drop_data; -}; - static int ocfs2_drop_lock(struct ocfs2_super *osb, - struct ocfs2_lock_res *lockres, - struct drop_lock_cb *dcb) + struct ocfs2_lock_res *lockres) { enum dlm_status status; unsigned long flags; + int lkm_flags = 0; /* We didn't get anywhere near actually using this lockres. */ if (!(lockres->l_flags & OCFS2_LOCK_INITIALIZED)) goto out; + if (lockres->l_ops->flags & LOCK_TYPE_USES_LVB) + lkm_flags |= LKM_VALBLK; + spin_lock_irqsave(&lockres->l_lock, flags); mlog_bug_on_msg(!(lockres->l_flags & OCFS2_LOCK_FREEING), @@ -2231,8 +2283,12 @@ static int ocfs2_drop_lock(struct ocfs2_super *osb, spin_lock_irqsave(&lockres->l_lock, flags); } - if (dcb) - dcb->drop_func(lockres, dcb->drop_data); + if (lockres->l_ops->flags & LOCK_TYPE_USES_LVB) { + if (lockres->l_flags & OCFS2_LOCK_ATTACHED && + lockres->l_level == LKM_EXMODE && + !(lockres->l_flags & OCFS2_LOCK_NEEDS_REFRESH)) + lockres->l_ops->set_lvb(lockres); + } if (lockres->l_flags & OCFS2_LOCK_BUSY) mlog(ML_ERROR, "destroying busy lock: \"%s\"\n", @@ -2258,8 +2314,8 @@ static int ocfs2_drop_lock(struct ocfs2_super *osb, mlog(0, "lock %s\n", lockres->l_name); - status = dlmunlock(osb->dlm, &lockres->l_lksb, LKM_VALBLK, - lockres->l_ops->unlock_ast, lockres); + status = dlmunlock(osb->dlm, &lockres->l_lksb, lkm_flags, + ocfs2_unlock_ast, lockres); if (status != DLM_NORMAL) { ocfs2_log_dlm_error("dlmunlock", status, lockres); mlog(ML_ERROR, "lockres flags: %lu\n", lockres->l_flags); @@ -2306,43 +2362,26 @@ void ocfs2_mark_lockres_freeing(struct ocfs2_lock_res *lockres) spin_unlock_irqrestore(&lockres->l_lock, flags); } -static void ocfs2_drop_osb_locks(struct ocfs2_super *osb) +void ocfs2_simple_drop_lockres(struct ocfs2_super *osb, + struct ocfs2_lock_res *lockres) { - int status; - - mlog_entry_void(); - - ocfs2_mark_lockres_freeing(&osb->osb_super_lockres); - - status = ocfs2_drop_lock(osb, &osb->osb_super_lockres, NULL); - if (status < 0) - mlog_errno(status); - - ocfs2_mark_lockres_freeing(&osb->osb_rename_lockres); - - status = ocfs2_drop_lock(osb, &osb->osb_rename_lockres, NULL); - if (status < 0) - mlog_errno(status); + int ret; - mlog_exit(status); + ocfs2_mark_lockres_freeing(lockres); + ret = ocfs2_drop_lock(osb, lockres); + if (ret) + mlog_errno(ret); } -static void ocfs2_meta_pre_drop(struct ocfs2_lock_res *lockres, void *data) +static void ocfs2_drop_osb_locks(struct ocfs2_super *osb) { - struct inode *inode = data; - - /* the metadata lock requires a bit more work as we have an - * LVB to worry about. */ - if (lockres->l_flags & OCFS2_LOCK_ATTACHED && - lockres->l_level == LKM_EXMODE && - !(lockres->l_flags & OCFS2_LOCK_NEEDS_REFRESH)) - __ocfs2_stuff_meta_lvb(inode); + ocfs2_simple_drop_lockres(osb, &osb->osb_super_lockres); + ocfs2_simple_drop_lockres(osb, &osb->osb_rename_lockres); } int ocfs2_drop_inode_locks(struct inode *inode) { int status, err; - struct drop_lock_cb meta_dcb = { ocfs2_meta_pre_drop, inode, }; mlog_entry_void(); @@ -2350,24 +2389,21 @@ int ocfs2_drop_inode_locks(struct inode *inode) * ocfs2_clear_inode has done it for us. */ err = ocfs2_drop_lock(OCFS2_SB(inode->i_sb), - &OCFS2_I(inode)->ip_data_lockres, - NULL); + &OCFS2_I(inode)->ip_data_lockres); if (err < 0) mlog_errno(err); status = err; err = ocfs2_drop_lock(OCFS2_SB(inode->i_sb), - &OCFS2_I(inode)->ip_meta_lockres, - &meta_dcb); + &OCFS2_I(inode)->ip_meta_lockres); if (err < 0) mlog_errno(err); if (err < 0 && !status) status = err; err = ocfs2_drop_lock(OCFS2_SB(inode->i_sb), - &OCFS2_I(inode)->ip_rw_lockres, - NULL); + &OCFS2_I(inode)->ip_rw_lockres); if (err < 0) mlog_errno(err); if (err < 0 && !status) @@ -2416,9 +2452,10 @@ static int ocfs2_downconvert_lock(struct ocfs2_super *osb, &lockres->l_lksb, dlm_flags, lockres->l_name, - lockres->l_ops->ast, + OCFS2_LOCK_ID_MAX_LEN - 1, + ocfs2_locking_ast, lockres, - lockres->l_ops->bast); + ocfs2_blocking_ast); if (status != DLM_NORMAL) { ocfs2_log_dlm_error("dlmlock", status, lockres); ret = -EINVAL; @@ -2477,7 +2514,7 @@ static int ocfs2_cancel_convert(struct ocfs2_super *osb, status = dlmunlock(osb->dlm, &lockres->l_lksb, LKM_CANCEL, - lockres->l_ops->unlock_ast, + ocfs2_unlock_ast, lockres); if (status != DLM_NORMAL) { ocfs2_log_dlm_error("dlmunlock", status, lockres); @@ -2491,115 +2528,15 @@ static int ocfs2_cancel_convert(struct ocfs2_super *osb, return ret; } -static inline int ocfs2_can_downconvert_meta_lock(struct inode *inode, - struct ocfs2_lock_res *lockres, - int new_level) -{ - int ret; - - mlog_entry_void(); - - BUG_ON(new_level != LKM_NLMODE && new_level != LKM_PRMODE); - - if (lockres->l_flags & OCFS2_LOCK_REFRESHING) { - ret = 0; - mlog(0, "lockres %s currently being refreshed -- backing " - "off!\n", lockres->l_name); - } else if (new_level == LKM_PRMODE) - ret = !lockres->l_ex_holders && - ocfs2_inode_fully_checkpointed(inode); - else /* Must be NLMODE we're converting to. */ - ret = !lockres->l_ro_holders && !lockres->l_ex_holders && - ocfs2_inode_fully_checkpointed(inode); - - mlog_exit(ret); - return ret; -} - -static int ocfs2_do_unblock_meta(struct inode *inode, - int *requeue) -{ - int new_level; - int set_lvb = 0; - int ret = 0; - struct ocfs2_lock_res *lockres = &OCFS2_I(inode)->ip_meta_lockres; - unsigned long flags; - - struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); - - mlog_entry_void(); - - spin_lock_irqsave(&lockres->l_lock, flags); - - BUG_ON(!(lockres->l_flags & OCFS2_LOCK_BLOCKED)); - - mlog(0, "l_level=%d, l_blocking=%d\n", lockres->l_level, - lockres->l_blocking); - - BUG_ON(lockres->l_level != LKM_EXMODE && - lockres->l_level != LKM_PRMODE); - - if (lockres->l_flags & OCFS2_LOCK_BUSY) { - *requeue = 1; - ret = ocfs2_prepare_cancel_convert(osb, lockres); - spin_unlock_irqrestore(&lockres->l_lock, flags); - if (ret) { - ret = ocfs2_cancel_convert(osb, lockres); - if (ret < 0) - mlog_errno(ret); - } - goto leave; - } - - new_level = ocfs2_highest_compat_lock_level(lockres->l_blocking); - - mlog(0, "l_level=%d, l_blocking=%d, new_level=%d\n", - lockres->l_level, lockres->l_blocking, new_level); - - if (ocfs2_can_downconvert_meta_lock(inode, lockres, new_level)) { - if (lockres->l_level == LKM_EXMODE) - set_lvb = 1; - - /* If the lock hasn't been refreshed yet (rare), then - * our memory inode values are old and we skip - * stuffing the lvb. There's no need to actually clear - * out the lvb here as it's value is still valid. */ - if (!(lockres->l_flags & OCFS2_LOCK_NEEDS_REFRESH)) { - if (set_lvb) - __ocfs2_stuff_meta_lvb(inode); - } else - mlog(0, "lockres %s: downconverting stale lock!\n", - lockres->l_name); - - mlog(0, "calling ocfs2_downconvert_lock with l_level=%d, " - "l_blocking=%d, new_level=%d\n", - lockres->l_level, lockres->l_blocking, new_level); - - ocfs2_prepare_downconvert(lockres, new_level); - spin_unlock_irqrestore(&lockres->l_lock, flags); - ret = ocfs2_downconvert_lock(osb, lockres, new_level, set_lvb); - goto leave; - } - if (!ocfs2_inode_fully_checkpointed(inode)) - ocfs2_start_checkpoint(osb); - - *requeue = 1; - spin_unlock_irqrestore(&lockres->l_lock, flags); - ret = 0; -leave: - mlog_exit(ret); - return ret; -} - -static int ocfs2_generic_unblock_lock(struct ocfs2_super *osb, - struct ocfs2_lock_res *lockres, - int *requeue, - ocfs2_convert_worker_t *worker) +static int ocfs2_unblock_lock(struct ocfs2_super *osb, + struct ocfs2_lock_res *lockres, + struct ocfs2_unblock_ctl *ctl) { unsigned long flags; int blocking; int new_level; int ret = 0; + int set_lvb = 0; mlog_entry_void(); @@ -2609,7 +2546,7 @@ static int ocfs2_generic_unblock_lock(struct ocfs2_super *osb, recheck: if (lockres->l_flags & OCFS2_LOCK_BUSY) { - *requeue = 1; + ctl->requeue = 1; ret = ocfs2_prepare_cancel_convert(osb, lockres); spin_unlock_irqrestore(&lockres->l_lock, flags); if (ret) { @@ -2623,27 +2560,33 @@ recheck: /* if we're blocking an exclusive and we have *any* holders, * then requeue. */ if ((lockres->l_blocking == LKM_EXMODE) - && (lockres->l_ex_holders || lockres->l_ro_holders)) { - spin_unlock_irqrestore(&lockres->l_lock, flags); - *requeue = 1; - ret = 0; - goto leave; - } + && (lockres->l_ex_holders || lockres->l_ro_holders)) + goto leave_requeue; /* If it's a PR we're blocking, then only * requeue if we've got any EX holders */ if (lockres->l_blocking == LKM_PRMODE && - lockres->l_ex_holders) { - spin_unlock_irqrestore(&lockres->l_lock, flags); - *requeue = 1; - ret = 0; - goto leave; - } + lockres->l_ex_holders) + goto leave_requeue; + + /* + * Can we get a lock in this state if the holder counts are + * zero? The meta data unblock code used to check this. + */ + if ((lockres->l_ops->flags & LOCK_TYPE_REQUIRES_REFRESH) + && (lockres->l_flags & OCFS2_LOCK_REFRESHING)) + goto leave_requeue; + + new_level = ocfs2_highest_compat_lock_level(lockres->l_blocking); + + if (lockres->l_ops->check_downconvert + && !lockres->l_ops->check_downconvert(lockres, new_level)) + goto leave_requeue; /* If we get here, then we know that there are no more * incompatible holders (and anyone asking for an incompatible * lock is blocked). We can now downconvert the lock */ - if (!worker) + if (!lockres->l_ops->downconvert_worker) goto downconvert; /* Some lockres types want to do a bit of work before @@ -2653,7 +2596,10 @@ recheck: blocking = lockres->l_blocking; spin_unlock_irqrestore(&lockres->l_lock, flags); - worker(lockres, blocking); + ctl->unblock_action = lockres->l_ops->downconvert_worker(lockres, blocking); + + if (ctl->unblock_action == UNBLOCK_STOP_POST) + goto leave; spin_lock_irqsave(&lockres->l_lock, flags); if (blocking != lockres->l_blocking) { @@ -2663,25 +2609,43 @@ recheck: } downconvert: - *requeue = 0; - new_level = ocfs2_highest_compat_lock_level(lockres->l_blocking); + ctl->requeue = 0; + + if (lockres->l_ops->flags & LOCK_TYPE_USES_LVB) { + if (lockres->l_level == LKM_EXMODE) + set_lvb = 1; + + /* + * We only set the lvb if the lock has been fully + * refreshed - otherwise we risk setting stale + * data. Otherwise, there's no need to actually clear + * out the lvb here as it's value is still valid. + */ + if (set_lvb && !(lockres->l_flags & OCFS2_LOCK_NEEDS_REFRESH)) + lockres->l_ops->set_lvb(lockres); + } ocfs2_prepare_downconvert(lockres, new_level); spin_unlock_irqrestore(&lockres->l_lock, flags); - ret = ocfs2_downconvert_lock(osb, lockres, new_level, 0); + ret = ocfs2_downconvert_lock(osb, lockres, new_level, set_lvb); leave: mlog_exit(ret); return ret; + +leave_requeue: + spin_unlock_irqrestore(&lockres->l_lock, flags); + ctl->requeue = 1; + + mlog_exit(0); + return 0; } -static void ocfs2_data_convert_worker(struct ocfs2_lock_res *lockres, - int blocking) +static int ocfs2_data_convert_worker(struct ocfs2_lock_res *lockres, + int blocking) { struct inode *inode; struct address_space *mapping; - mlog_entry_void(); - inode = ocfs2_lock_res_inode(lockres); mapping = inode->i_mapping; @@ -2702,116 +2666,159 @@ static void ocfs2_data_convert_worker(struct ocfs2_lock_res *lockres, filemap_fdatawait(mapping); } - mlog_exit_void(); + return UNBLOCK_CONTINUE; } -int ocfs2_unblock_data(struct ocfs2_lock_res *lockres, - int *requeue) +static int ocfs2_check_meta_downconvert(struct ocfs2_lock_res *lockres, + int new_level) { - int status; - struct inode *inode; - struct ocfs2_super *osb; - - mlog_entry_void(); - - inode = ocfs2_lock_res_inode(lockres); - osb = OCFS2_SB(inode->i_sb); - - mlog(0, "unblock inode %llu\n", - (unsigned long long)OCFS2_I(inode)->ip_blkno); + struct inode *inode = ocfs2_lock_res_inode(lockres); + int checkpointed = ocfs2_inode_fully_checkpointed(inode); - status = ocfs2_generic_unblock_lock(osb, - lockres, - requeue, - ocfs2_data_convert_worker); - if (status < 0) - mlog_errno(status); + BUG_ON(new_level != LKM_NLMODE && new_level != LKM_PRMODE); + BUG_ON(lockres->l_level != LKM_EXMODE && !checkpointed); - mlog(0, "inode %llu, requeue = %d\n", - (unsigned long long)OCFS2_I(inode)->ip_blkno, *requeue); + if (checkpointed) + return 1; - mlog_exit(status); - return status; + ocfs2_start_checkpoint(OCFS2_SB(inode->i_sb)); + return 0; } -static int ocfs2_unblock_inode_lock(struct ocfs2_lock_res *lockres, - int *requeue) +static void ocfs2_set_meta_lvb(struct ocfs2_lock_res *lockres) { - int status; - struct inode *inode; - - mlog_entry_void(); - - mlog(0, "Unblock lockres %s\n", lockres->l_name); - - inode = ocfs2_lock_res_inode(lockres); + struct inode *inode = ocfs2_lock_res_inode(lockres); - status = ocfs2_generic_unblock_lock(OCFS2_SB(inode->i_sb), - lockres, - requeue, - NULL); - if (status < 0) - mlog_errno(status); - - mlog_exit(status); - return status; + __ocfs2_stuff_meta_lvb(inode); } - -int ocfs2_unblock_meta(struct ocfs2_lock_res *lockres, - int *requeue) +/* + * Does the final reference drop on our dentry lock. Right now this + * happens in the vote thread, but we could choose to simplify the + * dlmglue API and push these off to the ocfs2_wq in the future. + */ +static void ocfs2_dentry_post_unlock(struct ocfs2_super *osb, + struct ocfs2_lock_res *lockres) { - int status; - struct inode *inode; - - mlog_entry_void(); + struct ocfs2_dentry_lock *dl = ocfs2_lock_res_dl(lockres); + ocfs2_dentry_lock_put(osb, dl); +} - inode = ocfs2_lock_res_inode(lockres); +/* + * d_delete() matching dentries before the lock downconvert. + * + * At this point, any process waiting to destroy the + * dentry_lock due to last ref count is stopped by the + * OCFS2_LOCK_QUEUED flag. + * + * We have two potential problems + * + * 1) If we do the last reference drop on our dentry_lock (via dput) + * we'll wind up in ocfs2_release_dentry_lock(), waiting on + * the downconvert to finish. Instead we take an elevated + * reference and push the drop until after we've completed our + * unblock processing. + * + * 2) There might be another process with a final reference, + * waiting on us to finish processing. If this is the case, we + * detect it and exit out - there's no more dentries anyway. + */ +static int ocfs2_dentry_convert_worker(struct ocfs2_lock_res *lockres, + int blocking) +{ + struct ocfs2_dentry_lock *dl = ocfs2_lock_res_dl(lockres); + struct ocfs2_inode_info *oi = OCFS2_I(dl->dl_inode); + struct dentry *dentry; + unsigned long flags; + int extra_ref = 0; - mlog(0, "unblock inode %llu\n", - (unsigned long long)OCFS2_I(inode)->ip_blkno); + /* + * This node is blocking another node from getting a read + * lock. This happens when we've renamed within a + * directory. We've forced the other nodes to d_delete(), but + * we never actually dropped our lock because it's still + * valid. The downconvert code will retain a PR for this node, + * so there's no further work to do. + */ + if (blocking == LKM_PRMODE) + return UNBLOCK_CONTINUE; - status = ocfs2_do_unblock_meta(inode, requeue); - if (status < 0) - mlog_errno(status); + /* + * Mark this inode as potentially orphaned. The code in + * ocfs2_delete_inode() will figure out whether it actually + * needs to be freed or not. + */ + spin_lock(&oi->ip_lock); + oi->ip_flags |= OCFS2_INODE_MAYBE_ORPHANED; + spin_unlock(&oi->ip_lock); - mlog(0, "inode %llu, requeue = %d\n", - (unsigned long long)OCFS2_I(inode)->ip_blkno, *requeue); + /* + * Yuck. We need to make sure however that the check of + * OCFS2_LOCK_FREEING and the extra reference are atomic with + * respect to a reference decrement or the setting of that + * flag. + */ + spin_lock_irqsave(&lockres->l_lock, flags); + spin_lock(&dentry_attach_lock); + if (!(lockres->l_flags & OCFS2_LOCK_FREEING) + && dl->dl_count) { + dl->dl_count++; + extra_ref = 1; + } + spin_unlock(&dentry_attach_lock); + spin_unlock_irqrestore(&lockres->l_lock, flags); - mlog_exit(status); - return status; -} + mlog(0, "extra_ref = %d\n", extra_ref); -/* Generic unblock function for any lockres whose private data is an - * ocfs2_super pointer. */ -static int ocfs2_unblock_osb_lock(struct ocfs2_lock_res *lockres, - int *requeue) -{ - int status; - struct ocfs2_super *osb; + /* + * We have a process waiting on us in ocfs2_dentry_iput(), + * which means we can't have any more outstanding + * aliases. There's no need to do any more work. + */ + if (!extra_ref) + return UNBLOCK_CONTINUE; + + spin_lock(&dentry_attach_lock); + while (1) { + dentry = ocfs2_find_local_alias(dl->dl_inode, + dl->dl_parent_blkno, 1); + if (!dentry) + break; + spin_unlock(&dentry_attach_lock); - mlog_entry_void(); + mlog(0, "d_delete(%.*s);\n", dentry->d_name.len, + dentry->d_name.name); - mlog(0, "Unblock lockres %s\n", lockres->l_name); + /* + * The following dcache calls may do an + * iput(). Normally we don't want that from the + * downconverting thread, but in this case it's ok + * because the requesting node already has an + * exclusive lock on the inode, so it can't be queued + * for a downconvert. + */ + d_delete(dentry); + dput(dentry); - osb = ocfs2_lock_res_super(lockres); + spin_lock(&dentry_attach_lock); + } + spin_unlock(&dentry_attach_lock); - status = ocfs2_generic_unblock_lock(osb, - lockres, - requeue, - NULL); - if (status < 0) - mlog_errno(status); + /* + * If we are the last holder of this dentry lock, there is no + * reason to downconvert so skip straight to the unlock. + */ + if (dl->dl_count == 1) + return UNBLOCK_STOP_POST; - mlog_exit(status); - return status; + return UNBLOCK_CONTINUE_POST; } void ocfs2_process_blocked_lock(struct ocfs2_super *osb, struct ocfs2_lock_res *lockres) { int status; - int requeue = 0; + struct ocfs2_unblock_ctl ctl = {0, 0,}; unsigned long flags; /* Our reference to the lockres in this function can be @@ -2822,7 +2829,6 @@ void ocfs2_process_blocked_lock(struct ocfs2_super *osb, BUG_ON(!lockres); BUG_ON(!lockres->l_ops); - BUG_ON(!lockres->l_ops->unblock); mlog(0, "lockres %s blocked.\n", lockres->l_name); @@ -2836,21 +2842,25 @@ void ocfs2_process_blocked_lock(struct ocfs2_super *osb, goto unqueue; spin_unlock_irqrestore(&lockres->l_lock, flags); - status = lockres->l_ops->unblock(lockres, &requeue); + status = ocfs2_unblock_lock(osb, lockres, &ctl); if (status < 0) mlog_errno(status); spin_lock_irqsave(&lockres->l_lock, flags); unqueue: - if (lockres->l_flags & OCFS2_LOCK_FREEING || !requeue) { + if (lockres->l_flags & OCFS2_LOCK_FREEING || !ctl.requeue) { lockres_clear_flags(lockres, OCFS2_LOCK_QUEUED); } else ocfs2_schedule_blocked_lock(osb, lockres); mlog(0, "lockres %s, requeue = %s.\n", lockres->l_name, - requeue ? "yes" : "no"); + ctl.requeue ? "yes" : "no"); spin_unlock_irqrestore(&lockres->l_lock, flags); + if (ctl.unblock_action != UNBLOCK_CONTINUE + && lockres->l_ops->post_unlock) + lockres->l_ops->post_unlock(osb, lockres); + mlog_exit_void(); } @@ -2893,15 +2903,17 @@ void ocfs2_dump_meta_lvb_info(u64 level, mlog(level, "LVB information for %s (called from %s:%u):\n", lockres->l_name, function, line); - mlog(level, "version: %u, clusters: %u\n", - be32_to_cpu(lvb->lvb_version), be32_to_cpu(lvb->lvb_iclusters)); + mlog(level, "version: %u, clusters: %u, generation: 0x%x\n", + lvb->lvb_version, be32_to_cpu(lvb->lvb_iclusters), + be32_to_cpu(lvb->lvb_igeneration)); mlog(level, "size: %llu, uid %u, gid %u, mode 0x%x\n", (unsigned long long)be64_to_cpu(lvb->lvb_isize), be32_to_cpu(lvb->lvb_iuid), be32_to_cpu(lvb->lvb_igid), be16_to_cpu(lvb->lvb_imode)); mlog(level, "nlink %u, atime_packed 0x%llx, ctime_packed 0x%llx, " - "mtime_packed 0x%llx\n", be16_to_cpu(lvb->lvb_inlink), + "mtime_packed 0x%llx iattr 0x%x\n", be16_to_cpu(lvb->lvb_inlink), (long long)be64_to_cpu(lvb->lvb_iatime_packed), (long long)be64_to_cpu(lvb->lvb_ictime_packed), - (long long)be64_to_cpu(lvb->lvb_imtime_packed)); + (long long)be64_to_cpu(lvb->lvb_imtime_packed), + be32_to_cpu(lvb->lvb_iattr)); } diff --git a/fs/ocfs2/dlmglue.h b/fs/ocfs2/dlmglue.h index 8f2d1db2d9e..4a276938722 100644 --- a/fs/ocfs2/dlmglue.h +++ b/fs/ocfs2/dlmglue.h @@ -27,10 +27,14 @@ #ifndef DLMGLUE_H #define DLMGLUE_H -#define OCFS2_LVB_VERSION 2 +#include "dcache.h" + +#define OCFS2_LVB_VERSION 4 struct ocfs2_meta_lvb { - __be32 lvb_version; + __u8 lvb_version; + __u8 lvb_reserved0; + __be16 lvb_reserved1; __be32 lvb_iclusters; __be32 lvb_iuid; __be32 lvb_igid; @@ -40,7 +44,9 @@ struct ocfs2_meta_lvb { __be64 lvb_isize; __be16 lvb_imode; __be16 lvb_inlink; - __be32 lvb_reserved[3]; + __be32 lvb_iattr; + __be32 lvb_igeneration; + __be32 lvb_reserved2; }; /* ocfs2_meta_lock_full() and ocfs2_data_lock_full() 'arg_flags' flags */ @@ -56,9 +62,14 @@ void ocfs2_dlm_shutdown(struct ocfs2_super *osb); void ocfs2_lock_res_init_once(struct ocfs2_lock_res *res); void ocfs2_inode_lock_res_init(struct ocfs2_lock_res *res, enum ocfs2_lock_type type, + unsigned int generation, struct inode *inode); +void ocfs2_dentry_lock_res_init(struct ocfs2_dentry_lock *dl, + u64 parent, struct inode *inode); void ocfs2_lock_res_free(struct ocfs2_lock_res *res); int ocfs2_create_new_inode_locks(struct inode *inode); +int ocfs2_create_new_lock(struct ocfs2_super *osb, + struct ocfs2_lock_res *lockres, int ex, int local); int ocfs2_drop_inode_locks(struct inode *inode); int ocfs2_data_lock_full(struct inode *inode, int write, @@ -92,7 +103,12 @@ void ocfs2_super_unlock(struct ocfs2_super *osb, int ex); int ocfs2_rename_lock(struct ocfs2_super *osb); void ocfs2_rename_unlock(struct ocfs2_super *osb); +int ocfs2_dentry_lock(struct dentry *dentry, int ex); +void ocfs2_dentry_unlock(struct dentry *dentry, int ex); + void ocfs2_mark_lockres_freeing(struct ocfs2_lock_res *lockres); +void ocfs2_simple_drop_lockres(struct ocfs2_super *osb, + struct ocfs2_lock_res *lockres); /* for the vote thread */ void ocfs2_process_blocked_lock(struct ocfs2_super *osb, diff --git a/fs/ocfs2/export.c b/fs/ocfs2/export.c index ec55ab3c121..fb91089a60a 100644 --- a/fs/ocfs2/export.c +++ b/fs/ocfs2/export.c @@ -33,6 +33,7 @@ #include "dir.h" #include "dlmglue.h" +#include "dcache.h" #include "export.h" #include "inode.h" @@ -57,7 +58,7 @@ static struct dentry *ocfs2_get_dentry(struct super_block *sb, void *vobjp) return ERR_PTR(-ESTALE); } - inode = ocfs2_iget(OCFS2_SB(sb), handle->ih_blkno); + inode = ocfs2_iget(OCFS2_SB(sb), handle->ih_blkno, 0); if (IS_ERR(inode)) { mlog_errno(PTR_ERR(inode)); @@ -77,6 +78,7 @@ static struct dentry *ocfs2_get_dentry(struct super_block *sb, void *vobjp) mlog_errno(-ENOMEM); return ERR_PTR(-ENOMEM); } + result->d_op = &ocfs2_dentry_ops; mlog_exit_ptr(result); return result; @@ -113,7 +115,7 @@ static struct dentry *ocfs2_get_parent(struct dentry *child) goto bail_unlock; } - inode = ocfs2_iget(OCFS2_SB(dir->i_sb), blkno); + inode = ocfs2_iget(OCFS2_SB(dir->i_sb), blkno, 0); if (IS_ERR(inode)) { mlog(ML_ERROR, "Unable to create inode %llu\n", (unsigned long long)blkno); @@ -127,6 +129,8 @@ static struct dentry *ocfs2_get_parent(struct dentry *child) parent = ERR_PTR(-ENOMEM); } + parent->d_op = &ocfs2_dentry_ops; + bail_unlock: ocfs2_meta_unlock(dir, 0); diff --git a/fs/ocfs2/extent_map.c b/fs/ocfs2/extent_map.c index 1a5c69071df..fcd4475d1f8 100644 --- a/fs/ocfs2/extent_map.c +++ b/fs/ocfs2/extent_map.c @@ -298,7 +298,7 @@ static int ocfs2_extent_map_find_leaf(struct inode *inode, ret = ocfs2_extent_map_insert(inode, rec, le16_to_cpu(el->l_tree_depth)); - if (ret) { + if (ret && (ret != -EEXIST)) { mlog_errno(ret); goto out_free; } @@ -427,6 +427,11 @@ static int ocfs2_extent_map_insert_entry(struct ocfs2_extent_map *em, /* * Simple rule: on any return code other than -EAGAIN, anything left * in the insert_context will be freed. + * + * Simple rule #2: A return code of -EEXIST from this function or + * its calls to ocfs2_extent_map_insert_entry() signifies that another + * thread beat us to the insert. It is not an actual error, but it + * tells the caller we have no more work to do. */ static int ocfs2_extent_map_try_insert(struct inode *inode, struct ocfs2_extent_rec *rec, @@ -448,22 +453,32 @@ static int ocfs2_extent_map_try_insert(struct inode *inode, goto out_unlock; } + /* Since insert_entry failed, the map MUST have old_ent */ old_ent = ocfs2_extent_map_lookup(em, le32_to_cpu(rec->e_cpos), - le32_to_cpu(rec->e_clusters), NULL, - NULL); + le32_to_cpu(rec->e_clusters), + NULL, NULL); BUG_ON(!old_ent); - ret = -EEXIST; - if (old_ent->e_tree_depth < tree_depth) + if (old_ent->e_tree_depth < tree_depth) { + /* Another thread beat us to the lower tree_depth */ + ret = -EEXIST; goto out_unlock; + } if (old_ent->e_tree_depth == tree_depth) { + /* + * Another thread beat us to this tree_depth. + * Let's make sure we agree with that thread (the + * extent_rec should be identical). + */ if (!memcmp(rec, &old_ent->e_rec, sizeof(struct ocfs2_extent_rec))) ret = 0; + else + /* FIXME: Should this be ESRCH/EBADR??? */ + ret = -EEXIST; - /* FIXME: Should this be ESRCH/EBADR??? */ goto out_unlock; } @@ -599,7 +614,7 @@ static int ocfs2_extent_map_insert(struct inode *inode, tree_depth, &ctxt); } while (ret == -EAGAIN); - if (ret < 0) + if ((ret < 0) && (ret != -EEXIST)) mlog_errno(ret); if (ctxt.left_ent) diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c index a9559c87453..2bbfa17090c 100644 --- a/fs/ocfs2/file.c +++ b/fs/ocfs2/file.c @@ -44,6 +44,7 @@ #include "file.h" #include "sysfile.h" #include "inode.h" +#include "ioctl.h" #include "journal.h" #include "mmap.h" #include "suballoc.h" @@ -1227,10 +1228,12 @@ const struct file_operations ocfs2_fops = { .open = ocfs2_file_open, .aio_read = ocfs2_file_aio_read, .aio_write = ocfs2_file_aio_write, + .ioctl = ocfs2_ioctl, }; const struct file_operations ocfs2_dops = { .read = generic_read_dir, .readdir = ocfs2_readdir, .fsync = ocfs2_sync_file, + .ioctl = ocfs2_ioctl, }; diff --git a/fs/ocfs2/inode.c b/fs/ocfs2/inode.c index 327a5b7b86e..16e8e74dc96 100644 --- a/fs/ocfs2/inode.c +++ b/fs/ocfs2/inode.c @@ -54,8 +54,6 @@ #include "buffer_head_io.h" -#define OCFS2_FI_FLAG_NOWAIT 0x1 -#define OCFS2_FI_FLAG_DELETE 0x2 struct ocfs2_find_inode_args { u64 fi_blkno; @@ -71,6 +69,26 @@ static int ocfs2_truncate_for_delete(struct ocfs2_super *osb, struct inode *inode, struct buffer_head *fe_bh); +void ocfs2_set_inode_flags(struct inode *inode) +{ + unsigned int flags = OCFS2_I(inode)->ip_attr; + + inode->i_flags &= ~(S_IMMUTABLE | + S_SYNC | S_APPEND | S_NOATIME | S_DIRSYNC); + + if (flags & OCFS2_IMMUTABLE_FL) + inode->i_flags |= S_IMMUTABLE; + + if (flags & OCFS2_SYNC_FL) + inode->i_flags |= S_SYNC; + if (flags & OCFS2_APPEND_FL) + inode->i_flags |= S_APPEND; + if (flags & OCFS2_NOATIME_FL) + inode->i_flags |= S_NOATIME; + if (flags & OCFS2_DIRSYNC_FL) + inode->i_flags |= S_DIRSYNC; +} + struct inode *ocfs2_ilookup_for_vote(struct ocfs2_super *osb, u64 blkno, int delete_vote) @@ -89,7 +107,7 @@ struct inode *ocfs2_ilookup_for_vote(struct ocfs2_super *osb, return ilookup5(osb->sb, args.fi_ino, ocfs2_find_actor, &args); } -struct inode *ocfs2_iget(struct ocfs2_super *osb, u64 blkno) +struct inode *ocfs2_iget(struct ocfs2_super *osb, u64 blkno, int flags) { struct inode *inode = NULL; struct super_block *sb = osb->sb; @@ -107,7 +125,7 @@ struct inode *ocfs2_iget(struct ocfs2_super *osb, u64 blkno) } args.fi_blkno = blkno; - args.fi_flags = 0; + args.fi_flags = flags; args.fi_ino = ino_from_blkno(sb, blkno); inode = iget5_locked(sb, args.fi_ino, ocfs2_find_actor, @@ -251,7 +269,6 @@ int ocfs2_populate_inode(struct inode *inode, struct ocfs2_dinode *fe, inode->i_mode = le16_to_cpu(fe->i_mode); inode->i_uid = le32_to_cpu(fe->i_uid); inode->i_gid = le32_to_cpu(fe->i_gid); - inode->i_blksize = (u32)osb->s_clustersize; /* Fast symlinks will have i_size but no allocated clusters. */ if (S_ISLNK(inode->i_mode) && !fe->i_clusters) @@ -260,7 +277,6 @@ int ocfs2_populate_inode(struct inode *inode, struct ocfs2_dinode *fe, inode->i_blocks = ocfs2_align_bytes_to_sectors(le64_to_cpu(fe->i_size)); inode->i_mapping->a_ops = &ocfs2_aops; - inode->i_flags |= S_NOATIME; inode->i_atime.tv_sec = le64_to_cpu(fe->i_atime); inode->i_atime.tv_nsec = le32_to_cpu(fe->i_atime_nsec); inode->i_mtime.tv_sec = le64_to_cpu(fe->i_mtime); @@ -276,16 +292,13 @@ int ocfs2_populate_inode(struct inode *inode, struct ocfs2_dinode *fe, OCFS2_I(inode)->ip_clusters = le32_to_cpu(fe->i_clusters); OCFS2_I(inode)->ip_orphaned_slot = OCFS2_INVALID_SLOT; - - if (create_ino) - inode->i_ino = ino_from_blkno(inode->i_sb, - le64_to_cpu(fe->i_blkno)); - - mlog(0, "blkno = %llu, ino = %lu, create_ino = %s\n", - (unsigned long long)fe->i_blkno, inode->i_ino, create_ino ? "true" : "false"); + OCFS2_I(inode)->ip_attr = le32_to_cpu(fe->i_attr); inode->i_nlink = le16_to_cpu(fe->i_links_count); + if (fe->i_flags & cpu_to_le32(OCFS2_SYSTEM_FL)) + OCFS2_I(inode)->ip_flags |= OCFS2_INODE_SYSTEM_FILE; + if (fe->i_flags & cpu_to_le32(OCFS2_LOCAL_ALLOC_FL)) { OCFS2_I(inode)->ip_flags |= OCFS2_INODE_BITMAP; mlog(0, "local alloc inode: i_ino=%lu\n", inode->i_ino); @@ -323,12 +336,31 @@ int ocfs2_populate_inode(struct inode *inode, struct ocfs2_dinode *fe, break; } + if (create_ino) { + inode->i_ino = ino_from_blkno(inode->i_sb, + le64_to_cpu(fe->i_blkno)); + + /* + * If we ever want to create system files from kernel, + * the generation argument to + * ocfs2_inode_lock_res_init() will have to change. + */ + BUG_ON(fe->i_flags & cpu_to_le32(OCFS2_SYSTEM_FL)); + + ocfs2_inode_lock_res_init(&OCFS2_I(inode)->ip_meta_lockres, + OCFS2_LOCK_TYPE_META, 0, inode); + } + ocfs2_inode_lock_res_init(&OCFS2_I(inode)->ip_rw_lockres, - OCFS2_LOCK_TYPE_RW, inode); - ocfs2_inode_lock_res_init(&OCFS2_I(inode)->ip_meta_lockres, - OCFS2_LOCK_TYPE_META, inode); + OCFS2_LOCK_TYPE_RW, inode->i_generation, + inode); + ocfs2_inode_lock_res_init(&OCFS2_I(inode)->ip_data_lockres, - OCFS2_LOCK_TYPE_DATA, inode); + OCFS2_LOCK_TYPE_DATA, inode->i_generation, + inode); + + ocfs2_set_inode_flags(inode); + inode->i_flags |= S_NOATIME; status = 0; bail: @@ -343,15 +375,15 @@ static int ocfs2_read_locked_inode(struct inode *inode, struct ocfs2_super *osb; struct ocfs2_dinode *fe; struct buffer_head *bh = NULL; - int status; - int sysfile = 0; + int status, can_lock; + u32 generation = 0; mlog_entry("(0x%p, 0x%p)\n", inode, args); status = -EINVAL; if (inode == NULL || inode->i_sb == NULL) { mlog(ML_ERROR, "bad inode\n"); - goto bail; + return status; } sb = inode->i_sb; osb = OCFS2_SB(sb); @@ -359,50 +391,110 @@ static int ocfs2_read_locked_inode(struct inode *inode, if (!args) { mlog(ML_ERROR, "bad inode args\n"); make_bad_inode(inode); - goto bail; + return status; + } + + /* + * To improve performance of cold-cache inode stats, we take + * the cluster lock here if possible. + * + * Generally, OCFS2 never trusts the contents of an inode + * unless it's holding a cluster lock, so taking it here isn't + * a correctness issue as much as it is a performance + * improvement. + * + * There are three times when taking the lock is not a good idea: + * + * 1) During startup, before we have initialized the DLM. + * + * 2) If we are reading certain system files which never get + * cluster locks (local alloc, truncate log). + * + * 3) If the process doing the iget() is responsible for + * orphan dir recovery. We're holding the orphan dir lock and + * can get into a deadlock with another process on another + * node in ->delete_inode(). + * + * #1 and #2 can be simply solved by never taking the lock + * here for system files (which are the only type we read + * during mount). It's a heavier approach, but our main + * concern is user-accesible files anyway. + * + * #3 works itself out because we'll eventually take the + * cluster lock before trusting anything anyway. + */ + can_lock = !(args->fi_flags & OCFS2_FI_FLAG_SYSFILE) + && !(args->fi_flags & OCFS2_FI_FLAG_NOLOCK); + + /* + * To maintain backwards compatibility with older versions of + * ocfs2-tools, we still store the generation value for system + * files. The only ones that actually matter to userspace are + * the journals, but it's easier and inexpensive to just flag + * all system files similarly. + */ + if (args->fi_flags & OCFS2_FI_FLAG_SYSFILE) + generation = osb->fs_generation; + + ocfs2_inode_lock_res_init(&OCFS2_I(inode)->ip_meta_lockres, + OCFS2_LOCK_TYPE_META, + generation, inode); + + if (can_lock) { + status = ocfs2_meta_lock(inode, NULL, NULL, 0); + if (status) { + make_bad_inode(inode); + mlog_errno(status); + return status; + } } - /* Read the FE off disk. This is safe because the kernel only - * does one read_inode2 for a new inode, and if it doesn't - * exist yet then nobody can be working on it! */ - status = ocfs2_read_block(osb, args->fi_blkno, &bh, 0, NULL); + status = ocfs2_read_block(osb, args->fi_blkno, &bh, 0, + can_lock ? inode : NULL); if (status < 0) { mlog_errno(status); - make_bad_inode(inode); goto bail; } + status = -EINVAL; fe = (struct ocfs2_dinode *) bh->b_data; if (!OCFS2_IS_VALID_DINODE(fe)) { mlog(ML_ERROR, "Invalid dinode #%llu: signature = %.*s\n", (unsigned long long)fe->i_blkno, 7, fe->i_signature); - make_bad_inode(inode); goto bail; } - if (fe->i_flags & cpu_to_le32(OCFS2_SYSTEM_FL)) - sysfile = 1; + /* + * This is a code bug. Right now the caller needs to + * understand whether it is asking for a system file inode or + * not so the proper lock names can be built. + */ + mlog_bug_on_msg(!!(fe->i_flags & cpu_to_le32(OCFS2_SYSTEM_FL)) != + !!(args->fi_flags & OCFS2_FI_FLAG_SYSFILE), + "Inode %llu: system file state is ambigous\n", + (unsigned long long)args->fi_blkno); if (S_ISCHR(le16_to_cpu(fe->i_mode)) || S_ISBLK(le16_to_cpu(fe->i_mode))) inode->i_rdev = huge_decode_dev(le64_to_cpu(fe->id1.dev1.i_rdev)); - status = -EINVAL; if (ocfs2_populate_inode(inode, fe, 0) < 0) { mlog(ML_ERROR, "populate failed! i_blkno=%llu, i_ino=%lu\n", (unsigned long long)fe->i_blkno, inode->i_ino); - make_bad_inode(inode); goto bail; } BUG_ON(args->fi_blkno != le64_to_cpu(fe->i_blkno)); - if (sysfile) - OCFS2_I(inode)->ip_flags |= OCFS2_INODE_SYSTEM_FILE; - status = 0; bail: + if (can_lock) + ocfs2_meta_unlock(inode, 0); + + if (status < 0) + make_bad_inode(inode); + if (args && bh) brelse(bh); @@ -875,9 +967,15 @@ void ocfs2_delete_inode(struct inode *inode) goto bail_unlock_inode; } - /* Mark the inode as successfully deleted. This is important - * for ocfs2_clear_inode as it will check this flag and skip - * any checkpointing work */ + /* + * Mark the inode as successfully deleted. + * + * This is important for ocfs2_clear_inode() as it will check + * this flag and skip any checkpointing work + * + * ocfs2_stuff_meta_lvb() also uses this flag to invalidate + * the LVB for other nodes. + */ OCFS2_I(inode)->ip_flags |= OCFS2_INODE_DELETED; bail_unlock_inode: @@ -1002,12 +1100,10 @@ void ocfs2_drop_inode(struct inode *inode) /* Testing ip_orphaned_slot here wouldn't work because we may * not have gotten a delete_inode vote from any other nodes * yet. */ - if (oi->ip_flags & OCFS2_INODE_MAYBE_ORPHANED) { - mlog(0, "Inode was orphaned on another node, clearing nlink.\n"); - inode->i_nlink = 0; - } - - generic_drop_inode(inode); + if (oi->ip_flags & OCFS2_INODE_MAYBE_ORPHANED) + generic_delete_inode(inode); + else + generic_drop_inode(inode); mlog_exit_void(); } @@ -1027,12 +1123,8 @@ struct buffer_head *ocfs2_bread(struct inode *inode, u64 p_blkno; int readflags = OCFS2_BH_CACHED; -#if 0 - /* only turn this on if we know we can deal with read_block - * returning nothing */ if (reada) readflags |= OCFS2_BH_READAHEAD; -#endif if (((u64)block << inode->i_sb->s_blocksize_bits) >= i_size_read(inode)) { @@ -1131,6 +1223,7 @@ int ocfs2_mark_inode_dirty(struct ocfs2_journal_handle *handle, spin_lock(&OCFS2_I(inode)->ip_lock); fe->i_clusters = cpu_to_le32(OCFS2_I(inode)->ip_clusters); + fe->i_attr = cpu_to_le32(OCFS2_I(inode)->ip_attr); spin_unlock(&OCFS2_I(inode)->ip_lock); fe->i_size = cpu_to_le64(i_size_read(inode)); @@ -1164,17 +1257,16 @@ leave: void ocfs2_refresh_inode(struct inode *inode, struct ocfs2_dinode *fe) { - struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); - spin_lock(&OCFS2_I(inode)->ip_lock); OCFS2_I(inode)->ip_clusters = le32_to_cpu(fe->i_clusters); + OCFS2_I(inode)->ip_attr = le32_to_cpu(fe->i_attr); + ocfs2_set_inode_flags(inode); i_size_write(inode, le64_to_cpu(fe->i_size)); inode->i_nlink = le16_to_cpu(fe->i_links_count); inode->i_uid = le32_to_cpu(fe->i_uid); inode->i_gid = le32_to_cpu(fe->i_gid); inode->i_mode = le16_to_cpu(fe->i_mode); - inode->i_blksize = (u32) osb->s_clustersize; if (S_ISLNK(inode->i_mode) && le32_to_cpu(fe->i_clusters) == 0) inode->i_blocks = 0; else diff --git a/fs/ocfs2/inode.h b/fs/ocfs2/inode.h index 84c50796128..9957810fdf8 100644 --- a/fs/ocfs2/inode.h +++ b/fs/ocfs2/inode.h @@ -56,6 +56,7 @@ struct ocfs2_inode_info struct ocfs2_journal_handle *ip_handle; u32 ip_flags; /* see below */ + u32 ip_attr; /* inode attributes */ /* protected by recovery_lock. */ struct inode *ip_next_orphan; @@ -114,14 +115,20 @@ static inline struct ocfs2_inode_info *OCFS2_I(struct inode *inode) extern kmem_cache_t *ocfs2_inode_cache; -extern struct address_space_operations ocfs2_aops; +extern const struct address_space_operations ocfs2_aops; struct buffer_head *ocfs2_bread(struct inode *inode, int block, int *err, int reada); void ocfs2_clear_inode(struct inode *inode); void ocfs2_delete_inode(struct inode *inode); void ocfs2_drop_inode(struct inode *inode); -struct inode *ocfs2_iget(struct ocfs2_super *osb, u64 feoff); + +/* Flags for ocfs2_iget() */ +#define OCFS2_FI_FLAG_NOWAIT 0x1 +#define OCFS2_FI_FLAG_DELETE 0x2 +#define OCFS2_FI_FLAG_SYSFILE 0x4 +#define OCFS2_FI_FLAG_NOLOCK 0x8 +struct inode *ocfs2_iget(struct ocfs2_super *osb, u64 feoff, int flags); struct inode *ocfs2_ilookup_for_vote(struct ocfs2_super *osb, u64 blkno, int delete_vote); @@ -142,4 +149,6 @@ int ocfs2_mark_inode_dirty(struct ocfs2_journal_handle *handle, int ocfs2_aio_read(struct file *file, struct kiocb *req, struct iocb *iocb); int ocfs2_aio_write(struct file *file, struct kiocb *req, struct iocb *iocb); +void ocfs2_set_inode_flags(struct inode *inode); + #endif /* OCFS2_INODE_H */ diff --git a/fs/ocfs2/ioctl.c b/fs/ocfs2/ioctl.c new file mode 100644 index 00000000000..3663cef8068 --- /dev/null +++ b/fs/ocfs2/ioctl.c @@ -0,0 +1,136 @@ +/* + * linux/fs/ocfs2/ioctl.c + * + * Copyright (C) 2006 Herbert Poetzl + * adapted from Remy Card's ext2/ioctl.c + */ + +#include <linux/fs.h> +#include <linux/mount.h> + +#define MLOG_MASK_PREFIX ML_INODE +#include <cluster/masklog.h> + +#include "ocfs2.h" +#include "alloc.h" +#include "dlmglue.h" +#include "inode.h" +#include "journal.h" + +#include "ocfs2_fs.h" +#include "ioctl.h" + +#include <linux/ext2_fs.h> + +static int ocfs2_get_inode_attr(struct inode *inode, unsigned *flags) +{ + int status; + + status = ocfs2_meta_lock(inode, NULL, NULL, 0); + if (status < 0) { + mlog_errno(status); + return status; + } + *flags = OCFS2_I(inode)->ip_attr; + ocfs2_meta_unlock(inode, 0); + + mlog_exit(status); + return status; +} + +static int ocfs2_set_inode_attr(struct inode *inode, unsigned flags, + unsigned mask) +{ + struct ocfs2_inode_info *ocfs2_inode = OCFS2_I(inode); + struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); + struct ocfs2_journal_handle *handle = NULL; + struct buffer_head *bh = NULL; + unsigned oldflags; + int status; + + mutex_lock(&inode->i_mutex); + + status = ocfs2_meta_lock(inode, NULL, &bh, 1); + if (status < 0) { + mlog_errno(status); + goto bail; + } + + status = -EROFS; + if (IS_RDONLY(inode)) + goto bail_unlock; + + status = -EACCES; + if ((current->fsuid != inode->i_uid) && !capable(CAP_FOWNER)) + goto bail_unlock; + + if (!S_ISDIR(inode->i_mode)) + flags &= ~OCFS2_DIRSYNC_FL; + + handle = ocfs2_start_trans(osb, NULL, OCFS2_INODE_UPDATE_CREDITS); + if (IS_ERR(handle)) { + status = PTR_ERR(handle); + mlog_errno(status); + goto bail_unlock; + } + + oldflags = ocfs2_inode->ip_attr; + flags = flags & mask; + flags |= oldflags & ~mask; + + /* + * The IMMUTABLE and APPEND_ONLY flags can only be changed by + * the relevant capability. + */ + status = -EPERM; + if ((oldflags & OCFS2_IMMUTABLE_FL) || ((flags ^ oldflags) & + (OCFS2_APPEND_FL | OCFS2_IMMUTABLE_FL))) { + if (!capable(CAP_LINUX_IMMUTABLE)) + goto bail_unlock; + } + + ocfs2_inode->ip_attr = flags; + ocfs2_set_inode_flags(inode); + + status = ocfs2_mark_inode_dirty(handle, inode, bh); + if (status < 0) + mlog_errno(status); + + ocfs2_commit_trans(handle); +bail_unlock: + ocfs2_meta_unlock(inode, 1); +bail: + mutex_unlock(&inode->i_mutex); + + if (bh) + brelse(bh); + + mlog_exit(status); + return status; +} + +int ocfs2_ioctl(struct inode * inode, struct file * filp, + unsigned int cmd, unsigned long arg) +{ + unsigned int flags; + int status; + + switch (cmd) { + case OCFS2_IOC_GETFLAGS: + status = ocfs2_get_inode_attr(inode, &flags); + if (status < 0) + return status; + + flags &= OCFS2_FL_VISIBLE; + return put_user(flags, (int __user *) arg); + case OCFS2_IOC_SETFLAGS: + if (get_user(flags, (int __user *) arg)) + return -EFAULT; + + return ocfs2_set_inode_attr(inode, flags, + OCFS2_FL_MODIFIABLE); + default: + return -ENOTTY; + } +} + diff --git a/fs/ocfs2/ioctl.h b/fs/ocfs2/ioctl.h new file mode 100644 index 00000000000..4a7c82931db --- /dev/null +++ b/fs/ocfs2/ioctl.h @@ -0,0 +1,16 @@ +/* + * ioctl.h + * + * Function prototypes + * + * Copyright (C) 2006 Herbert Poetzl + * + */ + +#ifndef OCFS2_IOCTL_H +#define OCFS2_IOCTL_H + +int ocfs2_ioctl(struct inode * inode, struct file * filp, + unsigned int cmd, unsigned long arg); + +#endif /* OCFS2_IOCTL_H */ diff --git a/fs/ocfs2/journal.c b/fs/ocfs2/journal.c index eebc3cfa6be..fd9734def55 100644 --- a/fs/ocfs2/journal.c +++ b/fs/ocfs2/journal.c @@ -49,7 +49,7 @@ #include "buffer_head_io.h" -spinlock_t trans_inc_lock = SPIN_LOCK_UNLOCKED; +DEFINE_SPINLOCK(trans_inc_lock); static int ocfs2_force_read_journal(struct inode *inode); static int ocfs2_recover_node(struct ocfs2_super *osb, @@ -222,8 +222,7 @@ void ocfs2_handle_add_inode(struct ocfs2_journal_handle *handle, BUG_ON(!list_empty(&OCFS2_I(inode)->ip_handle_list)); OCFS2_I(inode)->ip_handle = handle; - list_del(&(OCFS2_I(inode)->ip_handle_list)); - list_add_tail(&(OCFS2_I(inode)->ip_handle_list), &(handle->inode_list)); + list_move_tail(&(OCFS2_I(inode)->ip_handle_list), &(handle->inode_list)); } static void ocfs2_handle_unlock_inodes(struct ocfs2_journal_handle *handle) @@ -785,8 +784,7 @@ int ocfs2_journal_load(struct ocfs2_journal *journal) } /* Launch the commit thread */ - osb->commit_task = kthread_run(ocfs2_commit_thread, osb, "ocfs2cmt-%d", - osb->osb_id); + osb->commit_task = kthread_run(ocfs2_commit_thread, osb, "ocfs2cmt"); if (IS_ERR(osb->commit_task)) { status = PTR_ERR(osb->commit_task); osb->commit_task = NULL; @@ -1119,7 +1117,7 @@ void ocfs2_recovery_thread(struct ocfs2_super *osb, int node_num) goto out; osb->recovery_thread_task = kthread_run(__ocfs2_recovery_thread, osb, - "ocfs2rec-%d", osb->osb_id); + "ocfs2rec"); if (IS_ERR(osb->recovery_thread_task)) { mlog_errno((int)PTR_ERR(osb->recovery_thread_task)); osb->recovery_thread_task = NULL; @@ -1495,7 +1493,8 @@ static int ocfs2_queue_orphans(struct ocfs2_super *osb, if (de->name_len == 2 && !strncmp("..", de->name, 2)) continue; - iter = ocfs2_iget(osb, le64_to_cpu(de->inode)); + iter = ocfs2_iget(osb, le64_to_cpu(de->inode), + OCFS2_FI_FLAG_NOLOCK); if (IS_ERR(iter)) continue; diff --git a/fs/ocfs2/localalloc.c b/fs/ocfs2/localalloc.c index 0d1973ea32b..1f17a4d0828 100644 --- a/fs/ocfs2/localalloc.c +++ b/fs/ocfs2/localalloc.c @@ -840,6 +840,12 @@ static int ocfs2_local_alloc_new_window(struct ocfs2_super *osb, mlog(0, "Allocating %u clusters for a new window.\n", ocfs2_local_alloc_window_bits(osb)); + + /* Instruct the allocation code to try the most recently used + * cluster group. We'll re-record the group used this pass + * below. */ + ac->ac_last_group = osb->la_last_gd; + /* we used the generic suballoc reserve function, but we set * everything up nicely, so there's no reason why we can't use * the more specific cluster api to claim bits. */ @@ -852,6 +858,8 @@ static int ocfs2_local_alloc_new_window(struct ocfs2_super *osb, goto bail; } + osb->la_last_gd = ac->ac_last_group; + la->la_bm_off = cpu_to_le32(cluster_off); alloc->id1.bitmap1.i_total = cpu_to_le32(cluster_count); /* just in case... In the future when we find space ourselves, diff --git a/fs/ocfs2/mmap.c b/fs/ocfs2/mmap.c index 843cf9ddefe..83934e33e5b 100644 --- a/fs/ocfs2/mmap.c +++ b/fs/ocfs2/mmap.c @@ -46,12 +46,12 @@ static struct page *ocfs2_nopage(struct vm_area_struct * area, unsigned long address, int *type) { - struct inode *inode = area->vm_file->f_dentry->d_inode; struct page *page = NOPAGE_SIGBUS; sigset_t blocked, oldset; int ret; - mlog_entry("(inode %lu, address %lu)\n", inode->i_ino, address); + mlog_entry("(area=%p, address=%lu, type=%p)\n", area, address, + type); /* The best way to deal with signals in this path is * to block them upfront, rather than allowing the diff --git a/fs/ocfs2/namei.c b/fs/ocfs2/namei.c index 0673862c8bd..849c3b4bb94 100644 --- a/fs/ocfs2/namei.c +++ b/fs/ocfs2/namei.c @@ -56,6 +56,7 @@ #include "journal.h" #include "namei.h" #include "suballoc.h" +#include "super.h" #include "symlink.h" #include "sysfile.h" #include "uptodate.h" @@ -178,7 +179,7 @@ static struct dentry *ocfs2_lookup(struct inode *dir, struct dentry *dentry, if (status < 0) goto bail_add; - inode = ocfs2_iget(OCFS2_SB(dir->i_sb), blkno); + inode = ocfs2_iget(OCFS2_SB(dir->i_sb), blkno, 0); if (IS_ERR(inode)) { mlog(ML_ERROR, "Unable to create inode %llu\n", (unsigned long long)blkno); @@ -198,10 +199,32 @@ static struct dentry *ocfs2_lookup(struct inode *dir, struct dentry *dentry, spin_unlock(&oi->ip_lock); bail_add: - dentry->d_op = &ocfs2_dentry_ops; ret = d_splice_alias(inode, dentry); + if (inode) { + /* + * If d_splice_alias() finds a DCACHE_DISCONNECTED + * dentry, it will d_move() it on top of ourse. The + * return value will indicate this however, so in + * those cases, we switch them around for the locking + * code. + * + * NOTE: This dentry already has ->d_op set from + * ocfs2_get_parent() and ocfs2_get_dentry() + */ + if (ret) + dentry = ret; + + status = ocfs2_dentry_attach_lock(dentry, inode, + OCFS2_I(dir)->ip_blkno); + if (status) { + mlog_errno(status); + ret = ERR_PTR(status); + goto bail_unlock; + } + } + bail_unlock: /* Don't drop the cluster lock until *after* the d_add -- * unlink on another node will message us to remove that @@ -310,13 +333,6 @@ static int ocfs2_mknod(struct inode *dir, /* get our super block */ osb = OCFS2_SB(dir->i_sb); - if (S_ISDIR(mode) && (dir->i_nlink >= OCFS2_LINK_MAX)) { - mlog(ML_ERROR, "inode %llu has i_nlink of %u\n", - (unsigned long long)OCFS2_I(dir)->ip_blkno, dir->i_nlink); - status = -EMLINK; - goto leave; - } - handle = ocfs2_alloc_handle(osb); if (handle == NULL) { status = -ENOMEM; @@ -331,6 +347,11 @@ static int ocfs2_mknod(struct inode *dir, goto leave; } + if (S_ISDIR(mode) && (dir->i_nlink >= OCFS2_LINK_MAX)) { + status = -EMLINK; + goto leave; + } + dirfe = (struct ocfs2_dinode *) parent_fe_bh->b_data; if (!dirfe->i_links_count) { /* can't make a file in a deleted directory. */ @@ -419,6 +440,13 @@ static int ocfs2_mknod(struct inode *dir, goto leave; } + status = ocfs2_dentry_attach_lock(dentry, inode, + OCFS2_I(dir)->ip_blkno); + if (status) { + mlog_errno(status); + goto leave; + } + insert_inode_hash(inode); dentry->d_op = &ocfs2_dentry_ops; d_instantiate(dentry, inode); @@ -643,11 +671,6 @@ static int ocfs2_link(struct dentry *old_dentry, goto bail; } - if (inode->i_nlink >= OCFS2_LINK_MAX) { - err = -EMLINK; - goto bail; - } - handle = ocfs2_alloc_handle(osb); if (handle == NULL) { err = -ENOMEM; @@ -661,6 +684,11 @@ static int ocfs2_link(struct dentry *old_dentry, goto bail; } + if (!dir->i_nlink) { + err = -ENOENT; + goto bail; + } + err = ocfs2_check_dir_for_entry(dir, dentry->d_name.name, dentry->d_name.len); if (err) @@ -726,6 +754,12 @@ static int ocfs2_link(struct dentry *old_dentry, goto bail; } + err = ocfs2_dentry_attach_lock(dentry, inode, OCFS2_I(dir)->ip_blkno); + if (err) { + mlog_errno(err); + goto bail; + } + atomic_inc(&inode->i_count); dentry->d_op = &ocfs2_dentry_ops; d_instantiate(dentry, inode); @@ -744,6 +778,23 @@ bail: return err; } +/* + * Takes and drops an exclusive lock on the given dentry. This will + * force other nodes to drop it. + */ +static int ocfs2_remote_dentry_delete(struct dentry *dentry) +{ + int ret; + + ret = ocfs2_dentry_lock(dentry, 1); + if (ret) + mlog_errno(ret); + else + ocfs2_dentry_unlock(dentry, 1); + + return ret; +} + static int ocfs2_unlink(struct inode *dir, struct dentry *dentry) { @@ -833,8 +884,7 @@ static int ocfs2_unlink(struct inode *dir, else inode->i_nlink--; - status = ocfs2_request_unlink_vote(inode, dentry, - (unsigned int) inode->i_nlink); + status = ocfs2_remote_dentry_delete(dentry); if (status < 0) { /* This vote should succeed under all normal * circumstances. */ @@ -1020,7 +1070,6 @@ static int ocfs2_rename(struct inode *old_dir, struct buffer_head *old_inode_de_bh = NULL; // if old_dentry is a dir, // this is the 1st dirent bh nlink_t old_dir_nlink = old_dir->i_nlink, new_dir_nlink = new_dir->i_nlink; - unsigned int links_count; /* At some point it might be nice to break this function up a * bit. */ @@ -1094,23 +1143,26 @@ static int ocfs2_rename(struct inode *old_dir, } } - if (S_ISDIR(old_inode->i_mode)) { - /* Directories actually require metadata updates to - * the directory info so we can't get away with not - * doing node locking on it. */ - status = ocfs2_meta_lock(old_inode, handle, NULL, 1); - if (status < 0) { - if (status != -ENOENT) - mlog_errno(status); - goto bail; - } - - status = ocfs2_request_rename_vote(old_inode, old_dentry); - if (status < 0) { + /* + * Though we don't require an inode meta data update if + * old_inode is not a directory, we lock anyway here to ensure + * the vote thread on other nodes won't have to concurrently + * downconvert the inode and the dentry locks. + */ + status = ocfs2_meta_lock(old_inode, handle, NULL, 1); + if (status < 0) { + if (status != -ENOENT) mlog_errno(status); - goto bail; - } + goto bail; + } + + status = ocfs2_remote_dentry_delete(old_dentry); + if (status < 0) { + mlog_errno(status); + goto bail; + } + if (S_ISDIR(old_inode->i_mode)) { status = -EIO; old_inode_de_bh = ocfs2_bread(old_inode, 0, &status, 0); if (!old_inode_de_bh) @@ -1124,14 +1176,6 @@ static int ocfs2_rename(struct inode *old_dir, if (!new_inode && new_dir!=old_dir && new_dir->i_nlink >= OCFS2_LINK_MAX) goto bail; - } else { - /* Ah, the simple case - we're a file so just send a - * message. */ - status = ocfs2_request_rename_vote(old_inode, old_dentry); - if (status < 0) { - mlog_errno(status); - goto bail; - } } status = -ENOENT; @@ -1203,13 +1247,7 @@ static int ocfs2_rename(struct inode *old_dir, goto bail; } - if (S_ISDIR(new_inode->i_mode)) - links_count = 0; - else - links_count = (unsigned int) (new_inode->i_nlink - 1); - - status = ocfs2_request_unlink_vote(new_inode, new_dentry, - links_count); + status = ocfs2_remote_dentry_delete(new_dentry); if (status < 0) { mlog_errno(status); goto bail; @@ -1388,6 +1426,7 @@ static int ocfs2_rename(struct inode *old_dir, } } + ocfs2_dentry_move(old_dentry, new_dentry, old_dir, new_dir); status = 0; bail: if (rename_lock) @@ -1676,6 +1715,12 @@ static int ocfs2_symlink(struct inode *dir, goto bail; } + status = ocfs2_dentry_attach_lock(dentry, inode, OCFS2_I(dir)->ip_blkno); + if (status) { + mlog_errno(status); + goto bail; + } + insert_inode_hash(inode); dentry->d_op = &ocfs2_dentry_ops; d_instantiate(dentry, inode); @@ -1964,13 +2009,8 @@ restart: } num++; - /* XXX: questionable readahead stuff here */ bh = ocfs2_bread(dir, b++, &err, 1); bh_use[ra_max] = bh; -#if 0 // ??? - if (bh) - ll_rw_block(READ, 1, &bh); -#endif } } if ((bh = bh_use[ra_ptr++]) == NULL) @@ -1978,6 +2018,10 @@ restart: wait_on_buffer(bh); if (!buffer_uptodate(bh)) { /* read error, skip block & hope for the best */ + ocfs2_error(dir->i_sb, "reading directory %llu, " + "offset %lu\n", + (unsigned long long)OCFS2_I(dir)->ip_blkno, + block); brelse(bh); goto next; } diff --git a/fs/ocfs2/ocfs2.h b/fs/ocfs2/ocfs2.h index da1093039c0..0462a7f4e21 100644 --- a/fs/ocfs2/ocfs2.h +++ b/fs/ocfs2/ocfs2.h @@ -184,7 +184,6 @@ struct ocfs2_journal; struct ocfs2_journal_handle; struct ocfs2_super { - u32 osb_id; /* id used by the proc interface */ struct task_struct *commit_task; struct super_block *sb; struct inode *root_inode; @@ -198,7 +197,6 @@ struct ocfs2_super struct ocfs2_node_map recovery_map; struct ocfs2_node_map umount_map; - u32 num_clusters; u64 root_blkno; u64 system_dir_blkno; u64 bitmap_blkno; @@ -222,13 +220,11 @@ struct ocfs2_super unsigned long s_mount_opt; u16 max_slots; - u16 num_nodes; s16 node_num; s16 slot_num; int s_sectsize_bits; int s_clustersize; int s_clustersize_bits; - struct proc_dir_entry *proc_sub_dir; /* points to /proc/fs/ocfs2/<maj_min> */ atomic_t vol_state; struct mutex recovery_lock; @@ -240,6 +236,7 @@ struct ocfs2_super enum ocfs2_local_alloc_state local_alloc_state; struct buffer_head *local_alloc_bh; + u64 la_last_gd; /* Next two fields are for local node slot recovery during * mount. */ @@ -294,7 +291,6 @@ struct ocfs2_super }; #define OCFS2_SB(sb) ((struct ocfs2_super *)(sb)->s_fs_info) -#define OCFS2_MAX_OSB_ID 65536 static inline int ocfs2_should_order_data(struct inode *inode) { diff --git a/fs/ocfs2/ocfs2_fs.h b/fs/ocfs2/ocfs2_fs.h index c5b1ac547c1..3330a5dc6be 100644 --- a/fs/ocfs2/ocfs2_fs.h +++ b/fs/ocfs2/ocfs2_fs.h @@ -114,6 +114,26 @@ #define OCFS2_CHAIN_FL (0x00000400) /* Chain allocator */ #define OCFS2_DEALLOC_FL (0x00000800) /* Truncate log */ +/* Inode attributes, keep in sync with EXT2 */ +#define OCFS2_SECRM_FL (0x00000001) /* Secure deletion */ +#define OCFS2_UNRM_FL (0x00000002) /* Undelete */ +#define OCFS2_COMPR_FL (0x00000004) /* Compress file */ +#define OCFS2_SYNC_FL (0x00000008) /* Synchronous updates */ +#define OCFS2_IMMUTABLE_FL (0x00000010) /* Immutable file */ +#define OCFS2_APPEND_FL (0x00000020) /* writes to file may only append */ +#define OCFS2_NODUMP_FL (0x00000040) /* do not dump file */ +#define OCFS2_NOATIME_FL (0x00000080) /* do not update atime */ +#define OCFS2_DIRSYNC_FL (0x00010000) /* dirsync behaviour (directories only) */ + +#define OCFS2_FL_VISIBLE (0x000100FF) /* User visible flags */ +#define OCFS2_FL_MODIFIABLE (0x000100FF) /* User modifiable flags */ + +/* + * ioctl commands + */ +#define OCFS2_IOC_GETFLAGS _IOR('f', 1, long) +#define OCFS2_IOC_SETFLAGS _IOW('f', 2, long) + /* * Journal Flags (ocfs2_dinode.id1.journal1.i_flags) */ @@ -399,7 +419,9 @@ struct ocfs2_dinode { __le32 i_atime_nsec; __le32 i_ctime_nsec; __le32 i_mtime_nsec; -/*70*/ __le64 i_reserved1[9]; + __le32 i_attr; + __le32 i_reserved1; +/*70*/ __le64 i_reserved2[8]; /*B8*/ union { __le64 i_pad1; /* Generic way to refer to this 64bit union */ diff --git a/fs/ocfs2/ocfs2_lockid.h b/fs/ocfs2/ocfs2_lockid.h index 7dd9e1e705b..4d5d5655c18 100644 --- a/fs/ocfs2/ocfs2_lockid.h +++ b/fs/ocfs2/ocfs2_lockid.h @@ -35,12 +35,15 @@ #define OCFS2_LOCK_ID_MAX_LEN 32 #define OCFS2_LOCK_ID_PAD "000000" +#define OCFS2_DENTRY_LOCK_INO_START 18 + enum ocfs2_lock_type { OCFS2_LOCK_TYPE_META = 0, OCFS2_LOCK_TYPE_DATA, OCFS2_LOCK_TYPE_SUPER, OCFS2_LOCK_TYPE_RENAME, OCFS2_LOCK_TYPE_RW, + OCFS2_LOCK_TYPE_DENTRY, OCFS2_NUM_LOCK_TYPES }; @@ -63,6 +66,9 @@ static inline char ocfs2_lock_type_char(enum ocfs2_lock_type type) case OCFS2_LOCK_TYPE_RW: c = 'W'; break; + case OCFS2_LOCK_TYPE_DENTRY: + c = 'N'; + break; default: c = '\0'; } @@ -70,4 +76,23 @@ static inline char ocfs2_lock_type_char(enum ocfs2_lock_type type) return c; } +static char *ocfs2_lock_type_strings[] = { + [OCFS2_LOCK_TYPE_META] = "Meta", + [OCFS2_LOCK_TYPE_DATA] = "Data", + [OCFS2_LOCK_TYPE_SUPER] = "Super", + [OCFS2_LOCK_TYPE_RENAME] = "Rename", + /* Need to differntiate from [R]ename.. serializing writes is the + * important job it does, anyway. */ + [OCFS2_LOCK_TYPE_RW] = "Write/Read", + [OCFS2_LOCK_TYPE_DENTRY] = "Dentry", +}; + +static inline const char *ocfs2_lock_type_string(enum ocfs2_lock_type type) +{ +#ifdef __KERNEL__ + mlog_bug_on_msg(type >= OCFS2_NUM_LOCK_TYPES, "%d\n", type); +#endif + return ocfs2_lock_type_strings[type]; +} + #endif /* OCFS2_LOCKID_H */ diff --git a/fs/ocfs2/slot_map.c b/fs/ocfs2/slot_map.c index 871627961d6..aa6f5aadedc 100644 --- a/fs/ocfs2/slot_map.c +++ b/fs/ocfs2/slot_map.c @@ -264,7 +264,7 @@ int ocfs2_find_slot(struct ocfs2_super *osb) osb->slot_num = slot; spin_unlock(&si->si_lock); - mlog(ML_NOTICE, "taking node slot %d\n", osb->slot_num); + mlog(0, "taking node slot %d\n", osb->slot_num); status = ocfs2_update_disk_slots(osb, si); if (status < 0) diff --git a/fs/ocfs2/suballoc.c b/fs/ocfs2/suballoc.c index 195523090c8..9d91e66f51a 100644 --- a/fs/ocfs2/suballoc.c +++ b/fs/ocfs2/suballoc.c @@ -70,12 +70,6 @@ static int ocfs2_block_group_search(struct inode *inode, struct buffer_head *group_bh, u32 bits_wanted, u32 min_bits, u16 *bit_off, u16 *bits_found); -static int ocfs2_search_chain(struct ocfs2_alloc_context *ac, - u32 bits_wanted, - u32 min_bits, - u16 *bit_off, - unsigned int *num_bits, - u64 *bg_blkno); static int ocfs2_claim_suballoc_bits(struct ocfs2_super *osb, struct ocfs2_alloc_context *ac, u32 bits_wanted, @@ -85,11 +79,6 @@ static int ocfs2_claim_suballoc_bits(struct ocfs2_super *osb, u64 *bg_blkno); static int ocfs2_test_bg_bit_allocatable(struct buffer_head *bg_bh, int nr); -static int ocfs2_block_group_find_clear_bits(struct ocfs2_super *osb, - struct buffer_head *bg_bh, - unsigned int bits_wanted, - u16 *bit_off, - u16 *bits_found); static inline int ocfs2_block_group_set_bits(struct ocfs2_journal_handle *handle, struct inode *alloc_inode, struct ocfs2_group_desc *bg, @@ -143,6 +132,64 @@ static u32 ocfs2_bits_per_group(struct ocfs2_chain_list *cl) return (u32)le16_to_cpu(cl->cl_cpg) * (u32)le16_to_cpu(cl->cl_bpc); } +/* somewhat more expensive than our other checks, so use sparingly. */ +static int ocfs2_check_group_descriptor(struct super_block *sb, + struct ocfs2_dinode *di, + struct ocfs2_group_desc *gd) +{ + unsigned int max_bits; + + if (!OCFS2_IS_VALID_GROUP_DESC(gd)) { + OCFS2_RO_ON_INVALID_GROUP_DESC(sb, gd); + return -EIO; + } + + if (di->i_blkno != gd->bg_parent_dinode) { + ocfs2_error(sb, "Group descriptor # %llu has bad parent " + "pointer (%llu, expected %llu)", + (unsigned long long)le64_to_cpu(gd->bg_blkno), + (unsigned long long)le64_to_cpu(gd->bg_parent_dinode), + (unsigned long long)le64_to_cpu(di->i_blkno)); + return -EIO; + } + + max_bits = le16_to_cpu(di->id2.i_chain.cl_cpg) * le16_to_cpu(di->id2.i_chain.cl_bpc); + if (le16_to_cpu(gd->bg_bits) > max_bits) { + ocfs2_error(sb, "Group descriptor # %llu has bit count of %u", + (unsigned long long)le64_to_cpu(gd->bg_blkno), + le16_to_cpu(gd->bg_bits)); + return -EIO; + } + + if (le16_to_cpu(gd->bg_chain) >= + le16_to_cpu(di->id2.i_chain.cl_next_free_rec)) { + ocfs2_error(sb, "Group descriptor # %llu has bad chain %u", + (unsigned long long)le64_to_cpu(gd->bg_blkno), + le16_to_cpu(gd->bg_chain)); + return -EIO; + } + + if (le16_to_cpu(gd->bg_free_bits_count) > le16_to_cpu(gd->bg_bits)) { + ocfs2_error(sb, "Group descriptor # %llu has bit count %u but " + "claims that %u are free", + (unsigned long long)le64_to_cpu(gd->bg_blkno), + le16_to_cpu(gd->bg_bits), + le16_to_cpu(gd->bg_free_bits_count)); + return -EIO; + } + + if (le16_to_cpu(gd->bg_bits) > (8 * le16_to_cpu(gd->bg_size))) { + ocfs2_error(sb, "Group descriptor # %llu has bit count %u but " + "max bitmap bits of %u", + (unsigned long long)le64_to_cpu(gd->bg_blkno), + le16_to_cpu(gd->bg_bits), + 8 * le16_to_cpu(gd->bg_size)); + return -EIO; + } + + return 0; +} + static int ocfs2_block_group_fill(struct ocfs2_journal_handle *handle, struct inode *alloc_inode, struct buffer_head *bg_bh, @@ -663,6 +710,7 @@ static int ocfs2_test_bg_bit_allocatable(struct buffer_head *bg_bh, static int ocfs2_block_group_find_clear_bits(struct ocfs2_super *osb, struct buffer_head *bg_bh, unsigned int bits_wanted, + unsigned int total_bits, u16 *bit_off, u16 *bits_found) { @@ -679,10 +727,8 @@ static int ocfs2_block_group_find_clear_bits(struct ocfs2_super *osb, found = start = best_offset = best_size = 0; bitmap = bg->bg_bitmap; - while((offset = ocfs2_find_next_zero_bit(bitmap, - le16_to_cpu(bg->bg_bits), - start)) != -1) { - if (offset == le16_to_cpu(bg->bg_bits)) + while((offset = ocfs2_find_next_zero_bit(bitmap, total_bits, start)) != -1) { + if (offset == total_bits) break; if (!ocfs2_test_bg_bit_allocatable(bg_bh, offset)) { @@ -911,14 +957,35 @@ static int ocfs2_cluster_group_search(struct inode *inode, { int search = -ENOSPC; int ret; - struct ocfs2_group_desc *bg = (struct ocfs2_group_desc *) group_bh->b_data; + struct ocfs2_group_desc *gd = (struct ocfs2_group_desc *) group_bh->b_data; u16 tmp_off, tmp_found; + unsigned int max_bits, gd_cluster_off; BUG_ON(!ocfs2_is_cluster_bitmap(inode)); - if (bg->bg_free_bits_count) { + if (gd->bg_free_bits_count) { + max_bits = le16_to_cpu(gd->bg_bits); + + /* Tail groups in cluster bitmaps which aren't cpg + * aligned are prone to partial extention by a failed + * fs resize. If the file system resize never got to + * update the dinode cluster count, then we don't want + * to trust any clusters past it, regardless of what + * the group descriptor says. */ + gd_cluster_off = ocfs2_blocks_to_clusters(inode->i_sb, + le64_to_cpu(gd->bg_blkno)); + if ((gd_cluster_off + max_bits) > + OCFS2_I(inode)->ip_clusters) { + max_bits = OCFS2_I(inode)->ip_clusters - gd_cluster_off; + mlog(0, "Desc %llu, bg_bits %u, clusters %u, use %u\n", + (unsigned long long)le64_to_cpu(gd->bg_blkno), + le16_to_cpu(gd->bg_bits), + OCFS2_I(inode)->ip_clusters, max_bits); + } + ret = ocfs2_block_group_find_clear_bits(OCFS2_SB(inode->i_sb), group_bh, bits_wanted, + max_bits, &tmp_off, &tmp_found); if (ret) return ret; @@ -951,17 +1018,109 @@ static int ocfs2_block_group_search(struct inode *inode, if (bg->bg_free_bits_count) ret = ocfs2_block_group_find_clear_bits(OCFS2_SB(inode->i_sb), group_bh, bits_wanted, + le16_to_cpu(bg->bg_bits), bit_off, bits_found); return ret; } +static int ocfs2_alloc_dinode_update_counts(struct inode *inode, + struct ocfs2_journal_handle *handle, + struct buffer_head *di_bh, + u32 num_bits, + u16 chain) +{ + int ret; + u32 tmp_used; + struct ocfs2_dinode *di = (struct ocfs2_dinode *) di_bh->b_data; + struct ocfs2_chain_list *cl = (struct ocfs2_chain_list *) &di->id2.i_chain; + + ret = ocfs2_journal_access(handle, inode, di_bh, + OCFS2_JOURNAL_ACCESS_WRITE); + if (ret < 0) { + mlog_errno(ret); + goto out; + } + + tmp_used = le32_to_cpu(di->id1.bitmap1.i_used); + di->id1.bitmap1.i_used = cpu_to_le32(num_bits + tmp_used); + le32_add_cpu(&cl->cl_recs[chain].c_free, -num_bits); + + ret = ocfs2_journal_dirty(handle, di_bh); + if (ret < 0) + mlog_errno(ret); + +out: + return ret; +} + +static int ocfs2_search_one_group(struct ocfs2_alloc_context *ac, + u32 bits_wanted, + u32 min_bits, + u16 *bit_off, + unsigned int *num_bits, + u64 gd_blkno, + u16 *bits_left) +{ + int ret; + u16 found; + struct buffer_head *group_bh = NULL; + struct ocfs2_group_desc *gd; + struct inode *alloc_inode = ac->ac_inode; + struct ocfs2_journal_handle *handle = ac->ac_handle; + + ret = ocfs2_read_block(OCFS2_SB(alloc_inode->i_sb), gd_blkno, + &group_bh, OCFS2_BH_CACHED, alloc_inode); + if (ret < 0) { + mlog_errno(ret); + return ret; + } + + gd = (struct ocfs2_group_desc *) group_bh->b_data; + if (!OCFS2_IS_VALID_GROUP_DESC(gd)) { + OCFS2_RO_ON_INVALID_GROUP_DESC(alloc_inode->i_sb, gd); + ret = -EIO; + goto out; + } + + ret = ac->ac_group_search(alloc_inode, group_bh, bits_wanted, min_bits, + bit_off, &found); + if (ret < 0) { + if (ret != -ENOSPC) + mlog_errno(ret); + goto out; + } + + *num_bits = found; + + ret = ocfs2_alloc_dinode_update_counts(alloc_inode, handle, ac->ac_bh, + *num_bits, + le16_to_cpu(gd->bg_chain)); + if (ret < 0) { + mlog_errno(ret); + goto out; + } + + ret = ocfs2_block_group_set_bits(handle, alloc_inode, gd, group_bh, + *bit_off, *num_bits); + if (ret < 0) + mlog_errno(ret); + + *bits_left = le16_to_cpu(gd->bg_free_bits_count); + +out: + brelse(group_bh); + + return ret; +} + static int ocfs2_search_chain(struct ocfs2_alloc_context *ac, u32 bits_wanted, u32 min_bits, u16 *bit_off, unsigned int *num_bits, - u64 *bg_blkno) + u64 *bg_blkno, + u16 *bits_left) { int status; u16 chain, tmp_bits; @@ -988,9 +1147,9 @@ static int ocfs2_search_chain(struct ocfs2_alloc_context *ac, goto bail; } bg = (struct ocfs2_group_desc *) group_bh->b_data; - if (!OCFS2_IS_VALID_GROUP_DESC(bg)) { - OCFS2_RO_ON_INVALID_GROUP_DESC(alloc_inode->i_sb, bg); - status = -EIO; + status = ocfs2_check_group_descriptor(alloc_inode->i_sb, fe, bg); + if (status) { + mlog_errno(status); goto bail; } @@ -1018,9 +1177,9 @@ static int ocfs2_search_chain(struct ocfs2_alloc_context *ac, goto bail; } bg = (struct ocfs2_group_desc *) group_bh->b_data; - if (!OCFS2_IS_VALID_GROUP_DESC(bg)) { - OCFS2_RO_ON_INVALID_GROUP_DESC(alloc_inode->i_sb, bg); - status = -EIO; + status = ocfs2_check_group_descriptor(alloc_inode->i_sb, fe, bg); + if (status) { + mlog_errno(status); goto bail; } } @@ -1099,6 +1258,7 @@ static int ocfs2_search_chain(struct ocfs2_alloc_context *ac, (unsigned long long)fe->i_blkno); *bg_blkno = le64_to_cpu(bg->bg_blkno); + *bits_left = le16_to_cpu(bg->bg_free_bits_count); bail: if (group_bh) brelse(group_bh); @@ -1120,6 +1280,8 @@ static int ocfs2_claim_suballoc_bits(struct ocfs2_super *osb, { int status; u16 victim, i; + u16 bits_left = 0; + u64 hint_blkno = ac->ac_last_group; struct ocfs2_chain_list *cl; struct ocfs2_dinode *fe; @@ -1146,6 +1308,28 @@ static int ocfs2_claim_suballoc_bits(struct ocfs2_super *osb, goto bail; } + if (hint_blkno) { + /* Attempt to short-circuit the usual search mechanism + * by jumping straight to the most recently used + * allocation group. This helps us mantain some + * contiguousness across allocations. */ + status = ocfs2_search_one_group(ac, bits_wanted, min_bits, + bit_off, num_bits, + hint_blkno, &bits_left); + if (!status) { + /* Be careful to update *bg_blkno here as the + * caller is expecting it to be filled in, and + * ocfs2_search_one_group() won't do that for + * us. */ + *bg_blkno = hint_blkno; + goto set_hint; + } + if (status < 0 && status != -ENOSPC) { + mlog_errno(status); + goto bail; + } + } + cl = (struct ocfs2_chain_list *) &fe->id2.i_chain; victim = ocfs2_find_victim_chain(cl); @@ -1153,9 +1337,9 @@ static int ocfs2_claim_suballoc_bits(struct ocfs2_super *osb, ac->ac_allow_chain_relink = 1; status = ocfs2_search_chain(ac, bits_wanted, min_bits, bit_off, - num_bits, bg_blkno); + num_bits, bg_blkno, &bits_left); if (!status) - goto bail; + goto set_hint; if (status < 0 && status != -ENOSPC) { mlog_errno(status); goto bail; @@ -1177,8 +1361,8 @@ static int ocfs2_claim_suballoc_bits(struct ocfs2_super *osb, ac->ac_chain = i; status = ocfs2_search_chain(ac, bits_wanted, min_bits, - bit_off, num_bits, - bg_blkno); + bit_off, num_bits, bg_blkno, + &bits_left); if (!status) break; if (status < 0 && status != -ENOSPC) { @@ -1186,8 +1370,19 @@ static int ocfs2_claim_suballoc_bits(struct ocfs2_super *osb, goto bail; } } -bail: +set_hint: + if (status != -ENOSPC) { + /* If the next search of this group is not likely to + * yield a suitable extent, then we reset the last + * group hint so as to not waste a disk read */ + if (bits_left < min_bits) + ac->ac_last_group = 0; + else + ac->ac_last_group = *bg_blkno; + } + +bail: mlog_exit(status); return status; } @@ -1341,7 +1536,7 @@ int ocfs2_claim_clusters(struct ocfs2_super *osb, { int status; unsigned int bits_wanted = ac->ac_bits_wanted - ac->ac_bits_given; - u64 bg_blkno; + u64 bg_blkno = 0; u16 bg_bit_off; mlog_entry_void(); @@ -1494,9 +1689,9 @@ static int ocfs2_free_suballoc_bits(struct ocfs2_journal_handle *handle, } group = (struct ocfs2_group_desc *) group_bh->b_data; - if (!OCFS2_IS_VALID_GROUP_DESC(group)) { - OCFS2_RO_ON_INVALID_GROUP_DESC(alloc_inode->i_sb, group); - status = -EIO; + status = ocfs2_check_group_descriptor(alloc_inode->i_sb, fe, group); + if (status) { + mlog_errno(status); goto bail; } BUG_ON((count + start_bit) > le16_to_cpu(group->bg_bits)); diff --git a/fs/ocfs2/suballoc.h b/fs/ocfs2/suballoc.h index a76c82a7cea..c787838d105 100644 --- a/fs/ocfs2/suballoc.h +++ b/fs/ocfs2/suballoc.h @@ -49,6 +49,8 @@ struct ocfs2_alloc_context { u16 ac_chain; int ac_allow_chain_relink; group_search_t *ac_group_search; + + u64 ac_last_group; }; void ocfs2_free_alloc_context(struct ocfs2_alloc_context *ac); diff --git a/fs/ocfs2/super.c b/fs/ocfs2/super.c index cdf73393f09..4c29cd7cc8e 100644 --- a/fs/ocfs2/super.c +++ b/fs/ocfs2/super.c @@ -68,13 +68,6 @@ #include "buffer_head_io.h" -/* - * Globals - */ -static spinlock_t ocfs2_globals_lock = SPIN_LOCK_UNLOCKED; - -static u32 osb_id; /* Keeps track of next available OSB Id */ - static kmem_cache_t *ocfs2_inode_cachep = NULL; kmem_cache_t *ocfs2_lock_cache = NULL; @@ -209,7 +202,7 @@ static int ocfs2_init_global_system_inodes(struct ocfs2_super *osb) mlog_entry_void(); - new = ocfs2_iget(osb, osb->root_blkno); + new = ocfs2_iget(osb, osb->root_blkno, OCFS2_FI_FLAG_SYSFILE); if (IS_ERR(new)) { status = PTR_ERR(new); mlog_errno(status); @@ -217,7 +210,7 @@ static int ocfs2_init_global_system_inodes(struct ocfs2_super *osb) } osb->root_inode = new; - new = ocfs2_iget(osb, osb->system_dir_blkno); + new = ocfs2_iget(osb, osb->system_dir_blkno, OCFS2_FI_FLAG_SYSFILE); if (IS_ERR(new)) { status = PTR_ERR(new); mlog_errno(status); @@ -642,10 +635,9 @@ static int ocfs2_fill_super(struct super_block *sb, void *data, int silent) ocfs2_complete_mount_recovery(osb); - printk("ocfs2: Mounting device (%u,%u) on (node %d, slot %d) with %s " - "data mode.\n", - MAJOR(sb->s_dev), MINOR(sb->s_dev), osb->node_num, - osb->slot_num, + printk(KERN_INFO "ocfs2: Mounting device (%s) on (node %d, slot %d) " + "with %s data mode.\n", + osb->dev_str, osb->node_num, osb->slot_num, osb->s_mount_opt & OCFS2_MOUNT_DATA_WRITEBACK ? "writeback" : "ordered"); @@ -690,7 +682,7 @@ static struct file_system_type ocfs2_fs_type = { .kill_sb = kill_block_super, /* set to the generic one * right now, but do we * need to change that? */ - .fs_flags = FS_REQUIRES_DEV, + .fs_flags = FS_REQUIRES_DEV|FS_RENAME_DOES_D_MOVE, .next = NULL }; @@ -800,10 +792,6 @@ static int __init ocfs2_init(void) goto leave; } - spin_lock(&ocfs2_globals_lock); - osb_id = 0; - spin_unlock(&ocfs2_globals_lock); - ocfs2_debugfs_root = debugfs_create_dir("ocfs2", NULL); if (!ocfs2_debugfs_root) { status = -EFAULT; @@ -1020,7 +1008,7 @@ static int ocfs2_fill_local_node_info(struct ocfs2_super *osb) goto bail; } - mlog(ML_NOTICE, "I am node %d\n", osb->node_num); + mlog(0, "I am node %d\n", osb->node_num); status = 0; bail: @@ -1191,8 +1179,8 @@ static void ocfs2_dismount_volume(struct super_block *sb, int mnt_err) atomic_set(&osb->vol_state, VOLUME_DISMOUNTED); - printk("ocfs2: Unmounting device (%u,%u) on (node %d)\n", - MAJOR(osb->sb->s_dev), MINOR(osb->sb->s_dev), osb->node_num); + printk(KERN_INFO "ocfs2: Unmounting device (%s) on (node %d)\n", + osb->dev_str, osb->node_num); ocfs2_delete_osb(osb); kfree(osb); @@ -1212,8 +1200,6 @@ static int ocfs2_setup_osb_uuid(struct ocfs2_super *osb, const unsigned char *uu if (osb->uuid_str == NULL) return -ENOMEM; - memcpy(osb->uuid, uuid, OCFS2_VOL_UUID_LEN); - for (i = 0, ptr = osb->uuid_str; i < OCFS2_VOL_UUID_LEN; i++) { /* print with null */ ret = snprintf(ptr, 3, "%02X", uuid[i]); @@ -1311,13 +1297,6 @@ static int ocfs2_initialize_super(struct super_block *sb, goto bail; } - osb->uuid = kmalloc(OCFS2_VOL_UUID_LEN, GFP_KERNEL); - if (!osb->uuid) { - mlog(ML_ERROR, "unable to alloc uuid\n"); - status = -ENOMEM; - goto bail; - } - di = (struct ocfs2_dinode *)bh->b_data; osb->max_slots = le16_to_cpu(di->id2.i_super.s_max_slots); @@ -1327,7 +1306,7 @@ static int ocfs2_initialize_super(struct super_block *sb, status = -EINVAL; goto bail; } - mlog(ML_NOTICE, "max_slots for this device: %u\n", osb->max_slots); + mlog(0, "max_slots for this device: %u\n", osb->max_slots); init_waitqueue_head(&osb->osb_wipe_event); osb->osb_orphan_wipes = kcalloc(osb->max_slots, @@ -1418,7 +1397,7 @@ static int ocfs2_initialize_super(struct super_block *sb, goto bail; } - memcpy(&uuid_net_key, &osb->uuid[i], sizeof(osb->net_key)); + memcpy(&uuid_net_key, di->id2.i_super.s_uuid, sizeof(uuid_net_key)); osb->net_key = le32_to_cpu(uuid_net_key); strncpy(osb->vol_label, di->id2.i_super.s_label, 63); @@ -1463,8 +1442,13 @@ static int ocfs2_initialize_super(struct super_block *sb, osb->bitmap_blkno = OCFS2_I(inode)->ip_blkno; + /* We don't have a cluster lock on the bitmap here because + * we're only interested in static information and the extra + * complexity at mount time isn't worht it. Don't pass the + * inode in to the read function though as we don't want it to + * be put in the cache. */ status = ocfs2_read_block(osb, osb->bitmap_blkno, &bitmap_bh, 0, - inode); + NULL); iput(inode); if (status < 0) { mlog_errno(status); @@ -1473,7 +1457,6 @@ static int ocfs2_initialize_super(struct super_block *sb, di = (struct ocfs2_dinode *) bitmap_bh->b_data; osb->bitmap_cpg = le16_to_cpu(di->id2.i_chain.cl_cpg); - osb->num_clusters = le32_to_cpu(di->id1.bitmap1.i_total); brelse(bitmap_bh); mlog(0, "cluster bitmap inode: %llu, clusters per group: %u\n", (unsigned long long)osb->bitmap_blkno, osb->bitmap_cpg); @@ -1484,18 +1467,6 @@ static int ocfs2_initialize_super(struct super_block *sb, goto bail; } - /* Link this osb onto the global linked list of all osb structures. */ - /* The Global Link List is mainted for the whole driver . */ - spin_lock(&ocfs2_globals_lock); - osb->osb_id = osb_id; - if (osb_id < OCFS2_MAX_OSB_ID) - osb_id++; - else { - mlog(ML_ERROR, "Too many volumes mounted\n"); - status = -ENOMEM; - } - spin_unlock(&ocfs2_globals_lock); - bail: mlog_exit(status); return status; diff --git a/fs/ocfs2/symlink.c b/fs/ocfs2/symlink.c index 0c8a1294ec9..c0f68aa6c17 100644 --- a/fs/ocfs2/symlink.c +++ b/fs/ocfs2/symlink.c @@ -154,7 +154,7 @@ static void *ocfs2_follow_link(struct dentry *dentry, } status = vfs_follow_link(nd, link); - if (status) + if (status && status != -ENOENT) mlog_errno(status); bail: if (page) { diff --git a/fs/ocfs2/sysfile.c b/fs/ocfs2/sysfile.c index fc29cb7a437..5df6e35d09b 100644 --- a/fs/ocfs2/sysfile.c +++ b/fs/ocfs2/sysfile.c @@ -28,11 +28,11 @@ #include <linux/slab.h> #include <linux/highmem.h> -#include "ocfs2.h" - #define MLOG_MASK_PREFIX ML_INODE #include <cluster/masklog.h> +#include "ocfs2.h" + #include "alloc.h" #include "dir.h" #include "inode.h" @@ -115,7 +115,7 @@ static struct inode * _ocfs2_get_system_file_inode(struct ocfs2_super *osb, goto bail; } - inode = ocfs2_iget(osb, blkno); + inode = ocfs2_iget(osb, blkno, OCFS2_FI_FLAG_SYSFILE); if (IS_ERR(inode)) { mlog_errno(PTR_ERR(inode)); inode = NULL; diff --git a/fs/ocfs2/uptodate.c b/fs/ocfs2/uptodate.c index b8a00a79332..9707ed7a320 100644 --- a/fs/ocfs2/uptodate.c +++ b/fs/ocfs2/uptodate.c @@ -206,7 +206,10 @@ static int ocfs2_buffer_cached(struct ocfs2_inode_info *oi, } /* Warning: even if it returns true, this does *not* guarantee that - * the block is stored in our inode metadata cache. */ + * the block is stored in our inode metadata cache. + * + * This can be called under lock_buffer() + */ int ocfs2_buffer_uptodate(struct inode *inode, struct buffer_head *bh) { @@ -226,6 +229,16 @@ int ocfs2_buffer_uptodate(struct inode *inode, return ocfs2_buffer_cached(OCFS2_I(inode), bh); } +/* + * Determine whether a buffer is currently out on a read-ahead request. + * ip_io_sem should be held to serialize submitters with the logic here. + */ +int ocfs2_buffer_read_ahead(struct inode *inode, + struct buffer_head *bh) +{ + return buffer_locked(bh) && ocfs2_buffer_cached(OCFS2_I(inode), bh); +} + /* Requires ip_lock */ static void ocfs2_append_cache_array(struct ocfs2_caching_info *ci, sector_t block) @@ -403,7 +416,11 @@ out_free: * * Note that this function may actually fail to insert the block if * memory cannot be allocated. This is not fatal however (but may - * result in a performance penalty) */ + * result in a performance penalty) + * + * Readahead buffers can be passed in here before the I/O request is + * completed. + */ void ocfs2_set_buffer_uptodate(struct inode *inode, struct buffer_head *bh) { diff --git a/fs/ocfs2/uptodate.h b/fs/ocfs2/uptodate.h index 01cd32d26b0..2e73206059a 100644 --- a/fs/ocfs2/uptodate.h +++ b/fs/ocfs2/uptodate.h @@ -40,5 +40,7 @@ void ocfs2_set_new_buffer_uptodate(struct inode *inode, struct buffer_head *bh); void ocfs2_remove_from_cache(struct inode *inode, struct buffer_head *bh); +int ocfs2_buffer_read_ahead(struct inode *inode, + struct buffer_head *bh); #endif /* OCFS2_UPTODATE_H */ diff --git a/fs/ocfs2/vote.c b/fs/ocfs2/vote.c index ee42765a855..5b4dca79990 100644 --- a/fs/ocfs2/vote.c +++ b/fs/ocfs2/vote.c @@ -74,9 +74,6 @@ struct ocfs2_vote_msg __be32 v_orphaned_slot; /* Used during delete votes */ __be32 v_nlink; /* Used during unlink votes */ } md1; /* Message type dependant 1 */ - __be32 v_unlink_namelen; - __be64 v_unlink_parent; - u8 v_unlink_dirent[OCFS2_VOTE_FILENAME_LEN]; }; /* Responses are given these values to maintain backwards @@ -100,8 +97,6 @@ struct ocfs2_vote_work { enum ocfs2_vote_request { OCFS2_VOTE_REQ_INVALID = 0, OCFS2_VOTE_REQ_DELETE, - OCFS2_VOTE_REQ_UNLINK, - OCFS2_VOTE_REQ_RENAME, OCFS2_VOTE_REQ_MOUNT, OCFS2_VOTE_REQ_UMOUNT, OCFS2_VOTE_REQ_LAST @@ -261,103 +256,13 @@ done: return response; } -static int ocfs2_match_dentry(struct dentry *dentry, - u64 parent_blkno, - unsigned int namelen, - const char *name) -{ - struct inode *parent; - - if (!dentry->d_parent) { - mlog(0, "Detached from parent.\n"); - return 0; - } - - parent = dentry->d_parent->d_inode; - /* Negative parent dentry? */ - if (!parent) - return 0; - - /* Name is in a different directory. */ - if (OCFS2_I(parent)->ip_blkno != parent_blkno) - return 0; - - if (dentry->d_name.len != namelen) - return 0; - - /* comparison above guarantees this is safe. */ - if (memcmp(dentry->d_name.name, name, namelen)) - return 0; - - return 1; -} - -static void ocfs2_process_dentry_request(struct inode *inode, - int rename, - unsigned int new_nlink, - u64 parent_blkno, - unsigned int namelen, - const char *name) -{ - struct dentry *dentry = NULL; - struct list_head *p; - struct ocfs2_inode_info *oi = OCFS2_I(inode); - - mlog(0, "parent %llu, namelen = %u, name = %.*s\n", - (unsigned long long)parent_blkno, namelen, namelen, name); - - spin_lock(&dcache_lock); - - /* Another node is removing this name from the system. It is - * up to us to find the corresponding dentry and if it exists, - * unhash it from the dcache. */ - list_for_each(p, &inode->i_dentry) { - dentry = list_entry(p, struct dentry, d_alias); - - if (ocfs2_match_dentry(dentry, parent_blkno, namelen, name)) { - mlog(0, "dentry found: %.*s\n", - dentry->d_name.len, dentry->d_name.name); - - dget_locked(dentry); - break; - } - - dentry = NULL; - } - - spin_unlock(&dcache_lock); - - if (dentry) { - d_delete(dentry); - dput(dentry); - } - - /* rename votes don't send link counts */ - if (!rename) { - mlog(0, "new_nlink = %u\n", new_nlink); - - /* We don't have the proper locks here to directly - * change i_nlink and besides, the vote is sent - * *before* the operation so it may have failed on the - * other node. This passes a hint to ocfs2_drop_inode - * to force ocfs2_delete_inode, who will take the - * proper cluster locks to sort things out. */ - if (new_nlink == 0) { - spin_lock(&oi->ip_lock); - oi->ip_flags |= OCFS2_INODE_MAYBE_ORPHANED; - spin_unlock(&OCFS2_I(inode)->ip_lock); - } - } -} - static void ocfs2_process_vote(struct ocfs2_super *osb, struct ocfs2_vote_msg *msg) { int net_status, vote_response; int orphaned_slot = 0; - int rename = 0; - unsigned int node_num, generation, new_nlink, namelen; - u64 blkno, parent_blkno; + unsigned int node_num, generation; + u64 blkno; enum ocfs2_vote_request request; struct inode *inode = NULL; struct ocfs2_msg_hdr *hdr = &msg->v_hdr; @@ -437,18 +342,6 @@ static void ocfs2_process_vote(struct ocfs2_super *osb, vote_response = ocfs2_process_delete_request(inode, &orphaned_slot); break; - case OCFS2_VOTE_REQ_RENAME: - rename = 1; - /* fall through */ - case OCFS2_VOTE_REQ_UNLINK: - parent_blkno = be64_to_cpu(msg->v_unlink_parent); - namelen = be32_to_cpu(msg->v_unlink_namelen); - /* new_nlink will be ignored in case of a rename vote */ - new_nlink = be32_to_cpu(msg->md1.v_nlink); - ocfs2_process_dentry_request(inode, rename, new_nlink, - parent_blkno, namelen, - msg->v_unlink_dirent); - break; default: mlog(ML_ERROR, "node %u, invalid request: %u\n", node_num, request); @@ -889,75 +782,6 @@ int ocfs2_request_delete_vote(struct inode *inode) return status; } -static void ocfs2_setup_unlink_vote(struct ocfs2_vote_msg *request, - struct dentry *dentry) -{ - struct inode *parent = dentry->d_parent->d_inode; - - /* We need some values which will uniquely identify a dentry - * on the other nodes so that they can find it and run - * d_delete against it. Parent directory block and full name - * should suffice. */ - - mlog(0, "unlink/rename request: parent: %llu name: %.*s\n", - (unsigned long long)OCFS2_I(parent)->ip_blkno, dentry->d_name.len, - dentry->d_name.name); - - request->v_unlink_parent = cpu_to_be64(OCFS2_I(parent)->ip_blkno); - request->v_unlink_namelen = cpu_to_be32(dentry->d_name.len); - memcpy(request->v_unlink_dirent, dentry->d_name.name, - dentry->d_name.len); -} - -int ocfs2_request_unlink_vote(struct inode *inode, - struct dentry *dentry, - unsigned int nlink) -{ - int status; - struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); - struct ocfs2_vote_msg *request; - - if (dentry->d_name.len > OCFS2_VOTE_FILENAME_LEN) - return -ENAMETOOLONG; - - status = -ENOMEM; - request = ocfs2_new_vote_request(osb, OCFS2_I(inode)->ip_blkno, - inode->i_generation, - OCFS2_VOTE_REQ_UNLINK, nlink); - if (request) { - ocfs2_setup_unlink_vote(request, dentry); - - status = ocfs2_request_vote(inode, request, NULL); - - kfree(request); - } - return status; -} - -int ocfs2_request_rename_vote(struct inode *inode, - struct dentry *dentry) -{ - int status; - struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); - struct ocfs2_vote_msg *request; - - if (dentry->d_name.len > OCFS2_VOTE_FILENAME_LEN) - return -ENAMETOOLONG; - - status = -ENOMEM; - request = ocfs2_new_vote_request(osb, OCFS2_I(inode)->ip_blkno, - inode->i_generation, - OCFS2_VOTE_REQ_RENAME, 0); - if (request) { - ocfs2_setup_unlink_vote(request, dentry); - - status = ocfs2_request_vote(inode, request, NULL); - - kfree(request); - } - return status; -} - int ocfs2_request_mount_vote(struct ocfs2_super *osb) { int status; @@ -988,9 +812,7 @@ int ocfs2_request_mount_vote(struct ocfs2_super *osb) } bail: - if (request) - kfree(request); - + kfree(request); return status; } @@ -1021,9 +843,7 @@ int ocfs2_request_umount_vote(struct ocfs2_super *osb) } bail: - if (request) - kfree(request); - + kfree(request); return status; } diff --git a/fs/ocfs2/vote.h b/fs/ocfs2/vote.h index 9cce6070346..53ebc1c69e5 100644 --- a/fs/ocfs2/vote.h +++ b/fs/ocfs2/vote.h @@ -39,11 +39,6 @@ static inline void ocfs2_kick_vote_thread(struct ocfs2_super *osb) } int ocfs2_request_delete_vote(struct inode *inode); -int ocfs2_request_unlink_vote(struct inode *inode, - struct dentry *dentry, - unsigned int nlink); -int ocfs2_request_rename_vote(struct inode *inode, - struct dentry *dentry); int ocfs2_request_mount_vote(struct ocfs2_super *osb); int ocfs2_request_umount_vote(struct ocfs2_super *osb); int ocfs2_register_net_handlers(struct ocfs2_super *osb); diff --git a/fs/open.c b/fs/open.c index 5fb16e5267d..304c1c7814c 100644 --- a/fs/open.c +++ b/fs/open.c @@ -322,7 +322,7 @@ static long do_sys_ftruncate(unsigned int fd, loff_t length, int small) error = locks_verify_truncate(inode, file, length); if (!error) - error = do_truncate(dentry, length, 0, file); + error = do_truncate(dentry, length, ATTR_MTIME|ATTR_CTIME, file); out_putf: fput(file); out: @@ -546,7 +546,8 @@ asmlinkage long sys_chdir(const char __user * filename) struct nameidata nd; int error; - error = __user_walk(filename, LOOKUP_FOLLOW|LOOKUP_DIRECTORY, &nd); + error = __user_walk(filename, + LOOKUP_FOLLOW|LOOKUP_DIRECTORY|LOOKUP_CHDIR, &nd); if (error) goto out; @@ -1172,6 +1173,7 @@ asmlinkage long sys_close(unsigned int fd) struct file * filp; struct files_struct *files = current->files; struct fdtable *fdt; + int retval; spin_lock(&files->file_lock); fdt = files_fdtable(files); @@ -1184,7 +1186,16 @@ asmlinkage long sys_close(unsigned int fd) FD_CLR(fd, fdt->close_on_exec); __put_unused_fd(files, fd); spin_unlock(&files->file_lock); - return filp_close(filp, files); + retval = filp_close(filp, files); + + /* can't restart close syscall because file table entry was cleared */ + if (unlikely(retval == -ERESTARTSYS || + retval == -ERESTARTNOINTR || + retval == -ERESTARTNOHAND || + retval == -ERESTART_RESTARTBLOCK)) + retval = -EINTR; + + return retval; out_unlock: spin_unlock(&files->file_lock); diff --git a/fs/openpromfs/inode.c b/fs/openpromfs/inode.c index 464e2bce020..592a6402e85 100644 --- a/fs/openpromfs/inode.c +++ b/fs/openpromfs/inode.c @@ -1,5 +1,4 @@ -/* $Id: inode.c,v 1.15 2001/11/12 09:43:39 davem Exp $ - * openpromfs.c: /proc/openprom handling routines +/* inode.c: /proc/openprom handling routines * * Copyright (C) 1996-1999 Jakub Jelinek (jakub@redhat.com) * Copyright (C) 1998 Eddie C. Dost (ecd@skynet.be) @@ -9,759 +8,248 @@ #include <linux/types.h> #include <linux/string.h> #include <linux/fs.h> -#include <linux/openprom_fs.h> #include <linux/init.h> #include <linux/slab.h> -#include <linux/smp_lock.h> +#include <linux/seq_file.h> +#include <linux/magic.h> #include <asm/openprom.h> #include <asm/oplib.h> +#include <asm/prom.h> #include <asm/uaccess.h> -#define ALIASES_NNODES 64 - -typedef struct { - u16 parent; - u16 next; - u16 child; - u16 first_prop; - u32 node; -} openpromfs_node; - -typedef struct { -#define OPP_STRING 0x10 -#define OPP_STRINGLIST 0x20 -#define OPP_BINARY 0x40 -#define OPP_HEXSTRING 0x80 -#define OPP_DIRTY 0x01 -#define OPP_QUOTED 0x02 -#define OPP_NOTQUOTED 0x04 -#define OPP_ASCIIZ 0x08 - u32 flag; - u32 alloclen; - u32 len; - char *value; - char name[8]; -} openprom_property; - -static openpromfs_node *nodes; -static int alloced; -static u16 last_node; -static u16 first_prop; -static u16 options = 0xffff; -static u16 aliases = 0xffff; -static int aliases_nodes; -static char *alias_names [ALIASES_NNODES]; - -#define OPENPROM_ROOT_INO 16 -#define OPENPROM_FIRST_INO OPENPROM_ROOT_INO -#define NODE(ino) nodes[ino - OPENPROM_FIRST_INO] -#define NODE2INO(node) (node + OPENPROM_FIRST_INO) -#define NODEP2INO(no) (no + OPENPROM_FIRST_INO + last_node) - -static int openpromfs_create (struct inode *, struct dentry *, int, struct nameidata *); -static int openpromfs_readdir(struct file *, void *, filldir_t); -static struct dentry *openpromfs_lookup(struct inode *, struct dentry *dentry, struct nameidata *nd); -static int openpromfs_unlink (struct inode *, struct dentry *dentry); +static DEFINE_MUTEX(op_mutex); + +#define OPENPROM_ROOT_INO 0 + +enum op_inode_type { + op_inode_node, + op_inode_prop, +}; + +union op_inode_data { + struct device_node *node; + struct property *prop; +}; -static ssize_t nodenum_read(struct file *file, char __user *buf, - size_t count, loff_t *ppos) +struct op_inode_info { + struct inode vfs_inode; + enum op_inode_type type; + union op_inode_data u; +}; + +static inline struct op_inode_info *OP_I(struct inode *inode) { - struct inode *inode = file->f_dentry->d_inode; - char buffer[10]; - - if (count < 0 || !inode->u.generic_ip) - return -EINVAL; - sprintf (buffer, "%8.8x\n", (u32)(long)(inode->u.generic_ip)); - if (file->f_pos >= 9) - return 0; - if (count > 9 - file->f_pos) - count = 9 - file->f_pos; - if (copy_to_user(buf, buffer + file->f_pos, count)) - return -EFAULT; - *ppos += count; - return count; + return container_of(inode, struct op_inode_info, vfs_inode); } -static ssize_t property_read(struct file *filp, char __user *buf, - size_t count, loff_t *ppos) +static int is_string(unsigned char *p, int len) { - struct inode *inode = filp->f_dentry->d_inode; - int i, j, k; - u32 node; - char *p, *s; - u32 *q; - openprom_property *op; - char buffer[64]; - - if (!filp->private_data) { - node = nodes[(u16)((long)inode->u.generic_ip)].node; - i = ((u32)(long)inode->u.generic_ip) >> 16; - if ((u16)((long)inode->u.generic_ip) == aliases) { - if (i >= aliases_nodes) - p = NULL; - else - p = alias_names [i]; - } else - for (p = prom_firstprop (node, buffer); - i && p && *p; - p = prom_nextprop (node, p, buffer), i--) - /* nothing */ ; - if (!p || !*p) - return -EIO; - i = prom_getproplen (node, p); - if (i < 0) { - if ((u16)((long)inode->u.generic_ip) == aliases) - i = 0; - else - return -EIO; - } - k = i; - if (i < 64) i = 64; - filp->private_data = kmalloc (sizeof (openprom_property) - + (j = strlen (p)) + 2 * i, - GFP_KERNEL); - if (!filp->private_data) - return -ENOMEM; - op = (openprom_property *)filp->private_data; - op->flag = 0; - op->alloclen = 2 * i; - strcpy (op->name, p); - op->value = (char *)(((unsigned long)(op->name + j + 4)) & ~3); - op->len = k; - if (k && prom_getproperty (node, p, op->value, i) < 0) - return -EIO; - op->value [k] = 0; - if (k) { - for (s = NULL, p = op->value; p < op->value + k; p++) { - if ((*p >= ' ' && *p <= '~') || *p == '\n') { - op->flag |= OPP_STRING; - s = p; - continue; - } - if (p > op->value && !*p && s == p - 1) { - if (p < op->value + k - 1) - op->flag |= OPP_STRINGLIST; - else - op->flag |= OPP_ASCIIZ; - continue; - } - if (k == 1 && !*p) { - op->flag |= (OPP_STRING|OPP_ASCIIZ); - break; - } - op->flag &= ~(OPP_STRING|OPP_STRINGLIST); - if (k & 3) - op->flag |= OPP_HEXSTRING; - else - op->flag |= OPP_BINARY; - break; - } - if (op->flag & OPP_STRINGLIST) - op->flag &= ~(OPP_STRING); - if (op->flag & OPP_ASCIIZ) - op->len--; - } - } else - op = (openprom_property *)filp->private_data; - if (!count || !(op->len || (op->flag & OPP_ASCIIZ))) - return 0; - if (*ppos >= 0xffffff || count >= 0xffffff) - return -EINVAL; - if (op->flag & OPP_STRINGLIST) { - for (k = 0, p = op->value; p < op->value + op->len; p++) - if (!*p) - k++; - i = op->len + 4 * k + 3; - } else if (op->flag & OPP_STRING) { - i = op->len + 3; - } else if (op->flag & OPP_BINARY) { - i = (op->len * 9) >> 2; - } else { - i = (op->len << 1) + 1; - } - k = *ppos; - if (k >= i) return 0; - if (count > i - k) count = i - k; - if (op->flag & OPP_STRING) { - if (!k) { - if (put_user('\'', buf)) - return -EFAULT; - k++; - count--; - } + int i; - if (k + count >= i - 2) - j = i - 2 - k; - else - j = count; - - if (j >= 0) { - if (copy_to_user(buf + k - *ppos, - op->value + k - 1, j)) - return -EFAULT; - count -= j; - k += j; - } + for (i = 0; i < len; i++) { + unsigned char val = p[i]; - if (count) { - if (put_user('\'', &buf [k++ - *ppos])) - return -EFAULT; - } - if (count > 1) { - if (put_user('\n', &buf [k++ - *ppos])) - return -EFAULT; - } - } else if (op->flag & OPP_STRINGLIST) { - char *tmp; - - tmp = kmalloc (i, GFP_KERNEL); - if (!tmp) - return -ENOMEM; - - s = tmp; - *s++ = '\''; - for (p = op->value; p < op->value + op->len; p++) { - if (!*p) { - strcpy(s, "' + '"); - s += 5; - continue; - } - *s++ = *p; - } - strcpy(s, "'\n"); - - if (copy_to_user(buf, tmp + k, count)) - return -EFAULT; - - kfree(tmp); - k += count; - - } else if (op->flag & OPP_BINARY) { - char buffer[10]; - u32 *first, *last; - int first_off, last_cnt; - - first = ((u32 *)op->value) + k / 9; - first_off = k % 9; - last = ((u32 *)op->value) + (k + count - 1) / 9; - last_cnt = (k + count) % 9; - if (!last_cnt) last_cnt = 9; - - if (first == last) { - sprintf (buffer, "%08x.", *first); - if (copy_to_user(buf, buffer + first_off, - last_cnt - first_off)) - return -EFAULT; - buf += last_cnt - first_off; - } else { - for (q = first; q <= last; q++) { - sprintf (buffer, "%08x.", *q); - if (q == first) { - if (copy_to_user(buf, buffer + first_off, - 9 - first_off)) - return -EFAULT; - buf += 9 - first_off; - } else if (q == last) { - if (copy_to_user(buf, buffer, last_cnt)) - return -EFAULT; - buf += last_cnt; - } else { - if (copy_to_user(buf, buffer, 9)) - return -EFAULT; - buf += 9; - } - } - } + if ((i && !val) || + (val >= ' ' && val <= '~')) + continue; - if (last == (u32 *)(op->value + op->len - 4) && last_cnt == 9) { - if (put_user('\n', (buf - 1))) - return -EFAULT; - } + return 0; + } - k += count; + return 1; +} - } else if (op->flag & OPP_HEXSTRING) { - char buffer[3]; +static int property_show(struct seq_file *f, void *v) +{ + struct property *prop = f->private; + void *pval; + int len; - if ((k < i - 1) && (k & 1)) { - sprintf (buffer, "%02x", - (unsigned char) *(op->value + (k >> 1)) & 0xff); - if (put_user(buffer[1], &buf[k++ - *ppos])) - return -EFAULT; - count--; - } + len = prop->length; + pval = prop->value; - for (; (count > 1) && (k < i - 1); k += 2) { - sprintf (buffer, "%02x", - (unsigned char) *(op->value + (k >> 1)) & 0xff); - if (copy_to_user(buf + k - *ppos, buffer, 2)) - return -EFAULT; - count -= 2; - } + if (is_string(pval, len)) { + while (len > 0) { + int n = strlen(pval); - if (count && (k < i - 1)) { - sprintf (buffer, "%02x", - (unsigned char) *(op->value + (k >> 1)) & 0xff); - if (put_user(buffer[0], &buf[k++ - *ppos])) - return -EFAULT; - count--; - } + seq_printf(f, "%s", (char *) pval); - if (count) { - if (put_user('\n', &buf [k++ - *ppos])) - return -EFAULT; - } - } - count = k - *ppos; - *ppos = k; - return count; -} + /* Skip over the NULL byte too. */ + pval += n + 1; + len -= n + 1; -static ssize_t property_write(struct file *filp, const char __user *buf, - size_t count, loff_t *ppos) -{ - int i, j, k; - char *p; - u32 *q; - void *b; - openprom_property *op; - - if (*ppos >= 0xffffff || count >= 0xffffff) - return -EINVAL; - if (!filp->private_data) { - i = property_read (filp, NULL, 0, NULL); - if (i) - return i; - } - k = *ppos; - op = (openprom_property *)filp->private_data; - if (!(op->flag & OPP_STRING)) { - u32 *first, *last; - int first_off, last_cnt; - u32 mask, mask2; - char tmp [9]; - int forcelen = 0; - - j = k % 9; - for (i = 0; i < count; i++, j++) { - if (j == 9) j = 0; - if (!j) { - char ctmp; - if (get_user(ctmp, &buf[i])) - return -EFAULT; - if (ctmp != '.') { - if (ctmp != '\n') { - if (op->flag & OPP_BINARY) - return -EINVAL; - else - goto write_try_string; - } else { - count = i + 1; - forcelen = 1; - break; - } - } - } else { - char ctmp; - if (get_user(ctmp, &buf[i])) - return -EFAULT; - if (ctmp < '0' || - (ctmp > '9' && ctmp < 'A') || - (ctmp > 'F' && ctmp < 'a') || - ctmp > 'f') { - if (op->flag & OPP_BINARY) - return -EINVAL; - else - goto write_try_string; - } - } - } - op->flag |= OPP_BINARY; - tmp [8] = 0; - i = ((count + k + 8) / 9) << 2; - if (op->alloclen <= i) { - b = kmalloc (sizeof (openprom_property) + 2 * i, - GFP_KERNEL); - if (!b) - return -ENOMEM; - memcpy (b, filp->private_data, - sizeof (openprom_property) - + strlen (op->name) + op->alloclen); - memset (((char *)b) + sizeof (openprom_property) - + strlen (op->name) + op->alloclen, - 0, 2 * i - op->alloclen); - op = (openprom_property *)b; - op->alloclen = 2*i; - b = filp->private_data; - filp->private_data = (void *)op; - kfree (b); + if (len > 0) + seq_printf(f, " + "); } - first = ((u32 *)op->value) + (k / 9); - first_off = k % 9; - last = (u32 *)(op->value + i); - last_cnt = (k + count) % 9; - if (first + 1 == last) { - memset (tmp, '0', 8); - if (copy_from_user(tmp + first_off, buf, - (count + first_off > 8) ? - 8 - first_off : count)) - return -EFAULT; - mask = 0xffffffff; - mask2 = 0xffffffff; - for (j = 0; j < first_off; j++) - mask >>= 1; - for (j = 8 - count - first_off; j > 0; j--) - mask2 <<= 1; - mask &= mask2; - if (mask) { - *first &= ~mask; - *first |= simple_strtoul (tmp, NULL, 16); - op->flag |= OPP_DIRTY; + } else { + if (len & 3) { + while (len) { + len--; + if (len) + seq_printf(f, "%02x.", + *(unsigned char *) pval); + else + seq_printf(f, "%02x", + *(unsigned char *) pval); + pval++; } } else { - op->flag |= OPP_DIRTY; - for (q = first; q < last; q++) { - if (q == first) { - if (first_off < 8) { - memset (tmp, '0', 8); - if (copy_from_user(tmp + first_off, - buf, - 8 - first_off)) - return -EFAULT; - mask = 0xffffffff; - for (j = 0; j < first_off; j++) - mask >>= 1; - *q &= ~mask; - *q |= simple_strtoul (tmp,NULL,16); - } - buf += 9; - } else if ((q == last - 1) && last_cnt - && (last_cnt < 8)) { - memset (tmp, '0', 8); - if (copy_from_user(tmp, buf, last_cnt)) - return -EFAULT; - mask = 0xffffffff; - for (j = 0; j < 8 - last_cnt; j++) - mask <<= 1; - *q &= ~mask; - *q |= simple_strtoul (tmp, NULL, 16); - buf += last_cnt; - } else { - char tchars[17]; /* XXX yuck... */ - - if (copy_from_user(tchars, buf, 16)) - return -EFAULT; - *q = simple_strtoul (tchars, NULL, 16); - buf += 9; - } - } - } - if (!forcelen) { - if (op->len < i) - op->len = i; - } else - op->len = i; - *ppos += count; - } -write_try_string: - if (!(op->flag & OPP_BINARY)) { - if (!(op->flag & (OPP_QUOTED | OPP_NOTQUOTED))) { - char ctmp; - - /* No way, if somebody starts writing from the middle, - * we don't know whether he uses quotes around or not - */ - if (k > 0) - return -EINVAL; - if (get_user(ctmp, buf)) - return -EFAULT; - if (ctmp == '\'') { - op->flag |= OPP_QUOTED; - buf++; - count--; - (*ppos)++; - if (!count) { - op->flag |= OPP_STRING; - return 1; - } - } else - op->flag |= OPP_NOTQUOTED; - } - op->flag |= OPP_STRING; - if (op->alloclen <= count + *ppos) { - b = kmalloc (sizeof (openprom_property) - + 2 * (count + *ppos), GFP_KERNEL); - if (!b) - return -ENOMEM; - memcpy (b, filp->private_data, - sizeof (openprom_property) - + strlen (op->name) + op->alloclen); - memset (((char *)b) + sizeof (openprom_property) - + strlen (op->name) + op->alloclen, - 0, 2*(count - *ppos) - op->alloclen); - op = (openprom_property *)b; - op->alloclen = 2*(count + *ppos); - b = filp->private_data; - filp->private_data = (void *)op; - kfree (b); - } - p = op->value + *ppos - ((op->flag & OPP_QUOTED) ? 1 : 0); - if (copy_from_user(p, buf, count)) - return -EFAULT; - op->flag |= OPP_DIRTY; - for (i = 0; i < count; i++, p++) - if (*p == '\n') { - *p = 0; - break; + while (len >= 4) { + len -= 4; + + if (len) + seq_printf(f, "%08x.", + *(unsigned int *) pval); + else + seq_printf(f, "%08x", + *(unsigned int *) pval); + pval += 4; } - if (i < count) { - op->len = p - op->value; - *ppos += i + 1; - if ((p > op->value) && (op->flag & OPP_QUOTED) - && (*(p - 1) == '\'')) - op->len--; - } else { - if (p - op->value > op->len) - op->len = p - op->value; - *ppos += count; } } - return *ppos - k; + seq_printf(f, "\n"); + + return 0; } -int property_release (struct inode *inode, struct file *filp) +static void *property_start(struct seq_file *f, loff_t *pos) { - openprom_property *op = (openprom_property *)filp->private_data; - int error; - u32 node; - - if (!op) - return 0; - lock_kernel(); - node = nodes[(u16)((long)inode->u.generic_ip)].node; - if ((u16)((long)inode->u.generic_ip) == aliases) { - if ((op->flag & OPP_DIRTY) && (op->flag & OPP_STRING)) { - char *p = op->name; - int i = (op->value - op->name) - strlen (op->name) - 1; - op->value [op->len] = 0; - *(op->value - 1) = ' '; - if (i) { - for (p = op->value - i - 2; p >= op->name; p--) - p[i] = *p; - p = op->name + i; - } - memcpy (p - 8, "nvalias ", 8); - prom_feval (p - 8); - } - } else if (op->flag & OPP_DIRTY) { - if (op->flag & OPP_STRING) { - op->value [op->len] = 0; - error = prom_setprop (node, op->name, - op->value, op->len + 1); - if (error <= 0) - printk (KERN_WARNING "openpromfs: " - "Couldn't write property %s\n", - op->name); - } else if ((op->flag & OPP_BINARY) || !op->len) { - error = prom_setprop (node, op->name, - op->value, op->len); - if (error <= 0) - printk (KERN_WARNING "openpromfs: " - "Couldn't write property %s\n", - op->name); - } else { - printk (KERN_WARNING "openpromfs: " - "Unknown property type of %s\n", - op->name); - } + if (*pos == 0) + return pos; + return NULL; +} + +static void *property_next(struct seq_file *f, void *v, loff_t *pos) +{ + (*pos)++; + return NULL; +} + +static void property_stop(struct seq_file *f, void *v) +{ + /* Nothing to do */ +} + +static struct seq_operations property_op = { + .start = property_start, + .next = property_next, + .stop = property_stop, + .show = property_show +}; + +static int property_open(struct inode *inode, struct file *file) +{ + struct op_inode_info *oi = OP_I(inode); + int ret; + + BUG_ON(oi->type != op_inode_prop); + + ret = seq_open(file, &property_op); + if (!ret) { + struct seq_file *m = file->private_data; + m->private = oi->u.prop; } - unlock_kernel(); - kfree (filp->private_data); - return 0; + return ret; } static const struct file_operations openpromfs_prop_ops = { - .read = property_read, - .write = property_write, - .release = property_release, + .open = property_open, + .read = seq_read, + .llseek = seq_lseek, + .release = seq_release, }; -static const struct file_operations openpromfs_nodenum_ops = { - .read = nodenum_read, -}; +static int openpromfs_readdir(struct file *, void *, filldir_t); static const struct file_operations openprom_operations = { .read = generic_read_dir, .readdir = openpromfs_readdir, }; -static struct inode_operations openprom_alias_inode_operations = { - .create = openpromfs_create, - .lookup = openpromfs_lookup, - .unlink = openpromfs_unlink, -}; +static struct dentry *openpromfs_lookup(struct inode *, struct dentry *, struct nameidata *); static struct inode_operations openprom_inode_operations = { .lookup = openpromfs_lookup, }; -static int lookup_children(u16 n, const char * name, int len) -{ - int ret; - u16 node; - for (; n != 0xffff; n = nodes[n].next) { - node = nodes[n].child; - if (node != 0xffff) { - char buffer[128]; - int i; - char *p; - - while (node != 0xffff) { - if (prom_getname (nodes[node].node, - buffer, 128) >= 0) { - i = strlen (buffer); - if ((len == i) - && !strncmp (buffer, name, len)) - return NODE2INO(node); - p = strchr (buffer, '@'); - if (p && (len == p - buffer) - && !strncmp (buffer, name, len)) - return NODE2INO(node); - } - node = nodes[node].next; - } - } else - continue; - ret = lookup_children (nodes[n].child, name, len); - if (ret) return ret; - } - return 0; -} - -static struct dentry *openpromfs_lookup(struct inode * dir, struct dentry *dentry, struct nameidata *nd) +static struct dentry *openpromfs_lookup(struct inode *dir, struct dentry *dentry, struct nameidata *nd) { - int ino = 0; -#define OPFSL_DIR 0 -#define OPFSL_PROPERTY 1 -#define OPFSL_NODENUM 2 - int type = 0; - char buffer[128]; - char *p; + struct op_inode_info *ent_oi, *oi = OP_I(dir); + struct device_node *dp, *child; + struct property *prop; + enum op_inode_type ent_type; + union op_inode_data ent_data; const char *name; - u32 n; - u16 dirnode; - unsigned int len; - int i; struct inode *inode; - char buffer2[64]; + unsigned int ino; + int len; - inode = NULL; + BUG_ON(oi->type != op_inode_node); + + dp = oi->u.node; + name = dentry->d_name.name; len = dentry->d_name.len; - lock_kernel(); - if (name [0] == '.' && len == 5 && !strncmp (name + 1, "node", 4)) { - ino = NODEP2INO(NODE(dir->i_ino).first_prop); - type = OPFSL_NODENUM; - } - if (!ino) { - u16 node = NODE(dir->i_ino).child; - while (node != 0xffff) { - if (prom_getname (nodes[node].node, buffer, 128) >= 0) { - i = strlen (buffer); - if (len == i && !strncmp (buffer, name, len)) { - ino = NODE2INO(node); - type = OPFSL_DIR; - break; - } - p = strchr (buffer, '@'); - if (p && (len == p - buffer) - && !strncmp (buffer, name, len)) { - ino = NODE2INO(node); - type = OPFSL_DIR; - break; - } - } - node = nodes[node].next; - } - } - n = NODE(dir->i_ino).node; - dirnode = dir->i_ino - OPENPROM_FIRST_INO; - if (!ino) { - int j = NODEP2INO(NODE(dir->i_ino).first_prop); - if (dirnode != aliases) { - for (p = prom_firstprop (n, buffer2); - p && *p; - p = prom_nextprop (n, p, buffer2)) { - j++; - if ((len == strlen (p)) - && !strncmp (p, name, len)) { - ino = j; - type = OPFSL_PROPERTY; - break; - } - } - } else { - int k; - for (k = 0; k < aliases_nodes; k++) { - j++; - if (alias_names [k] - && (len == strlen (alias_names [k])) - && !strncmp (alias_names [k], name, len)) { - ino = j; - type = OPFSL_PROPERTY; - break; - } - } + + mutex_lock(&op_mutex); + + child = dp->child; + while (child) { + int n = strlen(child->path_component_name); + + if (len == n && + !strncmp(child->path_component_name, name, len)) { + ent_type = op_inode_node; + ent_data.node = child; + ino = child->unique_id; + goto found; } + child = child->sibling; } - if (!ino) { - ino = lookup_children (NODE(dir->i_ino).child, name, len); - if (ino) - type = OPFSL_DIR; - else { - unlock_kernel(); - return ERR_PTR(-ENOENT); + + prop = dp->properties; + while (prop) { + int n = strlen(prop->name); + + if (len == n && !strncmp(prop->name, name, len)) { + ent_type = op_inode_prop; + ent_data.prop = prop; + ino = prop->unique_id; + goto found; } + + prop = prop->next; } - inode = iget (dir->i_sb, ino); - unlock_kernel(); + + mutex_unlock(&op_mutex); + return ERR_PTR(-ENOENT); + +found: + inode = iget(dir->i_sb, ino); + mutex_unlock(&op_mutex); if (!inode) return ERR_PTR(-EINVAL); - switch (type) { - case OPFSL_DIR: + ent_oi = OP_I(inode); + ent_oi->type = ent_type; + ent_oi->u = ent_data; + + switch (ent_type) { + case op_inode_node: inode->i_mode = S_IFDIR | S_IRUGO | S_IXUGO; - if (ino == OPENPROM_FIRST_INO + aliases) { - inode->i_mode |= S_IWUSR; - inode->i_op = &openprom_alias_inode_operations; - } else - inode->i_op = &openprom_inode_operations; + inode->i_op = &openprom_inode_operations; inode->i_fop = &openprom_operations; inode->i_nlink = 2; break; - case OPFSL_NODENUM: - inode->i_mode = S_IFREG | S_IRUGO; - inode->i_fop = &openpromfs_nodenum_ops; - inode->i_nlink = 1; - inode->u.generic_ip = (void *)(long)(n); - break; - case OPFSL_PROPERTY: - if ((dirnode == options) && (len == 17) - && !strncmp (name, "security-password", 17)) + case op_inode_prop: + if (!strcmp(dp->name, "options") && (len == 17) && + !strncmp (name, "security-password", 17)) inode->i_mode = S_IFREG | S_IRUSR | S_IWUSR; - else { + else inode->i_mode = S_IFREG | S_IRUGO; - if (dirnode == options || dirnode == aliases) { - if (len != 4 || strncmp (name, "name", 4)) - inode->i_mode |= S_IWUSR; - } - } inode->i_fop = &openpromfs_prop_ops; inode->i_nlink = 1; - if (inode->i_size < 0) - inode->i_size = 0; - inode->u.generic_ip = (void *)(long)(((u16)dirnode) | - (((u16)(ino - NODEP2INO(NODE(dir->i_ino).first_prop) - 1)) << 16)); + inode->i_size = ent_oi->u.prop->length; break; } @@ -775,237 +263,89 @@ static struct dentry *openpromfs_lookup(struct inode * dir, struct dentry *dentr static int openpromfs_readdir(struct file * filp, void * dirent, filldir_t filldir) { struct inode *inode = filp->f_dentry->d_inode; + struct op_inode_info *oi = OP_I(inode); + struct device_node *dp = oi->u.node; + struct device_node *child; + struct property *prop; unsigned int ino; - u32 n; - int i, j; - char buffer[128]; - u16 node; - char *p; - char buffer2[64]; - - lock_kernel(); + int i; + + mutex_lock(&op_mutex); ino = inode->i_ino; i = filp->f_pos; switch (i) { case 0: - if (filldir(dirent, ".", 1, i, ino, DT_DIR) < 0) goto out; + if (filldir(dirent, ".", 1, i, ino, DT_DIR) < 0) + goto out; i++; filp->f_pos++; /* fall thru */ case 1: - if (filldir(dirent, "..", 2, i, - (NODE(ino).parent == 0xffff) ? - OPENPROM_ROOT_INO : NODE2INO(NODE(ino).parent), DT_DIR) < 0) + if (filldir(dirent, "..", 2, i, + (dp->parent == NULL ? + OPENPROM_ROOT_INO : + dp->parent->unique_id), DT_DIR) < 0) goto out; i++; filp->f_pos++; /* fall thru */ default: i -= 2; - node = NODE(ino).child; - while (i && node != 0xffff) { - node = nodes[node].next; + + /* First, the children nodes as directories. */ + child = dp->child; + while (i && child) { + child = child->sibling; i--; } - while (node != 0xffff) { - if (prom_getname (nodes[node].node, buffer, 128) < 0) - goto out; - if (filldir(dirent, buffer, strlen(buffer), - filp->f_pos, NODE2INO(node), DT_DIR) < 0) + while (child) { + if (filldir(dirent, + child->path_component_name, + strlen(child->path_component_name), + filp->f_pos, child->unique_id, DT_DIR) < 0) goto out; + filp->f_pos++; - node = nodes[node].next; + child = child->sibling; } - j = NODEP2INO(NODE(ino).first_prop); - if (!i) { - if (filldir(dirent, ".node", 5, filp->f_pos, j, DT_REG) < 0) + + /* Next, the properties as files. */ + prop = dp->properties; + while (i && prop) { + prop = prop->next; + i--; + } + while (prop) { + if (filldir(dirent, prop->name, strlen(prop->name), + filp->f_pos, prop->unique_id, DT_REG) < 0) goto out; + filp->f_pos++; - } else - i--; - n = NODE(ino).node; - if (ino == OPENPROM_FIRST_INO + aliases) { - for (j++; i < aliases_nodes; i++, j++) { - if (alias_names [i]) { - if (filldir (dirent, alias_names [i], - strlen (alias_names [i]), - filp->f_pos, j, DT_REG) < 0) goto out; - filp->f_pos++; - } - } - } else { - for (p = prom_firstprop (n, buffer2); - p && *p; - p = prom_nextprop (n, p, buffer2)) { - j++; - if (i) i--; - else { - if (filldir(dirent, p, strlen(p), - filp->f_pos, j, DT_REG) < 0) - goto out; - filp->f_pos++; - } - } + prop = prop->next; } } out: - unlock_kernel(); - return 0; -} - -static int openpromfs_create (struct inode *dir, struct dentry *dentry, int mode, - struct nameidata *nd) -{ - char *p; - struct inode *inode; - - if (!dir) - return -ENOENT; - if (dentry->d_name.len > 256) - return -EINVAL; - p = kmalloc (dentry->d_name.len + 1, GFP_KERNEL); - if (!p) - return -ENOMEM; - strncpy (p, dentry->d_name.name, dentry->d_name.len); - p [dentry->d_name.len] = 0; - lock_kernel(); - if (aliases_nodes == ALIASES_NNODES) { - kfree(p); - unlock_kernel(); - return -EIO; - } - alias_names [aliases_nodes++] = p; - inode = iget (dir->i_sb, - NODEP2INO(NODE(dir->i_ino).first_prop) + aliases_nodes); - if (!inode) { - unlock_kernel(); - return -EINVAL; - } - inode->i_mode = S_IFREG | S_IRUGO | S_IWUSR; - inode->i_fop = &openpromfs_prop_ops; - inode->i_nlink = 1; - if (inode->i_size < 0) inode->i_size = 0; - inode->u.generic_ip = (void *)(long)(((u16)aliases) | - (((u16)(aliases_nodes - 1)) << 16)); - unlock_kernel(); - d_instantiate(dentry, inode); + mutex_unlock(&op_mutex); return 0; } -static int openpromfs_unlink (struct inode *dir, struct dentry *dentry) -{ - unsigned int len; - char *p; - const char *name; - int i; - - name = dentry->d_name.name; - len = dentry->d_name.len; - lock_kernel(); - for (i = 0; i < aliases_nodes; i++) - if ((strlen (alias_names [i]) == len) - && !strncmp (name, alias_names[i], len)) { - char buffer[512]; - - p = alias_names [i]; - alias_names [i] = NULL; - kfree (p); - strcpy (buffer, "nvunalias "); - memcpy (buffer + 10, name, len); - buffer [10 + len] = 0; - prom_feval (buffer); - } - unlock_kernel(); - return 0; -} +static kmem_cache_t *op_inode_cachep; -/* {{{ init section */ -static int __init check_space (u16 n) +static struct inode *openprom_alloc_inode(struct super_block *sb) { - unsigned long pages; + struct op_inode_info *oi; - if ((1 << alloced) * PAGE_SIZE < (n + 2) * sizeof(openpromfs_node)) { - pages = __get_free_pages (GFP_KERNEL, alloced + 1); - if (!pages) - return -1; + oi = kmem_cache_alloc(op_inode_cachep, SLAB_KERNEL); + if (!oi) + return NULL; - if (nodes) { - memcpy ((char *)pages, (char *)nodes, - (1 << alloced) * PAGE_SIZE); - free_pages ((unsigned long)nodes, alloced); - } - alloced++; - nodes = (openpromfs_node *)pages; - } - return 0; + return &oi->vfs_inode; } -static u16 __init get_nodes (u16 parent, u32 node) +static void openprom_destroy_inode(struct inode *inode) { - char *p; - u16 n = last_node++, i; - char buffer[64]; - - if (check_space (n) < 0) - return 0xffff; - nodes[n].parent = parent; - nodes[n].node = node; - nodes[n].next = 0xffff; - nodes[n].child = 0xffff; - nodes[n].first_prop = first_prop++; - if (!parent) { - char buffer[8]; - int j; - - if ((j = prom_getproperty (node, "name", buffer, 8)) >= 0) { - buffer[j] = 0; - if (!strcmp (buffer, "options")) - options = n; - else if (!strcmp (buffer, "aliases")) - aliases = n; - } - } - if (n != aliases) - for (p = prom_firstprop (node, buffer); - p && p != (char *)-1 && *p; - p = prom_nextprop (node, p, buffer)) - first_prop++; - else { - char *q; - for (p = prom_firstprop (node, buffer); - p && p != (char *)-1 && *p; - p = prom_nextprop (node, p, buffer)) { - if (aliases_nodes == ALIASES_NNODES) - break; - for (i = 0; i < aliases_nodes; i++) - if (!strcmp (p, alias_names [i])) - break; - if (i < aliases_nodes) - continue; - q = kmalloc (strlen (p) + 1, GFP_KERNEL); - if (!q) - return 0xffff; - strcpy (q, p); - alias_names [aliases_nodes++] = q; - } - first_prop += ALIASES_NNODES; - } - node = prom_getchild (node); - if (node) { - parent = get_nodes (n, node); - if (parent == 0xffff) - return 0xffff; - nodes[n].child = parent; - while ((node = prom_getsibling (node)) != 0) { - i = get_nodes (n, node); - if (i == 0xffff) - return 0xffff; - nodes[parent].next = i; - parent = i; - } - } - return n; + kmem_cache_free(op_inode_cachep, OP_I(inode)); } static void openprom_read_inode(struct inode * inode) @@ -1025,6 +365,8 @@ static int openprom_remount(struct super_block *sb, int *flags, char *data) } static struct super_operations openprom_sops = { + .alloc_inode = openprom_alloc_inode, + .destroy_inode = openprom_destroy_inode, .read_inode = openprom_read_inode, .statfs = simple_statfs, .remount_fs = openprom_remount, @@ -1032,7 +374,8 @@ static struct super_operations openprom_sops = { static int openprom_fill_super(struct super_block *s, void *data, int silent) { - struct inode * root_inode; + struct inode *root_inode; + struct op_inode_info *oi; s->s_flags |= MS_NOATIME; s->s_blocksize = 1024; @@ -1043,6 +386,11 @@ static int openprom_fill_super(struct super_block *s, void *data, int silent) root_inode = iget(s, OPENPROM_ROOT_INO); if (!root_inode) goto out_no_root; + + oi = OP_I(root_inode); + oi->type = op_inode_node; + oi->u.node = of_find_node_by_path("/"); + s->s_root = d_alloc_root(root_inode); if (!s->s_root) goto out_no_root; @@ -1067,29 +415,39 @@ static struct file_system_type openprom_fs_type = { .kill_sb = kill_anon_super, }; +static void op_inode_init_once(void *data, kmem_cache_t * cachep, unsigned long flags) +{ + struct op_inode_info *oi = (struct op_inode_info *) data; + + if ((flags & (SLAB_CTOR_VERIFY|SLAB_CTOR_CONSTRUCTOR)) == + SLAB_CTOR_CONSTRUCTOR) + inode_init_once(&oi->vfs_inode); +} + static int __init init_openprom_fs(void) { - nodes = (openpromfs_node *)__get_free_pages(GFP_KERNEL, 0); - if (!nodes) { - printk (KERN_WARNING "openpromfs: can't get free page\n"); - return -EIO; - } - if (get_nodes (0xffff, prom_root_node) == 0xffff) { - printk (KERN_WARNING "openpromfs: couldn't setup tree\n"); - return -EIO; - } - nodes[last_node].first_prop = first_prop; - return register_filesystem(&openprom_fs_type); + int err; + + op_inode_cachep = kmem_cache_create("op_inode_cache", + sizeof(struct op_inode_info), + 0, + (SLAB_RECLAIM_ACCOUNT | + SLAB_MEM_SPREAD), + op_inode_init_once, NULL); + if (!op_inode_cachep) + return -ENOMEM; + + err = register_filesystem(&openprom_fs_type); + if (err) + kmem_cache_destroy(op_inode_cachep); + + return err; } static void __exit exit_openprom_fs(void) { - int i; unregister_filesystem(&openprom_fs_type); - free_pages ((unsigned long)nodes, alloced); - for (i = 0; i < aliases_nodes; i++) - kfree (alias_names [i]); - nodes = NULL; + kmem_cache_destroy(op_inode_cachep); } module_init(init_openprom_fs) diff --git a/fs/partitions/Kconfig b/fs/partitions/Kconfig index c9a47809928..e478f194183 100644 --- a/fs/partitions/Kconfig +++ b/fs/partitions/Kconfig @@ -99,7 +99,7 @@ config IBM_PARTITION config MAC_PARTITION bool "Macintosh partition map support" if PARTITION_ADVANCED - default y if MAC + default y if (MAC || PPC_PMAC) help Say Y here if you would like to use hard disks under Linux which were partitioned on a Macintosh. diff --git a/fs/partitions/Makefile b/fs/partitions/Makefile index 42c7d3878ed..d713ce6b3e1 100644 --- a/fs/partitions/Makefile +++ b/fs/partitions/Makefile @@ -4,7 +4,6 @@ obj-y := check.o -obj-$(CONFIG_DEVFS_FS) += devfs.o obj-$(CONFIG_ACORN_PARTITION) += acorn.o obj-$(CONFIG_AMIGA_PARTITION) += amiga.o obj-$(CONFIG_ATARI_PARTITION) += atari.o diff --git a/fs/partitions/acorn.c b/fs/partitions/acorn.c index c05085710fc..1bc9f372c7d 100644 --- a/fs/partitions/acorn.c +++ b/fs/partitions/acorn.c @@ -12,7 +12,6 @@ * every single manufacturer of SCSI and IDE cards created their own * method. */ -#include <linux/config.h> #include <linux/buffer_head.h> #include <linux/adfs_fs.h> diff --git a/fs/partitions/check.c b/fs/partitions/check.c index 2ef313a96b6..51c6a748df4 100644 --- a/fs/partitions/check.c +++ b/fs/partitions/check.c @@ -18,10 +18,8 @@ #include <linux/fs.h> #include <linux/kmod.h> #include <linux/ctype.h> -#include <linux/devfs_fs_kernel.h> #include "check.h" -#include "devfs.h" #include "acorn.h" #include "amiga.h" @@ -161,18 +159,11 @@ check_partition(struct gendisk *hd, struct block_device *bdev) if (!state) return NULL; -#ifdef CONFIG_DEVFS_FS - if (hd->devfs_name[0] != '\0') { - printk(KERN_INFO " /dev/%s:", hd->devfs_name); + disk_name(hd, 0, state->name); + printk(KERN_INFO " %s:", state->name); + if (isdigit(state->name[strlen(state->name)-1])) sprintf(state->name, "p"); - } -#endif - else { - disk_name(hd, 0, state->name); - printk(KERN_INFO " %s:", state->name); - if (isdigit(state->name[strlen(state->name)-1])) - sprintf(state->name, "p"); - } + state->limit = hd->minors; i = res = 0; while (!res && check_part[i]) { @@ -328,7 +319,6 @@ void delete_partition(struct gendisk *disk, int part) p->nr_sects = 0; p->ios[0] = p->ios[1] = 0; p->sectors[0] = p->sectors[1] = 0; - devfs_remove("%s/part%d", disk->devfs_name, part); sysfs_remove_link(&p->kobj, "subsystem"); if (p->holder_dir) kobject_unregister(p->holder_dir); @@ -349,10 +339,7 @@ void add_partition(struct gendisk *disk, int part, sector_t start, sector_t len) p->start_sect = start; p->nr_sects = len; p->partno = part; - - devfs_mk_bdev(MKDEV(disk->major, disk->first_minor + part), - S_IFBLK|S_IRUSR|S_IWUSR, - "%s/part%d", disk->devfs_name, part); + p->policy = disk->policy; if (isdigit(disk->kobj.name[strlen(disk->kobj.name)-1])) snprintf(p->kobj.name,KOBJ_NAME_LEN,"%sp%d",disk->kobj.name,part); @@ -423,14 +410,8 @@ void register_disk(struct gendisk *disk) disk_sysfs_add_subdirs(disk); /* No minors to use for partitions */ - if (disk->minors == 1) { - if (disk->devfs_name[0] != '\0') - devfs_add_disk(disk); + if (disk->minors == 1) goto exit; - } - - /* always add handle for the whole disk */ - devfs_add_partitioned(disk); /* No such device (e.g., media were just removed) */ if (!get_capacity(disk)) @@ -538,8 +519,6 @@ void del_gendisk(struct gendisk *disk) disk_stat_set_all(disk, 0); disk->stamp = 0; - devfs_remove_disk(disk); - kobject_uevent(&disk->kobj, KOBJ_REMOVE); if (disk->holder_dir) kobject_unregister(disk->holder_dir); diff --git a/fs/partitions/devfs.c b/fs/partitions/devfs.c deleted file mode 100644 index 3f0a780c9ce..00000000000 --- a/fs/partitions/devfs.c +++ /dev/null @@ -1,130 +0,0 @@ -/* - * This tries to keep block devices away from devfs as much as possible. - */ -#include <linux/fs.h> -#include <linux/devfs_fs_kernel.h> -#include <linux/vmalloc.h> -#include <linux/genhd.h> -#include <linux/bitops.h> -#include <linux/mutex.h> - - -struct unique_numspace { - u32 num_free; /* Num free in bits */ - u32 length; /* Array length in bytes */ - unsigned long *bits; - struct semaphore mutex; -}; - -static DEFINE_MUTEX(numspace_mutex); - -static int expand_numspace(struct unique_numspace *s) -{ - u32 length; - void *bits; - - if (s->length < 16) - length = 16; - else - length = s->length << 1; - - bits = vmalloc(length); - if (!bits) - return -ENOMEM; - if (s->bits) { - memcpy(bits, s->bits, s->length); - vfree(s->bits); - } - - s->num_free = (length - s->length) << 3; - s->bits = bits; - memset(bits + s->length, 0, length - s->length); - s->length = length; - - return 0; -} - -static int alloc_unique_number(struct unique_numspace *s) -{ - int rval = 0; - - mutex_lock(&numspace_mutex); - if (s->num_free < 1) - rval = expand_numspace(s); - if (!rval) { - rval = find_first_zero_bit(s->bits, s->length << 3); - --s->num_free; - __set_bit(rval, s->bits); - } - mutex_unlock(&numspace_mutex); - - return rval; -} - -static void dealloc_unique_number(struct unique_numspace *s, int number) -{ - int old_val; - - if (number >= 0) { - mutex_lock(&numspace_mutex); - old_val = __test_and_clear_bit(number, s->bits); - if (old_val) - ++s->num_free; - mutex_unlock(&numspace_mutex); - } -} - -static struct unique_numspace disc_numspace; -static struct unique_numspace cdrom_numspace; - -void devfs_add_partitioned(struct gendisk *disk) -{ - char dirname[64], symlink[16]; - - devfs_mk_dir(disk->devfs_name); - devfs_mk_bdev(MKDEV(disk->major, disk->first_minor), - S_IFBLK|S_IRUSR|S_IWUSR, - "%s/disc", disk->devfs_name); - - disk->number = alloc_unique_number(&disc_numspace); - - sprintf(symlink, "discs/disc%d", disk->number); - sprintf(dirname, "../%s", disk->devfs_name); - devfs_mk_symlink(symlink, dirname); - -} - -void devfs_add_disk(struct gendisk *disk) -{ - devfs_mk_bdev(MKDEV(disk->major, disk->first_minor), - (disk->flags & GENHD_FL_CD) ? - S_IFBLK|S_IRUGO|S_IWUGO : - S_IFBLK|S_IRUSR|S_IWUSR, - "%s", disk->devfs_name); - - if (disk->flags & GENHD_FL_CD) { - char dirname[64], symlink[16]; - - disk->number = alloc_unique_number(&cdrom_numspace); - - sprintf(symlink, "cdroms/cdrom%d", disk->number); - sprintf(dirname, "../%s", disk->devfs_name); - devfs_mk_symlink(symlink, dirname); - } -} - -void devfs_remove_disk(struct gendisk *disk) -{ - if (disk->minors != 1) { - devfs_remove("discs/disc%d", disk->number); - dealloc_unique_number(&disc_numspace, disk->number); - devfs_remove("%s/disc", disk->devfs_name); - } - if (disk->flags & GENHD_FL_CD) { - devfs_remove("cdroms/cdrom%d", disk->number); - dealloc_unique_number(&cdrom_numspace, disk->number); - } - devfs_remove(disk->devfs_name); -} - - diff --git a/fs/partitions/devfs.h b/fs/partitions/devfs.h deleted file mode 100644 index 176118b4e49..00000000000 --- a/fs/partitions/devfs.h +++ /dev/null @@ -1,10 +0,0 @@ - -#ifdef CONFIG_DEVFS_FS -void devfs_add_disk(struct gendisk *dev); -void devfs_add_partitioned(struct gendisk *dev); -void devfs_remove_disk(struct gendisk *dev); -#else -# define devfs_add_disk(disk) do { } while (0) -# define devfs_add_partitioned(disk) do { } while (0) -# define devfs_remove_disk(disk) do { } while (0) -#endif diff --git a/fs/partitions/efi.c b/fs/partitions/efi.c index 0f5b017aeba..1bea610078b 100644 --- a/fs/partitions/efi.c +++ b/fs/partitions/efi.c @@ -91,7 +91,6 @@ * - Code works, detects all the partitions. * ************************************************************/ -#include <linux/config.h> #include <linux/crc32.h> #include "check.h" #include "efi.h" @@ -239,10 +238,9 @@ alloc_read_gpt_entries(struct block_device *bdev, gpt_header *gpt) le32_to_cpu(gpt->sizeof_partition_entry); if (!count) return NULL; - pte = kmalloc(count, GFP_KERNEL); + pte = kzalloc(count, GFP_KERNEL); if (!pte) return NULL; - memset(pte, 0, count); if (read_lba(bdev, le64_to_cpu(gpt->partition_entry_lba), (u8 *) pte, @@ -270,10 +268,9 @@ alloc_read_gpt_header(struct block_device *bdev, u64 lba) if (!bdev) return NULL; - gpt = kmalloc(sizeof (gpt_header), GFP_KERNEL); + gpt = kzalloc(sizeof (gpt_header), GFP_KERNEL); if (!gpt) return NULL; - memset(gpt, 0, sizeof (gpt_header)); if (read_lba(bdev, lba, (u8 *) gpt, sizeof (gpt_header)) < sizeof (gpt_header)) { @@ -527,9 +524,8 @@ find_valid_gpt(struct block_device *bdev, gpt_header **gpt, gpt_entry **ptes) lastlba = last_lba(bdev); if (!force_gpt) { /* This will be added to the EFI Spec. per Intel after v1.02. */ - legacymbr = kmalloc(sizeof (*legacymbr), GFP_KERNEL); + legacymbr = kzalloc(sizeof (*legacymbr), GFP_KERNEL); if (legacymbr) { - memset(legacymbr, 0, sizeof (*legacymbr)); read_lba(bdev, 0, (u8 *) legacymbr, sizeof (*legacymbr)); good_pmbr = is_pmbr_valid(legacymbr, lastlba); diff --git a/fs/partitions/efi.h b/fs/partitions/efi.h index c44fb056144..2cc89d0475b 100644 --- a/fs/partitions/efi.h +++ b/fs/partitions/efi.h @@ -26,7 +26,6 @@ #define FS_PART_EFI_H_INCLUDED #include <linux/types.h> -#include <linux/config.h> #include <linux/fs.h> #include <linux/genhd.h> #include <linux/kernel.h> diff --git a/fs/partitions/ibm.c b/fs/partitions/ibm.c index 830c55d86ab..d352a7381fe 100644 --- a/fs/partitions/ibm.c +++ b/fs/partitions/ibm.c @@ -6,7 +6,6 @@ * (C) IBM Corporation, IBM Deutschland Entwicklung GmbH, 1999,2000 */ -#include <linux/config.h> #include <linux/buffer_head.h> #include <linux/hdreg.h> #include <linux/slab.h> diff --git a/fs/partitions/mac.c b/fs/partitions/mac.c index 813292f2121..c0871002d00 100644 --- a/fs/partitions/mac.c +++ b/fs/partitions/mac.c @@ -6,7 +6,6 @@ * Re-organised Feb 1998 Russell King */ -#include <linux/config.h> #include <linux/ctype.h> #include "check.h" #include "mac.h" diff --git a/fs/partitions/msdos.c b/fs/partitions/msdos.c index 9935d254186..4f8df71e49d 100644 --- a/fs/partitions/msdos.c +++ b/fs/partitions/msdos.c @@ -19,7 +19,6 @@ * Re-organised Feb 1998 Russell King */ -#include <linux/config.h> #include "check.h" #include "msdos.h" @@ -59,6 +58,31 @@ msdos_magic_present(unsigned char *p) return (p[0] == MSDOS_LABEL_MAGIC1 && p[1] == MSDOS_LABEL_MAGIC2); } +/* Value is EBCDIC 'IBMA' */ +#define AIX_LABEL_MAGIC1 0xC9 +#define AIX_LABEL_MAGIC2 0xC2 +#define AIX_LABEL_MAGIC3 0xD4 +#define AIX_LABEL_MAGIC4 0xC1 +static int aix_magic_present(unsigned char *p, struct block_device *bdev) +{ + Sector sect; + unsigned char *d; + int ret = 0; + + if (p[0] != AIX_LABEL_MAGIC1 && + p[1] != AIX_LABEL_MAGIC2 && + p[2] != AIX_LABEL_MAGIC3 && + p[3] != AIX_LABEL_MAGIC4) + return 0; + d = read_dev_sector(bdev, 7, §); + if (d) { + if (d[0] == '_' && d[1] == 'L' && d[2] == 'V' && d[3] == 'M') + ret = 1; + put_dev_sector(sect); + }; + return ret; +} + /* * Create devices for each logical partition in an extended partition. * The logical partitions form a linked list, with each entry being @@ -394,6 +418,12 @@ int msdos_partition(struct parsed_partitions *state, struct block_device *bdev) return 0; } + if (aix_magic_present(data, bdev)) { + put_dev_sector(sect); + printk( " [AIX]"); + return 0; + } + /* * Now that the 55aa signature is present, this is probably * either the boot sector of a FAT filesystem or a DOS-type diff --git a/fs/partitions/sun.c b/fs/partitions/sun.c index abe91ca03ed..0a5927c806c 100644 --- a/fs/partitions/sun.c +++ b/fs/partitions/sun.c @@ -74,7 +74,7 @@ int sun_partition(struct parsed_partitions *state, struct block_device *bdev) spc = be16_to_cpu(label->ntrks) * be16_to_cpu(label->nsect); for (i = 0; i < 8; i++, p++) { unsigned long st_sector; - int num_sectors; + unsigned int num_sectors; st_sector = be32_to_cpu(p->start_cylinder) * spc; num_sectors = be32_to_cpu(p->num_sectors); diff --git a/fs/pipe.c b/fs/pipe.c index 20352573e02..f3b6f71e9d0 100644 --- a/fs/pipe.c +++ b/fs/pipe.c @@ -879,7 +879,6 @@ static struct inode * get_pipe_inode(void) inode->i_uid = current->fsuid; inode->i_gid = current->fsgid; inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME; - inode->i_blksize = PAGE_SIZE; return inode; diff --git a/fs/pnode.c b/fs/pnode.c index 37b568ed0e0..da42ee61c1d 100644 --- a/fs/pnode.c +++ b/fs/pnode.c @@ -53,8 +53,7 @@ static int do_make_slave(struct vfsmount *mnt) if (master) { list_for_each_entry(slave_mnt, &mnt->mnt_slave_list, mnt_slave) slave_mnt->mnt_master = master; - list_del(&mnt->mnt_slave); - list_add(&mnt->mnt_slave, &master->mnt_slave_list); + list_move(&mnt->mnt_slave, &master->mnt_slave_list); list_splice(&mnt->mnt_slave_list, master->mnt_slave_list.prev); INIT_LIST_HEAD(&mnt->mnt_slave_list); } else { @@ -283,10 +282,8 @@ static void __propagate_umount(struct vfsmount *mnt) * umount the child only if the child has no * other children */ - if (child && list_empty(&child->mnt_mounts)) { - list_del(&child->mnt_hash); - list_add_tail(&child->mnt_hash, &mnt->mnt_hash); - } + if (child && list_empty(&child->mnt_mounts)) + list_move_tail(&child->mnt_hash, &mnt->mnt_hash); } } diff --git a/fs/proc/array.c b/fs/proc/array.c index 7a76ad57023..c0e554971df 100644 --- a/fs/proc/array.c +++ b/fs/proc/array.c @@ -52,7 +52,6 @@ * : base.c too. */ -#include <linux/config.h> #include <linux/types.h> #include <linux/errno.h> #include <linux/time.h> @@ -75,6 +74,7 @@ #include <linux/times.h> #include <linux/cpuset.h> #include <linux/rcupdate.h> +#include <linux/delayacct.h> #include <asm/uaccess.h> #include <asm/pgtable.h> @@ -347,6 +347,8 @@ static int do_task_stat(struct task_struct *task, char * buffer, int whole) sigemptyset(&sigign); sigemptyset(&sigcatch); cutime = cstime = utime = stime = cputime_zero; + + mutex_lock(&tty_mutex); read_lock(&tasklist_lock); if (task->sighand) { spin_lock_irq(&task->sighand->siglock); @@ -388,6 +390,7 @@ static int do_task_stat(struct task_struct *task, char * buffer, int whole) } ppid = pid_alive(task) ? task->group_leader->real_parent->tgid : 0; read_unlock(&tasklist_lock); + mutex_unlock(&tty_mutex); if (!whole || num_threads<2) wchan = get_wchan(task); @@ -412,7 +415,7 @@ static int do_task_stat(struct task_struct *task, char * buffer, int whole) res = sprintf(buffer,"%d (%s) %c %d %d %d %d %d %lu %lu \ %lu %lu %lu %lu %lu %ld %ld %ld %ld %d 0 %llu %lu %ld %lu %lu %lu %lu %lu \ -%lu %lu %lu %lu %lu %lu %lu %lu %d %d %lu %lu\n", +%lu %lu %lu %lu %lu %lu %lu %lu %d %d %lu %lu %llu\n", task->pid, tcomm, state, @@ -456,7 +459,8 @@ static int do_task_stat(struct task_struct *task, char * buffer, int whole) task->exit_signal, task_cpu(task), task->rt_priority, - task->policy); + task->policy, + (unsigned long long)delayacct_blkio_ticks(task)); if(mm) mmput(mm); return res; diff --git a/fs/proc/base.c b/fs/proc/base.c index 6afff725a8c..89c20d9d50b 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -49,7 +49,6 @@ #include <asm/uaccess.h> -#include <linux/config.h> #include <linux/errno.h> #include <linux/time.h> #include <linux/proc_fs.h> @@ -74,6 +73,16 @@ #include <linux/poll.h> #include "internal.h" +/* NOTE: + * Implementing inode permission operations in /proc is almost + * certainly an error. Permission checks need to happen during + * each system call not at open time. The reason is that most of + * what we wish to check for permissions in /proc varies at runtime. + * + * The classic example of a problem is opening file descriptors + * in /proc for a task before it execs a suid executable. + */ + /* * For hysterical raisins we keep the same inumbers as in the old procfs. * Feel free to change the macro below - just keep the range distinct from @@ -121,6 +130,8 @@ enum pid_directory_inos { PROC_TGID_ATTR_PREV, PROC_TGID_ATTR_EXEC, PROC_TGID_ATTR_FSCREATE, + PROC_TGID_ATTR_KEYCREATE, + PROC_TGID_ATTR_SOCKCREATE, #endif #ifdef CONFIG_AUDITSYSCALL PROC_TGID_LOGINUID, @@ -162,6 +173,8 @@ enum pid_directory_inos { PROC_TID_ATTR_PREV, PROC_TID_ATTR_EXEC, PROC_TID_ATTR_FSCREATE, + PROC_TID_ATTR_KEYCREATE, + PROC_TID_ATTR_SOCKCREATE, #endif #ifdef CONFIG_AUDITSYSCALL PROC_TID_LOGINUID, @@ -173,6 +186,9 @@ enum pid_directory_inos { PROC_TID_FD_DIR = 0x8000, /* 0x8000-0xffff */ }; +/* Worst case buffer size needed for holding an integer. */ +#define PROC_NUMBUF 10 + struct pid_entry { int type; int len; @@ -275,6 +291,8 @@ static struct pid_entry tgid_attr_stuff[] = { E(PROC_TGID_ATTR_PREV, "prev", S_IFREG|S_IRUGO), E(PROC_TGID_ATTR_EXEC, "exec", S_IFREG|S_IRUGO|S_IWUGO), E(PROC_TGID_ATTR_FSCREATE, "fscreate", S_IFREG|S_IRUGO|S_IWUGO), + E(PROC_TGID_ATTR_KEYCREATE, "keycreate", S_IFREG|S_IRUGO|S_IWUGO), + E(PROC_TGID_ATTR_SOCKCREATE, "sockcreate", S_IFREG|S_IRUGO|S_IWUGO), {0,0,NULL,0} }; static struct pid_entry tid_attr_stuff[] = { @@ -282,6 +300,8 @@ static struct pid_entry tid_attr_stuff[] = { E(PROC_TID_ATTR_PREV, "prev", S_IFREG|S_IRUGO), E(PROC_TID_ATTR_EXEC, "exec", S_IFREG|S_IRUGO|S_IWUGO), E(PROC_TID_ATTR_FSCREATE, "fscreate", S_IFREG|S_IRUGO|S_IWUGO), + E(PROC_TID_ATTR_KEYCREATE, "keycreate", S_IFREG|S_IRUGO|S_IWUGO), + E(PROC_TID_ATTR_SOCKCREATE, "sockcreate", S_IFREG|S_IRUGO|S_IWUGO), {0,0,NULL,0} }; #endif @@ -290,12 +310,15 @@ static struct pid_entry tid_attr_stuff[] = { static int proc_fd_link(struct inode *inode, struct dentry **dentry, struct vfsmount **mnt) { - struct task_struct *task = proc_task(inode); - struct files_struct *files; + struct task_struct *task = get_proc_task(inode); + struct files_struct *files = NULL; struct file *file; - int fd = proc_type(inode) - PROC_TID_FD_DIR; + int fd = proc_fd(inode); - files = get_files_struct(task); + if (task) { + files = get_files_struct(task); + put_task_struct(task); + } if (files) { /* * We are not taking a ref to the file structure, so we must @@ -327,29 +350,33 @@ static struct fs_struct *get_fs_struct(struct task_struct *task) return fs; } -static int proc_cwd_link(struct inode *inode, struct dentry **dentry, struct vfsmount **mnt) +static int get_nr_threads(struct task_struct *tsk) { - struct fs_struct *fs = get_fs_struct(proc_task(inode)); - int result = -ENOENT; - if (fs) { - read_lock(&fs->lock); - *mnt = mntget(fs->pwdmnt); - *dentry = dget(fs->pwd); - read_unlock(&fs->lock); - result = 0; - put_fs_struct(fs); + /* Must be called with the rcu_read_lock held */ + unsigned long flags; + int count = 0; + + if (lock_task_sighand(tsk, &flags)) { + count = atomic_read(&tsk->signal->count); + unlock_task_sighand(tsk, &flags); } - return result; + return count; } -static int proc_root_link(struct inode *inode, struct dentry **dentry, struct vfsmount **mnt) +static int proc_cwd_link(struct inode *inode, struct dentry **dentry, struct vfsmount **mnt) { - struct fs_struct *fs = get_fs_struct(proc_task(inode)); + struct task_struct *task = get_proc_task(inode); + struct fs_struct *fs = NULL; int result = -ENOENT; + + if (task) { + fs = get_fs_struct(task); + put_task_struct(task); + } if (fs) { read_lock(&fs->lock); - *mnt = mntget(fs->rootmnt); - *dentry = dget(fs->root); + *mnt = mntget(fs->pwdmnt); + *dentry = dget(fs->pwd); read_unlock(&fs->lock); result = 0; put_fs_struct(fs); @@ -357,42 +384,16 @@ static int proc_root_link(struct inode *inode, struct dentry **dentry, struct vf return result; } - -/* Same as proc_root_link, but this addionally tries to get fs from other - * threads in the group */ -static int proc_task_root_link(struct inode *inode, struct dentry **dentry, - struct vfsmount **mnt) +static int proc_root_link(struct inode *inode, struct dentry **dentry, struct vfsmount **mnt) { - struct fs_struct *fs; + struct task_struct *task = get_proc_task(inode); + struct fs_struct *fs = NULL; int result = -ENOENT; - struct task_struct *leader = proc_task(inode); - task_lock(leader); - fs = leader->fs; - if (fs) { - atomic_inc(&fs->count); - task_unlock(leader); - } else { - /* Try to get fs from other threads */ - task_unlock(leader); - read_lock(&tasklist_lock); - if (pid_alive(leader)) { - struct task_struct *task = leader; - - while ((task = next_thread(task)) != leader) { - task_lock(task); - fs = task->fs; - if (fs) { - atomic_inc(&fs->count); - task_unlock(task); - break; - } - task_unlock(task); - } - } - read_unlock(&tasklist_lock); + if (task) { + fs = get_fs_struct(task); + put_task_struct(task); } - if (fs) { read_lock(&fs->lock); *mnt = mntget(fs->rootmnt); @@ -404,7 +405,6 @@ static int proc_task_root_link(struct inode *inode, struct dentry **dentry, return result; } - #define MAY_PTRACE(task) \ (task == current || \ (task->parent == current && \ @@ -535,141 +535,42 @@ static int proc_oom_score(struct task_struct *task, char *buffer) /************************************************************************/ /* permission checks */ - -/* If the process being read is separated by chroot from the reading process, - * don't let the reader access the threads. - * - * note: this does dput(root) and mntput(vfsmnt) on exit. - */ -static int proc_check_chroot(struct dentry *root, struct vfsmount *vfsmnt) +static int proc_fd_access_allowed(struct inode *inode) { - struct dentry *de, *base; - struct vfsmount *our_vfsmnt, *mnt; - int res = 0; - - read_lock(¤t->fs->lock); - our_vfsmnt = mntget(current->fs->rootmnt); - base = dget(current->fs->root); - read_unlock(¤t->fs->lock); - - spin_lock(&vfsmount_lock); - de = root; - mnt = vfsmnt; - - while (mnt != our_vfsmnt) { - if (mnt == mnt->mnt_parent) - goto out; - de = mnt->mnt_mountpoint; - mnt = mnt->mnt_parent; - } - - if (!is_subdir(de, base)) - goto out; - spin_unlock(&vfsmount_lock); - -exit: - dput(base); - mntput(our_vfsmnt); - dput(root); - mntput(vfsmnt); - return res; -out: - spin_unlock(&vfsmount_lock); - res = -EACCES; - goto exit; -} - -static int proc_check_root(struct inode *inode) -{ - struct dentry *root; - struct vfsmount *vfsmnt; - - if (proc_root_link(inode, &root, &vfsmnt)) /* Ewww... */ - return -ENOENT; - return proc_check_chroot(root, vfsmnt); -} - -static int proc_permission(struct inode *inode, int mask, struct nameidata *nd) -{ - if (generic_permission(inode, mask, NULL) != 0) - return -EACCES; - return proc_check_root(inode); -} - -static int proc_task_permission(struct inode *inode, int mask, struct nameidata *nd) -{ - struct dentry *root; - struct vfsmount *vfsmnt; - - if (generic_permission(inode, mask, NULL) != 0) - return -EACCES; - - if (proc_task_root_link(inode, &root, &vfsmnt)) - return -ENOENT; - - return proc_check_chroot(root, vfsmnt); -} - -extern struct seq_operations proc_pid_maps_op; -static int maps_open(struct inode *inode, struct file *file) -{ - struct task_struct *task = proc_task(inode); - int ret = seq_open(file, &proc_pid_maps_op); - if (!ret) { - struct seq_file *m = file->private_data; - m->private = task; + struct task_struct *task; + int allowed = 0; + /* Allow access to a task's file descriptors if it is us or we + * may use ptrace attach to the process and find out that + * information. + */ + task = get_proc_task(inode); + if (task) { + allowed = ptrace_may_attach(task); + put_task_struct(task); } - return ret; + return allowed; } -static struct file_operations proc_maps_operations = { - .open = maps_open, - .read = seq_read, - .llseek = seq_lseek, - .release = seq_release, -}; - -#ifdef CONFIG_NUMA -extern struct seq_operations proc_pid_numa_maps_op; -static int numa_maps_open(struct inode *inode, struct file *file) +static int proc_setattr(struct dentry *dentry, struct iattr *attr) { - struct task_struct *task = proc_task(inode); - int ret = seq_open(file, &proc_pid_numa_maps_op); - if (!ret) { - struct seq_file *m = file->private_data; - m->private = task; - } - return ret; -} + int error; + struct inode *inode = dentry->d_inode; -static struct file_operations proc_numa_maps_operations = { - .open = numa_maps_open, - .read = seq_read, - .llseek = seq_lseek, - .release = seq_release, -}; -#endif + if (attr->ia_valid & ATTR_MODE) + return -EPERM; -#ifdef CONFIG_MMU -extern struct seq_operations proc_pid_smaps_op; -static int smaps_open(struct inode *inode, struct file *file) -{ - struct task_struct *task = proc_task(inode); - int ret = seq_open(file, &proc_pid_smaps_op); - if (!ret) { - struct seq_file *m = file->private_data; - m->private = task; + error = inode_change_ok(inode, attr); + if (!error) { + error = security_inode_setattr(dentry, attr); + if (!error) + error = inode_setattr(inode, attr); } - return ret; + return error; } -static struct file_operations proc_smaps_operations = { - .open = smaps_open, - .read = seq_read, - .llseek = seq_lseek, - .release = seq_release, +static struct inode_operations proc_def_inode_operations = { + .setattr = proc_setattr, }; -#endif extern struct seq_operations mounts_op; struct proc_mounts { @@ -679,16 +580,19 @@ struct proc_mounts { static int mounts_open(struct inode *inode, struct file *file) { - struct task_struct *task = proc_task(inode); - struct namespace *namespace; + struct task_struct *task = get_proc_task(inode); + struct namespace *namespace = NULL; struct proc_mounts *p; int ret = -EINVAL; - task_lock(task); - namespace = task->namespace; - if (namespace) - get_namespace(namespace); - task_unlock(task); + if (task) { + task_lock(task); + namespace = task->namespace; + if (namespace) + get_namespace(namespace); + task_unlock(task); + put_task_struct(task); + } if (namespace) { ret = -ENOMEM; @@ -745,17 +649,21 @@ static struct file_operations proc_mounts_operations = { extern struct seq_operations mountstats_op; static int mountstats_open(struct inode *inode, struct file *file) { - struct task_struct *task = proc_task(inode); int ret = seq_open(file, &mountstats_op); if (!ret) { struct seq_file *m = file->private_data; - struct namespace *namespace; - task_lock(task); - namespace = task->namespace; - if (namespace) - get_namespace(namespace); - task_unlock(task); + struct namespace *namespace = NULL; + struct task_struct *task = get_proc_task(inode); + + if (task) { + task_lock(task); + namespace = task->namespace; + if (namespace) + get_namespace(namespace); + task_unlock(task); + put_task_struct(task); + } if (namespace) m->private = namespace; @@ -782,18 +690,27 @@ static ssize_t proc_info_read(struct file * file, char __user * buf, struct inode * inode = file->f_dentry->d_inode; unsigned long page; ssize_t length; - struct task_struct *task = proc_task(inode); + struct task_struct *task = get_proc_task(inode); + + length = -ESRCH; + if (!task) + goto out_no_task; if (count > PROC_BLOCK_SIZE) count = PROC_BLOCK_SIZE; + + length = -ENOMEM; if (!(page = __get_free_page(GFP_KERNEL))) - return -ENOMEM; + goto out; length = PROC_I(inode)->op.proc_read(task, (char*)page); if (length >= 0) length = simple_read_from_buffer(buf, count, ppos, (char *)page, length); free_page(page); +out: + put_task_struct(task); +out_no_task: return length; } @@ -810,12 +727,15 @@ static int mem_open(struct inode* inode, struct file* file) static ssize_t mem_read(struct file * file, char __user * buf, size_t count, loff_t *ppos) { - struct task_struct *task = proc_task(file->f_dentry->d_inode); + struct task_struct *task = get_proc_task(file->f_dentry->d_inode); char *page; unsigned long src = *ppos; int ret = -ESRCH; struct mm_struct *mm; + if (!task) + goto out_no_task; + if (!MAY_PTRACE(task) || !ptrace_may_attach(task)) goto out; @@ -865,6 +785,8 @@ out_put: out_free: free_page((unsigned long) page); out: + put_task_struct(task); +out_no_task: return ret; } @@ -875,18 +797,24 @@ out: static ssize_t mem_write(struct file * file, const char * buf, size_t count, loff_t *ppos) { - int copied = 0; + int copied; char *page; - struct task_struct *task = proc_task(file->f_dentry->d_inode); + struct task_struct *task = get_proc_task(file->f_dentry->d_inode); unsigned long dst = *ppos; + copied = -ESRCH; + if (!task) + goto out_no_task; + if (!MAY_PTRACE(task) || !ptrace_may_attach(task)) - return -ESRCH; + goto out; + copied = -ENOMEM; page = (char *)__get_free_page(GFP_USER); if (!page) - return -ENOMEM; + goto out; + copied = 0; while (count > 0) { int this_len, retval; @@ -908,6 +836,9 @@ static ssize_t mem_write(struct file * file, const char * buf, } *ppos = dst; free_page((unsigned long) page); +out: + put_task_struct(task); +out_no_task: return copied; } #endif @@ -938,13 +869,18 @@ static struct file_operations proc_mem_operations = { static ssize_t oom_adjust_read(struct file *file, char __user *buf, size_t count, loff_t *ppos) { - struct task_struct *task = proc_task(file->f_dentry->d_inode); - char buffer[8]; + struct task_struct *task = get_proc_task(file->f_dentry->d_inode); + char buffer[PROC_NUMBUF]; size_t len; - int oom_adjust = task->oomkilladj; + int oom_adjust; loff_t __ppos = *ppos; - len = sprintf(buffer, "%i\n", oom_adjust); + if (!task) + return -ESRCH; + oom_adjust = task->oomkilladj; + put_task_struct(task); + + len = snprintf(buffer, sizeof(buffer), "%i\n", oom_adjust); if (__ppos >= len) return 0; if (count > len-__ppos) @@ -958,15 +894,15 @@ static ssize_t oom_adjust_read(struct file *file, char __user *buf, static ssize_t oom_adjust_write(struct file *file, const char __user *buf, size_t count, loff_t *ppos) { - struct task_struct *task = proc_task(file->f_dentry->d_inode); - char buffer[8], *end; + struct task_struct *task; + char buffer[PROC_NUMBUF], *end; int oom_adjust; if (!capable(CAP_SYS_RESOURCE)) return -EPERM; - memset(buffer, 0, 8); - if (count > 6) - count = 6; + memset(buffer, 0, sizeof(buffer)); + if (count > sizeof(buffer) - 1) + count = sizeof(buffer) - 1; if (copy_from_user(buffer, buf, count)) return -EFAULT; oom_adjust = simple_strtol(buffer, &end, 0); @@ -974,7 +910,11 @@ static ssize_t oom_adjust_write(struct file *file, const char __user *buf, return -EINVAL; if (*end == '\n') end++; + task = get_proc_task(file->f_dentry->d_inode); + if (!task) + return -ESRCH; task->oomkilladj = oom_adjust; + put_task_struct(task); if (end - buffer == 0) return -EIO; return end - buffer; @@ -985,22 +925,21 @@ static struct file_operations proc_oom_adjust_operations = { .write = oom_adjust_write, }; -static struct inode_operations proc_mem_inode_operations = { - .permission = proc_permission, -}; - #ifdef CONFIG_AUDITSYSCALL #define TMPBUFLEN 21 static ssize_t proc_loginuid_read(struct file * file, char __user * buf, size_t count, loff_t *ppos) { struct inode * inode = file->f_dentry->d_inode; - struct task_struct *task = proc_task(inode); + struct task_struct *task = get_proc_task(inode); ssize_t length; char tmpbuf[TMPBUFLEN]; + if (!task) + return -ESRCH; length = scnprintf(tmpbuf, TMPBUFLEN, "%u", audit_get_loginuid(task->audit_context)); + put_task_struct(task); return simple_read_from_buffer(buf, count, ppos, tmpbuf, length); } @@ -1010,13 +949,12 @@ static ssize_t proc_loginuid_write(struct file * file, const char __user * buf, struct inode * inode = file->f_dentry->d_inode; char *page, *tmp; ssize_t length; - struct task_struct *task = proc_task(inode); uid_t loginuid; if (!capable(CAP_AUDIT_CONTROL)) return -EPERM; - if (current != task) + if (current != pid_task(proc_pid(inode), PIDTYPE_PID)) return -EPERM; if (count >= PAGE_SIZE) @@ -1040,7 +978,7 @@ static ssize_t proc_loginuid_write(struct file * file, const char __user * buf, goto out_free_page; } - length = audit_set_loginuid(task, loginuid); + length = audit_set_loginuid(current, loginuid); if (likely(length == 0)) length = count; @@ -1059,13 +997,16 @@ static struct file_operations proc_loginuid_operations = { static ssize_t seccomp_read(struct file *file, char __user *buf, size_t count, loff_t *ppos) { - struct task_struct *tsk = proc_task(file->f_dentry->d_inode); + struct task_struct *tsk = get_proc_task(file->f_dentry->d_inode); char __buf[20]; loff_t __ppos = *ppos; size_t len; + if (!tsk) + return -ESRCH; /* no need to print the trailing zero, so use only len */ len = sprintf(__buf, "%u\n", tsk->seccomp.mode); + put_task_struct(tsk); if (__ppos >= len) return 0; if (count > len - __ppos) @@ -1079,29 +1020,43 @@ static ssize_t seccomp_read(struct file *file, char __user *buf, static ssize_t seccomp_write(struct file *file, const char __user *buf, size_t count, loff_t *ppos) { - struct task_struct *tsk = proc_task(file->f_dentry->d_inode); + struct task_struct *tsk = get_proc_task(file->f_dentry->d_inode); char __buf[20], *end; unsigned int seccomp_mode; + ssize_t result; + + result = -ESRCH; + if (!tsk) + goto out_no_task; /* can set it only once to be even more secure */ + result = -EPERM; if (unlikely(tsk->seccomp.mode)) - return -EPERM; + goto out; + result = -EFAULT; memset(__buf, 0, sizeof(__buf)); count = min(count, sizeof(__buf) - 1); if (copy_from_user(__buf, buf, count)) - return -EFAULT; + goto out; + seccomp_mode = simple_strtoul(__buf, &end, 0); if (*end == '\n') end++; + result = -EINVAL; if (seccomp_mode && seccomp_mode <= NR_SECCOMP_MODES) { tsk->seccomp.mode = seccomp_mode; set_tsk_thread_flag(tsk, TIF_SECCOMP); } else - return -EINVAL; + goto out; + result = -EIO; if (unlikely(!(end - __buf))) - return -EIO; - return end - __buf; + goto out; + result = end - __buf; +out: + put_task_struct(tsk); +out_no_task: + return result; } static struct file_operations proc_seccomp_operations = { @@ -1118,10 +1073,8 @@ static void *proc_pid_follow_link(struct dentry *dentry, struct nameidata *nd) /* We don't need a base pointer in the /proc filesystem */ path_release(nd); - if (current->fsuid != inode->i_uid && !capable(CAP_DAC_OVERRIDE)) - goto out; - error = proc_check_root(inode); - if (error) + /* Are we allowed to snoop on the tasks file descriptors? */ + if (!proc_fd_access_allowed(inode)) goto out; error = PROC_I(inode)->op.proc_get_link(inode, &nd->dentry, &nd->mnt); @@ -1163,12 +1116,8 @@ static int proc_pid_readlink(struct dentry * dentry, char __user * buffer, int b struct dentry *de; struct vfsmount *mnt = NULL; - lock_kernel(); - - if (current->fsuid != inode->i_uid && !capable(CAP_DAC_OVERRIDE)) - goto out; - error = proc_check_root(inode); - if (error) + /* Are we allowed to snoop on the tasks file descriptors? */ + if (!proc_fd_access_allowed(inode)) goto out; error = PROC_I(inode)->op.proc_get_link(inode, &de, &mnt); @@ -1179,30 +1128,29 @@ static int proc_pid_readlink(struct dentry * dentry, char __user * buffer, int b dput(de); mntput(mnt); out: - unlock_kernel(); return error; } static struct inode_operations proc_pid_link_inode_operations = { .readlink = proc_pid_readlink, - .follow_link = proc_pid_follow_link + .follow_link = proc_pid_follow_link, + .setattr = proc_setattr, }; -#define NUMBUF 10 - static int proc_readfd(struct file * filp, void * dirent, filldir_t filldir) { - struct inode *inode = filp->f_dentry->d_inode; - struct task_struct *p = proc_task(inode); + struct dentry *dentry = filp->f_dentry; + struct inode *inode = dentry->d_inode; + struct task_struct *p = get_proc_task(inode); unsigned int fd, tid, ino; int retval; - char buf[NUMBUF]; + char buf[PROC_NUMBUF]; struct files_struct * files; struct fdtable *fdt; retval = -ENOENT; - if (!pid_alive(p)) - goto out; + if (!p) + goto out_no_task; retval = 0; tid = p->pid; @@ -1213,7 +1161,7 @@ static int proc_readfd(struct file * filp, void * dirent, filldir_t filldir) goto out; filp->f_pos++; case 1: - ino = fake_ino(tid, PROC_TID_INO); + ino = parent_ino(dentry); if (filldir(dirent, "..", 2, 1, ino, DT_DIR) < 0) goto out; filp->f_pos++; @@ -1232,7 +1180,7 @@ static int proc_readfd(struct file * filp, void * dirent, filldir_t filldir) continue; rcu_read_unlock(); - j = NUMBUF; + j = PROC_NUMBUF; i = fd; do { j--; @@ -1241,7 +1189,7 @@ static int proc_readfd(struct file * filp, void * dirent, filldir_t filldir) } while (i); ino = fake_ino(tid, PROC_TID_FD_DIR + fd); - if (filldir(dirent, buf+j, NUMBUF-j, fd+2, ino, DT_LNK) < 0) { + if (filldir(dirent, buf+j, PROC_NUMBUF-j, fd+2, ino, DT_LNK) < 0) { rcu_read_lock(); break; } @@ -1251,6 +1199,8 @@ static int proc_readfd(struct file * filp, void * dirent, filldir_t filldir) put_files_struct(files); } out: + put_task_struct(p); +out_no_task: return retval; } @@ -1262,16 +1212,18 @@ static int proc_pident_readdir(struct file *filp, int pid; struct dentry *dentry = filp->f_dentry; struct inode *inode = dentry->d_inode; + struct task_struct *task = get_proc_task(inode); struct pid_entry *p; ino_t ino; int ret; ret = -ENOENT; - if (!pid_alive(proc_task(inode))) + if (!task) goto out; ret = 0; - pid = proc_task(inode)->pid; + pid = task->pid; + put_task_struct(task); i = filp->f_pos; switch (i) { case 0: @@ -1354,22 +1306,20 @@ static struct inode *proc_pid_make_inode(struct super_block * sb, struct task_st /* Common stuff */ ei = PROC_I(inode); - ei->task = NULL; inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME; inode->i_ino = fake_ino(task->pid, ino); - - if (!pid_alive(task)) - goto out_unlock; + inode->i_op = &proc_def_inode_operations; /* * grab the reference to task. */ - get_task_struct(task); - ei->task = task; - ei->type = ino; + ei->pid = get_pid(task->pids[PIDTYPE_PID].pid); + if (!ei->pid) + goto out_unlock; + inode->i_uid = 0; inode->i_gid = 0; - if (ino == PROC_TGID_INO || ino == PROC_TID_INO || task_dumpable(task)) { + if (task_dumpable(task)) { inode->i_uid = task->euid; inode->i_gid = task->egid; } @@ -1379,7 +1329,6 @@ out: return inode; out_unlock: - ei->pde = NULL; iput(inode); return NULL; } @@ -1393,73 +1342,99 @@ out_unlock: * * Rewrite the inode's ownerships here because the owning task may have * performed a setuid(), etc. + * + * Before the /proc/pid/status file was created the only way to read + * the effective uid of a /process was to stat /proc/pid. Reading + * /proc/pid/status is slow enough that procps and other packages + * kept stating /proc/pid. To keep the rules in /proc simple I have + * made this apply to all per process world readable and executable + * directories. */ static int pid_revalidate(struct dentry *dentry, struct nameidata *nd) { struct inode *inode = dentry->d_inode; - struct task_struct *task = proc_task(inode); - if (pid_alive(task)) { - if (proc_type(inode) == PROC_TGID_INO || proc_type(inode) == PROC_TID_INO || task_dumpable(task)) { + struct task_struct *task = get_proc_task(inode); + if (task) { + if ((inode->i_mode == (S_IFDIR|S_IRUGO|S_IXUGO)) || + task_dumpable(task)) { inode->i_uid = task->euid; inode->i_gid = task->egid; } else { inode->i_uid = 0; inode->i_gid = 0; } + inode->i_mode &= ~(S_ISUID | S_ISGID); security_task_to_inode(task, inode); + put_task_struct(task); return 1; } d_drop(dentry); return 0; } +static int pid_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat) +{ + struct inode *inode = dentry->d_inode; + struct task_struct *task; + generic_fillattr(inode, stat); + + rcu_read_lock(); + stat->uid = 0; + stat->gid = 0; + task = pid_task(proc_pid(inode), PIDTYPE_PID); + if (task) { + if ((inode->i_mode == (S_IFDIR|S_IRUGO|S_IXUGO)) || + task_dumpable(task)) { + stat->uid = task->euid; + stat->gid = task->egid; + } + } + rcu_read_unlock(); + return 0; +} + static int tid_fd_revalidate(struct dentry *dentry, struct nameidata *nd) { struct inode *inode = dentry->d_inode; - struct task_struct *task = proc_task(inode); - int fd = proc_type(inode) - PROC_TID_FD_DIR; + struct task_struct *task = get_proc_task(inode); + int fd = proc_fd(inode); struct files_struct *files; - files = get_files_struct(task); - if (files) { - rcu_read_lock(); - if (fcheck_files(files, fd)) { + if (task) { + files = get_files_struct(task); + if (files) { + rcu_read_lock(); + if (fcheck_files(files, fd)) { + rcu_read_unlock(); + put_files_struct(files); + if (task_dumpable(task)) { + inode->i_uid = task->euid; + inode->i_gid = task->egid; + } else { + inode->i_uid = 0; + inode->i_gid = 0; + } + inode->i_mode &= ~(S_ISUID | S_ISGID); + security_task_to_inode(task, inode); + put_task_struct(task); + return 1; + } rcu_read_unlock(); put_files_struct(files); - if (task_dumpable(task)) { - inode->i_uid = task->euid; - inode->i_gid = task->egid; - } else { - inode->i_uid = 0; - inode->i_gid = 0; - } - security_task_to_inode(task, inode); - return 1; } - rcu_read_unlock(); - put_files_struct(files); + put_task_struct(task); } d_drop(dentry); return 0; } -static void pid_base_iput(struct dentry *dentry, struct inode *inode) -{ - struct task_struct *task = proc_task(inode); - spin_lock(&task->proc_lock); - if (task->proc_dentry == dentry) - task->proc_dentry = NULL; - spin_unlock(&task->proc_lock); - iput(inode); -} - static int pid_delete_dentry(struct dentry * dentry) { /* Is the task we represent dead? * If so, then don't put the dentry on the lru list, * kill it immediately. */ - return !pid_alive(proc_task(dentry->d_inode)); + return !proc_pid(dentry->d_inode)->tasks[PIDTYPE_PID].first; } static struct dentry_operations tid_fd_dentry_operations = @@ -1474,13 +1449,6 @@ static struct dentry_operations pid_dentry_operations = .d_delete = pid_delete_dentry, }; -static struct dentry_operations pid_base_dentry_operations = -{ - .d_revalidate = pid_revalidate, - .d_iput = pid_base_iput, - .d_delete = pid_delete_dentry, -}; - /* Lookups */ static unsigned name_to_int(struct dentry *dentry) @@ -1508,22 +1476,24 @@ out: /* SMP-safe */ static struct dentry *proc_lookupfd(struct inode * dir, struct dentry * dentry, struct nameidata *nd) { - struct task_struct *task = proc_task(dir); + struct task_struct *task = get_proc_task(dir); unsigned fd = name_to_int(dentry); + struct dentry *result = ERR_PTR(-ENOENT); struct file * file; struct files_struct * files; struct inode *inode; struct proc_inode *ei; + if (!task) + goto out_no_task; if (fd == ~0U) goto out; - if (!pid_alive(task)) - goto out; inode = proc_pid_make_inode(dir->i_sb, task, PROC_TID_FD_DIR+fd); if (!inode) goto out; ei = PROC_I(inode); + ei->fd = fd; files = get_files_struct(task); if (!files) goto out_unlock; @@ -1548,19 +1518,25 @@ static struct dentry *proc_lookupfd(struct inode * dir, struct dentry * dentry, ei->op.proc_get_link = proc_fd_link; dentry->d_op = &tid_fd_dentry_operations; d_add(dentry, inode); - return NULL; + /* Close the race of the process dying before we return the dentry */ + if (tid_fd_revalidate(dentry, NULL)) + result = NULL; +out: + put_task_struct(task); +out_no_task: + return result; out_unlock2: spin_unlock(&files->file_lock); put_files_struct(files); out_unlock: iput(inode); -out: - return ERR_PTR(-ENOENT); + goto out; } static int proc_task_readdir(struct file * filp, void * dirent, filldir_t filldir); static struct dentry *proc_task_lookup(struct inode *dir, struct dentry * dentry, struct nameidata *nd); +static int proc_task_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat); static struct file_operations proc_fd_operations = { .read = generic_read_dir, @@ -1577,12 +1553,13 @@ static struct file_operations proc_task_operations = { */ static struct inode_operations proc_fd_inode_operations = { .lookup = proc_lookupfd, - .permission = proc_permission, + .setattr = proc_setattr, }; static struct inode_operations proc_task_inode_operations = { .lookup = proc_task_lookup, - .permission = proc_task_permission, + .getattr = proc_task_getattr, + .setattr = proc_setattr, }; #ifdef CONFIG_SECURITY @@ -1592,12 +1569,17 @@ static ssize_t proc_pid_attr_read(struct file * file, char __user * buf, struct inode * inode = file->f_dentry->d_inode; unsigned long page; ssize_t length; - struct task_struct *task = proc_task(inode); + struct task_struct *task = get_proc_task(inode); + + length = -ESRCH; + if (!task) + goto out_no_task; if (count > PAGE_SIZE) count = PAGE_SIZE; + length = -ENOMEM; if (!(page = __get_free_page(GFP_KERNEL))) - return -ENOMEM; + goto out; length = security_getprocattr(task, (char*)file->f_dentry->d_name.name, @@ -1605,6 +1587,9 @@ static ssize_t proc_pid_attr_read(struct file * file, char __user * buf, if (length >= 0) length = simple_read_from_buffer(buf, count, ppos, (char *)page, length); free_page(page); +out: + put_task_struct(task); +out_no_task: return length; } @@ -1614,26 +1599,36 @@ static ssize_t proc_pid_attr_write(struct file * file, const char __user * buf, struct inode * inode = file->f_dentry->d_inode; char *page; ssize_t length; - struct task_struct *task = proc_task(inode); + struct task_struct *task = get_proc_task(inode); + length = -ESRCH; + if (!task) + goto out_no_task; if (count > PAGE_SIZE) count = PAGE_SIZE; - if (*ppos != 0) { - /* No partial writes. */ - return -EINVAL; - } + + /* No partial writes. */ + length = -EINVAL; + if (*ppos != 0) + goto out; + + length = -ENOMEM; page = (char*)__get_free_page(GFP_USER); if (!page) - return -ENOMEM; + goto out; + length = -EFAULT; if (copy_from_user(page, buf, count)) - goto out; + goto out_free; length = security_setprocattr(task, (char*)file->f_dentry->d_name.name, (void*)page, count); -out: +out_free: free_page((unsigned long) page); +out: + put_task_struct(task); +out_no_task: return length; } @@ -1648,24 +1643,22 @@ static struct file_operations proc_tgid_attr_operations; static struct inode_operations proc_tgid_attr_inode_operations; #endif -static int get_tid_list(int index, unsigned int *tids, struct inode *dir); - /* SMP-safe */ static struct dentry *proc_pident_lookup(struct inode *dir, struct dentry *dentry, struct pid_entry *ents) { struct inode *inode; - int error; - struct task_struct *task = proc_task(dir); + struct dentry *error; + struct task_struct *task = get_proc_task(dir); struct pid_entry *p; struct proc_inode *ei; - error = -ENOENT; + error = ERR_PTR(-ENOENT); inode = NULL; - if (!pid_alive(task)) - goto out; + if (!task) + goto out_no_task; for (p = ents; p->name; p++) { if (p->len != dentry->d_name.len) @@ -1676,7 +1669,7 @@ static struct dentry *proc_pident_lookup(struct inode *dir, if (!p->name) goto out; - error = -EINVAL; + error = ERR_PTR(-EINVAL); inode = proc_pid_make_inode(dir->i_sb, task, p->type); if (!inode) goto out; @@ -1689,7 +1682,7 @@ static struct dentry *proc_pident_lookup(struct inode *dir, */ switch(p->type) { case PROC_TGID_TASK: - inode->i_nlink = 2 + get_tid_list(2, NULL, dir); + inode->i_nlink = 2; inode->i_op = &proc_task_inode_operations; inode->i_fop = &proc_task_operations; break; @@ -1759,7 +1752,6 @@ static struct dentry *proc_pident_lookup(struct inode *dir, #endif case PROC_TID_MEM: case PROC_TGID_MEM: - inode->i_op = &proc_mem_inode_operations; inode->i_fop = &proc_mem_operations; break; #ifdef CONFIG_SECCOMP @@ -1801,6 +1793,10 @@ static struct dentry *proc_pident_lookup(struct inode *dir, case PROC_TGID_ATTR_EXEC: case PROC_TID_ATTR_FSCREATE: case PROC_TGID_ATTR_FSCREATE: + case PROC_TID_ATTR_KEYCREATE: + case PROC_TGID_ATTR_KEYCREATE: + case PROC_TID_ATTR_SOCKCREATE: + case PROC_TGID_ATTR_SOCKCREATE: inode->i_fop = &proc_pid_attr_operations; break; #endif @@ -1842,14 +1838,18 @@ static struct dentry *proc_pident_lookup(struct inode *dir, default: printk("procfs: impossible type (%d)",p->type); iput(inode); - return ERR_PTR(-EINVAL); + error = ERR_PTR(-EINVAL); + goto out; } dentry->d_op = &pid_dentry_operations; d_add(dentry, inode); - return NULL; - + /* Close the race of the process dying before we return the dentry */ + if (pid_revalidate(dentry, NULL)) + error = NULL; out: - return ERR_PTR(error); + put_task_struct(task); +out_no_task: + return error; } static struct dentry *proc_tgid_base_lookup(struct inode *dir, struct dentry *dentry, struct nameidata *nd){ @@ -1872,10 +1872,14 @@ static struct file_operations proc_tid_base_operations = { static struct inode_operations proc_tgid_base_inode_operations = { .lookup = proc_tgid_base_lookup, + .getattr = pid_getattr, + .setattr = proc_setattr, }; static struct inode_operations proc_tid_base_inode_operations = { .lookup = proc_tid_base_lookup, + .getattr = pid_getattr, + .setattr = proc_setattr, }; #ifdef CONFIG_SECURITY @@ -1917,10 +1921,14 @@ static struct dentry *proc_tid_attr_lookup(struct inode *dir, static struct inode_operations proc_tgid_attr_inode_operations = { .lookup = proc_tgid_attr_lookup, + .getattr = pid_getattr, + .setattr = proc_setattr, }; static struct inode_operations proc_tid_attr_inode_operations = { .lookup = proc_tid_attr_lookup, + .getattr = pid_getattr, + .setattr = proc_setattr, }; #endif @@ -1930,14 +1938,14 @@ static struct inode_operations proc_tid_attr_inode_operations = { static int proc_self_readlink(struct dentry *dentry, char __user *buffer, int buflen) { - char tmp[30]; + char tmp[PROC_NUMBUF]; sprintf(tmp, "%d", current->tgid); return vfs_readlink(dentry,buffer,buflen,tmp); } static void *proc_self_follow_link(struct dentry *dentry, struct nameidata *nd) { - char tmp[30]; + char tmp[PROC_NUMBUF]; sprintf(tmp, "%d", current->tgid); return ERR_PTR(vfs_follow_link(nd,tmp)); } @@ -1948,67 +1956,80 @@ static struct inode_operations proc_self_inode_operations = { }; /** - * proc_pid_unhash - Unhash /proc/@pid entry from the dcache. - * @p: task that should be flushed. + * proc_flush_task - Remove dcache entries for @task from the /proc dcache. + * + * @task: task that should be flushed. * - * Drops the /proc/@pid dcache entry from the hash chains. + * Looks in the dcache for + * /proc/@pid + * /proc/@tgid/task/@pid + * if either directory is present flushes it and all of it'ts children + * from the dcache. * - * Dropping /proc/@pid entries and detach_pid must be synchroneous, - * otherwise e.g. /proc/@pid/exe might point to the wrong executable, - * if the pid value is immediately reused. This is enforced by - * - caller must acquire spin_lock(p->proc_lock) - * - must be called before detach_pid() - * - proc_pid_lookup acquires proc_lock, and checks that - * the target is not dead by looking at the attach count - * of PIDTYPE_PID. + * It is safe and reasonable to cache /proc entries for a task until + * that task exits. After that they just clog up the dcache with + * useless entries, possibly causing useful dcache entries to be + * flushed instead. This routine is proved to flush those useless + * dcache entries at process exit time. + * + * NOTE: This routine is just an optimization so it does not guarantee + * that no dcache entries will exist at process exit time it + * just makes it very unlikely that any will persist. */ - -struct dentry *proc_pid_unhash(struct task_struct *p) +void proc_flush_task(struct task_struct *task) { - struct dentry *proc_dentry; + struct dentry *dentry, *leader, *dir; + char buf[PROC_NUMBUF]; + struct qstr name; + + name.name = buf; + name.len = snprintf(buf, sizeof(buf), "%d", task->pid); + dentry = d_hash_and_lookup(proc_mnt->mnt_root, &name); + if (dentry) { + shrink_dcache_parent(dentry); + d_drop(dentry); + dput(dentry); + } - proc_dentry = p->proc_dentry; - if (proc_dentry != NULL) { + if (thread_group_leader(task)) + goto out; - spin_lock(&dcache_lock); - spin_lock(&proc_dentry->d_lock); - if (!d_unhashed(proc_dentry)) { - dget_locked(proc_dentry); - __d_drop(proc_dentry); - spin_unlock(&proc_dentry->d_lock); - } else { - spin_unlock(&proc_dentry->d_lock); - proc_dentry = NULL; - } - spin_unlock(&dcache_lock); - } - return proc_dentry; -} + name.name = buf; + name.len = snprintf(buf, sizeof(buf), "%d", task->tgid); + leader = d_hash_and_lookup(proc_mnt->mnt_root, &name); + if (!leader) + goto out; -/** - * proc_pid_flush - recover memory used by stale /proc/@pid/x entries - * @proc_dentry: directoy to prune. - * - * Shrink the /proc directory that was used by the just killed thread. - */ - -void proc_pid_flush(struct dentry *proc_dentry) -{ - might_sleep(); - if(proc_dentry != NULL) { - shrink_dcache_parent(proc_dentry); - dput(proc_dentry); + name.name = "task"; + name.len = strlen(name.name); + dir = d_hash_and_lookup(leader, &name); + if (!dir) + goto out_put_leader; + + name.name = buf; + name.len = snprintf(buf, sizeof(buf), "%d", task->pid); + dentry = d_hash_and_lookup(dir, &name); + if (dentry) { + shrink_dcache_parent(dentry); + d_drop(dentry); + dput(dentry); } + + dput(dir); +out_put_leader: + dput(leader); +out: + return; } /* SMP-safe */ struct dentry *proc_pid_lookup(struct inode *dir, struct dentry * dentry, struct nameidata *nd) { + struct dentry *result = ERR_PTR(-ENOENT); struct task_struct *task; struct inode *inode; struct proc_inode *ei; unsigned tgid; - int died; if (dentry->d_name.len == 4 && !memcmp(dentry->d_name.name,"self",4)) { inode = new_inode(dir->i_sb); @@ -2029,21 +2050,18 @@ struct dentry *proc_pid_lookup(struct inode *dir, struct dentry * dentry, struct if (tgid == ~0U) goto out; - read_lock(&tasklist_lock); + rcu_read_lock(); task = find_task_by_pid(tgid); if (task) get_task_struct(task); - read_unlock(&tasklist_lock); + rcu_read_unlock(); if (!task) goto out; inode = proc_pid_make_inode(dir->i_sb, task, PROC_TGID_INO); + if (!inode) + goto out_put_task; - - if (!inode) { - put_task_struct(task); - goto out; - } inode->i_mode = S_IFDIR|S_IRUGO|S_IXUGO; inode->i_op = &proc_tgid_base_inode_operations; inode->i_fop = &proc_tgid_base_operations; @@ -2054,45 +2072,40 @@ struct dentry *proc_pid_lookup(struct inode *dir, struct dentry * dentry, struct inode->i_nlink = 4; #endif - dentry->d_op = &pid_base_dentry_operations; + dentry->d_op = &pid_dentry_operations; - died = 0; d_add(dentry, inode); - spin_lock(&task->proc_lock); - task->proc_dentry = dentry; - if (!pid_alive(task)) { - dentry = proc_pid_unhash(task); - died = 1; - } - spin_unlock(&task->proc_lock); + /* Close the race of the process dying before we return the dentry */ + if (pid_revalidate(dentry, NULL)) + result = NULL; +out_put_task: put_task_struct(task); - if (died) { - proc_pid_flush(dentry); - goto out; - } - return NULL; out: - return ERR_PTR(-ENOENT); + return result; } /* SMP-safe */ static struct dentry *proc_task_lookup(struct inode *dir, struct dentry * dentry, struct nameidata *nd) { + struct dentry *result = ERR_PTR(-ENOENT); struct task_struct *task; - struct task_struct *leader = proc_task(dir); + struct task_struct *leader = get_proc_task(dir); struct inode *inode; unsigned tid; + if (!leader) + goto out_no_task; + tid = name_to_int(dentry); if (tid == ~0U) goto out; - read_lock(&tasklist_lock); + rcu_read_lock(); task = find_task_by_pid(tid); if (task) get_task_struct(task); - read_unlock(&tasklist_lock); + rcu_read_unlock(); if (!task) goto out; if (leader->tgid != task->tgid) @@ -2113,101 +2126,95 @@ static struct dentry *proc_task_lookup(struct inode *dir, struct dentry * dentry inode->i_nlink = 3; #endif - dentry->d_op = &pid_base_dentry_operations; + dentry->d_op = &pid_dentry_operations; d_add(dentry, inode); + /* Close the race of the process dying before we return the dentry */ + if (pid_revalidate(dentry, NULL)) + result = NULL; - put_task_struct(task); - return NULL; out_drop_task: put_task_struct(task); out: - return ERR_PTR(-ENOENT); + put_task_struct(leader); +out_no_task: + return result; } -#define PROC_NUMBUF 10 -#define PROC_MAXPIDS 20 - /* - * Get a few tgid's to return for filldir - we need to hold the - * tasklist lock while doing this, and we must release it before - * we actually do the filldir itself, so we use a temp buffer.. + * Find the first tgid to return to user space. + * + * Usually this is just whatever follows &init_task, but if the users + * buffer was too small to hold the full list or there was a seek into + * the middle of the directory we have more work to do. + * + * In the case of a short read we start with find_task_by_pid. + * + * In the case of a seek we start with &init_task and walk nr + * threads past it. */ -static int get_tgid_list(int index, unsigned long version, unsigned int *tgids) -{ - struct task_struct *p; - int nr_tgids = 0; - - index--; - read_lock(&tasklist_lock); - p = NULL; - if (version) { - p = find_task_by_pid(version); - if (p && !thread_group_leader(p)) - p = NULL; +static struct task_struct *first_tgid(int tgid, unsigned int nr) +{ + struct task_struct *pos; + rcu_read_lock(); + if (tgid && nr) { + pos = find_task_by_pid(tgid); + if (pos && thread_group_leader(pos)) + goto found; } + /* If nr exceeds the number of processes get out quickly */ + pos = NULL; + if (nr && nr >= nr_processes()) + goto done; - if (p) - index = 0; - else - p = next_task(&init_task); - - for ( ; p != &init_task; p = next_task(p)) { - int tgid = p->pid; - if (!pid_alive(p)) - continue; - if (--index >= 0) - continue; - tgids[nr_tgids] = tgid; - nr_tgids++; - if (nr_tgids >= PROC_MAXPIDS) - break; + /* If we haven't found our starting place yet start with + * the init_task and walk nr tasks forward. + */ + for (pos = next_task(&init_task); nr > 0; --nr) { + pos = next_task(pos); + if (pos == &init_task) { + pos = NULL; + goto done; + } } - read_unlock(&tasklist_lock); - return nr_tgids; +found: + get_task_struct(pos); +done: + rcu_read_unlock(); + return pos; } /* - * Get a few tid's to return for filldir - we need to hold the - * tasklist lock while doing this, and we must release it before - * we actually do the filldir itself, so we use a temp buffer.. + * Find the next task in the task list. + * Return NULL if we loop or there is any error. + * + * The reference to the input task_struct is released. */ -static int get_tid_list(int index, unsigned int *tids, struct inode *dir) -{ - struct task_struct *leader_task = proc_task(dir); - struct task_struct *task = leader_task; - int nr_tids = 0; - - index -= 2; - read_lock(&tasklist_lock); - /* - * The starting point task (leader_task) might be an already - * unlinked task, which cannot be used to access the task-list - * via next_thread(). - */ - if (pid_alive(task)) do { - int tid = task->pid; - - if (--index >= 0) - continue; - if (tids != NULL) - tids[nr_tids] = tid; - nr_tids++; - if (nr_tids >= PROC_MAXPIDS) - break; - } while ((task = next_thread(task)) != leader_task); - read_unlock(&tasklist_lock); - return nr_tids; +static struct task_struct *next_tgid(struct task_struct *start) +{ + struct task_struct *pos; + rcu_read_lock(); + pos = start; + if (pid_alive(start)) + pos = next_task(start); + if (pid_alive(pos) && (pos != &init_task)) { + get_task_struct(pos); + goto done; + } + pos = NULL; +done: + rcu_read_unlock(); + put_task_struct(start); + return pos; } /* for the /proc/ directory itself, after non-process stuff has been done */ int proc_pid_readdir(struct file * filp, void * dirent, filldir_t filldir) { - unsigned int tgid_array[PROC_MAXPIDS]; char buf[PROC_NUMBUF]; unsigned int nr = filp->f_pos - FIRST_PROCESS_ENTRY; - unsigned int nr_tgids, i; - int next_tgid; + struct task_struct *task; + int tgid; if (!nr) { ino_t ino = fake_ino(0,PROC_TGID_INO); @@ -2216,63 +2223,116 @@ int proc_pid_readdir(struct file * filp, void * dirent, filldir_t filldir) filp->f_pos++; nr++; } + nr -= 1; /* f_version caches the tgid value that the last readdir call couldn't * return. lseek aka telldir automagically resets f_version to 0. */ - next_tgid = filp->f_version; + tgid = filp->f_version; filp->f_version = 0; - for (;;) { - nr_tgids = get_tgid_list(nr, next_tgid, tgid_array); - if (!nr_tgids) { - /* no more entries ! */ + for (task = first_tgid(tgid, nr); + task; + task = next_tgid(task), filp->f_pos++) { + int len; + ino_t ino; + tgid = task->pid; + len = snprintf(buf, sizeof(buf), "%d", tgid); + ino = fake_ino(tgid, PROC_TGID_INO); + if (filldir(dirent, buf, len, filp->f_pos, ino, DT_DIR) < 0) { + /* returning this tgid failed, save it as the first + * pid for the next readir call */ + filp->f_version = tgid; + put_task_struct(task); break; } - next_tgid = 0; + } + return 0; +} - /* do not use the last found pid, reserve it for next_tgid */ - if (nr_tgids == PROC_MAXPIDS) { - nr_tgids--; - next_tgid = tgid_array[nr_tgids]; - } +/* + * Find the first tid of a thread group to return to user space. + * + * Usually this is just the thread group leader, but if the users + * buffer was too small or there was a seek into the middle of the + * directory we have more work todo. + * + * In the case of a short read we start with find_task_by_pid. + * + * In the case of a seek we start with the leader and walk nr + * threads past it. + */ +static struct task_struct *first_tid(struct task_struct *leader, + int tid, int nr) +{ + struct task_struct *pos; - for (i=0;i<nr_tgids;i++) { - int tgid = tgid_array[i]; - ino_t ino = fake_ino(tgid,PROC_TGID_INO); - unsigned long j = PROC_NUMBUF; + rcu_read_lock(); + /* Attempt to start with the pid of a thread */ + if (tid && (nr > 0)) { + pos = find_task_by_pid(tid); + if (pos && (pos->group_leader == leader)) + goto found; + } - do - buf[--j] = '0' + (tgid % 10); - while ((tgid /= 10) != 0); + /* If nr exceeds the number of threads there is nothing todo */ + pos = NULL; + if (nr && nr >= get_nr_threads(leader)) + goto out; - if (filldir(dirent, buf+j, PROC_NUMBUF-j, filp->f_pos, ino, DT_DIR) < 0) { - /* returning this tgid failed, save it as the first - * pid for the next readir call */ - filp->f_version = tgid_array[i]; - goto out; - } - filp->f_pos++; - nr++; + /* If we haven't found our starting place yet start + * with the leader and walk nr threads forward. + */ + for (pos = leader; nr > 0; --nr) { + pos = next_thread(pos); + if (pos == leader) { + pos = NULL; + goto out; } } +found: + get_task_struct(pos); out: - return 0; + rcu_read_unlock(); + return pos; +} + +/* + * Find the next thread in the thread list. + * Return NULL if there is an error or no next thread. + * + * The reference to the input task_struct is released. + */ +static struct task_struct *next_tid(struct task_struct *start) +{ + struct task_struct *pos = NULL; + rcu_read_lock(); + if (pid_alive(start)) { + pos = next_thread(start); + if (thread_group_leader(pos)) + pos = NULL; + else + get_task_struct(pos); + } + rcu_read_unlock(); + put_task_struct(start); + return pos; } /* for the /proc/TGID/task/ directories */ static int proc_task_readdir(struct file * filp, void * dirent, filldir_t filldir) { - unsigned int tid_array[PROC_MAXPIDS]; char buf[PROC_NUMBUF]; - unsigned int nr_tids, i; struct dentry *dentry = filp->f_dentry; struct inode *inode = dentry->d_inode; + struct task_struct *leader = get_proc_task(inode); + struct task_struct *task; int retval = -ENOENT; ino_t ino; + int tid; unsigned long pos = filp->f_pos; /* avoiding "long long" filp->f_pos */ - if (!pid_alive(proc_task(inode))) - goto out; + if (!leader) + goto out_no_task; retval = 0; switch (pos) { @@ -2290,24 +2350,45 @@ static int proc_task_readdir(struct file * filp, void * dirent, filldir_t filldi /* fall through */ } - nr_tids = get_tid_list(pos, tid_array, inode); - inode->i_nlink = pos + nr_tids; - - for (i = 0; i < nr_tids; i++) { - unsigned long j = PROC_NUMBUF; - int tid = tid_array[i]; - - ino = fake_ino(tid,PROC_TID_INO); - - do - buf[--j] = '0' + (tid % 10); - while ((tid /= 10) != 0); - - if (filldir(dirent, buf+j, PROC_NUMBUF-j, pos, ino, DT_DIR) < 0) + /* f_version caches the tgid value that the last readdir call couldn't + * return. lseek aka telldir automagically resets f_version to 0. + */ + tid = filp->f_version; + filp->f_version = 0; + for (task = first_tid(leader, tid, pos - 2); + task; + task = next_tid(task), pos++) { + int len; + tid = task->pid; + len = snprintf(buf, sizeof(buf), "%d", tid); + ino = fake_ino(tid, PROC_TID_INO); + if (filldir(dirent, buf, len, pos, ino, DT_DIR < 0)) { + /* returning this tgid failed, save it as the first + * pid for the next readir call */ + filp->f_version = tid; + put_task_struct(task); break; - pos++; + } } out: filp->f_pos = pos; + put_task_struct(leader); +out_no_task: return retval; } + +static int proc_task_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat) +{ + struct inode *inode = dentry->d_inode; + struct task_struct *p = get_proc_task(inode); + generic_fillattr(inode, stat); + + if (p) { + rcu_read_lock(); + stat->nlink += get_nr_threads(p); + rcu_read_unlock(); + put_task_struct(p); + } + + return 0; +} diff --git a/fs/proc/inode.c b/fs/proc/inode.c index 722b9c46311..49dfb2ab783 100644 --- a/fs/proc/inode.c +++ b/fs/proc/inode.c @@ -58,14 +58,11 @@ static void de_put(struct proc_dir_entry *de) static void proc_delete_inode(struct inode *inode) { struct proc_dir_entry *de; - struct task_struct *tsk; truncate_inode_pages(&inode->i_data, 0); - /* Let go of any associated process */ - tsk = PROC_I(inode)->task; - if (tsk) - put_task_struct(tsk); + /* Stop tracking associated processes */ + put_pid(PROC_I(inode)->pid); /* Let go of any associated proc directory entry */ de = PROC_I(inode)->pde; @@ -94,8 +91,8 @@ static struct inode *proc_alloc_inode(struct super_block *sb) ei = (struct proc_inode *)kmem_cache_alloc(proc_inode_cachep, SLAB_KERNEL); if (!ei) return NULL; - ei->task = NULL; - ei->type = 0; + ei->pid = NULL; + ei->fd = 0; ei->op.proc_get_link = NULL; ei->pde = NULL; inode = &ei->vfs_inode; @@ -195,7 +192,7 @@ int proc_fill_super(struct super_block *s, void *data, int silent) { struct inode * root_inode; - s->s_flags |= MS_NODIRATIME; + s->s_flags |= MS_NODIRATIME | MS_NOSUID | MS_NOEXEC; s->s_blocksize = 1024; s->s_blocksize_bits = 10; s->s_magic = PROC_SUPER_MAGIC; diff --git a/fs/proc/internal.h b/fs/proc/internal.h index 0502f17b860..987c773dbb2 100644 --- a/fs/proc/internal.h +++ b/fs/proc/internal.h @@ -28,6 +28,7 @@ do { \ (vmi)->largest_chunk = 0; \ } while(0) +extern int nommu_vma_show(struct seq_file *, struct vm_area_struct *); #endif extern void create_seq_entry(char *name, mode_t mode, const struct file_operations *f); @@ -37,16 +38,30 @@ extern int proc_tgid_stat(struct task_struct *, char *); extern int proc_pid_status(struct task_struct *, char *); extern int proc_pid_statm(struct task_struct *, char *); +extern struct file_operations proc_maps_operations; +extern struct file_operations proc_numa_maps_operations; +extern struct file_operations proc_smaps_operations; + +extern struct file_operations proc_maps_operations; +extern struct file_operations proc_numa_maps_operations; +extern struct file_operations proc_smaps_operations; + + void free_proc_entry(struct proc_dir_entry *de); int proc_init_inodecache(void); -static inline struct task_struct *proc_task(struct inode *inode) +static inline struct pid *proc_pid(struct inode *inode) +{ + return PROC_I(inode)->pid; +} + +static inline struct task_struct *get_proc_task(struct inode *inode) { - return PROC_I(inode)->task; + return get_pid_task(proc_pid(inode), PIDTYPE_PID); } -static inline int proc_type(struct inode *inode) +static inline int proc_fd(struct inode *inode) { - return PROC_I(inode)->type; + return PROC_I(inode)->fd; } diff --git a/fs/proc/kcore.c b/fs/proc/kcore.c index 17f6e8fa139..1294eda4aca 100644 --- a/fs/proc/kcore.c +++ b/fs/proc/kcore.c @@ -9,7 +9,6 @@ * Safe accesses to vmalloc/direct-mapped discontiguous areas, Kanoj Sarcar <kanoj@sgi.com> */ -#include <linux/config.h> #include <linux/mm.h> #include <linux/proc_fs.h> #include <linux/user.h> @@ -43,8 +42,6 @@ const struct file_operations proc_kcore_operations = { #define kc_offset_to_vaddr(o) ((o) + PAGE_OFFSET) #endif -#define roundup(x, y) ((((x)+((y)-1))/(y))*(y)) - /* An ELF note in memory */ struct memelfnote { @@ -103,7 +100,7 @@ static int notesize(struct memelfnote *en) int sz; sz = sizeof(struct elf_note); - sz += roundup(strlen(en->name), 4); + sz += roundup((strlen(en->name) + 1), 4); sz += roundup(en->datasz, 4); return sz; @@ -119,7 +116,7 @@ static char *storenote(struct memelfnote *men, char *bufp) #define DUMP_WRITE(addr,nr) do { memcpy(bufp,addr,nr); bufp += nr; } while(0) - en.n_namesz = strlen(men->name); + en.n_namesz = strlen(men->name) + 1; en.n_descsz = men->datasz; en.n_type = men->type; @@ -282,12 +279,11 @@ read_kcore(struct file *file, char __user *buffer, size_t buflen, loff_t *fpos) tsz = elf_buflen - *fpos; if (buflen < tsz) tsz = buflen; - elf_buf = kmalloc(elf_buflen, GFP_ATOMIC); + elf_buf = kzalloc(elf_buflen, GFP_ATOMIC); if (!elf_buf) { read_unlock(&kclist_lock); return -ENOMEM; } - memset(elf_buf, 0, elf_buflen); elf_kcore_store_hdr(elf_buf, nphdr, elf_buflen); read_unlock(&kclist_lock); if (copy_to_user(buffer, elf_buf + *fpos, tsz)) { @@ -333,10 +329,9 @@ read_kcore(struct file *file, char __user *buffer, size_t buflen, loff_t *fpos) unsigned long curstart = start; unsigned long cursize = tsz; - elf_buf = kmalloc(tsz, GFP_KERNEL); + elf_buf = kzalloc(tsz, GFP_KERNEL); if (!elf_buf) return -ENOMEM; - memset(elf_buf, 0, tsz); read_lock(&vmlist_lock); for (m=vmlist; m && cursize; m=m->next) { @@ -385,7 +380,7 @@ read_kcore(struct file *file, char __user *buffer, size_t buflen, loff_t *fpos) */ if (n) { if (clear_user(buffer + tsz - n, - tsz - n)) + n)) return -EFAULT; } } else { diff --git a/fs/proc/nommu.c b/fs/proc/nommu.c index cff10ab1af6..d7dbdf9e0f4 100644 --- a/fs/proc/nommu.c +++ b/fs/proc/nommu.c @@ -33,19 +33,15 @@ #include "internal.h" /* - * display a list of all the VMAs the kernel knows about - * - nommu kernals have a single flat list + * display a single VMA to a sequenced file */ -static int nommu_vma_list_show(struct seq_file *m, void *v) +int nommu_vma_show(struct seq_file *m, struct vm_area_struct *vma) { - struct vm_area_struct *vma; unsigned long ino = 0; struct file *file; dev_t dev = 0; int flags, len; - vma = rb_entry((struct rb_node *) v, struct vm_area_struct, vm_rb); - flags = vma->vm_flags; file = vma->vm_file; @@ -78,6 +74,18 @@ static int nommu_vma_list_show(struct seq_file *m, void *v) return 0; } +/* + * display a list of all the VMAs the kernel knows about + * - nommu kernals have a single flat list + */ +static int nommu_vma_list_show(struct seq_file *m, void *v) +{ + struct vm_area_struct *vma; + + vma = rb_entry((struct rb_node *) v, struct vm_area_struct, vm_rb); + return nommu_vma_show(m, vma); +} + static void *nommu_vma_list_start(struct seq_file *m, loff_t *_pos) { struct rb_node *_rb; diff --git a/fs/proc/proc_misc.c b/fs/proc/proc_misc.c index 5c10ea15742..5bbd6089605 100644 --- a/fs/proc/proc_misc.c +++ b/fs/proc/proc_misc.c @@ -26,7 +26,6 @@ #include <linux/mman.h> #include <linux/proc_fs.h> #include <linux/ioport.h> -#include <linux/config.h> #include <linux/mm.h> #include <linux/mmzone.h> #include <linux/pagemap.h> @@ -120,7 +119,6 @@ static int meminfo_read_proc(char *page, char **start, off_t off, { struct sysinfo i; int len; - struct page_state ps; unsigned long inactive; unsigned long active; unsigned long free; @@ -129,7 +127,6 @@ static int meminfo_read_proc(char *page, char **start, off_t off, struct vmalloc_info vmi; long cached; - get_page_state(&ps); get_zone_counts(&active, &inactive, &free); /* @@ -142,7 +139,8 @@ static int meminfo_read_proc(char *page, char **start, off_t off, allowed = ((totalram_pages - hugetlb_total_pages()) * sysctl_overcommit_ratio / 100) + total_swap_pages; - cached = get_page_cache_size() - total_swapcache_pages - i.bufferram; + cached = global_page_state(NR_FILE_PAGES) - + total_swapcache_pages - i.bufferram; if (cached < 0) cached = 0; @@ -159,19 +157,26 @@ static int meminfo_read_proc(char *page, char **start, off_t off, "SwapCached: %8lu kB\n" "Active: %8lu kB\n" "Inactive: %8lu kB\n" +#ifdef CONFIG_HIGHMEM "HighTotal: %8lu kB\n" "HighFree: %8lu kB\n" "LowTotal: %8lu kB\n" "LowFree: %8lu kB\n" +#endif "SwapTotal: %8lu kB\n" "SwapFree: %8lu kB\n" "Dirty: %8lu kB\n" "Writeback: %8lu kB\n" + "AnonPages: %8lu kB\n" "Mapped: %8lu kB\n" "Slab: %8lu kB\n" + "SReclaimable: %8lu kB\n" + "SUnreclaim: %8lu kB\n" + "PageTables: %8lu kB\n" + "NFS_Unstable: %8lu kB\n" + "Bounce: %8lu kB\n" "CommitLimit: %8lu kB\n" "Committed_AS: %8lu kB\n" - "PageTables: %8lu kB\n" "VmallocTotal: %8lu kB\n" "VmallocUsed: %8lu kB\n" "VmallocChunk: %8lu kB\n", @@ -182,19 +187,27 @@ static int meminfo_read_proc(char *page, char **start, off_t off, K(total_swapcache_pages), K(active), K(inactive), +#ifdef CONFIG_HIGHMEM K(i.totalhigh), K(i.freehigh), K(i.totalram-i.totalhigh), K(i.freeram-i.freehigh), +#endif K(i.totalswap), K(i.freeswap), - K(ps.nr_dirty), - K(ps.nr_writeback), - K(ps.nr_mapped), - K(ps.nr_slab), + K(global_page_state(NR_FILE_DIRTY)), + K(global_page_state(NR_WRITEBACK)), + K(global_page_state(NR_ANON_PAGES)), + K(global_page_state(NR_FILE_MAPPED)), + K(global_page_state(NR_SLAB_RECLAIMABLE) + + global_page_state(NR_SLAB_UNRECLAIMABLE)), + K(global_page_state(NR_SLAB_RECLAIMABLE)), + K(global_page_state(NR_SLAB_UNRECLAIMABLE)), + K(global_page_state(NR_PAGETABLE)), + K(global_page_state(NR_UNSTABLE_NFS)), + K(global_page_state(NR_BOUNCE)), K(allowed), K(committed), - K(ps.nr_page_table_pages), (unsigned long)VMALLOC_TOTAL >> 10, vmi.used >> 10, vmi.largest_chunk >> 10 diff --git a/fs/proc/root.c b/fs/proc/root.c index 9995356ce73..8901c65caca 100644 --- a/fs/proc/root.c +++ b/fs/proc/root.c @@ -12,7 +12,6 @@ #include <linux/time.h> #include <linux/proc_fs.h> #include <linux/stat.h> -#include <linux/config.h> #include <linux/init.h> #include <linux/module.h> #include <linux/bitops.h> diff --git a/fs/proc/task_mmu.c b/fs/proc/task_mmu.c index 91b7c15ab37..6b769afac55 100644 --- a/fs/proc/task_mmu.c +++ b/fs/proc/task_mmu.c @@ -75,9 +75,13 @@ int proc_exe_link(struct inode *inode, struct dentry **dentry, struct vfsmount * { struct vm_area_struct * vma; int result = -ENOENT; - struct task_struct *task = proc_task(inode); - struct mm_struct * mm = get_task_mm(task); + struct task_struct *task = get_proc_task(inode); + struct mm_struct * mm = NULL; + if (task) { + mm = get_task_mm(task); + put_task_struct(task); + } if (!mm) goto out; down_read(&mm->mmap_sem); @@ -120,7 +124,8 @@ struct mem_size_stats static int show_map_internal(struct seq_file *m, void *v, struct mem_size_stats *mss) { - struct task_struct *task = m->private; + struct proc_maps_private *priv = m->private; + struct task_struct *task = priv->task; struct vm_area_struct *vma = v; struct mm_struct *mm = vma->vm_mm; struct file *file = vma->vm_file; @@ -153,22 +158,23 @@ static int show_map_internal(struct seq_file *m, void *v, struct mem_size_stats pad_len_spaces(m, len); seq_path(m, file->f_vfsmnt, file->f_dentry, "\n"); } else { - if (mm) { - if (vma->vm_start <= mm->start_brk && + const char *name = arch_vma_name(vma); + if (!name) { + if (mm) { + if (vma->vm_start <= mm->start_brk && vma->vm_end >= mm->brk) { - pad_len_spaces(m, len); - seq_puts(m, "[heap]"); - } else { - if (vma->vm_start <= mm->start_stack && - vma->vm_end >= mm->start_stack) { - - pad_len_spaces(m, len); - seq_puts(m, "[stack]"); + name = "[heap]"; + } else if (vma->vm_start <= mm->start_stack && + vma->vm_end >= mm->start_stack) { + name = "[stack]"; } + } else { + name = "[vdso]"; } - } else { + } + if (name) { pad_len_spaces(m, len); - seq_puts(m, "[vdso]"); + seq_puts(m, name); } } seq_putc(m, '\n'); @@ -295,12 +301,16 @@ static int show_smap(struct seq_file *m, void *v) static void *m_start(struct seq_file *m, loff_t *pos) { - struct task_struct *task = m->private; + struct proc_maps_private *priv = m->private; unsigned long last_addr = m->version; struct mm_struct *mm; - struct vm_area_struct *vma, *tail_vma; + struct vm_area_struct *vma, *tail_vma = NULL; loff_t l = *pos; + /* Clear the per syscall fields in priv */ + priv->task = NULL; + priv->tail_vma = NULL; + /* * We remember last_addr rather than next_addr to hit with * mmap_cache most of the time. We have zero last_addr at @@ -311,11 +321,15 @@ static void *m_start(struct seq_file *m, loff_t *pos) if (last_addr == -1UL) return NULL; - mm = get_task_mm(task); + priv->task = get_pid_task(priv->pid, PIDTYPE_PID); + if (!priv->task) + return NULL; + + mm = get_task_mm(priv->task); if (!mm) return NULL; - tail_vma = get_gate_vma(task); + priv->tail_vma = tail_vma = get_gate_vma(priv->task); down_read(&mm->mmap_sem); /* Start with last addr hint */ @@ -350,11 +364,9 @@ out: return tail_vma; } -static void m_stop(struct seq_file *m, void *v) +static void vma_stop(struct proc_maps_private *priv, struct vm_area_struct *vma) { - struct task_struct *task = m->private; - struct vm_area_struct *vma = v; - if (vma && vma != get_gate_vma(task)) { + if (vma && vma != priv->tail_vma) { struct mm_struct *mm = vma->vm_mm; up_read(&mm->mmap_sem); mmput(mm); @@ -363,38 +375,103 @@ static void m_stop(struct seq_file *m, void *v) static void *m_next(struct seq_file *m, void *v, loff_t *pos) { - struct task_struct *task = m->private; + struct proc_maps_private *priv = m->private; struct vm_area_struct *vma = v; - struct vm_area_struct *tail_vma = get_gate_vma(task); + struct vm_area_struct *tail_vma = priv->tail_vma; (*pos)++; if (vma && (vma != tail_vma) && vma->vm_next) return vma->vm_next; - m_stop(m, v); + vma_stop(priv, vma); return (vma != tail_vma)? tail_vma: NULL; } -struct seq_operations proc_pid_maps_op = { +static void m_stop(struct seq_file *m, void *v) +{ + struct proc_maps_private *priv = m->private; + struct vm_area_struct *vma = v; + + vma_stop(priv, vma); + if (priv->task) + put_task_struct(priv->task); +} + +static struct seq_operations proc_pid_maps_op = { .start = m_start, .next = m_next, .stop = m_stop, .show = show_map }; -struct seq_operations proc_pid_smaps_op = { +static struct seq_operations proc_pid_smaps_op = { .start = m_start, .next = m_next, .stop = m_stop, .show = show_smap }; +static int do_maps_open(struct inode *inode, struct file *file, + struct seq_operations *ops) +{ + struct proc_maps_private *priv; + int ret = -ENOMEM; + priv = kzalloc(sizeof(*priv), GFP_KERNEL); + if (priv) { + priv->pid = proc_pid(inode); + ret = seq_open(file, ops); + if (!ret) { + struct seq_file *m = file->private_data; + m->private = priv; + } else { + kfree(priv); + } + } + return ret; +} + +static int maps_open(struct inode *inode, struct file *file) +{ + return do_maps_open(inode, file, &proc_pid_maps_op); +} + +struct file_operations proc_maps_operations = { + .open = maps_open, + .read = seq_read, + .llseek = seq_lseek, + .release = seq_release_private, +}; + #ifdef CONFIG_NUMA extern int show_numa_map(struct seq_file *m, void *v); -struct seq_operations proc_pid_numa_maps_op = { +static struct seq_operations proc_pid_numa_maps_op = { .start = m_start, .next = m_next, .stop = m_stop, .show = show_numa_map }; + +static int numa_maps_open(struct inode *inode, struct file *file) +{ + return do_maps_open(inode, file, &proc_pid_numa_maps_op); +} + +struct file_operations proc_numa_maps_operations = { + .open = numa_maps_open, + .read = seq_read, + .llseek = seq_lseek, + .release = seq_release_private, +}; #endif + +static int smaps_open(struct inode *inode, struct file *file) +{ + return do_maps_open(inode, file, &proc_pid_smaps_op); +} + +struct file_operations proc_smaps_operations = { + .open = smaps_open, + .read = seq_read, + .llseek = seq_lseek, + .release = seq_release_private, +}; diff --git a/fs/proc/task_nommu.c b/fs/proc/task_nommu.c index 8f68827ed10..091aa8e48e0 100644 --- a/fs/proc/task_nommu.c +++ b/fs/proc/task_nommu.c @@ -107,7 +107,7 @@ int proc_exe_link(struct inode *inode, struct dentry **dentry, struct vfsmount * { struct vm_list_struct *vml; struct vm_area_struct *vma; - struct task_struct *task = proc_task(inode); + struct task_struct *task = get_proc_task(inode); struct mm_struct *mm = get_task_mm(task); int result = -ENOENT; @@ -138,27 +138,92 @@ out: } /* - * Albert D. Cahalan suggested to fake entries for the traditional - * sections here. This might be worth investigating. + * display mapping lines for a particular process's /proc/pid/maps */ -static int show_map(struct seq_file *m, void *v) +static int show_map(struct seq_file *m, void *_vml) { - return 0; + struct vm_list_struct *vml = _vml; + return nommu_vma_show(m, vml->vma); } + static void *m_start(struct seq_file *m, loff_t *pos) { + struct proc_maps_private *priv = m->private; + struct vm_list_struct *vml; + struct mm_struct *mm; + loff_t n = *pos; + + /* pin the task and mm whilst we play with them */ + priv->task = get_pid_task(priv->pid, PIDTYPE_PID); + if (!priv->task) + return NULL; + + mm = get_task_mm(priv->task); + if (!mm) { + put_task_struct(priv->task); + priv->task = NULL; + return NULL; + } + + down_read(&mm->mmap_sem); + + /* start from the Nth VMA */ + for (vml = mm->context.vmlist; vml; vml = vml->next) + if (n-- == 0) + return vml; return NULL; } -static void m_stop(struct seq_file *m, void *v) + +static void m_stop(struct seq_file *m, void *_vml) { + struct proc_maps_private *priv = m->private; + + if (priv->task) { + struct mm_struct *mm = priv->task->mm; + up_read(&mm->mmap_sem); + mmput(mm); + put_task_struct(priv->task); + } } -static void *m_next(struct seq_file *m, void *v, loff_t *pos) + +static void *m_next(struct seq_file *m, void *_vml, loff_t *pos) { - return NULL; + struct vm_list_struct *vml = _vml; + + (*pos)++; + return vml ? vml->next : NULL; } -struct seq_operations proc_pid_maps_op = { + +static struct seq_operations proc_pid_maps_ops = { .start = m_start, .next = m_next, .stop = m_stop, .show = show_map }; + +static int maps_open(struct inode *inode, struct file *file) +{ + struct proc_maps_private *priv; + int ret = -ENOMEM; + + priv = kzalloc(sizeof(*priv), GFP_KERNEL); + if (priv) { + priv->pid = proc_pid(inode); + ret = seq_open(file, &proc_pid_maps_ops); + if (!ret) { + struct seq_file *m = file->private_data; + m->private = priv; + } else { + kfree(priv); + } + } + return ret; +} + +struct file_operations proc_maps_operations = { + .open = maps_open, + .read = seq_read, + .llseek = seq_lseek, + .release = seq_release_private, +}; + diff --git a/fs/proc/vmcore.c b/fs/proc/vmcore.c index 20d4b2237fc..d96050728c4 100644 --- a/fs/proc/vmcore.c +++ b/fs/proc/vmcore.c @@ -7,7 +7,6 @@ * */ -#include <linux/config.h> #include <linux/mm.h> #include <linux/proc_fs.h> #include <linux/user.h> diff --git a/fs/qnx4/bitmap.c b/fs/qnx4/bitmap.c index 46efbf52cbe..8425cf6e962 100644 --- a/fs/qnx4/bitmap.c +++ b/fs/qnx4/bitmap.c @@ -13,7 +13,6 @@ * 28-06-1998 by Frank Denis : qnx4_free_inode (to be fixed) . */ -#include <linux/config.h> #include <linux/time.h> #include <linux/fs.h> #include <linux/qnx4_fs.h> diff --git a/fs/qnx4/dir.c b/fs/qnx4/dir.c index 9031948fefd..0d7103fa0df 100644 --- a/fs/qnx4/dir.c +++ b/fs/qnx4/dir.c @@ -11,7 +11,6 @@ * 20-06-1998 by Frank Denis : Linux 2.1.99+ & dcache support. */ -#include <linux/config.h> #include <linux/string.h> #include <linux/errno.h> #include <linux/fs.h> diff --git a/fs/qnx4/fsync.c b/fs/qnx4/fsync.c index df5bc75d541..aa3b19544be 100644 --- a/fs/qnx4/fsync.c +++ b/fs/qnx4/fsync.c @@ -10,7 +10,6 @@ * 24-03-1998 by Richard Frowijn : first release. */ -#include <linux/config.h> #include <linux/errno.h> #include <linux/time.h> #include <linux/stat.h> diff --git a/fs/qnx4/inode.c b/fs/qnx4/inode.c index 2f24c46f72a..5a41db2a218 100644 --- a/fs/qnx4/inode.c +++ b/fs/qnx4/inode.c @@ -12,7 +12,6 @@ * 30-06-1998 by Frank Denis : first step to write inodes. */ -#include <linux/config.h> #include <linux/module.h> #include <linux/types.h> #include <linux/string.h> @@ -359,11 +358,10 @@ static int qnx4_fill_super(struct super_block *s, void *data, int silent) const char *errmsg; struct qnx4_sb_info *qs; - qs = kmalloc(sizeof(struct qnx4_sb_info), GFP_KERNEL); + qs = kzalloc(sizeof(struct qnx4_sb_info), GFP_KERNEL); if (!qs) return -ENOMEM; s->s_fs_info = qs; - memset(qs, 0, sizeof(struct qnx4_sb_info)); sb_set_blocksize(s, QNX4_BLOCK_SIZE); @@ -450,7 +448,7 @@ static sector_t qnx4_bmap(struct address_space *mapping, sector_t block) { return generic_block_bmap(mapping,block,qnx4_get_block); } -static struct address_space_operations qnx4_aops = { +static const struct address_space_operations qnx4_aops = { .readpage = qnx4_readpage, .writepage = qnx4_writepage, .sync_page = block_sync_page, @@ -498,7 +496,6 @@ static void qnx4_read_inode(struct inode *inode) inode->i_ctime.tv_sec = le32_to_cpu(raw_inode->di_ctime); inode->i_ctime.tv_nsec = 0; inode->i_blocks = le32_to_cpu(raw_inode->di_first_xtnt.xtnt_size); - inode->i_blksize = QNX4_DIR_ENTRY_SIZE; memcpy(qnx4_inode, raw_inode, QNX4_DIR_ENTRY_SIZE); if (S_ISREG(inode->i_mode)) { @@ -558,9 +555,7 @@ static int init_inodecache(void) static void destroy_inodecache(void) { - if (kmem_cache_destroy(qnx4_inode_cachep)) - printk(KERN_INFO - "qnx4_inode_cache: not all structures were freed\n"); + kmem_cache_destroy(qnx4_inode_cachep); } static int qnx4_get_sb(struct file_system_type *fs_type, diff --git a/fs/qnx4/namei.c b/fs/qnx4/namei.c index 4af4951d7f5..c3d83f67154 100644 --- a/fs/qnx4/namei.c +++ b/fs/qnx4/namei.c @@ -12,7 +12,6 @@ * 04-07-1998 by Frank Denis : first step for rmdir/unlink. */ -#include <linux/config.h> #include <linux/time.h> #include <linux/fs.h> #include <linux/qnx4_fs.h> diff --git a/fs/qnx4/truncate.c b/fs/qnx4/truncate.c index 86563ec01b3..6437c1c3d1d 100644 --- a/fs/qnx4/truncate.c +++ b/fs/qnx4/truncate.c @@ -10,7 +10,6 @@ * 30-06-1998 by Frank DENIS : ugly filler. */ -#include <linux/config.h> #include <linux/types.h> #include <linux/errno.h> #include <linux/fs.h> diff --git a/fs/ramfs/file-mmu.c b/fs/ramfs/file-mmu.c index 00a933eb820..86f14cacf64 100644 --- a/fs/ramfs/file-mmu.c +++ b/fs/ramfs/file-mmu.c @@ -26,7 +26,7 @@ #include <linux/fs.h> -struct address_space_operations ramfs_aops = { +const struct address_space_operations ramfs_aops = { .readpage = simple_readpage, .prepare_write = simple_prepare_write, .commit_write = simple_commit_write diff --git a/fs/ramfs/file-nommu.c b/fs/ramfs/file-nommu.c index f443a84b98a..677139b48e0 100644 --- a/fs/ramfs/file-nommu.c +++ b/fs/ramfs/file-nommu.c @@ -27,7 +27,7 @@ static int ramfs_nommu_setattr(struct dentry *, struct iattr *); -struct address_space_operations ramfs_aops = { +const struct address_space_operations ramfs_aops = { .readpage = simple_readpage, .prepare_write = simple_prepare_write, .commit_write = simple_commit_write @@ -283,9 +283,9 @@ unsigned long ramfs_nommu_get_unmapped_area(struct file *file, /*****************************************************************************/ /* - * set up a mapping + * set up a mapping for shared memory segments */ int ramfs_nommu_mmap(struct file *file, struct vm_area_struct *vma) { - return 0; + return vma->vm_flags & VM_SHARED ? 0 : -ENOSYS; } diff --git a/fs/ramfs/inode.c b/fs/ramfs/inode.c index b9677335cc8..bc0e5166242 100644 --- a/fs/ramfs/inode.c +++ b/fs/ramfs/inode.c @@ -58,7 +58,6 @@ struct inode *ramfs_get_inode(struct super_block *sb, int mode, dev_t dev) inode->i_mode = mode; inode->i_uid = current->fsuid; inode->i_gid = current->fsgid; - inode->i_blksize = PAGE_CACHE_SIZE; inode->i_blocks = 0; inode->i_mapping->a_ops = &ramfs_aops; inode->i_mapping->backing_dev_info = &ramfs_backing_dev_info; diff --git a/fs/ramfs/internal.h b/fs/ramfs/internal.h index 313237631b4..c2bb58e7465 100644 --- a/fs/ramfs/internal.h +++ b/fs/ramfs/internal.h @@ -10,6 +10,6 @@ */ -extern struct address_space_operations ramfs_aops; +extern const struct address_space_operations ramfs_aops; extern const struct file_operations ramfs_file_operations; extern struct inode_operations ramfs_file_inode_operations; diff --git a/fs/read_write.c b/fs/read_write.c index 5bc0e9234f9..d4cb3183c99 100644 --- a/fs/read_write.c +++ b/fs/read_write.c @@ -436,7 +436,7 @@ unsigned long iov_shorten(struct iovec *iov, unsigned long nr_segs, size_t to) return seg; } -EXPORT_SYMBOL(iov_shorten); +EXPORT_UNUSED_SYMBOL(iov_shorten); /* June 2006 */ /* A write operation does a read from user space and vice versa */ #define vrfy_dir(type) ((type) == READ ? VERIFY_WRITE : VERIFY_READ) diff --git a/fs/reiserfs/Makefile b/fs/reiserfs/Makefile index 3a59309f3ca..0eb7ac08048 100644 --- a/fs/reiserfs/Makefile +++ b/fs/reiserfs/Makefile @@ -28,7 +28,7 @@ endif # will work around it. If any other architecture displays this behavior, # add it here. ifeq ($(CONFIG_PPC32),y) -EXTRA_CFLAGS := -O1 +EXTRA_CFLAGS := $(call cc-ifversion, -lt, 0400, -O1) endif TAGS: diff --git a/fs/reiserfs/bitmap.c b/fs/reiserfs/bitmap.c index 909f71e9a30..4a7dbdee1b6 100644 --- a/fs/reiserfs/bitmap.c +++ b/fs/reiserfs/bitmap.c @@ -3,7 +3,6 @@ */ /* Reiserfs block (de)allocator, bitmap-based. */ -#include <linux/config.h> #include <linux/time.h> #include <linux/reiserfs_fs.h> #include <linux/errno.h> diff --git a/fs/reiserfs/dir.c b/fs/reiserfs/dir.c index 973c819f803..9aabcc0ccd2 100644 --- a/fs/reiserfs/dir.c +++ b/fs/reiserfs/dir.c @@ -2,7 +2,6 @@ * Copyright 2000 by Hans Reiser, licensing governed by reiserfs/README */ -#include <linux/config.h> #include <linux/string.h> #include <linux/errno.h> #include <linux/fs.h> diff --git a/fs/reiserfs/do_balan.c b/fs/reiserfs/do_balan.c index b2264ba3cc5..fba304e64de 100644 --- a/fs/reiserfs/do_balan.c +++ b/fs/reiserfs/do_balan.c @@ -15,7 +15,6 @@ ** **/ -#include <linux/config.h> #include <asm/uaccess.h> #include <linux/time.h> #include <linux/reiserfs_fs.h> diff --git a/fs/reiserfs/file.c b/fs/reiserfs/file.c index cf6e1cf4035..1cfbe857ba2 100644 --- a/fs/reiserfs/file.c +++ b/fs/reiserfs/file.c @@ -48,8 +48,8 @@ static int reiserfs_file_release(struct inode *inode, struct file *filp) return 0; } - reiserfs_write_lock(inode->i_sb); mutex_lock(&inode->i_mutex); + reiserfs_write_lock(inode->i_sb); /* freeing preallocation only involves relogging blocks that * are already in the current transaction. preallocation gets * freed at the end of each transaction, so it is impossible for @@ -130,7 +130,7 @@ static int reiserfs_sync_file(struct file *p_s_filp, reiserfs_write_lock(p_s_inode->i_sb); barrier_done = reiserfs_commit_for_inode(p_s_inode); reiserfs_write_unlock(p_s_inode->i_sb); - if (barrier_done != 1) + if (barrier_done != 1 && reiserfs_barrier_flush(p_s_inode->i_sb)) blkdev_issue_flush(p_s_inode->i_sb->s_bdev, NULL); if (barrier_done < 0) return barrier_done; @@ -860,8 +860,12 @@ static int reiserfs_submit_file_region_for_write(struct reiserfs_transaction_han // this sets the proper flags for O_SYNC to trigger a commit mark_inode_dirty(inode); reiserfs_write_unlock(inode->i_sb); - } else + } else { + reiserfs_write_lock(inode->i_sb); + reiserfs_update_inode_transaction(inode); mark_inode_dirty(inode); + reiserfs_write_unlock(inode->i_sb); + } sd_update = 1; } @@ -1560,12 +1564,6 @@ static ssize_t reiserfs_file_write(struct file *file, /* the file we are going t return res; } -static ssize_t reiserfs_aio_write(struct kiocb *iocb, const char __user * buf, - size_t count, loff_t pos) -{ - return generic_file_aio_write(iocb, buf, count, pos); -} - const struct file_operations reiserfs_file_operations = { .read = generic_file_read, .write = reiserfs_file_write, @@ -1575,7 +1573,7 @@ const struct file_operations reiserfs_file_operations = { .fsync = reiserfs_sync_file, .sendfile = generic_file_sendfile, .aio_read = generic_file_aio_read, - .aio_write = reiserfs_aio_write, + .aio_write = generic_file_aio_write, .splice_read = generic_file_splice_read, .splice_write = generic_file_splice_write, }; diff --git a/fs/reiserfs/fix_node.c b/fs/reiserfs/fix_node.c index 5600d3d60cf..6d0e554daa9 100644 --- a/fs/reiserfs/fix_node.c +++ b/fs/reiserfs/fix_node.c @@ -34,7 +34,6 @@ ** **/ -#include <linux/config.h> #include <linux/time.h> #include <linux/string.h> #include <linux/reiserfs_fs.h> diff --git a/fs/reiserfs/ibalance.c b/fs/reiserfs/ibalance.c index 6c5a726fd34..de391a82b99 100644 --- a/fs/reiserfs/ibalance.c +++ b/fs/reiserfs/ibalance.c @@ -2,7 +2,6 @@ * Copyright 2000 by Hans Reiser, licensing governed by reiserfs/README */ -#include <linux/config.h> #include <asm/uaccess.h> #include <linux/string.h> #include <linux/time.h> diff --git a/fs/reiserfs/inode.c b/fs/reiserfs/inode.c index 9857e50f85e..7e5a2f5ebeb 100644 --- a/fs/reiserfs/inode.c +++ b/fs/reiserfs/inode.c @@ -2,7 +2,6 @@ * Copyright 2000 by Hans Reiser, licensing governed by reiserfs/README */ -#include <linux/config.h> #include <linux/time.h> #include <linux/fs.h> #include <linux/reiserfs_fs.h> @@ -18,8 +17,6 @@ #include <linux/writeback.h> #include <linux/quotaops.h> -extern int reiserfs_default_io_size; /* default io size devuned in super.c */ - static int reiserfs_commit_write(struct file *f, struct page *page, unsigned from, unsigned to); static int reiserfs_prepare_write(struct file *f, struct page *page, @@ -40,14 +37,10 @@ void reiserfs_delete_inode(struct inode *inode) /* The = 0 happens when we abort creating a new inode for some reason like lack of space.. */ if (!(inode->i_state & I_NEW) && INODE_PKEY(inode)->k_objectid != 0) { /* also handles bad_inode case */ - mutex_lock(&inode->i_mutex); - reiserfs_delete_xattrs(inode); - if (journal_begin(&th, inode->i_sb, jbegin_count)) { - mutex_unlock(&inode->i_mutex); + if (journal_begin(&th, inode->i_sb, jbegin_count)) goto out; - } reiserfs_update_inode_transaction(inode); err = reiserfs_delete_object(&th, inode); @@ -58,12 +51,8 @@ void reiserfs_delete_inode(struct inode *inode) if (!err) DQUOT_FREE_INODE(inode); - if (journal_end(&th, inode->i_sb, jbegin_count)) { - mutex_unlock(&inode->i_mutex); + if (journal_end(&th, inode->i_sb, jbegin_count)) goto out; - } - - mutex_unlock(&inode->i_mutex); /* check return value from reiserfs_delete_object after * ending the transaction @@ -1131,7 +1120,6 @@ static void init_inode(struct inode *inode, struct path *path) ih = PATH_PITEM_HEAD(path); copy_key(INODE_PKEY(inode), &(ih->ih_key)); - inode->i_blksize = reiserfs_default_io_size; INIT_LIST_HEAD(&(REISERFS_I(inode)->i_prealloc_list)); REISERFS_I(inode)->i_flags = 0; @@ -1139,9 +1127,9 @@ static void init_inode(struct inode *inode, struct path *path) REISERFS_I(inode)->i_prealloc_count = 0; REISERFS_I(inode)->i_trans_id = 0; REISERFS_I(inode)->i_jl = NULL; - REISERFS_I(inode)->i_acl_access = NULL; - REISERFS_I(inode)->i_acl_default = NULL; - init_rwsem(&REISERFS_I(inode)->xattr_sem); + reiserfs_init_acl_access(inode); + reiserfs_init_acl_default(inode); + reiserfs_init_xattr_rwsem(inode); if (stat_data_v1(ih)) { struct stat_data_v1 *sd = @@ -1846,9 +1834,9 @@ int reiserfs_new_inode(struct reiserfs_transaction_handle *th, REISERFS_I(inode)->i_attrs = REISERFS_I(dir)->i_attrs & REISERFS_INHERIT_MASK; sd_attrs_to_i_attrs(REISERFS_I(inode)->i_attrs, inode); - REISERFS_I(inode)->i_acl_access = NULL; - REISERFS_I(inode)->i_acl_default = NULL; - init_rwsem(&REISERFS_I(inode)->xattr_sem); + reiserfs_init_acl_access(inode); + reiserfs_init_acl_default(inode); + reiserfs_init_xattr_rwsem(inode); if (old_format_only(sb)) make_le_item_head(&ih, NULL, KEY_FORMAT_3_5, SD_OFFSET, @@ -1886,7 +1874,6 @@ int reiserfs_new_inode(struct reiserfs_transaction_handle *th, } // these do not go to on-disk stat data inode->i_ino = le32_to_cpu(ih.ih_key.k_objectid); - inode->i_blksize = reiserfs_default_io_size; // store in in-core inode the key of stat data and version all // object items will have (directory items will have old offset @@ -1987,11 +1974,13 @@ int reiserfs_new_inode(struct reiserfs_transaction_handle *th, * iput doesn't deadlock in reiserfs_delete_xattrs. The locking * code really needs to be reworked, but this will take care of it * for now. -jeffm */ +#ifdef CONFIG_REISERFS_FS_POSIX_ACL if (REISERFS_I(dir)->i_acl_default && !IS_ERR(REISERFS_I(dir)->i_acl_default)) { reiserfs_write_unlock_xattrs(dir->i_sb); iput(inode); reiserfs_write_lock_xattrs(dir->i_sb); } else +#endif iput(inode); return err; } @@ -2349,6 +2338,7 @@ static int reiserfs_write_full_page(struct page *page, unsigned long end_index = inode->i_size >> PAGE_CACHE_SHIFT; int error = 0; unsigned long block; + sector_t last_block; struct buffer_head *head, *bh; int partial = 0; int nr = 0; @@ -2396,10 +2386,19 @@ static int reiserfs_write_full_page(struct page *page, } bh = head; block = page->index << (PAGE_CACHE_SHIFT - s->s_blocksize_bits); + last_block = (i_size_read(inode) - 1) >> inode->i_blkbits; /* first map all the buffers, logging any direct items we find */ do { - if ((checked || buffer_dirty(bh)) && (!buffer_mapped(bh) || - (buffer_mapped(bh) + if (block > last_block) { + /* + * This can happen when the block size is less than + * the page size. The corresponding bytes in the page + * were zero filled above + */ + clear_buffer_dirty(bh); + set_buffer_uptodate(bh); + } else if ((checked || buffer_dirty(bh)) && + (!buffer_mapped(bh) || (buffer_mapped(bh) && bh->b_blocknr == 0))) { /* not mapped yet, or it points to a direct item, search @@ -2933,6 +2932,11 @@ int reiserfs_setattr(struct dentry *dentry, struct iattr *attr) } if (error) goto out; + /* + * file size is changed, ctime and mtime are + * to be updated + */ + attr->ia_valid |= (ATTR_MTIME | ATTR_CTIME); } } @@ -2996,7 +3000,7 @@ int reiserfs_setattr(struct dentry *dentry, struct iattr *attr) return error; } -struct address_space_operations reiserfs_address_space_operations = { +const struct address_space_operations reiserfs_address_space_operations = { .writepage = reiserfs_writepage, .readpage = reiserfs_readpage, .readpages = reiserfs_readpages, diff --git a/fs/reiserfs/ioctl.c b/fs/reiserfs/ioctl.c index 745c8810089..a986b5e1e28 100644 --- a/fs/reiserfs/ioctl.c +++ b/fs/reiserfs/ioctl.c @@ -116,12 +116,12 @@ static int reiserfs_unpack(struct inode *inode, struct file *filp) if (REISERFS_I(inode)->i_flags & i_nopack_mask) { return 0; } - reiserfs_write_lock(inode->i_sb); /* we need to make sure nobody is changing the file size beneath ** us */ mutex_lock(&inode->i_mutex); + reiserfs_write_lock(inode->i_sb); write_from = inode->i_size & (blocksize - 1); /* if we are on a block boundary, we are already unpacked. */ diff --git a/fs/reiserfs/journal.c b/fs/reiserfs/journal.c index 1b73529b809..e6b5ccf23f1 100644 --- a/fs/reiserfs/journal.c +++ b/fs/reiserfs/journal.c @@ -34,7 +34,6 @@ ** from within kupdate, it will ignore the immediate flag */ -#include <linux/config.h> #include <asm/uaccess.h> #include <asm/system.h> @@ -834,8 +833,7 @@ static int write_ordered_buffers(spinlock_t * lock, get_bh(bh); if (test_set_buffer_locked(bh)) { if (!buffer_dirty(bh)) { - list_del_init(&jh->list); - list_add(&jh->list, &tmp); + list_move(&jh->list, &tmp); goto loop_next; } spin_unlock(lock); @@ -855,8 +853,7 @@ static int write_ordered_buffers(spinlock_t * lock, ret = -EIO; } if (buffer_dirty(bh)) { - list_del_init(&jh->list); - list_add(&jh->list, &tmp); + list_move(&jh->list, &tmp); add_to_chunk(&chunk, bh, lock, write_ordered_chunk); } else { reiserfs_free_jh(bh); @@ -1189,6 +1186,21 @@ static struct reiserfs_journal_list *find_newer_jl_for_cn(struct return NULL; } +static int newer_jl_done(struct reiserfs_journal_cnode *cn) +{ + struct super_block *sb = cn->sb; + b_blocknr_t blocknr = cn->blocknr; + + cn = cn->hprev; + while (cn) { + if (cn->sb == sb && cn->blocknr == blocknr && cn->jlist && + atomic_read(&cn->jlist->j_commit_left) != 0) + return 0; + cn = cn->hprev; + } + return 1; +} + static void remove_journal_hash(struct super_block *, struct reiserfs_journal_cnode **, struct reiserfs_journal_list *, unsigned long, @@ -1607,6 +1619,31 @@ static int flush_journal_list(struct super_block *s, return err; } +static int test_transaction(struct super_block *s, + struct reiserfs_journal_list *jl) +{ + struct reiserfs_journal_cnode *cn; + + if (jl->j_len == 0 || atomic_read(&jl->j_nonzerolen) == 0) + return 1; + + cn = jl->j_realblock; + while (cn) { + /* if the blocknr == 0, this has been cleared from the hash, + ** skip it + */ + if (cn->blocknr == 0) { + goto next; + } + if (cn->bh && !newer_jl_done(cn)) + return 0; + next: + cn = cn->next; + cond_resched(); + } + return 0; +} + static int write_one_transaction(struct super_block *s, struct reiserfs_journal_list *jl, struct buffer_chunk *chunk) @@ -3436,16 +3473,6 @@ static void flush_async_commits(void *p) flush_commit_list(p_s_sb, jl, 1); } unlock_kernel(); - /* - * this is a little racey, but there's no harm in missing - * the filemap_fdata_write - */ - if (!atomic_read(&journal->j_async_throttle) - && !reiserfs_is_journal_aborted(journal)) { - atomic_inc(&journal->j_async_throttle); - filemap_fdatawrite(p_s_sb->s_bdev->bd_inode->i_mapping); - atomic_dec(&journal->j_async_throttle); - } } /* @@ -3847,7 +3874,9 @@ static void flush_old_journal_lists(struct super_block *s) entry = journal->j_journal_list.next; jl = JOURNAL_LIST_ENTRY(entry); /* this check should always be run, to send old lists to disk */ - if (jl->j_timestamp < (now - (JOURNAL_MAX_TRANS_AGE * 4))) { + if (jl->j_timestamp < (now - (JOURNAL_MAX_TRANS_AGE * 4)) && + atomic_read(&jl->j_commit_left) == 0 && + test_transaction(s, jl)) { flush_used_journal_lists(s, jl); } else { break; diff --git a/fs/reiserfs/lbalance.c b/fs/reiserfs/lbalance.c index 2533c1f64ab..281f8061ac5 100644 --- a/fs/reiserfs/lbalance.c +++ b/fs/reiserfs/lbalance.c @@ -2,7 +2,6 @@ * Copyright 2000 by Hans Reiser, licensing governed by reiserfs/README */ -#include <linux/config.h> #include <asm/uaccess.h> #include <linux/string.h> #include <linux/time.h> diff --git a/fs/reiserfs/namei.c b/fs/reiserfs/namei.c index 284f7852de8..c61710e49c6 100644 --- a/fs/reiserfs/namei.c +++ b/fs/reiserfs/namei.c @@ -11,7 +11,6 @@ * NO WARRANTY */ -#include <linux/config.h> #include <linux/time.h> #include <linux/bitops.h> #include <linux/reiserfs_fs.h> diff --git a/fs/reiserfs/objectid.c b/fs/reiserfs/objectid.c index f62590aa9c9..65feba4deb6 100644 --- a/fs/reiserfs/objectid.c +++ b/fs/reiserfs/objectid.c @@ -2,7 +2,6 @@ * Copyright 2000 by Hans Reiser, licensing governed by reiserfs/README */ -#include <linux/config.h> #include <linux/string.h> #include <linux/random.h> #include <linux/time.h> diff --git a/fs/reiserfs/prints.c b/fs/reiserfs/prints.c index 27bd3a1df2a..bc808a91eea 100644 --- a/fs/reiserfs/prints.c +++ b/fs/reiserfs/prints.c @@ -2,7 +2,6 @@ * Copyright 2000 by Hans Reiser, licensing governed by reiserfs/README */ -#include <linux/config.h> #include <linux/time.h> #include <linux/fs.h> #include <linux/reiserfs_fs.h> diff --git a/fs/reiserfs/procfs.c b/fs/reiserfs/procfs.c index 731688e1cfe..c533ec1bcae 100644 --- a/fs/reiserfs/procfs.c +++ b/fs/reiserfs/procfs.c @@ -10,7 +10,6 @@ /* $Id: procfs.c,v 1.1.8.2 2001/07/15 17:08:42 god Exp $ */ -#include <linux/config.h> #include <linux/module.h> #include <linux/time.h> #include <linux/seq_file.h> @@ -493,9 +492,17 @@ static void add_file(struct super_block *sb, char *name, int reiserfs_proc_info_init(struct super_block *sb) { + char b[BDEVNAME_SIZE]; + char *s; + + /* Some block devices use /'s */ + strlcpy(b, reiserfs_bdevname(sb), BDEVNAME_SIZE); + s = strchr(b, '/'); + if (s) + *s = '!'; + spin_lock_init(&__PINFO(sb).lock); - REISERFS_SB(sb)->procdir = - proc_mkdir(reiserfs_bdevname(sb), proc_info_root); + REISERFS_SB(sb)->procdir = proc_mkdir(b, proc_info_root); if (REISERFS_SB(sb)->procdir) { REISERFS_SB(sb)->procdir->owner = THIS_MODULE; REISERFS_SB(sb)->procdir->data = sb; @@ -509,13 +516,22 @@ int reiserfs_proc_info_init(struct super_block *sb) return 0; } reiserfs_warning(sb, "reiserfs: cannot create /proc/%s/%s", - proc_info_root_name, reiserfs_bdevname(sb)); + proc_info_root_name, b); return 1; } int reiserfs_proc_info_done(struct super_block *sb) { struct proc_dir_entry *de = REISERFS_SB(sb)->procdir; + char b[BDEVNAME_SIZE]; + char *s; + + /* Some block devices use /'s */ + strlcpy(b, reiserfs_bdevname(sb), BDEVNAME_SIZE); + s = strchr(b, '/'); + if (s) + *s = '!'; + if (de) { remove_proc_entry("journal", de); remove_proc_entry("oidmap", de); @@ -529,7 +545,7 @@ int reiserfs_proc_info_done(struct super_block *sb) __PINFO(sb).exiting = 1; spin_unlock(&__PINFO(sb).lock); if (proc_info_root) { - remove_proc_entry(reiserfs_bdevname(sb), proc_info_root); + remove_proc_entry(b, proc_info_root); REISERFS_SB(sb)->procdir = NULL; } return 0; diff --git a/fs/reiserfs/stree.c b/fs/reiserfs/stree.c index d2b25e1ba6e..8b9b1312713 100644 --- a/fs/reiserfs/stree.c +++ b/fs/reiserfs/stree.c @@ -49,7 +49,6 @@ * reiserfs_insert_item */ -#include <linux/config.h> #include <linux/time.h> #include <linux/string.h> #include <linux/pagemap.h> diff --git a/fs/reiserfs/super.c b/fs/reiserfs/super.c index 00f1321e920..80fc3b32802 100644 --- a/fs/reiserfs/super.c +++ b/fs/reiserfs/super.c @@ -11,7 +11,6 @@ * NO WARRANTY */ -#include <linux/config.h> #include <linux/module.h> #include <linux/vmalloc.h> #include <linux/time.h> @@ -511,8 +510,10 @@ static void init_once(void *foo, kmem_cache_t * cachep, unsigned long flags) SLAB_CTOR_CONSTRUCTOR) { INIT_LIST_HEAD(&ei->i_prealloc_list); inode_init_once(&ei->vfs_inode); +#ifdef CONFIG_REISERFS_FS_POSIX_ACL ei->i_acl_access = NULL; ei->i_acl_default = NULL; +#endif } } @@ -531,9 +532,7 @@ static int init_inodecache(void) static void destroy_inodecache(void) { - if (kmem_cache_destroy(reiserfs_inode_cachep)) - reiserfs_warning(NULL, - "reiserfs_inode_cache: not all structures were freed"); + kmem_cache_destroy(reiserfs_inode_cachep); } /* we don't mark inodes dirty, we just log them */ @@ -563,6 +562,7 @@ static void reiserfs_dirty_inode(struct inode *inode) reiserfs_write_unlock(inode->i_sb); } +#ifdef CONFIG_REISERFS_FS_POSIX_ACL static void reiserfs_clear_inode(struct inode *inode) { struct posix_acl *acl; @@ -577,6 +577,9 @@ static void reiserfs_clear_inode(struct inode *inode) posix_acl_release(acl); REISERFS_I(inode)->i_acl_default = NULL; } +#else +#define reiserfs_clear_inode NULL +#endif #ifdef CONFIG_QUOTA static ssize_t reiserfs_quota_write(struct super_block *, int, const char *, @@ -726,12 +729,6 @@ static const arg_desc_t error_actions[] = { {NULL, 0, 0}, }; -int reiserfs_default_io_size = 128 * 1024; /* Default recommended I/O size is 128k. - There might be broken applications that are - confused by this. Use nolargeio mount option - to get usual i/o size = PAGE_SIZE. - */ - /* proceed only one option from a list *cur - string containing of mount options opts - array of options which are accepted opt_arg - if option is found and requires an argument and if it is specifed @@ -960,19 +957,8 @@ static int reiserfs_parse_options(struct super_block *s, char *options, /* strin } if (c == 'w') { - char *p = NULL; - int val = simple_strtoul(arg, &p, 0); - - if (*p != '\0') { - reiserfs_warning(s, - "reiserfs_parse_options: non-numeric value %s for nolargeio option", - arg); - return 0; - } - if (val) - reiserfs_default_io_size = PAGE_SIZE; - else - reiserfs_default_io_size = 128 * 1024; + reiserfs_warning(s, "reiserfs: nolargeio option is no longer supported"); + return 0; } if (c == 'j') { @@ -2204,7 +2190,7 @@ static ssize_t reiserfs_quota_write(struct super_block *sb, int type, size_t towrite = len; struct buffer_head tmp_bh, *bh; - mutex_lock(&inode->i_mutex); + mutex_lock_nested(&inode->i_mutex, I_MUTEX_QUOTA); while (towrite > 0) { tocopy = sb->s_blocksize - offset < towrite ? sb->s_blocksize - offset : towrite; diff --git a/fs/reiserfs/tail_conversion.c b/fs/reiserfs/tail_conversion.c index 196e971c03c..36f108fc1cf 100644 --- a/fs/reiserfs/tail_conversion.c +++ b/fs/reiserfs/tail_conversion.c @@ -2,7 +2,6 @@ * Copyright 1999 Hans Reiser, see reiserfs/README for licensing and copyright details */ -#include <linux/config.h> #include <linux/time.h> #include <linux/pagemap.h> #include <linux/buffer_head.h> diff --git a/fs/reiserfs/xattr.c b/fs/reiserfs/xattr.c index 39fedaa88a0..d935fb9394e 100644 --- a/fs/reiserfs/xattr.c +++ b/fs/reiserfs/xattr.c @@ -424,7 +424,7 @@ int xattr_readdir(struct file *file, filldir_t filler, void *buf) int res = -ENOTDIR; if (!file->f_op || !file->f_op->readdir) goto out; - mutex_lock(&inode->i_mutex); + mutex_lock_nested(&inode->i_mutex, I_MUTEX_XATTR); // down(&inode->i_zombie); res = -ENOENT; if (!IS_DEADDIR(inode)) { diff --git a/fs/romfs/inode.c b/fs/romfs/inode.c index 283fbc6b8ee..ddcd9e1ef28 100644 --- a/fs/romfs/inode.c +++ b/fs/romfs/inode.c @@ -459,7 +459,7 @@ err_out: /* Mapping from our types to the kernel */ -static struct address_space_operations romfs_aops = { +static const struct address_space_operations romfs_aops = { .readpage = romfs_readpage }; @@ -589,8 +589,7 @@ static int init_inodecache(void) static void destroy_inodecache(void) { - if (kmem_cache_destroy(romfs_inode_cachep)) - printk(KERN_INFO "romfs_inode_cache: not all structures were freed\n"); + kmem_cache_destroy(romfs_inode_cachep); } static int romfs_remount(struct super_block *sb, int *flags, char *data) diff --git a/fs/select.c b/fs/select.c index 9c4f0f2604f..dcbc1112b7e 100644 --- a/fs/select.c +++ b/fs/select.c @@ -658,8 +658,6 @@ int do_sys_poll(struct pollfd __user *ufds, unsigned int nfds, s64 *timeout) unsigned int i; struct poll_list *head; struct poll_list *walk; - struct fdtable *fdt; - int max_fdset; /* Allocate small arguments on the stack to save memory and be faster - use long to make sure the buffer is aligned properly on 64 bit archs to avoid unaligned access */ @@ -667,11 +665,7 @@ int do_sys_poll(struct pollfd __user *ufds, unsigned int nfds, s64 *timeout) struct poll_list *stack_pp = NULL; /* Do a sanity check on nfds ... */ - rcu_read_lock(); - fdt = files_fdtable(current->files); - max_fdset = fdt->max_fdset; - rcu_read_unlock(); - if (nfds > max_fdset && nfds > OPEN_MAX) + if (nfds > current->signal->rlim[RLIMIT_NOFILE].rlim_cur) return -EINVAL; poll_initwait(&table); @@ -746,9 +740,9 @@ out_fds: asmlinkage long sys_poll(struct pollfd __user *ufds, unsigned int nfds, long timeout_msecs) { - s64 timeout_jiffies = 0; + s64 timeout_jiffies; - if (timeout_msecs) { + if (timeout_msecs > 0) { #if HZ > 1000 /* We can only overflow if HZ > 1000 */ if (timeout_msecs / 1000 > (s64)0x7fffffffffffffffULL / (s64)HZ) @@ -756,6 +750,9 @@ asmlinkage long sys_poll(struct pollfd __user *ufds, unsigned int nfds, else #endif timeout_jiffies = msecs_to_jiffies(timeout_msecs); + } else { + /* Infinite (< 0) or no (0) timeout */ + timeout_jiffies = timeout_msecs; } return do_sys_poll(ufds, nfds, &timeout_jiffies); diff --git a/fs/smbfs/file.c b/fs/smbfs/file.c index ed9a24d19d7..dae67048bab 100644 --- a/fs/smbfs/file.c +++ b/fs/smbfs/file.c @@ -306,7 +306,7 @@ static int smb_commit_write(struct file *file, struct page *page, return status; } -struct address_space_operations smb_file_aops = { +const struct address_space_operations smb_file_aops = { .readpage = smb_readpage, .writepage = smb_writepage, .prepare_write = smb_prepare_write, diff --git a/fs/smbfs/inode.c b/fs/smbfs/inode.c index 506ff87c1d4..2c122ee83ad 100644 --- a/fs/smbfs/inode.c +++ b/fs/smbfs/inode.c @@ -7,7 +7,6 @@ * Please add a note about your changes to smbfs in the ChangeLog file. */ -#include <linux/config.h> #include <linux/module.h> #include <linux/time.h> #include <linux/kernel.h> @@ -90,8 +89,7 @@ static int init_inodecache(void) static void destroy_inodecache(void) { - if (kmem_cache_destroy(smb_inode_cachep)) - printk(KERN_INFO "smb_inode_cache: not all structures were freed\n"); + kmem_cache_destroy(smb_inode_cachep); } static int smb_remount(struct super_block *sb, int *flags, char *data) @@ -168,7 +166,6 @@ smb_get_inode_attr(struct inode *inode, struct smb_fattr *fattr) fattr->f_mtime = inode->i_mtime; fattr->f_ctime = inode->i_ctime; fattr->f_atime = inode->i_atime; - fattr->f_blksize= inode->i_blksize; fattr->f_blocks = inode->i_blocks; fattr->attr = SMB_I(inode)->attr; @@ -202,7 +199,6 @@ smb_set_inode_attr(struct inode *inode, struct smb_fattr *fattr) inode->i_uid = fattr->f_uid; inode->i_gid = fattr->f_gid; inode->i_ctime = fattr->f_ctime; - inode->i_blksize= fattr->f_blksize; inode->i_blocks = fattr->f_blocks; inode->i_size = fattr->f_size; inode->i_mtime = fattr->f_mtime; diff --git a/fs/smbfs/proc.c b/fs/smbfs/proc.c index c3495059889..40e174db987 100644 --- a/fs/smbfs/proc.c +++ b/fs/smbfs/proc.c @@ -1826,7 +1826,6 @@ smb_init_dirent(struct smb_sb_info *server, struct smb_fattr *fattr) fattr->f_nlink = 1; fattr->f_uid = server->mnt->uid; fattr->f_gid = server->mnt->gid; - fattr->f_blksize = SMB_ST_BLKSIZE; fattr->f_unix = 0; } diff --git a/fs/smbfs/proto.h b/fs/smbfs/proto.h index 972ed7dad38..34fb462b237 100644 --- a/fs/smbfs/proto.h +++ b/fs/smbfs/proto.h @@ -63,7 +63,7 @@ extern int smb_revalidate_inode(struct dentry *dentry); extern int smb_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat); extern int smb_notify_change(struct dentry *dentry, struct iattr *attr); /* file.c */ -extern struct address_space_operations smb_file_aops; +extern const struct address_space_operations smb_file_aops; extern const struct file_operations smb_file_operations; extern struct inode_operations smb_file_inode_operations; /* ioctl.c */ diff --git a/fs/smbfs/request.c b/fs/smbfs/request.c index c71dd2760d3..0fb74697abc 100644 --- a/fs/smbfs/request.c +++ b/fs/smbfs/request.c @@ -49,8 +49,7 @@ int smb_init_request_cache(void) void smb_destroy_request_cache(void) { - if (kmem_cache_destroy(req_cachep)) - printk(KERN_INFO "smb_destroy_request_cache: not all structures were freed\n"); + kmem_cache_destroy(req_cachep); } /* @@ -400,8 +399,7 @@ static int smb_request_send_req(struct smb_request *req) if (!(req->rq_flags & SMB_REQ_TRANSMITTED)) goto out; - list_del_init(&req->rq_queue); - list_add_tail(&req->rq_queue, &server->recvq); + list_move_tail(&req->rq_queue, &server->recvq); result = 1; out: return result; @@ -435,8 +433,7 @@ int smb_request_send_server(struct smb_sb_info *server) result = smb_request_send_req(req); if (result < 0) { server->conn_error = result; - list_del_init(&req->rq_queue); - list_add(&req->rq_queue, &server->xmitq); + list_move(&req->rq_queue, &server->xmitq); result = -EIO; goto out; } diff --git a/fs/smbfs/smbiod.c b/fs/smbfs/smbiod.c index 481a97a423f..e6754044128 100644 --- a/fs/smbfs/smbiod.c +++ b/fs/smbfs/smbiod.c @@ -5,7 +5,6 @@ * Copyright (C) 2001, Urban Widmark */ -#include <linux/config.h> #include <linux/sched.h> #include <linux/kernel.h> @@ -20,6 +19,7 @@ #include <linux/smp_lock.h> #include <linux/module.h> #include <linux/net.h> +#include <linux/kthread.h> #include <net/ip.h> #include <linux/smb_fs.h> @@ -40,7 +40,7 @@ enum smbiod_state { }; static enum smbiod_state smbiod_state = SMBIOD_DEAD; -static pid_t smbiod_pid; +static struct task_struct *smbiod_thread; static DECLARE_WAIT_QUEUE_HEAD(smbiod_wait); static LIST_HEAD(smb_servers); static DEFINE_SPINLOCK(servers_lock); @@ -67,20 +67,29 @@ void smbiod_wake_up(void) */ static int smbiod_start(void) { - pid_t pid; + struct task_struct *tsk; + int err = 0; + if (smbiod_state != SMBIOD_DEAD) return 0; smbiod_state = SMBIOD_STARTING; __module_get(THIS_MODULE); spin_unlock(&servers_lock); - pid = kernel_thread(smbiod, NULL, 0); - if (pid < 0) + tsk = kthread_run(smbiod, NULL, "smbiod"); + if (IS_ERR(tsk)) { + err = PTR_ERR(tsk); module_put(THIS_MODULE); + } spin_lock(&servers_lock); - smbiod_state = pid < 0 ? SMBIOD_DEAD : SMBIOD_RUNNING; - smbiod_pid = pid; - return pid; + if (err < 0) { + smbiod_state = SMBIOD_DEAD; + smbiod_thread = NULL; + } else { + smbiod_state = SMBIOD_RUNNING; + smbiod_thread = tsk; + } + return err; } /* @@ -183,8 +192,7 @@ int smbiod_retry(struct smb_sb_info *server) if (req->rq_flags & SMB_REQ_RETRY) { /* must move the request to the xmitq */ VERBOSE("retrying request %p on recvq\n", req); - list_del(&req->rq_queue); - list_add(&req->rq_queue, &server->xmitq); + list_move(&req->rq_queue, &server->xmitq); continue; } #endif @@ -290,8 +298,6 @@ out: */ static int smbiod(void *unused) { - daemonize("smbiod"); - allow_signal(SIGKILL); VERBOSE("SMB Kernel thread starting (%d) ...\n", current->pid); diff --git a/fs/splice.c b/fs/splice.c index 05fd2787be9..684bca3d3a1 100644 --- a/fs/splice.c +++ b/fs/splice.c @@ -1307,6 +1307,85 @@ asmlinkage long sys_splice(int fd_in, loff_t __user *off_in, } /* + * Make sure there's data to read. Wait for input if we can, otherwise + * return an appropriate error. + */ +static int link_ipipe_prep(struct pipe_inode_info *pipe, unsigned int flags) +{ + int ret; + + /* + * Check ->nrbufs without the inode lock first. This function + * is speculative anyways, so missing one is ok. + */ + if (pipe->nrbufs) + return 0; + + ret = 0; + mutex_lock(&pipe->inode->i_mutex); + + while (!pipe->nrbufs) { + if (signal_pending(current)) { + ret = -ERESTARTSYS; + break; + } + if (!pipe->writers) + break; + if (!pipe->waiting_writers) { + if (flags & SPLICE_F_NONBLOCK) { + ret = -EAGAIN; + break; + } + } + pipe_wait(pipe); + } + + mutex_unlock(&pipe->inode->i_mutex); + return ret; +} + +/* + * Make sure there's writeable room. Wait for room if we can, otherwise + * return an appropriate error. + */ +static int link_opipe_prep(struct pipe_inode_info *pipe, unsigned int flags) +{ + int ret; + + /* + * Check ->nrbufs without the inode lock first. This function + * is speculative anyways, so missing one is ok. + */ + if (pipe->nrbufs < PIPE_BUFFERS) + return 0; + + ret = 0; + mutex_lock(&pipe->inode->i_mutex); + + while (pipe->nrbufs >= PIPE_BUFFERS) { + if (!pipe->readers) { + send_sig(SIGPIPE, current, 0); + ret = -EPIPE; + break; + } + if (flags & SPLICE_F_NONBLOCK) { + ret = -EAGAIN; + break; + } + if (signal_pending(current)) { + ret = -ERESTARTSYS; + break; + } + pipe->waiting_writers++; + pipe_wait(pipe); + pipe->waiting_writers--; + } + + mutex_unlock(&pipe->inode->i_mutex); + return ret; +} + +/* * Link contents of ipipe to opipe. */ static int link_pipe(struct pipe_inode_info *ipipe, @@ -1314,9 +1393,7 @@ static int link_pipe(struct pipe_inode_info *ipipe, size_t len, unsigned int flags) { struct pipe_buffer *ibuf, *obuf; - int ret, do_wakeup, i, ipipe_first; - - ret = do_wakeup = ipipe_first = 0; + int ret = 0, i = 0, nbuf; /* * Potential ABBA deadlock, work around it by ordering lock @@ -1324,126 +1401,62 @@ static int link_pipe(struct pipe_inode_info *ipipe, * could deadlock (one doing tee from A -> B, the other from B -> A). */ if (ipipe->inode < opipe->inode) { - ipipe_first = 1; - mutex_lock(&ipipe->inode->i_mutex); - mutex_lock(&opipe->inode->i_mutex); + mutex_lock_nested(&ipipe->inode->i_mutex, I_MUTEX_PARENT); + mutex_lock_nested(&opipe->inode->i_mutex, I_MUTEX_CHILD); } else { - mutex_lock(&opipe->inode->i_mutex); - mutex_lock(&ipipe->inode->i_mutex); + mutex_lock_nested(&opipe->inode->i_mutex, I_MUTEX_PARENT); + mutex_lock_nested(&ipipe->inode->i_mutex, I_MUTEX_CHILD); } - for (i = 0;; i++) { + do { if (!opipe->readers) { send_sig(SIGPIPE, current, 0); if (!ret) ret = -EPIPE; break; } - if (ipipe->nrbufs - i) { - ibuf = ipipe->bufs + ((ipipe->curbuf + i) & (PIPE_BUFFERS - 1)); - /* - * If we have room, fill this buffer - */ - if (opipe->nrbufs < PIPE_BUFFERS) { - int nbuf = (opipe->curbuf + opipe->nrbufs) & (PIPE_BUFFERS - 1); - - /* - * Get a reference to this pipe buffer, - * so we can copy the contents over. - */ - ibuf->ops->get(ipipe, ibuf); - - obuf = opipe->bufs + nbuf; - *obuf = *ibuf; - - /* - * Don't inherit the gift flag, we need to - * prevent multiple steals of this page. - */ - obuf->flags &= ~PIPE_BUF_FLAG_GIFT; - - if (obuf->len > len) - obuf->len = len; - - opipe->nrbufs++; - do_wakeup = 1; - ret += obuf->len; - len -= obuf->len; - - if (!len) - break; - if (opipe->nrbufs < PIPE_BUFFERS) - continue; - } - - /* - * We have input available, but no output room. - * If we already copied data, return that. If we - * need to drop the opipe lock, it must be ordered - * last to avoid deadlocks. - */ - if ((flags & SPLICE_F_NONBLOCK) || !ipipe_first) { - if (!ret) - ret = -EAGAIN; - break; - } - if (signal_pending(current)) { - if (!ret) - ret = -ERESTARTSYS; - break; - } - if (do_wakeup) { - smp_mb(); - if (waitqueue_active(&opipe->wait)) - wake_up_interruptible(&opipe->wait); - kill_fasync(&opipe->fasync_readers, SIGIO, POLL_IN); - do_wakeup = 0; - } + /* + * If we have iterated all input buffers or ran out of + * output room, break. + */ + if (i >= ipipe->nrbufs || opipe->nrbufs >= PIPE_BUFFERS) + break; - opipe->waiting_writers++; - pipe_wait(opipe); - opipe->waiting_writers--; - continue; - } + ibuf = ipipe->bufs + ((ipipe->curbuf + i) & (PIPE_BUFFERS - 1)); + nbuf = (opipe->curbuf + opipe->nrbufs) & (PIPE_BUFFERS - 1); /* - * No input buffers, do the usual checks for available - * writers and blocking and wait if necessary + * Get a reference to this pipe buffer, + * so we can copy the contents over. */ - if (!ipipe->writers) - break; - if (!ipipe->waiting_writers) { - if (ret) - break; - } + ibuf->ops->get(ipipe, ibuf); + + obuf = opipe->bufs + nbuf; + *obuf = *ibuf; + /* - * pipe_wait() drops the ipipe mutex. To avoid deadlocks - * with another process, we can only safely do that if - * the ipipe lock is ordered last. + * Don't inherit the gift flag, we need to + * prevent multiple steals of this page. */ - if ((flags & SPLICE_F_NONBLOCK) || ipipe_first) { - if (!ret) - ret = -EAGAIN; - break; - } - if (signal_pending(current)) { - if (!ret) - ret = -ERESTARTSYS; - break; - } + obuf->flags &= ~PIPE_BUF_FLAG_GIFT; - if (waitqueue_active(&ipipe->wait)) - wake_up_interruptible_sync(&ipipe->wait); - kill_fasync(&ipipe->fasync_writers, SIGIO, POLL_OUT); + if (obuf->len > len) + obuf->len = len; - pipe_wait(ipipe); - } + opipe->nrbufs++; + ret += obuf->len; + len -= obuf->len; + i++; + } while (len); mutex_unlock(&ipipe->inode->i_mutex); mutex_unlock(&opipe->inode->i_mutex); - if (do_wakeup) { + /* + * If we put data in the output pipe, wakeup any potential readers. + */ + if (ret > 0) { smp_mb(); if (waitqueue_active(&opipe->wait)) wake_up_interruptible(&opipe->wait); @@ -1464,14 +1477,29 @@ static long do_tee(struct file *in, struct file *out, size_t len, { struct pipe_inode_info *ipipe = in->f_dentry->d_inode->i_pipe; struct pipe_inode_info *opipe = out->f_dentry->d_inode->i_pipe; + int ret = -EINVAL; /* - * Link ipipe to the two output pipes, consuming as we go along. + * Duplicate the contents of ipipe to opipe without actually + * copying the data. */ - if (ipipe && opipe) - return link_pipe(ipipe, opipe, len, flags); + if (ipipe && opipe && ipipe != opipe) { + /* + * Keep going, unless we encounter an error. The ipipe/opipe + * ordering doesn't really matter. + */ + ret = link_ipipe_prep(ipipe, flags); + if (!ret) { + ret = link_opipe_prep(opipe, flags); + if (!ret) { + ret = link_pipe(ipipe, opipe, len, flags); + if (!ret && (flags & SPLICE_F_NONBLOCK)) + ret = -EAGAIN; + } + } + } - return -EINVAL; + return ret; } asmlinkage long sys_tee(int fdin, int fdout, size_t len, unsigned int flags) diff --git a/fs/stat.c b/fs/stat.c index 0f282face32..60a31d5e596 100644 --- a/fs/stat.c +++ b/fs/stat.c @@ -4,7 +4,6 @@ * Copyright (C) 1991, 1992 Linus Torvalds */ -#include <linux/config.h> #include <linux/module.h> #include <linux/mm.h> #include <linux/errno.h> @@ -15,6 +14,7 @@ #include <linux/namei.h> #include <linux/security.h> #include <linux/syscalls.h> +#include <linux/pagemap.h> #include <asm/uaccess.h> #include <asm/unistd.h> @@ -33,7 +33,7 @@ void generic_fillattr(struct inode *inode, struct kstat *stat) stat->ctime = inode->i_ctime; stat->size = i_size_read(inode); stat->blocks = inode->i_blocks; - stat->blksize = inode->i_blksize; + stat->blksize = (1 << inode->i_blkbits); } EXPORT_SYMBOL(generic_fillattr); diff --git a/fs/super.c b/fs/super.c index 057b5325b7e..6987824d0dc 100644 --- a/fs/super.c +++ b/fs/super.c @@ -20,7 +20,6 @@ * Heavily rewritten for 'one fs - one tree' dcache architecture. AV, Mar 2000 */ -#include <linux/config.h> #include <linux/module.h> #include <linux/slab.h> #include <linux/init.h> @@ -50,11 +49,12 @@ DEFINE_SPINLOCK(sb_lock); /** * alloc_super - create new superblock + * @type: filesystem type superblock should belong to * * Allocates and initializes a new &struct super_block. alloc_super() * returns a pointer new superblock or %NULL if allocation had failed. */ -static struct super_block *alloc_super(void) +static struct super_block *alloc_super(struct file_system_type *type) { struct super_block *s = kzalloc(sizeof(struct super_block), GFP_USER); static struct super_operations default_op; @@ -73,6 +73,13 @@ static struct super_block *alloc_super(void) INIT_LIST_HEAD(&s->s_inodes); init_rwsem(&s->s_umount); mutex_init(&s->s_lock); + lockdep_set_class(&s->s_umount, &type->s_umount_key); + /* + * The locking rules for s_lock are up to the + * filesystem. For example ext3fs has different + * lock ordering than usbfs: + */ + lockdep_set_class(&s->s_lock, &type->s_lock_key); down_write(&s->s_umount); s->s_count = S_BIAS; atomic_set(&s->s_active, 1); @@ -192,7 +199,7 @@ EXPORT_SYMBOL(deactivate_super); * success, 0 if we had failed (superblock contents was already dead or * dying when grab_super() had been called). */ -static int grab_super(struct super_block *s) +static int grab_super(struct super_block *s) __releases(sb_lock) { s->s_count++; spin_unlock(&sb_lock); @@ -296,7 +303,7 @@ retry: } if (!s) { spin_unlock(&sb_lock); - s = alloc_super(); + s = alloc_super(type); if (!s) return ERR_PTR(-ENOMEM); goto retry; @@ -871,8 +878,6 @@ do_kern_mount(const char *fstype, int flags, const char *name, void *data) return mnt; } -EXPORT_SYMBOL_GPL(do_kern_mount); - struct vfsmount *kern_mount(struct file_system_type *type) { return vfs_kern_mount(type, 0, type->name, NULL); diff --git a/fs/sysfs/bin.c b/fs/sysfs/bin.c index c16a93c353c..98022e41cda 100644 --- a/fs/sysfs/bin.c +++ b/fs/sysfs/bin.c @@ -10,6 +10,7 @@ #include <linux/errno.h> #include <linux/fs.h> +#include <linux/kernel.h> #include <linux/kobject.h> #include <linux/module.h> #include <linux/slab.h> @@ -176,7 +177,6 @@ const struct file_operations bin_fops = { * sysfs_create_bin_file - create binary file for object. * @kobj: object. * @attr: attribute descriptor. - * */ int sysfs_create_bin_file(struct kobject * kobj, struct bin_attribute * attr) @@ -191,13 +191,16 @@ int sysfs_create_bin_file(struct kobject * kobj, struct bin_attribute * attr) * sysfs_remove_bin_file - remove binary file for object. * @kobj: object. * @attr: attribute descriptor. - * */ -int sysfs_remove_bin_file(struct kobject * kobj, struct bin_attribute * attr) +void sysfs_remove_bin_file(struct kobject * kobj, struct bin_attribute * attr) { - sysfs_hash_and_remove(kobj->dentry,attr->attr.name); - return 0; + if (sysfs_hash_and_remove(kobj->dentry, attr->attr.name) < 0) { + printk(KERN_ERR "%s: " + "bad dentry or inode or no such file: \"%s\"\n", + __FUNCTION__, attr->attr.name); + dump_stack(); + } } EXPORT_SYMBOL_GPL(sysfs_create_bin_file); diff --git a/fs/sysfs/dir.c b/fs/sysfs/dir.c index 610b5bdbe75..5f3d725d112 100644 --- a/fs/sysfs/dir.c +++ b/fs/sysfs/dir.c @@ -43,7 +43,7 @@ static struct sysfs_dirent * sysfs_new_dirent(struct sysfs_dirent * parent_sd, memset(sd, 0, sizeof(*sd)); atomic_set(&sd->s_count, 1); - atomic_set(&sd->s_event, 0); + atomic_set(&sd->s_event, 1); INIT_LIST_HEAD(&sd->s_children); list_add(&sd->s_sibling, &parent_sd->s_children); sd->s_element = element; @@ -430,10 +430,9 @@ static int sysfs_readdir(struct file * filp, void * dirent, filldir_t filldir) i++; /* fallthrough */ default: - if (filp->f_pos == 2) { - list_del(q); - list_add(q, &parent_sd->s_children); - } + if (filp->f_pos == 2) + list_move(q, &parent_sd->s_children); + for (p=q->next; p!= &parent_sd->s_children; p=p->next) { struct sysfs_dirent *next; const char * name; @@ -455,8 +454,7 @@ static int sysfs_readdir(struct file * filp, void * dirent, filldir_t filldir) dt_type(next)) < 0) return 0; - list_del(q); - list_add(q, p); + list_move(q, p); p = q; filp->f_pos++; } diff --git a/fs/sysfs/inode.c b/fs/sysfs/inode.c index f0b347bd12c..e79e38d52c0 100644 --- a/fs/sysfs/inode.c +++ b/fs/sysfs/inode.c @@ -12,11 +12,12 @@ #include <linux/namei.h> #include <linux/backing-dev.h> #include <linux/capability.h> +#include <linux/errno.h> #include "sysfs.h" extern struct super_block * sysfs_sb; -static struct address_space_operations sysfs_aops = { +static const struct address_space_operations sysfs_aops = { .readpage = simple_readpage, .prepare_write = simple_prepare_write, .commit_write = simple_commit_write @@ -109,15 +110,26 @@ static inline void set_inode_attr(struct inode * inode, struct iattr * iattr) inode->i_ctime = iattr->ia_ctime; } + +/* + * sysfs has a different i_mutex lock order behavior for i_mutex than other + * filesystems; sysfs i_mutex is called in many places with subsystem locks + * held. At the same time, many of the VFS locking rules do not apply to + * sysfs at all (cross directory rename for example). To untangle this mess + * (which gives false positives in lockdep), we're giving sysfs inodes their + * own class for i_mutex. + */ +static struct lock_class_key sysfs_inode_imutex_key; + struct inode * sysfs_new_inode(mode_t mode, struct sysfs_dirent * sd) { struct inode * inode = new_inode(sysfs_sb); if (inode) { - inode->i_blksize = PAGE_CACHE_SIZE; inode->i_blocks = 0; inode->i_mapping->a_ops = &sysfs_aops; inode->i_mapping->backing_dev_info = &sysfs_backing_dev_info; inode->i_op = &sysfs_inode_operations; + lockdep_set_class(&inode->i_mutex, &sysfs_inode_imutex_key); if (sd->s_iattr) { /* sysfs_dirent has non-default attributes @@ -222,17 +234,18 @@ void sysfs_drop_dentry(struct sysfs_dirent * sd, struct dentry * parent) } } -void sysfs_hash_and_remove(struct dentry * dir, const char * name) +int sysfs_hash_and_remove(struct dentry * dir, const char * name) { struct sysfs_dirent * sd; struct sysfs_dirent * parent_sd; + int found = 0; if (!dir) - return; + return -ENOENT; if (dir->d_inode == NULL) /* no inode means this hasn't been made visible yet */ - return; + return -ENOENT; parent_sd = dir->d_fsdata; mutex_lock(&dir->d_inode->i_mutex); @@ -243,8 +256,11 @@ void sysfs_hash_and_remove(struct dentry * dir, const char * name) list_del_init(&sd->s_sibling); sysfs_drop_dentry(sd, dir); sysfs_put(sd); + found = 1; break; } } mutex_unlock(&dir->d_inode->i_mutex); + + return found ? 0 : -ENOENT; } diff --git a/fs/sysfs/symlink.c b/fs/sysfs/symlink.c index d2eac3ceed5..f50e3cc2ded 100644 --- a/fs/sysfs/symlink.c +++ b/fs/sysfs/symlink.c @@ -3,6 +3,7 @@ */ #include <linux/fs.h> +#include <linux/mount.h> #include <linux/module.h> #include <linux/kobject.h> #include <linux/namei.h> @@ -82,10 +83,19 @@ exit1: */ int sysfs_create_link(struct kobject * kobj, struct kobject * target, const char * name) { - struct dentry * dentry = kobj->dentry; + struct dentry *dentry = NULL; int error = -EEXIST; - BUG_ON(!kobj || !kobj->dentry || !name); + BUG_ON(!name); + + if (!kobj) { + if (sysfs_mount && sysfs_mount->mnt_sb) + dentry = sysfs_mount->mnt_sb->s_root; + } else + dentry = kobj->dentry; + + if (!dentry) + return -EFAULT; mutex_lock(&dentry->d_inode->i_mutex); if (!sysfs_dirent_exist(dentry->d_fsdata, name)) diff --git a/fs/sysfs/sysfs.h b/fs/sysfs/sysfs.h index 3651ffb5ec0..6f3d6bd5288 100644 --- a/fs/sysfs/sysfs.h +++ b/fs/sysfs/sysfs.h @@ -10,7 +10,7 @@ extern int sysfs_make_dirent(struct sysfs_dirent *, struct dentry *, void *, umode_t, int); extern int sysfs_add_file(struct dentry *, const struct attribute *, int); -extern void sysfs_hash_and_remove(struct dentry * dir, const char * name); +extern int sysfs_hash_and_remove(struct dentry * dir, const char * name); extern struct sysfs_dirent *sysfs_find(struct sysfs_dirent *dir, const char * name); extern int sysfs_create_subdir(struct kobject *, const char *, struct dentry **); diff --git a/fs/sysv/ialloc.c b/fs/sysv/ialloc.c index 9b585d1081c..115ab0d6f4b 100644 --- a/fs/sysv/ialloc.c +++ b/fs/sysv/ialloc.c @@ -170,7 +170,7 @@ struct inode * sysv_new_inode(const struct inode * dir, mode_t mode) inode->i_uid = current->fsuid; inode->i_ino = fs16_to_cpu(sbi, ino); inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME_SEC; - inode->i_blocks = inode->i_blksize = 0; + inode->i_blocks = 0; memset(SYSV_I(inode)->i_data, 0, sizeof(SYSV_I(inode)->i_data)); SYSV_I(inode)->i_dir_start_lookup = 0; insert_inode_hash(inode); diff --git a/fs/sysv/inode.c b/fs/sysv/inode.c index 58b2d22142b..d63c5e48b05 100644 --- a/fs/sysv/inode.c +++ b/fs/sysv/inode.c @@ -201,7 +201,7 @@ static void sysv_read_inode(struct inode *inode) inode->i_ctime.tv_nsec = 0; inode->i_atime.tv_nsec = 0; inode->i_mtime.tv_nsec = 0; - inode->i_blocks = inode->i_blksize = 0; + inode->i_blocks = 0; si = SYSV_I(inode); for (block = 0; block < 10+1+1+1; block++) diff --git a/fs/sysv/itree.c b/fs/sysv/itree.c index 86f5f8d43d0..f2bcccd1d6f 100644 --- a/fs/sysv/itree.c +++ b/fs/sysv/itree.c @@ -465,7 +465,7 @@ static sector_t sysv_bmap(struct address_space *mapping, sector_t block) { return generic_block_bmap(mapping,block,get_block); } -struct address_space_operations sysv_aops = { +const struct address_space_operations sysv_aops = { .readpage = sysv_readpage, .writepage = sysv_writepage, .sync_page = block_sync_page, diff --git a/fs/sysv/super.c b/fs/sysv/super.c index 876639b9332..350cba5d680 100644 --- a/fs/sysv/super.c +++ b/fs/sysv/super.c @@ -369,10 +369,9 @@ static int sysv_fill_super(struct super_block *sb, void *data, int silent) if (64 != sizeof (struct sysv_inode)) panic("sysv fs: bad inode size"); - sbi = kmalloc(sizeof(struct sysv_sb_info), GFP_KERNEL); + sbi = kzalloc(sizeof(struct sysv_sb_info), GFP_KERNEL); if (!sbi) return -ENOMEM; - memset(sbi, 0, sizeof(struct sysv_sb_info)); sbi->s_sb = sb; sbi->s_block_base = 0; @@ -453,10 +452,9 @@ static int v7_fill_super(struct super_block *sb, void *data, int silent) if (64 != sizeof (struct sysv_inode)) panic("sysv fs: bad i-node size"); - sbi = kmalloc(sizeof(struct sysv_sb_info), GFP_KERNEL); + sbi = kzalloc(sizeof(struct sysv_sb_info), GFP_KERNEL); if (!sbi) return -ENOMEM; - memset(sbi, 0, sizeof(struct sysv_sb_info)); sbi->s_sb = sb; sbi->s_block_base = 0; diff --git a/fs/sysv/sysv.h b/fs/sysv/sysv.h index 393a480e4de..9dcc8212093 100644 --- a/fs/sysv/sysv.h +++ b/fs/sysv/sysv.h @@ -161,7 +161,7 @@ extern struct inode_operations sysv_dir_inode_operations; extern struct inode_operations sysv_fast_symlink_inode_operations; extern const struct file_operations sysv_file_operations; extern const struct file_operations sysv_dir_operations; -extern struct address_space_operations sysv_aops; +extern const struct address_space_operations sysv_aops; extern struct super_operations sysv_sops; extern struct dentry_operations sysv_dentry_operations; diff --git a/fs/udf/file.c b/fs/udf/file.c index e34b00e303f..a59e5f33daf 100644 --- a/fs/udf/file.c +++ b/fs/udf/file.c @@ -95,7 +95,7 @@ static int udf_adinicb_commit_write(struct file *file, struct page *page, unsign return 0; } -struct address_space_operations udf_adinicb_aops = { +const struct address_space_operations udf_adinicb_aops = { .readpage = udf_adinicb_readpage, .writepage = udf_adinicb_writepage, .sync_page = block_sync_page, diff --git a/fs/udf/ialloc.c b/fs/udf/ialloc.c index 3873c672cb4..8206983f2eb 100644 --- a/fs/udf/ialloc.c +++ b/fs/udf/ialloc.c @@ -75,6 +75,12 @@ struct inode * udf_new_inode (struct inode *dir, int mode, int * err) } *err = -ENOSPC; + UDF_I_UNIQUE(inode) = 0; + UDF_I_LENEXTENTS(inode) = 0; + UDF_I_NEXT_ALLOC_BLOCK(inode) = 0; + UDF_I_NEXT_ALLOC_GOAL(inode) = 0; + UDF_I_STRAT4096(inode) = 0; + block = udf_new_block(dir->i_sb, NULL, UDF_I_LOCATION(dir).partitionReferenceNum, start, err); if (*err) @@ -84,11 +90,6 @@ struct inode * udf_new_inode (struct inode *dir, int mode, int * err) } mutex_lock(&sbi->s_alloc_mutex); - UDF_I_UNIQUE(inode) = 0; - UDF_I_LENEXTENTS(inode) = 0; - UDF_I_NEXT_ALLOC_BLOCK(inode) = 0; - UDF_I_NEXT_ALLOC_GOAL(inode) = 0; - UDF_I_STRAT4096(inode) = 0; if (UDF_SB_LVIDBH(sb)) { struct logicalVolHeaderDesc *lvhd; @@ -120,7 +121,6 @@ struct inode * udf_new_inode (struct inode *dir, int mode, int * err) UDF_I_LOCATION(inode).logicalBlockNum = block; UDF_I_LOCATION(inode).partitionReferenceNum = UDF_I_LOCATION(dir).partitionReferenceNum; inode->i_ino = udf_get_lb_pblock(sb, UDF_I_LOCATION(inode), 0); - inode->i_blksize = PAGE_SIZE; inode->i_blocks = 0; UDF_I_LENEATTR(inode) = 0; UDF_I_LENALLOC(inode) = 0; @@ -129,14 +129,12 @@ struct inode * udf_new_inode (struct inode *dir, int mode, int * err) { UDF_I_EFE(inode) = 1; UDF_UPDATE_UDFREV(inode->i_sb, UDF_VERS_USE_EXTENDED_FE); - UDF_I_DATA(inode) = kmalloc(inode->i_sb->s_blocksize - sizeof(struct extendedFileEntry), GFP_KERNEL); - memset(UDF_I_DATA(inode), 0x00, inode->i_sb->s_blocksize - sizeof(struct extendedFileEntry)); + UDF_I_DATA(inode) = kzalloc(inode->i_sb->s_blocksize - sizeof(struct extendedFileEntry), GFP_KERNEL); } else { UDF_I_EFE(inode) = 0; - UDF_I_DATA(inode) = kmalloc(inode->i_sb->s_blocksize - sizeof(struct fileEntry), GFP_KERNEL); - memset(UDF_I_DATA(inode), 0x00, inode->i_sb->s_blocksize - sizeof(struct fileEntry)); + UDF_I_DATA(inode) = kzalloc(inode->i_sb->s_blocksize - sizeof(struct fileEntry), GFP_KERNEL); } if (UDF_QUERY_FLAG(inode->i_sb, UDF_FLAG_USE_AD_IN_ICB)) UDF_I_ALLOCTYPE(inode) = ICBTAG_FLAG_AD_IN_ICB; diff --git a/fs/udf/inode.c b/fs/udf/inode.c index 2983afd5e7f..b223b32db99 100644 --- a/fs/udf/inode.c +++ b/fs/udf/inode.c @@ -132,7 +132,7 @@ static sector_t udf_bmap(struct address_space *mapping, sector_t block) return generic_block_bmap(mapping,block,udf_get_block); } -struct address_space_operations udf_aops = { +const struct address_space_operations udf_aops = { .readpage = udf_readpage, .writepage = udf_writepage, .sync_page = block_sync_page, @@ -916,8 +916,6 @@ __udf_read_inode(struct inode *inode) * i_nlink = 1 * i_op = NULL; */ - inode->i_blksize = PAGE_SIZE; - bh = udf_read_ptagged(inode->i_sb, UDF_I_LOCATION(inode), 0, &ident); if (!bh) diff --git a/fs/udf/super.c b/fs/udf/super.c index 44fe2cb0bbb..1d3b5d2070e 100644 --- a/fs/udf/super.c +++ b/fs/udf/super.c @@ -40,7 +40,6 @@ #include "udfdecl.h" -#include <linux/config.h> #include <linux/blkdev.h> #include <linux/slab.h> #include <linux/kernel.h> @@ -116,6 +115,13 @@ static struct inode *udf_alloc_inode(struct super_block *sb) ei = (struct udf_inode_info *)kmem_cache_alloc(udf_inode_cachep, SLAB_KERNEL); if (!ei) return NULL; + + ei->i_unique = 0; + ei->i_lenExtents = 0; + ei->i_next_alloc_block = 0; + ei->i_next_alloc_goal = 0; + ei->i_strat4096 = 0; + return &ei->vfs_inode; } @@ -150,8 +156,7 @@ static int init_inodecache(void) static void destroy_inodecache(void) { - if (kmem_cache_destroy(udf_inode_cachep)) - printk(KERN_INFO "udf_inode_cache: not all structures were freed\n"); + kmem_cache_destroy(udf_inode_cachep); } /* Superblock operations */ @@ -1616,6 +1621,10 @@ static int udf_fill_super(struct super_block *sb, void *options, int silent) goto error_out; } + if (UDF_SB_PARTFLAGS(sb, UDF_SB_PARTITION(sb)) & UDF_PART_FLAG_READ_ONLY) + printk("UDF-fs: Partition marked readonly; forcing readonly mount\n"); + sb->s_flags |= MS_RDONLY; + if ( udf_find_fileset(sb, &fileset, &rootdir) ) { printk("UDF-fs: No fileset found\n"); @@ -1653,7 +1662,7 @@ static int udf_fill_super(struct super_block *sb, void *options, int silent) iput(inode); goto error_out; } - sb->s_maxbytes = MAX_LFS_FILESIZE; + sb->s_maxbytes = 1<<30; return 0; error_out: diff --git a/fs/udf/symlink.c b/fs/udf/symlink.c index 674bb40edc8..ba068a78656 100644 --- a/fs/udf/symlink.c +++ b/fs/udf/symlink.c @@ -113,6 +113,6 @@ out: /* * symlinks can't do much... */ -struct address_space_operations udf_symlink_aops = { +const struct address_space_operations udf_symlink_aops = { .readpage = udf_symlink_filler, }; diff --git a/fs/udf/truncate.c b/fs/udf/truncate.c index e1b0e8cfecb..0abd66ce36e 100644 --- a/fs/udf/truncate.c +++ b/fs/udf/truncate.c @@ -239,37 +239,51 @@ void udf_truncate_extents(struct inode * inode) { if (offset) { - extoffset -= adsize; - etype = udf_next_aext(inode, &bloc, &extoffset, &eloc, &elen, &bh, 1); - if (etype == (EXT_NOT_RECORDED_NOT_ALLOCATED >> 30)) - { - extoffset -= adsize; - elen = EXT_NOT_RECORDED_NOT_ALLOCATED | (elen + offset); - udf_write_aext(inode, bloc, &extoffset, eloc, elen, bh, 0); + /* + * OK, there is not extent covering inode->i_size and + * no extent above inode->i_size => truncate is + * extending the file by 'offset'. + */ + if ((!bh && extoffset == udf_file_entry_alloc_offset(inode)) || + (bh && extoffset == sizeof(struct allocExtDesc))) { + /* File has no extents at all! */ + memset(&eloc, 0x00, sizeof(kernel_lb_addr)); + elen = EXT_NOT_RECORDED_NOT_ALLOCATED | offset; + udf_add_aext(inode, &bloc, &extoffset, eloc, elen, &bh, 1); } - else if (etype == (EXT_NOT_RECORDED_ALLOCATED >> 30)) - { - kernel_lb_addr neloc = { 0, 0 }; + else { extoffset -= adsize; - nelen = EXT_NOT_RECORDED_NOT_ALLOCATED | - ((elen + offset + inode->i_sb->s_blocksize - 1) & - ~(inode->i_sb->s_blocksize - 1)); - udf_write_aext(inode, bloc, &extoffset, neloc, nelen, bh, 1); - udf_add_aext(inode, &bloc, &extoffset, eloc, (etype << 30) | elen, &bh, 1); - } - else - { - if (elen & (inode->i_sb->s_blocksize - 1)) + etype = udf_next_aext(inode, &bloc, &extoffset, &eloc, &elen, &bh, 1); + if (etype == (EXT_NOT_RECORDED_NOT_ALLOCATED >> 30)) + { + extoffset -= adsize; + elen = EXT_NOT_RECORDED_NOT_ALLOCATED | (elen + offset); + udf_write_aext(inode, bloc, &extoffset, eloc, elen, bh, 0); + } + else if (etype == (EXT_NOT_RECORDED_ALLOCATED >> 30)) { + kernel_lb_addr neloc = { 0, 0 }; extoffset -= adsize; - elen = EXT_RECORDED_ALLOCATED | - ((elen + inode->i_sb->s_blocksize - 1) & + nelen = EXT_NOT_RECORDED_NOT_ALLOCATED | + ((elen + offset + inode->i_sb->s_blocksize - 1) & ~(inode->i_sb->s_blocksize - 1)); - udf_write_aext(inode, bloc, &extoffset, eloc, elen, bh, 1); + udf_write_aext(inode, bloc, &extoffset, neloc, nelen, bh, 1); + udf_add_aext(inode, &bloc, &extoffset, eloc, (etype << 30) | elen, &bh, 1); + } + else + { + if (elen & (inode->i_sb->s_blocksize - 1)) + { + extoffset -= adsize; + elen = EXT_RECORDED_ALLOCATED | + ((elen + inode->i_sb->s_blocksize - 1) & + ~(inode->i_sb->s_blocksize - 1)); + udf_write_aext(inode, bloc, &extoffset, eloc, elen, bh, 1); + } + memset(&eloc, 0x00, sizeof(kernel_lb_addr)); + elen = EXT_NOT_RECORDED_NOT_ALLOCATED | offset; + udf_add_aext(inode, &bloc, &extoffset, eloc, elen, &bh, 1); } - memset(&eloc, 0x00, sizeof(kernel_lb_addr)); - elen = EXT_NOT_RECORDED_NOT_ALLOCATED | offset; - udf_add_aext(inode, &bloc, &extoffset, eloc, elen, &bh, 1); } } } diff --git a/fs/udf/udfdecl.h b/fs/udf/udfdecl.h index 023e19ba5a2..1033b7cf293 100644 --- a/fs/udf/udfdecl.h +++ b/fs/udf/udfdecl.h @@ -6,7 +6,6 @@ #include "osta_udf.h" #include <linux/fs.h> -#include <linux/config.h> #include <linux/types.h> #include <linux/udf_fs_i.h> #include <linux/udf_fs_sb.h> @@ -47,9 +46,9 @@ extern struct inode_operations udf_dir_inode_operations; extern const struct file_operations udf_dir_operations; extern struct inode_operations udf_file_inode_operations; extern const struct file_operations udf_file_operations; -extern struct address_space_operations udf_aops; -extern struct address_space_operations udf_adinicb_aops; -extern struct address_space_operations udf_symlink_aops; +extern const struct address_space_operations udf_aops; +extern const struct address_space_operations udf_adinicb_aops; +extern const struct address_space_operations udf_symlink_aops; struct udf_fileident_bh { diff --git a/fs/ufs/balloc.c b/fs/ufs/balloc.c index 3ada9dcf55b..b8238147577 100644 --- a/fs/ufs/balloc.c +++ b/fs/ufs/balloc.c @@ -21,14 +21,6 @@ #include "swab.h" #include "util.h" -#undef UFS_BALLOC_DEBUG - -#ifdef UFS_BALLOC_DEBUG -#define UFSD(x) printk("(%s, %d), %s:", __FILE__, __LINE__, __FUNCTION__); printk x; -#else -#define UFSD(x) -#endif - static unsigned ufs_add_fragments (struct inode *, unsigned, unsigned, unsigned, int *); static unsigned ufs_alloc_fragments (struct inode *, unsigned, unsigned, unsigned, int *); static unsigned ufs_alloccg_block (struct inode *, struct ufs_cg_private_info *, unsigned, int *); @@ -39,7 +31,8 @@ static void ufs_clusteracct(struct super_block *, struct ufs_cg_private_info *, /* * Free 'count' fragments from fragment number 'fragment' */ -void ufs_free_fragments (struct inode * inode, unsigned fragment, unsigned count) { +void ufs_free_fragments(struct inode *inode, unsigned fragment, unsigned count) +{ struct super_block * sb; struct ufs_sb_private_info * uspi; struct ufs_super_block_first * usb1; @@ -51,7 +44,7 @@ void ufs_free_fragments (struct inode * inode, unsigned fragment, unsigned count uspi = UFS_SB(sb)->s_uspi; usb1 = ubh_get_usb_first(uspi); - UFSD(("ENTER, fragment %u, count %u\n", fragment, count)) + UFSD("ENTER, fragment %u, count %u\n", fragment, count); if (ufs_fragnum(fragment) + count > uspi->s_fpg) ufs_error (sb, "ufs_free_fragments", "internal error"); @@ -68,7 +61,7 @@ void ufs_free_fragments (struct inode * inode, unsigned fragment, unsigned count ucpi = ufs_load_cylinder (sb, cgno); if (!ucpi) goto failed; - ucg = ubh_get_ucg (UCPI_UBH); + ucg = ubh_get_ucg (UCPI_UBH(ucpi)); if (!ufs_cg_chkmagic(sb, ucg)) { ufs_panic (sb, "ufs_free_fragments", "internal error, bad magic number on cg %u", cgno); goto failed; @@ -76,11 +69,11 @@ void ufs_free_fragments (struct inode * inode, unsigned fragment, unsigned count end_bit = bit + count; bbase = ufs_blknum (bit); - blkmap = ubh_blkmap (UCPI_UBH, ucpi->c_freeoff, bbase); + blkmap = ubh_blkmap (UCPI_UBH(ucpi), ucpi->c_freeoff, bbase); ufs_fragacct (sb, blkmap, ucg->cg_frsum, -1); for (i = bit; i < end_bit; i++) { - if (ubh_isclr (UCPI_UBH, ucpi->c_freeoff, i)) - ubh_setbit (UCPI_UBH, ucpi->c_freeoff, i); + if (ubh_isclr (UCPI_UBH(ucpi), ucpi->c_freeoff, i)) + ubh_setbit (UCPI_UBH(ucpi), ucpi->c_freeoff, i); else ufs_error (sb, "ufs_free_fragments", "bit already cleared for fragment %u", i); @@ -90,51 +83,52 @@ void ufs_free_fragments (struct inode * inode, unsigned fragment, unsigned count fs32_add(sb, &ucg->cg_cs.cs_nffree, count); - fs32_add(sb, &usb1->fs_cstotal.cs_nffree, count); + uspi->cs_total.cs_nffree += count; fs32_add(sb, &UFS_SB(sb)->fs_cs(cgno).cs_nffree, count); - blkmap = ubh_blkmap (UCPI_UBH, ucpi->c_freeoff, bbase); + blkmap = ubh_blkmap (UCPI_UBH(ucpi), ucpi->c_freeoff, bbase); ufs_fragacct(sb, blkmap, ucg->cg_frsum, 1); /* * Trying to reassemble free fragments into block */ blkno = ufs_fragstoblks (bbase); - if (ubh_isblockset(UCPI_UBH, ucpi->c_freeoff, blkno)) { + if (ubh_isblockset(UCPI_UBH(ucpi), ucpi->c_freeoff, blkno)) { fs32_sub(sb, &ucg->cg_cs.cs_nffree, uspi->s_fpb); - fs32_sub(sb, &usb1->fs_cstotal.cs_nffree, uspi->s_fpb); + uspi->cs_total.cs_nffree -= uspi->s_fpb; fs32_sub(sb, &UFS_SB(sb)->fs_cs(cgno).cs_nffree, uspi->s_fpb); if ((UFS_SB(sb)->s_flags & UFS_CG_MASK) == UFS_CG_44BSD) ufs_clusteracct (sb, ucpi, blkno, 1); fs32_add(sb, &ucg->cg_cs.cs_nbfree, 1); - fs32_add(sb, &usb1->fs_cstotal.cs_nbfree, 1); + uspi->cs_total.cs_nbfree++; fs32_add(sb, &UFS_SB(sb)->fs_cs(cgno).cs_nbfree, 1); cylno = ufs_cbtocylno (bbase); fs16_add(sb, &ubh_cg_blks(ucpi, cylno, ufs_cbtorpos(bbase)), 1); fs32_add(sb, &ubh_cg_blktot(ucpi, cylno), 1); } - ubh_mark_buffer_dirty (USPI_UBH); - ubh_mark_buffer_dirty (UCPI_UBH); + ubh_mark_buffer_dirty (USPI_UBH(uspi)); + ubh_mark_buffer_dirty (UCPI_UBH(ucpi)); if (sb->s_flags & MS_SYNCHRONOUS) { - ubh_ll_rw_block (SWRITE, 1, (struct ufs_buffer_head **)&ucpi); - ubh_wait_on_buffer (UCPI_UBH); + ubh_ll_rw_block(SWRITE, UCPI_UBH(ucpi)); + ubh_wait_on_buffer (UCPI_UBH(ucpi)); } sb->s_dirt = 1; unlock_super (sb); - UFSD(("EXIT\n")) + UFSD("EXIT\n"); return; failed: unlock_super (sb); - UFSD(("EXIT (FAILED)\n")) + UFSD("EXIT (FAILED)\n"); return; } /* * Free 'count' fragments from fragment number 'fragment' (free whole blocks) */ -void ufs_free_blocks (struct inode * inode, unsigned fragment, unsigned count) { +void ufs_free_blocks(struct inode *inode, unsigned fragment, unsigned count) +{ struct super_block * sb; struct ufs_sb_private_info * uspi; struct ufs_super_block_first * usb1; @@ -146,7 +140,7 @@ void ufs_free_blocks (struct inode * inode, unsigned fragment, unsigned count) { uspi = UFS_SB(sb)->s_uspi; usb1 = ubh_get_usb_first(uspi); - UFSD(("ENTER, fragment %u, count %u\n", fragment, count)) + UFSD("ENTER, fragment %u, count %u\n", fragment, count); if ((fragment & uspi->s_fpbmask) || (count & uspi->s_fpbmask)) { ufs_error (sb, "ufs_free_blocks", "internal error, " @@ -162,7 +156,7 @@ do_more: bit = ufs_dtogd (fragment); if (cgno >= uspi->s_ncg) { ufs_panic (sb, "ufs_free_blocks", "freeing blocks are outside device"); - goto failed; + goto failed_unlock; } end_bit = bit + count; if (end_bit > uspi->s_fpg) { @@ -173,36 +167,36 @@ do_more: ucpi = ufs_load_cylinder (sb, cgno); if (!ucpi) - goto failed; - ucg = ubh_get_ucg (UCPI_UBH); + goto failed_unlock; + ucg = ubh_get_ucg (UCPI_UBH(ucpi)); if (!ufs_cg_chkmagic(sb, ucg)) { ufs_panic (sb, "ufs_free_blocks", "internal error, bad magic number on cg %u", cgno); - goto failed; + goto failed_unlock; } for (i = bit; i < end_bit; i += uspi->s_fpb) { blkno = ufs_fragstoblks(i); - if (ubh_isblockset(UCPI_UBH, ucpi->c_freeoff, blkno)) { + if (ubh_isblockset(UCPI_UBH(ucpi), ucpi->c_freeoff, blkno)) { ufs_error(sb, "ufs_free_blocks", "freeing free fragment"); } - ubh_setblock(UCPI_UBH, ucpi->c_freeoff, blkno); + ubh_setblock(UCPI_UBH(ucpi), ucpi->c_freeoff, blkno); if ((UFS_SB(sb)->s_flags & UFS_CG_MASK) == UFS_CG_44BSD) ufs_clusteracct (sb, ucpi, blkno, 1); DQUOT_FREE_BLOCK(inode, uspi->s_fpb); fs32_add(sb, &ucg->cg_cs.cs_nbfree, 1); - fs32_add(sb, &usb1->fs_cstotal.cs_nbfree, 1); + uspi->cs_total.cs_nbfree++; fs32_add(sb, &UFS_SB(sb)->fs_cs(cgno).cs_nbfree, 1); cylno = ufs_cbtocylno(i); fs16_add(sb, &ubh_cg_blks(ucpi, cylno, ufs_cbtorpos(i)), 1); fs32_add(sb, &ubh_cg_blktot(ucpi, cylno), 1); } - ubh_mark_buffer_dirty (USPI_UBH); - ubh_mark_buffer_dirty (UCPI_UBH); + ubh_mark_buffer_dirty (USPI_UBH(uspi)); + ubh_mark_buffer_dirty (UCPI_UBH(ucpi)); if (sb->s_flags & MS_SYNCHRONOUS) { - ubh_ll_rw_block (SWRITE, 1, (struct ufs_buffer_head **)&ucpi); - ubh_wait_on_buffer (UCPI_UBH); + ubh_ll_rw_block(SWRITE, UCPI_UBH(ucpi)); + ubh_wait_on_buffer (UCPI_UBH(ucpi)); } if (overflow) { @@ -213,38 +207,83 @@ do_more: sb->s_dirt = 1; unlock_super (sb); - UFSD(("EXIT\n")) + UFSD("EXIT\n"); return; -failed: +failed_unlock: unlock_super (sb); - UFSD(("EXIT (FAILED)\n")) +failed: + UFSD("EXIT (FAILED)\n"); return; } +/* + * Modify inode page cache in such way: + * have - blocks with b_blocknr equal to oldb...oldb+count-1 + * get - blocks with b_blocknr equal to newb...newb+count-1 + * also we suppose that oldb...oldb+count-1 blocks + * situated at the end of file. + * + * We can come here from ufs_writepage or ufs_prepare_write, + * locked_page is argument of these functions, so we already lock it. + */ +static void ufs_change_blocknr(struct inode *inode, unsigned int baseblk, + unsigned int count, unsigned int oldb, + unsigned int newb, struct page *locked_page) +{ + unsigned int blk_per_page = 1 << (PAGE_CACHE_SHIFT - inode->i_blkbits); + struct address_space *mapping = inode->i_mapping; + pgoff_t index, cur_index = locked_page->index; + unsigned int i, j; + struct page *page; + struct buffer_head *head, *bh; + + UFSD("ENTER, ino %lu, count %u, oldb %u, newb %u\n", + inode->i_ino, count, oldb, newb); + + BUG_ON(!PageLocked(locked_page)); + + for (i = 0; i < count; i += blk_per_page) { + index = (baseblk+i) >> (PAGE_CACHE_SHIFT - inode->i_blkbits); + + if (likely(cur_index != index)) { + page = ufs_get_locked_page(mapping, index); + if (!page || IS_ERR(page)) /* it was truncated or EIO */ + continue; + } else + page = locked_page; + + j = i; + head = page_buffers(page); + bh = head; + do { + if (likely(bh->b_blocknr == j + oldb && j < count)) { + unmap_underlying_metadata(bh->b_bdev, + bh->b_blocknr); + bh->b_blocknr = newb + j++; + mark_buffer_dirty(bh); + } + bh = bh->b_this_page; + } while (bh != head); -#define NULLIFY_FRAGMENTS \ - for (i = oldcount; i < newcount; i++) { \ - bh = sb_getblk(sb, result + i); \ - memset (bh->b_data, 0, sb->s_blocksize); \ - set_buffer_uptodate(bh); \ - mark_buffer_dirty (bh); \ - if (IS_SYNC(inode)) \ - sync_dirty_buffer(bh); \ - brelse (bh); \ - } + set_page_dirty(page); + + if (likely(cur_index != index)) + ufs_put_locked_page(page); + } + UFSD("EXIT\n"); +} -unsigned ufs_new_fragments (struct inode * inode, __fs32 * p, unsigned fragment, - unsigned goal, unsigned count, int * err ) +unsigned ufs_new_fragments(struct inode * inode, __fs32 * p, unsigned fragment, + unsigned goal, unsigned count, int * err, struct page *locked_page) { struct super_block * sb; struct ufs_sb_private_info * uspi; struct ufs_super_block_first * usb1; - struct buffer_head * bh; - unsigned cgno, oldcount, newcount, tmp, request, i, result; + unsigned cgno, oldcount, newcount, tmp, request, result; - UFSD(("ENTER, ino %lu, fragment %u, goal %u, count %u\n", inode->i_ino, fragment, goal, count)) + UFSD("ENTER, ino %lu, fragment %u, goal %u, count %u\n", inode->i_ino, fragment, goal, count); sb = inode->i_sb; uspi = UFS_SB(sb)->s_uspi; @@ -273,14 +312,14 @@ unsigned ufs_new_fragments (struct inode * inode, __fs32 * p, unsigned fragment, return (unsigned)-1; } if (fragment < UFS_I(inode)->i_lastfrag) { - UFSD(("EXIT (ALREADY ALLOCATED)\n")) + UFSD("EXIT (ALREADY ALLOCATED)\n"); unlock_super (sb); return 0; } } else { if (tmp) { - UFSD(("EXIT (ALREADY ALLOCATED)\n")) + UFSD("EXIT (ALREADY ALLOCATED)\n"); unlock_super(sb); return 0; } @@ -289,9 +328,9 @@ unsigned ufs_new_fragments (struct inode * inode, __fs32 * p, unsigned fragment, /* * There is not enough space for user on the device */ - if (!capable(CAP_SYS_RESOURCE) && ufs_freespace(usb1, UFS_MINFREE) <= 0) { + if (!capable(CAP_SYS_RESOURCE) && ufs_freespace(uspi, UFS_MINFREE) <= 0) { unlock_super (sb); - UFSD(("EXIT (FAILED)\n")) + UFSD("EXIT (FAILED)\n"); return 0; } @@ -310,12 +349,10 @@ unsigned ufs_new_fragments (struct inode * inode, __fs32 * p, unsigned fragment, if (result) { *p = cpu_to_fs32(sb, result); *err = 0; - inode->i_blocks += count << uspi->s_nspfshift; UFS_I(inode)->i_lastfrag = max_t(u32, UFS_I(inode)->i_lastfrag, fragment + count); - NULLIFY_FRAGMENTS } unlock_super(sb); - UFSD(("EXIT, result %u\n", result)) + UFSD("EXIT, result %u\n", result); return result; } @@ -325,11 +362,9 @@ unsigned ufs_new_fragments (struct inode * inode, __fs32 * p, unsigned fragment, result = ufs_add_fragments (inode, tmp, oldcount, newcount, err); if (result) { *err = 0; - inode->i_blocks += count << uspi->s_nspfshift; UFS_I(inode)->i_lastfrag = max_t(u32, UFS_I(inode)->i_lastfrag, fragment + count); - NULLIFY_FRAGMENTS unlock_super(sb); - UFSD(("EXIT, result %u\n", result)) + UFSD("EXIT, result %u\n", result); return result; } @@ -339,8 +374,8 @@ unsigned ufs_new_fragments (struct inode * inode, __fs32 * p, unsigned fragment, switch (fs32_to_cpu(sb, usb1->fs_optim)) { case UFS_OPTSPACE: request = newcount; - if (uspi->s_minfree < 5 || fs32_to_cpu(sb, usb1->fs_cstotal.cs_nffree) - > uspi->s_dsize * uspi->s_minfree / (2 * 100) ) + if (uspi->s_minfree < 5 || uspi->cs_total.cs_nffree + > uspi->s_dsize * uspi->s_minfree / (2 * 100)) break; usb1->fs_optim = cpu_to_fs32(sb, UFS_OPTTIME); break; @@ -349,7 +384,7 @@ unsigned ufs_new_fragments (struct inode * inode, __fs32 * p, unsigned fragment, case UFS_OPTTIME: request = uspi->s_fpb; - if (fs32_to_cpu(sb, usb1->fs_cstotal.cs_nffree) < uspi->s_dsize * + if (uspi->cs_total.cs_nffree < uspi->s_dsize * (uspi->s_minfree - 2) / 100) break; usb1->fs_optim = cpu_to_fs32(sb, UFS_OPTTIME); @@ -357,39 +392,22 @@ unsigned ufs_new_fragments (struct inode * inode, __fs32 * p, unsigned fragment, } result = ufs_alloc_fragments (inode, cgno, goal, request, err); if (result) { - for (i = 0; i < oldcount; i++) { - bh = sb_bread(sb, tmp + i); - if(bh) - { - clear_buffer_dirty(bh); - bh->b_blocknr = result + i; - mark_buffer_dirty (bh); - if (IS_SYNC(inode)) - sync_dirty_buffer(bh); - brelse (bh); - } - else - { - printk(KERN_ERR "ufs_new_fragments: bread fail\n"); - unlock_super(sb); - return 0; - } - } + ufs_change_blocknr(inode, fragment - oldcount, oldcount, tmp, + result, locked_page); + *p = cpu_to_fs32(sb, result); *err = 0; - inode->i_blocks += count << uspi->s_nspfshift; UFS_I(inode)->i_lastfrag = max_t(u32, UFS_I(inode)->i_lastfrag, fragment + count); - NULLIFY_FRAGMENTS unlock_super(sb); if (newcount < request) ufs_free_fragments (inode, result + newcount, request - newcount); ufs_free_fragments (inode, tmp, oldcount); - UFSD(("EXIT, result %u\n", result)) + UFSD("EXIT, result %u\n", result); return result; } unlock_super(sb); - UFSD(("EXIT (FAILED)\n")) + UFSD("EXIT (FAILED)\n"); return 0; } @@ -404,7 +422,7 @@ ufs_add_fragments (struct inode * inode, unsigned fragment, struct ufs_cylinder_group * ucg; unsigned cgno, fragno, fragoff, count, fragsize, i; - UFSD(("ENTER, fragment %u, oldcount %u, newcount %u\n", fragment, oldcount, newcount)) + UFSD("ENTER, fragment %u, oldcount %u, newcount %u\n", fragment, oldcount, newcount); sb = inode->i_sb; uspi = UFS_SB(sb)->s_uspi; @@ -419,7 +437,7 @@ ufs_add_fragments (struct inode * inode, unsigned fragment, ucpi = ufs_load_cylinder (sb, cgno); if (!ucpi) return 0; - ucg = ubh_get_ucg (UCPI_UBH); + ucg = ubh_get_ucg (UCPI_UBH(ucpi)); if (!ufs_cg_chkmagic(sb, ucg)) { ufs_panic (sb, "ufs_add_fragments", "internal error, bad magic number on cg %u", cgno); @@ -429,14 +447,14 @@ ufs_add_fragments (struct inode * inode, unsigned fragment, fragno = ufs_dtogd (fragment); fragoff = ufs_fragnum (fragno); for (i = oldcount; i < newcount; i++) - if (ubh_isclr (UCPI_UBH, ucpi->c_freeoff, fragno + i)) + if (ubh_isclr (UCPI_UBH(ucpi), ucpi->c_freeoff, fragno + i)) return 0; /* * Block can be extended */ ucg->cg_time = cpu_to_fs32(sb, get_seconds()); for (i = newcount; i < (uspi->s_fpb - fragoff); i++) - if (ubh_isclr (UCPI_UBH, ucpi->c_freeoff, fragno + i)) + if (ubh_isclr (UCPI_UBH(ucpi), ucpi->c_freeoff, fragno + i)) break; fragsize = i - oldcount; if (!fs32_to_cpu(sb, ucg->cg_frsum[fragsize])) @@ -446,7 +464,7 @@ ufs_add_fragments (struct inode * inode, unsigned fragment, if (fragsize != count) fs32_add(sb, &ucg->cg_frsum[fragsize - count], 1); for (i = oldcount; i < newcount; i++) - ubh_clrbit (UCPI_UBH, ucpi->c_freeoff, fragno + i); + ubh_clrbit (UCPI_UBH(ucpi), ucpi->c_freeoff, fragno + i); if(DQUOT_ALLOC_BLOCK(inode, count)) { *err = -EDQUOT; return 0; @@ -454,17 +472,17 @@ ufs_add_fragments (struct inode * inode, unsigned fragment, fs32_sub(sb, &ucg->cg_cs.cs_nffree, count); fs32_sub(sb, &UFS_SB(sb)->fs_cs(cgno).cs_nffree, count); - fs32_sub(sb, &usb1->fs_cstotal.cs_nffree, count); + uspi->cs_total.cs_nffree -= count; - ubh_mark_buffer_dirty (USPI_UBH); - ubh_mark_buffer_dirty (UCPI_UBH); + ubh_mark_buffer_dirty (USPI_UBH(uspi)); + ubh_mark_buffer_dirty (UCPI_UBH(ucpi)); if (sb->s_flags & MS_SYNCHRONOUS) { - ubh_ll_rw_block (SWRITE, 1, (struct ufs_buffer_head **)&ucpi); - ubh_wait_on_buffer (UCPI_UBH); + ubh_ll_rw_block(SWRITE, UCPI_UBH(ucpi)); + ubh_wait_on_buffer (UCPI_UBH(ucpi)); } sb->s_dirt = 1; - UFSD(("EXIT, fragment %u\n", fragment)) + UFSD("EXIT, fragment %u\n", fragment); return fragment; } @@ -487,7 +505,7 @@ static unsigned ufs_alloc_fragments (struct inode * inode, unsigned cgno, struct ufs_cylinder_group * ucg; unsigned oldcg, i, j, k, result, allocsize; - UFSD(("ENTER, ino %lu, cgno %u, goal %u, count %u\n", inode->i_ino, cgno, goal, count)) + UFSD("ENTER, ino %lu, cgno %u, goal %u, count %u\n", inode->i_ino, cgno, goal, count); sb = inode->i_sb; uspi = UFS_SB(sb)->s_uspi; @@ -521,14 +539,14 @@ static unsigned ufs_alloc_fragments (struct inode * inode, unsigned cgno, UFS_TEST_FREE_SPACE_CG } - UFSD(("EXIT (FAILED)\n")) + UFSD("EXIT (FAILED)\n"); return 0; cg_found: ucpi = ufs_load_cylinder (sb, cgno); if (!ucpi) return 0; - ucg = ubh_get_ucg (UCPI_UBH); + ucg = ubh_get_ucg (UCPI_UBH(ucpi)); if (!ufs_cg_chkmagic(sb, ucg)) ufs_panic (sb, "ufs_alloc_fragments", "internal error, bad magic number on cg %u", cgno); @@ -551,12 +569,12 @@ cg_found: return 0; goal = ufs_dtogd (result); for (i = count; i < uspi->s_fpb; i++) - ubh_setbit (UCPI_UBH, ucpi->c_freeoff, goal + i); + ubh_setbit (UCPI_UBH(ucpi), ucpi->c_freeoff, goal + i); i = uspi->s_fpb - count; DQUOT_FREE_BLOCK(inode, i); fs32_add(sb, &ucg->cg_cs.cs_nffree, i); - fs32_add(sb, &usb1->fs_cstotal.cs_nffree, i); + uspi->cs_total.cs_nffree += i; fs32_add(sb, &UFS_SB(sb)->fs_cs(cgno).cs_nffree, i); fs32_add(sb, &ucg->cg_frsum[i], 1); goto succed; @@ -570,10 +588,10 @@ cg_found: return 0; } for (i = 0; i < count; i++) - ubh_clrbit (UCPI_UBH, ucpi->c_freeoff, result + i); + ubh_clrbit (UCPI_UBH(ucpi), ucpi->c_freeoff, result + i); fs32_sub(sb, &ucg->cg_cs.cs_nffree, count); - fs32_sub(sb, &usb1->fs_cstotal.cs_nffree, count); + uspi->cs_total.cs_nffree -= count; fs32_sub(sb, &UFS_SB(sb)->fs_cs(cgno).cs_nffree, count); fs32_sub(sb, &ucg->cg_frsum[allocsize], 1); @@ -581,16 +599,16 @@ cg_found: fs32_add(sb, &ucg->cg_frsum[allocsize - count], 1); succed: - ubh_mark_buffer_dirty (USPI_UBH); - ubh_mark_buffer_dirty (UCPI_UBH); + ubh_mark_buffer_dirty (USPI_UBH(uspi)); + ubh_mark_buffer_dirty (UCPI_UBH(ucpi)); if (sb->s_flags & MS_SYNCHRONOUS) { - ubh_ll_rw_block (SWRITE, 1, (struct ufs_buffer_head **)&ucpi); - ubh_wait_on_buffer (UCPI_UBH); + ubh_ll_rw_block(SWRITE, UCPI_UBH(ucpi)); + ubh_wait_on_buffer (UCPI_UBH(ucpi)); } sb->s_dirt = 1; result += cgno * uspi->s_fpg; - UFSD(("EXIT3, result %u\n", result)) + UFSD("EXIT3, result %u\n", result); return result; } @@ -603,12 +621,12 @@ static unsigned ufs_alloccg_block (struct inode * inode, struct ufs_cylinder_group * ucg; unsigned result, cylno, blkno; - UFSD(("ENTER, goal %u\n", goal)) + UFSD("ENTER, goal %u\n", goal); sb = inode->i_sb; uspi = UFS_SB(sb)->s_uspi; usb1 = ubh_get_usb_first(uspi); - ucg = ubh_get_ucg(UCPI_UBH); + ucg = ubh_get_ucg(UCPI_UBH(ucpi)); if (goal == 0) { goal = ucpi->c_rotor; @@ -620,7 +638,7 @@ static unsigned ufs_alloccg_block (struct inode * inode, /* * If the requested block is available, use it. */ - if (ubh_isblockset(UCPI_UBH, ucpi->c_freeoff, ufs_fragstoblks(goal))) { + if (ubh_isblockset(UCPI_UBH(ucpi), ucpi->c_freeoff, ufs_fragstoblks(goal))) { result = goal; goto gotit; } @@ -632,7 +650,7 @@ norot: ucpi->c_rotor = result; gotit: blkno = ufs_fragstoblks(result); - ubh_clrblock (UCPI_UBH, ucpi->c_freeoff, blkno); + ubh_clrblock (UCPI_UBH(ucpi), ucpi->c_freeoff, blkno); if ((UFS_SB(sb)->s_flags & UFS_CG_MASK) == UFS_CG_44BSD) ufs_clusteracct (sb, ucpi, blkno, -1); if(DQUOT_ALLOC_BLOCK(inode, uspi->s_fpb)) { @@ -641,31 +659,76 @@ gotit: } fs32_sub(sb, &ucg->cg_cs.cs_nbfree, 1); - fs32_sub(sb, &usb1->fs_cstotal.cs_nbfree, 1); + uspi->cs_total.cs_nbfree--; fs32_sub(sb, &UFS_SB(sb)->fs_cs(ucpi->c_cgx).cs_nbfree, 1); cylno = ufs_cbtocylno(result); fs16_sub(sb, &ubh_cg_blks(ucpi, cylno, ufs_cbtorpos(result)), 1); fs32_sub(sb, &ubh_cg_blktot(ucpi, cylno), 1); - UFSD(("EXIT, result %u\n", result)) + UFSD("EXIT, result %u\n", result); return result; } -static unsigned ufs_bitmap_search (struct super_block * sb, - struct ufs_cg_private_info * ucpi, unsigned goal, unsigned count) +static unsigned ubh_scanc(struct ufs_sb_private_info *uspi, + struct ufs_buffer_head *ubh, + unsigned begin, unsigned size, + unsigned char *table, unsigned char mask) { - struct ufs_sb_private_info * uspi; - struct ufs_super_block_first * usb1; - struct ufs_cylinder_group * ucg; - unsigned start, length, location, result; - unsigned possition, fragsize, blockmap, mask; - - UFSD(("ENTER, cg %u, goal %u, count %u\n", ucpi->c_cgx, goal, count)) + unsigned rest, offset; + unsigned char *cp; + + + offset = begin & ~uspi->s_fmask; + begin >>= uspi->s_fshift; + for (;;) { + if ((offset + size) < uspi->s_fsize) + rest = size; + else + rest = uspi->s_fsize - offset; + size -= rest; + cp = ubh->bh[begin]->b_data + offset; + while ((table[*cp++] & mask) == 0 && --rest) + ; + if (rest || !size) + break; + begin++; + offset = 0; + } + return (size + rest); +} + +/* + * Find a block of the specified size in the specified cylinder group. + * @sp: pointer to super block + * @ucpi: pointer to cylinder group info + * @goal: near which block we want find new one + * @count: specified size + */ +static unsigned ufs_bitmap_search(struct super_block *sb, + struct ufs_cg_private_info *ucpi, + unsigned goal, unsigned count) +{ + /* + * Bit patterns for identifying fragments in the block map + * used as ((map & mask_arr) == want_arr) + */ + static const int mask_arr[9] = { + 0x3, 0x7, 0xf, 0x1f, 0x3f, 0x7f, 0xff, 0x1ff, 0x3ff + }; + static const int want_arr[9] = { + 0x0, 0x2, 0x6, 0xe, 0x1e, 0x3e, 0x7e, 0xfe, 0x1fe + }; + struct ufs_sb_private_info *uspi = UFS_SB(sb)->s_uspi; + struct ufs_super_block_first *usb1; + struct ufs_cylinder_group *ucg; + unsigned start, length, loc, result; + unsigned pos, want, blockmap, mask, end; + + UFSD("ENTER, cg %u, goal %u, count %u\n", ucpi->c_cgx, goal, count); - uspi = UFS_SB(sb)->s_uspi; usb1 = ubh_get_usb_first (uspi); - ucg = ubh_get_ucg(UCPI_UBH); + ucg = ubh_get_ucg(UCPI_UBH(ucpi)); if (goal) start = ufs_dtogd(goal) >> 3; @@ -673,53 +736,50 @@ static unsigned ufs_bitmap_search (struct super_block * sb, start = ucpi->c_frotor >> 3; length = ((uspi->s_fpg + 7) >> 3) - start; - location = ubh_scanc(UCPI_UBH, ucpi->c_freeoff + start, length, + loc = ubh_scanc(uspi, UCPI_UBH(ucpi), ucpi->c_freeoff + start, length, (uspi->s_fpb == 8) ? ufs_fragtable_8fpb : ufs_fragtable_other, 1 << (count - 1 + (uspi->s_fpb & 7))); - if (location == 0) { + if (loc == 0) { length = start + 1; - location = ubh_scanc(UCPI_UBH, ucpi->c_freeoff, length, - (uspi->s_fpb == 8) ? ufs_fragtable_8fpb : ufs_fragtable_other, - 1 << (count - 1 + (uspi->s_fpb & 7))); - if (location == 0) { - ufs_error (sb, "ufs_bitmap_search", - "bitmap corrupted on cg %u, start %u, length %u, count %u, freeoff %u\n", - ucpi->c_cgx, start, length, count, ucpi->c_freeoff); + loc = ubh_scanc(uspi, UCPI_UBH(ucpi), ucpi->c_freeoff, length, + (uspi->s_fpb == 8) ? ufs_fragtable_8fpb : + ufs_fragtable_other, + 1 << (count - 1 + (uspi->s_fpb & 7))); + if (loc == 0) { + ufs_error(sb, "ufs_bitmap_search", + "bitmap corrupted on cg %u, start %u," + " length %u, count %u, freeoff %u\n", + ucpi->c_cgx, start, length, count, + ucpi->c_freeoff); return (unsigned)-1; } start = 0; } - result = (start + length - location) << 3; + result = (start + length - loc) << 3; ucpi->c_frotor = result; /* * found the byte in the map */ - blockmap = ubh_blkmap(UCPI_UBH, ucpi->c_freeoff, result); - fragsize = 0; - for (possition = 0, mask = 1; possition < 8; possition++, mask <<= 1) { - if (blockmap & mask) { - if (!(possition & uspi->s_fpbmask)) - fragsize = 1; - else - fragsize++; - } - else { - if (fragsize == count) { - result += possition - count; - UFSD(("EXIT, result %u\n", result)) - return result; - } - fragsize = 0; - } - } - if (fragsize == count) { - result += possition - count; - UFSD(("EXIT, result %u\n", result)) - return result; - } - ufs_error (sb, "ufs_bitmap_search", "block not in map on cg %u\n", ucpi->c_cgx); - UFSD(("EXIT (FAILED)\n")) + + for (end = result + 8; result < end; result += uspi->s_fpb) { + blockmap = ubh_blkmap(UCPI_UBH(ucpi), ucpi->c_freeoff, result); + blockmap <<= 1; + mask = mask_arr[count]; + want = want_arr[count]; + for (pos = 0; pos <= uspi->s_fpb - count; pos++) { + if ((blockmap & mask) == want) { + UFSD("EXIT, result %u\n", result); + return result + pos; + } + mask <<= 1; + want <<= 1; + } + } + + ufs_error(sb, "ufs_bitmap_search", "block not in map on cg %u\n", + ucpi->c_cgx); + UFSD("EXIT (FAILED)\n"); return (unsigned)-1; } @@ -734,9 +794,9 @@ static void ufs_clusteracct(struct super_block * sb, return; if (cnt > 0) - ubh_setbit(UCPI_UBH, ucpi->c_clusteroff, blkno); + ubh_setbit(UCPI_UBH(ucpi), ucpi->c_clusteroff, blkno); else - ubh_clrbit(UCPI_UBH, ucpi->c_clusteroff, blkno); + ubh_clrbit(UCPI_UBH(ucpi), ucpi->c_clusteroff, blkno); /* * Find the size of the cluster going forward. @@ -745,7 +805,7 @@ static void ufs_clusteracct(struct super_block * sb, end = start + uspi->s_contigsumsize; if ( end >= ucpi->c_nclusterblks) end = ucpi->c_nclusterblks; - i = ubh_find_next_zero_bit (UCPI_UBH, ucpi->c_clusteroff, end, start); + i = ubh_find_next_zero_bit (UCPI_UBH(ucpi), ucpi->c_clusteroff, end, start); if (i > end) i = end; forw = i - start; @@ -757,7 +817,7 @@ static void ufs_clusteracct(struct super_block * sb, end = start - uspi->s_contigsumsize; if (end < 0 ) end = -1; - i = ubh_find_last_zero_bit (UCPI_UBH, ucpi->c_clusteroff, start, end); + i = ubh_find_last_zero_bit (UCPI_UBH(ucpi), ucpi->c_clusteroff, start, end); if ( i < end) i = end; back = start - i; @@ -769,11 +829,11 @@ static void ufs_clusteracct(struct super_block * sb, i = back + forw + 1; if (i > uspi->s_contigsumsize) i = uspi->s_contigsumsize; - fs32_add(sb, (__fs32*)ubh_get_addr(UCPI_UBH, ucpi->c_clustersumoff + (i << 2)), cnt); + fs32_add(sb, (__fs32*)ubh_get_addr(UCPI_UBH(ucpi), ucpi->c_clustersumoff + (i << 2)), cnt); if (back > 0) - fs32_sub(sb, (__fs32*)ubh_get_addr(UCPI_UBH, ucpi->c_clustersumoff + (back << 2)), cnt); + fs32_sub(sb, (__fs32*)ubh_get_addr(UCPI_UBH(ucpi), ucpi->c_clustersumoff + (back << 2)), cnt); if (forw > 0) - fs32_sub(sb, (__fs32*)ubh_get_addr(UCPI_UBH, ucpi->c_clustersumoff + (forw << 2)), cnt); + fs32_sub(sb, (__fs32*)ubh_get_addr(UCPI_UBH(ucpi), ucpi->c_clustersumoff + (forw << 2)), cnt); } diff --git a/fs/ufs/cylinder.c b/fs/ufs/cylinder.c index 14abb8b835f..09c39e5e638 100644 --- a/fs/ufs/cylinder.c +++ b/fs/ufs/cylinder.c @@ -20,15 +20,6 @@ #include "swab.h" #include "util.h" -#undef UFS_CYLINDER_DEBUG - -#ifdef UFS_CYLINDER_DEBUG -#define UFSD(x) printk("(%s, %d), %s:", __FILE__, __LINE__, __FUNCTION__); printk x; -#else -#define UFSD(x) -#endif - - /* * Read cylinder group into cache. The memory space for ufs_cg_private_info * structure is already allocated during ufs_read_super. @@ -42,19 +33,19 @@ static void ufs_read_cylinder (struct super_block * sb, struct ufs_cylinder_group * ucg; unsigned i, j; - UFSD(("ENTER, cgno %u, bitmap_nr %u\n", cgno, bitmap_nr)) + UFSD("ENTER, cgno %u, bitmap_nr %u\n", cgno, bitmap_nr); uspi = sbi->s_uspi; ucpi = sbi->s_ucpi[bitmap_nr]; ucg = (struct ufs_cylinder_group *)sbi->s_ucg[cgno]->b_data; - UCPI_UBH->fragment = ufs_cgcmin(cgno); - UCPI_UBH->count = uspi->s_cgsize >> sb->s_blocksize_bits; + UCPI_UBH(ucpi)->fragment = ufs_cgcmin(cgno); + UCPI_UBH(ucpi)->count = uspi->s_cgsize >> sb->s_blocksize_bits; /* * We have already the first fragment of cylinder group block in buffer */ - UCPI_UBH->bh[0] = sbi->s_ucg[cgno]; - for (i = 1; i < UCPI_UBH->count; i++) - if (!(UCPI_UBH->bh[i] = sb_bread(sb, UCPI_UBH->fragment + i))) + UCPI_UBH(ucpi)->bh[0] = sbi->s_ucg[cgno]; + for (i = 1; i < UCPI_UBH(ucpi)->count; i++) + if (!(UCPI_UBH(ucpi)->bh[i] = sb_bread(sb, UCPI_UBH(ucpi)->fragment + i))) goto failed; sbi->s_cgno[bitmap_nr] = cgno; @@ -73,7 +64,7 @@ static void ufs_read_cylinder (struct super_block * sb, ucpi->c_clustersumoff = fs32_to_cpu(sb, ucg->cg_u.cg_44.cg_clustersumoff); ucpi->c_clusteroff = fs32_to_cpu(sb, ucg->cg_u.cg_44.cg_clusteroff); ucpi->c_nclusterblks = fs32_to_cpu(sb, ucg->cg_u.cg_44.cg_nclusterblks); - UFSD(("EXIT\n")) + UFSD("EXIT\n"); return; failed: @@ -95,15 +86,15 @@ void ufs_put_cylinder (struct super_block * sb, unsigned bitmap_nr) struct ufs_cylinder_group * ucg; unsigned i; - UFSD(("ENTER, bitmap_nr %u\n", bitmap_nr)) + UFSD("ENTER, bitmap_nr %u\n", bitmap_nr); uspi = sbi->s_uspi; if (sbi->s_cgno[bitmap_nr] == UFS_CGNO_EMPTY) { - UFSD(("EXIT\n")) + UFSD("EXIT\n"); return; } ucpi = sbi->s_ucpi[bitmap_nr]; - ucg = ubh_get_ucg(UCPI_UBH); + ucg = ubh_get_ucg(UCPI_UBH(ucpi)); if (uspi->s_ncg > UFS_MAX_GROUP_LOADED && bitmap_nr >= sbi->s_cg_loaded) { ufs_panic (sb, "ufs_put_cylinder", "internal error"); @@ -116,13 +107,13 @@ void ufs_put_cylinder (struct super_block * sb, unsigned bitmap_nr) ucg->cg_rotor = cpu_to_fs32(sb, ucpi->c_rotor); ucg->cg_frotor = cpu_to_fs32(sb, ucpi->c_frotor); ucg->cg_irotor = cpu_to_fs32(sb, ucpi->c_irotor); - ubh_mark_buffer_dirty (UCPI_UBH); - for (i = 1; i < UCPI_UBH->count; i++) { - brelse (UCPI_UBH->bh[i]); + ubh_mark_buffer_dirty (UCPI_UBH(ucpi)); + for (i = 1; i < UCPI_UBH(ucpi)->count; i++) { + brelse (UCPI_UBH(ucpi)->bh[i]); } sbi->s_cgno[bitmap_nr] = UFS_CGNO_EMPTY; - UFSD(("EXIT\n")) + UFSD("EXIT\n"); } /* @@ -139,7 +130,7 @@ struct ufs_cg_private_info * ufs_load_cylinder ( struct ufs_cg_private_info * ucpi; unsigned cg, i, j; - UFSD(("ENTER, cgno %u\n", cgno)) + UFSD("ENTER, cgno %u\n", cgno); uspi = sbi->s_uspi; if (cgno >= uspi->s_ncg) { @@ -150,7 +141,7 @@ struct ufs_cg_private_info * ufs_load_cylinder ( * Cylinder group number cg it in cache and it was last used */ if (sbi->s_cgno[0] == cgno) { - UFSD(("EXIT\n")) + UFSD("EXIT\n"); return sbi->s_ucpi[0]; } /* @@ -160,16 +151,16 @@ struct ufs_cg_private_info * ufs_load_cylinder ( if (sbi->s_cgno[cgno] != UFS_CGNO_EMPTY) { if (sbi->s_cgno[cgno] != cgno) { ufs_panic (sb, "ufs_load_cylinder", "internal error, wrong number of cg in cache"); - UFSD(("EXIT (FAILED)\n")) + UFSD("EXIT (FAILED)\n"); return NULL; } else { - UFSD(("EXIT\n")) + UFSD("EXIT\n"); return sbi->s_ucpi[cgno]; } } else { ufs_read_cylinder (sb, cgno, cgno); - UFSD(("EXIT\n")) + UFSD("EXIT\n"); return sbi->s_ucpi[cgno]; } } @@ -204,6 +195,6 @@ struct ufs_cg_private_info * ufs_load_cylinder ( sbi->s_ucpi[0] = ucpi; ufs_read_cylinder (sb, cgno, 0); } - UFSD(("EXIT\n")) + UFSD("EXIT\n"); return sbi->s_ucpi[0]; } diff --git a/fs/ufs/dir.c b/fs/ufs/dir.c index 1a561202d3f..7f0a0aa6358 100644 --- a/fs/ufs/dir.c +++ b/fs/ufs/dir.c @@ -11,31 +11,20 @@ * 4.4BSD (FreeBSD) support added on February 1st 1998 by * Niels Kristian Bech Jensen <nkbj@image.dk> partially based * on code by Martin von Loewis <martin@mira.isdn.cs.tu-berlin.de>. + * + * Migration to usage of "page cache" on May 2006 by + * Evgeniy Dushistov <dushistov@mail.ru> based on ext2 code base. */ #include <linux/time.h> #include <linux/fs.h> #include <linux/ufs_fs.h> #include <linux/smp_lock.h> -#include <linux/buffer_head.h> #include <linux/sched.h> #include "swab.h" #include "util.h" -#undef UFS_DIR_DEBUG - -#ifdef UFS_DIR_DEBUG -#define UFSD(x) printk("(%s, %d), %s: ", __FILE__, __LINE__, __FUNCTION__); printk x; -#else -#define UFSD(x) -#endif - -static int -ufs_check_dir_entry (const char *, struct inode *, struct ufs_dir_entry *, - struct buffer_head *, unsigned long); - - /* * NOTE! unlike strncmp, ufs_match returns 1 for success, 0 for failure. * @@ -51,495 +40,541 @@ static inline int ufs_match(struct super_block *sb, int len, return !memcmp(name, de->d_name, len); } -/* - * This is blatantly stolen from ext2fs - */ -static int -ufs_readdir (struct file * filp, void * dirent, filldir_t filldir) +static int ufs_commit_chunk(struct page *page, unsigned from, unsigned to) { - struct inode *inode = filp->f_dentry->d_inode; - int error = 0; - unsigned long offset, lblk; - int i, stored; - struct buffer_head * bh; - struct ufs_dir_entry * de; - struct super_block * sb; - int de_reclen; - unsigned flags; - u64 blk= 0L; - - lock_kernel(); - - sb = inode->i_sb; - flags = UFS_SB(sb)->s_flags; - - UFSD(("ENTER, ino %lu f_pos %lu\n", inode->i_ino, (unsigned long) filp->f_pos)) - - stored = 0; - bh = NULL; - offset = filp->f_pos & (sb->s_blocksize - 1); - - while (!error && !stored && filp->f_pos < inode->i_size) { - lblk = (filp->f_pos) >> sb->s_blocksize_bits; - blk = ufs_frag_map(inode, lblk); - if (!blk || !(bh = sb_bread(sb, blk))) { - /* XXX - error - skip to the next block */ - printk("ufs_readdir: " - "dir inode %lu has a hole at offset %lu\n", - inode->i_ino, (unsigned long int)filp->f_pos); - filp->f_pos += sb->s_blocksize - offset; - continue; - } - -revalidate: - /* If the dir block has changed since the last call to - * readdir(2), then we might be pointing to an invalid - * dirent right now. Scan from the start of the block - * to make sure. */ - if (filp->f_version != inode->i_version) { - for (i = 0; i < sb->s_blocksize && i < offset; ) { - de = (struct ufs_dir_entry *)(bh->b_data + i); - /* It's too expensive to do a full - * dirent test each time round this - * loop, but we do have to test at - * least that it is non-zero. A - * failure will be detected in the - * dirent test below. */ - de_reclen = fs16_to_cpu(sb, de->d_reclen); - if (de_reclen < 1) - break; - i += de_reclen; - } - offset = i; - filp->f_pos = (filp->f_pos & ~(sb->s_blocksize - 1)) - | offset; - filp->f_version = inode->i_version; - } + struct inode *dir = page->mapping->host; + int err = 0; + dir->i_version++; + page->mapping->a_ops->commit_write(NULL, page, from, to); + if (IS_DIRSYNC(dir)) + err = write_one_page(page, 1); + else + unlock_page(page); + return err; +} - while (!error && filp->f_pos < inode->i_size - && offset < sb->s_blocksize) { - de = (struct ufs_dir_entry *) (bh->b_data + offset); - /* XXX - put in a real ufs_check_dir_entry() */ - if ((de->d_reclen == 0) || (ufs_get_de_namlen(sb, de) == 0)) { - filp->f_pos = (filp->f_pos & - (sb->s_blocksize - 1)) + - sb->s_blocksize; - brelse(bh); - unlock_kernel(); - return stored; - } - if (!ufs_check_dir_entry ("ufs_readdir", inode, de, - bh, offset)) { - /* On error, skip the f_pos to the - next block. */ - filp->f_pos = (filp->f_pos | - (sb->s_blocksize - 1)) + - 1; - brelse (bh); - unlock_kernel(); - return stored; - } - offset += fs16_to_cpu(sb, de->d_reclen); - if (de->d_ino) { - /* We might block in the next section - * if the data destination is - * currently swapped out. So, use a - * version stamp to detect whether or - * not the directory has been modified - * during the copy operation. */ - unsigned long version = filp->f_version; - unsigned char d_type = DT_UNKNOWN; +static inline void ufs_put_page(struct page *page) +{ + kunmap(page); + page_cache_release(page); +} - UFSD(("filldir(%s,%u)\n", de->d_name, - fs32_to_cpu(sb, de->d_ino))) - UFSD(("namlen %u\n", ufs_get_de_namlen(sb, de))) +static inline unsigned long ufs_dir_pages(struct inode *inode) +{ + return (inode->i_size+PAGE_CACHE_SIZE-1)>>PAGE_CACHE_SHIFT; +} - if ((flags & UFS_DE_MASK) == UFS_DE_44BSD) - d_type = de->d_u.d_44.d_type; - error = filldir(dirent, de->d_name, - ufs_get_de_namlen(sb, de), filp->f_pos, - fs32_to_cpu(sb, de->d_ino), d_type); - if (error) - break; - if (version != filp->f_version) - goto revalidate; - stored ++; - } - filp->f_pos += fs16_to_cpu(sb, de->d_reclen); - } - offset = 0; - brelse (bh); +ino_t ufs_inode_by_name(struct inode *dir, struct dentry *dentry) +{ + ino_t res = 0; + struct ufs_dir_entry *de; + struct page *page; + + de = ufs_find_entry(dir, dentry, &page); + if (de) { + res = fs32_to_cpu(dir->i_sb, de->d_ino); + ufs_put_page(page); } - unlock_kernel(); - return 0; + return res; } -/* - * define how far ahead to read directories while searching them. - */ -#define NAMEI_RA_CHUNKS 2 -#define NAMEI_RA_BLOCKS 4 -#define NAMEI_RA_SIZE (NAMEI_RA_CHUNKS * NAMEI_RA_BLOCKS) -#define NAMEI_RA_INDEX(c,b) (((c) * NAMEI_RA_BLOCKS) + (b)) -/* - * ufs_find_entry() - * - * finds an entry in the specified directory with the wanted name. It - * returns the cache buffer in which the entry was found, and the entry - * itself (as a parameter - res_bh). It does NOT read the inode of the - * entry - you'll have to do that yourself if you want to. - */ -struct ufs_dir_entry * ufs_find_entry (struct dentry *dentry, - struct buffer_head ** res_bh) +/* Releases the page */ +void ufs_set_link(struct inode *dir, struct ufs_dir_entry *de, + struct page *page, struct inode *inode) { - struct super_block * sb; - struct buffer_head * bh_use[NAMEI_RA_SIZE]; - struct buffer_head * bh_read[NAMEI_RA_SIZE]; - unsigned long offset; - int block, toread, i, err; - struct inode *dir = dentry->d_parent->d_inode; - const char *name = dentry->d_name.name; - int namelen = dentry->d_name.len; + unsigned from = (char *) de - (char *) page_address(page); + unsigned to = from + fs16_to_cpu(dir->i_sb, de->d_reclen); + int err; - UFSD(("ENTER, dir_ino %lu, name %s, namlen %u\n", dir->i_ino, name, namelen)) - - *res_bh = NULL; - - sb = dir->i_sb; - - if (namelen > UFS_MAXNAMLEN) - return NULL; + lock_page(page); + err = page->mapping->a_ops->prepare_write(NULL, page, from, to); + BUG_ON(err); + de->d_ino = cpu_to_fs32(dir->i_sb, inode->i_ino); + ufs_set_de_type(dir->i_sb, de, inode->i_mode); + err = ufs_commit_chunk(page, from, to); + ufs_put_page(page); + dir->i_mtime = dir->i_ctime = CURRENT_TIME_SEC; + mark_inode_dirty(dir); +} - memset (bh_use, 0, sizeof (bh_use)); - toread = 0; - for (block = 0; block < NAMEI_RA_SIZE; ++block) { - struct buffer_head * bh; - if ((block << sb->s_blocksize_bits) >= dir->i_size) - break; - bh = ufs_getfrag (dir, block, 0, &err); - bh_use[block] = bh; - if (bh && !buffer_uptodate(bh)) - bh_read[toread++] = bh; +static void ufs_check_page(struct page *page) +{ + struct inode *dir = page->mapping->host; + struct super_block *sb = dir->i_sb; + char *kaddr = page_address(page); + unsigned offs, rec_len; + unsigned limit = PAGE_CACHE_SIZE; + struct ufs_dir_entry *p; + char *error; + + if ((dir->i_size >> PAGE_CACHE_SHIFT) == page->index) { + limit = dir->i_size & ~PAGE_CACHE_MASK; + if (limit & (UFS_SECTOR_SIZE - 1)) + goto Ebadsize; + if (!limit) + goto out; } + for (offs = 0; offs <= limit - UFS_DIR_REC_LEN(1); offs += rec_len) { + p = (struct ufs_dir_entry *)(kaddr + offs); + rec_len = fs16_to_cpu(sb, p->d_reclen); + + if (rec_len < UFS_DIR_REC_LEN(1)) + goto Eshort; + if (rec_len & 3) + goto Ealign; + if (rec_len < UFS_DIR_REC_LEN(ufs_get_de_namlen(sb, p))) + goto Enamelen; + if (((offs + rec_len - 1) ^ offs) & ~(UFS_SECTOR_SIZE-1)) + goto Espan; + if (fs32_to_cpu(sb, p->d_ino) > (UFS_SB(sb)->s_uspi->s_ipg * + UFS_SB(sb)->s_uspi->s_ncg)) + goto Einumber; + } + if (offs != limit) + goto Eend; +out: + SetPageChecked(page); + return; + + /* Too bad, we had an error */ + +Ebadsize: + ufs_error(sb, "ufs_check_page", + "size of directory #%lu is not a multiple of chunk size", + dir->i_ino + ); + goto fail; +Eshort: + error = "rec_len is smaller than minimal"; + goto bad_entry; +Ealign: + error = "unaligned directory entry"; + goto bad_entry; +Enamelen: + error = "rec_len is too small for name_len"; + goto bad_entry; +Espan: + error = "directory entry across blocks"; + goto bad_entry; +Einumber: + error = "inode out of bounds"; +bad_entry: + ufs_error (sb, "ufs_check_page", "bad entry in directory #%lu: %s - " + "offset=%lu, rec_len=%d, name_len=%d", + dir->i_ino, error, (page->index<<PAGE_CACHE_SHIFT)+offs, + rec_len, ufs_get_de_namlen(sb, p)); + goto fail; +Eend: + p = (struct ufs_dir_entry *)(kaddr + offs); + ufs_error (sb, "ext2_check_page", + "entry in directory #%lu spans the page boundary" + "offset=%lu", + dir->i_ino, (page->index<<PAGE_CACHE_SHIFT)+offs); +fail: + SetPageChecked(page); + SetPageError(page); +} - for (block = 0, offset = 0; offset < dir->i_size; block++) { - struct buffer_head * bh; - struct ufs_dir_entry * de; - char * dlimit; - - if ((block % NAMEI_RA_BLOCKS) == 0 && toread) { - ll_rw_block (READ, toread, bh_read); - toread = 0; - } - bh = bh_use[block % NAMEI_RA_SIZE]; - if (!bh) { - ufs_error (sb, "ufs_find_entry", - "directory #%lu contains a hole at offset %lu", - dir->i_ino, offset); - offset += sb->s_blocksize; - continue; - } - wait_on_buffer (bh); - if (!buffer_uptodate(bh)) { - /* - * read error: all bets are off - */ - break; - } - - de = (struct ufs_dir_entry *) bh->b_data; - dlimit = bh->b_data + sb->s_blocksize; - while ((char *) de < dlimit && offset < dir->i_size) { - /* this code is executed quadratically often */ - /* do minimal checking by hand */ - int de_len; - - if ((char *) de + namelen <= dlimit && - ufs_match(sb, namelen, name, de)) { - /* found a match - - just to be sure, do a full check */ - if (!ufs_check_dir_entry("ufs_find_entry", - dir, de, bh, offset)) - goto failed; - for (i = 0; i < NAMEI_RA_SIZE; ++i) { - if (bh_use[i] != bh) - brelse (bh_use[i]); - } - *res_bh = bh; - return de; - } - /* prevent looping on a bad block */ - de_len = fs16_to_cpu(sb, de->d_reclen); - if (de_len <= 0) - goto failed; - offset += de_len; - de = (struct ufs_dir_entry *) ((char *) de + de_len); - } - - brelse (bh); - if (((block + NAMEI_RA_SIZE) << sb->s_blocksize_bits ) >= - dir->i_size) - bh = NULL; - else - bh = ufs_getfrag (dir, block + NAMEI_RA_SIZE, 0, &err); - bh_use[block % NAMEI_RA_SIZE] = bh; - if (bh && !buffer_uptodate(bh)) - bh_read[toread++] = bh; +static struct page *ufs_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); + if (!IS_ERR(page)) { + wait_on_page_locked(page); + kmap(page); + if (!PageUptodate(page)) + goto fail; + if (!PageChecked(page)) + ufs_check_page(page); + if (PageError(page)) + goto fail; } + return page; -failed: - for (i = 0; i < NAMEI_RA_SIZE; ++i) brelse (bh_use[i]); - UFSD(("EXIT\n")) - return NULL; +fail: + ufs_put_page(page); + return ERR_PTR(-EIO); } -static int -ufs_check_dir_entry (const char *function, struct inode *dir, - struct ufs_dir_entry *de, struct buffer_head *bh, - unsigned long offset) +/* + * Return the offset into page `page_nr' of the last valid + * byte in that page, plus one. + */ +static unsigned +ufs_last_byte(struct inode *inode, unsigned long page_nr) { - struct super_block *sb = dir->i_sb; - const char *error_msg = NULL; - int rlen = fs16_to_cpu(sb, de->d_reclen); - - if (rlen < UFS_DIR_REC_LEN(1)) - error_msg = "reclen is smaller than minimal"; - else if (rlen % 4 != 0) - error_msg = "reclen % 4 != 0"; - else if (rlen < UFS_DIR_REC_LEN(ufs_get_de_namlen(sb, de))) - error_msg = "reclen is too small for namlen"; - else if (((char *) de - bh->b_data) + rlen > dir->i_sb->s_blocksize) - error_msg = "directory entry across blocks"; - else if (fs32_to_cpu(sb, de->d_ino) > (UFS_SB(sb)->s_uspi->s_ipg * - UFS_SB(sb)->s_uspi->s_ncg)) - error_msg = "inode out of bounds"; - - if (error_msg != NULL) - ufs_error (sb, function, "bad entry in directory #%lu, size %Lu: %s - " - "offset=%lu, inode=%lu, reclen=%d, namlen=%d", - dir->i_ino, dir->i_size, error_msg, offset, - (unsigned long)fs32_to_cpu(sb, de->d_ino), - rlen, ufs_get_de_namlen(sb, de)); - - return (error_msg == NULL ? 1 : 0); + unsigned last_byte = inode->i_size; + + last_byte -= page_nr << PAGE_CACHE_SHIFT; + if (last_byte > PAGE_CACHE_SIZE) + last_byte = PAGE_CACHE_SIZE; + return last_byte; } -struct ufs_dir_entry *ufs_dotdot(struct inode *dir, struct buffer_head **p) +static inline struct ufs_dir_entry * +ufs_next_entry(struct super_block *sb, struct ufs_dir_entry *p) { - int err; - struct buffer_head *bh = ufs_bread (dir, 0, 0, &err); - struct ufs_dir_entry *res = NULL; - - if (bh) { - res = (struct ufs_dir_entry *) bh->b_data; - res = (struct ufs_dir_entry *)((char *)res + - fs16_to_cpu(dir->i_sb, res->d_reclen)); - } - *p = bh; - return res; + return (struct ufs_dir_entry *)((char *)p + + fs16_to_cpu(sb, p->d_reclen)); } -ino_t ufs_inode_by_name(struct inode * dir, struct dentry *dentry) + +struct ufs_dir_entry *ufs_dotdot(struct inode *dir, struct page **p) { - ino_t res = 0; - struct ufs_dir_entry * de; - struct buffer_head *bh; + struct page *page = ufs_get_page(dir, 0); + struct ufs_dir_entry *de = NULL; - de = ufs_find_entry (dentry, &bh); - if (de) { - res = fs32_to_cpu(dir->i_sb, de->d_ino); - brelse(bh); + if (!IS_ERR(page)) { + de = ufs_next_entry(dir->i_sb, + (struct ufs_dir_entry *)page_address(page)); + *p = page; } - return res; + return de; } -void ufs_set_link(struct inode *dir, struct ufs_dir_entry *de, - struct buffer_head *bh, struct inode *inode) +/* + * ufs_find_entry() + * + * finds an entry in the specified directory with the wanted name. It + * returns the page in which the entry was found, and the entry itself + * (as a parameter - res_dir). Page is returned mapped and unlocked. + * Entry is guaranteed to be valid. + */ +struct ufs_dir_entry *ufs_find_entry(struct inode *dir, struct dentry *dentry, + struct page **res_page) { - dir->i_version++; - de->d_ino = cpu_to_fs32(dir->i_sb, inode->i_ino); - mark_buffer_dirty(bh); - if (IS_DIRSYNC(dir)) - sync_dirty_buffer(bh); - brelse (bh); + struct super_block *sb = dir->i_sb; + const char *name = dentry->d_name.name; + int namelen = dentry->d_name.len; + unsigned reclen = UFS_DIR_REC_LEN(namelen); + unsigned long start, n; + unsigned long npages = ufs_dir_pages(dir); + struct page *page = NULL; + struct ufs_inode_info *ui = UFS_I(dir); + struct ufs_dir_entry *de; + + UFSD("ENTER, dir_ino %lu, name %s, namlen %u\n", dir->i_ino, name, namelen); + + if (npages == 0 || namelen > UFS_MAXNAMLEN) + goto out; + + /* OFFSET_CACHE */ + *res_page = NULL; + + start = ui->i_dir_start_lookup; + + if (start >= npages) + start = 0; + n = start; + do { + char *kaddr; + page = ufs_get_page(dir, n); + if (!IS_ERR(page)) { + kaddr = page_address(page); + de = (struct ufs_dir_entry *) kaddr; + kaddr += ufs_last_byte(dir, n) - reclen; + while ((char *) de <= kaddr) { + if (de->d_reclen == 0) { + ufs_error(dir->i_sb, __FUNCTION__, + "zero-length directory entry"); + ufs_put_page(page); + goto out; + } + if (ufs_match(sb, namelen, name, de)) + goto found; + de = ufs_next_entry(sb, de); + } + ufs_put_page(page); + } + if (++n >= npages) + n = 0; + } while (n != start); +out: + return NULL; + +found: + *res_page = page; + ui->i_dir_start_lookup = n; + return de; } /* - * ufs_add_entry() - * - * adds a file entry to the specified directory, using the same - * semantics as ufs_find_entry(). It returns NULL if it failed. + * Parent is locked. */ int ufs_add_link(struct dentry *dentry, struct inode *inode) { - struct super_block * sb; - struct ufs_sb_private_info * uspi; - unsigned long offset; - unsigned fragoff; - unsigned short rec_len; - struct buffer_head * bh; - struct ufs_dir_entry * de, * de1; struct inode *dir = dentry->d_parent->d_inode; const char *name = dentry->d_name.name; int namelen = dentry->d_name.len; + struct super_block *sb = dir->i_sb; + unsigned reclen = UFS_DIR_REC_LEN(namelen); + unsigned short rec_len, name_len; + struct page *page = NULL; + struct ufs_dir_entry *de; + unsigned long npages = ufs_dir_pages(dir); + unsigned long n; + char *kaddr; + unsigned from, to; int err; - UFSD(("ENTER, name %s, namelen %u\n", name, namelen)) - - sb = dir->i_sb; - uspi = UFS_SB(sb)->s_uspi; - - if (!namelen) - return -EINVAL; - bh = ufs_bread (dir, 0, 0, &err); - if (!bh) - return err; - rec_len = UFS_DIR_REC_LEN(namelen); - offset = 0; - de = (struct ufs_dir_entry *) bh->b_data; - while (1) { - if ((char *)de >= UFS_SECTOR_SIZE + bh->b_data) { - fragoff = offset & ~uspi->s_fmask; - if (fragoff != 0 && fragoff != UFS_SECTOR_SIZE) - ufs_error (sb, "ufs_add_entry", "internal error" - " fragoff %u", fragoff); - if (!fragoff) { - brelse (bh); - bh = ufs_bread (dir, offset >> sb->s_blocksize_bits, 1, &err); - if (!bh) - return err; - } - if (dir->i_size <= offset) { - if (dir->i_size == 0) { - brelse(bh); - return -ENOENT; - } - de = (struct ufs_dir_entry *) (bh->b_data + fragoff); - de->d_ino = 0; + UFSD("ENTER, name %s, namelen %u\n", name, namelen); + + /* + * We take care of directory expansion in the same loop. + * This code plays outside i_size, so it locks the page + * to protect that region. + */ + for (n = 0; n <= npages; n++) { + char *dir_end; + + page = ufs_get_page(dir, n); + err = PTR_ERR(page); + if (IS_ERR(page)) + goto out; + lock_page(page); + kaddr = page_address(page); + dir_end = kaddr + ufs_last_byte(dir, n); + de = (struct ufs_dir_entry *)kaddr; + kaddr += PAGE_CACHE_SIZE - reclen; + while ((char *)de <= kaddr) { + if ((char *)de == dir_end) { + /* We hit i_size */ + name_len = 0; + rec_len = UFS_SECTOR_SIZE; de->d_reclen = cpu_to_fs16(sb, UFS_SECTOR_SIZE); - ufs_set_de_namlen(sb, de, 0); - dir->i_size = offset + UFS_SECTOR_SIZE; - mark_inode_dirty(dir); - } else { - de = (struct ufs_dir_entry *) bh->b_data; + de->d_ino = 0; + goto got_it; } + if (de->d_reclen == 0) { + ufs_error(dir->i_sb, __FUNCTION__, + "zero-length directory entry"); + err = -EIO; + goto out_unlock; + } + err = -EEXIST; + if (ufs_match(sb, namelen, name, de)) + goto out_unlock; + name_len = UFS_DIR_REC_LEN(ufs_get_de_namlen(sb, de)); + rec_len = fs16_to_cpu(sb, de->d_reclen); + if (!de->d_ino && rec_len >= reclen) + goto got_it; + if (rec_len >= name_len + reclen) + goto got_it; + de = (struct ufs_dir_entry *) ((char *) de + rec_len); } - if (!ufs_check_dir_entry ("ufs_add_entry", dir, de, bh, offset)) { - brelse (bh); - return -ENOENT; - } - if (ufs_match(sb, namelen, name, de)) { - brelse (bh); - return -EEXIST; - } - if (de->d_ino == 0 && fs16_to_cpu(sb, de->d_reclen) >= rec_len) - break; - - if (fs16_to_cpu(sb, de->d_reclen) >= - UFS_DIR_REC_LEN(ufs_get_de_namlen(sb, de)) + rec_len) - break; - offset += fs16_to_cpu(sb, de->d_reclen); - de = (struct ufs_dir_entry *) ((char *) de + fs16_to_cpu(sb, de->d_reclen)); + unlock_page(page); + ufs_put_page(page); } - + BUG(); + return -EINVAL; + +got_it: + from = (char*)de - (char*)page_address(page); + to = from + rec_len; + err = page->mapping->a_ops->prepare_write(NULL, page, from, to); + if (err) + goto out_unlock; if (de->d_ino) { - de1 = (struct ufs_dir_entry *) ((char *) de + - UFS_DIR_REC_LEN(ufs_get_de_namlen(sb, de))); - de1->d_reclen = - cpu_to_fs16(sb, fs16_to_cpu(sb, de->d_reclen) - - UFS_DIR_REC_LEN(ufs_get_de_namlen(sb, de))); - de->d_reclen = - cpu_to_fs16(sb, UFS_DIR_REC_LEN(ufs_get_de_namlen(sb, de))); + struct ufs_dir_entry *de1 = + (struct ufs_dir_entry *) ((char *) de + name_len); + de1->d_reclen = cpu_to_fs16(sb, rec_len - name_len); + de->d_reclen = cpu_to_fs16(sb, name_len); + de = de1; } - de->d_ino = 0; + ufs_set_de_namlen(sb, de, namelen); - memcpy (de->d_name, name, namelen + 1); + memcpy(de->d_name, name, namelen + 1); de->d_ino = cpu_to_fs32(sb, inode->i_ino); ufs_set_de_type(sb, de, inode->i_mode); - mark_buffer_dirty(bh); - if (IS_DIRSYNC(dir)) - sync_dirty_buffer(bh); - brelse (bh); + + err = ufs_commit_chunk(page, from, to); dir->i_mtime = dir->i_ctime = CURRENT_TIME_SEC; - dir->i_version++; + mark_inode_dirty(dir); + /* OFFSET_CACHE */ +out_put: + ufs_put_page(page); +out: + return err; +out_unlock: + unlock_page(page); + goto out_put; +} - UFSD(("EXIT\n")) +static inline unsigned +ufs_validate_entry(struct super_block *sb, char *base, + unsigned offset, unsigned mask) +{ + struct ufs_dir_entry *de = (struct ufs_dir_entry*)(base + offset); + struct ufs_dir_entry *p = (struct ufs_dir_entry*)(base + (offset&mask)); + while ((char*)p < (char*)de) { + if (p->d_reclen == 0) + break; + p = ufs_next_entry(sb, p); + } + return (char *)p - base; +} + + +/* + * This is blatantly stolen from ext2fs + */ +static int +ufs_readdir(struct file *filp, void *dirent, filldir_t filldir) +{ + loff_t pos = filp->f_pos; + struct inode *inode = filp->f_dentry->d_inode; + struct super_block *sb = inode->i_sb; + unsigned int offset = pos & ~PAGE_CACHE_MASK; + unsigned long n = pos >> PAGE_CACHE_SHIFT; + unsigned long npages = ufs_dir_pages(inode); + unsigned chunk_mask = ~(UFS_SECTOR_SIZE - 1); + int need_revalidate = filp->f_version != inode->i_version; + unsigned flags = UFS_SB(sb)->s_flags; + + UFSD("BEGIN\n"); + + if (pos > inode->i_size - UFS_DIR_REC_LEN(1)) + return 0; + + for ( ; n < npages; n++, offset = 0) { + char *kaddr, *limit; + struct ufs_dir_entry *de; + + struct page *page = ufs_get_page(inode, n); + + if (IS_ERR(page)) { + ufs_error(sb, __FUNCTION__, + "bad page in #%lu", + inode->i_ino); + filp->f_pos += PAGE_CACHE_SIZE - offset; + return -EIO; + } + kaddr = page_address(page); + if (unlikely(need_revalidate)) { + if (offset) { + offset = ufs_validate_entry(sb, kaddr, offset, chunk_mask); + filp->f_pos = (n<<PAGE_CACHE_SHIFT) + offset; + } + filp->f_version = inode->i_version; + need_revalidate = 0; + } + de = (struct ufs_dir_entry *)(kaddr+offset); + limit = kaddr + ufs_last_byte(inode, n) - UFS_DIR_REC_LEN(1); + for ( ;(char*)de <= limit; de = ufs_next_entry(sb, de)) { + if (de->d_reclen == 0) { + ufs_error(sb, __FUNCTION__, + "zero-length directory entry"); + ufs_put_page(page); + return -EIO; + } + if (de->d_ino) { + int over; + unsigned char d_type = DT_UNKNOWN; + + offset = (char *)de - kaddr; + + UFSD("filldir(%s,%u)\n", de->d_name, + fs32_to_cpu(sb, de->d_ino)); + UFSD("namlen %u\n", ufs_get_de_namlen(sb, de)); + + if ((flags & UFS_DE_MASK) == UFS_DE_44BSD) + d_type = de->d_u.d_44.d_type; + + over = filldir(dirent, de->d_name, + ufs_get_de_namlen(sb, de), + (n<<PAGE_CACHE_SHIFT) | offset, + fs32_to_cpu(sb, de->d_ino), d_type); + if (over) { + ufs_put_page(page); + return 0; + } + } + filp->f_pos += fs16_to_cpu(sb, de->d_reclen); + } + ufs_put_page(page); + } return 0; } + /* * ufs_delete_entry deletes a directory entry by merging it with the * previous entry. */ -int ufs_delete_entry (struct inode * inode, struct ufs_dir_entry * dir, - struct buffer_head * bh ) - +int ufs_delete_entry(struct inode *inode, struct ufs_dir_entry *dir, + struct page * page) { - struct super_block * sb; - struct ufs_dir_entry * de, * pde; - unsigned i; - - UFSD(("ENTER\n")) + struct super_block *sb = inode->i_sb; + struct address_space *mapping = page->mapping; + char *kaddr = page_address(page); + unsigned from = ((char*)dir - kaddr) & ~(UFS_SECTOR_SIZE - 1); + unsigned to = ((char*)dir - kaddr) + fs16_to_cpu(sb, dir->d_reclen); + struct ufs_dir_entry *pde = NULL; + struct ufs_dir_entry *de = (struct ufs_dir_entry *) (kaddr + from); + int err; - sb = inode->i_sb; - i = 0; - pde = NULL; - de = (struct ufs_dir_entry *) bh->b_data; - - UFSD(("ino %u, reclen %u, namlen %u, name %s\n", - fs32_to_cpu(sb, de->d_ino), - fs16_to_cpu(sb, de->d_reclen), - ufs_get_de_namlen(sb, de), de->d_name)) - - while (i < bh->b_size) { - if (!ufs_check_dir_entry ("ufs_delete_entry", inode, de, bh, i)) { - brelse(bh); - return -EIO; - } - if (de == dir) { - if (pde) - fs16_add(sb, &pde->d_reclen, - fs16_to_cpu(sb, dir->d_reclen)); - dir->d_ino = 0; - inode->i_version++; - inode->i_ctime = inode->i_mtime = CURRENT_TIME_SEC; - mark_inode_dirty(inode); - mark_buffer_dirty(bh); - if (IS_DIRSYNC(inode)) - sync_dirty_buffer(bh); - brelse(bh); - UFSD(("EXIT\n")) - return 0; + UFSD("ENTER\n"); + + UFSD("ino %u, reclen %u, namlen %u, name %s\n", + fs32_to_cpu(sb, de->d_ino), + fs16_to_cpu(sb, de->d_reclen), + ufs_get_de_namlen(sb, de), de->d_name); + + while ((char*)de < (char*)dir) { + if (de->d_reclen == 0) { + ufs_error(inode->i_sb, __FUNCTION__, + "zero-length directory entry"); + err = -EIO; + goto out; } - i += fs16_to_cpu(sb, de->d_reclen); - if (i == UFS_SECTOR_SIZE) pde = NULL; - else pde = de; - de = (struct ufs_dir_entry *) - ((char *) de + fs16_to_cpu(sb, de->d_reclen)); - if (i == UFS_SECTOR_SIZE && de->d_reclen == 0) - break; + pde = de; + de = ufs_next_entry(sb, de); } - UFSD(("EXIT\n")) - brelse(bh); - return -ENOENT; + if (pde) + from = (char*)pde - (char*)page_address(page); + lock_page(page); + err = mapping->a_ops->prepare_write(NULL, page, from, to); + BUG_ON(err); + if (pde) + pde->d_reclen = cpu_to_fs16(sb, to-from); + dir->d_ino = 0; + err = ufs_commit_chunk(page, from, to); + inode->i_ctime = inode->i_mtime = CURRENT_TIME_SEC; + mark_inode_dirty(inode); +out: + ufs_put_page(page); + UFSD("EXIT\n"); + return err; } int ufs_make_empty(struct inode * inode, struct inode *dir) { struct super_block * sb = dir->i_sb; - struct buffer_head * dir_block; + struct address_space *mapping = inode->i_mapping; + struct page *page = grab_cache_page(mapping, 0); struct ufs_dir_entry * de; + char *base; int err; - dir_block = ufs_bread (inode, 0, 1, &err); - if (!dir_block) - return err; + if (!page) + return -ENOMEM; + kmap(page); + err = mapping->a_ops->prepare_write(NULL, page, 0, UFS_SECTOR_SIZE); + if (err) { + unlock_page(page); + goto fail; + } + + + base = (char*)page_address(page); + memset(base, 0, PAGE_CACHE_SIZE); + + de = (struct ufs_dir_entry *) base; - inode->i_blocks = sb->s_blocksize / UFS_SECTOR_SIZE; - de = (struct ufs_dir_entry *) dir_block->b_data; de->d_ino = cpu_to_fs32(sb, inode->i_ino); ufs_set_de_type(sb, de, inode->i_mode); ufs_set_de_namlen(sb, de, 1); @@ -552,72 +587,65 @@ int ufs_make_empty(struct inode * inode, struct inode *dir) de->d_reclen = cpu_to_fs16(sb, UFS_SECTOR_SIZE - UFS_DIR_REC_LEN(1)); ufs_set_de_namlen(sb, de, 2); strcpy (de->d_name, ".."); - mark_buffer_dirty(dir_block); - brelse (dir_block); - mark_inode_dirty(inode); - return 0; + + err = ufs_commit_chunk(page, 0, UFS_SECTOR_SIZE); +fail: + kunmap(page); + page_cache_release(page); + return err; } /* * routine to check that the specified directory is empty (for rmdir) */ -int ufs_empty_dir (struct inode * inode) +int ufs_empty_dir(struct inode * inode) { - struct super_block * sb; - unsigned long offset; - struct buffer_head * bh; - struct ufs_dir_entry * de, * de1; - int err; - - sb = inode->i_sb; - - if (inode->i_size < UFS_DIR_REC_LEN(1) + UFS_DIR_REC_LEN(2) || - !(bh = ufs_bread (inode, 0, 0, &err))) { - ufs_warning (inode->i_sb, "empty_dir", - "bad directory (dir #%lu) - no data block", - inode->i_ino); - return 1; - } - de = (struct ufs_dir_entry *) bh->b_data; - de1 = (struct ufs_dir_entry *) - ((char *)de + fs16_to_cpu(sb, de->d_reclen)); - if (fs32_to_cpu(sb, de->d_ino) != inode->i_ino || de1->d_ino == 0 || - strcmp (".", de->d_name) || strcmp ("..", de1->d_name)) { - ufs_warning (inode->i_sb, "empty_dir", - "bad directory (dir #%lu) - no `.' or `..'", - inode->i_ino); - return 1; - } - offset = fs16_to_cpu(sb, de->d_reclen) + fs16_to_cpu(sb, de1->d_reclen); - de = (struct ufs_dir_entry *) - ((char *)de1 + fs16_to_cpu(sb, de1->d_reclen)); - while (offset < inode->i_size ) { - if (!bh || (void *) de >= (void *) (bh->b_data + sb->s_blocksize)) { - brelse (bh); - bh = ufs_bread (inode, offset >> sb->s_blocksize_bits, 1, &err); - if (!bh) { - ufs_error (sb, "empty_dir", - "directory #%lu contains a hole at offset %lu", - inode->i_ino, offset); - offset += sb->s_blocksize; - continue; + struct super_block *sb = inode->i_sb; + struct page *page = NULL; + unsigned long i, npages = ufs_dir_pages(inode); + + for (i = 0; i < npages; i++) { + char *kaddr; + struct ufs_dir_entry *de; + page = ufs_get_page(inode, i); + + if (IS_ERR(page)) + continue; + + kaddr = page_address(page); + de = (struct ufs_dir_entry *)kaddr; + kaddr += ufs_last_byte(inode, i) - UFS_DIR_REC_LEN(1); + + while ((char *)de <= kaddr) { + if (de->d_reclen == 0) { + ufs_error(inode->i_sb, __FUNCTION__, + "zero-length directory entry: " + "kaddr=%p, de=%p\n", kaddr, de); + goto not_empty; } - de = (struct ufs_dir_entry *) bh->b_data; - } - if (!ufs_check_dir_entry ("empty_dir", inode, de, bh, offset)) { - brelse (bh); - return 1; - } - if (de->d_ino) { - brelse (bh); - return 0; + if (de->d_ino) { + u16 namelen=ufs_get_de_namlen(sb, de); + /* check for . and .. */ + if (de->d_name[0] != '.') + goto not_empty; + if (namelen > 2) + goto not_empty; + if (namelen < 2) { + if (inode->i_ino != + fs32_to_cpu(sb, de->d_ino)) + goto not_empty; + } else if (de->d_name[1] != '.') + goto not_empty; + } + de = ufs_next_entry(sb, de); } - offset += fs16_to_cpu(sb, de->d_reclen); - de = (struct ufs_dir_entry *) - ((char *)de + fs16_to_cpu(sb, de->d_reclen)); + ufs_put_page(page); } - brelse (bh); return 1; + +not_empty: + ufs_put_page(page); + return 0; } const struct file_operations ufs_dir_operations = { diff --git a/fs/ufs/file.c b/fs/ufs/file.c index 312fd3f8631..a9c6e5f04fa 100644 --- a/fs/ufs/file.c +++ b/fs/ufs/file.c @@ -25,6 +25,26 @@ #include <linux/fs.h> #include <linux/ufs_fs.h> +#include <linux/buffer_head.h> /* for sync_mapping_buffers() */ + +static int ufs_sync_file(struct file *file, struct dentry *dentry, int datasync) +{ + struct inode *inode = dentry->d_inode; + int err; + int ret; + + ret = sync_mapping_buffers(inode->i_mapping); + if (!(inode->i_state & I_DIRTY)) + return ret; + if (datasync && !(inode->i_state & I_DIRTY_DATASYNC)) + return ret; + + err = ufs_sync_inode(inode); + if (ret == 0) + ret = err; + return ret; +} + /* * We have mostly NULL's here: the current defaults are ok for @@ -37,9 +57,6 @@ const struct file_operations ufs_file_operations = { .write = generic_file_write, .mmap = generic_file_mmap, .open = generic_file_open, + .fsync = ufs_sync_file, .sendfile = generic_file_sendfile, }; - -struct inode_operations ufs_file_inode_operations = { - .truncate = ufs_truncate, -}; diff --git a/fs/ufs/ialloc.c b/fs/ufs/ialloc.c index c7a47ed4f43..2ad1259c6ec 100644 --- a/fs/ufs/ialloc.c +++ b/fs/ufs/ialloc.c @@ -34,14 +34,6 @@ #include "swab.h" #include "util.h" -#undef UFS_IALLOC_DEBUG - -#ifdef UFS_IALLOC_DEBUG -#define UFSD(x) printk("(%s, %d), %s: ", __FILE__, __LINE__, __FUNCTION__); printk x; -#else -#define UFSD(x) -#endif - /* * NOTE! When we get the inode, we're the only people * that have access to it, and as such there are no @@ -68,7 +60,7 @@ void ufs_free_inode (struct inode * inode) int is_directory; unsigned ino, cg, bit; - UFSD(("ENTER, ino %lu\n", inode->i_ino)) + UFSD("ENTER, ino %lu\n", inode->i_ino); sb = inode->i_sb; uspi = UFS_SB(sb)->s_uspi; @@ -91,7 +83,7 @@ void ufs_free_inode (struct inode * inode) unlock_super (sb); return; } - ucg = ubh_get_ucg(UCPI_UBH); + ucg = ubh_get_ucg(UCPI_UBH(ucpi)); if (!ufs_cg_chkmagic(sb, ucg)) ufs_panic (sb, "ufs_free_fragments", "internal error, bad cg magic number"); @@ -104,33 +96,33 @@ void ufs_free_inode (struct inode * inode) clear_inode (inode); - if (ubh_isclr (UCPI_UBH, ucpi->c_iusedoff, bit)) + if (ubh_isclr (UCPI_UBH(ucpi), ucpi->c_iusedoff, bit)) ufs_error(sb, "ufs_free_inode", "bit already cleared for inode %u", ino); else { - ubh_clrbit (UCPI_UBH, ucpi->c_iusedoff, bit); + ubh_clrbit (UCPI_UBH(ucpi), ucpi->c_iusedoff, bit); if (ino < ucpi->c_irotor) ucpi->c_irotor = ino; fs32_add(sb, &ucg->cg_cs.cs_nifree, 1); - fs32_add(sb, &usb1->fs_cstotal.cs_nifree, 1); + uspi->cs_total.cs_nifree++; fs32_add(sb, &UFS_SB(sb)->fs_cs(cg).cs_nifree, 1); if (is_directory) { fs32_sub(sb, &ucg->cg_cs.cs_ndir, 1); - fs32_sub(sb, &usb1->fs_cstotal.cs_ndir, 1); + uspi->cs_total.cs_ndir--; fs32_sub(sb, &UFS_SB(sb)->fs_cs(cg).cs_ndir, 1); } } - ubh_mark_buffer_dirty (USPI_UBH); - ubh_mark_buffer_dirty (UCPI_UBH); + ubh_mark_buffer_dirty (USPI_UBH(uspi)); + ubh_mark_buffer_dirty (UCPI_UBH(ucpi)); if (sb->s_flags & MS_SYNCHRONOUS) { - ubh_ll_rw_block (SWRITE, 1, (struct ufs_buffer_head **) &ucpi); - ubh_wait_on_buffer (UCPI_UBH); + ubh_ll_rw_block(SWRITE, UCPI_UBH(ucpi)); + ubh_wait_on_buffer (UCPI_UBH(ucpi)); } sb->s_dirt = 1; unlock_super (sb); - UFSD(("EXIT\n")) + UFSD("EXIT\n"); } /* @@ -155,7 +147,7 @@ struct inode * ufs_new_inode(struct inode * dir, int mode) unsigned cg, bit, i, j, start; struct ufs_inode_info *ufsi; - UFSD(("ENTER\n")) + UFSD("ENTER\n"); /* Cannot create files in a deleted directory */ if (!dir || !dir->i_nlink) @@ -213,43 +205,43 @@ cg_found: ucpi = ufs_load_cylinder (sb, cg); if (!ucpi) goto failed; - ucg = ubh_get_ucg(UCPI_UBH); + ucg = ubh_get_ucg(UCPI_UBH(ucpi)); if (!ufs_cg_chkmagic(sb, ucg)) ufs_panic (sb, "ufs_new_inode", "internal error, bad cg magic number"); start = ucpi->c_irotor; - bit = ubh_find_next_zero_bit (UCPI_UBH, ucpi->c_iusedoff, uspi->s_ipg, start); + bit = ubh_find_next_zero_bit (UCPI_UBH(ucpi), ucpi->c_iusedoff, uspi->s_ipg, start); if (!(bit < uspi->s_ipg)) { - bit = ubh_find_first_zero_bit (UCPI_UBH, ucpi->c_iusedoff, start); + bit = ubh_find_first_zero_bit (UCPI_UBH(ucpi), ucpi->c_iusedoff, start); if (!(bit < start)) { ufs_error (sb, "ufs_new_inode", "cylinder group %u corrupted - error in inode bitmap\n", cg); goto failed; } } - UFSD(("start = %u, bit = %u, ipg = %u\n", start, bit, uspi->s_ipg)) - if (ubh_isclr (UCPI_UBH, ucpi->c_iusedoff, bit)) - ubh_setbit (UCPI_UBH, ucpi->c_iusedoff, bit); + UFSD("start = %u, bit = %u, ipg = %u\n", start, bit, uspi->s_ipg); + if (ubh_isclr (UCPI_UBH(ucpi), ucpi->c_iusedoff, bit)) + ubh_setbit (UCPI_UBH(ucpi), ucpi->c_iusedoff, bit); else { ufs_panic (sb, "ufs_new_inode", "internal error"); goto failed; } fs32_sub(sb, &ucg->cg_cs.cs_nifree, 1); - fs32_sub(sb, &usb1->fs_cstotal.cs_nifree, 1); + uspi->cs_total.cs_nifree--; fs32_sub(sb, &sbi->fs_cs(cg).cs_nifree, 1); if (S_ISDIR(mode)) { fs32_add(sb, &ucg->cg_cs.cs_ndir, 1); - fs32_add(sb, &usb1->fs_cstotal.cs_ndir, 1); + uspi->cs_total.cs_ndir++; fs32_add(sb, &sbi->fs_cs(cg).cs_ndir, 1); } - ubh_mark_buffer_dirty (USPI_UBH); - ubh_mark_buffer_dirty (UCPI_UBH); + ubh_mark_buffer_dirty (USPI_UBH(uspi)); + ubh_mark_buffer_dirty (UCPI_UBH(ucpi)); if (sb->s_flags & MS_SYNCHRONOUS) { - ubh_ll_rw_block (SWRITE, 1, (struct ufs_buffer_head **) &ucpi); - ubh_wait_on_buffer (UCPI_UBH); + ubh_ll_rw_block(SWRITE, UCPI_UBH(ucpi)); + ubh_wait_on_buffer (UCPI_UBH(ucpi)); } sb->s_dirt = 1; @@ -263,7 +255,6 @@ cg_found: inode->i_gid = current->fsgid; inode->i_ino = cg * uspi->s_ipg + bit; - inode->i_blksize = PAGE_SIZE; /* This is the optimal IO size (for stat), not the fs block size */ inode->i_blocks = 0; inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME_SEC; ufsi->i_flags = UFS_I(dir)->i_flags; @@ -272,6 +263,7 @@ cg_found: ufsi->i_shadow = 0; ufsi->i_osync = 0; ufsi->i_oeftflag = 0; + ufsi->i_dir_start_lookup = 0; memset(&ufsi->i_u1, 0, sizeof(ufsi->i_u1)); insert_inode_hash(inode); @@ -287,14 +279,14 @@ cg_found: return ERR_PTR(-EDQUOT); } - UFSD(("allocating inode %lu\n", inode->i_ino)) - UFSD(("EXIT\n")) + UFSD("allocating inode %lu\n", inode->i_ino); + UFSD("EXIT\n"); return inode; failed: unlock_super (sb); make_bad_inode(inode); iput (inode); - UFSD(("EXIT (FAILED)\n")) + UFSD("EXIT (FAILED)\n"); return ERR_PTR(-ENOSPC); } diff --git a/fs/ufs/inode.c b/fs/ufs/inode.c index 3c3f62ce2ad..ee1eaa6f4ec 100644 --- a/fs/ufs/inode.c +++ b/fs/ufs/inode.c @@ -41,14 +41,7 @@ #include "swab.h" #include "util.h" -#undef UFS_INODE_DEBUG -#undef UFS_INODE_DEBUG_MORE - -#ifdef UFS_INODE_DEBUG -#define UFSD(x) printk("(%s, %d), %s: ", __FILE__, __LINE__, __FUNCTION__); printk x; -#else -#define UFSD(x) -#endif +static u64 ufs_frag_map(struct inode *inode, sector_t frag); static int ufs_block_to_path(struct inode *inode, sector_t i_block, sector_t offsets[4]) { @@ -61,7 +54,7 @@ static int ufs_block_to_path(struct inode *inode, sector_t i_block, sector_t off int n = 0; - UFSD(("ptrs=uspi->s_apb = %d,double_blocks=%ld \n",ptrs,double_blocks)); + UFSD("ptrs=uspi->s_apb = %d,double_blocks=%ld \n",ptrs,double_blocks); if (i_block < 0) { ufs_warning(inode->i_sb, "ufs_block_to_path", "block < 0"); } else if (i_block < direct_blocks) { @@ -89,7 +82,7 @@ static int ufs_block_to_path(struct inode *inode, sector_t i_block, sector_t off * the begining of the filesystem. */ -u64 ufs_frag_map(struct inode *inode, sector_t frag) +static u64 ufs_frag_map(struct inode *inode, sector_t frag) { struct ufs_inode_info *ufsi = UFS_I(inode); struct super_block *sb = inode->i_sb; @@ -104,8 +97,10 @@ u64 ufs_frag_map(struct inode *inode, sector_t frag) unsigned flags = UFS_SB(sb)->s_flags; u64 temp = 0L; - UFSD((": frag = %llu depth = %d\n", (unsigned long long)frag, depth)); - UFSD((": uspi->s_fpbshift = %d ,uspi->s_apbmask = %x, mask=%llx\n",uspi->s_fpbshift,uspi->s_apbmask,mask)); + UFSD(": frag = %llu depth = %d\n", (unsigned long long)frag, depth); + UFSD(": uspi->s_fpbshift = %d ,uspi->s_apbmask = %x, mask=%llx\n", + uspi->s_fpbshift, uspi->s_apbmask, + (unsigned long long)mask); if (depth == 0) return 0; @@ -161,26 +156,66 @@ out: return ret; } -static struct buffer_head * ufs_inode_getfrag (struct inode *inode, - unsigned int fragment, unsigned int new_fragment, - unsigned int required, int *err, int metadata, long *phys, int *new) +static void ufs_clear_frag(struct inode *inode, struct buffer_head *bh) +{ + lock_buffer(bh); + memset(bh->b_data, 0, inode->i_sb->s_blocksize); + set_buffer_uptodate(bh); + mark_buffer_dirty(bh); + unlock_buffer(bh); + if (IS_SYNC(inode)) + sync_dirty_buffer(bh); +} + +static struct buffer_head * +ufs_clear_frags(struct inode *inode, sector_t beg, + unsigned int n, sector_t want) +{ + struct buffer_head *res = NULL, *bh; + sector_t end = beg + n; + + for (; beg < end; ++beg) { + bh = sb_getblk(inode->i_sb, beg); + ufs_clear_frag(inode, bh); + if (want != beg) + brelse(bh); + else + res = bh; + } + BUG_ON(!res); + return res; +} + +/** + * ufs_inode_getfrag() - allocate new fragment(s) + * @inode - pointer to inode + * @fragment - number of `fragment' which hold pointer + * to new allocated fragment(s) + * @new_fragment - number of new allocated fragment(s) + * @required - how many fragment(s) we require + * @err - we set it if something wrong + * @phys - pointer to where we save physical number of new allocated fragments, + * NULL if we allocate not data(indirect blocks for example). + * @new - we set it if we allocate new block + * @locked_page - for ufs_new_fragments() + */ +static struct buffer_head * +ufs_inode_getfrag(struct inode *inode, unsigned int fragment, + sector_t new_fragment, unsigned int required, int *err, + long *phys, int *new, struct page *locked_page) { struct ufs_inode_info *ufsi = UFS_I(inode); - struct super_block * sb; - struct ufs_sb_private_info * uspi; + struct super_block *sb = inode->i_sb; + struct ufs_sb_private_info *uspi = UFS_SB(sb)->s_uspi; struct buffer_head * result; unsigned block, blockoff, lastfrag, lastblock, lastblockoff; unsigned tmp, goal; __fs32 * p, * p2; - unsigned flags = 0; - UFSD(("ENTER, ino %lu, fragment %u, new_fragment %u, required %u\n", - inode->i_ino, fragment, new_fragment, required)) + UFSD("ENTER, ino %lu, fragment %u, new_fragment %llu, required %u, " + "metadata %d\n", inode->i_ino, fragment, + (unsigned long long)new_fragment, required, !phys); - sb = inode->i_sb; - uspi = UFS_SB(sb)->s_uspi; - - flags = UFS_SB(sb)->s_flags; /* TODO : to be done for write support if ( (flags & UFS_TYPE_MASK) == UFS_TYPE_UFS2) goto ufs2; @@ -195,16 +230,16 @@ repeat: tmp = fs32_to_cpu(sb, *p); lastfrag = ufsi->i_lastfrag; if (tmp && fragment < lastfrag) { - if (metadata) { + if (!phys) { result = sb_getblk(sb, uspi->s_sbbase + tmp + blockoff); if (tmp == fs32_to_cpu(sb, *p)) { - UFSD(("EXIT, result %u\n", tmp + blockoff)) + UFSD("EXIT, result %u\n", tmp + blockoff); return result; } brelse (result); goto repeat; } else { - *phys = tmp; + *phys = tmp + blockoff; return NULL; } } @@ -221,7 +256,8 @@ repeat: if (lastblockoff) { p2 = ufsi->i_u1.i_data + lastblock; tmp = ufs_new_fragments (inode, p2, lastfrag, - fs32_to_cpu(sb, *p2), uspi->s_fpb - lastblockoff, err); + fs32_to_cpu(sb, *p2), uspi->s_fpb - lastblockoff, + err, locked_page); if (!tmp) { if (lastfrag != ufsi->i_lastfrag) goto repeat; @@ -231,25 +267,31 @@ repeat: lastfrag = ufsi->i_lastfrag; } - goal = fs32_to_cpu(sb, ufsi->i_u1.i_data[lastblock]) + uspi->s_fpb; + tmp = fs32_to_cpu(sb, ufsi->i_u1.i_data[lastblock]); + if (tmp) + goal = tmp + uspi->s_fpb; tmp = ufs_new_fragments (inode, p, fragment - blockoff, - goal, required + blockoff, err); + goal, required + blockoff, + err, locked_page); } /* * We will extend last allocated block */ else if (lastblock == block) { - tmp = ufs_new_fragments (inode, p, fragment - (blockoff - lastblockoff), - fs32_to_cpu(sb, *p), required + (blockoff - lastblockoff), err); - } + tmp = ufs_new_fragments(inode, p, fragment - (blockoff - lastblockoff), + fs32_to_cpu(sb, *p), required + (blockoff - lastblockoff), + err, locked_page); + } else /* (lastblock > block) */ { /* * We will allocate new block before last allocated block */ - else /* (lastblock > block) */ { - if (lastblock && (tmp = fs32_to_cpu(sb, ufsi->i_u1.i_data[lastblock-1]))) - goal = tmp + uspi->s_fpb; - tmp = ufs_new_fragments (inode, p, fragment - blockoff, - goal, uspi->s_fpb, err); + if (block) { + tmp = fs32_to_cpu(sb, ufsi->i_u1.i_data[block-1]); + if (tmp) + goal = tmp + uspi->s_fpb; + } + tmp = ufs_new_fragments(inode, p, fragment - blockoff, + goal, uspi->s_fpb, err, locked_page); } if (!tmp) { if ((!blockoff && *p) || @@ -259,14 +301,10 @@ repeat: return NULL; } - /* The nullification of framgents done in ufs/balloc.c is - * something I don't have the stomache to move into here right - * now. -DaveM - */ - if (metadata) { - result = sb_getblk(inode->i_sb, tmp + blockoff); + if (!phys) { + result = ufs_clear_frags(inode, tmp, required, tmp + blockoff); } else { - *phys = tmp; + *phys = tmp + blockoff; result = NULL; *err = 0; *new = 1; @@ -276,7 +314,7 @@ repeat: if (IS_SYNC(inode)) ufs_sync_inode (inode); mark_inode_dirty(inode); - UFSD(("EXIT, result %u\n", tmp + blockoff)) + UFSD("EXIT, result %u\n", tmp + blockoff); return result; /* This part : To be implemented .... @@ -295,22 +333,35 @@ repeat2: */ } -static struct buffer_head * ufs_block_getfrag (struct inode *inode, - struct buffer_head *bh, unsigned int fragment, unsigned int new_fragment, - unsigned int blocksize, int * err, int metadata, long *phys, int *new) +/** + * ufs_inode_getblock() - allocate new block + * @inode - pointer to inode + * @bh - pointer to block which hold "pointer" to new allocated block + * @fragment - number of `fragment' which hold pointer + * to new allocated block + * @new_fragment - number of new allocated fragment + * (block will hold this fragment and also uspi->s_fpb-1) + * @err - see ufs_inode_getfrag() + * @phys - see ufs_inode_getfrag() + * @new - see ufs_inode_getfrag() + * @locked_page - see ufs_inode_getfrag() + */ +static struct buffer_head * +ufs_inode_getblock(struct inode *inode, struct buffer_head *bh, + unsigned int fragment, sector_t new_fragment, int *err, + long *phys, int *new, struct page *locked_page) { - struct super_block * sb; - struct ufs_sb_private_info * uspi; + struct super_block *sb = inode->i_sb; + struct ufs_sb_private_info *uspi = UFS_SB(sb)->s_uspi; struct buffer_head * result; unsigned tmp, goal, block, blockoff; __fs32 * p; - sb = inode->i_sb; - uspi = UFS_SB(sb)->s_uspi; block = ufs_fragstoblks (fragment); blockoff = ufs_fragnum (fragment); - UFSD(("ENTER, ino %lu, fragment %u, new_fragment %u\n", inode->i_ino, fragment, new_fragment)) + UFSD("ENTER, ino %lu, fragment %u, new_fragment %llu, metadata %d\n", + inode->i_ino, fragment, (unsigned long long)new_fragment, !phys); result = NULL; if (!bh) @@ -326,37 +377,36 @@ static struct buffer_head * ufs_block_getfrag (struct inode *inode, repeat: tmp = fs32_to_cpu(sb, *p); if (tmp) { - if (metadata) { + if (!phys) { result = sb_getblk(sb, uspi->s_sbbase + tmp + blockoff); if (tmp == fs32_to_cpu(sb, *p)) goto out; brelse (result); goto repeat; } else { - *phys = tmp; + *phys = tmp + blockoff; goto out; } } - if (block && (tmp = fs32_to_cpu(sb, ((__fs32*)bh->b_data)[block-1]) + uspi->s_fpb)) + if (block && (tmp = fs32_to_cpu(sb, ((__fs32*)bh->b_data)[block-1]))) goal = tmp + uspi->s_fpb; else goal = bh->b_blocknr + uspi->s_fpb; - tmp = ufs_new_fragments (inode, p, ufs_blknum(new_fragment), goal, uspi->s_fpb, err); + tmp = ufs_new_fragments(inode, p, ufs_blknum(new_fragment), goal, + uspi->s_fpb, err, locked_page); if (!tmp) { if (fs32_to_cpu(sb, *p)) goto repeat; goto out; } - /* The nullification of framgents done in ufs/balloc.c is - * something I don't have the stomache to move into here right - * now. -DaveM - */ - if (metadata) { - result = sb_getblk(sb, tmp + blockoff); + + if (!phys) { + result = ufs_clear_frags(inode, tmp, uspi->s_fpb, + tmp + blockoff); } else { - *phys = tmp; + *phys = tmp + blockoff; *new = 1; } @@ -365,18 +415,19 @@ repeat: sync_dirty_buffer(bh); inode->i_ctime = CURRENT_TIME_SEC; mark_inode_dirty(inode); - UFSD(("result %u\n", tmp + blockoff)); + UFSD("result %u\n", tmp + blockoff); out: brelse (bh); - UFSD(("EXIT\n")); + UFSD("EXIT\n"); return result; } -/* - * This function gets the block which contains the fragment. +/** + * ufs_getfrag_bloc() - `get_block_t' function, interface between UFS and + * readpage, writepage and so on */ -int ufs_getfrag_block (struct inode *inode, sector_t fragment, struct buffer_head *bh_result, int create) +int ufs_getfrag_block(struct inode *inode, sector_t fragment, struct buffer_head *bh_result, int create) { struct super_block * sb = inode->i_sb; struct ufs_sb_private_info * uspi = UFS_SB(sb)->s_uspi; @@ -387,7 +438,7 @@ int ufs_getfrag_block (struct inode *inode, sector_t fragment, struct buffer_hea if (!create) { phys64 = ufs_frag_map(inode, fragment); - UFSD(("phys64 = %llu \n",phys64)); + UFSD("phys64 = %llu\n", (unsigned long long)phys64); if (phys64) map_bh(bh_result, sb, phys64); return 0; @@ -402,7 +453,7 @@ int ufs_getfrag_block (struct inode *inode, sector_t fragment, struct buffer_hea lock_kernel(); - UFSD(("ENTER, ino %lu, fragment %llu\n", inode->i_ino, (unsigned long long)fragment)) + UFSD("ENTER, ino %lu, fragment %llu\n", inode->i_ino, (unsigned long long)fragment); if (fragment < 0) goto abort_negative; if (fragment > @@ -418,15 +469,15 @@ int ufs_getfrag_block (struct inode *inode, sector_t fragment, struct buffer_hea * it much more readable: */ #define GET_INODE_DATABLOCK(x) \ - ufs_inode_getfrag(inode, x, fragment, 1, &err, 0, &phys, &new) + ufs_inode_getfrag(inode, x, fragment, 1, &err, &phys, &new, bh_result->b_page) #define GET_INODE_PTR(x) \ - ufs_inode_getfrag(inode, x, fragment, uspi->s_fpb, &err, 1, NULL, NULL) + ufs_inode_getfrag(inode, x, fragment, uspi->s_fpb, &err, NULL, NULL, bh_result->b_page) #define GET_INDIRECT_DATABLOCK(x) \ - ufs_block_getfrag(inode, bh, x, fragment, sb->s_blocksize, \ - &err, 0, &phys, &new); + ufs_inode_getblock(inode, bh, x, fragment, \ + &err, &phys, &new, bh_result->b_page); #define GET_INDIRECT_PTR(x) \ - ufs_block_getfrag(inode, bh, x, fragment, sb->s_blocksize, \ - &err, 1, NULL, NULL); + ufs_inode_getblock(inode, bh, x, fragment, \ + &err, NULL, NULL, bh_result->b_page); if (ptr < UFS_NDIR_FRAGMENT) { bh = GET_INODE_DATABLOCK(ptr); @@ -474,8 +525,9 @@ abort_too_big: goto abort; } -struct buffer_head *ufs_getfrag(struct inode *inode, unsigned int fragment, - int create, int *err) +static struct buffer_head *ufs_getfrag(struct inode *inode, + unsigned int fragment, + int create, int *err) { struct buffer_head dummy; int error; @@ -502,7 +554,7 @@ struct buffer_head * ufs_bread (struct inode * inode, unsigned fragment, { struct buffer_head * bh; - UFSD(("ENTER, ino %lu, fragment %u\n", inode->i_ino, fragment)) + UFSD("ENTER, ino %lu, fragment %u\n", inode->i_ino, fragment); bh = ufs_getfrag (inode, fragment, create, err); if (!bh || buffer_uptodate(bh)) return bh; @@ -531,7 +583,7 @@ static sector_t ufs_bmap(struct address_space *mapping, sector_t block) { return generic_block_bmap(mapping,block,ufs_getfrag_block); } -struct address_space_operations ufs_aops = { +const struct address_space_operations ufs_aops = { .readpage = ufs_readpage, .writepage = ufs_writepage, .sync_page = block_sync_page, @@ -540,39 +592,34 @@ struct address_space_operations ufs_aops = { .bmap = ufs_bmap }; -void ufs_read_inode (struct inode * inode) +static void ufs_set_inode_ops(struct inode *inode) +{ + if (S_ISREG(inode->i_mode)) { + inode->i_op = &ufs_file_inode_operations; + inode->i_fop = &ufs_file_operations; + inode->i_mapping->a_ops = &ufs_aops; + } else if (S_ISDIR(inode->i_mode)) { + inode->i_op = &ufs_dir_inode_operations; + inode->i_fop = &ufs_dir_operations; + inode->i_mapping->a_ops = &ufs_aops; + } else if (S_ISLNK(inode->i_mode)) { + if (!inode->i_blocks) + inode->i_op = &ufs_fast_symlink_inode_operations; + else { + inode->i_op = &page_symlink_inode_operations; + inode->i_mapping->a_ops = &ufs_aops; + } + } else + init_special_inode(inode, inode->i_mode, + ufs_get_inode_dev(inode->i_sb, UFS_I(inode))); +} + +static void ufs1_read_inode(struct inode *inode, struct ufs_inode *ufs_inode) { struct ufs_inode_info *ufsi = UFS_I(inode); - struct super_block * sb; - struct ufs_sb_private_info * uspi; - struct ufs_inode * ufs_inode; - struct ufs2_inode *ufs2_inode; - struct buffer_head * bh; + struct super_block *sb = inode->i_sb; mode_t mode; unsigned i; - unsigned flags; - - UFSD(("ENTER, ino %lu\n", inode->i_ino)) - - sb = inode->i_sb; - uspi = UFS_SB(sb)->s_uspi; - flags = UFS_SB(sb)->s_flags; - - if (inode->i_ino < UFS_ROOTINO || - inode->i_ino > (uspi->s_ncg * uspi->s_ipg)) { - ufs_warning (sb, "ufs_read_inode", "bad inode number (%lu)\n", inode->i_ino); - goto bad_inode; - } - - bh = sb_bread(sb, uspi->s_sbbase + ufs_inotofsba(inode->i_ino)); - if (!bh) { - ufs_warning (sb, "ufs_read_inode", "unable to read inode %lu\n", inode->i_ino); - goto bad_inode; - } - if ((flags & UFS_TYPE_MASK) == UFS_TYPE_UFS2) - goto ufs2_inode; - - ufs_inode = (struct ufs_inode *) (bh->b_data + sizeof(struct ufs_inode) * ufs_inotofsbo(inode->i_ino)); /* * Copy data to the in-core inode. @@ -596,56 +643,29 @@ void ufs_read_inode (struct inode * inode) inode->i_atime.tv_nsec = 0; inode->i_ctime.tv_nsec = 0; inode->i_blocks = fs32_to_cpu(sb, ufs_inode->ui_blocks); - inode->i_blksize = PAGE_SIZE; /* This is the optimal IO size (for stat) */ - inode->i_version++; ufsi->i_flags = fs32_to_cpu(sb, ufs_inode->ui_flags); ufsi->i_gen = fs32_to_cpu(sb, ufs_inode->ui_gen); ufsi->i_shadow = fs32_to_cpu(sb, ufs_inode->ui_u3.ui_sun.ui_shadow); ufsi->i_oeftflag = fs32_to_cpu(sb, ufs_inode->ui_u3.ui_sun.ui_oeftflag); - ufsi->i_lastfrag = (inode->i_size + uspi->s_fsize - 1) >> uspi->s_fshift; + if (S_ISCHR(mode) || S_ISBLK(mode) || inode->i_blocks) { for (i = 0; i < (UFS_NDADDR + UFS_NINDIR); i++) ufsi->i_u1.i_data[i] = ufs_inode->ui_u2.ui_addr.ui_db[i]; - } - else { + } else { for (i = 0; i < (UFS_NDADDR + UFS_NINDIR) * 4; i++) ufsi->i_u1.i_symlink[i] = ufs_inode->ui_u2.ui_symlink[i]; } - ufsi->i_osync = 0; - - if (S_ISREG(inode->i_mode)) { - inode->i_op = &ufs_file_inode_operations; - inode->i_fop = &ufs_file_operations; - inode->i_mapping->a_ops = &ufs_aops; - } else if (S_ISDIR(inode->i_mode)) { - inode->i_op = &ufs_dir_inode_operations; - inode->i_fop = &ufs_dir_operations; - } else if (S_ISLNK(inode->i_mode)) { - if (!inode->i_blocks) - inode->i_op = &ufs_fast_symlink_inode_operations; - else { - inode->i_op = &page_symlink_inode_operations; - inode->i_mapping->a_ops = &ufs_aops; - } - } else - init_special_inode(inode, inode->i_mode, - ufs_get_inode_dev(sb, ufsi)); - - brelse (bh); - - UFSD(("EXIT\n")) - return; - -bad_inode: - make_bad_inode(inode); - return; - -ufs2_inode : - UFSD(("Reading ufs2 inode, ino %lu\n", inode->i_ino)) +} - ufs2_inode = (struct ufs2_inode *)(bh->b_data + sizeof(struct ufs2_inode) * ufs_inotofsbo(inode->i_ino)); +static void ufs2_read_inode(struct inode *inode, struct ufs2_inode *ufs2_inode) +{ + struct ufs_inode_info *ufsi = UFS_I(inode); + struct super_block *sb = inode->i_sb; + mode_t mode; + unsigned i; + UFSD("Reading ufs2 inode, ino %lu\n", inode->i_ino); /* * Copy data to the in-core inode. */ @@ -668,50 +688,74 @@ ufs2_inode : inode->i_atime.tv_nsec = 0; inode->i_ctime.tv_nsec = 0; inode->i_blocks = fs64_to_cpu(sb, ufs2_inode->ui_blocks); - inode->i_blksize = PAGE_SIZE; /*This is the optimal IO size(for stat)*/ - - inode->i_version++; ufsi->i_flags = fs32_to_cpu(sb, ufs2_inode->ui_flags); ufsi->i_gen = fs32_to_cpu(sb, ufs2_inode->ui_gen); /* ufsi->i_shadow = fs32_to_cpu(sb, ufs_inode->ui_u3.ui_sun.ui_shadow); ufsi->i_oeftflag = fs32_to_cpu(sb, ufs_inode->ui_u3.ui_sun.ui_oeftflag); */ - ufsi->i_lastfrag= (inode->i_size + uspi->s_fsize- 1) >> uspi->s_fshift; if (S_ISCHR(mode) || S_ISBLK(mode) || inode->i_blocks) { for (i = 0; i < (UFS_NDADDR + UFS_NINDIR); i++) ufsi->i_u1.u2_i_data[i] = ufs2_inode->ui_u2.ui_addr.ui_db[i]; - } - else { + } else { for (i = 0; i < (UFS_NDADDR + UFS_NINDIR) * 4; i++) ufsi->i_u1.i_symlink[i] = ufs2_inode->ui_u2.ui_symlink[i]; } +} + +void ufs_read_inode(struct inode * inode) +{ + struct ufs_inode_info *ufsi = UFS_I(inode); + struct super_block * sb; + struct ufs_sb_private_info * uspi; + struct buffer_head * bh; + + UFSD("ENTER, ino %lu\n", inode->i_ino); + + sb = inode->i_sb; + uspi = UFS_SB(sb)->s_uspi; + + if (inode->i_ino < UFS_ROOTINO || + inode->i_ino > (uspi->s_ncg * uspi->s_ipg)) { + ufs_warning(sb, "ufs_read_inode", "bad inode number (%lu)\n", + inode->i_ino); + goto bad_inode; + } + + bh = sb_bread(sb, uspi->s_sbbase + ufs_inotofsba(inode->i_ino)); + if (!bh) { + ufs_warning(sb, "ufs_read_inode", "unable to read inode %lu\n", + inode->i_ino); + goto bad_inode; + } + if ((UFS_SB(sb)->s_flags & UFS_TYPE_MASK) == UFS_TYPE_UFS2) { + struct ufs2_inode *ufs2_inode = (struct ufs2_inode *)bh->b_data; + + ufs2_read_inode(inode, + ufs2_inode + ufs_inotofsbo(inode->i_ino)); + } else { + struct ufs_inode *ufs_inode = (struct ufs_inode *)bh->b_data; + + ufs1_read_inode(inode, ufs_inode + ufs_inotofsbo(inode->i_ino)); + } + + inode->i_version++; + ufsi->i_lastfrag = + (inode->i_size + uspi->s_fsize - 1) >> uspi->s_fshift; + ufsi->i_dir_start_lookup = 0; ufsi->i_osync = 0; - if (S_ISREG(inode->i_mode)) { - inode->i_op = &ufs_file_inode_operations; - inode->i_fop = &ufs_file_operations; - inode->i_mapping->a_ops = &ufs_aops; - } else if (S_ISDIR(inode->i_mode)) { - inode->i_op = &ufs_dir_inode_operations; - inode->i_fop = &ufs_dir_operations; - } else if (S_ISLNK(inode->i_mode)) { - if (!inode->i_blocks) - inode->i_op = &ufs_fast_symlink_inode_operations; - else { - inode->i_op = &page_symlink_inode_operations; - inode->i_mapping->a_ops = &ufs_aops; - } - } else /* TODO : here ...*/ - init_special_inode(inode, inode->i_mode, - ufs_get_inode_dev(sb, ufsi)); + ufs_set_inode_ops(inode); brelse(bh); - UFSD(("EXIT\n")) + UFSD("EXIT\n"); return; + +bad_inode: + make_bad_inode(inode); } static int ufs_update_inode(struct inode * inode, int do_sync) @@ -724,7 +768,7 @@ static int ufs_update_inode(struct inode * inode, int do_sync) unsigned i; unsigned flags; - UFSD(("ENTER, ino %lu\n", inode->i_ino)) + UFSD("ENTER, ino %lu\n", inode->i_ino); sb = inode->i_sb; uspi = UFS_SB(sb)->s_uspi; @@ -785,7 +829,7 @@ static int ufs_update_inode(struct inode * inode, int do_sync) sync_dirty_buffer(bh); brelse (bh); - UFSD(("EXIT\n")) + UFSD("EXIT\n"); return 0; } @@ -805,14 +849,17 @@ int ufs_sync_inode (struct inode *inode) void ufs_delete_inode (struct inode * inode) { + loff_t old_i_size; + truncate_inode_pages(&inode->i_data, 0); /*UFS_I(inode)->i_dtime = CURRENT_TIME;*/ lock_kernel(); mark_inode_dirty(inode); ufs_update_inode(inode, IS_SYNC(inode)); + old_i_size = inode->i_size; inode->i_size = 0; - if (inode->i_blocks) - ufs_truncate (inode); + if (inode->i_blocks && ufs_truncate(inode, old_i_size)) + ufs_warning(inode->i_sb, __FUNCTION__, "ufs_truncate failed\n"); ufs_free_inode (inode); unlock_kernel(); } diff --git a/fs/ufs/namei.c b/fs/ufs/namei.c index 8d5f98a01c7..d344b411e26 100644 --- a/fs/ufs/namei.c +++ b/fs/ufs/namei.c @@ -1,6 +1,9 @@ /* * linux/fs/ufs/namei.c * + * Migration to usage of "page cache" on May 2006 by + * Evgeniy Dushistov <dushistov@mail.ru> based on ext2 code base. + * * Copyright (C) 1998 * Daniel Pirkl <daniel.pirkl@email.cz> * Charles University, Faculty of Mathematics and Physics @@ -28,21 +31,9 @@ #include <linux/fs.h> #include <linux/ufs_fs.h> #include <linux/smp_lock.h> -#include <linux/buffer_head.h> #include "swab.h" /* will go away - see comment in mknod() */ #include "util.h" -/* -#undef UFS_NAMEI_DEBUG -*/ -#define UFS_NAMEI_DEBUG - -#ifdef UFS_NAMEI_DEBUG -#define UFSD(x) printk("(%s, %d), %s: ", __FILE__, __LINE__, __FUNCTION__); printk x; -#else -#define UFSD(x) -#endif - static inline int ufs_add_nondir(struct dentry *dentry, struct inode *inode) { int err = ufs_add_link(dentry, inode); @@ -88,8 +79,13 @@ static struct dentry *ufs_lookup(struct inode * dir, struct dentry *dentry, stru static int ufs_create (struct inode * dir, struct dentry * dentry, int mode, struct nameidata *nd) { - struct inode * inode = ufs_new_inode(dir, mode); - int err = PTR_ERR(inode); + struct inode *inode; + int err; + + UFSD("BEGIN\n"); + inode = ufs_new_inode(dir, mode); + err = PTR_ERR(inode); + if (!IS_ERR(inode)) { inode->i_op = &ufs_file_inode_operations; inode->i_fop = &ufs_file_operations; @@ -99,6 +95,7 @@ static int ufs_create (struct inode * dir, struct dentry * dentry, int mode, err = ufs_add_nondir(dentry, inode); unlock_kernel(); } + UFSD("END: err=%d\n", err); return err; } @@ -132,7 +129,7 @@ static int ufs_symlink (struct inode * dir, struct dentry * dentry, struct inode * inode; if (l > sb->s_blocksize) - goto out; + goto out_notlocked; lock_kernel(); inode = ufs_new_inode(dir, S_IFLNK | S_IRWXUGO); @@ -158,6 +155,7 @@ static int ufs_symlink (struct inode * dir, struct dentry * dentry, err = ufs_add_nondir(dentry, inode); out: unlock_kernel(); +out_notlocked: return err; out_fail: @@ -205,6 +203,7 @@ static int ufs_mkdir(struct inode * dir, struct dentry * dentry, int mode) inode->i_op = &ufs_dir_inode_operations; inode->i_fop = &ufs_dir_operations; + inode->i_mapping->a_ops = &ufs_aops; inode_inc_link_count(inode); @@ -231,19 +230,18 @@ out_dir: goto out; } -static int ufs_unlink(struct inode * dir, struct dentry *dentry) +static int ufs_unlink(struct inode *dir, struct dentry *dentry) { struct inode * inode = dentry->d_inode; - struct buffer_head * bh; - struct ufs_dir_entry * de; + struct ufs_dir_entry *de; + struct page *page; int err = -ENOENT; - lock_kernel(); - de = ufs_find_entry (dentry, &bh); + de = ufs_find_entry(dir, dentry, &page); if (!de) goto out; - err = ufs_delete_entry (dir, de, bh); + err = ufs_delete_entry(dir, de, page); if (err) goto out; @@ -251,7 +249,6 @@ static int ufs_unlink(struct inode * dir, struct dentry *dentry) inode_dec_link_count(inode); err = 0; out: - unlock_kernel(); return err; } @@ -273,42 +270,42 @@ static int ufs_rmdir (struct inode * dir, struct dentry *dentry) return err; } -static int ufs_rename (struct inode * old_dir, struct dentry * old_dentry, - struct inode * new_dir, struct dentry * new_dentry ) +static int ufs_rename(struct inode *old_dir, struct dentry *old_dentry, + struct inode *new_dir, struct dentry *new_dentry) { struct inode *old_inode = old_dentry->d_inode; struct inode *new_inode = new_dentry->d_inode; - struct buffer_head *dir_bh = NULL; - struct ufs_dir_entry *dir_de = NULL; - struct buffer_head *old_bh; + struct page *dir_page = NULL; + struct ufs_dir_entry * dir_de = NULL; + struct page *old_page; struct ufs_dir_entry *old_de; int err = -ENOENT; - lock_kernel(); - old_de = ufs_find_entry (old_dentry, &old_bh); + old_de = ufs_find_entry(old_dir, old_dentry, &old_page); if (!old_de) goto out; if (S_ISDIR(old_inode->i_mode)) { err = -EIO; - dir_de = ufs_dotdot(old_inode, &dir_bh); + dir_de = ufs_dotdot(old_inode, &dir_page); if (!dir_de) goto out_old; } if (new_inode) { - struct buffer_head *new_bh; + struct page *new_page; struct ufs_dir_entry *new_de; err = -ENOTEMPTY; - if (dir_de && !ufs_empty_dir (new_inode)) + if (dir_de && !ufs_empty_dir(new_inode)) goto out_dir; + err = -ENOENT; - new_de = ufs_find_entry (new_dentry, &new_bh); + new_de = ufs_find_entry(new_dir, new_dentry, &new_page); if (!new_de) goto out_dir; inode_inc_link_count(old_inode); - ufs_set_link(new_dir, new_de, new_bh, old_inode); + ufs_set_link(new_dir, new_de, new_page, old_inode); new_inode->i_ctime = CURRENT_TIME_SEC; if (dir_de) new_inode->i_nlink--; @@ -329,24 +326,32 @@ static int ufs_rename (struct inode * old_dir, struct dentry * old_dentry, inode_inc_link_count(new_dir); } - ufs_delete_entry (old_dir, old_de, old_bh); + /* + * Like most other Unix systems, set the ctime for inodes on a + * rename. + * inode_dec_link_count() will mark the inode dirty. + */ + old_inode->i_ctime = CURRENT_TIME_SEC; + ufs_delete_entry(old_dir, old_de, old_page); inode_dec_link_count(old_inode); if (dir_de) { - ufs_set_link(old_inode, dir_de, dir_bh, new_dir); + ufs_set_link(old_inode, dir_de, dir_page, new_dir); inode_dec_link_count(old_dir); } - unlock_kernel(); return 0; + out_dir: - if (dir_de) - brelse(dir_bh); + if (dir_de) { + kunmap(dir_page); + page_cache_release(dir_page); + } out_old: - brelse (old_bh); + kunmap(old_page); + page_cache_release(old_page); out: - unlock_kernel(); return err; } diff --git a/fs/ufs/super.c b/fs/ufs/super.c index fe5ab2aa289..ec79e3091d1 100644 --- a/fs/ufs/super.c +++ b/fs/ufs/super.c @@ -64,7 +64,6 @@ */ -#include <linux/config.h> #include <linux/module.h> #include <linux/bitops.h> @@ -90,95 +89,84 @@ #include "swab.h" #include "util.h" -#undef UFS_SUPER_DEBUG -#undef UFS_SUPER_DEBUG_MORE - - -#undef UFS_SUPER_DEBUG_MORE -#ifdef UFS_SUPER_DEBUG -#define UFSD(x) printk("(%s, %d), %s: ", __FILE__, __LINE__, __FUNCTION__); printk x; -#else -#define UFSD(x) -#endif - -#ifdef UFS_SUPER_DEBUG_MORE +#ifdef CONFIG_UFS_DEBUG /* * Print contents of ufs_super_block, useful for debugging */ -void ufs_print_super_stuff(struct super_block *sb, - struct ufs_super_block_first * usb1, - struct ufs_super_block_second * usb2, - struct ufs_super_block_third * usb3) +static void ufs_print_super_stuff(struct super_block *sb, unsigned flags, + struct ufs_super_block_first *usb1, + struct ufs_super_block_second *usb2, + struct ufs_super_block_third *usb3) { printk("ufs_print_super_stuff\n"); - printk("size of usb: %u\n", sizeof(struct ufs_super_block)); - printk(" magic: 0x%x\n", fs32_to_cpu(sb, usb3->fs_magic)); - printk(" sblkno: %u\n", fs32_to_cpu(sb, usb1->fs_sblkno)); - printk(" cblkno: %u\n", fs32_to_cpu(sb, usb1->fs_cblkno)); - printk(" iblkno: %u\n", fs32_to_cpu(sb, usb1->fs_iblkno)); - printk(" dblkno: %u\n", fs32_to_cpu(sb, usb1->fs_dblkno)); - printk(" cgoffset: %u\n", fs32_to_cpu(sb, usb1->fs_cgoffset)); - printk(" ~cgmask: 0x%x\n", ~fs32_to_cpu(sb, usb1->fs_cgmask)); - printk(" size: %u\n", fs32_to_cpu(sb, usb1->fs_size)); - printk(" dsize: %u\n", fs32_to_cpu(sb, usb1->fs_dsize)); - printk(" ncg: %u\n", fs32_to_cpu(sb, usb1->fs_ncg)); - printk(" bsize: %u\n", fs32_to_cpu(sb, usb1->fs_bsize)); - printk(" fsize: %u\n", fs32_to_cpu(sb, usb1->fs_fsize)); - printk(" frag: %u\n", fs32_to_cpu(sb, usb1->fs_frag)); - printk(" fragshift: %u\n", fs32_to_cpu(sb, usb1->fs_fragshift)); - printk(" ~fmask: %u\n", ~fs32_to_cpu(sb, usb1->fs_fmask)); - printk(" fshift: %u\n", fs32_to_cpu(sb, usb1->fs_fshift)); - printk(" sbsize: %u\n", fs32_to_cpu(sb, usb1->fs_sbsize)); - printk(" spc: %u\n", fs32_to_cpu(sb, usb1->fs_spc)); - printk(" cpg: %u\n", fs32_to_cpu(sb, usb1->fs_cpg)); - printk(" ipg: %u\n", fs32_to_cpu(sb, usb1->fs_ipg)); - printk(" fpg: %u\n", fs32_to_cpu(sb, usb1->fs_fpg)); - printk(" csaddr: %u\n", fs32_to_cpu(sb, usb1->fs_csaddr)); - printk(" cssize: %u\n", fs32_to_cpu(sb, usb1->fs_cssize)); - printk(" cgsize: %u\n", fs32_to_cpu(sb, usb1->fs_cgsize)); - printk(" fstodb: %u\n", fs32_to_cpu(sb, usb1->fs_fsbtodb)); - printk(" contigsumsize: %d\n", fs32_to_cpu(sb, usb3->fs_u2.fs_44.fs_contigsumsize)); - printk(" postblformat: %u\n", fs32_to_cpu(sb, usb3->fs_postblformat)); - printk(" nrpos: %u\n", fs32_to_cpu(sb, usb3->fs_nrpos)); - printk(" ndir %u\n", fs32_to_cpu(sb, usb1->fs_cstotal.cs_ndir)); - printk(" nifree %u\n", fs32_to_cpu(sb, usb1->fs_cstotal.cs_nifree)); - printk(" nbfree %u\n", fs32_to_cpu(sb, usb1->fs_cstotal.cs_nbfree)); - printk(" nffree %u\n", fs32_to_cpu(sb, usb1->fs_cstotal.cs_nffree)); - printk("\n"); -} - -/* - * Print contents of ufs2 ufs_super_block, useful for debugging - */ -void ufs2_print_super_stuff( - struct super_block *sb, - struct ufs_super_block *usb) -{ - printk("ufs_print_super_stuff\n"); - printk("size of usb: %u\n", sizeof(struct ufs_super_block)); - printk(" magic: 0x%x\n", fs32_to_cpu(sb, usb->fs_magic)); - printk(" fs_size: %u\n",fs64_to_cpu(sb, usb->fs_u11.fs_u2.fs_size)); - printk(" fs_dsize: %u\n",fs64_to_cpu(sb, usb->fs_u11.fs_u2.fs_dsize)); - printk(" bsize: %u\n", fs32_to_cpu(usb, usb->fs_bsize)); - printk(" fsize: %u\n", fs32_to_cpu(usb, usb->fs_fsize)); - printk(" fs_volname: %s\n", usb->fs_u11.fs_u2.fs_volname); - printk(" fs_fsmnt: %s\n", usb->fs_u11.fs_u2.fs_fsmnt); - printk(" fs_sblockloc: %u\n",fs64_to_cpu(sb, - usb->fs_u11.fs_u2.fs_sblockloc)); - printk(" cs_ndir(No of dirs): %u\n",fs64_to_cpu(sb, - usb->fs_u11.fs_u2.fs_cstotal.cs_ndir)); - printk(" cs_nbfree(No of free blocks): %u\n",fs64_to_cpu(sb, - usb->fs_u11.fs_u2.fs_cstotal.cs_nbfree)); + printk(" magic: 0x%x\n", fs32_to_cpu(sb, usb3->fs_magic)); + if ((flags & UFS_TYPE_MASK) == UFS_TYPE_UFS2) { + printk(" fs_size: %llu\n", (unsigned long long) + fs64_to_cpu(sb, usb3->fs_un1.fs_u2.fs_size)); + printk(" fs_dsize: %llu\n", (unsigned long long) + fs64_to_cpu(sb, usb3->fs_un1.fs_u2.fs_dsize)); + printk(" bsize: %u\n", + fs32_to_cpu(sb, usb1->fs_bsize)); + printk(" fsize: %u\n", + fs32_to_cpu(sb, usb1->fs_fsize)); + printk(" fs_volname: %s\n", usb2->fs_un.fs_u2.fs_volname); + printk(" fs_sblockloc: %llu\n", (unsigned long long) + fs64_to_cpu(sb, usb2->fs_un.fs_u2.fs_sblockloc)); + printk(" cs_ndir(No of dirs): %llu\n", (unsigned long long) + fs64_to_cpu(sb, usb2->fs_un.fs_u2.cs_ndir)); + printk(" cs_nbfree(No of free blocks): %llu\n", + (unsigned long long) + fs64_to_cpu(sb, usb2->fs_un.fs_u2.cs_nbfree)); + } else { + printk(" sblkno: %u\n", fs32_to_cpu(sb, usb1->fs_sblkno)); + printk(" cblkno: %u\n", fs32_to_cpu(sb, usb1->fs_cblkno)); + printk(" iblkno: %u\n", fs32_to_cpu(sb, usb1->fs_iblkno)); + printk(" dblkno: %u\n", fs32_to_cpu(sb, usb1->fs_dblkno)); + printk(" cgoffset: %u\n", + fs32_to_cpu(sb, usb1->fs_cgoffset)); + printk(" ~cgmask: 0x%x\n", + ~fs32_to_cpu(sb, usb1->fs_cgmask)); + printk(" size: %u\n", fs32_to_cpu(sb, usb1->fs_size)); + printk(" dsize: %u\n", fs32_to_cpu(sb, usb1->fs_dsize)); + printk(" ncg: %u\n", fs32_to_cpu(sb, usb1->fs_ncg)); + printk(" bsize: %u\n", fs32_to_cpu(sb, usb1->fs_bsize)); + printk(" fsize: %u\n", fs32_to_cpu(sb, usb1->fs_fsize)); + printk(" frag: %u\n", fs32_to_cpu(sb, usb1->fs_frag)); + printk(" fragshift: %u\n", + fs32_to_cpu(sb, usb1->fs_fragshift)); + printk(" ~fmask: %u\n", ~fs32_to_cpu(sb, usb1->fs_fmask)); + printk(" fshift: %u\n", fs32_to_cpu(sb, usb1->fs_fshift)); + printk(" sbsize: %u\n", fs32_to_cpu(sb, usb1->fs_sbsize)); + printk(" spc: %u\n", fs32_to_cpu(sb, usb1->fs_spc)); + printk(" cpg: %u\n", fs32_to_cpu(sb, usb1->fs_cpg)); + printk(" ipg: %u\n", fs32_to_cpu(sb, usb1->fs_ipg)); + printk(" fpg: %u\n", fs32_to_cpu(sb, usb1->fs_fpg)); + printk(" csaddr: %u\n", fs32_to_cpu(sb, usb1->fs_csaddr)); + printk(" cssize: %u\n", fs32_to_cpu(sb, usb1->fs_cssize)); + printk(" cgsize: %u\n", fs32_to_cpu(sb, usb1->fs_cgsize)); + printk(" fstodb: %u\n", + fs32_to_cpu(sb, usb1->fs_fsbtodb)); + printk(" nrpos: %u\n", fs32_to_cpu(sb, usb3->fs_nrpos)); + printk(" ndir %u\n", + fs32_to_cpu(sb, usb1->fs_cstotal.cs_ndir)); + printk(" nifree %u\n", + fs32_to_cpu(sb, usb1->fs_cstotal.cs_nifree)); + printk(" nbfree %u\n", + fs32_to_cpu(sb, usb1->fs_cstotal.cs_nbfree)); + printk(" nffree %u\n", + fs32_to_cpu(sb, usb1->fs_cstotal.cs_nffree)); + } printk("\n"); } /* * Print contents of ufs_cylinder_group, useful for debugging */ -void ufs_print_cylinder_stuff(struct super_block *sb, struct ufs_cylinder_group *cg) +static void ufs_print_cylinder_stuff(struct super_block *sb, + struct ufs_cylinder_group *cg) { printk("\nufs_print_cylinder_stuff\n"); - printk("size of ucg: %u\n", sizeof(struct ufs_cylinder_group)); + printk("size of ucg: %zu\n", sizeof(struct ufs_cylinder_group)); printk(" magic: %x\n", fs32_to_cpu(sb, cg->cg_magic)); printk(" time: %u\n", fs32_to_cpu(sb, cg->cg_time)); printk(" cgx: %u\n", fs32_to_cpu(sb, cg->cg_cgx)); @@ -202,12 +190,18 @@ void ufs_print_cylinder_stuff(struct super_block *sb, struct ufs_cylinder_group printk(" iuseoff: %u\n", fs32_to_cpu(sb, cg->cg_iusedoff)); printk(" freeoff: %u\n", fs32_to_cpu(sb, cg->cg_freeoff)); printk(" nextfreeoff: %u\n", fs32_to_cpu(sb, cg->cg_nextfreeoff)); - printk(" clustersumoff %u\n", fs32_to_cpu(sb, cg->cg_u.cg_44.cg_clustersumoff)); - printk(" clusteroff %u\n", fs32_to_cpu(sb, cg->cg_u.cg_44.cg_clusteroff)); - printk(" nclusterblks %u\n", fs32_to_cpu(sb, cg->cg_u.cg_44.cg_nclusterblks)); + printk(" clustersumoff %u\n", + fs32_to_cpu(sb, cg->cg_u.cg_44.cg_clustersumoff)); + printk(" clusteroff %u\n", + fs32_to_cpu(sb, cg->cg_u.cg_44.cg_clusteroff)); + printk(" nclusterblks %u\n", + fs32_to_cpu(sb, cg->cg_u.cg_44.cg_nclusterblks)); printk("\n"); } -#endif /* UFS_SUPER_DEBUG_MORE */ +#else +# define ufs_print_super_stuff(sb, flags, usb1, usb2, usb3) /**/ +# define ufs_print_cylinder_stuff(sb, cg) /**/ +#endif /* CONFIG_UFS_DEBUG */ static struct super_operations ufs_super_ops; @@ -225,7 +219,7 @@ void ufs_error (struct super_block * sb, const char * function, if (!(sb->s_flags & MS_RDONLY)) { usb1->fs_clean = UFS_FSBAD; - ubh_mark_buffer_dirty(USPI_UBH); + ubh_mark_buffer_dirty(USPI_UBH(uspi)); sb->s_dirt = 1; sb->s_flags |= MS_RDONLY; } @@ -257,7 +251,7 @@ void ufs_panic (struct super_block * sb, const char * function, if (!(sb->s_flags & MS_RDONLY)) { usb1->fs_clean = UFS_FSBAD; - ubh_mark_buffer_dirty(USPI_UBH); + ubh_mark_buffer_dirty(USPI_UBH(uspi)); sb->s_dirt = 1; } va_start (args, fmt); @@ -309,7 +303,7 @@ static int ufs_parse_options (char * options, unsigned * mount_options) { char * p; - UFSD(("ENTER\n")) + UFSD("ENTER\n"); if (!options) return 1; @@ -386,27 +380,57 @@ static int ufs_parse_options (char * options, unsigned * mount_options) } /* + * Diffrent types of UFS hold fs_cstotal in different + * places, and use diffrent data structure for it. + * To make things simplier we just copy fs_cstotal to ufs_sb_private_info + */ +static void ufs_setup_cstotal(struct super_block *sb) +{ + struct ufs_sb_info *sbi = UFS_SB(sb); + struct ufs_sb_private_info *uspi = sbi->s_uspi; + struct ufs_super_block_first *usb1; + struct ufs_super_block_second *usb2; + struct ufs_super_block_third *usb3; + unsigned mtype = sbi->s_mount_opt & UFS_MOUNT_UFSTYPE; + + UFSD("ENTER, mtype=%u\n", mtype); + usb1 = ubh_get_usb_first(uspi); + usb2 = ubh_get_usb_second(uspi); + usb3 = ubh_get_usb_third(uspi); + + if ((mtype == UFS_MOUNT_UFSTYPE_44BSD && + (usb1->fs_flags & UFS_FLAGS_UPDATED)) || + mtype == UFS_MOUNT_UFSTYPE_UFS2) { + /*we have statistic in different place, then usual*/ + uspi->cs_total.cs_ndir = fs64_to_cpu(sb, usb2->fs_un.fs_u2.cs_ndir); + uspi->cs_total.cs_nbfree = fs64_to_cpu(sb, usb2->fs_un.fs_u2.cs_nbfree); + uspi->cs_total.cs_nifree = fs64_to_cpu(sb, usb3->fs_un1.fs_u2.cs_nifree); + uspi->cs_total.cs_nffree = fs64_to_cpu(sb, usb3->fs_un1.fs_u2.cs_nffree); + } else { + uspi->cs_total.cs_ndir = fs32_to_cpu(sb, usb1->fs_cstotal.cs_ndir); + uspi->cs_total.cs_nbfree = fs32_to_cpu(sb, usb1->fs_cstotal.cs_nbfree); + uspi->cs_total.cs_nifree = fs32_to_cpu(sb, usb1->fs_cstotal.cs_nifree); + uspi->cs_total.cs_nffree = fs32_to_cpu(sb, usb1->fs_cstotal.cs_nffree); + } + UFSD("EXIT\n"); +} + +/* * Read on-disk structures associated with cylinder groups */ -static int ufs_read_cylinder_structures (struct super_block *sb) +static int ufs_read_cylinder_structures(struct super_block *sb) { - struct ufs_sb_info * sbi = UFS_SB(sb); - struct ufs_sb_private_info * uspi; - struct ufs_super_block *usb; + struct ufs_sb_info *sbi = UFS_SB(sb); + struct ufs_sb_private_info *uspi = sbi->s_uspi; + unsigned flags = sbi->s_flags; struct ufs_buffer_head * ubh; unsigned char * base, * space; unsigned size, blks, i; - unsigned flags = 0; - - UFSD(("ENTER\n")) - - uspi = sbi->s_uspi; + struct ufs_super_block_third *usb3; - usb = (struct ufs_super_block *) - ((struct ufs_buffer_head *)uspi)->bh[0]->b_data; + UFSD("ENTER\n"); - flags = UFS_SB(sb)->s_flags; - + usb3 = ubh_get_usb_third(uspi); /* * Read cs structures from (usually) first data block * on the device. @@ -424,7 +448,7 @@ static int ufs_read_cylinder_structures (struct super_block *sb) if ((flags & UFS_TYPE_MASK) == UFS_TYPE_UFS2) ubh = ubh_bread(sb, - fs64_to_cpu(sb, usb->fs_u11.fs_u2.fs_csaddr) + i, size); + fs64_to_cpu(sb, usb3->fs_un1.fs_u2.fs_csaddr) + i, size); else ubh = ubh_bread(sb, uspi->s_csaddr + i, size); @@ -451,14 +475,13 @@ static int ufs_read_cylinder_structures (struct super_block *sb) sbi->s_cgno[i] = UFS_CGNO_EMPTY; } for (i = 0; i < uspi->s_ncg; i++) { - UFSD(("read cg %u\n", i)) + UFSD("read cg %u\n", i); if (!(sbi->s_ucg[i] = sb_bread(sb, ufs_cgcmin(i)))) goto failed; if (!ufs_cg_chkmagic (sb, (struct ufs_cylinder_group *) sbi->s_ucg[i]->b_data)) goto failed; -#ifdef UFS_SUPER_DEBUG_MORE + ufs_print_cylinder_stuff(sb, (struct ufs_cylinder_group *) sbi->s_ucg[i]->b_data); -#endif } for (i = 0; i < UFS_MAX_GROUP_LOADED; i++) { if (!(sbi->s_ucpi[i] = kmalloc (sizeof(struct ufs_cg_private_info), GFP_KERNEL))) @@ -466,7 +489,7 @@ static int ufs_read_cylinder_structures (struct super_block *sb) sbi->s_cgno[i] = UFS_CGNO_EMPTY; } sbi->s_cg_loaded = 0; - UFSD(("EXIT\n")) + UFSD("EXIT\n"); return 1; failed: @@ -479,26 +502,69 @@ failed: for (i = 0; i < UFS_MAX_GROUP_LOADED; i++) kfree (sbi->s_ucpi[i]); } - UFSD(("EXIT (FAILED)\n")) + UFSD("EXIT (FAILED)\n"); return 0; } /* - * Put on-disk structures associated with cylinder groups and - * write them back to disk + * Sync our internal copy of fs_cstotal with disk */ -static void ufs_put_cylinder_structures (struct super_block *sb) +static void ufs_put_cstotal(struct super_block *sb) { - struct ufs_sb_info * sbi = UFS_SB(sb); - struct ufs_sb_private_info * uspi; + unsigned mtype = UFS_SB(sb)->s_mount_opt & UFS_MOUNT_UFSTYPE; + struct ufs_sb_private_info *uspi = UFS_SB(sb)->s_uspi; + struct ufs_super_block_first *usb1; + struct ufs_super_block_second *usb2; + struct ufs_super_block_third *usb3; + + UFSD("ENTER\n"); + usb1 = ubh_get_usb_first(uspi); + usb2 = ubh_get_usb_second(uspi); + usb3 = ubh_get_usb_third(uspi); + + if ((mtype == UFS_MOUNT_UFSTYPE_44BSD && + (usb1->fs_flags & UFS_FLAGS_UPDATED)) || + mtype == UFS_MOUNT_UFSTYPE_UFS2) { + /*we have statistic in different place, then usual*/ + usb2->fs_un.fs_u2.cs_ndir = + cpu_to_fs64(sb, uspi->cs_total.cs_ndir); + usb2->fs_un.fs_u2.cs_nbfree = + cpu_to_fs64(sb, uspi->cs_total.cs_nbfree); + usb3->fs_un1.fs_u2.cs_nifree = + cpu_to_fs64(sb, uspi->cs_total.cs_nifree); + usb3->fs_un1.fs_u2.cs_nffree = + cpu_to_fs64(sb, uspi->cs_total.cs_nffree); + } else { + usb1->fs_cstotal.cs_ndir = + cpu_to_fs32(sb, uspi->cs_total.cs_ndir); + usb1->fs_cstotal.cs_nbfree = + cpu_to_fs32(sb, uspi->cs_total.cs_nbfree); + usb1->fs_cstotal.cs_nifree = + cpu_to_fs32(sb, uspi->cs_total.cs_nifree); + usb1->fs_cstotal.cs_nffree = + cpu_to_fs32(sb, uspi->cs_total.cs_nffree); + } + ubh_mark_buffer_dirty(USPI_UBH(uspi)); + UFSD("EXIT\n"); +} + +/** + * ufs_put_super_internal() - put on-disk intrenal structures + * @sb: pointer to super_block structure + * Put on-disk structures associated with cylinder groups + * and write them back to disk, also update cs_total on disk + */ +static void ufs_put_super_internal(struct super_block *sb) +{ + struct ufs_sb_info *sbi = UFS_SB(sb); + struct ufs_sb_private_info *uspi = sbi->s_uspi; struct ufs_buffer_head * ubh; unsigned char * base, * space; unsigned blks, size, i; - - UFSD(("ENTER\n")) - - uspi = sbi->s_uspi; + + UFSD("ENTER\n"); + ufs_put_cstotal(sb); size = uspi->s_cssize; blks = (size + uspi->s_fsize - 1) >> uspi->s_fshift; base = space = (char*) sbi->s_csp; @@ -523,7 +589,7 @@ static void ufs_put_cylinder_structures (struct super_block *sb) brelse (sbi->s_ucg[i]); kfree (sbi->s_ucg); kfree (base); - UFSD(("EXIT\n")) + UFSD("EXIT\n"); } static int ufs_fill_super(struct super_block *sb, void *data, int silent) @@ -533,7 +599,6 @@ static int ufs_fill_super(struct super_block *sb, void *data, int silent) struct ufs_super_block_first * usb1; struct ufs_super_block_second * usb2; struct ufs_super_block_third * usb3; - struct ufs_super_block *usb; struct ufs_buffer_head * ubh; struct inode *inode; unsigned block_size, super_block_size; @@ -544,15 +609,14 @@ static int ufs_fill_super(struct super_block *sb, void *data, int silent) ubh = NULL; flags = 0; - UFSD(("ENTER\n")) + UFSD("ENTER\n"); - sbi = kmalloc(sizeof(struct ufs_sb_info), GFP_KERNEL); + sbi = kzalloc(sizeof(struct ufs_sb_info), GFP_KERNEL); if (!sbi) goto failed_nomem; sb->s_fs_info = sbi; - memset(sbi, 0, sizeof(struct ufs_sb_info)); - UFSD(("flag %u\n", (int)(sb->s_flags & MS_RDONLY))) + UFSD("flag %u\n", (int)(sb->s_flags & MS_RDONLY)); #ifndef CONFIG_UFS_FS_WRITE if (!(sb->s_flags & MS_RDONLY)) { @@ -593,7 +657,7 @@ static int ufs_fill_super(struct super_block *sb, void *data, int silent) the rules */ switch (sbi->s_mount_opt & UFS_MOUNT_UFSTYPE) { case UFS_MOUNT_UFSTYPE_44BSD: - UFSD(("ufstype=44bsd\n")) + UFSD("ufstype=44bsd\n"); uspi->s_fsize = block_size = 512; uspi->s_fmask = ~(512 - 1); uspi->s_fshift = 9; @@ -602,7 +666,7 @@ static int ufs_fill_super(struct super_block *sb, void *data, int silent) flags |= UFS_DE_44BSD | UFS_UID_44BSD | UFS_ST_44BSD | UFS_CG_44BSD; break; case UFS_MOUNT_UFSTYPE_UFS2: - UFSD(("ufstype=ufs2\n")); + UFSD("ufstype=ufs2\n"); super_block_offset=SBLOCK_UFS2; uspi->s_fsize = block_size = 512; uspi->s_fmask = ~(512 - 1); @@ -617,7 +681,7 @@ static int ufs_fill_super(struct super_block *sb, void *data, int silent) break; case UFS_MOUNT_UFSTYPE_SUN: - UFSD(("ufstype=sun\n")) + UFSD("ufstype=sun\n"); uspi->s_fsize = block_size = 1024; uspi->s_fmask = ~(1024 - 1); uspi->s_fshift = 10; @@ -628,7 +692,7 @@ static int ufs_fill_super(struct super_block *sb, void *data, int silent) break; case UFS_MOUNT_UFSTYPE_SUNx86: - UFSD(("ufstype=sunx86\n")) + UFSD("ufstype=sunx86\n"); uspi->s_fsize = block_size = 1024; uspi->s_fmask = ~(1024 - 1); uspi->s_fshift = 10; @@ -639,7 +703,7 @@ static int ufs_fill_super(struct super_block *sb, void *data, int silent) break; case UFS_MOUNT_UFSTYPE_OLD: - UFSD(("ufstype=old\n")) + UFSD("ufstype=old\n"); uspi->s_fsize = block_size = 1024; uspi->s_fmask = ~(1024 - 1); uspi->s_fshift = 10; @@ -654,7 +718,7 @@ static int ufs_fill_super(struct super_block *sb, void *data, int silent) break; case UFS_MOUNT_UFSTYPE_NEXTSTEP: - UFSD(("ufstype=nextstep\n")) + UFSD("ufstype=nextstep\n"); uspi->s_fsize = block_size = 1024; uspi->s_fmask = ~(1024 - 1); uspi->s_fshift = 10; @@ -669,7 +733,7 @@ static int ufs_fill_super(struct super_block *sb, void *data, int silent) break; case UFS_MOUNT_UFSTYPE_NEXTSTEP_CD: - UFSD(("ufstype=nextstep-cd\n")) + UFSD("ufstype=nextstep-cd\n"); uspi->s_fsize = block_size = 2048; uspi->s_fmask = ~(2048 - 1); uspi->s_fshift = 11; @@ -684,7 +748,7 @@ static int ufs_fill_super(struct super_block *sb, void *data, int silent) break; case UFS_MOUNT_UFSTYPE_OPENSTEP: - UFSD(("ufstype=openstep\n")) + UFSD("ufstype=openstep\n"); uspi->s_fsize = block_size = 1024; uspi->s_fmask = ~(1024 - 1); uspi->s_fshift = 10; @@ -699,7 +763,7 @@ static int ufs_fill_super(struct super_block *sb, void *data, int silent) break; case UFS_MOUNT_UFSTYPE_HP: - UFSD(("ufstype=hp\n")) + UFSD("ufstype=hp\n"); uspi->s_fsize = block_size = 1024; uspi->s_fmask = ~(1024 - 1); uspi->s_fshift = 10; @@ -737,8 +801,6 @@ again: usb1 = ubh_get_usb_first(uspi); usb2 = ubh_get_usb_second(uspi); usb3 = ubh_get_usb_third(uspi); - usb = (struct ufs_super_block *) - ((struct ufs_buffer_head *)uspi)->bh[0]->b_data ; /* * Check ufs magic number @@ -820,16 +882,12 @@ magic_found: ubh = NULL; block_size = uspi->s_fsize; super_block_size = uspi->s_sbsize; - UFSD(("another value of block_size or super_block_size %u, %u\n", block_size, super_block_size)) + UFSD("another value of block_size or super_block_size %u, %u\n", block_size, super_block_size); goto again; } -#ifdef UFS_SUPER_DEBUG_MORE - if ((flags & UFS_TYPE_MASK) == UFS_TYPE_UFS2) - ufs2_print_super_stuff(sb,usb); - else - ufs_print_super_stuff(sb, usb1, usb2, usb3); -#endif + + ufs_print_super_stuff(sb, flags, usb1, usb2, usb3); /* * Check, if file system was correctly unmounted. @@ -842,13 +900,13 @@ magic_found: (ufs_get_fs_state(sb, usb1, usb3) == (UFS_FSOK - fs32_to_cpu(sb, usb1->fs_time))))) { switch(usb1->fs_clean) { case UFS_FSCLEAN: - UFSD(("fs is clean\n")) + UFSD("fs is clean\n"); break; case UFS_FSSTABLE: - UFSD(("fs is stable\n")) + UFSD("fs is stable\n"); break; case UFS_FSOSF1: - UFSD(("fs is DEC OSF/1\n")) + UFSD("fs is DEC OSF/1\n"); break; case UFS_FSACTIVE: printk("ufs_read_super: fs is active\n"); @@ -863,8 +921,7 @@ magic_found: sb->s_flags |= MS_RDONLY; break; } - } - else { + } else { printk("ufs_read_super: fs needs fsck\n"); sb->s_flags |= MS_RDONLY; } @@ -884,10 +941,9 @@ magic_found: uspi->s_cgmask = fs32_to_cpu(sb, usb1->fs_cgmask); if ((flags & UFS_TYPE_MASK) == UFS_TYPE_UFS2) { - uspi->s_u2_size = fs64_to_cpu(sb, usb->fs_u11.fs_u2.fs_size); - uspi->s_u2_dsize = fs64_to_cpu(sb, usb->fs_u11.fs_u2.fs_dsize); - } - else { + uspi->s_u2_size = fs64_to_cpu(sb, usb3->fs_un1.fs_u2.fs_size); + uspi->s_u2_dsize = fs64_to_cpu(sb, usb3->fs_un1.fs_u2.fs_dsize); + } else { uspi->s_size = fs32_to_cpu(sb, usb1->fs_size); uspi->s_dsize = fs32_to_cpu(sb, usb1->fs_dsize); } @@ -901,8 +957,8 @@ magic_found: uspi->s_fmask = fs32_to_cpu(sb, usb1->fs_fmask); uspi->s_bshift = fs32_to_cpu(sb, usb1->fs_bshift); uspi->s_fshift = fs32_to_cpu(sb, usb1->fs_fshift); - UFSD(("uspi->s_bshift = %d,uspi->s_fshift = %d", uspi->s_bshift, - uspi->s_fshift)); + UFSD("uspi->s_bshift = %d,uspi->s_fshift = %d", uspi->s_bshift, + uspi->s_fshift); uspi->s_fpbshift = fs32_to_cpu(sb, usb1->fs_fragshift); uspi->s_fsbtodb = fs32_to_cpu(sb, usb1->fs_fsbtodb); /* s_sbsize already set */ @@ -922,8 +978,8 @@ magic_found: uspi->s_spc = fs32_to_cpu(sb, usb1->fs_spc); uspi->s_ipg = fs32_to_cpu(sb, usb1->fs_ipg); uspi->s_fpg = fs32_to_cpu(sb, usb1->fs_fpg); - uspi->s_cpc = fs32_to_cpu(sb, usb2->fs_cpc); - uspi->s_contigsumsize = fs32_to_cpu(sb, usb3->fs_u2.fs_44.fs_contigsumsize); + uspi->s_cpc = fs32_to_cpu(sb, usb2->fs_un.fs_u1.fs_cpc); + uspi->s_contigsumsize = fs32_to_cpu(sb, usb3->fs_un2.fs_44.fs_contigsumsize); uspi->s_qbmask = ufs_get_fs_qbmask(sb, usb3); uspi->s_qfmask = ufs_get_fs_qfmask(sb, usb3); uspi->s_postblformat = fs32_to_cpu(sb, usb3->fs_postblformat); @@ -935,12 +991,11 @@ magic_found: * Compute another frequently used values */ uspi->s_fpbmask = uspi->s_fpb - 1; - if ((flags & UFS_TYPE_MASK) == UFS_TYPE_UFS2) { + if ((flags & UFS_TYPE_MASK) == UFS_TYPE_UFS2) uspi->s_apbshift = uspi->s_bshift - 3; - } - else { + else uspi->s_apbshift = uspi->s_bshift - 2; - } + uspi->s_2apbshift = uspi->s_apbshift * 2; uspi->s_3apbshift = uspi->s_apbshift * 3; uspi->s_apb = 1 << uspi->s_apbshift; @@ -956,7 +1011,7 @@ magic_found: if ((sbi->s_mount_opt & UFS_MOUNT_UFSTYPE) == UFS_MOUNT_UFSTYPE_44BSD) uspi->s_maxsymlinklen = - fs32_to_cpu(sb, usb3->fs_u2.fs_44.fs_maxsymlinklen); + fs32_to_cpu(sb, usb3->fs_un2.fs_44.fs_maxsymlinklen); sbi->s_flags = flags; @@ -967,7 +1022,7 @@ magic_found: if (!sb->s_root) goto dalloc_failed; - + ufs_setup_cstotal(sb); /* * Read cylinder group structures */ @@ -975,7 +1030,7 @@ magic_found: if (!ufs_read_cylinder_structures(sb)) goto failed; - UFSD(("EXIT\n")) + UFSD("EXIT\n"); return 0; dalloc_failed: @@ -986,15 +1041,16 @@ failed: kfree (uspi); kfree(sbi); sb->s_fs_info = NULL; - UFSD(("EXIT (FAILED)\n")) + UFSD("EXIT (FAILED)\n"); return -EINVAL; failed_nomem: - UFSD(("EXIT (NOMEM)\n")) + UFSD("EXIT (NOMEM)\n"); return -ENOMEM; } -static void ufs_write_super (struct super_block *sb) { +static void ufs_write_super(struct super_block *sb) +{ struct ufs_sb_private_info * uspi; struct ufs_super_block_first * usb1; struct ufs_super_block_third * usb3; @@ -1002,7 +1058,7 @@ static void ufs_write_super (struct super_block *sb) { lock_kernel(); - UFSD(("ENTER\n")) + UFSD("ENTER\n"); flags = UFS_SB(sb)->s_flags; uspi = UFS_SB(sb)->s_uspi; usb1 = ubh_get_usb_first(uspi); @@ -1014,26 +1070,27 @@ static void ufs_write_super (struct super_block *sb) { || (flags & UFS_ST_MASK) == UFS_ST_SUNx86) ufs_set_fs_state(sb, usb1, usb3, UFS_FSOK - fs32_to_cpu(sb, usb1->fs_time)); - ubh_mark_buffer_dirty (USPI_UBH); + ufs_put_cstotal(sb); } sb->s_dirt = 0; - UFSD(("EXIT\n")) + UFSD("EXIT\n"); unlock_kernel(); } -static void ufs_put_super (struct super_block *sb) +static void ufs_put_super(struct super_block *sb) { struct ufs_sb_info * sbi = UFS_SB(sb); - UFSD(("ENTER\n")) + UFSD("ENTER\n"); if (!(sb->s_flags & MS_RDONLY)) - ufs_put_cylinder_structures (sb); + ufs_put_super_internal(sb); ubh_brelse_uspi (sbi->s_uspi); kfree (sbi->s_uspi); kfree (sbi); sb->s_fs_info = NULL; + UFSD("EXIT\n"); return; } @@ -1062,8 +1119,7 @@ static int ufs_remount (struct super_block *sb, int *mount_flags, char *data) return -EINVAL; if (!(new_mount_opt & UFS_MOUNT_UFSTYPE)) { new_mount_opt |= ufstype; - } - else if ((new_mount_opt & UFS_MOUNT_UFSTYPE) != ufstype) { + } else if ((new_mount_opt & UFS_MOUNT_UFSTYPE) != ufstype) { printk("ufstype can't be changed during remount\n"); return -EINVAL; } @@ -1077,20 +1133,19 @@ static int ufs_remount (struct super_block *sb, int *mount_flags, char *data) * fs was mouted as rw, remounting ro */ if (*mount_flags & MS_RDONLY) { - ufs_put_cylinder_structures(sb); + ufs_put_super_internal(sb); usb1->fs_time = cpu_to_fs32(sb, get_seconds()); if ((flags & UFS_ST_MASK) == UFS_ST_SUN || (flags & UFS_ST_MASK) == UFS_ST_SUNx86) ufs_set_fs_state(sb, usb1, usb3, UFS_FSOK - fs32_to_cpu(sb, usb1->fs_time)); - ubh_mark_buffer_dirty (USPI_UBH); + ubh_mark_buffer_dirty (USPI_UBH(uspi)); sb->s_dirt = 0; sb->s_flags |= MS_RDONLY; - } + } else { /* * fs was mounted as ro, remounting rw */ - else { #ifndef CONFIG_UFS_FS_WRITE printk("ufs was compiled with read-only support, " "can't be mounted as read-write\n"); @@ -1102,7 +1157,7 @@ static int ufs_remount (struct super_block *sb, int *mount_flags, char *data) printk("this ufstype is read-only supported\n"); return -EINVAL; } - if (!ufs_read_cylinder_structures (sb)) { + if (!ufs_read_cylinder_structures(sb)) { printk("failed during remounting\n"); return -EPERM; } @@ -1113,37 +1168,31 @@ static int ufs_remount (struct super_block *sb, int *mount_flags, char *data) return 0; } -static int ufs_statfs (struct dentry *dentry, struct kstatfs *buf) +static int ufs_statfs(struct dentry *dentry, struct kstatfs *buf) { struct super_block *sb = dentry->d_sb; - struct ufs_sb_private_info * uspi; - struct ufs_super_block_first * usb1; - struct ufs_super_block * usb; - unsigned flags = 0; + struct ufs_sb_private_info *uspi= UFS_SB(sb)->s_uspi; + unsigned flags = UFS_SB(sb)->s_flags; + struct ufs_super_block_first *usb1; + struct ufs_super_block_second *usb2; + struct ufs_super_block_third *usb3; lock_kernel(); - uspi = UFS_SB(sb)->s_uspi; - usb1 = ubh_get_usb_first (uspi); - usb = (struct ufs_super_block *) - ((struct ufs_buffer_head *)uspi)->bh[0]->b_data ; + usb1 = ubh_get_usb_first(uspi); + usb2 = ubh_get_usb_second(uspi); + usb3 = ubh_get_usb_third(uspi); - flags = UFS_SB(sb)->s_flags; if ((flags & UFS_TYPE_MASK) == UFS_TYPE_UFS2) { buf->f_type = UFS2_MAGIC; - buf->f_blocks = fs64_to_cpu(sb, usb->fs_u11.fs_u2.fs_dsize); - buf->f_bfree = ufs_blkstofrags(fs64_to_cpu(sb, usb->fs_u11.fs_u2.fs_cstotal.cs_nbfree)) + - fs64_to_cpu(sb, usb->fs_u11.fs_u2.fs_cstotal.cs_nffree); - buf->f_ffree = fs64_to_cpu(sb, - usb->fs_u11.fs_u2.fs_cstotal.cs_nifree); - } - else { + buf->f_blocks = fs64_to_cpu(sb, usb3->fs_un1.fs_u2.fs_dsize); + } else { buf->f_type = UFS_MAGIC; buf->f_blocks = uspi->s_dsize; - buf->f_bfree = ufs_blkstofrags(fs32_to_cpu(sb, usb1->fs_cstotal.cs_nbfree)) + - fs32_to_cpu(sb, usb1->fs_cstotal.cs_nffree); - buf->f_ffree = fs32_to_cpu(sb, usb1->fs_cstotal.cs_nifree); } + buf->f_bfree = ufs_blkstofrags(uspi->cs_total.cs_nbfree) + + uspi->cs_total.cs_nffree; + buf->f_ffree = uspi->cs_total.cs_nifree; buf->f_bsize = sb->s_blocksize; buf->f_bavail = (buf->f_bfree > (((long)buf->f_blocks / 100) * uspi->s_minfree)) ? (buf->f_bfree - (((long)buf->f_blocks / 100) * uspi->s_minfree)) : 0; @@ -1195,8 +1244,7 @@ static int init_inodecache(void) static void destroy_inodecache(void) { - if (kmem_cache_destroy(ufs_inode_cachep)) - printk(KERN_INFO "ufs_inode_cache: not all structures were freed\n"); + kmem_cache_destroy(ufs_inode_cachep); } #ifdef CONFIG_QUOTA @@ -1276,7 +1324,7 @@ static ssize_t ufs_quota_write(struct super_block *sb, int type, size_t towrite = len; struct buffer_head *bh; - mutex_lock(&inode->i_mutex); + mutex_lock_nested(&inode->i_mutex, I_MUTEX_QUOTA); while (towrite > 0) { tocopy = sb->s_blocksize - offset < towrite ? sb->s_blocksize - offset : towrite; diff --git a/fs/ufs/truncate.c b/fs/ufs/truncate.c index 02e86291ef8..ea11d04c41a 100644 --- a/fs/ufs/truncate.c +++ b/fs/ufs/truncate.c @@ -49,14 +49,6 @@ #include "swab.h" #include "util.h" -#undef UFS_TRUNCATE_DEBUG - -#ifdef UFS_TRUNCATE_DEBUG -#define UFSD(x) printk("(%s, %d), %s: ", __FILE__, __LINE__, __FUNCTION__); printk x; -#else -#define UFSD(x) -#endif - /* * Secure deletion currently doesn't work. It interacts very badly * with buffers shared with memory mappings, and for that reason @@ -82,7 +74,7 @@ static int ufs_trunc_direct (struct inode * inode) unsigned i, tmp; int retry; - UFSD(("ENTER\n")) + UFSD("ENTER\n"); sb = inode->i_sb; uspi = UFS_SB(sb)->s_uspi; @@ -105,7 +97,7 @@ static int ufs_trunc_direct (struct inode * inode) block2 = ufs_fragstoblks (frag3); } - UFSD(("frag1 %u, frag2 %u, block1 %u, block2 %u, frag3 %u, frag4 %u\n", frag1, frag2, block1, block2, frag3, frag4)) + UFSD("frag1 %u, frag2 %u, block1 %u, block2 %u, frag3 %u, frag4 %u\n", frag1, frag2, block1, block2, frag3, frag4); if (frag1 >= frag2) goto next1; @@ -120,9 +112,8 @@ static int ufs_trunc_direct (struct inode * inode) frag1 = ufs_fragnum (frag1); frag2 = ufs_fragnum (frag2); - inode->i_blocks -= (frag2-frag1) << uspi->s_nspfshift; - mark_inode_dirty(inode); ufs_free_fragments (inode, tmp + frag1, frag2 - frag1); + mark_inode_dirty(inode); frag_to_free = tmp + frag1; next1: @@ -136,8 +127,7 @@ next1: continue; *p = 0; - inode->i_blocks -= uspi->s_nspb; - mark_inode_dirty(inode); + if (free_count == 0) { frag_to_free = tmp; free_count = uspi->s_fpb; @@ -148,6 +138,7 @@ next1: frag_to_free = tmp; free_count = uspi->s_fpb; } + mark_inode_dirty(inode); } if (free_count > 0) @@ -166,12 +157,12 @@ next1: frag4 = ufs_fragnum (frag4); *p = 0; - inode->i_blocks -= frag4 << uspi->s_nspfshift; - mark_inode_dirty(inode); + ufs_free_fragments (inode, tmp, frag4); + mark_inode_dirty(inode); next3: - UFSD(("EXIT\n")) + UFSD("EXIT\n"); return retry; } @@ -186,7 +177,7 @@ static int ufs_trunc_indirect (struct inode * inode, unsigned offset, __fs32 *p) unsigned frag_to_free, free_count; int retry; - UFSD(("ENTER\n")) + UFSD("ENTER\n"); sb = inode->i_sb; uspi = UFS_SB(sb)->s_uspi; @@ -227,7 +218,7 @@ static int ufs_trunc_indirect (struct inode * inode, unsigned offset, __fs32 *p) frag_to_free = tmp; free_count = uspi->s_fpb; } - inode->i_blocks -= uspi->s_nspb; + mark_inode_dirty(inode); } @@ -238,26 +229,21 @@ static int ufs_trunc_indirect (struct inode * inode, unsigned offset, __fs32 *p) if (*ubh_get_addr32(ind_ubh,i)) break; if (i >= uspi->s_apb) { - if (ubh_max_bcount(ind_ubh) != 1) { - retry = 1; - } - else { - tmp = fs32_to_cpu(sb, *p); - *p = 0; - inode->i_blocks -= uspi->s_nspb; - mark_inode_dirty(inode); - ufs_free_blocks (inode, tmp, uspi->s_fpb); - ubh_bforget(ind_ubh); - ind_ubh = NULL; - } + tmp = fs32_to_cpu(sb, *p); + *p = 0; + + ufs_free_blocks (inode, tmp, uspi->s_fpb); + mark_inode_dirty(inode); + ubh_bforget(ind_ubh); + ind_ubh = NULL; } if (IS_SYNC(inode) && ind_ubh && ubh_buffer_dirty(ind_ubh)) { - ubh_ll_rw_block (SWRITE, 1, &ind_ubh); + ubh_ll_rw_block(SWRITE, ind_ubh); ubh_wait_on_buffer (ind_ubh); } ubh_brelse (ind_ubh); - UFSD(("EXIT\n")) + UFSD("EXIT\n"); return retry; } @@ -271,7 +257,7 @@ static int ufs_trunc_dindirect (struct inode *inode, unsigned offset, __fs32 *p) __fs32 * dind; int retry = 0; - UFSD(("ENTER\n")) + UFSD("ENTER\n"); sb = inode->i_sb; uspi = UFS_SB(sb)->s_uspi; @@ -306,25 +292,21 @@ static int ufs_trunc_dindirect (struct inode *inode, unsigned offset, __fs32 *p) if (*ubh_get_addr32 (dind_bh, i)) break; if (i >= uspi->s_apb) { - if (ubh_max_bcount(dind_bh) != 1) - retry = 1; - else { - tmp = fs32_to_cpu(sb, *p); - *p = 0; - inode->i_blocks -= uspi->s_nspb; - mark_inode_dirty(inode); - ufs_free_blocks (inode, tmp, uspi->s_fpb); - ubh_bforget(dind_bh); - dind_bh = NULL; - } + tmp = fs32_to_cpu(sb, *p); + *p = 0; + + ufs_free_blocks(inode, tmp, uspi->s_fpb); + mark_inode_dirty(inode); + ubh_bforget(dind_bh); + dind_bh = NULL; } if (IS_SYNC(inode) && dind_bh && ubh_buffer_dirty(dind_bh)) { - ubh_ll_rw_block (SWRITE, 1, &dind_bh); + ubh_ll_rw_block(SWRITE, dind_bh); ubh_wait_on_buffer (dind_bh); } ubh_brelse (dind_bh); - UFSD(("EXIT\n")) + UFSD("EXIT\n"); return retry; } @@ -339,7 +321,7 @@ static int ufs_trunc_tindirect (struct inode * inode) __fs32 * tind, * p; int retry; - UFSD(("ENTER\n")) + UFSD("ENTER\n"); sb = inode->i_sb; uspi = UFS_SB(sb)->s_uspi; @@ -370,45 +352,100 @@ static int ufs_trunc_tindirect (struct inode * inode) if (*ubh_get_addr32 (tind_bh, i)) break; if (i >= uspi->s_apb) { - if (ubh_max_bcount(tind_bh) != 1) - retry = 1; - else { - tmp = fs32_to_cpu(sb, *p); - *p = 0; - inode->i_blocks -= uspi->s_nspb; - mark_inode_dirty(inode); - ufs_free_blocks (inode, tmp, uspi->s_fpb); - ubh_bforget(tind_bh); - tind_bh = NULL; - } + tmp = fs32_to_cpu(sb, *p); + *p = 0; + + ufs_free_blocks(inode, tmp, uspi->s_fpb); + mark_inode_dirty(inode); + ubh_bforget(tind_bh); + tind_bh = NULL; } if (IS_SYNC(inode) && tind_bh && ubh_buffer_dirty(tind_bh)) { - ubh_ll_rw_block (SWRITE, 1, &tind_bh); + ubh_ll_rw_block(SWRITE, tind_bh); ubh_wait_on_buffer (tind_bh); } ubh_brelse (tind_bh); - UFSD(("EXIT\n")) + UFSD("EXIT\n"); return retry; } - -void ufs_truncate (struct inode * inode) + +static int ufs_alloc_lastblock(struct inode *inode) +{ + int err = 0; + struct address_space *mapping = inode->i_mapping; + struct ufs_sb_private_info *uspi = UFS_SB(inode->i_sb)->s_uspi; + unsigned lastfrag, i, end; + struct page *lastpage; + struct buffer_head *bh; + + lastfrag = (i_size_read(inode) + uspi->s_fsize - 1) >> uspi->s_fshift; + + if (!lastfrag) + goto out; + + lastfrag--; + + lastpage = ufs_get_locked_page(mapping, lastfrag >> + (PAGE_CACHE_SHIFT - inode->i_blkbits)); + if (IS_ERR(lastpage)) { + err = -EIO; + goto out; + } + + end = lastfrag & ((1 << (PAGE_CACHE_SHIFT - inode->i_blkbits)) - 1); + bh = page_buffers(lastpage); + for (i = 0; i < end; ++i) + bh = bh->b_this_page; + + + err = ufs_getfrag_block(inode, lastfrag, bh, 1); + + if (unlikely(err)) + goto out_unlock; + + if (buffer_new(bh)) { + clear_buffer_new(bh); + unmap_underlying_metadata(bh->b_bdev, + bh->b_blocknr); + /* + * we do not zeroize fragment, because of + * if it maped to hole, it already contains zeroes + */ + set_buffer_uptodate(bh); + mark_buffer_dirty(bh); + set_page_dirty(lastpage); + } + +out_unlock: + ufs_put_locked_page(lastpage); +out: + return err; +} + +int ufs_truncate(struct inode *inode, loff_t old_i_size) { struct ufs_inode_info *ufsi = UFS_I(inode); - struct super_block * sb; - struct ufs_sb_private_info * uspi; - int retry; + struct super_block *sb = inode->i_sb; + struct ufs_sb_private_info *uspi = UFS_SB(sb)->s_uspi; + int retry, err = 0; - UFSD(("ENTER\n")) - sb = inode->i_sb; - uspi = UFS_SB(sb)->s_uspi; + UFSD("ENTER\n"); - if (!(S_ISREG(inode->i_mode) || S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode))) - return; + if (!(S_ISREG(inode->i_mode) || S_ISDIR(inode->i_mode) || + S_ISLNK(inode->i_mode))) + return -EINVAL; if (IS_APPEND(inode) || IS_IMMUTABLE(inode)) - return; + return -EPERM; + + err = ufs_alloc_lastblock(inode); + + if (err) { + i_size_write(inode, old_i_size); + goto out; + } - block_truncate_page(inode->i_mapping, inode->i_size, ufs_getfrag_block); + block_truncate_page(inode->i_mapping, inode->i_size, ufs_getfrag_block); lock_kernel(); while (1) { @@ -430,5 +467,41 @@ void ufs_truncate (struct inode * inode) ufsi->i_lastfrag = DIRECT_FRAGMENT; unlock_kernel(); mark_inode_dirty(inode); - UFSD(("EXIT\n")) +out: + UFSD("EXIT: err %d\n", err); + return err; +} + + +/* + * We don't define our `inode->i_op->truncate', and call it here, + * because of: + * - there is no way to know old size + * - there is no way inform user about error, if it happens in `truncate' + */ +static int ufs_setattr(struct dentry *dentry, struct iattr *attr) +{ + struct inode *inode = dentry->d_inode; + unsigned int ia_valid = attr->ia_valid; + int error; + + error = inode_change_ok(inode, attr); + if (error) + return error; + + if (ia_valid & ATTR_SIZE && + attr->ia_size != i_size_read(inode)) { + loff_t old_i_size = inode->i_size; + error = vmtruncate(inode, attr->ia_size); + if (error) + return error; + error = ufs_truncate(inode, old_i_size); + if (error) + return error; + } + return inode_setattr(inode, attr); } + +struct inode_operations ufs_file_inode_operations = { + .setattr = ufs_setattr, +}; diff --git a/fs/ufs/util.c b/fs/ufs/util.c index 59acc8f073a..22f820a9b15 100644 --- a/fs/ufs/util.c +++ b/fs/ufs/util.c @@ -14,15 +14,6 @@ #include "swab.h" #include "util.h" -#undef UFS_UTILS_DEBUG - -#ifdef UFS_UTILS_DEBUG -#define UFSD(x) printk("(%s, %d), %s: ", __FILE__, __LINE__, __FUNCTION__); printk x; -#else -#define UFSD(x) -#endif - - struct ufs_buffer_head * _ubh_bread_ (struct ufs_sb_private_info * uspi, struct super_block *sb, u64 fragment, u64 size) { @@ -63,17 +54,17 @@ struct ufs_buffer_head * ubh_bread_uspi (struct ufs_sb_private_info * uspi, count = size >> uspi->s_fshift; if (count <= 0 || count > UFS_MAXFRAG) return NULL; - USPI_UBH->fragment = fragment; - USPI_UBH->count = count; + USPI_UBH(uspi)->fragment = fragment; + USPI_UBH(uspi)->count = count; for (i = 0; i < count; i++) - if (!(USPI_UBH->bh[i] = sb_bread(sb, fragment + i))) + if (!(USPI_UBH(uspi)->bh[i] = sb_bread(sb, fragment + i))) goto failed; for (; i < UFS_MAXFRAG; i++) - USPI_UBH->bh[i] = NULL; - return USPI_UBH; + USPI_UBH(uspi)->bh[i] = NULL; + return USPI_UBH(uspi); failed: for (j = 0; j < i; j++) - brelse (USPI_UBH->bh[j]); + brelse (USPI_UBH(uspi)->bh[j]); return NULL; } @@ -90,11 +81,11 @@ void ubh_brelse (struct ufs_buffer_head * ubh) void ubh_brelse_uspi (struct ufs_sb_private_info * uspi) { unsigned i; - if (!USPI_UBH) + if (!USPI_UBH(uspi)) return; - for ( i = 0; i < USPI_UBH->count; i++ ) { - brelse (USPI_UBH->bh[i]); - USPI_UBH->bh[i] = NULL; + for ( i = 0; i < USPI_UBH(uspi)->count; i++ ) { + brelse (USPI_UBH(uspi)->bh[i]); + USPI_UBH(uspi)->bh[i] = NULL; } } @@ -121,13 +112,12 @@ void ubh_mark_buffer_uptodate (struct ufs_buffer_head * ubh, int flag) } } -void ubh_ll_rw_block (int rw, unsigned nr, struct ufs_buffer_head * ubh[]) +void ubh_ll_rw_block(int rw, struct ufs_buffer_head *ubh) { - unsigned i; if (!ubh) return; - for ( i = 0; i < nr; i++ ) - ll_rw_block (rw, ubh[i]->count, ubh[i]->bh); + + ll_rw_block(rw, ubh->count, ubh->bh); } void ubh_wait_on_buffer (struct ufs_buffer_head * ubh) @@ -139,18 +129,6 @@ void ubh_wait_on_buffer (struct ufs_buffer_head * ubh) wait_on_buffer (ubh->bh[i]); } -unsigned ubh_max_bcount (struct ufs_buffer_head * ubh) -{ - unsigned i; - unsigned max = 0; - if (!ubh) - return 0; - for ( i = 0; i < ubh->count; i++ ) - if ( atomic_read(&ubh->bh[i]->b_count) > max ) - max = atomic_read(&ubh->bh[i]->b_count); - return max; -} - void ubh_bforget (struct ufs_buffer_head * ubh) { unsigned i; @@ -255,3 +233,58 @@ ufs_set_inode_dev(struct super_block *sb, struct ufs_inode_info *ufsi, dev_t dev else ufsi->i_u1.i_data[0] = fs32; } + +/** + * ufs_get_locked_page() - locate, pin and lock a pagecache page, if not exist + * read it from disk. + * @mapping: the address_space to search + * @index: the page index + * + * Locates the desired pagecache page, if not exist we'll read it, + * locks it, increments its reference + * count and returns its address. + * + */ + +struct page *ufs_get_locked_page(struct address_space *mapping, + pgoff_t index) +{ + struct page *page; + + page = find_lock_page(mapping, index); + if (!page) { + page = read_cache_page(mapping, index, + (filler_t*)mapping->a_ops->readpage, + NULL); + + if (IS_ERR(page)) { + printk(KERN_ERR "ufs_change_blocknr: " + "read_cache_page error: ino %lu, index: %lu\n", + mapping->host->i_ino, index); + goto out; + } + + lock_page(page); + + if (unlikely(page->mapping == NULL)) { + /* Truncate got there first */ + unlock_page(page); + page_cache_release(page); + page = NULL; + goto out; + } + + if (!PageUptodate(page) || PageError(page)) { + unlock_page(page); + page_cache_release(page); + + printk(KERN_ERR "ufs_change_blocknr: " + "can not read page: ino %lu, index: %lu\n", + mapping->host->i_ino, index); + + page = ERR_PTR(-EIO); + } + } +out: + return page; +} diff --git a/fs/ufs/util.h b/fs/ufs/util.h index 48d6d9bcc15..28fce6c239b 100644 --- a/fs/ufs/util.h +++ b/fs/ufs/util.h @@ -17,10 +17,16 @@ #define in_range(b,first,len) ((b)>=(first)&&(b)<(first)+(len)) /* - * macros used for retyping + * functions used for retyping */ -#define UCPI_UBH ((struct ufs_buffer_head *)ucpi) -#define USPI_UBH ((struct ufs_buffer_head *)uspi) +static inline struct ufs_buffer_head *UCPI_UBH(struct ufs_cg_private_info *cpi) +{ + return &cpi->c_ubh; +} +static inline struct ufs_buffer_head *USPI_UBH(struct ufs_sb_private_info *spi) +{ + return &spi->s_ubh; +} @@ -33,12 +39,12 @@ ufs_get_fs_state(struct super_block *sb, struct ufs_super_block_first *usb1, { switch (UFS_SB(sb)->s_flags & UFS_ST_MASK) { case UFS_ST_SUN: - return fs32_to_cpu(sb, usb3->fs_u2.fs_sun.fs_state); + return fs32_to_cpu(sb, usb3->fs_un2.fs_sun.fs_state); case UFS_ST_SUNx86: return fs32_to_cpu(sb, usb1->fs_u1.fs_sunx86.fs_state); case UFS_ST_44BSD: default: - return fs32_to_cpu(sb, usb3->fs_u2.fs_44.fs_state); + return fs32_to_cpu(sb, usb3->fs_un2.fs_44.fs_state); } } @@ -48,13 +54,13 @@ ufs_set_fs_state(struct super_block *sb, struct ufs_super_block_first *usb1, { switch (UFS_SB(sb)->s_flags & UFS_ST_MASK) { case UFS_ST_SUN: - usb3->fs_u2.fs_sun.fs_state = cpu_to_fs32(sb, value); + usb3->fs_un2.fs_sun.fs_state = cpu_to_fs32(sb, value); break; case UFS_ST_SUNx86: usb1->fs_u1.fs_sunx86.fs_state = cpu_to_fs32(sb, value); break; case UFS_ST_44BSD: - usb3->fs_u2.fs_44.fs_state = cpu_to_fs32(sb, value); + usb3->fs_un2.fs_44.fs_state = cpu_to_fs32(sb, value); break; } } @@ -64,7 +70,7 @@ ufs_get_fs_npsect(struct super_block *sb, struct ufs_super_block_first *usb1, struct ufs_super_block_third *usb3) { if ((UFS_SB(sb)->s_flags & UFS_ST_MASK) == UFS_ST_SUNx86) - return fs32_to_cpu(sb, usb3->fs_u2.fs_sunx86.fs_npsect); + return fs32_to_cpu(sb, usb3->fs_un2.fs_sunx86.fs_npsect); else return fs32_to_cpu(sb, usb1->fs_u1.fs_sun.fs_npsect); } @@ -76,16 +82,16 @@ ufs_get_fs_qbmask(struct super_block *sb, struct ufs_super_block_third *usb3) switch (UFS_SB(sb)->s_flags & UFS_ST_MASK) { case UFS_ST_SUN: - ((__fs32 *)&tmp)[0] = usb3->fs_u2.fs_sun.fs_qbmask[0]; - ((__fs32 *)&tmp)[1] = usb3->fs_u2.fs_sun.fs_qbmask[1]; + ((__fs32 *)&tmp)[0] = usb3->fs_un2.fs_sun.fs_qbmask[0]; + ((__fs32 *)&tmp)[1] = usb3->fs_un2.fs_sun.fs_qbmask[1]; break; case UFS_ST_SUNx86: - ((__fs32 *)&tmp)[0] = usb3->fs_u2.fs_sunx86.fs_qbmask[0]; - ((__fs32 *)&tmp)[1] = usb3->fs_u2.fs_sunx86.fs_qbmask[1]; + ((__fs32 *)&tmp)[0] = usb3->fs_un2.fs_sunx86.fs_qbmask[0]; + ((__fs32 *)&tmp)[1] = usb3->fs_un2.fs_sunx86.fs_qbmask[1]; break; case UFS_ST_44BSD: - ((__fs32 *)&tmp)[0] = usb3->fs_u2.fs_44.fs_qbmask[0]; - ((__fs32 *)&tmp)[1] = usb3->fs_u2.fs_44.fs_qbmask[1]; + ((__fs32 *)&tmp)[0] = usb3->fs_un2.fs_44.fs_qbmask[0]; + ((__fs32 *)&tmp)[1] = usb3->fs_un2.fs_44.fs_qbmask[1]; break; } @@ -99,16 +105,16 @@ ufs_get_fs_qfmask(struct super_block *sb, struct ufs_super_block_third *usb3) switch (UFS_SB(sb)->s_flags & UFS_ST_MASK) { case UFS_ST_SUN: - ((__fs32 *)&tmp)[0] = usb3->fs_u2.fs_sun.fs_qfmask[0]; - ((__fs32 *)&tmp)[1] = usb3->fs_u2.fs_sun.fs_qfmask[1]; + ((__fs32 *)&tmp)[0] = usb3->fs_un2.fs_sun.fs_qfmask[0]; + ((__fs32 *)&tmp)[1] = usb3->fs_un2.fs_sun.fs_qfmask[1]; break; case UFS_ST_SUNx86: - ((__fs32 *)&tmp)[0] = usb3->fs_u2.fs_sunx86.fs_qfmask[0]; - ((__fs32 *)&tmp)[1] = usb3->fs_u2.fs_sunx86.fs_qfmask[1]; + ((__fs32 *)&tmp)[0] = usb3->fs_un2.fs_sunx86.fs_qfmask[0]; + ((__fs32 *)&tmp)[1] = usb3->fs_un2.fs_sunx86.fs_qfmask[1]; break; case UFS_ST_44BSD: - ((__fs32 *)&tmp)[0] = usb3->fs_u2.fs_44.fs_qfmask[0]; - ((__fs32 *)&tmp)[1] = usb3->fs_u2.fs_44.fs_qfmask[1]; + ((__fs32 *)&tmp)[0] = usb3->fs_un2.fs_44.fs_qfmask[0]; + ((__fs32 *)&tmp)[1] = usb3->fs_un2.fs_44.fs_qfmask[1]; break; } @@ -236,9 +242,8 @@ extern void ubh_brelse (struct ufs_buffer_head *); extern void ubh_brelse_uspi (struct ufs_sb_private_info *); extern void ubh_mark_buffer_dirty (struct ufs_buffer_head *); extern void ubh_mark_buffer_uptodate (struct ufs_buffer_head *, int); -extern void ubh_ll_rw_block (int, unsigned, struct ufs_buffer_head **); +extern void ubh_ll_rw_block(int, struct ufs_buffer_head *); extern void ubh_wait_on_buffer (struct ufs_buffer_head *); -extern unsigned ubh_max_bcount (struct ufs_buffer_head *); extern void ubh_bforget (struct ufs_buffer_head *); extern int ubh_buffer_dirty (struct ufs_buffer_head *); #define ubh_ubhcpymem(mem,ubh,size) _ubh_ubhcpymem_(uspi,mem,ubh,size) @@ -246,6 +251,14 @@ extern void _ubh_ubhcpymem_(struct ufs_sb_private_info *, unsigned char *, struc #define ubh_memcpyubh(ubh,mem,size) _ubh_memcpyubh_(uspi,ubh,mem,size) extern void _ubh_memcpyubh_(struct ufs_sb_private_info *, struct ufs_buffer_head *, unsigned char *, unsigned); +/* This functions works with cache pages*/ +extern struct page *ufs_get_locked_page(struct address_space *mapping, + pgoff_t index); +static inline void ufs_put_locked_page(struct page *page) +{ + unlock_page(page); + page_cache_release(page); +} /* @@ -297,40 +310,26 @@ static inline void *get_usb_offset(struct ufs_sb_private_info *uspi, #define ubh_blkmap(ubh,begin,bit) \ ((*ubh_get_addr(ubh, (begin) + ((bit) >> 3)) >> ((bit) & 7)) & (0xff >> (UFS_MAXFRAG - uspi->s_fpb))) - -/* - * Macros for access to superblock array structures - */ -#define ubh_postbl(ubh,cylno,i) \ - ((uspi->s_postblformat != UFS_DYNAMICPOSTBLFMT) \ - ? (*(__s16*)(ubh_get_addr(ubh, \ - (unsigned)(&((struct ufs_super_block *)0)->fs_opostbl) \ - + (((cylno) * 16 + (i)) << 1) ) )) \ - : (*(__s16*)(ubh_get_addr(ubh, \ - uspi->s_postbloff + (((cylno) * uspi->s_nrpos + (i)) << 1) )))) - -#define ubh_rotbl(ubh,i) \ - ((uspi->s_postblformat != UFS_DYNAMICPOSTBLFMT) \ - ? (*(__u8*)(ubh_get_addr(ubh, \ - (unsigned)(&((struct ufs_super_block *)0)->fs_space) + (i)))) \ - : (*(__u8*)(ubh_get_addr(ubh, uspi->s_rotbloff + (i))))) - /* * Determine the number of available frags given a * percentage to hold in reserve. */ -#define ufs_freespace(usb, percentreserved) \ - (ufs_blkstofrags(fs32_to_cpu(sb, (usb)->fs_cstotal.cs_nbfree)) + \ - fs32_to_cpu(sb, (usb)->fs_cstotal.cs_nffree) - (uspi->s_dsize * (percentreserved) / 100)) +static inline u64 +ufs_freespace(struct ufs_sb_private_info *uspi, int percentreserved) +{ + return ufs_blkstofrags(uspi->cs_total.cs_nbfree) + + uspi->cs_total.cs_nffree - + (uspi->s_dsize * (percentreserved) / 100); +} /* * Macros to access cylinder group array structures */ #define ubh_cg_blktot(ucpi,cylno) \ - (*((__fs32*)ubh_get_addr(UCPI_UBH, (ucpi)->c_btotoff + ((cylno) << 2)))) + (*((__fs32*)ubh_get_addr(UCPI_UBH(ucpi), (ucpi)->c_btotoff + ((cylno) << 2)))) #define ubh_cg_blks(ucpi,cylno,rpos) \ - (*((__fs16*)ubh_get_addr(UCPI_UBH, \ + (*((__fs16*)ubh_get_addr(UCPI_UBH(ucpi), \ (ucpi)->c_boff + (((cylno) * uspi->s_nrpos + (rpos)) << 1 )))) /* @@ -508,29 +507,3 @@ static inline void ufs_fragacct (struct super_block * sb, unsigned blockmap, if (fragsize > 0 && fragsize < uspi->s_fpb) fs32_add(sb, &fraglist[fragsize], cnt); } - -#define ubh_scanc(ubh,begin,size,table,mask) _ubh_scanc_(uspi,ubh,begin,size,table,mask) -static inline unsigned _ubh_scanc_(struct ufs_sb_private_info * uspi, struct ufs_buffer_head * ubh, - unsigned begin, unsigned size, unsigned char * table, unsigned char mask) -{ - unsigned rest, offset; - unsigned char * cp; - - - offset = begin & ~uspi->s_fmask; - begin >>= uspi->s_fshift; - for (;;) { - if ((offset + size) < uspi->s_fsize) - rest = size; - else - rest = uspi->s_fsize - offset; - size -= rest; - cp = ubh->bh[begin]->b_data + offset; - while ((table[*cp++] & mask) == 0 && --rest); - if (rest || !size) - break; - begin++; - offset = 0; - } - return (size + rest); -} diff --git a/fs/xfs/Makefile-linux-2.6 b/fs/xfs/Makefile-linux-2.6 index 9e7f85986d0..291948d5085 100644 --- a/fs/xfs/Makefile-linux-2.6 +++ b/fs/xfs/Makefile-linux-2.6 @@ -30,7 +30,6 @@ ifeq ($(CONFIG_XFS_TRACE),y) EXTRA_CFLAGS += -DXFS_BLI_TRACE EXTRA_CFLAGS += -DXFS_BMAP_TRACE EXTRA_CFLAGS += -DXFS_BMBT_TRACE - EXTRA_CFLAGS += -DXFS_DIR_TRACE EXTRA_CFLAGS += -DXFS_DIR2_TRACE EXTRA_CFLAGS += -DXFS_DQUOT_TRACE EXTRA_CFLAGS += -DXFS_ILOCK_TRACE diff --git a/fs/xfs/linux-2.6/kmem.c b/fs/xfs/linux-2.6/kmem.c index aba7fcf881a..d5973758981 100644 --- a/fs/xfs/linux-2.6/kmem.c +++ b/fs/xfs/linux-2.6/kmem.c @@ -34,6 +34,14 @@ kmem_alloc(size_t size, unsigned int __nocast flags) gfp_t lflags = kmem_flags_convert(flags); void *ptr; +#ifdef DEBUG + if (unlikely(!(flags & KM_LARGE) && (size > PAGE_SIZE))) { + printk(KERN_WARNING "Large %s attempt, size=%ld\n", + __FUNCTION__, (long)size); + dump_stack(); + } +#endif + do { if (size < MAX_SLAB_SIZE || retries > MAX_VMALLOCS) ptr = kmalloc(size, lflags); @@ -60,6 +68,27 @@ kmem_zalloc(size_t size, unsigned int __nocast flags) return ptr; } +void * +kmem_zalloc_greedy(size_t *size, size_t minsize, size_t maxsize, + unsigned int __nocast flags) +{ + void *ptr; + size_t kmsize = maxsize; + unsigned int kmflags = (flags & ~KM_SLEEP) | KM_NOSLEEP; + + while (!(ptr = kmem_zalloc(kmsize, kmflags))) { + if ((kmsize <= minsize) && (flags & KM_NOSLEEP)) + break; + if ((kmsize >>= 1) <= minsize) { + kmsize = minsize; + kmflags = flags; + } + } + if (ptr) + *size = kmsize; + return ptr; +} + void kmem_free(void *ptr, size_t size) { diff --git a/fs/xfs/linux-2.6/kmem.h b/fs/xfs/linux-2.6/kmem.h index 939bd84bc7e..9ebabdf7829 100644 --- a/fs/xfs/linux-2.6/kmem.h +++ b/fs/xfs/linux-2.6/kmem.h @@ -30,6 +30,7 @@ #define KM_NOSLEEP 0x0002u #define KM_NOFS 0x0004u #define KM_MAYFAIL 0x0008u +#define KM_LARGE 0x0010u /* * We use a special process flag to avoid recursive callbacks into @@ -41,7 +42,7 @@ kmem_flags_convert(unsigned int __nocast flags) { gfp_t lflags; - BUG_ON(flags & ~(KM_SLEEP|KM_NOSLEEP|KM_NOFS|KM_MAYFAIL)); + BUG_ON(flags & ~(KM_SLEEP|KM_NOSLEEP|KM_NOFS|KM_MAYFAIL|KM_LARGE)); if (flags & KM_NOSLEEP) { lflags = GFP_ATOMIC | __GFP_NOWARN; @@ -54,8 +55,9 @@ kmem_flags_convert(unsigned int __nocast flags) } extern void *kmem_alloc(size_t, unsigned int __nocast); -extern void *kmem_realloc(void *, size_t, size_t, unsigned int __nocast); extern void *kmem_zalloc(size_t, unsigned int __nocast); +extern void *kmem_zalloc_greedy(size_t *, size_t, size_t, unsigned int __nocast); +extern void *kmem_realloc(void *, size_t, size_t, unsigned int __nocast); extern void kmem_free(void *, size_t); /* @@ -91,8 +93,8 @@ kmem_zone_free(kmem_zone_t *zone, void *ptr) static inline void kmem_zone_destroy(kmem_zone_t *zone) { - if (zone && kmem_cache_destroy(zone)) - BUG(); + if (zone) + kmem_cache_destroy(zone); } extern void *kmem_zone_alloc(kmem_zone_t *, unsigned int __nocast); diff --git a/fs/xfs/linux-2.6/sema.h b/fs/xfs/linux-2.6/sema.h index b25090094cc..2009e6d922c 100644 --- a/fs/xfs/linux-2.6/sema.h +++ b/fs/xfs/linux-2.6/sema.h @@ -29,8 +29,6 @@ typedef struct semaphore sema_t; -#define init_sema(sp, val, c, d) sema_init(sp, val) -#define initsema(sp, val) sema_init(sp, val) #define initnsema(sp, val, name) sema_init(sp, val) #define psema(sp, b) down(sp) #define vsema(sp) up(sp) diff --git a/fs/xfs/linux-2.6/sv.h b/fs/xfs/linux-2.6/sv.h index 9a8ad481b00..351a8f454bd 100644 --- a/fs/xfs/linux-2.6/sv.h +++ b/fs/xfs/linux-2.6/sv.h @@ -53,8 +53,6 @@ static inline void _sv_wait(sv_t *sv, spinlock_t *lock, int state, remove_wait_queue(&sv->waiters, &wait); } -#define init_sv(sv,type,name,flag) \ - init_waitqueue_head(&(sv)->waiters) #define sv_init(sv,flag,name) \ init_waitqueue_head(&(sv)->waiters) #define sv_destroy(sv) \ diff --git a/fs/xfs/linux-2.6/xfs_aops.c b/fs/xfs/linux-2.6/xfs_aops.c index 3e807b828e2..09360cf1e1f 100644 --- a/fs/xfs/linux-2.6/xfs_aops.c +++ b/fs/xfs/linux-2.6/xfs_aops.c @@ -71,7 +71,7 @@ xfs_page_trace( int tag, struct inode *inode, struct page *page, - int mask) + unsigned long pgoff) { xfs_inode_t *ip; bhv_vnode_t *vp = vn_from_inode(inode); @@ -91,7 +91,7 @@ xfs_page_trace( (void *)ip, (void *)inode, (void *)page, - (void *)((unsigned long)mask), + (void *)pgoff, (void *)((unsigned long)((ip->i_d.di_size >> 32) & 0xffffffff)), (void *)((unsigned long)(ip->i_d.di_size & 0xffffffff)), (void *)((unsigned long)((isize >> 32) & 0xffffffff)), @@ -105,7 +105,7 @@ xfs_page_trace( (void *)NULL); } #else -#define xfs_page_trace(tag, inode, page, mask) +#define xfs_page_trace(tag, inode, page, pgoff) #endif /* @@ -1197,7 +1197,7 @@ xfs_vm_releasepage( .nr_to_write = 1, }; - xfs_page_trace(XFS_RELEASEPAGE_ENTER, inode, page, gfp_mask); + xfs_page_trace(XFS_RELEASEPAGE_ENTER, inode, page, 0); if (!page_has_buffers(page)) return 0; @@ -1356,7 +1356,6 @@ xfs_end_io_direct( ioend->io_size = size; xfs_finish_ioend(ioend); } else { - ASSERT(size >= 0); xfs_destroy_ioend(ioend); } @@ -1390,11 +1389,19 @@ xfs_vm_direct_IO( iocb->private = xfs_alloc_ioend(inode, IOMAP_UNWRITTEN); - ret = blockdev_direct_IO_own_locking(rw, iocb, inode, - iomap.iomap_target->bt_bdev, - iov, offset, nr_segs, - xfs_get_blocks_direct, - xfs_end_io_direct); + if (rw == WRITE) { + ret = blockdev_direct_IO_own_locking(rw, iocb, inode, + iomap.iomap_target->bt_bdev, + iov, offset, nr_segs, + xfs_get_blocks_direct, + xfs_end_io_direct); + } else { + ret = blockdev_direct_IO_no_locking(rw, iocb, inode, + iomap.iomap_target->bt_bdev, + iov, offset, nr_segs, + xfs_get_blocks_direct, + xfs_end_io_direct); + } if (unlikely(ret <= 0 && iocb->private)) xfs_destroy_ioend(iocb->private); @@ -1454,7 +1461,7 @@ xfs_vm_invalidatepage( block_invalidatepage(page, offset); } -struct address_space_operations xfs_address_space_operations = { +const struct address_space_operations xfs_address_space_operations = { .readpage = xfs_vm_readpage, .readpages = xfs_vm_readpages, .writepage = xfs_vm_writepage, diff --git a/fs/xfs/linux-2.6/xfs_aops.h b/fs/xfs/linux-2.6/xfs_aops.h index 706d8c781b8..2244e516b66 100644 --- a/fs/xfs/linux-2.6/xfs_aops.h +++ b/fs/xfs/linux-2.6/xfs_aops.h @@ -40,7 +40,7 @@ typedef struct xfs_ioend { struct work_struct io_work; /* xfsdatad work queue */ } xfs_ioend_t; -extern struct address_space_operations xfs_address_space_operations; +extern const struct address_space_operations xfs_address_space_operations; extern int xfs_get_blocks(struct inode *, sector_t, struct buffer_head *, 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 26fed0756f0..9bbadafdcb0 100644 --- a/fs/xfs/linux-2.6/xfs_buf.c +++ b/fs/xfs/linux-2.6/xfs_buf.c @@ -1,5 +1,5 @@ /* - * Copyright (c) 2000-2005 Silicon Graphics, Inc. + * Copyright (c) 2000-2006 Silicon Graphics, Inc. * All Rights Reserved. * * This program is free software; you can redistribute it and/or @@ -318,8 +318,12 @@ xfs_buf_free( if ((bp->b_flags & XBF_MAPPED) && (bp->b_page_count > 1)) free_address(bp->b_addr - bp->b_offset); - for (i = 0; i < bp->b_page_count; i++) - page_cache_release(bp->b_pages[i]); + for (i = 0; i < bp->b_page_count; i++) { + struct page *page = bp->b_pages[i]; + + ASSERT(!PagePrivate(page)); + page_cache_release(page); + } _xfs_buf_free_pages(bp); } else if (bp->b_flags & _XBF_KMEM_ALLOC) { /* @@ -400,6 +404,7 @@ _xfs_buf_lookup_pages( nbytes = min_t(size_t, size, PAGE_CACHE_SIZE - offset); size -= nbytes; + ASSERT(!PagePrivate(page)); if (!PageUptodate(page)) { page_count--; if (blocksize >= PAGE_CACHE_SIZE) { @@ -768,7 +773,7 @@ xfs_buf_get_noaddr( _xfs_buf_initialize(bp, target, 0, len, 0); try_again: - data = kmem_alloc(malloc_len, KM_SLEEP | KM_MAYFAIL); + data = kmem_alloc(malloc_len, KM_SLEEP | KM_MAYFAIL | KM_LARGE); if (unlikely(data == NULL)) goto fail_free_buf; @@ -1117,10 +1122,10 @@ xfs_buf_bio_end_io( do { struct page *page = bvec->bv_page; + ASSERT(!PagePrivate(page)); if (unlikely(bp->b_error)) { if (bp->b_flags & XBF_READ) ClearPageUptodate(page); - SetPageError(page); } else if (blocksize >= PAGE_CACHE_SIZE) { SetPageUptodate(page); } else if (!PagePrivate(page) && @@ -1156,16 +1161,16 @@ _xfs_buf_ioapply( total_nr_pages = bp->b_page_count; map_i = 0; - if (bp->b_flags & _XBF_RUN_QUEUES) { - bp->b_flags &= ~_XBF_RUN_QUEUES; - rw = (bp->b_flags & XBF_READ) ? READ_SYNC : WRITE_SYNC; - } else { - rw = (bp->b_flags & XBF_READ) ? READ : WRITE; - } - if (bp->b_flags & XBF_ORDERED) { ASSERT(!(bp->b_flags & XBF_READ)); rw = WRITE_BARRIER; + } else if (bp->b_flags & _XBF_RUN_QUEUES) { + ASSERT(!(bp->b_flags & XBF_READ_AHEAD)); + bp->b_flags &= ~_XBF_RUN_QUEUES; + rw = (bp->b_flags & XBF_WRITE) ? WRITE_SYNC : READ_SYNC; + } else { + rw = (bp->b_flags & XBF_WRITE) ? WRITE : + (bp->b_flags & XBF_READ_AHEAD) ? READA : READ; } /* Special code path for reading a sub page size buffer in -- @@ -1520,7 +1525,7 @@ xfs_mapping_buftarg( struct backing_dev_info *bdi; struct inode *inode; struct address_space *mapping; - static struct address_space_operations mapping_aops = { + static const struct address_space_operations mapping_aops = { .sync_page = block_sync_page, .migratepage = fail_migrate_page, }; @@ -1681,6 +1686,7 @@ xfsbufd( xfs_buf_t *bp, *n; struct list_head *dwq = &target->bt_delwrite_queue; spinlock_t *dwlk = &target->bt_delwrite_lock; + int count; current->flags |= PF_MEMALLOC; @@ -1696,6 +1702,7 @@ xfsbufd( schedule_timeout_interruptible( xfs_buf_timer_centisecs * msecs_to_jiffies(10)); + count = 0; age = xfs_buf_age_centisecs * msecs_to_jiffies(10); spin_lock(dwlk); list_for_each_entry_safe(bp, n, dwq, b_list) { @@ -1711,9 +1718,11 @@ xfsbufd( break; } - bp->b_flags &= ~(XBF_DELWRI|_XBF_DELWRI_Q); + bp->b_flags &= ~(XBF_DELWRI|_XBF_DELWRI_Q| + _XBF_RUN_QUEUES); bp->b_flags |= XBF_WRITE; - list_move(&bp->b_list, &tmp); + list_move_tail(&bp->b_list, &tmp); + count++; } } spin_unlock(dwlk); @@ -1724,12 +1733,12 @@ xfsbufd( list_del_init(&bp->b_list); xfs_buf_iostrategy(bp); - - blk_run_address_space(target->bt_mapping); } if (as_list_len > 0) purge_addresses(); + if (count) + blk_run_address_space(target->bt_mapping); clear_bit(XBT_FORCE_FLUSH, &target->bt_flags); } while (!kthread_should_stop()); @@ -1767,7 +1776,7 @@ xfs_flush_buftarg( continue; } - list_move(&bp->b_list, &tmp); + list_move_tail(&bp->b_list, &tmp); } spin_unlock(dwlk); @@ -1776,7 +1785,7 @@ xfs_flush_buftarg( */ list_for_each_entry_safe(bp, n, &tmp, b_list) { xfs_buf_lock(bp); - bp->b_flags &= ~(XBF_DELWRI|_XBF_DELWRI_Q); + bp->b_flags &= ~(XBF_DELWRI|_XBF_DELWRI_Q|_XBF_RUN_QUEUES); bp->b_flags |= XBF_WRITE; if (wait) bp->b_flags &= ~XBF_ASYNC; @@ -1786,6 +1795,9 @@ xfs_flush_buftarg( xfs_buf_iostrategy(bp); } + if (wait) + blk_run_address_space(target->bt_mapping); + /* * Remaining list items must be flushed before returning */ @@ -1797,9 +1809,6 @@ xfs_flush_buftarg( xfs_buf_relse(bp); } - if (wait) - blk_run_address_space(target->bt_mapping); - return pincount; } diff --git a/fs/xfs/linux-2.6/xfs_buf.h b/fs/xfs/linux-2.6/xfs_buf.h index 4dd6592d5a4..9dd235cb010 100644 --- a/fs/xfs/linux-2.6/xfs_buf.h +++ b/fs/xfs/linux-2.6/xfs_buf.h @@ -18,7 +18,6 @@ #ifndef __XFS_BUF_H__ #define __XFS_BUF_H__ -#include <linux/config.h> #include <linux/list.h> #include <linux/types.h> #include <linux/spinlock.h> @@ -247,8 +246,8 @@ extern void xfs_buf_trace(xfs_buf_t *, char *, void *, void *); #define BUF_BUSY XBF_DONT_BLOCK #define XFS_BUF_BFLAGS(bp) ((bp)->b_flags) -#define XFS_BUF_ZEROFLAGS(bp) \ - ((bp)->b_flags &= ~(XBF_READ|XBF_WRITE|XBF_ASYNC|XBF_DELWRI)) +#define XFS_BUF_ZEROFLAGS(bp) ((bp)->b_flags &= \ + ~(XBF_READ|XBF_WRITE|XBF_ASYNC|XBF_DELWRI|XBF_ORDERED)) #define XFS_BUF_STALE(bp) ((bp)->b_flags |= XFS_B_STALE) #define XFS_BUF_UNSTALE(bp) ((bp)->b_flags &= ~XFS_B_STALE) @@ -299,11 +298,6 @@ extern void xfs_buf_trace(xfs_buf_t *, char *, void *, void *); #define XFS_BUF_UNWRITE(bp) ((bp)->b_flags &= ~XBF_WRITE) #define XFS_BUF_ISWRITE(bp) ((bp)->b_flags & XBF_WRITE) -#define XFS_BUF_ISUNINITIAL(bp) (0) -#define XFS_BUF_UNUNINITIAL(bp) (0) - -#define XFS_BUF_BP_ISMAPPED(bp) (1) - #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) @@ -394,8 +388,6 @@ static inline int XFS_bwrite(xfs_buf_t *bp) return error; } -#define XFS_bdwrite(bp) xfs_buf_iostart(bp, XBF_DELWRI | XBF_ASYNC) - static inline int xfs_bdwrite(void *mp, xfs_buf_t *bp) { bp->b_strat = xfs_bdstrat_cb; diff --git a/fs/xfs/linux-2.6/xfs_file.c b/fs/xfs/linux-2.6/xfs_file.c index 70662371bb1..41cfcba7ce4 100644 --- a/fs/xfs/linux-2.6/xfs_file.c +++ b/fs/xfs/linux-2.6/xfs_file.c @@ -299,7 +299,8 @@ xfs_file_open( STATIC int xfs_file_close( - struct file *filp) + struct file *filp, + fl_owner_t id) { return -bhv_vop_close(vn_from_inode(filp->f_dentry->d_inode), 0, file_count(filp) > 1 ? L_FALSE : L_TRUE, NULL); @@ -369,7 +370,7 @@ xfs_file_readdir( /* Try fairly hard to get memory */ do { - if ((read_buf = (caddr_t)kmalloc(rlen, GFP_KERNEL))) + if ((read_buf = kmalloc(rlen, GFP_KERNEL))) break; rlen >>= 1; } while (rlen >= 1024); diff --git a/fs/xfs/linux-2.6/xfs_globals.c b/fs/xfs/linux-2.6/xfs_globals.c index 6c162c3dde7..ed3a5e1b4b6 100644 --- a/fs/xfs/linux-2.6/xfs_globals.c +++ b/fs/xfs/linux-2.6/xfs_globals.c @@ -34,7 +34,7 @@ xfs_param_t xfs_params = { .restrict_chown = { 0, 1, 1 }, .sgid_inherit = { 0, 0, 1 }, .symlink_mode = { 0, 0, 1 }, - .panic_mask = { 0, 0, 127 }, + .panic_mask = { 0, 0, 255 }, .error_level = { 0, 3, 11 }, .syncd_timer = { 1*100, 30*100, 7200*100}, .stats_clear = { 0, 0, 1 }, diff --git a/fs/xfs/linux-2.6/xfs_ioctl.c b/fs/xfs/linux-2.6/xfs_ioctl.c index 6e52a5dd38d..a74f854d91e 100644 --- a/fs/xfs/linux-2.6/xfs_ioctl.c +++ b/fs/xfs/linux-2.6/xfs_ioctl.c @@ -653,7 +653,7 @@ xfs_attrmulti_by_handle( STATIC int xfs_ioc_space( bhv_desc_t *bdp, - bhv_vnode_t *vp, + struct inode *inode, struct file *filp, int flags, unsigned int cmd, @@ -735,7 +735,7 @@ xfs_ioctl( !capable(CAP_SYS_ADMIN)) return -EPERM; - return xfs_ioc_space(bdp, vp, filp, ioflags, cmd, arg); + return xfs_ioc_space(bdp, inode, filp, ioflags, cmd, arg); case XFS_IOC_DIOINFO: { struct dioattr da; @@ -763,6 +763,8 @@ xfs_ioctl( return xfs_ioc_fsgeometry(mp, arg); case XFS_IOC_GETVERSION: + return put_user(inode->i_generation, (int __user *)arg); + case XFS_IOC_GETXFLAGS: case XFS_IOC_SETXFLAGS: case XFS_IOC_FSGETXATTR: @@ -957,7 +959,7 @@ xfs_ioctl( STATIC int xfs_ioc_space( bhv_desc_t *bdp, - bhv_vnode_t *vp, + struct inode *inode, struct file *filp, int ioflags, unsigned int cmd, @@ -967,13 +969,13 @@ xfs_ioc_space( int attr_flags = 0; int error; - if (vp->v_inode.i_flags & (S_IMMUTABLE|S_APPEND)) + if (inode->i_flags & (S_IMMUTABLE|S_APPEND)) return -XFS_ERROR(EPERM); if (!(filp->f_mode & FMODE_WRITE)) return -XFS_ERROR(EBADF); - if (!VN_ISREG(vp)) + if (!S_ISREG(inode->i_mode)) return -XFS_ERROR(EINVAL); if (copy_from_user(&bf, arg, sizeof(bf))) @@ -1264,13 +1266,6 @@ xfs_ioc_xattr( break; } - case XFS_IOC_GETVERSION: { - flags = vn_to_inode(vp)->i_generation; - if (copy_to_user(arg, &flags, sizeof(flags))) - error = -EFAULT; - break; - } - default: error = -ENOTTY; break; diff --git a/fs/xfs/linux-2.6/xfs_ioctl32.c b/fs/xfs/linux-2.6/xfs_ioctl32.c index 601f01c92f7..270db0f3861 100644 --- a/fs/xfs/linux-2.6/xfs_ioctl32.c +++ b/fs/xfs/linux-2.6/xfs_ioctl32.c @@ -15,7 +15,6 @@ * along with this program; if not, write the Free Software Foundation, * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA */ -#include <linux/config.h> #include <linux/compat.h> #include <linux/init.h> #include <linux/ioctl.h> diff --git a/fs/xfs/linux-2.6/xfs_iops.c b/fs/xfs/linux-2.6/xfs_iops.c index 12810baeb5d..3ba814ae3bb 100644 --- a/fs/xfs/linux-2.6/xfs_iops.c +++ b/fs/xfs/linux-2.6/xfs_iops.c @@ -419,16 +419,15 @@ xfs_vn_link( int error; ip = old_dentry->d_inode; /* inode being linked to */ - if (S_ISDIR(ip->i_mode)) - return -EPERM; - tdvp = vn_from_inode(dir); vp = vn_from_inode(ip); + VN_HOLD(vp); error = bhv_vop_link(tdvp, vp, dentry, NULL); - if (likely(!error)) { + if (unlikely(error)) { + VN_RELE(vp); + } else { VMODIFY(tdvp); - VN_HOLD(vp); xfs_validate_fields(ip, &vattr); d_instantiate(dentry, ip); } @@ -554,13 +553,13 @@ xfs_vn_follow_link( ASSERT(dentry); ASSERT(nd); - link = (char *)kmalloc(MAXPATHLEN+1, GFP_KERNEL); + link = kmalloc(MAXPATHLEN+1, GFP_KERNEL); if (!link) { nd_set_link(nd, ERR_PTR(-ENOMEM)); return NULL; } - uio = (uio_t *)kmalloc(sizeof(uio_t), GFP_KERNEL); + uio = kmalloc(sizeof(uio_t), GFP_KERNEL); if (!uio) { kfree(link); nd_set_link(nd, ERR_PTR(-ENOMEM)); @@ -624,12 +623,27 @@ xfs_vn_getattr( { struct inode *inode = dentry->d_inode; bhv_vnode_t *vp = vn_from_inode(inode); - int error = 0; + bhv_vattr_t vattr = { .va_mask = XFS_AT_STAT }; + int error; - if (unlikely(vp->v_flag & VMODIFIED)) - error = vn_revalidate(vp); - if (!error) - generic_fillattr(inode, stat); + error = bhv_vop_getattr(vp, &vattr, ATTR_LAZY, NULL); + if (likely(!error)) { + stat->size = i_size_read(inode); + stat->dev = inode->i_sb->s_dev; + stat->rdev = (vattr.va_rdev == 0) ? 0 : + MKDEV(sysv_major(vattr.va_rdev) & 0x1ff, + sysv_minor(vattr.va_rdev)); + stat->mode = vattr.va_mode; + stat->nlink = vattr.va_nlink; + stat->uid = vattr.va_uid; + stat->gid = vattr.va_gid; + stat->ino = vattr.va_nodeid; + stat->atime = vattr.va_atime; + stat->mtime = vattr.va_mtime; + stat->ctime = vattr.va_ctime; + stat->blocks = vattr.va_nblocks; + stat->blksize = vattr.va_blocksize; + } return -error; } diff --git a/fs/xfs/linux-2.6/xfs_linux.h b/fs/xfs/linux-2.6/xfs_linux.h index aa26ab906c8..2b0e0018738 100644 --- a/fs/xfs/linux-2.6/xfs_linux.h +++ b/fs/xfs/linux-2.6/xfs_linux.h @@ -19,7 +19,6 @@ #define __XFS_LINUX__ #include <linux/types.h> -#include <linux/config.h> /* * Some types are conditional depending on the target system. @@ -140,9 +139,7 @@ BUFFER_FNS(PrivateStart, unwritten); #define current_pid() (current->pid) #define current_fsuid(cred) (current->fsuid) #define current_fsgid(cred) (current->fsgid) -#define current_set_flags(f) (current->flags |= (f)) #define current_test_flags(f) (current->flags & (f)) -#define current_clear_flags(f) (current->flags & ~(f)) #define current_set_flags_nested(sp, f) \ (*(sp) = current->flags, current->flags |= (f)) #define current_clear_flags_nested(sp, f) \ @@ -151,11 +148,7 @@ BUFFER_FNS(PrivateStart, unwritten); (current->flags = ((current->flags & ~(f)) | (*(sp) & (f)))) #define NBPP PAGE_SIZE -#define DPPSHFT (PAGE_SHIFT - 9) #define NDPP (1 << (PAGE_SHIFT - 9)) -#define dtop(DD) (((DD) + NDPP - 1) >> DPPSHFT) -#define dtopt(DD) ((DD) >> DPPSHFT) -#define dpoff(DD) ((DD) & (NDPP-1)) #define NBBY 8 /* number of bits per byte */ #define NBPC PAGE_SIZE /* Number of bytes per click */ @@ -175,8 +168,6 @@ BUFFER_FNS(PrivateStart, unwritten); #define btoct(x) ((__psunsigned_t)(x)>>BPCSHIFT) #define btoc64(x) (((__uint64_t)(x)+(NBPC-1))>>BPCSHIFT) #define btoct64(x) ((__uint64_t)(x)>>BPCSHIFT) -#define io_btoc(x) (((__psunsigned_t)(x)+(IO_NBPC-1))>>IO_BPCSHIFT) -#define io_btoct(x) ((__psunsigned_t)(x)>>IO_BPCSHIFT) /* off_t bytes to clicks */ #define offtoc(x) (((__uint64_t)(x)+(NBPC-1))>>BPCSHIFT) @@ -189,7 +180,6 @@ BUFFER_FNS(PrivateStart, unwritten); #define ctob(x) ((__psunsigned_t)(x)<<BPCSHIFT) #define btoct(x) ((__psunsigned_t)(x)>>BPCSHIFT) #define ctob64(x) ((__uint64_t)(x)<<BPCSHIFT) -#define io_ctob(x) ((__psunsigned_t)(x)<<IO_BPCSHIFT) /* bytes to clicks */ #define btoc(x) (((__psunsigned_t)(x)+(NBPC-1))>>BPCSHIFT) @@ -218,7 +208,6 @@ BUFFER_FNS(PrivateStart, unwritten); #define MIN(a,b) (min(a,b)) #define MAX(a,b) (max(a,b)) #define howmany(x, y) (((x)+((y)-1))/(y)) -#define roundup(x, y) ((((x)+((y)-1))/(y))*(y)) /* * Various platform dependent calls that don't fit anywhere else @@ -343,4 +332,11 @@ static inline __uint64_t roundup_64(__uint64_t x, __uint32_t y) return(x * y); } +static inline __uint64_t howmany_64(__uint64_t x, __uint32_t y) +{ + x += y - 1; + do_div(x, y); + return x; +} + #endif /* __XFS_LINUX__ */ diff --git a/fs/xfs/linux-2.6/xfs_lrw.c b/fs/xfs/linux-2.6/xfs_lrw.c index 5d9cfd91ad0..55992b40353 100644 --- a/fs/xfs/linux-2.6/xfs_lrw.c +++ b/fs/xfs/linux-2.6/xfs_lrw.c @@ -264,13 +264,18 @@ xfs_read( dmflags, &locktype); if (ret) { xfs_iunlock(ip, XFS_IOLOCK_SHARED); - goto unlock_mutex; + if (unlikely(ioflags & IO_ISDIRECT)) + mutex_unlock(&inode->i_mutex); + return ret; } } - if (unlikely((ioflags & IO_ISDIRECT) && VN_CACHED(vp))) - bhv_vop_flushinval_pages(vp, ctooff(offtoct(*offset)), - -1, FI_REMAPF_LOCKED); + if (unlikely(ioflags & IO_ISDIRECT)) { + if (VN_CACHED(vp)) + bhv_vop_flushinval_pages(vp, ctooff(offtoct(*offset)), + -1, FI_REMAPF_LOCKED); + mutex_unlock(&inode->i_mutex); + } xfs_rw_enter_trace(XFS_READ_ENTER, &ip->i_iocore, (void *)iovp, segs, *offset, ioflags); @@ -281,10 +286,6 @@ xfs_read( XFS_STATS_ADD(xs_read_bytes, ret); xfs_iunlock(ip, XFS_IOLOCK_SHARED); - -unlock_mutex: - if (unlikely(ioflags & IO_ISDIRECT)) - mutex_unlock(&inode->i_mutex); return ret; } @@ -390,6 +391,8 @@ xfs_splice_write( xfs_inode_t *ip = XFS_BHVTOI(bdp); xfs_mount_t *mp = ip->i_mount; ssize_t ret; + struct inode *inode = outfilp->f_mapping->host; + xfs_fsize_t isize; XFS_STATS_INC(xs_write_calls); if (XFS_FORCED_SHUTDOWN(ip->i_mount)) @@ -416,6 +419,20 @@ xfs_splice_write( if (ret > 0) XFS_STATS_ADD(xs_write_bytes, ret); + isize = i_size_read(inode); + if (unlikely(ret < 0 && ret != -EFAULT && *ppos > isize)) + *ppos = isize; + + if (*ppos > ip->i_d.di_size) { + xfs_ilock(ip, XFS_ILOCK_EXCL); + if (*ppos > ip->i_d.di_size) { + ip->i_d.di_size = *ppos; + i_size_write(inode, *ppos); + ip->i_update_core = 1; + ip->i_update_size = 1; + } + xfs_iunlock(ip, XFS_ILOCK_EXCL); + } xfs_iunlock(ip, XFS_IOLOCK_EXCL); return ret; } diff --git a/fs/xfs/linux-2.6/xfs_super.c b/fs/xfs/linux-2.6/xfs_super.c index 9bdef9d5190..38c4d128a8c 100644 --- a/fs/xfs/linux-2.6/xfs_super.c +++ b/fs/xfs/linux-2.6/xfs_super.c @@ -171,7 +171,6 @@ xfs_revalidate_inode( break; } - inode->i_blksize = xfs_preferred_iosize(mp); inode->i_generation = ip->i_d.di_gen; i_size_write(inode, ip->i_d.di_size); inode->i_blocks = @@ -228,7 +227,9 @@ xfs_initialize_vnode( xfs_revalidate_inode(XFS_BHVTOM(bdp), vp, ip); xfs_set_inodeops(inode); + spin_lock(&ip->i_flags_lock); ip->i_flags &= ~XFS_INEW; + spin_unlock(&ip->i_flags_lock); barrier(); unlock_new_inode(inode); @@ -314,6 +315,13 @@ xfs_mountfs_check_barriers(xfs_mount_t *mp) return; } + if (xfs_readonly_buftarg(mp->m_ddev_targp)) { + xfs_fs_cmn_err(CE_NOTE, mp, + "Disabling barriers, underlying device is readonly"); + mp->m_flags &= ~XFS_MOUNT_BARRIER; + return; + } + error = xfs_barrier_test(mp); if (error) { xfs_fs_cmn_err(CE_NOTE, mp, diff --git a/fs/xfs/linux-2.6/xfs_vfs.h b/fs/xfs/linux-2.6/xfs_vfs.h index 91fc2c4b335..da255bdf526 100644 --- a/fs/xfs/linux-2.6/xfs_vfs.h +++ b/fs/xfs/linux-2.6/xfs_vfs.h @@ -79,7 +79,7 @@ typedef enum { #define VFS_RDONLY 0x0001 /* read-only vfs */ #define VFS_GRPID 0x0002 /* group-ID assigned from directory */ #define VFS_DMI 0x0004 /* filesystem has the DMI enabled */ -#define VFS_UMOUNT 0x0008 /* unmount in progress */ +/* ---- VFS_UMOUNT ---- 0x0008 -- unneeded, fixed via kthread APIs */ #define VFS_32BITINODES 0x0010 /* do not use inums above 32 bits */ #define VFS_END 0x0010 /* max flag */ diff --git a/fs/xfs/linux-2.6/xfs_vnode.c b/fs/xfs/linux-2.6/xfs_vnode.c index 6628d96b6fd..553fa731ade 100644 --- a/fs/xfs/linux-2.6/xfs_vnode.c +++ b/fs/xfs/linux-2.6/xfs_vnode.c @@ -122,7 +122,6 @@ vn_revalidate_core( inode->i_blocks = vap->va_nblocks; inode->i_mtime = vap->va_mtime; inode->i_ctime = vap->va_ctime; - inode->i_blksize = vap->va_blocksize; if (vap->va_xflags & XFS_XFLAG_IMMUTABLE) inode->i_flags |= S_IMMUTABLE; else diff --git a/fs/xfs/linux-2.6/xfs_vnode.h b/fs/xfs/linux-2.6/xfs_vnode.h index 35c6a01963a..515f5fdea57 100644 --- a/fs/xfs/linux-2.6/xfs_vnode.h +++ b/fs/xfs/linux-2.6/xfs_vnode.h @@ -85,15 +85,13 @@ typedef enum { #define VN_BHV_HEAD(vp) ((bhv_head_t *)(&((vp)->v_bh))) #define vn_bhv_head_init(bhp,name) bhv_head_init(bhp,name) #define vn_bhv_remove(bhp,bdp) bhv_remove(bhp,bdp) -#define vn_bhv_lookup(bhp,ops) bhv_lookup(bhp,ops) -#define vn_bhv_lookup_unlocked(bhp,ops) bhv_lookup_unlocked(bhp,ops) /* * Vnode to Linux inode mapping. */ static inline struct bhv_vnode *vn_from_inode(struct inode *inode) { - return (bhv_vnode_t *)list_entry(inode, bhv_vnode_t, v_inode); + return container_of(inode, bhv_vnode_t, v_inode); } static inline struct inode *vn_to_inode(struct bhv_vnode *vnode) { diff --git a/fs/xfs/quota/xfs_dquot_item.c b/fs/xfs/quota/xfs_dquot_item.c index 5b2dcc58b24..33ad5af386e 100644 --- a/fs/xfs/quota/xfs_dquot_item.c +++ b/fs/xfs/quota/xfs_dquot_item.c @@ -382,18 +382,6 @@ xfs_qm_dquot_logitem_unlock( /* - * The transaction with the dquot locked has aborted. The dquot - * must not be dirty within the transaction. We simply unlock just - * as if the transaction had been cancelled. - */ -STATIC void -xfs_qm_dquot_logitem_abort( - xfs_dq_logitem_t *ql) -{ - xfs_qm_dquot_logitem_unlock(ql); -} - -/* * 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 @@ -426,7 +414,6 @@ STATIC struct xfs_item_ops xfs_dquot_item_ops = { .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_abort = (void(*)(xfs_log_item_t*))xfs_qm_dquot_logitem_abort, .iop_pushbuf = (void(*)(xfs_log_item_t*)) xfs_qm_dquot_logitem_pushbuf, .iop_committing = (void(*)(xfs_log_item_t*, xfs_lsn_t)) @@ -559,17 +546,6 @@ xfs_qm_qoff_logitem_committed(xfs_qoff_logitem_t *qf, xfs_lsn_t lsn) } /* - * The transaction of which this QUOTAOFF is a part has been aborted. - * Just clean up after ourselves. - * Shouldn't this never happen in the case of qoffend logitems? XXX - */ -STATIC void -xfs_qm_qoff_logitem_abort(xfs_qoff_logitem_t *qf) -{ - kmem_free(qf, sizeof(xfs_qoff_logitem_t)); -} - -/* * 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. */ @@ -644,7 +620,6 @@ STATIC struct xfs_item_ops xfs_qm_qoffend_logitem_ops = { .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_abort = (void(*)(xfs_log_item_t*))xfs_qm_qoff_logitem_abort, .iop_pushbuf = NULL, .iop_committing = (void(*)(xfs_log_item_t*, xfs_lsn_t)) xfs_qm_qoffend_logitem_committing @@ -667,7 +642,6 @@ STATIC struct xfs_item_ops xfs_qm_qoff_logitem_ops = { .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_abort = (void(*)(xfs_log_item_t*))xfs_qm_qoff_logitem_abort, .iop_pushbuf = NULL, .iop_committing = (void(*)(xfs_log_item_t*, xfs_lsn_t)) xfs_qm_qoff_logitem_committing diff --git a/fs/xfs/quota/xfs_qm.c b/fs/xfs/quota/xfs_qm.c index e23e45535c4..7c6a3a50379 100644 --- a/fs/xfs/quota/xfs_qm.c +++ b/fs/xfs/quota/xfs_qm.c @@ -112,17 +112,17 @@ xfs_Gqm_init(void) { xfs_dqhash_t *udqhash, *gdqhash; xfs_qm_t *xqm; - uint i, hsize, flags = KM_SLEEP | KM_MAYFAIL; + size_t hsize; + uint i; /* * Initialize the dquot hash tables. */ - hsize = XFS_QM_HASHSIZE_HIGH; - while (!(udqhash = kmem_zalloc(hsize * sizeof(xfs_dqhash_t), flags))) { - if ((hsize >>= 1) <= XFS_QM_HASHSIZE_LOW) - flags = KM_SLEEP; - } - gdqhash = kmem_zalloc(hsize * sizeof(xfs_dqhash_t), KM_SLEEP); + udqhash = kmem_zalloc_greedy(&hsize, + XFS_QM_HASHSIZE_LOW, XFS_QM_HASHSIZE_HIGH, + KM_SLEEP | KM_MAYFAIL | KM_LARGE); + gdqhash = kmem_zalloc(hsize, KM_SLEEP | KM_LARGE); + hsize /= sizeof(xfs_dqhash_t); ndquot = hsize << 8; xqm = kmem_zalloc(sizeof(xfs_qm_t), KM_SLEEP); diff --git a/fs/xfs/quota/xfs_qm.h b/fs/xfs/quota/xfs_qm.h index 4568deb6da8..689407de0a2 100644 --- a/fs/xfs/quota/xfs_qm.h +++ b/fs/xfs/quota/xfs_qm.h @@ -56,12 +56,6 @@ extern kmem_zone_t *qm_dqtrxzone; #define XFS_QM_HASHSIZE_HIGH ((NBPP * 4) / sizeof(xfs_dqhash_t)) /* - * We output a cmn_err when quotachecking a quota file with more than - * this many fsbs. - */ -#define XFS_QM_BIG_QCHECK_NBLKS 500 - -/* * This defines the unit of allocation of dquots. * Currently, it is just one file system block, and a 4K blk contains 30 * (136 * 30 = 4080) dquots. It's probably not worth trying to make diff --git a/fs/xfs/quota/xfs_qm_bhv.c b/fs/xfs/quota/xfs_qm_bhv.c index e95e99f7168..db8872be8c8 100644 --- a/fs/xfs/quota/xfs_qm_bhv.c +++ b/fs/xfs/quota/xfs_qm_bhv.c @@ -203,7 +203,7 @@ xfs_qm_statvfs( if (error || !vnode) return error; - mp = XFS_BHVTOM(bhv); + mp = xfs_vfstom(bhvtovfs(bhv)); ip = xfs_vtoi(vnode); if (!(ip->i_d.di_flags & XFS_DIFLAG_PROJINHERIT)) @@ -217,17 +217,24 @@ xfs_qm_statvfs( return 0; dp = &dqp->q_core; - limit = dp->d_blk_softlimit ? dp->d_blk_softlimit : dp->d_blk_hardlimit; + limit = dp->d_blk_softlimit ? + be64_to_cpu(dp->d_blk_softlimit) : + be64_to_cpu(dp->d_blk_hardlimit); if (limit && statp->f_blocks > limit) { statp->f_blocks = limit; - statp->f_bfree = (statp->f_blocks > dp->d_bcount) ? - (statp->f_blocks - dp->d_bcount) : 0; + statp->f_bfree = + (statp->f_blocks > be64_to_cpu(dp->d_bcount)) ? + (statp->f_blocks - be64_to_cpu(dp->d_bcount)) : 0; } - limit = dp->d_ino_softlimit ? dp->d_ino_softlimit : dp->d_ino_hardlimit; + + limit = dp->d_ino_softlimit ? + be64_to_cpu(dp->d_ino_softlimit) : + be64_to_cpu(dp->d_ino_hardlimit); if (limit && statp->f_files > limit) { statp->f_files = limit; - statp->f_ffree = (statp->f_files > dp->d_icount) ? - (statp->f_ffree - dp->d_icount) : 0; + statp->f_ffree = + (statp->f_files > be64_to_cpu(dp->d_icount)) ? + (statp->f_ffree - be64_to_cpu(dp->d_icount)) : 0; } xfs_qm_dqput(dqp); diff --git a/fs/xfs/quota/xfs_quota_priv.h b/fs/xfs/quota/xfs_quota_priv.h index b7ddd04aae3..a8b85e2be9d 100644 --- a/fs/xfs/quota/xfs_quota_priv.h +++ b/fs/xfs/quota/xfs_quota_priv.h @@ -75,7 +75,6 @@ static inline int XQMISLCKD(struct xfs_dqhash *h) #define xfs_qm_freelist_lock(qm) XQMLCK(&((qm)->qm_dqfreelist)) #define xfs_qm_freelist_unlock(qm) XQMUNLCK(&((qm)->qm_dqfreelist)) -#define XFS_QM_IS_FREELIST_LOCKED(qm) XQMISLCKD(&((qm)->qm_dqfreelist)) /* * Hash into a bucket in the dquot hash table, based on <mp, id>. @@ -170,6 +169,5 @@ for ((dqp) = (qlist)->qh_next; (dqp) != (xfs_dquot_t *)(qlist); \ #define DQFLAGTO_TYPESTR(d) (((d)->dq_flags & XFS_DQ_USER) ? "USR" : \ (((d)->dq_flags & XFS_DQ_GROUP) ? "GRP" : \ (((d)->dq_flags & XFS_DQ_PROJ) ? "PRJ":"???"))) -#define DQFLAGTO_DIRTYSTR(d) (XFS_DQ_IS_DIRTY(d) ? "DIRTY" : "NOTDIRTY") #endif /* __XFS_QUOTA_PRIV_H__ */ diff --git a/fs/xfs/support/ktrace.c b/fs/xfs/support/ktrace.c index addf5a7ea06..5cf2e86caa7 100644 --- a/fs/xfs/support/ktrace.c +++ b/fs/xfs/support/ktrace.c @@ -75,7 +75,7 @@ ktrace_alloc(int nentries, unsigned int __nocast sleep) sleep); } else { ktep = (ktrace_entry_t*)kmem_zalloc((nentries * sizeof(*ktep)), - sleep); + sleep | KM_LARGE); } if (ktep == NULL) { diff --git a/fs/xfs/xfs_ag.h b/fs/xfs/xfs_ag.h index dc2361dd740..9ece7f87ec5 100644 --- a/fs/xfs/xfs_ag.h +++ b/fs/xfs/xfs_ag.h @@ -150,7 +150,7 @@ typedef struct xfs_agi { #define XFS_BUF_TO_AGFL(bp) ((xfs_agfl_t *)XFS_BUF_PTR(bp)) typedef struct xfs_agfl { - xfs_agblock_t agfl_bno[1]; /* actually XFS_AGFL_SIZE(mp) */ + __be32 agfl_bno[1]; /* actually XFS_AGFL_SIZE(mp) */ } xfs_agfl_t; /* diff --git a/fs/xfs/xfs_alloc.c b/fs/xfs/xfs_alloc.c index eef6763f3a6..e80dda3437d 100644 --- a/fs/xfs/xfs_alloc.c +++ b/fs/xfs/xfs_alloc.c @@ -1477,8 +1477,10 @@ xfs_alloc_ag_vextent_small( /* * Can't allocate from the freelist for some reason. */ - else + else { + fbno = NULLAGBLOCK; flen = 0; + } /* * Can't do the allocation, give up. */ @@ -1835,40 +1837,47 @@ xfs_alloc_fix_freelist( &agbp))) return error; if (!pag->pagf_init) { + ASSERT(flags & XFS_ALLOC_FLAG_TRYLOCK); + ASSERT(!(flags & XFS_ALLOC_FLAG_FREEING)); args->agbp = NULL; return 0; } } else agbp = NULL; - /* If this is a metadata preferred pag and we are user data + /* + * If this is a metadata preferred pag and we are user data * then try somewhere else if we are not being asked to * try harder at this point */ - if (pag->pagf_metadata && args->userdata && flags) { + if (pag->pagf_metadata && args->userdata && + (flags & XFS_ALLOC_FLAG_TRYLOCK)) { + ASSERT(!(flags & XFS_ALLOC_FLAG_FREEING)); args->agbp = NULL; return 0; } - need = XFS_MIN_FREELIST_PAG(pag, mp); - delta = need > pag->pagf_flcount ? need - pag->pagf_flcount : 0; - /* - * If it looks like there isn't a long enough extent, or enough - * total blocks, reject it. - */ - longest = (pag->pagf_longest > delta) ? - (pag->pagf_longest - delta) : - (pag->pagf_flcount > 0 || pag->pagf_longest > 0); - if (args->minlen + args->alignment + args->minalignslop - 1 > longest || - (!(flags & XFS_ALLOC_FLAG_FREEING) && - (int)(pag->pagf_freeblks + pag->pagf_flcount - - need - args->total) < - (int)args->minleft)) { - if (agbp) - xfs_trans_brelse(tp, agbp); - args->agbp = NULL; - return 0; + if (!(flags & XFS_ALLOC_FLAG_FREEING)) { + need = XFS_MIN_FREELIST_PAG(pag, mp); + delta = need > pag->pagf_flcount ? need - pag->pagf_flcount : 0; + /* + * If it looks like there isn't a long enough extent, or enough + * total blocks, reject it. + */ + longest = (pag->pagf_longest > delta) ? + (pag->pagf_longest - delta) : + (pag->pagf_flcount > 0 || pag->pagf_longest > 0); + if ((args->minlen + args->alignment + args->minalignslop - 1) > + longest || + ((int)(pag->pagf_freeblks + pag->pagf_flcount - + need - args->total) < (int)args->minleft)) { + if (agbp) + xfs_trans_brelse(tp, agbp); + args->agbp = NULL; + return 0; + } } + /* * Get the a.g. freespace buffer. * Can fail if we're not blocking on locks, and it's held. @@ -1878,6 +1887,8 @@ xfs_alloc_fix_freelist( &agbp))) return error; if (agbp == NULL) { + ASSERT(flags & XFS_ALLOC_FLAG_TRYLOCK); + ASSERT(!(flags & XFS_ALLOC_FLAG_FREEING)); args->agbp = NULL; return 0; } @@ -1887,22 +1898,24 @@ xfs_alloc_fix_freelist( */ agf = XFS_BUF_TO_AGF(agbp); need = XFS_MIN_FREELIST(agf, mp); - delta = need > be32_to_cpu(agf->agf_flcount) ? - (need - be32_to_cpu(agf->agf_flcount)) : 0; /* * If there isn't enough total or single-extent, reject it. */ - longest = be32_to_cpu(agf->agf_longest); - longest = (longest > delta) ? (longest - delta) : - (be32_to_cpu(agf->agf_flcount) > 0 || longest > 0); - if (args->minlen + args->alignment + args->minalignslop - 1 > longest || - (!(flags & XFS_ALLOC_FLAG_FREEING) && - (int)(be32_to_cpu(agf->agf_freeblks) + - be32_to_cpu(agf->agf_flcount) - need - args->total) < - (int)args->minleft)) { - xfs_trans_brelse(tp, agbp); - args->agbp = NULL; - return 0; + if (!(flags & XFS_ALLOC_FLAG_FREEING)) { + delta = need > be32_to_cpu(agf->agf_flcount) ? + (need - be32_to_cpu(agf->agf_flcount)) : 0; + longest = be32_to_cpu(agf->agf_longest); + longest = (longest > delta) ? (longest - delta) : + (be32_to_cpu(agf->agf_flcount) > 0 || longest > 0); + if ((args->minlen + args->alignment + args->minalignslop - 1) > + longest || + ((int)(be32_to_cpu(agf->agf_freeblks) + + be32_to_cpu(agf->agf_flcount) - need - args->total) < + (int)args->minleft)) { + xfs_trans_brelse(tp, agbp); + args->agbp = NULL; + return 0; + } } /* * Make the freelist shorter if it's too long. @@ -1950,12 +1963,11 @@ xfs_alloc_fix_freelist( * on a completely full ag. */ if (targs.agbno == NULLAGBLOCK) { - if (!(flags & XFS_ALLOC_FLAG_FREEING)) { - xfs_trans_brelse(tp, agflbp); - args->agbp = NULL; - return 0; - } - break; + if (flags & XFS_ALLOC_FLAG_FREEING) + break; + xfs_trans_brelse(tp, agflbp); + args->agbp = NULL; + return 0; } /* * Put each allocated block on the list. @@ -2011,7 +2023,7 @@ xfs_alloc_get_freelist( /* * Get the block number and update the data structures. */ - bno = INT_GET(agfl->agfl_bno[be32_to_cpu(agf->agf_flfirst)], ARCH_CONVERT); + bno = be32_to_cpu(agfl->agfl_bno[be32_to_cpu(agf->agf_flfirst)]); be32_add(&agf->agf_flfirst, 1); xfs_trans_brelse(tp, agflbp); if (be32_to_cpu(agf->agf_flfirst) == XFS_AGFL_SIZE(mp)) @@ -2098,7 +2110,7 @@ xfs_alloc_put_freelist( { xfs_agf_t *agf; /* a.g. freespace structure */ xfs_agfl_t *agfl; /* a.g. free block array */ - xfs_agblock_t *blockp;/* pointer to array entry */ + __be32 *blockp;/* pointer to array entry */ int error; #ifdef XFS_ALLOC_TRACE static char fname[] = "xfs_alloc_put_freelist"; @@ -2122,7 +2134,7 @@ xfs_alloc_put_freelist( pag->pagf_flcount++; ASSERT(be32_to_cpu(agf->agf_flcount) <= XFS_AGFL_SIZE(mp)); blockp = &agfl->agfl_bno[be32_to_cpu(agf->agf_fllast)]; - INT_SET(*blockp, ARCH_CONVERT, bno); + *blockp = cpu_to_be32(bno); TRACE_MODAGF(NULL, agf, XFS_AGF_FLLAST | XFS_AGF_FLCOUNT); xfs_alloc_log_agf(tp, agbp, XFS_AGF_FLLAST | XFS_AGF_FLCOUNT); xfs_trans_log_buf(tp, agflbp, @@ -2442,31 +2454,26 @@ xfs_free_extent( xfs_fsblock_t bno, /* starting block number of extent */ xfs_extlen_t len) /* length of extent */ { -#ifdef DEBUG - xfs_agf_t *agf; /* a.g. freespace header */ -#endif - xfs_alloc_arg_t args; /* allocation argument structure */ + xfs_alloc_arg_t args; int error; ASSERT(len != 0); + memset(&args, 0, sizeof(xfs_alloc_arg_t)); args.tp = tp; args.mp = tp->t_mountp; args.agno = XFS_FSB_TO_AGNO(args.mp, bno); ASSERT(args.agno < args.mp->m_sb.sb_agcount); args.agbno = XFS_FSB_TO_AGBNO(args.mp, bno); - args.alignment = 1; - args.minlen = args.minleft = args.minalignslop = 0; down_read(&args.mp->m_peraglock); args.pag = &args.mp->m_perag[args.agno]; if ((error = xfs_alloc_fix_freelist(&args, XFS_ALLOC_FLAG_FREEING))) goto error0; #ifdef DEBUG ASSERT(args.agbp != NULL); - agf = XFS_BUF_TO_AGF(args.agbp); - ASSERT(args.agbno + len <= be32_to_cpu(agf->agf_length)); + ASSERT((args.agbno + len) <= + be32_to_cpu(XFS_BUF_TO_AGF(args.agbp)->agf_length)); #endif - error = xfs_free_ag_extent(tp, args.agbp, args.agno, args.agbno, - len, 0); + error = xfs_free_ag_extent(tp, args.agbp, args.agno, args.agbno, len, 0); error0: up_read(&args.mp->m_peraglock); return error; diff --git a/fs/xfs/xfs_alloc.h b/fs/xfs/xfs_alloc.h index 650591f999a..5a4256120cc 100644 --- a/fs/xfs/xfs_alloc.h +++ b/fs/xfs/xfs_alloc.h @@ -44,6 +44,26 @@ typedef enum xfs_alloctype #define XFS_ALLOC_FLAG_FREEING 0x00000002 /* indicate caller is freeing extents*/ /* + * In order to avoid ENOSPC-related deadlock caused by + * out-of-order locking of AGF buffer (PV 947395), we place + * constraints on the relationship among actual allocations for + * data blocks, freelist blocks, and potential file data bmap + * btree blocks. However, these restrictions may result in no + * actual space allocated for a delayed extent, for example, a data + * block in a certain AG is allocated but there is no additional + * block for the additional bmap btree block due to a split of the + * bmap btree of the file. The result of this may lead to an + * infinite loop in xfssyncd when the file gets flushed to disk and + * all delayed extents need to be actually allocated. To get around + * this, we explicitly set aside a few blocks which will not be + * reserved in delayed allocation. Considering the minimum number of + * needed freelist blocks is 4 fsbs _per AG_, a potential split of file's bmap + * btree requires 1 fsb, so we set the number of set-aside blocks + * to 4 + 4*agcount. + */ +#define XFS_ALLOC_SET_ASIDE(mp) (4 + ((mp)->m_sb.sb_agcount * 4)) + +/* * Argument structure for xfs_alloc routines. * This is turned into a structure to avoid having 20 arguments passed * down several levels of the stack. diff --git a/fs/xfs/xfs_alloc_btree.c b/fs/xfs/xfs_alloc_btree.c index 7446556e802..74cadf95d4e 100644 --- a/fs/xfs/xfs_alloc_btree.c +++ b/fs/xfs/xfs_alloc_btree.c @@ -92,6 +92,7 @@ xfs_alloc_delrec( xfs_alloc_key_t *rkp; /* right block key pointer */ xfs_alloc_ptr_t *rpp; /* right block address pointer */ int rrecs=0; /* number of records in right block */ + int numrecs; xfs_alloc_rec_t *rrp; /* right block record pointer */ xfs_btree_cur_t *tcur; /* temporary btree cursor */ @@ -115,7 +116,8 @@ xfs_alloc_delrec( /* * Fail if we're off the end of the block. */ - if (ptr > be16_to_cpu(block->bb_numrecs)) { + numrecs = be16_to_cpu(block->bb_numrecs); + if (ptr > numrecs) { *stat = 0; return 0; } @@ -129,18 +131,18 @@ xfs_alloc_delrec( lkp = XFS_ALLOC_KEY_ADDR(block, 1, cur); lpp = XFS_ALLOC_PTR_ADDR(block, 1, cur); #ifdef DEBUG - for (i = ptr; i < be16_to_cpu(block->bb_numrecs); i++) { + for (i = ptr; i < numrecs; i++) { if ((error = xfs_btree_check_sptr(cur, be32_to_cpu(lpp[i]), level))) return error; } #endif - if (ptr < be16_to_cpu(block->bb_numrecs)) { + if (ptr < numrecs) { memmove(&lkp[ptr - 1], &lkp[ptr], - (be16_to_cpu(block->bb_numrecs) - ptr) * sizeof(*lkp)); + (numrecs - ptr) * sizeof(*lkp)); memmove(&lpp[ptr - 1], &lpp[ptr], - (be16_to_cpu(block->bb_numrecs) - ptr) * sizeof(*lpp)); - xfs_alloc_log_ptrs(cur, bp, ptr, be16_to_cpu(block->bb_numrecs) - 1); - xfs_alloc_log_keys(cur, bp, ptr, be16_to_cpu(block->bb_numrecs) - 1); + (numrecs - ptr) * sizeof(*lpp)); + xfs_alloc_log_ptrs(cur, bp, ptr, numrecs - 1); + xfs_alloc_log_keys(cur, bp, ptr, numrecs - 1); } } /* @@ -149,10 +151,10 @@ xfs_alloc_delrec( */ else { lrp = XFS_ALLOC_REC_ADDR(block, 1, cur); - if (ptr < be16_to_cpu(block->bb_numrecs)) { + if (ptr < numrecs) { memmove(&lrp[ptr - 1], &lrp[ptr], - (be16_to_cpu(block->bb_numrecs) - ptr) * sizeof(*lrp)); - xfs_alloc_log_recs(cur, bp, ptr, be16_to_cpu(block->bb_numrecs) - 1); + (numrecs - ptr) * sizeof(*lrp)); + xfs_alloc_log_recs(cur, bp, ptr, numrecs - 1); } /* * If it's the first record in the block, we'll need a key @@ -167,7 +169,8 @@ xfs_alloc_delrec( /* * Decrement and log the number of entries in the block. */ - be16_add(&block->bb_numrecs, -1); + numrecs--; + block->bb_numrecs = cpu_to_be16(numrecs); xfs_alloc_log_block(cur->bc_tp, bp, XFS_BB_NUMRECS); /* * See if the longest free extent in the allocation group was @@ -181,14 +184,14 @@ xfs_alloc_delrec( if (level == 0 && cur->bc_btnum == XFS_BTNUM_CNT && be32_to_cpu(block->bb_rightsib) == NULLAGBLOCK && - ptr > be16_to_cpu(block->bb_numrecs)) { - ASSERT(ptr == be16_to_cpu(block->bb_numrecs) + 1); + ptr > numrecs) { + ASSERT(ptr == numrecs + 1); /* * There are still records in the block. Grab the size * from the last one. */ - if (be16_to_cpu(block->bb_numrecs)) { - rrp = XFS_ALLOC_REC_ADDR(block, be16_to_cpu(block->bb_numrecs), cur); + if (numrecs) { + rrp = XFS_ALLOC_REC_ADDR(block, numrecs, cur); agf->agf_longest = rrp->ar_blockcount; } /* @@ -211,7 +214,7 @@ xfs_alloc_delrec( * and it's NOT the leaf level, * then we can get rid of this level. */ - if (be16_to_cpu(block->bb_numrecs) == 1 && level > 0) { + if (numrecs == 1 && level > 0) { /* * lpp is still set to the first pointer in the block. * Make it the new root of the btree. @@ -267,7 +270,7 @@ xfs_alloc_delrec( * If the number of records remaining in the block is at least * the minimum, we're done. */ - if (be16_to_cpu(block->bb_numrecs) >= XFS_ALLOC_BLOCK_MINRECS(level, cur)) { + if (numrecs >= XFS_ALLOC_BLOCK_MINRECS(level, cur)) { if (level > 0 && (error = xfs_alloc_decrement(cur, level, &i))) return error; *stat = 1; @@ -419,19 +422,21 @@ xfs_alloc_delrec( * See if we can join with the left neighbor block. */ if (lbno != NULLAGBLOCK && - lrecs + be16_to_cpu(block->bb_numrecs) <= XFS_ALLOC_BLOCK_MAXRECS(level, cur)) { + lrecs + numrecs <= XFS_ALLOC_BLOCK_MAXRECS(level, cur)) { /* * Set "right" to be the starting block, * "left" to be the left neighbor. */ rbno = bno; right = block; + rrecs = be16_to_cpu(right->bb_numrecs); rbp = bp; if ((error = xfs_btree_read_bufs(mp, cur->bc_tp, cur->bc_private.a.agno, lbno, 0, &lbp, XFS_ALLOC_BTREE_REF))) return error; left = XFS_BUF_TO_ALLOC_BLOCK(lbp); + lrecs = be16_to_cpu(left->bb_numrecs); if ((error = xfs_btree_check_sblock(cur, left, level, lbp))) return error; } @@ -439,20 +444,21 @@ xfs_alloc_delrec( * If that won't work, see if we can join with the right neighbor block. */ else if (rbno != NULLAGBLOCK && - rrecs + be16_to_cpu(block->bb_numrecs) <= - XFS_ALLOC_BLOCK_MAXRECS(level, cur)) { + rrecs + numrecs <= XFS_ALLOC_BLOCK_MAXRECS(level, cur)) { /* * Set "left" to be the starting block, * "right" to be the right neighbor. */ lbno = bno; left = block; + lrecs = be16_to_cpu(left->bb_numrecs); lbp = bp; if ((error = xfs_btree_read_bufs(mp, cur->bc_tp, cur->bc_private.a.agno, rbno, 0, &rbp, XFS_ALLOC_BTREE_REF))) return error; right = XFS_BUF_TO_ALLOC_BLOCK(rbp); + rrecs = be16_to_cpu(right->bb_numrecs); if ((error = xfs_btree_check_sblock(cur, right, level, rbp))) return error; } @@ -474,34 +480,28 @@ xfs_alloc_delrec( /* * It's a non-leaf. Move keys and pointers. */ - lkp = XFS_ALLOC_KEY_ADDR(left, be16_to_cpu(left->bb_numrecs) + 1, cur); - lpp = XFS_ALLOC_PTR_ADDR(left, be16_to_cpu(left->bb_numrecs) + 1, cur); + lkp = XFS_ALLOC_KEY_ADDR(left, lrecs + 1, cur); + lpp = XFS_ALLOC_PTR_ADDR(left, lrecs + 1, cur); rkp = XFS_ALLOC_KEY_ADDR(right, 1, cur); rpp = XFS_ALLOC_PTR_ADDR(right, 1, cur); #ifdef DEBUG - for (i = 0; i < be16_to_cpu(right->bb_numrecs); i++) { + for (i = 0; i < rrecs; i++) { if ((error = xfs_btree_check_sptr(cur, be32_to_cpu(rpp[i]), level))) return error; } #endif - memcpy(lkp, rkp, be16_to_cpu(right->bb_numrecs) * sizeof(*lkp)); - memcpy(lpp, rpp, be16_to_cpu(right->bb_numrecs) * sizeof(*lpp)); - xfs_alloc_log_keys(cur, lbp, be16_to_cpu(left->bb_numrecs) + 1, - be16_to_cpu(left->bb_numrecs) + - be16_to_cpu(right->bb_numrecs)); - xfs_alloc_log_ptrs(cur, lbp, be16_to_cpu(left->bb_numrecs) + 1, - be16_to_cpu(left->bb_numrecs) + - be16_to_cpu(right->bb_numrecs)); + memcpy(lkp, rkp, rrecs * sizeof(*lkp)); + memcpy(lpp, rpp, rrecs * sizeof(*lpp)); + xfs_alloc_log_keys(cur, lbp, lrecs + 1, lrecs + rrecs); + xfs_alloc_log_ptrs(cur, lbp, lrecs + 1, lrecs + rrecs); } else { /* * It's a leaf. Move records. */ - lrp = XFS_ALLOC_REC_ADDR(left, be16_to_cpu(left->bb_numrecs) + 1, cur); + lrp = XFS_ALLOC_REC_ADDR(left, lrecs + 1, cur); rrp = XFS_ALLOC_REC_ADDR(right, 1, cur); - memcpy(lrp, rrp, be16_to_cpu(right->bb_numrecs) * sizeof(*lrp)); - xfs_alloc_log_recs(cur, lbp, be16_to_cpu(left->bb_numrecs) + 1, - be16_to_cpu(left->bb_numrecs) + - be16_to_cpu(right->bb_numrecs)); + memcpy(lrp, rrp, rrecs * sizeof(*lrp)); + xfs_alloc_log_recs(cur, lbp, lrecs + 1, lrecs + rrecs); } /* * If we joined with the left neighbor, set the buffer in the @@ -509,7 +509,7 @@ xfs_alloc_delrec( */ if (bp != lbp) { xfs_btree_setbuf(cur, level, lbp); - cur->bc_ptrs[level] += be16_to_cpu(left->bb_numrecs); + cur->bc_ptrs[level] += lrecs; } /* * If we joined with the right neighbor and there's a level above @@ -521,7 +521,8 @@ xfs_alloc_delrec( /* * Fix up the number of records in the surviving block. */ - be16_add(&left->bb_numrecs, be16_to_cpu(right->bb_numrecs)); + lrecs += rrecs; + left->bb_numrecs = cpu_to_be16(lrecs); /* * Fix up the right block pointer in the surviving block, and log it. */ @@ -608,6 +609,7 @@ xfs_alloc_insrec( xfs_btree_cur_t *ncur; /* new cursor to be used at next lvl */ xfs_alloc_key_t nkey; /* new key value, from split */ xfs_alloc_rec_t nrec; /* new record value, for caller */ + int numrecs; int optr; /* old ptr value */ xfs_alloc_ptr_t *pp; /* pointer to btree addresses */ int ptr; /* index in btree block for this rec */ @@ -653,13 +655,14 @@ xfs_alloc_insrec( */ bp = cur->bc_bufs[level]; block = XFS_BUF_TO_ALLOC_BLOCK(bp); + numrecs = be16_to_cpu(block->bb_numrecs); #ifdef DEBUG if ((error = xfs_btree_check_sblock(cur, block, level, bp))) return error; /* * Check that the new entry is being inserted in the right place. */ - if (ptr <= be16_to_cpu(block->bb_numrecs)) { + if (ptr <= numrecs) { if (level == 0) { rp = XFS_ALLOC_REC_ADDR(block, ptr, cur); xfs_btree_check_rec(cur->bc_btnum, recp, rp); @@ -670,12 +673,12 @@ xfs_alloc_insrec( } #endif nbno = NULLAGBLOCK; - ncur = (xfs_btree_cur_t *)0; + ncur = NULL; /* * If the block is full, we can't insert the new entry until we * make the block un-full. */ - if (be16_to_cpu(block->bb_numrecs) == XFS_ALLOC_BLOCK_MAXRECS(level, cur)) { + if (numrecs == XFS_ALLOC_BLOCK_MAXRECS(level, cur)) { /* * First, try shifting an entry to the right neighbor. */ @@ -729,6 +732,7 @@ xfs_alloc_insrec( * At this point we know there's room for our new entry in the block * we're pointing at. */ + numrecs = be16_to_cpu(block->bb_numrecs); if (level > 0) { /* * It's a non-leaf entry. Make a hole for the new data @@ -737,15 +741,15 @@ xfs_alloc_insrec( kp = XFS_ALLOC_KEY_ADDR(block, 1, cur); pp = XFS_ALLOC_PTR_ADDR(block, 1, cur); #ifdef DEBUG - for (i = be16_to_cpu(block->bb_numrecs); i >= ptr; i--) { + for (i = numrecs; i >= ptr; i--) { if ((error = xfs_btree_check_sptr(cur, be32_to_cpu(pp[i - 1]), level))) return error; } #endif memmove(&kp[ptr], &kp[ptr - 1], - (be16_to_cpu(block->bb_numrecs) - ptr + 1) * sizeof(*kp)); + (numrecs - ptr + 1) * sizeof(*kp)); memmove(&pp[ptr], &pp[ptr - 1], - (be16_to_cpu(block->bb_numrecs) - ptr + 1) * sizeof(*pp)); + (numrecs - ptr + 1) * sizeof(*pp)); #ifdef DEBUG if ((error = xfs_btree_check_sptr(cur, *bnop, level))) return error; @@ -755,11 +759,12 @@ xfs_alloc_insrec( */ kp[ptr - 1] = key; pp[ptr - 1] = cpu_to_be32(*bnop); - be16_add(&block->bb_numrecs, 1); - xfs_alloc_log_keys(cur, bp, ptr, be16_to_cpu(block->bb_numrecs)); - xfs_alloc_log_ptrs(cur, bp, ptr, be16_to_cpu(block->bb_numrecs)); + numrecs++; + block->bb_numrecs = cpu_to_be16(numrecs); + xfs_alloc_log_keys(cur, bp, ptr, numrecs); + xfs_alloc_log_ptrs(cur, bp, ptr, numrecs); #ifdef DEBUG - if (ptr < be16_to_cpu(block->bb_numrecs)) + if (ptr < numrecs) xfs_btree_check_key(cur->bc_btnum, kp + ptr - 1, kp + ptr); #endif @@ -769,16 +774,17 @@ xfs_alloc_insrec( */ rp = XFS_ALLOC_REC_ADDR(block, 1, cur); memmove(&rp[ptr], &rp[ptr - 1], - (be16_to_cpu(block->bb_numrecs) - ptr + 1) * sizeof(*rp)); + (numrecs - ptr + 1) * sizeof(*rp)); /* * Now stuff the new record in, bump numrecs * and log the new data. */ - rp[ptr - 1] = *recp; /* INT_: struct copy */ - be16_add(&block->bb_numrecs, 1); - xfs_alloc_log_recs(cur, bp, ptr, be16_to_cpu(block->bb_numrecs)); + rp[ptr - 1] = *recp; + numrecs++; + block->bb_numrecs = cpu_to_be16(numrecs); + xfs_alloc_log_recs(cur, bp, ptr, numrecs); #ifdef DEBUG - if (ptr < be16_to_cpu(block->bb_numrecs)) + if (ptr < numrecs) xfs_btree_check_rec(cur->bc_btnum, rp + ptr - 1, rp + ptr); #endif @@ -819,8 +825,8 @@ xfs_alloc_insrec( */ *bnop = nbno; if (nbno != NULLAGBLOCK) { - *recp = nrec; /* INT_: struct copy */ - *curp = ncur; /* INT_: struct copy */ + *recp = nrec; + *curp = ncur; } *stat = 1; return 0; @@ -981,7 +987,7 @@ xfs_alloc_lookup( */ bp = cur->bc_bufs[level]; if (bp && XFS_BUF_ADDR(bp) != d) - bp = (xfs_buf_t *)0; + bp = NULL; if (!bp) { /* * Need to get a new buffer. Read it, then @@ -1229,7 +1235,7 @@ xfs_alloc_lshift( if ((error = xfs_btree_check_sptr(cur, be32_to_cpu(*rpp), level))) return error; #endif - *lpp = *rpp; /* INT_: copy */ + *lpp = *rpp; xfs_alloc_log_ptrs(cur, lbp, nrec, nrec); xfs_btree_check_key(cur->bc_btnum, lkp - 1, lkp); } @@ -1406,8 +1412,8 @@ xfs_alloc_newroot( kp = XFS_ALLOC_KEY_ADDR(new, 1, cur); if (be16_to_cpu(left->bb_level) > 0) { - kp[0] = *XFS_ALLOC_KEY_ADDR(left, 1, cur); /* INT_: structure copy */ - kp[1] = *XFS_ALLOC_KEY_ADDR(right, 1, cur);/* INT_: structure copy */ + kp[0] = *XFS_ALLOC_KEY_ADDR(left, 1, cur); + kp[1] = *XFS_ALLOC_KEY_ADDR(right, 1, cur); } else { xfs_alloc_rec_t *rp; /* btree record pointer */ @@ -1527,8 +1533,8 @@ xfs_alloc_rshift( if ((error = xfs_btree_check_sptr(cur, be32_to_cpu(*lpp), level))) return error; #endif - *rkp = *lkp; /* INT_: copy */ - *rpp = *lpp; /* INT_: copy */ + *rkp = *lkp; + *rpp = *lpp; xfs_alloc_log_keys(cur, rbp, 1, be16_to_cpu(right->bb_numrecs) + 1); xfs_alloc_log_ptrs(cur, rbp, 1, be16_to_cpu(right->bb_numrecs) + 1); xfs_btree_check_key(cur->bc_btnum, rkp, rkp + 1); @@ -2044,7 +2050,7 @@ xfs_alloc_insert( nbno = NULLAGBLOCK; nrec.ar_startblock = cpu_to_be32(cur->bc_rec.a.ar_startblock); nrec.ar_blockcount = cpu_to_be32(cur->bc_rec.a.ar_blockcount); - ncur = (xfs_btree_cur_t *)0; + ncur = NULL; pcur = cur; /* * Loop going up the tree, starting at the leaf level. @@ -2076,7 +2082,7 @@ xfs_alloc_insert( */ if (ncur) { pcur = ncur; - ncur = (xfs_btree_cur_t *)0; + ncur = NULL; } } while (nbno != NULLAGBLOCK); *stat = i; diff --git a/fs/xfs/xfs_attr.c b/fs/xfs/xfs_attr.c index 1a210104327..9ada7bdbae5 100644 --- a/fs/xfs/xfs_attr.c +++ b/fs/xfs/xfs_attr.c @@ -91,7 +91,6 @@ STATIC int xfs_attr_refillstate(xfs_da_state_t *state); /* * Routines to manipulate out-of-line attribute values. */ -STATIC int xfs_attr_rmtval_get(xfs_da_args_t *args); STATIC int xfs_attr_rmtval_set(xfs_da_args_t *args); STATIC int xfs_attr_rmtval_remove(xfs_da_args_t *args); @@ -180,7 +179,7 @@ xfs_attr_get(bhv_desc_t *bdp, const char *name, char *value, int *valuelenp, return(error); } -STATIC int +int xfs_attr_set_int(xfs_inode_t *dp, const char *name, int namelen, char *value, int valuelen, int flags) { @@ -440,7 +439,7 @@ xfs_attr_set(bhv_desc_t *bdp, const char *name, char *value, int valuelen, int f * Generic handler routine to remove a name from an attribute list. * Transitions attribute list from Btree to shortform as necessary. */ -STATIC int +int xfs_attr_remove_int(xfs_inode_t *dp, const char *name, int namelen, int flags) { xfs_da_args_t args; @@ -591,6 +590,110 @@ xfs_attr_remove(bhv_desc_t *bdp, const char *name, int flags, struct cred *cred) return xfs_attr_remove_int(dp, name, namelen, flags); } +int /* error */ +xfs_attr_list_int(xfs_attr_list_context_t *context) +{ + int error; + xfs_inode_t *dp = context->dp; + + /* + * Decide on what work routines to call based on the inode size. + */ + if (XFS_IFORK_Q(dp) == 0 || + (dp->i_d.di_aformat == XFS_DINODE_FMT_EXTENTS && + dp->i_d.di_anextents == 0)) { + error = 0; + } else if (dp->i_d.di_aformat == XFS_DINODE_FMT_LOCAL) { + error = xfs_attr_shortform_list(context); + } else if (xfs_bmap_one_block(dp, XFS_ATTR_FORK)) { + error = xfs_attr_leaf_list(context); + } else { + error = xfs_attr_node_list(context); + } + return error; +} + +#define ATTR_ENTBASESIZE /* minimum bytes used by an attr */ \ + (((struct attrlist_ent *) 0)->a_name - (char *) 0) +#define ATTR_ENTSIZE(namelen) /* actual bytes used by an attr */ \ + ((ATTR_ENTBASESIZE + (namelen) + 1 + sizeof(u_int32_t)-1) \ + & ~(sizeof(u_int32_t)-1)) + +/* + * Format an attribute and copy it out to the user's buffer. + * Take care to check values and protect against them changing later, + * we may be reading them directly out of a user buffer. + */ +/*ARGSUSED*/ +STATIC int +xfs_attr_put_listent(xfs_attr_list_context_t *context, attrnames_t *namesp, + char *name, int namelen, + int valuelen, char *value) +{ + attrlist_ent_t *aep; + int arraytop; + + ASSERT(!(context->flags & ATTR_KERNOVAL)); + ASSERT(context->count >= 0); + ASSERT(context->count < (ATTR_MAX_VALUELEN/8)); + ASSERT(context->firstu >= sizeof(*context->alist)); + ASSERT(context->firstu <= context->bufsize); + + arraytop = sizeof(*context->alist) + + context->count * sizeof(context->alist->al_offset[0]); + context->firstu -= ATTR_ENTSIZE(namelen); + if (context->firstu < arraytop) { + xfs_attr_trace_l_c("buffer full", context); + context->alist->al_more = 1; + context->seen_enough = 1; + return 1; + } + + aep = (attrlist_ent_t *)&(((char *)context->alist)[ context->firstu ]); + aep->a_valuelen = valuelen; + memcpy(aep->a_name, name, namelen); + aep->a_name[ namelen ] = 0; + context->alist->al_offset[ context->count++ ] = context->firstu; + context->alist->al_count = context->count; + xfs_attr_trace_l_c("add", context); + return 0; +} + +STATIC int +xfs_attr_kern_list(xfs_attr_list_context_t *context, attrnames_t *namesp, + char *name, int namelen, + int valuelen, char *value) +{ + char *offset; + int arraytop; + + ASSERT(context->count >= 0); + + arraytop = context->count + namesp->attr_namelen + namelen + 1; + if (arraytop > context->firstu) { + context->count = -1; /* insufficient space */ + return 1; + } + offset = (char *)context->alist + context->count; + strncpy(offset, namesp->attr_name, namesp->attr_namelen); + offset += namesp->attr_namelen; + strncpy(offset, name, namelen); /* real name */ + offset += namelen; + *offset = '\0'; + context->count += namesp->attr_namelen + namelen + 1; + return 0; +} + +/*ARGSUSED*/ +STATIC int +xfs_attr_kern_list_sizes(xfs_attr_list_context_t *context, attrnames_t *namesp, + char *name, int namelen, + int valuelen, char *value) +{ + context->count += namesp->attr_namelen + namelen + 1; + return 0; +} + /* * Generate a list of extended attribute names and optionally * also value lengths. Positive return value follows the XFS @@ -615,13 +718,13 @@ xfs_attr_list(bhv_desc_t *bdp, char *buffer, int bufsize, int flags, return(XFS_ERROR(EINVAL)); if ((cursor->initted == 0) && (cursor->hashval || cursor->blkno || cursor->offset)) - return(XFS_ERROR(EINVAL)); + return XFS_ERROR(EINVAL); /* * Check for a properly aligned buffer. */ if (((long)buffer) & (sizeof(int)-1)) - return(XFS_ERROR(EFAULT)); + return XFS_ERROR(EFAULT); if (flags & ATTR_KERNOVAL) bufsize = 0; @@ -634,53 +737,47 @@ xfs_attr_list(bhv_desc_t *bdp, char *buffer, int bufsize, int flags, context.dupcnt = 0; context.resynch = 1; context.flags = flags; - if (!(flags & ATTR_KERNAMELS)) { + context.seen_enough = 0; + context.alist = (attrlist_t *)buffer; + context.put_value = 0; + + if (flags & ATTR_KERNAMELS) { + context.bufsize = bufsize; + context.firstu = context.bufsize; + if (flags & ATTR_KERNOVAL) + context.put_listent = xfs_attr_kern_list_sizes; + else + context.put_listent = xfs_attr_kern_list; + } else { context.bufsize = (bufsize & ~(sizeof(int)-1)); /* align */ context.firstu = context.bufsize; - context.alist = (attrlist_t *)buffer; context.alist->al_count = 0; context.alist->al_more = 0; context.alist->al_offset[0] = context.bufsize; - } - else { - context.bufsize = bufsize; - context.firstu = context.bufsize; - context.alist = (attrlist_t *)buffer; + context.put_listent = xfs_attr_put_listent; } if (XFS_FORCED_SHUTDOWN(dp->i_mount)) - return (EIO); + return EIO; xfs_ilock(dp, XFS_ILOCK_SHARED); - /* - * Decide on what work routines to call based on the inode size. - */ xfs_attr_trace_l_c("syscall start", &context); - if (XFS_IFORK_Q(dp) == 0 || - (dp->i_d.di_aformat == XFS_DINODE_FMT_EXTENTS && - dp->i_d.di_anextents == 0)) { - error = 0; - } else if (dp->i_d.di_aformat == XFS_DINODE_FMT_LOCAL) { - error = xfs_attr_shortform_list(&context); - } else if (xfs_bmap_one_block(dp, XFS_ATTR_FORK)) { - error = xfs_attr_leaf_list(&context); - } else { - error = xfs_attr_node_list(&context); - } + + error = xfs_attr_list_int(&context); + xfs_iunlock(dp, XFS_ILOCK_SHARED); xfs_attr_trace_l_c("syscall end", &context); - if (!(context.flags & (ATTR_KERNOVAL|ATTR_KERNAMELS))) { - ASSERT(error >= 0); - } - else { /* must return negated buffer size or the error */ + if (context.flags & (ATTR_KERNOVAL|ATTR_KERNAMELS)) { + /* must return negated buffer size or the error */ if (context.count < 0) error = XFS_ERROR(ERANGE); else error = -context.count; - } + } else + ASSERT(error >= 0); - return(error); + return error; } int /* error */ @@ -1122,19 +1219,19 @@ xfs_attr_leaf_list(xfs_attr_list_context_t *context) context->cursor->blkno = 0; error = xfs_da_read_buf(NULL, context->dp, 0, -1, &bp, XFS_ATTR_FORK); if (error) - return(error); + return XFS_ERROR(error); ASSERT(bp != NULL); leaf = bp->data; if (unlikely(be16_to_cpu(leaf->hdr.info.magic) != XFS_ATTR_LEAF_MAGIC)) { XFS_CORRUPTION_ERROR("xfs_attr_leaf_list", XFS_ERRLEVEL_LOW, context->dp->i_mount, leaf); xfs_da_brelse(NULL, bp); - return(XFS_ERROR(EFSCORRUPTED)); + return XFS_ERROR(EFSCORRUPTED); } - (void)xfs_attr_leaf_list_int(bp, context); + error = xfs_attr_leaf_list_int(bp, context); xfs_da_brelse(NULL, bp); - return(0); + return XFS_ERROR(error); } @@ -1858,8 +1955,12 @@ xfs_attr_node_list(xfs_attr_list_context_t *context) return(XFS_ERROR(EFSCORRUPTED)); } error = xfs_attr_leaf_list_int(bp, context); - if (error || !leaf->hdr.info.forw) - break; /* not really an error, buffer full or EOF */ + if (error) { + xfs_da_brelse(NULL, bp); + return error; + } + if (context->seen_enough || leaf->hdr.info.forw == 0) + break; cursor->blkno = be32_to_cpu(leaf->hdr.info.forw); xfs_da_brelse(NULL, bp); error = xfs_da_read_buf(NULL, context->dp, cursor->blkno, -1, @@ -1886,7 +1987,7 @@ xfs_attr_node_list(xfs_attr_list_context_t *context) * Read the value associated with an attribute from the out-of-line buffer * that we stored it in. */ -STATIC int +int xfs_attr_rmtval_get(xfs_da_args_t *args) { xfs_bmbt_irec_t map[ATTR_RMTVALUE_MAPSIZE]; diff --git a/fs/xfs/xfs_attr.h b/fs/xfs/xfs_attr.h index 981633f6c07..783977d3ea7 100644 --- a/fs/xfs/xfs_attr.h +++ b/fs/xfs/xfs_attr.h @@ -37,6 +37,7 @@ struct cred; struct bhv_vnode; +struct xfs_attr_list_context; typedef int (*attrset_t)(struct bhv_vnode *, char *, void *, size_t, int); typedef int (*attrget_t)(struct bhv_vnode *, char *, void *, size_t, int); @@ -160,13 +161,16 @@ struct xfs_da_args; */ int xfs_attr_get(bhv_desc_t *, const char *, char *, int *, int, struct cred *); int xfs_attr_set(bhv_desc_t *, const char *, char *, int, int, struct cred *); +int xfs_attr_set_int(struct xfs_inode *, const char *, int, char *, int, int); int xfs_attr_remove(bhv_desc_t *, const char *, int, struct cred *); -int xfs_attr_list(bhv_desc_t *, char *, int, int, - struct attrlist_cursor_kern *, struct cred *); +int xfs_attr_remove_int(struct xfs_inode *, const char *, int, int); +int xfs_attr_list(bhv_desc_t *, char *, int, int, struct attrlist_cursor_kern *, struct cred *); +int xfs_attr_list_int(struct xfs_attr_list_context *); int xfs_attr_inactive(struct xfs_inode *dp); int xfs_attr_shortform_getvalue(struct xfs_da_args *); int xfs_attr_fetch(struct xfs_inode *, const char *, int, char *, int *, int, struct cred *); +int xfs_attr_rmtval_get(struct xfs_da_args *args); #endif /* __XFS_ATTR_H__ */ diff --git a/fs/xfs/xfs_attr_leaf.c b/fs/xfs/xfs_attr_leaf.c index 9455051f012..9719bbef122 100644 --- a/fs/xfs/xfs_attr_leaf.c +++ b/fs/xfs/xfs_attr_leaf.c @@ -89,9 +89,46 @@ STATIC void xfs_attr_leaf_moveents(xfs_attr_leafblock_t *src_leaf, int dst_start, int move_count, xfs_mount_t *mp); STATIC int xfs_attr_leaf_entsize(xfs_attr_leafblock_t *leaf, int index); -STATIC int xfs_attr_put_listent(xfs_attr_list_context_t *context, - attrnames_t *, char *name, int namelen, - int valuelen); + +/*======================================================================== + * Namespace helper routines + *========================================================================*/ + +STATIC inline attrnames_t * +xfs_attr_flags_namesp(int flags) +{ + return ((flags & XFS_ATTR_SECURE) ? &attr_secure: + ((flags & XFS_ATTR_ROOT) ? &attr_trusted : &attr_user)); +} + +/* + * If namespace bits don't match return 0. + * If all match then return 1. + */ +STATIC inline int +xfs_attr_namesp_match(int arg_flags, int ondisk_flags) +{ + return XFS_ATTR_NSP_ONDISK(ondisk_flags) == XFS_ATTR_NSP_ARGS_TO_ONDISK(arg_flags); +} + +/* + * If namespace bits don't match and we don't have an override for it + * then return 0. + * If all match or are overridable then return 1. + */ +STATIC inline int +xfs_attr_namesp_match_overrides(int arg_flags, int ondisk_flags) +{ + if (((arg_flags & ATTR_SECURE) == 0) != + ((ondisk_flags & XFS_ATTR_SECURE) == 0) && + !(arg_flags & ATTR_KERNORMALS)) + return 0; + if (((arg_flags & ATTR_ROOT) == 0) != + ((ondisk_flags & XFS_ATTR_ROOT) == 0) && + !(arg_flags & ATTR_KERNROOTLS)) + return 0; + return 1; +} /*======================================================================== @@ -228,11 +265,7 @@ xfs_attr_shortform_add(xfs_da_args_t *args, int forkoff) continue; if (memcmp(args->name, sfe->nameval, args->namelen) != 0) continue; - if (((args->flags & ATTR_SECURE) != 0) != - ((sfe->flags & XFS_ATTR_SECURE) != 0)) - continue; - if (((args->flags & ATTR_ROOT) != 0) != - ((sfe->flags & XFS_ATTR_ROOT) != 0)) + if (!xfs_attr_namesp_match(args->flags, sfe->flags)) continue; ASSERT(0); #endif @@ -246,8 +279,7 @@ xfs_attr_shortform_add(xfs_da_args_t *args, int forkoff) sfe->namelen = args->namelen; sfe->valuelen = args->valuelen; - sfe->flags = (args->flags & ATTR_SECURE) ? XFS_ATTR_SECURE : - ((args->flags & ATTR_ROOT) ? XFS_ATTR_ROOT : 0); + sfe->flags = XFS_ATTR_NSP_ARGS_TO_ONDISK(args->flags); memcpy(sfe->nameval, args->name, args->namelen); memcpy(&sfe->nameval[args->namelen], args->value, args->valuelen); sf->hdr.count++; @@ -282,11 +314,7 @@ xfs_attr_shortform_remove(xfs_da_args_t *args) continue; if (memcmp(sfe->nameval, args->name, args->namelen) != 0) continue; - if (((args->flags & ATTR_SECURE) != 0) != - ((sfe->flags & XFS_ATTR_SECURE) != 0)) - continue; - if (((args->flags & ATTR_ROOT) != 0) != - ((sfe->flags & XFS_ATTR_ROOT) != 0)) + if (!xfs_attr_namesp_match(args->flags, sfe->flags)) continue; break; } @@ -363,11 +391,7 @@ xfs_attr_shortform_lookup(xfs_da_args_t *args) continue; if (memcmp(args->name, sfe->nameval, args->namelen) != 0) continue; - if (((args->flags & ATTR_SECURE) != 0) != - ((sfe->flags & XFS_ATTR_SECURE) != 0)) - continue; - if (((args->flags & ATTR_ROOT) != 0) != - ((sfe->flags & XFS_ATTR_ROOT) != 0)) + if (!xfs_attr_namesp_match(args->flags, sfe->flags)) continue; return(XFS_ERROR(EEXIST)); } @@ -394,11 +418,7 @@ xfs_attr_shortform_getvalue(xfs_da_args_t *args) continue; if (memcmp(args->name, sfe->nameval, args->namelen) != 0) continue; - if (((args->flags & ATTR_SECURE) != 0) != - ((sfe->flags & XFS_ATTR_SECURE) != 0)) - continue; - if (((args->flags & ATTR_ROOT) != 0) != - ((sfe->flags & XFS_ATTR_ROOT) != 0)) + if (!xfs_attr_namesp_match(args->flags, sfe->flags)) continue; if (args->flags & ATTR_KERNOVAL) { args->valuelen = sfe->valuelen; @@ -485,8 +505,7 @@ xfs_attr_shortform_to_leaf(xfs_da_args_t *args) nargs.valuelen = sfe->valuelen; nargs.hashval = xfs_da_hashname((char *)sfe->nameval, sfe->namelen); - nargs.flags = (sfe->flags & XFS_ATTR_SECURE) ? ATTR_SECURE : - ((sfe->flags & XFS_ATTR_ROOT) ? ATTR_ROOT : 0); + nargs.flags = XFS_ATTR_NSP_ONDISK_TO_ARGS(sfe->flags); error = xfs_attr_leaf_lookup_int(bp, &nargs); /* set a->index */ ASSERT(error == ENOATTR); error = xfs_attr_leaf_add(bp, &nargs); @@ -520,6 +539,10 @@ xfs_attr_shortform_compare(const void *a, const void *b) } } + +#define XFS_ISRESET_CURSOR(cursor) \ + (!((cursor)->initted) && !((cursor)->hashval) && \ + !((cursor)->blkno) && !((cursor)->offset)) /* * Copy out entries of shortform attribute lists for attr_list(). * Shortform attribute lists are not stored in hashval sorted order. @@ -537,6 +560,7 @@ xfs_attr_shortform_list(xfs_attr_list_context_t *context) xfs_attr_sf_entry_t *sfe; xfs_inode_t *dp; int sbsize, nsbuf, count, i; + int error; ASSERT(context != NULL); dp = context->dp; @@ -552,46 +576,51 @@ xfs_attr_shortform_list(xfs_attr_list_context_t *context) xfs_attr_trace_l_c("sf start", context); /* - * If the buffer is large enough, do not bother with sorting. + * If the buffer is large enough and the cursor is at the start, + * do not bother with sorting since we will return everything in + * one buffer and another call using the cursor won't need to be + * made. * Note the generous fudge factor of 16 overhead bytes per entry. + * If bufsize is zero then put_listent must be a search function + * and can just scan through what we have. */ - if ((dp->i_afp->if_bytes + sf->hdr.count * 16) < context->bufsize) { + if (context->bufsize == 0 || + (XFS_ISRESET_CURSOR(cursor) && + (dp->i_afp->if_bytes + sf->hdr.count * 16) < context->bufsize)) { for (i = 0, sfe = &sf->list[0]; i < sf->hdr.count; i++) { attrnames_t *namesp; - if (((context->flags & ATTR_SECURE) != 0) != - ((sfe->flags & XFS_ATTR_SECURE) != 0) && - !(context->flags & ATTR_KERNORMALS)) { - sfe = XFS_ATTR_SF_NEXTENTRY(sfe); - continue; - } - if (((context->flags & ATTR_ROOT) != 0) != - ((sfe->flags & XFS_ATTR_ROOT) != 0) && - !(context->flags & ATTR_KERNROOTLS)) { + if (!xfs_attr_namesp_match_overrides(context->flags, sfe->flags)) { sfe = XFS_ATTR_SF_NEXTENTRY(sfe); continue; } - namesp = (sfe->flags & XFS_ATTR_SECURE) ? &attr_secure: - ((sfe->flags & XFS_ATTR_ROOT) ? &attr_trusted : - &attr_user); - if (context->flags & ATTR_KERNOVAL) { - ASSERT(context->flags & ATTR_KERNAMELS); - context->count += namesp->attr_namelen + - sfe->namelen + 1; - } - else { - if (xfs_attr_put_listent(context, namesp, - (char *)sfe->nameval, - (int)sfe->namelen, - (int)sfe->valuelen)) - break; - } + namesp = xfs_attr_flags_namesp(sfe->flags); + error = context->put_listent(context, + namesp, + (char *)sfe->nameval, + (int)sfe->namelen, + (int)sfe->valuelen, + (char*)&sfe->nameval[sfe->namelen]); + + /* + * Either search callback finished early or + * didn't fit it all in the buffer after all. + */ + if (context->seen_enough) + break; + + if (error) + return error; sfe = XFS_ATTR_SF_NEXTENTRY(sfe); } xfs_attr_trace_l_c("sf big-gulp", context); return(0); } + /* do no more for a search callback */ + if (context->bufsize == 0) + return 0; + /* * It didn't all fit, so we have to sort everything on hashval. */ @@ -614,15 +643,7 @@ xfs_attr_shortform_list(xfs_attr_list_context_t *context) kmem_free(sbuf, sbsize); return XFS_ERROR(EFSCORRUPTED); } - if (((context->flags & ATTR_SECURE) != 0) != - ((sfe->flags & XFS_ATTR_SECURE) != 0) && - !(context->flags & ATTR_KERNORMALS)) { - sfe = XFS_ATTR_SF_NEXTENTRY(sfe); - continue; - } - if (((context->flags & ATTR_ROOT) != 0) != - ((sfe->flags & XFS_ATTR_ROOT) != 0) && - !(context->flags & ATTR_KERNROOTLS)) { + if (!xfs_attr_namesp_match_overrides(context->flags, sfe->flags)) { sfe = XFS_ATTR_SF_NEXTENTRY(sfe); continue; } @@ -671,24 +692,22 @@ xfs_attr_shortform_list(xfs_attr_list_context_t *context) for ( ; i < nsbuf; i++, sbp++) { attrnames_t *namesp; - namesp = (sbp->flags & XFS_ATTR_SECURE) ? &attr_secure : - ((sbp->flags & XFS_ATTR_ROOT) ? &attr_trusted : - &attr_user); + namesp = xfs_attr_flags_namesp(sbp->flags); if (cursor->hashval != sbp->hash) { cursor->hashval = sbp->hash; cursor->offset = 0; } - if (context->flags & ATTR_KERNOVAL) { - ASSERT(context->flags & ATTR_KERNAMELS); - context->count += namesp->attr_namelen + - sbp->namelen + 1; - } else { - if (xfs_attr_put_listent(context, namesp, - sbp->name, sbp->namelen, - sbp->valuelen)) - break; - } + error = context->put_listent(context, + namesp, + sbp->name, + sbp->namelen, + sbp->valuelen, + &sbp->name[sbp->namelen]); + if (error) + return error; + if (context->seen_enough) + break; cursor->offset++; } @@ -810,8 +829,7 @@ xfs_attr_leaf_to_shortform(xfs_dabuf_t *bp, xfs_da_args_t *args, int forkoff) nargs.value = (char *)&name_loc->nameval[nargs.namelen]; nargs.valuelen = be16_to_cpu(name_loc->valuelen); nargs.hashval = be32_to_cpu(entry->hashval); - nargs.flags = (entry->flags & XFS_ATTR_SECURE) ? ATTR_SECURE : - ((entry->flags & XFS_ATTR_ROOT) ? ATTR_ROOT : 0); + nargs.flags = XFS_ATTR_NSP_ONDISK_TO_ARGS(entry->flags); xfs_attr_shortform_add(&nargs, forkoff); } error = 0; @@ -1098,8 +1116,7 @@ xfs_attr_leaf_add_work(xfs_dabuf_t *bp, xfs_da_args_t *args, int mapindex) be16_to_cpu(map->size)); entry->hashval = cpu_to_be32(args->hashval); entry->flags = tmp ? XFS_ATTR_LOCAL : 0; - entry->flags |= (args->flags & ATTR_SECURE) ? XFS_ATTR_SECURE : - ((args->flags & ATTR_ROOT) ? XFS_ATTR_ROOT : 0); + entry->flags |= XFS_ATTR_NSP_ARGS_TO_ONDISK(args->flags); if (args->rename) { entry->flags |= XFS_ATTR_INCOMPLETE; if ((args->blkno2 == args->blkno) && @@ -1926,7 +1943,7 @@ xfs_attr_leaf_lookup_int(xfs_dabuf_t *bp, xfs_da_args_t *args) else break; } - ASSERT((probe >= 0) && + ASSERT((probe >= 0) && (!leaf->hdr.count || (probe < be16_to_cpu(leaf->hdr.count)))); ASSERT((span <= 4) || (be32_to_cpu(entry->hashval) == hashval)); @@ -1971,14 +1988,9 @@ xfs_attr_leaf_lookup_int(xfs_dabuf_t *bp, xfs_da_args_t *args) name_loc = XFS_ATTR_LEAF_NAME_LOCAL(leaf, probe); if (name_loc->namelen != args->namelen) continue; - if (memcmp(args->name, (char *)name_loc->nameval, - args->namelen) != 0) + if (memcmp(args->name, (char *)name_loc->nameval, args->namelen) != 0) continue; - if (((args->flags & ATTR_SECURE) != 0) != - ((entry->flags & XFS_ATTR_SECURE) != 0)) - continue; - if (((args->flags & ATTR_ROOT) != 0) != - ((entry->flags & XFS_ATTR_ROOT) != 0)) + if (!xfs_attr_namesp_match(args->flags, entry->flags)) continue; args->index = probe; return(XFS_ERROR(EEXIST)); @@ -1989,11 +2001,7 @@ xfs_attr_leaf_lookup_int(xfs_dabuf_t *bp, xfs_da_args_t *args) if (memcmp(args->name, (char *)name_rmt->name, args->namelen) != 0) continue; - if (((args->flags & ATTR_SECURE) != 0) != - ((entry->flags & XFS_ATTR_SECURE) != 0)) - continue; - if (((args->flags & ATTR_ROOT) != 0) != - ((entry->flags & XFS_ATTR_ROOT) != 0)) + if (!xfs_attr_namesp_match(args->flags, entry->flags)) continue; args->index = probe; args->rmtblkno = be32_to_cpu(name_rmt->valueblk); @@ -2312,8 +2320,6 @@ xfs_attr_leaf_list_int(xfs_dabuf_t *bp, xfs_attr_list_context_t *context) attrlist_cursor_kern_t *cursor; xfs_attr_leafblock_t *leaf; xfs_attr_leaf_entry_t *entry; - xfs_attr_leaf_name_local_t *name_loc; - xfs_attr_leaf_name_remote_t *name_rmt; int retval, i; ASSERT(bp != NULL); @@ -2355,9 +2361,8 @@ xfs_attr_leaf_list_int(xfs_dabuf_t *bp, xfs_attr_list_context_t *context) * We have found our place, start copying out the new attributes. */ retval = 0; - for ( ; (i < be16_to_cpu(leaf->hdr.count)) - && (retval == 0); entry++, i++) { - attrnames_t *namesp; + for ( ; (i < be16_to_cpu(leaf->hdr.count)); entry++, i++) { + attrnames_t *namesp; if (be32_to_cpu(entry->hashval) != cursor->hashval) { cursor->hashval = be32_to_cpu(entry->hashval); @@ -2366,115 +2371,69 @@ xfs_attr_leaf_list_int(xfs_dabuf_t *bp, xfs_attr_list_context_t *context) if (entry->flags & XFS_ATTR_INCOMPLETE) continue; /* skip incomplete entries */ - if (((context->flags & ATTR_SECURE) != 0) != - ((entry->flags & XFS_ATTR_SECURE) != 0) && - !(context->flags & ATTR_KERNORMALS)) - continue; /* skip non-matching entries */ - if (((context->flags & ATTR_ROOT) != 0) != - ((entry->flags & XFS_ATTR_ROOT) != 0) && - !(context->flags & ATTR_KERNROOTLS)) - continue; /* skip non-matching entries */ - - namesp = (entry->flags & XFS_ATTR_SECURE) ? &attr_secure : - ((entry->flags & XFS_ATTR_ROOT) ? &attr_trusted : - &attr_user); + if (!xfs_attr_namesp_match_overrides(context->flags, entry->flags)) + continue; + + namesp = xfs_attr_flags_namesp(entry->flags); if (entry->flags & XFS_ATTR_LOCAL) { - name_loc = XFS_ATTR_LEAF_NAME_LOCAL(leaf, i); - if (context->flags & ATTR_KERNOVAL) { - ASSERT(context->flags & ATTR_KERNAMELS); - context->count += namesp->attr_namelen + - (int)name_loc->namelen + 1; - } else { - retval = xfs_attr_put_listent(context, namesp, - (char *)name_loc->nameval, - (int)name_loc->namelen, - be16_to_cpu(name_loc->valuelen)); - } + xfs_attr_leaf_name_local_t *name_loc = + XFS_ATTR_LEAF_NAME_LOCAL(leaf, i); + + retval = context->put_listent(context, + namesp, + (char *)name_loc->nameval, + (int)name_loc->namelen, + be16_to_cpu(name_loc->valuelen), + (char *)&name_loc->nameval[name_loc->namelen]); + if (retval) + return retval; } else { - name_rmt = XFS_ATTR_LEAF_NAME_REMOTE(leaf, i); - if (context->flags & ATTR_KERNOVAL) { - ASSERT(context->flags & ATTR_KERNAMELS); - context->count += namesp->attr_namelen + - (int)name_rmt->namelen + 1; - } else { - retval = xfs_attr_put_listent(context, namesp, - (char *)name_rmt->name, - (int)name_rmt->namelen, - be32_to_cpu(name_rmt->valuelen)); + xfs_attr_leaf_name_remote_t *name_rmt = + XFS_ATTR_LEAF_NAME_REMOTE(leaf, i); + + int valuelen = be32_to_cpu(name_rmt->valuelen); + + if (context->put_value) { + xfs_da_args_t args; + + memset((char *)&args, 0, sizeof(args)); + args.dp = context->dp; + args.whichfork = XFS_ATTR_FORK; + args.valuelen = valuelen; + args.value = kmem_alloc(valuelen, KM_SLEEP); + args.rmtblkno = be32_to_cpu(name_rmt->valueblk); + args.rmtblkcnt = XFS_B_TO_FSB(args.dp->i_mount, valuelen); + retval = xfs_attr_rmtval_get(&args); + if (retval) + return retval; + retval = context->put_listent(context, + namesp, + (char *)name_rmt->name, + (int)name_rmt->namelen, + valuelen, + (char*)args.value); + kmem_free(args.value, valuelen); } + else { + retval = context->put_listent(context, + namesp, + (char *)name_rmt->name, + (int)name_rmt->namelen, + valuelen, + NULL); + } + if (retval) + return retval; } - if (retval == 0) { - cursor->offset++; - } + if (context->seen_enough) + break; + cursor->offset++; } xfs_attr_trace_l_cl("blk end", context, leaf); return(retval); } -#define ATTR_ENTBASESIZE /* minimum bytes used by an attr */ \ - (((struct attrlist_ent *) 0)->a_name - (char *) 0) -#define ATTR_ENTSIZE(namelen) /* actual bytes used by an attr */ \ - ((ATTR_ENTBASESIZE + (namelen) + 1 + sizeof(u_int32_t)-1) \ - & ~(sizeof(u_int32_t)-1)) - -/* - * Format an attribute and copy it out to the user's buffer. - * Take care to check values and protect against them changing later, - * we may be reading them directly out of a user buffer. - */ -/*ARGSUSED*/ -STATIC int -xfs_attr_put_listent(xfs_attr_list_context_t *context, - attrnames_t *namesp, char *name, int namelen, int valuelen) -{ - attrlist_ent_t *aep; - int arraytop; - - ASSERT(!(context->flags & ATTR_KERNOVAL)); - if (context->flags & ATTR_KERNAMELS) { - char *offset; - - ASSERT(context->count >= 0); - - arraytop = context->count + namesp->attr_namelen + namelen + 1; - if (arraytop > context->firstu) { - context->count = -1; /* insufficient space */ - return(1); - } - offset = (char *)context->alist + context->count; - strncpy(offset, namesp->attr_name, namesp->attr_namelen); - offset += namesp->attr_namelen; - strncpy(offset, name, namelen); /* real name */ - offset += namelen; - *offset = '\0'; - context->count += namesp->attr_namelen + namelen + 1; - return(0); - } - - ASSERT(context->count >= 0); - ASSERT(context->count < (ATTR_MAX_VALUELEN/8)); - ASSERT(context->firstu >= sizeof(*context->alist)); - ASSERT(context->firstu <= context->bufsize); - - arraytop = sizeof(*context->alist) + - context->count * sizeof(context->alist->al_offset[0]); - context->firstu -= ATTR_ENTSIZE(namelen); - if (context->firstu < arraytop) { - xfs_attr_trace_l_c("buffer full", context); - context->alist->al_more = 1; - return(1); - } - - aep = (attrlist_ent_t *)&(((char *)context->alist)[ context->firstu ]); - aep->a_valuelen = valuelen; - memcpy(aep->a_name, name, namelen); - aep->a_name[ namelen ] = 0; - context->alist->al_offset[ context->count++ ] = context->firstu; - context->alist->al_count = context->count; - xfs_attr_trace_l_c("add", context); - return(0); -} /*======================================================================== * Manage the INCOMPLETE flag in a leaf entry diff --git a/fs/xfs/xfs_attr_leaf.h b/fs/xfs/xfs_attr_leaf.h index 51c3ee156b2..040f732ce1e 100644 --- a/fs/xfs/xfs_attr_leaf.h +++ b/fs/xfs/xfs_attr_leaf.h @@ -130,6 +130,19 @@ typedef struct xfs_attr_leafblock { #define XFS_ATTR_INCOMPLETE (1 << XFS_ATTR_INCOMPLETE_BIT) /* + * Conversion macros for converting namespace bits from argument flags + * to ondisk flags. + */ +#define XFS_ATTR_NSP_ARGS_MASK (ATTR_ROOT | ATTR_SECURE) +#define XFS_ATTR_NSP_ONDISK_MASK (XFS_ATTR_ROOT | XFS_ATTR_SECURE) +#define XFS_ATTR_NSP_ONDISK(flags) ((flags) & XFS_ATTR_NSP_ONDISK_MASK) +#define XFS_ATTR_NSP_ARGS(flags) ((flags) & XFS_ATTR_NSP_ARGS_MASK) +#define XFS_ATTR_NSP_ARGS_TO_ONDISK(x) (((x) & ATTR_ROOT ? XFS_ATTR_ROOT : 0) |\ + ((x) & ATTR_SECURE ? XFS_ATTR_SECURE : 0)) +#define XFS_ATTR_NSP_ONDISK_TO_ARGS(x) (((x) & XFS_ATTR_ROOT ? ATTR_ROOT : 0) |\ + ((x) & XFS_ATTR_SECURE ? ATTR_SECURE : 0)) + +/* * Alignment for namelist and valuelist entries (since they are mixed * there can be only one alignment value) */ @@ -196,16 +209,26 @@ static inline int xfs_attr_leaf_entsize_local_max(int bsize) * Structure used to pass context around among the routines. *========================================================================*/ + +struct xfs_attr_list_context; + +typedef int (*put_listent_func_t)(struct xfs_attr_list_context *, struct attrnames *, + char *, int, int, char *); + typedef struct xfs_attr_list_context { - struct xfs_inode *dp; /* inode */ - struct attrlist_cursor_kern *cursor;/* position in list */ - struct attrlist *alist; /* output buffer */ - int count; /* num used entries */ - int dupcnt; /* count dup hashvals seen */ - int bufsize;/* total buffer size */ - int firstu; /* first used byte in buffer */ - int flags; /* from VOP call */ - int resynch;/* T/F: resynch with cursor */ + struct xfs_inode *dp; /* inode */ + struct attrlist_cursor_kern *cursor; /* position in list */ + struct attrlist *alist; /* output buffer */ + int seen_enough; /* T/F: seen enough of list? */ + int count; /* num used entries */ + int dupcnt; /* count dup hashvals seen */ + int bufsize; /* total buffer size */ + int firstu; /* first used byte in buffer */ + int flags; /* from VOP call */ + int resynch; /* T/F: resynch with cursor */ + int put_value; /* T/F: need value for listent */ + put_listent_func_t put_listent; /* list output fmt function */ + int index; /* index into output buffer */ } xfs_attr_list_context_t; /* diff --git a/fs/xfs/xfs_behavior.c b/fs/xfs/xfs_behavior.c index f4fe3715a80..0dc17219d41 100644 --- a/fs/xfs/xfs_behavior.c +++ b/fs/xfs/xfs_behavior.c @@ -110,26 +110,6 @@ bhv_remove_not_first(bhv_head_t *bhp, bhv_desc_t *bdp) } /* - * Look for a specific ops vector on the specified behavior chain. - * Return the associated behavior descriptor. Or NULL, if not found. - */ -bhv_desc_t * -bhv_lookup(bhv_head_t *bhp, void *ops) -{ - bhv_desc_t *curdesc; - - for (curdesc = bhp->bh_first; - curdesc != NULL; - curdesc = curdesc->bd_next) { - - if (curdesc->bd_ops == ops) - return curdesc; - } - - return NULL; -} - -/* * Looks for the first behavior within a specified range of positions. * Return the associated behavior descriptor. Or NULL, if none found. */ diff --git a/fs/xfs/xfs_behavior.h b/fs/xfs/xfs_behavior.h index 1d8ff103201..e7ca1fed955 100644 --- a/fs/xfs/xfs_behavior.h +++ b/fs/xfs/xfs_behavior.h @@ -78,15 +78,12 @@ * */ -struct bhv_head_lock; - /* * Behavior head. Head of the chain of behaviors. * Contained within each virtualized object data structure. */ typedef struct bhv_head { struct bhv_desc *bh_first; /* first behavior in chain */ - struct bhv_head_lock *bh_lockp; /* pointer to lock info struct */ } bhv_head_t; /* @@ -179,12 +176,10 @@ extern void bhv_insert_initial(bhv_head_t *, bhv_desc_t *); * Behavior module prototypes. */ extern void bhv_remove_not_first(bhv_head_t *bhp, bhv_desc_t *bdp); -extern bhv_desc_t * bhv_lookup(bhv_head_t *bhp, void *ops); extern bhv_desc_t * bhv_lookup_range(bhv_head_t *bhp, int low, int high); extern bhv_desc_t * bhv_base(bhv_head_t *bhp); /* No bhv locking on Linux */ -#define bhv_lookup_unlocked bhv_lookup #define bhv_base_unlocked bhv_base #endif /* __XFS_BEHAVIOR_H__ */ diff --git a/fs/xfs/xfs_bmap.c b/fs/xfs/xfs_bmap.c index 3a613753906..5b050c06795 100644 --- a/fs/xfs/xfs_bmap.c +++ b/fs/xfs/xfs_bmap.c @@ -2999,7 +2999,7 @@ xfs_bmap_btree_to_extents( int error; /* error return value */ xfs_ifork_t *ifp; /* inode fork data */ xfs_mount_t *mp; /* mount point structure */ - xfs_bmbt_ptr_t *pp; /* ptr to block address */ + __be64 *pp; /* ptr to block address */ xfs_bmbt_block_t *rblock;/* root btree block */ ifp = XFS_IFORK_PTR(ip, whichfork); @@ -3011,12 +3011,12 @@ xfs_bmap_btree_to_extents( ASSERT(XFS_BMAP_BROOT_MAXRECS(ifp->if_broot_bytes) == 1); mp = ip->i_mount; pp = XFS_BMAP_BROOT_PTR_ADDR(rblock, 1, ifp->if_broot_bytes); + cbno = be64_to_cpu(*pp); *logflagsp = 0; #ifdef DEBUG - if ((error = xfs_btree_check_lptr(cur, INT_GET(*pp, ARCH_CONVERT), 1))) + if ((error = xfs_btree_check_lptr(cur, cbno, 1))) return error; #endif - cbno = INT_GET(*pp, ARCH_CONVERT); if ((error = xfs_btree_read_bufl(mp, tp, cbno, 0, &cbp, XFS_BMAP_BTREE_REF))) return error; @@ -3512,9 +3512,9 @@ xfs_bmap_extents_to_btree( */ kp = XFS_BMAP_KEY_IADDR(block, 1, cur); arp = XFS_BMAP_REC_IADDR(ablock, 1, cur); - INT_SET(kp->br_startoff, ARCH_CONVERT, xfs_bmbt_disk_get_startoff(arp)); + kp->br_startoff = cpu_to_be64(xfs_bmbt_disk_get_startoff(arp)); pp = XFS_BMAP_PTR_IADDR(block, 1, cur); - INT_SET(*pp, ARCH_CONVERT, args.fsbno); + *pp = cpu_to_be64(args.fsbno); /* * Do all this logging at the end so that * the root is at the right level. @@ -3705,7 +3705,7 @@ STATIC xfs_bmbt_rec_t * /* pointer to found extent entry */ xfs_bmap_search_extents( xfs_inode_t *ip, /* incore inode pointer */ xfs_fileoff_t bno, /* block number searched for */ - int whichfork, /* data or attr fork */ + int fork, /* data or attr fork */ int *eofp, /* out: end of file found */ xfs_extnum_t *lastxp, /* out: last extent index */ xfs_bmbt_irec_t *gotp, /* out: extent entry found */ @@ -3713,25 +3713,28 @@ xfs_bmap_search_extents( { xfs_ifork_t *ifp; /* inode fork pointer */ xfs_bmbt_rec_t *ep; /* extent record pointer */ - int rt; /* realtime flag */ XFS_STATS_INC(xs_look_exlist); - ifp = XFS_IFORK_PTR(ip, whichfork); + ifp = XFS_IFORK_PTR(ip, fork); ep = xfs_bmap_search_multi_extents(ifp, bno, eofp, lastxp, gotp, prevp); - rt = (whichfork == XFS_DATA_FORK) && XFS_IS_REALTIME_INODE(ip); - if (unlikely(!rt && !gotp->br_startblock && (*lastxp != NULLEXTNUM))) { - cmn_err(CE_PANIC,"Access to block zero: fs: <%s> inode: %lld " - "start_block : %llx start_off : %llx blkcnt : %llx " - "extent-state : %x \n", - (ip->i_mount)->m_fsname, (long long)ip->i_ino, + if (unlikely(!(gotp->br_startblock) && (*lastxp != NULLEXTNUM) && + !(XFS_IS_REALTIME_INODE(ip) && fork == XFS_DATA_FORK))) { + xfs_cmn_err(XFS_PTAG_FSBLOCK_ZERO, CE_ALERT, ip->i_mount, + "Access to block zero in inode %llu " + "start_block: %llx start_off: %llx " + "blkcnt: %llx extent-state: %x lastx: %x\n", + (unsigned long long)ip->i_ino, (unsigned long long)gotp->br_startblock, (unsigned long long)gotp->br_startoff, (unsigned long long)gotp->br_blockcount, - gotp->br_state); - } - return ep; + gotp->br_state, *lastxp); + *lastxp = NULLEXTNUM; + *eofp = 1; + return NULL; + } + return ep; } @@ -4494,7 +4497,7 @@ xfs_bmap_read_extents( xfs_ifork_t *ifp; /* fork structure */ int level; /* btree level, for checking */ xfs_mount_t *mp; /* file system mount structure */ - xfs_bmbt_ptr_t *pp; /* pointer to block address */ + __be64 *pp; /* pointer to block address */ /* REFERENCED */ xfs_extnum_t room; /* number of entries there's room for */ @@ -4510,10 +4513,10 @@ xfs_bmap_read_extents( level = be16_to_cpu(block->bb_level); ASSERT(level > 0); pp = XFS_BMAP_BROOT_PTR_ADDR(block, 1, ifp->if_broot_bytes); - ASSERT(INT_GET(*pp, ARCH_CONVERT) != NULLDFSBNO); - ASSERT(XFS_FSB_TO_AGNO(mp, INT_GET(*pp, ARCH_CONVERT)) < mp->m_sb.sb_agcount); - ASSERT(XFS_FSB_TO_AGBNO(mp, INT_GET(*pp, ARCH_CONVERT)) < mp->m_sb.sb_agblocks); - bno = INT_GET(*pp, ARCH_CONVERT); + bno = be64_to_cpu(*pp); + ASSERT(bno != NULLDFSBNO); + ASSERT(XFS_FSB_TO_AGNO(mp, bno) < mp->m_sb.sb_agcount); + ASSERT(XFS_FSB_TO_AGBNO(mp, bno) < mp->m_sb.sb_agblocks); /* * Go down the tree until leaf level is reached, following the first * pointer (leftmost) at each level. @@ -4530,10 +4533,8 @@ xfs_bmap_read_extents( break; pp = XFS_BTREE_PTR_ADDR(mp->m_sb.sb_blocksize, xfs_bmbt, block, 1, mp->m_bmap_dmxr[1]); - XFS_WANT_CORRUPTED_GOTO( - XFS_FSB_SANITY_CHECK(mp, INT_GET(*pp, ARCH_CONVERT)), - error0); - bno = INT_GET(*pp, ARCH_CONVERT); + bno = be64_to_cpu(*pp); + XFS_WANT_CORRUPTED_GOTO(XFS_FSB_SANITY_CHECK(mp, bno), error0); xfs_trans_brelse(tp, bp); } /* @@ -4993,7 +4994,7 @@ xfs_bmapi( bma.firstblock = *firstblock; bma.alen = alen; bma.off = aoff; - bma.conv = (flags & XFS_BMAPI_CONVERT); + bma.conv = !!(flags & XFS_BMAPI_CONVERT); bma.wasdel = wasdelay; bma.minlen = minlen; bma.low = flist->xbf_low; @@ -6141,7 +6142,7 @@ xfs_check_block( short sz) { int i, j, dmxr; - xfs_bmbt_ptr_t *pp, *thispa; /* pointer to block address */ + __be64 *pp, *thispa; /* pointer to block address */ xfs_bmbt_key_t *prevp, *keyp; ASSERT(be16_to_cpu(block->bb_level) > 0); @@ -6179,11 +6180,10 @@ xfs_check_block( thispa = XFS_BTREE_PTR_ADDR(mp->m_sb.sb_blocksize, xfs_bmbt, block, j, dmxr); } - if (INT_GET(*thispa, ARCH_CONVERT) == - INT_GET(*pp, ARCH_CONVERT)) { + if (*thispa == *pp) { cmn_err(CE_WARN, "%s: thispa(%d) == pp(%d) %Ld", __FUNCTION__, j, i, - INT_GET(*thispa, ARCH_CONVERT)); + (unsigned long long)be64_to_cpu(*thispa)); panic("%s: ptrs are equal in node\n", __FUNCTION__); } @@ -6210,7 +6210,7 @@ xfs_bmap_check_leaf_extents( xfs_ifork_t *ifp; /* fork structure */ int level; /* btree level, for checking */ xfs_mount_t *mp; /* file system mount structure */ - xfs_bmbt_ptr_t *pp; /* pointer to block address */ + __be64 *pp; /* pointer to block address */ xfs_bmbt_rec_t *ep; /* pointer to current extent */ xfs_bmbt_rec_t *lastp; /* pointer to previous extent */ xfs_bmbt_rec_t *nextp; /* pointer to next extent */ @@ -6231,10 +6231,12 @@ xfs_bmap_check_leaf_extents( ASSERT(level > 0); xfs_check_block(block, mp, 1, ifp->if_broot_bytes); pp = XFS_BMAP_BROOT_PTR_ADDR(block, 1, ifp->if_broot_bytes); - ASSERT(INT_GET(*pp, ARCH_CONVERT) != NULLDFSBNO); - ASSERT(XFS_FSB_TO_AGNO(mp, INT_GET(*pp, ARCH_CONVERT)) < mp->m_sb.sb_agcount); - ASSERT(XFS_FSB_TO_AGBNO(mp, INT_GET(*pp, ARCH_CONVERT)) < mp->m_sb.sb_agblocks); - bno = INT_GET(*pp, ARCH_CONVERT); + bno = be64_to_cpu(*pp); + + ASSERT(bno != NULLDFSBNO); + ASSERT(XFS_FSB_TO_AGNO(mp, bno) < mp->m_sb.sb_agcount); + ASSERT(XFS_FSB_TO_AGBNO(mp, bno) < mp->m_sb.sb_agblocks); + /* * Go down the tree until leaf level is reached, following the first * pointer (leftmost) at each level. @@ -6265,8 +6267,8 @@ xfs_bmap_check_leaf_extents( xfs_check_block(block, mp, 0, 0); pp = XFS_BTREE_PTR_ADDR(mp->m_sb.sb_blocksize, xfs_bmbt, block, 1, mp->m_bmap_dmxr[1]); - XFS_WANT_CORRUPTED_GOTO(XFS_FSB_SANITY_CHECK(mp, INT_GET(*pp, ARCH_CONVERT)), error0); - bno = INT_GET(*pp, ARCH_CONVERT); + bno = be64_to_cpu(*pp); + XFS_WANT_CORRUPTED_GOTO(XFS_FSB_SANITY_CHECK(mp, bno), error0); if (bp_release) { bp_release = 0; xfs_trans_brelse(NULL, bp); @@ -6372,7 +6374,7 @@ xfs_bmap_count_blocks( xfs_ifork_t *ifp; /* fork structure */ int level; /* btree level, for checking */ xfs_mount_t *mp; /* file system mount structure */ - xfs_bmbt_ptr_t *pp; /* pointer to block address */ + __be64 *pp; /* pointer to block address */ bno = NULLFSBLOCK; mp = ip->i_mount; @@ -6395,10 +6397,10 @@ xfs_bmap_count_blocks( level = be16_to_cpu(block->bb_level); ASSERT(level > 0); pp = XFS_BMAP_BROOT_PTR_ADDR(block, 1, ifp->if_broot_bytes); - ASSERT(INT_GET(*pp, ARCH_CONVERT) != NULLDFSBNO); - ASSERT(XFS_FSB_TO_AGNO(mp, INT_GET(*pp, ARCH_CONVERT)) < mp->m_sb.sb_agcount); - ASSERT(XFS_FSB_TO_AGBNO(mp, INT_GET(*pp, ARCH_CONVERT)) < mp->m_sb.sb_agblocks); - bno = INT_GET(*pp, ARCH_CONVERT); + bno = be64_to_cpu(*pp); + ASSERT(bno != NULLDFSBNO); + ASSERT(XFS_FSB_TO_AGNO(mp, bno) < mp->m_sb.sb_agcount); + ASSERT(XFS_FSB_TO_AGBNO(mp, bno) < mp->m_sb.sb_agblocks); if (unlikely(xfs_bmap_count_tree(mp, tp, ifp, bno, level, count) < 0)) { XFS_ERROR_REPORT("xfs_bmap_count_blocks(2)", XFS_ERRLEVEL_LOW, @@ -6425,7 +6427,7 @@ xfs_bmap_count_tree( int error; xfs_buf_t *bp, *nbp; int level = levelin; - xfs_bmbt_ptr_t *pp; + __be64 *pp; xfs_fsblock_t bno = blockno; xfs_fsblock_t nextbno; xfs_bmbt_block_t *block, *nextblock; @@ -6452,7 +6454,7 @@ xfs_bmap_count_tree( /* Dive to the next level */ pp = XFS_BTREE_PTR_ADDR(mp->m_sb.sb_blocksize, xfs_bmbt, block, 1, mp->m_bmap_dmxr[1]); - bno = INT_GET(*pp, ARCH_CONVERT); + bno = be64_to_cpu(*pp); if (unlikely((error = xfs_bmap_count_tree(mp, tp, ifp, bno, level, count)) < 0)) { xfs_trans_brelse(tp, bp); diff --git a/fs/xfs/xfs_bmap_btree.c b/fs/xfs/xfs_bmap_btree.c index 18fb7385d71..a7b835bf870 100644 --- a/fs/xfs/xfs_bmap_btree.c +++ b/fs/xfs/xfs_bmap_btree.c @@ -58,7 +58,7 @@ STATIC void xfs_bmbt_log_ptrs(xfs_btree_cur_t *, xfs_buf_t *, int, int); STATIC int xfs_bmbt_lshift(xfs_btree_cur_t *, int, int *); STATIC int xfs_bmbt_rshift(xfs_btree_cur_t *, int, int *); STATIC int xfs_bmbt_split(xfs_btree_cur_t *, int, xfs_fsblock_t *, - xfs_bmbt_key_t *, xfs_btree_cur_t **, int *); + __uint64_t *, xfs_btree_cur_t **, int *); STATIC int xfs_bmbt_updkey(xfs_btree_cur_t *, xfs_bmbt_key_t *, int); @@ -192,16 +192,11 @@ xfs_bmbt_trace_argifk( xfs_btree_cur_t *cur, int i, xfs_fsblock_t f, - xfs_bmbt_key_t *k, + xfs_dfiloff_t o, int line) { - xfs_dfsbno_t d; - xfs_dfiloff_t o; - - d = (xfs_dfsbno_t)f; - o = INT_GET(k->br_startoff, ARCH_CONVERT); xfs_bmbt_trace_enter(func, cur, ARGS, XFS_BMBT_KTRACE_ARGIFK, line, - i, d >> 32, (int)d, o >> 32, + i, (xfs_dfsbno_t)f >> 32, (int)f, o >> 32, (int)o, 0, 0, 0, 0, 0, 0); } @@ -248,7 +243,7 @@ xfs_bmbt_trace_argik( { xfs_dfiloff_t o; - o = INT_GET(k->br_startoff, ARCH_CONVERT); + o = be64_to_cpu(k->br_startoff); xfs_bmbt_trace_enter(func, cur, ARGS, XFS_BMBT_KTRACE_ARGIFK, line, i, o >> 32, (int)o, 0, 0, 0, 0, 0, @@ -286,8 +281,8 @@ xfs_bmbt_trace_cursor( xfs_bmbt_trace_argfffi(fname, c, o, b, i, j, __LINE__) #define XFS_BMBT_TRACE_ARGI(c,i) \ xfs_bmbt_trace_argi(fname, c, i, __LINE__) -#define XFS_BMBT_TRACE_ARGIFK(c,i,f,k) \ - xfs_bmbt_trace_argifk(fname, c, i, f, k, __LINE__) +#define XFS_BMBT_TRACE_ARGIFK(c,i,f,s) \ + xfs_bmbt_trace_argifk(fname, c, i, f, s, __LINE__) #define XFS_BMBT_TRACE_ARGIFR(c,i,f,r) \ xfs_bmbt_trace_argifr(fname, c, i, f, r, __LINE__) #define XFS_BMBT_TRACE_ARGIK(c,i,k) \ @@ -299,7 +294,7 @@ xfs_bmbt_trace_cursor( #define XFS_BMBT_TRACE_ARGBII(c,b,i,j) #define XFS_BMBT_TRACE_ARGFFFI(c,o,b,i,j) #define XFS_BMBT_TRACE_ARGI(c,i) -#define XFS_BMBT_TRACE_ARGIFK(c,i,f,k) +#define XFS_BMBT_TRACE_ARGIFK(c,i,f,s) #define XFS_BMBT_TRACE_ARGIFR(c,i,f,r) #define XFS_BMBT_TRACE_ARGIK(c,i,k) #define XFS_BMBT_TRACE_CURSOR(c,s) @@ -357,7 +352,7 @@ xfs_bmbt_delrec( XFS_BMBT_TRACE_CURSOR(cur, ENTRY); XFS_BMBT_TRACE_ARGI(cur, level); ptr = cur->bc_ptrs[level]; - tcur = (xfs_btree_cur_t *)0; + tcur = NULL; if (ptr == 0) { XFS_BMBT_TRACE_CURSOR(cur, EXIT); *stat = 0; @@ -382,7 +377,7 @@ xfs_bmbt_delrec( pp = XFS_BMAP_PTR_IADDR(block, 1, cur); #ifdef DEBUG for (i = ptr; i < numrecs; i++) { - if ((error = xfs_btree_check_lptr(cur, INT_GET(pp[i], ARCH_CONVERT), level))) { + if ((error = xfs_btree_check_lptr_disk(cur, pp[i], level))) { XFS_BMBT_TRACE_CURSOR(cur, ERROR); goto error0; } @@ -404,7 +399,8 @@ xfs_bmbt_delrec( xfs_bmbt_log_recs(cur, bp, ptr, numrecs - 1); } if (ptr == 1) { - INT_SET(key.br_startoff, ARCH_CONVERT, xfs_bmbt_disk_get_startoff(rp)); + key.br_startoff = + cpu_to_be64(xfs_bmbt_disk_get_startoff(rp)); kp = &key; } } @@ -621,7 +617,7 @@ xfs_bmbt_delrec( rpp = XFS_BMAP_PTR_IADDR(right, 1, cur); #ifdef DEBUG for (i = 0; i < numrrecs; i++) { - if ((error = xfs_btree_check_lptr(cur, INT_GET(rpp[i], ARCH_CONVERT), level))) { + if ((error = xfs_btree_check_lptr_disk(cur, rpp[i], level))) { XFS_BMBT_TRACE_CURSOR(cur, ERROR); goto error0; } @@ -748,7 +744,7 @@ xfs_bmbt_insrec( int logflags; /* inode logging flags */ xfs_fsblock_t nbno; /* new block number */ struct xfs_btree_cur *ncur; /* new btree cursor */ - xfs_bmbt_key_t nkey; /* new btree key value */ + __uint64_t startoff; /* new btree key value */ xfs_bmbt_rec_t nrec; /* new record count */ int optr; /* old key/record index */ xfs_bmbt_ptr_t *pp; /* pointer to bmap block addr */ @@ -759,9 +755,8 @@ xfs_bmbt_insrec( ASSERT(level < cur->bc_nlevels); XFS_BMBT_TRACE_CURSOR(cur, ENTRY); XFS_BMBT_TRACE_ARGIFR(cur, level, *bnop, recp); - ncur = (xfs_btree_cur_t *)0; - INT_SET(key.br_startoff, ARCH_CONVERT, - xfs_bmbt_disk_get_startoff(recp)); + ncur = NULL; + key.br_startoff = cpu_to_be64(xfs_bmbt_disk_get_startoff(recp)); optr = ptr = cur->bc_ptrs[level]; if (ptr == 0) { XFS_BMBT_TRACE_CURSOR(cur, EXIT); @@ -820,7 +815,7 @@ xfs_bmbt_insrec( optr = ptr = cur->bc_ptrs[level]; } else { if ((error = xfs_bmbt_split(cur, level, - &nbno, &nkey, &ncur, + &nbno, &startoff, &ncur, &i))) { XFS_BMBT_TRACE_CURSOR(cur, ERROR); @@ -840,7 +835,7 @@ xfs_bmbt_insrec( #endif ptr = cur->bc_ptrs[level]; xfs_bmbt_disk_set_allf(&nrec, - nkey.br_startoff, 0, 0, + startoff, 0, 0, XFS_EXT_NORM); } else { XFS_BMBT_TRACE_CURSOR(cur, @@ -858,7 +853,7 @@ xfs_bmbt_insrec( pp = XFS_BMAP_PTR_IADDR(block, 1, cur); #ifdef DEBUG for (i = numrecs; i >= ptr; i--) { - if ((error = xfs_btree_check_lptr(cur, INT_GET(pp[i - 1], ARCH_CONVERT), + if ((error = xfs_btree_check_lptr_disk(cur, pp[i - 1], level))) { XFS_BMBT_TRACE_CURSOR(cur, ERROR); return error; @@ -870,14 +865,13 @@ xfs_bmbt_insrec( memmove(&pp[ptr], &pp[ptr - 1], /* INT_: direct copy */ (numrecs - ptr + 1) * sizeof(*pp)); #ifdef DEBUG - if ((error = xfs_btree_check_lptr(cur, (xfs_bmbt_ptr_t)*bnop, - level))) { + if ((error = xfs_btree_check_lptr(cur, *bnop, level))) { XFS_BMBT_TRACE_CURSOR(cur, ERROR); return error; } #endif kp[ptr - 1] = key; - INT_SET(pp[ptr - 1], ARCH_CONVERT, *bnop); + pp[ptr - 1] = cpu_to_be64(*bnop); numrecs++; block->bb_numrecs = cpu_to_be16(numrecs); xfs_bmbt_log_keys(cur, bp, ptr, numrecs); @@ -988,7 +982,7 @@ xfs_bmbt_killroot( cpp = XFS_BMAP_PTR_IADDR(cblock, 1, cur); #ifdef DEBUG for (i = 0; i < be16_to_cpu(cblock->bb_numrecs); i++) { - if ((error = xfs_btree_check_lptr(cur, INT_GET(cpp[i], ARCH_CONVERT), level - 1))) { + if ((error = xfs_btree_check_lptr_disk(cur, cpp[i], level - 1))) { XFS_BMBT_TRACE_CURSOR(cur, ERROR); return error; } @@ -1132,7 +1126,7 @@ xfs_bmbt_lookup( d = XFS_FSB_TO_DADDR(mp, fsbno); bp = cur->bc_bufs[level]; if (bp && XFS_BUF_ADDR(bp) != d) - bp = (xfs_buf_t *)0; + bp = NULL; if (!bp) { if ((error = xfs_btree_read_bufl(mp, tp, fsbno, 0, &bp, XFS_BMAP_BTREE_REF))) { @@ -1170,7 +1164,7 @@ xfs_bmbt_lookup( keyno = (low + high) >> 1; if (level > 0) { kkp = kkbase + keyno - 1; - startoff = INT_GET(kkp->br_startoff, ARCH_CONVERT); + startoff = be64_to_cpu(kkp->br_startoff); } else { krp = krbase + keyno - 1; startoff = xfs_bmbt_disk_get_startoff(krp); @@ -1189,13 +1183,13 @@ xfs_bmbt_lookup( if (diff > 0 && --keyno < 1) keyno = 1; pp = XFS_BMAP_PTR_IADDR(block, keyno, cur); + fsbno = be64_to_cpu(*pp); #ifdef DEBUG - if ((error = xfs_btree_check_lptr(cur, INT_GET(*pp, ARCH_CONVERT), level))) { + if ((error = xfs_btree_check_lptr(cur, fsbno, level))) { XFS_BMBT_TRACE_CURSOR(cur, ERROR); return error; } #endif - fsbno = INT_GET(*pp, ARCH_CONVERT); cur->bc_ptrs[level] = keyno; } } @@ -1313,7 +1307,7 @@ xfs_bmbt_lshift( lpp = XFS_BMAP_PTR_IADDR(left, lrecs, cur); rpp = XFS_BMAP_PTR_IADDR(right, 1, cur); #ifdef DEBUG - if ((error = xfs_btree_check_lptr(cur, INT_GET(*rpp, ARCH_CONVERT), level))) { + if ((error = xfs_btree_check_lptr_disk(cur, *rpp, level))) { XFS_BMBT_TRACE_CURSOR(cur, ERROR); return error; } @@ -1340,7 +1334,7 @@ xfs_bmbt_lshift( if (level > 0) { #ifdef DEBUG for (i = 0; i < rrecs; i++) { - if ((error = xfs_btree_check_lptr(cur, INT_GET(rpp[i + 1], ARCH_CONVERT), + if ((error = xfs_btree_check_lptr_disk(cur, rpp[i + 1], level))) { XFS_BMBT_TRACE_CURSOR(cur, ERROR); return error; @@ -1354,8 +1348,7 @@ xfs_bmbt_lshift( } else { memmove(rrp, rrp + 1, rrecs * sizeof(*rrp)); xfs_bmbt_log_recs(cur, rbp, 1, rrecs); - INT_SET(key.br_startoff, ARCH_CONVERT, - xfs_bmbt_disk_get_startoff(rrp)); + key.br_startoff = cpu_to_be64(xfs_bmbt_disk_get_startoff(rrp)); rkp = &key; } if ((error = xfs_bmbt_updkey(cur, rkp, level + 1))) { @@ -1445,7 +1438,7 @@ xfs_bmbt_rshift( rpp = XFS_BMAP_PTR_IADDR(right, 1, cur); #ifdef DEBUG for (i = be16_to_cpu(right->bb_numrecs) - 1; i >= 0; i--) { - if ((error = xfs_btree_check_lptr(cur, INT_GET(rpp[i], ARCH_CONVERT), level))) { + if ((error = xfs_btree_check_lptr_disk(cur, rpp[i], level))) { XFS_BMBT_TRACE_CURSOR(cur, ERROR); return error; } @@ -1454,7 +1447,7 @@ xfs_bmbt_rshift( memmove(rkp + 1, rkp, be16_to_cpu(right->bb_numrecs) * sizeof(*rkp)); memmove(rpp + 1, rpp, be16_to_cpu(right->bb_numrecs) * sizeof(*rpp)); #ifdef DEBUG - if ((error = xfs_btree_check_lptr(cur, INT_GET(*lpp, ARCH_CONVERT), level))) { + if ((error = xfs_btree_check_lptr_disk(cur, *lpp, level))) { XFS_BMBT_TRACE_CURSOR(cur, ERROR); return error; } @@ -1469,8 +1462,7 @@ xfs_bmbt_rshift( memmove(rrp + 1, rrp, be16_to_cpu(right->bb_numrecs) * sizeof(*rrp)); *rrp = *lrp; xfs_bmbt_log_recs(cur, rbp, 1, be16_to_cpu(right->bb_numrecs) + 1); - INT_SET(key.br_startoff, ARCH_CONVERT, - xfs_bmbt_disk_get_startoff(rrp)); + key.br_startoff = cpu_to_be64(xfs_bmbt_disk_get_startoff(rrp)); rkp = &key; } be16_add(&left->bb_numrecs, -1); @@ -1535,7 +1527,7 @@ xfs_bmbt_split( xfs_btree_cur_t *cur, int level, xfs_fsblock_t *bnop, - xfs_bmbt_key_t *keyp, + __uint64_t *startoff, xfs_btree_cur_t **curp, int *stat) /* success/failure */ { @@ -1560,7 +1552,7 @@ xfs_bmbt_split( xfs_bmbt_rec_t *rrp; /* right record pointer */ XFS_BMBT_TRACE_CURSOR(cur, ENTRY); - XFS_BMBT_TRACE_ARGIFK(cur, level, *bnop, keyp); + XFS_BMBT_TRACE_ARGIFK(cur, level, *bnop, *startoff); args.tp = cur->bc_tp; args.mp = cur->bc_mp; lbp = cur->bc_bufs[level]; @@ -1619,7 +1611,7 @@ xfs_bmbt_split( rpp = XFS_BMAP_PTR_IADDR(right, 1, cur); #ifdef DEBUG for (i = 0; i < be16_to_cpu(right->bb_numrecs); i++) { - if ((error = xfs_btree_check_lptr(cur, INT_GET(lpp[i], ARCH_CONVERT), level))) { + if ((error = xfs_btree_check_lptr_disk(cur, lpp[i], level))) { XFS_BMBT_TRACE_CURSOR(cur, ERROR); return error; } @@ -1629,13 +1621,13 @@ xfs_bmbt_split( memcpy(rpp, lpp, be16_to_cpu(right->bb_numrecs) * sizeof(*rpp)); xfs_bmbt_log_keys(cur, rbp, 1, be16_to_cpu(right->bb_numrecs)); xfs_bmbt_log_ptrs(cur, rbp, 1, be16_to_cpu(right->bb_numrecs)); - keyp->br_startoff = INT_GET(rkp->br_startoff, ARCH_CONVERT); + *startoff = be64_to_cpu(rkp->br_startoff); } else { lrp = XFS_BMAP_REC_IADDR(left, i, cur); rrp = XFS_BMAP_REC_IADDR(right, 1, cur); memcpy(rrp, lrp, be16_to_cpu(right->bb_numrecs) * sizeof(*rrp)); xfs_bmbt_log_recs(cur, rbp, 1, be16_to_cpu(right->bb_numrecs)); - keyp->br_startoff = xfs_bmbt_disk_get_startoff(rrp); + *startoff = xfs_bmbt_disk_get_startoff(rrp); } be16_add(&left->bb_numrecs, -(be16_to_cpu(right->bb_numrecs))); right->bb_rightsib = left->bb_rightsib; @@ -1728,9 +1720,9 @@ xfs_bmdr_to_bmbt( { int dmxr; xfs_bmbt_key_t *fkp; - xfs_bmbt_ptr_t *fpp; + __be64 *fpp; xfs_bmbt_key_t *tkp; - xfs_bmbt_ptr_t *tpp; + __be64 *tpp; rblock->bb_magic = cpu_to_be32(XFS_BMAP_MAGIC); rblock->bb_level = dblock->bb_level; @@ -1745,7 +1737,7 @@ xfs_bmdr_to_bmbt( tpp = XFS_BMAP_BROOT_PTR_ADDR(rblock, 1, rblocklen); dmxr = be16_to_cpu(dblock->bb_numrecs); memcpy(tkp, fkp, sizeof(*fkp) * dmxr); - memcpy(tpp, fpp, sizeof(*fpp) * dmxr); /* INT_: direct copy */ + memcpy(tpp, fpp, sizeof(*fpp) * dmxr); } /* @@ -1805,7 +1797,7 @@ xfs_bmbt_decrement( tp = cur->bc_tp; mp = cur->bc_mp; for (block = xfs_bmbt_get_block(cur, lev, &bp); lev > level; ) { - fsbno = INT_GET(*XFS_BMAP_PTR_IADDR(block, cur->bc_ptrs[lev], cur), ARCH_CONVERT); + fsbno = be64_to_cpu(*XFS_BMAP_PTR_IADDR(block, cur->bc_ptrs[lev], cur)); if ((error = xfs_btree_read_bufl(mp, tp, fsbno, 0, &bp, XFS_BMAP_BTREE_REF))) { XFS_BMBT_TRACE_CURSOR(cur, ERROR); @@ -2135,7 +2127,7 @@ xfs_bmbt_increment( tp = cur->bc_tp; mp = cur->bc_mp; for (block = xfs_bmbt_get_block(cur, lev, &bp); lev > level; ) { - fsbno = INT_GET(*XFS_BMAP_PTR_IADDR(block, cur->bc_ptrs[lev], cur), ARCH_CONVERT); + fsbno = be64_to_cpu(*XFS_BMAP_PTR_IADDR(block, cur->bc_ptrs[lev], cur)); if ((error = xfs_btree_read_bufl(mp, tp, fsbno, 0, &bp, XFS_BMAP_BTREE_REF))) { XFS_BMBT_TRACE_CURSOR(cur, ERROR); @@ -2178,7 +2170,7 @@ xfs_bmbt_insert( level = 0; nbno = NULLFSBLOCK; xfs_bmbt_disk_set_all(&nrec, &cur->bc_rec.b); - ncur = (xfs_btree_cur_t *)0; + ncur = NULL; pcur = cur; do { if ((error = xfs_bmbt_insrec(pcur, level++, &nbno, &nrec, &ncur, @@ -2205,7 +2197,7 @@ xfs_bmbt_insert( } if (ncur) { pcur = ncur; - ncur = (xfs_btree_cur_t *)0; + ncur = NULL; } } while (nbno != NULLFSBLOCK); XFS_BMBT_TRACE_CURSOR(cur, EXIT); @@ -2356,12 +2348,12 @@ xfs_bmbt_newroot( args.firstblock = args.fsbno; if (args.fsbno == NULLFSBLOCK) { #ifdef DEBUG - if ((error = xfs_btree_check_lptr(cur, INT_GET(*pp, ARCH_CONVERT), level))) { + if ((error = xfs_btree_check_lptr_disk(cur, *pp, level))) { XFS_BMBT_TRACE_CURSOR(cur, ERROR); return error; } #endif - args.fsbno = INT_GET(*pp, ARCH_CONVERT); + args.fsbno = be64_to_cpu(*pp); args.type = XFS_ALLOCTYPE_START_BNO; } else args.type = XFS_ALLOCTYPE_NEAR_BNO; @@ -2393,7 +2385,7 @@ xfs_bmbt_newroot( cpp = XFS_BMAP_PTR_IADDR(cblock, 1, cur); #ifdef DEBUG for (i = 0; i < be16_to_cpu(cblock->bb_numrecs); i++) { - if ((error = xfs_btree_check_lptr(cur, INT_GET(pp[i], ARCH_CONVERT), level))) { + if ((error = xfs_btree_check_lptr_disk(cur, pp[i], level))) { XFS_BMBT_TRACE_CURSOR(cur, ERROR); return error; } @@ -2401,13 +2393,12 @@ xfs_bmbt_newroot( #endif memcpy(cpp, pp, be16_to_cpu(cblock->bb_numrecs) * sizeof(*pp)); #ifdef DEBUG - if ((error = xfs_btree_check_lptr(cur, (xfs_bmbt_ptr_t)args.fsbno, - level))) { + if ((error = xfs_btree_check_lptr(cur, args.fsbno, level))) { XFS_BMBT_TRACE_CURSOR(cur, ERROR); return error; } #endif - INT_SET(*pp, ARCH_CONVERT, args.fsbno); + *pp = cpu_to_be64(args.fsbno); xfs_iroot_realloc(cur->bc_private.b.ip, 1 - be16_to_cpu(cblock->bb_numrecs), cur->bc_private.b.whichfork); xfs_btree_setbuf(cur, level, bp); @@ -2681,9 +2672,9 @@ xfs_bmbt_to_bmdr( { int dmxr; xfs_bmbt_key_t *fkp; - xfs_bmbt_ptr_t *fpp; + __be64 *fpp; xfs_bmbt_key_t *tkp; - xfs_bmbt_ptr_t *tpp; + __be64 *tpp; ASSERT(be32_to_cpu(rblock->bb_magic) == XFS_BMAP_MAGIC); ASSERT(be64_to_cpu(rblock->bb_leftsib) == NULLDFSBNO); @@ -2698,7 +2689,7 @@ xfs_bmbt_to_bmdr( tpp = XFS_BTREE_PTR_ADDR(dblocklen, xfs_bmdr, dblock, 1, dmxr); dmxr = be16_to_cpu(dblock->bb_numrecs); memcpy(tkp, fkp, sizeof(*fkp) * dmxr); - memcpy(tpp, fpp, sizeof(*fpp) * dmxr); /* INT_: direct copy */ + memcpy(tpp, fpp, sizeof(*fpp) * dmxr); } /* @@ -2740,7 +2731,7 @@ xfs_bmbt_update( XFS_BMBT_TRACE_CURSOR(cur, EXIT); return 0; } - INT_SET(key.br_startoff, ARCH_CONVERT, off); + key.br_startoff = cpu_to_be64(off); if ((error = xfs_bmbt_updkey(cur, &key, 1))) { XFS_BMBT_TRACE_CURSOR(cur, ERROR); return error; diff --git a/fs/xfs/xfs_bmap_btree.h b/fs/xfs/xfs_bmap_btree.h index 6478cfa0e53..49539de9525 100644 --- a/fs/xfs/xfs_bmap_btree.h +++ b/fs/xfs/xfs_bmap_btree.h @@ -163,13 +163,14 @@ typedef struct xfs_bmbt_irec /* * Key structure for non-leaf levels of the tree. */ -typedef struct xfs_bmbt_key -{ - xfs_dfiloff_t br_startoff; /* starting file offset */ +typedef struct xfs_bmbt_key { + __be64 br_startoff; /* starting file offset */ } xfs_bmbt_key_t, xfs_bmdr_key_t; -typedef xfs_dfsbno_t xfs_bmbt_ptr_t, xfs_bmdr_ptr_t; /* btree pointer type */ - /* btree block header type */ +/* btree pointer type */ +typedef __be64 xfs_bmbt_ptr_t, xfs_bmdr_ptr_t; + +/* btree block header type */ typedef struct xfs_btree_lblock xfs_bmbt_block_t; #define XFS_BUF_TO_BMBT_BLOCK(bp) ((xfs_bmbt_block_t *)XFS_BUF_PTR(bp)) diff --git a/fs/xfs/xfs_btree.c b/fs/xfs/xfs_btree.c index ee2255bd656..aeb87ca69fc 100644 --- a/fs/xfs/xfs_btree.c +++ b/fs/xfs/xfs_btree.c @@ -161,7 +161,7 @@ xfs_btree_check_key( k1 = ak1; k2 = ak2; - ASSERT(INT_GET(k1->br_startoff, ARCH_CONVERT) < INT_GET(k2->br_startoff, ARCH_CONVERT)); + ASSERT(be64_to_cpu(k1->br_startoff) < be64_to_cpu(k2->br_startoff)); break; } case XFS_BTNUM_INO: { @@ -170,7 +170,7 @@ xfs_btree_check_key( k1 = ak1; k2 = ak2; - ASSERT(INT_GET(k1->ir_startino, ARCH_CONVERT) < INT_GET(k2->ir_startino, ARCH_CONVERT)); + ASSERT(be32_to_cpu(k1->ir_startino) < be32_to_cpu(k2->ir_startino)); break; } default: @@ -285,8 +285,8 @@ xfs_btree_check_rec( r1 = ar1; r2 = ar2; - ASSERT(INT_GET(r1->ir_startino, ARCH_CONVERT) + XFS_INODES_PER_CHUNK <= - INT_GET(r2->ir_startino, ARCH_CONVERT)); + ASSERT(be32_to_cpu(r1->ir_startino) + XFS_INODES_PER_CHUNK <= + be32_to_cpu(r2->ir_startino)); break; } default: diff --git a/fs/xfs/xfs_btree.h b/fs/xfs/xfs_btree.h index 44f1bd98064..892b06c5426 100644 --- a/fs/xfs/xfs_btree.h +++ b/fs/xfs/xfs_btree.h @@ -145,7 +145,7 @@ typedef struct xfs_btree_cur union { xfs_alloc_rec_incore_t a; xfs_bmbt_irec_t b; - xfs_inobt_rec_t i; + xfs_inobt_rec_incore_t i; } bc_rec; /* current insert/search record value */ struct xfs_buf *bc_bufs[XFS_BTREE_MAXLEVELS]; /* buf ptr per level */ int bc_ptrs[XFS_BTREE_MAXLEVELS]; /* key/record # */ @@ -243,6 +243,9 @@ xfs_btree_check_lptr( xfs_dfsbno_t ptr, /* btree block disk address */ int level); /* btree block level */ +#define xfs_btree_check_lptr_disk(cur, ptr, level) \ + xfs_btree_check_lptr(cur, be64_to_cpu(ptr), level) + /* * Checking routine: check that short form block header is ok. */ diff --git a/fs/xfs/xfs_buf_item.c b/fs/xfs/xfs_buf_item.c index a4aa53974f7..7a55c248ea7 100644 --- a/fs/xfs/xfs_buf_item.c +++ b/fs/xfs/xfs_buf_item.c @@ -234,7 +234,6 @@ xfs_buf_item_format( ASSERT((bip->bli_flags & XFS_BLI_LOGGED) || (bip->bli_flags & XFS_BLI_STALE)); bp = bip->bli_buf; - ASSERT(XFS_BUF_BP_ISMAPPED(bp)); vecp = log_vector; /* @@ -628,25 +627,6 @@ xfs_buf_item_committed( } /* - * This is called when the transaction holding the buffer is aborted. - * Just behave as if the transaction had been cancelled. If we're shutting down - * and have aborted this transaction, we'll trap this buffer when it tries to - * get written out. - */ -STATIC void -xfs_buf_item_abort( - xfs_buf_log_item_t *bip) -{ - xfs_buf_t *bp; - - bp = bip->bli_buf; - xfs_buftrace("XFS_ABORT", bp); - XFS_BUF_SUPER_STALE(bp); - xfs_buf_item_unlock(bip); - return; -} - -/* * This is called to asynchronously write the buffer associated with this * buf log item out to disk. The buffer will already have been locked by * a successful call to xfs_buf_item_trylock(). If the buffer still has @@ -693,7 +673,6 @@ STATIC struct xfs_item_ops xfs_buf_item_ops = { .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_abort = (void(*)(xfs_log_item_t*))xfs_buf_item_abort, .iop_pushbuf = NULL, .iop_committing = (void(*)(xfs_log_item_t*, xfs_lsn_t)) xfs_buf_item_committing @@ -901,7 +880,6 @@ xfs_buf_item_relse( XFS_BUF_SET_FSPRIVATE(bp, bip->bli_item.li_bio_list); if ((XFS_BUF_FSPRIVATE(bp, void *) == NULL) && (XFS_BUF_IODONE_FUNC(bp) != NULL)) { - ASSERT((XFS_BUF_ISUNINITIAL(bp)) == 0); XFS_BUF_CLR_IODONE_FUNC(bp); } diff --git a/fs/xfs/xfs_da_btree.c b/fs/xfs/xfs_da_btree.c index 32ab61d17ac..a68bc1f1a31 100644 --- a/fs/xfs/xfs_da_btree.c +++ b/fs/xfs/xfs_da_btree.c @@ -1054,7 +1054,7 @@ xfs_da_node_lookup_int(xfs_da_state_t *state, int *result) xfs_da_node_entry_t *btree; xfs_dablk_t blkno; int probe, span, max, error, retval; - xfs_dahash_t hashval; + xfs_dahash_t hashval, btreehashval; xfs_da_args_t *args; args = state->args; @@ -1079,30 +1079,32 @@ xfs_da_node_lookup_int(xfs_da_state_t *state, int *result) return(error); } curr = blk->bp->data; - ASSERT(be16_to_cpu(curr->magic) == XFS_DA_NODE_MAGIC || - be16_to_cpu(curr->magic) == XFS_DIR2_LEAFN_MAGIC || - be16_to_cpu(curr->magic) == XFS_ATTR_LEAF_MAGIC); + blk->magic = be16_to_cpu(curr->magic); + ASSERT(blk->magic == XFS_DA_NODE_MAGIC || + blk->magic == XFS_DIR2_LEAFN_MAGIC || + blk->magic == XFS_ATTR_LEAF_MAGIC); /* * Search an intermediate node for a match. */ - blk->magic = be16_to_cpu(curr->magic); if (blk->magic == XFS_DA_NODE_MAGIC) { node = blk->bp->data; - blk->hashval = be32_to_cpu(node->btree[be16_to_cpu(node->hdr.count)-1].hashval); + max = be16_to_cpu(node->hdr.count); + btreehashval = node->btree[max-1].hashval; + blk->hashval = be32_to_cpu(btreehashval); /* * Binary search. (note: small blocks will skip loop) */ - max = be16_to_cpu(node->hdr.count); probe = span = max / 2; hashval = args->hashval; for (btree = &node->btree[probe]; span > 4; btree = &node->btree[probe]) { span /= 2; - if (be32_to_cpu(btree->hashval) < hashval) + btreehashval = be32_to_cpu(btree->hashval); + if (btreehashval < hashval) probe += span; - else if (be32_to_cpu(btree->hashval) > hashval) + else if (btreehashval > hashval) probe -= span; else break; @@ -1133,10 +1135,10 @@ xfs_da_node_lookup_int(xfs_da_state_t *state, int *result) blk->index = probe; blkno = be32_to_cpu(btree->before); } - } else if (be16_to_cpu(curr->magic) == XFS_ATTR_LEAF_MAGIC) { + } else if (blk->magic == XFS_ATTR_LEAF_MAGIC) { blk->hashval = xfs_attr_leaf_lasthash(blk->bp, NULL); break; - } else if (be16_to_cpu(curr->magic) == XFS_DIR2_LEAFN_MAGIC) { + } else if (blk->magic == XFS_DIR2_LEAFN_MAGIC) { blk->hashval = xfs_dir2_leafn_lasthash(blk->bp, NULL); break; } @@ -1152,11 +1154,13 @@ xfs_da_node_lookup_int(xfs_da_state_t *state, int *result) if (blk->magic == XFS_DIR2_LEAFN_MAGIC) { retval = xfs_dir2_leafn_lookup_int(blk->bp, args, &blk->index, state); - } - else if (blk->magic == XFS_ATTR_LEAF_MAGIC) { + } else if (blk->magic == XFS_ATTR_LEAF_MAGIC) { retval = xfs_attr_leaf_lookup_int(blk->bp, args); blk->index = args->index; args->blkno = blk->blkno; + } else { + ASSERT(0); + return XFS_ERROR(EFSCORRUPTED); } if (((retval == ENOENT) || (retval == ENOATTR)) && (blk->hashval == args->hashval)) { @@ -1166,8 +1170,7 @@ xfs_da_node_lookup_int(xfs_da_state_t *state, int *result) return(error); if (retval == 0) { continue; - } - else if (blk->magic == XFS_ATTR_LEAF_MAGIC) { + } else if (blk->magic == XFS_ATTR_LEAF_MAGIC) { /* path_shift() gives ENOENT */ retval = XFS_ERROR(ENOATTR); } diff --git a/fs/xfs/xfs_error.h b/fs/xfs/xfs_error.h index bc43163456e..0893e16b7d8 100644 --- a/fs/xfs/xfs_error.h +++ b/fs/xfs/xfs_error.h @@ -18,14 +18,6 @@ #ifndef __XFS_ERROR_H__ #define __XFS_ERROR_H__ -#define XFS_ERECOVER 1 /* Failure to recover log */ -#define XFS_ELOGSTAT 2 /* Failure to stat log in user space */ -#define XFS_ENOLOGSPACE 3 /* Reservation too large */ -#define XFS_ENOTSUP 4 /* Operation not supported */ -#define XFS_ENOLSN 5 /* Can't find the lsn you asked for */ -#define XFS_ENOTFOUND 6 -#define XFS_ENOTXFS 7 /* Not XFS filesystem */ - #ifdef DEBUG #define XFS_ERROR_NTRAP 10 extern int xfs_etrap[XFS_ERROR_NTRAP]; @@ -175,6 +167,7 @@ extern int xfs_errortag_clearall_umount(int64_t fsid, char *fsname, int loud); #define XFS_PTAG_SHUTDOWN_CORRUPT 0x00000010 #define XFS_PTAG_SHUTDOWN_IOERROR 0x00000020 #define XFS_PTAG_SHUTDOWN_LOGERROR 0x00000040 +#define XFS_PTAG_FSBLOCK_ZERO 0x00000080 struct xfs_mount; /* PRINTFLIKE4 */ diff --git a/fs/xfs/xfs_extfree_item.c b/fs/xfs/xfs_extfree_item.c index 6cf6d8769b9..6dba78199fa 100644 --- a/fs/xfs/xfs_extfree_item.c +++ b/fs/xfs/xfs_extfree_item.c @@ -33,9 +33,6 @@ kmem_zone_t *xfs_efi_zone; kmem_zone_t *xfs_efd_zone; STATIC void xfs_efi_item_unlock(xfs_efi_log_item_t *); -STATIC void xfs_efi_item_abort(xfs_efi_log_item_t *); -STATIC void xfs_efd_item_abort(xfs_efd_log_item_t *); - void xfs_efi_item_free(xfs_efi_log_item_t *efip) @@ -184,7 +181,7 @@ STATIC void xfs_efi_item_unlock(xfs_efi_log_item_t *efip) { if (efip->efi_item.li_flags & XFS_LI_ABORTED) - xfs_efi_item_abort(efip); + xfs_efi_item_free(efip); return; } @@ -202,18 +199,6 @@ xfs_efi_item_committed(xfs_efi_log_item_t *efip, xfs_lsn_t lsn) } /* - * This is called when the transaction logging the EFI is aborted. - * Free up the EFI and return. No need to clean up the slot for - * the item in the transaction. That was done by the unpin code - * which is called prior to this routine in the abort/fs-shutdown path. - */ -STATIC void -xfs_efi_item_abort(xfs_efi_log_item_t *efip) -{ - xfs_efi_item_free(efip); -} - -/* * There isn't much you can do to push on an efi item. It is simply * stuck waiting for all of its corresponding efd items to be * committed to disk. @@ -255,7 +240,6 @@ STATIC struct xfs_item_ops xfs_efi_item_ops = { .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_abort = (void(*)(xfs_log_item_t*))xfs_efi_item_abort, .iop_pushbuf = NULL, .iop_committing = (void(*)(xfs_log_item_t*, xfs_lsn_t)) xfs_efi_item_committing @@ -386,33 +370,6 @@ xfs_efi_release(xfs_efi_log_item_t *efip, } } -/* - * This is called when the transaction that should be committing the - * EFD corresponding to the given EFI is aborted. The committed and - * canceled flags are used to coordinate the freeing of the EFI and - * the references by the transaction that committed it. - */ -STATIC void -xfs_efi_cancel( - xfs_efi_log_item_t *efip) -{ - xfs_mount_t *mp; - SPLDECL(s); - - mp = efip->efi_item.li_mountp; - AIL_LOCK(mp, s); - if (efip->efi_flags & XFS_EFI_COMMITTED) { - /* - * xfs_trans_delete_ail() drops the AIL lock. - */ - xfs_trans_delete_ail(mp, (xfs_log_item_t *)efip, s); - xfs_efi_item_free(efip); - } else { - efip->efi_flags |= XFS_EFI_CANCELED; - AIL_UNLOCK(mp, s); - } -} - STATIC void xfs_efd_item_free(xfs_efd_log_item_t *efdp) { @@ -514,7 +471,7 @@ STATIC void xfs_efd_item_unlock(xfs_efd_log_item_t *efdp) { if (efdp->efd_item.li_flags & XFS_LI_ABORTED) - xfs_efd_item_abort(efdp); + xfs_efd_item_free(efdp); return; } @@ -541,27 +498,6 @@ xfs_efd_item_committed(xfs_efd_log_item_t *efdp, xfs_lsn_t lsn) } /* - * The transaction of which this EFD is a part has been aborted. - * Inform its companion EFI of this fact and then clean up after - * ourselves. No need to clean up the slot for the item in the - * transaction. That was done by the unpin code which is called - * prior to this routine in the abort/fs-shutdown path. - */ -STATIC void -xfs_efd_item_abort(xfs_efd_log_item_t *efdp) -{ - /* - * 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. So don't - * reference the EFI at all in that case. - */ - if ((efdp->efd_item.li_flags & XFS_LI_ABORTED) == 0) - xfs_efi_cancel(efdp->efd_efip); - - xfs_efd_item_free(efdp); -} - -/* * 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. */ @@ -602,7 +538,6 @@ STATIC struct xfs_item_ops xfs_efd_item_ops = { .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_abort = (void(*)(xfs_log_item_t*))xfs_efd_item_abort, .iop_pushbuf = NULL, .iop_committing = (void(*)(xfs_log_item_t*, xfs_lsn_t)) xfs_efd_item_committing diff --git a/fs/xfs/xfs_extfree_item.h b/fs/xfs/xfs_extfree_item.h index 0ea45edaab0..2f049f63e85 100644 --- a/fs/xfs/xfs_extfree_item.h +++ b/fs/xfs/xfs_extfree_item.h @@ -33,14 +33,16 @@ typedef struct xfs_extent { * conversion routine. */ +#ifndef HAVE_FORMAT32 typedef struct xfs_extent_32 { - xfs_dfsbno_t ext_start; - xfs_extlen_t ext_len; + __uint64_t ext_start; + __uint32_t ext_len; } __attribute__((packed)) xfs_extent_32_t; +#endif typedef struct xfs_extent_64 { - xfs_dfsbno_t ext_start; - xfs_extlen_t ext_len; + __uint64_t ext_start; + __uint32_t ext_len; __uint32_t ext_pad; } xfs_extent_64_t; @@ -50,25 +52,27 @@ typedef struct xfs_extent_64 { * size is given by efi_nextents. */ typedef struct xfs_efi_log_format { - unsigned short efi_type; /* efi log item type */ - unsigned short efi_size; /* size of this item */ - uint efi_nextents; /* # extents to free */ + __uint16_t efi_type; /* efi log item type */ + __uint16_t efi_size; /* size of this item */ + __uint32_t efi_nextents; /* # extents to free */ __uint64_t efi_id; /* efi identifier */ xfs_extent_t efi_extents[1]; /* array of extents to free */ } xfs_efi_log_format_t; +#ifndef HAVE_FORMAT32 typedef struct xfs_efi_log_format_32 { - unsigned short efi_type; /* efi log item type */ - unsigned short efi_size; /* size of this item */ - uint efi_nextents; /* # extents to free */ + __uint16_t efi_type; /* efi log item type */ + __uint16_t efi_size; /* size of this item */ + __uint32_t efi_nextents; /* # extents to free */ __uint64_t efi_id; /* efi identifier */ xfs_extent_32_t efi_extents[1]; /* array of extents to free */ } __attribute__((packed)) xfs_efi_log_format_32_t; +#endif typedef struct xfs_efi_log_format_64 { - unsigned short efi_type; /* efi log item type */ - unsigned short efi_size; /* size of this item */ - uint efi_nextents; /* # extents to free */ + __uint16_t efi_type; /* efi log item type */ + __uint16_t efi_size; /* size of this item */ + __uint32_t efi_nextents; /* # extents to free */ __uint64_t efi_id; /* efi identifier */ xfs_extent_64_t efi_extents[1]; /* array of extents to free */ } xfs_efi_log_format_64_t; @@ -79,25 +83,27 @@ typedef struct xfs_efi_log_format_64 { * size is given by efd_nextents; */ typedef struct xfs_efd_log_format { - unsigned short efd_type; /* efd log item type */ - unsigned short efd_size; /* size of this item */ - uint efd_nextents; /* # of extents freed */ + __uint16_t efd_type; /* efd log item type */ + __uint16_t efd_size; /* size of this item */ + __uint32_t efd_nextents; /* # of extents freed */ __uint64_t efd_efi_id; /* id of corresponding efi */ xfs_extent_t efd_extents[1]; /* array of extents freed */ } xfs_efd_log_format_t; +#ifndef HAVE_FORMAT32 typedef struct xfs_efd_log_format_32 { - unsigned short efd_type; /* efd log item type */ - unsigned short efd_size; /* size of this item */ - uint efd_nextents; /* # of extents freed */ + __uint16_t efd_type; /* efd log item type */ + __uint16_t efd_size; /* size of this item */ + __uint32_t efd_nextents; /* # of extents freed */ __uint64_t efd_efi_id; /* id of corresponding efi */ xfs_extent_32_t efd_extents[1]; /* array of extents freed */ } __attribute__((packed)) xfs_efd_log_format_32_t; +#endif typedef struct xfs_efd_log_format_64 { - unsigned short efd_type; /* efd log item type */ - unsigned short efd_size; /* size of this item */ - uint efd_nextents; /* # of extents freed */ + __uint16_t efd_type; /* efd log item type */ + __uint16_t efd_size; /* size of this item */ + __uint32_t efd_nextents; /* # of extents freed */ __uint64_t efd_efi_id; /* id of corresponding efi */ xfs_extent_64_t efd_extents[1]; /* array of extents freed */ } xfs_efd_log_format_64_t; diff --git a/fs/xfs/xfs_fs.h b/fs/xfs/xfs_fs.h index 0f0ad153595..1335449841c 100644 --- a/fs/xfs/xfs_fs.h +++ b/fs/xfs/xfs_fs.h @@ -22,8 +22,6 @@ * SGI's XFS filesystem's major stuff (constants, structures) */ -#define XFS_NAME "xfs" - /* * Direct I/O attribute record used with XFS_IOC_DIOINFO * d_miniosz is the min xfer size, xfer size multiple and file seek offset @@ -426,11 +424,7 @@ typedef struct xfs_handle { - (char *) &(handle)) \ + (handle).ha_fid.xfs_fid_len) -#define XFS_HANDLE_CMP(h1, h2) memcmp(h1, h2, sizeof(xfs_handle_t)) - -#define FSHSIZE sizeof(fsid_t) - -/* +/* * Flags for going down operation */ #define XFS_FSOP_GOING_FLAGS_DEFAULT 0x0 /* going down */ diff --git a/fs/xfs/xfs_fsops.c b/fs/xfs/xfs_fsops.c index 077629bab53..c064e72ada9 100644 --- a/fs/xfs/xfs_fsops.c +++ b/fs/xfs/xfs_fsops.c @@ -462,7 +462,7 @@ xfs_fs_counts( xfs_icsb_sync_counters_lazy(mp); s = XFS_SB_LOCK(mp); - cnt->freedata = mp->m_sb.sb_fdblocks; + cnt->freedata = mp->m_sb.sb_fdblocks - XFS_ALLOC_SET_ASIDE(mp); cnt->freertx = mp->m_sb.sb_frextents; cnt->freeino = mp->m_sb.sb_ifree; cnt->allocino = mp->m_sb.sb_icount; @@ -519,15 +519,19 @@ xfs_reserve_blocks( } mp->m_resblks = request; } else { + __int64_t free; + + free = mp->m_sb.sb_fdblocks - XFS_ALLOC_SET_ASIDE(mp); delta = request - mp->m_resblks; - lcounter = mp->m_sb.sb_fdblocks - delta; + lcounter = free - delta; if (lcounter < 0) { /* We can't satisfy the request, just get what we can */ - mp->m_resblks += mp->m_sb.sb_fdblocks; - mp->m_resblks_avail += mp->m_sb.sb_fdblocks; - mp->m_sb.sb_fdblocks = 0; + mp->m_resblks += free; + mp->m_resblks_avail += free; + mp->m_sb.sb_fdblocks = XFS_ALLOC_SET_ASIDE(mp); } else { - mp->m_sb.sb_fdblocks = lcounter; + mp->m_sb.sb_fdblocks = + lcounter + XFS_ALLOC_SET_ASIDE(mp); mp->m_resblks = request; mp->m_resblks_avail += delta; } diff --git a/fs/xfs/xfs_ialloc.c b/fs/xfs/xfs_ialloc.c index 33164a85aa9..a446e5a115c 100644 --- a/fs/xfs/xfs_ialloc.c +++ b/fs/xfs/xfs_ialloc.c @@ -458,7 +458,7 @@ nextag: */ if (XFS_FORCED_SHUTDOWN(mp)) { up_read(&mp->m_peraglock); - return (xfs_buf_t *)0; + return NULL; } agno++; if (agno >= agcount) @@ -466,7 +466,7 @@ nextag: if (agno == pagno) { if (flags == 0) { up_read(&mp->m_peraglock); - return (xfs_buf_t *)0; + return NULL; } flags = 0; } @@ -529,10 +529,10 @@ xfs_dialloc( int offset; /* index of inode in chunk */ xfs_agino_t pagino; /* parent's a.g. relative inode # */ xfs_agnumber_t pagno; /* parent's allocation group number */ - xfs_inobt_rec_t rec; /* inode allocation record */ + xfs_inobt_rec_incore_t rec; /* inode allocation record */ xfs_agnumber_t tagno; /* testing allocation group number */ xfs_btree_cur_t *tcur; /* temp cursor */ - xfs_inobt_rec_t trec; /* temp inode allocation record */ + xfs_inobt_rec_incore_t trec; /* temp inode allocation record */ if (*IO_agbp == NULL) { @@ -945,7 +945,7 @@ xfs_difree( int ilen; /* inodes in an inode cluster */ xfs_mount_t *mp; /* mount structure for filesystem */ int off; /* offset of inode in inode chunk */ - xfs_inobt_rec_t rec; /* btree record */ + xfs_inobt_rec_incore_t rec; /* btree record */ mp = tp->t_mountp; @@ -1195,6 +1195,7 @@ xfs_dilocate( "(0x%llx)", ino, XFS_AGINO_TO_INO(mp, agno, agino)); } + xfs_stack_trace(); #endif /* DEBUG */ return XFS_ERROR(EINVAL); } diff --git a/fs/xfs/xfs_ialloc_btree.c b/fs/xfs/xfs_ialloc_btree.c index 616eeeb6953..8cdeeaf8632 100644 --- a/fs/xfs/xfs_ialloc_btree.c +++ b/fs/xfs/xfs_ialloc_btree.c @@ -568,7 +568,7 @@ xfs_inobt_insrec( /* * Make a key out of the record data to be inserted, and save it. */ - key.ir_startino = recp->ir_startino; /* INT_: direct copy */ + key.ir_startino = recp->ir_startino; optr = ptr = cur->bc_ptrs[level]; /* * If we're off the left edge, return failure. @@ -600,7 +600,7 @@ xfs_inobt_insrec( } #endif nbno = NULLAGBLOCK; - ncur = (xfs_btree_cur_t *)0; + ncur = NULL; /* * If the block is full, we can't insert the new entry until we * make the block un-full. @@ -641,7 +641,7 @@ xfs_inobt_insrec( return error; #endif ptr = cur->bc_ptrs[level]; - nrec.ir_startino = nkey.ir_startino; /* INT_: direct copy */ + nrec.ir_startino = nkey.ir_startino; } else { /* * Otherwise the insert fails. @@ -681,7 +681,7 @@ xfs_inobt_insrec( if ((error = xfs_btree_check_sptr(cur, *bnop, level))) return error; #endif - kp[ptr - 1] = key; /* INT_: struct copy */ + kp[ptr - 1] = key; pp[ptr - 1] = cpu_to_be32(*bnop); numrecs++; block->bb_numrecs = cpu_to_be16(numrecs); @@ -698,7 +698,7 @@ xfs_inobt_insrec( * Now stuff the new record in, bump numrecs * and log the new data. */ - rp[ptr - 1] = *recp; /* INT_: struct copy */ + rp[ptr - 1] = *recp; numrecs++; block->bb_numrecs = cpu_to_be16(numrecs); xfs_inobt_log_recs(cur, bp, ptr, numrecs); @@ -731,7 +731,7 @@ xfs_inobt_insrec( */ *bnop = nbno; if (nbno != NULLAGBLOCK) { - *recp = nrec; /* INT_: struct copy */ + *recp = nrec; *curp = ncur; } *stat = 1; @@ -878,7 +878,7 @@ xfs_inobt_lookup( */ bp = cur->bc_bufs[level]; if (bp && XFS_BUF_ADDR(bp) != d) - bp = (xfs_buf_t *)0; + bp = NULL; if (!bp) { /* * Need to get a new buffer. Read it, then @@ -950,12 +950,12 @@ xfs_inobt_lookup( xfs_inobt_key_t *kkp; kkp = kkbase + keyno - 1; - startino = INT_GET(kkp->ir_startino, ARCH_CONVERT); + startino = be32_to_cpu(kkp->ir_startino); } else { xfs_inobt_rec_t *krp; krp = krbase + keyno - 1; - startino = INT_GET(krp->ir_startino, ARCH_CONVERT); + startino = be32_to_cpu(krp->ir_startino); } /* * Compute difference to get next direction. @@ -1117,7 +1117,7 @@ xfs_inobt_lshift( if ((error = xfs_btree_check_sptr(cur, be32_to_cpu(*rpp), level))) return error; #endif - *lpp = *rpp; /* INT_: no-change copy */ + *lpp = *rpp; xfs_inobt_log_ptrs(cur, lbp, nrec, nrec); } /* @@ -1160,7 +1160,7 @@ xfs_inobt_lshift( } else { memmove(rrp, rrp + 1, be16_to_cpu(right->bb_numrecs) * sizeof(*rrp)); xfs_inobt_log_recs(cur, rbp, 1, be16_to_cpu(right->bb_numrecs)); - key.ir_startino = rrp->ir_startino; /* INT_: direct copy */ + key.ir_startino = rrp->ir_startino; rkp = &key; } /* @@ -1297,13 +1297,13 @@ xfs_inobt_newroot( */ kp = XFS_INOBT_KEY_ADDR(new, 1, cur); if (be16_to_cpu(left->bb_level) > 0) { - kp[0] = *XFS_INOBT_KEY_ADDR(left, 1, cur); /* INT_: struct copy */ - kp[1] = *XFS_INOBT_KEY_ADDR(right, 1, cur); /* INT_: struct copy */ + kp[0] = *XFS_INOBT_KEY_ADDR(left, 1, cur); + kp[1] = *XFS_INOBT_KEY_ADDR(right, 1, cur); } else { rp = XFS_INOBT_REC_ADDR(left, 1, cur); - INT_COPY(kp[0].ir_startino, rp->ir_startino, ARCH_CONVERT); + kp[0].ir_startino = rp->ir_startino; rp = XFS_INOBT_REC_ADDR(right, 1, cur); - INT_COPY(kp[1].ir_startino, rp->ir_startino, ARCH_CONVERT); + kp[1].ir_startino = rp->ir_startino; } xfs_inobt_log_keys(cur, nbp, 1, 2); /* @@ -1410,8 +1410,8 @@ xfs_inobt_rshift( if ((error = xfs_btree_check_sptr(cur, be32_to_cpu(*lpp), level))) return error; #endif - *rkp = *lkp; /* INT_: no change copy */ - *rpp = *lpp; /* INT_: no change copy */ + *rkp = *lkp; + *rpp = *lpp; xfs_inobt_log_keys(cur, rbp, 1, be16_to_cpu(right->bb_numrecs) + 1); xfs_inobt_log_ptrs(cur, rbp, 1, be16_to_cpu(right->bb_numrecs) + 1); } else { @@ -1420,7 +1420,7 @@ xfs_inobt_rshift( memmove(rrp + 1, rrp, be16_to_cpu(right->bb_numrecs) * sizeof(*rrp)); *rrp = *lrp; xfs_inobt_log_recs(cur, rbp, 1, be16_to_cpu(right->bb_numrecs) + 1); - key.ir_startino = rrp->ir_startino; /* INT_: direct copy */ + key.ir_startino = rrp->ir_startino; rkp = &key; } /* @@ -1559,7 +1559,7 @@ xfs_inobt_split( rrp = XFS_INOBT_REC_ADDR(right, 1, cur); memcpy(rrp, lrp, be16_to_cpu(right->bb_numrecs) * sizeof(*rrp)); xfs_inobt_log_recs(cur, rbp, 1, be16_to_cpu(right->bb_numrecs)); - keyp->ir_startino = rrp->ir_startino; /* INT_: direct copy */ + keyp->ir_startino = rrp->ir_startino; } /* * Find the left block number by looking in the buffer. @@ -1813,9 +1813,9 @@ xfs_inobt_get_rec( * Point to the record and extract its data. */ rec = XFS_INOBT_REC_ADDR(block, ptr, cur); - *ino = INT_GET(rec->ir_startino, ARCH_CONVERT); - *fcnt = INT_GET(rec->ir_freecount, ARCH_CONVERT); - *free = INT_GET(rec->ir_free, ARCH_CONVERT); + *ino = be32_to_cpu(rec->ir_startino); + *fcnt = be32_to_cpu(rec->ir_freecount); + *free = be64_to_cpu(rec->ir_free); *stat = 1; return 0; } @@ -1930,10 +1930,10 @@ xfs_inobt_insert( level = 0; nbno = NULLAGBLOCK; - INT_SET(nrec.ir_startino, ARCH_CONVERT, cur->bc_rec.i.ir_startino); - INT_SET(nrec.ir_freecount, ARCH_CONVERT, cur->bc_rec.i.ir_freecount); - INT_SET(nrec.ir_free, ARCH_CONVERT, cur->bc_rec.i.ir_free); - ncur = (xfs_btree_cur_t *)0; + nrec.ir_startino = cpu_to_be32(cur->bc_rec.i.ir_startino); + nrec.ir_freecount = cpu_to_be32(cur->bc_rec.i.ir_freecount); + nrec.ir_free = cpu_to_be64(cur->bc_rec.i.ir_free); + ncur = NULL; pcur = cur; /* * Loop going up the tree, starting at the leaf level. @@ -1965,7 +1965,7 @@ xfs_inobt_insert( */ if (ncur) { pcur = ncur; - ncur = (xfs_btree_cur_t *)0; + ncur = NULL; } } while (nbno != NULLAGBLOCK); *stat = i; @@ -2060,9 +2060,9 @@ xfs_inobt_update( /* * Fill in the new contents and log them. */ - INT_SET(rp->ir_startino, ARCH_CONVERT, ino); - INT_SET(rp->ir_freecount, ARCH_CONVERT, fcnt); - INT_SET(rp->ir_free, ARCH_CONVERT, free); + rp->ir_startino = cpu_to_be32(ino); + rp->ir_freecount = cpu_to_be32(fcnt); + rp->ir_free = cpu_to_be64(free); xfs_inobt_log_recs(cur, bp, ptr, ptr); /* * Updating first record in leaf. Pass new key value up to our parent. @@ -2070,7 +2070,7 @@ xfs_inobt_update( if (ptr == 1) { xfs_inobt_key_t key; /* key containing [ino] */ - INT_SET(key.ir_startino, ARCH_CONVERT, ino); + key.ir_startino = cpu_to_be32(ino); if ((error = xfs_inobt_updkey(cur, &key, 1))) return error; } diff --git a/fs/xfs/xfs_ialloc_btree.h b/fs/xfs/xfs_ialloc_btree.h index ae3904cb1ee..2c0e49893ff 100644 --- a/fs/xfs/xfs_ialloc_btree.h +++ b/fs/xfs/xfs_ialloc_btree.h @@ -47,19 +47,24 @@ static inline xfs_inofree_t xfs_inobt_maskn(int i, int n) /* * Data record structure */ -typedef struct xfs_inobt_rec -{ +typedef struct xfs_inobt_rec { + __be32 ir_startino; /* starting inode number */ + __be32 ir_freecount; /* count of free inodes (set bits) */ + __be64 ir_free; /* free inode mask */ +} xfs_inobt_rec_t; + +typedef struct xfs_inobt_rec_incore { xfs_agino_t ir_startino; /* starting inode number */ __int32_t ir_freecount; /* count of free inodes (set bits) */ xfs_inofree_t ir_free; /* free inode mask */ -} xfs_inobt_rec_t; +} xfs_inobt_rec_incore_t; + /* * Key structure */ -typedef struct xfs_inobt_key -{ - xfs_agino_t ir_startino; /* starting inode number */ +typedef struct xfs_inobt_key { + __be32 ir_startino; /* starting inode number */ } xfs_inobt_key_t; /* btree pointer type */ @@ -77,7 +82,7 @@ typedef struct xfs_btree_sblock xfs_inobt_block_t; #define XFS_INOBT_IS_FREE(rp,i) \ (((rp)->ir_free & XFS_INOBT_MASK(i)) != 0) #define XFS_INOBT_IS_FREE_DISK(rp,i) \ - ((INT_GET((rp)->ir_free,ARCH_CONVERT) & XFS_INOBT_MASK(i)) != 0) + ((be64_to_cpu((rp)->ir_free) & XFS_INOBT_MASK(i)) != 0) #define XFS_INOBT_SET_FREE(rp,i) ((rp)->ir_free |= XFS_INOBT_MASK(i)) #define XFS_INOBT_CLR_FREE(rp,i) ((rp)->ir_free &= ~XFS_INOBT_MASK(i)) diff --git a/fs/xfs/xfs_iget.c b/fs/xfs/xfs_iget.c index 0724df7fabb..b73d216ecaf 100644 --- a/fs/xfs/xfs_iget.c +++ b/fs/xfs/xfs_iget.c @@ -50,7 +50,7 @@ void xfs_ihash_init(xfs_mount_t *mp) { __uint64_t icount; - uint i, flags = KM_SLEEP | KM_MAYFAIL; + uint i; if (!mp->m_ihsize) { icount = mp->m_maxicount ? mp->m_maxicount : @@ -61,14 +61,13 @@ xfs_ihash_init(xfs_mount_t *mp) (64 * NBPP) / sizeof(xfs_ihash_t)); } - while (!(mp->m_ihash = (xfs_ihash_t *)kmem_zalloc(mp->m_ihsize * - sizeof(xfs_ihash_t), flags))) { - if ((mp->m_ihsize >>= 1) <= NBPP) - flags = KM_SLEEP; - } - for (i = 0; i < mp->m_ihsize; i++) { + mp->m_ihash = kmem_zalloc_greedy(&mp->m_ihsize, + NBPC * sizeof(xfs_ihash_t), + mp->m_ihsize * sizeof(xfs_ihash_t), + KM_SLEEP | KM_MAYFAIL | KM_LARGE); + mp->m_ihsize /= sizeof(xfs_ihash_t); + for (i = 0; i < mp->m_ihsize; i++) rwlock_init(&(mp->m_ihash[i].ih_lock)); - } } /* @@ -77,7 +76,7 @@ xfs_ihash_init(xfs_mount_t *mp) void xfs_ihash_free(xfs_mount_t *mp) { - kmem_free(mp->m_ihash, mp->m_ihsize*sizeof(xfs_ihash_t)); + kmem_free(mp->m_ihash, mp->m_ihsize * sizeof(xfs_ihash_t)); mp->m_ihash = NULL; } @@ -95,7 +94,7 @@ xfs_chash_init(xfs_mount_t *mp) mp->m_chsize = min_t(uint, mp->m_chsize, mp->m_ihsize); mp->m_chash = (xfs_chash_t *)kmem_zalloc(mp->m_chsize * sizeof(xfs_chash_t), - KM_SLEEP); + KM_SLEEP | KM_LARGE); for (i = 0; i < mp->m_chsize; i++) { spinlock_init(&mp->m_chash[i].ch_lock,"xfshash"); } @@ -244,7 +243,9 @@ again: XFS_STATS_INC(xs_ig_found); + spin_lock(&ip->i_flags_lock); ip->i_flags &= ~XFS_IRECLAIMABLE; + spin_unlock(&ip->i_flags_lock); version = ih->ih_version; read_unlock(&ih->ih_lock); xfs_ihash_promote(ih, ip, version); @@ -290,15 +291,17 @@ again: finish_inode: if (ip->i_d.di_mode == 0) { - if (!(flags & IGET_CREATE)) + if (!(flags & XFS_IGET_CREATE)) return ENOENT; xfs_iocore_inode_reinit(ip); } - + if (lock_flags != 0) xfs_ilock(ip, lock_flags); + spin_lock(&ip->i_flags_lock); ip->i_flags &= ~XFS_ISTALE; + spin_unlock(&ip->i_flags_lock); vn_trace_exit(vp, "xfs_iget.found", (inst_t *)__return_address); @@ -320,21 +323,20 @@ finish_inode: * Read the disk inode attributes into a new inode structure and get * a new vnode for it. This should also initialize i_ino and i_mount. */ - error = xfs_iread(mp, tp, ino, &ip, bno); - if (error) { + error = xfs_iread(mp, tp, ino, &ip, bno, + (flags & XFS_IGET_BULKSTAT) ? XFS_IMAP_BULKSTAT : 0); + if (error) return error; - } vn_trace_exit(vp, "xfs_iget.alloc", (inst_t *)__return_address); xfs_inode_lock_init(ip, vp); xfs_iocore_inode_init(ip); - if (lock_flags != 0) { + if (lock_flags) xfs_ilock(ip, lock_flags); - } - - if ((ip->i_d.di_mode == 0) && !(flags & IGET_CREATE)) { + + if ((ip->i_d.di_mode == 0) && !(flags & XFS_IGET_CREATE)) { xfs_idestroy(ip); return ENOENT; } @@ -369,7 +371,9 @@ finish_inode: ih->ih_next = ip; ip->i_udquot = ip->i_gdquot = NULL; ih->ih_version++; + spin_lock(&ip->i_flags_lock); ip->i_flags |= XFS_INEW; + spin_unlock(&ip->i_flags_lock); write_unlock(&ih->ih_lock); @@ -548,7 +552,7 @@ xfs_inode_lock_init( mrlock_init(&ip->i_iolock, MRLOCK_BARRIER, "xfsio", vp->v_number); init_waitqueue_head(&ip->i_ipin_wait); atomic_set(&ip->i_pincount, 0); - init_sema(&ip->i_flock, 1, "xfsfino", vp->v_number); + initnsema(&ip->i_flock, 1, "xfsfino"); } /* diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c index 5fa0adb7e17..c27d7d495aa 100644 --- a/fs/xfs/xfs_inode.c +++ b/fs/xfs/xfs_inode.c @@ -334,10 +334,9 @@ xfs_itobp( #if !defined(__KERNEL__) ni = 0; #elif defined(DEBUG) - ni = (imap_flags & XFS_IMAP_BULKSTAT) ? 0 : - (BBTOB(imap.im_len) >> mp->m_sb.sb_inodelog); + ni = BBTOB(imap.im_len) >> mp->m_sb.sb_inodelog; #else /* usual case */ - ni = (imap_flags & XFS_IMAP_BULKSTAT) ? 0 : 1; + ni = 1; #endif for (i = 0; i < ni; i++) { @@ -348,11 +347,15 @@ xfs_itobp( (i << mp->m_sb.sb_inodelog)); di_ok = INT_GET(dip->di_core.di_magic, ARCH_CONVERT) == XFS_DINODE_MAGIC && XFS_DINODE_GOOD_VERSION(INT_GET(dip->di_core.di_version, ARCH_CONVERT)); - if (unlikely(XFS_TEST_ERROR(!di_ok, mp, XFS_ERRTAG_ITOBP_INOTOBP, - XFS_RANDOM_ITOBP_INOTOBP))) { + if (unlikely(XFS_TEST_ERROR(!di_ok, mp, + XFS_ERRTAG_ITOBP_INOTOBP, + XFS_RANDOM_ITOBP_INOTOBP))) { + if (imap_flags & XFS_IMAP_BULKSTAT) { + xfs_trans_brelse(tp, bp); + return XFS_ERROR(EINVAL); + } #ifdef DEBUG - if (!(imap_flags & XFS_IMAP_BULKSTAT)) - cmn_err(CE_ALERT, + cmn_err(CE_ALERT, "Device %s - bad inode magic/vsn " "daddr %lld #%d (magic=%x)", XFS_BUFTARG_NAME(mp->m_ddev_targp), @@ -851,7 +854,8 @@ xfs_iread( xfs_trans_t *tp, xfs_ino_t ino, xfs_inode_t **ipp, - xfs_daddr_t bno) + xfs_daddr_t bno, + uint imap_flags) { xfs_buf_t *bp; xfs_dinode_t *dip; @@ -863,6 +867,7 @@ xfs_iread( ip = kmem_zone_zalloc(xfs_inode_zone, KM_SLEEP); ip->i_ino = ino; ip->i_mount = mp; + spin_lock_init(&ip->i_flags_lock); /* * Get pointer's to the on-disk inode and the buffer containing it. @@ -871,7 +876,7 @@ xfs_iread( * return NULL as well. Set i_blkno to 0 so that xfs_itobp() will * know that this is a new incore inode. */ - error = xfs_itobp(mp, tp, ip, &dip, &bp, bno, 0); + error = xfs_itobp(mp, tp, ip, &dip, &bp, bno, imap_flags); if (error) { kmem_zone_free(xfs_inode_zone, ip); return error; @@ -1110,7 +1115,7 @@ xfs_ialloc( * to prevent others from looking at until we're done. */ error = xfs_trans_iget(tp->t_mountp, tp, ino, - IGET_CREATE, XFS_ILOCK_EXCL, &ip); + XFS_IGET_CREATE, XFS_ILOCK_EXCL, &ip); if (error != 0) { return error; } @@ -1961,9 +1966,9 @@ xfs_iunlink_remove( xfs_agino_t agino; xfs_agino_t next_agino; xfs_buf_t *last_ibp; - xfs_dinode_t *last_dip; + xfs_dinode_t *last_dip = NULL; short bucket_index; - int offset, last_offset; + int offset, last_offset = 0; int error; int agi_ok; @@ -2210,7 +2215,9 @@ xfs_ifree_cluster( if (ip == free_ip) { if (xfs_iflock_nowait(ip)) { + spin_lock(&ip->i_flags_lock); ip->i_flags |= XFS_ISTALE; + spin_unlock(&ip->i_flags_lock); if (xfs_inode_clean(ip)) { xfs_ifunlock(ip); @@ -2224,7 +2231,9 @@ xfs_ifree_cluster( if (xfs_ilock_nowait(ip, XFS_ILOCK_EXCL)) { if (xfs_iflock_nowait(ip)) { + spin_lock(&ip->i_flags_lock); ip->i_flags |= XFS_ISTALE; + spin_unlock(&ip->i_flags_lock); if (xfs_inode_clean(ip)) { xfs_ifunlock(ip); @@ -2254,7 +2263,9 @@ xfs_ifree_cluster( AIL_LOCK(mp,s); iip->ili_flush_lsn = iip->ili_item.li_lsn; AIL_UNLOCK(mp, s); + spin_lock(&iip->ili_inode->i_flags_lock); iip->ili_inode->i_flags |= XFS_ISTALE; + spin_unlock(&iip->ili_inode->i_flags_lock); pre_flushed++; } lip = lip->li_bio_list; @@ -2750,19 +2761,29 @@ xfs_iunpin( * call as the inode reclaim may be blocked waiting for * the inode to become unpinned. */ + struct inode *inode = NULL; + + spin_lock(&ip->i_flags_lock); if (!(ip->i_flags & (XFS_IRECLAIM|XFS_IRECLAIMABLE))) { bhv_vnode_t *vp = XFS_ITOV_NULL(ip); /* make sync come back and flush this inode */ if (vp) { - struct inode *inode = vn_to_inode(vp); + inode = vn_to_inode(vp); if (!(inode->i_state & - (I_NEW|I_FREEING|I_CLEAR))) - mark_inode_dirty_sync(inode); + (I_NEW|I_FREEING|I_CLEAR))) { + inode = igrab(inode); + if (inode) + mark_inode_dirty_sync(inode); + } else + inode = NULL; } } + spin_unlock(&ip->i_flags_lock); wake_up(&ip->i_ipin_wait); + if (inode) + iput(inode); } } diff --git a/fs/xfs/xfs_inode.h b/fs/xfs/xfs_inode.h index d10b76ed1e5..e96eb0835fe 100644 --- a/fs/xfs/xfs_inode.h +++ b/fs/xfs/xfs_inode.h @@ -267,6 +267,7 @@ typedef struct xfs_inode { sema_t i_flock; /* inode flush lock */ atomic_t i_pincount; /* inode pin count */ wait_queue_head_t i_ipin_wait; /* inode pinning wait queue */ + spinlock_t i_flags_lock; /* inode i_flags lock */ #ifdef HAVE_REFCACHE struct xfs_inode **i_refcache; /* ptr to entry in ref cache */ struct xfs_inode *i_release; /* inode to unref */ @@ -389,11 +390,14 @@ typedef struct xfs_inode { (((vfsp)->vfs_flag & VFS_GRPID) || ((pip)->i_d.di_mode & S_ISGID)) /* - * xfs_iget.c prototypes. + * Flags for xfs_iget() */ +#define XFS_IGET_CREATE 0x1 +#define XFS_IGET_BULKSTAT 0x2 -#define IGET_CREATE 1 - +/* + * xfs_iget.c prototypes. + */ void xfs_ihash_init(struct xfs_mount *); void xfs_ihash_free(struct xfs_mount *); void xfs_chash_init(struct xfs_mount *); @@ -425,7 +429,7 @@ int xfs_itobp(struct xfs_mount *, struct xfs_trans *, xfs_inode_t *, xfs_dinode_t **, struct xfs_buf **, xfs_daddr_t, uint); int xfs_iread(struct xfs_mount *, struct xfs_trans *, xfs_ino_t, - xfs_inode_t **, xfs_daddr_t); + xfs_inode_t **, xfs_daddr_t, uint); int xfs_iread_extents(struct xfs_trans *, xfs_inode_t *, int); int xfs_ialloc(struct xfs_trans *, xfs_inode_t *, mode_t, xfs_nlink_t, xfs_dev_t, struct cred *, xfs_prid_t, diff --git a/fs/xfs/xfs_inode_item.c b/fs/xfs/xfs_inode_item.c index f8e80d8e723..a7a92251eb5 100644 --- a/fs/xfs/xfs_inode_item.c +++ b/fs/xfs/xfs_inode_item.c @@ -743,21 +743,6 @@ xfs_inode_item_committed( } /* - * The transaction with the inode locked has aborted. The inode - * must not be dirty within the transaction (unless we're forcibly - * shutting down). We simply unlock just as if the transaction - * had been cancelled. - */ -STATIC void -xfs_inode_item_abort( - xfs_inode_log_item_t *iip) -{ - xfs_inode_item_unlock(iip); - return; -} - - -/* * This gets called by xfs_trans_push_ail(), when IOP_TRYLOCK * failed to get the inode flush lock but did get the inode locked SHARED. * Here we're trying to see if the inode buffer is incore, and if so whether it's @@ -915,7 +900,6 @@ STATIC struct xfs_item_ops xfs_inode_item_ops = { .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_abort = (void(*)(xfs_log_item_t*))xfs_inode_item_abort, .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 diff --git a/fs/xfs/xfs_inode_item.h b/fs/xfs/xfs_inode_item.h index 5db6cd1b4cf..bfe92ea1795 100644 --- a/fs/xfs/xfs_inode_item.h +++ b/fs/xfs/xfs_inode_item.h @@ -25,52 +25,54 @@ * must be added on to the end. */ typedef struct xfs_inode_log_format { - unsigned short ilf_type; /* inode log item type */ - unsigned short ilf_size; /* size of this item */ - uint ilf_fields; /* flags for fields logged */ - ushort ilf_asize; /* size of attr d/ext/root */ - ushort ilf_dsize; /* size of data/ext/root */ - xfs_ino_t ilf_ino; /* inode number */ + __uint16_t ilf_type; /* inode log item type */ + __uint16_t ilf_size; /* size of this item */ + __uint32_t ilf_fields; /* flags for fields logged */ + __uint16_t ilf_asize; /* size of attr d/ext/root */ + __uint16_t ilf_dsize; /* size of data/ext/root */ + __uint64_t ilf_ino; /* inode number */ union { - xfs_dev_t ilfu_rdev; /* rdev value for dev inode*/ + __uint32_t ilfu_rdev; /* rdev value for dev inode*/ uuid_t ilfu_uuid; /* mount point value */ } ilf_u; __int64_t ilf_blkno; /* blkno of inode buffer */ - int ilf_len; /* len of inode buffer */ - int ilf_boffset; /* off of inode in buffer */ + __int32_t ilf_len; /* len of inode buffer */ + __int32_t ilf_boffset; /* off of inode in buffer */ } xfs_inode_log_format_t; +#ifndef HAVE_FORMAT32 typedef struct xfs_inode_log_format_32 { - unsigned short ilf_type; /* 16: inode log item type */ - unsigned short ilf_size; /* 16: size of this item */ - uint ilf_fields; /* 32: flags for fields logged */ - ushort ilf_asize; /* 32: size of attr d/ext/root */ - ushort ilf_dsize; /* 32: size of data/ext/root */ - xfs_ino_t ilf_ino; /* 64: inode number */ + __uint16_t ilf_type; /* inode log item type */ + __uint16_t ilf_size; /* size of this item */ + __uint32_t ilf_fields; /* flags for fields logged */ + __uint16_t ilf_asize; /* size of attr d/ext/root */ + __uint16_t ilf_dsize; /* size of data/ext/root */ + __uint64_t ilf_ino; /* inode number */ union { - xfs_dev_t ilfu_rdev; /* 32: rdev value for dev inode*/ - uuid_t ilfu_uuid; /* 128: mount point value */ + __uint32_t ilfu_rdev; /* rdev value for dev inode*/ + uuid_t ilfu_uuid; /* mount point value */ } ilf_u; - __int64_t ilf_blkno; /* 64: blkno of inode buffer */ - int ilf_len; /* 32: len of inode buffer */ - int ilf_boffset; /* 32: off of inode in buffer */ + __int64_t ilf_blkno; /* blkno of inode buffer */ + __int32_t ilf_len; /* len of inode buffer */ + __int32_t ilf_boffset; /* off of inode in buffer */ } __attribute__((packed)) xfs_inode_log_format_32_t; +#endif typedef struct xfs_inode_log_format_64 { - unsigned short ilf_type; /* 16: inode log item type */ - unsigned short ilf_size; /* 16: size of this item */ - uint ilf_fields; /* 32: flags for fields logged */ - ushort ilf_asize; /* 32: size of attr d/ext/root */ - ushort ilf_dsize; /* 32: size of data/ext/root */ - __uint32_t ilf_pad; /* 32: pad for 64 bit boundary */ - xfs_ino_t ilf_ino; /* 64: inode number */ + __uint16_t ilf_type; /* inode log item type */ + __uint16_t ilf_size; /* size of this item */ + __uint32_t ilf_fields; /* flags for fields logged */ + __uint16_t ilf_asize; /* size of attr d/ext/root */ + __uint16_t ilf_dsize; /* size of data/ext/root */ + __uint32_t ilf_pad; /* pad for 64 bit boundary */ + __uint64_t ilf_ino; /* inode number */ union { - xfs_dev_t ilfu_rdev; /* 32: rdev value for dev inode*/ - uuid_t ilfu_uuid; /* 128: mount point value */ + __uint32_t ilfu_rdev; /* rdev value for dev inode*/ + uuid_t ilfu_uuid; /* mount point value */ } ilf_u; - __int64_t ilf_blkno; /* 64: blkno of inode buffer */ - int ilf_len; /* 32: len of inode buffer */ - int ilf_boffset; /* 32: off of inode in buffer */ + __int64_t ilf_blkno; /* blkno of inode buffer */ + __int32_t ilf_len; /* len of inode buffer */ + __int32_t ilf_boffset; /* off of inode in buffer */ } xfs_inode_log_format_64_t; /* diff --git a/fs/xfs/xfs_iomap.c b/fs/xfs/xfs_iomap.c index f1949c16df1..19655124da7 100644 --- a/fs/xfs/xfs_iomap.c +++ b/fs/xfs/xfs_iomap.c @@ -398,6 +398,23 @@ xfs_flush_space( return 1; } +STATIC int +xfs_cmn_err_fsblock_zero( + xfs_inode_t *ip, + xfs_bmbt_irec_t *imap) +{ + xfs_cmn_err(XFS_PTAG_FSBLOCK_ZERO, CE_ALERT, ip->i_mount, + "Access to block zero in inode %llu " + "start_block: %llx start_off: %llx " + "blkcnt: %llx extent-state: %x\n", + (unsigned long long)ip->i_ino, + (unsigned long long)imap->br_startblock, + (unsigned long long)imap->br_startoff, + (unsigned long long)imap->br_blockcount, + imap->br_state); + return EFSCORRUPTED; +} + int xfs_iomap_write_direct( xfs_inode_t *ip, @@ -536,23 +553,17 @@ xfs_iomap_write_direct( * Copy any maps to caller's array and return any error. */ if (nimaps == 0) { - error = (ENOSPC); + error = ENOSPC; + goto error_out; + } + + if (unlikely(!imap.br_startblock && !(io->io_flags & XFS_IOCORE_RT))) { + error = xfs_cmn_err_fsblock_zero(ip, &imap); goto error_out; } *ret_imap = imap; *nmaps = 1; - if ( !(io->io_flags & XFS_IOCORE_RT) && !ret_imap->br_startblock) { - cmn_err(CE_PANIC,"Access to block zero: fs <%s> inode: %lld " - "start_block : %llx start_off : %llx blkcnt : %llx " - "extent-state : %x \n", - (ip->i_mount)->m_fsname, - (long long)ip->i_ino, - (unsigned long long)ret_imap->br_startblock, - (unsigned long long)ret_imap->br_startoff, - (unsigned long long)ret_imap->br_blockcount, - ret_imap->br_state); - } return 0; error0: /* Cancel bmap, unlock inode, unreserve quota blocks, cancel trans */ @@ -715,17 +726,8 @@ retry: goto retry; } - if (!(io->io_flags & XFS_IOCORE_RT) && !ret_imap->br_startblock) { - cmn_err(CE_PANIC,"Access to block zero: fs <%s> inode: %lld " - "start_block : %llx start_off : %llx blkcnt : %llx " - "extent-state : %x \n", - (ip->i_mount)->m_fsname, - (long long)ip->i_ino, - (unsigned long long)ret_imap->br_startblock, - (unsigned long long)ret_imap->br_startoff, - (unsigned long long)ret_imap->br_blockcount, - ret_imap->br_state); - } + if (unlikely(!imap[0].br_startblock && !(io->io_flags & XFS_IOCORE_RT))) + return xfs_cmn_err_fsblock_zero(ip, &imap[0]); *ret_imap = imap[0]; *nmaps = 1; @@ -853,24 +855,10 @@ xfs_iomap_write_allocate( * See if we were able to allocate an extent that * covers at least part of the callers request */ - for (i = 0; i < nimaps; i++) { - if (!(io->io_flags & XFS_IOCORE_RT) && - !imap[i].br_startblock) { - cmn_err(CE_PANIC,"Access to block zero: " - "fs <%s> inode: %lld " - "start_block : %llx start_off : %llx " - "blkcnt : %llx extent-state : %x \n", - (ip->i_mount)->m_fsname, - (long long)ip->i_ino, - (unsigned long long) - imap[i].br_startblock, - (unsigned long long) - imap[i].br_startoff, - (unsigned long long) - imap[i].br_blockcount, - imap[i].br_state); - } + if (unlikely(!imap[i].br_startblock && + !(io->io_flags & XFS_IOCORE_RT))) + return xfs_cmn_err_fsblock_zero(ip, &imap[i]); if ((offset_fsb >= imap[i].br_startoff) && (offset_fsb < (imap[i].br_startoff + imap[i].br_blockcount))) { @@ -941,7 +929,7 @@ xfs_iomap_write_unwritten( XFS_WRITE_LOG_COUNT); if (error) { xfs_trans_cancel(tp, 0); - goto error0; + return XFS_ERROR(error); } xfs_ilock(ip, XFS_ILOCK_EXCL); @@ -967,19 +955,11 @@ xfs_iomap_write_unwritten( error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES, NULL); xfs_iunlock(ip, XFS_ILOCK_EXCL); if (error) - goto error0; - - if ( !(io->io_flags & XFS_IOCORE_RT) && !imap.br_startblock) { - cmn_err(CE_PANIC,"Access to block zero: fs <%s> " - "inode: %lld start_block : %llx start_off : " - "%llx blkcnt : %llx extent-state : %x \n", - (ip->i_mount)->m_fsname, - (long long)ip->i_ino, - (unsigned long long)imap.br_startblock, - (unsigned long long)imap.br_startoff, - (unsigned long long)imap.br_blockcount, - imap.br_state); - } + return XFS_ERROR(error); + + if (unlikely(!imap.br_startblock && + !(io->io_flags & XFS_IOCORE_RT))) + return xfs_cmn_err_fsblock_zero(ip, &imap); if ((numblks_fsb = imap.br_blockcount) == 0) { /* @@ -999,6 +979,5 @@ error_on_bmapi_transaction: xfs_bmap_cancel(&free_list); xfs_trans_cancel(tp, (XFS_TRANS_RELEASE_LOG_RES | XFS_TRANS_ABORT)); xfs_iunlock(ip, XFS_ILOCK_EXCL); -error0: return XFS_ERROR(error); } diff --git a/fs/xfs/xfs_itable.c b/fs/xfs/xfs_itable.c index 46249e4d1fe..7775ddc0b3c 100644 --- a/fs/xfs/xfs_itable.c +++ b/fs/xfs/xfs_itable.c @@ -39,6 +39,16 @@ #include "xfs_error.h" #include "xfs_btree.h" +int +xfs_internal_inum( + xfs_mount_t *mp, + xfs_ino_t ino) +{ + return (ino == mp->m_sb.sb_rbmino || ino == mp->m_sb.sb_rsumino || + (XFS_SB_VERSION_HASQUOTA(&mp->m_sb) && + (ino == mp->m_sb.sb_uquotino || ino == mp->m_sb.sb_gquotino))); +} + STATIC int xfs_bulkstat_one_iget( xfs_mount_t *mp, /* mount point for filesystem */ @@ -52,7 +62,8 @@ xfs_bulkstat_one_iget( bhv_vnode_t *vp; int error; - error = xfs_iget(mp, NULL, ino, 0, XFS_ILOCK_SHARED, &ip, bno); + error = xfs_iget(mp, NULL, ino, + XFS_IGET_BULKSTAT, XFS_ILOCK_SHARED, &ip, bno); if (error) { *stat = BULKSTAT_RV_NOTHING; return error; @@ -212,17 +223,12 @@ xfs_bulkstat_one( xfs_dinode_t *dip; /* dinode inode pointer */ dip = (xfs_dinode_t *)dibuff; + *stat = BULKSTAT_RV_NOTHING; - if (!buffer || ino == mp->m_sb.sb_rbmino || ino == mp->m_sb.sb_rsumino || - (XFS_SB_VERSION_HASQUOTA(&mp->m_sb) && - (ino == mp->m_sb.sb_uquotino || ino == mp->m_sb.sb_gquotino))) { - *stat = BULKSTAT_RV_NOTHING; + if (!buffer || xfs_internal_inum(mp, ino)) return XFS_ERROR(EINVAL); - } - if (ubsize < sizeof(*buf)) { - *stat = BULKSTAT_RV_NOTHING; + if (ubsize < sizeof(*buf)) return XFS_ERROR(ENOMEM); - } buf = kmem_alloc(sizeof(*buf), KM_SLEEP); @@ -238,8 +244,7 @@ xfs_bulkstat_one( } if (copy_to_user(buffer, buf, sizeof(*buf))) { - *stat = BULKSTAT_RV_NOTHING; - error = EFAULT; + error = EFAULT; goto out_free; } @@ -253,6 +258,46 @@ xfs_bulkstat_one( } /* + * Test to see whether we can use the ondisk inode directly, based + * on the given bulkstat flags, filling in dipp accordingly. + * Returns zero if the inode is dodgey. + */ +STATIC int +xfs_bulkstat_use_dinode( + xfs_mount_t *mp, + int flags, + xfs_buf_t *bp, + int clustidx, + xfs_dinode_t **dipp) +{ + xfs_dinode_t *dip; + unsigned int aformat; + + *dipp = NULL; + if (!bp || (flags & BULKSTAT_FG_IGET)) + return 1; + dip = (xfs_dinode_t *) + xfs_buf_offset(bp, clustidx << mp->m_sb.sb_inodelog); + if (INT_GET(dip->di_core.di_magic, ARCH_CONVERT) != XFS_DINODE_MAGIC || + !XFS_DINODE_GOOD_VERSION( + INT_GET(dip->di_core.di_version, ARCH_CONVERT))) + return 0; + if (flags & BULKSTAT_FG_QUICK) { + *dipp = dip; + return 1; + } + /* BULKSTAT_FG_INLINE: if attr fork is local, or not there, use it */ + aformat = INT_GET(dip->di_core.di_aformat, ARCH_CONVERT); + if ((XFS_CFORK_Q(&dip->di_core) == 0) || + (aformat == XFS_DINODE_FMT_LOCAL) || + (aformat == XFS_DINODE_FMT_EXTENTS && !dip->di_core.di_anextents)) { + *dipp = dip; + return 1; + } + return 1; +} + +/* * Return stat information in bulk (by-inode) for the filesystem. */ int /* error status */ @@ -284,10 +329,11 @@ xfs_bulkstat( xfs_agino_t gino; /* current btree rec's start inode */ int i; /* loop index */ int icount; /* count of inodes good in irbuf */ + size_t irbsize; /* size of irec buffer in bytes */ xfs_ino_t ino; /* inode number (filesystem) */ - xfs_inobt_rec_t *irbp; /* current irec buffer pointer */ - xfs_inobt_rec_t *irbuf; /* start of irec buffer */ - xfs_inobt_rec_t *irbufend; /* end of good irec buffer entries */ + xfs_inobt_rec_incore_t *irbp; /* current irec buffer pointer */ + xfs_inobt_rec_incore_t *irbuf; /* start of irec buffer */ + xfs_inobt_rec_incore_t *irbufend; /* end of good irec buffer entries */ xfs_ino_t lastino=0; /* last inode number returned */ int nbcluster; /* # of blocks in a cluster */ int nicluster; /* # of inodes in a cluster */ @@ -328,13 +374,10 @@ xfs_bulkstat( (XFS_INODE_CLUSTER_SIZE(mp) >> mp->m_sb.sb_inodelog); nimask = ~(nicluster - 1); nbcluster = nicluster >> mp->m_sb.sb_inopblog; - /* - * Allocate a page-sized buffer for inode btree records. - * We could try allocating something smaller, but for normal - * calls we'll always (potentially) need the whole page. - */ - irbuf = kmem_alloc(NBPC, KM_SLEEP); - nirbuf = NBPC / sizeof(*irbuf); + irbuf = kmem_zalloc_greedy(&irbsize, NBPC, NBPC * 4, + KM_SLEEP | KM_MAYFAIL | KM_LARGE); + nirbuf = irbsize / sizeof(*irbuf); + /* * Loop over the allocation groups, starting from the last * inode returned; 0 means start of the allocation group. @@ -358,7 +401,7 @@ xfs_bulkstat( * Allocate and initialize a btree cursor for ialloc btree. */ cur = xfs_btree_init_cursor(mp, NULL, agbp, agno, XFS_BTNUM_INO, - (xfs_inode_t *)0, 0); + (xfs_inode_t *)0, 0); irbp = irbuf; irbufend = irbuf + nirbuf; end_of_ag = 0; @@ -395,9 +438,9 @@ xfs_bulkstat( gcnt++; } gfree |= XFS_INOBT_MASKN(0, chunkidx); - INT_SET(irbp->ir_startino, ARCH_CONVERT, gino); - INT_SET(irbp->ir_freecount, ARCH_CONVERT, gcnt); - INT_SET(irbp->ir_free, ARCH_CONVERT, gfree); + irbp->ir_startino = gino; + irbp->ir_freecount = gcnt; + irbp->ir_free = gfree; irbp++; agino = gino + XFS_INODES_PER_CHUNK; icount = XFS_INODES_PER_CHUNK - gcnt; @@ -451,11 +494,27 @@ xfs_bulkstat( } /* * If this chunk has any allocated inodes, save it. + * Also start read-ahead now for this chunk. */ if (gcnt < XFS_INODES_PER_CHUNK) { - INT_SET(irbp->ir_startino, ARCH_CONVERT, gino); - INT_SET(irbp->ir_freecount, ARCH_CONVERT, gcnt); - INT_SET(irbp->ir_free, ARCH_CONVERT, gfree); + /* + * Loop over all clusters in the next chunk. + * Do a readahead if there are any allocated + * inodes in that cluster. + */ + for (agbno = XFS_AGINO_TO_AGBNO(mp, gino), + chunkidx = 0; + chunkidx < XFS_INODES_PER_CHUNK; + chunkidx += nicluster, + agbno += nbcluster) { + if (XFS_INOBT_MASKN(chunkidx, + nicluster) & ~gfree) + xfs_btree_reada_bufs(mp, agno, + agbno, nbcluster); + } + irbp->ir_startino = gino; + irbp->ir_freecount = gcnt; + irbp->ir_free = gfree; irbp++; icount += XFS_INODES_PER_CHUNK - gcnt; } @@ -479,33 +538,11 @@ xfs_bulkstat( for (irbp = irbuf; irbp < irbufend && ubleft >= statstruct_size; irbp++) { /* - * Read-ahead the next chunk's worth of inodes. - */ - if (&irbp[1] < irbufend) { - /* - * Loop over all clusters in the next chunk. - * Do a readahead if there are any allocated - * inodes in that cluster. - */ - for (agbno = XFS_AGINO_TO_AGBNO(mp, - INT_GET(irbp[1].ir_startino, ARCH_CONVERT)), - chunkidx = 0; - chunkidx < XFS_INODES_PER_CHUNK; - chunkidx += nicluster, - agbno += nbcluster) { - if (XFS_INOBT_MASKN(chunkidx, - nicluster) & - ~(INT_GET(irbp[1].ir_free, ARCH_CONVERT))) - xfs_btree_reada_bufs(mp, agno, - agbno, nbcluster); - } - } - /* * Now process this chunk of inodes. */ - for (agino = INT_GET(irbp->ir_startino, ARCH_CONVERT), chunkidx = 0, clustidx = 0; + for (agino = irbp->ir_startino, chunkidx = clustidx = 0; ubleft > 0 && - INT_GET(irbp->ir_freecount, ARCH_CONVERT) < XFS_INODES_PER_CHUNK; + irbp->ir_freecount < XFS_INODES_PER_CHUNK; chunkidx++, clustidx++, agino++) { ASSERT(chunkidx < XFS_INODES_PER_CHUNK); /* @@ -525,11 +562,12 @@ xfs_bulkstat( */ if ((chunkidx & (nicluster - 1)) == 0) { agbno = XFS_AGINO_TO_AGBNO(mp, - INT_GET(irbp->ir_startino, ARCH_CONVERT)) + + irbp->ir_startino) + ((chunkidx & nimask) >> mp->m_sb.sb_inopblog); - if (flags & BULKSTAT_FG_QUICK) { + if (flags & (BULKSTAT_FG_QUICK | + BULKSTAT_FG_INLINE)) { ino = XFS_AGINO_TO_INO(mp, agno, agino); bno = XFS_AGB_TO_DADDR(mp, agno, @@ -543,6 +581,7 @@ xfs_bulkstat( KM_SLEEP); ip->i_ino = ino; ip->i_mount = mp; + spin_lock_init(&ip->i_flags_lock); if (bp) xfs_buf_relse(bp); error = xfs_itobp(mp, NULL, ip, @@ -564,30 +603,34 @@ xfs_bulkstat( /* * Skip if this inode is free. */ - if (XFS_INOBT_MASK(chunkidx) & INT_GET(irbp->ir_free, ARCH_CONVERT)) + if (XFS_INOBT_MASK(chunkidx) & irbp->ir_free) continue; /* * Count used inodes as free so we can tell * when the chunk is used up. */ - INT_MOD(irbp->ir_freecount, ARCH_CONVERT, +1); + irbp->ir_freecount++; ino = XFS_AGINO_TO_INO(mp, agno, agino); bno = XFS_AGB_TO_DADDR(mp, agno, agbno); - if (flags & BULKSTAT_FG_QUICK) { - dip = (xfs_dinode_t *)xfs_buf_offset(bp, - (clustidx << mp->m_sb.sb_inodelog)); - - if (INT_GET(dip->di_core.di_magic, ARCH_CONVERT) - != XFS_DINODE_MAGIC - || !XFS_DINODE_GOOD_VERSION( - INT_GET(dip->di_core.di_version, ARCH_CONVERT))) - continue; + if (!xfs_bulkstat_use_dinode(mp, flags, bp, + clustidx, &dip)) + continue; + /* + * If we need to do an iget, cannot hold bp. + * Drop it, until starting the next cluster. + */ + if ((flags & BULKSTAT_FG_INLINE) && !dip) { + if (bp) + xfs_buf_relse(bp); + bp = NULL; } /* * Get the inode and fill in a single buffer. * BULKSTAT_FG_QUICK uses dip to fill it in. * BULKSTAT_FG_IGET uses igets. + * BULKSTAT_FG_INLINE uses dip if we have an + * inline attr fork, else igets. * See: xfs_bulkstat_one & xfs_dm_bulkstat_one. * This is also used to count inodes/blks, etc * in xfs_qm_quotacheck. @@ -597,8 +640,15 @@ xfs_bulkstat( ubleft, private_data, bno, &ubused, dip, &fmterror); if (fmterror == BULKSTAT_RV_NOTHING) { - if (error == ENOMEM) + if (error == EFAULT) { + ubleft = 0; + rval = error; + break; + } + else if (error == ENOMEM) ubleft = 0; + else + lastino = ino; continue; } if (fmterror == BULKSTAT_RV_GIVEUP) { @@ -633,7 +683,7 @@ xfs_bulkstat( /* * Done, we're either out of filesystem or space to put the data. */ - kmem_free(irbuf, NBPC); + kmem_free(irbuf, irbsize); *ubcountp = ubelem; if (agno >= mp->m_sb.sb_agcount) { /* diff --git a/fs/xfs/xfs_itable.h b/fs/xfs/xfs_itable.h index be5f12e07d2..f25a28862a1 100644 --- a/fs/xfs/xfs_itable.h +++ b/fs/xfs/xfs_itable.h @@ -36,15 +36,16 @@ typedef int (*bulkstat_one_pf)(struct xfs_mount *mp, /* * Values for stat return value. */ -#define BULKSTAT_RV_NOTHING 0 -#define BULKSTAT_RV_DIDONE 1 -#define BULKSTAT_RV_GIVEUP 2 +#define BULKSTAT_RV_NOTHING 0 +#define BULKSTAT_RV_DIDONE 1 +#define BULKSTAT_RV_GIVEUP 2 /* * Values for bulkstat flag argument. */ -#define BULKSTAT_FG_IGET 0x1 /* Go through the buffer cache */ -#define BULKSTAT_FG_QUICK 0x2 /* No iget, walk the dinode cluster */ +#define BULKSTAT_FG_IGET 0x1 /* Go through the buffer cache */ +#define BULKSTAT_FG_QUICK 0x2 /* No iget, walk the dinode cluster */ +#define BULKSTAT_FG_INLINE 0x4 /* No iget if inline attrs */ /* * Return stat information in bulk (by-inode) for the filesystem. @@ -80,6 +81,11 @@ xfs_bulkstat_one( void *dibuff, int *stat); +int +xfs_internal_inum( + xfs_mount_t *mp, + xfs_ino_t ino); + int /* error status */ xfs_inumbers( xfs_mount_t *mp, /* mount point for filesystem */ diff --git a/fs/xfs/xfs_log.c b/fs/xfs/xfs_log.c index d8f5d4cbe8b..c48bf61f17b 100644 --- a/fs/xfs/xfs_log.c +++ b/fs/xfs/xfs_log.c @@ -617,7 +617,8 @@ xfs_log_unmount_write(xfs_mount_t *mp) reg[0].i_len = sizeof(magic); XLOG_VEC_SET_TYPE(®[0], XLOG_REG_TYPE_UNMOUNT); - error = xfs_log_reserve(mp, 600, 1, &tic, XFS_LOG, 0, 0); + error = xfs_log_reserve(mp, 600, 1, &tic, + XFS_LOG, 0, XLOG_UNMOUNT_REC_TYPE); if (!error) { /* remove inited flag */ ((xlog_ticket_t *)tic)->t_flags = 0; @@ -655,8 +656,11 @@ xfs_log_unmount_write(xfs_mount_t *mp) } else { LOG_UNLOCK(log, s); } - if (tic) + if (tic) { + xlog_trace_loggrant(log, tic, "unmount rec"); + xlog_ungrant_log_space(log, tic); xlog_state_put_ticket(log, tic); + } } else { /* * We're already in forced_shutdown mode, couldn't @@ -1196,7 +1200,7 @@ xlog_alloc_log(xfs_mount_t *mp, kmem_zalloc(sizeof(xlog_in_core_t), KM_SLEEP); iclog = *iclogp; iclog->hic_data = (xlog_in_core_2_t *) - kmem_zalloc(iclogsize, KM_SLEEP); + kmem_zalloc(iclogsize, KM_SLEEP | KM_LARGE); iclog->ic_prev = prev_iclog; prev_iclog = iclog; @@ -1413,7 +1417,7 @@ xlog_sync(xlog_t *log, ops = iclog->ic_header.h_num_logops; INT_SET(iclog->ic_header.h_num_logops, ARCH_CONVERT, ops); - bp = iclog->ic_bp; + bp = iclog->ic_bp; ASSERT(XFS_BUF_FSPRIVATE2(bp, unsigned long) == (unsigned long)1); XFS_BUF_SET_FSPRIVATE2(bp, (unsigned long)2); XFS_BUF_SET_ADDR(bp, BLOCK_LSN(INT_GET(iclog->ic_header.h_lsn, ARCH_CONVERT))); @@ -1430,15 +1434,14 @@ xlog_sync(xlog_t *log, } XFS_BUF_SET_PTR(bp, (xfs_caddr_t) &(iclog->ic_header), count); XFS_BUF_SET_FSPRIVATE(bp, iclog); /* save for later */ + XFS_BUF_ZEROFLAGS(bp); XFS_BUF_BUSY(bp); XFS_BUF_ASYNC(bp); /* * Do an ordered write for the log block. - * - * It may not be needed to flush the first split block in the log wrap - * case, but do it anyways to be safe -AK + * Its unnecessary to flush the first split block in the log wrap case. */ - if (log->l_mp->m_flags & XFS_MOUNT_BARRIER) + if (!split && (log->l_mp->m_flags & XFS_MOUNT_BARRIER)) XFS_BUF_ORDERED(bp); ASSERT(XFS_BUF_ADDR(bp) <= log->l_logBBsize-1); @@ -1460,7 +1463,7 @@ xlog_sync(xlog_t *log, return error; } if (split) { - bp = iclog->ic_log->l_xbuf; + bp = iclog->ic_log->l_xbuf; ASSERT(XFS_BUF_FSPRIVATE2(bp, unsigned long) == (unsigned long)1); XFS_BUF_SET_FSPRIVATE2(bp, (unsigned long)2); @@ -1468,6 +1471,7 @@ xlog_sync(xlog_t *log, XFS_BUF_SET_PTR(bp, (xfs_caddr_t)((__psint_t)&(iclog->ic_header)+ (__psint_t)count), split); XFS_BUF_SET_FSPRIVATE(bp, iclog); + XFS_BUF_ZEROFLAGS(bp); XFS_BUF_BUSY(bp); XFS_BUF_ASYNC(bp); if (log->l_mp->m_flags & XFS_MOUNT_BARRIER) @@ -1740,10 +1744,10 @@ xlog_write(xfs_mount_t * mp, xlog_in_core_t **commit_iclog, uint flags) { - xlog_t *log = mp->m_log; + xlog_t *log = mp->m_log; xlog_ticket_t *ticket = (xlog_ticket_t *)tic; + xlog_in_core_t *iclog = NULL; /* ptr to current in-core log */ xlog_op_header_t *logop_head; /* ptr to log operation header */ - xlog_in_core_t *iclog; /* ptr to current in-core log */ __psint_t ptr; /* copy address into data region */ int len; /* # xlog_write() bytes 2 still copy */ int index; /* region index currently copying */ @@ -2212,9 +2216,13 @@ xlog_state_do_callback( iclog = iclog->ic_next; } while (first_iclog != iclog); - if (repeats && (repeats % 10) == 0) { + + if (repeats > 5000) { + flushcnt += repeats; + repeats = 0; xfs_fs_cmn_err(CE_WARN, log->l_mp, - "xlog_state_do_callback: looping %d", repeats); + "%s: possible infinite loop (%d iterations)", + __FUNCTION__, flushcnt); } } while (!ioerrors && loopdidcallbacks); @@ -2246,6 +2254,7 @@ xlog_state_do_callback( } #endif + flushcnt = 0; if (log->l_iclog->ic_state & (XLOG_STATE_ACTIVE|XLOG_STATE_IOERROR)) { flushcnt = log->l_flushcnt; log->l_flushcnt = 0; diff --git a/fs/xfs/xfs_log.h b/fs/xfs/xfs_log.h index eacb3d4987f..ebbe93f4f97 100644 --- a/fs/xfs/xfs_log.h +++ b/fs/xfs/xfs_log.h @@ -48,16 +48,10 @@ static inline xfs_lsn_t _lsn_cmp(xfs_lsn_t lsn1, xfs_lsn_t lsn2) */ /* - * Flags to xfs_log_mount - */ -#define XFS_LOG_RECOVER 0x1 - -/* * Flags to xfs_log_done() */ #define XFS_LOG_REL_PERM_RESERV 0x1 - /* * Flags to xfs_log_reserve() * @@ -70,8 +64,6 @@ static inline xfs_lsn_t _lsn_cmp(xfs_lsn_t lsn1, xfs_lsn_t lsn2) #define XFS_LOG_SLEEP 0x0 #define XFS_LOG_NOSLEEP 0x1 #define XFS_LOG_PERM_RESERV 0x2 -#define XFS_LOG_RESV_ALL (XFS_LOG_NOSLEEP|XFS_LOG_PERM_RESERV) - /* * Flags to xfs_log_force() diff --git a/fs/xfs/xfs_log_priv.h b/fs/xfs/xfs_log_priv.h index 34bcbf50789..9bd3cdf11a8 100644 --- a/fs/xfs/xfs_log_priv.h +++ b/fs/xfs/xfs_log_priv.h @@ -32,7 +32,6 @@ struct xfs_mount; #define XLOG_MIN_ICLOGS 2 #define XLOG_MED_ICLOGS 4 #define XLOG_MAX_ICLOGS 8 -#define XLOG_CALLBACK_SIZE 10 #define XLOG_HEADER_MAGIC_NUM 0xFEEDbabe /* Invalid cycle number */ #define XLOG_VERSION_1 1 #define XLOG_VERSION_2 2 /* Large IClogs, Log sunit */ @@ -149,9 +148,6 @@ struct xfs_mount; #define XLOG_WAS_CONT_TRANS 0x08 /* Cont this trans into new region */ #define XLOG_END_TRANS 0x10 /* End a continued transaction */ #define XLOG_UNMOUNT_TRANS 0x20 /* Unmount a filesystem transaction */ -#define XLOG_SKIP_TRANS (XLOG_COMMIT_TRANS | XLOG_CONTINUE_TRANS | \ - XLOG_WAS_CONT_TRANS | XLOG_END_TRANS | \ - XLOG_UNMOUNT_TRANS) #ifdef __KERNEL__ /* @@ -506,6 +502,12 @@ extern int xlog_bread(xlog_t *, xfs_daddr_t, int, struct xfs_buf *); #define XLOG_TRACE_SLEEP_FLUSH 3 #define XLOG_TRACE_WAKE_FLUSH 4 +/* + * Unmount record type is used as a pseudo transaction type for the ticket. + * It's value must be outside the range of XFS_TRANS_* values. + */ +#define XLOG_UNMOUNT_REC_TYPE (-1U) + #endif /* __KERNEL__ */ #endif /* __XFS_LOG_PRIV_H__ */ diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c index 55b4237c215..3cb678e3a13 100644 --- a/fs/xfs/xfs_log_recover.c +++ b/fs/xfs/xfs_log_recover.c @@ -990,6 +990,8 @@ xlog_find_zeroed( xfs_daddr_t num_scan_bblks; int error, log_bbnum = log->l_logBBsize; + *blk_no = 0; + /* check totally zeroed log */ bp = xlog_get_bp(log, 1); if (!bp) diff --git a/fs/xfs/xfs_mount.c b/fs/xfs/xfs_mount.c index 10dbf203c62..9dfae18d995 100644 --- a/fs/xfs/xfs_mount.c +++ b/fs/xfs/xfs_mount.c @@ -1243,24 +1243,6 @@ xfs_mod_sb(xfs_trans_t *tp, __int64_t fields) xfs_trans_log_buf(tp, bp, first, last); } -/* - * In order to avoid ENOSPC-related deadlock caused by - * out-of-order locking of AGF buffer (PV 947395), we place - * constraints on the relationship among actual allocations for - * data blocks, freelist blocks, and potential file data bmap - * btree blocks. However, these restrictions may result in no - * actual space allocated for a delayed extent, for example, a data - * block in a certain AG is allocated but there is no additional - * block for the additional bmap btree block due to a split of the - * bmap btree of the file. The result of this may lead to an - * infinite loop in xfssyncd when the file gets flushed to disk and - * all delayed extents need to be actually allocated. To get around - * this, we explicitly set aside a few blocks which will not be - * reserved in delayed allocation. Considering the minimum number of - * needed freelist blocks is 4 fsbs, a potential split of file's bmap - * btree requires 1 fsb, so we set the number of set-aside blocks to 8. -*/ -#define SET_ASIDE_BLOCKS 8 /* * xfs_mod_incore_sb_unlocked() is a utility routine common used to apply @@ -1306,7 +1288,8 @@ xfs_mod_incore_sb_unlocked(xfs_mount_t *mp, xfs_sb_field_t field, return 0; case XFS_SBS_FDBLOCKS: - lcounter = (long long)mp->m_sb.sb_fdblocks - SET_ASIDE_BLOCKS; + lcounter = (long long) + mp->m_sb.sb_fdblocks - XFS_ALLOC_SET_ASIDE(mp); res_used = (long long)(mp->m_resblks - mp->m_resblks_avail); if (delta > 0) { /* Putting blocks back */ @@ -1340,7 +1323,7 @@ xfs_mod_incore_sb_unlocked(xfs_mount_t *mp, xfs_sb_field_t field, } } - mp->m_sb.sb_fdblocks = lcounter + SET_ASIDE_BLOCKS; + mp->m_sb.sb_fdblocks = lcounter + XFS_ALLOC_SET_ASIDE(mp); return 0; case XFS_SBS_FREXTENTS: lcounter = (long long)mp->m_sb.sb_frextents; @@ -1721,15 +1704,14 @@ xfs_mount_log_sbunit( * is present to prevent thrashing). */ +#ifdef CONFIG_HOTPLUG_CPU /* * hot-plug CPU notifier support. * - * We cannot use the hotcpu_register() function because it does - * not allow notifier instances. We need a notifier per filesystem - * as we need to be able to identify the filesystem to balance - * the counters out. This is achieved by having a notifier block - * embedded in the xfs_mount_t and doing pointer magic to get the - * mount pointer from the notifier block address. + * We need a notifier per filesystem as we need to be able to identify + * the filesystem to balance the counters out. This is achieved by + * having a notifier block embedded in the xfs_mount_t and doing pointer + * magic to get the mount pointer from the notifier block address. */ STATIC int xfs_icsb_cpu_notify( @@ -1779,6 +1761,7 @@ xfs_icsb_cpu_notify( return NOTIFY_OK; } +#endif /* CONFIG_HOTPLUG_CPU */ int xfs_icsb_init_counters( @@ -1791,9 +1774,11 @@ xfs_icsb_init_counters( if (mp->m_sb_cnts == NULL) return -ENOMEM; +#ifdef CONFIG_HOTPLUG_CPU mp->m_icsb_notifier.notifier_call = xfs_icsb_cpu_notify; mp->m_icsb_notifier.priority = 0; - register_cpu_notifier(&mp->m_icsb_notifier); + register_hotcpu_notifier(&mp->m_icsb_notifier); +#endif /* CONFIG_HOTPLUG_CPU */ for_each_online_cpu(i) { cntp = (xfs_icsb_cnts_t *)per_cpu_ptr(mp->m_sb_cnts, i); @@ -1812,7 +1797,7 @@ xfs_icsb_destroy_counters( xfs_mount_t *mp) { if (mp->m_sb_cnts) { - unregister_cpu_notifier(&mp->m_icsb_notifier); + unregister_hotcpu_notifier(&mp->m_icsb_notifier); free_percpu(mp->m_sb_cnts); } } @@ -2019,14 +2004,15 @@ xfs_icsb_sync_counters_lazy( * when we get near ENOSPC. */ #define XFS_ICSB_INO_CNTR_REENABLE 64 -#define XFS_ICSB_FDBLK_CNTR_REENABLE 512 +#define XFS_ICSB_FDBLK_CNTR_REENABLE(mp) \ + (512 + XFS_ALLOC_SET_ASIDE(mp)) STATIC void xfs_icsb_balance_counter( xfs_mount_t *mp, xfs_sb_field_t field, int flags) { - uint64_t count, resid = 0; + uint64_t count, resid; int weight = num_online_cpus(); int s; @@ -2053,11 +2039,12 @@ xfs_icsb_balance_counter( case XFS_SBS_FDBLOCKS: count = mp->m_sb.sb_fdblocks; resid = do_div(count, weight); - if (count < XFS_ICSB_FDBLK_CNTR_REENABLE) + if (count < XFS_ICSB_FDBLK_CNTR_REENABLE(mp)) goto out; break; default: BUG(); + count = resid = 0; /* quiet, gcc */ break; } @@ -2107,11 +2094,11 @@ again: case XFS_SBS_FDBLOCKS: BUG_ON((mp->m_resblks - mp->m_resblks_avail) != 0); - lcounter = icsbp->icsb_fdblocks; + lcounter = icsbp->icsb_fdblocks - XFS_ALLOC_SET_ASIDE(mp); lcounter += delta; if (unlikely(lcounter < 0)) goto slow_path; - icsbp->icsb_fdblocks = lcounter; + icsbp->icsb_fdblocks = lcounter + XFS_ALLOC_SET_ASIDE(mp); break; default: BUG(); diff --git a/fs/xfs/xfs_mount.h b/fs/xfs/xfs_mount.h index b2bd4be4200..e5f396ff9a3 100644 --- a/fs/xfs/xfs_mount.h +++ b/fs/xfs/xfs_mount.h @@ -331,7 +331,7 @@ typedef struct xfs_mount { xfs_agnumber_t m_agirotor; /* last ag dir inode alloced */ lock_t m_agirotor_lock;/* .. and lock protecting it */ xfs_agnumber_t m_maxagi; /* highest inode alloc group */ - uint m_ihsize; /* size of next field */ + size_t m_ihsize; /* size of next field */ struct xfs_ihash *m_ihash; /* fs private inode hash table*/ struct xfs_inode *m_inodes; /* active inode list */ struct list_head m_del_inodes; /* inodes to reclaim */ @@ -541,7 +541,8 @@ static inline xfs_mount_t *xfs_bhvtom(bhv_desc_t *bdp) #define XFS_VFSTOM(vfs) xfs_vfstom(vfs) static inline xfs_mount_t *xfs_vfstom(bhv_vfs_t *vfs) { - return XFS_BHVTOM(bhv_lookup(VFS_BHVHEAD(vfs), &xfs_vfsops)); + return XFS_BHVTOM(bhv_lookup_range(VFS_BHVHEAD(vfs), + VFS_POSITION_XFS, VFS_POSITION_XFS)); } #define XFS_DADDR_TO_AGNO(mp,d) xfs_daddr_to_agno(mp,d) diff --git a/fs/xfs/xfs_quota.h b/fs/xfs/xfs_quota.h index acb853b33eb..9dcb32aa4e2 100644 --- a/fs/xfs/xfs_quota.h +++ b/fs/xfs/xfs_quota.h @@ -281,8 +281,6 @@ typedef struct xfs_qoff_logformat { XFS_UQUOTA_CHKD|XFS_PQUOTA_ACCT|\ XFS_OQUOTA_ENFD|XFS_OQUOTA_CHKD|\ XFS_GQUOTA_ACCT) -#define XFS_MOUNT_QUOTA_MASK (XFS_MOUNT_QUOTA_ALL | XFS_UQUOTA_ACTIVE | \ - XFS_GQUOTA_ACTIVE | XFS_PQUOTA_ACTIVE) /* diff --git a/fs/xfs/xfs_rtalloc.c b/fs/xfs/xfs_rtalloc.c index 0c1e42b037e..880c73271c0 100644 --- a/fs/xfs/xfs_rtalloc.c +++ b/fs/xfs/xfs_rtalloc.c @@ -1929,7 +1929,7 @@ xfs_growfs_rt( /* * Initial error checking. */ - if (mp->m_rtdev_targp || mp->m_rbmip == NULL || + if (mp->m_rtdev_targp == NULL || mp->m_rbmip == NULL || (nrblocks = in->newblocks) <= sbp->sb_rblocks || (sbp->sb_rblocks && (in->extsize != sbp->sb_rextsize))) return XFS_ERROR(EINVAL); @@ -1948,7 +1948,7 @@ xfs_growfs_rt( */ nrextents = nrblocks; do_div(nrextents, in->extsize); - nrbmblocks = roundup_64(nrextents, NBBY * sbp->sb_blocksize); + nrbmblocks = howmany_64(nrextents, NBBY * sbp->sb_blocksize); nrextslog = xfs_highbit32(nrextents); nrsumlevels = nrextslog + 1; nrsumsize = (uint)sizeof(xfs_suminfo_t) * nrsumlevels * nrbmblocks; @@ -1976,7 +1976,10 @@ xfs_growfs_rt( if ((error = xfs_growfs_rt_alloc(mp, rsumblocks, nrsumblocks, mp->m_sb.sb_rsumino))) return error; - nmp = NULL; + /* + * Allocate a new (fake) mount/sb. + */ + nmp = kmem_alloc(sizeof(*nmp), KM_SLEEP); /* * Loop over the bitmap blocks. * We will do everything one bitmap block at a time. @@ -1987,10 +1990,6 @@ xfs_growfs_rt( ((sbp->sb_rextents & ((1 << mp->m_blkbit_log) - 1)) != 0); bmbno < nrbmblocks; bmbno++) { - /* - * Allocate a new (fake) mount/sb. - */ - nmp = kmem_alloc(sizeof(*nmp), KM_SLEEP); *nmp = *mp; nsbp = &nmp->m_sb; /* @@ -2018,13 +2017,13 @@ xfs_growfs_rt( cancelflags = 0; if ((error = xfs_trans_reserve(tp, 0, XFS_GROWRTFREE_LOG_RES(nmp), 0, 0, 0))) - goto error_exit; + break; /* * Lock out other callers by grabbing the bitmap inode lock. */ if ((error = xfs_trans_iget(mp, tp, mp->m_sb.sb_rbmino, 0, XFS_ILOCK_EXCL, &ip))) - goto error_exit; + break; ASSERT(ip == mp->m_rbmip); /* * Update the bitmap inode's size. @@ -2038,7 +2037,7 @@ xfs_growfs_rt( */ if ((error = xfs_trans_iget(mp, tp, mp->m_sb.sb_rsumino, 0, XFS_ILOCK_EXCL, &ip))) - goto error_exit; + break; ASSERT(ip == mp->m_rsumip); /* * Update the summary inode's size. @@ -2053,7 +2052,7 @@ xfs_growfs_rt( mp->m_rsumlevels != nmp->m_rsumlevels) { error = xfs_rtcopy_summary(mp, nmp, tp); if (error) - goto error_exit; + break; } /* * Update superblock fields. @@ -2080,18 +2079,13 @@ xfs_growfs_rt( error = xfs_rtfree_range(nmp, tp, sbp->sb_rextents, nsbp->sb_rextents - sbp->sb_rextents, &bp, &sumbno); if (error) - goto error_exit; + break; /* * Mark more blocks free in the superblock. */ xfs_trans_mod_sb(tp, XFS_TRANS_SB_FREXTENTS, nsbp->sb_rextents - sbp->sb_rextents); /* - * Free the fake mp structure. - */ - kmem_free(nmp, sizeof(*nmp)); - nmp = NULL; - /* * Update mp values into the real mp structure. */ mp->m_rsumlevels = nrsumlevels; @@ -2101,15 +2095,15 @@ xfs_growfs_rt( */ xfs_trans_commit(tp, 0, NULL); } - return 0; + + if (error) + xfs_trans_cancel(tp, cancelflags); /* - * Error paths come here. + * Free the fake mp structure. */ -error_exit: - if (nmp) - kmem_free(nmp, sizeof(*nmp)); - xfs_trans_cancel(tp, cancelflags); + kmem_free(nmp, sizeof(*nmp)); + return error; } diff --git a/fs/xfs/xfs_sb.h b/fs/xfs/xfs_sb.h index bf168a91ddb..467854b45c8 100644 --- a/fs/xfs/xfs_sb.h +++ b/fs/xfs/xfs_sb.h @@ -60,10 +60,6 @@ struct xfs_mount; XFS_SB_VERSION_LOGV2BIT | \ XFS_SB_VERSION_SECTORBIT | \ XFS_SB_VERSION_MOREBITSBIT) -#define XFS_SB_VERSION_OKSASHBITS \ - (XFS_SB_VERSION_NUMBITS | \ - XFS_SB_VERSION_REALFBITS | \ - XFS_SB_VERSION_OKSASHFBITS) #define XFS_SB_VERSION_OKREALBITS \ (XFS_SB_VERSION_NUMBITS | \ XFS_SB_VERSION_OKREALFBITS | \ @@ -81,9 +77,6 @@ struct xfs_mount; #define XFS_SB_VERSION2_RESERVED2BIT 0x00000002 #define XFS_SB_VERSION2_RESERVED4BIT 0x00000004 #define XFS_SB_VERSION2_ATTR2BIT 0x00000008 /* Inline attr rework */ -#define XFS_SB_VERSION2_SASHFBITS 0xff000000 /* Mask: features that - require changing - PROM and SASH */ #define XFS_SB_VERSION2_OKREALFBITS \ (XFS_SB_VERSION2_ATTR2BIT) @@ -238,12 +231,6 @@ static inline int xfs_sb_good_version(xfs_sb_t *sbp) } #endif /* __KERNEL__ */ -#define XFS_SB_GOOD_SASH_VERSION(sbp) \ - ((((sbp)->sb_versionnum >= XFS_SB_VERSION_1) && \ - ((sbp)->sb_versionnum <= XFS_SB_VERSION_3)) || \ - ((XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4) && \ - !((sbp)->sb_versionnum & ~XFS_SB_VERSION_OKSASHBITS))) - #define XFS_SB_VERSION_TONEW(v) xfs_sb_version_tonew(v) static inline unsigned xfs_sb_version_tonew(unsigned v) { @@ -461,15 +448,6 @@ static inline void xfs_sb_version_addattr2(xfs_sb_t *sbp) * File system sector to basic block conversions. */ #define XFS_FSS_TO_BB(mp,sec) ((sec) << (mp)->m_sectbb_log) -#define XFS_BB_TO_FSS(mp,bb) \ - (((bb) + (XFS_FSS_TO_BB(mp,1) - 1)) >> (mp)->m_sectbb_log) -#define XFS_BB_TO_FSST(mp,bb) ((bb) >> (mp)->m_sectbb_log) - -/* - * File system sector to byte conversions. - */ -#define XFS_FSS_TO_B(mp,sectno) ((xfs_fsize_t)(sectno) << (mp)->m_sb.sb_sectlog) -#define XFS_B_TO_FSST(mp,b) (((__uint64_t)(b)) >> (mp)->m_sb.sb_sectlog) /* * File system block to basic block conversions. diff --git a/fs/xfs/xfs_trans.h b/fs/xfs/xfs_trans.h index cb65c3a603f..c68e00105d2 100644 --- a/fs/xfs/xfs_trans.h +++ b/fs/xfs/xfs_trans.h @@ -149,7 +149,6 @@ typedef struct xfs_item_ops { void (*iop_unlock)(xfs_log_item_t *); xfs_lsn_t (*iop_committed)(xfs_log_item_t *, xfs_lsn_t); void (*iop_push)(xfs_log_item_t *); - void (*iop_abort)(xfs_log_item_t *); void (*iop_pushbuf)(xfs_log_item_t *); void (*iop_committing)(xfs_log_item_t *, xfs_lsn_t); } xfs_item_ops_t; @@ -163,7 +162,6 @@ typedef struct xfs_item_ops { #define IOP_UNLOCK(ip) (*(ip)->li_ops->iop_unlock)(ip) #define IOP_COMMITTED(ip, lsn) (*(ip)->li_ops->iop_committed)(ip, lsn) #define IOP_PUSH(ip) (*(ip)->li_ops->iop_push)(ip) -#define IOP_ABORT(ip) (*(ip)->li_ops->iop_abort)(ip) #define IOP_PUSHBUF(ip) (*(ip)->li_ops->iop_pushbuf)(ip) #define IOP_COMMITTING(ip, lsn) (*(ip)->li_ops->iop_committing)(ip, lsn) @@ -338,8 +336,6 @@ typedef void (*xfs_trans_callback_t)(struct xfs_trans *, void *); typedef struct xfs_trans { unsigned int t_magic; /* magic number */ xfs_log_callback_t t_logcb; /* log callback struct */ - struct xfs_trans *t_forw; /* async list pointers */ - struct xfs_trans *t_back; /* async list pointers */ unsigned int t_type; /* transaction type */ unsigned int t_log_res; /* amt of log space resvd */ unsigned int t_log_count; /* count for perm log res */ @@ -364,9 +360,11 @@ typedef struct xfs_trans { long t_res_fdblocks_delta; /* on-disk only chg */ long t_frextents_delta;/* superblock freextents chg*/ long t_res_frextents_delta; /* on-disk only chg */ +#ifdef DEBUG long t_ag_freeblks_delta; /* debugging counter */ long t_ag_flist_delta; /* debugging counter */ long t_ag_btree_delta; /* debugging counter */ +#endif long t_dblocks_delta;/* superblock dblocks change */ long t_agcount_delta;/* superblock agcount change */ long t_imaxpct_delta;/* superblock imaxpct change */ diff --git a/fs/xfs/xfs_trans_ail.c b/fs/xfs/xfs_trans_ail.c index 558c87ff0c4..fc39b166d40 100644 --- a/fs/xfs/xfs_trans_ail.c +++ b/fs/xfs/xfs_trans_ail.c @@ -276,7 +276,7 @@ xfs_trans_update_ail( xfs_mount_t *mp, xfs_log_item_t *lip, xfs_lsn_t lsn, - unsigned long s) + unsigned long s) __releases(mp->m_ail_lock) { xfs_ail_entry_t *ailp; xfs_log_item_t *dlip=NULL; @@ -328,7 +328,7 @@ void xfs_trans_delete_ail( xfs_mount_t *mp, xfs_log_item_t *lip, - unsigned long s) + unsigned long s) __releases(mp->m_ail_lock) { xfs_ail_entry_t *ailp; xfs_log_item_t *dlip; diff --git a/fs/xfs/xfs_trans_priv.h b/fs/xfs/xfs_trans_priv.h index 13edab8a9e9..447ac4308c9 100644 --- a/fs/xfs/xfs_trans_priv.h +++ b/fs/xfs/xfs_trans_priv.h @@ -46,11 +46,13 @@ xfs_log_busy_slot_t *xfs_trans_add_busy(xfs_trans_t *tp, /* * From xfs_trans_ail.c */ -void xfs_trans_update_ail(struct xfs_mount *, - struct xfs_log_item *, xfs_lsn_t, - unsigned long); -void xfs_trans_delete_ail(struct xfs_mount *, - struct xfs_log_item *, unsigned long); +void xfs_trans_update_ail(struct xfs_mount *mp, + struct xfs_log_item *lip, xfs_lsn_t lsn, + unsigned long s) + __releases(mp->m_ail_lock); +void xfs_trans_delete_ail(struct xfs_mount *mp, + struct xfs_log_item *lip, unsigned long s) + __releases(mp->m_ail_lock); struct xfs_log_item *xfs_trans_first_ail(struct xfs_mount *, int *); struct xfs_log_item *xfs_trans_next_ail(struct xfs_mount *, struct xfs_log_item *, int *, int *); diff --git a/fs/xfs/xfs_vfsops.c b/fs/xfs/xfs_vfsops.c index 6c96391f3f1..62336a4cc5a 100644 --- a/fs/xfs/xfs_vfsops.c +++ b/fs/xfs/xfs_vfsops.c @@ -515,7 +515,7 @@ xfs_mount( if (error) goto error2; - if ((mp->m_flags & XFS_MOUNT_BARRIER) && !(vfsp->vfs_flag & VFS_RDONLY)) + if (mp->m_flags & XFS_MOUNT_BARRIER) xfs_mountfs_check_barriers(mp); error = XFS_IOINIT(vfsp, args, flags); @@ -811,7 +811,8 @@ xfs_statvfs( statp->f_bsize = sbp->sb_blocksize; lsize = sbp->sb_logstart ? sbp->sb_logblocks : 0; statp->f_blocks = sbp->sb_dblocks - lsize; - statp->f_bfree = statp->f_bavail = sbp->sb_fdblocks; + statp->f_bfree = statp->f_bavail = + sbp->sb_fdblocks - XFS_ALLOC_SET_ASIDE(mp); fakeinos = statp->f_bfree << sbp->sb_inopblog; #if XFS_BIG_INUMS fakeinos += mp->m_inoadd; @@ -1921,7 +1922,7 @@ xfs_showargs( } if (mp->m_flags & XFS_MOUNT_IHASHSIZE) - seq_printf(m, "," MNTOPT_IHASHSIZE "=%d", mp->m_ihsize); + seq_printf(m, "," MNTOPT_IHASHSIZE "=%d", (int)mp->m_ihsize); if (mp->m_flags & XFS_MOUNT_DFLT_IOSIZE) seq_printf(m, "," MNTOPT_ALLOCSIZE "=%dk", diff --git a/fs/xfs/xfs_vnodeops.c b/fs/xfs/xfs_vnodeops.c index 00a6b7dc24a..061e2ffdd1d 100644 --- a/fs/xfs/xfs_vnodeops.c +++ b/fs/xfs/xfs_vnodeops.c @@ -2366,10 +2366,15 @@ xfs_remove( namelen = VNAMELEN(dentry); + if (!xfs_get_dir_entry(dentry, &ip)) { + dm_di_mode = ip->i_d.di_mode; + IRELE(ip); + } + if (DM_EVENT_ENABLED(dir_vp->v_vfsp, dp, DM_EVENT_REMOVE)) { error = XFS_SEND_NAMESP(mp, DM_EVENT_REMOVE, dir_vp, DM_RIGHT_NULL, NULL, DM_RIGHT_NULL, - name, NULL, 0, 0, 0); + name, NULL, dm_di_mode, 0, 0); if (error) return error; } @@ -2603,8 +2608,7 @@ xfs_link( vn_trace_entry(src_vp, __FUNCTION__, (inst_t *)__return_address); target_namelen = VNAMELEN(dentry); - if (VN_ISDIR(src_vp)) - return XFS_ERROR(EPERM); + ASSERT(!VN_ISDIR(src_vp)); sip = xfs_vtoi(src_vp); tdp = XFS_BHVTOI(target_dir_bdp); @@ -2699,9 +2703,8 @@ xfs_link( xfs_trans_log_inode(tp, tdp, XFS_ILOG_CORE); error = xfs_bumplink(tp, sip); - if (error) { + if (error) goto abort_return; - } /* * If this is a synchronous mount, make sure that the @@ -2719,9 +2722,8 @@ xfs_link( } error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES, NULL); - if (error) { + if (error) goto std_return; - } /* Fall through to std_return with error = 0. */ std_return: @@ -2742,6 +2744,8 @@ std_return: xfs_trans_cancel(tp, cancel_flags); goto std_return; } + + /* * xfs_mkdir * @@ -2996,7 +3000,7 @@ xfs_rmdir( int cancel_flags; int committed; bhv_vnode_t *dir_vp; - int dm_di_mode = 0; + int dm_di_mode = S_IFDIR; int last_cdp_link; int namelen; uint resblks; @@ -3011,11 +3015,16 @@ xfs_rmdir( return XFS_ERROR(EIO); namelen = VNAMELEN(dentry); + if (!xfs_get_dir_entry(dentry, &cdp)) { + dm_di_mode = cdp->i_d.di_mode; + IRELE(cdp); + } + if (DM_EVENT_ENABLED(dir_vp->v_vfsp, dp, DM_EVENT_REMOVE)) { error = XFS_SEND_NAMESP(mp, DM_EVENT_REMOVE, dir_vp, DM_RIGHT_NULL, NULL, DM_RIGHT_NULL, - name, NULL, 0, 0, 0); + name, NULL, dm_di_mode, 0, 0); if (error) return XFS_ERROR(error); } @@ -3835,7 +3844,9 @@ xfs_reclaim( XFS_MOUNT_ILOCK(mp); vn_bhv_remove(VN_BHV_HEAD(vp), XFS_ITOBHV(ip)); list_add_tail(&ip->i_reclaim, &mp->m_del_inodes); + spin_lock(&ip->i_flags_lock); ip->i_flags |= XFS_IRECLAIMABLE; + spin_unlock(&ip->i_flags_lock); XFS_MOUNT_IUNLOCK(mp); } return 0; @@ -3860,8 +3871,10 @@ xfs_finish_reclaim( * us. */ write_lock(&ih->ih_lock); + spin_lock(&ip->i_flags_lock); if ((ip->i_flags & XFS_IRECLAIM) || (!(ip->i_flags & XFS_IRECLAIMABLE) && vp == NULL)) { + spin_unlock(&ip->i_flags_lock); write_unlock(&ih->ih_lock); if (locked) { xfs_ifunlock(ip); @@ -3870,6 +3883,7 @@ xfs_finish_reclaim( return 1; } ip->i_flags |= XFS_IRECLAIM; + spin_unlock(&ip->i_flags_lock); write_unlock(&ih->ih_lock); /* @@ -4273,7 +4287,7 @@ xfs_free_file_space( xfs_mount_t *mp; int nimap; uint resblks; - int rounding; + uint rounding; int rt; xfs_fileoff_t startoffset_fsb; xfs_trans_t *tp; @@ -4314,8 +4328,7 @@ xfs_free_file_space( vn_iowait(vp); /* wait for the completion of any pending DIOs */ } - rounding = MAX((__uint8_t)(1 << mp->m_sb.sb_blocklog), - (__uint8_t)NBPP); + rounding = max_t(uint, 1 << mp->m_sb.sb_blocklog, NBPP); ilen = len + (offset & (rounding - 1)); ioffset = offset & ~(rounding - 1); if (ilen & (rounding - 1)) |