diff options
Diffstat (limited to 'fs')
95 files changed, 2424 insertions, 2155 deletions
diff --git a/fs/afs/write.c b/fs/afs/write.c index c63a3c8beb7..5e15a21dbf9 100644 --- a/fs/afs/write.c +++ b/fs/afs/write.c @@ -671,7 +671,6 @@ ssize_t afs_file_write(struct kiocb *iocb, const struct iovec *iov, struct afs_vnode *vnode = AFS_FS_I(dentry->d_inode); ssize_t result; size_t count = iov_length(iov, nr_segs); - int ret; _enter("{%x.%u},{%zu},%lu,", vnode->fid.vid, vnode->fid.vnode, count, nr_segs); @@ -691,13 +690,6 @@ ssize_t afs_file_write(struct kiocb *iocb, const struct iovec *iov, return result; } - /* return error values for O_SYNC and IS_SYNC() */ - if (IS_SYNC(&vnode->vfs_inode) || iocb->ki_filp->f_flags & O_SYNC) { - ret = afs_fsync(iocb->ki_filp, dentry, 1); - if (ret < 0) - result = ret; - } - _leave(" = %zd", result); return result; } diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index 06550affbd2..77f759302e1 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -909,7 +909,7 @@ static ssize_t btrfs_file_write(struct file *file, const char __user *buf, unsigned long last_index; int will_write; - will_write = ((file->f_flags & O_SYNC) || IS_SYNC(inode) || + will_write = ((file->f_flags & O_DSYNC) || IS_SYNC(inode) || (file->f_flags & O_DIRECT)); nrptrs = min((count + PAGE_CACHE_SIZE - 1) / PAGE_CACHE_SIZE, @@ -1076,7 +1076,7 @@ out_nolock: if (err) num_written = err; - if ((file->f_flags & O_SYNC) || IS_SYNC(inode)) { + if ((file->f_flags & O_DSYNC) || IS_SYNC(inode)) { trans = btrfs_start_transaction(root, 1); ret = btrfs_log_dentry_safe(trans, root, file->f_dentry); diff --git a/fs/cifs/dir.c b/fs/cifs/dir.c index 1f42f772865..6ccf7262d1b 100644 --- a/fs/cifs/dir.c +++ b/fs/cifs/dir.c @@ -214,7 +214,8 @@ int cifs_posix_open(char *full_path, struct inode **pinode, posix_flags |= SMB_O_EXCL; if (oflags & O_TRUNC) posix_flags |= SMB_O_TRUNC; - if (oflags & O_SYNC) + /* be safe and imply O_SYNC for O_DSYNC */ + if (oflags & O_DSYNC) posix_flags |= SMB_O_SYNC; if (oflags & O_DIRECTORY) posix_flags |= SMB_O_DIRECTORY; diff --git a/fs/cifs/file.c b/fs/cifs/file.c index 429337eb7af..057e1dae12a 100644 --- a/fs/cifs/file.c +++ b/fs/cifs/file.c @@ -76,8 +76,10 @@ static inline fmode_t cifs_posix_convert_flags(unsigned int flags) reopening a file. They had their effect on the original open */ if (flags & O_APPEND) posix_flags |= (fmode_t)O_APPEND; - if (flags & O_SYNC) - posix_flags |= (fmode_t)O_SYNC; + if (flags & O_DSYNC) + posix_flags |= (fmode_t)O_DSYNC; + if (flags & __O_SYNC) + posix_flags |= (fmode_t)__O_SYNC; if (flags & O_DIRECTORY) posix_flags |= (fmode_t)O_DIRECTORY; if (flags & O_NOFOLLOW) diff --git a/fs/debugfs/inode.c b/fs/debugfs/inode.c index 0d23b52dd22..b486169f42b 100644 --- a/fs/debugfs/inode.c +++ b/fs/debugfs/inode.c @@ -32,7 +32,9 @@ static struct vfsmount *debugfs_mount; static int debugfs_mount_count; static bool debugfs_registered; -static struct inode *debugfs_get_inode(struct super_block *sb, int mode, dev_t dev) +static struct inode *debugfs_get_inode(struct super_block *sb, int mode, dev_t dev, + void *data, const struct file_operations *fops) + { struct inode *inode = new_inode(sb); @@ -44,14 +46,18 @@ static struct inode *debugfs_get_inode(struct super_block *sb, int mode, dev_t d init_special_inode(inode, mode, dev); break; case S_IFREG: - inode->i_fop = &debugfs_file_operations; + inode->i_fop = fops ? fops : &debugfs_file_operations; + inode->i_private = data; break; case S_IFLNK: inode->i_op = &debugfs_link_operations; + inode->i_fop = fops; + inode->i_private = data; break; case S_IFDIR: inode->i_op = &simple_dir_inode_operations; - inode->i_fop = &simple_dir_operations; + inode->i_fop = fops ? fops : &simple_dir_operations; + inode->i_private = data; /* directory inodes start off with i_nlink == 2 * (for "." entry) */ @@ -64,7 +70,8 @@ static struct inode *debugfs_get_inode(struct super_block *sb, int mode, dev_t d /* SMP-safe */ static int debugfs_mknod(struct inode *dir, struct dentry *dentry, - int mode, dev_t dev) + int mode, dev_t dev, void *data, + const struct file_operations *fops) { struct inode *inode; int error = -EPERM; @@ -72,7 +79,7 @@ static int debugfs_mknod(struct inode *dir, struct dentry *dentry, if (dentry->d_inode) return -EEXIST; - inode = debugfs_get_inode(dir->i_sb, mode, dev); + inode = debugfs_get_inode(dir->i_sb, mode, dev, data, fops); if (inode) { d_instantiate(dentry, inode); dget(dentry); @@ -81,12 +88,13 @@ static int debugfs_mknod(struct inode *dir, struct dentry *dentry, return error; } -static int debugfs_mkdir(struct inode *dir, struct dentry *dentry, int mode) +static int debugfs_mkdir(struct inode *dir, struct dentry *dentry, int mode, + void *data, const struct file_operations *fops) { int res; mode = (mode & (S_IRWXUGO | S_ISVTX)) | S_IFDIR; - res = debugfs_mknod(dir, dentry, mode, 0); + res = debugfs_mknod(dir, dentry, mode, 0, data, fops); if (!res) { inc_nlink(dir); fsnotify_mkdir(dir, dentry); @@ -94,18 +102,20 @@ static int debugfs_mkdir(struct inode *dir, struct dentry *dentry, int mode) return res; } -static int debugfs_link(struct inode *dir, struct dentry *dentry, int mode) +static int debugfs_link(struct inode *dir, struct dentry *dentry, int mode, + void *data, const struct file_operations *fops) { mode = (mode & S_IALLUGO) | S_IFLNK; - return debugfs_mknod(dir, dentry, mode, 0); + return debugfs_mknod(dir, dentry, mode, 0, data, fops); } -static int debugfs_create(struct inode *dir, struct dentry *dentry, int mode) +static int debugfs_create(struct inode *dir, struct dentry *dentry, int mode, + void *data, const struct file_operations *fops) { int res; mode = (mode & S_IALLUGO) | S_IFREG; - res = debugfs_mknod(dir, dentry, mode, 0); + res = debugfs_mknod(dir, dentry, mode, 0, data, fops); if (!res) fsnotify_create(dir, dentry); return res; @@ -139,7 +149,9 @@ static struct file_system_type debug_fs_type = { static int debugfs_create_by_name(const char *name, mode_t mode, struct dentry *parent, - struct dentry **dentry) + struct dentry **dentry, + void *data, + const struct file_operations *fops) { int error = 0; @@ -164,13 +176,16 @@ static int debugfs_create_by_name(const char *name, mode_t mode, if (!IS_ERR(*dentry)) { switch (mode & S_IFMT) { case S_IFDIR: - error = debugfs_mkdir(parent->d_inode, *dentry, mode); + error = debugfs_mkdir(parent->d_inode, *dentry, mode, + data, fops); break; case S_IFLNK: - error = debugfs_link(parent->d_inode, *dentry, mode); + error = debugfs_link(parent->d_inode, *dentry, mode, + data, fops); break; default: - error = debugfs_create(parent->d_inode, *dentry, mode); + error = debugfs_create(parent->d_inode, *dentry, mode, + data, fops); break; } dput(*dentry); @@ -221,19 +236,13 @@ struct dentry *debugfs_create_file(const char *name, mode_t mode, if (error) goto exit; - error = debugfs_create_by_name(name, mode, parent, &dentry); + error = debugfs_create_by_name(name, mode, parent, &dentry, + data, fops); if (error) { dentry = NULL; simple_release_fs(&debugfs_mount, &debugfs_mount_count); goto exit; } - - if (dentry->d_inode) { - if (data) - dentry->d_inode->i_private = data; - if (fops) - dentry->d_inode->i_fop = fops; - } exit: return dentry; } diff --git a/fs/ext2/dir.c b/fs/ext2/dir.c index 6cde970b0a1..fc2bd05d355 100644 --- a/fs/ext2/dir.c +++ b/fs/ext2/dir.c @@ -353,8 +353,8 @@ ext2_readdir (struct file * filp, void * dirent, filldir_t filldir) * ext2_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. + * returns the page in which the entry was found (as a parameter - res_page), + * and the entry itself. Page is returned mapped and unlocked. * Entry is guaranteed to be valid. */ struct ext2_dir_entry_2 *ext2_find_entry (struct inode * dir, diff --git a/fs/ext2/ext2.h b/fs/ext2/ext2.h index 9a8a8e27a06..da318b0fa63 100644 --- a/fs/ext2/ext2.h +++ b/fs/ext2/ext2.h @@ -142,7 +142,7 @@ struct dentry *ext2_get_parent(struct dentry *child); /* super.c */ extern void ext2_error (struct super_block *, const char *, const char *, ...) __attribute__ ((format (printf, 3, 4))); -extern void ext2_warning (struct super_block *, const char *, const char *, ...) +extern void ext2_msg(struct super_block *, const char *, const char *, ...) __attribute__ ((format (printf, 3, 4))); extern void ext2_update_dynamic_rev (struct super_block *sb); extern void ext2_write_super (struct super_block *); diff --git a/fs/ext2/inode.c b/fs/ext2/inode.c index ade634076d0..71b032c65a0 100644 --- a/fs/ext2/inode.c +++ b/fs/ext2/inode.c @@ -137,7 +137,8 @@ static int ext2_block_to_path(struct inode *inode, int final = 0; if (i_block < 0) { - ext2_warning (inode->i_sb, "ext2_block_to_path", "block < 0"); + ext2_msg(inode->i_sb, KERN_WARNING, + "warning: %s: block < 0", __func__); } else if (i_block < direct_blocks) { offsets[n++] = i_block; final = direct_blocks; @@ -157,7 +158,8 @@ static int ext2_block_to_path(struct inode *inode, offsets[n++] = i_block & (ptrs - 1); final = ptrs; } else { - ext2_warning (inode->i_sb, "ext2_block_to_path", "block > big"); + ext2_msg(inode->i_sb, KERN_WARNING, + "warning: %s: block is too big", __func__); } if (boundary) *boundary = final - 1 - (i_block & (ptrs - 1)); diff --git a/fs/ext2/super.c b/fs/ext2/super.c index 1a9ffee47d5..1388802b780 100644 --- a/fs/ext2/super.c +++ b/fs/ext2/super.c @@ -58,27 +58,27 @@ void ext2_error (struct super_block * sb, const char * function, } va_start(args, fmt); - printk(KERN_CRIT "EXT2-fs error (device %s): %s: ",sb->s_id, function); + printk(KERN_CRIT "EXT2-fs (%s): error: %s: ", sb->s_id, function); vprintk(fmt, args); printk("\n"); va_end(args); if (test_opt(sb, ERRORS_PANIC)) - panic("EXT2-fs panic from previous error\n"); + panic("EXT2-fs: panic from previous error\n"); if (test_opt(sb, ERRORS_RO)) { - printk("Remounting filesystem read-only\n"); + ext2_msg(sb, KERN_CRIT, + "error: remounting filesystem read-only"); sb->s_flags |= MS_RDONLY; } } -void ext2_warning (struct super_block * sb, const char * function, - const char * fmt, ...) +void ext2_msg(struct super_block *sb, const char *prefix, + const char *fmt, ...) { va_list args; va_start(args, fmt); - printk(KERN_WARNING "EXT2-fs warning (device %s): %s: ", - sb->s_id, function); + printk("%sEXT2-fs (%s): ", prefix, sb->s_id); vprintk(fmt, args); printk("\n"); va_end(args); @@ -91,9 +91,9 @@ void ext2_update_dynamic_rev(struct super_block *sb) if (le32_to_cpu(es->s_rev_level) > EXT2_GOOD_OLD_REV) return; - ext2_warning(sb, __func__, - "updating to rev %d because of new feature flag, " - "running e2fsck is recommended", + ext2_msg(sb, KERN_WARNING, + "warning: updating to rev %d because of " + "new feature flag, running e2fsck is recommended", EXT2_DYNAMIC_REV); es->s_first_ino = cpu_to_le32(EXT2_GOOD_OLD_FIRST_INO); @@ -419,10 +419,10 @@ static const match_table_t tokens = { {Opt_err, NULL} }; -static int parse_options (char * options, - struct ext2_sb_info *sbi) +static int parse_options(char *options, struct super_block *sb) { - char * p; + char *p; + struct ext2_sb_info *sbi = EXT2_SB(sb); substring_t args[MAX_OPT_ARGS]; int option; @@ -505,7 +505,8 @@ static int parse_options (char * options, #else case Opt_user_xattr: case Opt_nouser_xattr: - printk("EXT2 (no)user_xattr options not supported\n"); + ext2_msg(sb, KERN_INFO, "(no)user_xattr options" + "not supported"); break; #endif #ifdef CONFIG_EXT2_FS_POSIX_ACL @@ -518,14 +519,15 @@ static int parse_options (char * options, #else case Opt_acl: case Opt_noacl: - printk("EXT2 (no)acl options not supported\n"); + ext2_msg(sb, KERN_INFO, + "(no)acl options not supported"); break; #endif case Opt_xip: #ifdef CONFIG_EXT2_FS_XIP set_opt (sbi->s_mount_opt, XIP); #else - printk("EXT2 xip option not supported\n"); + ext2_msg(sb, KERN_INFO, "xip option not supported"); #endif break; @@ -542,19 +544,18 @@ static int parse_options (char * options, case Opt_quota: case Opt_usrquota: case Opt_grpquota: - printk(KERN_ERR - "EXT2-fs: quota operations not supported.\n"); - + ext2_msg(sb, KERN_INFO, + "quota operations not supported"); break; #endif case Opt_reservation: set_opt(sbi->s_mount_opt, RESERVATION); - printk("reservations ON\n"); + ext2_msg(sb, KERN_INFO, "reservations ON"); break; case Opt_noreservation: clear_opt(sbi->s_mount_opt, RESERVATION); - printk("reservations OFF\n"); + ext2_msg(sb, KERN_INFO, "reservations OFF"); break; case Opt_ignore: break; @@ -573,34 +574,40 @@ static int ext2_setup_super (struct super_block * sb, struct ext2_sb_info *sbi = EXT2_SB(sb); if (le32_to_cpu(es->s_rev_level) > EXT2_MAX_SUPP_REV) { - printk ("EXT2-fs warning: revision level too high, " - "forcing read-only mode\n"); + ext2_msg(sb, KERN_ERR, + "error: revision level too high, " + "forcing read-only mode"); res = MS_RDONLY; } if (read_only) return res; if (!(sbi->s_mount_state & EXT2_VALID_FS)) - printk ("EXT2-fs warning: mounting unchecked fs, " - "running e2fsck is recommended\n"); + ext2_msg(sb, KERN_WARNING, + "warning: mounting unchecked fs, " + "running e2fsck is recommended"); else if ((sbi->s_mount_state & EXT2_ERROR_FS)) - printk ("EXT2-fs warning: mounting fs with errors, " - "running e2fsck is recommended\n"); + ext2_msg(sb, KERN_WARNING, + "warning: mounting fs with errors, " + "running e2fsck is recommended"); else if ((__s16) le16_to_cpu(es->s_max_mnt_count) >= 0 && le16_to_cpu(es->s_mnt_count) >= (unsigned short) (__s16) le16_to_cpu(es->s_max_mnt_count)) - printk ("EXT2-fs warning: maximal mount count reached, " - "running e2fsck is recommended\n"); + ext2_msg(sb, KERN_WARNING, + "warning: maximal mount count reached, " + "running e2fsck is recommended"); else if (le32_to_cpu(es->s_checkinterval) && - (le32_to_cpu(es->s_lastcheck) + le32_to_cpu(es->s_checkinterval) <= get_seconds())) - printk ("EXT2-fs warning: checktime reached, " - "running e2fsck is recommended\n"); + (le32_to_cpu(es->s_lastcheck) + + le32_to_cpu(es->s_checkinterval) <= get_seconds())) + ext2_msg(sb, KERN_WARNING, + "warning: checktime reached, " + "running e2fsck is recommended"); if (!le16_to_cpu(es->s_max_mnt_count)) es->s_max_mnt_count = cpu_to_le16(EXT2_DFL_MAX_MNT_COUNT); le16_add_cpu(&es->s_mnt_count, 1); ext2_write_super(sb); if (test_opt (sb, DEBUG)) - printk ("[EXT II FS %s, %s, bs=%lu, fs=%lu, gc=%lu, " - "bpg=%lu, ipg=%lu, mo=%04lx]\n", + ext2_msg(sb, KERN_INFO, "%s, %s, bs=%lu, fs=%lu, gc=%lu, " + "bpg=%lu, ipg=%lu, mo=%04lx]", EXT2FS_VERSION, EXT2FS_DATE, sb->s_blocksize, sbi->s_frag_size, sbi->s_groups_count, @@ -767,7 +774,7 @@ static int ext2_fill_super(struct super_block *sb, void *data, int silent) */ blocksize = sb_min_blocksize(sb, BLOCK_SIZE); if (!blocksize) { - printk ("EXT2-fs: unable to set blocksize\n"); + ext2_msg(sb, KERN_ERR, "error: unable to set blocksize"); goto failed_sbi; } @@ -783,7 +790,7 @@ static int ext2_fill_super(struct super_block *sb, void *data, int silent) } if (!(bh = sb_bread(sb, logic_sb_block))) { - printk ("EXT2-fs: unable to read superblock\n"); + ext2_msg(sb, KERN_ERR, "error: unable to read superblock"); goto failed_sbi; } /* @@ -826,7 +833,7 @@ static int ext2_fill_super(struct super_block *sb, void *data, int silent) set_opt(sbi->s_mount_opt, RESERVATION); - if (!parse_options ((char *) data, sbi)) + if (!parse_options((char *) data, sb)) goto failed_mount; sb->s_flags = (sb->s_flags & ~MS_POSIXACL) | @@ -840,8 +847,9 @@ static int ext2_fill_super(struct super_block *sb, void *data, int silent) (EXT2_HAS_COMPAT_FEATURE(sb, ~0U) || EXT2_HAS_RO_COMPAT_FEATURE(sb, ~0U) || EXT2_HAS_INCOMPAT_FEATURE(sb, ~0U))) - printk("EXT2-fs warning: feature flags set on rev 0 fs, " - "running e2fsck is recommended\n"); + ext2_msg(sb, KERN_WARNING, + "warning: feature flags set on rev 0 fs, " + "running e2fsck is recommended"); /* * Check feature flags regardless of the revision level, since we * previously didn't change the revision level when setting the flags, @@ -849,16 +857,16 @@ static int ext2_fill_super(struct super_block *sb, void *data, int silent) */ features = EXT2_HAS_INCOMPAT_FEATURE(sb, ~EXT2_FEATURE_INCOMPAT_SUPP); if (features) { - printk("EXT2-fs: %s: couldn't mount because of " - "unsupported optional features (%x).\n", - sb->s_id, le32_to_cpu(features)); + ext2_msg(sb, KERN_ERR, "error: couldn't mount because of " + "unsupported optional features (%x)", + le32_to_cpu(features)); goto failed_mount; } if (!(sb->s_flags & MS_RDONLY) && (features = EXT2_HAS_RO_COMPAT_FEATURE(sb, ~EXT2_FEATURE_RO_COMPAT_SUPP))){ - printk("EXT2-fs: %s: couldn't mount RDWR because of " - "unsupported optional features (%x).\n", - sb->s_id, le32_to_cpu(features)); + ext2_msg(sb, KERN_ERR, "error: couldn't mount RDWR because of " + "unsupported optional features (%x)", + le32_to_cpu(features)); goto failed_mount; } @@ -866,7 +874,8 @@ static int ext2_fill_super(struct super_block *sb, void *data, int silent) if (ext2_use_xip(sb) && blocksize != PAGE_SIZE) { if (!silent) - printk("XIP: Unsupported blocksize\n"); + ext2_msg(sb, KERN_ERR, + "error: unsupported blocksize for xip"); goto failed_mount; } @@ -875,7 +884,7 @@ static int ext2_fill_super(struct super_block *sb, void *data, int silent) brelse(bh); if (!sb_set_blocksize(sb, blocksize)) { - printk(KERN_ERR "EXT2-fs: blocksize too small for device.\n"); + ext2_msg(sb, KERN_ERR, "error: blocksize is too small"); goto failed_sbi; } @@ -883,14 +892,14 @@ static int ext2_fill_super(struct super_block *sb, void *data, int silent) offset = (sb_block*BLOCK_SIZE) % blocksize; bh = sb_bread(sb, logic_sb_block); if(!bh) { - printk("EXT2-fs: Couldn't read superblock on " - "2nd try.\n"); + ext2_msg(sb, KERN_ERR, "error: couldn't read" + "superblock on 2nd try"); goto failed_sbi; } es = (struct ext2_super_block *) (((char *)bh->b_data) + offset); sbi->s_es = es; if (es->s_magic != cpu_to_le16(EXT2_SUPER_MAGIC)) { - printk ("EXT2-fs: Magic mismatch, very weird !\n"); + ext2_msg(sb, KERN_ERR, "error: magic mismatch"); goto failed_mount; } } @@ -906,7 +915,8 @@ static int ext2_fill_super(struct super_block *sb, void *data, int silent) if ((sbi->s_inode_size < EXT2_GOOD_OLD_INODE_SIZE) || !is_power_of_2(sbi->s_inode_size) || (sbi->s_inode_size > blocksize)) { - printk ("EXT2-fs: unsupported inode size: %d\n", + ext2_msg(sb, KERN_ERR, + "error: unsupported inode size: %d", sbi->s_inode_size); goto failed_mount; } @@ -943,29 +953,33 @@ static int ext2_fill_super(struct super_block *sb, void *data, int silent) if (sb->s_blocksize != bh->b_size) { if (!silent) - printk ("VFS: Unsupported blocksize on dev " - "%s.\n", sb->s_id); + ext2_msg(sb, KERN_ERR, "error: unsupported blocksize"); goto failed_mount; } if (sb->s_blocksize != sbi->s_frag_size) { - printk ("EXT2-fs: fragsize %lu != blocksize %lu (not supported yet)\n", + ext2_msg(sb, KERN_ERR, + "error: fragsize %lu != blocksize %lu" + "(not supported yet)", sbi->s_frag_size, sb->s_blocksize); goto failed_mount; } if (sbi->s_blocks_per_group > sb->s_blocksize * 8) { - printk ("EXT2-fs: #blocks per group too big: %lu\n", + ext2_msg(sb, KERN_ERR, + "error: #blocks per group too big: %lu", sbi->s_blocks_per_group); goto failed_mount; } if (sbi->s_frags_per_group > sb->s_blocksize * 8) { - printk ("EXT2-fs: #fragments per group too big: %lu\n", + ext2_msg(sb, KERN_ERR, + "error: #fragments per group too big: %lu", sbi->s_frags_per_group); goto failed_mount; } if (sbi->s_inodes_per_group > sb->s_blocksize * 8) { - printk ("EXT2-fs: #inodes per group too big: %lu\n", + ext2_msg(sb, KERN_ERR, + "error: #inodes per group too big: %lu", sbi->s_inodes_per_group); goto failed_mount; } @@ -979,13 +993,13 @@ static int ext2_fill_super(struct super_block *sb, void *data, int silent) EXT2_DESC_PER_BLOCK(sb); sbi->s_group_desc = kmalloc (db_count * sizeof (struct buffer_head *), GFP_KERNEL); if (sbi->s_group_desc == NULL) { - printk ("EXT2-fs: not enough memory\n"); + ext2_msg(sb, KERN_ERR, "error: not enough memory"); goto failed_mount; } bgl_lock_init(sbi->s_blockgroup_lock); sbi->s_debts = kcalloc(sbi->s_groups_count, sizeof(*sbi->s_debts), GFP_KERNEL); if (!sbi->s_debts) { - printk ("EXT2-fs: not enough memory\n"); + ext2_msg(sb, KERN_ERR, "error: not enough memory"); goto failed_mount_group_desc; } for (i = 0; i < db_count; i++) { @@ -994,12 +1008,13 @@ static int ext2_fill_super(struct super_block *sb, void *data, int silent) if (!sbi->s_group_desc[i]) { for (j = 0; j < i; j++) brelse (sbi->s_group_desc[j]); - printk ("EXT2-fs: unable to read group descriptors\n"); + ext2_msg(sb, KERN_ERR, + "error: unable to read group descriptors"); goto failed_mount_group_desc; } } if (!ext2_check_descriptors (sb)) { - printk ("EXT2-fs: group descriptors corrupted!\n"); + ext2_msg(sb, KERN_ERR, "group descriptors corrupted"); goto failed_mount2; } sbi->s_gdb_count = db_count; @@ -1032,7 +1047,7 @@ static int ext2_fill_super(struct super_block *sb, void *data, int silent) ext2_count_dirs(sb)); } if (err) { - printk(KERN_ERR "EXT2-fs: insufficient memory\n"); + ext2_msg(sb, KERN_ERR, "error: insufficient memory"); goto failed_mount3; } /* @@ -1048,27 +1063,28 @@ static int ext2_fill_super(struct super_block *sb, void *data, int silent) } if (!S_ISDIR(root->i_mode) || !root->i_blocks || !root->i_size) { iput(root); - printk(KERN_ERR "EXT2-fs: corrupt root inode, run e2fsck\n"); + ext2_msg(sb, KERN_ERR, "error: corrupt root inode, run e2fsck"); goto failed_mount3; } sb->s_root = d_alloc_root(root); if (!sb->s_root) { iput(root); - printk(KERN_ERR "EXT2-fs: get root inode failed\n"); + ext2_msg(sb, KERN_ERR, "error: get root inode failed"); ret = -ENOMEM; goto failed_mount3; } if (EXT2_HAS_COMPAT_FEATURE(sb, EXT3_FEATURE_COMPAT_HAS_JOURNAL)) - ext2_warning(sb, __func__, - "mounting ext3 filesystem as ext2"); + ext2_msg(sb, KERN_WARNING, + "warning: mounting ext3 filesystem as ext2"); ext2_setup_super (sb, es, sb->s_flags & MS_RDONLY); return 0; cantfind_ext2: if (!silent) - printk("VFS: Can't find an ext2 filesystem on dev %s.\n", - sb->s_id); + ext2_msg(sb, KERN_ERR, + "error: can't find an ext2 filesystem on dev %s.", + sb->s_id); goto failed_mount; failed_mount3: percpu_counter_destroy(&sbi->s_freeblocks_counter); @@ -1121,8 +1137,24 @@ static void ext2_sync_super(struct super_block *sb, struct ext2_super_block *es) static int ext2_sync_fs(struct super_block *sb, int wait) { struct ext2_super_block *es = EXT2_SB(sb)->s_es; + struct buffer_head *sbh = EXT2_SB(sb)->s_sbh; lock_kernel(); + if (buffer_write_io_error(sbh)) { + /* + * Oh, dear. A previous attempt to write the + * superblock failed. This could happen because the + * USB device was yanked out. Or it could happen to + * be a transient write error and maybe the block will + * be remapped. Nothing we can do but to retry the + * write and hope for the best. + */ + ext2_msg(sb, KERN_ERR, + "previous I/O error to superblock detected\n"); + clear_buffer_write_io_error(sbh); + set_buffer_uptodate(sbh); + } + if (es->s_state & cpu_to_le16(EXT2_VALID_FS)) { ext2_debug("setting valid to 0\n"); es->s_state &= cpu_to_le16(~EXT2_VALID_FS); @@ -1170,7 +1202,7 @@ static int ext2_remount (struct super_block * sb, int * flags, char * data) /* * Allow the "check" option to be passed as a remount option. */ - if (!parse_options (data, sbi)) { + if (!parse_options(data, sb)) { err = -EINVAL; goto restore_opts; } @@ -1182,7 +1214,8 @@ static int ext2_remount (struct super_block * sb, int * flags, char * data) EXT2_MOUNT_XIP if not */ if ((ext2_use_xip(sb)) && (sb->s_blocksize != PAGE_SIZE)) { - printk("XIP: Unsupported blocksize\n"); + ext2_msg(sb, KERN_WARNING, + "warning: unsupported blocksize for xip"); err = -EINVAL; goto restore_opts; } @@ -1191,8 +1224,8 @@ static int ext2_remount (struct super_block * sb, int * flags, char * data) if (((sbi->s_mount_opt & EXT2_MOUNT_XIP) != (old_mount_opt & EXT2_MOUNT_XIP)) && invalidate_inodes(sb)) { - ext2_warning(sb, __func__, "refusing change of xip flag " - "with busy inodes while remounting"); + ext2_msg(sb, KERN_WARNING, "warning: refusing change of " + "xip flag with busy inodes while remounting"); sbi->s_mount_opt &= ~EXT2_MOUNT_XIP; sbi->s_mount_opt |= old_mount_opt & EXT2_MOUNT_XIP; } @@ -1216,9 +1249,10 @@ static int ext2_remount (struct super_block * sb, int * flags, char * data) __le32 ret = EXT2_HAS_RO_COMPAT_FEATURE(sb, ~EXT2_FEATURE_RO_COMPAT_SUPP); if (ret) { - printk("EXT2-fs: %s: couldn't remount RDWR because of " - "unsupported optional features (%x).\n", - sb->s_id, le32_to_cpu(ret)); + ext2_msg(sb, KERN_WARNING, + "warning: couldn't remount RDWR because of " + "unsupported optional features (%x).", + le32_to_cpu(ret)); err = -EROFS; goto restore_opts; } diff --git a/fs/ext2/xip.c b/fs/ext2/xip.c index c18fbf3e406..322a56b2dfb 100644 --- a/fs/ext2/xip.c +++ b/fs/ext2/xip.c @@ -69,8 +69,9 @@ void ext2_xip_verify_sb(struct super_block *sb) if ((sbi->s_mount_opt & EXT2_MOUNT_XIP) && !sb->s_bdev->bd_disk->fops->direct_access) { sbi->s_mount_opt &= (~EXT2_MOUNT_XIP); - ext2_warning(sb, __func__, - "ignoring xip option - not supported by bdev"); + ext2_msg(sb, KERN_WARNING, + "warning: ignoring xip option - " + "not supported by bdev"); } } diff --git a/fs/ext3/inode.c b/fs/ext3/inode.c index 2db95777890..ad14227f509 100644 --- a/fs/ext3/inode.c +++ b/fs/ext3/inode.c @@ -1151,6 +1151,16 @@ static int do_journal_get_write_access(handle_t *handle, return ext3_journal_get_write_access(handle, bh); } +/* + * Truncate blocks that were not used by write. We have to truncate the + * pagecache as well so that corresponding buffers get properly unmapped. + */ +static void ext3_truncate_failed_write(struct inode *inode) +{ + truncate_inode_pages(inode->i_mapping, inode->i_size); + ext3_truncate(inode); +} + static int ext3_write_begin(struct file *file, struct address_space *mapping, loff_t pos, unsigned len, unsigned flags, struct page **pagep, void **fsdata) @@ -1209,7 +1219,7 @@ write_begin_failed: unlock_page(page); page_cache_release(page); if (pos + len > inode->i_size) - ext3_truncate(inode); + ext3_truncate_failed_write(inode); } if (ret == -ENOSPC && ext3_should_retry_alloc(inode->i_sb, &retries)) goto retry; @@ -1304,7 +1314,7 @@ static int ext3_ordered_write_end(struct file *file, page_cache_release(page); if (pos + len > inode->i_size) - ext3_truncate(inode); + ext3_truncate_failed_write(inode); return ret ? ret : copied; } @@ -1330,7 +1340,7 @@ static int ext3_writeback_write_end(struct file *file, page_cache_release(page); if (pos + len > inode->i_size) - ext3_truncate(inode); + ext3_truncate_failed_write(inode); return ret ? ret : copied; } @@ -1383,7 +1393,7 @@ static int ext3_journalled_write_end(struct file *file, page_cache_release(page); if (pos + len > inode->i_size) - ext3_truncate(inode); + ext3_truncate_failed_write(inode); return ret ? ret : copied; } diff --git a/fs/ext3/resize.c b/fs/ext3/resize.c index 8359e7b3dc8..5f83b617917 100644 --- a/fs/ext3/resize.c +++ b/fs/ext3/resize.c @@ -266,7 +266,7 @@ static int setup_new_group_blocks(struct super_block *sb, goto exit_bh; if (IS_ERR(gdb = bclean(handle, sb, block))) { - err = PTR_ERR(bh); + err = PTR_ERR(gdb); goto exit_bh; } ext3_journal_dirty_metadata(handle, gdb); diff --git a/fs/ext3/super.c b/fs/ext3/super.c index 427496c4767..7ad1e8c30bd 100644 --- a/fs/ext3/super.c +++ b/fs/ext3/super.c @@ -135,12 +135,24 @@ void ext3_journal_abort_handle(const char *caller, const char *err_fn, if (is_handle_aborted(handle)) return; - printk(KERN_ERR "%s: aborting transaction: %s in %s\n", - caller, errstr, err_fn); + printk(KERN_ERR "EXT3-fs: %s: aborting transaction: %s in %s\n", + caller, errstr, err_fn); journal_abort_handle(handle); } +void ext3_msg(struct super_block *sb, const char *prefix, + const char *fmt, ...) +{ + va_list args; + + va_start(args, fmt); + printk("%sEXT3-fs (%s): ", prefix, sb->s_id); + vprintk(fmt, args); + printk("\n"); + va_end(args); +} + /* Deal with the reporting of failure conditions on a filesystem such as * inconsistencies detected or read IO failures. * @@ -174,12 +186,13 @@ static void ext3_handle_error(struct super_block *sb) journal_abort(journal, -EIO); } if (test_opt (sb, ERRORS_RO)) { - printk (KERN_CRIT "Remounting filesystem read-only\n"); + ext3_msg(sb, KERN_CRIT, + "error: remounting filesystem read-only"); 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", + panic("EXT3-fs (%s): panic forced after error\n", sb->s_id); } @@ -247,8 +260,7 @@ void __ext3_std_error (struct super_block * sb, const char * function, return; errstr = ext3_decode_error(sb, errno, nbuf); - printk (KERN_CRIT "EXT3-fs error (device %s) in %s: %s\n", - sb->s_id, function, errstr); + ext3_msg(sb, KERN_CRIT, "error in %s: %s", function, errstr); ext3_handle_error(sb); } @@ -268,21 +280,20 @@ void ext3_abort (struct super_block * sb, const char * function, { va_list args; - printk (KERN_CRIT "ext3_abort called.\n"); - va_start(args, fmt); - printk(KERN_CRIT "EXT3-fs error (device %s): %s: ",sb->s_id, function); + printk(KERN_CRIT "EXT3-fs (%s): error: %s: ", sb->s_id, function); vprintk(fmt, args); printk("\n"); va_end(args); if (test_opt(sb, ERRORS_PANIC)) - panic("EXT3-fs panic from previous error\n"); + panic("EXT3-fs: panic from previous error\n"); if (sb->s_flags & MS_RDONLY) return; - printk(KERN_CRIT "Remounting filesystem read-only\n"); + ext3_msg(sb, KERN_CRIT, + "error: remounting filesystem read-only"); EXT3_SB(sb)->s_mount_state |= EXT3_ERROR_FS; sb->s_flags |= MS_RDONLY; EXT3_SB(sb)->s_mount_opt |= EXT3_MOUNT_ABORT; @@ -296,7 +307,7 @@ void ext3_warning (struct super_block * sb, const char * function, va_list args; va_start(args, fmt); - printk(KERN_WARNING "EXT3-fs warning (device %s): %s: ", + printk(KERN_WARNING "EXT3-fs (%s): warning: %s: ", sb->s_id, function); vprintk(fmt, args); printk("\n"); @@ -310,10 +321,10 @@ void ext3_update_dynamic_rev(struct super_block *sb) if (le32_to_cpu(es->s_rev_level) > EXT3_GOOD_OLD_REV) return; - ext3_warning(sb, __func__, - "updating to rev %d because of new feature flag, " - "running e2fsck is recommended", - EXT3_DYNAMIC_REV); + ext3_msg(sb, KERN_WARNING, + "warning: updating to rev %d because of " + "new feature flag, running e2fsck is recommended", + EXT3_DYNAMIC_REV); es->s_first_ino = cpu_to_le32(EXT3_GOOD_OLD_FIRST_INO); es->s_inode_size = cpu_to_le16(EXT3_GOOD_OLD_INODE_SIZE); @@ -331,7 +342,7 @@ void ext3_update_dynamic_rev(struct super_block *sb) /* * Open the external journal device */ -static struct block_device *ext3_blkdev_get(dev_t dev) +static struct block_device *ext3_blkdev_get(dev_t dev, struct super_block *sb) { struct block_device *bdev; char b[BDEVNAME_SIZE]; @@ -342,8 +353,9 @@ static struct block_device *ext3_blkdev_get(dev_t dev) return bdev; fail: - printk(KERN_ERR "EXT3: failed to open journal device %s: %ld\n", - __bdevname(dev, b), PTR_ERR(bdev)); + ext3_msg(sb, "error: failed to open journal device %s: %ld", + __bdevname(dev, b), PTR_ERR(bdev)); + return NULL; } @@ -378,13 +390,13 @@ 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", + ext3_msg(sb, KERN_ERR, "error: sb orphan head is %d", le32_to_cpu(sbi->s_es->s_last_orphan)); - printk(KERN_ERR "sb_info orphan list:\n"); + ext3_msg(sb, KERN_ERR, "sb_info orphan list:"); list_for_each(l, &sbi->s_orphan) { struct inode *inode = orphan_list_entry(l); - printk(KERN_ERR " " + ext3_msg(sb, KERN_ERR, " " "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, @@ -527,9 +539,22 @@ static inline void ext3_show_quota_options(struct seq_file *seq, struct super_bl #if defined(CONFIG_QUOTA) struct ext3_sb_info *sbi = EXT3_SB(sb); - if (sbi->s_jquota_fmt) - seq_printf(seq, ",jqfmt=%s", - (sbi->s_jquota_fmt == QFMT_VFS_OLD) ? "vfsold": "vfsv0"); + if (sbi->s_jquota_fmt) { + char *fmtname = ""; + + switch (sbi->s_jquota_fmt) { + case QFMT_VFS_OLD: + fmtname = "vfsold"; + break; + case QFMT_VFS_V0: + fmtname = "vfsv0"; + break; + case QFMT_VFS_V1: + fmtname = "vfsv1"; + break; + } + seq_printf(seq, ",jqfmt=%s", fmtname); + } if (sbi->s_qf_names[USRQUOTA]) seq_printf(seq, ",usrjquota=%s", sbi->s_qf_names[USRQUOTA]); @@ -636,6 +661,9 @@ static int ext3_show_options(struct seq_file *seq, struct vfsmount *vfs) if (test_opt(sb, DATA_ERR_ABORT)) seq_puts(seq, ",data_err=abort"); + if (test_opt(sb, NOLOAD)) + seq_puts(seq, ",norecovery"); + ext3_show_quota_options(seq, sb); return 0; @@ -787,9 +815,9 @@ enum { Opt_abort, Opt_data_journal, Opt_data_ordered, Opt_data_writeback, Opt_data_err_abort, Opt_data_err_ignore, Opt_usrjquota, Opt_grpjquota, Opt_offusrjquota, Opt_offgrpjquota, - Opt_jqfmt_vfsold, Opt_jqfmt_vfsv0, Opt_quota, Opt_noquota, - Opt_ignore, Opt_barrier, Opt_err, Opt_resize, Opt_usrquota, - Opt_grpquota + Opt_jqfmt_vfsold, Opt_jqfmt_vfsv0, Opt_jqfmt_vfsv1, Opt_quota, + Opt_noquota, Opt_ignore, Opt_barrier, Opt_err, Opt_resize, + Opt_usrquota, Opt_grpquota }; static const match_table_t tokens = { @@ -818,6 +846,7 @@ static const match_table_t tokens = { {Opt_reservation, "reservation"}, {Opt_noreservation, "noreservation"}, {Opt_noload, "noload"}, + {Opt_noload, "norecovery"}, {Opt_nobh, "nobh"}, {Opt_bh, "bh"}, {Opt_commit, "commit=%u"}, @@ -836,6 +865,7 @@ static const match_table_t tokens = { {Opt_grpjquota, "grpjquota=%s"}, {Opt_jqfmt_vfsold, "jqfmt=vfsold"}, {Opt_jqfmt_vfsv0, "jqfmt=vfsv0"}, + {Opt_jqfmt_vfsv1, "jqfmt=vfsv1"}, {Opt_grpquota, "grpquota"}, {Opt_noquota, "noquota"}, {Opt_quota, "quota"}, @@ -845,7 +875,7 @@ static const match_table_t tokens = { {Opt_err, NULL}, }; -static ext3_fsblk_t get_sb_block(void **data) +static ext3_fsblk_t get_sb_block(void **data, struct super_block *sb) { ext3_fsblk_t sb_block; char *options = (char *) *data; @@ -856,7 +886,7 @@ static ext3_fsblk_t get_sb_block(void **data) /*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", + ext3_msg(sb, "error: invalid sb specification: %s", (char *) *data); return 1; } @@ -956,7 +986,8 @@ static int parse_options (char *options, struct super_block *sb, #else case Opt_user_xattr: case Opt_nouser_xattr: - printk("EXT3 (no)user_xattr options not supported\n"); + ext3_msg(sb, KERN_INFO, + "(no)user_xattr options not supported"); break; #endif #ifdef CONFIG_EXT3_FS_POSIX_ACL @@ -969,7 +1000,8 @@ static int parse_options (char *options, struct super_block *sb, #else case Opt_acl: case Opt_noacl: - printk("EXT3 (no)acl options not supported\n"); + ext3_msg(sb, KERN_INFO, + "(no)acl options not supported"); break; #endif case Opt_reservation: @@ -985,16 +1017,16 @@ static int parse_options (char *options, struct super_block *sb, user to specify an existing inode to be the journal file. */ if (is_remount) { - printk(KERN_ERR "EXT3-fs: cannot specify " - "journal on remount\n"); + ext3_msg(sb, KERN_ERR, "error: cannot specify " + "journal on remount"); return 0; } set_opt (sbi->s_mount_opt, UPDATE_JOURNAL); break; case Opt_journal_inum: if (is_remount) { - printk(KERN_ERR "EXT3-fs: cannot specify " - "journal on remount\n"); + ext3_msg(sb, KERN_ERR, "error: cannot specify " + "journal on remount"); return 0; } if (match_int(&args[0], &option)) @@ -1003,8 +1035,8 @@ static int parse_options (char *options, struct super_block *sb, break; case Opt_journal_dev: if (is_remount) { - printk(KERN_ERR "EXT3-fs: cannot specify " - "journal on remount\n"); + ext3_msg(sb, KERN_ERR, "error: cannot specify " + "journal on remount"); return 0; } if (match_int(&args[0], &option)) @@ -1036,12 +1068,11 @@ static int parse_options (char *options, struct super_block *sb, if ((sbi->s_mount_opt & EXT3_MOUNT_DATA_FLAGS) == data_opt) break; - printk(KERN_ERR - "EXT3-fs (device %s): Cannot change " + ext3_msg(sb, KERN_ERR, + "error: cannot change " "data mode on remount. The filesystem " "is mounted in data=%s mode and you " - "try to remount it in data=%s mode.\n", - sb->s_id, + "try to remount it in data=%s mode.", data_mode_string(sbi->s_mount_opt & EXT3_MOUNT_DATA_FLAGS), data_mode_string(data_opt)); @@ -1066,31 +1097,31 @@ static int parse_options (char *options, struct super_block *sb, set_qf_name: if (sb_any_quota_loaded(sb) && !sbi->s_qf_names[qtype]) { - printk(KERN_ERR - "EXT3-fs: Cannot change journaled " - "quota options when quota turned on.\n"); + ext3_msg(sb, KERN_ERR, + "error: cannot change journaled " + "quota options when quota turned on."); return 0; } qname = match_strdup(&args[0]); if (!qname) { - printk(KERN_ERR - "EXT3-fs: not enough memory for " - "storing quotafile name.\n"); + ext3_msg(sb, KERN_ERR, + "error: not enough memory for " + "storing quotafile name."); return 0; } if (sbi->s_qf_names[qtype] && strcmp(sbi->s_qf_names[qtype], qname)) { - printk(KERN_ERR - "EXT3-fs: %s quota file already " - "specified.\n", QTYPE2NAME(qtype)); + ext3_msg(sb, KERN_ERR, + "error: %s quota file already " + "specified.", QTYPE2NAME(qtype)); kfree(qname); return 0; } sbi->s_qf_names[qtype] = qname; if (strchr(sbi->s_qf_names[qtype], '/')) { - printk(KERN_ERR - "EXT3-fs: quotafile must be on " - "filesystem root.\n"); + ext3_msg(sb, KERN_ERR, + "error: quotafile must be on " + "filesystem root."); kfree(sbi->s_qf_names[qtype]); sbi->s_qf_names[qtype] = NULL; return 0; @@ -1105,9 +1136,9 @@ set_qf_name: clear_qf_name: if (sb_any_quota_loaded(sb) && sbi->s_qf_names[qtype]) { - printk(KERN_ERR "EXT3-fs: Cannot change " + ext3_msg(sb, KERN_ERR, "error: cannot change " "journaled quota options when " - "quota turned on.\n"); + "quota turned on."); return 0; } /* @@ -1121,12 +1152,15 @@ clear_qf_name: goto set_qf_format; case Opt_jqfmt_vfsv0: qfmt = QFMT_VFS_V0; + goto set_qf_format; + case Opt_jqfmt_vfsv1: + qfmt = QFMT_VFS_V1; set_qf_format: if (sb_any_quota_loaded(sb) && sbi->s_jquota_fmt != qfmt) { - printk(KERN_ERR "EXT3-fs: Cannot change " + ext3_msg(sb, KERN_ERR, "error: cannot change " "journaled quota options when " - "quota turned on.\n"); + "quota turned on."); return 0; } sbi->s_jquota_fmt = qfmt; @@ -1142,8 +1176,8 @@ set_qf_format: break; case Opt_noquota: if (sb_any_quota_loaded(sb)) { - printk(KERN_ERR "EXT3-fs: Cannot change quota " - "options when quota turned on.\n"); + ext3_msg(sb, KERN_ERR, "error: cannot change " + "quota options when quota turned on."); return 0; } clear_opt(sbi->s_mount_opt, QUOTA); @@ -1154,8 +1188,8 @@ set_qf_format: case Opt_quota: case Opt_usrquota: case Opt_grpquota: - printk(KERN_ERR - "EXT3-fs: quota options not supported.\n"); + ext3_msg(sb, KERN_ERR, + "error: quota options not supported."); break; case Opt_usrjquota: case Opt_grpjquota: @@ -1163,9 +1197,10 @@ set_qf_format: case Opt_offgrpjquota: case Opt_jqfmt_vfsold: case Opt_jqfmt_vfsv0: - printk(KERN_ERR - "EXT3-fs: journaled quota options not " - "supported.\n"); + case Opt_jqfmt_vfsv1: + ext3_msg(sb, KERN_ERR, + "error: journaled quota options not " + "supported."); break; case Opt_noquota: break; @@ -1185,8 +1220,9 @@ set_qf_format: break; case Opt_resize: if (!is_remount) { - printk("EXT3-fs: resize option only available " - "for remount\n"); + ext3_msg(sb, KERN_ERR, + "error: resize option only available " + "for remount"); return 0; } if (match_int(&args[0], &option) != 0) @@ -1200,9 +1236,9 @@ set_qf_format: clear_opt(sbi->s_mount_opt, NOBH); break; default: - printk (KERN_ERR - "EXT3-fs: Unrecognized mount option \"%s\" " - "or missing value\n", p); + ext3_msg(sb, KERN_ERR, + "error: unrecognized mount option \"%s\" " + "or missing value", p); return 0; } } @@ -1220,21 +1256,21 @@ set_qf_format: (sbi->s_mount_opt & EXT3_MOUNT_GRPQUOTA)) || (sbi->s_qf_names[GRPQUOTA] && (sbi->s_mount_opt & EXT3_MOUNT_USRQUOTA))) { - printk(KERN_ERR "EXT3-fs: old and new quota " - "format mixing.\n"); + ext3_msg(sb, KERN_ERR, "error: old and new quota " + "format mixing."); return 0; } if (!sbi->s_jquota_fmt) { - printk(KERN_ERR "EXT3-fs: journaled quota format " - "not specified.\n"); + ext3_msg(sb, KERN_ERR, "error: journaled quota format " + "not specified."); return 0; } } else { if (sbi->s_jquota_fmt) { - printk(KERN_ERR "EXT3-fs: journaled quota format " + ext3_msg(sb, KERN_ERR, "error: journaled quota format " "specified with no journaling " - "enabled.\n"); + "enabled."); return 0; } } @@ -1249,31 +1285,33 @@ static int ext3_setup_super(struct super_block *sb, struct ext3_super_block *es, int res = 0; if (le32_to_cpu(es->s_rev_level) > EXT3_MAX_SUPP_REV) { - printk (KERN_ERR "EXT3-fs warning: revision level too high, " - "forcing read-only mode\n"); + ext3_msg(sb, KERN_ERR, + "error: revision level too high, " + "forcing read-only mode"); res = MS_RDONLY; } if (read_only) return res; if (!(sbi->s_mount_state & EXT3_VALID_FS)) - printk (KERN_WARNING "EXT3-fs warning: mounting unchecked fs, " - "running e2fsck is recommended\n"); + ext3_msg(sb, KERN_WARNING, + "warning: mounting unchecked fs, " + "running e2fsck is recommended"); else if ((sbi->s_mount_state & EXT3_ERROR_FS)) - printk (KERN_WARNING - "EXT3-fs warning: mounting fs with errors, " - "running e2fsck is recommended\n"); + ext3_msg(sb, KERN_WARNING, + "warning: mounting fs with errors, " + "running e2fsck is recommended"); else if ((__s16) le16_to_cpu(es->s_max_mnt_count) >= 0 && le16_to_cpu(es->s_mnt_count) >= (unsigned short) (__s16) le16_to_cpu(es->s_max_mnt_count)) - printk (KERN_WARNING - "EXT3-fs warning: maximal mount count reached, " - "running e2fsck is recommended\n"); + ext3_msg(sb, KERN_WARNING, + "warning: maximal mount count reached, " + "running e2fsck is recommended"); else if (le32_to_cpu(es->s_checkinterval) && (le32_to_cpu(es->s_lastcheck) + le32_to_cpu(es->s_checkinterval) <= get_seconds())) - printk (KERN_WARNING - "EXT3-fs warning: checktime reached, " - "running e2fsck is recommended\n"); + ext3_msg(sb, KERN_WARNING, + "warning: checktime reached, " + "running e2fsck is recommended"); #if 0 /* @@@ We _will_ want to clear the valid bit if we find inconsistencies, to force a fsck at reboot. But for @@ -1290,22 +1328,20 @@ static int ext3_setup_super(struct super_block *sb, struct ext3_super_block *es, ext3_commit_super(sb, es, 1); if (test_opt(sb, DEBUG)) - printk(KERN_INFO "[EXT3 FS bs=%lu, gc=%lu, " - "bpg=%lu, ipg=%lu, mo=%04lx]\n", + ext3_msg(sb, KERN_INFO, "[bs=%lu, gc=%lu, " + "bpg=%lu, ipg=%lu, mo=%04lx]", sb->s_blocksize, sbi->s_groups_count, EXT3_BLOCKS_PER_GROUP(sb), EXT3_INODES_PER_GROUP(sb), sbi->s_mount_opt); - printk(KERN_INFO "EXT3 FS on %s, ", sb->s_id); if (EXT3_SB(sb)->s_journal->j_inode == NULL) { char b[BDEVNAME_SIZE]; - - printk("external journal on %s\n", + ext3_msg(sb, KERN_INFO, "using external journal on %s", bdevname(EXT3_SB(sb)->s_journal->j_dev, b)); } else { - printk("internal journal\n"); + ext3_msg(sb, KERN_INFO, "using internal journal"); } return res; } @@ -1399,8 +1435,8 @@ static void ext3_orphan_cleanup (struct super_block * sb, } if (bdev_read_only(sb->s_bdev)) { - printk(KERN_ERR "EXT3-fs: write access " - "unavailable, skipping orphan cleanup.\n"); + ext3_msg(sb, KERN_ERR, "error: write access " + "unavailable, skipping orphan cleanup."); return; } @@ -1414,8 +1450,7 @@ static void ext3_orphan_cleanup (struct super_block * sb, } if (s_flags & MS_RDONLY) { - printk(KERN_INFO "EXT3-fs: %s: orphan cleanup on readonly fs\n", - sb->s_id); + ext3_msg(sb, KERN_INFO, "orphan cleanup on readonly fs"); sb->s_flags &= ~MS_RDONLY; } #ifdef CONFIG_QUOTA @@ -1426,9 +1461,9 @@ static void ext3_orphan_cleanup (struct super_block * sb, if (EXT3_SB(sb)->s_qf_names[i]) { int ret = ext3_quota_on_mount(sb, i); if (ret < 0) - printk(KERN_ERR - "EXT3-fs: Cannot turn on journaled " - "quota: error %d\n", ret); + ext3_msg(sb, KERN_ERR, + "error: cannot turn on journaled " + "quota: %d", ret); } } #endif @@ -1466,11 +1501,11 @@ static void ext3_orphan_cleanup (struct super_block * sb, #define PLURAL(x) (x), ((x)==1) ? "" : "s" if (nr_orphans) - printk(KERN_INFO "EXT3-fs: %s: %d orphan inode%s deleted\n", - sb->s_id, PLURAL(nr_orphans)); + ext3_msg(sb, KERN_INFO, "%d orphan inode%s deleted", + PLURAL(nr_orphans)); if (nr_truncates) - printk(KERN_INFO "EXT3-fs: %s: %d truncate%s cleaned up\n", - sb->s_id, PLURAL(nr_truncates)); + ext3_msg(sb, KERN_INFO, "%d truncate%s cleaned up", + PLURAL(nr_truncates)); #ifdef CONFIG_QUOTA /* Turn quotas off */ for (i = 0; i < MAXQUOTAS; i++) { @@ -1554,7 +1589,7 @@ static int ext3_fill_super (struct super_block *sb, void *data, int silent) struct ext3_super_block *es = NULL; struct ext3_sb_info *sbi; ext3_fsblk_t block; - ext3_fsblk_t sb_block = get_sb_block(&data); + ext3_fsblk_t sb_block = get_sb_block(&data, sb); ext3_fsblk_t logic_sb_block; unsigned long offset = 0; unsigned int journal_inum = 0; @@ -1590,7 +1625,7 @@ static int ext3_fill_super (struct super_block *sb, void *data, int silent) blocksize = sb_min_blocksize(sb, EXT3_MIN_BLOCK_SIZE); if (!blocksize) { - printk(KERN_ERR "EXT3-fs: unable to set blocksize\n"); + ext3_msg(sb, KERN_ERR, "error: unable to set blocksize"); goto out_fail; } @@ -1606,7 +1641,7 @@ static int ext3_fill_super (struct super_block *sb, void *data, int silent) } if (!(bh = sb_bread(sb, logic_sb_block))) { - printk (KERN_ERR "EXT3-fs: unable to read superblock\n"); + ext3_msg(sb, KERN_ERR, "error: unable to read superblock"); goto out_fail; } /* @@ -1665,9 +1700,9 @@ 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 - "EXT3-fs warning: feature flags set on rev 0 fs, " - "running e2fsck is recommended\n"); + ext3_msg(sb, KERN_WARNING, + "warning: feature flags set on rev 0 fs, " + "running e2fsck is recommended"); /* * Check feature flags regardless of the revision level, since we * previously didn't change the revision level when setting the flags, @@ -1675,25 +1710,25 @@ static int ext3_fill_super (struct super_block *sb, void *data, int silent) */ features = EXT3_HAS_INCOMPAT_FEATURE(sb, ~EXT3_FEATURE_INCOMPAT_SUPP); if (features) { - printk(KERN_ERR "EXT3-fs: %s: couldn't mount because of " - "unsupported optional features (%x).\n", - sb->s_id, le32_to_cpu(features)); + ext3_msg(sb, KERN_ERR, + "error: couldn't mount because of unsupported " + "optional features (%x)", le32_to_cpu(features)); goto failed_mount; } features = EXT3_HAS_RO_COMPAT_FEATURE(sb, ~EXT3_FEATURE_RO_COMPAT_SUPP); if (!(sb->s_flags & MS_RDONLY) && features) { - printk(KERN_ERR "EXT3-fs: %s: couldn't mount RDWR because of " - "unsupported optional features (%x).\n", - sb->s_id, le32_to_cpu(features)); + ext3_msg(sb, KERN_ERR, + "error: couldn't mount RDWR because of unsupported " + "optional features (%x)", le32_to_cpu(features)); goto failed_mount; } blocksize = BLOCK_SIZE << le32_to_cpu(es->s_log_block_size); if (blocksize < EXT3_MIN_BLOCK_SIZE || blocksize > EXT3_MAX_BLOCK_SIZE) { - printk(KERN_ERR - "EXT3-fs: Unsupported filesystem blocksize %d on %s.\n", - blocksize, sb->s_id); + ext3_msg(sb, KERN_ERR, + "error: couldn't mount because of unsupported " + "filesystem blocksize %d", blocksize); goto failed_mount; } @@ -1704,30 +1739,31 @@ static int ext3_fill_super (struct super_block *sb, void *data, int silent) * than the hardware sectorsize for the machine. */ if (blocksize < hblock) { - printk(KERN_ERR "EXT3-fs: blocksize %d too small for " - "device blocksize %d.\n", blocksize, hblock); + ext3_msg(sb, KERN_ERR, + "error: fsblocksize %d too small for " + "hardware sectorsize %d", blocksize, hblock); goto failed_mount; } brelse (bh); if (!sb_set_blocksize(sb, blocksize)) { - printk(KERN_ERR "EXT3-fs: bad blocksize %d.\n", - blocksize); + ext3_msg(sb, KERN_ERR, + "error: bad blocksize %d", blocksize); goto out_fail; } logic_sb_block = (sb_block * EXT3_MIN_BLOCK_SIZE) / blocksize; offset = (sb_block * EXT3_MIN_BLOCK_SIZE) % blocksize; bh = sb_bread(sb, logic_sb_block); if (!bh) { - printk(KERN_ERR - "EXT3-fs: Can't read superblock on 2nd try.\n"); + ext3_msg(sb, KERN_ERR, + "error: can't read superblock on 2nd try"); 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 - "EXT3-fs: Magic mismatch, very weird !\n"); + ext3_msg(sb, KERN_ERR, + "error: magic mismatch"); goto failed_mount; } } @@ -1743,8 +1779,8 @@ static int ext3_fill_super (struct super_block *sb, void *data, int silent) if ((sbi->s_inode_size < EXT3_GOOD_OLD_INODE_SIZE) || (!is_power_of_2(sbi->s_inode_size)) || (sbi->s_inode_size > blocksize)) { - printk (KERN_ERR - "EXT3-fs: unsupported inode size: %d\n", + ext3_msg(sb, KERN_ERR, + "error: unsupported inode size: %d", sbi->s_inode_size); goto failed_mount; } @@ -1752,8 +1788,8 @@ static int ext3_fill_super (struct super_block *sb, void *data, int silent) sbi->s_frag_size = EXT3_MIN_FRAG_SIZE << le32_to_cpu(es->s_log_frag_size); if (blocksize != sbi->s_frag_size) { - printk(KERN_ERR - "EXT3-fs: fragsize %lu != blocksize %u (unsupported)\n", + ext3_msg(sb, KERN_ERR, + "error: fragsize %lu != blocksize %u (unsupported)", sbi->s_frag_size, blocksize); goto failed_mount; } @@ -1789,31 +1825,31 @@ static int ext3_fill_super (struct super_block *sb, void *data, int silent) } if (sbi->s_blocks_per_group > blocksize * 8) { - printk (KERN_ERR - "EXT3-fs: #blocks per group too big: %lu\n", + ext3_msg(sb, KERN_ERR, + "#blocks per group too big: %lu", sbi->s_blocks_per_group); goto failed_mount; } if (sbi->s_frags_per_group > blocksize * 8) { - printk (KERN_ERR - "EXT3-fs: #fragments per group too big: %lu\n", + ext3_msg(sb, KERN_ERR, + "error: #fragments per group too big: %lu", sbi->s_frags_per_group); goto failed_mount; } if (sbi->s_inodes_per_group > blocksize * 8) { - printk (KERN_ERR - "EXT3-fs: #inodes per group too big: %lu\n", + ext3_msg(sb, KERN_ERR, + "error: #inodes per group too big: %lu", sbi->s_inodes_per_group); 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); + ext3_msg(sb, KERN_ERR, + "error: filesystem is too large to mount safely"); if (sizeof(sector_t) < 8) - printk(KERN_WARNING "EXT3-fs: CONFIG_LBDAF not " - "enabled\n"); + ext3_msg(sb, KERN_ERR, + "error: CONFIG_LBDAF not enabled"); goto failed_mount; } @@ -1827,7 +1863,8 @@ static int ext3_fill_super (struct super_block *sb, void *data, int silent) sbi->s_group_desc = kmalloc(db_count * sizeof (struct buffer_head *), GFP_KERNEL); if (sbi->s_group_desc == NULL) { - printk (KERN_ERR "EXT3-fs: not enough memory\n"); + ext3_msg(sb, KERN_ERR, + "error: not enough memory"); goto failed_mount; } @@ -1837,14 +1874,15 @@ static int ext3_fill_super (struct super_block *sb, void *data, int silent) block = descriptor_loc(sb, logic_sb_block, i); sbi->s_group_desc[i] = sb_bread(sb, block); if (!sbi->s_group_desc[i]) { - printk (KERN_ERR "EXT3-fs: " - "can't read group descriptor %d\n", i); + ext3_msg(sb, KERN_ERR, + "error: can't read group descriptor %d", i); db_count = i; goto failed_mount2; } } if (!ext3_check_descriptors (sb)) { - printk(KERN_ERR "EXT3-fs: group descriptors corrupted!\n"); + ext3_msg(sb, KERN_ERR, + "error: group descriptors corrupted"); goto failed_mount2; } sbi->s_gdb_count = db_count; @@ -1862,7 +1900,7 @@ static int ext3_fill_super (struct super_block *sb, void *data, int silent) ext3_count_dirs(sb)); } if (err) { - printk(KERN_ERR "EXT3-fs: insufficient memory\n"); + ext3_msg(sb, KERN_ERR, "error: insufficient memory"); goto failed_mount3; } @@ -1910,9 +1948,9 @@ static int ext3_fill_super (struct super_block *sb, void *data, int silent) goto failed_mount3; } else { if (!silent) - printk (KERN_ERR - "ext3: No journal on filesystem on %s\n", - sb->s_id); + ext3_msg(sb, KERN_ERR, + "error: no journal found. " + "mounting ext3 over ext2?"); goto failed_mount3; } @@ -1934,8 +1972,9 @@ static int ext3_fill_super (struct super_block *sb, void *data, int silent) case EXT3_MOUNT_WRITEBACK_DATA: if (!journal_check_available_features (sbi->s_journal, 0, 0, JFS_FEATURE_INCOMPAT_REVOKE)) { - printk(KERN_ERR "EXT3-fs: Journal does not support " - "requested data journaling mode\n"); + ext3_msg(sb, KERN_ERR, + "error: journal does not support " + "requested data journaling mode"); goto failed_mount4; } default: @@ -1944,8 +1983,9 @@ static int ext3_fill_super (struct super_block *sb, void *data, int silent) if (test_opt(sb, NOBH)) { if (!(test_opt(sb, DATA_FLAGS) == EXT3_MOUNT_WRITEBACK_DATA)) { - printk(KERN_WARNING "EXT3-fs: Ignoring nobh option - " - "its supported only with writeback mode\n"); + ext3_msg(sb, KERN_WARNING, + "warning: ignoring nobh option - " + "it is supported only with writeback mode"); clear_opt(sbi->s_mount_opt, NOBH); } } @@ -1956,18 +1996,18 @@ static int ext3_fill_super (struct super_block *sb, void *data, int silent) root = ext3_iget(sb, EXT3_ROOT_INO); if (IS_ERR(root)) { - printk(KERN_ERR "EXT3-fs: get root inode failed\n"); + ext3_msg(sb, KERN_ERR, "error: get root inode failed"); ret = PTR_ERR(root); goto failed_mount4; } if (!S_ISDIR(root->i_mode) || !root->i_blocks || !root->i_size) { iput(root); - printk(KERN_ERR "EXT3-fs: corrupt root inode, run e2fsck\n"); + ext3_msg(sb, KERN_ERR, "error: corrupt root inode, run e2fsck"); goto failed_mount4; } sb->s_root = d_alloc_root(root); if (!sb->s_root) { - printk(KERN_ERR "EXT3-fs: get root dentry failed\n"); + ext3_msg(sb, KERN_ERR, "error: get root dentry failed"); iput(root); ret = -ENOMEM; goto failed_mount4; @@ -1986,9 +2026,9 @@ static int ext3_fill_super (struct super_block *sb, void *data, int silent) ext3_orphan_cleanup(sb, es); EXT3_SB(sb)->s_mount_state &= ~EXT3_ORPHAN_FS; if (needs_recovery) - printk (KERN_INFO "EXT3-fs: recovery complete.\n"); + ext3_msg(sb, KERN_INFO, "recovery complete"); ext3_mark_recovery_complete(sb, es); - printk (KERN_INFO "EXT3-fs: mounted filesystem with %s data mode.\n", + ext3_msg(sb, KERN_INFO, "mounted filesystem with %s data mode", test_opt(sb,DATA_FLAGS) == EXT3_MOUNT_JOURNAL_DATA ? "journal": test_opt(sb,DATA_FLAGS) == EXT3_MOUNT_ORDERED_DATA ? "ordered": "writeback"); @@ -1998,7 +2038,8 @@ static int ext3_fill_super (struct super_block *sb, void *data, int silent) cantfind_ext3: if (!silent) - printk(KERN_ERR "VFS: Can't find ext3 filesystem on dev %s.\n", + ext3_msg(sb, KERN_INFO, + "error: can't find ext3 filesystem on dev %s.", sb->s_id); goto failed_mount; @@ -2066,27 +2107,27 @@ static journal_t *ext3_get_journal(struct super_block *sb, journal_inode = ext3_iget(sb, journal_inum); if (IS_ERR(journal_inode)) { - printk(KERN_ERR "EXT3-fs: no journal found.\n"); + ext3_msg(sb, KERN_ERR, "error: no journal found"); return NULL; } if (!journal_inode->i_nlink) { make_bad_inode(journal_inode); iput(journal_inode); - printk(KERN_ERR "EXT3-fs: journal inode is deleted.\n"); + ext3_msg(sb, KERN_ERR, "error: journal inode is deleted"); return NULL; } jbd_debug(2, "Journal inode found at %p: %Ld bytes\n", journal_inode, journal_inode->i_size); if (!S_ISREG(journal_inode->i_mode)) { - printk(KERN_ERR "EXT3-fs: invalid journal inode.\n"); + ext3_msg(sb, KERN_ERR, "error: invalid journal inode"); iput(journal_inode); return NULL; } journal = journal_init_inode(journal_inode); if (!journal) { - printk(KERN_ERR "EXT3-fs: Could not load journal inode\n"); + ext3_msg(sb, KERN_ERR, "error: could not load journal inode"); iput(journal_inode); return NULL; } @@ -2108,13 +2149,13 @@ static journal_t *ext3_get_dev_journal(struct super_block *sb, struct ext3_super_block * es; struct block_device *bdev; - bdev = ext3_blkdev_get(j_dev); + bdev = ext3_blkdev_get(j_dev, sb); if (bdev == NULL) return NULL; if (bd_claim(bdev, sb)) { - printk(KERN_ERR - "EXT3: failed to claim external journal device.\n"); + ext3_msg(sb, KERN_ERR, + "error: failed to claim external journal device"); blkdev_put(bdev, FMODE_READ|FMODE_WRITE); return NULL; } @@ -2122,8 +2163,8 @@ static journal_t *ext3_get_dev_journal(struct super_block *sb, blocksize = sb->s_blocksize; hblock = bdev_logical_block_size(bdev); if (blocksize < hblock) { - printk(KERN_ERR - "EXT3-fs: blocksize too small for journal device.\n"); + ext3_msg(sb, KERN_ERR, + "error: blocksize too small for journal device"); goto out_bdev; } @@ -2131,8 +2172,8 @@ static journal_t *ext3_get_dev_journal(struct super_block *sb, offset = EXT3_MIN_BLOCK_SIZE % blocksize; set_blocksize(bdev, blocksize); if (!(bh = __bread(bdev, sb_block, blocksize))) { - printk(KERN_ERR "EXT3-fs: couldn't read superblock of " - "external journal\n"); + ext3_msg(sb, KERN_ERR, "error: couldn't read superblock of " + "external journal"); goto out_bdev; } @@ -2140,14 +2181,14 @@ static journal_t *ext3_get_dev_journal(struct super_block *sb, if ((le16_to_cpu(es->s_magic) != EXT3_SUPER_MAGIC) || !(le32_to_cpu(es->s_feature_incompat) & EXT3_FEATURE_INCOMPAT_JOURNAL_DEV)) { - printk(KERN_ERR "EXT3-fs: external journal has " - "bad superblock\n"); + ext3_msg(sb, KERN_ERR, "error: external journal has " + "bad superblock"); brelse(bh); goto out_bdev; } if (memcmp(EXT3_SB(sb)->s_es->s_journal_uuid, es->s_uuid, 16)) { - printk(KERN_ERR "EXT3-fs: journal UUID does not match\n"); + ext3_msg(sb, KERN_ERR, "error: journal UUID does not match"); brelse(bh); goto out_bdev; } @@ -2159,19 +2200,21 @@ static journal_t *ext3_get_dev_journal(struct super_block *sb, journal = journal_init_dev(bdev, sb->s_bdev, start, len, blocksize); if (!journal) { - printk(KERN_ERR "EXT3-fs: failed to create device journal\n"); + ext3_msg(sb, KERN_ERR, + "error: failed to create device journal"); goto out_bdev; } journal->j_private = sb; ll_rw_block(READ, 1, &journal->j_sb_buffer); wait_on_buffer(journal->j_sb_buffer); if (!buffer_uptodate(journal->j_sb_buffer)) { - printk(KERN_ERR "EXT3-fs: I/O error on journal device\n"); + ext3_msg(sb, KERN_ERR, "I/O error on journal device"); goto out_journal; } if (be32_to_cpu(journal->j_superblock->s_nr_users) != 1) { - printk(KERN_ERR "EXT3-fs: External journal has more than one " - "user (unsupported) - %d\n", + ext3_msg(sb, KERN_ERR, + "error: external journal has more than one " + "user (unsupported) - %d", be32_to_cpu(journal->j_superblock->s_nr_users)); goto out_journal; } @@ -2197,8 +2240,8 @@ static int ext3_load_journal(struct super_block *sb, if (journal_devnum && journal_devnum != le32_to_cpu(es->s_journal_dev)) { - printk(KERN_INFO "EXT3-fs: external journal device major/minor " - "numbers have changed\n"); + ext3_msg(sb, KERN_INFO, "external journal device major/minor " + "numbers have changed"); journal_dev = new_decode_dev(journal_devnum); } else journal_dev = new_decode_dev(le32_to_cpu(es->s_journal_dev)); @@ -2213,21 +2256,21 @@ static int ext3_load_journal(struct super_block *sb, if (EXT3_HAS_INCOMPAT_FEATURE(sb, EXT3_FEATURE_INCOMPAT_RECOVER)) { if (sb->s_flags & MS_RDONLY) { - printk(KERN_INFO "EXT3-fs: INFO: recovery " - "required on readonly filesystem.\n"); + ext3_msg(sb, KERN_INFO, + "recovery required on readonly filesystem"); if (really_read_only) { - printk(KERN_ERR "EXT3-fs: write access " - "unavailable, cannot proceed.\n"); + ext3_msg(sb, KERN_ERR, "error: write access " + "unavailable, cannot proceed"); return -EROFS; } - printk (KERN_INFO "EXT3-fs: write access will " - "be enabled during recovery.\n"); + ext3_msg(sb, KERN_INFO, + "write access will be enabled during recovery"); } } if (journal_inum && journal_dev) { - printk(KERN_ERR "EXT3-fs: filesystem has both journal " - "and inode journals!\n"); + ext3_msg(sb, KERN_ERR, "error: filesystem has both journal " + "and inode journals"); return -EINVAL; } @@ -2242,7 +2285,7 @@ static int ext3_load_journal(struct super_block *sb, if (!really_read_only && test_opt(sb, UPDATE_JOURNAL)) { err = journal_update_format(journal); if (err) { - printk(KERN_ERR "EXT3-fs: error updating journal.\n"); + ext3_msg(sb, KERN_ERR, "error updating journal"); journal_destroy(journal); return err; } @@ -2254,7 +2297,7 @@ static int ext3_load_journal(struct super_block *sb, err = journal_load(journal); if (err) { - printk(KERN_ERR "EXT3-fs: error loading journal.\n"); + ext3_msg(sb, KERN_ERR, "error loading journal"); journal_destroy(journal); return err; } @@ -2273,16 +2316,17 @@ static int ext3_load_journal(struct super_block *sb, return 0; } -static int ext3_create_journal(struct super_block * sb, - struct ext3_super_block * es, +static int ext3_create_journal(struct super_block *sb, + struct ext3_super_block *es, unsigned int journal_inum) { journal_t *journal; int err; if (sb->s_flags & MS_RDONLY) { - printk(KERN_ERR "EXT3-fs: readonly filesystem when trying to " - "create journal.\n"); + ext3_msg(sb, KERN_ERR, + "error: readonly filesystem when trying to " + "create journal"); return -EROFS; } @@ -2290,12 +2334,12 @@ static int ext3_create_journal(struct super_block * sb, if (!journal) return -EINVAL; - printk(KERN_INFO "EXT3-fs: creating new journal on inode %u\n", + ext3_msg(sb, KERN_INFO, "creating new journal on inode %u", journal_inum); err = journal_create(journal); if (err) { - printk(KERN_ERR "EXT3-fs: error creating journal.\n"); + ext3_msg(sb, KERN_ERR, "error creating journal"); journal_destroy(journal); return -EIO; } @@ -2376,8 +2420,8 @@ out: * has recorded an error from a previous lifetime, move that error to the * main filesystem now. */ -static void ext3_clear_journal_err(struct super_block * sb, - struct ext3_super_block * es) +static void ext3_clear_journal_err(struct super_block *sb, + struct ext3_super_block *es) { journal_t *journal; int j_errno; @@ -2568,10 +2612,10 @@ static int ext3_remount (struct super_block * sb, int * flags, char * data) __le32 ret; if ((ret = EXT3_HAS_RO_COMPAT_FEATURE(sb, ~EXT3_FEATURE_RO_COMPAT_SUPP))) { - printk(KERN_WARNING "EXT3-fs: %s: couldn't " - "remount RDWR because of unsupported " - "optional features (%x).\n", - sb->s_id, le32_to_cpu(ret)); + ext3_msg(sb, KERN_WARNING, + "warning: couldn't remount RDWR " + "because of unsupported optional " + "features (%x)", le32_to_cpu(ret)); err = -EROFS; goto restore_opts; } @@ -2582,11 +2626,10 @@ static int ext3_remount (struct super_block * sb, int * flags, char * data) * require a full umount/remount for now. */ if (es->s_last_orphan) { - printk(KERN_WARNING "EXT3-fs: %s: couldn't " + ext3_msg(sb, KERN_WARNING, "warning: couldn't " "remount RDWR because of unprocessed " "orphan inode list. Please " - "umount/remount instead.\n", - sb->s_id); + "umount/remount instead."); err = -EINVAL; goto restore_opts; } @@ -2686,13 +2729,11 @@ static int ext3_statfs (struct dentry * dentry, struct kstatfs * buf) buf->f_bsize = sb->s_blocksize; buf->f_blocks = le32_to_cpu(es->s_blocks_count) - sbi->s_overhead_last; buf->f_bfree = percpu_counter_sum_positive(&sbi->s_freeblocks_counter); - es->s_free_blocks_count = cpu_to_le32(buf->f_bfree); buf->f_bavail = buf->f_bfree - le32_to_cpu(es->s_r_blocks_count); if (buf->f_bfree < le32_to_cpu(es->s_r_blocks_count)) buf->f_bavail = 0; buf->f_files = le32_to_cpu(es->s_inodes_count); buf->f_ffree = percpu_counter_sum_positive(&sbi->s_freeinodes_counter); - es->s_free_inodes_count = cpu_to_le32(buf->f_ffree); buf->f_namelen = EXT3_NAME_LEN; fsid = le64_to_cpup((void *)es->s_uuid) ^ le64_to_cpup((void *)es->s_uuid + sizeof(u64)); @@ -2837,9 +2878,9 @@ static int ext3_quota_on(struct super_block *sb, int type, int format_id, if (EXT3_SB(sb)->s_qf_names[type]) { /* Quotafile not of fs root? */ if (path.dentry->d_parent != sb->s_root) - printk(KERN_WARNING - "EXT3-fs: Quota file not on filesystem root. " - "Journaled quota will not work.\n"); + ext3_msg(sb, KERN_WARNING, + "warning: Quota file not on filesystem root. " + "Journaled quota will not work."); } /* @@ -2921,8 +2962,9 @@ static ssize_t ext3_quota_write(struct super_block *sb, int type, handle_t *handle = journal_current_handle(); if (!handle) { - printk(KERN_WARNING "EXT3-fs: Quota write (off=%Lu, len=%Lu)" - " cancelled because transaction is not started.\n", + ext3_msg(sb, KERN_WARNING, + "warning: quota write (off=%llu, len=%llu)" + " cancelled because transaction is not started.", (unsigned long long)off, (unsigned long long)len); return -EIO; } diff --git a/fs/ext3/xattr.c b/fs/ext3/xattr.c index 545e37c4b91..387d92d00b9 100644 --- a/fs/ext3/xattr.c +++ b/fs/ext3/xattr.c @@ -960,6 +960,10 @@ ext3_xattr_set_handle(handle_t *handle, struct inode *inode, int name_index, if (error) goto cleanup; + error = ext3_journal_get_write_access(handle, is.iloc.bh); + if (error) + goto cleanup; + if (EXT3_I(inode)->i_state & EXT3_STATE_NEW) { struct ext3_inode *raw_inode = ext3_raw_inode(&is.iloc); memset(raw_inode, 0, EXT3_SB(inode->i_sb)->s_inode_size); @@ -985,9 +989,6 @@ ext3_xattr_set_handle(handle_t *handle, struct inode *inode, int name_index, if (flags & XATTR_CREATE) goto cleanup; } - error = ext3_journal_get_write_access(handle, is.iloc.bh); - if (error) - goto cleanup; if (!value) { if (!is.s.not_found) error = ext3_xattr_ibody_set(handle, inode, &i, &is); diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 8b58a144c31..768c111a77e 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -769,9 +769,22 @@ static inline void ext4_show_quota_options(struct seq_file *seq, #if defined(CONFIG_QUOTA) struct ext4_sb_info *sbi = EXT4_SB(sb); - if (sbi->s_jquota_fmt) - seq_printf(seq, ",jqfmt=%s", - (sbi->s_jquota_fmt == QFMT_VFS_OLD) ? "vfsold" : "vfsv0"); + if (sbi->s_jquota_fmt) { + char *fmtname = ""; + + switch (sbi->s_jquota_fmt) { + case QFMT_VFS_OLD: + fmtname = "vfsold"; + break; + case QFMT_VFS_V0: + fmtname = "vfsv0"; + break; + case QFMT_VFS_V1: + fmtname = "vfsv1"; + break; + } + seq_printf(seq, ",jqfmt=%s", fmtname); + } if (sbi->s_qf_names[USRQUOTA]) seq_printf(seq, ",usrjquota=%s", sbi->s_qf_names[USRQUOTA]); @@ -1084,9 +1097,9 @@ enum { Opt_abort, Opt_data_journal, Opt_data_ordered, Opt_data_writeback, Opt_data_err_abort, Opt_data_err_ignore, Opt_usrjquota, Opt_grpjquota, Opt_offusrjquota, Opt_offgrpjquota, - Opt_jqfmt_vfsold, Opt_jqfmt_vfsv0, Opt_quota, Opt_noquota, - Opt_ignore, Opt_barrier, Opt_nobarrier, Opt_err, Opt_resize, - Opt_usrquota, Opt_grpquota, Opt_i_version, + Opt_jqfmt_vfsold, Opt_jqfmt_vfsv0, Opt_jqfmt_vfsv1, Opt_quota, + Opt_noquota, Opt_ignore, Opt_barrier, Opt_nobarrier, Opt_err, + Opt_resize, Opt_usrquota, Opt_grpquota, Opt_i_version, Opt_stripe, Opt_delalloc, Opt_nodelalloc, Opt_block_validity, Opt_noblock_validity, Opt_inode_readahead_blks, Opt_journal_ioprio, @@ -1137,6 +1150,7 @@ static const match_table_t tokens = { {Opt_grpjquota, "grpjquota=%s"}, {Opt_jqfmt_vfsold, "jqfmt=vfsold"}, {Opt_jqfmt_vfsv0, "jqfmt=vfsv0"}, + {Opt_jqfmt_vfsv1, "jqfmt=vfsv1"}, {Opt_grpquota, "grpquota"}, {Opt_noquota, "noquota"}, {Opt_quota, "quota"}, @@ -1439,6 +1453,9 @@ clear_qf_name: goto set_qf_format; case Opt_jqfmt_vfsv0: qfmt = QFMT_VFS_V0; + goto set_qf_format; + case Opt_jqfmt_vfsv1: + qfmt = QFMT_VFS_V1; set_qf_format: if (sb_any_quota_loaded(sb) && sbi->s_jquota_fmt != qfmt) { @@ -1481,6 +1498,7 @@ set_qf_format: case Opt_offgrpjquota: case Opt_jqfmt_vfsold: case Opt_jqfmt_vfsv0: + case Opt_jqfmt_vfsv1: ext4_msg(sb, KERN_ERR, "journaled quota options not supported"); break; diff --git a/fs/isofs/compress.c b/fs/isofs/compress.c index defb932eee9..0b3fa7974fa 100644 --- a/fs/isofs/compress.c +++ b/fs/isofs/compress.c @@ -36,286 +36,323 @@ static void *zisofs_zlib_workspace; static DEFINE_MUTEX(zisofs_zlib_lock); /* - * When decompressing, we typically obtain more than one page - * per reference. We inject the additional pages into the page - * cache as a form of readahead. + * Read data of @inode from @block_start to @block_end and uncompress + * to one zisofs block. Store the data in the @pages array with @pcount + * entries. Start storing at offset @poffset of the first page. */ -static int zisofs_readpage(struct file *file, struct page *page) +static loff_t zisofs_uncompress_block(struct inode *inode, loff_t block_start, + loff_t block_end, int pcount, + struct page **pages, unsigned poffset, + int *errp) { - struct inode *inode = file->f_path.dentry->d_inode; - struct address_space *mapping = inode->i_mapping; - unsigned int maxpage, xpage, fpage, blockindex; - unsigned long offset; - unsigned long blockptr, blockendptr, cstart, cend, csize; - struct buffer_head *bh, *ptrbh[2]; - unsigned long bufsize = ISOFS_BUFFER_SIZE(inode); - unsigned int bufshift = ISOFS_BUFFER_BITS(inode); - unsigned long bufmask = bufsize - 1; - int err = -EIO; - int i; - unsigned int header_size = ISOFS_I(inode)->i_format_parm[0]; unsigned int zisofs_block_shift = ISOFS_I(inode)->i_format_parm[1]; - /* unsigned long zisofs_block_size = 1UL << zisofs_block_shift; */ - unsigned int zisofs_block_page_shift = zisofs_block_shift-PAGE_CACHE_SHIFT; - unsigned long zisofs_block_pages = 1UL << zisofs_block_page_shift; - unsigned long zisofs_block_page_mask = zisofs_block_pages-1; - struct page *pages[zisofs_block_pages]; - unsigned long index = page->index; - int indexblocks; - - /* We have already been given one page, this is the one - we must do. */ - xpage = index & zisofs_block_page_mask; - pages[xpage] = page; - - /* The remaining pages need to be allocated and inserted */ - offset = index & ~zisofs_block_page_mask; - blockindex = offset >> zisofs_block_page_shift; - maxpage = (inode->i_size + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT; - - /* - * If this page is wholly outside i_size we just return zero; - * do_generic_file_read() will handle this for us - */ - if (page->index >= maxpage) { - SetPageUptodate(page); - unlock_page(page); + unsigned int bufsize = ISOFS_BUFFER_SIZE(inode); + unsigned int bufshift = ISOFS_BUFFER_BITS(inode); + unsigned int bufmask = bufsize - 1; + int i, block_size = block_end - block_start; + z_stream stream = { .total_out = 0, + .avail_in = 0, + .avail_out = 0, }; + int zerr; + int needblocks = (block_size + (block_start & bufmask) + bufmask) + >> bufshift; + int haveblocks; + blkcnt_t blocknum; + struct buffer_head *bhs[needblocks + 1]; + int curbh, curpage; + + if (block_size > deflateBound(1UL << zisofs_block_shift)) { + *errp = -EIO; return 0; } - - maxpage = min(zisofs_block_pages, maxpage-offset); - - for ( i = 0 ; i < maxpage ; i++, offset++ ) { - if ( i != xpage ) { - pages[i] = grab_cache_page_nowait(mapping, offset); - } - page = pages[i]; - if ( page ) { - ClearPageError(page); - kmap(page); + /* Empty block? */ + if (block_size == 0) { + for ( i = 0 ; i < pcount ; i++ ) { + if (!pages[i]) + continue; + memset(page_address(pages[i]), 0, PAGE_CACHE_SIZE); + flush_dcache_page(pages[i]); + SetPageUptodate(pages[i]); } + return ((loff_t)pcount) << PAGE_CACHE_SHIFT; } - /* This is the last page filled, plus one; used in case of abort. */ - fpage = 0; + /* Because zlib is not thread-safe, do all the I/O at the top. */ + blocknum = block_start >> bufshift; + memset(bhs, 0, (needblocks + 1) * sizeof(struct buffer_head *)); + haveblocks = isofs_get_blocks(inode, blocknum, bhs, needblocks); + ll_rw_block(READ, haveblocks, bhs); - /* Find the pointer to this specific chunk */ - /* Note: we're not using isonum_731() here because the data is known aligned */ - /* Note: header_size is in 32-bit words (4 bytes) */ - blockptr = (header_size + blockindex) << 2; - blockendptr = blockptr + 4; + curbh = 0; + curpage = 0; + /* + * First block is special since it may be fractional. We also wait for + * it before grabbing the zlib mutex; odds are that the subsequent + * blocks are going to come in in short order so we don't hold the zlib + * mutex longer than necessary. + */ - indexblocks = ((blockptr^blockendptr) >> bufshift) ? 2 : 1; - ptrbh[0] = ptrbh[1] = NULL; + if (!bhs[0]) + goto b_eio; - if ( isofs_get_blocks(inode, blockptr >> bufshift, ptrbh, indexblocks) != indexblocks ) { - if ( ptrbh[0] ) brelse(ptrbh[0]); - printk(KERN_DEBUG "zisofs: Null buffer on reading block table, inode = %lu, block = %lu\n", - inode->i_ino, blockptr >> bufshift); - goto eio; - } - ll_rw_block(READ, indexblocks, ptrbh); - - bh = ptrbh[0]; - if ( !bh || (wait_on_buffer(bh), !buffer_uptodate(bh)) ) { - printk(KERN_DEBUG "zisofs: Failed to read block table, inode = %lu, block = %lu\n", - inode->i_ino, blockptr >> bufshift); - if ( ptrbh[1] ) - brelse(ptrbh[1]); - goto eio; - } - cstart = le32_to_cpu(*(__le32 *)(bh->b_data + (blockptr & bufmask))); - - if ( indexblocks == 2 ) { - /* We just crossed a block boundary. Switch to the next block */ - brelse(bh); - bh = ptrbh[1]; - if ( !bh || (wait_on_buffer(bh), !buffer_uptodate(bh)) ) { - printk(KERN_DEBUG "zisofs: Failed to read block table, inode = %lu, block = %lu\n", - inode->i_ino, blockendptr >> bufshift); - goto eio; - } + wait_on_buffer(bhs[0]); + if (!buffer_uptodate(bhs[0])) { + *errp = -EIO; + goto b_eio; } - cend = le32_to_cpu(*(__le32 *)(bh->b_data + (blockendptr & bufmask))); - brelse(bh); - if (cstart > cend) - goto eio; + stream.workspace = zisofs_zlib_workspace; + mutex_lock(&zisofs_zlib_lock); - csize = cend-cstart; - - if (csize > deflateBound(1UL << zisofs_block_shift)) - goto eio; - - /* Now page[] contains an array of pages, any of which can be NULL, - and the locks on which we hold. We should now read the data and - release the pages. If the pages are NULL the decompressed data - for that particular page should be discarded. */ - - if ( csize == 0 ) { - /* This data block is empty. */ - - for ( fpage = 0 ; fpage < maxpage ; fpage++ ) { - if ( (page = pages[fpage]) != NULL ) { - memset(page_address(page), 0, PAGE_CACHE_SIZE); - - flush_dcache_page(page); - SetPageUptodate(page); - kunmap(page); - unlock_page(page); - if ( fpage == xpage ) - err = 0; /* The critical page */ - else - page_cache_release(page); + zerr = zlib_inflateInit(&stream); + if (zerr != Z_OK) { + if (zerr == Z_MEM_ERROR) + *errp = -ENOMEM; + else + *errp = -EIO; + printk(KERN_DEBUG "zisofs: zisofs_inflateInit returned %d\n", + zerr); + goto z_eio; + } + + while (curpage < pcount && curbh < haveblocks && + zerr != Z_STREAM_END) { + if (!stream.avail_out) { + if (pages[curpage]) { + stream.next_out = page_address(pages[curpage]) + + poffset; + stream.avail_out = PAGE_CACHE_SIZE - poffset; + poffset = 0; + } else { + stream.next_out = (void *)&zisofs_sink_page; + stream.avail_out = PAGE_CACHE_SIZE; } } - } else { - /* This data block is compressed. */ - z_stream stream; - int bail = 0, left_out = -1; - int zerr; - int needblocks = (csize + (cstart & bufmask) + bufmask) >> bufshift; - int haveblocks; - struct buffer_head *bhs[needblocks+1]; - struct buffer_head **bhptr; - - /* Because zlib is not thread-safe, do all the I/O at the top. */ - - blockptr = cstart >> bufshift; - memset(bhs, 0, (needblocks+1)*sizeof(struct buffer_head *)); - haveblocks = isofs_get_blocks(inode, blockptr, bhs, needblocks); - ll_rw_block(READ, haveblocks, bhs); - - bhptr = &bhs[0]; - bh = *bhptr++; - - /* First block is special since it may be fractional. - We also wait for it before grabbing the zlib - mutex; odds are that the subsequent blocks are - going to come in in short order so we don't hold - the zlib mutex longer than necessary. */ - - if ( !bh || (wait_on_buffer(bh), !buffer_uptodate(bh)) ) { - printk(KERN_DEBUG "zisofs: Hit null buffer, fpage = %d, xpage = %d, csize = %ld\n", - fpage, xpage, csize); - goto b_eio; - } - stream.next_in = bh->b_data + (cstart & bufmask); - stream.avail_in = min(bufsize-(cstart & bufmask), csize); - csize -= stream.avail_in; - - stream.workspace = zisofs_zlib_workspace; - mutex_lock(&zisofs_zlib_lock); - - zerr = zlib_inflateInit(&stream); - if ( zerr != Z_OK ) { - if ( err && zerr == Z_MEM_ERROR ) - err = -ENOMEM; - printk(KERN_DEBUG "zisofs: zisofs_inflateInit returned %d\n", - zerr); - goto z_eio; + if (!stream.avail_in) { + wait_on_buffer(bhs[curbh]); + if (!buffer_uptodate(bhs[curbh])) { + *errp = -EIO; + break; + } + stream.next_in = bhs[curbh]->b_data + + (block_start & bufmask); + stream.avail_in = min_t(unsigned, bufsize - + (block_start & bufmask), + block_size); + block_size -= stream.avail_in; + block_start = 0; } - while ( !bail && fpage < maxpage ) { - page = pages[fpage]; - if ( page ) - stream.next_out = page_address(page); - else - stream.next_out = (void *)&zisofs_sink_page; - stream.avail_out = PAGE_CACHE_SIZE; - - while ( stream.avail_out ) { - int ao, ai; - if ( stream.avail_in == 0 && left_out ) { - if ( !csize ) { - printk(KERN_WARNING "zisofs: ZF read beyond end of input\n"); - bail = 1; - break; - } else { - bh = *bhptr++; - if ( !bh || - (wait_on_buffer(bh), !buffer_uptodate(bh)) ) { - /* Reached an EIO */ - printk(KERN_DEBUG "zisofs: Hit null buffer, fpage = %d, xpage = %d, csize = %ld\n", - fpage, xpage, csize); - - bail = 1; - break; - } - stream.next_in = bh->b_data; - stream.avail_in = min(csize,bufsize); - csize -= stream.avail_in; - } - } - ao = stream.avail_out; ai = stream.avail_in; - zerr = zlib_inflate(&stream, Z_SYNC_FLUSH); - left_out = stream.avail_out; - if ( zerr == Z_BUF_ERROR && stream.avail_in == 0 ) - continue; - if ( zerr != Z_OK ) { - /* EOF, error, or trying to read beyond end of input */ - if ( err && zerr == Z_MEM_ERROR ) - err = -ENOMEM; - if ( zerr != Z_STREAM_END ) - printk(KERN_DEBUG "zisofs: zisofs_inflate returned %d, inode = %lu, index = %lu, fpage = %d, xpage = %d, avail_in = %d, avail_out = %d, ai = %d, ao = %d\n", - zerr, inode->i_ino, index, - fpage, xpage, - stream.avail_in, stream.avail_out, - ai, ao); - bail = 1; - break; + while (stream.avail_out && stream.avail_in) { + zerr = zlib_inflate(&stream, Z_SYNC_FLUSH); + if (zerr == Z_BUF_ERROR && stream.avail_in == 0) + break; + if (zerr == Z_STREAM_END) + break; + if (zerr != Z_OK) { + /* EOF, error, or trying to read beyond end of input */ + if (zerr == Z_MEM_ERROR) + *errp = -ENOMEM; + else { + printk(KERN_DEBUG + "zisofs: zisofs_inflate returned" + " %d, inode = %lu," + " page idx = %d, bh idx = %d," + " avail_in = %d," + " avail_out = %d\n", + zerr, inode->i_ino, curpage, + curbh, stream.avail_in, + stream.avail_out); + *errp = -EIO; } + goto inflate_out; } + } - if ( stream.avail_out && zerr == Z_STREAM_END ) { - /* Fractional page written before EOF. This may - be the last page in the file. */ - memset(stream.next_out, 0, stream.avail_out); - stream.avail_out = 0; + if (!stream.avail_out) { + /* This page completed */ + if (pages[curpage]) { + flush_dcache_page(pages[curpage]); + SetPageUptodate(pages[curpage]); } + curpage++; + } + if (!stream.avail_in) + curbh++; + } +inflate_out: + zlib_inflateEnd(&stream); - if ( !stream.avail_out ) { - /* This page completed */ - if ( page ) { - flush_dcache_page(page); - SetPageUptodate(page); - kunmap(page); - unlock_page(page); - if ( fpage == xpage ) - err = 0; /* The critical page */ - else - page_cache_release(page); - } - fpage++; - } +z_eio: + mutex_unlock(&zisofs_zlib_lock); + +b_eio: + for (i = 0; i < haveblocks; i++) + brelse(bhs[i]); + return stream.total_out; +} + +/* + * Uncompress data so that pages[full_page] is fully uptodate and possibly + * fills in other pages if we have data for them. + */ +static int zisofs_fill_pages(struct inode *inode, int full_page, int pcount, + struct page **pages) +{ + loff_t start_off, end_off; + loff_t block_start, block_end; + unsigned int header_size = ISOFS_I(inode)->i_format_parm[0]; + unsigned int zisofs_block_shift = ISOFS_I(inode)->i_format_parm[1]; + unsigned int blockptr; + loff_t poffset = 0; + blkcnt_t cstart_block, cend_block; + struct buffer_head *bh; + unsigned int blkbits = ISOFS_BUFFER_BITS(inode); + unsigned int blksize = 1 << blkbits; + int err; + loff_t ret; + + BUG_ON(!pages[full_page]); + + /* + * We want to read at least 'full_page' page. Because we have to + * uncompress the whole compression block anyway, fill the surrounding + * pages with the data we have anyway... + */ + start_off = page_offset(pages[full_page]); + end_off = min_t(loff_t, start_off + PAGE_CACHE_SIZE, inode->i_size); + + cstart_block = start_off >> zisofs_block_shift; + cend_block = (end_off + (1 << zisofs_block_shift) - 1) + >> zisofs_block_shift; + + WARN_ON(start_off - (full_page << PAGE_CACHE_SHIFT) != + ((cstart_block << zisofs_block_shift) & PAGE_CACHE_MASK)); + + /* Find the pointer to this specific chunk */ + /* Note: we're not using isonum_731() here because the data is known aligned */ + /* Note: header_size is in 32-bit words (4 bytes) */ + blockptr = (header_size + cstart_block) << 2; + bh = isofs_bread(inode, blockptr >> blkbits); + if (!bh) + return -EIO; + block_start = le32_to_cpu(*(__le32 *) + (bh->b_data + (blockptr & (blksize - 1)))); + + while (cstart_block < cend_block && pcount > 0) { + /* Load end of the compressed block in the file */ + blockptr += 4; + /* Traversed to next block? */ + if (!(blockptr & (blksize - 1))) { + brelse(bh); + + bh = isofs_bread(inode, blockptr >> blkbits); + if (!bh) + return -EIO; + } + block_end = le32_to_cpu(*(__le32 *) + (bh->b_data + (blockptr & (blksize - 1)))); + if (block_start > block_end) { + brelse(bh); + return -EIO; + } + err = 0; + ret = zisofs_uncompress_block(inode, block_start, block_end, + pcount, pages, poffset, &err); + poffset += ret; + pages += poffset >> PAGE_CACHE_SHIFT; + pcount -= poffset >> PAGE_CACHE_SHIFT; + full_page -= poffset >> PAGE_CACHE_SHIFT; + poffset &= ~PAGE_CACHE_MASK; + + if (err) { + brelse(bh); + /* + * Did we finish reading the page we really wanted + * to read? + */ + if (full_page < 0) + return 0; + return err; } - zlib_inflateEnd(&stream); - z_eio: - mutex_unlock(&zisofs_zlib_lock); + block_start = block_end; + cstart_block++; + } + + if (poffset && *pages) { + memset(page_address(*pages) + poffset, 0, + PAGE_CACHE_SIZE - poffset); + flush_dcache_page(*pages); + SetPageUptodate(*pages); + } + return 0; +} - b_eio: - for ( i = 0 ; i < haveblocks ; i++ ) { - if ( bhs[i] ) - brelse(bhs[i]); +/* + * When decompressing, we typically obtain more than one page + * per reference. We inject the additional pages into the page + * cache as a form of readahead. + */ +static int zisofs_readpage(struct file *file, struct page *page) +{ + struct inode *inode = file->f_path.dentry->d_inode; + struct address_space *mapping = inode->i_mapping; + int err; + int i, pcount, full_page; + unsigned int zisofs_block_shift = ISOFS_I(inode)->i_format_parm[1]; + unsigned int zisofs_pages_per_cblock = + PAGE_CACHE_SHIFT <= zisofs_block_shift ? + (1 << (zisofs_block_shift - PAGE_CACHE_SHIFT)) : 0; + struct page *pages[max_t(unsigned, zisofs_pages_per_cblock, 1)]; + pgoff_t index = page->index, end_index; + + end_index = (inode->i_size + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT; + /* + * If this page is wholly outside i_size we just return zero; + * do_generic_file_read() will handle this for us + */ + if (index >= end_index) { + SetPageUptodate(page); + unlock_page(page); + return 0; + } + + if (PAGE_CACHE_SHIFT <= zisofs_block_shift) { + /* We have already been given one page, this is the one + we must do. */ + full_page = index & (zisofs_pages_per_cblock - 1); + pcount = min_t(int, zisofs_pages_per_cblock, + end_index - (index & ~(zisofs_pages_per_cblock - 1))); + index -= full_page; + } else { + full_page = 0; + pcount = 1; + } + pages[full_page] = page; + + for (i = 0; i < pcount; i++, index++) { + if (i != full_page) + pages[i] = grab_cache_page_nowait(mapping, index); + if (pages[i]) { + ClearPageError(pages[i]); + kmap(pages[i]); } } -eio: + err = zisofs_fill_pages(inode, full_page, pcount, pages); /* Release any residual pages, do not SetPageUptodate */ - while ( fpage < maxpage ) { - page = pages[fpage]; - if ( page ) { - flush_dcache_page(page); - if ( fpage == xpage ) - SetPageError(page); - kunmap(page); - unlock_page(page); - if ( fpage != xpage ) - page_cache_release(page); + for (i = 0; i < pcount; i++) { + if (pages[i]) { + flush_dcache_page(pages[i]); + if (i == full_page && err) + SetPageError(pages[i]); + kunmap(pages[i]); + unlock_page(pages[i]); + if (i != full_page) + page_cache_release(pages[i]); } - fpage++; } /* At this point, err contains 0 or -EIO depending on the "critical" page */ diff --git a/fs/isofs/rock.c b/fs/isofs/rock.c index c2fb2dd0131..96a685c550f 100644 --- a/fs/isofs/rock.c +++ b/fs/isofs/rock.c @@ -518,8 +518,7 @@ repeat: if (algo == SIG('p', 'z')) { int block_shift = isonum_711(&rr->u.ZF.parms[1]); - if (block_shift < PAGE_CACHE_SHIFT - || block_shift > 17) { + if (block_shift > 17) { printk(KERN_WARNING "isofs: " "Can't handle ZF block " "size of 2^%d\n", diff --git a/fs/jbd2/commit.c b/fs/jbd2/commit.c index 8896c1d4feb..6a10238d2c6 100644 --- a/fs/jbd2/commit.c +++ b/fs/jbd2/commit.c @@ -286,7 +286,7 @@ static int journal_finish_inode_data_buffers(journal_t *journal, if (err) { /* * Because AS_EIO is cleared by - * wait_on_page_writeback_range(), set it again so + * filemap_fdatawait_range(), set it again so * that user process can get -EIO from fsync(). */ set_bit(AS_EIO, diff --git a/fs/namei.c b/fs/namei.c index d11f404667e..87f97ba90ad 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -1279,28 +1279,6 @@ struct dentry *lookup_one_len(const char *name, struct dentry *base, int len) return __lookup_hash(&this, base, NULL); } -/** - * lookup_one_noperm - bad hack for sysfs - * @name: pathname component to lookup - * @base: base directory to lookup from - * - * This is a variant of lookup_one_len that doesn't perform any permission - * checks. It's a horrible hack to work around the braindead sysfs - * architecture and should not be used anywhere else. - * - * DON'T USE THIS FUNCTION EVER, thanks. - */ -struct dentry *lookup_one_noperm(const char *name, struct dentry *base) -{ - int err; - struct qstr this; - - err = __lookup_one_len(name, &this, base, strlen(name)); - if (err) - return ERR_PTR(err); - return __lookup_hash(&this, base, NULL); -} - int user_path_at(int dfd, const char __user *name, unsigned flags, struct path *path) { @@ -1678,6 +1656,15 @@ struct file *do_filp_open(int dfd, const char *pathname, int will_write; int flag = open_to_namei_flags(open_flag); + /* + * O_SYNC is implemented as __O_SYNC|O_DSYNC. As many places only + * check for O_DSYNC if the need any syncing at all we enforce it's + * always set instead of having to deal with possibly weird behaviour + * for malicious applications setting only __O_SYNC. + */ + if (open_flag & __O_SYNC) + open_flag |= O_DSYNC; + if (!acc_mode) acc_mode = MAY_OPEN | ACC_MODE(flag); diff --git a/fs/nfs/file.c b/fs/nfs/file.c index f5fdd39e037..6b891328f33 100644 --- a/fs/nfs/file.c +++ b/fs/nfs/file.c @@ -581,7 +581,7 @@ static int nfs_need_sync_write(struct file *filp, struct inode *inode) { struct nfs_open_context *ctx; - if (IS_SYNC(inode) || (filp->f_flags & O_SYNC)) + if (IS_SYNC(inode) || (filp->f_flags & O_DSYNC)) return 1; ctx = nfs_file_open_context(filp); if (test_bit(NFS_CONTEXT_ERROR_WRITE, &ctx->flags)) @@ -622,7 +622,7 @@ static ssize_t nfs_file_write(struct kiocb *iocb, const struct iovec *iov, nfs_add_stats(inode, NFSIOS_NORMALWRITTENBYTES, count); result = generic_file_aio_write(iocb, iov, nr_segs, pos); - /* Return error values for O_SYNC and IS_SYNC() */ + /* Return error values for O_DSYNC and IS_SYNC() */ if (result >= 0 && nfs_need_sync_write(iocb->ki_filp, inode)) { int err = nfs_do_fsync(nfs_file_open_context(iocb->ki_filp), inode); if (err < 0) diff --git a/fs/nfs/write.c b/fs/nfs/write.c index c84b5cc1a94..b1ce2ea9b93 100644 --- a/fs/nfs/write.c +++ b/fs/nfs/write.c @@ -774,7 +774,7 @@ int nfs_updatepage(struct file *file, struct page *page, */ if (nfs_write_pageuptodate(page, inode) && inode->i_flock == NULL && - !(file->f_flags & O_SYNC)) { + !(file->f_flags & O_DSYNC)) { count = max(count + offset, nfs_page_length(page)); offset = 0; } diff --git a/fs/nilfs2/alloc.c b/fs/nilfs2/alloc.c index d69e6ae5925..3f959f1879d 100644 --- a/fs/nilfs2/alloc.c +++ b/fs/nilfs2/alloc.c @@ -142,29 +142,75 @@ static void nilfs_palloc_desc_block_init(struct inode *inode, } } +static int nilfs_palloc_get_block(struct inode *inode, unsigned long blkoff, + int create, + void (*init_block)(struct inode *, + struct buffer_head *, + void *), + struct buffer_head **bhp, + struct nilfs_bh_assoc *prev, + spinlock_t *lock) +{ + int ret; + + spin_lock(lock); + if (prev->bh && blkoff == prev->blkoff) { + get_bh(prev->bh); + *bhp = prev->bh; + spin_unlock(lock); + return 0; + } + spin_unlock(lock); + + ret = nilfs_mdt_get_block(inode, blkoff, create, init_block, bhp); + if (!ret) { + spin_lock(lock); + /* + * The following code must be safe for change of the + * cache contents during the get block call. + */ + brelse(prev->bh); + get_bh(*bhp); + prev->bh = *bhp; + prev->blkoff = blkoff; + spin_unlock(lock); + } + return ret; +} + static int nilfs_palloc_get_desc_block(struct inode *inode, unsigned long group, int create, struct buffer_head **bhp) { - return nilfs_mdt_get_block(inode, - nilfs_palloc_desc_blkoff(inode, group), - create, nilfs_palloc_desc_block_init, bhp); + struct nilfs_palloc_cache *cache = NILFS_MDT(inode)->mi_palloc_cache; + + return nilfs_palloc_get_block(inode, + nilfs_palloc_desc_blkoff(inode, group), + create, nilfs_palloc_desc_block_init, + bhp, &cache->prev_desc, &cache->lock); } static int nilfs_palloc_get_bitmap_block(struct inode *inode, unsigned long group, int create, struct buffer_head **bhp) { - return nilfs_mdt_get_block(inode, - nilfs_palloc_bitmap_blkoff(inode, group), - create, NULL, bhp); + struct nilfs_palloc_cache *cache = NILFS_MDT(inode)->mi_palloc_cache; + + return nilfs_palloc_get_block(inode, + nilfs_palloc_bitmap_blkoff(inode, group), + create, NULL, bhp, + &cache->prev_bitmap, &cache->lock); } int nilfs_palloc_get_entry_block(struct inode *inode, __u64 nr, int create, struct buffer_head **bhp) { - return nilfs_mdt_get_block(inode, nilfs_palloc_entry_blkoff(inode, nr), - create, NULL, bhp); + struct nilfs_palloc_cache *cache = NILFS_MDT(inode)->mi_palloc_cache; + + return nilfs_palloc_get_block(inode, + nilfs_palloc_entry_blkoff(inode, nr), + create, NULL, bhp, + &cache->prev_entry, &cache->lock); } static struct nilfs_palloc_group_desc * @@ -176,13 +222,6 @@ nilfs_palloc_block_get_group_desc(const struct inode *inode, group % nilfs_palloc_groups_per_desc_block(inode); } -static unsigned char * -nilfs_palloc_block_get_bitmap(const struct inode *inode, - const struct buffer_head *bh, void *kaddr) -{ - return (unsigned char *)(kaddr + bh_offset(bh)); -} - void *nilfs_palloc_block_get_entry(const struct inode *inode, __u64 nr, const struct buffer_head *bh, void *kaddr) { @@ -289,8 +328,7 @@ int nilfs_palloc_prepare_alloc_entry(struct inode *inode, if (ret < 0) goto out_desc; bitmap_kaddr = kmap(bitmap_bh->b_page); - bitmap = nilfs_palloc_block_get_bitmap( - inode, bitmap_bh, bitmap_kaddr); + bitmap = bitmap_kaddr + bh_offset(bitmap_bh); pos = nilfs_palloc_find_available_slot( inode, group, group_offset, bitmap, entries_per_group); @@ -351,8 +389,7 @@ void nilfs_palloc_commit_free_entry(struct inode *inode, desc = nilfs_palloc_block_get_group_desc(inode, group, req->pr_desc_bh, desc_kaddr); bitmap_kaddr = kmap(req->pr_bitmap_bh->b_page); - bitmap = nilfs_palloc_block_get_bitmap(inode, req->pr_bitmap_bh, - bitmap_kaddr); + bitmap = bitmap_kaddr + bh_offset(req->pr_bitmap_bh); if (!nilfs_clear_bit_atomic(nilfs_mdt_bgl_lock(inode, group), group_offset, bitmap)) @@ -385,8 +422,7 @@ void nilfs_palloc_abort_alloc_entry(struct inode *inode, desc = nilfs_palloc_block_get_group_desc(inode, group, req->pr_desc_bh, desc_kaddr); bitmap_kaddr = kmap(req->pr_bitmap_bh->b_page); - bitmap = nilfs_palloc_block_get_bitmap(inode, req->pr_bitmap_bh, - bitmap_kaddr); + bitmap = bitmap_kaddr + bh_offset(req->pr_bitmap_bh); if (!nilfs_clear_bit_atomic(nilfs_mdt_bgl_lock(inode, group), group_offset, bitmap)) printk(KERN_WARNING "%s: entry numer %llu already freed\n", @@ -472,8 +508,7 @@ int nilfs_palloc_freev(struct inode *inode, __u64 *entry_nrs, size_t nitems) desc = nilfs_palloc_block_get_group_desc( inode, group, desc_bh, desc_kaddr); bitmap_kaddr = kmap(bitmap_bh->b_page); - bitmap = nilfs_palloc_block_get_bitmap( - inode, bitmap_bh, bitmap_kaddr); + bitmap = bitmap_kaddr + bh_offset(bitmap_bh); for (j = i, n = 0; (j < nitems) && nilfs_palloc_group_is_in(inode, group, entry_nrs[j]); @@ -502,3 +537,30 @@ int nilfs_palloc_freev(struct inode *inode, __u64 *entry_nrs, size_t nitems) } return 0; } + +void nilfs_palloc_setup_cache(struct inode *inode, + struct nilfs_palloc_cache *cache) +{ + NILFS_MDT(inode)->mi_palloc_cache = cache; + spin_lock_init(&cache->lock); +} + +void nilfs_palloc_clear_cache(struct inode *inode) +{ + struct nilfs_palloc_cache *cache = NILFS_MDT(inode)->mi_palloc_cache; + + spin_lock(&cache->lock); + brelse(cache->prev_desc.bh); + brelse(cache->prev_bitmap.bh); + brelse(cache->prev_entry.bh); + cache->prev_desc.bh = NULL; + cache->prev_bitmap.bh = NULL; + cache->prev_entry.bh = NULL; + spin_unlock(&cache->lock); +} + +void nilfs_palloc_destroy_cache(struct inode *inode) +{ + nilfs_palloc_clear_cache(inode); + NILFS_MDT(inode)->mi_palloc_cache = NULL; +} diff --git a/fs/nilfs2/alloc.h b/fs/nilfs2/alloc.h index 4ace5475c2c..f4543ac4f56 100644 --- a/fs/nilfs2/alloc.h +++ b/fs/nilfs2/alloc.h @@ -69,4 +69,25 @@ int nilfs_palloc_freev(struct inode *, __u64 *, size_t); #define nilfs_clear_bit_atomic ext2_clear_bit_atomic #define nilfs_find_next_zero_bit ext2_find_next_zero_bit +/* + * persistent object allocator cache + */ + +struct nilfs_bh_assoc { + unsigned long blkoff; + struct buffer_head *bh; +}; + +struct nilfs_palloc_cache { + spinlock_t lock; + struct nilfs_bh_assoc prev_desc; + struct nilfs_bh_assoc prev_bitmap; + struct nilfs_bh_assoc prev_entry; +}; + +void nilfs_palloc_setup_cache(struct inode *inode, + struct nilfs_palloc_cache *cache); +void nilfs_palloc_clear_cache(struct inode *inode); +void nilfs_palloc_destroy_cache(struct inode *inode); + #endif /* _NILFS_ALLOC_H */ diff --git a/fs/nilfs2/bmap.c b/fs/nilfs2/bmap.c index 08834df6ec6..f4a14ea2ed9 100644 --- a/fs/nilfs2/bmap.c +++ b/fs/nilfs2/bmap.c @@ -402,19 +402,11 @@ int nilfs_bmap_test_and_clear_dirty(struct nilfs_bmap *bmap) void nilfs_bmap_add_blocks(const struct nilfs_bmap *bmap, int n) { inode_add_bytes(bmap->b_inode, (1 << bmap->b_inode->i_blkbits) * n); - if (NILFS_MDT(bmap->b_inode)) - nilfs_mdt_mark_dirty(bmap->b_inode); - else - mark_inode_dirty(bmap->b_inode); } void nilfs_bmap_sub_blocks(const struct nilfs_bmap *bmap, int n) { inode_sub_bytes(bmap->b_inode, (1 << bmap->b_inode->i_blkbits) * n); - if (NILFS_MDT(bmap->b_inode)) - nilfs_mdt_mark_dirty(bmap->b_inode); - else - mark_inode_dirty(bmap->b_inode); } __u64 nilfs_bmap_data_get_key(const struct nilfs_bmap *bmap, diff --git a/fs/nilfs2/btnode.c b/fs/nilfs2/btnode.c index 84c25382f8e..471e269536a 100644 --- a/fs/nilfs2/btnode.c +++ b/fs/nilfs2/btnode.c @@ -68,9 +68,34 @@ void nilfs_btnode_cache_clear(struct address_space *btnc) truncate_inode_pages(btnc, 0); } +struct buffer_head * +nilfs_btnode_create_block(struct address_space *btnc, __u64 blocknr) +{ + struct inode *inode = NILFS_BTNC_I(btnc); + struct buffer_head *bh; + + bh = nilfs_grab_buffer(inode, btnc, blocknr, 1 << BH_NILFS_Node); + if (unlikely(!bh)) + return NULL; + + if (unlikely(buffer_mapped(bh) || buffer_uptodate(bh) || + buffer_dirty(bh))) { + brelse(bh); + BUG(); + } + memset(bh->b_data, 0, 1 << inode->i_blkbits); + bh->b_bdev = NILFS_I_NILFS(inode)->ns_bdev; + bh->b_blocknr = blocknr; + set_buffer_mapped(bh); + set_buffer_uptodate(bh); + + unlock_page(bh->b_page); + page_cache_release(bh->b_page); + return bh; +} + int nilfs_btnode_submit_block(struct address_space *btnc, __u64 blocknr, - sector_t pblocknr, struct buffer_head **pbh, - int newblk) + sector_t pblocknr, struct buffer_head **pbh) { struct buffer_head *bh; struct inode *inode = NILFS_BTNC_I(btnc); @@ -81,19 +106,6 @@ int nilfs_btnode_submit_block(struct address_space *btnc, __u64 blocknr, return -ENOMEM; err = -EEXIST; /* internal code */ - if (newblk) { - if (unlikely(buffer_mapped(bh) || buffer_uptodate(bh) || - buffer_dirty(bh))) { - brelse(bh); - BUG(); - } - memset(bh->b_data, 0, 1 << inode->i_blkbits); - bh->b_bdev = NILFS_I_NILFS(inode)->ns_bdev; - bh->b_blocknr = blocknr; - set_buffer_mapped(bh); - set_buffer_uptodate(bh); - goto found; - } if (buffer_uptodate(bh) || buffer_dirty(bh)) goto found; @@ -135,27 +147,6 @@ out_locked: return err; } -int nilfs_btnode_get(struct address_space *btnc, __u64 blocknr, - sector_t pblocknr, struct buffer_head **pbh, int newblk) -{ - struct buffer_head *bh; - int err; - - err = nilfs_btnode_submit_block(btnc, blocknr, pblocknr, pbh, newblk); - if (err == -EEXIST) /* internal code (cache hit) */ - return 0; - if (unlikely(err)) - return err; - - bh = *pbh; - wait_on_buffer(bh); - if (!buffer_uptodate(bh)) { - brelse(bh); - return -EIO; - } - return 0; -} - /** * nilfs_btnode_delete - delete B-tree node buffer * @bh: buffer to be deleted @@ -244,12 +235,13 @@ retry: unlock_page(obh->b_page); } - err = nilfs_btnode_get(btnc, newkey, 0, &nbh, 1); - if (likely(!err)) { - BUG_ON(nbh == obh); - ctxt->newbh = nbh; - } - return err; + nbh = nilfs_btnode_create_block(btnc, newkey); + if (!nbh) + return -ENOMEM; + + BUG_ON(nbh == obh); + ctxt->newbh = nbh; + return 0; failed_unlock: unlock_page(obh->b_page); diff --git a/fs/nilfs2/btnode.h b/fs/nilfs2/btnode.h index 3e2275172ed..07da83f0771 100644 --- a/fs/nilfs2/btnode.h +++ b/fs/nilfs2/btnode.h @@ -40,10 +40,10 @@ struct nilfs_btnode_chkey_ctxt { void nilfs_btnode_cache_init_once(struct address_space *); void nilfs_btnode_cache_init(struct address_space *, struct backing_dev_info *); void nilfs_btnode_cache_clear(struct address_space *); +struct buffer_head *nilfs_btnode_create_block(struct address_space *btnc, + __u64 blocknr); int nilfs_btnode_submit_block(struct address_space *, __u64, sector_t, - struct buffer_head **, int); -int nilfs_btnode_get(struct address_space *, __u64, sector_t, - struct buffer_head **, int); + struct buffer_head **); void nilfs_btnode_delete(struct buffer_head *); int nilfs_btnode_prepare_change_key(struct address_space *, struct nilfs_btnode_chkey_ctxt *); diff --git a/fs/nilfs2/btree.c b/fs/nilfs2/btree.c index e25b507a474..7cdd98b8d51 100644 --- a/fs/nilfs2/btree.c +++ b/fs/nilfs2/btree.c @@ -114,7 +114,18 @@ static int nilfs_btree_get_block(const struct nilfs_btree *btree, __u64 ptr, { struct address_space *btnc = &NILFS_BMAP_I((struct nilfs_bmap *)btree)->i_btnode_cache; - return nilfs_btnode_get(btnc, ptr, 0, bhp, 0); + int err; + + err = nilfs_btnode_submit_block(btnc, ptr, 0, bhp); + if (err) + return err == -EEXIST ? 0 : err; + + wait_on_buffer(*bhp); + if (!buffer_uptodate(*bhp)) { + brelse(*bhp); + return -EIO; + } + return 0; } static int nilfs_btree_get_new_block(const struct nilfs_btree *btree, @@ -122,12 +133,15 @@ static int nilfs_btree_get_new_block(const struct nilfs_btree *btree, { struct address_space *btnc = &NILFS_BMAP_I((struct nilfs_bmap *)btree)->i_btnode_cache; - int ret; + struct buffer_head *bh; - ret = nilfs_btnode_get(btnc, ptr, 0, bhp, 1); - if (!ret) - set_buffer_nilfs_volatile(*bhp); - return ret; + bh = nilfs_btnode_create_block(btnc, ptr); + if (!bh) + return -ENOMEM; + + set_buffer_nilfs_volatile(bh); + *bhp = bh; + return 0; } static inline int @@ -444,6 +458,18 @@ nilfs_btree_get_node(const struct nilfs_btree *btree, nilfs_btree_get_nonroot_node(path, level); } +static inline int +nilfs_btree_bad_node(struct nilfs_btree_node *node, int level) +{ + if (unlikely(nilfs_btree_node_get_level(node) != level)) { + dump_stack(); + printk(KERN_CRIT "NILFS: btree level mismatch: %d != %d\n", + nilfs_btree_node_get_level(node), level); + return 1; + } + return 0; +} + static int nilfs_btree_do_lookup(const struct nilfs_btree *btree, struct nilfs_btree_path *path, __u64 key, __u64 *ptrp, int minlevel) @@ -467,7 +493,8 @@ static int nilfs_btree_do_lookup(const struct nilfs_btree *btree, if (ret < 0) return ret; node = nilfs_btree_get_nonroot_node(path, level); - BUG_ON(level != nilfs_btree_node_get_level(node)); + if (nilfs_btree_bad_node(node, level)) + return -EINVAL; if (!found) found = nilfs_btree_node_lookup(node, key, &index); else @@ -512,7 +539,8 @@ static int nilfs_btree_do_lookup_last(const struct nilfs_btree *btree, if (ret < 0) return ret; node = nilfs_btree_get_nonroot_node(path, level); - BUG_ON(level != nilfs_btree_node_get_level(node)); + if (nilfs_btree_bad_node(node, level)) + return -EINVAL; index = nilfs_btree_node_get_nchildren(node) - 1; ptr = nilfs_btree_node_get_ptr(btree, node, index); path[level].bp_index = index; @@ -638,13 +666,11 @@ static void nilfs_btree_promote_key(struct nilfs_btree *btree, { if (level < nilfs_btree_height(btree) - 1) { do { - lock_buffer(path[level].bp_bh); nilfs_btree_node_set_key( nilfs_btree_get_nonroot_node(path, level), path[level].bp_index, key); if (!buffer_dirty(path[level].bp_bh)) nilfs_btnode_mark_dirty(path[level].bp_bh); - unlock_buffer(path[level].bp_bh); } while ((path[level].bp_index == 0) && (++level < nilfs_btree_height(btree) - 1)); } @@ -663,13 +689,11 @@ static void nilfs_btree_do_insert(struct nilfs_btree *btree, struct nilfs_btree_node *node; if (level < nilfs_btree_height(btree) - 1) { - lock_buffer(path[level].bp_bh); node = nilfs_btree_get_nonroot_node(path, level); nilfs_btree_node_insert(btree, node, *keyp, *ptrp, path[level].bp_index); if (!buffer_dirty(path[level].bp_bh)) nilfs_btnode_mark_dirty(path[level].bp_bh); - unlock_buffer(path[level].bp_bh); if (path[level].bp_index == 0) nilfs_btree_promote_key(btree, path, level + 1, @@ -689,9 +713,6 @@ static void nilfs_btree_carry_left(struct nilfs_btree *btree, struct nilfs_btree_node *node, *left; int nchildren, lnchildren, n, move; - lock_buffer(path[level].bp_bh); - lock_buffer(path[level].bp_sib_bh); - node = nilfs_btree_get_nonroot_node(path, level); left = nilfs_btree_get_sib_node(path, level); nchildren = nilfs_btree_node_get_nchildren(node); @@ -712,9 +733,6 @@ static void nilfs_btree_carry_left(struct nilfs_btree *btree, if (!buffer_dirty(path[level].bp_sib_bh)) nilfs_btnode_mark_dirty(path[level].bp_sib_bh); - unlock_buffer(path[level].bp_bh); - unlock_buffer(path[level].bp_sib_bh); - nilfs_btree_promote_key(btree, path, level + 1, nilfs_btree_node_get_key(node, 0)); @@ -740,9 +758,6 @@ static void nilfs_btree_carry_right(struct nilfs_btree *btree, struct nilfs_btree_node *node, *right; int nchildren, rnchildren, n, move; - lock_buffer(path[level].bp_bh); - lock_buffer(path[level].bp_sib_bh); - node = nilfs_btree_get_nonroot_node(path, level); right = nilfs_btree_get_sib_node(path, level); nchildren = nilfs_btree_node_get_nchildren(node); @@ -763,9 +778,6 @@ static void nilfs_btree_carry_right(struct nilfs_btree *btree, if (!buffer_dirty(path[level].bp_sib_bh)) nilfs_btnode_mark_dirty(path[level].bp_sib_bh); - unlock_buffer(path[level].bp_bh); - unlock_buffer(path[level].bp_sib_bh); - path[level + 1].bp_index++; nilfs_btree_promote_key(btree, path, level + 1, nilfs_btree_node_get_key(right, 0)); @@ -794,9 +806,6 @@ static void nilfs_btree_split(struct nilfs_btree *btree, __u64 newptr; int nchildren, n, move; - lock_buffer(path[level].bp_bh); - lock_buffer(path[level].bp_sib_bh); - node = nilfs_btree_get_nonroot_node(path, level); right = nilfs_btree_get_sib_node(path, level); nchildren = nilfs_btree_node_get_nchildren(node); @@ -815,9 +824,6 @@ static void nilfs_btree_split(struct nilfs_btree *btree, if (!buffer_dirty(path[level].bp_sib_bh)) nilfs_btnode_mark_dirty(path[level].bp_sib_bh); - unlock_buffer(path[level].bp_bh); - unlock_buffer(path[level].bp_sib_bh); - newkey = nilfs_btree_node_get_key(right, 0); newptr = path[level].bp_newreq.bpr_ptr; @@ -852,8 +858,6 @@ static void nilfs_btree_grow(struct nilfs_btree *btree, struct nilfs_btree_node *root, *child; int n; - lock_buffer(path[level].bp_sib_bh); - root = nilfs_btree_get_root(btree); child = nilfs_btree_get_sib_node(path, level); @@ -865,8 +869,6 @@ static void nilfs_btree_grow(struct nilfs_btree *btree, if (!buffer_dirty(path[level].bp_sib_bh)) nilfs_btnode_mark_dirty(path[level].bp_sib_bh); - unlock_buffer(path[level].bp_sib_bh); - path[level].bp_bh = path[level].bp_sib_bh; path[level].bp_sib_bh = NULL; @@ -1023,11 +1025,9 @@ static int nilfs_btree_prepare_insert(struct nilfs_btree *btree, stats->bs_nblocks++; - lock_buffer(bh); nilfs_btree_node_init(btree, (struct nilfs_btree_node *)bh->b_data, 0, level, 0, NULL, NULL); - unlock_buffer(bh); path[level].bp_sib_bh = bh; path[level].bp_op = nilfs_btree_split; } @@ -1052,10 +1052,8 @@ static int nilfs_btree_prepare_insert(struct nilfs_btree *btree, if (ret < 0) goto err_out_curr_node; - lock_buffer(bh); nilfs_btree_node_init(btree, (struct nilfs_btree_node *)bh->b_data, 0, level, 0, NULL, NULL); - unlock_buffer(bh); path[level].bp_sib_bh = bh; path[level].bp_op = nilfs_btree_grow; @@ -1154,13 +1152,11 @@ static void nilfs_btree_do_delete(struct nilfs_btree *btree, struct nilfs_btree_node *node; if (level < nilfs_btree_height(btree) - 1) { - lock_buffer(path[level].bp_bh); node = nilfs_btree_get_nonroot_node(path, level); nilfs_btree_node_delete(btree, node, keyp, ptrp, path[level].bp_index); if (!buffer_dirty(path[level].bp_bh)) nilfs_btnode_mark_dirty(path[level].bp_bh); - unlock_buffer(path[level].bp_bh); if (path[level].bp_index == 0) nilfs_btree_promote_key(btree, path, level + 1, nilfs_btree_node_get_key(node, 0)); @@ -1180,9 +1176,6 @@ static void nilfs_btree_borrow_left(struct nilfs_btree *btree, nilfs_btree_do_delete(btree, path, level, keyp, ptrp); - lock_buffer(path[level].bp_bh); - lock_buffer(path[level].bp_sib_bh); - node = nilfs_btree_get_nonroot_node(path, level); left = nilfs_btree_get_sib_node(path, level); nchildren = nilfs_btree_node_get_nchildren(node); @@ -1197,9 +1190,6 @@ static void nilfs_btree_borrow_left(struct nilfs_btree *btree, if (!buffer_dirty(path[level].bp_sib_bh)) nilfs_btnode_mark_dirty(path[level].bp_sib_bh); - unlock_buffer(path[level].bp_bh); - unlock_buffer(path[level].bp_sib_bh); - nilfs_btree_promote_key(btree, path, level + 1, nilfs_btree_node_get_key(node, 0)); @@ -1217,9 +1207,6 @@ static void nilfs_btree_borrow_right(struct nilfs_btree *btree, nilfs_btree_do_delete(btree, path, level, keyp, ptrp); - lock_buffer(path[level].bp_bh); - lock_buffer(path[level].bp_sib_bh); - node = nilfs_btree_get_nonroot_node(path, level); right = nilfs_btree_get_sib_node(path, level); nchildren = nilfs_btree_node_get_nchildren(node); @@ -1234,9 +1221,6 @@ static void nilfs_btree_borrow_right(struct nilfs_btree *btree, if (!buffer_dirty(path[level].bp_sib_bh)) nilfs_btnode_mark_dirty(path[level].bp_sib_bh); - unlock_buffer(path[level].bp_bh); - unlock_buffer(path[level].bp_sib_bh); - path[level + 1].bp_index++; nilfs_btree_promote_key(btree, path, level + 1, nilfs_btree_node_get_key(right, 0)); @@ -1255,9 +1239,6 @@ static void nilfs_btree_concat_left(struct nilfs_btree *btree, nilfs_btree_do_delete(btree, path, level, keyp, ptrp); - lock_buffer(path[level].bp_bh); - lock_buffer(path[level].bp_sib_bh); - node = nilfs_btree_get_nonroot_node(path, level); left = nilfs_btree_get_sib_node(path, level); @@ -1268,9 +1249,6 @@ static void nilfs_btree_concat_left(struct nilfs_btree *btree, if (!buffer_dirty(path[level].bp_sib_bh)) nilfs_btnode_mark_dirty(path[level].bp_sib_bh); - unlock_buffer(path[level].bp_bh); - unlock_buffer(path[level].bp_sib_bh); - nilfs_btnode_delete(path[level].bp_bh); path[level].bp_bh = path[level].bp_sib_bh; path[level].bp_sib_bh = NULL; @@ -1286,9 +1264,6 @@ static void nilfs_btree_concat_right(struct nilfs_btree *btree, nilfs_btree_do_delete(btree, path, level, keyp, ptrp); - lock_buffer(path[level].bp_bh); - lock_buffer(path[level].bp_sib_bh); - node = nilfs_btree_get_nonroot_node(path, level); right = nilfs_btree_get_sib_node(path, level); @@ -1299,9 +1274,6 @@ static void nilfs_btree_concat_right(struct nilfs_btree *btree, if (!buffer_dirty(path[level].bp_bh)) nilfs_btnode_mark_dirty(path[level].bp_bh); - unlock_buffer(path[level].bp_bh); - unlock_buffer(path[level].bp_sib_bh); - nilfs_btnode_delete(path[level].bp_sib_bh); path[level].bp_sib_bh = NULL; path[level + 1].bp_index++; @@ -1316,7 +1288,6 @@ static void nilfs_btree_shrink(struct nilfs_btree *btree, nilfs_btree_do_delete(btree, path, level, keyp, ptrp); - lock_buffer(path[level].bp_bh); root = nilfs_btree_get_root(btree); child = nilfs_btree_get_nonroot_node(path, level); @@ -1324,7 +1295,6 @@ static void nilfs_btree_shrink(struct nilfs_btree *btree, nilfs_btree_node_set_level(root, level); n = nilfs_btree_node_get_nchildren(child); nilfs_btree_node_move_left(btree, root, child, n); - unlock_buffer(path[level].bp_bh); nilfs_btnode_delete(path[level].bp_bh); path[level].bp_bh = NULL; @@ -1699,7 +1669,6 @@ nilfs_btree_commit_convert_and_insert(struct nilfs_bmap *bmap, nilfs_bmap_commit_alloc_ptr(bmap, nreq, dat); /* create child node at level 1 */ - lock_buffer(bh); node = (struct nilfs_btree_node *)bh->b_data; nilfs_btree_node_init(btree, node, 0, 1, n, keys, ptrs); nilfs_btree_node_insert(btree, node, @@ -1709,7 +1678,6 @@ nilfs_btree_commit_convert_and_insert(struct nilfs_bmap *bmap, if (!nilfs_bmap_dirty(bmap)) nilfs_bmap_set_dirty(bmap); - unlock_buffer(bh); brelse(bh); /* create root node at level 2 */ @@ -2050,7 +2018,7 @@ static void nilfs_btree_lookup_dirty_buffers(struct nilfs_bmap *bmap, for (level = NILFS_BTREE_LEVEL_NODE_MIN; level < NILFS_BTREE_LEVEL_MAX; level++) - list_splice(&lists[level], listp->prev); + list_splice_tail(&lists[level], listp); } static int nilfs_btree_assign_p(struct nilfs_btree *btree, diff --git a/fs/nilfs2/btree.h b/fs/nilfs2/btree.h index 0e72bbbc6b6..4b82d84ade7 100644 --- a/fs/nilfs2/btree.h +++ b/fs/nilfs2/btree.h @@ -34,28 +34,6 @@ struct nilfs_btree; struct nilfs_btree_path; /** - * struct nilfs_btree_node - B-tree node - * @bn_flags: flags - * @bn_level: level - * @bn_nchildren: number of children - * @bn_pad: padding - */ -struct nilfs_btree_node { - __u8 bn_flags; - __u8 bn_level; - __le16 bn_nchildren; - __le32 bn_pad; -}; - -/* flags */ -#define NILFS_BTREE_NODE_ROOT 0x01 - -/* level */ -#define NILFS_BTREE_LEVEL_DATA 0 -#define NILFS_BTREE_LEVEL_NODE_MIN (NILFS_BTREE_LEVEL_DATA + 1) -#define NILFS_BTREE_LEVEL_MAX 14 - -/** * struct nilfs_btree - B-tree structure * @bt_bmap: bmap base structure */ diff --git a/fs/nilfs2/cpfile.c b/fs/nilfs2/cpfile.c index 3f5d5d06f53..d5ad54e204a 100644 --- a/fs/nilfs2/cpfile.c +++ b/fs/nilfs2/cpfile.c @@ -926,3 +926,29 @@ int nilfs_cpfile_get_stat(struct inode *cpfile, struct nilfs_cpstat *cpstat) up_read(&NILFS_MDT(cpfile)->mi_sem); return ret; } + +/** + * nilfs_cpfile_read - read cpfile inode + * @cpfile: cpfile inode + * @raw_inode: on-disk cpfile inode + */ +int nilfs_cpfile_read(struct inode *cpfile, struct nilfs_inode *raw_inode) +{ + return nilfs_read_inode_common(cpfile, raw_inode); +} + +/** + * nilfs_cpfile_new - create cpfile + * @nilfs: nilfs object + * @cpsize: size of a checkpoint entry + */ +struct inode *nilfs_cpfile_new(struct the_nilfs *nilfs, size_t cpsize) +{ + struct inode *cpfile; + + cpfile = nilfs_mdt_new(nilfs, NULL, NILFS_CPFILE_INO, 0); + if (cpfile) + nilfs_mdt_set_entry_size(cpfile, cpsize, + sizeof(struct nilfs_cpfile_header)); + return cpfile; +} diff --git a/fs/nilfs2/cpfile.h b/fs/nilfs2/cpfile.h index debea896e70..bc0809e0ab4 100644 --- a/fs/nilfs2/cpfile.h +++ b/fs/nilfs2/cpfile.h @@ -40,4 +40,7 @@ int nilfs_cpfile_get_stat(struct inode *, struct nilfs_cpstat *); ssize_t nilfs_cpfile_get_cpinfo(struct inode *, __u64 *, int, void *, unsigned, size_t); +int nilfs_cpfile_read(struct inode *cpfile, struct nilfs_inode *raw_inode); +struct inode *nilfs_cpfile_new(struct the_nilfs *nilfs, size_t cpsize); + #endif /* _NILFS_CPFILE_H */ diff --git a/fs/nilfs2/dat.c b/fs/nilfs2/dat.c index 1ff8e15bd36..187dd07ba86 100644 --- a/fs/nilfs2/dat.c +++ b/fs/nilfs2/dat.c @@ -33,6 +33,16 @@ #define NILFS_CNO_MIN ((__u64)1) #define NILFS_CNO_MAX (~(__u64)0) +struct nilfs_dat_info { + struct nilfs_mdt_info mi; + struct nilfs_palloc_cache palloc_cache; +}; + +static inline struct nilfs_dat_info *NILFS_DAT_I(struct inode *dat) +{ + return (struct nilfs_dat_info *)NILFS_MDT(dat); +} + static int nilfs_dat_prepare_entry(struct inode *dat, struct nilfs_palloc_req *req, int create) { @@ -425,3 +435,40 @@ ssize_t nilfs_dat_get_vinfo(struct inode *dat, void *buf, unsigned visz, return nvi; } + +/** + * nilfs_dat_read - read dat inode + * @dat: dat inode + * @raw_inode: on-disk dat inode + */ +int nilfs_dat_read(struct inode *dat, struct nilfs_inode *raw_inode) +{ + return nilfs_read_inode_common(dat, raw_inode); +} + +/** + * nilfs_dat_new - create dat file + * @nilfs: nilfs object + * @entry_size: size of a dat entry + */ +struct inode *nilfs_dat_new(struct the_nilfs *nilfs, size_t entry_size) +{ + static struct lock_class_key dat_lock_key; + struct inode *dat; + struct nilfs_dat_info *di; + int err; + + dat = nilfs_mdt_new(nilfs, NULL, NILFS_DAT_INO, sizeof(*di)); + if (dat) { + err = nilfs_palloc_init_blockgroup(dat, entry_size); + if (unlikely(err)) { + nilfs_mdt_destroy(dat); + return NULL; + } + + di = NILFS_DAT_I(dat); + lockdep_set_class(&di->mi.mi_sem, &dat_lock_key); + nilfs_palloc_setup_cache(dat, &di->palloc_cache); + } + return dat; +} diff --git a/fs/nilfs2/dat.h b/fs/nilfs2/dat.h index 406070d3ff4..d31c3aab0ef 100644 --- a/fs/nilfs2/dat.h +++ b/fs/nilfs2/dat.h @@ -53,4 +53,7 @@ int nilfs_dat_freev(struct inode *, __u64 *, size_t); int nilfs_dat_move(struct inode *, __u64, sector_t); ssize_t nilfs_dat_get_vinfo(struct inode *, void *, unsigned, size_t); +int nilfs_dat_read(struct inode *dat, struct nilfs_inode *raw_inode); +struct inode *nilfs_dat_new(struct the_nilfs *nilfs, size_t entry_size); + #endif /* _NILFS_DAT_H */ diff --git a/fs/nilfs2/dir.c b/fs/nilfs2/dir.c index e097099bfc8..76d803e060a 100644 --- a/fs/nilfs2/dir.c +++ b/fs/nilfs2/dir.c @@ -99,9 +99,9 @@ static int nilfs_prepare_chunk(struct page *page, NULL, nilfs_get_block); } -static int nilfs_commit_chunk(struct page *page, - struct address_space *mapping, - unsigned from, unsigned to) +static void nilfs_commit_chunk(struct page *page, + struct address_space *mapping, + unsigned from, unsigned to) { struct inode *dir = mapping->host; struct nilfs_sb_info *sbi = NILFS_SB(dir->i_sb); @@ -112,15 +112,13 @@ static int nilfs_commit_chunk(struct page *page, nr_dirty = nilfs_page_count_clean_buffers(page, from, to); copied = block_write_end(NULL, mapping, pos, len, len, page, NULL); - if (pos + copied > dir->i_size) { + if (pos + copied > dir->i_size) i_size_write(dir, pos + copied); - mark_inode_dirty(dir); - } if (IS_DIRSYNC(dir)) nilfs_set_transaction_flag(NILFS_TI_SYNC); err = nilfs_set_file_dirty(sbi, dir, nr_dirty); + WARN_ON(err); /* do not happen */ unlock_page(page); - return err; } static void nilfs_check_page(struct page *page) @@ -455,11 +453,10 @@ void nilfs_set_link(struct inode *dir, struct nilfs_dir_entry *de, BUG_ON(err); de->inode = cpu_to_le64(inode->i_ino); nilfs_set_de_type(de, inode); - err = nilfs_commit_chunk(page, mapping, from, to); + nilfs_commit_chunk(page, mapping, from, to); nilfs_put_page(page); dir->i_mtime = dir->i_ctime = CURRENT_TIME; /* NILFS_I(dir)->i_flags &= ~NILFS_BTREE_FL; */ - mark_inode_dirty(dir); } /* @@ -548,10 +545,10 @@ got_it: memcpy(de->name, name, namelen); de->inode = cpu_to_le64(inode->i_ino); nilfs_set_de_type(de, inode); - err = nilfs_commit_chunk(page, page->mapping, from, to); + nilfs_commit_chunk(page, page->mapping, from, to); dir->i_mtime = dir->i_ctime = CURRENT_TIME; /* NILFS_I(dir)->i_flags &= ~NILFS_BTREE_FL; */ - mark_inode_dirty(dir); + nilfs_mark_inode_dirty(dir); /* OFFSET_CACHE */ out_put: nilfs_put_page(page); @@ -595,10 +592,9 @@ int nilfs_delete_entry(struct nilfs_dir_entry *dir, struct page *page) if (pde) pde->rec_len = cpu_to_le16(to - from); dir->inode = 0; - err = nilfs_commit_chunk(page, mapping, from, to); + nilfs_commit_chunk(page, mapping, from, to); inode->i_ctime = inode->i_mtime = CURRENT_TIME; /* NILFS_I(inode)->i_flags &= ~NILFS_BTREE_FL; */ - mark_inode_dirty(inode); out: nilfs_put_page(page); return err; @@ -640,7 +636,7 @@ int nilfs_make_empty(struct inode *inode, struct inode *parent) memcpy(de->name, "..\0", 4); nilfs_set_de_type(de, inode); kunmap_atomic(kaddr, KM_USER0); - err = nilfs_commit_chunk(page, mapping, 0, chunk_size); + nilfs_commit_chunk(page, mapping, 0, chunk_size); fail: page_cache_release(page); return err; diff --git a/fs/nilfs2/gcdat.c b/fs/nilfs2/gcdat.c index 93383c5cee9..dd5f7e0a95f 100644 --- a/fs/nilfs2/gcdat.c +++ b/fs/nilfs2/gcdat.c @@ -61,6 +61,8 @@ void nilfs_commit_gcdat_inode(struct the_nilfs *nilfs) nilfs_bmap_commit_gcdat(gii->i_bmap, dii->i_bmap); + nilfs_palloc_clear_cache(dat); + nilfs_palloc_clear_cache(gcdat); nilfs_clear_dirty_pages(mapping); nilfs_copy_back_pages(mapping, gmapping); /* note: mdt dirty flags should be cleared by segctor. */ @@ -79,6 +81,7 @@ void nilfs_clear_gcdat_inode(struct the_nilfs *nilfs) gcdat->i_state = I_CLEAR; gii->i_flags = 0; + nilfs_palloc_clear_cache(gcdat); truncate_inode_pages(gcdat->i_mapping, 0); truncate_inode_pages(&gii->i_btnode_cache, 0); } diff --git a/fs/nilfs2/gcinode.c b/fs/nilfs2/gcinode.c index e6de0a27ab5..e16a6664dfa 100644 --- a/fs/nilfs2/gcinode.c +++ b/fs/nilfs2/gcinode.c @@ -149,7 +149,7 @@ int nilfs_gccache_submit_read_node(struct inode *inode, sector_t pbn, __u64 vbn, struct buffer_head **out_bh) { int ret = nilfs_btnode_submit_block(&NILFS_I(inode)->i_btnode_cache, - vbn ? : pbn, pbn, out_bh, 0); + vbn ? : pbn, pbn, out_bh); if (ret == -EEXIST) /* internal code (cache hit) */ ret = 0; return ret; @@ -212,9 +212,10 @@ void nilfs_destroy_gccache(struct the_nilfs *nilfs) static struct inode *alloc_gcinode(struct the_nilfs *nilfs, ino_t ino, __u64 cno) { - struct inode *inode = nilfs_mdt_new_common(nilfs, NULL, ino, GFP_NOFS); + struct inode *inode; struct nilfs_inode_info *ii; + inode = nilfs_mdt_new_common(nilfs, NULL, ino, GFP_NOFS, 0); if (!inode) return NULL; @@ -265,7 +266,6 @@ struct inode *nilfs_gc_iget(struct the_nilfs *nilfs, ino_t ino, __u64 cno) */ void nilfs_clear_gcinode(struct inode *inode) { - nilfs_mdt_clear(inode); nilfs_mdt_destroy(inode); } diff --git a/fs/nilfs2/ifile.c b/fs/nilfs2/ifile.c index de86401f209..922d9dd42c8 100644 --- a/fs/nilfs2/ifile.c +++ b/fs/nilfs2/ifile.c @@ -29,6 +29,17 @@ #include "alloc.h" #include "ifile.h" + +struct nilfs_ifile_info { + struct nilfs_mdt_info mi; + struct nilfs_palloc_cache palloc_cache; +}; + +static inline struct nilfs_ifile_info *NILFS_IFILE_I(struct inode *ifile) +{ + return (struct nilfs_ifile_info *)NILFS_MDT(ifile); +} + /** * nilfs_ifile_create_inode - create a new disk inode * @ifile: ifile inode @@ -148,3 +159,27 @@ int nilfs_ifile_get_inode_block(struct inode *ifile, ino_t ino, } return err; } + +/** + * nilfs_ifile_new - create inode file + * @sbi: nilfs_sb_info struct + * @inode_size: size of an inode + */ +struct inode *nilfs_ifile_new(struct nilfs_sb_info *sbi, size_t inode_size) +{ + struct inode *ifile; + int err; + + ifile = nilfs_mdt_new(sbi->s_nilfs, sbi->s_super, NILFS_IFILE_INO, + sizeof(struct nilfs_ifile_info)); + if (ifile) { + err = nilfs_palloc_init_blockgroup(ifile, inode_size); + if (unlikely(err)) { + nilfs_mdt_destroy(ifile); + return NULL; + } + nilfs_palloc_setup_cache(ifile, + &NILFS_IFILE_I(ifile)->palloc_cache); + } + return ifile; +} diff --git a/fs/nilfs2/ifile.h b/fs/nilfs2/ifile.h index ecc3ba76db4..cbca32e498f 100644 --- a/fs/nilfs2/ifile.h +++ b/fs/nilfs2/ifile.h @@ -49,4 +49,6 @@ int nilfs_ifile_create_inode(struct inode *, ino_t *, struct buffer_head **); int nilfs_ifile_delete_inode(struct inode *, ino_t); int nilfs_ifile_get_inode_block(struct inode *, ino_t, struct buffer_head **); +struct inode *nilfs_ifile_new(struct nilfs_sb_info *sbi, size_t inode_size); + #endif /* _NILFS_IFILE_H */ diff --git a/fs/nilfs2/inode.c b/fs/nilfs2/inode.c index 2a0a5a3ac13..7868cc122ac 100644 --- a/fs/nilfs2/inode.c +++ b/fs/nilfs2/inode.c @@ -97,6 +97,7 @@ int nilfs_get_block(struct inode *inode, sector_t blkoff, nilfs_transaction_abort(inode->i_sb); goto out; } + nilfs_mark_inode_dirty(inode); nilfs_transaction_commit(inode->i_sb); /* never fails */ /* Error handling should be detailed */ set_buffer_new(bh_result); @@ -322,7 +323,6 @@ struct inode *nilfs_new_inode(struct inode *dir, int mode) nilfs_init_acl(), proper cancellation of above jobs should be considered */ - mark_inode_dirty(inode); return inode; failed_acl: @@ -525,7 +525,6 @@ void nilfs_update_inode(struct inode *inode, struct buffer_head *ibh) raw_inode = nilfs_ifile_map_inode(sbi->s_ifile, ino, ibh); - /* The buffer is guarded with lock_buffer() by the caller */ if (test_and_clear_bit(NILFS_I_NEW, &ii->i_state)) memset(raw_inode, 0, NILFS_MDT(sbi->s_ifile)->mi_entry_size); set_bit(NILFS_I_INODE_DIRTY, &ii->i_state); @@ -599,6 +598,7 @@ void nilfs_truncate(struct inode *inode) if (IS_SYNC(inode)) nilfs_set_transaction_flag(NILFS_TI_SYNC); + nilfs_mark_inode_dirty(inode); nilfs_set_file_dirty(NILFS_SB(sb), inode, 0); nilfs_transaction_commit(sb); /* May construct a logical segment and may fail in sync mode. @@ -623,6 +623,7 @@ void nilfs_delete_inode(struct inode *inode) truncate_inode_pages(&inode->i_data, 0); nilfs_truncate_bmap(ii, 0); + nilfs_mark_inode_dirty(inode); nilfs_free_inode(inode); /* nilfs_free_inode() marks inode buffer dirty */ if (IS_SYNC(inode)) @@ -745,9 +746,7 @@ int nilfs_mark_inode_dirty(struct inode *inode) "failed to reget inode block.\n"); return err; } - lock_buffer(ibh); nilfs_update_inode(inode, ibh); - unlock_buffer(ibh); nilfs_mdt_mark_buffer_dirty(ibh); nilfs_mdt_mark_dirty(sbi->s_ifile); brelse(ibh); diff --git a/fs/nilfs2/mdt.c b/fs/nilfs2/mdt.c index f6326112d64..06713ffcc7f 100644 --- a/fs/nilfs2/mdt.c +++ b/fs/nilfs2/mdt.c @@ -186,7 +186,7 @@ nilfs_mdt_submit_block(struct inode *inode, unsigned long blkoff, } static int nilfs_mdt_read_block(struct inode *inode, unsigned long block, - struct buffer_head **out_bh) + int readahead, struct buffer_head **out_bh) { struct buffer_head *first_bh, *bh; unsigned long blkoff; @@ -200,16 +200,18 @@ static int nilfs_mdt_read_block(struct inode *inode, unsigned long block, if (unlikely(err)) goto failed; - blkoff = block + 1; - for (i = 0; i < nr_ra_blocks; i++, blkoff++) { - err = nilfs_mdt_submit_block(inode, blkoff, READA, &bh); - if (likely(!err || err == -EEXIST)) - brelse(bh); - else if (err != -EBUSY) - break; /* abort readahead if bmap lookup failed */ - - if (!buffer_locked(first_bh)) - goto out_no_wait; + if (readahead) { + blkoff = block + 1; + for (i = 0; i < nr_ra_blocks; i++, blkoff++) { + err = nilfs_mdt_submit_block(inode, blkoff, READA, &bh); + if (likely(!err || err == -EEXIST)) + brelse(bh); + else if (err != -EBUSY) + break; + /* abort readahead if bmap lookup failed */ + if (!buffer_locked(first_bh)) + goto out_no_wait; + } } wait_on_buffer(first_bh); @@ -263,7 +265,7 @@ int nilfs_mdt_get_block(struct inode *inode, unsigned long blkoff, int create, /* Should be rewritten with merging nilfs_mdt_read_block() */ retry: - ret = nilfs_mdt_read_block(inode, blkoff, out_bh); + ret = nilfs_mdt_read_block(inode, blkoff, !create, out_bh); if (!create || ret != -ENOENT) return ret; @@ -371,7 +373,7 @@ int nilfs_mdt_mark_block_dirty(struct inode *inode, unsigned long block) struct buffer_head *bh; int err; - err = nilfs_mdt_read_block(inode, block, &bh); + err = nilfs_mdt_read_block(inode, block, 0, &bh); if (unlikely(err)) return err; nilfs_mark_buffer_dirty(bh); @@ -445,9 +447,17 @@ static const struct file_operations def_mdt_fops; * longer than those of the super block structs; they may continue for * several consecutive mounts/umounts. This would need discussions. */ +/** + * nilfs_mdt_new_common - allocate a pseudo inode for metadata file + * @nilfs: nilfs object + * @sb: super block instance the metadata file belongs to + * @ino: inode number + * @gfp_mask: gfp mask for data pages + * @objsz: size of the private object attached to inode->i_private + */ struct inode * nilfs_mdt_new_common(struct the_nilfs *nilfs, struct super_block *sb, - ino_t ino, gfp_t gfp_mask) + ino_t ino, gfp_t gfp_mask, size_t objsz) { struct inode *inode = nilfs_alloc_inode_common(nilfs); @@ -455,8 +465,9 @@ nilfs_mdt_new_common(struct the_nilfs *nilfs, struct super_block *sb, return NULL; else { struct address_space * const mapping = &inode->i_data; - struct nilfs_mdt_info *mi = kzalloc(sizeof(*mi), GFP_NOFS); + struct nilfs_mdt_info *mi; + mi = kzalloc(max(sizeof(*mi), objsz), GFP_NOFS); if (!mi) { nilfs_destroy_inode(inode); return NULL; @@ -513,11 +524,11 @@ nilfs_mdt_new_common(struct the_nilfs *nilfs, struct super_block *sb, } struct inode *nilfs_mdt_new(struct the_nilfs *nilfs, struct super_block *sb, - ino_t ino) + ino_t ino, size_t objsz) { - struct inode *inode = nilfs_mdt_new_common(nilfs, sb, ino, - NILFS_MDT_GFP); + struct inode *inode; + inode = nilfs_mdt_new_common(nilfs, sb, ino, NILFS_MDT_GFP, objsz); if (!inode) return NULL; @@ -544,14 +555,15 @@ void nilfs_mdt_set_shadow(struct inode *orig, struct inode *shadow) &NILFS_I(orig)->i_btnode_cache; } -void nilfs_mdt_clear(struct inode *inode) +static void nilfs_mdt_clear(struct inode *inode) { struct nilfs_inode_info *ii = NILFS_I(inode); invalidate_mapping_pages(inode->i_mapping, 0, -1); truncate_inode_pages(inode->i_mapping, 0); - nilfs_bmap_clear(ii->i_bmap); + if (test_bit(NILFS_I_BMAP, &ii->i_state)) + nilfs_bmap_clear(ii->i_bmap); nilfs_btnode_cache_clear(&ii->i_btnode_cache); } @@ -559,6 +571,10 @@ void nilfs_mdt_destroy(struct inode *inode) { struct nilfs_mdt_info *mdi = NILFS_MDT(inode); + if (mdi->mi_palloc_cache) + nilfs_palloc_destroy_cache(inode); + nilfs_mdt_clear(inode); + kfree(mdi->mi_bgl); /* kfree(NULL) is safe */ kfree(mdi); nilfs_destroy_inode(inode); diff --git a/fs/nilfs2/mdt.h b/fs/nilfs2/mdt.h index 431599733c9..6c4bbb0470f 100644 --- a/fs/nilfs2/mdt.h +++ b/fs/nilfs2/mdt.h @@ -36,6 +36,7 @@ * @mi_entry_size: size of an entry * @mi_first_entry_offset: offset to the first entry * @mi_entries_per_block: number of entries in a block + * @mi_palloc_cache: persistent object allocator cache * @mi_blocks_per_group: number of blocks in a group * @mi_blocks_per_desc_block: number of blocks per descriptor block */ @@ -46,6 +47,7 @@ struct nilfs_mdt_info { unsigned mi_entry_size; unsigned mi_first_entry_offset; unsigned long mi_entries_per_block; + struct nilfs_palloc_cache *mi_palloc_cache; unsigned long mi_blocks_per_group; unsigned long mi_blocks_per_desc_block; }; @@ -74,11 +76,11 @@ int nilfs_mdt_forget_block(struct inode *, unsigned long); int nilfs_mdt_mark_block_dirty(struct inode *, unsigned long); int nilfs_mdt_fetch_dirty(struct inode *); -struct inode *nilfs_mdt_new(struct the_nilfs *, struct super_block *, ino_t); +struct inode *nilfs_mdt_new(struct the_nilfs *, struct super_block *, ino_t, + size_t); struct inode *nilfs_mdt_new_common(struct the_nilfs *, struct super_block *, - ino_t, gfp_t); + ino_t, gfp_t, size_t); void nilfs_mdt_destroy(struct inode *); -void nilfs_mdt_clear(struct inode *); void nilfs_mdt_set_entry_size(struct inode *, unsigned, unsigned); void nilfs_mdt_set_shadow(struct inode *, struct inode *); @@ -104,21 +106,4 @@ static inline __u64 nilfs_mdt_cno(struct inode *inode) #define nilfs_mdt_bgl_lock(inode, bg) \ (&NILFS_MDT(inode)->mi_bgl->locks[(bg) & (NR_BG_LOCKS-1)].lock) - -static inline int -nilfs_mdt_read_inode_direct(struct inode *inode, struct buffer_head *bh, - unsigned n) -{ - return nilfs_read_inode_common( - inode, (struct nilfs_inode *)(bh->b_data + n)); -} - -static inline void -nilfs_mdt_write_inode_direct(struct inode *inode, struct buffer_head *bh, - unsigned n) -{ - nilfs_write_inode_common( - inode, (struct nilfs_inode *)(bh->b_data + n), 1); -} - #endif /* _NILFS_MDT_H */ diff --git a/fs/nilfs2/namei.c b/fs/nilfs2/namei.c index ed02e886fa7..07ba838ef08 100644 --- a/fs/nilfs2/namei.c +++ b/fs/nilfs2/namei.c @@ -120,7 +120,7 @@ static int nilfs_create(struct inode *dir, struct dentry *dentry, int mode, inode->i_op = &nilfs_file_inode_operations; inode->i_fop = &nilfs_file_operations; inode->i_mapping->a_ops = &nilfs_aops; - mark_inode_dirty(inode); + nilfs_mark_inode_dirty(inode); err = nilfs_add_nondir(dentry, inode); } if (!err) @@ -148,7 +148,7 @@ nilfs_mknod(struct inode *dir, struct dentry *dentry, int mode, dev_t rdev) err = PTR_ERR(inode); if (!IS_ERR(inode)) { init_special_inode(inode, inode->i_mode, rdev); - mark_inode_dirty(inode); + nilfs_mark_inode_dirty(inode); err = nilfs_add_nondir(dentry, inode); } if (!err) @@ -188,7 +188,7 @@ static int nilfs_symlink(struct inode *dir, struct dentry *dentry, goto out_fail; /* mark_inode_dirty(inode); */ - /* nilfs_new_inode() and page_symlink() do this */ + /* page_symlink() do this */ err = nilfs_add_nondir(dentry, inode); out: @@ -200,7 +200,8 @@ out: return err; out_fail: - inode_dec_link_count(inode); + drop_nlink(inode); + nilfs_mark_inode_dirty(inode); iput(inode); goto out; } @@ -245,7 +246,7 @@ static int nilfs_mkdir(struct inode *dir, struct dentry *dentry, int mode) if (err) return err; - inode_inc_link_count(dir); + inc_nlink(dir); inode = nilfs_new_inode(dir, S_IFDIR | mode); err = PTR_ERR(inode); @@ -256,7 +257,7 @@ static int nilfs_mkdir(struct inode *dir, struct dentry *dentry, int mode) inode->i_fop = &nilfs_dir_operations; inode->i_mapping->a_ops = &nilfs_aops; - inode_inc_link_count(inode); + inc_nlink(inode); err = nilfs_make_empty(inode, dir); if (err) @@ -266,6 +267,7 @@ static int nilfs_mkdir(struct inode *dir, struct dentry *dentry, int mode) if (err) goto out_fail; + nilfs_mark_inode_dirty(inode); d_instantiate(dentry, inode); out: if (!err) @@ -276,26 +278,23 @@ out: return err; out_fail: - inode_dec_link_count(inode); - inode_dec_link_count(inode); + drop_nlink(inode); + drop_nlink(inode); + nilfs_mark_inode_dirty(inode); iput(inode); out_dir: - inode_dec_link_count(dir); + drop_nlink(dir); + nilfs_mark_inode_dirty(dir); goto out; } -static int nilfs_unlink(struct inode *dir, struct dentry *dentry) +static int nilfs_do_unlink(struct inode *dir, struct dentry *dentry) { struct inode *inode; struct nilfs_dir_entry *de; struct page *page; - struct nilfs_transaction_info ti; int err; - err = nilfs_transaction_begin(dir->i_sb, &ti, 0); - if (err) - return err; - err = -ENOENT; de = nilfs_find_entry(dir, dentry, &page); if (!de) @@ -317,12 +316,28 @@ static int nilfs_unlink(struct inode *dir, struct dentry *dentry) goto out; inode->i_ctime = dir->i_ctime; - inode_dec_link_count(inode); + drop_nlink(inode); err = 0; out: - if (!err) + return err; +} + +static int nilfs_unlink(struct inode *dir, struct dentry *dentry) +{ + struct nilfs_transaction_info ti; + int err; + + err = nilfs_transaction_begin(dir->i_sb, &ti, 0); + if (err) + return err; + + err = nilfs_do_unlink(dir, dentry); + + if (!err) { + nilfs_mark_inode_dirty(dir); + nilfs_mark_inode_dirty(dentry->d_inode); err = nilfs_transaction_commit(dir->i_sb); - else + } else nilfs_transaction_abort(dir->i_sb); return err; @@ -340,11 +355,13 @@ static int nilfs_rmdir(struct inode *dir, struct dentry *dentry) err = -ENOTEMPTY; if (nilfs_empty_dir(inode)) { - err = nilfs_unlink(dir, dentry); + err = nilfs_do_unlink(dir, dentry); if (!err) { inode->i_size = 0; - inode_dec_link_count(inode); - inode_dec_link_count(dir); + drop_nlink(inode); + nilfs_mark_inode_dirty(inode); + drop_nlink(dir); + nilfs_mark_inode_dirty(dir); } } if (!err) @@ -395,42 +412,48 @@ static int nilfs_rename(struct inode *old_dir, struct dentry *old_dentry, new_de = nilfs_find_entry(new_dir, new_dentry, &new_page); if (!new_de) goto out_dir; - inode_inc_link_count(old_inode); + inc_nlink(old_inode); nilfs_set_link(new_dir, new_de, new_page, old_inode); + nilfs_mark_inode_dirty(new_dir); new_inode->i_ctime = CURRENT_TIME; if (dir_de) drop_nlink(new_inode); - inode_dec_link_count(new_inode); + drop_nlink(new_inode); + nilfs_mark_inode_dirty(new_inode); } else { if (dir_de) { err = -EMLINK; if (new_dir->i_nlink >= NILFS_LINK_MAX) goto out_dir; } - inode_inc_link_count(old_inode); + inc_nlink(old_inode); err = nilfs_add_link(new_dentry, old_inode); if (err) { - inode_dec_link_count(old_inode); + drop_nlink(old_inode); + nilfs_mark_inode_dirty(old_inode); goto out_dir; } - if (dir_de) - inode_inc_link_count(new_dir); + if (dir_de) { + inc_nlink(new_dir); + nilfs_mark_inode_dirty(new_dir); + } } /* * 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; nilfs_delete_entry(old_de, old_page); - inode_dec_link_count(old_inode); + drop_nlink(old_inode); if (dir_de) { nilfs_set_link(old_inode, dir_de, dir_page, new_dir); - inode_dec_link_count(old_dir); + drop_nlink(old_dir); } + nilfs_mark_inode_dirty(old_dir); + nilfs_mark_inode_dirty(old_inode); err = nilfs_transaction_commit(old_dir->i_sb); return err; diff --git a/fs/nilfs2/recovery.c b/fs/nilfs2/recovery.c index 6dc83591d11..c9c96c7825d 100644 --- a/fs/nilfs2/recovery.c +++ b/fs/nilfs2/recovery.c @@ -770,14 +770,8 @@ int nilfs_recover_logical_segments(struct the_nilfs *nilfs, nilfs_finish_roll_forward(nilfs, sbi, ri); } - nilfs_detach_checkpoint(sbi); - return 0; - failed: nilfs_detach_checkpoint(sbi); - nilfs_mdt_clear(nilfs->ns_cpfile); - nilfs_mdt_clear(nilfs->ns_sufile); - nilfs_mdt_clear(nilfs->ns_dat); return err; } @@ -804,6 +798,7 @@ int nilfs_search_super_root(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi, struct nilfs_segsum_info ssi; sector_t pseg_start, pseg_end, sr_pseg_start = 0; sector_t seg_start, seg_end; /* range of full segment (block number) */ + sector_t b, end; u64 seg_seq; __u64 segnum, nextnum = 0; __u64 cno; @@ -819,6 +814,11 @@ int nilfs_search_super_root(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi, /* Calculate range of segment */ nilfs_get_segment_range(nilfs, segnum, &seg_start, &seg_end); + /* Read ahead segment */ + b = seg_start; + while (b <= seg_end) + sb_breadahead(sbi->s_super, b++); + for (;;) { /* Load segment summary */ ret = load_segment_summary(sbi, pseg_start, seg_seq, &ssi, 1); @@ -841,14 +841,20 @@ int nilfs_search_super_root(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi, ri->ri_nextnum = nextnum; empty_seg = 0; + if (!NILFS_SEG_HAS_SR(&ssi) && !scan_newer) { + /* This will never happen because a superblock + (last_segment) always points to a pseg + having a super root. */ + ret = NILFS_SEG_FAIL_CONSISTENCY; + goto failed; + } + + if (pseg_start == seg_start) { + nilfs_get_segment_range(nilfs, nextnum, &b, &end); + while (b <= end) + sb_breadahead(sbi->s_super, b++); + } if (!NILFS_SEG_HAS_SR(&ssi)) { - if (!scan_newer) { - /* This will never happen because a superblock - (last_segment) always points to a pseg - having a super root. */ - ret = NILFS_SEG_FAIL_CONSISTENCY; - goto failed; - } if (!ri->ri_lsegs_start && NILFS_SEG_LOGBGN(&ssi)) { ri->ri_lsegs_start = pseg_start; ri->ri_lsegs_start_seq = seg_seq; @@ -919,7 +925,7 @@ int nilfs_search_super_root(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi, super_root_found: /* Updating pointers relating to the latest checkpoint */ - list_splice(&segments, ri->ri_used_segments.prev); + list_splice_tail(&segments, &ri->ri_used_segments); nilfs->ns_last_pseg = sr_pseg_start; nilfs->ns_last_seq = nilfs->ns_seg_seq; nilfs->ns_last_cno = ri->ri_cno; diff --git a/fs/nilfs2/segbuf.c b/fs/nilfs2/segbuf.c index e6d9e37fa24..645c78656aa 100644 --- a/fs/nilfs2/segbuf.c +++ b/fs/nilfs2/segbuf.c @@ -24,10 +24,22 @@ #include <linux/buffer_head.h> #include <linux/writeback.h> #include <linux/crc32.h> +#include <linux/backing-dev.h> #include "page.h" #include "segbuf.h" +struct nilfs_write_info { + struct the_nilfs *nilfs; + struct bio *bio; + int start, end; /* The region to be submitted */ + int rest_blocks; + int max_pages; + int nr_vecs; + sector_t blocknr; +}; + + static struct kmem_cache *nilfs_segbuf_cachep; static void nilfs_segbuf_init_once(void *obj) @@ -63,6 +75,11 @@ struct nilfs_segment_buffer *nilfs_segbuf_new(struct super_block *sb) INIT_LIST_HEAD(&segbuf->sb_list); INIT_LIST_HEAD(&segbuf->sb_segsum_buffers); INIT_LIST_HEAD(&segbuf->sb_payload_buffers); + + init_completion(&segbuf->sb_bio_event); + atomic_set(&segbuf->sb_err, 0); + segbuf->sb_nbio = 0; + return segbuf; } @@ -83,6 +100,22 @@ void nilfs_segbuf_map(struct nilfs_segment_buffer *segbuf, __u64 segnum, segbuf->sb_fseg_end - segbuf->sb_pseg_start + 1; } +/** + * nilfs_segbuf_map_cont - map a new log behind a given log + * @segbuf: new segment buffer + * @prev: segment buffer containing a log to be continued + */ +void nilfs_segbuf_map_cont(struct nilfs_segment_buffer *segbuf, + struct nilfs_segment_buffer *prev) +{ + segbuf->sb_segnum = prev->sb_segnum; + segbuf->sb_fseg_start = prev->sb_fseg_start; + segbuf->sb_fseg_end = prev->sb_fseg_end; + segbuf->sb_pseg_start = prev->sb_pseg_start + prev->sb_sum.nblocks; + segbuf->sb_rest_blocks = + segbuf->sb_fseg_end - segbuf->sb_pseg_start + 1; +} + void nilfs_segbuf_set_next_segnum(struct nilfs_segment_buffer *segbuf, __u64 nextnum, struct the_nilfs *nilfs) { @@ -132,8 +165,6 @@ int nilfs_segbuf_reset(struct nilfs_segment_buffer *segbuf, unsigned flags, segbuf->sb_sum.sumbytes = sizeof(struct nilfs_segment_summary); segbuf->sb_sum.nfinfo = segbuf->sb_sum.nfileblk = 0; segbuf->sb_sum.ctime = ctime; - - segbuf->sb_io_error = 0; return 0; } @@ -219,7 +250,7 @@ void nilfs_segbuf_fill_in_data_crc(struct nilfs_segment_buffer *segbuf, raw_sum->ss_datasum = cpu_to_le32(crc); } -void nilfs_release_buffers(struct list_head *list) +static void nilfs_release_buffers(struct list_head *list) { struct buffer_head *bh, *n; @@ -241,13 +272,56 @@ void nilfs_release_buffers(struct list_head *list) } } +static void nilfs_segbuf_clear(struct nilfs_segment_buffer *segbuf) +{ + nilfs_release_buffers(&segbuf->sb_segsum_buffers); + nilfs_release_buffers(&segbuf->sb_payload_buffers); +} + +/* + * Iterators for segment buffers + */ +void nilfs_clear_logs(struct list_head *logs) +{ + struct nilfs_segment_buffer *segbuf; + + list_for_each_entry(segbuf, logs, sb_list) + nilfs_segbuf_clear(segbuf); +} + +void nilfs_truncate_logs(struct list_head *logs, + struct nilfs_segment_buffer *last) +{ + struct nilfs_segment_buffer *n, *segbuf; + + segbuf = list_prepare_entry(last, logs, sb_list); + list_for_each_entry_safe_continue(segbuf, n, logs, sb_list) { + list_del_init(&segbuf->sb_list); + nilfs_segbuf_clear(segbuf); + nilfs_segbuf_free(segbuf); + } +} + +int nilfs_wait_on_logs(struct list_head *logs) +{ + struct nilfs_segment_buffer *segbuf; + int err; + + list_for_each_entry(segbuf, logs, sb_list) { + err = nilfs_segbuf_wait(segbuf); + if (err) + return err; + } + return 0; +} + /* * BIO operations */ static void nilfs_end_bio_write(struct bio *bio, int err) { const int uptodate = test_bit(BIO_UPTODATE, &bio->bi_flags); - struct nilfs_write_info *wi = bio->bi_private; + struct nilfs_segment_buffer *segbuf = bio->bi_private; if (err == -EOPNOTSUPP) { set_bit(BIO_EOPNOTSUPP, &bio->bi_flags); @@ -256,21 +330,22 @@ static void nilfs_end_bio_write(struct bio *bio, int err) } if (!uptodate) - atomic_inc(&wi->err); + atomic_inc(&segbuf->sb_err); bio_put(bio); - complete(&wi->bio_event); + complete(&segbuf->sb_bio_event); } -static int nilfs_submit_seg_bio(struct nilfs_write_info *wi, int mode) +static int nilfs_segbuf_submit_bio(struct nilfs_segment_buffer *segbuf, + struct nilfs_write_info *wi, int mode) { struct bio *bio = wi->bio; int err; - if (wi->nbio > 0 && bdi_write_congested(wi->bdi)) { - wait_for_completion(&wi->bio_event); - wi->nbio--; - if (unlikely(atomic_read(&wi->err))) { + if (segbuf->sb_nbio > 0 && bdi_write_congested(wi->nilfs->ns_bdi)) { + wait_for_completion(&segbuf->sb_bio_event); + segbuf->sb_nbio--; + if (unlikely(atomic_read(&segbuf->sb_err))) { bio_put(bio); err = -EIO; goto failed; @@ -278,7 +353,7 @@ static int nilfs_submit_seg_bio(struct nilfs_write_info *wi, int mode) } bio->bi_end_io = nilfs_end_bio_write; - bio->bi_private = wi; + bio->bi_private = segbuf; bio_get(bio); submit_bio(mode, bio); if (bio_flagged(bio, BIO_EOPNOTSUPP)) { @@ -286,7 +361,7 @@ static int nilfs_submit_seg_bio(struct nilfs_write_info *wi, int mode) err = -EOPNOTSUPP; goto failed; } - wi->nbio++; + segbuf->sb_nbio++; bio_put(bio); wi->bio = NULL; @@ -301,17 +376,15 @@ static int nilfs_submit_seg_bio(struct nilfs_write_info *wi, int mode) } /** - * nilfs_alloc_seg_bio - allocate a bio for writing segment. - * @sb: super block - * @start: beginning disk block number of this BIO. + * nilfs_alloc_seg_bio - allocate a new bio for writing log + * @nilfs: nilfs object + * @start: start block number of the bio * @nr_vecs: request size of page vector. * - * alloc_seg_bio() allocates a new BIO structure and initialize it. - * * Return Value: On success, pointer to the struct bio is returned. * On error, NULL is returned. */ -static struct bio *nilfs_alloc_seg_bio(struct super_block *sb, sector_t start, +static struct bio *nilfs_alloc_seg_bio(struct the_nilfs *nilfs, sector_t start, int nr_vecs) { struct bio *bio; @@ -322,36 +395,33 @@ static struct bio *nilfs_alloc_seg_bio(struct super_block *sb, sector_t start, bio = bio_alloc(GFP_NOIO, nr_vecs); } if (likely(bio)) { - bio->bi_bdev = sb->s_bdev; - bio->bi_sector = (sector_t)start << (sb->s_blocksize_bits - 9); + bio->bi_bdev = nilfs->ns_bdev; + bio->bi_sector = start << (nilfs->ns_blocksize_bits - 9); } return bio; } -void nilfs_segbuf_prepare_write(struct nilfs_segment_buffer *segbuf, - struct nilfs_write_info *wi) +static void nilfs_segbuf_prepare_write(struct nilfs_segment_buffer *segbuf, + struct nilfs_write_info *wi) { wi->bio = NULL; wi->rest_blocks = segbuf->sb_sum.nblocks; - wi->max_pages = bio_get_nr_vecs(wi->sb->s_bdev); + wi->max_pages = bio_get_nr_vecs(wi->nilfs->ns_bdev); wi->nr_vecs = min(wi->max_pages, wi->rest_blocks); wi->start = wi->end = 0; - wi->nbio = 0; wi->blocknr = segbuf->sb_pseg_start; - - atomic_set(&wi->err, 0); - init_completion(&wi->bio_event); } -static int nilfs_submit_bh(struct nilfs_write_info *wi, struct buffer_head *bh, - int mode) +static int nilfs_segbuf_submit_bh(struct nilfs_segment_buffer *segbuf, + struct nilfs_write_info *wi, + struct buffer_head *bh, int mode) { int len, err; BUG_ON(wi->nr_vecs <= 0); repeat: if (!wi->bio) { - wi->bio = nilfs_alloc_seg_bio(wi->sb, wi->blocknr + wi->end, + wi->bio = nilfs_alloc_seg_bio(wi->nilfs, wi->blocknr + wi->end, wi->nr_vecs); if (unlikely(!wi->bio)) return -ENOMEM; @@ -363,76 +433,83 @@ static int nilfs_submit_bh(struct nilfs_write_info *wi, struct buffer_head *bh, return 0; } /* bio is FULL */ - err = nilfs_submit_seg_bio(wi, mode); + err = nilfs_segbuf_submit_bio(segbuf, wi, mode); /* never submit current bh */ if (likely(!err)) goto repeat; return err; } +/** + * nilfs_segbuf_write - submit write requests of a log + * @segbuf: buffer storing a log to be written + * @nilfs: nilfs object + * + * Return Value: On Success, 0 is returned. On Error, one of the following + * negative error code is returned. + * + * %-EIO - I/O error + * + * %-ENOMEM - Insufficient memory available. + */ int nilfs_segbuf_write(struct nilfs_segment_buffer *segbuf, - struct nilfs_write_info *wi) + struct the_nilfs *nilfs) { + struct nilfs_write_info wi; struct buffer_head *bh; - int res, rw = WRITE; + int res = 0, rw = WRITE; + + wi.nilfs = nilfs; + nilfs_segbuf_prepare_write(segbuf, &wi); list_for_each_entry(bh, &segbuf->sb_segsum_buffers, b_assoc_buffers) { - res = nilfs_submit_bh(wi, bh, rw); + res = nilfs_segbuf_submit_bh(segbuf, &wi, bh, rw); if (unlikely(res)) goto failed_bio; } list_for_each_entry(bh, &segbuf->sb_payload_buffers, b_assoc_buffers) { - res = nilfs_submit_bh(wi, bh, rw); + res = nilfs_segbuf_submit_bh(segbuf, &wi, bh, rw); if (unlikely(res)) goto failed_bio; } - if (wi->bio) { + if (wi.bio) { /* * Last BIO is always sent through the following * submission. */ rw |= (1 << BIO_RW_SYNCIO) | (1 << BIO_RW_UNPLUG); - res = nilfs_submit_seg_bio(wi, rw); - if (unlikely(res)) - goto failed_bio; + res = nilfs_segbuf_submit_bio(segbuf, &wi, rw); } - res = 0; - out: - return res; - failed_bio: - atomic_inc(&wi->err); - goto out; + return res; } /** * nilfs_segbuf_wait - wait for completion of requested BIOs - * @wi: nilfs_write_info + * @segbuf: segment buffer * * Return Value: On Success, 0 is returned. On Error, one of the following * negative error code is returned. * * %-EIO - I/O error */ -int nilfs_segbuf_wait(struct nilfs_segment_buffer *segbuf, - struct nilfs_write_info *wi) +int nilfs_segbuf_wait(struct nilfs_segment_buffer *segbuf) { int err = 0; - if (!wi->nbio) + if (!segbuf->sb_nbio) return 0; do { - wait_for_completion(&wi->bio_event); - } while (--wi->nbio > 0); + wait_for_completion(&segbuf->sb_bio_event); + } while (--segbuf->sb_nbio > 0); - if (unlikely(atomic_read(&wi->err) > 0)) { + if (unlikely(atomic_read(&segbuf->sb_err) > 0)) { printk(KERN_ERR "NILFS: IO error writing segment\n"); err = -EIO; - segbuf->sb_io_error = 1; } return err; } diff --git a/fs/nilfs2/segbuf.h b/fs/nilfs2/segbuf.h index 0c3076f4e59..6af1630fb40 100644 --- a/fs/nilfs2/segbuf.h +++ b/fs/nilfs2/segbuf.h @@ -27,7 +27,6 @@ #include <linux/buffer_head.h> #include <linux/bio.h> #include <linux/completion.h> -#include <linux/backing-dev.h> /** * struct nilfs_segsum_info - On-memory segment summary @@ -77,7 +76,9 @@ struct nilfs_segsum_info { * @sb_rest_blocks: Number of residual blocks in the current segment * @sb_segsum_buffers: List of buffers for segment summaries * @sb_payload_buffers: List of buffers for segment payload - * @sb_io_error: I/O error status + * @sb_nbio: Number of flying bio requests + * @sb_err: I/O error status + * @sb_bio_event: Completion event of log writing */ struct nilfs_segment_buffer { struct super_block *sb_super; @@ -96,7 +97,9 @@ struct nilfs_segment_buffer { struct list_head sb_payload_buffers; /* including super root */ /* io status */ - int sb_io_error; + int sb_nbio; + atomic_t sb_err; + struct completion sb_bio_event; }; #define NILFS_LIST_SEGBUF(head) \ @@ -125,6 +128,8 @@ struct nilfs_segment_buffer *nilfs_segbuf_new(struct super_block *); void nilfs_segbuf_free(struct nilfs_segment_buffer *); void nilfs_segbuf_map(struct nilfs_segment_buffer *, __u64, unsigned long, struct the_nilfs *); +void nilfs_segbuf_map_cont(struct nilfs_segment_buffer *segbuf, + struct nilfs_segment_buffer *prev); void nilfs_segbuf_set_next_segnum(struct nilfs_segment_buffer *, __u64, struct the_nilfs *); int nilfs_segbuf_reset(struct nilfs_segment_buffer *, unsigned, time_t); @@ -161,41 +166,18 @@ nilfs_segbuf_add_file_buffer(struct nilfs_segment_buffer *segbuf, segbuf->sb_sum.nfileblk++; } -void nilfs_release_buffers(struct list_head *); +int nilfs_segbuf_write(struct nilfs_segment_buffer *segbuf, + struct the_nilfs *nilfs); +int nilfs_segbuf_wait(struct nilfs_segment_buffer *segbuf); -static inline void nilfs_segbuf_clear(struct nilfs_segment_buffer *segbuf) +void nilfs_clear_logs(struct list_head *logs); +void nilfs_truncate_logs(struct list_head *logs, + struct nilfs_segment_buffer *last); +int nilfs_wait_on_logs(struct list_head *logs); + +static inline void nilfs_destroy_logs(struct list_head *logs) { - nilfs_release_buffers(&segbuf->sb_segsum_buffers); - nilfs_release_buffers(&segbuf->sb_payload_buffers); + nilfs_truncate_logs(logs, NULL); } -struct nilfs_write_info { - struct bio *bio; - int start, end; /* The region to be submitted */ - int rest_blocks; - int max_pages; - int nr_vecs; - sector_t blocknr; - - int nbio; - atomic_t err; - struct completion bio_event; - /* completion event of segment write */ - - /* - * The following fields must be set explicitly - */ - struct super_block *sb; - struct backing_dev_info *bdi; /* backing dev info */ - struct buffer_head *bh_sr; -}; - - -void nilfs_segbuf_prepare_write(struct nilfs_segment_buffer *, - struct nilfs_write_info *); -int nilfs_segbuf_write(struct nilfs_segment_buffer *, - struct nilfs_write_info *); -int nilfs_segbuf_wait(struct nilfs_segment_buffer *, - struct nilfs_write_info *); - #endif /* _NILFS_SEGBUF_H */ diff --git a/fs/nilfs2/segment.c b/fs/nilfs2/segment.c index 6eff66a070d..17584c52448 100644 --- a/fs/nilfs2/segment.c +++ b/fs/nilfs2/segment.c @@ -974,12 +974,12 @@ static void nilfs_segctor_fill_in_super_root(struct nilfs_sc_info *sci, nilfs->ns_nongc_ctime : sci->sc_seg_ctime); raw_sr->sr_flags = 0; - nilfs_mdt_write_inode_direct( - nilfs_dat_inode(nilfs), bh_sr, NILFS_SR_DAT_OFFSET(isz)); - nilfs_mdt_write_inode_direct( - nilfs->ns_cpfile, bh_sr, NILFS_SR_CPFILE_OFFSET(isz)); - nilfs_mdt_write_inode_direct( - nilfs->ns_sufile, bh_sr, NILFS_SR_SUFILE_OFFSET(isz)); + nilfs_write_inode_common(nilfs_dat_inode(nilfs), (void *)raw_sr + + NILFS_SR_DAT_OFFSET(isz), 1); + nilfs_write_inode_common(nilfs->ns_cpfile, (void *)raw_sr + + NILFS_SR_CPFILE_OFFSET(isz), 1); + nilfs_write_inode_common(nilfs->ns_sufile, (void *)raw_sr + + NILFS_SR_SUFILE_OFFSET(isz), 1); } static void nilfs_redirty_inodes(struct list_head *head) @@ -1273,73 +1273,75 @@ static int nilfs_segctor_collect_blocks(struct nilfs_sc_info *sci, int mode) return err; } -static int nilfs_touch_segusage(struct inode *sufile, __u64 segnum) -{ - struct buffer_head *bh_su; - struct nilfs_segment_usage *raw_su; - int err; - - err = nilfs_sufile_get_segment_usage(sufile, segnum, &raw_su, &bh_su); - if (unlikely(err)) - return err; - nilfs_mdt_mark_buffer_dirty(bh_su); - nilfs_mdt_mark_dirty(sufile); - nilfs_sufile_put_segment_usage(sufile, segnum, bh_su); - return 0; -} - +/** + * nilfs_segctor_begin_construction - setup segment buffer to make a new log + * @sci: nilfs_sc_info + * @nilfs: nilfs object + */ static int nilfs_segctor_begin_construction(struct nilfs_sc_info *sci, struct the_nilfs *nilfs) { - struct nilfs_segment_buffer *segbuf, *n; + struct nilfs_segment_buffer *segbuf, *prev; __u64 nextnum; - int err; + int err, alloc = 0; - if (list_empty(&sci->sc_segbufs)) { - segbuf = nilfs_segbuf_new(sci->sc_super); - if (unlikely(!segbuf)) - return -ENOMEM; - list_add(&segbuf->sb_list, &sci->sc_segbufs); - } else - segbuf = NILFS_FIRST_SEGBUF(&sci->sc_segbufs); + segbuf = nilfs_segbuf_new(sci->sc_super); + if (unlikely(!segbuf)) + return -ENOMEM; + + if (list_empty(&sci->sc_write_logs)) { + nilfs_segbuf_map(segbuf, nilfs->ns_segnum, + nilfs->ns_pseg_offset, nilfs); + if (segbuf->sb_rest_blocks < NILFS_PSEG_MIN_BLOCKS) { + nilfs_shift_to_next_segment(nilfs); + nilfs_segbuf_map(segbuf, nilfs->ns_segnum, 0, nilfs); + } - nilfs_segbuf_map(segbuf, nilfs->ns_segnum, nilfs->ns_pseg_offset, - nilfs); + segbuf->sb_sum.seg_seq = nilfs->ns_seg_seq; + nextnum = nilfs->ns_nextnum; - if (segbuf->sb_rest_blocks < NILFS_PSEG_MIN_BLOCKS) { - nilfs_shift_to_next_segment(nilfs); - nilfs_segbuf_map(segbuf, nilfs->ns_segnum, 0, nilfs); + if (nilfs->ns_segnum == nilfs->ns_nextnum) + /* Start from the head of a new full segment */ + alloc++; + } else { + /* Continue logs */ + prev = NILFS_LAST_SEGBUF(&sci->sc_write_logs); + nilfs_segbuf_map_cont(segbuf, prev); + segbuf->sb_sum.seg_seq = prev->sb_sum.seg_seq; + nextnum = prev->sb_nextnum; + + if (segbuf->sb_rest_blocks < NILFS_PSEG_MIN_BLOCKS) { + nilfs_segbuf_map(segbuf, prev->sb_nextnum, 0, nilfs); + segbuf->sb_sum.seg_seq++; + alloc++; + } } - sci->sc_segbuf_nblocks = segbuf->sb_rest_blocks; - err = nilfs_touch_segusage(nilfs->ns_sufile, segbuf->sb_segnum); - if (unlikely(err)) - return err; + err = nilfs_sufile_mark_dirty(nilfs->ns_sufile, segbuf->sb_segnum); + if (err) + goto failed; - if (nilfs->ns_segnum == nilfs->ns_nextnum) { - /* Start from the head of a new full segment */ + if (alloc) { err = nilfs_sufile_alloc(nilfs->ns_sufile, &nextnum); - if (unlikely(err)) - return err; - } else - nextnum = nilfs->ns_nextnum; - - segbuf->sb_sum.seg_seq = nilfs->ns_seg_seq; + if (err) + goto failed; + } nilfs_segbuf_set_next_segnum(segbuf, nextnum, nilfs); - /* truncating segment buffers */ - list_for_each_entry_safe_continue(segbuf, n, &sci->sc_segbufs, - sb_list) { - list_del_init(&segbuf->sb_list); - nilfs_segbuf_free(segbuf); - } + BUG_ON(!list_empty(&sci->sc_segbufs)); + list_add_tail(&segbuf->sb_list, &sci->sc_segbufs); + sci->sc_segbuf_nblocks = segbuf->sb_rest_blocks; return 0; + + failed: + nilfs_segbuf_free(segbuf); + return err; } static int nilfs_segctor_extend_segments(struct nilfs_sc_info *sci, struct the_nilfs *nilfs, int nadd) { - struct nilfs_segment_buffer *segbuf, *prev, *n; + struct nilfs_segment_buffer *segbuf, *prev; struct inode *sufile = nilfs->ns_sufile; __u64 nextnextnum; LIST_HEAD(list); @@ -1352,7 +1354,7 @@ static int nilfs_segctor_extend_segments(struct nilfs_sc_info *sci, * not be dirty. The following call ensures that the buffer is dirty * and will pin the buffer on memory until the sufile is written. */ - err = nilfs_touch_segusage(sufile, prev->sb_nextnum); + err = nilfs_sufile_mark_dirty(sufile, prev->sb_nextnum); if (unlikely(err)) return err; @@ -1378,33 +1380,33 @@ static int nilfs_segctor_extend_segments(struct nilfs_sc_info *sci, list_add_tail(&segbuf->sb_list, &list); prev = segbuf; } - list_splice(&list, sci->sc_segbufs.prev); + list_splice_tail(&list, &sci->sc_segbufs); return 0; failed_segbuf: nilfs_segbuf_free(segbuf); failed: - list_for_each_entry_safe(segbuf, n, &list, sb_list) { + list_for_each_entry(segbuf, &list, sb_list) { ret = nilfs_sufile_free(sufile, segbuf->sb_nextnum); WARN_ON(ret); /* never fails */ - list_del_init(&segbuf->sb_list); - nilfs_segbuf_free(segbuf); } + nilfs_destroy_logs(&list); return err; } -static void nilfs_segctor_free_incomplete_segments(struct nilfs_sc_info *sci, - struct the_nilfs *nilfs) +static void nilfs_free_incomplete_logs(struct list_head *logs, + struct the_nilfs *nilfs) { - struct nilfs_segment_buffer *segbuf; - int ret, done = 0; + struct nilfs_segment_buffer *segbuf, *prev; + struct inode *sufile = nilfs->ns_sufile; + int ret; - segbuf = NILFS_FIRST_SEGBUF(&sci->sc_segbufs); + segbuf = NILFS_FIRST_SEGBUF(logs); if (nilfs->ns_nextnum != segbuf->sb_nextnum) { - ret = nilfs_sufile_free(nilfs->ns_sufile, segbuf->sb_nextnum); + ret = nilfs_sufile_free(sufile, segbuf->sb_nextnum); WARN_ON(ret); /* never fails */ } - if (segbuf->sb_io_error) { + if (atomic_read(&segbuf->sb_err)) { /* Case 1: The first segment failed */ if (segbuf->sb_pseg_start != segbuf->sb_fseg_start) /* Case 1a: Partial segment appended into an existing @@ -1413,106 +1415,54 @@ static void nilfs_segctor_free_incomplete_segments(struct nilfs_sc_info *sci, segbuf->sb_fseg_end); else /* Case 1b: New full segment */ set_nilfs_discontinued(nilfs); - done++; } - list_for_each_entry_continue(segbuf, &sci->sc_segbufs, sb_list) { - ret = nilfs_sufile_free(nilfs->ns_sufile, segbuf->sb_nextnum); - WARN_ON(ret); /* never fails */ - if (!done && segbuf->sb_io_error) { - if (segbuf->sb_segnum != nilfs->ns_nextnum) - /* Case 2: extended segment (!= next) failed */ - nilfs_sufile_set_error(nilfs->ns_sufile, - segbuf->sb_segnum); - done++; - } - } -} - -static void nilfs_segctor_clear_segment_buffers(struct nilfs_sc_info *sci) -{ - struct nilfs_segment_buffer *segbuf; - - list_for_each_entry(segbuf, &sci->sc_segbufs, sb_list) - nilfs_segbuf_clear(segbuf); - sci->sc_super_root = NULL; -} - -static void nilfs_segctor_destroy_segment_buffers(struct nilfs_sc_info *sci) -{ - struct nilfs_segment_buffer *segbuf; - - while (!list_empty(&sci->sc_segbufs)) { - segbuf = NILFS_FIRST_SEGBUF(&sci->sc_segbufs); - list_del_init(&segbuf->sb_list); - nilfs_segbuf_free(segbuf); - } - /* sci->sc_curseg = NULL; */ -} - -static void nilfs_segctor_end_construction(struct nilfs_sc_info *sci, - struct the_nilfs *nilfs, int err) -{ - if (unlikely(err)) { - nilfs_segctor_free_incomplete_segments(sci, nilfs); - if (sci->sc_stage.flags & NILFS_CF_SUFREED) { - int ret; - - ret = nilfs_sufile_cancel_freev(nilfs->ns_sufile, - sci->sc_freesegs, - sci->sc_nfreesegs, - NULL); - WARN_ON(ret); /* do not happen */ + prev = segbuf; + list_for_each_entry_continue(segbuf, logs, sb_list) { + if (prev->sb_nextnum != segbuf->sb_nextnum) { + ret = nilfs_sufile_free(sufile, segbuf->sb_nextnum); + WARN_ON(ret); /* never fails */ } + if (atomic_read(&segbuf->sb_err) && + segbuf->sb_segnum != nilfs->ns_nextnum) + /* Case 2: extended segment (!= next) failed */ + nilfs_sufile_set_error(sufile, segbuf->sb_segnum); + prev = segbuf; } - nilfs_segctor_clear_segment_buffers(sci); } static void nilfs_segctor_update_segusage(struct nilfs_sc_info *sci, struct inode *sufile) { struct nilfs_segment_buffer *segbuf; - struct buffer_head *bh_su; - struct nilfs_segment_usage *raw_su; unsigned long live_blocks; int ret; list_for_each_entry(segbuf, &sci->sc_segbufs, sb_list) { - ret = nilfs_sufile_get_segment_usage(sufile, segbuf->sb_segnum, - &raw_su, &bh_su); - WARN_ON(ret); /* always succeed because bh_su is dirty */ live_blocks = segbuf->sb_sum.nblocks + (segbuf->sb_pseg_start - segbuf->sb_fseg_start); - raw_su->su_lastmod = cpu_to_le64(sci->sc_seg_ctime); - raw_su->su_nblocks = cpu_to_le32(live_blocks); - nilfs_sufile_put_segment_usage(sufile, segbuf->sb_segnum, - bh_su); + ret = nilfs_sufile_set_segment_usage(sufile, segbuf->sb_segnum, + live_blocks, + sci->sc_seg_ctime); + WARN_ON(ret); /* always succeed because the segusage is dirty */ } } -static void nilfs_segctor_cancel_segusage(struct nilfs_sc_info *sci, - struct inode *sufile) +static void nilfs_cancel_segusage(struct list_head *logs, struct inode *sufile) { struct nilfs_segment_buffer *segbuf; - struct buffer_head *bh_su; - struct nilfs_segment_usage *raw_su; int ret; - segbuf = NILFS_FIRST_SEGBUF(&sci->sc_segbufs); - ret = nilfs_sufile_get_segment_usage(sufile, segbuf->sb_segnum, - &raw_su, &bh_su); - WARN_ON(ret); /* always succeed because bh_su is dirty */ - raw_su->su_nblocks = cpu_to_le32(segbuf->sb_pseg_start - - segbuf->sb_fseg_start); - nilfs_sufile_put_segment_usage(sufile, segbuf->sb_segnum, bh_su); + segbuf = NILFS_FIRST_SEGBUF(logs); + ret = nilfs_sufile_set_segment_usage(sufile, segbuf->sb_segnum, + segbuf->sb_pseg_start - + segbuf->sb_fseg_start, 0); + WARN_ON(ret); /* always succeed because the segusage is dirty */ - list_for_each_entry_continue(segbuf, &sci->sc_segbufs, sb_list) { - ret = nilfs_sufile_get_segment_usage(sufile, segbuf->sb_segnum, - &raw_su, &bh_su); + list_for_each_entry_continue(segbuf, logs, sb_list) { + ret = nilfs_sufile_set_segment_usage(sufile, segbuf->sb_segnum, + 0, 0); WARN_ON(ret); /* always succeed */ - raw_su->su_nblocks = 0; - nilfs_sufile_put_segment_usage(sufile, segbuf->sb_segnum, - bh_su); } } @@ -1520,17 +1470,15 @@ static void nilfs_segctor_truncate_segments(struct nilfs_sc_info *sci, struct nilfs_segment_buffer *last, struct inode *sufile) { - struct nilfs_segment_buffer *segbuf = last, *n; + struct nilfs_segment_buffer *segbuf = last; int ret; - list_for_each_entry_safe_continue(segbuf, n, &sci->sc_segbufs, - sb_list) { - list_del_init(&segbuf->sb_list); + list_for_each_entry_continue(segbuf, &sci->sc_segbufs, sb_list) { sci->sc_segbuf_nblocks -= segbuf->sb_rest_blocks; ret = nilfs_sufile_free(sufile, segbuf->sb_nextnum); WARN_ON(ret); - nilfs_segbuf_free(segbuf); } + nilfs_truncate_logs(&sci->sc_segbufs, last); } @@ -1569,7 +1517,7 @@ static int nilfs_segctor_collect(struct nilfs_sc_info *sci, NULL); WARN_ON(err); /* do not happen */ } - nilfs_segctor_clear_segment_buffers(sci); + nilfs_clear_logs(&sci->sc_segbufs); err = nilfs_segctor_extend_segments(sci, nilfs, nadd); if (unlikely(err)) @@ -1814,26 +1762,18 @@ static int nilfs_segctor_prepare_write(struct nilfs_sc_info *sci, } static int nilfs_segctor_write(struct nilfs_sc_info *sci, - struct backing_dev_info *bdi) + struct the_nilfs *nilfs) { struct nilfs_segment_buffer *segbuf; - struct nilfs_write_info wi; - int err, res; - - wi.sb = sci->sc_super; - wi.bh_sr = sci->sc_super_root; - wi.bdi = bdi; + int ret = 0; list_for_each_entry(segbuf, &sci->sc_segbufs, sb_list) { - nilfs_segbuf_prepare_write(segbuf, &wi); - err = nilfs_segbuf_write(segbuf, &wi); - - res = nilfs_segbuf_wait(segbuf, &wi); - err = err ? : res; - if (err) - return err; + ret = nilfs_segbuf_write(segbuf, nilfs); + if (ret) + break; } - return 0; + list_splice_tail_init(&sci->sc_segbufs, &sci->sc_write_logs); + return ret; } static void __nilfs_end_page_io(struct page *page, int err) @@ -1911,15 +1851,17 @@ static void nilfs_clear_copied_buffers(struct list_head *list, int err) } } -static void nilfs_segctor_abort_write(struct nilfs_sc_info *sci, - struct page *failed_page, int err) +static void nilfs_abort_logs(struct list_head *logs, struct page *failed_page, + struct buffer_head *bh_sr, int err) { struct nilfs_segment_buffer *segbuf; struct page *bd_page = NULL, *fs_page = NULL; + struct buffer_head *bh; - list_for_each_entry(segbuf, &sci->sc_segbufs, sb_list) { - struct buffer_head *bh; + if (list_empty(logs)) + return; + list_for_each_entry(segbuf, logs, sb_list) { list_for_each_entry(bh, &segbuf->sb_segsum_buffers, b_assoc_buffers) { if (bh->b_page != bd_page) { @@ -1931,7 +1873,7 @@ static void nilfs_segctor_abort_write(struct nilfs_sc_info *sci, list_for_each_entry(bh, &segbuf->sb_payload_buffers, b_assoc_buffers) { - if (bh == sci->sc_super_root) { + if (bh == bh_sr) { if (bh->b_page != bd_page) { end_page_writeback(bd_page); bd_page = bh->b_page; @@ -1941,7 +1883,7 @@ static void nilfs_segctor_abort_write(struct nilfs_sc_info *sci, if (bh->b_page != fs_page) { nilfs_end_page_io(fs_page, err); if (fs_page && fs_page == failed_page) - goto done; + return; fs_page = bh->b_page; } } @@ -1950,8 +1892,34 @@ static void nilfs_segctor_abort_write(struct nilfs_sc_info *sci, end_page_writeback(bd_page); nilfs_end_page_io(fs_page, err); - done: +} + +static void nilfs_segctor_abort_construction(struct nilfs_sc_info *sci, + struct the_nilfs *nilfs, int err) +{ + LIST_HEAD(logs); + int ret; + + list_splice_tail_init(&sci->sc_write_logs, &logs); + ret = nilfs_wait_on_logs(&logs); + if (ret) + nilfs_abort_logs(&logs, NULL, sci->sc_super_root, ret); + + list_splice_tail_init(&sci->sc_segbufs, &logs); + nilfs_cancel_segusage(&logs, nilfs->ns_sufile); + nilfs_free_incomplete_logs(&logs, nilfs); nilfs_clear_copied_buffers(&sci->sc_copied_buffers, err); + + if (sci->sc_stage.flags & NILFS_CF_SUFREED) { + ret = nilfs_sufile_cancel_freev(nilfs->ns_sufile, + sci->sc_freesegs, + sci->sc_nfreesegs, + NULL); + WARN_ON(ret); /* do not happen */ + } + + nilfs_destroy_logs(&logs); + sci->sc_super_root = NULL; } static void nilfs_set_next_segment(struct the_nilfs *nilfs, @@ -1973,7 +1941,7 @@ static void nilfs_segctor_complete_write(struct nilfs_sc_info *sci) struct the_nilfs *nilfs = sbi->s_nilfs; int update_sr = (sci->sc_super_root != NULL); - list_for_each_entry(segbuf, &sci->sc_segbufs, sb_list) { + list_for_each_entry(segbuf, &sci->sc_write_logs, sb_list) { struct buffer_head *bh; list_for_each_entry(bh, &segbuf->sb_segsum_buffers, @@ -2046,7 +2014,7 @@ static void nilfs_segctor_complete_write(struct nilfs_sc_info *sci) sci->sc_nblk_inc += sci->sc_nblk_this_inc; - segbuf = NILFS_LAST_SEGBUF(&sci->sc_segbufs); + segbuf = NILFS_LAST_SEGBUF(&sci->sc_write_logs); nilfs_set_next_segment(nilfs, segbuf); if (update_sr) { @@ -2057,10 +2025,23 @@ static void nilfs_segctor_complete_write(struct nilfs_sc_info *sci) clear_bit(NILFS_SC_HAVE_DELTA, &sci->sc_flags); clear_bit(NILFS_SC_DIRTY, &sci->sc_flags); set_bit(NILFS_SC_SUPER_ROOT, &sci->sc_flags); + nilfs_segctor_clear_metadata_dirty(sci); } else clear_bit(NILFS_SC_SUPER_ROOT, &sci->sc_flags); } +static int nilfs_segctor_wait(struct nilfs_sc_info *sci) +{ + int ret; + + ret = nilfs_wait_on_logs(&sci->sc_write_logs); + if (!ret) { + nilfs_segctor_complete_write(sci); + nilfs_destroy_logs(&sci->sc_write_logs); + } + return ret; +} + static int nilfs_segctor_check_in_files(struct nilfs_sc_info *sci, struct nilfs_sb_info *sbi) { @@ -2173,7 +2154,7 @@ static int nilfs_segctor_do_construct(struct nilfs_sc_info *sci, int mode) /* Avoid empty segment */ if (sci->sc_stage.scnt == NILFS_ST_DONE && NILFS_SEG_EMPTY(&sci->sc_curseg->sb_sum)) { - nilfs_segctor_end_construction(sci, nilfs, 1); + nilfs_segctor_abort_construction(sci, nilfs, 1); goto out; } @@ -2187,7 +2168,7 @@ static int nilfs_segctor_do_construct(struct nilfs_sc_info *sci, int mode) if (has_sr) { err = nilfs_segctor_fill_in_checkpoint(sci); if (unlikely(err)) - goto failed_to_make_up; + goto failed_to_write; nilfs_segctor_fill_in_super_root(sci, nilfs); } @@ -2195,42 +2176,46 @@ static int nilfs_segctor_do_construct(struct nilfs_sc_info *sci, int mode) /* Write partial segments */ err = nilfs_segctor_prepare_write(sci, &failed_page); - if (unlikely(err)) + if (err) { + nilfs_abort_logs(&sci->sc_segbufs, failed_page, + sci->sc_super_root, err); goto failed_to_write; - + } nilfs_segctor_fill_in_checksums(sci, nilfs->ns_crc_seed); - err = nilfs_segctor_write(sci, nilfs->ns_bdi); + err = nilfs_segctor_write(sci, nilfs); if (unlikely(err)) goto failed_to_write; - nilfs_segctor_complete_write(sci); - - /* Commit segments */ - if (has_sr) - nilfs_segctor_clear_metadata_dirty(sci); - - nilfs_segctor_end_construction(sci, nilfs, 0); - + if (sci->sc_stage.scnt == NILFS_ST_DONE || + nilfs->ns_blocksize_bits != PAGE_CACHE_SHIFT) { + /* + * At this point, we avoid double buffering + * for blocksize < pagesize because page dirty + * flag is turned off during write and dirty + * buffers are not properly collected for + * pages crossing over segments. + */ + err = nilfs_segctor_wait(sci); + if (err) + goto failed_to_write; + } } while (sci->sc_stage.scnt != NILFS_ST_DONE); + sci->sc_super_root = NULL; + out: - nilfs_segctor_destroy_segment_buffers(sci); nilfs_segctor_check_out_files(sci, sbi); return err; failed_to_write: - nilfs_segctor_abort_write(sci, failed_page, err); - nilfs_segctor_cancel_segusage(sci, nilfs->ns_sufile); - - failed_to_make_up: if (sci->sc_stage.flags & NILFS_CF_IFILE_STARTED) nilfs_redirty_inodes(&sci->sc_dirty_files); failed: if (nilfs_doing_gc()) nilfs_redirty_inodes(&sci->sc_gc_inodes); - nilfs_segctor_end_construction(sci, nilfs, err); + nilfs_segctor_abort_construction(sci, nilfs, err); goto out; } @@ -2559,7 +2544,7 @@ int nilfs_clean_segments(struct super_block *sb, struct nilfs_argv *argv, sci->sc_freesegs = kbufs[4]; sci->sc_nfreesegs = argv[4].v_nmembs; - list_splice_init(&nilfs->ns_gc_inodes, sci->sc_gc_inodes.prev); + list_splice_tail_init(&nilfs->ns_gc_inodes, &sci->sc_gc_inodes); for (;;) { nilfs_segctor_accept(sci, &req); @@ -2788,6 +2773,7 @@ static struct nilfs_sc_info *nilfs_segctor_new(struct nilfs_sb_info *sbi) spin_lock_init(&sci->sc_state_lock); INIT_LIST_HEAD(&sci->sc_dirty_files); INIT_LIST_HEAD(&sci->sc_segbufs); + INIT_LIST_HEAD(&sci->sc_write_logs); INIT_LIST_HEAD(&sci->sc_gc_inodes); INIT_LIST_HEAD(&sci->sc_copied_buffers); @@ -2855,6 +2841,7 @@ static void nilfs_segctor_destroy(struct nilfs_sc_info *sci) } WARN_ON(!list_empty(&sci->sc_segbufs)); + WARN_ON(!list_empty(&sci->sc_write_logs)); down_write(&sbi->s_nilfs->ns_segctor_sem); diff --git a/fs/nilfs2/segment.h b/fs/nilfs2/segment.h index 0d2a475a741..3d3ab2f9864 100644 --- a/fs/nilfs2/segment.h +++ b/fs/nilfs2/segment.h @@ -97,6 +97,7 @@ struct nilfs_segsum_pointer { * @sc_dsync_start: start byte offset of data pages * @sc_dsync_end: end byte offset of data pages (inclusive) * @sc_segbufs: List of segment buffers + * @sc_write_logs: List of segment buffers to hold logs under writing * @sc_segbuf_nblocks: Number of available blocks in segment buffers. * @sc_curseg: Current segment buffer * @sc_super_root: Pointer to the super root buffer @@ -143,6 +144,7 @@ struct nilfs_sc_info { /* Segment buffers */ struct list_head sc_segbufs; + struct list_head sc_write_logs; unsigned long sc_segbuf_nblocks; struct nilfs_segment_buffer *sc_curseg; struct buffer_head *sc_super_root; diff --git a/fs/nilfs2/sufile.c b/fs/nilfs2/sufile.c index 37994d4a59c..b6c36d0cc33 100644 --- a/fs/nilfs2/sufile.c +++ b/fs/nilfs2/sufile.c @@ -31,6 +31,16 @@ #include "sufile.h" +struct nilfs_sufile_info { + struct nilfs_mdt_info mi; + unsigned long ncleansegs; +}; + +static inline struct nilfs_sufile_info *NILFS_SUI(struct inode *sufile) +{ + return (struct nilfs_sufile_info *)NILFS_MDT(sufile); +} + static inline unsigned long nilfs_sufile_segment_usages_per_block(const struct inode *sufile) { @@ -62,14 +72,6 @@ nilfs_sufile_segment_usages_in_block(const struct inode *sufile, __u64 curr, max - curr + 1); } -static inline struct nilfs_sufile_header * -nilfs_sufile_block_get_header(const struct inode *sufile, - struct buffer_head *bh, - void *kaddr) -{ - return kaddr + bh_offset(bh); -} - static struct nilfs_segment_usage * nilfs_sufile_block_get_segment_usage(const struct inode *sufile, __u64 segnum, struct buffer_head *bh, void *kaddr) @@ -110,6 +112,15 @@ static void nilfs_sufile_mod_counter(struct buffer_head *header_bh, } /** + * nilfs_sufile_get_ncleansegs - return the number of clean segments + * @sufile: inode of segment usage file + */ +unsigned long nilfs_sufile_get_ncleansegs(struct inode *sufile) +{ + return NILFS_SUI(sufile)->ncleansegs; +} + +/** * nilfs_sufile_updatev - modify multiple segment usages at a time * @sufile: inode of segment usage file * @segnumv: array of segment numbers @@ -270,7 +281,7 @@ int nilfs_sufile_alloc(struct inode *sufile, __u64 *segnump) if (ret < 0) goto out_sem; kaddr = kmap_atomic(header_bh->b_page, KM_USER0); - header = nilfs_sufile_block_get_header(sufile, header_bh, kaddr); + header = kaddr + bh_offset(header_bh); ncleansegs = le64_to_cpu(header->sh_ncleansegs); last_alloc = le64_to_cpu(header->sh_last_alloc); kunmap_atomic(kaddr, KM_USER0); @@ -302,13 +313,13 @@ int nilfs_sufile_alloc(struct inode *sufile, __u64 *segnump) kunmap_atomic(kaddr, KM_USER0); kaddr = kmap_atomic(header_bh->b_page, KM_USER0); - header = nilfs_sufile_block_get_header( - sufile, header_bh, kaddr); + header = kaddr + bh_offset(header_bh); le64_add_cpu(&header->sh_ncleansegs, -1); le64_add_cpu(&header->sh_ndirtysegs, 1); header->sh_last_alloc = cpu_to_le64(segnum); kunmap_atomic(kaddr, KM_USER0); + NILFS_SUI(sufile)->ncleansegs--; nilfs_mdt_mark_buffer_dirty(header_bh); nilfs_mdt_mark_buffer_dirty(su_bh); nilfs_mdt_mark_dirty(sufile); @@ -351,6 +362,8 @@ void nilfs_sufile_do_cancel_free(struct inode *sufile, __u64 segnum, kunmap_atomic(kaddr, KM_USER0); nilfs_sufile_mod_counter(header_bh, -1, 1); + NILFS_SUI(sufile)->ncleansegs--; + nilfs_mdt_mark_buffer_dirty(su_bh); nilfs_mdt_mark_dirty(sufile); } @@ -380,6 +393,8 @@ void nilfs_sufile_do_scrap(struct inode *sufile, __u64 segnum, kunmap_atomic(kaddr, KM_USER0); nilfs_sufile_mod_counter(header_bh, clean ? (u64)-1 : 0, dirty ? 0 : 1); + NILFS_SUI(sufile)->ncleansegs -= clean; + nilfs_mdt_mark_buffer_dirty(su_bh); nilfs_mdt_mark_dirty(sufile); } @@ -409,79 +424,65 @@ void nilfs_sufile_do_free(struct inode *sufile, __u64 segnum, nilfs_mdt_mark_buffer_dirty(su_bh); nilfs_sufile_mod_counter(header_bh, 1, sudirty ? (u64)-1 : 0); + NILFS_SUI(sufile)->ncleansegs++; + nilfs_mdt_mark_dirty(sufile); } /** - * nilfs_sufile_get_segment_usage - get a segment usage + * nilfs_sufile_mark_dirty - mark the buffer having a segment usage dirty * @sufile: inode of segment usage file * @segnum: segment number - * @sup: pointer to segment usage - * @bhp: pointer to buffer head - * - * Description: nilfs_sufile_get_segment_usage() acquires the segment usage - * specified by @segnum. - * - * Return Value: On success, 0 is returned, and the segment usage and the - * buffer head of the buffer on which the segment usage is located are stored - * in the place pointed by @sup and @bhp, respectively. On error, one of the - * following negative error codes is returned. - * - * %-EIO - I/O error. - * - * %-ENOMEM - Insufficient amount of memory available. - * - * %-EINVAL - Invalid segment usage number. */ -int nilfs_sufile_get_segment_usage(struct inode *sufile, __u64 segnum, - struct nilfs_segment_usage **sup, - struct buffer_head **bhp) +int nilfs_sufile_mark_dirty(struct inode *sufile, __u64 segnum) { struct buffer_head *bh; - struct nilfs_segment_usage *su; - void *kaddr; int ret; - /* segnum is 0 origin */ - if (segnum >= nilfs_sufile_get_nsegments(sufile)) - return -EINVAL; - down_write(&NILFS_MDT(sufile)->mi_sem); - ret = nilfs_sufile_get_segment_usage_block(sufile, segnum, 1, &bh); - if (ret < 0) - goto out_sem; - kaddr = kmap(bh->b_page); - su = nilfs_sufile_block_get_segment_usage(sufile, segnum, bh, kaddr); - if (nilfs_segment_usage_error(su)) { - kunmap(bh->b_page); + ret = nilfs_sufile_get_segment_usage_block(sufile, segnum, 0, &bh); + if (!ret) { + nilfs_mdt_mark_buffer_dirty(bh); + nilfs_mdt_mark_dirty(sufile); brelse(bh); - ret = -EINVAL; - goto out_sem; } - - if (sup != NULL) - *sup = su; - *bhp = bh; - - out_sem: - up_write(&NILFS_MDT(sufile)->mi_sem); return ret; } /** - * nilfs_sufile_put_segment_usage - put a segment usage + * nilfs_sufile_set_segment_usage - set usage of a segment * @sufile: inode of segment usage file * @segnum: segment number - * @bh: buffer head - * - * Description: nilfs_sufile_put_segment_usage() releases the segment usage - * specified by @segnum. @bh must be the buffer head which have been returned - * by a previous call to nilfs_sufile_get_segment_usage() with @segnum. + * @nblocks: number of live blocks in the segment + * @modtime: modification time (option) */ -void nilfs_sufile_put_segment_usage(struct inode *sufile, __u64 segnum, - struct buffer_head *bh) +int nilfs_sufile_set_segment_usage(struct inode *sufile, __u64 segnum, + unsigned long nblocks, time_t modtime) { - kunmap(bh->b_page); + struct buffer_head *bh; + struct nilfs_segment_usage *su; + void *kaddr; + int ret; + + down_write(&NILFS_MDT(sufile)->mi_sem); + ret = nilfs_sufile_get_segment_usage_block(sufile, segnum, 0, &bh); + if (ret < 0) + goto out_sem; + + kaddr = kmap_atomic(bh->b_page, KM_USER0); + su = nilfs_sufile_block_get_segment_usage(sufile, segnum, bh, kaddr); + WARN_ON(nilfs_segment_usage_error(su)); + if (modtime) + su->su_lastmod = cpu_to_le64(modtime); + su->su_nblocks = cpu_to_le32(nblocks); + kunmap_atomic(kaddr, KM_USER0); + + nilfs_mdt_mark_buffer_dirty(bh); + nilfs_mdt_mark_dirty(sufile); brelse(bh); + + out_sem: + up_write(&NILFS_MDT(sufile)->mi_sem); + return ret; } /** @@ -515,7 +516,7 @@ int nilfs_sufile_get_stat(struct inode *sufile, struct nilfs_sustat *sustat) goto out_sem; kaddr = kmap_atomic(header_bh->b_page, KM_USER0); - header = nilfs_sufile_block_get_header(sufile, header_bh, kaddr); + header = kaddr + bh_offset(header_bh); sustat->ss_nsegs = nilfs_sufile_get_nsegments(sufile); sustat->ss_ncleansegs = le64_to_cpu(header->sh_ncleansegs); sustat->ss_ndirtysegs = le64_to_cpu(header->sh_ndirtysegs); @@ -532,33 +533,6 @@ int nilfs_sufile_get_stat(struct inode *sufile, struct nilfs_sustat *sustat) return ret; } -/** - * nilfs_sufile_get_ncleansegs - get the number of clean segments - * @sufile: inode of segment usage file - * @nsegsp: pointer to the number of clean segments - * - * Description: nilfs_sufile_get_ncleansegs() acquires the number of clean - * segments. - * - * Return Value: On success, 0 is returned and the number of clean segments is - * stored in the place pointed by @nsegsp. On error, one of the following - * negative error codes is returned. - * - * %-EIO - I/O error. - * - * %-ENOMEM - Insufficient amount of memory available. - */ -int nilfs_sufile_get_ncleansegs(struct inode *sufile, unsigned long *nsegsp) -{ - struct nilfs_sustat sustat; - int ret; - - ret = nilfs_sufile_get_stat(sufile, &sustat); - if (ret == 0) - *nsegsp = sustat.ss_ncleansegs; - return ret; -} - void nilfs_sufile_do_set_error(struct inode *sufile, __u64 segnum, struct buffer_head *header_bh, struct buffer_head *su_bh) @@ -577,8 +551,10 @@ void nilfs_sufile_do_set_error(struct inode *sufile, __u64 segnum, nilfs_segment_usage_set_error(su); kunmap_atomic(kaddr, KM_USER0); - if (suclean) + if (suclean) { nilfs_sufile_mod_counter(header_bh, -1, 0); + NILFS_SUI(sufile)->ncleansegs--; + } nilfs_mdt_mark_buffer_dirty(su_bh); nilfs_mdt_mark_dirty(sufile); } @@ -657,3 +633,48 @@ ssize_t nilfs_sufile_get_suinfo(struct inode *sufile, __u64 segnum, void *buf, up_read(&NILFS_MDT(sufile)->mi_sem); return ret; } + +/** + * nilfs_sufile_read - read sufile inode + * @sufile: sufile inode + * @raw_inode: on-disk sufile inode + */ +int nilfs_sufile_read(struct inode *sufile, struct nilfs_inode *raw_inode) +{ + struct nilfs_sufile_info *sui = NILFS_SUI(sufile); + struct buffer_head *header_bh; + struct nilfs_sufile_header *header; + void *kaddr; + int ret; + + ret = nilfs_read_inode_common(sufile, raw_inode); + if (ret < 0) + return ret; + + ret = nilfs_sufile_get_header_block(sufile, &header_bh); + if (!ret) { + kaddr = kmap_atomic(header_bh->b_page, KM_USER0); + header = kaddr + bh_offset(header_bh); + sui->ncleansegs = le64_to_cpu(header->sh_ncleansegs); + kunmap_atomic(kaddr, KM_USER0); + brelse(header_bh); + } + return ret; +} + +/** + * nilfs_sufile_new - create sufile + * @nilfs: nilfs object + * @susize: size of a segment usage entry + */ +struct inode *nilfs_sufile_new(struct the_nilfs *nilfs, size_t susize) +{ + struct inode *sufile; + + sufile = nilfs_mdt_new(nilfs, NULL, NILFS_SUFILE_INO, + sizeof(struct nilfs_sufile_info)); + if (sufile) + nilfs_mdt_set_entry_size(sufile, susize, + sizeof(struct nilfs_sufile_header)); + return sufile; +} diff --git a/fs/nilfs2/sufile.h b/fs/nilfs2/sufile.h index 0e99e5c0bd0..15163b8aff7 100644 --- a/fs/nilfs2/sufile.h +++ b/fs/nilfs2/sufile.h @@ -34,14 +34,13 @@ static inline unsigned long nilfs_sufile_get_nsegments(struct inode *sufile) return NILFS_MDT(sufile)->mi_nilfs->ns_nsegments; } +unsigned long nilfs_sufile_get_ncleansegs(struct inode *sufile); + int nilfs_sufile_alloc(struct inode *, __u64 *); -int nilfs_sufile_get_segment_usage(struct inode *, __u64, - struct nilfs_segment_usage **, - struct buffer_head **); -void nilfs_sufile_put_segment_usage(struct inode *, __u64, - struct buffer_head *); +int nilfs_sufile_mark_dirty(struct inode *sufile, __u64 segnum); +int nilfs_sufile_set_segment_usage(struct inode *sufile, __u64 segnum, + unsigned long nblocks, time_t modtime); int nilfs_sufile_get_stat(struct inode *, struct nilfs_sustat *); -int nilfs_sufile_get_ncleansegs(struct inode *, unsigned long *); ssize_t nilfs_sufile_get_suinfo(struct inode *, __u64, void *, unsigned, size_t); @@ -62,6 +61,9 @@ void nilfs_sufile_do_cancel_free(struct inode *, __u64, struct buffer_head *, void nilfs_sufile_do_set_error(struct inode *, __u64, struct buffer_head *, struct buffer_head *); +int nilfs_sufile_read(struct inode *sufile, struct nilfs_inode *raw_inode); +struct inode *nilfs_sufile_new(struct the_nilfs *nilfs, size_t susize); + /** * nilfs_sufile_scrap - make a segment garbage * @sufile: inode of segment usage file diff --git a/fs/nilfs2/super.c b/fs/nilfs2/super.c index 644e66727dd..5403b3ef3a4 100644 --- a/fs/nilfs2/super.c +++ b/fs/nilfs2/super.c @@ -363,14 +363,10 @@ int nilfs_attach_checkpoint(struct nilfs_sb_info *sbi, __u64 cno) list_add(&sbi->s_list, &nilfs->ns_supers); up_write(&nilfs->ns_super_sem); - sbi->s_ifile = nilfs_mdt_new(nilfs, sbi->s_super, NILFS_IFILE_INO); + sbi->s_ifile = nilfs_ifile_new(sbi, nilfs->ns_inode_size); if (!sbi->s_ifile) return -ENOMEM; - err = nilfs_palloc_init_blockgroup(sbi->s_ifile, nilfs->ns_inode_size); - if (unlikely(err)) - goto failed; - down_read(&nilfs->ns_segctor_sem); err = nilfs_cpfile_get_checkpoint(nilfs->ns_cpfile, cno, 0, &raw_cp, &bh_cp); @@ -411,7 +407,6 @@ void nilfs_detach_checkpoint(struct nilfs_sb_info *sbi) { struct the_nilfs *nilfs = sbi->s_nilfs; - nilfs_mdt_clear(sbi->s_ifile); nilfs_mdt_destroy(sbi->s_ifile); sbi->s_ifile = NULL; down_write(&nilfs->ns_super_sem); @@ -419,22 +414,6 @@ void nilfs_detach_checkpoint(struct nilfs_sb_info *sbi) up_write(&nilfs->ns_super_sem); } -static int nilfs_mark_recovery_complete(struct nilfs_sb_info *sbi) -{ - struct the_nilfs *nilfs = sbi->s_nilfs; - int err = 0; - - down_write(&nilfs->ns_sem); - if (!(nilfs->ns_mount_state & NILFS_VALID_FS)) { - nilfs->ns_mount_state |= NILFS_VALID_FS; - err = nilfs_commit_super(sbi, 1); - if (likely(!err)) - printk(KERN_INFO "NILFS: recovery complete.\n"); - } - up_write(&nilfs->ns_sem); - return err; -} - static int nilfs_statfs(struct dentry *dentry, struct kstatfs *buf) { struct super_block *sb = dentry->d_sb; @@ -490,7 +469,7 @@ static int nilfs_show_options(struct seq_file *seq, struct vfsmount *vfs) struct nilfs_sb_info *sbi = NILFS_SB(sb); if (!nilfs_test_opt(sbi, BARRIER)) - seq_printf(seq, ",barrier=off"); + seq_printf(seq, ",nobarrier"); if (nilfs_test_opt(sbi, SNAPSHOT)) seq_printf(seq, ",cp=%llu", (unsigned long long int)sbi->s_snapshot_cno); @@ -500,6 +479,8 @@ static int nilfs_show_options(struct seq_file *seq, struct vfsmount *vfs) seq_printf(seq, ",errors=panic"); if (nilfs_test_opt(sbi, STRICT_ORDER)) seq_printf(seq, ",order=strict"); + if (nilfs_test_opt(sbi, NORECOVERY)) + seq_printf(seq, ",norecovery"); return 0; } @@ -568,7 +549,7 @@ static const struct export_operations nilfs_export_ops = { enum { Opt_err_cont, Opt_err_panic, Opt_err_ro, - Opt_barrier, Opt_snapshot, Opt_order, + Opt_nobarrier, Opt_snapshot, Opt_order, Opt_norecovery, Opt_err, }; @@ -576,25 +557,13 @@ static match_table_t tokens = { {Opt_err_cont, "errors=continue"}, {Opt_err_panic, "errors=panic"}, {Opt_err_ro, "errors=remount-ro"}, - {Opt_barrier, "barrier=%s"}, + {Opt_nobarrier, "nobarrier"}, {Opt_snapshot, "cp=%u"}, {Opt_order, "order=%s"}, + {Opt_norecovery, "norecovery"}, {Opt_err, NULL} }; -static int match_bool(substring_t *s, int *result) -{ - int len = s->to - s->from; - - if (strncmp(s->from, "on", len) == 0) - *result = 1; - else if (strncmp(s->from, "off", len) == 0) - *result = 0; - else - return 1; - return 0; -} - static int parse_options(char *options, struct super_block *sb) { struct nilfs_sb_info *sbi = NILFS_SB(sb); @@ -612,13 +581,8 @@ static int parse_options(char *options, struct super_block *sb) token = match_token(p, tokens, args); switch (token) { - case Opt_barrier: - if (match_bool(&args[0], &option)) - return 0; - if (option) - nilfs_set_opt(sbi, BARRIER); - else - nilfs_clear_opt(sbi, BARRIER); + case Opt_nobarrier: + nilfs_clear_opt(sbi, BARRIER); break; case Opt_order: if (strcmp(args[0].from, "relaxed") == 0) @@ -647,6 +611,9 @@ static int parse_options(char *options, struct super_block *sb) sbi->s_snapshot_cno = option; nilfs_set_opt(sbi, SNAPSHOT); break; + case Opt_norecovery: + nilfs_set_opt(sbi, NORECOVERY); + break; default: printk(KERN_ERR "NILFS: Unrecognized mount option \"%s\"\n", p); @@ -672,9 +639,7 @@ static int nilfs_setup_super(struct nilfs_sb_info *sbi) int mnt_count = le16_to_cpu(sbp->s_mnt_count); /* nilfs->sem must be locked by the caller. */ - if (!(nilfs->ns_mount_state & NILFS_VALID_FS)) { - printk(KERN_WARNING "NILFS warning: mounting unchecked fs\n"); - } else if (nilfs->ns_mount_state & NILFS_ERROR_FS) { + if (nilfs->ns_mount_state & NILFS_ERROR_FS) { printk(KERN_WARNING "NILFS warning: mounting fs with errors\n"); #if 0 @@ -782,11 +747,10 @@ nilfs_fill_super(struct super_block *sb, void *data, int silent, sb->s_root = NULL; sb->s_time_gran = 1; - if (!nilfs_loaded(nilfs)) { - err = load_nilfs(nilfs, sbi); - if (err) - goto failed_sbi; - } + err = load_nilfs(nilfs, sbi); + if (err) + goto failed_sbi; + cno = nilfs_last_cno(nilfs); if (sb->s_flags & MS_RDONLY) { @@ -854,12 +818,6 @@ nilfs_fill_super(struct super_block *sb, void *data, int silent, up_write(&nilfs->ns_sem); } - err = nilfs_mark_recovery_complete(sbi); - if (unlikely(err)) { - printk(KERN_ERR "NILFS: recovery failed.\n"); - goto failed_root; - } - down_write(&nilfs->ns_super_sem); if (!nilfs_test_opt(sbi, SNAPSHOT)) nilfs->ns_current = sbi; @@ -867,10 +825,6 @@ nilfs_fill_super(struct super_block *sb, void *data, int silent, return 0; - failed_root: - dput(sb->s_root); - sb->s_root = NULL; - failed_segctor: nilfs_detach_segment_constructor(sbi); @@ -915,6 +869,14 @@ static int nilfs_remount(struct super_block *sb, int *flags, char *data) goto restore_opts; } + if (!nilfs_valid_fs(nilfs)) { + printk(KERN_WARNING "NILFS (device %s): couldn't " + "remount because the filesystem is in an " + "incomplete recovery state.\n", sb->s_id); + err = -EINVAL; + goto restore_opts; + } + if ((*flags & MS_RDONLY) == (sb->s_flags & MS_RDONLY)) goto out; if (*flags & MS_RDONLY) { diff --git a/fs/nilfs2/the_nilfs.c b/fs/nilfs2/the_nilfs.c index ad391a8c3e7..6241e1722ef 100644 --- a/fs/nilfs2/the_nilfs.c +++ b/fs/nilfs2/the_nilfs.c @@ -146,13 +146,9 @@ void put_nilfs(struct the_nilfs *nilfs) might_sleep(); if (nilfs_loaded(nilfs)) { - nilfs_mdt_clear(nilfs->ns_sufile); nilfs_mdt_destroy(nilfs->ns_sufile); - nilfs_mdt_clear(nilfs->ns_cpfile); nilfs_mdt_destroy(nilfs->ns_cpfile); - nilfs_mdt_clear(nilfs->ns_dat); nilfs_mdt_destroy(nilfs->ns_dat); - /* XXX: how and when to clear nilfs->ns_gc_dat? */ nilfs_mdt_destroy(nilfs->ns_gc_dat); } if (nilfs_init(nilfs)) { @@ -166,7 +162,6 @@ void put_nilfs(struct the_nilfs *nilfs) static int nilfs_load_super_root(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi, sector_t sr_block) { - static struct lock_class_key dat_lock_key; struct buffer_head *bh_sr; struct nilfs_super_root *raw_sr; struct nilfs_super_block **sbp = nilfs->ns_sbp; @@ -187,51 +182,36 @@ static int nilfs_load_super_root(struct the_nilfs *nilfs, inode_size = nilfs->ns_inode_size; err = -ENOMEM; - nilfs->ns_dat = nilfs_mdt_new(nilfs, NULL, NILFS_DAT_INO); + nilfs->ns_dat = nilfs_dat_new(nilfs, dat_entry_size); if (unlikely(!nilfs->ns_dat)) goto failed; - nilfs->ns_gc_dat = nilfs_mdt_new(nilfs, NULL, NILFS_DAT_INO); + nilfs->ns_gc_dat = nilfs_dat_new(nilfs, dat_entry_size); if (unlikely(!nilfs->ns_gc_dat)) goto failed_dat; - nilfs->ns_cpfile = nilfs_mdt_new(nilfs, NULL, NILFS_CPFILE_INO); + nilfs->ns_cpfile = nilfs_cpfile_new(nilfs, checkpoint_size); if (unlikely(!nilfs->ns_cpfile)) goto failed_gc_dat; - nilfs->ns_sufile = nilfs_mdt_new(nilfs, NULL, NILFS_SUFILE_INO); + nilfs->ns_sufile = nilfs_sufile_new(nilfs, segment_usage_size); if (unlikely(!nilfs->ns_sufile)) goto failed_cpfile; - err = nilfs_palloc_init_blockgroup(nilfs->ns_dat, dat_entry_size); - if (unlikely(err)) - goto failed_sufile; - - err = nilfs_palloc_init_blockgroup(nilfs->ns_gc_dat, dat_entry_size); - if (unlikely(err)) - goto failed_sufile; - - lockdep_set_class(&NILFS_MDT(nilfs->ns_dat)->mi_sem, &dat_lock_key); - lockdep_set_class(&NILFS_MDT(nilfs->ns_gc_dat)->mi_sem, &dat_lock_key); - nilfs_mdt_set_shadow(nilfs->ns_dat, nilfs->ns_gc_dat); - nilfs_mdt_set_entry_size(nilfs->ns_cpfile, checkpoint_size, - sizeof(struct nilfs_cpfile_header)); - nilfs_mdt_set_entry_size(nilfs->ns_sufile, segment_usage_size, - sizeof(struct nilfs_sufile_header)); - err = nilfs_mdt_read_inode_direct( - nilfs->ns_dat, bh_sr, NILFS_SR_DAT_OFFSET(inode_size)); + err = nilfs_dat_read(nilfs->ns_dat, (void *)bh_sr->b_data + + NILFS_SR_DAT_OFFSET(inode_size)); if (unlikely(err)) goto failed_sufile; - err = nilfs_mdt_read_inode_direct( - nilfs->ns_cpfile, bh_sr, NILFS_SR_CPFILE_OFFSET(inode_size)); + err = nilfs_cpfile_read(nilfs->ns_cpfile, (void *)bh_sr->b_data + + NILFS_SR_CPFILE_OFFSET(inode_size)); if (unlikely(err)) goto failed_sufile; - err = nilfs_mdt_read_inode_direct( - nilfs->ns_sufile, bh_sr, NILFS_SR_SUFILE_OFFSET(inode_size)); + err = nilfs_sufile_read(nilfs->ns_sufile, (void *)bh_sr->b_data + + NILFS_SR_SUFILE_OFFSET(inode_size)); if (unlikely(err)) goto failed_sufile; @@ -281,29 +261,30 @@ int load_nilfs(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi) struct nilfs_recovery_info ri; unsigned int s_flags = sbi->s_super->s_flags; int really_read_only = bdev_read_only(nilfs->ns_bdev); - unsigned valid_fs; - int err = 0; - - nilfs_init_recovery_info(&ri); + int valid_fs = nilfs_valid_fs(nilfs); + int err; - down_write(&nilfs->ns_sem); - valid_fs = (nilfs->ns_mount_state & NILFS_VALID_FS); - up_write(&nilfs->ns_sem); + if (nilfs_loaded(nilfs)) { + if (valid_fs || + ((s_flags & MS_RDONLY) && nilfs_test_opt(sbi, NORECOVERY))) + return 0; + printk(KERN_ERR "NILFS: the filesystem is in an incomplete " + "recovery state.\n"); + return -EINVAL; + } - if (!valid_fs && (s_flags & MS_RDONLY)) { - printk(KERN_INFO "NILFS: INFO: recovery " - "required for readonly filesystem.\n"); - if (really_read_only) { - printk(KERN_ERR "NILFS: write access " - "unavailable, cannot proceed.\n"); - err = -EROFS; - goto failed; + if (!valid_fs) { + printk(KERN_WARNING "NILFS warning: mounting unchecked fs\n"); + if (s_flags & MS_RDONLY) { + printk(KERN_INFO "NILFS: INFO: recovery " + "required for readonly filesystem.\n"); + printk(KERN_INFO "NILFS: write access will " + "be enabled during recovery.\n"); } - printk(KERN_INFO "NILFS: write access will " - "be enabled during recovery.\n"); - sbi->s_super->s_flags &= ~MS_RDONLY; } + nilfs_init_recovery_info(&ri); + err = nilfs_search_super_root(nilfs, sbi, &ri); if (unlikely(err)) { printk(KERN_ERR "NILFS: error searching super root.\n"); @@ -316,19 +297,56 @@ int load_nilfs(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi) goto failed; } - if (!valid_fs) { - err = nilfs_recover_logical_segments(nilfs, sbi, &ri); - if (unlikely(err)) { - nilfs_mdt_destroy(nilfs->ns_cpfile); - nilfs_mdt_destroy(nilfs->ns_sufile); - nilfs_mdt_destroy(nilfs->ns_dat); - goto failed; + if (valid_fs) + goto skip_recovery; + + if (s_flags & MS_RDONLY) { + if (nilfs_test_opt(sbi, NORECOVERY)) { + printk(KERN_INFO "NILFS: norecovery option specified. " + "skipping roll-forward recovery\n"); + goto skip_recovery; } - if (ri.ri_need_recovery == NILFS_RECOVERY_SR_UPDATED) - sbi->s_super->s_dirt = 1; + if (really_read_only) { + printk(KERN_ERR "NILFS: write access " + "unavailable, cannot proceed.\n"); + err = -EROFS; + goto failed_unload; + } + sbi->s_super->s_flags &= ~MS_RDONLY; + } else if (nilfs_test_opt(sbi, NORECOVERY)) { + printk(KERN_ERR "NILFS: recovery cancelled because norecovery " + "option was specified for a read/write mount\n"); + err = -EINVAL; + goto failed_unload; } + err = nilfs_recover_logical_segments(nilfs, sbi, &ri); + if (err) + goto failed_unload; + + down_write(&nilfs->ns_sem); + nilfs->ns_mount_state |= NILFS_VALID_FS; + nilfs->ns_sbp[0]->s_state = cpu_to_le16(nilfs->ns_mount_state); + err = nilfs_commit_super(sbi, 1); + up_write(&nilfs->ns_sem); + + if (err) { + printk(KERN_ERR "NILFS: failed to update super block. " + "recovery unfinished.\n"); + goto failed_unload; + } + printk(KERN_INFO "NILFS: recovery complete.\n"); + + skip_recovery: set_nilfs_loaded(nilfs); + nilfs_clear_recovery_info(&ri); + sbi->s_super->s_flags = s_flags; + return 0; + + failed_unload: + nilfs_mdt_destroy(nilfs->ns_cpfile); + nilfs_mdt_destroy(nilfs->ns_sufile); + nilfs_mdt_destroy(nilfs->ns_dat); failed: nilfs_clear_recovery_info(&ri); @@ -632,30 +650,23 @@ int nilfs_count_free_blocks(struct the_nilfs *nilfs, sector_t *nblocks) { struct inode *dat = nilfs_dat_inode(nilfs); unsigned long ncleansegs; - int err; down_read(&NILFS_MDT(dat)->mi_sem); /* XXX */ - err = nilfs_sufile_get_ncleansegs(nilfs->ns_sufile, &ncleansegs); + ncleansegs = nilfs_sufile_get_ncleansegs(nilfs->ns_sufile); up_read(&NILFS_MDT(dat)->mi_sem); /* XXX */ - if (likely(!err)) - *nblocks = (sector_t)ncleansegs * nilfs->ns_blocks_per_segment; - return err; + *nblocks = (sector_t)ncleansegs * nilfs->ns_blocks_per_segment; + return 0; } int nilfs_near_disk_full(struct the_nilfs *nilfs) { - struct inode *sufile = nilfs->ns_sufile; unsigned long ncleansegs, nincsegs; - int ret; - ret = nilfs_sufile_get_ncleansegs(sufile, &ncleansegs); - if (likely(!ret)) { - nincsegs = atomic_read(&nilfs->ns_ndirtyblks) / - nilfs->ns_blocks_per_segment + 1; - if (ncleansegs <= nilfs->ns_nrsvsegs + nincsegs) - ret++; - } - return ret; + ncleansegs = nilfs_sufile_get_ncleansegs(nilfs->ns_sufile); + nincsegs = atomic_read(&nilfs->ns_ndirtyblks) / + nilfs->ns_blocks_per_segment + 1; + + return ncleansegs <= nilfs->ns_nrsvsegs + nincsegs; } /** diff --git a/fs/nilfs2/the_nilfs.h b/fs/nilfs2/the_nilfs.h index 20abd55881e..589786e3346 100644 --- a/fs/nilfs2/the_nilfs.h +++ b/fs/nilfs2/the_nilfs.h @@ -258,6 +258,16 @@ static inline void nilfs_put_sbinfo(struct nilfs_sb_info *sbi) kfree(sbi); } +static inline int nilfs_valid_fs(struct the_nilfs *nilfs) +{ + unsigned valid_fs; + + down_read(&nilfs->ns_sem); + valid_fs = (nilfs->ns_mount_state & NILFS_VALID_FS); + up_read(&nilfs->ns_sem); + return valid_fs; +} + static inline void nilfs_get_segment_range(struct the_nilfs *nilfs, __u64 segnum, sector_t *seg_start, sector_t *seg_end) diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c index de059f49058..3d30a1c974a 100644 --- a/fs/ocfs2/file.c +++ b/fs/ocfs2/file.c @@ -2006,7 +2006,7 @@ out_dio: /* buffered aio wouldn't have proper lock coverage today */ BUG_ON(ret == -EIOCBQUEUED && !(file->f_flags & O_DIRECT)); - if ((file->f_flags & O_SYNC && !direct_io) || IS_SYNC(inode)) { + if ((file->f_flags & O_DSYNC && !direct_io) || IS_SYNC(inode)) { ret = filemap_fdatawrite_range(file->f_mapping, pos, pos + count - 1); if (ret < 0) diff --git a/fs/ocfs2/quota.h b/fs/ocfs2/quota.h index e5df9d170b0..123bc520a2c 100644 --- a/fs/ocfs2/quota.h +++ b/fs/ocfs2/quota.h @@ -17,10 +17,6 @@ #include "ocfs2.h" -/* Common stuff */ -/* id number of quota format */ -#define QFMT_OCFS2 3 - /* * In-memory structures */ diff --git a/fs/ocfs2/quota_local.c b/fs/ocfs2/quota_local.c index 1a2c50a759f..21f9e71223c 100644 --- a/fs/ocfs2/quota_local.c +++ b/fs/ocfs2/quota_local.c @@ -1325,7 +1325,7 @@ out: return status; } -static struct quota_format_ops ocfs2_format_ops = { +static const struct quota_format_ops ocfs2_format_ops = { .check_quota_file = ocfs2_local_check_quota_file, .read_file_info = ocfs2_local_read_info, .write_file_info = ocfs2_global_write_info, diff --git a/fs/quota/Kconfig b/fs/quota/Kconfig index 353e78a9ebe..efc02ebb8c7 100644 --- a/fs/quota/Kconfig +++ b/fs/quota/Kconfig @@ -46,12 +46,14 @@ config QFMT_V1 format say Y here. config QFMT_V2 - tristate "Quota format v2 support" + tristate "Quota format vfsv0 and vfsv1 support" depends on QUOTA select QUOTA_TREE help - This quota format allows using quotas with 32-bit UIDs/GIDs. If you - need this functionality say Y here. + This config option enables kernel support for vfsv0 and vfsv1 quota + formats. Both these formats support 32-bit UIDs/GIDs and vfsv1 format + also supports 64-bit inode and block quota limits. If you need this + functionality say Y here. config QUOTACTL bool diff --git a/fs/quota/dquot.c b/fs/quota/dquot.c index eb5a755718f..cd6bb9a33c1 100644 --- a/fs/quota/dquot.c +++ b/fs/quota/dquot.c @@ -2164,7 +2164,9 @@ int vfs_quota_on_mount(struct super_block *sb, char *qf_name, struct dentry *dentry; int error; + mutex_lock(&sb->s_root->d_inode->i_mutex); dentry = lookup_one_len(qf_name, sb->s_root, strlen(qf_name)); + mutex_unlock(&sb->s_root->d_inode->i_mutex); if (IS_ERR(dentry)) return PTR_ERR(dentry); diff --git a/fs/quota/quota_v1.c b/fs/quota/quota_v1.c index 0edcf42b177..2ae757e9c00 100644 --- a/fs/quota/quota_v1.c +++ b/fs/quota/quota_v1.c @@ -204,7 +204,7 @@ out: return ret; } -static struct quota_format_ops v1_format_ops = { +static const struct quota_format_ops v1_format_ops = { .check_quota_file = v1_check_quota_file, .read_file_info = v1_read_file_info, .write_file_info = v1_write_file_info, diff --git a/fs/quota/quota_v2.c b/fs/quota/quota_v2.c index a5475fb1ae4..3dfc23e0213 100644 --- a/fs/quota/quota_v2.c +++ b/fs/quota/quota_v2.c @@ -23,14 +23,23 @@ MODULE_LICENSE("GPL"); #define __QUOTA_V2_PARANOIA -static void v2_mem2diskdqb(void *dp, struct dquot *dquot); -static void v2_disk2memdqb(struct dquot *dquot, void *dp); -static int v2_is_id(void *dp, struct dquot *dquot); - -static struct qtree_fmt_operations v2_qtree_ops = { - .mem2disk_dqblk = v2_mem2diskdqb, - .disk2mem_dqblk = v2_disk2memdqb, - .is_id = v2_is_id, +static void v2r0_mem2diskdqb(void *dp, struct dquot *dquot); +static void v2r0_disk2memdqb(struct dquot *dquot, void *dp); +static int v2r0_is_id(void *dp, struct dquot *dquot); +static void v2r1_mem2diskdqb(void *dp, struct dquot *dquot); +static void v2r1_disk2memdqb(struct dquot *dquot, void *dp); +static int v2r1_is_id(void *dp, struct dquot *dquot); + +static struct qtree_fmt_operations v2r0_qtree_ops = { + .mem2disk_dqblk = v2r0_mem2diskdqb, + .disk2mem_dqblk = v2r0_disk2memdqb, + .is_id = v2r0_is_id, +}; + +static struct qtree_fmt_operations v2r1_qtree_ops = { + .mem2disk_dqblk = v2r1_mem2diskdqb, + .disk2mem_dqblk = v2r1_disk2memdqb, + .is_id = v2r1_is_id, }; #define QUOTABLOCK_BITS 10 @@ -46,23 +55,33 @@ static inline qsize_t v2_qbtos(qsize_t blocks) return blocks << QUOTABLOCK_BITS; } +static int v2_read_header(struct super_block *sb, int type, + struct v2_disk_dqheader *dqhead) +{ + ssize_t size; + + size = sb->s_op->quota_read(sb, type, (char *)dqhead, + sizeof(struct v2_disk_dqheader), 0); + if (size != sizeof(struct v2_disk_dqheader)) { + printk(KERN_WARNING "quota_v2: Failed header read:" + " expected=%zd got=%zd\n", + sizeof(struct v2_disk_dqheader), size); + return 0; + } + return 1; +} + /* Check whether given file is really vfsv0 quotafile */ static int v2_check_quota_file(struct super_block *sb, int type) { struct v2_disk_dqheader dqhead; - ssize_t size; static const uint quota_magics[] = V2_INITQMAGICS; static const uint quota_versions[] = V2_INITQVERSIONS; - size = sb->s_op->quota_read(sb, type, (char *)&dqhead, - sizeof(struct v2_disk_dqheader), 0); - if (size != sizeof(struct v2_disk_dqheader)) { - printk("quota_v2: failed read expected=%zd got=%zd\n", - sizeof(struct v2_disk_dqheader), size); + if (!v2_read_header(sb, type, &dqhead)) return 0; - } if (le32_to_cpu(dqhead.dqh_magic) != quota_magics[type] || - le32_to_cpu(dqhead.dqh_version) != quota_versions[type]) + le32_to_cpu(dqhead.dqh_version) > quota_versions[type]) return 0; return 1; } @@ -71,14 +90,20 @@ static int v2_check_quota_file(struct super_block *sb, int type) static int v2_read_file_info(struct super_block *sb, int type) { struct v2_disk_dqinfo dinfo; + struct v2_disk_dqheader dqhead; struct mem_dqinfo *info = sb_dqinfo(sb, type); struct qtree_mem_dqinfo *qinfo; ssize_t size; + unsigned int version; + + if (!v2_read_header(sb, type, &dqhead)) + return 0; + version = le32_to_cpu(dqhead.dqh_version); size = sb->s_op->quota_read(sb, type, (char *)&dinfo, sizeof(struct v2_disk_dqinfo), V2_DQINFOOFF); if (size != sizeof(struct v2_disk_dqinfo)) { - printk(KERN_WARNING "Can't read info structure on device %s.\n", + printk(KERN_WARNING "quota_v2: Can't read info structure on device %s.\n", sb->s_id); return -1; } @@ -89,9 +114,15 @@ static int v2_read_file_info(struct super_block *sb, int type) return -1; } qinfo = info->dqi_priv; - /* limits are stored as unsigned 32-bit data */ - info->dqi_maxblimit = 0xffffffff; - info->dqi_maxilimit = 0xffffffff; + if (version == 0) { + /* limits are stored as unsigned 32-bit data */ + info->dqi_maxblimit = 0xffffffff; + info->dqi_maxilimit = 0xffffffff; + } else { + /* used space is stored as unsigned 64-bit value */ + info->dqi_maxblimit = 0xffffffffffffffff; /* 2^64-1 */ + info->dqi_maxilimit = 0xffffffffffffffff; + } info->dqi_bgrace = le32_to_cpu(dinfo.dqi_bgrace); info->dqi_igrace = le32_to_cpu(dinfo.dqi_igrace); info->dqi_flags = le32_to_cpu(dinfo.dqi_flags); @@ -103,8 +134,13 @@ static int v2_read_file_info(struct super_block *sb, int type) qinfo->dqi_blocksize_bits = V2_DQBLKSIZE_BITS; qinfo->dqi_usable_bs = 1 << V2_DQBLKSIZE_BITS; qinfo->dqi_qtree_depth = qtree_depth(qinfo); - qinfo->dqi_entry_size = sizeof(struct v2_disk_dqblk); - qinfo->dqi_ops = &v2_qtree_ops; + if (version == 0) { + qinfo->dqi_entry_size = sizeof(struct v2r0_disk_dqblk); + qinfo->dqi_ops = &v2r0_qtree_ops; + } else { + qinfo->dqi_entry_size = sizeof(struct v2r1_disk_dqblk); + qinfo->dqi_ops = &v2r1_qtree_ops; + } return 0; } @@ -135,9 +171,9 @@ static int v2_write_file_info(struct super_block *sb, int type) return 0; } -static void v2_disk2memdqb(struct dquot *dquot, void *dp) +static void v2r0_disk2memdqb(struct dquot *dquot, void *dp) { - struct v2_disk_dqblk *d = dp, empty; + struct v2r0_disk_dqblk *d = dp, empty; struct mem_dqblk *m = &dquot->dq_dqb; m->dqb_ihardlimit = le32_to_cpu(d->dqb_ihardlimit); @@ -149,15 +185,15 @@ static void v2_disk2memdqb(struct dquot *dquot, void *dp) m->dqb_curspace = le64_to_cpu(d->dqb_curspace); m->dqb_btime = le64_to_cpu(d->dqb_btime); /* We need to escape back all-zero structure */ - memset(&empty, 0, sizeof(struct v2_disk_dqblk)); + memset(&empty, 0, sizeof(struct v2r0_disk_dqblk)); empty.dqb_itime = cpu_to_le64(1); - if (!memcmp(&empty, dp, sizeof(struct v2_disk_dqblk))) + if (!memcmp(&empty, dp, sizeof(struct v2r0_disk_dqblk))) m->dqb_itime = 0; } -static void v2_mem2diskdqb(void *dp, struct dquot *dquot) +static void v2r0_mem2diskdqb(void *dp, struct dquot *dquot) { - struct v2_disk_dqblk *d = dp; + struct v2r0_disk_dqblk *d = dp; struct mem_dqblk *m = &dquot->dq_dqb; struct qtree_mem_dqinfo *info = sb_dqinfo(dquot->dq_sb, dquot->dq_type)->dqi_priv; @@ -175,9 +211,60 @@ static void v2_mem2diskdqb(void *dp, struct dquot *dquot) d->dqb_itime = cpu_to_le64(1); } -static int v2_is_id(void *dp, struct dquot *dquot) +static int v2r0_is_id(void *dp, struct dquot *dquot) +{ + struct v2r0_disk_dqblk *d = dp; + struct qtree_mem_dqinfo *info = + sb_dqinfo(dquot->dq_sb, dquot->dq_type)->dqi_priv; + + if (qtree_entry_unused(info, dp)) + return 0; + return le32_to_cpu(d->dqb_id) == dquot->dq_id; +} + +static void v2r1_disk2memdqb(struct dquot *dquot, void *dp) +{ + struct v2r1_disk_dqblk *d = dp, empty; + struct mem_dqblk *m = &dquot->dq_dqb; + + m->dqb_ihardlimit = le64_to_cpu(d->dqb_ihardlimit); + m->dqb_isoftlimit = le64_to_cpu(d->dqb_isoftlimit); + m->dqb_curinodes = le64_to_cpu(d->dqb_curinodes); + m->dqb_itime = le64_to_cpu(d->dqb_itime); + m->dqb_bhardlimit = v2_qbtos(le64_to_cpu(d->dqb_bhardlimit)); + m->dqb_bsoftlimit = v2_qbtos(le64_to_cpu(d->dqb_bsoftlimit)); + m->dqb_curspace = le64_to_cpu(d->dqb_curspace); + m->dqb_btime = le64_to_cpu(d->dqb_btime); + /* We need to escape back all-zero structure */ + memset(&empty, 0, sizeof(struct v2r1_disk_dqblk)); + empty.dqb_itime = cpu_to_le64(1); + if (!memcmp(&empty, dp, sizeof(struct v2r1_disk_dqblk))) + m->dqb_itime = 0; +} + +static void v2r1_mem2diskdqb(void *dp, struct dquot *dquot) +{ + struct v2r1_disk_dqblk *d = dp; + struct mem_dqblk *m = &dquot->dq_dqb; + struct qtree_mem_dqinfo *info = + sb_dqinfo(dquot->dq_sb, dquot->dq_type)->dqi_priv; + + d->dqb_ihardlimit = cpu_to_le64(m->dqb_ihardlimit); + d->dqb_isoftlimit = cpu_to_le64(m->dqb_isoftlimit); + d->dqb_curinodes = cpu_to_le64(m->dqb_curinodes); + d->dqb_itime = cpu_to_le64(m->dqb_itime); + d->dqb_bhardlimit = cpu_to_le64(v2_stoqb(m->dqb_bhardlimit)); + d->dqb_bsoftlimit = cpu_to_le64(v2_stoqb(m->dqb_bsoftlimit)); + d->dqb_curspace = cpu_to_le64(m->dqb_curspace); + d->dqb_btime = cpu_to_le64(m->dqb_btime); + d->dqb_id = cpu_to_le32(dquot->dq_id); + if (qtree_entry_unused(info, dp)) + d->dqb_itime = cpu_to_le64(1); +} + +static int v2r1_is_id(void *dp, struct dquot *dquot) { - struct v2_disk_dqblk *d = dp; + struct v2r1_disk_dqblk *d = dp; struct qtree_mem_dqinfo *info = sb_dqinfo(dquot->dq_sb, dquot->dq_type)->dqi_priv; @@ -207,7 +294,7 @@ static int v2_free_file_info(struct super_block *sb, int type) return 0; } -static struct quota_format_ops v2_format_ops = { +static const struct quota_format_ops v2_format_ops = { .check_quota_file = v2_check_quota_file, .read_file_info = v2_read_file_info, .write_file_info = v2_write_file_info, @@ -217,20 +304,32 @@ static struct quota_format_ops v2_format_ops = { .release_dqblk = v2_release_dquot, }; -static struct quota_format_type v2_quota_format = { +static struct quota_format_type v2r0_quota_format = { .qf_fmt_id = QFMT_VFS_V0, .qf_ops = &v2_format_ops, .qf_owner = THIS_MODULE }; +static struct quota_format_type v2r1_quota_format = { + .qf_fmt_id = QFMT_VFS_V1, + .qf_ops = &v2_format_ops, + .qf_owner = THIS_MODULE +}; + static int __init init_v2_quota_format(void) { - return register_quota_format(&v2_quota_format); + int ret; + + ret = register_quota_format(&v2r0_quota_format); + if (ret) + return ret; + return register_quota_format(&v2r1_quota_format); } static void __exit exit_v2_quota_format(void) { - unregister_quota_format(&v2_quota_format); + unregister_quota_format(&v2r0_quota_format); + unregister_quota_format(&v2r1_quota_format); } module_init(init_v2_quota_format); diff --git a/fs/quota/quotaio_v2.h b/fs/quota/quotaio_v2.h index 530fe580685..f1966b42c2f 100644 --- a/fs/quota/quotaio_v2.h +++ b/fs/quota/quotaio_v2.h @@ -17,8 +17,8 @@ } #define V2_INITQVERSIONS {\ - 0, /* USRQUOTA */\ - 0 /* GRPQUOTA */\ + 1, /* USRQUOTA */\ + 1 /* GRPQUOTA */\ } /* First generic header */ @@ -32,7 +32,7 @@ struct v2_disk_dqheader { * (as it appears on disk) - the file is a radix tree whose leaves point * to blocks of these structures. */ -struct v2_disk_dqblk { +struct v2r0_disk_dqblk { __le32 dqb_id; /* id this quota applies to */ __le32 dqb_ihardlimit; /* absolute limit on allocated inodes */ __le32 dqb_isoftlimit; /* preferred inode limit */ @@ -44,6 +44,19 @@ struct v2_disk_dqblk { __le64 dqb_itime; /* time limit for excessive inode use */ }; +struct v2r1_disk_dqblk { + __le32 dqb_id; /* id this quota applies to */ + __le32 dqb_pad; + __le64 dqb_ihardlimit; /* absolute limit on allocated inodes */ + __le64 dqb_isoftlimit; /* preferred inode limit */ + __le64 dqb_curinodes; /* current # allocated inodes */ + __le64 dqb_bhardlimit; /* absolute limit on disk space (in QUOTABLOCK_SIZE) */ + __le64 dqb_bsoftlimit; /* preferred limit on disk space (in QUOTABLOCK_SIZE) */ + __le64 dqb_curspace; /* current space occupied (in bytes) */ + __le64 dqb_btime; /* time limit for excessive disk use */ + __le64 dqb_itime; /* time limit for excessive inode use */ +}; + /* Header with type and version specific information */ struct v2_disk_dqinfo { __le32 dqi_bgrace; /* Time before block soft limit becomes hard limit */ diff --git a/fs/sync.c b/fs/sync.c index d104591b066..36752a68348 100644 --- a/fs/sync.c +++ b/fs/sync.c @@ -295,10 +295,11 @@ SYSCALL_DEFINE1(fdatasync, unsigned int, fd) */ int generic_write_sync(struct file *file, loff_t pos, loff_t count) { - if (!(file->f_flags & O_SYNC) && !IS_SYNC(file->f_mapping->host)) + if (!(file->f_flags & O_DSYNC) && !IS_SYNC(file->f_mapping->host)) return 0; return vfs_fsync_range(file, file->f_path.dentry, pos, - pos + count - 1, 1); + pos + count - 1, + (file->f_flags & __O_SYNC) ? 0 : 1); } EXPORT_SYMBOL(generic_write_sync); @@ -452,9 +453,7 @@ int do_sync_mapping_range(struct address_space *mapping, loff_t offset, ret = 0; if (flags & SYNC_FILE_RANGE_WAIT_BEFORE) { - ret = wait_on_page_writeback_range(mapping, - offset >> PAGE_CACHE_SHIFT, - endbyte >> PAGE_CACHE_SHIFT); + ret = filemap_fdatawait_range(mapping, offset, endbyte); if (ret < 0) goto out; } @@ -467,9 +466,7 @@ int do_sync_mapping_range(struct address_space *mapping, loff_t offset, } if (flags & SYNC_FILE_RANGE_WAIT_AFTER) { - ret = wait_on_page_writeback_range(mapping, - offset >> PAGE_CACHE_SHIFT, - endbyte >> PAGE_CACHE_SHIFT); + ret = filemap_fdatawait_range(mapping, offset, endbyte); } out: return ret; diff --git a/fs/sysfs/dir.c b/fs/sysfs/dir.c index e0201837d24..f05f2303a8b 100644 --- a/fs/sysfs/dir.c +++ b/fs/sysfs/dir.c @@ -25,7 +25,6 @@ #include "sysfs.h" DEFINE_MUTEX(sysfs_mutex); -DEFINE_MUTEX(sysfs_rename_mutex); DEFINE_SPINLOCK(sysfs_assoc_lock); static DEFINE_SPINLOCK(sysfs_ino_lock); @@ -85,46 +84,6 @@ static void sysfs_unlink_sibling(struct sysfs_dirent *sd) } /** - * sysfs_get_dentry - get dentry for the given sysfs_dirent - * @sd: sysfs_dirent of interest - * - * Get dentry for @sd. Dentry is looked up if currently not - * present. This function descends from the root looking up - * dentry for each step. - * - * LOCKING: - * mutex_lock(sysfs_rename_mutex) - * - * RETURNS: - * Pointer to found dentry on success, ERR_PTR() value on error. - */ -struct dentry *sysfs_get_dentry(struct sysfs_dirent *sd) -{ - struct dentry *dentry = dget(sysfs_sb->s_root); - - while (dentry->d_fsdata != sd) { - struct sysfs_dirent *cur; - struct dentry *parent; - - /* find the first ancestor which hasn't been looked up */ - cur = sd; - while (cur->s_parent != dentry->d_fsdata) - cur = cur->s_parent; - - /* look it up */ - parent = dentry; - mutex_lock(&parent->d_inode->i_mutex); - dentry = lookup_one_noperm(cur->s_name, parent); - mutex_unlock(&parent->d_inode->i_mutex); - dput(parent); - - if (IS_ERR(dentry)) - break; - } - return dentry; -} - -/** * sysfs_get_active - get an active reference to sysfs_dirent * @sd: sysfs_dirent to get an active reference to * @@ -298,7 +257,61 @@ void release_sysfs_dirent(struct sysfs_dirent * sd) goto repeat; } -static void sysfs_d_iput(struct dentry * dentry, struct inode * inode) +static int sysfs_dentry_delete(struct dentry *dentry) +{ + struct sysfs_dirent *sd = dentry->d_fsdata; + return !!(sd->s_flags & SYSFS_FLAG_REMOVED); +} + +static int sysfs_dentry_revalidate(struct dentry *dentry, struct nameidata *nd) +{ + struct sysfs_dirent *sd = dentry->d_fsdata; + int is_dir; + + mutex_lock(&sysfs_mutex); + + /* The sysfs dirent has been deleted */ + if (sd->s_flags & SYSFS_FLAG_REMOVED) + goto out_bad; + + /* The sysfs dirent has been moved? */ + if (dentry->d_parent->d_fsdata != sd->s_parent) + goto out_bad; + + /* The sysfs dirent has been renamed */ + if (strcmp(dentry->d_name.name, sd->s_name) != 0) + goto out_bad; + + mutex_unlock(&sysfs_mutex); +out_valid: + return 1; +out_bad: + /* Remove the dentry from the dcache hashes. + * If this is a deleted dentry we use d_drop instead of d_delete + * so sysfs doesn't need to cope with negative dentries. + * + * If this is a dentry that has simply been renamed we + * use d_drop to remove it from the dcache lookup on its + * old parent. If this dentry persists later when a lookup + * is performed at its new name the dentry will be readded + * to the dcache hashes. + */ + is_dir = (sysfs_type(sd) == SYSFS_DIR); + mutex_unlock(&sysfs_mutex); + if (is_dir) { + /* If we have submounts we must allow the vfs caches + * to lie about the state of the filesystem to prevent + * leaks and other nasty things. + */ + if (have_submounts(dentry)) + goto out_valid; + shrink_dcache_parent(dentry); + } + d_drop(dentry); + return 0; +} + +static void sysfs_dentry_iput(struct dentry *dentry, struct inode *inode) { struct sysfs_dirent * sd = dentry->d_fsdata; @@ -307,7 +320,9 @@ static void sysfs_d_iput(struct dentry * dentry, struct inode * inode) } static const struct dentry_operations sysfs_dentry_ops = { - .d_iput = sysfs_d_iput, + .d_revalidate = sysfs_dentry_revalidate, + .d_delete = sysfs_dentry_delete, + .d_iput = sysfs_dentry_iput, }; struct sysfs_dirent *sysfs_new_dirent(const char *name, umode_t mode, int type) @@ -344,12 +359,6 @@ struct sysfs_dirent *sysfs_new_dirent(const char *name, umode_t mode, int type) return NULL; } -static int sysfs_ilookup_test(struct inode *inode, void *arg) -{ - struct sysfs_dirent *sd = arg; - return inode->i_ino == sd->s_ino; -} - /** * sysfs_addrm_start - prepare for sysfs_dirent add/remove * @acxt: pointer to sysfs_addrm_cxt to be used @@ -357,47 +366,20 @@ static int sysfs_ilookup_test(struct inode *inode, void *arg) * * This function is called when the caller is about to add or * remove sysfs_dirent under @parent_sd. This function acquires - * sysfs_mutex, grabs inode for @parent_sd if available and lock - * i_mutex of it. @acxt is used to keep and pass context to + * sysfs_mutex. @acxt is used to keep and pass context to * other addrm functions. * * LOCKING: * Kernel thread context (may sleep). sysfs_mutex is locked on - * return. i_mutex of parent inode is locked on return if - * available. + * return. */ void sysfs_addrm_start(struct sysfs_addrm_cxt *acxt, struct sysfs_dirent *parent_sd) { - struct inode *inode; - memset(acxt, 0, sizeof(*acxt)); acxt->parent_sd = parent_sd; - /* Lookup parent inode. inode initialization is protected by - * sysfs_mutex, so inode existence can be determined by - * looking up inode while holding sysfs_mutex. - */ mutex_lock(&sysfs_mutex); - - inode = ilookup5(sysfs_sb, parent_sd->s_ino, sysfs_ilookup_test, - parent_sd); - if (inode) { - WARN_ON(inode->i_state & I_NEW); - - /* parent inode available */ - acxt->parent_inode = inode; - - /* sysfs_mutex is below i_mutex in lock hierarchy. - * First, trylock i_mutex. If fails, unlock - * sysfs_mutex and lock them in order. - */ - if (!mutex_trylock(&inode->i_mutex)) { - mutex_unlock(&sysfs_mutex); - mutex_lock(&inode->i_mutex); - mutex_lock(&sysfs_mutex); - } - } } /** @@ -422,18 +404,22 @@ void sysfs_addrm_start(struct sysfs_addrm_cxt *acxt, */ int __sysfs_add_one(struct sysfs_addrm_cxt *acxt, struct sysfs_dirent *sd) { + struct sysfs_inode_attrs *ps_iattr; + if (sysfs_find_dirent(acxt->parent_sd, sd->s_name)) return -EEXIST; sd->s_parent = sysfs_get(acxt->parent_sd); - if (sysfs_type(sd) == SYSFS_DIR && acxt->parent_inode) - inc_nlink(acxt->parent_inode); - - acxt->cnt++; - sysfs_link_sibling(sd); + /* Update timestamps on the parent */ + ps_iattr = acxt->parent_sd->s_iattr; + if (ps_iattr) { + struct iattr *ps_iattrs = &ps_iattr->ia_iattr; + ps_iattrs->ia_ctime = ps_iattrs->ia_mtime = CURRENT_TIME; + } + return 0; } @@ -512,70 +498,22 @@ int sysfs_add_one(struct sysfs_addrm_cxt *acxt, struct sysfs_dirent *sd) */ void sysfs_remove_one(struct sysfs_addrm_cxt *acxt, struct sysfs_dirent *sd) { + struct sysfs_inode_attrs *ps_iattr; + BUG_ON(sd->s_flags & SYSFS_FLAG_REMOVED); sysfs_unlink_sibling(sd); + /* Update timestamps on the parent */ + ps_iattr = acxt->parent_sd->s_iattr; + if (ps_iattr) { + struct iattr *ps_iattrs = &ps_iattr->ia_iattr; + ps_iattrs->ia_ctime = ps_iattrs->ia_mtime = CURRENT_TIME; + } + sd->s_flags |= SYSFS_FLAG_REMOVED; sd->s_sibling = acxt->removed; acxt->removed = sd; - - if (sysfs_type(sd) == SYSFS_DIR && acxt->parent_inode) - drop_nlink(acxt->parent_inode); - - acxt->cnt++; -} - -/** - * sysfs_drop_dentry - drop dentry for the specified sysfs_dirent - * @sd: target sysfs_dirent - * - * Drop dentry for @sd. @sd must have been unlinked from its - * parent on entry to this function such that it can't be looked - * up anymore. - */ -static void sysfs_drop_dentry(struct sysfs_dirent *sd) -{ - struct inode *inode; - struct dentry *dentry; - - inode = ilookup(sysfs_sb, sd->s_ino); - if (!inode) - return; - - /* Drop any existing dentries associated with sd. - * - * For the dentry to be properly freed we need to grab a - * reference to the dentry under the dcache lock, unhash it, - * and then put it. The playing with the dentry count allows - * dput to immediately free the dentry if it is not in use. - */ -repeat: - spin_lock(&dcache_lock); - list_for_each_entry(dentry, &inode->i_dentry, d_alias) { - if (d_unhashed(dentry)) - continue; - dget_locked(dentry); - spin_lock(&dentry->d_lock); - __d_drop(dentry); - spin_unlock(&dentry->d_lock); - spin_unlock(&dcache_lock); - dput(dentry); - goto repeat; - } - spin_unlock(&dcache_lock); - - /* adjust nlink and update timestamp */ - mutex_lock(&inode->i_mutex); - - inode->i_ctime = CURRENT_TIME; - drop_nlink(inode); - if (sysfs_type(sd) == SYSFS_DIR) - drop_nlink(inode); - - mutex_unlock(&inode->i_mutex); - - iput(inode); } /** @@ -584,25 +522,15 @@ repeat: * * Finish up sysfs_dirent add/remove. Resources acquired by * sysfs_addrm_start() are released and removed sysfs_dirents are - * cleaned up. Timestamps on the parent inode are updated. + * cleaned up. * * LOCKING: - * All mutexes acquired by sysfs_addrm_start() are released. + * sysfs_mutex is released. */ void sysfs_addrm_finish(struct sysfs_addrm_cxt *acxt) { /* release resources acquired by sysfs_addrm_start() */ mutex_unlock(&sysfs_mutex); - if (acxt->parent_inode) { - struct inode *inode = acxt->parent_inode; - - /* if added/removed, update timestamps on the parent */ - if (acxt->cnt) - inode->i_ctime = inode->i_mtime = CURRENT_TIME; - - mutex_unlock(&inode->i_mutex); - iput(inode); - } /* kill removed sysfs_dirents */ while (acxt->removed) { @@ -611,7 +539,6 @@ void sysfs_addrm_finish(struct sysfs_addrm_cxt *acxt) acxt->removed = sd->s_sibling; sd->s_sibling = NULL; - sysfs_drop_dentry(sd); sysfs_deactivate(sd); unmap_bin_file(sd); sysfs_put(sd); @@ -751,10 +678,15 @@ static struct dentry * sysfs_lookup(struct inode *dir, struct dentry *dentry, } /* instantiate and hash dentry */ - dentry->d_op = &sysfs_dentry_ops; - dentry->d_fsdata = sysfs_get(sd); - d_instantiate(dentry, inode); - d_rehash(dentry); + ret = d_find_alias(inode); + if (!ret) { + dentry->d_op = &sysfs_dentry_ops; + dentry->d_fsdata = sysfs_get(sd); + d_add(dentry, inode); + } else { + d_move(ret, dentry); + iput(inode); + } out_unlock: mutex_unlock(&sysfs_mutex); @@ -763,7 +695,9 @@ static struct dentry * sysfs_lookup(struct inode *dir, struct dentry *dentry, const struct inode_operations sysfs_dir_inode_operations = { .lookup = sysfs_lookup, + .permission = sysfs_permission, .setattr = sysfs_setattr, + .getattr = sysfs_getattr, .setxattr = sysfs_setxattr, }; @@ -826,141 +760,65 @@ void sysfs_remove_dir(struct kobject * kobj) __sysfs_remove_dir(sd); } -int sysfs_rename_dir(struct kobject * kobj, const char *new_name) +int sysfs_rename(struct sysfs_dirent *sd, + struct sysfs_dirent *new_parent_sd, const char *new_name) { - struct sysfs_dirent *sd = kobj->sd; - struct dentry *parent = NULL; - struct dentry *old_dentry = NULL, *new_dentry = NULL; const char *dup_name = NULL; int error; - mutex_lock(&sysfs_rename_mutex); + mutex_lock(&sysfs_mutex); error = 0; - if (strcmp(sd->s_name, new_name) == 0) + if ((sd->s_parent == new_parent_sd) && + (strcmp(sd->s_name, new_name) == 0)) goto out; /* nothing to rename */ - /* get the original dentry */ - old_dentry = sysfs_get_dentry(sd); - if (IS_ERR(old_dentry)) { - error = PTR_ERR(old_dentry); - old_dentry = NULL; - goto out; - } - - parent = old_dentry->d_parent; - - /* lock parent and get dentry for new name */ - mutex_lock(&parent->d_inode->i_mutex); - mutex_lock(&sysfs_mutex); - error = -EEXIST; - if (sysfs_find_dirent(sd->s_parent, new_name)) - goto out_unlock; - - error = -ENOMEM; - new_dentry = d_alloc_name(parent, new_name); - if (!new_dentry) - goto out_unlock; + if (sysfs_find_dirent(new_parent_sd, new_name)) + goto out; /* rename sysfs_dirent */ - error = -ENOMEM; - new_name = dup_name = kstrdup(new_name, GFP_KERNEL); - if (!new_name) - goto out_unlock; - - dup_name = sd->s_name; - sd->s_name = new_name; + if (strcmp(sd->s_name, new_name) != 0) { + error = -ENOMEM; + new_name = dup_name = kstrdup(new_name, GFP_KERNEL); + if (!new_name) + goto out; + + dup_name = sd->s_name; + sd->s_name = new_name; + } - /* rename */ - d_add(new_dentry, NULL); - d_move(old_dentry, new_dentry); + /* Remove from old parent's list and insert into new parent's list. */ + if (sd->s_parent != new_parent_sd) { + sysfs_unlink_sibling(sd); + sysfs_get(new_parent_sd); + sysfs_put(sd->s_parent); + sd->s_parent = new_parent_sd; + sysfs_link_sibling(sd); + } error = 0; - out_unlock: + out: mutex_unlock(&sysfs_mutex); - mutex_unlock(&parent->d_inode->i_mutex); kfree(dup_name); - dput(old_dentry); - dput(new_dentry); - out: - mutex_unlock(&sysfs_rename_mutex); return error; } +int sysfs_rename_dir(struct kobject *kobj, const char *new_name) +{ + return sysfs_rename(kobj->sd, kobj->sd->s_parent, new_name); +} + int sysfs_move_dir(struct kobject *kobj, struct kobject *new_parent_kobj) { struct sysfs_dirent *sd = kobj->sd; struct sysfs_dirent *new_parent_sd; - struct dentry *old_parent, *new_parent = NULL; - struct dentry *old_dentry = NULL, *new_dentry = NULL; - int error; - mutex_lock(&sysfs_rename_mutex); BUG_ON(!sd->s_parent); - new_parent_sd = (new_parent_kobj && new_parent_kobj->sd) ? + new_parent_sd = new_parent_kobj && new_parent_kobj->sd ? new_parent_kobj->sd : &sysfs_root; - error = 0; - if (sd->s_parent == new_parent_sd) - goto out; /* nothing to move */ - - /* get dentries */ - old_dentry = sysfs_get_dentry(sd); - if (IS_ERR(old_dentry)) { - error = PTR_ERR(old_dentry); - old_dentry = NULL; - goto out; - } - old_parent = old_dentry->d_parent; - - new_parent = sysfs_get_dentry(new_parent_sd); - if (IS_ERR(new_parent)) { - error = PTR_ERR(new_parent); - new_parent = NULL; - goto out; - } - -again: - mutex_lock(&old_parent->d_inode->i_mutex); - if (!mutex_trylock(&new_parent->d_inode->i_mutex)) { - mutex_unlock(&old_parent->d_inode->i_mutex); - goto again; - } - mutex_lock(&sysfs_mutex); - - error = -EEXIST; - if (sysfs_find_dirent(new_parent_sd, sd->s_name)) - goto out_unlock; - - error = -ENOMEM; - new_dentry = d_alloc_name(new_parent, sd->s_name); - if (!new_dentry) - goto out_unlock; - - error = 0; - d_add(new_dentry, NULL); - d_move(old_dentry, new_dentry); - - /* Remove from old parent's list and insert into new parent's list. */ - sysfs_unlink_sibling(sd); - sysfs_get(new_parent_sd); - drop_nlink(old_parent->d_inode); - sysfs_put(sd->s_parent); - sd->s_parent = new_parent_sd; - inc_nlink(new_parent->d_inode); - sysfs_link_sibling(sd); - - out_unlock: - mutex_unlock(&sysfs_mutex); - mutex_unlock(&new_parent->d_inode->i_mutex); - mutex_unlock(&old_parent->d_inode->i_mutex); - out: - dput(new_parent); - dput(old_dentry); - dput(new_dentry); - mutex_unlock(&sysfs_rename_mutex); - return error; + return sysfs_rename(sd, new_parent_sd, sd->s_name); } /* Relationship between s_mode and the DT_xxx types */ diff --git a/fs/sysfs/file.c b/fs/sysfs/file.c index f5ea4680f15..dc30d9e3168 100644 --- a/fs/sysfs/file.c +++ b/fs/sysfs/file.c @@ -579,46 +579,23 @@ EXPORT_SYMBOL_GPL(sysfs_add_file_to_group); */ int sysfs_chmod_file(struct kobject *kobj, struct attribute *attr, mode_t mode) { - struct sysfs_dirent *victim_sd = NULL; - struct dentry *victim = NULL; - struct inode * inode; + struct sysfs_dirent *sd; struct iattr newattrs; int rc; - rc = -ENOENT; - victim_sd = sysfs_get_dirent(kobj->sd, attr->name); - if (!victim_sd) - goto out; + mutex_lock(&sysfs_mutex); - mutex_lock(&sysfs_rename_mutex); - victim = sysfs_get_dentry(victim_sd); - mutex_unlock(&sysfs_rename_mutex); - if (IS_ERR(victim)) { - rc = PTR_ERR(victim); - victim = NULL; + rc = -ENOENT; + sd = sysfs_find_dirent(kobj->sd, attr->name); + if (!sd) goto out; - } - - inode = victim->d_inode; - - mutex_lock(&inode->i_mutex); - newattrs.ia_mode = (mode & S_IALLUGO) | (inode->i_mode & ~S_IALLUGO); - newattrs.ia_valid = ATTR_MODE | ATTR_CTIME; - newattrs.ia_ctime = current_fs_time(inode->i_sb); - rc = sysfs_setattr(victim, &newattrs); + newattrs.ia_mode = (mode & S_IALLUGO) | (sd->s_mode & ~S_IALLUGO); + newattrs.ia_valid = ATTR_MODE; + rc = sysfs_sd_setattr(sd, &newattrs); - if (rc == 0) { - fsnotify_change(victim, newattrs.ia_valid); - mutex_lock(&sysfs_mutex); - victim_sd->s_mode = newattrs.ia_mode; - mutex_unlock(&sysfs_mutex); - } - - mutex_unlock(&inode->i_mutex); out: - dput(victim); - sysfs_put(victim_sd); + mutex_unlock(&sysfs_mutex); return rc; } EXPORT_SYMBOL_GPL(sysfs_chmod_file); diff --git a/fs/sysfs/inode.c b/fs/sysfs/inode.c index e28cecf179f..220b758523a 100644 --- a/fs/sysfs/inode.c +++ b/fs/sysfs/inode.c @@ -37,7 +37,9 @@ static struct backing_dev_info sysfs_backing_dev_info = { }; static const struct inode_operations sysfs_inode_operations ={ + .permission = sysfs_permission, .setattr = sysfs_setattr, + .getattr = sysfs_getattr, .setxattr = sysfs_setxattr, }; @@ -46,7 +48,7 @@ int __init sysfs_inode_init(void) return bdi_init(&sysfs_backing_dev_info); } -struct sysfs_inode_attrs *sysfs_init_inode_attrs(struct sysfs_dirent *sd) +static struct sysfs_inode_attrs *sysfs_init_inode_attrs(struct sysfs_dirent *sd) { struct sysfs_inode_attrs *attrs; struct iattr *iattrs; @@ -64,30 +66,15 @@ struct sysfs_inode_attrs *sysfs_init_inode_attrs(struct sysfs_dirent *sd) return attrs; } -int sysfs_setattr(struct dentry * dentry, struct iattr * iattr) + +int sysfs_sd_setattr(struct sysfs_dirent *sd, struct iattr * iattr) { - struct inode * inode = dentry->d_inode; - struct sysfs_dirent * sd = dentry->d_fsdata; struct sysfs_inode_attrs *sd_attrs; struct iattr *iattrs; unsigned int ia_valid = iattr->ia_valid; - int error; - - if (!sd) - return -EINVAL; sd_attrs = sd->s_iattr; - error = inode_change_ok(inode, iattr); - if (error) - return error; - - iattr->ia_valid &= ~ATTR_SIZE; /* ignore size changes */ - - error = inode_setattr(inode, iattr); - if (error) - return error; - if (!sd_attrs) { /* setting attributes for the first time, allocate now */ sd_attrs = sysfs_init_inode_attrs(sd); @@ -103,42 +90,78 @@ int sysfs_setattr(struct dentry * dentry, struct iattr * iattr) if (ia_valid & ATTR_GID) iattrs->ia_gid = iattr->ia_gid; if (ia_valid & ATTR_ATIME) - iattrs->ia_atime = timespec_trunc(iattr->ia_atime, - inode->i_sb->s_time_gran); + iattrs->ia_atime = iattr->ia_atime; if (ia_valid & ATTR_MTIME) - iattrs->ia_mtime = timespec_trunc(iattr->ia_mtime, - inode->i_sb->s_time_gran); + iattrs->ia_mtime = iattr->ia_mtime; if (ia_valid & ATTR_CTIME) - iattrs->ia_ctime = timespec_trunc(iattr->ia_ctime, - inode->i_sb->s_time_gran); + iattrs->ia_ctime = iattr->ia_ctime; if (ia_valid & ATTR_MODE) { umode_t mode = iattr->ia_mode; - - if (!in_group_p(inode->i_gid) && !capable(CAP_FSETID)) - mode &= ~S_ISGID; iattrs->ia_mode = sd->s_mode = mode; } } + return 0; +} + +int sysfs_setattr(struct dentry *dentry, struct iattr *iattr) +{ + struct inode *inode = dentry->d_inode; + struct sysfs_dirent *sd = dentry->d_fsdata; + int error; + + if (!sd) + return -EINVAL; + + error = inode_change_ok(inode, iattr); + if (error) + return error; + + iattr->ia_valid &= ~ATTR_SIZE; /* ignore size changes */ + + error = inode_setattr(inode, iattr); + if (error) + return error; + + mutex_lock(&sysfs_mutex); + error = sysfs_sd_setattr(sd, iattr); + mutex_unlock(&sysfs_mutex); + return error; } +static int sysfs_sd_setsecdata(struct sysfs_dirent *sd, void **secdata, u32 *secdata_len) +{ + struct sysfs_inode_attrs *iattrs; + void *old_secdata; + size_t old_secdata_len; + + iattrs = sd->s_iattr; + if (!iattrs) + iattrs = sysfs_init_inode_attrs(sd); + if (!iattrs) + return -ENOMEM; + + old_secdata = iattrs->ia_secdata; + old_secdata_len = iattrs->ia_secdata_len; + + iattrs->ia_secdata = *secdata; + iattrs->ia_secdata_len = *secdata_len; + + *secdata = old_secdata; + *secdata_len = old_secdata_len; + return 0; +} + int sysfs_setxattr(struct dentry *dentry, const char *name, const void *value, size_t size, int flags) { struct sysfs_dirent *sd = dentry->d_fsdata; - struct sysfs_inode_attrs *iattrs; void *secdata; int error; u32 secdata_len = 0; if (!sd) return -EINVAL; - if (!sd->s_iattr) - sd->s_iattr = sysfs_init_inode_attrs(sd); - if (!sd->s_iattr) - return -ENOMEM; - - iattrs = sd->s_iattr; if (!strncmp(name, XATTR_SECURITY_PREFIX, XATTR_SECURITY_PREFIX_LEN)) { const char *suffix = name + XATTR_SECURITY_PREFIX_LEN; @@ -150,12 +173,13 @@ int sysfs_setxattr(struct dentry *dentry, const char *name, const void *value, &secdata, &secdata_len); if (error) goto out; - if (iattrs->ia_secdata) - security_release_secctx(iattrs->ia_secdata, - iattrs->ia_secdata_len); - iattrs->ia_secdata = secdata; - iattrs->ia_secdata_len = secdata_len; + mutex_lock(&sysfs_mutex); + error = sysfs_sd_setsecdata(sd, &secdata, &secdata_len); + mutex_unlock(&sysfs_mutex); + + if (secdata) + security_release_secctx(secdata, secdata_len); } else return -EINVAL; out: @@ -170,7 +194,6 @@ static inline void set_default_inode_attr(struct inode * inode, mode_t mode) static inline void set_inode_attr(struct inode * inode, struct iattr * iattr) { - inode->i_mode = iattr->ia_mode; inode->i_uid = iattr->ia_uid; inode->i_gid = iattr->ia_gid; inode->i_atime = iattr->ia_atime; @@ -178,17 +201,6 @@ 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; - static int sysfs_count_nlink(struct sysfs_dirent *sd) { struct sysfs_dirent *child; @@ -201,38 +213,55 @@ static int sysfs_count_nlink(struct sysfs_dirent *sd) return nr + 2; } +static void sysfs_refresh_inode(struct sysfs_dirent *sd, struct inode *inode) +{ + struct sysfs_inode_attrs *iattrs = sd->s_iattr; + + inode->i_mode = sd->s_mode; + if (iattrs) { + /* sysfs_dirent has non-default attributes + * get them from persistent copy in sysfs_dirent + */ + set_inode_attr(inode, &iattrs->ia_iattr); + security_inode_notifysecctx(inode, + iattrs->ia_secdata, + iattrs->ia_secdata_len); + } + + if (sysfs_type(sd) == SYSFS_DIR) + inode->i_nlink = sysfs_count_nlink(sd); +} + +int sysfs_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat) +{ + struct sysfs_dirent *sd = dentry->d_fsdata; + struct inode *inode = dentry->d_inode; + + mutex_lock(&sysfs_mutex); + sysfs_refresh_inode(sd, inode); + mutex_unlock(&sysfs_mutex); + + generic_fillattr(inode, stat); + return 0; +} + static void sysfs_init_inode(struct sysfs_dirent *sd, struct inode *inode) { struct bin_attribute *bin_attr; - struct sysfs_inode_attrs *iattrs; inode->i_private = sysfs_get(sd); inode->i_mapping->a_ops = &sysfs_aops; inode->i_mapping->backing_dev_info = &sysfs_backing_dev_info; inode->i_op = &sysfs_inode_operations; - inode->i_ino = sd->s_ino; - lockdep_set_class(&inode->i_mutex, &sysfs_inode_imutex_key); - iattrs = sd->s_iattr; - if (iattrs) { - /* sysfs_dirent has non-default attributes - * get them for the new inode from persistent copy - * in sysfs_dirent - */ - set_inode_attr(inode, &iattrs->ia_iattr); - if (iattrs->ia_secdata) - security_inode_notifysecctx(inode, - iattrs->ia_secdata, - iattrs->ia_secdata_len); - } else - set_default_inode_attr(inode, sd->s_mode); + set_default_inode_attr(inode, sd->s_mode); + sysfs_refresh_inode(sd, inode); /* initialize inode according to type */ switch (sysfs_type(sd)) { case SYSFS_DIR: inode->i_op = &sysfs_dir_inode_operations; inode->i_fop = &sysfs_dir_operations; - inode->i_nlink = sysfs_count_nlink(sd); break; case SYSFS_KOBJ_ATTR: inode->i_size = PAGE_SIZE; @@ -315,3 +344,14 @@ int sysfs_hash_and_remove(struct sysfs_dirent *dir_sd, const char *name) else return -ENOENT; } + +int sysfs_permission(struct inode *inode, int mask) +{ + struct sysfs_dirent *sd = inode->i_private; + + mutex_lock(&sysfs_mutex); + sysfs_refresh_inode(sd, inode); + mutex_unlock(&sysfs_mutex); + + return generic_permission(inode, mask, NULL); +} diff --git a/fs/sysfs/symlink.c b/fs/sysfs/symlink.c index c5081ad7702..c5eff49fa41 100644 --- a/fs/sysfs/symlink.c +++ b/fs/sysfs/symlink.c @@ -210,10 +210,13 @@ static void sysfs_put_link(struct dentry *dentry, struct nameidata *nd, void *co } const struct inode_operations sysfs_symlink_inode_operations = { - .setxattr = sysfs_setxattr, - .readlink = generic_readlink, - .follow_link = sysfs_follow_link, - .put_link = sysfs_put_link, + .setxattr = sysfs_setxattr, + .readlink = generic_readlink, + .follow_link = sysfs_follow_link, + .put_link = sysfs_put_link, + .setattr = sysfs_setattr, + .getattr = sysfs_getattr, + .permission = sysfs_permission, }; diff --git a/fs/sysfs/sysfs.h b/fs/sysfs/sysfs.h index af4c4e7482a..ca52e7b9d8f 100644 --- a/fs/sysfs/sysfs.h +++ b/fs/sysfs/sysfs.h @@ -89,9 +89,7 @@ static inline unsigned int sysfs_type(struct sysfs_dirent *sd) */ struct sysfs_addrm_cxt { struct sysfs_dirent *parent_sd; - struct inode *parent_inode; struct sysfs_dirent *removed; - int cnt; }; /* @@ -105,7 +103,6 @@ extern struct kmem_cache *sysfs_dir_cachep; * dir.c */ extern struct mutex sysfs_mutex; -extern struct mutex sysfs_rename_mutex; extern spinlock_t sysfs_assoc_lock; extern const struct file_operations sysfs_dir_operations; @@ -133,6 +130,9 @@ int sysfs_create_subdir(struct kobject *kobj, const char *name, struct sysfs_dirent **p_sd); void sysfs_remove_subdir(struct sysfs_dirent *sd); +int sysfs_rename(struct sysfs_dirent *sd, + struct sysfs_dirent *new_parent_sd, const char *new_name); + static inline struct sysfs_dirent *__sysfs_get(struct sysfs_dirent *sd) { if (sd) { @@ -155,7 +155,10 @@ static inline void __sysfs_put(struct sysfs_dirent *sd) */ struct inode *sysfs_get_inode(struct sysfs_dirent *sd); void sysfs_delete_inode(struct inode *inode); +int sysfs_sd_setattr(struct sysfs_dirent *sd, struct iattr *iattr); +int sysfs_permission(struct inode *inode, int mask); int sysfs_setattr(struct dentry *dentry, struct iattr *iattr); +int sysfs_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat); int sysfs_setxattr(struct dentry *dentry, const char *name, const void *value, size_t size, int flags); int sysfs_hash_and_remove(struct sysfs_dirent *dir_sd, const char *name); diff --git a/fs/xfs/linux-2.6/xfs_aops.c b/fs/xfs/linux-2.6/xfs_aops.c index 70f989895d1..87813e405ce 100644 --- a/fs/xfs/linux-2.6/xfs_aops.c +++ b/fs/xfs/linux-2.6/xfs_aops.c @@ -235,71 +235,36 @@ xfs_setfilesize( } /* - * Buffered IO write completion for delayed allocate extents. + * IO write completion. */ STATIC void -xfs_end_bio_delalloc( - struct work_struct *work) -{ - xfs_ioend_t *ioend = - container_of(work, xfs_ioend_t, io_work); - - xfs_setfilesize(ioend); - xfs_destroy_ioend(ioend); -} - -/* - * Buffered IO write completion for regular, written extents. - */ -STATIC void -xfs_end_bio_written( - struct work_struct *work) -{ - xfs_ioend_t *ioend = - container_of(work, xfs_ioend_t, io_work); - - xfs_setfilesize(ioend); - xfs_destroy_ioend(ioend); -} - -/* - * IO write completion for unwritten extents. - * - * Issue transactions to convert a buffer range from unwritten - * to written extents. - */ -STATIC void -xfs_end_bio_unwritten( +xfs_end_io( struct work_struct *work) { xfs_ioend_t *ioend = container_of(work, xfs_ioend_t, io_work); struct xfs_inode *ip = XFS_I(ioend->io_inode); - xfs_off_t offset = ioend->io_offset; - size_t size = ioend->io_size; - - if (likely(!ioend->io_error)) { - if (!XFS_FORCED_SHUTDOWN(ip->i_mount)) { - int error; - error = xfs_iomap_write_unwritten(ip, offset, size); - if (error) - ioend->io_error = error; - } - xfs_setfilesize(ioend); - } - xfs_destroy_ioend(ioend); -} -/* - * IO read completion for regular, written extents. - */ -STATIC void -xfs_end_bio_read( - struct work_struct *work) -{ - xfs_ioend_t *ioend = - container_of(work, xfs_ioend_t, io_work); + /* + * For unwritten extents we need to issue transactions to convert a + * range to normal written extens after the data I/O has finished. + */ + if (ioend->io_type == IOMAP_UNWRITTEN && + likely(!ioend->io_error && !XFS_FORCED_SHUTDOWN(ip->i_mount))) { + int error; + + error = xfs_iomap_write_unwritten(ip, ioend->io_offset, + ioend->io_size); + if (error) + ioend->io_error = error; + } + /* + * We might have to update the on-disk file size after extending + * writes. + */ + if (ioend->io_type != IOMAP_READ) + xfs_setfilesize(ioend); xfs_destroy_ioend(ioend); } @@ -314,10 +279,10 @@ xfs_finish_ioend( int wait) { if (atomic_dec_and_test(&ioend->io_remaining)) { - struct workqueue_struct *wq = xfsdatad_workqueue; - if (ioend->io_work.func == xfs_end_bio_unwritten) - wq = xfsconvertd_workqueue; + struct workqueue_struct *wq; + wq = (ioend->io_type == IOMAP_UNWRITTEN) ? + xfsconvertd_workqueue : xfsdatad_workqueue; queue_work(wq, &ioend->io_work); if (wait) flush_workqueue(wq); @@ -355,15 +320,7 @@ xfs_alloc_ioend( ioend->io_offset = 0; ioend->io_size = 0; - if (type == IOMAP_UNWRITTEN) - INIT_WORK(&ioend->io_work, xfs_end_bio_unwritten); - else if (type == IOMAP_DELAY) - INIT_WORK(&ioend->io_work, xfs_end_bio_delalloc); - else if (type == IOMAP_READ) - INIT_WORK(&ioend->io_work, xfs_end_bio_read); - else - INIT_WORK(&ioend->io_work, xfs_end_bio_written); - + INIT_WORK(&ioend->io_work, xfs_end_io); return ioend; } @@ -380,7 +337,7 @@ xfs_map_blocks( return -xfs_iomap(XFS_I(inode), offset, count, flags, mapp, &nmaps); } -STATIC_INLINE int +STATIC int xfs_iomap_valid( xfs_iomap_t *iomapp, loff_t offset) @@ -412,8 +369,9 @@ xfs_end_bio( STATIC void xfs_submit_ioend_bio( - xfs_ioend_t *ioend, - struct bio *bio) + struct writeback_control *wbc, + xfs_ioend_t *ioend, + struct bio *bio) { atomic_inc(&ioend->io_remaining); bio->bi_private = ioend; @@ -426,7 +384,8 @@ xfs_submit_ioend_bio( if (xfs_ioend_new_eof(ioend)) xfs_mark_inode_dirty_sync(XFS_I(ioend->io_inode)); - submit_bio(WRITE, bio); + submit_bio(wbc->sync_mode == WB_SYNC_ALL ? + WRITE_SYNC_PLUG : WRITE, bio); ASSERT(!bio_flagged(bio, BIO_EOPNOTSUPP)); bio_put(bio); } @@ -505,6 +464,7 @@ static inline int bio_add_buffer(struct bio *bio, struct buffer_head *bh) */ STATIC void xfs_submit_ioend( + struct writeback_control *wbc, xfs_ioend_t *ioend) { xfs_ioend_t *head = ioend; @@ -533,19 +493,19 @@ xfs_submit_ioend( retry: bio = xfs_alloc_ioend_bio(bh); } else if (bh->b_blocknr != lastblock + 1) { - xfs_submit_ioend_bio(ioend, bio); + xfs_submit_ioend_bio(wbc, ioend, bio); goto retry; } if (bio_add_buffer(bio, bh) != bh->b_size) { - xfs_submit_ioend_bio(ioend, bio); + xfs_submit_ioend_bio(wbc, ioend, bio); goto retry; } lastblock = bh->b_blocknr; } if (bio) - xfs_submit_ioend_bio(ioend, bio); + xfs_submit_ioend_bio(wbc, ioend, bio); xfs_finish_ioend(ioend, 0); } while ((ioend = next) != NULL); } @@ -1191,7 +1151,7 @@ xfs_page_state_convert( } if (iohead) - xfs_submit_ioend(iohead); + xfs_submit_ioend(wbc, iohead); return page_dirty; @@ -1528,7 +1488,7 @@ xfs_end_io_direct( * didn't map an unwritten extent so switch it's completion * handler. */ - INIT_WORK(&ioend->io_work, xfs_end_bio_written); + ioend->io_type = IOMAP_NEW; xfs_finish_ioend(ioend, 0); } diff --git a/fs/xfs/linux-2.6/xfs_buf.c b/fs/xfs/linux-2.6/xfs_buf.c index 965df1227d6..4ddc973aea7 100644 --- a/fs/xfs/linux-2.6/xfs_buf.c +++ b/fs/xfs/linux-2.6/xfs_buf.c @@ -149,7 +149,7 @@ page_region_mask( return mask; } -STATIC_INLINE void +STATIC void set_page_region( struct page *page, size_t offset, @@ -161,7 +161,7 @@ set_page_region( SetPageUptodate(page); } -STATIC_INLINE int +STATIC int test_page_region( struct page *page, size_t offset, @@ -582,7 +582,7 @@ found: * although backing storage may not be. */ xfs_buf_t * -xfs_buf_get_flags( +xfs_buf_get( xfs_buftarg_t *target,/* target for buffer */ xfs_off_t ioff, /* starting offset of range */ size_t isize, /* length of range */ @@ -661,7 +661,7 @@ _xfs_buf_read( } xfs_buf_t * -xfs_buf_read_flags( +xfs_buf_read( xfs_buftarg_t *target, xfs_off_t ioff, size_t isize, @@ -671,7 +671,7 @@ xfs_buf_read_flags( flags |= XBF_READ; - bp = xfs_buf_get_flags(target, ioff, isize, flags); + bp = xfs_buf_get(target, ioff, isize, flags); if (bp) { if (!XFS_BUF_ISDONE(bp)) { XB_TRACE(bp, "read", (unsigned long)flags); @@ -718,7 +718,7 @@ xfs_buf_readahead( return; flags |= (XBF_TRYLOCK|XBF_ASYNC|XBF_READ_AHEAD); - xfs_buf_read_flags(target, ioff, isize, flags); + xfs_buf_read(target, ioff, isize, flags); } xfs_buf_t * @@ -1113,7 +1113,7 @@ xfs_bdwrite( xfs_buf_delwri_queue(bp, 1); } -STATIC_INLINE void +STATIC void _xfs_buf_ioend( xfs_buf_t *bp, int schedule) diff --git a/fs/xfs/linux-2.6/xfs_buf.h b/fs/xfs/linux-2.6/xfs_buf.h index 9b4d666ad31..5f07dd91c5f 100644 --- a/fs/xfs/linux-2.6/xfs_buf.h +++ b/fs/xfs/linux-2.6/xfs_buf.h @@ -186,15 +186,10 @@ extern xfs_buf_t *_xfs_buf_find(xfs_buftarg_t *, xfs_off_t, size_t, #define xfs_incore(buftarg,blkno,len,lockit) \ _xfs_buf_find(buftarg, blkno ,len, lockit, NULL) -extern xfs_buf_t *xfs_buf_get_flags(xfs_buftarg_t *, xfs_off_t, size_t, +extern xfs_buf_t *xfs_buf_get(xfs_buftarg_t *, xfs_off_t, size_t, xfs_buf_flags_t); -#define xfs_buf_get(target, blkno, len, flags) \ - xfs_buf_get_flags((target), (blkno), (len), XBF_LOCK | XBF_MAPPED) - -extern xfs_buf_t *xfs_buf_read_flags(xfs_buftarg_t *, xfs_off_t, size_t, +extern xfs_buf_t *xfs_buf_read(xfs_buftarg_t *, xfs_off_t, size_t, xfs_buf_flags_t); -#define xfs_buf_read(target, blkno, len, flags) \ - xfs_buf_read_flags((target), (blkno), (len), XBF_LOCK | XBF_MAPPED) extern xfs_buf_t *xfs_buf_get_empty(size_t, xfs_buftarg_t *); extern xfs_buf_t *xfs_buf_get_noaddr(size_t, xfs_buftarg_t *); diff --git a/fs/xfs/linux-2.6/xfs_file.c b/fs/xfs/linux-2.6/xfs_file.c index eff61e2732a..e4caeb28ce2 100644 --- a/fs/xfs/linux-2.6/xfs_file.c +++ b/fs/xfs/linux-2.6/xfs_file.c @@ -52,7 +52,7 @@ xfs_file_aio_read( loff_t pos) { struct file *file = iocb->ki_filp; - int ioflags = IO_ISAIO; + int ioflags = 0; BUG_ON(iocb->ki_pos != pos); if (unlikely(file->f_flags & O_DIRECT)) @@ -71,7 +71,7 @@ xfs_file_aio_write( loff_t pos) { struct file *file = iocb->ki_filp; - int ioflags = IO_ISAIO; + int ioflags = 0; BUG_ON(iocb->ki_pos != pos); if (unlikely(file->f_flags & O_DIRECT)) diff --git a/fs/xfs/linux-2.6/xfs_iops.c b/fs/xfs/linux-2.6/xfs_iops.c index cd42ef78f6b..1f3b4b8f7dd 100644 --- a/fs/xfs/linux-2.6/xfs_iops.c +++ b/fs/xfs/linux-2.6/xfs_iops.c @@ -573,8 +573,8 @@ xfs_vn_fallocate( bf.l_len = len; xfs_ilock(ip, XFS_IOLOCK_EXCL); - error = xfs_change_file_space(ip, XFS_IOC_RESVSP, &bf, - 0, XFS_ATTR_NOLOCK); + error = -xfs_change_file_space(ip, XFS_IOC_RESVSP, &bf, + 0, XFS_ATTR_NOLOCK); if (!error && !(mode & FALLOC_FL_KEEP_SIZE) && offset + len > i_size_read(inode)) new_size = offset + len; @@ -585,7 +585,7 @@ xfs_vn_fallocate( iattr.ia_valid = ATTR_SIZE; iattr.ia_size = new_size; - error = xfs_setattr(ip, &iattr, XFS_ATTR_NOLOCK); + error = -xfs_setattr(ip, &iattr, XFS_ATTR_NOLOCK); } xfs_iunlock(ip, XFS_IOLOCK_EXCL); diff --git a/fs/xfs/linux-2.6/xfs_lrw.c b/fs/xfs/linux-2.6/xfs_lrw.c index 072050f8d34..1bf47f219c9 100644 --- a/fs/xfs/linux-2.6/xfs_lrw.c +++ b/fs/xfs/linux-2.6/xfs_lrw.c @@ -255,8 +255,6 @@ xfs_read( iocb->ki_pos = *offset; ret = generic_file_aio_read(iocb, iovp, segs, *offset); - if (ret == -EIOCBQUEUED && !(ioflags & IO_ISAIO)) - ret = wait_on_sync_kiocb(iocb); if (ret > 0) XFS_STATS_ADD(xs_read_bytes, ret); @@ -774,9 +772,6 @@ write_retry: current->backing_dev_info = NULL; - if (ret == -EIOCBQUEUED && !(ioflags & IO_ISAIO)) - ret = wait_on_sync_kiocb(iocb); - isize = i_size_read(inode); if (unlikely(ret < 0 && ret != -EFAULT && *offset > isize)) *offset = isize; @@ -811,7 +806,7 @@ write_retry: XFS_STATS_ADD(xs_write_bytes, ret); /* Handle various SYNC-type writes */ - if ((file->f_flags & O_SYNC) || IS_SYNC(inode)) { + if ((file->f_flags & O_DSYNC) || IS_SYNC(inode)) { loff_t end = pos + ret - 1; int error2; diff --git a/fs/xfs/linux-2.6/xfs_super.c b/fs/xfs/linux-2.6/xfs_super.c index 18a4b8e11df..1bfb0e98019 100644 --- a/fs/xfs/linux-2.6/xfs_super.c +++ b/fs/xfs/linux-2.6/xfs_super.c @@ -930,13 +930,39 @@ xfs_fs_alloc_inode( */ STATIC void xfs_fs_destroy_inode( - struct inode *inode) + struct inode *inode) { - xfs_inode_t *ip = XFS_I(inode); + struct xfs_inode *ip = XFS_I(inode); + + xfs_itrace_entry(ip); XFS_STATS_INC(vn_reclaim); - if (xfs_reclaim(ip)) - panic("%s: cannot reclaim 0x%p\n", __func__, inode); + + /* bad inode, get out here ASAP */ + if (is_bad_inode(inode)) + goto out_reclaim; + + xfs_ioend_wait(ip); + + ASSERT(XFS_FORCED_SHUTDOWN(ip->i_mount) || ip->i_delayed_blks == 0); + + /* + * We should never get here with one of the reclaim flags already set. + */ + ASSERT_ALWAYS(!xfs_iflags_test(ip, XFS_IRECLAIMABLE)); + ASSERT_ALWAYS(!xfs_iflags_test(ip, XFS_IRECLAIM)); + + /* + * If we have nothing to flush with this inode then complete the + * teardown now, otherwise delay the flush operation. + */ + if (!xfs_inode_clean(ip)) { + xfs_inode_set_reclaim_tag(ip); + return; + } + +out_reclaim: + xfs_ireclaim(ip); } /* @@ -973,7 +999,6 @@ xfs_fs_inode_init_once( mrlock_init(&ip->i_lock, MRLOCK_ALLOW_EQUAL_PRI|MRLOCK_BARRIER, "xfsino", ip->i_ino); - mrlock_init(&ip->i_iolock, MRLOCK_BARRIER, "xfsio", ip->i_ino); } /* @@ -1075,6 +1100,20 @@ xfs_fs_clear_inode( XFS_STATS_INC(vn_remove); XFS_STATS_DEC(vn_active); + /* + * The iolock is used by the file system to coordinate reads, + * writes, and block truncates. Up to this point the lock + * protected concurrent accesses by users of the inode. But + * from here forward we're doing some final processing of the + * inode because we're done with it, and although we reuse the + * iolock for protection it is really a distinct lock class + * (in the lockdep sense) from before. To keep lockdep happy + * (and basically indicate what we are doing), we explicitly + * re-init the iolock here. + */ + ASSERT(!rwsem_is_locked(&ip->i_iolock.mr_lock)); + mrlock_init(&ip->i_iolock, MRLOCK_BARRIER, "xfsio", ip->i_ino); + xfs_inactive(ip); } @@ -1092,8 +1131,6 @@ xfs_fs_put_super( struct super_block *sb) { struct xfs_mount *mp = XFS_M(sb); - struct xfs_inode *rip = mp->m_rootip; - int unmount_event_flags = 0; xfs_syncd_stop(mp); @@ -1109,20 +1146,7 @@ xfs_fs_put_super( xfs_sync_attr(mp, 0); } -#ifdef HAVE_DMAPI - if (mp->m_flags & XFS_MOUNT_DMAPI) { - unmount_event_flags = - (mp->m_dmevmask & (1 << DM_EVENT_UNMOUNT)) ? - 0 : DM_FLAGS_UNWANTED; - /* - * Ignore error from dmapi here, first unmount is not allowed - * to fail anyway, and second we wouldn't want to fail a - * unmount because of dmapi. - */ - XFS_SEND_PREUNMOUNT(mp, rip, DM_RIGHT_NULL, rip, DM_RIGHT_NULL, - NULL, NULL, 0, 0, unmount_event_flags); - } -#endif + XFS_SEND_PREUNMOUNT(mp); /* * Blow away any referenced inode in the filestreams cache. @@ -1133,10 +1157,7 @@ xfs_fs_put_super( XFS_bflush(mp->m_ddev_targp); - if (mp->m_flags & XFS_MOUNT_DMAPI) { - XFS_SEND_UNMOUNT(mp, rip, DM_RIGHT_NULL, 0, 0, - unmount_event_flags); - } + XFS_SEND_UNMOUNT(mp); xfs_unmountfs(mp); xfs_freesb(mp); diff --git a/fs/xfs/linux-2.6/xfs_sync.c b/fs/xfs/linux-2.6/xfs_sync.c index 961df0a22c7..d895a3a960f 100644 --- a/fs/xfs/linux-2.6/xfs_sync.c +++ b/fs/xfs/linux-2.6/xfs_sync.c @@ -663,10 +663,9 @@ xfs_syncd_stop( kthread_stop(mp->m_sync_task); } -int +STATIC int xfs_reclaim_inode( xfs_inode_t *ip, - int locked, int sync_mode) { xfs_perag_t *pag = xfs_get_perag(ip->i_mount, ip->i_ino); @@ -682,10 +681,6 @@ xfs_reclaim_inode( !__xfs_iflags_test(ip, XFS_IRECLAIMABLE)) { spin_unlock(&ip->i_flags_lock); write_unlock(&pag->pag_ici_lock); - if (locked) { - xfs_ifunlock(ip); - xfs_iunlock(ip, XFS_ILOCK_EXCL); - } return -EAGAIN; } __xfs_iflags_set(ip, XFS_IRECLAIM); @@ -704,10 +699,8 @@ xfs_reclaim_inode( * We get the flush lock regardless, though, just to make sure * we don't free it while it is being flushed. */ - if (!locked) { - xfs_ilock(ip, XFS_ILOCK_EXCL); - xfs_iflock(ip); - } + xfs_ilock(ip, XFS_ILOCK_EXCL); + xfs_iflock(ip); /* * In the case of a forced shutdown we rely on xfs_iflush() to @@ -778,7 +771,7 @@ xfs_reclaim_inode_now( } read_unlock(&pag->pag_ici_lock); - return xfs_reclaim_inode(ip, 0, flags); + return xfs_reclaim_inode(ip, flags); } int diff --git a/fs/xfs/linux-2.6/xfs_sync.h b/fs/xfs/linux-2.6/xfs_sync.h index 27920eb7a82..a500b4d9183 100644 --- a/fs/xfs/linux-2.6/xfs_sync.h +++ b/fs/xfs/linux-2.6/xfs_sync.h @@ -44,7 +44,6 @@ void xfs_quiesce_attr(struct xfs_mount *mp); void xfs_flush_inodes(struct xfs_inode *ip); -int xfs_reclaim_inode(struct xfs_inode *ip, int locked, int sync_mode); int xfs_reclaim_inodes(struct xfs_mount *mp, int mode); void xfs_inode_set_reclaim_tag(struct xfs_inode *ip); diff --git a/fs/xfs/linux-2.6/xfs_vnode.h b/fs/xfs/linux-2.6/xfs_vnode.h index ad7fbead4c9..00cabf5354d 100644 --- a/fs/xfs/linux-2.6/xfs_vnode.h +++ b/fs/xfs/linux-2.6/xfs_vnode.h @@ -36,7 +36,6 @@ struct attrlist_cursor_kern; /* * Flags for read/write calls - same values as IRIX */ -#define IO_ISAIO 0x00001 /* don't wait for completion */ #define IO_ISDIRECT 0x00004 /* bypass page cache */ #define IO_INVIS 0x00020 /* don't update inode timestamps */ diff --git a/fs/xfs/support/debug.h b/fs/xfs/support/debug.h index 6f4fd37c67a..d2d20462fd4 100644 --- a/fs/xfs/support/debug.h +++ b/fs/xfs/support/debug.h @@ -41,10 +41,6 @@ extern void assfail(char *expr, char *f, int l); # define STATIC static noinline #endif -#ifndef STATIC_INLINE -# define STATIC_INLINE static inline -#endif - #else /* DEBUG */ #define ASSERT(expr) \ @@ -54,19 +50,5 @@ extern void assfail(char *expr, char *f, int l); # define STATIC noinline #endif -/* - * We stop inlining of inline functions in debug mode. - * Unfortunately, this means static inline in header files - * get multiple definitions, so they need to remain static. - * This then gives tonnes of warnings about unused but defined - * functions, so we need to add the unused attribute to prevent - * these spurious warnings. - */ -#ifndef STATIC_INLINE -# define STATIC_INLINE static __attribute__ ((unused)) noinline -#endif - #endif /* DEBUG */ - - #endif /* __XFS_SUPPORT_DEBUG_H__ */ diff --git a/fs/xfs/xfs_attr.c b/fs/xfs/xfs_attr.c index 4ece1906bd4..8fe6f6b78a4 100644 --- a/fs/xfs/xfs_attr.c +++ b/fs/xfs/xfs_attr.c @@ -123,9 +123,13 @@ xfs_inode_hasattr( * Overall external interface routines. *========================================================================*/ -int -xfs_attr_fetch(xfs_inode_t *ip, struct xfs_name *name, - char *value, int *valuelenp, int flags) +STATIC int +xfs_attr_get_int( + struct xfs_inode *ip, + struct xfs_name *name, + char *value, + int *valuelenp, + int flags) { xfs_da_args_t args; int error; @@ -188,7 +192,7 @@ xfs_attr_get( return error; xfs_ilock(ip, XFS_ILOCK_SHARED); - error = xfs_attr_fetch(ip, &xname, value, valuelenp, flags); + error = xfs_attr_get_int(ip, &xname, value, valuelenp, flags); xfs_iunlock(ip, XFS_ILOCK_SHARED); return(error); } @@ -2143,8 +2147,8 @@ xfs_attr_rmtval_set(xfs_da_args_t *args) dblkno = XFS_FSB_TO_DADDR(mp, map.br_startblock), blkcnt = XFS_FSB_TO_BB(mp, map.br_blockcount); - bp = xfs_buf_get_flags(mp->m_ddev_targp, dblkno, blkcnt, - XFS_BUF_LOCK | XBF_DONT_BLOCK); + bp = xfs_buf_get(mp->m_ddev_targp, dblkno, blkcnt, + XFS_BUF_LOCK | XBF_DONT_BLOCK); ASSERT(bp); ASSERT(!XFS_BUF_GETERROR(bp)); diff --git a/fs/xfs/xfs_attr.h b/fs/xfs/xfs_attr.h index fb3b2a68b9b..12f0be3a73d 100644 --- a/fs/xfs/xfs_attr.h +++ b/fs/xfs/xfs_attr.h @@ -131,7 +131,6 @@ typedef struct xfs_attr_list_context { */ int xfs_attr_calc_size(struct xfs_inode *, int, int, int *); int xfs_attr_inactive(struct xfs_inode *dp); -int xfs_attr_fetch(struct xfs_inode *, struct xfs_name *, char *, int *, int); int xfs_attr_rmtval_get(struct xfs_da_args *args); int xfs_attr_list_int(struct xfs_attr_list_context *); diff --git a/fs/xfs/xfs_attr_leaf.c b/fs/xfs/xfs_attr_leaf.c index afdc8911637..0b687351293 100644 --- a/fs/xfs/xfs_attr_leaf.c +++ b/fs/xfs/xfs_attr_leaf.c @@ -98,7 +98,7 @@ STATIC int xfs_attr_leaf_entsize(xfs_attr_leafblock_t *leaf, int index); * If namespace bits don't match return 0. * If all match then return 1. */ -STATIC_INLINE int +STATIC 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); diff --git a/fs/xfs/xfs_bmap_btree.c b/fs/xfs/xfs_bmap_btree.c index eb7b702d069..6f5ccede63f 100644 --- a/fs/xfs/xfs_bmap_btree.c +++ b/fs/xfs/xfs_bmap_btree.c @@ -98,8 +98,7 @@ xfs_bmdr_to_bmbt( * This code must be in sync with the routines xfs_bmbt_get_startoff, * xfs_bmbt_get_startblock, xfs_bmbt_get_blockcount and xfs_bmbt_get_state. */ - -STATIC_INLINE void +STATIC void __xfs_bmbt_get_all( __uint64_t l0, __uint64_t l1, diff --git a/fs/xfs/xfs_filestream.h b/fs/xfs/xfs_filestream.h index f655f7dc334..4aba67c5f64 100644 --- a/fs/xfs/xfs_filestream.h +++ b/fs/xfs/xfs_filestream.h @@ -79,7 +79,7 @@ extern ktrace_t *xfs_filestreams_trace_buf; * the cache that reference per-ag array elements that have since been * reallocated. */ -STATIC_INLINE int +static inline int xfs_filestream_peek_ag( xfs_mount_t *mp, xfs_agnumber_t agno) @@ -87,7 +87,7 @@ xfs_filestream_peek_ag( return atomic_read(&mp->m_perag[agno].pagf_fstrms); } -STATIC_INLINE int +static inline int xfs_filestream_get_ag( xfs_mount_t *mp, xfs_agnumber_t agno) @@ -95,7 +95,7 @@ xfs_filestream_get_ag( return atomic_inc_return(&mp->m_perag[agno].pagf_fstrms); } -STATIC_INLINE int +static inline int xfs_filestream_put_ag( xfs_mount_t *mp, xfs_agnumber_t agno) @@ -122,7 +122,7 @@ int xfs_filestream_new_ag(struct xfs_bmalloca *ap, xfs_agnumber_t *agp); /* filestreams for the inode? */ -STATIC_INLINE int +static inline int xfs_inode_is_filestream( struct xfs_inode *ip) { diff --git a/fs/xfs/xfs_fsops.c b/fs/xfs/xfs_fsops.c index 2d0b3e1da9e..36079aa9134 100644 --- a/fs/xfs/xfs_fsops.c +++ b/fs/xfs/xfs_fsops.c @@ -201,8 +201,8 @@ xfs_growfs_data_private( * AG freelist header block */ bp = xfs_buf_get(mp->m_ddev_targp, - XFS_AG_DADDR(mp, agno, XFS_AGF_DADDR(mp)), - XFS_FSS_TO_BB(mp, 1), 0); + XFS_AG_DADDR(mp, agno, XFS_AGF_DADDR(mp)), + XFS_FSS_TO_BB(mp, 1), XBF_LOCK | XBF_MAPPED); agf = XFS_BUF_TO_AGF(bp); memset(agf, 0, mp->m_sb.sb_sectsize); agf->agf_magicnum = cpu_to_be32(XFS_AGF_MAGIC); @@ -233,8 +233,8 @@ xfs_growfs_data_private( * AG inode header block */ bp = xfs_buf_get(mp->m_ddev_targp, - XFS_AG_DADDR(mp, agno, XFS_AGI_DADDR(mp)), - XFS_FSS_TO_BB(mp, 1), 0); + XFS_AG_DADDR(mp, agno, XFS_AGI_DADDR(mp)), + XFS_FSS_TO_BB(mp, 1), XBF_LOCK | XBF_MAPPED); agi = XFS_BUF_TO_AGI(bp); memset(agi, 0, mp->m_sb.sb_sectsize); agi->agi_magicnum = cpu_to_be32(XFS_AGI_MAGIC); @@ -257,8 +257,9 @@ xfs_growfs_data_private( * BNO btree root block */ bp = xfs_buf_get(mp->m_ddev_targp, - XFS_AGB_TO_DADDR(mp, agno, XFS_BNO_BLOCK(mp)), - BTOBB(mp->m_sb.sb_blocksize), 0); + XFS_AGB_TO_DADDR(mp, agno, XFS_BNO_BLOCK(mp)), + BTOBB(mp->m_sb.sb_blocksize), + XBF_LOCK | XBF_MAPPED); block = XFS_BUF_TO_BLOCK(bp); memset(block, 0, mp->m_sb.sb_blocksize); block->bb_magic = cpu_to_be32(XFS_ABTB_MAGIC); @@ -278,8 +279,9 @@ xfs_growfs_data_private( * CNT btree root block */ bp = xfs_buf_get(mp->m_ddev_targp, - XFS_AGB_TO_DADDR(mp, agno, XFS_CNT_BLOCK(mp)), - BTOBB(mp->m_sb.sb_blocksize), 0); + XFS_AGB_TO_DADDR(mp, agno, XFS_CNT_BLOCK(mp)), + BTOBB(mp->m_sb.sb_blocksize), + XBF_LOCK | XBF_MAPPED); block = XFS_BUF_TO_BLOCK(bp); memset(block, 0, mp->m_sb.sb_blocksize); block->bb_magic = cpu_to_be32(XFS_ABTC_MAGIC); @@ -300,8 +302,9 @@ xfs_growfs_data_private( * INO btree root block */ bp = xfs_buf_get(mp->m_ddev_targp, - XFS_AGB_TO_DADDR(mp, agno, XFS_IBT_BLOCK(mp)), - BTOBB(mp->m_sb.sb_blocksize), 0); + XFS_AGB_TO_DADDR(mp, agno, XFS_IBT_BLOCK(mp)), + BTOBB(mp->m_sb.sb_blocksize), + XBF_LOCK | XBF_MAPPED); block = XFS_BUF_TO_BLOCK(bp); memset(block, 0, mp->m_sb.sb_blocksize); block->bb_magic = cpu_to_be32(XFS_IBT_MAGIC); @@ -611,7 +614,7 @@ xfs_fs_log_dummy( xfs_inode_t *ip; int error; - tp = _xfs_trans_alloc(mp, XFS_TRANS_DUMMY1); + tp = _xfs_trans_alloc(mp, XFS_TRANS_DUMMY1, KM_SLEEP); error = xfs_trans_reserve(tp, 0, XFS_ICHANGE_LOG_RES(mp), 0, 0, 0); if (error) { xfs_trans_cancel(tp, 0); diff --git a/fs/xfs/xfs_ialloc.c b/fs/xfs/xfs_ialloc.c index 0785797db82..cb907ba69c4 100644 --- a/fs/xfs/xfs_ialloc.c +++ b/fs/xfs/xfs_ialloc.c @@ -425,7 +425,7 @@ xfs_ialloc_ag_alloc( return 0; } -STATIC_INLINE xfs_agnumber_t +STATIC xfs_agnumber_t xfs_ialloc_next_ag( xfs_mount_t *mp) { diff --git a/fs/xfs/xfs_iget.c b/fs/xfs/xfs_iget.c index 80e526489be..073bb4a26b1 100644 --- a/fs/xfs/xfs_iget.c +++ b/fs/xfs/xfs_iget.c @@ -73,6 +73,9 @@ xfs_inode_alloc( ASSERT(atomic_read(&ip->i_pincount) == 0); ASSERT(!spin_is_locked(&ip->i_flags_lock)); ASSERT(completion_done(&ip->i_flush)); + ASSERT(!rwsem_is_locked(&ip->i_iolock.mr_lock)); + + mrlock_init(&ip->i_iolock, MRLOCK_BARRIER, "xfsio", ip->i_ino); /* initialise the xfs inode */ ip->i_ino = ino; @@ -290,7 +293,7 @@ xfs_iget_cache_miss( struct xfs_inode **ipp, xfs_daddr_t bno, int flags, - int lock_flags) __releases(pag->pag_ici_lock) + int lock_flags) { struct xfs_inode *ip; int error; diff --git a/fs/xfs/xfs_iomap.c b/fs/xfs/xfs_iomap.c index 67ae5555a30..7294abce6ef 100644 --- a/fs/xfs/xfs_iomap.c +++ b/fs/xfs/xfs_iomap.c @@ -860,8 +860,15 @@ xfs_iomap_write_unwritten( * set up a transaction to convert the range of extents * from unwritten to real. Do allocations in a loop until * we have covered the range passed in. + * + * Note that we open code the transaction allocation here + * to pass KM_NOFS--we can't risk to recursing back into + * the filesystem here as we might be asked to write out + * the same inode that we complete here and might deadlock + * on the iolock. */ - tp = xfs_trans_alloc(mp, XFS_TRANS_STRAT_WRITE); + xfs_wait_for_freeze(mp, SB_FREEZE_TRANS); + tp = _xfs_trans_alloc(mp, XFS_TRANS_STRAT_WRITE, KM_NOFS); tp->t_flags |= XFS_TRANS_RESERVE; error = xfs_trans_reserve(tp, resblks, XFS_WRITE_LOG_RES(mp), 0, diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c index fb17f8226b0..1ec98ed914d 100644 --- a/fs/xfs/xfs_log_recover.c +++ b/fs/xfs/xfs_log_recover.c @@ -2206,6 +2206,7 @@ xlog_recover_do_buffer_trans( xfs_daddr_t blkno; int len; ushort flags; + uint buf_flags; buf_f = (xfs_buf_log_format_t *)item->ri_buf[0].i_addr; @@ -2246,12 +2247,11 @@ xlog_recover_do_buffer_trans( } mp = log->l_mp; - if (flags & XFS_BLI_INODE_BUF) { - bp = xfs_buf_read_flags(mp->m_ddev_targp, blkno, len, - XFS_BUF_LOCK); - } else { - bp = xfs_buf_read(mp->m_ddev_targp, blkno, len, 0); - } + buf_flags = XFS_BUF_LOCK; + if (!(flags & XFS_BLI_INODE_BUF)) + buf_flags |= XFS_BUF_MAPPED; + + bp = xfs_buf_read(mp->m_ddev_targp, blkno, len, buf_flags); if (XFS_BUF_ISERROR(bp)) { xfs_ioerror_alert("xlog_recover_do..(read#1)", log->l_mp, bp, blkno); @@ -2350,8 +2350,8 @@ xlog_recover_do_inode_trans( goto error; } - bp = xfs_buf_read_flags(mp->m_ddev_targp, in_f->ilf_blkno, - in_f->ilf_len, XFS_BUF_LOCK); + bp = xfs_buf_read(mp->m_ddev_targp, in_f->ilf_blkno, in_f->ilf_len, + XFS_BUF_LOCK); if (XFS_BUF_ISERROR(bp)) { xfs_ioerror_alert("xlog_recover_do..(read#2)", mp, bp, in_f->ilf_blkno); @@ -3517,7 +3517,7 @@ xlog_do_recovery_pass( { xlog_rec_header_t *rhead; xfs_daddr_t blk_no; - xfs_caddr_t bufaddr, offset; + xfs_caddr_t offset; xfs_buf_t *hbp, *dbp; int error = 0, h_size; int bblks, split_bblks; @@ -3610,7 +3610,7 @@ xlog_do_recovery_pass( /* * Check for header wrapping around physical end-of-log */ - offset = NULL; + offset = XFS_BUF_PTR(hbp); split_hblks = 0; wrapped_hblks = 0; if (blk_no + hblks <= log->l_logBBsize) { @@ -3646,9 +3646,8 @@ xlog_do_recovery_pass( * - order is important. */ wrapped_hblks = hblks - split_hblks; - bufaddr = XFS_BUF_PTR(hbp); error = XFS_BUF_SET_PTR(hbp, - bufaddr + BBTOB(split_hblks), + offset + BBTOB(split_hblks), BBTOB(hblks - split_hblks)); if (error) goto bread_err2; @@ -3658,14 +3657,10 @@ xlog_do_recovery_pass( if (error) goto bread_err2; - error = XFS_BUF_SET_PTR(hbp, bufaddr, + error = XFS_BUF_SET_PTR(hbp, offset, BBTOB(hblks)); if (error) goto bread_err2; - - if (!offset) - offset = xlog_align(log, 0, - wrapped_hblks, hbp); } rhead = (xlog_rec_header_t *)offset; error = xlog_valid_rec_header(log, rhead, @@ -3685,7 +3680,7 @@ xlog_do_recovery_pass( } else { /* This log record is split across the * physical end of log */ - offset = NULL; + offset = XFS_BUF_PTR(dbp); split_bblks = 0; if (blk_no != log->l_logBBsize) { /* some data is before the physical @@ -3714,9 +3709,8 @@ xlog_do_recovery_pass( * _first_, then the log start (LR header end) * - order is important. */ - bufaddr = XFS_BUF_PTR(dbp); error = XFS_BUF_SET_PTR(dbp, - bufaddr + BBTOB(split_bblks), + offset + BBTOB(split_bblks), BBTOB(bblks - split_bblks)); if (error) goto bread_err2; @@ -3727,13 +3721,9 @@ xlog_do_recovery_pass( if (error) goto bread_err2; - error = XFS_BUF_SET_PTR(dbp, bufaddr, h_size); + error = XFS_BUF_SET_PTR(dbp, offset, h_size); if (error) goto bread_err2; - - if (!offset) - offset = xlog_align(log, wrapped_hblks, - bblks - split_bblks, dbp); } xlog_unpack_data(rhead, offset, log); if ((error = xlog_recover_process_data(log, rhash, diff --git a/fs/xfs/xfs_mount.c b/fs/xfs/xfs_mount.c index 8b6c9e807ef..66a888a9ad6 100644 --- a/fs/xfs/xfs_mount.c +++ b/fs/xfs/xfs_mount.c @@ -583,8 +583,8 @@ xfs_readsb(xfs_mount_t *mp, int flags) sector_size = xfs_getsize_buftarg(mp->m_ddev_targp); extra_flags = XFS_BUF_LOCK | XFS_BUF_MANAGE | XFS_BUF_MAPPED; - bp = xfs_buf_read_flags(mp->m_ddev_targp, XFS_SB_DADDR, - BTOBB(sector_size), extra_flags); + bp = xfs_buf_read(mp->m_ddev_targp, XFS_SB_DADDR, BTOBB(sector_size), + extra_flags); if (!bp || XFS_BUF_ISERROR(bp)) { xfs_fs_mount_cmn_err(flags, "SB read failed"); error = bp ? XFS_BUF_GETERROR(bp) : ENOMEM; @@ -624,8 +624,8 @@ xfs_readsb(xfs_mount_t *mp, int flags) XFS_BUF_UNMANAGE(bp); xfs_buf_relse(bp); sector_size = mp->m_sb.sb_sectsize; - bp = xfs_buf_read_flags(mp->m_ddev_targp, XFS_SB_DADDR, - BTOBB(sector_size), extra_flags); + bp = xfs_buf_read(mp->m_ddev_targp, XFS_SB_DADDR, + BTOBB(sector_size), extra_flags); if (!bp || XFS_BUF_ISERROR(bp)) { xfs_fs_mount_cmn_err(flags, "SB re-read failed"); error = bp ? XFS_BUF_GETERROR(bp) : ENOMEM; @@ -1471,7 +1471,7 @@ xfs_log_sbcount( if (!xfs_sb_version_haslazysbcount(&mp->m_sb)) return 0; - tp = _xfs_trans_alloc(mp, XFS_TRANS_SB_COUNT); + tp = _xfs_trans_alloc(mp, XFS_TRANS_SB_COUNT, KM_SLEEP); error = xfs_trans_reserve(tp, 0, mp->m_sb.sb_sectsize + 128, 0, 0, XFS_DEFAULT_LOG_COUNT); if (error) { @@ -2123,7 +2123,7 @@ xfs_icsb_destroy_counters( mutex_destroy(&mp->m_icsb_mutex); } -STATIC_INLINE void +STATIC void xfs_icsb_lock_cntr( xfs_icsb_cnts_t *icsbp) { @@ -2132,7 +2132,7 @@ xfs_icsb_lock_cntr( } } -STATIC_INLINE void +STATIC void xfs_icsb_unlock_cntr( xfs_icsb_cnts_t *icsbp) { @@ -2140,7 +2140,7 @@ xfs_icsb_unlock_cntr( } -STATIC_INLINE void +STATIC void xfs_icsb_lock_all_counters( xfs_mount_t *mp) { @@ -2153,7 +2153,7 @@ xfs_icsb_lock_all_counters( } } -STATIC_INLINE void +STATIC void xfs_icsb_unlock_all_counters( xfs_mount_t *mp) { diff --git a/fs/xfs/xfs_mount.h b/fs/xfs/xfs_mount.h index a6c023bc0fb..1df7e450296 100644 --- a/fs/xfs/xfs_mount.h +++ b/fs/xfs/xfs_mount.h @@ -93,6 +93,9 @@ typedef struct xfs_dmops { xfs_send_unmount_t xfs_send_unmount; } xfs_dmops_t; +#define XFS_DMAPI_UNMOUNT_FLAGS(mp) \ + (((mp)->m_dmevmask & (1 << DM_EVENT_UNMOUNT)) ? 0 : DM_FLAGS_UNWANTED) + #define XFS_SEND_DATA(mp, ev,ip,off,len,fl,lock) \ (*(mp)->m_dm_ops->xfs_send_data)(ev,ip,off,len,fl,lock) #define XFS_SEND_MMAP(mp, vma,fl) \ @@ -101,12 +104,24 @@ typedef struct xfs_dmops { (*(mp)->m_dm_ops->xfs_send_destroy)(ip,right) #define XFS_SEND_NAMESP(mp, ev,b1,r1,b2,r2,n1,n2,mode,rval,fl) \ (*(mp)->m_dm_ops->xfs_send_namesp)(ev,NULL,b1,r1,b2,r2,n1,n2,mode,rval,fl) -#define XFS_SEND_PREUNMOUNT(mp,b1,r1,b2,r2,n1,n2,mode,rval,fl) \ - (*(mp)->m_dm_ops->xfs_send_namesp)(DM_EVENT_PREUNMOUNT,mp,b1,r1,b2,r2,n1,n2,mode,rval,fl) #define XFS_SEND_MOUNT(mp,right,path,name) \ (*(mp)->m_dm_ops->xfs_send_mount)(mp,right,path,name) -#define XFS_SEND_UNMOUNT(mp, ip,right,mode,rval,fl) \ - (*(mp)->m_dm_ops->xfs_send_unmount)(mp,ip,right,mode,rval,fl) +#define XFS_SEND_PREUNMOUNT(mp) \ +do { \ + if (mp->m_flags & XFS_MOUNT_DMAPI) { \ + (*(mp)->m_dm_ops->xfs_send_namesp)(DM_EVENT_PREUNMOUNT, mp, \ + (mp)->m_rootip, DM_RIGHT_NULL, \ + (mp)->m_rootip, DM_RIGHT_NULL, \ + NULL, NULL, 0, 0, XFS_DMAPI_UNMOUNT_FLAGS(mp)); \ + } \ +} while (0) +#define XFS_SEND_UNMOUNT(mp) \ +do { \ + if (mp->m_flags & XFS_MOUNT_DMAPI) { \ + (*(mp)->m_dm_ops->xfs_send_unmount)(mp, (mp)->m_rootip, \ + DM_RIGHT_NULL, 0, 0, XFS_DMAPI_UNMOUNT_FLAGS(mp)); \ + } \ +} while (0) #ifdef HAVE_PERCPU_SB @@ -387,13 +402,13 @@ xfs_put_perag(struct xfs_mount *mp, xfs_perag_t *pag) * Per-cpu superblock locking functions */ #ifdef HAVE_PERCPU_SB -STATIC_INLINE void +static inline void xfs_icsb_lock(xfs_mount_t *mp) { mutex_lock(&mp->m_icsb_mutex); } -STATIC_INLINE void +static inline void xfs_icsb_unlock(xfs_mount_t *mp) { mutex_unlock(&mp->m_icsb_mutex); diff --git a/fs/xfs/xfs_rw.c b/fs/xfs/xfs_rw.c index 3f816ad7ff1..4c199d18f85 100644 --- a/fs/xfs/xfs_rw.c +++ b/fs/xfs/xfs_rw.c @@ -277,10 +277,10 @@ xfs_read_buf( xfs_buf_t *bp; int error; - if (flags) - bp = xfs_buf_read_flags(target, blkno, len, flags); - else - bp = xfs_buf_read(target, blkno, len, flags); + if (!flags) + flags = XBF_LOCK | XBF_MAPPED; + + bp = xfs_buf_read(target, blkno, len, flags); if (!bp) return XFS_ERROR(EIO); error = XFS_BUF_GETERROR(bp); @@ -336,3 +336,25 @@ xfs_bwrite( } return (error); } + +/* + * helper function to extract extent size hint from inode + */ +xfs_extlen_t +xfs_get_extsz_hint( + struct xfs_inode *ip) +{ + xfs_extlen_t extsz; + + if (unlikely(XFS_IS_REALTIME_INODE(ip))) { + extsz = (ip->i_d.di_flags & XFS_DIFLAG_EXTSIZE) + ? ip->i_d.di_extsize + : ip->i_mount->m_sb.sb_rextsize; + ASSERT(extsz); + } else { + extsz = (ip->i_d.di_flags & XFS_DIFLAG_EXTSIZE) + ? ip->i_d.di_extsize : 0; + } + + return extsz; +} diff --git a/fs/xfs/xfs_rw.h b/fs/xfs/xfs_rw.h index f5e4874c37d..571f2174435 100644 --- a/fs/xfs/xfs_rw.h +++ b/fs/xfs/xfs_rw.h @@ -37,34 +37,6 @@ xfs_fsb_to_db(struct xfs_inode *ip, xfs_fsblock_t fsb) } /* - * Flags for xfs_free_eofblocks - */ -#define XFS_FREE_EOF_LOCK (1<<0) -#define XFS_FREE_EOF_NOLOCK (1<<1) - - -/* - * helper function to extract extent size hint from inode - */ -STATIC_INLINE xfs_extlen_t -xfs_get_extsz_hint( - xfs_inode_t *ip) -{ - xfs_extlen_t extsz; - - if (unlikely(XFS_IS_REALTIME_INODE(ip))) { - extsz = (ip->i_d.di_flags & XFS_DIFLAG_EXTSIZE) - ? ip->i_d.di_extsize - : ip->i_mount->m_sb.sb_rextsize; - ASSERT(extsz); - } else { - extsz = (ip->i_d.di_flags & XFS_DIFLAG_EXTSIZE) - ? ip->i_d.di_extsize : 0; - } - return extsz; -} - -/* * Prototypes for functions in xfs_rw.c. */ extern int xfs_write_clear_setuid(struct xfs_inode *ip); @@ -76,5 +48,6 @@ extern int xfs_read_buf(struct xfs_mount *mp, xfs_buftarg_t *btp, struct xfs_buf **bpp); extern void xfs_ioerror_alert(char *func, struct xfs_mount *mp, xfs_buf_t *bp, xfs_daddr_t blkno); +extern xfs_extlen_t xfs_get_extsz_hint(struct xfs_inode *ip); #endif /* __XFS_RW_H__ */ diff --git a/fs/xfs/xfs_trans.c b/fs/xfs/xfs_trans.c index 66b849358e6..237badcbac3 100644 --- a/fs/xfs/xfs_trans.c +++ b/fs/xfs/xfs_trans.c @@ -236,19 +236,20 @@ xfs_trans_alloc( uint type) { xfs_wait_for_freeze(mp, SB_FREEZE_TRANS); - return _xfs_trans_alloc(mp, type); + return _xfs_trans_alloc(mp, type, KM_SLEEP); } xfs_trans_t * _xfs_trans_alloc( xfs_mount_t *mp, - uint type) + uint type, + uint memflags) { xfs_trans_t *tp; atomic_inc(&mp->m_active_trans); - tp = kmem_zone_zalloc(xfs_trans_zone, KM_SLEEP); + tp = kmem_zone_zalloc(xfs_trans_zone, memflags); tp->t_magic = XFS_TRANS_MAGIC; tp->t_type = type; tp->t_mountp = mp; diff --git a/fs/xfs/xfs_trans.h b/fs/xfs/xfs_trans.h index ed47fc77759..a0574f593f5 100644 --- a/fs/xfs/xfs_trans.h +++ b/fs/xfs/xfs_trans.h @@ -924,7 +924,7 @@ typedef struct xfs_trans { * XFS transaction mechanism exported interfaces. */ xfs_trans_t *xfs_trans_alloc(struct xfs_mount *, uint); -xfs_trans_t *_xfs_trans_alloc(struct xfs_mount *, uint); +xfs_trans_t *_xfs_trans_alloc(struct xfs_mount *, uint, uint); xfs_trans_t *xfs_trans_dup(xfs_trans_t *); int xfs_trans_reserve(xfs_trans_t *, uint, uint, uint, uint, uint); diff --git a/fs/xfs/xfs_trans_buf.c b/fs/xfs/xfs_trans_buf.c index 218829e6a15..03a1f701fea 100644 --- a/fs/xfs/xfs_trans_buf.c +++ b/fs/xfs/xfs_trans_buf.c @@ -79,11 +79,8 @@ xfs_trans_get_buf(xfs_trans_t *tp, /* * Default to a normal get_buf() call if the tp is NULL. */ - if (tp == NULL) { - bp = xfs_buf_get_flags(target_dev, blkno, len, - flags | BUF_BUSY); - return(bp); - } + if (tp == NULL) + return xfs_buf_get(target_dev, blkno, len, flags | BUF_BUSY); /* * If we find the buffer in the cache with this transaction @@ -129,7 +126,7 @@ xfs_trans_get_buf(xfs_trans_t *tp, * easily deadlock with our current transaction as well as cause * us to run out of stack space. */ - bp = xfs_buf_get_flags(target_dev, blkno, len, flags | BUF_BUSY); + bp = xfs_buf_get(target_dev, blkno, len, flags | BUF_BUSY); if (bp == NULL) { return NULL; } @@ -302,7 +299,7 @@ xfs_trans_read_buf( * Default to a normal get_buf() call if the tp is NULL. */ if (tp == NULL) { - bp = xfs_buf_read_flags(target, blkno, len, flags | BUF_BUSY); + bp = xfs_buf_read(target, blkno, len, flags | BUF_BUSY); if (!bp) return (flags & XFS_BUF_TRYLOCK) ? EAGAIN : XFS_ERROR(ENOMEM); @@ -398,7 +395,7 @@ xfs_trans_read_buf( * easily deadlock with our current transaction as well as cause * us to run out of stack space. */ - bp = xfs_buf_read_flags(target, blkno, len, flags | BUF_BUSY); + bp = xfs_buf_read(target, blkno, len, flags | BUF_BUSY); if (bp == NULL) { *bpp = NULL; return 0; diff --git a/fs/xfs/xfs_vnodeops.c b/fs/xfs/xfs_vnodeops.c index b572f7e840e..578f3f59b78 100644 --- a/fs/xfs/xfs_vnodeops.c +++ b/fs/xfs/xfs_vnodeops.c @@ -538,9 +538,8 @@ xfs_readlink_bmap( d = XFS_FSB_TO_DADDR(mp, mval[n].br_startblock); byte_cnt = XFS_FSB_TO_B(mp, mval[n].br_blockcount); - bp = xfs_buf_read_flags(mp->m_ddev_targp, d, BTOBB(byte_cnt), - XBF_LOCK | XBF_MAPPED | - XBF_DONT_BLOCK); + bp = xfs_buf_read(mp->m_ddev_targp, d, BTOBB(byte_cnt), + XBF_LOCK | XBF_MAPPED | XBF_DONT_BLOCK); error = XFS_BUF_GETERROR(bp); if (error) { xfs_ioerror_alert("xfs_readlink", @@ -709,6 +708,11 @@ xfs_fsync( } /* + * Flags for xfs_free_eofblocks + */ +#define XFS_FREE_EOF_TRYLOCK (1<<0) + +/* * This is called by xfs_inactive to free any blocks beyond eof * when the link count isn't zero and by xfs_dm_punch_hole() when * punching a hole to EOF. @@ -726,7 +730,6 @@ xfs_free_eofblocks( xfs_filblks_t map_len; int nimaps; xfs_bmbt_irec_t imap; - int use_iolock = (flags & XFS_FREE_EOF_LOCK); /* * Figure out if there are any blocks beyond the end @@ -768,14 +771,19 @@ xfs_free_eofblocks( * cache and we can't * do that within a transaction. */ - if (use_iolock) + if (flags & XFS_FREE_EOF_TRYLOCK) { + if (!xfs_ilock_nowait(ip, XFS_IOLOCK_EXCL)) { + xfs_trans_cancel(tp, 0); + return 0; + } + } else { xfs_ilock(ip, XFS_IOLOCK_EXCL); + } error = xfs_itruncate_start(ip, XFS_ITRUNC_DEFINITE, ip->i_size); if (error) { xfs_trans_cancel(tp, 0); - if (use_iolock) - xfs_iunlock(ip, XFS_IOLOCK_EXCL); + xfs_iunlock(ip, XFS_IOLOCK_EXCL); return error; } @@ -812,8 +820,7 @@ xfs_free_eofblocks( error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES); } - xfs_iunlock(ip, (use_iolock ? (XFS_IOLOCK_EXCL|XFS_ILOCK_EXCL) - : XFS_ILOCK_EXCL)); + xfs_iunlock(ip, XFS_IOLOCK_EXCL|XFS_ILOCK_EXCL); } return error; } @@ -1113,7 +1120,17 @@ xfs_release( (ip->i_df.if_flags & XFS_IFEXTENTS)) && (!(ip->i_d.di_flags & (XFS_DIFLAG_PREALLOC | XFS_DIFLAG_APPEND)))) { - error = xfs_free_eofblocks(mp, ip, XFS_FREE_EOF_LOCK); + + /* + * If we can't get the iolock just skip truncating + * the blocks past EOF because we could deadlock + * with the mmap_sem otherwise. We'll get another + * chance to drop them once the last reference to + * the inode is dropped, so we'll never leak blocks + * permanently. + */ + error = xfs_free_eofblocks(mp, ip, + XFS_FREE_EOF_TRYLOCK); if (error) return error; } @@ -1184,7 +1201,7 @@ xfs_inactive( (!(ip->i_d.di_flags & (XFS_DIFLAG_PREALLOC | XFS_DIFLAG_APPEND)) || (ip->i_delayed_blks != 0)))) { - error = xfs_free_eofblocks(mp, ip, XFS_FREE_EOF_LOCK); + error = xfs_free_eofblocks(mp, ip, 0); if (error) return VN_INACTIVE_CACHE; } @@ -2456,46 +2473,6 @@ xfs_set_dmattrs( return error; } -int -xfs_reclaim( - xfs_inode_t *ip) -{ - - xfs_itrace_entry(ip); - - ASSERT(!VN_MAPPED(VFS_I(ip))); - - /* bad inode, get out here ASAP */ - if (is_bad_inode(VFS_I(ip))) { - xfs_ireclaim(ip); - return 0; - } - - xfs_ioend_wait(ip); - - ASSERT(XFS_FORCED_SHUTDOWN(ip->i_mount) || ip->i_delayed_blks == 0); - - /* - * If we have nothing to flush with this inode then complete the - * teardown now, otherwise break the link between the xfs inode and the - * linux inode and clean up the xfs inode later. This avoids flushing - * the inode to disk during the delete operation itself. - * - * When breaking the link, we need to set the XFS_IRECLAIMABLE flag - * first to ensure that xfs_iunpin() will never see an xfs inode - * that has a linux inode being reclaimed. Synchronisation is provided - * by the i_flags_lock. - */ - if (!ip->i_update_core && (ip->i_itemp == NULL)) { - xfs_ilock(ip, XFS_ILOCK_EXCL); - xfs_iflock(ip); - xfs_iflags_set(ip, XFS_IRECLAIMABLE); - return xfs_reclaim_inode(ip, 1, XFS_IFLUSH_DELWRI_ELSE_SYNC); - } - xfs_inode_set_reclaim_tag(ip); - return 0; -} - /* * xfs_alloc_file_space() * This routine allocates disk space for the given file. diff --git a/fs/xfs/xfs_vnodeops.h b/fs/xfs/xfs_vnodeops.h index a9e102de71a..167a467403a 100644 --- a/fs/xfs/xfs_vnodeops.h +++ b/fs/xfs/xfs_vnodeops.h @@ -38,7 +38,6 @@ int xfs_symlink(struct xfs_inode *dp, struct xfs_name *link_name, const char *target_path, mode_t mode, struct xfs_inode **ipp, cred_t *credp); int xfs_set_dmattrs(struct xfs_inode *ip, u_int evmask, u_int16_t state); -int xfs_reclaim(struct xfs_inode *ip); int xfs_change_file_space(struct xfs_inode *ip, int cmd, xfs_flock64_t *bf, xfs_off_t offset, int attr_flags); int xfs_rename(struct xfs_inode *src_dp, struct xfs_name *src_name, |