From 8de52778798fe39660a8d6b26f290e0c93202761 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Mon, 6 Feb 2012 12:45:27 -0500 Subject: [PATCH 01/64] vfs: check i_nlink limits in vfs_{mkdir,rename_dir,link} New field of struct super_block - ->s_max_links. Maximal allowed value of ->i_nlink or 0; in the latter case all checks still need to be done in ->link/->mkdir/->rename instances. Note that this limit applies both to directoris and to non-directories. Signed-off-by: Al Viro --- fs/exofs/namei.c | 13 +------------ fs/exofs/super.c | 1 + fs/ext2/namei.c | 13 +------------ fs/ext2/super.c | 1 + fs/jfs/namei.c | 13 ------------- fs/jfs/super.c | 1 + fs/logfs/dir.c | 3 --- fs/logfs/super.c | 1 + fs/minix/inode.c | 10 +++++----- fs/minix/minix.h | 1 - fs/minix/namei.c | 14 +------------- fs/namei.c | 13 +++++++++++++ fs/nilfs2/namei.c | 11 ----------- fs/nilfs2/super.c | 1 + fs/sysv/namei.c | 12 +----------- fs/sysv/super.c | 24 ++++++++++++------------ fs/sysv/sysv.h | 1 - fs/udf/namei.c | 13 ------------- fs/udf/super.c | 3 +++ fs/ufs/namei.c | 14 +------------- fs/ufs/super.c | 1 + fs/xfs/xfs_rename.c | 11 ----------- fs/xfs/xfs_super.c | 1 + fs/xfs/xfs_utils.c | 2 -- fs/xfs/xfs_vnodeops.c | 16 ---------------- include/linux/fs.h | 1 + 26 files changed, 46 insertions(+), 149 deletions(-) diff --git a/fs/exofs/namei.c b/fs/exofs/namei.c index 9dbf0c301030..fc7161d6bf6b 100644 --- a/fs/exofs/namei.c +++ b/fs/exofs/namei.c @@ -143,9 +143,6 @@ static int exofs_link(struct dentry *old_dentry, struct inode *dir, { struct inode *inode = old_dentry->d_inode; - if (inode->i_nlink >= EXOFS_LINK_MAX) - return -EMLINK; - inode->i_ctime = CURRENT_TIME; inode_inc_link_count(inode); ihold(inode); @@ -156,10 +153,7 @@ static int exofs_link(struct dentry *old_dentry, struct inode *dir, static int exofs_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode) { struct inode *inode; - int err = -EMLINK; - - if (dir->i_nlink >= EXOFS_LINK_MAX) - goto out; + int err; inode_inc_link_count(dir); @@ -275,11 +269,6 @@ static int exofs_rename(struct inode *old_dir, struct dentry *old_dentry, if (err) goto out_dir; } else { - if (dir_de) { - err = -EMLINK; - if (new_dir->i_nlink >= EXOFS_LINK_MAX) - goto out_dir; - } err = exofs_add_link(new_dentry, old_inode); if (err) goto out_dir; diff --git a/fs/exofs/super.c b/fs/exofs/super.c index d22cd168c6ee..6cafcadfc3c8 100644 --- a/fs/exofs/super.c +++ b/fs/exofs/super.c @@ -754,6 +754,7 @@ static int exofs_fill_super(struct super_block *sb, void *data, int silent) sb->s_blocksize = EXOFS_BLKSIZE; sb->s_blocksize_bits = EXOFS_BLKSHIFT; sb->s_maxbytes = MAX_LFS_FILESIZE; + sb->s_max_links = EXOFS_LINK_MAX; atomic_set(&sbi->s_curr_pending, 0); sb->s_bdev = NULL; sb->s_dev = 0; diff --git a/fs/ext2/namei.c b/fs/ext2/namei.c index 080419814bae..dffb86536285 100644 --- a/fs/ext2/namei.c +++ b/fs/ext2/namei.c @@ -195,9 +195,6 @@ static int ext2_link (struct dentry * old_dentry, struct inode * dir, struct inode *inode = old_dentry->d_inode; int err; - if (inode->i_nlink >= EXT2_LINK_MAX) - return -EMLINK; - dquot_initialize(dir); inode->i_ctime = CURRENT_TIME_SEC; @@ -217,10 +214,7 @@ static int ext2_link (struct dentry * old_dentry, struct inode * dir, static int ext2_mkdir(struct inode * dir, struct dentry * dentry, umode_t mode) { struct inode * inode; - int err = -EMLINK; - - if (dir->i_nlink >= EXT2_LINK_MAX) - goto out; + int err; dquot_initialize(dir); @@ -346,11 +340,6 @@ static int ext2_rename (struct inode * old_dir, struct dentry * old_dentry, drop_nlink(new_inode); inode_dec_link_count(new_inode); } else { - if (dir_de) { - err = -EMLINK; - if (new_dir->i_nlink >= EXT2_LINK_MAX) - goto out_dir; - } err = ext2_add_link(new_dentry, old_inode); if (err) goto out_dir; diff --git a/fs/ext2/super.c b/fs/ext2/super.c index 0090595beb28..9f6766a3ac1e 100644 --- a/fs/ext2/super.c +++ b/fs/ext2/super.c @@ -919,6 +919,7 @@ static int ext2_fill_super(struct super_block *sb, void *data, int silent) } sb->s_maxbytes = ext2_max_size(sb->s_blocksize_bits); + sb->s_max_links = EXT2_LINK_MAX; if (le32_to_cpu(es->s_rev_level) == EXT2_GOOD_OLD_REV) { sbi->s_inode_size = EXT2_GOOD_OLD_INODE_SIZE; diff --git a/fs/jfs/namei.c b/fs/jfs/namei.c index 5f7c160ea64f..07c91ca6017d 100644 --- a/fs/jfs/namei.c +++ b/fs/jfs/namei.c @@ -220,12 +220,6 @@ static int jfs_mkdir(struct inode *dip, struct dentry *dentry, umode_t mode) dquot_initialize(dip); - /* link count overflow on parent directory ? */ - if (dip->i_nlink == JFS_LINK_MAX) { - rc = -EMLINK; - goto out1; - } - /* * search parent directory for entry/freespace * (dtSearch() returns parent directory page pinned) @@ -806,9 +800,6 @@ static int jfs_link(struct dentry *old_dentry, jfs_info("jfs_link: %s %s", old_dentry->d_name.name, dentry->d_name.name); - if (ip->i_nlink == JFS_LINK_MAX) - return -EMLINK; - dquot_initialize(dir); tid = txBegin(ip->i_sb, 0); @@ -1138,10 +1129,6 @@ static int jfs_rename(struct inode *old_dir, struct dentry *old_dentry, rc = -ENOTEMPTY; goto out3; } - } else if ((new_dir != old_dir) && - (new_dir->i_nlink == JFS_LINK_MAX)) { - rc = -EMLINK; - goto out3; } } else if (new_ip) { IWRITE_LOCK(new_ip, RDWRLOCK_NORMAL); diff --git a/fs/jfs/super.c b/fs/jfs/super.c index 682bca642f38..4661ad705130 100644 --- a/fs/jfs/super.c +++ b/fs/jfs/super.c @@ -441,6 +441,7 @@ static int jfs_fill_super(struct super_block *sb, void *data, int silent) return -ENOMEM; sb->s_fs_info = sbi; + sb->s_max_links = JFS_LINK_MAX; sbi->sb = sb; sbi->uid = sbi->gid = sbi->umask = -1; diff --git a/fs/logfs/dir.c b/fs/logfs/dir.c index 3de7a32cadbe..4aea231fc9e6 100644 --- a/fs/logfs/dir.c +++ b/fs/logfs/dir.c @@ -558,9 +558,6 @@ static int logfs_link(struct dentry *old_dentry, struct inode *dir, { struct inode *inode = old_dentry->d_inode; - if (inode->i_nlink >= LOGFS_LINK_MAX) - return -EMLINK; - inode->i_ctime = dir->i_ctime = dir->i_mtime = CURRENT_TIME; ihold(inode); inc_nlink(inode); diff --git a/fs/logfs/super.c b/fs/logfs/super.c index c9ee7f5d1caf..b1a491a5fe78 100644 --- a/fs/logfs/super.c +++ b/fs/logfs/super.c @@ -542,6 +542,7 @@ static struct dentry *logfs_get_sb_device(struct logfs_super *super, * the filesystem incompatible with 32bit systems. */ sb->s_maxbytes = (1ull << 43) - 1; + sb->s_max_links = LOGFS_LINK_MAX; sb->s_op = &logfs_super_operations; sb->s_flags = flags | MS_NOATIME; diff --git a/fs/minix/inode.c b/fs/minix/inode.c index fa8b612b8ce2..62c697caffb9 100644 --- a/fs/minix/inode.c +++ b/fs/minix/inode.c @@ -190,24 +190,24 @@ static int minix_fill_super(struct super_block *s, void *data, int silent) sbi->s_version = MINIX_V1; sbi->s_dirsize = 16; sbi->s_namelen = 14; - sbi->s_link_max = MINIX_LINK_MAX; + s->s_max_links = MINIX_LINK_MAX; } else if (s->s_magic == MINIX_SUPER_MAGIC2) { sbi->s_version = MINIX_V1; sbi->s_dirsize = 32; sbi->s_namelen = 30; - sbi->s_link_max = MINIX_LINK_MAX; + s->s_max_links = MINIX_LINK_MAX; } else if (s->s_magic == MINIX2_SUPER_MAGIC) { sbi->s_version = MINIX_V2; sbi->s_nzones = ms->s_zones; sbi->s_dirsize = 16; sbi->s_namelen = 14; - sbi->s_link_max = MINIX2_LINK_MAX; + s->s_max_links = MINIX2_LINK_MAX; } else if (s->s_magic == MINIX2_SUPER_MAGIC2) { sbi->s_version = MINIX_V2; sbi->s_nzones = ms->s_zones; sbi->s_dirsize = 32; sbi->s_namelen = 30; - sbi->s_link_max = MINIX2_LINK_MAX; + s->s_max_links = MINIX2_LINK_MAX; } else if ( *(__u16 *)(bh->b_data + 24) == MINIX3_SUPER_MAGIC) { m3s = (struct minix3_super_block *) bh->b_data; s->s_magic = m3s->s_magic; @@ -221,9 +221,9 @@ static int minix_fill_super(struct super_block *s, void *data, int silent) sbi->s_dirsize = 64; sbi->s_namelen = 60; sbi->s_version = MINIX_V3; - sbi->s_link_max = MINIX2_LINK_MAX; sbi->s_mount_state = MINIX_VALID_FS; sb_set_blocksize(s, m3s->s_blocksize); + s->s_max_links = MINIX2_LINK_MAX; } else goto out_no_fs; diff --git a/fs/minix/minix.h b/fs/minix/minix.h index c889ef0aa571..1ebd11854622 100644 --- a/fs/minix/minix.h +++ b/fs/minix/minix.h @@ -34,7 +34,6 @@ struct minix_sb_info { unsigned long s_max_size; int s_dirsize; int s_namelen; - int s_link_max; struct buffer_head ** s_imap; struct buffer_head ** s_zmap; struct buffer_head * s_sbh; diff --git a/fs/minix/namei.c b/fs/minix/namei.c index 2f76e38c2065..2d0ee1786305 100644 --- a/fs/minix/namei.c +++ b/fs/minix/namei.c @@ -94,9 +94,6 @@ static int minix_link(struct dentry * old_dentry, struct inode * dir, { struct inode *inode = old_dentry->d_inode; - if (inode->i_nlink >= minix_sb(inode->i_sb)->s_link_max) - return -EMLINK; - inode->i_ctime = CURRENT_TIME_SEC; inode_inc_link_count(inode); ihold(inode); @@ -106,10 +103,7 @@ static int minix_link(struct dentry * old_dentry, struct inode * dir, static int minix_mkdir(struct inode * dir, struct dentry *dentry, umode_t mode) { struct inode * inode; - int err = -EMLINK; - - if (dir->i_nlink >= minix_sb(dir->i_sb)->s_link_max) - goto out; + int err; inode_inc_link_count(dir); @@ -181,7 +175,6 @@ static int minix_rmdir(struct inode * dir, struct dentry *dentry) static int minix_rename(struct inode * old_dir, struct dentry *old_dentry, struct inode * new_dir, struct dentry *new_dentry) { - struct minix_sb_info * info = minix_sb(old_dir->i_sb); struct inode * old_inode = old_dentry->d_inode; struct inode * new_inode = new_dentry->d_inode; struct page * dir_page = NULL; @@ -219,11 +212,6 @@ static int minix_rename(struct inode * old_dir, struct dentry *old_dentry, drop_nlink(new_inode); inode_dec_link_count(new_inode); } else { - if (dir_de) { - err = -EMLINK; - if (new_dir->i_nlink >= info->s_link_max) - goto out_dir; - } err = minix_add_link(new_dentry, old_inode); if (err) goto out_dir; diff --git a/fs/namei.c b/fs/namei.c index 46ea9cc16647..a0b82762e8fc 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -2569,6 +2569,7 @@ SYSCALL_DEFINE3(mknod, const char __user *, filename, umode_t, mode, unsigned, d int vfs_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode) { int error = may_create(dir, dentry); + unsigned max_links = dir->i_sb->s_max_links; if (error) return error; @@ -2581,6 +2582,9 @@ int vfs_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode) if (error) return error; + if (max_links && dir->i_nlink >= max_links) + return -EMLINK; + error = dir->i_op->mkdir(dir, dentry, mode); if (!error) fsnotify_mkdir(dir, dentry); @@ -2911,6 +2915,7 @@ SYSCALL_DEFINE2(symlink, const char __user *, oldname, const char __user *, newn int vfs_link(struct dentry *old_dentry, struct inode *dir, struct dentry *new_dentry) { struct inode *inode = old_dentry->d_inode; + unsigned max_links = dir->i_sb->s_max_links; int error; if (!inode) @@ -2941,6 +2946,8 @@ int vfs_link(struct dentry *old_dentry, struct inode *dir, struct dentry *new_de /* Make sure we don't allow creating hardlink to an unlinked file */ if (inode->i_nlink == 0) error = -ENOENT; + else if (max_links && inode->i_nlink >= max_links) + error = -EMLINK; else error = dir->i_op->link(old_dentry, dir, new_dentry); mutex_unlock(&inode->i_mutex); @@ -3050,6 +3057,7 @@ static int vfs_rename_dir(struct inode *old_dir, struct dentry *old_dentry, { int error = 0; struct inode *target = new_dentry->d_inode; + unsigned max_links = new_dir->i_sb->s_max_links; /* * If we are going to change the parent - check write permissions, @@ -3073,6 +3081,11 @@ static int vfs_rename_dir(struct inode *old_dir, struct dentry *old_dentry, if (d_mountpoint(old_dentry) || d_mountpoint(new_dentry)) goto out; + error = -EMLINK; + if (max_links && !target && new_dir != old_dir && + new_dir->i_nlink >= max_links) + goto out; + if (target) shrink_dcache_parent(new_dentry); error = old_dir->i_op->rename(old_dir, old_dentry, new_dir, new_dentry); diff --git a/fs/nilfs2/namei.c b/fs/nilfs2/namei.c index 1cd3f624dffc..fce2bbee66d4 100644 --- a/fs/nilfs2/namei.c +++ b/fs/nilfs2/namei.c @@ -193,9 +193,6 @@ static int nilfs_link(struct dentry *old_dentry, struct inode *dir, struct nilfs_transaction_info ti; int err; - if (inode->i_nlink >= NILFS_LINK_MAX) - return -EMLINK; - err = nilfs_transaction_begin(dir->i_sb, &ti, 1); if (err) return err; @@ -219,9 +216,6 @@ static int nilfs_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode) struct nilfs_transaction_info ti; int err; - if (dir->i_nlink >= NILFS_LINK_MAX) - return -EMLINK; - err = nilfs_transaction_begin(dir->i_sb, &ti, 1); if (err) return err; @@ -400,11 +394,6 @@ static int nilfs_rename(struct inode *old_dir, struct dentry *old_dentry, 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; - } err = nilfs_add_link(new_dentry, old_inode); if (err) goto out_dir; diff --git a/fs/nilfs2/super.c b/fs/nilfs2/super.c index 08e3d4f9df18..1fc9ad3c1d14 100644 --- a/fs/nilfs2/super.c +++ b/fs/nilfs2/super.c @@ -1059,6 +1059,7 @@ nilfs_fill_super(struct super_block *sb, void *data, int silent) sb->s_export_op = &nilfs_export_ops; sb->s_root = NULL; sb->s_time_gran = 1; + sb->s_max_links = NILFS_LINK_MAX; bdi = sb->s_bdev->bd_inode->i_mapping->backing_dev_info; sb->s_bdi = bdi ? : &default_backing_dev_info; diff --git a/fs/sysv/namei.c b/fs/sysv/namei.c index b217797e621b..d7466e293614 100644 --- a/fs/sysv/namei.c +++ b/fs/sysv/namei.c @@ -121,9 +121,6 @@ static int sysv_link(struct dentry * old_dentry, struct inode * dir, { struct inode *inode = old_dentry->d_inode; - if (inode->i_nlink >= SYSV_SB(inode->i_sb)->s_link_max) - return -EMLINK; - inode->i_ctime = CURRENT_TIME_SEC; inode_inc_link_count(inode); ihold(inode); @@ -134,10 +131,8 @@ static int sysv_link(struct dentry * old_dentry, struct inode * dir, static int sysv_mkdir(struct inode * dir, struct dentry *dentry, umode_t mode) { struct inode * inode; - int err = -EMLINK; + int err; - if (dir->i_nlink >= SYSV_SB(dir->i_sb)->s_link_max) - goto out; inode_inc_link_count(dir); inode = sysv_new_inode(dir, S_IFDIR|mode); @@ -251,11 +246,6 @@ static int sysv_rename(struct inode * old_dir, struct dentry * old_dentry, drop_nlink(new_inode); inode_dec_link_count(new_inode); } else { - if (dir_de) { - err = -EMLINK; - if (new_dir->i_nlink >= SYSV_SB(new_dir->i_sb)->s_link_max) - goto out_dir; - } err = sysv_add_link(new_dentry, old_inode); if (err) goto out_dir; diff --git a/fs/sysv/super.c b/fs/sysv/super.c index f60c196913ea..f467740e088c 100644 --- a/fs/sysv/super.c +++ b/fs/sysv/super.c @@ -44,7 +44,7 @@ enum { JAN_1_1980 = (10*365 + 2) * 24 * 60 * 60 }; -static void detected_xenix(struct sysv_sb_info *sbi) +static void detected_xenix(struct sysv_sb_info *sbi, unsigned *max_links) { struct buffer_head *bh1 = sbi->s_bh1; struct buffer_head *bh2 = sbi->s_bh2; @@ -59,7 +59,7 @@ static void detected_xenix(struct sysv_sb_info *sbi) sbd2 = (struct xenix_super_block *) (bh2->b_data - 512); } - sbi->s_link_max = XENIX_LINK_MAX; + *max_links = XENIX_LINK_MAX; sbi->s_fic_size = XENIX_NICINOD; sbi->s_flc_size = XENIX_NICFREE; sbi->s_sbd1 = (char *)sbd1; @@ -75,7 +75,7 @@ static void detected_xenix(struct sysv_sb_info *sbi) sbi->s_nzones = fs32_to_cpu(sbi, sbd1->s_fsize); } -static void detected_sysv4(struct sysv_sb_info *sbi) +static void detected_sysv4(struct sysv_sb_info *sbi, unsigned *max_links) { struct sysv4_super_block * sbd; struct buffer_head *bh1 = sbi->s_bh1; @@ -86,7 +86,7 @@ static void detected_sysv4(struct sysv_sb_info *sbi) else sbd = (struct sysv4_super_block *) bh2->b_data; - sbi->s_link_max = SYSV_LINK_MAX; + *max_links = SYSV_LINK_MAX; sbi->s_fic_size = SYSV_NICINOD; sbi->s_flc_size = SYSV_NICFREE; sbi->s_sbd1 = (char *)sbd; @@ -103,7 +103,7 @@ static void detected_sysv4(struct sysv_sb_info *sbi) sbi->s_nzones = fs32_to_cpu(sbi, sbd->s_fsize); } -static void detected_sysv2(struct sysv_sb_info *sbi) +static void detected_sysv2(struct sysv_sb_info *sbi, unsigned *max_links) { struct sysv2_super_block *sbd; struct buffer_head *bh1 = sbi->s_bh1; @@ -114,7 +114,7 @@ static void detected_sysv2(struct sysv_sb_info *sbi) else sbd = (struct sysv2_super_block *) bh2->b_data; - sbi->s_link_max = SYSV_LINK_MAX; + *max_links = SYSV_LINK_MAX; sbi->s_fic_size = SYSV_NICINOD; sbi->s_flc_size = SYSV_NICFREE; sbi->s_sbd1 = (char *)sbd; @@ -131,14 +131,14 @@ static void detected_sysv2(struct sysv_sb_info *sbi) sbi->s_nzones = fs32_to_cpu(sbi, sbd->s_fsize); } -static void detected_coherent(struct sysv_sb_info *sbi) +static void detected_coherent(struct sysv_sb_info *sbi, unsigned *max_links) { struct coh_super_block * sbd; struct buffer_head *bh1 = sbi->s_bh1; sbd = (struct coh_super_block *) bh1->b_data; - sbi->s_link_max = COH_LINK_MAX; + *max_links = COH_LINK_MAX; sbi->s_fic_size = COH_NICINOD; sbi->s_flc_size = COH_NICFREE; sbi->s_sbd1 = (char *)sbd; @@ -154,12 +154,12 @@ static void detected_coherent(struct sysv_sb_info *sbi) sbi->s_nzones = fs32_to_cpu(sbi, sbd->s_fsize); } -static void detected_v7(struct sysv_sb_info *sbi) +static void detected_v7(struct sysv_sb_info *sbi, unsigned *max_links) { struct buffer_head *bh2 = sbi->s_bh2; struct v7_super_block *sbd = (struct v7_super_block *)bh2->b_data; - sbi->s_link_max = V7_LINK_MAX; + *max_links = V7_LINK_MAX; sbi->s_fic_size = V7_NICINOD; sbi->s_flc_size = V7_NICFREE; sbi->s_sbd1 = (char *)sbd; @@ -290,7 +290,7 @@ static char *flavour_names[] = { [FSTYPE_AFS] = "AFS", }; -static void (*flavour_setup[])(struct sysv_sb_info *) = { +static void (*flavour_setup[])(struct sysv_sb_info *, unsigned *) = { [FSTYPE_XENIX] = detected_xenix, [FSTYPE_SYSV4] = detected_sysv4, [FSTYPE_SYSV2] = detected_sysv2, @@ -310,7 +310,7 @@ static int complete_read_super(struct super_block *sb, int silent, int size) sbi->s_firstinodezone = 2; - flavour_setup[sbi->s_type](sbi); + flavour_setup[sbi->s_type](sbi, &sb->s_max_links); sbi->s_truncate = 1; sbi->s_ndatazones = sbi->s_nzones - sbi->s_firstdatazone; diff --git a/fs/sysv/sysv.h b/fs/sysv/sysv.h index 0e4b821c5691..11b07672f6c5 100644 --- a/fs/sysv/sysv.h +++ b/fs/sysv/sysv.h @@ -24,7 +24,6 @@ struct sysv_sb_info { char s_bytesex; /* bytesex (le/be/pdp) */ char s_truncate; /* if 1: names > SYSV_NAMELEN chars are truncated */ /* if 0: they are disallowed (ENAMETOOLONG) */ - nlink_t s_link_max; /* max number of hard links to a file */ unsigned int s_inodes_per_block; /* number of inodes per block */ unsigned int s_inodes_per_block_1; /* inodes_per_block - 1 */ unsigned int s_inodes_per_block_bits; /* log2(inodes_per_block) */ diff --git a/fs/udf/namei.c b/fs/udf/namei.c index 08bf46edf9c4..38de8f234b94 100644 --- a/fs/udf/namei.c +++ b/fs/udf/namei.c @@ -32,8 +32,6 @@ #include #include -enum { UDF_MAX_LINKS = 0xffff }; - static inline int udf_match(int len1, const unsigned char *name1, int len2, const unsigned char *name2) { @@ -649,10 +647,6 @@ static int udf_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode) struct udf_inode_info *dinfo = UDF_I(dir); struct udf_inode_info *iinfo; - err = -EMLINK; - if (dir->i_nlink >= UDF_MAX_LINKS) - goto out; - err = -EIO; inode = udf_new_inode(dir, S_IFDIR | mode, &err); if (!inode) @@ -1032,9 +1026,6 @@ static int udf_link(struct dentry *old_dentry, struct inode *dir, struct fileIdentDesc cfi, *fi; int err; - if (inode->i_nlink >= UDF_MAX_LINKS) - return -EMLINK; - fi = udf_add_entry(dir, dentry, &fibh, &cfi, &err); if (!fi) { return err; @@ -1126,10 +1117,6 @@ static int udf_rename(struct inode *old_dir, struct dentry *old_dentry, if (udf_get_lb_pblock(old_inode->i_sb, &tloc, 0) != old_dir->i_ino) goto end_rename; - - retval = -EMLINK; - if (!new_inode && new_dir->i_nlink >= UDF_MAX_LINKS) - goto end_rename; } if (!nfi) { nfi = udf_add_entry(new_dir, new_dentry, &nfibh, &ncfi, diff --git a/fs/udf/super.c b/fs/udf/super.c index c09a84daaf50..8d8b25336fbb 100644 --- a/fs/udf/super.c +++ b/fs/udf/super.c @@ -75,6 +75,8 @@ #define UDF_DEFAULT_BLOCKSIZE 2048 +enum { UDF_MAX_LINKS = 0xffff }; + /* These are the "meat" - everything else is stuffing */ static int udf_fill_super(struct super_block *, void *, int); static void udf_put_super(struct super_block *); @@ -2042,6 +2044,7 @@ static int udf_fill_super(struct super_block *sb, void *options, int silent) goto error_out; } sb->s_maxbytes = MAX_LFS_FILESIZE; + sb->s_max_links = UDF_MAX_LINKS; return 0; error_out: diff --git a/fs/ufs/namei.c b/fs/ufs/namei.c index 38cac199edff..a2281cadefa1 100644 --- a/fs/ufs/namei.c +++ b/fs/ufs/namei.c @@ -166,10 +166,6 @@ static int ufs_link (struct dentry * old_dentry, struct inode * dir, int error; lock_ufs(dir->i_sb); - if (inode->i_nlink >= UFS_LINK_MAX) { - unlock_ufs(dir->i_sb); - return -EMLINK; - } inode->i_ctime = CURRENT_TIME_SEC; inode_inc_link_count(inode); @@ -183,10 +179,7 @@ static int ufs_link (struct dentry * old_dentry, struct inode * dir, static int ufs_mkdir(struct inode * dir, struct dentry * dentry, umode_t mode) { struct inode * inode; - int err = -EMLINK; - - if (dir->i_nlink >= UFS_LINK_MAX) - goto out; + int err; lock_ufs(dir->i_sb); inode_inc_link_count(dir); @@ -305,11 +298,6 @@ static int ufs_rename(struct inode *old_dir, struct dentry *old_dentry, drop_nlink(new_inode); inode_dec_link_count(new_inode); } else { - if (dir_de) { - err = -EMLINK; - if (new_dir->i_nlink >= UFS_LINK_MAX) - goto out_dir; - } err = ufs_add_link(new_dentry, old_inode); if (err) goto out_dir; diff --git a/fs/ufs/super.c b/fs/ufs/super.c index 5246ee3e5607..ec25d09fcaa8 100644 --- a/fs/ufs/super.c +++ b/fs/ufs/super.c @@ -1157,6 +1157,7 @@ magic_found: "fast symlink size (%u)\n", uspi->s_maxsymlinklen); uspi->s_maxsymlinklen = maxsymlen; } + sb->s_max_links = UFS_LINK_MAX; inode = ufs_iget(sb, UFS_ROOTINO); if (IS_ERR(inode)) { diff --git a/fs/xfs/xfs_rename.c b/fs/xfs/xfs_rename.c index 866de277079a..e44ef7ee8ce8 100644 --- a/fs/xfs/xfs_rename.c +++ b/fs/xfs/xfs_rename.c @@ -118,17 +118,6 @@ xfs_rename( new_parent = (src_dp != target_dp); src_is_directory = S_ISDIR(src_ip->i_d.di_mode); - if (src_is_directory) { - /* - * Check for link count overflow on target_dp - */ - if (target_ip == NULL && new_parent && - target_dp->i_d.di_nlink >= XFS_MAXLINK) { - error = XFS_ERROR(EMLINK); - goto std_return; - } - } - xfs_sort_for_rename(src_dp, target_dp, src_ip, target_ip, inodes, &num_inodes); diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c index ee5b695c99a7..0e4c5c017fba 100644 --- a/fs/xfs/xfs_super.c +++ b/fs/xfs/xfs_super.c @@ -1341,6 +1341,7 @@ xfs_fs_fill_super( sb->s_blocksize = mp->m_sb.sb_blocksize; sb->s_blocksize_bits = ffs(sb->s_blocksize) - 1; sb->s_maxbytes = xfs_max_file_offset(sb->s_blocksize_bits); + sb->s_max_links = XFS_MAXLINK; sb->s_time_gran = 1; set_posix_acl_flag(sb); diff --git a/fs/xfs/xfs_utils.c b/fs/xfs/xfs_utils.c index 89dbb4a50872..79c05ac85bfe 100644 --- a/fs/xfs/xfs_utils.c +++ b/fs/xfs/xfs_utils.c @@ -296,8 +296,6 @@ xfs_bumplink( xfs_trans_t *tp, xfs_inode_t *ip) { - if (ip->i_d.di_nlink >= XFS_MAXLINK) - return XFS_ERROR(EMLINK); xfs_trans_ichgtime(tp, ip, XFS_ICHGTIME_CHG); ASSERT(ip->i_d.di_nlink > 0); diff --git a/fs/xfs/xfs_vnodeops.c b/fs/xfs/xfs_vnodeops.c index ebdb88840a47..64981d7e7375 100644 --- a/fs/xfs/xfs_vnodeops.c +++ b/fs/xfs/xfs_vnodeops.c @@ -917,14 +917,6 @@ xfs_create( xfs_ilock(dp, XFS_ILOCK_EXCL | XFS_ILOCK_PARENT); unlock_dp_on_error = B_TRUE; - /* - * Check for directory link count overflow. - */ - if (is_dir && dp->i_d.di_nlink >= XFS_MAXLINK) { - error = XFS_ERROR(EMLINK); - goto out_trans_cancel; - } - xfs_bmap_init(&free_list, &first_block); /* @@ -1428,14 +1420,6 @@ xfs_link( xfs_trans_ijoin(tp, sip, XFS_ILOCK_EXCL); xfs_trans_ijoin(tp, tdp, XFS_ILOCK_EXCL); - /* - * If the source has too many links, we can't make any more to it. - */ - if (sip->i_d.di_nlink >= XFS_MAXLINK) { - error = XFS_ERROR(EMLINK); - goto error_return; - } - /* * If we are using project inheritance, we only allow hard link * creation in our tree when the project IDs are the same; else diff --git a/include/linux/fs.h b/include/linux/fs.h index 69cd5bb640f5..0ab89426ba55 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -1459,6 +1459,7 @@ struct super_block { u8 s_uuid[16]; /* UUID */ void *s_fs_info; /* Filesystem private info */ + unsigned int s_max_links; fmode_t s_mode; /* Granularity of c/m/atime in ns. From b57ce9694ec43dcb6ef6f189d6540e4b3d2c5e7a Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sun, 12 Feb 2012 02:38:16 -0500 Subject: [PATCH 02/64] vfs: drop_file_write_access() made static Signed-off-by: Al Viro --- fs/file_table.c | 3 +-- include/linux/file.h | 1 - 2 files changed, 1 insertion(+), 3 deletions(-) diff --git a/fs/file_table.c b/fs/file_table.c index 20002e39754d..70f2a0fd6aec 100644 --- a/fs/file_table.c +++ b/fs/file_table.c @@ -204,7 +204,7 @@ EXPORT_SYMBOL(alloc_file); * to write to @file, along with access to write through * its vfsmount. */ -void drop_file_write_access(struct file *file) +static void drop_file_write_access(struct file *file) { struct vfsmount *mnt = file->f_path.mnt; struct dentry *dentry = file->f_path.dentry; @@ -219,7 +219,6 @@ void drop_file_write_access(struct file *file) mnt_drop_write(mnt); file_release_write(file); } -EXPORT_SYMBOL_GPL(drop_file_write_access); /* the real guts of fput() - releasing the last reference to file */ diff --git a/include/linux/file.h b/include/linux/file.h index 21a79958541c..58bf158c53d9 100644 --- a/include/linux/file.h +++ b/include/linux/file.h @@ -12,7 +12,6 @@ struct file; extern void fput(struct file *); -extern void drop_file_write_access(struct file *file); struct file_operations; struct vfsmount; From e28e832c3e1e1197873cfd0b6ce86868cf5c391d Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sun, 12 Feb 2012 02:58:14 -0500 Subject: [PATCH 03/64] ecryptfs: don't bother with ->drop_inode() generic_drop_inode() is the default Signed-off-by: Al Viro --- fs/ecryptfs/super.c | 1 - 1 file changed, 1 deletion(-) diff --git a/fs/ecryptfs/super.c b/fs/ecryptfs/super.c index cf152823bbf4..2dd946b636d2 100644 --- a/fs/ecryptfs/super.c +++ b/fs/ecryptfs/super.c @@ -184,7 +184,6 @@ static int ecryptfs_show_options(struct seq_file *m, struct dentry *root) const struct super_operations ecryptfs_sops = { .alloc_inode = ecryptfs_alloc_inode, .destroy_inode = ecryptfs_destroy_inode, - .drop_inode = generic_drop_inode, .statfs = ecryptfs_statfs, .remount_fs = NULL, .evict_inode = ecryptfs_evict_inode, From 9bcb4b733c22b7dbc4cf847e707ac98f751e9180 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sun, 12 Feb 2012 19:43:17 -0500 Subject: [PATCH 04/64] vfs: turn generic_drop_inode() into static inline Once upon a time it used to be much bigger, but these days there's no point whatsoever keeping it in fs/inode.c, especially since it's not even needed as initializer for ->drop_inode() - it's the default and leaving ->drop_inode NULL will do just as well. Signed-off-by: Al Viro --- fs/inode.c | 11 ----------- include/linux/fs.h | 5 ++++- 2 files changed, 4 insertions(+), 12 deletions(-) diff --git a/fs/inode.c b/fs/inode.c index 83ab215baab1..92de04b0baa2 100644 --- a/fs/inode.c +++ b/fs/inode.c @@ -1368,17 +1368,6 @@ int generic_delete_inode(struct inode *inode) } EXPORT_SYMBOL(generic_delete_inode); -/* - * Normal UNIX filesystem behaviour: delete the - * inode when the usage count drops to zero, and - * i_nlink is zero. - */ -int generic_drop_inode(struct inode *inode) -{ - return !inode->i_nlink || inode_unhashed(inode); -} -EXPORT_SYMBOL_GPL(generic_drop_inode); - /* * Called when we're dropping the last reference * to an inode. diff --git a/include/linux/fs.h b/include/linux/fs.h index 0ab89426ba55..b89dc4d60d99 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -2305,7 +2305,10 @@ extern struct inode * igrab(struct inode *); extern ino_t iunique(struct super_block *, ino_t); extern int inode_needs_sync(struct inode *inode); extern int generic_delete_inode(struct inode *inode); -extern int generic_drop_inode(struct inode *inode); +static inline int generic_drop_inode(struct inode *inode) +{ + return !inode->i_nlink || inode_unhashed(inode); +} extern struct inode *ilookup5_nowait(struct super_block *sb, unsigned long hashval, int (*test)(struct inode *, void *), From 064326c0773af8a0e8bb82d895cceaedc8a51b9e Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sun, 12 Feb 2012 21:06:12 -0500 Subject: [PATCH 05/64] clean up the failure exits in cifs_read_super() no need to make that iput() conditional, just take it to the right place... Signed-off-by: Al Viro --- fs/cifs/cifsfs.c | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-) diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c index b1fd382d1952..8b7d7ff88792 100644 --- a/fs/cifs/cifsfs.c +++ b/fs/cifs/cifsfs.c @@ -119,7 +119,6 @@ cifs_read_super(struct super_block *sb) if (IS_ERR(inode)) { rc = PTR_ERR(inode); - inode = NULL; goto out_no_root; } @@ -127,6 +126,7 @@ cifs_read_super(struct super_block *sb) if (!sb->s_root) { rc = -ENOMEM; + iput(inode); goto out_no_root; } @@ -147,9 +147,6 @@ cifs_read_super(struct super_block *sb) out_no_root: cERROR(1, "cifs_read_super: get root inode failed"); - if (inode) - iput(inode); - return rc; } From f56b0fbc6477e50303a503ee1453ed94e20f154a Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sun, 12 Feb 2012 21:15:58 -0500 Subject: [PATCH 06/64] coda: clean failure exits in coda_fill_super() same as for cifs, move iput() to the right place, make it unconditional Signed-off-by: Al Viro --- fs/coda/inode.c | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-) diff --git a/fs/coda/inode.c b/fs/coda/inode.c index 5e2e1b3f068d..32dafc875c14 100644 --- a/fs/coda/inode.c +++ b/fs/coda/inode.c @@ -208,7 +208,6 @@ static int coda_fill_super(struct super_block *sb, void *data, int silent) if (IS_ERR(root)) { error = PTR_ERR(root); printk("Failure of coda_cnode_make for root: error %d\n", error); - root = NULL; goto error; } @@ -216,15 +215,13 @@ static int coda_fill_super(struct super_block *sb, void *data, int silent) root->i_ino, root->i_sb->s_id); sb->s_root = d_alloc_root(root); if (!sb->s_root) { + iput(root); error = -EINVAL; goto error; } return 0; error: - if (root) - iput(root); - mutex_lock(&vc->vc_mutex); bdi_destroy(&vc->bdi); vc->vc_sb = NULL; From be0d93f0aa5682a24a2a9ec0dd26fffaad608cce Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sun, 12 Feb 2012 21:46:49 -0500 Subject: [PATCH 07/64] ... and the same failure exits cleanup for ocfs2 Signed-off-by: Al Viro --- fs/ocfs2/super.c | 10 ++++------ 1 file changed, 4 insertions(+), 6 deletions(-) diff --git a/fs/ocfs2/super.c b/fs/ocfs2/super.c index 604e12c4e979..2b1184f7097f 100644 --- a/fs/ocfs2/super.c +++ b/fs/ocfs2/super.c @@ -1154,12 +1154,12 @@ static int ocfs2_fill_super(struct super_block *sb, void *data, int silent) } status = ocfs2_mount_volume(sb); - if (osb->root_inode) - inode = igrab(osb->root_inode); - if (status < 0) goto read_super_error; + if (osb->root_inode) + inode = igrab(osb->root_inode); + if (!inode) { status = -EIO; mlog_errno(status); @@ -1168,6 +1168,7 @@ static int ocfs2_fill_super(struct super_block *sb, void *data, int silent) root = d_alloc_root(inode); if (!root) { + iput(inode); status = -ENOMEM; mlog_errno(status); goto read_super_error; @@ -1220,9 +1221,6 @@ static int ocfs2_fill_super(struct super_block *sb, void *data, int silent) read_super_error: brelse(bh); - if (inode) - iput(inode); - if (osb) { atomic_set(&osb->vol_state, VOLUME_DISABLED); wake_up(&osb->osb_mount_event); From 6b4231e2f92adbcf96fb2a3fa751d7ca0a61b21f Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sun, 12 Feb 2012 21:56:08 -0500 Subject: [PATCH 08/64] procfs: clean proc_fill_super() up First of all, there's no need to zero ->i_uid/->i_gid on root inode - both had been set to zero already. Moreover, let's take the iput() on failure to the failure exit it belongs to... Signed-off-by: Al Viro --- fs/proc/inode.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) diff --git a/fs/proc/inode.c b/fs/proc/inode.c index 84fd3235a590..a70af3a44f45 100644 --- a/fs/proc/inode.c +++ b/fs/proc/inode.c @@ -499,16 +499,15 @@ int proc_fill_super(struct super_block *s) root_inode = proc_get_inode(s, &proc_root); if (!root_inode) goto out_no_root; - root_inode->i_uid = 0; - root_inode->i_gid = 0; s->s_root = d_alloc_root(root_inode); - if (!s->s_root) + if (!s->s_root) { + iput(root_inode); goto out_no_root; + } return 0; out_no_root: printk("proc_read_super: get root inode failed\n"); - iput(root_inode); pde_put(&proc_root); return -ENOMEM; } From 48fde701aff662559b38d9a609574068f22d00fe Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sun, 8 Jan 2012 22:15:13 -0500 Subject: [PATCH 09/64] switch open-coded instances of d_make_root() to new helper Signed-off-by: Al Viro --- arch/powerpc/platforms/cell/spufs/inode.c | 4 ++-- arch/s390/hypfs/inode.c | 6 ++---- drivers/misc/ibmasm/ibmasmfs.c | 6 ++---- drivers/oprofile/oprofilefs.c | 6 ++---- drivers/usb/core/inode.c | 9 +-------- drivers/usb/gadget/f_fs.c | 8 ++------ drivers/usb/gadget/inode.c | 4 +--- fs/9p/vfs_super.c | 3 +-- fs/adfs/super.c | 3 +-- fs/affs/super.c | 7 ++----- fs/afs/super.c | 7 ++----- fs/autofs4/inode.c | 10 ++-------- fs/befs/linuxvfs.c | 3 +-- fs/bfs/inode.c | 3 +-- fs/btrfs/super.c | 8 ++------ fs/ceph/super.c | 3 +-- fs/cifs/cifsfs.c | 4 +--- fs/coda/inode.c | 3 +-- fs/configfs/mount.c | 3 +-- fs/cramfs/inode.c | 6 ++---- fs/devpts/inode.c | 3 +-- fs/ecryptfs/main.c | 3 +-- fs/efs/super.c | 3 +-- fs/exofs/super.c | 3 +-- fs/ext2/super.c | 3 +-- fs/ext3/super.c | 3 +-- fs/ext4/super.c | 3 +-- fs/freevxfs/vxfs_super.c | 3 +-- fs/fuse/inode.c | 9 ++------- fs/gfs2/ops_fstype.c | 3 +-- fs/hfs/super.c | 6 ++---- fs/hostfs/hostfs_kern.c | 4 ++-- fs/hpfs/super.c | 6 ++---- fs/hppfs/hppfs.c | 9 ++------- fs/hugetlbfs/inode.c | 13 ++---------- fs/isofs/inode.c | 3 +-- fs/jffs2/fs.c | 6 ++---- fs/jfs/super.c | 3 +-- fs/libfs.c | 6 ++---- fs/logfs/super.c | 6 ++---- fs/ncpfs/inode.c | 6 ++---- fs/nfs/getroot.c | 6 ++---- fs/nilfs2/super.c | 3 +-- fs/ocfs2/dlmfs/dlmfs.c | 14 ++----------- fs/ocfs2/super.c | 3 +-- fs/omfs/inode.c | 6 ++---- fs/openpromfs/inode.c | 3 +-- fs/proc/inode.c | 15 +++----------- fs/pstore/inode.c | 3 +-- fs/qnx4/inode.c | 6 ++---- fs/ramfs/inode.c | 12 ++---------- fs/reiserfs/super.c | 6 ++---- fs/romfs/super.c | 6 ++---- fs/squashfs/super.c | 3 +-- fs/sysfs/mount.c | 3 +-- fs/sysv/super.c | 3 +-- fs/ubifs/super.c | 6 ++---- fs/udf/super.c | 3 +-- fs/ufs/super.c | 6 ++---- fs/xfs/xfs_super.c | 6 ++---- ipc/mqueue.c | 24 +++++++---------------- kernel/cgroup.c | 8 ++------ mm/shmem.c | 6 ++---- net/sunrpc/rpc_pipe.c | 8 ++------ 64 files changed, 105 insertions(+), 264 deletions(-) diff --git a/arch/powerpc/platforms/cell/spufs/inode.c b/arch/powerpc/platforms/cell/spufs/inode.c index d4a094ca96f3..17b3211e3641 100644 --- a/arch/powerpc/platforms/cell/spufs/inode.c +++ b/arch/powerpc/platforms/cell/spufs/inode.c @@ -757,9 +757,9 @@ spufs_create_root(struct super_block *sb, void *data) goto out_iput; ret = -ENOMEM; - sb->s_root = d_alloc_root(inode); + sb->s_root = d_make_root(inode); if (!sb->s_root) - goto out_iput; + goto out; return 0; out_iput: diff --git a/arch/s390/hypfs/inode.c b/arch/s390/hypfs/inode.c index 8a2a887478cc..6a2cb560e968 100644 --- a/arch/s390/hypfs/inode.c +++ b/arch/s390/hypfs/inode.c @@ -293,11 +293,9 @@ static int hypfs_fill_super(struct super_block *sb, void *data, int silent) return -ENOMEM; root_inode->i_op = &simple_dir_inode_operations; root_inode->i_fop = &simple_dir_operations; - sb->s_root = root_dentry = d_alloc_root(root_inode); - if (!root_dentry) { - iput(root_inode); + sb->s_root = root_dentry = d_make_root(root_inode); + if (!root_dentry) return -ENOMEM; - } if (MACHINE_IS_VM) rc = hypfs_vm_create_files(sb, root_dentry); else diff --git a/drivers/misc/ibmasm/ibmasmfs.c b/drivers/misc/ibmasm/ibmasmfs.c index 35361753b487..15f24f362208 100644 --- a/drivers/misc/ibmasm/ibmasmfs.c +++ b/drivers/misc/ibmasm/ibmasmfs.c @@ -129,11 +129,9 @@ static int ibmasmfs_fill_super (struct super_block *sb, void *data, int silent) root->i_op = &simple_dir_inode_operations; root->i_fop = ibmasmfs_dir_ops; - root_dentry = d_alloc_root(root); - if (!root_dentry) { - iput(root); + root_dentry = d_make_root(root); + if (!root_dentry) return -ENOMEM; - } sb->s_root = root_dentry; ibmasmfs_create_files(sb, root_dentry); diff --git a/drivers/oprofile/oprofilefs.c b/drivers/oprofile/oprofilefs.c index 2f0aa0f700e6..277bb70b8d75 100644 --- a/drivers/oprofile/oprofilefs.c +++ b/drivers/oprofile/oprofilefs.c @@ -251,11 +251,9 @@ static int oprofilefs_fill_super(struct super_block *sb, void *data, int silent) return -ENOMEM; root_inode->i_op = &simple_dir_inode_operations; root_inode->i_fop = &simple_dir_operations; - root_dentry = d_alloc_root(root_inode); - if (!root_dentry) { - iput(root_inode); + root_dentry = d_make_root(root_inode); + if (!root_dentry) return -ENOMEM; - } sb->s_root = root_dentry; diff --git a/drivers/usb/core/inode.c b/drivers/usb/core/inode.c index 9e186f3da839..bdaef8e36020 100644 --- a/drivers/usb/core/inode.c +++ b/drivers/usb/core/inode.c @@ -462,16 +462,9 @@ static int usbfs_fill_super(struct super_block *sb, void *data, int silent) sb->s_op = &usbfs_ops; sb->s_time_gran = 1; inode = usbfs_get_inode(sb, S_IFDIR | 0755, 0); - - if (!inode) { - dbg("%s: could not get inode!",__func__); - return -ENOMEM; - } - - root = d_alloc_root(inode); + root = d_make_root(inode); if (!root) { dbg("%s: could not get root dentry!",__func__); - iput(inode); return -ENOMEM; } sb->s_root = root; diff --git a/drivers/usb/gadget/f_fs.c b/drivers/usb/gadget/f_fs.c index f63dc6c150d2..d825b248728a 100644 --- a/drivers/usb/gadget/f_fs.c +++ b/drivers/usb/gadget/f_fs.c @@ -1063,13 +1063,9 @@ static int ffs_sb_fill(struct super_block *sb, void *_data, int silent) &simple_dir_operations, &simple_dir_inode_operations, &data->perms); - if (unlikely(!inode)) + sb->s_root = d_make_root(inode); + if (unlikely(!sb->s_root)) goto Enomem; - sb->s_root = d_alloc_root(inode); - if (unlikely(!sb->s_root)) { - iput(inode); - goto Enomem; - } /* EP0 file */ if (unlikely(!ffs_sb_create_file(sb, "ep0", ffs, diff --git a/drivers/usb/gadget/inode.c b/drivers/usb/gadget/inode.c index ae04266dba1b..c95eea43b637 100644 --- a/drivers/usb/gadget/inode.c +++ b/drivers/usb/gadget/inode.c @@ -2059,10 +2059,8 @@ gadgetfs_fill_super (struct super_block *sb, void *opts, int silent) if (!inode) goto Enomem; inode->i_op = &simple_dir_inode_operations; - if (!(sb->s_root = d_alloc_root (inode))) { - iput(inode); + if (!(sb->s_root = d_make_root (inode))) goto Enomem; - } /* the ep0 file is named after the controller we expect; * user mode code can use it for sanity checks, like we do. diff --git a/fs/9p/vfs_super.c b/fs/9p/vfs_super.c index 7b0cd87b07c2..10b7d3c9dba8 100644 --- a/fs/9p/vfs_super.c +++ b/fs/9p/vfs_super.c @@ -155,9 +155,8 @@ static struct dentry *v9fs_mount(struct file_system_type *fs_type, int flags, goto release_sb; } - root = d_alloc_root(inode); + root = d_make_root(inode); if (!root) { - iput(inode); retval = -ENOMEM; goto release_sb; } diff --git a/fs/adfs/super.c b/fs/adfs/super.c index 8e3b36ace305..06fdcc9382c4 100644 --- a/fs/adfs/super.c +++ b/fs/adfs/super.c @@ -483,10 +483,9 @@ static int adfs_fill_super(struct super_block *sb, void *data, int silent) sb->s_d_op = &adfs_dentry_operations; root = adfs_iget(sb, &root_obj); - sb->s_root = d_alloc_root(root); + sb->s_root = d_make_root(root); if (!sb->s_root) { int i; - iput(root); for (i = 0; i < asb->s_map_size; i++) brelse(asb->s_map[i].dm_bh); kfree(asb->s_map); diff --git a/fs/affs/super.c b/fs/affs/super.c index 8ba73fed7964..0782653a05a2 100644 --- a/fs/affs/super.c +++ b/fs/affs/super.c @@ -473,7 +473,7 @@ got_root: root_inode = affs_iget(sb, root_block); if (IS_ERR(root_inode)) { ret = PTR_ERR(root_inode); - goto out_error_noinode; + goto out_error; } if (AFFS_SB(sb)->s_flags & SF_INTL) @@ -481,7 +481,7 @@ got_root: else sb->s_d_op = &affs_dentry_operations; - sb->s_root = d_alloc_root(root_inode); + sb->s_root = d_make_root(root_inode); if (!sb->s_root) { printk(KERN_ERR "AFFS: Get root inode failed\n"); goto out_error; @@ -494,9 +494,6 @@ got_root: * Begin the cascaded cleanup ... */ out_error: - if (root_inode) - iput(root_inode); -out_error_noinode: kfree(sbi->s_bitmap); affs_brelse(root_bh); kfree(sbi->s_prefix); diff --git a/fs/afs/super.c b/fs/afs/super.c index 983ec59fc80d..f02b31e7e648 100644 --- a/fs/afs/super.c +++ b/fs/afs/super.c @@ -301,7 +301,6 @@ static int afs_fill_super(struct super_block *sb, { struct afs_super_info *as = sb->s_fs_info; struct afs_fid fid; - struct dentry *root = NULL; struct inode *inode = NULL; int ret; @@ -327,18 +326,16 @@ static int afs_fill_super(struct super_block *sb, set_bit(AFS_VNODE_AUTOCELL, &AFS_FS_I(inode)->flags); ret = -ENOMEM; - root = d_alloc_root(inode); - if (!root) + sb->s_root = d_make_root(inode); + if (!sb->s_root) goto error; sb->s_d_op = &afs_fs_dentry_operations; - sb->s_root = root; _leave(" = 0"); return 0; error: - iput(inode); _leave(" = %d", ret); return ret; } diff --git a/fs/autofs4/inode.c b/fs/autofs4/inode.c index 06858d955120..d8dc002e9cc3 100644 --- a/fs/autofs4/inode.c +++ b/fs/autofs4/inode.c @@ -247,12 +247,9 @@ int autofs4_fill_super(struct super_block *s, void *data, int silent) if (!ino) goto fail_free; root_inode = autofs4_get_inode(s, S_IFDIR | 0755); - if (!root_inode) - goto fail_ino; - - root = d_alloc_root(root_inode); + root = d_make_root(root_inode); if (!root) - goto fail_iput; + goto fail_ino; pipe = NULL; root->d_fsdata = ino; @@ -317,9 +314,6 @@ fail_fput: fail_dput: dput(root); goto fail_free; -fail_iput: - printk("autofs: get root dentry failed\n"); - iput(root_inode); fail_ino: kfree(ino); fail_free: diff --git a/fs/befs/linuxvfs.c b/fs/befs/linuxvfs.c index 6e6d536767fe..e18da23d42b5 100644 --- a/fs/befs/linuxvfs.c +++ b/fs/befs/linuxvfs.c @@ -852,9 +852,8 @@ befs_fill_super(struct super_block *sb, void *data, int silent) ret = PTR_ERR(root); goto unacquire_priv_sbp; } - sb->s_root = d_alloc_root(root); + sb->s_root = d_make_root(root); if (!sb->s_root) { - iput(root); befs_error(sb, "get root inode failed"); goto unacquire_priv_sbp; } diff --git a/fs/bfs/inode.c b/fs/bfs/inode.c index b0391bc402b1..e23dc7c8b884 100644 --- a/fs/bfs/inode.c +++ b/fs/bfs/inode.c @@ -367,9 +367,8 @@ static int bfs_fill_super(struct super_block *s, void *data, int silent) ret = PTR_ERR(inode); goto out2; } - s->s_root = d_alloc_root(inode); + s->s_root = d_make_root(inode); if (!s->s_root) { - iput(inode); ret = -ENOMEM; goto out2; } diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index 3ce97b217cbe..81df3fec6a6d 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -629,7 +629,6 @@ static int btrfs_fill_super(struct super_block *sb, void *data, int silent) { struct inode *inode; - struct dentry *root_dentry; struct btrfs_fs_info *fs_info = btrfs_sb(sb); struct btrfs_key key; int err; @@ -660,15 +659,12 @@ static int btrfs_fill_super(struct super_block *sb, goto fail_close; } - root_dentry = d_alloc_root(inode); - if (!root_dentry) { - iput(inode); + sb->s_root = d_make_root(inode); + if (!sb->s_root) { err = -ENOMEM; goto fail_close; } - sb->s_root = root_dentry; - save_mount_options(sb, data); cleancache_init_fs(sb); sb->s_flags |= MS_ACTIVE; diff --git a/fs/ceph/super.c b/fs/ceph/super.c index 00de2c9568cd..256f85221926 100644 --- a/fs/ceph/super.c +++ b/fs/ceph/super.c @@ -655,9 +655,8 @@ static struct dentry *open_root_dentry(struct ceph_fs_client *fsc, dout("open_root_inode success\n"); if (ceph_ino(inode) == CEPH_INO_ROOT && fsc->sb->s_root == NULL) { - root = d_alloc_root(inode); + root = d_make_root(inode); if (!root) { - iput(inode); root = ERR_PTR(-ENOMEM); goto out; } diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c index 8b7d7ff88792..418fc42fb8b2 100644 --- a/fs/cifs/cifsfs.c +++ b/fs/cifs/cifsfs.c @@ -122,11 +122,9 @@ cifs_read_super(struct super_block *sb) goto out_no_root; } - sb->s_root = d_alloc_root(inode); - + sb->s_root = d_make_root(inode); if (!sb->s_root) { rc = -ENOMEM; - iput(inode); goto out_no_root; } diff --git a/fs/coda/inode.c b/fs/coda/inode.c index 32dafc875c14..05156c17b551 100644 --- a/fs/coda/inode.c +++ b/fs/coda/inode.c @@ -213,9 +213,8 @@ static int coda_fill_super(struct super_block *sb, void *data, int silent) printk("coda_read_super: rootinode is %ld dev %s\n", root->i_ino, root->i_sb->s_id); - sb->s_root = d_alloc_root(root); + sb->s_root = d_make_root(root); if (!sb->s_root) { - iput(root); error = -EINVAL; goto error; } diff --git a/fs/configfs/mount.c b/fs/configfs/mount.c index 276e15cafd58..07f60455f1c1 100644 --- a/fs/configfs/mount.c +++ b/fs/configfs/mount.c @@ -91,10 +91,9 @@ static int configfs_fill_super(struct super_block *sb, void *data, int silent) return -ENOMEM; } - root = d_alloc_root(inode); + root = d_make_root(inode); if (!root) { pr_debug("%s: could not get root dentry!\n",__func__); - iput(inode); return -ENOMEM; } config_group_init(&configfs_root_group); diff --git a/fs/cramfs/inode.c b/fs/cramfs/inode.c index a2ee8f9f5a38..853480d2b3d1 100644 --- a/fs/cramfs/inode.c +++ b/fs/cramfs/inode.c @@ -318,11 +318,9 @@ static int cramfs_fill_super(struct super_block *sb, void *data, int silent) root = get_cramfs_inode(sb, &super.root, 0); if (IS_ERR(root)) goto out; - sb->s_root = d_alloc_root(root); - if (!sb->s_root) { - iput(root); + sb->s_root = d_make_root(root); + if (!sb->s_root) goto out; - } return 0; out: kfree(sbi); diff --git a/fs/devpts/inode.c b/fs/devpts/inode.c index c4e2a58a2e82..57dae0baedf2 100644 --- a/fs/devpts/inode.c +++ b/fs/devpts/inode.c @@ -309,12 +309,11 @@ devpts_fill_super(struct super_block *s, void *data, int silent) inode->i_fop = &simple_dir_operations; set_nlink(inode, 2); - s->s_root = d_alloc_root(inode); + s->s_root = d_make_root(inode); if (s->s_root) return 0; printk(KERN_ERR "devpts: get root dentry failed\n"); - iput(inode); fail: return -ENOMEM; diff --git a/fs/ecryptfs/main.c b/fs/ecryptfs/main.c index b4a6befb1216..6e0e017e6932 100644 --- a/fs/ecryptfs/main.c +++ b/fs/ecryptfs/main.c @@ -550,9 +550,8 @@ static struct dentry *ecryptfs_mount(struct file_system_type *fs_type, int flags if (IS_ERR(inode)) goto out_free; - s->s_root = d_alloc_root(inode); + s->s_root = d_make_root(inode); if (!s->s_root) { - iput(inode); rc = -ENOMEM; goto out_free; } diff --git a/fs/efs/super.c b/fs/efs/super.c index 981106429a9f..e755ec746c69 100644 --- a/fs/efs/super.c +++ b/fs/efs/super.c @@ -317,10 +317,9 @@ static int efs_fill_super(struct super_block *s, void *d, int silent) goto out_no_fs; } - s->s_root = d_alloc_root(root); + s->s_root = d_make_root(root); if (!(s->s_root)) { printk(KERN_ERR "EFS: get root dentry failed\n"); - iput(root); ret = -ENOMEM; goto out_no_fs; } diff --git a/fs/exofs/super.c b/fs/exofs/super.c index 6cafcadfc3c8..7f2b590a36b7 100644 --- a/fs/exofs/super.c +++ b/fs/exofs/super.c @@ -819,9 +819,8 @@ static int exofs_fill_super(struct super_block *sb, void *data, int silent) ret = PTR_ERR(root); goto free_sbi; } - sb->s_root = d_alloc_root(root); + sb->s_root = d_make_root(root); if (!sb->s_root) { - iput(root); EXOFS_ERR("ERROR: get root inode failed\n"); ret = -ENOMEM; goto free_sbi; diff --git a/fs/ext2/super.c b/fs/ext2/super.c index 9f6766a3ac1e..e1025c7a437a 100644 --- a/fs/ext2/super.c +++ b/fs/ext2/super.c @@ -1088,9 +1088,8 @@ static int ext2_fill_super(struct super_block *sb, void *data, int silent) goto failed_mount3; } - sb->s_root = d_alloc_root(root); + sb->s_root = d_make_root(root); if (!sb->s_root) { - iput(root); ext2_msg(sb, KERN_ERR, "error: get root inode failed"); ret = -ENOMEM; goto failed_mount3; diff --git a/fs/ext3/super.c b/fs/ext3/super.c index 726c7ef6cdf1..e0b45b93327b 100644 --- a/fs/ext3/super.c +++ b/fs/ext3/super.c @@ -2046,10 +2046,9 @@ static int ext3_fill_super (struct super_block *sb, void *data, int silent) ext3_msg(sb, KERN_ERR, "error: corrupt root inode, run e2fsck"); goto failed_mount3; } - sb->s_root = d_alloc_root(root); + sb->s_root = d_make_root(root); if (!sb->s_root) { ext3_msg(sb, KERN_ERR, "error: get root dentry failed"); - iput(root); ret = -ENOMEM; goto failed_mount3; } diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 502c61fd7392..d2baea7bcf30 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -3735,9 +3735,8 @@ no_journal: iput(root); goto failed_mount4; } - sb->s_root = d_alloc_root(root); + sb->s_root = d_make_root(root); if (!sb->s_root) { - iput(root); ext4_msg(sb, KERN_ERR, "get root dentry failed"); ret = -ENOMEM; goto failed_mount4; diff --git a/fs/freevxfs/vxfs_super.c b/fs/freevxfs/vxfs_super.c index 9d1c99558389..d4fabd26084e 100644 --- a/fs/freevxfs/vxfs_super.c +++ b/fs/freevxfs/vxfs_super.c @@ -224,9 +224,8 @@ static int vxfs_fill_super(struct super_block *sbp, void *dp, int silent) ret = PTR_ERR(root); goto out; } - sbp->s_root = d_alloc_root(root); + sbp->s_root = d_make_root(root); if (!sbp->s_root) { - iput(root); printk(KERN_WARNING "vxfs: unable to get root dentry.\n"); goto out_free_ilist; } diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c index 64cf8d07393e..4aec5995867e 100644 --- a/fs/fuse/inode.c +++ b/fs/fuse/inode.c @@ -988,14 +988,9 @@ static int fuse_fill_super(struct super_block *sb, void *data, int silent) err = -ENOMEM; root = fuse_get_root_inode(sb, d.rootmode); - if (!root) + root_dentry = d_make_root(root); + if (!root_dentry) goto err_put_conn; - - root_dentry = d_alloc_root(root); - if (!root_dentry) { - iput(root); - goto err_put_conn; - } /* only now - we want root dentry with NULL ->d_op */ sb->s_d_op = &fuse_dentry_operations; diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c index 24f609c9ef91..10e848c6d1b5 100644 --- a/fs/gfs2/ops_fstype.c +++ b/fs/gfs2/ops_fstype.c @@ -431,10 +431,9 @@ static int gfs2_lookup_root(struct super_block *sb, struct dentry **dptr, fs_err(sdp, "can't read in %s inode: %ld\n", name, PTR_ERR(inode)); return PTR_ERR(inode); } - dentry = d_alloc_root(inode); + dentry = d_make_root(inode); if (!dentry) { fs_err(sdp, "can't alloc %s dentry\n", name); - iput(inode); return -ENOMEM; } *dptr = dentry; diff --git a/fs/hfs/super.c b/fs/hfs/super.c index 8137fb3e6780..7b4c537d6e13 100644 --- a/fs/hfs/super.c +++ b/fs/hfs/super.c @@ -430,15 +430,13 @@ static int hfs_fill_super(struct super_block *sb, void *data, int silent) sb->s_d_op = &hfs_dentry_operations; res = -ENOMEM; - sb->s_root = d_alloc_root(root_inode); + sb->s_root = d_make_root(root_inode); if (!sb->s_root) - goto bail_iput; + goto bail_no_root; /* everything's okay */ return 0; -bail_iput: - iput(root_inode); bail_no_root: printk(KERN_ERR "hfs: get root inode failed.\n"); bail: diff --git a/fs/hostfs/hostfs_kern.c b/fs/hostfs/hostfs_kern.c index e130bd46d671..588d45885a6f 100644 --- a/fs/hostfs/hostfs_kern.c +++ b/fs/hostfs/hostfs_kern.c @@ -966,9 +966,9 @@ static int hostfs_fill_sb_common(struct super_block *sb, void *d, int silent) } err = -ENOMEM; - sb->s_root = d_alloc_root(root_inode); + sb->s_root = d_make_root(root_inode); if (sb->s_root == NULL) - goto out_put; + goto out; return 0; diff --git a/fs/hpfs/super.c b/fs/hpfs/super.c index 3690467c944e..54f6eccb79d9 100644 --- a/fs/hpfs/super.c +++ b/fs/hpfs/super.c @@ -625,11 +625,9 @@ static int hpfs_fill_super(struct super_block *s, void *options, int silent) hpfs_init_inode(root); hpfs_read_inode(root); unlock_new_inode(root); - s->s_root = d_alloc_root(root); - if (!s->s_root) { - iput(root); + s->s_root = d_make_root(root); + if (!s->s_root) goto bail0; - } /* * find the root directory's . pointer & finish filling in the inode diff --git a/fs/hppfs/hppfs.c b/fs/hppfs/hppfs.c index d92f4ce80925..a80e45a690ac 100644 --- a/fs/hppfs/hppfs.c +++ b/fs/hppfs/hppfs.c @@ -726,17 +726,12 @@ static int hppfs_fill_super(struct super_block *sb, void *d, int silent) err = -ENOMEM; root_inode = get_inode(sb, dget(proc_mnt->mnt_root)); - if (!root_inode) - goto out_mntput; - - sb->s_root = d_alloc_root(root_inode); + sb->s_root = d_make_root(root_inode); if (!sb->s_root) - goto out_iput; + goto out_mntput; return 0; - out_iput: - iput(root_inode); out_mntput: mntput(proc_mnt); out: diff --git a/fs/hugetlbfs/inode.c b/fs/hugetlbfs/inode.c index 1e85a7ac0217..81932fa1861a 100644 --- a/fs/hugetlbfs/inode.c +++ b/fs/hugetlbfs/inode.c @@ -831,8 +831,6 @@ bad_val: static int hugetlbfs_fill_super(struct super_block *sb, void *data, int silent) { - struct inode * inode; - struct dentry * root; int ret; struct hugetlbfs_config config; struct hugetlbfs_sb_info *sbinfo; @@ -865,16 +863,9 @@ hugetlbfs_fill_super(struct super_block *sb, void *data, int silent) sb->s_magic = HUGETLBFS_MAGIC; sb->s_op = &hugetlbfs_ops; sb->s_time_gran = 1; - inode = hugetlbfs_get_root(sb, &config); - if (!inode) + sb->s_root = d_make_root(hugetlbfs_get_root(sb, &config)); + if (!sb->s_root) goto out_free; - - root = d_alloc_root(inode); - if (!root) { - iput(inode); - goto out_free; - } - sb->s_root = root; return 0; out_free: kfree(sbinfo); diff --git a/fs/isofs/inode.c b/fs/isofs/inode.c index bd62c76fb5df..29037c365ba4 100644 --- a/fs/isofs/inode.c +++ b/fs/isofs/inode.c @@ -947,9 +947,8 @@ root_found: s->s_d_op = &isofs_dentry_ops[table]; /* get the root dentry */ - s->s_root = d_alloc_root(inode); + s->s_root = d_make_root(inode); if (!(s->s_root)) { - iput(inode); error = -ENOMEM; goto out_no_inode; } diff --git a/fs/jffs2/fs.c b/fs/jffs2/fs.c index 2e0123867cb1..c0d5c9d770da 100644 --- a/fs/jffs2/fs.c +++ b/fs/jffs2/fs.c @@ -561,9 +561,9 @@ int jffs2_do_fill_super(struct super_block *sb, void *data, int silent) ret = -ENOMEM; D1(printk(KERN_DEBUG "jffs2_do_fill_super(): d_alloc_root()\n")); - sb->s_root = d_alloc_root(root_i); + sb->s_root = d_make_root(root_i); if (!sb->s_root) - goto out_root_i; + goto out_root; sb->s_maxbytes = 0xFFFFFFFF; sb->s_blocksize = PAGE_CACHE_SIZE; @@ -573,8 +573,6 @@ int jffs2_do_fill_super(struct super_block *sb, void *data, int silent) jffs2_start_garbage_collect_thread(c); return 0; - out_root_i: - iput(root_i); out_root: jffs2_free_ino_caches(c); jffs2_free_raw_node_refs(c); diff --git a/fs/jfs/super.c b/fs/jfs/super.c index 4661ad705130..b3bb95504479 100644 --- a/fs/jfs/super.c +++ b/fs/jfs/super.c @@ -522,7 +522,7 @@ static int jfs_fill_super(struct super_block *sb, void *data, int silent) ret = PTR_ERR(inode); goto out_no_rw; } - sb->s_root = d_alloc_root(inode); + sb->s_root = d_make_root(inode); if (!sb->s_root) goto out_no_root; @@ -540,7 +540,6 @@ static int jfs_fill_super(struct super_block *sb, void *data, int silent) out_no_root: jfs_err("jfs_read_super: get root dentry failed"); - iput(inode); out_no_rw: rc = jfs_umount(sb); diff --git a/fs/libfs.c b/fs/libfs.c index 5b2dbb3ba4fc..7c895a763a1e 100644 --- a/fs/libfs.c +++ b/fs/libfs.c @@ -491,11 +491,9 @@ int simple_fill_super(struct super_block *s, unsigned long magic, inode->i_op = &simple_dir_inode_operations; inode->i_fop = &simple_dir_operations; set_nlink(inode, 2); - root = d_alloc_root(inode); - if (!root) { - iput(inode); + root = d_make_root(inode); + if (!root) return -ENOMEM; - } for (i = 0; !files->name || files->name[0]; i++, files++) { if (!files->name) continue; diff --git a/fs/logfs/super.c b/fs/logfs/super.c index b1a491a5fe78..7de18c3021fe 100644 --- a/fs/logfs/super.c +++ b/fs/logfs/super.c @@ -315,11 +315,9 @@ static int logfs_get_sb_final(struct super_block *sb) if (IS_ERR(rootdir)) goto fail; - sb->s_root = d_alloc_root(rootdir); - if (!sb->s_root) { - iput(rootdir); + sb->s_root = d_make_root(rootdir); + if (!sb->s_root) goto fail; - } /* at that point we know that ->put_super() will be called */ super->s_erase_page = alloc_pages(GFP_KERNEL, 0); diff --git a/fs/ncpfs/inode.c b/fs/ncpfs/inode.c index 3d1e34f8a68e..49df0e7f8379 100644 --- a/fs/ncpfs/inode.c +++ b/fs/ncpfs/inode.c @@ -716,13 +716,11 @@ static int ncp_fill_super(struct super_block *sb, void *raw_data, int silent) if (!root_inode) goto out_disconnect; DPRINTK("ncp_fill_super: root vol=%d\n", NCP_FINFO(root_inode)->volNumber); - sb->s_root = d_alloc_root(root_inode); + sb->s_root = d_make_root(root_inode); if (!sb->s_root) - goto out_no_root; + goto out_disconnect; return 0; -out_no_root: - iput(root_inode); out_disconnect: ncp_lock_server(server); ncp_disconnect(server); diff --git a/fs/nfs/getroot.c b/fs/nfs/getroot.c index dcb61548887f..801d6d830787 100644 --- a/fs/nfs/getroot.c +++ b/fs/nfs/getroot.c @@ -49,11 +49,9 @@ static int nfs_superblock_set_dummy_root(struct super_block *sb, struct inode *i { /* The mntroot acts as the dummy root dentry for this superblock */ if (sb->s_root == NULL) { - sb->s_root = d_alloc_root(inode); - if (sb->s_root == NULL) { - iput(inode); + sb->s_root = d_make_root(inode); + if (sb->s_root == NULL) return -ENOMEM; - } ihold(inode); /* * Ensure that this dentry is invisible to d_find_alias(). diff --git a/fs/nilfs2/super.c b/fs/nilfs2/super.c index 1fc9ad3c1d14..1099a76cee59 100644 --- a/fs/nilfs2/super.c +++ b/fs/nilfs2/super.c @@ -917,9 +917,8 @@ static int nilfs_get_root_dentry(struct super_block *sb, if (root->cno == NILFS_CPTREE_CURRENT_CNO) { dentry = d_find_alias(inode); if (!dentry) { - dentry = d_alloc_root(inode); + dentry = d_make_root(inode); if (!dentry) { - iput(inode); ret = -ENOMEM; goto failed_dentry; } diff --git a/fs/ocfs2/dlmfs/dlmfs.c b/fs/ocfs2/dlmfs/dlmfs.c index abfac0d7ae9c..3b5825ef3193 100644 --- a/fs/ocfs2/dlmfs/dlmfs.c +++ b/fs/ocfs2/dlmfs/dlmfs.c @@ -582,24 +582,14 @@ static int dlmfs_fill_super(struct super_block * sb, void * data, int silent) { - struct inode * inode; - struct dentry * root; - sb->s_maxbytes = MAX_LFS_FILESIZE; sb->s_blocksize = PAGE_CACHE_SIZE; sb->s_blocksize_bits = PAGE_CACHE_SHIFT; sb->s_magic = DLMFS_MAGIC; sb->s_op = &dlmfs_ops; - inode = dlmfs_get_root_inode(sb); - if (!inode) + sb->s_root = d_make_root(dlmfs_get_root_inode(sb)); + if (!sb->s_root) return -ENOMEM; - - root = d_alloc_root(inode); - if (!root) { - iput(inode); - return -ENOMEM; - } - sb->s_root = root; return 0; } diff --git a/fs/ocfs2/super.c b/fs/ocfs2/super.c index 2b1184f7097f..337687c3e233 100644 --- a/fs/ocfs2/super.c +++ b/fs/ocfs2/super.c @@ -1166,9 +1166,8 @@ static int ocfs2_fill_super(struct super_block *sb, void *data, int silent) goto read_super_error; } - root = d_alloc_root(inode); + root = d_make_root(inode); if (!root) { - iput(inode); status = -ENOMEM; mlog_errno(status); goto read_super_error; diff --git a/fs/omfs/inode.c b/fs/omfs/inode.c index 6065bb0ba207..dbc842222589 100644 --- a/fs/omfs/inode.c +++ b/fs/omfs/inode.c @@ -539,11 +539,9 @@ static int omfs_fill_super(struct super_block *sb, void *data, int silent) goto out_brelse_bh2; } - sb->s_root = d_alloc_root(root); - if (!sb->s_root) { - iput(root); + sb->s_root = d_make_root(root); + if (!sb->s_root) goto out_brelse_bh2; - } printk(KERN_DEBUG "omfs: Mounted volume %s\n", omfs_rb->r_name); ret = 0; diff --git a/fs/openpromfs/inode.c b/fs/openpromfs/inode.c index a88c03bc749d..bc49c975d501 100644 --- a/fs/openpromfs/inode.c +++ b/fs/openpromfs/inode.c @@ -408,13 +408,12 @@ static int openprom_fill_super(struct super_block *s, void *data, int silent) oi->type = op_inode_node; oi->u.node = of_find_node_by_path("/"); - s->s_root = d_alloc_root(root_inode); + s->s_root = d_make_root(root_inode); if (!s->s_root) goto out_no_root_dentry; return 0; out_no_root_dentry: - iput(root_inode); ret = -ENOMEM; out_no_root: printk("openprom_fill_super: get root inode failed\n"); diff --git a/fs/proc/inode.c b/fs/proc/inode.c index a70af3a44f45..8461a7b82fdb 100644 --- a/fs/proc/inode.c +++ b/fs/proc/inode.c @@ -486,8 +486,6 @@ struct inode *proc_get_inode(struct super_block *sb, struct proc_dir_entry *de) int proc_fill_super(struct super_block *s) { - struct inode * root_inode; - s->s_flags |= MS_NODIRATIME | MS_NOSUID | MS_NOEXEC; s->s_blocksize = 1024; s->s_blocksize_bits = 10; @@ -496,17 +494,10 @@ int proc_fill_super(struct super_block *s) s->s_time_gran = 1; pde_get(&proc_root); - root_inode = proc_get_inode(s, &proc_root); - if (!root_inode) - goto out_no_root; - s->s_root = d_alloc_root(root_inode); - if (!s->s_root) { - iput(root_inode); - goto out_no_root; - } - return 0; + s->s_root = d_make_root(proc_get_inode(s, &proc_root)); + if (s->s_root) + return 0; -out_no_root: printk("proc_read_super: get root inode failed\n"); pde_put(&proc_root); return -ENOMEM; diff --git a/fs/pstore/inode.c b/fs/pstore/inode.c index b3b426edb2fd..ec7d1fb6f35a 100644 --- a/fs/pstore/inode.c +++ b/fs/pstore/inode.c @@ -303,7 +303,7 @@ int pstore_fill_super(struct super_block *sb, void *data, int silent) /* override ramfs "dir" options so we catch unlink(2) */ inode->i_op = &pstore_dir_inode_operations; - root = d_alloc_root(inode); + root = d_make_root(inode); sb->s_root = root; if (!root) { err = -ENOMEM; @@ -314,7 +314,6 @@ int pstore_fill_super(struct super_block *sb, void *data, int silent) return 0; fail: - iput(inode); return err; } diff --git a/fs/qnx4/inode.c b/fs/qnx4/inode.c index 6b009548d2e0..db18d866d981 100644 --- a/fs/qnx4/inode.c +++ b/fs/qnx4/inode.c @@ -260,15 +260,13 @@ static int qnx4_fill_super(struct super_block *s, void *data, int silent) } ret = -ENOMEM; - s->s_root = d_alloc_root(root); + s->s_root = d_make_root(root); if (s->s_root == NULL) - goto outi; + goto outb; brelse(bh); return 0; - outi: - iput(root); outb: kfree(qs->BitMap); out: diff --git a/fs/ramfs/inode.c b/fs/ramfs/inode.c index aec766abe3af..b6612d2ed718 100644 --- a/fs/ramfs/inode.c +++ b/fs/ramfs/inode.c @@ -210,7 +210,6 @@ int ramfs_fill_super(struct super_block *sb, void *data, int silent) { struct ramfs_fs_info *fsi; struct inode *inode = NULL; - struct dentry *root; int err; save_mount_options(sb, data); @@ -234,14 +233,8 @@ int ramfs_fill_super(struct super_block *sb, void *data, int silent) sb->s_time_gran = 1; inode = ramfs_get_inode(sb, NULL, S_IFDIR | fsi->mount_opts.mode, 0); - if (!inode) { - err = -ENOMEM; - goto fail; - } - - root = d_alloc_root(inode); - sb->s_root = root; - if (!root) { + sb->s_root = d_make_root(inode); + if (!sb->s_root) { err = -ENOMEM; goto fail; } @@ -250,7 +243,6 @@ int ramfs_fill_super(struct super_block *sb, void *data, int silent) fail: kfree(fsi); sb->s_fs_info = NULL; - iput(inode); return err; } diff --git a/fs/reiserfs/super.c b/fs/reiserfs/super.c index e12d8b97cd4d..208dfd144409 100644 --- a/fs/reiserfs/super.c +++ b/fs/reiserfs/super.c @@ -1874,11 +1874,9 @@ static int reiserfs_fill_super(struct super_block *s, void *data, int silent) unlock_new_inode(root_inode); } - s->s_root = d_alloc_root(root_inode); - if (!s->s_root) { - iput(root_inode); + s->s_root = d_make_root(root_inode); + if (!s->s_root) goto error; - } // define and initialize hash function sbi->s_hash_function = hash_function(s); if (sbi->s_hash_function == NULL) { diff --git a/fs/romfs/super.c b/fs/romfs/super.c index bb36ab74eb45..e64f6b5f7ae5 100644 --- a/fs/romfs/super.c +++ b/fs/romfs/super.c @@ -538,14 +538,12 @@ static int romfs_fill_super(struct super_block *sb, void *data, int silent) if (IS_ERR(root)) goto error; - sb->s_root = d_alloc_root(root); + sb->s_root = d_make_root(root); if (!sb->s_root) - goto error_i; + goto error; return 0; -error_i: - iput(root); error: return -EINVAL; error_rsb_inval: diff --git a/fs/squashfs/super.c b/fs/squashfs/super.c index ecaa2f7bdb8f..970b1167e7cb 100644 --- a/fs/squashfs/super.c +++ b/fs/squashfs/super.c @@ -316,11 +316,10 @@ check_directory_table: } insert_inode_hash(root); - sb->s_root = d_alloc_root(root); + sb->s_root = d_make_root(root); if (sb->s_root == NULL) { ERROR("Root inode create failed\n"); err = -ENOMEM; - iput(root); goto failed_mount; } diff --git a/fs/sysfs/mount.c b/fs/sysfs/mount.c index e34f0d99ea4e..2243f8ec64d5 100644 --- a/fs/sysfs/mount.c +++ b/fs/sysfs/mount.c @@ -61,10 +61,9 @@ static int sysfs_fill_super(struct super_block *sb, void *data, int silent) } /* instantiate and link root dentry */ - root = d_alloc_root(inode); + root = d_make_root(inode); if (!root) { pr_debug("%s: could not get root dentry!\n",__func__); - iput(inode); return -ENOMEM; } root->d_fsdata = &sysfs_root; diff --git a/fs/sysv/super.c b/fs/sysv/super.c index f467740e088c..7491c33b6468 100644 --- a/fs/sysv/super.c +++ b/fs/sysv/super.c @@ -341,9 +341,8 @@ static int complete_read_super(struct super_block *sb, int silent, int size) printk("SysV FS: get root inode failed\n"); return 0; } - sb->s_root = d_alloc_root(root_inode); + sb->s_root = d_make_root(root_inode); if (!sb->s_root) { - iput(root_inode); printk("SysV FS: get root dentry failed\n"); return 0; } diff --git a/fs/ubifs/super.c b/fs/ubifs/super.c index 63765d58445b..76e4e0566ad6 100644 --- a/fs/ubifs/super.c +++ b/fs/ubifs/super.c @@ -2076,15 +2076,13 @@ static int ubifs_fill_super(struct super_block *sb, void *data, int silent) goto out_umount; } - sb->s_root = d_alloc_root(root); + sb->s_root = d_make_root(root); if (!sb->s_root) - goto out_iput; + goto out_umount; mutex_unlock(&c->umount_mutex); return 0; -out_iput: - iput(root); out_umount: ubifs_umount(c); out_unlock: diff --git a/fs/udf/super.c b/fs/udf/super.c index 8d8b25336fbb..85067b4c7e14 100644 --- a/fs/udf/super.c +++ b/fs/udf/super.c @@ -2037,10 +2037,9 @@ static int udf_fill_super(struct super_block *sb, void *options, int silent) } /* Allocate a dentry for the root inode */ - sb->s_root = d_alloc_root(inode); + sb->s_root = d_make_root(inode); if (!sb->s_root) { udf_err(sb, "Couldn't allocate root dentry\n"); - iput(inode); goto error_out; } sb->s_maxbytes = MAX_LFS_FILESIZE; diff --git a/fs/ufs/super.c b/fs/ufs/super.c index ec25d09fcaa8..f636f6b460d0 100644 --- a/fs/ufs/super.c +++ b/fs/ufs/super.c @@ -1164,10 +1164,10 @@ magic_found: ret = PTR_ERR(inode); goto failed; } - sb->s_root = d_alloc_root(inode); + sb->s_root = d_make_root(inode); if (!sb->s_root) { ret = -ENOMEM; - goto dalloc_failed; + goto failed; } ufs_setup_cstotal(sb); @@ -1181,8 +1181,6 @@ magic_found: UFSD("EXIT\n"); return 0; -dalloc_failed: - iput(inode); failed: if (ubh) ubh_brelse_uspi (uspi); diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c index 0e4c5c017fba..baf40e378d35 100644 --- a/fs/xfs/xfs_super.c +++ b/fs/xfs/xfs_super.c @@ -1362,10 +1362,10 @@ xfs_fs_fill_super( error = EINVAL; goto out_syncd_stop; } - sb->s_root = d_alloc_root(root); + sb->s_root = d_make_root(root); if (!sb->s_root) { error = ENOMEM; - goto out_iput; + goto out_syncd_stop; } return 0; @@ -1384,8 +1384,6 @@ xfs_fs_fill_super( out: return -error; - out_iput: - iput(root); out_syncd_stop: xfs_syncd_stop(mp); out_unmount: diff --git a/ipc/mqueue.c b/ipc/mqueue.c index 86ee272de210..28bd64ddeda3 100644 --- a/ipc/mqueue.c +++ b/ipc/mqueue.c @@ -188,30 +188,20 @@ static int mqueue_fill_super(struct super_block *sb, void *data, int silent) { struct inode *inode; struct ipc_namespace *ns = data; - int error; sb->s_blocksize = PAGE_CACHE_SIZE; sb->s_blocksize_bits = PAGE_CACHE_SHIFT; sb->s_magic = MQUEUE_MAGIC; sb->s_op = &mqueue_super_ops; - inode = mqueue_get_inode(sb, ns, S_IFDIR | S_ISVTX | S_IRWXUGO, - NULL); - if (IS_ERR(inode)) { - error = PTR_ERR(inode); - goto out; - } + inode = mqueue_get_inode(sb, ns, S_IFDIR | S_ISVTX | S_IRWXUGO, NULL); + if (IS_ERR(inode)) + return PTR_ERR(inode); - sb->s_root = d_alloc_root(inode); - if (!sb->s_root) { - iput(inode); - error = -ENOMEM; - goto out; - } - error = 0; - -out: - return error; + sb->s_root = d_make_root(inode); + if (!sb->s_root) + return -ENOMEM; + return 0; } static struct dentry *mqueue_mount(struct file_system_type *fs_type, diff --git a/kernel/cgroup.c b/kernel/cgroup.c index a5d3b5325f77..711c1a30ceaa 100644 --- a/kernel/cgroup.c +++ b/kernel/cgroup.c @@ -1472,7 +1472,6 @@ static int cgroup_get_rootdir(struct super_block *sb) struct inode *inode = cgroup_new_inode(S_IFDIR | S_IRUGO | S_IXUGO | S_IWUSR, sb); - struct dentry *dentry; if (!inode) return -ENOMEM; @@ -1481,12 +1480,9 @@ static int cgroup_get_rootdir(struct super_block *sb) inode->i_op = &cgroup_dir_inode_operations; /* directories start off with i_nlink == 2 (for "." entry) */ inc_nlink(inode); - dentry = d_alloc_root(inode); - if (!dentry) { - iput(inode); + sb->s_root = d_make_root(inode); + if (!sb->s_root) return -ENOMEM; - } - sb->s_root = dentry; /* for everything else we want ->d_op set */ sb->s_d_op = &cgroup_dops; return 0; diff --git a/mm/shmem.c b/mm/shmem.c index 269d049294ab..154243f0a27c 100644 --- a/mm/shmem.c +++ b/mm/shmem.c @@ -2232,14 +2232,12 @@ int shmem_fill_super(struct super_block *sb, void *data, int silent) goto failed; inode->i_uid = sbinfo->uid; inode->i_gid = sbinfo->gid; - root = d_alloc_root(inode); + root = d_make_root(inode); if (!root) - goto failed_iput; + goto failed; sb->s_root = root; return 0; -failed_iput: - iput(inode); failed: shmem_put_super(sb); return err; diff --git a/net/sunrpc/rpc_pipe.c b/net/sunrpc/rpc_pipe.c index 63a7a7add21e..7d6dd6efbdbe 100644 --- a/net/sunrpc/rpc_pipe.c +++ b/net/sunrpc/rpc_pipe.c @@ -1033,13 +1033,9 @@ rpc_fill_super(struct super_block *sb, void *data, int silent) sb->s_time_gran = 1; inode = rpc_get_inode(sb, S_IFDIR | 0755); - if (!inode) + sb->s_root = root = d_make_root(inode); + if (!root) return -ENOMEM; - sb->s_root = root = d_alloc_root(inode); - if (!root) { - iput(inode); - return -ENOMEM; - } if (rpc_populate(root, files, RPCAUTH_lockd, RPCAUTH_RootEOF, NULL)) return -ENOMEM; return 0; From ea29c6950a0305f2af7fd4a2021a6e946ada0174 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sun, 12 Feb 2012 22:04:09 -0500 Subject: [PATCH 10/64] ntfs: switch to d_make_root() Signed-off-by: Al Viro --- fs/ntfs/super.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/fs/ntfs/super.c b/fs/ntfs/super.c index f907611cca73..22020d8b1ed2 100644 --- a/fs/ntfs/super.c +++ b/fs/ntfs/super.c @@ -2908,9 +2908,10 @@ static int ntfs_fill_super(struct super_block *sb, void *opt, const int silent) ntfs_error(sb, "Failed to load system files."); goto unl_upcase_iput_tmp_ino_err_out_now; } - if ((sb->s_root = d_alloc_root(vol->root_ino))) { - /* We grab a reference, simulating an ntfs_iget(). */ - ihold(vol->root_ino); + + /* We grab a reference, simulating an ntfs_iget(). */ + ihold(vol->root_ino); + if ((sb->s_root = d_make_root(vol->root_ino))) { ntfs_debug("Exiting, status successful."); /* Release the default upcase if it has no users. */ mutex_lock(&ntfs_lock); From 1688f86046e5572623b2eacb685eb707fe21fb0a Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sun, 12 Feb 2012 22:06:33 -0500 Subject: [PATCH 11/64] fat: switch to d_make_root() Signed-off-by: Al Viro --- fs/fat/inode.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/fs/fat/inode.c b/fs/fat/inode.c index 3ab841054d53..21687e31acc0 100644 --- a/fs/fat/inode.c +++ b/fs/fat/inode.c @@ -1496,11 +1496,13 @@ int fat_fill_super(struct super_block *sb, void *data, int silent, int isvfat, root_inode->i_ino = MSDOS_ROOT_INO; root_inode->i_version = 1; error = fat_read_root(root_inode); - if (error < 0) + if (error < 0) { + iput(root_inode); goto out_fail; + } error = -ENOMEM; insert_inode_hash(root_inode); - sb->s_root = d_alloc_root(root_inode); + sb->s_root = d_make_root(root_inode); if (!sb->s_root) { fat_msg(sb, KERN_ERR, "get root inode failed"); goto out_fail; @@ -1516,8 +1518,6 @@ out_invalid: out_fail: if (fat_inode) iput(fat_inode); - if (root_inode) - iput(root_inode); unload_nls(sbi->nls_io); unload_nls(sbi->nls_disk); if (sbi->options.iocharset != fat_default_iocharset) From 68acb8e60d672cba0fd1d3545ba33343931c7a24 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sun, 12 Feb 2012 22:07:07 -0500 Subject: [PATCH 12/64] hfsplus: switch to d_make_root() Signed-off-by: Al Viro --- fs/hfsplus/super.c | 17 +++++++++-------- 1 file changed, 9 insertions(+), 8 deletions(-) diff --git a/fs/hfsplus/super.c b/fs/hfsplus/super.c index 427682ca9e48..ceb1c281eefb 100644 --- a/fs/hfsplus/super.c +++ b/fs/hfsplus/super.c @@ -465,6 +465,13 @@ static int hfsplus_fill_super(struct super_block *sb, void *data, int silent) goto out_put_alloc_file; } + sb->s_d_op = &hfsplus_dentry_operations; + sb->s_root = d_make_root(root); + if (!sb->s_root) { + err = -ENOMEM; + goto out_put_alloc_file; + } + str.len = sizeof(HFSP_HIDDENDIR_NAME) - 1; str.name = HFSP_HIDDENDIR_NAME; err = hfs_find_init(sbi->cat_tree, &fd); @@ -515,13 +522,6 @@ static int hfsplus_fill_super(struct super_block *sb, void *data, int silent) } } - sb->s_d_op = &hfsplus_dentry_operations; - sb->s_root = d_alloc_root(root); - if (!sb->s_root) { - err = -ENOMEM; - goto out_put_hidden_dir; - } - unload_nls(sbi->nls); sbi->nls = nls; return 0; @@ -529,7 +529,8 @@ static int hfsplus_fill_super(struct super_block *sb, void *data, int silent) out_put_hidden_dir: iput(sbi->hidden_dir); out_put_root: - iput(root); + dput(sb->s_root); + sb->s_root = NULL; out_put_alloc_file: iput(sbi->alloc_file); out_close_cat_tree: From ca85c07809ca19de3391cb79ee1198f3dd91fa8d Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sun, 12 Feb 2012 22:07:43 -0500 Subject: [PATCH 13/64] minixfs: switch to d_make_root() Signed-off-by: Al Viro --- fs/minix/inode.c | 28 ++++++++++++---------------- 1 file changed, 12 insertions(+), 16 deletions(-) diff --git a/fs/minix/inode.c b/fs/minix/inode.c index 62c697caffb9..fcb05d2c6b5f 100644 --- a/fs/minix/inode.c +++ b/fs/minix/inode.c @@ -254,14 +254,6 @@ static int minix_fill_super(struct super_block *s, void *data, int silent) minix_set_bit(0,sbi->s_imap[0]->b_data); minix_set_bit(0,sbi->s_zmap[0]->b_data); - /* set up enough so that it can read an inode */ - s->s_op = &minix_sops; - root_inode = minix_iget(s, MINIX_ROOT_INO); - if (IS_ERR(root_inode)) { - ret = PTR_ERR(root_inode); - goto out_no_root; - } - /* Apparently minix can create filesystems that allocate more blocks for * the bitmaps than needed. We simply ignore that, but verify it didn't * create one with not enough blocks and bail out if so. @@ -270,7 +262,7 @@ static int minix_fill_super(struct super_block *s, void *data, int silent) if (sbi->s_imap_blocks < block) { printk("MINIX-fs: file system does not have enough " "imap blocks allocated. Refusing to mount\n"); - goto out_iput; + goto out_no_bitmap; } block = minix_blocks_needed( @@ -279,13 +271,21 @@ static int minix_fill_super(struct super_block *s, void *data, int silent) if (sbi->s_zmap_blocks < block) { printk("MINIX-fs: file system does not have enough " "zmap blocks allocated. Refusing to mount.\n"); - goto out_iput; + goto out_no_bitmap; + } + + /* set up enough so that it can read an inode */ + s->s_op = &minix_sops; + root_inode = minix_iget(s, MINIX_ROOT_INO); + if (IS_ERR(root_inode)) { + ret = PTR_ERR(root_inode); + goto out_no_root; } ret = -ENOMEM; - s->s_root = d_alloc_root(root_inode); + s->s_root = d_make_root(root_inode); if (!s->s_root) - goto out_iput; + goto out_no_root; if (!(s->s_flags & MS_RDONLY)) { if (sbi->s_version != MINIX_V3) /* s_state is now out from V3 sb */ @@ -301,10 +301,6 @@ static int minix_fill_super(struct super_block *s, void *data, int silent) return 0; -out_iput: - iput(root_inode); - goto out_freemap; - out_no_root: if (!silent) printk("MINIX-fs: get root inode failed\n"); From 318ceed088497d1ca839b1172518ac4cc7096b82 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sun, 12 Feb 2012 22:08:01 -0500 Subject: [PATCH 14/64] tidy up after d_make_root() conversion Signed-off-by: Al Viro --- drivers/misc/ibmasm/ibmasmfs.c | 14 ++++++-------- drivers/oprofile/oprofilefs.c | 9 +++------ drivers/usb/core/inode.c | 6 ++---- fs/pstore/inode.c | 24 +++++++----------------- fs/ramfs/inode.c | 20 ++++++-------------- mm/shmem.c | 6 ++---- 6 files changed, 26 insertions(+), 53 deletions(-) diff --git a/drivers/misc/ibmasm/ibmasmfs.c b/drivers/misc/ibmasm/ibmasmfs.c index 15f24f362208..1c034b80d408 100644 --- a/drivers/misc/ibmasm/ibmasmfs.c +++ b/drivers/misc/ibmasm/ibmasmfs.c @@ -87,7 +87,7 @@ static LIST_HEAD(service_processors); static struct inode *ibmasmfs_make_inode(struct super_block *sb, int mode); -static void ibmasmfs_create_files (struct super_block *sb, struct dentry *root); +static void ibmasmfs_create_files (struct super_block *sb); static int ibmasmfs_fill_super (struct super_block *sb, void *data, int silent); @@ -114,7 +114,6 @@ static struct file_system_type ibmasmfs_type = { static int ibmasmfs_fill_super (struct super_block *sb, void *data, int silent) { struct inode *root; - struct dentry *root_dentry; sb->s_blocksize = PAGE_CACHE_SIZE; sb->s_blocksize_bits = PAGE_CACHE_SHIFT; @@ -129,12 +128,11 @@ static int ibmasmfs_fill_super (struct super_block *sb, void *data, int silent) root->i_op = &simple_dir_inode_operations; root->i_fop = ibmasmfs_dir_ops; - root_dentry = d_make_root(root); - if (!root_dentry) + sb->s_root = d_make_root(root); + if (!sb->s_root) return -ENOMEM; - sb->s_root = root_dentry; - ibmasmfs_create_files(sb, root_dentry); + ibmasmfs_create_files(sb); return 0; } @@ -610,7 +608,7 @@ static const struct file_operations remote_settings_fops = { }; -static void ibmasmfs_create_files (struct super_block *sb, struct dentry *root) +static void ibmasmfs_create_files (struct super_block *sb) { struct list_head *entry; struct service_processor *sp; @@ -619,7 +617,7 @@ static void ibmasmfs_create_files (struct super_block *sb, struct dentry *root) struct dentry *dir; struct dentry *remote_dir; sp = list_entry(entry, struct service_processor, node); - dir = ibmasmfs_create_dir(sb, root, sp->dirname); + dir = ibmasmfs_create_dir(sb, sb->s_root, sp->dirname); if (!dir) continue; diff --git a/drivers/oprofile/oprofilefs.c b/drivers/oprofile/oprofilefs.c index 277bb70b8d75..ee8fd037bb53 100644 --- a/drivers/oprofile/oprofilefs.c +++ b/drivers/oprofile/oprofilefs.c @@ -238,7 +238,6 @@ struct dentry *oprofilefs_mkdir(struct super_block *sb, static int oprofilefs_fill_super(struct super_block *sb, void *data, int silent) { struct inode *root_inode; - struct dentry *root_dentry; sb->s_blocksize = PAGE_CACHE_SIZE; sb->s_blocksize_bits = PAGE_CACHE_SHIFT; @@ -251,13 +250,11 @@ static int oprofilefs_fill_super(struct super_block *sb, void *data, int silent) return -ENOMEM; root_inode->i_op = &simple_dir_inode_operations; root_inode->i_fop = &simple_dir_operations; - root_dentry = d_make_root(root_inode); - if (!root_dentry) + sb->s_root = d_make_root(root_inode); + if (!sb->s_root) return -ENOMEM; - sb->s_root = root_dentry; - - oprofile_create_files(sb, root_dentry); + oprofile_create_files(sb, sb->s_root); // FIXME: verify kill_litter_super removes our dentries return 0; diff --git a/drivers/usb/core/inode.c b/drivers/usb/core/inode.c index bdaef8e36020..06c175aee6a3 100644 --- a/drivers/usb/core/inode.c +++ b/drivers/usb/core/inode.c @@ -454,7 +454,6 @@ static const struct super_operations usbfs_ops = { static int usbfs_fill_super(struct super_block *sb, void *data, int silent) { struct inode *inode; - struct dentry *root; sb->s_blocksize = PAGE_CACHE_SIZE; sb->s_blocksize_bits = PAGE_CACHE_SHIFT; @@ -462,12 +461,11 @@ static int usbfs_fill_super(struct super_block *sb, void *data, int silent) sb->s_op = &usbfs_ops; sb->s_time_gran = 1; inode = usbfs_get_inode(sb, S_IFDIR | 0755, 0); - root = d_make_root(inode); - if (!root) { + sb->s_root = d_make_root(inode); + if (!sb->s_root) { dbg("%s: could not get root dentry!",__func__); return -ENOMEM; } - sb->s_root = root; return 0; } diff --git a/fs/pstore/inode.c b/fs/pstore/inode.c index ec7d1fb6f35a..f37c32b94525 100644 --- a/fs/pstore/inode.c +++ b/fs/pstore/inode.c @@ -278,9 +278,7 @@ fail: int pstore_fill_super(struct super_block *sb, void *data, int silent) { - struct inode *inode = NULL; - struct dentry *root; - int err; + struct inode *inode; save_mount_options(sb, data); @@ -296,25 +294,17 @@ int pstore_fill_super(struct super_block *sb, void *data, int silent) parse_options(data); inode = pstore_get_inode(sb, NULL, S_IFDIR | 0755, 0); - if (!inode) { - err = -ENOMEM; - goto fail; - } - /* override ramfs "dir" options so we catch unlink(2) */ - inode->i_op = &pstore_dir_inode_operations; - - root = d_make_root(inode); - sb->s_root = root; - if (!root) { - err = -ENOMEM; - goto fail; + if (inode) { + /* override ramfs "dir" options so we catch unlink(2) */ + inode->i_op = &pstore_dir_inode_operations; } + sb->s_root = d_make_root(inode); + if (!sb->s_root) + return -ENOMEM; pstore_get_records(0); return 0; -fail: - return err; } static struct dentry *pstore_mount(struct file_system_type *fs_type, diff --git a/fs/ramfs/inode.c b/fs/ramfs/inode.c index b6612d2ed718..a1fdabe21dec 100644 --- a/fs/ramfs/inode.c +++ b/fs/ramfs/inode.c @@ -209,21 +209,19 @@ static int ramfs_parse_options(char *data, struct ramfs_mount_opts *opts) int ramfs_fill_super(struct super_block *sb, void *data, int silent) { struct ramfs_fs_info *fsi; - struct inode *inode = NULL; + struct inode *inode; int err; save_mount_options(sb, data); fsi = kzalloc(sizeof(struct ramfs_fs_info), GFP_KERNEL); sb->s_fs_info = fsi; - if (!fsi) { - err = -ENOMEM; - goto fail; - } + if (!fsi) + return -ENOMEM; err = ramfs_parse_options(data, &fsi->mount_opts); if (err) - goto fail; + return err; sb->s_maxbytes = MAX_LFS_FILESIZE; sb->s_blocksize = PAGE_CACHE_SIZE; @@ -234,16 +232,10 @@ int ramfs_fill_super(struct super_block *sb, void *data, int silent) inode = ramfs_get_inode(sb, NULL, S_IFDIR | fsi->mount_opts.mode, 0); sb->s_root = d_make_root(inode); - if (!sb->s_root) { - err = -ENOMEM; - goto fail; - } + if (!sb->s_root) + return -ENOMEM; return 0; -fail: - kfree(fsi); - sb->s_fs_info = NULL; - return err; } struct dentry *ramfs_mount(struct file_system_type *fs_type, diff --git a/mm/shmem.c b/mm/shmem.c index 154243f0a27c..9bc4a2353944 100644 --- a/mm/shmem.c +++ b/mm/shmem.c @@ -2175,7 +2175,6 @@ static void shmem_put_super(struct super_block *sb) int shmem_fill_super(struct super_block *sb, void *data, int silent) { struct inode *inode; - struct dentry *root; struct shmem_sb_info *sbinfo; int err = -ENOMEM; @@ -2232,10 +2231,9 @@ int shmem_fill_super(struct super_block *sb, void *data, int silent) goto failed; inode->i_uid = sbinfo->uid; inode->i_gid = sbinfo->gid; - root = d_make_root(inode); - if (!root) + sb->s_root = d_make_root(inode); + if (!sb->s_root) goto failed; - sb->s_root = root; return 0; failed: From 32991ab305ace7017c62f8eecbe5eb36dc32e13b Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sun, 12 Feb 2012 22:15:47 -0500 Subject: [PATCH 15/64] vfs: d_alloc_root() gone all callers converted to d_make_root() by now Signed-off-by: Al Viro --- Documentation/filesystems/porting | 6 ++++++ fs/dcache.c | 24 ------------------------ include/linux/dcache.h | 1 - 3 files changed, 6 insertions(+), 25 deletions(-) diff --git a/Documentation/filesystems/porting b/Documentation/filesystems/porting index b4a3d765ff9a..74acd9618819 100644 --- a/Documentation/filesystems/porting +++ b/Documentation/filesystems/porting @@ -429,3 +429,9 @@ filemap_write_and_wait_range() so that all dirty pages are synced out properly. You must also keep in mind that ->fsync() is not called with i_mutex held anymore, so if you require i_mutex locking you must make sure to take it and release it yourself. + +-- +[mandatory] + d_alloc_root() is gone, along with a lot of bugs caused by code +misusing it. Replacement: d_make_root(inode). The difference is, +d_make_root() drops the reference to inode if dentry allocation fails. diff --git a/fs/dcache.c b/fs/dcache.c index bcbdb33fcc20..a78e145a4357 100644 --- a/fs/dcache.c +++ b/fs/dcache.c @@ -1443,30 +1443,6 @@ struct dentry *d_instantiate_unique(struct dentry *entry, struct inode *inode) EXPORT_SYMBOL(d_instantiate_unique); -/** - * d_alloc_root - allocate root dentry - * @root_inode: inode to allocate the root for - * - * Allocate a root ("/") dentry for the inode given. The inode is - * instantiated and returned. %NULL is returned if there is insufficient - * memory or the inode passed is %NULL. - */ - -struct dentry * d_alloc_root(struct inode * root_inode) -{ - struct dentry *res = NULL; - - if (root_inode) { - static const struct qstr name = { .name = "/", .len = 1 }; - - res = __d_alloc(root_inode->i_sb, &name); - if (res) - d_instantiate(res, root_inode); - } - return res; -} -EXPORT_SYMBOL(d_alloc_root); - struct dentry *d_make_root(struct inode *root_inode) { struct dentry *res = NULL; diff --git a/include/linux/dcache.h b/include/linux/dcache.h index ff5f5256d175..7e11f1418203 100644 --- a/include/linux/dcache.h +++ b/include/linux/dcache.h @@ -222,7 +222,6 @@ extern void shrink_dcache_for_umount(struct super_block *); extern int d_invalidate(struct dentry *); /* only used at mount-time */ -extern struct dentry * d_alloc_root(struct inode *); extern struct dentry * d_make_root(struct inode *); /* - the ramfs-type tree */ From 516cdb68e5b44ca1bef31619f5da8d5e9e298f88 Mon Sep 17 00:00:00 2001 From: Kai Bankett Date: Mon, 13 Feb 2012 02:43:41 +0100 Subject: [PATCH 16/64] qnx4fs: small cleanup Small qnx4 cleanup patch. - removes .writepage, .write_begin and .write_end (+callback functions) - removes '.' path checking in namei.c (handled on upper layers) Signed-off-by: Kai Bankett Signed-off-by: Al Viro --- fs/qnx4/inode.c | 27 --------------------------- fs/qnx4/namei.c | 4 ---- 2 files changed, 31 deletions(-) diff --git a/fs/qnx4/inode.c b/fs/qnx4/inode.c index db18d866d981..14a0ba0f0c1c 100644 --- a/fs/qnx4/inode.c +++ b/fs/qnx4/inode.c @@ -286,44 +286,17 @@ static void qnx4_put_super(struct super_block *sb) return; } -static int qnx4_writepage(struct page *page, struct writeback_control *wbc) -{ - return block_write_full_page(page,qnx4_get_block, wbc); -} - static int qnx4_readpage(struct file *file, struct page *page) { return block_read_full_page(page,qnx4_get_block); } -static int qnx4_write_begin(struct file *file, struct address_space *mapping, - loff_t pos, unsigned len, unsigned flags, - struct page **pagep, void **fsdata) -{ - struct qnx4_inode_info *qnx4_inode = qnx4_i(mapping->host); - int ret; - - *pagep = NULL; - ret = cont_write_begin(file, mapping, pos, len, flags, pagep, fsdata, - qnx4_get_block, - &qnx4_inode->mmu_private); - if (unlikely(ret)) { - loff_t isize = mapping->host->i_size; - if (pos + len > isize) - vmtruncate(mapping->host, isize); - } - - return ret; -} static sector_t qnx4_bmap(struct address_space *mapping, sector_t block) { return generic_block_bmap(mapping,block,qnx4_get_block); } static const struct address_space_operations qnx4_aops = { .readpage = qnx4_readpage, - .writepage = qnx4_writepage, - .write_begin = qnx4_write_begin, - .write_end = generic_write_end, .bmap = qnx4_bmap }; diff --git a/fs/qnx4/namei.c b/fs/qnx4/namei.c index 275327b5615e..e8eb8784ed30 100644 --- a/fs/qnx4/namei.c +++ b/fs/qnx4/namei.c @@ -39,10 +39,6 @@ static int qnx4_match(int len, const char *name, } else { namelen = QNX4_SHORT_NAME_MAX; } - /* "" means "." ---> so paths like "/usr/lib//libc.a" work */ - if (!len && (de->di_fname[0] == '.') && (de->di_fname[1] == '\0')) { - return 1; - } thislen = strlen( de->di_fname ); if ( thislen > namelen ) thislen = namelen; From 5d026c7242201e7c9d0e12fcb2bcaffead9d59fd Mon Sep 17 00:00:00 2001 From: Kai Bankett Date: Fri, 17 Feb 2012 05:59:20 +0100 Subject: [PATCH 17/64] fs: initial qnx6fs addition Adds support for qnx6fs readonly support to the linux kernel. * Mount option The option mmi_fs can be used to mount Harman Becker/Audi MMI 3G HDD qnx6fs filesystems. * Documentation A high level filesystem stucture description can be found in the Documentation/filesystems directory. (qnx6.txt) * Additional features - Active (stable) superblock selection - Superblock checksum check (enforced) - Supports mount of qnx6 filesystems with to host different endianess - Automatic endianess detection - Longfilename support (with non-enfocing crc check) - All blocksizes (512, 1024, 2048 and 4096 supported) Signed-off-by: Kai Bankett Signed-off-by: Al Viro --- Documentation/filesystems/qnx6.txt | 174 +++++++ fs/Kconfig | 1 + fs/Makefile | 1 + fs/qnx6/Kconfig | 26 ++ fs/qnx6/Makefile | 7 + fs/qnx6/README | 8 + fs/qnx6/dir.c | 291 ++++++++++++ fs/qnx6/inode.c | 698 +++++++++++++++++++++++++++++ fs/qnx6/namei.c | 42 ++ fs/qnx6/qnx6.h | 135 ++++++ fs/qnx6/super_mmi.c | 150 +++++++ include/linux/magic.h | 1 + include/linux/qnx6_fs.h | 134 ++++++ 13 files changed, 1668 insertions(+) create mode 100644 Documentation/filesystems/qnx6.txt create mode 100644 fs/qnx6/Kconfig create mode 100644 fs/qnx6/Makefile create mode 100644 fs/qnx6/README create mode 100644 fs/qnx6/dir.c create mode 100644 fs/qnx6/inode.c create mode 100644 fs/qnx6/namei.c create mode 100644 fs/qnx6/qnx6.h create mode 100644 fs/qnx6/super_mmi.c create mode 100644 include/linux/qnx6_fs.h diff --git a/Documentation/filesystems/qnx6.txt b/Documentation/filesystems/qnx6.txt new file mode 100644 index 000000000000..050223ea03c7 --- /dev/null +++ b/Documentation/filesystems/qnx6.txt @@ -0,0 +1,174 @@ +The QNX6 Filesystem +=================== + +The qnx6fs is used by newer QNX operating system versions. (e.g. Neutrino) +It got introduced in QNX 6.4.0 and is used default since 6.4.1. + +Option +====== + +mmi_fs Mount filesystem as used for example by Audi MMI 3G system + +Specification +============= + +qnx6fs shares many properties with traditional Unix filesystems. It has the +concepts of blocks, inodes and directories. +On QNX it is possible to create little endian and big endian qnx6 filesystems. +This feature makes it possible to create and use a different endianness fs +for the target (QNX is used on quite a range of embedded systems) plattform +running on a different endianess. +The Linux driver handles endianness transparently. (LE and BE) + +Blocks +------ + +The space in the device or file is split up into blocks. These are a fixed +size of 512, 1024, 2048 or 4096, which is decided when the filesystem is +created. +Blockpointers are 32bit, so the maximum space that can be adressed is +2^32 * 4096 bytes or 16TB + +The superblocks +--------------- + +The superblock contains all global information about the filesystem. +Each qnx6fs got two superblocks, each one having a 64bit serial number. +That serial number is used to identify the "active" superblock. +In write mode with reach new snapshot (after each synchronous write), the +serial of the new master superblock is increased (old superblock serial + 1) + +So basically the snapshot functionality is realized by an atomic final +update of the serial number. Before updating that serial, all modifications +are done by copying all modified blocks during that specific write request +(or period) and building up a new (stable) filesystem structure under the +inactive superblock. + +Each superblock holds a set of root inodes for the different filesystem +parts. (Inode, Bitmap and Longfilenames) +Each of these root nodes holds information like total size of the stored +data and the adressing levels in that specific tree. +If the level value is 0, up to 16 direct blocks can be adressed by each +node. +Level 1 adds an additional indirect adressing level where each indirect +adressing block holds up to blocksize / 4 bytes pointers to data blocks. +Level 2 adds an additional indirect adressig block level (so, already up +to 16 * 256 * 256 = 1048576 blocks that can be adressed by such a tree)a + +Unused block pointers are always set to ~0 - regardless of root node, +indirect adressing blocks or inodes. +Data leaves are always on the lowest level. So no data is stored on upper +tree levels. + +The first Superblock is located at 0x2000. (0x2000 is the bootblock size) +The Audi MMI 3G first superblock directly starts at byte 0. +Second superblock position can either be calculated from the superblock +information (total number of filesystem blocks) or by taking the highest +device address, zeroing the last 3 bytes and then substracting 0x1000 from +that address. + +0x1000 is the size reserved for each superblock - regardless of the +blocksize of the filesystem. + +Inodes +------ + +Each object in the filesystem is represented by an inode. (index node) +The inode structure contains pointers to the filesystem blocks which contain +the data held in the object and all of the metadata about an object except +its longname. (filenames longer than 27 characters) +The metadata about an object includes the permissions, owner, group, flags, +size, number of blocks used, access time, change time and modification time. + +Object mode field is POSIX format. (which makes things easier) + +There are also pointers to the first 16 blocks, if the object data can be +adressed with 16 direct blocks. +For more than 16 blocks an indirect adressing in form of another tree is +used. (scheme is the same as the one used for the superblock root nodes) + +The filesize is stored 64bit. Inode counting starts with 1. (whilst long +filename inodes start with 0) + +Directories +----------- + +A directory is a filesystem object and has an inode just like a file. +It is a specially formatted file containing records which associate each +name with an inode number. +'.' inode number points to the directory inode +'..' inode number points to the parent directory inode +Eeach filename record additionally got a filename length field. + +One special case are long filenames or subdirectory names. +These got set a filename length field of 0xff in the corresponding directory +record plus the longfile inode number also stored in that record. +With that longfilename inode number, the longfilename tree can be walked +starting with the superblock longfilename root node pointers. + +Special files +------------- + +Symbolic links are also filesystem objects with inodes. They got a specific +bit in the inode mode field identifying them as symbolic link. +The directory entry file inode pointer points to the target file inode. + +Hard links got an inode, a directory entry, but a specific mode bit set, +no block pointers and the directory file record pointing to the target file +inode. + +Character and block special devices do not exist in QNX as those files +are handled by the QNX kernel/drivers and created in /dev independant of the +underlaying filesystem. + +Long filenames +-------------- + +Long filenames are stored in a seperate adressing tree. The staring point +is the longfilename root node in the active superblock. +Each data block (tree leaves) holds one long filename. That filename is +limited to 510 bytes. The first two starting bytes are used as length field +for the actual filename. +If that structure shall fit for all allowed blocksizes, it is clear why there +is a limit of 510 bytes for the actual filename stored. + +Bitmap +------ + +The qnx6fs filesystem allocation bitmap is stored in a tree under bitmap +root node in the superblock and each bit in the bitmap represents one +filesystem block. +The first block is block 0, which starts 0x1000 after superblock start. +So for a normal qnx6fs 0x3000 (bootblock + superblock) is the physical +address at which block 0 is located. + +Bits at the end of the last bitmap block are set to 1, if the device is +smaller than addressing space in the bitmap. + +Bitmap system area +------------------ + +The bitmap itself is devided into three parts. +First the system area, that is split into two halfs. +Then userspace. + +The requirement for a static, fixed preallocated system area comes from how +qnx6fs deals with writes. +Each superblock got it's own half of the system area. So superblock #1 +always uses blocks from the lower half whilst superblock #2 just writes to +blocks represented by the upper half bitmap system area bits. + +Bitmap blocks, Inode blocks and indirect addressing blocks for those two +tree structures are treated as system blocks. + +The rational behind that is that a write request can work on a new snapshot +(system area of the inactive - resp. lower serial numbered superblock) while +at the same time there is still a complete stable filesystem structer in the +other half of the system area. + +When finished with writing (a sync write is completed, the maximum sync leap +time or a filesystem sync is requested), serial of the previously inactive +superblock atomically is increased and the fs switches over to that - then +stable declared - superblock. + +For all data outside the system area, blocks are just copied while writing. diff --git a/fs/Kconfig b/fs/Kconfig index d621f02a3f9e..1497ddf27e91 100644 --- a/fs/Kconfig +++ b/fs/Kconfig @@ -210,6 +210,7 @@ source "fs/minix/Kconfig" source "fs/omfs/Kconfig" source "fs/hpfs/Kconfig" source "fs/qnx4/Kconfig" +source "fs/qnx6/Kconfig" source "fs/romfs/Kconfig" source "fs/pstore/Kconfig" source "fs/sysv/Kconfig" diff --git a/fs/Makefile b/fs/Makefile index 93804d4d66e1..2fb977934673 100644 --- a/fs/Makefile +++ b/fs/Makefile @@ -102,6 +102,7 @@ obj-$(CONFIG_UBIFS_FS) += ubifs/ obj-$(CONFIG_AFFS_FS) += affs/ obj-$(CONFIG_ROMFS_FS) += romfs/ obj-$(CONFIG_QNX4FS_FS) += qnx4/ +obj-$(CONFIG_QNX6FS_FS) += qnx6/ obj-$(CONFIG_AUTOFS4_FS) += autofs4/ obj-$(CONFIG_ADFS_FS) += adfs/ obj-$(CONFIG_FUSE_FS) += fuse/ diff --git a/fs/qnx6/Kconfig b/fs/qnx6/Kconfig new file mode 100644 index 000000000000..edbba5c17cc8 --- /dev/null +++ b/fs/qnx6/Kconfig @@ -0,0 +1,26 @@ +config QNX6FS_FS + tristate "QNX6 file system support (read only)" + depends on BLOCK && CRC32 + help + This is the file system used by the real-time operating systems + QNX 6 (also called QNX RTP). + Further information is available at . + Say Y if you intend to mount QNX hard disks or floppies formatted + with a mkqnx6fs. + However, keep in mind that this currently is a readonly driver! + + To compile this file system support as a module, choose M here: the + module will be called qnx6. + + If you don't know whether you need it, then you don't need it: + answer N. + +config QNX6FS_DEBUG + bool "QNX6 debugging information" + depends on QNX6FS_FS + help + Turns on extended debugging output. + + If you are not a developer working on the QNX6FS, you probably don't + want this: + answer N. diff --git a/fs/qnx6/Makefile b/fs/qnx6/Makefile new file mode 100644 index 000000000000..9dd06199afc9 --- /dev/null +++ b/fs/qnx6/Makefile @@ -0,0 +1,7 @@ +# +# Makefile for the linux qnx4-filesystem routines. +# + +obj-$(CONFIG_QNX6FS_FS) += qnx6.o + +qnx6-objs := inode.o dir.o namei.o super_mmi.o diff --git a/fs/qnx6/README b/fs/qnx6/README new file mode 100644 index 000000000000..116d622026cc --- /dev/null +++ b/fs/qnx6/README @@ -0,0 +1,8 @@ + + This is a snapshot of the QNX6 filesystem for Linux. + Please send diffs and remarks to . + +Credits : + +Al Viro (endless patience with me & support ;)) +Kai Bankett (Maintainer) diff --git a/fs/qnx6/dir.c b/fs/qnx6/dir.c new file mode 100644 index 000000000000..dc597353db3b --- /dev/null +++ b/fs/qnx6/dir.c @@ -0,0 +1,291 @@ +/* + * QNX6 file system, Linux implementation. + * + * Version : 1.0.0 + * + * History : + * + * 01-02-2012 by Kai Bankett (chaosman@ontika.net) : first release. + * 16-02-2012 pagemap extension by Al Viro + * + */ + +#include "qnx6.h" + +static unsigned qnx6_lfile_checksum(char *name, unsigned size) +{ + unsigned crc = 0; + char *end = name + size; + while (name < end) { + crc = ((crc >> 1) + *(name++)) ^ + ((crc & 0x00000001) ? 0x80000000 : 0); + } + return crc; +} + +static struct page *qnx6_get_page(struct inode *dir, unsigned long n) +{ + struct address_space *mapping = dir->i_mapping; + struct page *page = read_mapping_page(mapping, n, NULL); + if (!IS_ERR(page)) + kmap(page); + return page; +} + +static inline unsigned long dir_pages(struct inode *inode) +{ + return (inode->i_size+PAGE_CACHE_SIZE-1)>>PAGE_CACHE_SHIFT; +} + +static unsigned last_entry(struct inode *inode, unsigned long page_nr) +{ + unsigned long last_byte = inode->i_size; + last_byte -= page_nr << PAGE_CACHE_SHIFT; + if (last_byte > PAGE_CACHE_SIZE) + last_byte = PAGE_CACHE_SIZE; + return last_byte / QNX6_DIR_ENTRY_SIZE; +} + +static struct qnx6_long_filename *qnx6_longname(struct super_block *sb, + struct qnx6_long_dir_entry *de, + struct page **p) +{ + struct qnx6_sb_info *sbi = QNX6_SB(sb); + u32 s = fs32_to_cpu(sbi, de->de_long_inode); /* in block units */ + u32 n = s >> (PAGE_CACHE_SHIFT - sb->s_blocksize_bits); /* in pages */ + /* within page */ + u32 offs = (s << sb->s_blocksize_bits) & ~PAGE_CACHE_MASK; + struct address_space *mapping = sbi->longfile->i_mapping; + struct page *page = read_mapping_page(mapping, n, NULL); + if (IS_ERR(page)) + return ERR_CAST(page); + kmap(*p = page); + return (struct qnx6_long_filename *)(page_address(page) + offs); +} + +static int qnx6_dir_longfilename(struct inode *inode, + struct qnx6_long_dir_entry *de, + void *dirent, loff_t pos, + unsigned de_inode, filldir_t filldir) +{ + struct qnx6_long_filename *lf; + struct super_block *s = inode->i_sb; + struct qnx6_sb_info *sbi = QNX6_SB(s); + struct page *page; + int lf_size; + + if (de->de_size != 0xff) { + /* error - long filename entries always have size 0xff + in direntry */ + printk(KERN_ERR "qnx6: invalid direntry size (%i).\n", + de->de_size); + return 0; + } + lf = qnx6_longname(s, de, &page); + if (IS_ERR(lf)) { + printk(KERN_ERR "qnx6:Error reading longname\n"); + return 0; + } + + lf_size = fs16_to_cpu(sbi, lf->lf_size); + + if (lf_size > QNX6_LONG_NAME_MAX) { + QNX6DEBUG((KERN_INFO "file %s\n", lf->lf_fname)); + printk(KERN_ERR "qnx6:Filename too long (%i)\n", lf_size); + qnx6_put_page(page); + return 0; + } + + /* calc & validate longfilename checksum + mmi 3g filesystem does not have that checksum */ + if (!test_opt(s, MMI_FS) && fs32_to_cpu(sbi, de->de_checksum) != + qnx6_lfile_checksum(lf->lf_fname, lf_size)) + printk(KERN_INFO "qnx6: long filename checksum error.\n"); + + QNX6DEBUG((KERN_INFO "qnx6_readdir:%.*s inode:%u\n", + lf_size, lf->lf_fname, de_inode)); + if (filldir(dirent, lf->lf_fname, lf_size, pos, de_inode, + DT_UNKNOWN) < 0) { + qnx6_put_page(page); + return 0; + } + + qnx6_put_page(page); + /* success */ + return 1; +} + +static int qnx6_readdir(struct file *filp, void *dirent, filldir_t filldir) +{ + struct inode *inode = filp->f_path.dentry->d_inode; + struct super_block *s = inode->i_sb; + struct qnx6_sb_info *sbi = QNX6_SB(s); + loff_t pos = filp->f_pos & (QNX6_DIR_ENTRY_SIZE - 1); + unsigned long npages = dir_pages(inode); + unsigned long n = pos >> PAGE_CACHE_SHIFT; + unsigned start = (pos & ~PAGE_CACHE_MASK) / QNX6_DIR_ENTRY_SIZE; + bool done = false; + + if (filp->f_pos >= inode->i_size) + return 0; + + for ( ; !done && n < npages; n++, start = 0) { + struct page *page = qnx6_get_page(inode, n); + int limit = last_entry(inode, n); + struct qnx6_dir_entry *de; + int i = start; + + if (IS_ERR(page)) { + printk(KERN_ERR "qnx6_readdir: read failed\n"); + filp->f_pos = (n + 1) << PAGE_CACHE_SHIFT; + return PTR_ERR(page); + } + de = ((struct qnx6_dir_entry *)page_address(page)) + start; + for (; i < limit; i++, de++, pos += QNX6_DIR_ENTRY_SIZE) { + int size = de->de_size; + u32 no_inode = fs32_to_cpu(sbi, de->de_inode); + + if (!no_inode || !size) + continue; + + if (size > QNX6_SHORT_NAME_MAX) { + /* long filename detected + get the filename from long filename + structure / block */ + if (!qnx6_dir_longfilename(inode, + (struct qnx6_long_dir_entry *)de, + dirent, pos, no_inode, + filldir)) { + done = true; + break; + } + } else { + QNX6DEBUG((KERN_INFO "qnx6_readdir:%.*s" + " inode:%u\n", size, de->de_fname, + no_inode)); + if (filldir(dirent, de->de_fname, size, + pos, no_inode, DT_UNKNOWN) + < 0) { + done = true; + break; + } + } + } + qnx6_put_page(page); + } + filp->f_pos = pos; + return 0; +} + +/* + * check if the long filename is correct. + */ +static unsigned qnx6_long_match(int len, const char *name, + struct qnx6_long_dir_entry *de, struct inode *dir) +{ + struct super_block *s = dir->i_sb; + struct qnx6_sb_info *sbi = QNX6_SB(s); + struct page *page; + int thislen; + struct qnx6_long_filename *lf = qnx6_longname(s, de, &page); + + if (IS_ERR(lf)) + return 0; + + thislen = fs16_to_cpu(sbi, lf->lf_size); + if (len != thislen) { + qnx6_put_page(page); + return 0; + } + if (memcmp(name, lf->lf_fname, len) == 0) { + qnx6_put_page(page); + return fs32_to_cpu(sbi, de->de_inode); + } + qnx6_put_page(page); + return 0; +} + +/* + * check if the filename is correct. + */ +static unsigned qnx6_match(struct super_block *s, int len, const char *name, + struct qnx6_dir_entry *de) +{ + struct qnx6_sb_info *sbi = QNX6_SB(s); + if (memcmp(name, de->de_fname, len) == 0) + return fs32_to_cpu(sbi, de->de_inode); + return 0; +} + + +unsigned qnx6_find_entry(int len, struct inode *dir, const char *name, + struct page **res_page) +{ + struct super_block *s = dir->i_sb; + struct qnx6_inode_info *ei = QNX6_I(dir); + struct page *page = NULL; + unsigned long start, n; + unsigned long npages = dir_pages(dir); + unsigned ino; + struct qnx6_dir_entry *de; + struct qnx6_long_dir_entry *lde; + + *res_page = NULL; + + if (npages == 0) + return 0; + start = ei->i_dir_start_lookup; + if (start >= npages) + start = 0; + n = start; + + do { + page = qnx6_get_page(dir, n); + if (!IS_ERR(page)) { + int limit = last_entry(dir, n); + int i; + + de = (struct qnx6_dir_entry *)page_address(page); + for (i = 0; i < limit; i++, de++) { + if (len <= QNX6_SHORT_NAME_MAX) { + /* short filename */ + if (len != de->de_size) + continue; + ino = qnx6_match(s, len, name, de); + if (ino) + goto found; + } else if (de->de_size == 0xff) { + /* deal with long filename */ + lde = (struct qnx6_long_dir_entry *)de; + ino = qnx6_long_match(len, + name, lde, dir); + if (ino) + goto found; + } else + printk(KERN_ERR "qnx6: undefined " + "filename size in inode.\n"); + } + qnx6_put_page(page); + } + + if (++n >= npages) + n = 0; + } while (n != start); + return 0; + +found: + *res_page = page; + ei->i_dir_start_lookup = n; + return ino; +} + +const struct file_operations qnx6_dir_operations = { + .llseek = generic_file_llseek, + .read = generic_read_dir, + .readdir = qnx6_readdir, + .fsync = generic_file_fsync, +}; + +const struct inode_operations qnx6_dir_inode_operations = { + .lookup = qnx6_lookup, +}; diff --git a/fs/qnx6/inode.c b/fs/qnx6/inode.c new file mode 100644 index 000000000000..e44012dc5645 --- /dev/null +++ b/fs/qnx6/inode.c @@ -0,0 +1,698 @@ +/* + * QNX6 file system, Linux implementation. + * + * Version : 1.0.0 + * + * History : + * + * 01-02-2012 by Kai Bankett (chaosman@ontika.net) : first release. + * 16-02-2012 pagemap extension by Al Viro + * + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include "qnx6.h" + +static const struct super_operations qnx6_sops; + +static void qnx6_put_super(struct super_block *sb); +static struct inode *qnx6_alloc_inode(struct super_block *sb); +static void qnx6_destroy_inode(struct inode *inode); +static int qnx6_remount(struct super_block *sb, int *flags, char *data); +static int qnx6_statfs(struct dentry *dentry, struct kstatfs *buf); +static int qnx6_show_options(struct seq_file *seq, struct dentry *root); + +static const struct super_operations qnx6_sops = { + .alloc_inode = qnx6_alloc_inode, + .destroy_inode = qnx6_destroy_inode, + .put_super = qnx6_put_super, + .statfs = qnx6_statfs, + .remount_fs = qnx6_remount, + .show_options = qnx6_show_options, +}; + +static int qnx6_show_options(struct seq_file *seq, struct dentry *root) +{ + struct super_block *sb = root->d_sb; + struct qnx6_sb_info *sbi = QNX6_SB(sb); + + if (sbi->s_mount_opt & QNX6_MOUNT_MMI_FS) + seq_puts(seq, ",mmi_fs"); + return 0; +} + +static int qnx6_remount(struct super_block *sb, int *flags, char *data) +{ + *flags |= MS_RDONLY; + return 0; +} + +static unsigned qnx6_get_devblock(struct super_block *sb, __fs32 block) +{ + struct qnx6_sb_info *sbi = QNX6_SB(sb); + return fs32_to_cpu(sbi, block) + sbi->s_blks_off; +} + +static unsigned qnx6_block_map(struct inode *inode, unsigned iblock); + +static int qnx6_get_block(struct inode *inode, sector_t iblock, + struct buffer_head *bh, int create) +{ + unsigned phys; + + QNX6DEBUG((KERN_INFO "qnx6: qnx6_get_block inode=[%ld] iblock=[%ld]\n", + inode->i_ino, (unsigned long)iblock)); + + phys = qnx6_block_map(inode, iblock); + if (phys) { + /* logical block is before EOF */ + map_bh(bh, inode->i_sb, phys); + } + return 0; +} + +static int qnx6_check_blockptr(__fs32 ptr) +{ + if (ptr == ~(__fs32)0) { + printk(KERN_ERR "qnx6: hit unused blockpointer.\n"); + return 0; + } + return 1; +} + +static int qnx6_readpage(struct file *file, struct page *page) +{ + return mpage_readpage(page, qnx6_get_block); +} + +static int qnx6_readpages(struct file *file, struct address_space *mapping, + struct list_head *pages, unsigned nr_pages) +{ + return mpage_readpages(mapping, pages, nr_pages, qnx6_get_block); +} + +/* + * returns the block number for the no-th element in the tree + * inodebits requred as there are multiple inodes in one inode block + */ +static unsigned qnx6_block_map(struct inode *inode, unsigned no) +{ + struct super_block *s = inode->i_sb; + struct qnx6_sb_info *sbi = QNX6_SB(s); + struct qnx6_inode_info *ei = QNX6_I(inode); + unsigned block = 0; + struct buffer_head *bh; + __fs32 ptr; + int levelptr; + int ptrbits = sbi->s_ptrbits; + int bitdelta; + u32 mask = (1 << ptrbits) - 1; + int depth = ei->di_filelevels; + int i; + + bitdelta = ptrbits * depth; + levelptr = no >> bitdelta; + + if (levelptr > QNX6_NO_DIRECT_POINTERS - 1) { + printk(KERN_ERR "qnx6:Requested file block number (%u) too big.", + no); + return 0; + } + + block = qnx6_get_devblock(s, ei->di_block_ptr[levelptr]); + + for (i = 0; i < depth; i++) { + bh = sb_bread(s, block); + if (!bh) { + printk(KERN_ERR "qnx6:Error reading block (%u)\n", + block); + return 0; + } + bitdelta -= ptrbits; + levelptr = (no >> bitdelta) & mask; + ptr = ((__fs32 *)bh->b_data)[levelptr]; + + if (!qnx6_check_blockptr(ptr)) + return 0; + + block = qnx6_get_devblock(s, ptr); + brelse(bh); + } + return block; +} + +static int qnx6_statfs(struct dentry *dentry, struct kstatfs *buf) +{ + struct super_block *sb = dentry->d_sb; + struct qnx6_sb_info *sbi = QNX6_SB(sb); + u64 id = huge_encode_dev(sb->s_bdev->bd_dev); + + buf->f_type = sb->s_magic; + buf->f_bsize = sb->s_blocksize; + buf->f_blocks = fs32_to_cpu(sbi, sbi->sb->sb_num_blocks); + buf->f_bfree = fs32_to_cpu(sbi, sbi->sb->sb_free_blocks); + buf->f_files = fs32_to_cpu(sbi, sbi->sb->sb_num_inodes); + buf->f_ffree = fs32_to_cpu(sbi, sbi->sb->sb_free_inodes); + buf->f_bavail = buf->f_bfree; + buf->f_namelen = QNX6_LONG_NAME_MAX; + buf->f_fsid.val[0] = (u32)id; + buf->f_fsid.val[1] = (u32)(id >> 32); + + return 0; +} + +/* + * Check the root directory of the filesystem to make sure + * it really _is_ a qnx6 filesystem, and to check the size + * of the directory entry. + */ +static const char *qnx6_checkroot(struct super_block *s) +{ + static char match_root[2][3] = {".\0\0", "..\0"}; + int i, error = 0; + struct qnx6_dir_entry *dir_entry; + struct inode *root = s->s_root->d_inode; + struct address_space *mapping = root->i_mapping; + struct page *page = read_mapping_page(mapping, 0, NULL); + if (IS_ERR(page)) + return "error reading root directory"; + kmap(page); + dir_entry = page_address(page); + for (i = 0; i < 2; i++) { + /* maximum 3 bytes - due to match_root limitation */ + if (strncmp(dir_entry[i].de_fname, match_root[i], 3)) + error = 1; + } + qnx6_put_page(page); + if (error) + return "error reading root directory."; + return NULL; +} + +#ifdef CONFIG_QNX6FS_DEBUG +void qnx6_superblock_debug(struct qnx6_super_block *sb, struct super_block *s) +{ + struct qnx6_sb_info *sbi = QNX6_SB(s); + + QNX6DEBUG((KERN_INFO "magic: %08x\n", + fs32_to_cpu(sbi, sb->sb_magic))); + QNX6DEBUG((KERN_INFO "checksum: %08x\n", + fs32_to_cpu(sbi, sb->sb_checksum))); + QNX6DEBUG((KERN_INFO "serial: %llx\n", + fs64_to_cpu(sbi, sb->sb_serial))); + QNX6DEBUG((KERN_INFO "flags: %08x\n", + fs32_to_cpu(sbi, sb->sb_flags))); + QNX6DEBUG((KERN_INFO "blocksize: %08x\n", + fs32_to_cpu(sbi, sb->sb_blocksize))); + QNX6DEBUG((KERN_INFO "num_inodes: %08x\n", + fs32_to_cpu(sbi, sb->sb_num_inodes))); + QNX6DEBUG((KERN_INFO "free_inodes: %08x\n", + fs32_to_cpu(sbi, sb->sb_free_inodes))); + QNX6DEBUG((KERN_INFO "num_blocks: %08x\n", + fs32_to_cpu(sbi, sb->sb_num_blocks))); + QNX6DEBUG((KERN_INFO "free_blocks: %08x\n", + fs32_to_cpu(sbi, sb->sb_free_blocks))); + QNX6DEBUG((KERN_INFO "inode_levels: %02x\n", + sb->Inode.levels)); +} +#endif + +enum { + Opt_mmifs, + Opt_err +}; + +static const match_table_t tokens = { + {Opt_mmifs, "mmi_fs"}, + {Opt_err, NULL} +}; + +static int qnx6_parse_options(char *options, struct super_block *sb) +{ + char *p; + struct qnx6_sb_info *sbi = QNX6_SB(sb); + substring_t args[MAX_OPT_ARGS]; + + if (!options) + return 1; + + while ((p = strsep(&options, ",")) != NULL) { + int token; + if (!*p) + continue; + + token = match_token(p, tokens, args); + switch (token) { + case Opt_mmifs: + set_opt(sbi->s_mount_opt, MMI_FS); + break; + default: + return 0; + } + } + return 1; +} + +static struct buffer_head *qnx6_check_first_superblock(struct super_block *s, + int offset, int silent) +{ + struct qnx6_sb_info *sbi = QNX6_SB(s); + struct buffer_head *bh; + struct qnx6_super_block *sb; + + /* Check the superblock signatures + start with the first superblock */ + bh = sb_bread(s, offset); + if (!bh) { + printk(KERN_ERR "qnx6: unable to read the first superblock\n"); + return NULL; + } + sb = (struct qnx6_super_block *)bh->b_data; + if (fs32_to_cpu(sbi, sb->sb_magic) != QNX6_SUPER_MAGIC) { + sbi->s_bytesex = BYTESEX_BE; + if (fs32_to_cpu(sbi, sb->sb_magic) == QNX6_SUPER_MAGIC) { + /* we got a big endian fs */ + QNX6DEBUG((KERN_INFO "qnx6: fs got different" + " endianess.\n")); + return bh; + } else + sbi->s_bytesex = BYTESEX_LE; + if (!silent) { + if (offset == 0) { + printk(KERN_ERR "qnx6: wrong signature (magic)" + " in superblock #1.\n"); + } else { + printk(KERN_INFO "qnx6: wrong signature (magic)" + " at position (0x%lx) - will try" + " alternative position (0x0000).\n", + offset * s->s_blocksize); + } + } + brelse(bh); + return NULL; + } + return bh; +} + +static struct inode *qnx6_private_inode(struct super_block *s, + struct qnx6_root_node *p); + +static int qnx6_fill_super(struct super_block *s, void *data, int silent) +{ + struct buffer_head *bh1 = NULL, *bh2 = NULL; + struct qnx6_super_block *sb1 = NULL, *sb2 = NULL; + struct qnx6_sb_info *sbi; + struct inode *root; + const char *errmsg; + struct qnx6_sb_info *qs; + int ret = -EINVAL; + u64 offset; + int bootblock_offset = QNX6_BOOTBLOCK_SIZE; + + qs = kzalloc(sizeof(struct qnx6_sb_info), GFP_KERNEL); + if (!qs) + return -ENOMEM; + s->s_fs_info = qs; + + /* Superblock always is 512 Byte long */ + if (!sb_set_blocksize(s, QNX6_SUPERBLOCK_SIZE)) { + printk(KERN_ERR "qnx6: unable to set blocksize\n"); + goto outnobh; + } + + /* parse the mount-options */ + if (!qnx6_parse_options((char *) data, s)) { + printk(KERN_ERR "qnx6: invalid mount options.\n"); + goto outnobh; + } + if (test_opt(s, MMI_FS)) { + sb1 = qnx6_mmi_fill_super(s, silent); + if (sb1) + goto mmi_success; + else + goto outnobh; + } + sbi = QNX6_SB(s); + sbi->s_bytesex = BYTESEX_LE; + /* Check the superblock signatures + start with the first superblock */ + bh1 = qnx6_check_first_superblock(s, + bootblock_offset / QNX6_SUPERBLOCK_SIZE, silent); + if (!bh1) { + /* try again without bootblock offset */ + bh1 = qnx6_check_first_superblock(s, 0, silent); + if (!bh1) { + printk(KERN_ERR "qnx6: unable to read the first superblock\n"); + goto outnobh; + } + /* seems that no bootblock at partition start */ + bootblock_offset = 0; + } + sb1 = (struct qnx6_super_block *)bh1->b_data; + +#ifdef CONFIG_QNX6FS_DEBUG + qnx6_superblock_debug(sb1, s); +#endif + + /* checksum check - start at byte 8 and end at byte 512 */ + if (fs32_to_cpu(sbi, sb1->sb_checksum) != + crc32_be(0, (char *)(bh1->b_data + 8), 504)) { + printk(KERN_ERR "qnx6: superblock #1 checksum error\n"); + goto out; + } + + /* set new blocksize */ + if (!sb_set_blocksize(s, fs32_to_cpu(sbi, sb1->sb_blocksize))) { + printk(KERN_ERR "qnx6: unable to set blocksize\n"); + goto out; + } + /* blocksize invalidates bh - pull it back in */ + brelse(bh1); + bh1 = sb_bread(s, bootblock_offset >> s->s_blocksize_bits); + if (!bh1) + goto outnobh; + sb1 = (struct qnx6_super_block *)bh1->b_data; + + /* calculate second superblock blocknumber */ + offset = fs32_to_cpu(sbi, sb1->sb_num_blocks) + + (bootblock_offset >> s->s_blocksize_bits) + + (QNX6_SUPERBLOCK_AREA >> s->s_blocksize_bits); + + /* set bootblock offset */ + sbi->s_blks_off = (bootblock_offset >> s->s_blocksize_bits) + + (QNX6_SUPERBLOCK_AREA >> s->s_blocksize_bits); + + /* next the second superblock */ + bh2 = sb_bread(s, offset); + if (!bh2) { + printk(KERN_ERR "qnx6: unable to read the second superblock\n"); + goto out; + } + sb2 = (struct qnx6_super_block *)bh2->b_data; + if (fs32_to_cpu(sbi, sb2->sb_magic) != QNX6_SUPER_MAGIC) { + if (!silent) + printk(KERN_ERR "qnx6: wrong signature (magic)" + " in superblock #2.\n"); + goto out; + } + + /* checksum check - start at byte 8 and end at byte 512 */ + if (fs32_to_cpu(sbi, sb2->sb_checksum) != + crc32_be(0, (char *)(bh2->b_data + 8), 504)) { + printk(KERN_ERR "qnx6: superblock #2 checksum error\n"); + goto out; + } + + if (fs64_to_cpu(sbi, sb1->sb_serial) >= + fs64_to_cpu(sbi, sb2->sb_serial)) { + /* superblock #1 active */ + sbi->sb_buf = bh1; + sbi->sb = (struct qnx6_super_block *)bh1->b_data; + brelse(bh2); + printk(KERN_INFO "qnx6: superblock #1 active\n"); + } else { + /* superblock #2 active */ + sbi->sb_buf = bh2; + sbi->sb = (struct qnx6_super_block *)bh2->b_data; + brelse(bh1); + printk(KERN_INFO "qnx6: superblock #2 active\n"); + } +mmi_success: + /* sanity check - limit maximum indirect pointer levels */ + if (sb1->Inode.levels > QNX6_PTR_MAX_LEVELS) { + printk(KERN_ERR "qnx6: too many inode levels (max %i, sb %i)\n", + QNX6_PTR_MAX_LEVELS, sb1->Inode.levels); + goto out; + } + if (sb1->Longfile.levels > QNX6_PTR_MAX_LEVELS) { + printk(KERN_ERR "qnx6: too many longfilename levels" + " (max %i, sb %i)\n", + QNX6_PTR_MAX_LEVELS, sb1->Longfile.levels); + goto out; + } + s->s_op = &qnx6_sops; + s->s_magic = QNX6_SUPER_MAGIC; + s->s_flags |= MS_RDONLY; /* Yup, read-only yet */ + + /* ease the later tree level calculations */ + sbi = QNX6_SB(s); + sbi->s_ptrbits = ilog2(s->s_blocksize / 4); + sbi->inodes = qnx6_private_inode(s, &sb1->Inode); + if (!sbi->inodes) + goto out; + sbi->longfile = qnx6_private_inode(s, &sb1->Longfile); + if (!sbi->longfile) + goto out1; + + /* prefetch root inode */ + root = qnx6_iget(s, QNX6_ROOT_INO); + if (IS_ERR(root)) { + printk(KERN_ERR "qnx6: get inode failed\n"); + ret = PTR_ERR(root); + goto out2; + } + + ret = -ENOMEM; + s->s_root = d_make_root(root); + if (!s->s_root) + goto out2; + + ret = -EINVAL; + errmsg = qnx6_checkroot(s); + if (errmsg != NULL) { + if (!silent) + printk(KERN_ERR "qnx6: %s\n", errmsg); + goto out3; + } + return 0; + +out3: + dput(s->s_root); + s->s_root = NULL; +out2: + iput(sbi->longfile); +out1: + iput(sbi->inodes); +out: + if (bh1) + brelse(bh1); + if (bh2) + brelse(bh2); +outnobh: + kfree(qs); + s->s_fs_info = NULL; + return ret; +} + +static void qnx6_put_super(struct super_block *sb) +{ + struct qnx6_sb_info *qs = QNX6_SB(sb); + brelse(qs->sb_buf); + iput(qs->longfile); + iput(qs->inodes); + kfree(qs); + sb->s_fs_info = NULL; + return; +} + +static sector_t qnx6_bmap(struct address_space *mapping, sector_t block) +{ + return generic_block_bmap(mapping, block, qnx6_get_block); +} +static const struct address_space_operations qnx6_aops = { + .readpage = qnx6_readpage, + .readpages = qnx6_readpages, + .bmap = qnx6_bmap +}; + +static struct inode *qnx6_private_inode(struct super_block *s, + struct qnx6_root_node *p) +{ + struct inode *inode = new_inode(s); + if (inode) { + struct qnx6_inode_info *ei = QNX6_I(inode); + struct qnx6_sb_info *sbi = QNX6_SB(s); + inode->i_size = fs64_to_cpu(sbi, p->size); + memcpy(ei->di_block_ptr, p->ptr, sizeof(p->ptr)); + ei->di_filelevels = p->levels; + inode->i_mode = S_IFREG | S_IRUSR; /* probably wrong */ + inode->i_mapping->a_ops = &qnx6_aops; + } + return inode; +} + +struct inode *qnx6_iget(struct super_block *sb, unsigned ino) +{ + struct qnx6_sb_info *sbi = QNX6_SB(sb); + struct qnx6_inode_entry *raw_inode; + struct inode *inode; + struct qnx6_inode_info *ei; + struct address_space *mapping; + struct page *page; + u32 n, offs; + + inode = iget_locked(sb, ino); + if (!inode) + return ERR_PTR(-ENOMEM); + if (!(inode->i_state & I_NEW)) + return inode; + + ei = QNX6_I(inode); + + inode->i_mode = 0; + + if (ino == 0) { + printk(KERN_ERR "qnx6: bad inode number on dev %s: %u is " + "out of range\n", + sb->s_id, ino); + iget_failed(inode); + return ERR_PTR(-EIO); + } + n = (ino - 1) >> (PAGE_CACHE_SHIFT - QNX6_INODE_SIZE_BITS); + offs = (ino - 1) & (~PAGE_CACHE_MASK >> QNX6_INODE_SIZE_BITS); + mapping = sbi->inodes->i_mapping; + page = read_mapping_page(mapping, n, NULL); + if (IS_ERR(page)) { + printk(KERN_ERR "qnx6: major problem: unable to read inode from " + "dev %s\n", sb->s_id); + iget_failed(inode); + return ERR_CAST(page); + } + kmap(page); + raw_inode = ((struct qnx6_inode_entry *)page_address(page)) + offs; + + inode->i_mode = fs16_to_cpu(sbi, raw_inode->di_mode); + inode->i_uid = (uid_t)fs32_to_cpu(sbi, raw_inode->di_uid); + inode->i_gid = (gid_t)fs32_to_cpu(sbi, raw_inode->di_gid); + inode->i_size = fs64_to_cpu(sbi, raw_inode->di_size); + inode->i_mtime.tv_sec = fs32_to_cpu(sbi, raw_inode->di_mtime); + inode->i_mtime.tv_nsec = 0; + inode->i_atime.tv_sec = fs32_to_cpu(sbi, raw_inode->di_atime); + inode->i_atime.tv_nsec = 0; + inode->i_ctime.tv_sec = fs32_to_cpu(sbi, raw_inode->di_ctime); + inode->i_ctime.tv_nsec = 0; + + /* calc blocks based on 512 byte blocksize */ + inode->i_blocks = (inode->i_size + 511) >> 9; + + memcpy(&ei->di_block_ptr, &raw_inode->di_block_ptr, + sizeof(raw_inode->di_block_ptr)); + ei->di_filelevels = raw_inode->di_filelevels; + + if (S_ISREG(inode->i_mode)) { + inode->i_fop = &generic_ro_fops; + inode->i_mapping->a_ops = &qnx6_aops; + } else if (S_ISDIR(inode->i_mode)) { + inode->i_op = &qnx6_dir_inode_operations; + inode->i_fop = &qnx6_dir_operations; + inode->i_mapping->a_ops = &qnx6_aops; + } else if (S_ISLNK(inode->i_mode)) { + inode->i_op = &page_symlink_inode_operations; + inode->i_mapping->a_ops = &qnx6_aops; + } else + init_special_inode(inode, inode->i_mode, 0); + qnx6_put_page(page); + unlock_new_inode(inode); + return inode; +} + +static struct kmem_cache *qnx6_inode_cachep; + +static struct inode *qnx6_alloc_inode(struct super_block *sb) +{ + struct qnx6_inode_info *ei; + ei = kmem_cache_alloc(qnx6_inode_cachep, GFP_KERNEL); + if (!ei) + return NULL; + return &ei->vfs_inode; +} + +static void qnx6_i_callback(struct rcu_head *head) +{ + struct inode *inode = container_of(head, struct inode, i_rcu); + INIT_LIST_HEAD(&inode->i_dentry); + kmem_cache_free(qnx6_inode_cachep, QNX6_I(inode)); +} + +static void qnx6_destroy_inode(struct inode *inode) +{ + call_rcu(&inode->i_rcu, qnx6_i_callback); +} + +static void init_once(void *foo) +{ + struct qnx6_inode_info *ei = (struct qnx6_inode_info *) foo; + + inode_init_once(&ei->vfs_inode); +} + +static int init_inodecache(void) +{ + qnx6_inode_cachep = kmem_cache_create("qnx6_inode_cache", + sizeof(struct qnx6_inode_info), + 0, (SLAB_RECLAIM_ACCOUNT| + SLAB_MEM_SPREAD), + init_once); + if (!qnx6_inode_cachep) + return -ENOMEM; + return 0; +} + +static void destroy_inodecache(void) +{ + kmem_cache_destroy(qnx6_inode_cachep); +} + +static struct dentry *qnx6_mount(struct file_system_type *fs_type, + int flags, const char *dev_name, void *data) +{ + return mount_bdev(fs_type, flags, dev_name, data, qnx6_fill_super); +} + +static struct file_system_type qnx6_fs_type = { + .owner = THIS_MODULE, + .name = "qnx6", + .mount = qnx6_mount, + .kill_sb = kill_block_super, + .fs_flags = FS_REQUIRES_DEV, +}; + +static int __init init_qnx6_fs(void) +{ + int err; + + err = init_inodecache(); + if (err) + return err; + + err = register_filesystem(&qnx6_fs_type); + if (err) { + destroy_inodecache(); + return err; + } + + printk(KERN_INFO "QNX6 filesystem 1.0.0 registered.\n"); + return 0; +} + +static void __exit exit_qnx6_fs(void) +{ + unregister_filesystem(&qnx6_fs_type); + destroy_inodecache(); +} + +module_init(init_qnx6_fs) +module_exit(exit_qnx6_fs) +MODULE_LICENSE("GPL"); diff --git a/fs/qnx6/namei.c b/fs/qnx6/namei.c new file mode 100644 index 000000000000..8a97289e04ad --- /dev/null +++ b/fs/qnx6/namei.c @@ -0,0 +1,42 @@ +/* + * QNX6 file system, Linux implementation. + * + * Version : 1.0.0 + * + * History : + * + * 01-02-2012 by Kai Bankett (chaosman@ontika.net) : first release. + * 16-02-2012 pagemap extension by Al Viro + * + */ + +#include "qnx6.h" + +struct dentry *qnx6_lookup(struct inode *dir, struct dentry *dentry, + struct nameidata *nd) +{ + unsigned ino; + struct page *page; + struct inode *foundinode = NULL; + const char *name = dentry->d_name.name; + int len = dentry->d_name.len; + + if (len > QNX6_LONG_NAME_MAX) + return ERR_PTR(-ENAMETOOLONG); + + ino = qnx6_find_entry(len, dir, name, &page); + if (ino) { + foundinode = qnx6_iget(dir->i_sb, ino); + qnx6_put_page(page); + if (IS_ERR(foundinode)) { + QNX6DEBUG((KERN_ERR "qnx6: lookup->iget -> " + " error %ld\n", PTR_ERR(foundinode))); + return ERR_CAST(foundinode); + } + } else { + QNX6DEBUG((KERN_INFO "qnx6_lookup: not found %s\n", name)); + return NULL; + } + d_add(dentry, foundinode); + return NULL; +} diff --git a/fs/qnx6/qnx6.h b/fs/qnx6/qnx6.h new file mode 100644 index 000000000000..6c5e02a0b6a8 --- /dev/null +++ b/fs/qnx6/qnx6.h @@ -0,0 +1,135 @@ +/* + * QNX6 file system, Linux implementation. + * + * Version : 1.0.0 + * + * History : + * + * 01-02-2012 by Kai Bankett (chaosman@ontika.net) : first release. + * 16-02-2012 page map extension by Al Viro + * + */ + +#include +#include + +typedef __u16 __bitwise __fs16; +typedef __u32 __bitwise __fs32; +typedef __u64 __bitwise __fs64; + +#include + +#ifdef CONFIG_QNX6FS_DEBUG +#define QNX6DEBUG(X) printk X +#else +#define QNX6DEBUG(X) (void) 0 +#endif + +struct qnx6_sb_info { + struct buffer_head *sb_buf; /* superblock buffer */ + struct qnx6_super_block *sb; /* our superblock */ + int s_blks_off; /* blkoffset fs-startpoint */ + int s_ptrbits; /* indirect pointer bitfield */ + unsigned long s_mount_opt; /* all mount options */ + int s_bytesex; /* holds endianess info */ + struct inode * inodes; + struct inode * longfile; +}; + +struct qnx6_inode_info { + __fs32 di_block_ptr[QNX6_NO_DIRECT_POINTERS]; + __u8 di_filelevels; + __u32 i_dir_start_lookup; + struct inode vfs_inode; +}; + +extern struct inode *qnx6_iget(struct super_block *sb, unsigned ino); +extern struct dentry *qnx6_lookup(struct inode *dir, struct dentry *dentry, + struct nameidata *nd); + +#ifdef CONFIG_QNX6FS_DEBUG +extern void qnx6_superblock_debug(struct qnx6_super_block *, + struct super_block *); +#endif + +extern const struct inode_operations qnx6_dir_inode_operations; +extern const struct file_operations qnx6_dir_operations; + +static inline struct qnx6_sb_info *QNX6_SB(struct super_block *sb) +{ + return sb->s_fs_info; +} + +static inline struct qnx6_inode_info *QNX6_I(struct inode *inode) +{ + return container_of(inode, struct qnx6_inode_info, vfs_inode); +} + +#define clear_opt(o, opt) (o &= ~(QNX6_MOUNT_##opt)) +#define set_opt(o, opt) (o |= (QNX6_MOUNT_##opt)) +#define test_opt(sb, opt) (QNX6_SB(sb)->s_mount_opt & \ + QNX6_MOUNT_##opt) +enum { + BYTESEX_LE, + BYTESEX_BE, +}; + +static inline __u64 fs64_to_cpu(struct qnx6_sb_info *sbi, __fs64 n) +{ + if (sbi->s_bytesex == BYTESEX_LE) + return le64_to_cpu((__force __le64)n); + else + return be64_to_cpu((__force __be64)n); +} + +static inline __fs64 cpu_to_fs64(struct qnx6_sb_info *sbi, __u64 n) +{ + if (sbi->s_bytesex == BYTESEX_LE) + return (__force __fs64)cpu_to_le64(n); + else + return (__force __fs64)cpu_to_be64(n); +} + +static inline __u32 fs32_to_cpu(struct qnx6_sb_info *sbi, __fs32 n) +{ + if (sbi->s_bytesex == BYTESEX_LE) + return le32_to_cpu((__force __le32)n); + else + return be32_to_cpu((__force __be32)n); +} + +static inline __fs32 cpu_to_fs32(struct qnx6_sb_info *sbi, __u32 n) +{ + if (sbi->s_bytesex == BYTESEX_LE) + return (__force __fs32)cpu_to_le32(n); + else + return (__force __fs32)cpu_to_be32(n); +} + +static inline __u16 fs16_to_cpu(struct qnx6_sb_info *sbi, __fs16 n) +{ + if (sbi->s_bytesex == BYTESEX_LE) + return le16_to_cpu((__force __le16)n); + else + return be16_to_cpu((__force __be16)n); +} + +static inline __fs16 cpu_to_fs16(struct qnx6_sb_info *sbi, __u16 n) +{ + if (sbi->s_bytesex == BYTESEX_LE) + return (__force __fs16)cpu_to_le16(n); + else + return (__force __fs16)cpu_to_be16(n); +} + +extern struct qnx6_super_block *qnx6_mmi_fill_super(struct super_block *s, + int silent); + +static inline void qnx6_put_page(struct page *page) +{ + kunmap(page); + page_cache_release(page); +} + +extern unsigned qnx6_find_entry(int len, struct inode *dir, const char *name, + struct page **res_page); diff --git a/fs/qnx6/super_mmi.c b/fs/qnx6/super_mmi.c new file mode 100644 index 000000000000..29c32cba62d6 --- /dev/null +++ b/fs/qnx6/super_mmi.c @@ -0,0 +1,150 @@ +/* + * QNX6 file system, Linux implementation. + * + * Version : 1.0.0 + * + * History : + * + * 01-02-2012 by Kai Bankett (chaosman@ontika.net) : first release. + * + */ + +#include +#include +#include +#include "qnx6.h" + +static void qnx6_mmi_copy_sb(struct qnx6_super_block *qsb, + struct qnx6_mmi_super_block *sb) +{ + qsb->sb_magic = sb->sb_magic; + qsb->sb_checksum = sb->sb_checksum; + qsb->sb_serial = sb->sb_serial; + qsb->sb_blocksize = sb->sb_blocksize; + qsb->sb_num_inodes = sb->sb_num_inodes; + qsb->sb_free_inodes = sb->sb_free_inodes; + qsb->sb_num_blocks = sb->sb_num_blocks; + qsb->sb_free_blocks = sb->sb_free_blocks; + + /* the rest of the superblock is the same */ + memcpy(&qsb->Inode, &sb->Inode, sizeof(sb->Inode)); + memcpy(&qsb->Bitmap, &sb->Bitmap, sizeof(sb->Bitmap)); + memcpy(&qsb->Longfile, &sb->Longfile, sizeof(sb->Longfile)); +} + +struct qnx6_super_block *qnx6_mmi_fill_super(struct super_block *s, int silent) +{ + struct buffer_head *bh1, *bh2 = NULL; + struct qnx6_mmi_super_block *sb1, *sb2; + struct qnx6_super_block *qsb = NULL; + struct qnx6_sb_info *sbi; + __u64 offset; + + /* Check the superblock signatures + start with the first superblock */ + bh1 = sb_bread(s, 0); + if (!bh1) { + printk(KERN_ERR "qnx6: Unable to read first mmi superblock\n"); + return NULL; + } + sb1 = (struct qnx6_mmi_super_block *)bh1->b_data; + sbi = QNX6_SB(s); + if (fs32_to_cpu(sbi, sb1->sb_magic) != QNX6_SUPER_MAGIC) { + if (!silent) { + printk(KERN_ERR "qnx6: wrong signature (magic) in" + " superblock #1.\n"); + goto out; + } + } + + /* checksum check - start at byte 8 and end at byte 512 */ + if (fs32_to_cpu(sbi, sb1->sb_checksum) != + crc32_be(0, (char *)(bh1->b_data + 8), 504)) { + printk(KERN_ERR "qnx6: superblock #1 checksum error\n"); + goto out; + } + + /* calculate second superblock blocknumber */ + offset = fs32_to_cpu(sbi, sb1->sb_num_blocks) + QNX6_SUPERBLOCK_AREA / + fs32_to_cpu(sbi, sb1->sb_blocksize); + + /* set new blocksize */ + if (!sb_set_blocksize(s, fs32_to_cpu(sbi, sb1->sb_blocksize))) { + printk(KERN_ERR "qnx6: unable to set blocksize\n"); + goto out; + } + /* blocksize invalidates bh - pull it back in */ + brelse(bh1); + bh1 = sb_bread(s, 0); + if (!bh1) + goto out; + sb1 = (struct qnx6_mmi_super_block *)bh1->b_data; + + /* read second superblock */ + bh2 = sb_bread(s, offset); + if (!bh2) { + printk(KERN_ERR "qnx6: unable to read the second superblock\n"); + goto out; + } + sb2 = (struct qnx6_mmi_super_block *)bh2->b_data; + if (fs32_to_cpu(sbi, sb2->sb_magic) != QNX6_SUPER_MAGIC) { + if (!silent) + printk(KERN_ERR "qnx6: wrong signature (magic) in" + " superblock #2.\n"); + goto out; + } + + /* checksum check - start at byte 8 and end at byte 512 */ + if (fs32_to_cpu(sbi, sb2->sb_checksum) + != crc32_be(0, (char *)(bh2->b_data + 8), 504)) { + printk(KERN_ERR "qnx6: superblock #1 checksum error\n"); + goto out; + } + + qsb = kmalloc(sizeof(*qsb), GFP_KERNEL); + if (!qsb) { + printk(KERN_ERR "qnx6: unable to allocate memory.\n"); + goto out; + } + + if (fs64_to_cpu(sbi, sb1->sb_serial) > + fs64_to_cpu(sbi, sb2->sb_serial)) { + /* superblock #1 active */ + qnx6_mmi_copy_sb(qsb, sb1); +#ifdef CONFIG_QNX6FS_DEBUG + qnx6_superblock_debug(qsb, s); +#endif + memcpy(bh1->b_data, qsb, sizeof(struct qnx6_super_block)); + + sbi->sb_buf = bh1; + sbi->sb = (struct qnx6_super_block *)bh1->b_data; + brelse(bh2); + printk(KERN_INFO "qnx6: superblock #1 active\n"); + } else { + /* superblock #2 active */ + qnx6_mmi_copy_sb(qsb, sb2); +#ifdef CONFIG_QNX6FS_DEBUG + qnx6_superblock_debug(qsb, s); +#endif + memcpy(bh2->b_data, qsb, sizeof(struct qnx6_super_block)); + + sbi->sb_buf = bh2; + sbi->sb = (struct qnx6_super_block *)bh2->b_data; + brelse(bh1); + printk(KERN_INFO "qnx6: superblock #2 active\n"); + } + kfree(qsb); + + /* offset for mmi_fs is just SUPERBLOCK_AREA bytes */ + sbi->s_blks_off = QNX6_SUPERBLOCK_AREA / s->s_blocksize; + + /* success */ + return sbi->sb; + +out: + if (bh1 != NULL) + brelse(bh1); + if (bh2 != NULL) + brelse(bh2); + return NULL; +} diff --git a/include/linux/magic.h b/include/linux/magic.h index 2d4beab0d5b7..b7ed4759dbb2 100644 --- a/include/linux/magic.h +++ b/include/linux/magic.h @@ -42,6 +42,7 @@ #define OPENPROM_SUPER_MAGIC 0x9fa1 #define PROC_SUPER_MAGIC 0x9fa0 #define QNX4_SUPER_MAGIC 0x002f /* qnx4 fs detection */ +#define QNX6_SUPER_MAGIC 0x68191122 /* qnx6 fs detection */ #define REISERFS_SUPER_MAGIC 0x52654973 /* used by gcc */ /* used by file system utilities that diff --git a/include/linux/qnx6_fs.h b/include/linux/qnx6_fs.h new file mode 100644 index 000000000000..26049eab9010 --- /dev/null +++ b/include/linux/qnx6_fs.h @@ -0,0 +1,134 @@ +/* + * Name : qnx6_fs.h + * Author : Kai Bankett + * Function : qnx6 global filesystem definitions + * History : 17-01-2012 created + */ +#ifndef _LINUX_QNX6_FS_H +#define _LINUX_QNX6_FS_H + +#include +#include + +#define QNX6_ROOT_INO 1 + +/* for di_status */ +#define QNX6_FILE_DIRECTORY 0x01 +#define QNX6_FILE_DELETED 0x02 +#define QNX6_FILE_NORMAL 0x03 + +#define QNX6_SUPERBLOCK_SIZE 0x200 /* superblock always is 512 bytes */ +#define QNX6_SUPERBLOCK_AREA 0x1000 /* area reserved for superblock */ +#define QNX6_BOOTBLOCK_SIZE 0x2000 /* heading bootblock area */ +#define QNX6_DIR_ENTRY_SIZE 0x20 /* dir entry size of 32 bytes */ +#define QNX6_INODE_SIZE 0x80 /* each inode is 128 bytes */ +#define QNX6_INODE_SIZE_BITS 7 /* inode entry size shift */ + +#define QNX6_NO_DIRECT_POINTERS 16 /* 16 blockptrs in sbl/inode */ +#define QNX6_PTR_MAX_LEVELS 5 /* maximum indirect levels */ + +/* for filenames */ +#define QNX6_SHORT_NAME_MAX 27 +#define QNX6_LONG_NAME_MAX 510 + +/* list of mount options */ +#define QNX6_MOUNT_MMI_FS 0x010000 /* mount as Audi MMI 3G fs */ + +/* + * This is the original qnx6 inode layout on disk. + * Each inode is 128 byte long. + */ +struct qnx6_inode_entry { + __fs64 di_size; + __fs32 di_uid; + __fs32 di_gid; + __fs32 di_ftime; + __fs32 di_mtime; + __fs32 di_atime; + __fs32 di_ctime; + __fs16 di_mode; + __fs16 di_ext_mode; + __fs32 di_block_ptr[QNX6_NO_DIRECT_POINTERS]; + __u8 di_filelevels; + __u8 di_status; + __u8 di_unknown2[2]; + __fs32 di_zero2[6]; +}; + +/* + * Each directory entry is maximum 32 bytes long. + * If more characters or special characters required it is stored + * in the longfilenames structure. + */ +struct qnx6_dir_entry { + __fs32 de_inode; + __u8 de_size; + char de_fname[QNX6_SHORT_NAME_MAX]; +}; + +/* + * Longfilename direntries have a different structure + */ +struct qnx6_long_dir_entry { + __fs32 de_inode; + __u8 de_size; + __u8 de_unknown[3]; + __fs32 de_long_inode; + __fs32 de_checksum; +}; + +struct qnx6_long_filename { + __fs16 lf_size; + __u8 lf_fname[QNX6_LONG_NAME_MAX]; +}; + +struct qnx6_root_node { + __fs64 size; + __fs32 ptr[QNX6_NO_DIRECT_POINTERS]; + __u8 levels; + __u8 mode; + __u8 spare[6]; +}; + +struct qnx6_super_block { + __fs32 sb_magic; + __fs32 sb_checksum; + __fs64 sb_serial; + __fs32 sb_ctime; /* time the fs was created */ + __fs32 sb_atime; /* last access time */ + __fs32 sb_flags; + __fs16 sb_version1; /* filesystem version information */ + __fs16 sb_version2; /* filesystem version information */ + __u8 sb_volumeid[16]; + __fs32 sb_blocksize; + __fs32 sb_num_inodes; + __fs32 sb_free_inodes; + __fs32 sb_num_blocks; + __fs32 sb_free_blocks; + __fs32 sb_allocgroup; + struct qnx6_root_node Inode; + struct qnx6_root_node Bitmap; + struct qnx6_root_node Longfile; + struct qnx6_root_node Unknown; +}; + +/* Audi MMI 3G superblock layout is different to plain qnx6 */ +struct qnx6_mmi_super_block { + __fs32 sb_magic; + __fs32 sb_checksum; + __fs64 sb_serial; + __u8 sb_spare0[12]; + __u8 sb_id[12]; + __fs32 sb_blocksize; + __fs32 sb_num_inodes; + __fs32 sb_free_inodes; + __fs32 sb_num_blocks; + __fs32 sb_free_blocks; + __u8 sb_spare1[4]; + struct qnx6_root_node Inode; + struct qnx6_root_node Bitmap; + struct qnx6_root_node Longfile; + struct qnx6_root_node Unknown; +}; + +#endif From e23754f880f10124f0a2848f9d17e361a295378e Mon Sep 17 00:00:00 2001 From: Al Viro Date: Tue, 6 Mar 2012 14:33:22 -0500 Subject: [PATCH 18/64] aio: don't bother with async freeing on failure in ioctx_alloc() Signed-off-by: Al Viro --- fs/aio.c | 12 +++++------- 1 file changed, 5 insertions(+), 7 deletions(-) diff --git a/fs/aio.c b/fs/aio.c index b9d64d89a043..d09b56090aa5 100644 --- a/fs/aio.c +++ b/fs/aio.c @@ -248,6 +248,7 @@ static struct kioctx *ioctx_alloc(unsigned nr_events) struct mm_struct *mm; struct kioctx *ctx; int did_sync = 0; + int err = -ENOMEM; /* Prevent overflows */ if ((nr_events > (0x10000000U / sizeof(struct io_event))) || @@ -310,16 +311,13 @@ static struct kioctx *ioctx_alloc(unsigned nr_events) return ctx; out_cleanup: - __put_ioctx(ctx); - return ERR_PTR(-EAGAIN); - + err = -EAGAIN; + aio_free_ring(ctx); out_freectx: mmdrop(mm); kmem_cache_free(kioctx_cachep, ctx); - ctx = ERR_PTR(-ENOMEM); - - dprintk("aio: error allocating ioctx %p\n", ctx); - return ctx; + dprintk("aio: error allocating ioctx %d\n", err); + return ERR_PTR(err); } /* aio_cancel_all From 2dd542b7aeb1c222273cf0593a718d9b44998d9f Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 10 Mar 2012 23:10:35 -0500 Subject: [PATCH 19/64] aio: aio_nr decrements don't need to be delayed we can do that right in __put_ioctx(); as the result, the loop in ioctx_alloc() can be killed. Signed-off-by: Al Viro --- fs/aio.c | 42 ++++++++++++++---------------------------- 1 file changed, 14 insertions(+), 28 deletions(-) diff --git a/fs/aio.c b/fs/aio.c index d09b56090aa5..216eb37b2c76 100644 --- a/fs/aio.c +++ b/fs/aio.c @@ -199,16 +199,7 @@ static int aio_setup_ring(struct kioctx *ctx) static void ctx_rcu_free(struct rcu_head *head) { struct kioctx *ctx = container_of(head, struct kioctx, rcu_head); - unsigned nr_events = ctx->max_reqs; - kmem_cache_free(kioctx_cachep, ctx); - - if (nr_events) { - spin_lock(&aio_nr_lock); - BUG_ON(aio_nr - nr_events > aio_nr); - aio_nr -= nr_events; - spin_unlock(&aio_nr_lock); - } } /* __put_ioctx @@ -217,6 +208,7 @@ static void ctx_rcu_free(struct rcu_head *head) */ static void __put_ioctx(struct kioctx *ctx) { + unsigned nr_events = ctx->max_reqs; BUG_ON(ctx->reqs_active); cancel_delayed_work(&ctx->wq); @@ -224,6 +216,12 @@ static void __put_ioctx(struct kioctx *ctx) aio_free_ring(ctx); mmdrop(ctx->mm); ctx->mm = NULL; + if (nr_events) { + spin_lock(&aio_nr_lock); + BUG_ON(aio_nr - nr_events > aio_nr); + aio_nr -= nr_events; + spin_unlock(&aio_nr_lock); + } pr_debug("__put_ioctx: freeing %p\n", ctx); call_rcu(&ctx->rcu_head, ctx_rcu_free); } @@ -247,7 +245,6 @@ static struct kioctx *ioctx_alloc(unsigned nr_events) { struct mm_struct *mm; struct kioctx *ctx; - int did_sync = 0; int err = -ENOMEM; /* Prevent overflows */ @@ -257,7 +254,7 @@ static struct kioctx *ioctx_alloc(unsigned nr_events) return ERR_PTR(-EINVAL); } - if ((unsigned long)nr_events > aio_max_nr) + if (!nr_events || (unsigned long)nr_events > aio_max_nr) return ERR_PTR(-EAGAIN); ctx = kmem_cache_zalloc(kioctx_cachep, GFP_KERNEL); @@ -281,25 +278,14 @@ static struct kioctx *ioctx_alloc(unsigned nr_events) goto out_freectx; /* limit the number of system wide aios */ - do { - spin_lock_bh(&aio_nr_lock); - if (aio_nr + nr_events > aio_max_nr || - aio_nr + nr_events < aio_nr) - ctx->max_reqs = 0; - else - aio_nr += ctx->max_reqs; + spin_lock_bh(&aio_nr_lock); + if (aio_nr + nr_events > aio_max_nr || + aio_nr + nr_events < aio_nr) { spin_unlock_bh(&aio_nr_lock); - if (ctx->max_reqs || did_sync) - break; - - /* wait for rcu callbacks to have completed before giving up */ - synchronize_rcu(); - did_sync = 1; - ctx->max_reqs = nr_events; - } while (1); - - if (ctx->max_reqs == 0) goto out_cleanup; + } + aio_nr += ctx->max_reqs; + spin_unlock_bh(&aio_nr_lock); /* now link into global list. */ spin_lock(&mm->ioctx_lock); From 9fa1cb397fa052fc9acfaf5a9f2faff31e10f6b7 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 10 Mar 2012 23:14:05 -0500 Subject: [PATCH 20/64] aio: aio_nr_lock is taken only synchronously now Signed-off-by: Al Viro --- fs/aio.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/fs/aio.c b/fs/aio.c index 216eb37b2c76..9c3de88e2ead 100644 --- a/fs/aio.c +++ b/fs/aio.c @@ -278,14 +278,14 @@ static struct kioctx *ioctx_alloc(unsigned nr_events) goto out_freectx; /* limit the number of system wide aios */ - spin_lock_bh(&aio_nr_lock); + spin_lock(&aio_nr_lock); if (aio_nr + nr_events > aio_max_nr || aio_nr + nr_events < aio_nr) { - spin_unlock_bh(&aio_nr_lock); + spin_unlock(&aio_nr_lock); goto out_cleanup; } aio_nr += ctx->max_reqs; - spin_unlock_bh(&aio_nr_lock); + spin_unlock(&aio_nr_lock); /* now link into global list. */ spin_lock(&mm->ioctx_lock); From bf50722a3c4a83aae651dc20b708308a4f119eb9 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sun, 11 Mar 2012 00:58:40 -0500 Subject: [PATCH 21/64] aio: use cancel_delayed_work_sync() Signed-off-by: Al Viro --- fs/aio.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/fs/aio.c b/fs/aio.c index 9c3de88e2ead..a92d7547b6f6 100644 --- a/fs/aio.c +++ b/fs/aio.c @@ -211,8 +211,7 @@ static void __put_ioctx(struct kioctx *ctx) unsigned nr_events = ctx->max_reqs; BUG_ON(ctx->reqs_active); - cancel_delayed_work(&ctx->wq); - cancel_work_sync(&ctx->wq.work); + cancel_delayed_work_sync(&ctx->wq); aio_free_ring(ctx); mmdrop(ctx->mm); ctx->mm = NULL; From cd1ea261ac128479833b9f518bf788ee47ada2de Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sun, 11 Mar 2012 00:59:07 -0500 Subject: [PATCH 22/64] aio: don't bother with cancel_delayed_work() in exit_aio() __put_ioctx() will cover it anyway. Signed-off-by: Al Viro --- fs/aio.c | 4 ---- 1 file changed, 4 deletions(-) diff --git a/fs/aio.c b/fs/aio.c index a92d7547b6f6..3174cd969b01 100644 --- a/fs/aio.c +++ b/fs/aio.c @@ -390,10 +390,6 @@ void exit_aio(struct mm_struct *mm) aio_cancel_all(ctx); wait_for_all_aios(ctx); - /* - * Ensure we don't leave the ctx on the aio_wq - */ - cancel_work_sync(&ctx->wq.work); if (1 != atomic_read(&ctx->users)) printk(KERN_DEBUG From 9fcf03d0d6e845ed495fc8b1ec328b473ff298b3 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Tue, 13 Mar 2012 22:06:28 -0400 Subject: [PATCH 23/64] aio: fix the comment in aio_kick_handler() It should've been changed when queue_work() became queue_delayed_work(..., 0) in there. It's always had been about not needing a delay, not about not using specific function... Signed-off-by: Al Viro --- fs/aio.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/aio.c b/fs/aio.c index 3174cd969b01..59b431a2ae0c 100644 --- a/fs/aio.c +++ b/fs/aio.c @@ -899,7 +899,7 @@ static void aio_kick_handler(struct work_struct *work) unuse_mm(mm); set_fs(oldfs); /* - * we're in a worker thread already, don't use queue_delayed_work, + * we're in a worker thread already; no point using non-zero delay */ if (requeue) queue_delayed_work(aio_wq, &ctx->wq, 0); From 66b3fad3f4c535c92b6a1184d535a97d6aa5d82a Mon Sep 17 00:00:00 2001 From: Al Viro Date: Wed, 14 Mar 2012 21:48:20 -0400 Subject: [PATCH 24/64] constify path argument of audit_log_d_path() Signed-off-by: Al Viro --- include/linux/audit.h | 2 +- kernel/audit.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/include/linux/audit.h b/include/linux/audit.h index 9ff7a2c48b50..ed3ef1972496 100644 --- a/include/linux/audit.h +++ b/include/linux/audit.h @@ -684,7 +684,7 @@ extern void audit_log_untrustedstring(struct audit_buffer *ab, const char *string); extern void audit_log_d_path(struct audit_buffer *ab, const char *prefix, - struct path *path); + const struct path *path); extern void audit_log_key(struct audit_buffer *ab, char *key); extern void audit_log_lost(const char *message); diff --git a/kernel/audit.c b/kernel/audit.c index bb0eb5bb9a0a..1c7f2c61416b 100644 --- a/kernel/audit.c +++ b/kernel/audit.c @@ -1418,7 +1418,7 @@ void audit_log_untrustedstring(struct audit_buffer *ab, const char *string) /* This is a helper-function to print the escaped d_path */ void audit_log_d_path(struct audit_buffer *ab, const char *prefix, - struct path *path) + const struct path *path) { char *p, *pathname; From 38eff2892628fa5c4fc8962a17b7296f42833ebe Mon Sep 17 00:00:00 2001 From: Al Viro Date: Wed, 14 Mar 2012 21:51:10 -0400 Subject: [PATCH 25/64] constify path argument of trace_seq_path() Signed-off-by: Al Viro --- include/linux/trace_seq.h | 4 ++-- kernel/trace/trace_output.c | 2 +- 2 files changed, 3 insertions(+), 3 deletions(-) diff --git a/include/linux/trace_seq.h b/include/linux/trace_seq.h index 7dadc3df0c77..a32d86ec8bf2 100644 --- a/include/linux/trace_seq.h +++ b/include/linux/trace_seq.h @@ -44,7 +44,7 @@ extern int trace_seq_putmem(struct trace_seq *s, const void *mem, size_t len); extern int trace_seq_putmem_hex(struct trace_seq *s, const void *mem, size_t len); extern void *trace_seq_reserve(struct trace_seq *s, size_t len); -extern int trace_seq_path(struct trace_seq *s, struct path *path); +extern int trace_seq_path(struct trace_seq *s, const struct path *path); #else /* CONFIG_TRACING */ static inline int trace_seq_printf(struct trace_seq *s, const char *fmt, ...) @@ -88,7 +88,7 @@ static inline void *trace_seq_reserve(struct trace_seq *s, size_t len) { return NULL; } -static inline int trace_seq_path(struct trace_seq *s, struct path *path) +static inline int trace_seq_path(struct trace_seq *s, const struct path *path) { return 0; } diff --git a/kernel/trace/trace_output.c b/kernel/trace/trace_output.c index 0d6ff3555942..690987198ad7 100644 --- a/kernel/trace/trace_output.c +++ b/kernel/trace/trace_output.c @@ -264,7 +264,7 @@ void *trace_seq_reserve(struct trace_seq *s, size_t len) return ret; } -int trace_seq_path(struct trace_seq *s, struct path *path) +int trace_seq_path(struct trace_seq *s, const struct path *path) { unsigned char *p; From 40ffe67d2e89c7a475421d007becc11a2f88ea3d Mon Sep 17 00:00:00 2001 From: Al Viro Date: Wed, 14 Mar 2012 21:54:32 -0400 Subject: [PATCH 26/64] switch unix_sock to struct path Signed-off-by: Al Viro --- include/net/af_unix.h | 3 +-- net/unix/af_unix.c | 35 +++++++++++++++-------------------- net/unix/diag.c | 2 +- security/lsm_audit.c | 8 ++------ 4 files changed, 19 insertions(+), 29 deletions(-) diff --git a/include/net/af_unix.h b/include/net/af_unix.h index 5a4e29b168c9..ca68e2cef230 100644 --- a/include/net/af_unix.h +++ b/include/net/af_unix.h @@ -49,8 +49,7 @@ struct unix_sock { /* WARNING: sk has to be the first member */ struct sock sk; struct unix_address *addr; - struct dentry *dentry; - struct vfsmount *mnt; + struct path path; struct mutex readlock; struct sock *peer; struct sock *other; diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c index 85d3bb7490aa..ef4b780ef63d 100644 --- a/net/unix/af_unix.c +++ b/net/unix/af_unix.c @@ -293,7 +293,7 @@ static struct sock *unix_find_socket_byinode(struct inode *i) spin_lock(&unix_table_lock); sk_for_each(s, node, &unix_socket_table[i->i_ino & (UNIX_HASH_SIZE - 1)]) { - struct dentry *dentry = unix_sk(s)->dentry; + struct dentry *dentry = unix_sk(s)->path.dentry; if (dentry && dentry->d_inode == i) { sock_hold(s); @@ -377,8 +377,7 @@ static void unix_sock_destructor(struct sock *sk) static int unix_release_sock(struct sock *sk, int embrion) { struct unix_sock *u = unix_sk(sk); - struct dentry *dentry; - struct vfsmount *mnt; + struct path path; struct sock *skpair; struct sk_buff *skb; int state; @@ -389,10 +388,9 @@ static int unix_release_sock(struct sock *sk, int embrion) unix_state_lock(sk); sock_orphan(sk); sk->sk_shutdown = SHUTDOWN_MASK; - dentry = u->dentry; - u->dentry = NULL; - mnt = u->mnt; - u->mnt = NULL; + path = u->path; + u->path.dentry = NULL; + u->path.mnt = NULL; state = sk->sk_state; sk->sk_state = TCP_CLOSE; unix_state_unlock(sk); @@ -425,10 +423,8 @@ static int unix_release_sock(struct sock *sk, int embrion) kfree_skb(skb); } - if (dentry) { - dput(dentry); - mntput(mnt); - } + if (path.dentry) + path_put(&path); sock_put(sk); @@ -628,8 +624,8 @@ static struct sock *unix_create1(struct net *net, struct socket *sock) sk->sk_max_ack_backlog = net->unx.sysctl_max_dgram_qlen; sk->sk_destruct = unix_sock_destructor; u = unix_sk(sk); - u->dentry = NULL; - u->mnt = NULL; + u->path.dentry = NULL; + u->path.mnt = NULL; spin_lock_init(&u->lock); atomic_long_set(&u->inflight, 0); INIT_LIST_HEAD(&u->link); @@ -789,9 +785,9 @@ static struct sock *unix_find_other(struct net *net, u = unix_find_socket_byname(net, sunname, len, type, hash); if (u) { struct dentry *dentry; - dentry = unix_sk(u)->dentry; + dentry = unix_sk(u)->path.dentry; if (dentry) - touch_atime(unix_sk(u)->mnt, dentry); + touch_atime(unix_sk(u)->path.mnt, dentry); } else goto fail; } @@ -897,8 +893,7 @@ out_mknod_drop_write: list = &unix_socket_table[addr->hash]; } else { list = &unix_socket_table[dentry->d_inode->i_ino & (UNIX_HASH_SIZE-1)]; - u->dentry = path.dentry; - u->mnt = path.mnt; + u->path = path; } err = 0; @@ -1180,9 +1175,9 @@ restart: atomic_inc(&otheru->addr->refcnt); newu->addr = otheru->addr; } - if (otheru->dentry) { - newu->dentry = dget(otheru->dentry); - newu->mnt = mntget(otheru->mnt); + if (otheru->path.dentry) { + path_get(&otheru->path); + newu->path = otheru->path; } /* Set credentials */ diff --git a/net/unix/diag.c b/net/unix/diag.c index 6b7697fd911b..ffd86518e646 100644 --- a/net/unix/diag.c +++ b/net/unix/diag.c @@ -29,7 +29,7 @@ rtattr_failure: static int sk_diag_dump_vfs(struct sock *sk, struct sk_buff *nlskb) { - struct dentry *dentry = unix_sk(sk)->dentry; + struct dentry *dentry = unix_sk(sk)->path.dentry; struct unix_diag_vfs *uv; if (dentry) { diff --git a/security/lsm_audit.c b/security/lsm_audit.c index 293b8c45b1d1..8b8f0902f6e5 100644 --- a/security/lsm_audit.c +++ b/security/lsm_audit.c @@ -313,12 +313,8 @@ static void dump_common_audit_data(struct audit_buffer *ab, } case AF_UNIX: u = unix_sk(sk); - if (u->dentry) { - struct path path = { - .dentry = u->dentry, - .mnt = u->mnt - }; - audit_log_d_path(ab, " path=", &path); + if (u->path.dentry) { + audit_log_d_path(ab, " path=", &u->path); break; } if (!u->addr) From 68ac1234fb949b66941d94dce4157742799fc581 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 15 Mar 2012 08:21:57 -0400 Subject: [PATCH 27/64] switch touch_atime to struct path Signed-off-by: Al Viro --- fs/cachefiles/namei.c | 3 ++- fs/ecryptfs/file.c | 9 ++++----- fs/inode.c | 5 +++-- fs/namei.c | 2 +- fs/nfsd/vfs.c | 11 ++++++----- fs/stat.c | 2 +- include/linux/fs.h | 4 ++-- net/unix/af_unix.c | 4 ++-- 8 files changed, 21 insertions(+), 19 deletions(-) diff --git a/fs/cachefiles/namei.c b/fs/cachefiles/namei.c index a0358c2189cb..7f0771d3894e 100644 --- a/fs/cachefiles/namei.c +++ b/fs/cachefiles/namei.c @@ -646,7 +646,8 @@ lookup_again: * (this is used to keep track of culling, and atimes are only * updated by read, write and readdir but not lookup or * open) */ - touch_atime(cache->mnt, next); + path.dentry = next; + touch_atime(&path); } /* open a file interface onto a data file */ diff --git a/fs/ecryptfs/file.c b/fs/ecryptfs/file.c index d3f95f941c47..2b17f2f9b121 100644 --- a/fs/ecryptfs/file.c +++ b/fs/ecryptfs/file.c @@ -48,8 +48,7 @@ static ssize_t ecryptfs_read_update_atime(struct kiocb *iocb, unsigned long nr_segs, loff_t pos) { ssize_t rc; - struct dentry *lower_dentry; - struct vfsmount *lower_vfsmount; + struct path lower; struct file *file = iocb->ki_filp; rc = generic_file_aio_read(iocb, iov, nr_segs, pos); @@ -60,9 +59,9 @@ static ssize_t ecryptfs_read_update_atime(struct kiocb *iocb, if (-EIOCBQUEUED == rc) rc = wait_on_sync_kiocb(iocb); if (rc >= 0) { - lower_dentry = ecryptfs_dentry_to_lower(file->f_path.dentry); - lower_vfsmount = ecryptfs_dentry_to_lower_mnt(file->f_path.dentry); - touch_atime(lower_vfsmount, lower_dentry); + lower.dentry = ecryptfs_dentry_to_lower(file->f_path.dentry); + lower.mnt = ecryptfs_dentry_to_lower_mnt(file->f_path.dentry); + touch_atime(&lower); } return rc; } diff --git a/fs/inode.c b/fs/inode.c index 92de04b0baa2..8b612813a6a7 100644 --- a/fs/inode.c +++ b/fs/inode.c @@ -1499,9 +1499,10 @@ static int relatime_need_update(struct vfsmount *mnt, struct inode *inode, * This function automatically handles read only file systems and media, * as well as the "noatime" flag and inode specific "noatime" markers. */ -void touch_atime(struct vfsmount *mnt, struct dentry *dentry) +void touch_atime(struct path *path) { - struct inode *inode = dentry->d_inode; + struct vfsmount *mnt = path->mnt; + struct inode *inode = path->dentry->d_inode; struct timespec now; if (inode->i_flags & S_NOATIME) diff --git a/fs/namei.c b/fs/namei.c index a0b82762e8fc..0ccc74ee92fb 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -642,7 +642,7 @@ follow_link(struct path *link, struct nameidata *nd, void **p) cond_resched(); current->total_link_count++; - touch_atime(link->mnt, dentry); + touch_atime(link); nd_set_link(nd, NULL); error = security_inode_follow_link(link->dentry, nd); diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c index edf6d3ed8777..e59f71d0cf73 100644 --- a/fs/nfsd/vfs.c +++ b/fs/nfsd/vfs.c @@ -1541,30 +1541,31 @@ do_nfsd_create(struct svc_rqst *rqstp, struct svc_fh *fhp, __be32 nfsd_readlink(struct svc_rqst *rqstp, struct svc_fh *fhp, char *buf, int *lenp) { - struct dentry *dentry; struct inode *inode; mm_segment_t oldfs; __be32 err; int host_err; + struct path path; err = fh_verify(rqstp, fhp, S_IFLNK, NFSD_MAY_NOP); if (err) goto out; - dentry = fhp->fh_dentry; - inode = dentry->d_inode; + path.mnt = fhp->fh_export->ex_path.mnt; + path.dentry = fhp->fh_dentry; + inode = path.dentry->d_inode; err = nfserr_inval; if (!inode->i_op->readlink) goto out; - touch_atime(fhp->fh_export->ex_path.mnt, dentry); + touch_atime(&path); /* N.B. Why does this call need a get_fs()?? * Remove the set_fs and watch the fireworks:-) --okir */ oldfs = get_fs(); set_fs(KERNEL_DS); - host_err = inode->i_op->readlink(dentry, buf, *lenp); + host_err = inode->i_op->readlink(path.dentry, buf, *lenp); set_fs(oldfs); if (host_err < 0) diff --git a/fs/stat.c b/fs/stat.c index 8806b8997d2e..86f13563a463 100644 --- a/fs/stat.c +++ b/fs/stat.c @@ -307,7 +307,7 @@ SYSCALL_DEFINE4(readlinkat, int, dfd, const char __user *, pathname, if (inode->i_op->readlink) { error = security_inode_readlink(path.dentry); if (!error) { - touch_atime(path.mnt, path.dentry); + touch_atime(&path); error = inode->i_op->readlink(path.dentry, buf, bufsiz); } diff --git a/include/linux/fs.h b/include/linux/fs.h index b89dc4d60d99..9bbe1a9ac432 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -1812,11 +1812,11 @@ static inline void inode_inc_iversion(struct inode *inode) spin_unlock(&inode->i_lock); } -extern void touch_atime(struct vfsmount *mnt, struct dentry *dentry); +extern void touch_atime(struct path *); static inline void file_accessed(struct file *file) { if (!(file->f_flags & O_NOATIME)) - touch_atime(file->f_path.mnt, file->f_path.dentry); + touch_atime(&file->f_path); } int sync_inode(struct inode *inode, struct writeback_control *wbc); diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c index ef4b780ef63d..081679444a6e 100644 --- a/net/unix/af_unix.c +++ b/net/unix/af_unix.c @@ -771,7 +771,7 @@ static struct sock *unix_find_other(struct net *net, goto put_fail; if (u->sk_type == type) - touch_atime(path.mnt, path.dentry); + touch_atime(&path); path_put(&path); @@ -787,7 +787,7 @@ static struct sock *unix_find_other(struct net *net, struct dentry *dentry; dentry = unix_sk(u)->path.dentry; if (dentry) - touch_atime(unix_sk(u)->path.mnt, dentry); + touch_atime(&unix_sk(u)->path); } else goto fail; } From 3ee05ed0679f29ab19727067ff7c14f0c257fa9a Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 15 Mar 2012 14:48:29 -0400 Subject: [PATCH 28/64] no need to play with fs->seq in exit_fs() Signed-off-by: Al Viro --- fs/fs_struct.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/fs/fs_struct.c b/fs/fs_struct.c index 78b519c13536..27114b413622 100644 --- a/fs/fs_struct.c +++ b/fs/fs_struct.c @@ -107,10 +107,8 @@ void exit_fs(struct task_struct *tsk) int kill; task_lock(tsk); spin_lock(&fs->lock); - write_seqcount_begin(&fs->seq); tsk->fs = NULL; kill = !--fs->users; - write_seqcount_end(&fs->seq); spin_unlock(&fs->lock); task_unlock(tsk); if (kill) From 82234e61a8fc75599f29026c0805fc0cfe2a6c87 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 15 Mar 2012 14:48:55 -0400 Subject: [PATCH 29/64] vfs: take path_get_longterm() out of write_seqcount scope Signed-off-by: Al Viro --- fs/fs_struct.c | 31 +++++++++++++++++-------------- 1 file changed, 17 insertions(+), 14 deletions(-) diff --git a/fs/fs_struct.c b/fs/fs_struct.c index 27114b413622..6324c4274959 100644 --- a/fs/fs_struct.c +++ b/fs/fs_struct.c @@ -26,11 +26,11 @@ void set_fs_root(struct fs_struct *fs, struct path *path) { struct path old_root; + path_get_longterm(path); spin_lock(&fs->lock); write_seqcount_begin(&fs->seq); old_root = fs->root; fs->root = *path; - path_get_longterm(path); write_seqcount_end(&fs->seq); spin_unlock(&fs->lock); if (old_root.dentry) @@ -45,11 +45,11 @@ void set_fs_pwd(struct fs_struct *fs, struct path *path) { struct path old_pwd; + path_get_longterm(path); spin_lock(&fs->lock); write_seqcount_begin(&fs->seq); old_pwd = fs->pwd; fs->pwd = *path; - path_get_longterm(path); write_seqcount_end(&fs->seq); spin_unlock(&fs->lock); @@ -57,6 +57,14 @@ void set_fs_pwd(struct fs_struct *fs, struct path *path) path_put_longterm(&old_pwd); } +static inline int replace_path(struct path *p, const struct path *old, const struct path *new) +{ + if (likely(p->dentry != old->dentry || p->mnt != old->mnt)) + return 0; + *p = *new; + return 1; +} + void chroot_fs_refs(struct path *old_root, struct path *new_root) { struct task_struct *g, *p; @@ -68,21 +76,16 @@ void chroot_fs_refs(struct path *old_root, struct path *new_root) task_lock(p); fs = p->fs; if (fs) { + int hits = 0; spin_lock(&fs->lock); write_seqcount_begin(&fs->seq); - if (fs->root.dentry == old_root->dentry - && fs->root.mnt == old_root->mnt) { - path_get_longterm(new_root); - fs->root = *new_root; - count++; - } - if (fs->pwd.dentry == old_root->dentry - && fs->pwd.mnt == old_root->mnt) { - path_get_longterm(new_root); - fs->pwd = *new_root; - count++; - } + hits += replace_path(&fs->root, old_root, new_root); + hits += replace_path(&fs->pwd, old_root, new_root); write_seqcount_end(&fs->seq); + while (hits--) { + count++; + path_get_longterm(new_root); + } spin_unlock(&fs->lock); } task_unlock(p); From c45ac8887e778c4fa2b572c51a94a681a0955d4d Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 17 Mar 2012 00:59:06 -0400 Subject: [PATCH 30/64] take private bits of reiserfs_xattr.h to fs/reiserfs/xattr.h Signed-off-by: Al Viro --- fs/reiserfs/file.c | 2 +- fs/reiserfs/inode.c | 2 +- fs/reiserfs/namei.c | 2 +- fs/reiserfs/super.c | 2 +- fs/reiserfs/xattr.c | 2 +- fs/reiserfs/xattr.h | 124 ++++++++++++++++++++++++++++++++ fs/reiserfs/xattr_acl.c | 2 +- fs/reiserfs/xattr_security.c | 2 +- fs/reiserfs/xattr_trusted.c | 2 +- fs/reiserfs/xattr_user.c | 2 +- include/linux/reiserfs_xattr.h | 128 --------------------------------- 11 files changed, 133 insertions(+), 137 deletions(-) create mode 100644 fs/reiserfs/xattr.h diff --git a/fs/reiserfs/file.c b/fs/reiserfs/file.c index ace635053a36..e26ee4988e78 100644 --- a/fs/reiserfs/file.c +++ b/fs/reiserfs/file.c @@ -5,7 +5,7 @@ #include #include #include -#include +#include "xattr.h" #include #include #include diff --git a/fs/reiserfs/inode.c b/fs/reiserfs/inode.c index 9e8cd5acd79c..894c7316e9c6 100644 --- a/fs/reiserfs/inode.c +++ b/fs/reiserfs/inode.c @@ -6,7 +6,7 @@ #include #include #include -#include +#include "xattr.h" #include #include #include diff --git a/fs/reiserfs/namei.c b/fs/reiserfs/namei.c index 146378865239..cb67ebf882a9 100644 --- a/fs/reiserfs/namei.c +++ b/fs/reiserfs/namei.c @@ -16,7 +16,7 @@ #include #include #include -#include +#include "xattr.h" #include #define INC_DIR_INODE_NLINK(i) if (i->i_nlink != 1) { inc_nlink(i); if (i->i_nlink >= REISERFS_LINK_MAX) set_nlink(i, 1); } diff --git a/fs/reiserfs/super.c b/fs/reiserfs/super.c index 208dfd144409..a67fc7d28de5 100644 --- a/fs/reiserfs/super.c +++ b/fs/reiserfs/super.c @@ -18,7 +18,7 @@ #include #include #include -#include +#include "xattr.h" #include #include #include diff --git a/fs/reiserfs/xattr.c b/fs/reiserfs/xattr.c index c24deda8a8bc..ead5d8aab440 100644 --- a/fs/reiserfs/xattr.c +++ b/fs/reiserfs/xattr.c @@ -43,7 +43,7 @@ #include #include #include -#include +#include "xattr.h" #include #include #include diff --git a/fs/reiserfs/xattr.h b/fs/reiserfs/xattr.h new file mode 100644 index 000000000000..367c0459db6d --- /dev/null +++ b/fs/reiserfs/xattr.h @@ -0,0 +1,124 @@ +#include +#include +#include +#include +#include +#include + +struct inode; +struct dentry; +struct iattr; +struct super_block; +struct nameidata; + +int reiserfs_xattr_register_handlers(void) __init; +void reiserfs_xattr_unregister_handlers(void); +int reiserfs_xattr_init(struct super_block *sb, int mount_flags); +int reiserfs_lookup_privroot(struct super_block *sb); +int reiserfs_delete_xattrs(struct inode *inode); +int reiserfs_chown_xattrs(struct inode *inode, struct iattr *attrs); +int reiserfs_permission(struct inode *inode, int mask); + +#ifdef CONFIG_REISERFS_FS_XATTR +#define has_xattr_dir(inode) (REISERFS_I(inode)->i_flags & i_has_xattr_dir) +ssize_t reiserfs_getxattr(struct dentry *dentry, const char *name, + void *buffer, size_t size); +int reiserfs_setxattr(struct dentry *dentry, const char *name, + const void *value, size_t size, int flags); +ssize_t reiserfs_listxattr(struct dentry *dentry, char *buffer, size_t size); +int reiserfs_removexattr(struct dentry *dentry, const char *name); + +int reiserfs_xattr_get(struct inode *, const char *, void *, size_t); +int reiserfs_xattr_set(struct inode *, const char *, const void *, size_t, int); +int reiserfs_xattr_set_handle(struct reiserfs_transaction_handle *, + struct inode *, const char *, const void *, + size_t, int); + +extern const struct xattr_handler reiserfs_xattr_user_handler; +extern const struct xattr_handler reiserfs_xattr_trusted_handler; +extern const struct xattr_handler reiserfs_xattr_security_handler; +#ifdef CONFIG_REISERFS_FS_SECURITY +int reiserfs_security_init(struct inode *dir, struct inode *inode, + const struct qstr *qstr, + struct reiserfs_security_handle *sec); +int reiserfs_security_write(struct reiserfs_transaction_handle *th, + struct inode *inode, + struct reiserfs_security_handle *sec); +void reiserfs_security_free(struct reiserfs_security_handle *sec); +#endif + +static inline int reiserfs_xattrs_initialized(struct super_block *sb) +{ + return REISERFS_SB(sb)->priv_root != NULL; +} + +#define xattr_size(size) ((size) + sizeof(struct reiserfs_xattr_header)) +static inline loff_t reiserfs_xattr_nblocks(struct inode *inode, loff_t size) +{ + loff_t ret = 0; + if (reiserfs_file_data_log(inode)) { + ret = _ROUND_UP(xattr_size(size), inode->i_sb->s_blocksize); + ret >>= inode->i_sb->s_blocksize_bits; + } + return ret; +} + +/* We may have to create up to 3 objects: xattr root, xattr dir, xattr file. + * Let's try to be smart about it. + * xattr root: We cache it. If it's not cached, we may need to create it. + * xattr dir: If anything has been loaded for this inode, we can set a flag + * saying so. + * xattr file: Since we don't cache xattrs, we can't tell. We always include + * blocks for it. + * + * However, since root and dir can be created between calls - YOU MUST SAVE + * THIS VALUE. + */ +static inline size_t reiserfs_xattr_jcreate_nblocks(struct inode *inode) +{ + size_t nblocks = JOURNAL_BLOCKS_PER_OBJECT(inode->i_sb); + + if ((REISERFS_I(inode)->i_flags & i_has_xattr_dir) == 0) { + nblocks += JOURNAL_BLOCKS_PER_OBJECT(inode->i_sb); + if (!REISERFS_SB(inode->i_sb)->xattr_root->d_inode) + nblocks += JOURNAL_BLOCKS_PER_OBJECT(inode->i_sb); + } + + return nblocks; +} + +static inline void reiserfs_init_xattr_rwsem(struct inode *inode) +{ + init_rwsem(&REISERFS_I(inode)->i_xattr_sem); +} + +#else + +#define reiserfs_getxattr NULL +#define reiserfs_setxattr NULL +#define reiserfs_listxattr NULL +#define reiserfs_removexattr NULL + +static inline void reiserfs_init_xattr_rwsem(struct inode *inode) +{ +} +#endif /* CONFIG_REISERFS_FS_XATTR */ + +#ifndef CONFIG_REISERFS_FS_SECURITY +static inline int reiserfs_security_init(struct inode *dir, + struct inode *inode, + const struct qstr *qstr, + struct reiserfs_security_handle *sec) +{ + return 0; +} +static inline int +reiserfs_security_write(struct reiserfs_transaction_handle *th, + struct inode *inode, + struct reiserfs_security_handle *sec) +{ + return 0; +} +static inline void reiserfs_security_free(struct reiserfs_security_handle *sec) +{} +#endif diff --git a/fs/reiserfs/xattr_acl.c b/fs/reiserfs/xattr_acl.c index 6da0396e5052..c0a8c519b2e1 100644 --- a/fs/reiserfs/xattr_acl.c +++ b/fs/reiserfs/xattr_acl.c @@ -7,7 +7,7 @@ #include #include #include -#include +#include "xattr.h" #include #include diff --git a/fs/reiserfs/xattr_security.c b/fs/reiserfs/xattr_security.c index 534668fa41be..6104066e44c4 100644 --- a/fs/reiserfs/xattr_security.c +++ b/fs/reiserfs/xattr_security.c @@ -4,7 +4,7 @@ #include #include #include -#include +#include "xattr.h" #include #include diff --git a/fs/reiserfs/xattr_trusted.c b/fs/reiserfs/xattr_trusted.c index 9883736ce3ec..f995b258c141 100644 --- a/fs/reiserfs/xattr_trusted.c +++ b/fs/reiserfs/xattr_trusted.c @@ -4,7 +4,7 @@ #include #include #include -#include +#include "xattr.h" #include static int diff --git a/fs/reiserfs/xattr_user.c b/fs/reiserfs/xattr_user.c index 45ae1a00013a..748978db6f03 100644 --- a/fs/reiserfs/xattr_user.c +++ b/fs/reiserfs/xattr_user.c @@ -3,7 +3,7 @@ #include #include #include -#include +#include "xattr.h" #include static int diff --git a/include/linux/reiserfs_xattr.h b/include/linux/reiserfs_xattr.h index c2b71473266e..d8ce17c2459a 100644 --- a/include/linux/reiserfs_xattr.h +++ b/include/linux/reiserfs_xattr.h @@ -21,132 +21,4 @@ struct reiserfs_security_handle { size_t length; }; -#ifdef __KERNEL__ - -#include -#include -#include -#include -#include - -struct inode; -struct dentry; -struct iattr; -struct super_block; -struct nameidata; - -int reiserfs_xattr_register_handlers(void) __init; -void reiserfs_xattr_unregister_handlers(void); -int reiserfs_xattr_init(struct super_block *sb, int mount_flags); -int reiserfs_lookup_privroot(struct super_block *sb); -int reiserfs_delete_xattrs(struct inode *inode); -int reiserfs_chown_xattrs(struct inode *inode, struct iattr *attrs); -int reiserfs_permission(struct inode *inode, int mask); - -#ifdef CONFIG_REISERFS_FS_XATTR -#define has_xattr_dir(inode) (REISERFS_I(inode)->i_flags & i_has_xattr_dir) -ssize_t reiserfs_getxattr(struct dentry *dentry, const char *name, - void *buffer, size_t size); -int reiserfs_setxattr(struct dentry *dentry, const char *name, - const void *value, size_t size, int flags); -ssize_t reiserfs_listxattr(struct dentry *dentry, char *buffer, size_t size); -int reiserfs_removexattr(struct dentry *dentry, const char *name); - -int reiserfs_xattr_get(struct inode *, const char *, void *, size_t); -int reiserfs_xattr_set(struct inode *, const char *, const void *, size_t, int); -int reiserfs_xattr_set_handle(struct reiserfs_transaction_handle *, - struct inode *, const char *, const void *, - size_t, int); - -extern const struct xattr_handler reiserfs_xattr_user_handler; -extern const struct xattr_handler reiserfs_xattr_trusted_handler; -extern const struct xattr_handler reiserfs_xattr_security_handler; -#ifdef CONFIG_REISERFS_FS_SECURITY -int reiserfs_security_init(struct inode *dir, struct inode *inode, - const struct qstr *qstr, - struct reiserfs_security_handle *sec); -int reiserfs_security_write(struct reiserfs_transaction_handle *th, - struct inode *inode, - struct reiserfs_security_handle *sec); -void reiserfs_security_free(struct reiserfs_security_handle *sec); -#endif - -static inline int reiserfs_xattrs_initialized(struct super_block *sb) -{ - return REISERFS_SB(sb)->priv_root != NULL; -} - -#define xattr_size(size) ((size) + sizeof(struct reiserfs_xattr_header)) -static inline loff_t reiserfs_xattr_nblocks(struct inode *inode, loff_t size) -{ - loff_t ret = 0; - if (reiserfs_file_data_log(inode)) { - ret = _ROUND_UP(xattr_size(size), inode->i_sb->s_blocksize); - ret >>= inode->i_sb->s_blocksize_bits; - } - return ret; -} - -/* We may have to create up to 3 objects: xattr root, xattr dir, xattr file. - * Let's try to be smart about it. - * xattr root: We cache it. If it's not cached, we may need to create it. - * xattr dir: If anything has been loaded for this inode, we can set a flag - * saying so. - * xattr file: Since we don't cache xattrs, we can't tell. We always include - * blocks for it. - * - * However, since root and dir can be created between calls - YOU MUST SAVE - * THIS VALUE. - */ -static inline size_t reiserfs_xattr_jcreate_nblocks(struct inode *inode) -{ - size_t nblocks = JOURNAL_BLOCKS_PER_OBJECT(inode->i_sb); - - if ((REISERFS_I(inode)->i_flags & i_has_xattr_dir) == 0) { - nblocks += JOURNAL_BLOCKS_PER_OBJECT(inode->i_sb); - if (!REISERFS_SB(inode->i_sb)->xattr_root->d_inode) - nblocks += JOURNAL_BLOCKS_PER_OBJECT(inode->i_sb); - } - - return nblocks; -} - -static inline void reiserfs_init_xattr_rwsem(struct inode *inode) -{ - init_rwsem(&REISERFS_I(inode)->i_xattr_sem); -} - -#else - -#define reiserfs_getxattr NULL -#define reiserfs_setxattr NULL -#define reiserfs_listxattr NULL -#define reiserfs_removexattr NULL - -static inline void reiserfs_init_xattr_rwsem(struct inode *inode) -{ -} -#endif /* CONFIG_REISERFS_FS_XATTR */ - -#ifndef CONFIG_REISERFS_FS_SECURITY -static inline int reiserfs_security_init(struct inode *dir, - struct inode *inode, - const struct qstr *qstr, - struct reiserfs_security_handle *sec) -{ - return 0; -} -static inline int -reiserfs_security_write(struct reiserfs_transaction_handle *th, - struct inode *inode, - struct reiserfs_security_handle *sec) -{ - return 0; -} -static inline void reiserfs_security_free(struct reiserfs_security_handle *sec) -{} -#endif - -#endif /* __KERNEL__ */ - #endif /* _LINUX_REISERFS_XATTR_H */ From a3063ab88fcbe5249f841cb95dfd626b8bf2674f Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 17 Mar 2012 01:03:10 -0400 Subject: [PATCH 31/64] move reiserfs_acl.h to fs/reiserfs/acl.h Signed-off-by: Al Viro --- include/linux/reiserfs_acl.h => fs/reiserfs/acl.h | 0 fs/reiserfs/file.c | 2 +- fs/reiserfs/inode.c | 2 +- fs/reiserfs/namei.c | 2 +- fs/reiserfs/super.c | 2 +- fs/reiserfs/xattr.c | 2 +- fs/reiserfs/xattr_acl.c | 2 +- 7 files changed, 6 insertions(+), 6 deletions(-) rename include/linux/reiserfs_acl.h => fs/reiserfs/acl.h (100%) diff --git a/include/linux/reiserfs_acl.h b/fs/reiserfs/acl.h similarity index 100% rename from include/linux/reiserfs_acl.h rename to fs/reiserfs/acl.h diff --git a/fs/reiserfs/file.c b/fs/reiserfs/file.c index e26ee4988e78..3fa5915dea6e 100644 --- a/fs/reiserfs/file.c +++ b/fs/reiserfs/file.c @@ -4,7 +4,7 @@ #include #include -#include +#include "acl.h" #include "xattr.h" #include #include diff --git a/fs/reiserfs/inode.c b/fs/reiserfs/inode.c index 894c7316e9c6..b696493d6b66 100644 --- a/fs/reiserfs/inode.c +++ b/fs/reiserfs/inode.c @@ -5,7 +5,7 @@ #include #include #include -#include +#include "acl.h" #include "xattr.h" #include #include diff --git a/fs/reiserfs/namei.c b/fs/reiserfs/namei.c index cb67ebf882a9..34bdab29883b 100644 --- a/fs/reiserfs/namei.c +++ b/fs/reiserfs/namei.c @@ -15,7 +15,7 @@ #include #include #include -#include +#include "acl.h" #include "xattr.h" #include diff --git a/fs/reiserfs/super.c b/fs/reiserfs/super.c index a67fc7d28de5..cf68a6ba0ec6 100644 --- a/fs/reiserfs/super.c +++ b/fs/reiserfs/super.c @@ -17,7 +17,7 @@ #include #include #include -#include +#include "acl.h" #include "xattr.h" #include #include diff --git a/fs/reiserfs/xattr.c b/fs/reiserfs/xattr.c index ead5d8aab440..61c9b5633e27 100644 --- a/fs/reiserfs/xattr.c +++ b/fs/reiserfs/xattr.c @@ -44,7 +44,7 @@ #include #include #include "xattr.h" -#include +#include "acl.h" #include #include #include diff --git a/fs/reiserfs/xattr_acl.c b/fs/reiserfs/xattr_acl.c index c0a8c519b2e1..f09094057eaa 100644 --- a/fs/reiserfs/xattr_acl.c +++ b/fs/reiserfs/xattr_acl.c @@ -8,7 +8,7 @@ #include #include #include "xattr.h" -#include +#include "acl.h" #include static int reiserfs_set_acl(struct reiserfs_transaction_handle *th, From a8a4b79b53fc7cbb023afedf58b04dd4e9bbb114 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 17 Mar 2012 01:06:11 -0400 Subject: [PATCH 32/64] kill pointless includes of reiserfs_fs_{i,sb}.h Signed-off-by: Al Viro --- fs/reiserfs/bitmap.c | 2 -- fs/reiserfs/objectid.c | 1 - fs/reiserfs/procfs.c | 1 - fs/reiserfs/resize.c | 1 - fs/reiserfs/xattr.h | 1 - 5 files changed, 6 deletions(-) diff --git a/fs/reiserfs/bitmap.c b/fs/reiserfs/bitmap.c index 70de42f09f1d..3c4636162621 100644 --- a/fs/reiserfs/bitmap.c +++ b/fs/reiserfs/bitmap.c @@ -10,8 +10,6 @@ #include #include #include -#include -#include #include #include diff --git a/fs/reiserfs/objectid.c b/fs/reiserfs/objectid.c index 3a6de810bd61..efc929e6a323 100644 --- a/fs/reiserfs/objectid.c +++ b/fs/reiserfs/objectid.c @@ -6,7 +6,6 @@ #include #include #include -#include // find where objectid map starts #define objectid_map(s,rs) (old_format_only (s) ? \ diff --git a/fs/reiserfs/procfs.c b/fs/reiserfs/procfs.c index 7a9981196c1c..f931a089bbe7 100644 --- a/fs/reiserfs/procfs.c +++ b/fs/reiserfs/procfs.c @@ -13,7 +13,6 @@ #include #include #include -#include #include #include diff --git a/fs/reiserfs/resize.c b/fs/reiserfs/resize.c index 7483279b482d..e1415ad93251 100644 --- a/fs/reiserfs/resize.c +++ b/fs/reiserfs/resize.c @@ -14,7 +14,6 @@ #include #include #include -#include #include int reiserfs_resize(struct super_block *s, unsigned long block_count_new) diff --git a/fs/reiserfs/xattr.h b/fs/reiserfs/xattr.h index 367c0459db6d..ccd146bb0665 100644 --- a/fs/reiserfs/xattr.h +++ b/fs/reiserfs/xattr.h @@ -2,7 +2,6 @@ #include #include #include -#include #include struct inode; From f466c6fdb3b1f043ff1977a8d2a1d0cd4dc164fa Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 17 Mar 2012 01:16:43 -0400 Subject: [PATCH 33/64] move private bits of reiserfs_fs.h to fs/reiserfs/reiserfs.h Signed-off-by: Al Viro --- fs/reiserfs/bitmap.c | 2 +- fs/reiserfs/dir.c | 2 +- fs/reiserfs/do_balan.c | 2 +- fs/reiserfs/file.c | 2 +- fs/reiserfs/fix_node.c | 2 +- fs/reiserfs/hashes.c | 2 +- fs/reiserfs/ibalance.c | 2 +- fs/reiserfs/inode.c | 2 +- fs/reiserfs/ioctl.c | 2 +- fs/reiserfs/item_ops.c | 2 +- fs/reiserfs/journal.c | 2 +- fs/reiserfs/lbalance.c | 2 +- fs/reiserfs/lock.c | 2 +- fs/reiserfs/namei.c | 2 +- fs/reiserfs/objectid.c | 2 +- fs/reiserfs/prints.c | 4 +- fs/reiserfs/procfs.c | 2 +- fs/reiserfs/reiserfs.h | 2327 ++++++++++++++++++++++++++++++++ fs/reiserfs/resize.c | 2 +- fs/reiserfs/stree.c | 2 +- fs/reiserfs/super.c | 2 +- fs/reiserfs/tail_conversion.c | 2 +- fs/reiserfs/xattr.c | 2 +- fs/reiserfs/xattr.h | 1 - fs/reiserfs/xattr_acl.c | 2 +- fs/reiserfs/xattr_security.c | 2 +- fs/reiserfs/xattr_trusted.c | 2 +- fs/reiserfs/xattr_user.c | 2 +- include/linux/reiserfs_fs.h | 2334 --------------------------------- 29 files changed, 2354 insertions(+), 2362 deletions(-) create mode 100644 fs/reiserfs/reiserfs.h diff --git a/fs/reiserfs/bitmap.c b/fs/reiserfs/bitmap.c index 3c4636162621..4c0c7d163d15 100644 --- a/fs/reiserfs/bitmap.c +++ b/fs/reiserfs/bitmap.c @@ -4,7 +4,7 @@ /* Reiserfs block (de)allocator, bitmap-based. */ #include -#include +#include "reiserfs.h" #include #include #include diff --git a/fs/reiserfs/dir.c b/fs/reiserfs/dir.c index 133e9355dc6f..66c53b642a88 100644 --- a/fs/reiserfs/dir.c +++ b/fs/reiserfs/dir.c @@ -5,7 +5,7 @@ #include #include #include -#include +#include "reiserfs.h" #include #include #include diff --git a/fs/reiserfs/do_balan.c b/fs/reiserfs/do_balan.c index 60c080440661..2b7882b508db 100644 --- a/fs/reiserfs/do_balan.c +++ b/fs/reiserfs/do_balan.c @@ -17,7 +17,7 @@ #include #include -#include +#include "reiserfs.h" #include #include diff --git a/fs/reiserfs/file.c b/fs/reiserfs/file.c index 3fa5915dea6e..8375c922c0d5 100644 --- a/fs/reiserfs/file.c +++ b/fs/reiserfs/file.c @@ -3,7 +3,7 @@ */ #include -#include +#include "reiserfs.h" #include "acl.h" #include "xattr.h" #include diff --git a/fs/reiserfs/fix_node.c b/fs/reiserfs/fix_node.c index 1e4250bc3a6f..430e0658704c 100644 --- a/fs/reiserfs/fix_node.c +++ b/fs/reiserfs/fix_node.c @@ -37,7 +37,7 @@ #include #include #include -#include +#include "reiserfs.h" #include /* To make any changes in the tree we find a node, that contains item diff --git a/fs/reiserfs/hashes.c b/fs/reiserfs/hashes.c index 6471c670743e..91b0cc1242a2 100644 --- a/fs/reiserfs/hashes.c +++ b/fs/reiserfs/hashes.c @@ -19,7 +19,7 @@ // #include -#include +#include "reiserfs.h" #include #define DELTA 0x9E3779B9 diff --git a/fs/reiserfs/ibalance.c b/fs/reiserfs/ibalance.c index 2074fd95046b..e1978fd895f5 100644 --- a/fs/reiserfs/ibalance.c +++ b/fs/reiserfs/ibalance.c @@ -5,7 +5,7 @@ #include #include #include -#include +#include "reiserfs.h" #include /* this is one and only function that is used outside (do_balance.c) */ diff --git a/fs/reiserfs/inode.c b/fs/reiserfs/inode.c index b696493d6b66..494c315c7417 100644 --- a/fs/reiserfs/inode.c +++ b/fs/reiserfs/inode.c @@ -4,7 +4,7 @@ #include #include -#include +#include "reiserfs.h" #include "acl.h" #include "xattr.h" #include diff --git a/fs/reiserfs/ioctl.c b/fs/reiserfs/ioctl.c index 950e3d1b5c9e..0c2185042d5f 100644 --- a/fs/reiserfs/ioctl.c +++ b/fs/reiserfs/ioctl.c @@ -5,7 +5,7 @@ #include #include #include -#include +#include "reiserfs.h" #include #include #include diff --git a/fs/reiserfs/item_ops.c b/fs/reiserfs/item_ops.c index 72cb1cc51b87..ee382ef3d300 100644 --- a/fs/reiserfs/item_ops.c +++ b/fs/reiserfs/item_ops.c @@ -3,7 +3,7 @@ */ #include -#include +#include "reiserfs.h" // this contains item handlers for old item types: sd, direct, // indirect, directory diff --git a/fs/reiserfs/journal.c b/fs/reiserfs/journal.c index c3cf54fd4de3..cf9f4de00a95 100644 --- a/fs/reiserfs/journal.c +++ b/fs/reiserfs/journal.c @@ -37,7 +37,7 @@ #include #include #include -#include +#include "reiserfs.h" #include #include #include diff --git a/fs/reiserfs/lbalance.c b/fs/reiserfs/lbalance.c index 03d85cbf90bf..c4b73f9ccf85 100644 --- a/fs/reiserfs/lbalance.c +++ b/fs/reiserfs/lbalance.c @@ -5,7 +5,7 @@ #include #include #include -#include +#include "reiserfs.h" #include /* these are used in do_balance.c */ diff --git a/fs/reiserfs/lock.c b/fs/reiserfs/lock.c index 7df1ce48203a..d735bc8470e3 100644 --- a/fs/reiserfs/lock.c +++ b/fs/reiserfs/lock.c @@ -1,4 +1,4 @@ -#include +#include "reiserfs.h" #include /* diff --git a/fs/reiserfs/namei.c b/fs/reiserfs/namei.c index 34bdab29883b..84e8a69cee9d 100644 --- a/fs/reiserfs/namei.c +++ b/fs/reiserfs/namei.c @@ -14,7 +14,7 @@ #include #include #include -#include +#include "reiserfs.h" #include "acl.h" #include "xattr.h" #include diff --git a/fs/reiserfs/objectid.c b/fs/reiserfs/objectid.c index efc929e6a323..f732d6a5251d 100644 --- a/fs/reiserfs/objectid.c +++ b/fs/reiserfs/objectid.c @@ -5,7 +5,7 @@ #include #include #include -#include +#include "reiserfs.h" // find where objectid map starts #define objectid_map(s,rs) (old_format_only (s) ? \ diff --git a/fs/reiserfs/prints.c b/fs/reiserfs/prints.c index 45de98b59466..c0b1112ab7e3 100644 --- a/fs/reiserfs/prints.c +++ b/fs/reiserfs/prints.c @@ -4,7 +4,7 @@ #include #include -#include +#include "reiserfs.h" #include #include @@ -329,7 +329,7 @@ void reiserfs_debug(struct super_block *s, int level, const char *fmt, ...) Numbering scheme for panic used by Vladimir and Anatoly( Hans completely ignores this scheme, and considers it pointless complexity): - panics in reiserfs_fs.h have numbers from 1000 to 1999 + panics in reiserfs.h have numbers from 1000 to 1999 super.c 2000 to 2999 preserve.c (unused) 3000 to 3999 bitmap.c 4000 to 4999 diff --git a/fs/reiserfs/procfs.c b/fs/reiserfs/procfs.c index f931a089bbe7..2c1ade692cc8 100644 --- a/fs/reiserfs/procfs.c +++ b/fs/reiserfs/procfs.c @@ -12,7 +12,7 @@ #include #include #include -#include +#include "reiserfs.h" #include #include diff --git a/fs/reiserfs/reiserfs.h b/fs/reiserfs/reiserfs.h new file mode 100644 index 000000000000..b3865c84f54c --- /dev/null +++ b/fs/reiserfs/reiserfs.h @@ -0,0 +1,2327 @@ +/* + * Copyright 1996, 1997, 1998 Hans Reiser, see reiserfs/README for licensing and copyright details + */ + +#include + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +/* the 32 bit compat definitions with int argument */ +#define REISERFS_IOC32_UNPACK _IOW(0xCD, 1, int) +#define REISERFS_IOC32_GETFLAGS FS_IOC32_GETFLAGS +#define REISERFS_IOC32_SETFLAGS FS_IOC32_SETFLAGS +#define REISERFS_IOC32_GETVERSION FS_IOC32_GETVERSION +#define REISERFS_IOC32_SETVERSION FS_IOC32_SETVERSION + +/* + * Locking primitives. The write lock is a per superblock + * special mutex that has properties close to the Big Kernel Lock + * which was used in the previous locking scheme. + */ +void reiserfs_write_lock(struct super_block *s); +void reiserfs_write_unlock(struct super_block *s); +int reiserfs_write_lock_once(struct super_block *s); +void reiserfs_write_unlock_once(struct super_block *s, int lock_depth); + +#ifdef CONFIG_REISERFS_CHECK +void reiserfs_lock_check_recursive(struct super_block *s); +#else +static inline void reiserfs_lock_check_recursive(struct super_block *s) { } +#endif + +/* + * Several mutexes depend on the write lock. + * However sometimes we want to relax the write lock while we hold + * these mutexes, according to the release/reacquire on schedule() + * properties of the Bkl that were used. + * Reiserfs performances and locking were based on this scheme. + * Now that the write lock is a mutex and not the bkl anymore, doing so + * may result in a deadlock: + * + * A acquire write_lock + * A acquire j_commit_mutex + * A release write_lock and wait for something + * B acquire write_lock + * B can't acquire j_commit_mutex and sleep + * A can't acquire write lock anymore + * deadlock + * + * What we do here is avoiding such deadlock by playing the same game + * than the Bkl: if we can't acquire a mutex that depends on the write lock, + * we release the write lock, wait a bit and then retry. + * + * The mutexes concerned by this hack are: + * - The commit mutex of a journal list + * - The flush mutex + * - The journal lock + * - The inode mutex + */ +static inline void reiserfs_mutex_lock_safe(struct mutex *m, + struct super_block *s) +{ + reiserfs_lock_check_recursive(s); + reiserfs_write_unlock(s); + mutex_lock(m); + reiserfs_write_lock(s); +} + +static inline void +reiserfs_mutex_lock_nested_safe(struct mutex *m, unsigned int subclass, + struct super_block *s) +{ + reiserfs_lock_check_recursive(s); + reiserfs_write_unlock(s); + mutex_lock_nested(m, subclass); + reiserfs_write_lock(s); +} + +static inline void +reiserfs_down_read_safe(struct rw_semaphore *sem, struct super_block *s) +{ + reiserfs_lock_check_recursive(s); + reiserfs_write_unlock(s); + down_read(sem); + reiserfs_write_lock(s); +} + +/* + * When we schedule, we usually want to also release the write lock, + * according to the previous bkl based locking scheme of reiserfs. + */ +static inline void reiserfs_cond_resched(struct super_block *s) +{ + if (need_resched()) { + reiserfs_write_unlock(s); + schedule(); + reiserfs_write_lock(s); + } +} + +struct fid; + +/* in reading the #defines, it may help to understand that they employ + the following abbreviations: + + B = Buffer + I = Item header + H = Height within the tree (should be changed to LEV) + N = Number of the item in the node + STAT = stat data + DEH = Directory Entry Header + EC = Entry Count + E = Entry number + UL = Unsigned Long + BLKH = BLocK Header + UNFM = UNForMatted node + DC = Disk Child + P = Path + + These #defines are named by concatenating these abbreviations, + where first comes the arguments, and last comes the return value, + of the macro. + +*/ + +#define USE_INODE_GENERATION_COUNTER + +#define REISERFS_PREALLOCATE +#define DISPLACE_NEW_PACKING_LOCALITIES +#define PREALLOCATION_SIZE 9 + +/* n must be power of 2 */ +#define _ROUND_UP(x,n) (((x)+(n)-1u) & ~((n)-1u)) + +// to be ok for alpha and others we have to align structures to 8 byte +// boundary. +// FIXME: do not change 4 by anything else: there is code which relies on that +#define ROUND_UP(x) _ROUND_UP(x,8LL) + +/* debug levels. Right now, CONFIG_REISERFS_CHECK means print all debug +** messages. +*/ +#define REISERFS_DEBUG_CODE 5 /* extra messages to help find/debug errors */ + +void __reiserfs_warning(struct super_block *s, const char *id, + const char *func, const char *fmt, ...); +#define reiserfs_warning(s, id, fmt, args...) \ + __reiserfs_warning(s, id, __func__, fmt, ##args) +/* assertions handling */ + +/** always check a condition and panic if it's false. */ +#define __RASSERT(cond, scond, format, args...) \ +do { \ + if (!(cond)) \ + reiserfs_panic(NULL, "assertion failure", "(" #cond ") at " \ + __FILE__ ":%i:%s: " format "\n", \ + in_interrupt() ? -1 : task_pid_nr(current), \ + __LINE__, __func__ , ##args); \ +} while (0) + +#define RASSERT(cond, format, args...) __RASSERT(cond, #cond, format, ##args) + +#if defined( CONFIG_REISERFS_CHECK ) +#define RFALSE(cond, format, args...) __RASSERT(!(cond), "!(" #cond ")", format, ##args) +#else +#define RFALSE( cond, format, args... ) do {;} while( 0 ) +#endif + +#define CONSTF __attribute_const__ +/* + * Disk Data Structures + */ + +/***************************************************************************/ +/* SUPER BLOCK */ +/***************************************************************************/ + +/* + * Structure of super block on disk, a version of which in RAM is often accessed as REISERFS_SB(s)->s_rs + * the version in RAM is part of a larger structure containing fields never written to disk. + */ +#define UNSET_HASH 0 // read_super will guess about, what hash names + // in directories were sorted with +#define TEA_HASH 1 +#define YURA_HASH 2 +#define R5_HASH 3 +#define DEFAULT_HASH R5_HASH + +struct journal_params { + __le32 jp_journal_1st_block; /* where does journal start from on its + * device */ + __le32 jp_journal_dev; /* journal device st_rdev */ + __le32 jp_journal_size; /* size of the journal */ + __le32 jp_journal_trans_max; /* max number of blocks in a transaction. */ + __le32 jp_journal_magic; /* random value made on fs creation (this + * was sb_journal_block_count) */ + __le32 jp_journal_max_batch; /* max number of blocks to batch into a + * trans */ + __le32 jp_journal_max_commit_age; /* in seconds, how old can an async + * commit be */ + __le32 jp_journal_max_trans_age; /* in seconds, how old can a transaction + * be */ +}; + +/* this is the super from 3.5.X, where X >= 10 */ +struct reiserfs_super_block_v1 { + __le32 s_block_count; /* blocks count */ + __le32 s_free_blocks; /* free blocks count */ + __le32 s_root_block; /* root block number */ + struct journal_params s_journal; + __le16 s_blocksize; /* block size */ + __le16 s_oid_maxsize; /* max size of object id array, see + * get_objectid() commentary */ + __le16 s_oid_cursize; /* current size of object id array */ + __le16 s_umount_state; /* this is set to 1 when filesystem was + * umounted, to 2 - when not */ + char s_magic[10]; /* reiserfs magic string indicates that + * file system is reiserfs: + * "ReIsErFs" or "ReIsEr2Fs" or "ReIsEr3Fs" */ + __le16 s_fs_state; /* it is set to used by fsck to mark which + * phase of rebuilding is done */ + __le32 s_hash_function_code; /* indicate, what hash function is being use + * to sort names in a directory*/ + __le16 s_tree_height; /* height of disk tree */ + __le16 s_bmap_nr; /* amount of bitmap blocks needed to address + * each block of file system */ + __le16 s_version; /* this field is only reliable on filesystem + * with non-standard journal */ + __le16 s_reserved_for_journal; /* size in blocks of journal area on main + * device, we need to keep after + * making fs with non-standard journal */ +} __attribute__ ((__packed__)); + +#define SB_SIZE_V1 (sizeof(struct reiserfs_super_block_v1)) + +/* this is the on disk super block */ +struct reiserfs_super_block { + struct reiserfs_super_block_v1 s_v1; + __le32 s_inode_generation; + __le32 s_flags; /* Right now used only by inode-attributes, if enabled */ + unsigned char s_uuid[16]; /* filesystem unique identifier */ + unsigned char s_label[16]; /* filesystem volume label */ + __le16 s_mnt_count; /* Count of mounts since last fsck */ + __le16 s_max_mnt_count; /* Maximum mounts before check */ + __le32 s_lastcheck; /* Timestamp of last fsck */ + __le32 s_check_interval; /* Interval between checks */ + char s_unused[76]; /* zero filled by mkreiserfs and + * reiserfs_convert_objectid_map_v1() + * so any additions must be updated + * there as well. */ +} __attribute__ ((__packed__)); + +#define SB_SIZE (sizeof(struct reiserfs_super_block)) + +#define REISERFS_VERSION_1 0 +#define REISERFS_VERSION_2 2 + +// on-disk super block fields converted to cpu form +#define SB_DISK_SUPER_BLOCK(s) (REISERFS_SB(s)->s_rs) +#define SB_V1_DISK_SUPER_BLOCK(s) (&(SB_DISK_SUPER_BLOCK(s)->s_v1)) +#define SB_BLOCKSIZE(s) \ + le32_to_cpu ((SB_V1_DISK_SUPER_BLOCK(s)->s_blocksize)) +#define SB_BLOCK_COUNT(s) \ + le32_to_cpu ((SB_V1_DISK_SUPER_BLOCK(s)->s_block_count)) +#define SB_FREE_BLOCKS(s) \ + le32_to_cpu ((SB_V1_DISK_SUPER_BLOCK(s)->s_free_blocks)) +#define SB_REISERFS_MAGIC(s) \ + (SB_V1_DISK_SUPER_BLOCK(s)->s_magic) +#define SB_ROOT_BLOCK(s) \ + le32_to_cpu ((SB_V1_DISK_SUPER_BLOCK(s)->s_root_block)) +#define SB_TREE_HEIGHT(s) \ + le16_to_cpu ((SB_V1_DISK_SUPER_BLOCK(s)->s_tree_height)) +#define SB_REISERFS_STATE(s) \ + le16_to_cpu ((SB_V1_DISK_SUPER_BLOCK(s)->s_umount_state)) +#define SB_VERSION(s) le16_to_cpu ((SB_V1_DISK_SUPER_BLOCK(s)->s_version)) +#define SB_BMAP_NR(s) le16_to_cpu ((SB_V1_DISK_SUPER_BLOCK(s)->s_bmap_nr)) + +#define PUT_SB_BLOCK_COUNT(s, val) \ + do { SB_V1_DISK_SUPER_BLOCK(s)->s_block_count = cpu_to_le32(val); } while (0) +#define PUT_SB_FREE_BLOCKS(s, val) \ + do { SB_V1_DISK_SUPER_BLOCK(s)->s_free_blocks = cpu_to_le32(val); } while (0) +#define PUT_SB_ROOT_BLOCK(s, val) \ + do { SB_V1_DISK_SUPER_BLOCK(s)->s_root_block = cpu_to_le32(val); } while (0) +#define PUT_SB_TREE_HEIGHT(s, val) \ + do { SB_V1_DISK_SUPER_BLOCK(s)->s_tree_height = cpu_to_le16(val); } while (0) +#define PUT_SB_REISERFS_STATE(s, val) \ + do { SB_V1_DISK_SUPER_BLOCK(s)->s_umount_state = cpu_to_le16(val); } while (0) +#define PUT_SB_VERSION(s, val) \ + do { SB_V1_DISK_SUPER_BLOCK(s)->s_version = cpu_to_le16(val); } while (0) +#define PUT_SB_BMAP_NR(s, val) \ + do { SB_V1_DISK_SUPER_BLOCK(s)->s_bmap_nr = cpu_to_le16 (val); } while (0) + +#define SB_ONDISK_JP(s) (&SB_V1_DISK_SUPER_BLOCK(s)->s_journal) +#define SB_ONDISK_JOURNAL_SIZE(s) \ + le32_to_cpu ((SB_ONDISK_JP(s)->jp_journal_size)) +#define SB_ONDISK_JOURNAL_1st_BLOCK(s) \ + le32_to_cpu ((SB_ONDISK_JP(s)->jp_journal_1st_block)) +#define SB_ONDISK_JOURNAL_DEVICE(s) \ + le32_to_cpu ((SB_ONDISK_JP(s)->jp_journal_dev)) +#define SB_ONDISK_RESERVED_FOR_JOURNAL(s) \ + le16_to_cpu ((SB_V1_DISK_SUPER_BLOCK(s)->s_reserved_for_journal)) + +#define is_block_in_log_or_reserved_area(s, block) \ + block >= SB_JOURNAL_1st_RESERVED_BLOCK(s) \ + && block < SB_JOURNAL_1st_RESERVED_BLOCK(s) + \ + ((!is_reiserfs_jr(SB_DISK_SUPER_BLOCK(s)) ? \ + SB_ONDISK_JOURNAL_SIZE(s) + 1 : SB_ONDISK_RESERVED_FOR_JOURNAL(s))) + +int is_reiserfs_3_5(struct reiserfs_super_block *rs); +int is_reiserfs_3_6(struct reiserfs_super_block *rs); +int is_reiserfs_jr(struct reiserfs_super_block *rs); + +/* ReiserFS leaves the first 64k unused, so that partition labels have + enough space. If someone wants to write a fancy bootloader that + needs more than 64k, let us know, and this will be increased in size. + This number must be larger than than the largest block size on any + platform, or code will break. -Hans */ +#define REISERFS_DISK_OFFSET_IN_BYTES (64 * 1024) +#define REISERFS_FIRST_BLOCK unused_define +#define REISERFS_JOURNAL_OFFSET_IN_BYTES REISERFS_DISK_OFFSET_IN_BYTES + +/* the spot for the super in versions 3.5 - 3.5.10 (inclusive) */ +#define REISERFS_OLD_DISK_OFFSET_IN_BYTES (8 * 1024) + +/* reiserfs internal error code (used by search_by_key and fix_nodes)) */ +#define CARRY_ON 0 +#define REPEAT_SEARCH -1 +#define IO_ERROR -2 +#define NO_DISK_SPACE -3 +#define NO_BALANCING_NEEDED (-4) +#define NO_MORE_UNUSED_CONTIGUOUS_BLOCKS (-5) +#define QUOTA_EXCEEDED -6 + +typedef __u32 b_blocknr_t; +typedef __le32 unp_t; + +struct unfm_nodeinfo { + unp_t unfm_nodenum; + unsigned short unfm_freespace; +}; + +/* there are two formats of keys: 3.5 and 3.6 + */ +#define KEY_FORMAT_3_5 0 +#define KEY_FORMAT_3_6 1 + +/* there are two stat datas */ +#define STAT_DATA_V1 0 +#define STAT_DATA_V2 1 + +static inline struct reiserfs_inode_info *REISERFS_I(const struct inode *inode) +{ + return container_of(inode, struct reiserfs_inode_info, vfs_inode); +} + +static inline struct reiserfs_sb_info *REISERFS_SB(const struct super_block *sb) +{ + return sb->s_fs_info; +} + +/* Don't trust REISERFS_SB(sb)->s_bmap_nr, it's a u16 + * which overflows on large file systems. */ +static inline __u32 reiserfs_bmap_count(struct super_block *sb) +{ + return (SB_BLOCK_COUNT(sb) - 1) / (sb->s_blocksize * 8) + 1; +} + +static inline int bmap_would_wrap(unsigned bmap_nr) +{ + return bmap_nr > ((1LL << 16) - 1); +} + +/** this says about version of key of all items (but stat data) the + object consists of */ +#define get_inode_item_key_version( inode ) \ + ((REISERFS_I(inode)->i_flags & i_item_key_version_mask) ? KEY_FORMAT_3_6 : KEY_FORMAT_3_5) + +#define set_inode_item_key_version( inode, version ) \ + ({ if((version)==KEY_FORMAT_3_6) \ + REISERFS_I(inode)->i_flags |= i_item_key_version_mask; \ + else \ + REISERFS_I(inode)->i_flags &= ~i_item_key_version_mask; }) + +#define get_inode_sd_version(inode) \ + ((REISERFS_I(inode)->i_flags & i_stat_data_version_mask) ? STAT_DATA_V2 : STAT_DATA_V1) + +#define set_inode_sd_version(inode, version) \ + ({ if((version)==STAT_DATA_V2) \ + REISERFS_I(inode)->i_flags |= i_stat_data_version_mask; \ + else \ + REISERFS_I(inode)->i_flags &= ~i_stat_data_version_mask; }) + +/* This is an aggressive tail suppression policy, I am hoping it + improves our benchmarks. The principle behind it is that percentage + space saving is what matters, not absolute space saving. This is + non-intuitive, but it helps to understand it if you consider that the + cost to access 4 blocks is not much more than the cost to access 1 + block, if you have to do a seek and rotate. A tail risks a + non-linear disk access that is significant as a percentage of total + time cost for a 4 block file and saves an amount of space that is + less significant as a percentage of space, or so goes the hypothesis. + -Hans */ +#define STORE_TAIL_IN_UNFM_S1(n_file_size,n_tail_size,n_block_size) \ +(\ + (!(n_tail_size)) || \ + (((n_tail_size) > MAX_DIRECT_ITEM_LEN(n_block_size)) || \ + ( (n_file_size) >= (n_block_size) * 4 ) || \ + ( ( (n_file_size) >= (n_block_size) * 3 ) && \ + ( (n_tail_size) >= (MAX_DIRECT_ITEM_LEN(n_block_size))/4) ) || \ + ( ( (n_file_size) >= (n_block_size) * 2 ) && \ + ( (n_tail_size) >= (MAX_DIRECT_ITEM_LEN(n_block_size))/2) ) || \ + ( ( (n_file_size) >= (n_block_size) ) && \ + ( (n_tail_size) >= (MAX_DIRECT_ITEM_LEN(n_block_size) * 3)/4) ) ) \ +) + +/* Another strategy for tails, this one means only create a tail if all the + file would fit into one DIRECT item. + Primary intention for this one is to increase performance by decreasing + seeking. +*/ +#define STORE_TAIL_IN_UNFM_S2(n_file_size,n_tail_size,n_block_size) \ +(\ + (!(n_tail_size)) || \ + (((n_file_size) > MAX_DIRECT_ITEM_LEN(n_block_size)) ) \ +) + +/* + * values for s_umount_state field + */ +#define REISERFS_VALID_FS 1 +#define REISERFS_ERROR_FS 2 + +// +// there are 5 item types currently +// +#define TYPE_STAT_DATA 0 +#define TYPE_INDIRECT 1 +#define TYPE_DIRECT 2 +#define TYPE_DIRENTRY 3 +#define TYPE_MAXTYPE 3 +#define TYPE_ANY 15 // FIXME: comment is required + +/***************************************************************************/ +/* KEY & ITEM HEAD */ +/***************************************************************************/ + +// +// directories use this key as well as old files +// +struct offset_v1 { + __le32 k_offset; + __le32 k_uniqueness; +} __attribute__ ((__packed__)); + +struct offset_v2 { + __le64 v; +} __attribute__ ((__packed__)); + +static inline __u16 offset_v2_k_type(const struct offset_v2 *v2) +{ + __u8 type = le64_to_cpu(v2->v) >> 60; + return (type <= TYPE_MAXTYPE) ? type : TYPE_ANY; +} + +static inline void set_offset_v2_k_type(struct offset_v2 *v2, int type) +{ + v2->v = + (v2->v & cpu_to_le64(~0ULL >> 4)) | cpu_to_le64((__u64) type << 60); +} + +static inline loff_t offset_v2_k_offset(const struct offset_v2 *v2) +{ + return le64_to_cpu(v2->v) & (~0ULL >> 4); +} + +static inline void set_offset_v2_k_offset(struct offset_v2 *v2, loff_t offset) +{ + offset &= (~0ULL >> 4); + v2->v = (v2->v & cpu_to_le64(15ULL << 60)) | cpu_to_le64(offset); +} + +/* Key of an item determines its location in the S+tree, and + is composed of 4 components */ +struct reiserfs_key { + __le32 k_dir_id; /* packing locality: by default parent + directory object id */ + __le32 k_objectid; /* object identifier */ + union { + struct offset_v1 k_offset_v1; + struct offset_v2 k_offset_v2; + } __attribute__ ((__packed__)) u; +} __attribute__ ((__packed__)); + +struct in_core_key { + __u32 k_dir_id; /* packing locality: by default parent + directory object id */ + __u32 k_objectid; /* object identifier */ + __u64 k_offset; + __u8 k_type; +}; + +struct cpu_key { + struct in_core_key on_disk_key; + int version; + int key_length; /* 3 in all cases but direct2indirect and + indirect2direct conversion */ +}; + +/* Our function for comparing keys can compare keys of different + lengths. It takes as a parameter the length of the keys it is to + compare. These defines are used in determining what is to be passed + to it as that parameter. */ +#define REISERFS_FULL_KEY_LEN 4 +#define REISERFS_SHORT_KEY_LEN 2 + +/* The result of the key compare */ +#define FIRST_GREATER 1 +#define SECOND_GREATER -1 +#define KEYS_IDENTICAL 0 +#define KEY_FOUND 1 +#define KEY_NOT_FOUND 0 + +#define KEY_SIZE (sizeof(struct reiserfs_key)) +#define SHORT_KEY_SIZE (sizeof (__u32) + sizeof (__u32)) + +/* return values for search_by_key and clones */ +#define ITEM_FOUND 1 +#define ITEM_NOT_FOUND 0 +#define ENTRY_FOUND 1 +#define ENTRY_NOT_FOUND 0 +#define DIRECTORY_NOT_FOUND -1 +#define REGULAR_FILE_FOUND -2 +#define DIRECTORY_FOUND -3 +#define BYTE_FOUND 1 +#define BYTE_NOT_FOUND 0 +#define FILE_NOT_FOUND -1 + +#define POSITION_FOUND 1 +#define POSITION_NOT_FOUND 0 + +// return values for reiserfs_find_entry and search_by_entry_key +#define NAME_FOUND 1 +#define NAME_NOT_FOUND 0 +#define GOTO_PREVIOUS_ITEM 2 +#define NAME_FOUND_INVISIBLE 3 + +/* Everything in the filesystem is stored as a set of items. The + item head contains the key of the item, its free space (for + indirect items) and specifies the location of the item itself + within the block. */ + +struct item_head { + /* Everything in the tree is found by searching for it based on + * its key.*/ + struct reiserfs_key ih_key; + union { + /* The free space in the last unformatted node of an + indirect item if this is an indirect item. This + equals 0xFFFF iff this is a direct item or stat data + item. Note that the key, not this field, is used to + determine the item type, and thus which field this + union contains. */ + __le16 ih_free_space_reserved; + /* Iff this is a directory item, this field equals the + number of directory entries in the directory item. */ + __le16 ih_entry_count; + } __attribute__ ((__packed__)) u; + __le16 ih_item_len; /* total size of the item body */ + __le16 ih_item_location; /* an offset to the item body + * within the block */ + __le16 ih_version; /* 0 for all old items, 2 for new + ones. Highest bit is set by fsck + temporary, cleaned after all + done */ +} __attribute__ ((__packed__)); +/* size of item header */ +#define IH_SIZE (sizeof(struct item_head)) + +#define ih_free_space(ih) le16_to_cpu((ih)->u.ih_free_space_reserved) +#define ih_version(ih) le16_to_cpu((ih)->ih_version) +#define ih_entry_count(ih) le16_to_cpu((ih)->u.ih_entry_count) +#define ih_location(ih) le16_to_cpu((ih)->ih_item_location) +#define ih_item_len(ih) le16_to_cpu((ih)->ih_item_len) + +#define put_ih_free_space(ih, val) do { (ih)->u.ih_free_space_reserved = cpu_to_le16(val); } while(0) +#define put_ih_version(ih, val) do { (ih)->ih_version = cpu_to_le16(val); } while (0) +#define put_ih_entry_count(ih, val) do { (ih)->u.ih_entry_count = cpu_to_le16(val); } while (0) +#define put_ih_location(ih, val) do { (ih)->ih_item_location = cpu_to_le16(val); } while (0) +#define put_ih_item_len(ih, val) do { (ih)->ih_item_len = cpu_to_le16(val); } while (0) + +#define unreachable_item(ih) (ih_version(ih) & (1 << 15)) + +#define get_ih_free_space(ih) (ih_version (ih) == KEY_FORMAT_3_6 ? 0 : ih_free_space (ih)) +#define set_ih_free_space(ih,val) put_ih_free_space((ih), ((ih_version(ih) == KEY_FORMAT_3_6) ? 0 : (val))) + +/* these operate on indirect items, where you've got an array of ints +** at a possibly unaligned location. These are a noop on ia32 +** +** p is the array of __u32, i is the index into the array, v is the value +** to store there. +*/ +#define get_block_num(p, i) get_unaligned_le32((p) + (i)) +#define put_block_num(p, i, v) put_unaligned_le32((v), (p) + (i)) + +// +// in old version uniqueness field shows key type +// +#define V1_SD_UNIQUENESS 0 +#define V1_INDIRECT_UNIQUENESS 0xfffffffe +#define V1_DIRECT_UNIQUENESS 0xffffffff +#define V1_DIRENTRY_UNIQUENESS 500 +#define V1_ANY_UNIQUENESS 555 // FIXME: comment is required + +// +// here are conversion routines +// +static inline int uniqueness2type(__u32 uniqueness) CONSTF; +static inline int uniqueness2type(__u32 uniqueness) +{ + switch ((int)uniqueness) { + case V1_SD_UNIQUENESS: + return TYPE_STAT_DATA; + case V1_INDIRECT_UNIQUENESS: + return TYPE_INDIRECT; + case V1_DIRECT_UNIQUENESS: + return TYPE_DIRECT; + case V1_DIRENTRY_UNIQUENESS: + return TYPE_DIRENTRY; + case V1_ANY_UNIQUENESS: + default: + return TYPE_ANY; + } +} + +static inline __u32 type2uniqueness(int type) CONSTF; +static inline __u32 type2uniqueness(int type) +{ + switch (type) { + case TYPE_STAT_DATA: + return V1_SD_UNIQUENESS; + case TYPE_INDIRECT: + return V1_INDIRECT_UNIQUENESS; + case TYPE_DIRECT: + return V1_DIRECT_UNIQUENESS; + case TYPE_DIRENTRY: + return V1_DIRENTRY_UNIQUENESS; + case TYPE_ANY: + default: + return V1_ANY_UNIQUENESS; + } +} + +// +// key is pointer to on disk key which is stored in le, result is cpu, +// there is no way to get version of object from key, so, provide +// version to these defines +// +static inline loff_t le_key_k_offset(int version, + const struct reiserfs_key *key) +{ + return (version == KEY_FORMAT_3_5) ? + le32_to_cpu(key->u.k_offset_v1.k_offset) : + offset_v2_k_offset(&(key->u.k_offset_v2)); +} + +static inline loff_t le_ih_k_offset(const struct item_head *ih) +{ + return le_key_k_offset(ih_version(ih), &(ih->ih_key)); +} + +static inline loff_t le_key_k_type(int version, const struct reiserfs_key *key) +{ + return (version == KEY_FORMAT_3_5) ? + uniqueness2type(le32_to_cpu(key->u.k_offset_v1.k_uniqueness)) : + offset_v2_k_type(&(key->u.k_offset_v2)); +} + +static inline loff_t le_ih_k_type(const struct item_head *ih) +{ + return le_key_k_type(ih_version(ih), &(ih->ih_key)); +} + +static inline void set_le_key_k_offset(int version, struct reiserfs_key *key, + loff_t offset) +{ + (version == KEY_FORMAT_3_5) ? (void)(key->u.k_offset_v1.k_offset = cpu_to_le32(offset)) : /* jdm check */ + (void)(set_offset_v2_k_offset(&(key->u.k_offset_v2), offset)); +} + +static inline void set_le_ih_k_offset(struct item_head *ih, loff_t offset) +{ + set_le_key_k_offset(ih_version(ih), &(ih->ih_key), offset); +} + +static inline void set_le_key_k_type(int version, struct reiserfs_key *key, + int type) +{ + (version == KEY_FORMAT_3_5) ? + (void)(key->u.k_offset_v1.k_uniqueness = + cpu_to_le32(type2uniqueness(type))) + : (void)(set_offset_v2_k_type(&(key->u.k_offset_v2), type)); +} + +static inline void set_le_ih_k_type(struct item_head *ih, int type) +{ + set_le_key_k_type(ih_version(ih), &(ih->ih_key), type); +} + +static inline int is_direntry_le_key(int version, struct reiserfs_key *key) +{ + return le_key_k_type(version, key) == TYPE_DIRENTRY; +} + +static inline int is_direct_le_key(int version, struct reiserfs_key *key) +{ + return le_key_k_type(version, key) == TYPE_DIRECT; +} + +static inline int is_indirect_le_key(int version, struct reiserfs_key *key) +{ + return le_key_k_type(version, key) == TYPE_INDIRECT; +} + +static inline int is_statdata_le_key(int version, struct reiserfs_key *key) +{ + return le_key_k_type(version, key) == TYPE_STAT_DATA; +} + +// +// item header has version. +// +static inline int is_direntry_le_ih(struct item_head *ih) +{ + return is_direntry_le_key(ih_version(ih), &ih->ih_key); +} + +static inline int is_direct_le_ih(struct item_head *ih) +{ + return is_direct_le_key(ih_version(ih), &ih->ih_key); +} + +static inline int is_indirect_le_ih(struct item_head *ih) +{ + return is_indirect_le_key(ih_version(ih), &ih->ih_key); +} + +static inline int is_statdata_le_ih(struct item_head *ih) +{ + return is_statdata_le_key(ih_version(ih), &ih->ih_key); +} + +// +// key is pointer to cpu key, result is cpu +// +static inline loff_t cpu_key_k_offset(const struct cpu_key *key) +{ + return key->on_disk_key.k_offset; +} + +static inline loff_t cpu_key_k_type(const struct cpu_key *key) +{ + return key->on_disk_key.k_type; +} + +static inline void set_cpu_key_k_offset(struct cpu_key *key, loff_t offset) +{ + key->on_disk_key.k_offset = offset; +} + +static inline void set_cpu_key_k_type(struct cpu_key *key, int type) +{ + key->on_disk_key.k_type = type; +} + +static inline void cpu_key_k_offset_dec(struct cpu_key *key) +{ + key->on_disk_key.k_offset--; +} + +#define is_direntry_cpu_key(key) (cpu_key_k_type (key) == TYPE_DIRENTRY) +#define is_direct_cpu_key(key) (cpu_key_k_type (key) == TYPE_DIRECT) +#define is_indirect_cpu_key(key) (cpu_key_k_type (key) == TYPE_INDIRECT) +#define is_statdata_cpu_key(key) (cpu_key_k_type (key) == TYPE_STAT_DATA) + +/* are these used ? */ +#define is_direntry_cpu_ih(ih) (is_direntry_cpu_key (&((ih)->ih_key))) +#define is_direct_cpu_ih(ih) (is_direct_cpu_key (&((ih)->ih_key))) +#define is_indirect_cpu_ih(ih) (is_indirect_cpu_key (&((ih)->ih_key))) +#define is_statdata_cpu_ih(ih) (is_statdata_cpu_key (&((ih)->ih_key))) + +#define I_K_KEY_IN_ITEM(ih, key, n_blocksize) \ + (!COMP_SHORT_KEYS(ih, key) && \ + I_OFF_BYTE_IN_ITEM(ih, k_offset(key), n_blocksize)) + +/* maximal length of item */ +#define MAX_ITEM_LEN(block_size) (block_size - BLKH_SIZE - IH_SIZE) +#define MIN_ITEM_LEN 1 + +/* object identifier for root dir */ +#define REISERFS_ROOT_OBJECTID 2 +#define REISERFS_ROOT_PARENT_OBJECTID 1 + +extern struct reiserfs_key root_key; + +/* + * Picture represents a leaf of the S+tree + * ______________________________________________________ + * | | Array of | | | + * |Block | Object-Item | F r e e | Objects- | + * | head | Headers | S p a c e | Items | + * |______|_______________|___________________|___________| + */ + +/* Header of a disk block. More precisely, header of a formatted leaf + or internal node, and not the header of an unformatted node. */ +struct block_head { + __le16 blk_level; /* Level of a block in the tree. */ + __le16 blk_nr_item; /* Number of keys/items in a block. */ + __le16 blk_free_space; /* Block free space in bytes. */ + __le16 blk_reserved; + /* dump this in v4/planA */ + struct reiserfs_key blk_right_delim_key; /* kept only for compatibility */ +}; + +#define BLKH_SIZE (sizeof(struct block_head)) +#define blkh_level(p_blkh) (le16_to_cpu((p_blkh)->blk_level)) +#define blkh_nr_item(p_blkh) (le16_to_cpu((p_blkh)->blk_nr_item)) +#define blkh_free_space(p_blkh) (le16_to_cpu((p_blkh)->blk_free_space)) +#define blkh_reserved(p_blkh) (le16_to_cpu((p_blkh)->blk_reserved)) +#define set_blkh_level(p_blkh,val) ((p_blkh)->blk_level = cpu_to_le16(val)) +#define set_blkh_nr_item(p_blkh,val) ((p_blkh)->blk_nr_item = cpu_to_le16(val)) +#define set_blkh_free_space(p_blkh,val) ((p_blkh)->blk_free_space = cpu_to_le16(val)) +#define set_blkh_reserved(p_blkh,val) ((p_blkh)->blk_reserved = cpu_to_le16(val)) +#define blkh_right_delim_key(p_blkh) ((p_blkh)->blk_right_delim_key) +#define set_blkh_right_delim_key(p_blkh,val) ((p_blkh)->blk_right_delim_key = val) + +/* + * values for blk_level field of the struct block_head + */ + +#define FREE_LEVEL 0 /* when node gets removed from the tree its + blk_level is set to FREE_LEVEL. It is then + used to see whether the node is still in the + tree */ + +#define DISK_LEAF_NODE_LEVEL 1 /* Leaf node level. */ + +/* Given the buffer head of a formatted node, resolve to the block head of that node. */ +#define B_BLK_HEAD(bh) ((struct block_head *)((bh)->b_data)) +/* Number of items that are in buffer. */ +#define B_NR_ITEMS(bh) (blkh_nr_item(B_BLK_HEAD(bh))) +#define B_LEVEL(bh) (blkh_level(B_BLK_HEAD(bh))) +#define B_FREE_SPACE(bh) (blkh_free_space(B_BLK_HEAD(bh))) + +#define PUT_B_NR_ITEMS(bh, val) do { set_blkh_nr_item(B_BLK_HEAD(bh), val); } while (0) +#define PUT_B_LEVEL(bh, val) do { set_blkh_level(B_BLK_HEAD(bh), val); } while (0) +#define PUT_B_FREE_SPACE(bh, val) do { set_blkh_free_space(B_BLK_HEAD(bh), val); } while (0) + +/* Get right delimiting key. -- little endian */ +#define B_PRIGHT_DELIM_KEY(bh) (&(blk_right_delim_key(B_BLK_HEAD(bh)))) + +/* Does the buffer contain a disk leaf. */ +#define B_IS_ITEMS_LEVEL(bh) (B_LEVEL(bh) == DISK_LEAF_NODE_LEVEL) + +/* Does the buffer contain a disk internal node */ +#define B_IS_KEYS_LEVEL(bh) (B_LEVEL(bh) > DISK_LEAF_NODE_LEVEL \ + && B_LEVEL(bh) <= MAX_HEIGHT) + +/***************************************************************************/ +/* STAT DATA */ +/***************************************************************************/ + +// +// old stat data is 32 bytes long. We are going to distinguish new one by +// different size +// +struct stat_data_v1 { + __le16 sd_mode; /* file type, permissions */ + __le16 sd_nlink; /* number of hard links */ + __le16 sd_uid; /* owner */ + __le16 sd_gid; /* group */ + __le32 sd_size; /* file size */ + __le32 sd_atime; /* time of last access */ + __le32 sd_mtime; /* time file was last modified */ + __le32 sd_ctime; /* time inode (stat data) was last changed (except changes to sd_atime and sd_mtime) */ + union { + __le32 sd_rdev; + __le32 sd_blocks; /* number of blocks file uses */ + } __attribute__ ((__packed__)) u; + __le32 sd_first_direct_byte; /* first byte of file which is stored + in a direct item: except that if it + equals 1 it is a symlink and if it + equals ~(__u32)0 there is no + direct item. The existence of this + field really grates on me. Let's + replace it with a macro based on + sd_size and our tail suppression + policy. Someday. -Hans */ +} __attribute__ ((__packed__)); + +#define SD_V1_SIZE (sizeof(struct stat_data_v1)) +#define stat_data_v1(ih) (ih_version (ih) == KEY_FORMAT_3_5) +#define sd_v1_mode(sdp) (le16_to_cpu((sdp)->sd_mode)) +#define set_sd_v1_mode(sdp,v) ((sdp)->sd_mode = cpu_to_le16(v)) +#define sd_v1_nlink(sdp) (le16_to_cpu((sdp)->sd_nlink)) +#define set_sd_v1_nlink(sdp,v) ((sdp)->sd_nlink = cpu_to_le16(v)) +#define sd_v1_uid(sdp) (le16_to_cpu((sdp)->sd_uid)) +#define set_sd_v1_uid(sdp,v) ((sdp)->sd_uid = cpu_to_le16(v)) +#define sd_v1_gid(sdp) (le16_to_cpu((sdp)->sd_gid)) +#define set_sd_v1_gid(sdp,v) ((sdp)->sd_gid = cpu_to_le16(v)) +#define sd_v1_size(sdp) (le32_to_cpu((sdp)->sd_size)) +#define set_sd_v1_size(sdp,v) ((sdp)->sd_size = cpu_to_le32(v)) +#define sd_v1_atime(sdp) (le32_to_cpu((sdp)->sd_atime)) +#define set_sd_v1_atime(sdp,v) ((sdp)->sd_atime = cpu_to_le32(v)) +#define sd_v1_mtime(sdp) (le32_to_cpu((sdp)->sd_mtime)) +#define set_sd_v1_mtime(sdp,v) ((sdp)->sd_mtime = cpu_to_le32(v)) +#define sd_v1_ctime(sdp) (le32_to_cpu((sdp)->sd_ctime)) +#define set_sd_v1_ctime(sdp,v) ((sdp)->sd_ctime = cpu_to_le32(v)) +#define sd_v1_rdev(sdp) (le32_to_cpu((sdp)->u.sd_rdev)) +#define set_sd_v1_rdev(sdp,v) ((sdp)->u.sd_rdev = cpu_to_le32(v)) +#define sd_v1_blocks(sdp) (le32_to_cpu((sdp)->u.sd_blocks)) +#define set_sd_v1_blocks(sdp,v) ((sdp)->u.sd_blocks = cpu_to_le32(v)) +#define sd_v1_first_direct_byte(sdp) \ + (le32_to_cpu((sdp)->sd_first_direct_byte)) +#define set_sd_v1_first_direct_byte(sdp,v) \ + ((sdp)->sd_first_direct_byte = cpu_to_le32(v)) + +/* inode flags stored in sd_attrs (nee sd_reserved) */ + +/* we want common flags to have the same values as in ext2, + so chattr(1) will work without problems */ +#define REISERFS_IMMUTABLE_FL FS_IMMUTABLE_FL +#define REISERFS_APPEND_FL FS_APPEND_FL +#define REISERFS_SYNC_FL FS_SYNC_FL +#define REISERFS_NOATIME_FL FS_NOATIME_FL +#define REISERFS_NODUMP_FL FS_NODUMP_FL +#define REISERFS_SECRM_FL FS_SECRM_FL +#define REISERFS_UNRM_FL FS_UNRM_FL +#define REISERFS_COMPR_FL FS_COMPR_FL +#define REISERFS_NOTAIL_FL FS_NOTAIL_FL + +/* persistent flags that file inherits from the parent directory */ +#define REISERFS_INHERIT_MASK ( REISERFS_IMMUTABLE_FL | \ + REISERFS_SYNC_FL | \ + REISERFS_NOATIME_FL | \ + REISERFS_NODUMP_FL | \ + REISERFS_SECRM_FL | \ + REISERFS_COMPR_FL | \ + REISERFS_NOTAIL_FL ) + +/* Stat Data on disk (reiserfs version of UFS disk inode minus the + address blocks) */ +struct stat_data { + __le16 sd_mode; /* file type, permissions */ + __le16 sd_attrs; /* persistent inode flags */ + __le32 sd_nlink; /* number of hard links */ + __le64 sd_size; /* file size */ + __le32 sd_uid; /* owner */ + __le32 sd_gid; /* group */ + __le32 sd_atime; /* time of last access */ + __le32 sd_mtime; /* time file was last modified */ + __le32 sd_ctime; /* time inode (stat data) was last changed (except changes to sd_atime and sd_mtime) */ + __le32 sd_blocks; + union { + __le32 sd_rdev; + __le32 sd_generation; + //__le32 sd_first_direct_byte; + /* first byte of file which is stored in a + direct item: except that if it equals 1 + it is a symlink and if it equals + ~(__u32)0 there is no direct item. The + existence of this field really grates + on me. Let's replace it with a macro + based on sd_size and our tail + suppression policy? */ + } __attribute__ ((__packed__)) u; +} __attribute__ ((__packed__)); +// +// this is 44 bytes long +// +#define SD_SIZE (sizeof(struct stat_data)) +#define SD_V2_SIZE SD_SIZE +#define stat_data_v2(ih) (ih_version (ih) == KEY_FORMAT_3_6) +#define sd_v2_mode(sdp) (le16_to_cpu((sdp)->sd_mode)) +#define set_sd_v2_mode(sdp,v) ((sdp)->sd_mode = cpu_to_le16(v)) +/* sd_reserved */ +/* set_sd_reserved */ +#define sd_v2_nlink(sdp) (le32_to_cpu((sdp)->sd_nlink)) +#define set_sd_v2_nlink(sdp,v) ((sdp)->sd_nlink = cpu_to_le32(v)) +#define sd_v2_size(sdp) (le64_to_cpu((sdp)->sd_size)) +#define set_sd_v2_size(sdp,v) ((sdp)->sd_size = cpu_to_le64(v)) +#define sd_v2_uid(sdp) (le32_to_cpu((sdp)->sd_uid)) +#define set_sd_v2_uid(sdp,v) ((sdp)->sd_uid = cpu_to_le32(v)) +#define sd_v2_gid(sdp) (le32_to_cpu((sdp)->sd_gid)) +#define set_sd_v2_gid(sdp,v) ((sdp)->sd_gid = cpu_to_le32(v)) +#define sd_v2_atime(sdp) (le32_to_cpu((sdp)->sd_atime)) +#define set_sd_v2_atime(sdp,v) ((sdp)->sd_atime = cpu_to_le32(v)) +#define sd_v2_mtime(sdp) (le32_to_cpu((sdp)->sd_mtime)) +#define set_sd_v2_mtime(sdp,v) ((sdp)->sd_mtime = cpu_to_le32(v)) +#define sd_v2_ctime(sdp) (le32_to_cpu((sdp)->sd_ctime)) +#define set_sd_v2_ctime(sdp,v) ((sdp)->sd_ctime = cpu_to_le32(v)) +#define sd_v2_blocks(sdp) (le32_to_cpu((sdp)->sd_blocks)) +#define set_sd_v2_blocks(sdp,v) ((sdp)->sd_blocks = cpu_to_le32(v)) +#define sd_v2_rdev(sdp) (le32_to_cpu((sdp)->u.sd_rdev)) +#define set_sd_v2_rdev(sdp,v) ((sdp)->u.sd_rdev = cpu_to_le32(v)) +#define sd_v2_generation(sdp) (le32_to_cpu((sdp)->u.sd_generation)) +#define set_sd_v2_generation(sdp,v) ((sdp)->u.sd_generation = cpu_to_le32(v)) +#define sd_v2_attrs(sdp) (le16_to_cpu((sdp)->sd_attrs)) +#define set_sd_v2_attrs(sdp,v) ((sdp)->sd_attrs = cpu_to_le16(v)) + +/***************************************************************************/ +/* DIRECTORY STRUCTURE */ +/***************************************************************************/ +/* + Picture represents the structure of directory items + ________________________________________________ + | Array of | | | | | | + | directory |N-1| N-2 | .... | 1st |0th| + | entry headers | | | | | | + |_______________|___|_____|________|_______|___| + <---- directory entries ------> + + First directory item has k_offset component 1. We store "." and ".." + in one item, always, we never split "." and ".." into differing + items. This makes, among other things, the code for removing + directories simpler. */ +#define SD_OFFSET 0 +#define SD_UNIQUENESS 0 +#define DOT_OFFSET 1 +#define DOT_DOT_OFFSET 2 +#define DIRENTRY_UNIQUENESS 500 + +/* */ +#define FIRST_ITEM_OFFSET 1 + +/* + Q: How to get key of object pointed to by entry from entry? + + A: Each directory entry has its header. This header has deh_dir_id and deh_objectid fields, those are key + of object, entry points to */ + +/* NOT IMPLEMENTED: + Directory will someday contain stat data of object */ + +struct reiserfs_de_head { + __le32 deh_offset; /* third component of the directory entry key */ + __le32 deh_dir_id; /* objectid of the parent directory of the object, that is referenced + by directory entry */ + __le32 deh_objectid; /* objectid of the object, that is referenced by directory entry */ + __le16 deh_location; /* offset of name in the whole item */ + __le16 deh_state; /* whether 1) entry contains stat data (for future), and 2) whether + entry is hidden (unlinked) */ +} __attribute__ ((__packed__)); +#define DEH_SIZE sizeof(struct reiserfs_de_head) +#define deh_offset(p_deh) (le32_to_cpu((p_deh)->deh_offset)) +#define deh_dir_id(p_deh) (le32_to_cpu((p_deh)->deh_dir_id)) +#define deh_objectid(p_deh) (le32_to_cpu((p_deh)->deh_objectid)) +#define deh_location(p_deh) (le16_to_cpu((p_deh)->deh_location)) +#define deh_state(p_deh) (le16_to_cpu((p_deh)->deh_state)) + +#define put_deh_offset(p_deh,v) ((p_deh)->deh_offset = cpu_to_le32((v))) +#define put_deh_dir_id(p_deh,v) ((p_deh)->deh_dir_id = cpu_to_le32((v))) +#define put_deh_objectid(p_deh,v) ((p_deh)->deh_objectid = cpu_to_le32((v))) +#define put_deh_location(p_deh,v) ((p_deh)->deh_location = cpu_to_le16((v))) +#define put_deh_state(p_deh,v) ((p_deh)->deh_state = cpu_to_le16((v))) + +/* empty directory contains two entries "." and ".." and their headers */ +#define EMPTY_DIR_SIZE \ +(DEH_SIZE * 2 + ROUND_UP (strlen (".")) + ROUND_UP (strlen (".."))) + +/* old format directories have this size when empty */ +#define EMPTY_DIR_SIZE_V1 (DEH_SIZE * 2 + 3) + +#define DEH_Statdata 0 /* not used now */ +#define DEH_Visible 2 + +/* 64 bit systems (and the S/390) need to be aligned explicitly -jdm */ +#if BITS_PER_LONG == 64 || defined(__s390__) || defined(__hppa__) +# define ADDR_UNALIGNED_BITS (3) +#endif + +/* These are only used to manipulate deh_state. + * Because of this, we'll use the ext2_ bit routines, + * since they are little endian */ +#ifdef ADDR_UNALIGNED_BITS + +# define aligned_address(addr) ((void *)((long)(addr) & ~((1UL << ADDR_UNALIGNED_BITS) - 1))) +# define unaligned_offset(addr) (((int)((long)(addr) & ((1 << ADDR_UNALIGNED_BITS) - 1))) << 3) + +# define set_bit_unaligned(nr, addr) \ + __test_and_set_bit_le((nr) + unaligned_offset(addr), aligned_address(addr)) +# define clear_bit_unaligned(nr, addr) \ + __test_and_clear_bit_le((nr) + unaligned_offset(addr), aligned_address(addr)) +# define test_bit_unaligned(nr, addr) \ + test_bit_le((nr) + unaligned_offset(addr), aligned_address(addr)) + +#else + +# define set_bit_unaligned(nr, addr) __test_and_set_bit_le(nr, addr) +# define clear_bit_unaligned(nr, addr) __test_and_clear_bit_le(nr, addr) +# define test_bit_unaligned(nr, addr) test_bit_le(nr, addr) + +#endif + +#define mark_de_with_sd(deh) set_bit_unaligned (DEH_Statdata, &((deh)->deh_state)) +#define mark_de_without_sd(deh) clear_bit_unaligned (DEH_Statdata, &((deh)->deh_state)) +#define mark_de_visible(deh) set_bit_unaligned (DEH_Visible, &((deh)->deh_state)) +#define mark_de_hidden(deh) clear_bit_unaligned (DEH_Visible, &((deh)->deh_state)) + +#define de_with_sd(deh) test_bit_unaligned (DEH_Statdata, &((deh)->deh_state)) +#define de_visible(deh) test_bit_unaligned (DEH_Visible, &((deh)->deh_state)) +#define de_hidden(deh) !test_bit_unaligned (DEH_Visible, &((deh)->deh_state)) + +extern void make_empty_dir_item_v1(char *body, __le32 dirid, __le32 objid, + __le32 par_dirid, __le32 par_objid); +extern void make_empty_dir_item(char *body, __le32 dirid, __le32 objid, + __le32 par_dirid, __le32 par_objid); + +/* array of the entry headers */ + /* get item body */ +#define B_I_PITEM(bh,ih) ( (bh)->b_data + ih_location(ih) ) +#define B_I_DEH(bh,ih) ((struct reiserfs_de_head *)(B_I_PITEM(bh,ih))) + +/* length of the directory entry in directory item. This define + calculates length of i-th directory entry using directory entry + locations from dir entry head. When it calculates length of 0-th + directory entry, it uses length of whole item in place of entry + location of the non-existent following entry in the calculation. + See picture above.*/ +/* +#define I_DEH_N_ENTRY_LENGTH(ih,deh,i) \ +((i) ? (deh_location((deh)-1) - deh_location((deh))) : (ih_item_len((ih)) - deh_location((deh)))) +*/ +static inline int entry_length(const struct buffer_head *bh, + const struct item_head *ih, int pos_in_item) +{ + struct reiserfs_de_head *deh; + + deh = B_I_DEH(bh, ih) + pos_in_item; + if (pos_in_item) + return deh_location(deh - 1) - deh_location(deh); + + return ih_item_len(ih) - deh_location(deh); +} + +/* number of entries in the directory item, depends on ENTRY_COUNT being at the start of directory dynamic data. */ +#define I_ENTRY_COUNT(ih) (ih_entry_count((ih))) + +/* name by bh, ih and entry_num */ +#define B_I_E_NAME(bh,ih,entry_num) ((char *)(bh->b_data + ih_location(ih) + deh_location(B_I_DEH(bh,ih)+(entry_num)))) + +// two entries per block (at least) +#define REISERFS_MAX_NAME(block_size) 255 + +/* this structure is used for operations on directory entries. It is + not a disk structure. */ +/* When reiserfs_find_entry or search_by_entry_key find directory + entry, they return filled reiserfs_dir_entry structure */ +struct reiserfs_dir_entry { + struct buffer_head *de_bh; + int de_item_num; + struct item_head *de_ih; + int de_entry_num; + struct reiserfs_de_head *de_deh; + int de_entrylen; + int de_namelen; + char *de_name; + unsigned long *de_gen_number_bit_string; + + __u32 de_dir_id; + __u32 de_objectid; + + struct cpu_key de_entry_key; +}; + +/* these defines are useful when a particular member of a reiserfs_dir_entry is needed */ + +/* pointer to file name, stored in entry */ +#define B_I_DEH_ENTRY_FILE_NAME(bh,ih,deh) (B_I_PITEM (bh, ih) + deh_location(deh)) + +/* length of name */ +#define I_DEH_N_ENTRY_FILE_NAME_LENGTH(ih,deh,entry_num) \ +(I_DEH_N_ENTRY_LENGTH (ih, deh, entry_num) - (de_with_sd (deh) ? SD_SIZE : 0)) + +/* hash value occupies bits from 7 up to 30 */ +#define GET_HASH_VALUE(offset) ((offset) & 0x7fffff80LL) +/* generation number occupies 7 bits starting from 0 up to 6 */ +#define GET_GENERATION_NUMBER(offset) ((offset) & 0x7fLL) +#define MAX_GENERATION_NUMBER 127 + +#define SET_GENERATION_NUMBER(offset,gen_number) (GET_HASH_VALUE(offset)|(gen_number)) + +/* + * Picture represents an internal node of the reiserfs tree + * ______________________________________________________ + * | | Array of | Array of | Free | + * |block | keys | pointers | space | + * | head | N | N+1 | | + * |______|_______________|___________________|___________| + */ + +/***************************************************************************/ +/* DISK CHILD */ +/***************************************************************************/ +/* Disk child pointer: The pointer from an internal node of the tree + to a node that is on disk. */ +struct disk_child { + __le32 dc_block_number; /* Disk child's block number. */ + __le16 dc_size; /* Disk child's used space. */ + __le16 dc_reserved; +}; + +#define DC_SIZE (sizeof(struct disk_child)) +#define dc_block_number(dc_p) (le32_to_cpu((dc_p)->dc_block_number)) +#define dc_size(dc_p) (le16_to_cpu((dc_p)->dc_size)) +#define put_dc_block_number(dc_p, val) do { (dc_p)->dc_block_number = cpu_to_le32(val); } while(0) +#define put_dc_size(dc_p, val) do { (dc_p)->dc_size = cpu_to_le16(val); } while(0) + +/* Get disk child by buffer header and position in the tree node. */ +#define B_N_CHILD(bh, n_pos) ((struct disk_child *)\ +((bh)->b_data + BLKH_SIZE + B_NR_ITEMS(bh) * KEY_SIZE + DC_SIZE * (n_pos))) + +/* Get disk child number by buffer header and position in the tree node. */ +#define B_N_CHILD_NUM(bh, n_pos) (dc_block_number(B_N_CHILD(bh, n_pos))) +#define PUT_B_N_CHILD_NUM(bh, n_pos, val) \ + (put_dc_block_number(B_N_CHILD(bh, n_pos), val)) + + /* maximal value of field child_size in structure disk_child */ + /* child size is the combined size of all items and their headers */ +#define MAX_CHILD_SIZE(bh) ((int)( (bh)->b_size - BLKH_SIZE )) + +/* amount of used space in buffer (not including block head) */ +#define B_CHILD_SIZE(cur) (MAX_CHILD_SIZE(cur)-(B_FREE_SPACE(cur))) + +/* max and min number of keys in internal node */ +#define MAX_NR_KEY(bh) ( (MAX_CHILD_SIZE(bh)-DC_SIZE)/(KEY_SIZE+DC_SIZE) ) +#define MIN_NR_KEY(bh) (MAX_NR_KEY(bh)/2) + +/***************************************************************************/ +/* PATH STRUCTURES AND DEFINES */ +/***************************************************************************/ + +/* Search_by_key fills up the path from the root to the leaf as it descends the tree looking for the + key. It uses reiserfs_bread to try to find buffers in the cache given their block number. If it + does not find them in the cache it reads them from disk. For each node search_by_key finds using + reiserfs_bread it then uses bin_search to look through that node. bin_search will find the + position of the block_number of the next node if it is looking through an internal node. If it + is looking through a leaf node bin_search will find the position of the item which has key either + equal to given key, or which is the maximal key less than the given key. */ + +struct path_element { + struct buffer_head *pe_buffer; /* Pointer to the buffer at the path in the tree. */ + int pe_position; /* Position in the tree node which is placed in the */ + /* buffer above. */ +}; + +#define MAX_HEIGHT 5 /* maximal height of a tree. don't change this without changing JOURNAL_PER_BALANCE_CNT */ +#define EXTENDED_MAX_HEIGHT 7 /* Must be equals MAX_HEIGHT + FIRST_PATH_ELEMENT_OFFSET */ +#define FIRST_PATH_ELEMENT_OFFSET 2 /* Must be equal to at least 2. */ + +#define ILLEGAL_PATH_ELEMENT_OFFSET 1 /* Must be equal to FIRST_PATH_ELEMENT_OFFSET - 1 */ +#define MAX_FEB_SIZE 6 /* this MUST be MAX_HEIGHT + 1. See about FEB below */ + +/* We need to keep track of who the ancestors of nodes are. When we + perform a search we record which nodes were visited while + descending the tree looking for the node we searched for. This list + of nodes is called the path. This information is used while + performing balancing. Note that this path information may become + invalid, and this means we must check it when using it to see if it + is still valid. You'll need to read search_by_key and the comments + in it, especially about decrement_counters_in_path(), to understand + this structure. + +Paths make the code so much harder to work with and debug.... An +enormous number of bugs are due to them, and trying to write or modify +code that uses them just makes my head hurt. They are based on an +excessive effort to avoid disturbing the precious VFS code.:-( The +gods only know how we are going to SMP the code that uses them. +znodes are the way! */ + +#define PATH_READA 0x1 /* do read ahead */ +#define PATH_READA_BACK 0x2 /* read backwards */ + +struct treepath { + int path_length; /* Length of the array above. */ + int reada; + struct path_element path_elements[EXTENDED_MAX_HEIGHT]; /* Array of the path elements. */ + int pos_in_item; +}; + +#define pos_in_item(path) ((path)->pos_in_item) + +#define INITIALIZE_PATH(var) \ +struct treepath var = {.path_length = ILLEGAL_PATH_ELEMENT_OFFSET, .reada = 0,} + +/* Get path element by path and path position. */ +#define PATH_OFFSET_PELEMENT(path, n_offset) ((path)->path_elements + (n_offset)) + +/* Get buffer header at the path by path and path position. */ +#define PATH_OFFSET_PBUFFER(path, n_offset) (PATH_OFFSET_PELEMENT(path, n_offset)->pe_buffer) + +/* Get position in the element at the path by path and path position. */ +#define PATH_OFFSET_POSITION(path, n_offset) (PATH_OFFSET_PELEMENT(path, n_offset)->pe_position) + +#define PATH_PLAST_BUFFER(path) (PATH_OFFSET_PBUFFER((path), (path)->path_length)) + /* you know, to the person who didn't + write this the macro name does not + at first suggest what it does. + Maybe POSITION_FROM_PATH_END? Or + maybe we should just focus on + dumping paths... -Hans */ +#define PATH_LAST_POSITION(path) (PATH_OFFSET_POSITION((path), (path)->path_length)) + +#define PATH_PITEM_HEAD(path) B_N_PITEM_HEAD(PATH_PLAST_BUFFER(path), PATH_LAST_POSITION(path)) + +/* in do_balance leaf has h == 0 in contrast with path structure, + where root has level == 0. That is why we need these defines */ +#define PATH_H_PBUFFER(path, h) PATH_OFFSET_PBUFFER (path, path->path_length - (h)) /* tb->S[h] */ +#define PATH_H_PPARENT(path, h) PATH_H_PBUFFER (path, (h) + 1) /* tb->F[h] or tb->S[0]->b_parent */ +#define PATH_H_POSITION(path, h) PATH_OFFSET_POSITION (path, path->path_length - (h)) +#define PATH_H_B_ITEM_ORDER(path, h) PATH_H_POSITION(path, h + 1) /* tb->S[h]->b_item_order */ + +#define PATH_H_PATH_OFFSET(path, n_h) ((path)->path_length - (n_h)) + +#define get_last_bh(path) PATH_PLAST_BUFFER(path) +#define get_ih(path) PATH_PITEM_HEAD(path) +#define get_item_pos(path) PATH_LAST_POSITION(path) +#define get_item(path) ((void *)B_N_PITEM(PATH_PLAST_BUFFER(path), PATH_LAST_POSITION (path))) +#define item_moved(ih,path) comp_items(ih, path) +#define path_changed(ih,path) comp_items (ih, path) + +/***************************************************************************/ +/* MISC */ +/***************************************************************************/ + +/* Size of pointer to the unformatted node. */ +#define UNFM_P_SIZE (sizeof(unp_t)) +#define UNFM_P_SHIFT 2 + +// in in-core inode key is stored on le form +#define INODE_PKEY(inode) ((struct reiserfs_key *)(REISERFS_I(inode)->i_key)) + +#define MAX_UL_INT 0xffffffff +#define MAX_INT 0x7ffffff +#define MAX_US_INT 0xffff + +// reiserfs version 2 has max offset 60 bits. Version 1 - 32 bit offset +#define U32_MAX (~(__u32)0) + +static inline loff_t max_reiserfs_offset(struct inode *inode) +{ + if (get_inode_item_key_version(inode) == KEY_FORMAT_3_5) + return (loff_t) U32_MAX; + + return (loff_t) ((~(__u64) 0) >> 4); +} + +/*#define MAX_KEY_UNIQUENESS MAX_UL_INT*/ +#define MAX_KEY_OBJECTID MAX_UL_INT + +#define MAX_B_NUM MAX_UL_INT +#define MAX_FC_NUM MAX_US_INT + +/* the purpose is to detect overflow of an unsigned short */ +#define REISERFS_LINK_MAX (MAX_US_INT - 1000) + +/* The following defines are used in reiserfs_insert_item and reiserfs_append_item */ +#define REISERFS_KERNEL_MEM 0 /* reiserfs kernel memory mode */ +#define REISERFS_USER_MEM 1 /* reiserfs user memory mode */ + +#define fs_generation(s) (REISERFS_SB(s)->s_generation_counter) +#define get_generation(s) atomic_read (&fs_generation(s)) +#define FILESYSTEM_CHANGED_TB(tb) (get_generation((tb)->tb_sb) != (tb)->fs_gen) +#define __fs_changed(gen,s) (gen != get_generation (s)) +#define fs_changed(gen,s) \ +({ \ + reiserfs_cond_resched(s); \ + __fs_changed(gen, s); \ +}) + +/***************************************************************************/ +/* FIXATE NODES */ +/***************************************************************************/ + +#define VI_TYPE_LEFT_MERGEABLE 1 +#define VI_TYPE_RIGHT_MERGEABLE 2 + +/* To make any changes in the tree we always first find node, that + contains item to be changed/deleted or place to insert a new + item. We call this node S. To do balancing we need to decide what + we will shift to left/right neighbor, or to a new node, where new + item will be etc. To make this analysis simpler we build virtual + node. Virtual node is an array of items, that will replace items of + node S. (For instance if we are going to delete an item, virtual + node does not contain it). Virtual node keeps information about + item sizes and types, mergeability of first and last items, sizes + of all entries in directory item. We use this array of items when + calculating what we can shift to neighbors and how many nodes we + have to have if we do not any shiftings, if we shift to left/right + neighbor or to both. */ +struct virtual_item { + int vi_index; // index in the array of item operations + unsigned short vi_type; // left/right mergeability + unsigned short vi_item_len; /* length of item that it will have after balancing */ + struct item_head *vi_ih; + const char *vi_item; // body of item (old or new) + const void *vi_new_data; // 0 always but paste mode + void *vi_uarea; // item specific area +}; + +struct virtual_node { + char *vn_free_ptr; /* this is a pointer to the free space in the buffer */ + unsigned short vn_nr_item; /* number of items in virtual node */ + short vn_size; /* size of node , that node would have if it has unlimited size and no balancing is performed */ + short vn_mode; /* mode of balancing (paste, insert, delete, cut) */ + short vn_affected_item_num; + short vn_pos_in_item; + struct item_head *vn_ins_ih; /* item header of inserted item, 0 for other modes */ + const void *vn_data; + struct virtual_item *vn_vi; /* array of items (including a new one, excluding item to be deleted) */ +}; + +/* used by directory items when creating virtual nodes */ +struct direntry_uarea { + int flags; + __u16 entry_count; + __u16 entry_sizes[1]; +} __attribute__ ((__packed__)); + +/***************************************************************************/ +/* TREE BALANCE */ +/***************************************************************************/ + +/* This temporary structure is used in tree balance algorithms, and + constructed as we go to the extent that its various parts are + needed. It contains arrays of nodes that can potentially be + involved in the balancing of node S, and parameters that define how + each of the nodes must be balanced. Note that in these algorithms + for balancing the worst case is to need to balance the current node + S and the left and right neighbors and all of their parents plus + create a new node. We implement S1 balancing for the leaf nodes + and S0 balancing for the internal nodes (S1 and S0 are defined in + our papers.)*/ + +#define MAX_FREE_BLOCK 7 /* size of the array of buffers to free at end of do_balance */ + +/* maximum number of FEB blocknrs on a single level */ +#define MAX_AMOUNT_NEEDED 2 + +/* someday somebody will prefix every field in this struct with tb_ */ +struct tree_balance { + int tb_mode; + int need_balance_dirty; + struct super_block *tb_sb; + struct reiserfs_transaction_handle *transaction_handle; + struct treepath *tb_path; + struct buffer_head *L[MAX_HEIGHT]; /* array of left neighbors of nodes in the path */ + struct buffer_head *R[MAX_HEIGHT]; /* array of right neighbors of nodes in the path */ + struct buffer_head *FL[MAX_HEIGHT]; /* array of fathers of the left neighbors */ + struct buffer_head *FR[MAX_HEIGHT]; /* array of fathers of the right neighbors */ + struct buffer_head *CFL[MAX_HEIGHT]; /* array of common parents of center node and its left neighbor */ + struct buffer_head *CFR[MAX_HEIGHT]; /* array of common parents of center node and its right neighbor */ + + struct buffer_head *FEB[MAX_FEB_SIZE]; /* array of empty buffers. Number of buffers in array equals + cur_blknum. */ + struct buffer_head *used[MAX_FEB_SIZE]; + struct buffer_head *thrown[MAX_FEB_SIZE]; + int lnum[MAX_HEIGHT]; /* array of number of items which must be + shifted to the left in order to balance the + current node; for leaves includes item that + will be partially shifted; for internal + nodes, it is the number of child pointers + rather than items. It includes the new item + being created. The code sometimes subtracts + one to get the number of wholly shifted + items for other purposes. */ + int rnum[MAX_HEIGHT]; /* substitute right for left in comment above */ + int lkey[MAX_HEIGHT]; /* array indexed by height h mapping the key delimiting L[h] and + S[h] to its item number within the node CFL[h] */ + int rkey[MAX_HEIGHT]; /* substitute r for l in comment above */ + int insert_size[MAX_HEIGHT]; /* the number of bytes by we are trying to add or remove from + S[h]. A negative value means removing. */ + int blknum[MAX_HEIGHT]; /* number of nodes that will replace node S[h] after + balancing on the level h of the tree. If 0 then S is + being deleted, if 1 then S is remaining and no new nodes + are being created, if 2 or 3 then 1 or 2 new nodes is + being created */ + + /* fields that are used only for balancing leaves of the tree */ + int cur_blknum; /* number of empty blocks having been already allocated */ + int s0num; /* number of items that fall into left most node when S[0] splits */ + int s1num; /* number of items that fall into first new node when S[0] splits */ + int s2num; /* number of items that fall into second new node when S[0] splits */ + int lbytes; /* number of bytes which can flow to the left neighbor from the left */ + /* most liquid item that cannot be shifted from S[0] entirely */ + /* if -1 then nothing will be partially shifted */ + int rbytes; /* number of bytes which will flow to the right neighbor from the right */ + /* most liquid item that cannot be shifted from S[0] entirely */ + /* if -1 then nothing will be partially shifted */ + int s1bytes; /* number of bytes which flow to the first new node when S[0] splits */ + /* note: if S[0] splits into 3 nodes, then items do not need to be cut */ + int s2bytes; + struct buffer_head *buf_to_free[MAX_FREE_BLOCK]; /* buffers which are to be freed after do_balance finishes by unfix_nodes */ + char *vn_buf; /* kmalloced memory. Used to create + virtual node and keep map of + dirtied bitmap blocks */ + int vn_buf_size; /* size of the vn_buf */ + struct virtual_node *tb_vn; /* VN starts after bitmap of bitmap blocks */ + + int fs_gen; /* saved value of `reiserfs_generation' counter + see FILESYSTEM_CHANGED() macro in reiserfs_fs.h */ +#ifdef DISPLACE_NEW_PACKING_LOCALITIES + struct in_core_key key; /* key pointer, to pass to block allocator or + another low-level subsystem */ +#endif +}; + +/* These are modes of balancing */ + +/* When inserting an item. */ +#define M_INSERT 'i' +/* When inserting into (directories only) or appending onto an already + existent item. */ +#define M_PASTE 'p' +/* When deleting an item. */ +#define M_DELETE 'd' +/* When truncating an item or removing an entry from a (directory) item. */ +#define M_CUT 'c' + +/* used when balancing on leaf level skipped (in reiserfsck) */ +#define M_INTERNAL 'n' + +/* When further balancing is not needed, then do_balance does not need + to be called. */ +#define M_SKIP_BALANCING 's' +#define M_CONVERT 'v' + +/* modes of leaf_move_items */ +#define LEAF_FROM_S_TO_L 0 +#define LEAF_FROM_S_TO_R 1 +#define LEAF_FROM_R_TO_L 2 +#define LEAF_FROM_L_TO_R 3 +#define LEAF_FROM_S_TO_SNEW 4 + +#define FIRST_TO_LAST 0 +#define LAST_TO_FIRST 1 + +/* used in do_balance for passing parent of node information that has + been gotten from tb struct */ +struct buffer_info { + struct tree_balance *tb; + struct buffer_head *bi_bh; + struct buffer_head *bi_parent; + int bi_position; +}; + +static inline struct super_block *sb_from_tb(struct tree_balance *tb) +{ + return tb ? tb->tb_sb : NULL; +} + +static inline struct super_block *sb_from_bi(struct buffer_info *bi) +{ + return bi ? sb_from_tb(bi->tb) : NULL; +} + +/* there are 4 types of items: stat data, directory item, indirect, direct. ++-------------------+------------+--------------+------------+ +| | k_offset | k_uniqueness | mergeable? | ++-------------------+------------+--------------+------------+ +| stat data | 0 | 0 | no | ++-------------------+------------+--------------+------------+ +| 1st directory item| DOT_OFFSET |DIRENTRY_UNIQUENESS| no | +| non 1st directory | hash value | | yes | +| item | | | | ++-------------------+------------+--------------+------------+ +| indirect item | offset + 1 |TYPE_INDIRECT | if this is not the first indirect item of the object ++-------------------+------------+--------------+------------+ +| direct item | offset + 1 |TYPE_DIRECT | if not this is not the first direct item of the object ++-------------------+------------+--------------+------------+ +*/ + +struct item_operations { + int (*bytes_number) (struct item_head * ih, int block_size); + void (*decrement_key) (struct cpu_key *); + int (*is_left_mergeable) (struct reiserfs_key * ih, + unsigned long bsize); + void (*print_item) (struct item_head *, char *item); + void (*check_item) (struct item_head *, char *item); + + int (*create_vi) (struct virtual_node * vn, struct virtual_item * vi, + int is_affected, int insert_size); + int (*check_left) (struct virtual_item * vi, int free, + int start_skip, int end_skip); + int (*check_right) (struct virtual_item * vi, int free); + int (*part_size) (struct virtual_item * vi, int from, int to); + int (*unit_num) (struct virtual_item * vi); + void (*print_vi) (struct virtual_item * vi); +}; + +extern struct item_operations *item_ops[TYPE_ANY + 1]; + +#define op_bytes_number(ih,bsize) item_ops[le_ih_k_type (ih)]->bytes_number (ih, bsize) +#define op_is_left_mergeable(key,bsize) item_ops[le_key_k_type (le_key_version (key), key)]->is_left_mergeable (key, bsize) +#define op_print_item(ih,item) item_ops[le_ih_k_type (ih)]->print_item (ih, item) +#define op_check_item(ih,item) item_ops[le_ih_k_type (ih)]->check_item (ih, item) +#define op_create_vi(vn,vi,is_affected,insert_size) item_ops[le_ih_k_type ((vi)->vi_ih)]->create_vi (vn,vi,is_affected,insert_size) +#define op_check_left(vi,free,start_skip,end_skip) item_ops[(vi)->vi_index]->check_left (vi, free, start_skip, end_skip) +#define op_check_right(vi,free) item_ops[(vi)->vi_index]->check_right (vi, free) +#define op_part_size(vi,from,to) item_ops[(vi)->vi_index]->part_size (vi, from, to) +#define op_unit_num(vi) item_ops[(vi)->vi_index]->unit_num (vi) +#define op_print_vi(vi) item_ops[(vi)->vi_index]->print_vi (vi) + +#define COMP_SHORT_KEYS comp_short_keys + +/* number of blocks pointed to by the indirect item */ +#define I_UNFM_NUM(ih) (ih_item_len(ih) / UNFM_P_SIZE) + +/* the used space within the unformatted node corresponding to pos within the item pointed to by ih */ +#define I_POS_UNFM_SIZE(ih,pos,size) (((pos) == I_UNFM_NUM(ih) - 1 ) ? (size) - ih_free_space(ih) : (size)) + +/* number of bytes contained by the direct item or the unformatted nodes the indirect item points to */ + +/* get the item header */ +#define B_N_PITEM_HEAD(bh,item_num) ( (struct item_head * )((bh)->b_data + BLKH_SIZE) + (item_num) ) + +/* get key */ +#define B_N_PDELIM_KEY(bh,item_num) ( (struct reiserfs_key * )((bh)->b_data + BLKH_SIZE) + (item_num) ) + +/* get the key */ +#define B_N_PKEY(bh,item_num) ( &(B_N_PITEM_HEAD(bh,item_num)->ih_key) ) + +/* get item body */ +#define B_N_PITEM(bh,item_num) ( (bh)->b_data + ih_location(B_N_PITEM_HEAD((bh),(item_num)))) + +/* get the stat data by the buffer header and the item order */ +#define B_N_STAT_DATA(bh,nr) \ +( (struct stat_data *)((bh)->b_data + ih_location(B_N_PITEM_HEAD((bh),(nr))) ) ) + + /* following defines use reiserfs buffer header and item header */ + +/* get stat-data */ +#define B_I_STAT_DATA(bh, ih) ( (struct stat_data * )((bh)->b_data + ih_location(ih)) ) + +// this is 3976 for size==4096 +#define MAX_DIRECT_ITEM_LEN(size) ((size) - BLKH_SIZE - 2*IH_SIZE - SD_SIZE - UNFM_P_SIZE) + +/* indirect items consist of entries which contain blocknrs, pos + indicates which entry, and B_I_POS_UNFM_POINTER resolves to the + blocknr contained by the entry pos points to */ +#define B_I_POS_UNFM_POINTER(bh,ih,pos) le32_to_cpu(*(((unp_t *)B_I_PITEM(bh,ih)) + (pos))) +#define PUT_B_I_POS_UNFM_POINTER(bh,ih,pos, val) do {*(((unp_t *)B_I_PITEM(bh,ih)) + (pos)) = cpu_to_le32(val); } while (0) + +struct reiserfs_iget_args { + __u32 objectid; + __u32 dirid; +}; + +/***************************************************************************/ +/* FUNCTION DECLARATIONS */ +/***************************************************************************/ + +#define get_journal_desc_magic(bh) (bh->b_data + bh->b_size - 12) + +#define journal_trans_half(blocksize) \ + ((blocksize - sizeof (struct reiserfs_journal_desc) + sizeof (__u32) - 12) / sizeof (__u32)) + +/* journal.c see journal.c for all the comments here */ + +/* first block written in a commit. */ +struct reiserfs_journal_desc { + __le32 j_trans_id; /* id of commit */ + __le32 j_len; /* length of commit. len +1 is the commit block */ + __le32 j_mount_id; /* mount id of this trans */ + __le32 j_realblock[1]; /* real locations for each block */ +}; + +#define get_desc_trans_id(d) le32_to_cpu((d)->j_trans_id) +#define get_desc_trans_len(d) le32_to_cpu((d)->j_len) +#define get_desc_mount_id(d) le32_to_cpu((d)->j_mount_id) + +#define set_desc_trans_id(d,val) do { (d)->j_trans_id = cpu_to_le32 (val); } while (0) +#define set_desc_trans_len(d,val) do { (d)->j_len = cpu_to_le32 (val); } while (0) +#define set_desc_mount_id(d,val) do { (d)->j_mount_id = cpu_to_le32 (val); } while (0) + +/* last block written in a commit */ +struct reiserfs_journal_commit { + __le32 j_trans_id; /* must match j_trans_id from the desc block */ + __le32 j_len; /* ditto */ + __le32 j_realblock[1]; /* real locations for each block */ +}; + +#define get_commit_trans_id(c) le32_to_cpu((c)->j_trans_id) +#define get_commit_trans_len(c) le32_to_cpu((c)->j_len) +#define get_commit_mount_id(c) le32_to_cpu((c)->j_mount_id) + +#define set_commit_trans_id(c,val) do { (c)->j_trans_id = cpu_to_le32 (val); } while (0) +#define set_commit_trans_len(c,val) do { (c)->j_len = cpu_to_le32 (val); } while (0) + +/* this header block gets written whenever a transaction is considered fully flushed, and is more recent than the +** last fully flushed transaction. fully flushed means all the log blocks and all the real blocks are on disk, +** and this transaction does not need to be replayed. +*/ +struct reiserfs_journal_header { + __le32 j_last_flush_trans_id; /* id of last fully flushed transaction */ + __le32 j_first_unflushed_offset; /* offset in the log of where to start replay after a crash */ + __le32 j_mount_id; + /* 12 */ struct journal_params jh_journal; +}; + +/* biggest tunable defines are right here */ +#define JOURNAL_BLOCK_COUNT 8192 /* number of blocks in the journal */ +#define JOURNAL_TRANS_MAX_DEFAULT 1024 /* biggest possible single transaction, don't change for now (8/3/99) */ +#define JOURNAL_TRANS_MIN_DEFAULT 256 +#define JOURNAL_MAX_BATCH_DEFAULT 900 /* max blocks to batch into one transaction, don't make this any bigger than 900 */ +#define JOURNAL_MIN_RATIO 2 +#define JOURNAL_MAX_COMMIT_AGE 30 +#define JOURNAL_MAX_TRANS_AGE 30 +#define JOURNAL_PER_BALANCE_CNT (3 * (MAX_HEIGHT-2) + 9) +#define JOURNAL_BLOCKS_PER_OBJECT(sb) (JOURNAL_PER_BALANCE_CNT * 3 + \ + 2 * (REISERFS_QUOTA_INIT_BLOCKS(sb) + \ + REISERFS_QUOTA_TRANS_BLOCKS(sb))) + +#ifdef CONFIG_QUOTA +#define REISERFS_QUOTA_OPTS ((1 << REISERFS_USRQUOTA) | (1 << REISERFS_GRPQUOTA)) +/* We need to update data and inode (atime) */ +#define REISERFS_QUOTA_TRANS_BLOCKS(s) (REISERFS_SB(s)->s_mount_opt & REISERFS_QUOTA_OPTS ? 2 : 0) +/* 1 balancing, 1 bitmap, 1 data per write + stat data update */ +#define REISERFS_QUOTA_INIT_BLOCKS(s) (REISERFS_SB(s)->s_mount_opt & REISERFS_QUOTA_OPTS ? \ +(DQUOT_INIT_ALLOC*(JOURNAL_PER_BALANCE_CNT+2)+DQUOT_INIT_REWRITE+1) : 0) +/* same as with INIT */ +#define REISERFS_QUOTA_DEL_BLOCKS(s) (REISERFS_SB(s)->s_mount_opt & REISERFS_QUOTA_OPTS ? \ +(DQUOT_DEL_ALLOC*(JOURNAL_PER_BALANCE_CNT+2)+DQUOT_DEL_REWRITE+1) : 0) +#else +#define REISERFS_QUOTA_TRANS_BLOCKS(s) 0 +#define REISERFS_QUOTA_INIT_BLOCKS(s) 0 +#define REISERFS_QUOTA_DEL_BLOCKS(s) 0 +#endif + +/* both of these can be as low as 1, or as high as you want. The min is the +** number of 4k bitmap nodes preallocated on mount. New nodes are allocated +** as needed, and released when transactions are committed. On release, if +** the current number of nodes is > max, the node is freed, otherwise, +** it is put on a free list for faster use later. +*/ +#define REISERFS_MIN_BITMAP_NODES 10 +#define REISERFS_MAX_BITMAP_NODES 100 + +#define JBH_HASH_SHIFT 13 /* these are based on journal hash size of 8192 */ +#define JBH_HASH_MASK 8191 + +#define _jhashfn(sb,block) \ + (((unsigned long)sb>>L1_CACHE_SHIFT) ^ \ + (((block)<<(JBH_HASH_SHIFT - 6)) ^ ((block) >> 13) ^ ((block) << (JBH_HASH_SHIFT - 12)))) +#define journal_hash(t,sb,block) ((t)[_jhashfn((sb),(block)) & JBH_HASH_MASK]) + +// We need these to make journal.c code more readable +#define journal_find_get_block(s, block) __find_get_block(SB_JOURNAL(s)->j_dev_bd, block, s->s_blocksize) +#define journal_getblk(s, block) __getblk(SB_JOURNAL(s)->j_dev_bd, block, s->s_blocksize) +#define journal_bread(s, block) __bread(SB_JOURNAL(s)->j_dev_bd, block, s->s_blocksize) + +enum reiserfs_bh_state_bits { + BH_JDirty = BH_PrivateStart, /* buffer is in current transaction */ + BH_JDirty_wait, + BH_JNew, /* disk block was taken off free list before + * being in a finished transaction, or + * written to disk. Can be reused immed. */ + BH_JPrepared, + BH_JRestore_dirty, + BH_JTest, // debugging only will go away +}; + +BUFFER_FNS(JDirty, journaled); +TAS_BUFFER_FNS(JDirty, journaled); +BUFFER_FNS(JDirty_wait, journal_dirty); +TAS_BUFFER_FNS(JDirty_wait, journal_dirty); +BUFFER_FNS(JNew, journal_new); +TAS_BUFFER_FNS(JNew, journal_new); +BUFFER_FNS(JPrepared, journal_prepared); +TAS_BUFFER_FNS(JPrepared, journal_prepared); +BUFFER_FNS(JRestore_dirty, journal_restore_dirty); +TAS_BUFFER_FNS(JRestore_dirty, journal_restore_dirty); +BUFFER_FNS(JTest, journal_test); +TAS_BUFFER_FNS(JTest, journal_test); + +/* +** transaction handle which is passed around for all journal calls +*/ +struct reiserfs_transaction_handle { + struct super_block *t_super; /* super for this FS when journal_begin was + called. saves calls to reiserfs_get_super + also used by nested transactions to make + sure they are nesting on the right FS + _must_ be first in the handle + */ + int t_refcount; + int t_blocks_logged; /* number of blocks this writer has logged */ + int t_blocks_allocated; /* number of blocks this writer allocated */ + unsigned int t_trans_id; /* sanity check, equals the current trans id */ + void *t_handle_save; /* save existing current->journal_info */ + unsigned displace_new_blocks:1; /* if new block allocation occurres, that block + should be displaced from others */ + struct list_head t_list; +}; + +/* used to keep track of ordered and tail writes, attached to the buffer + * head through b_journal_head. + */ +struct reiserfs_jh { + struct reiserfs_journal_list *jl; + struct buffer_head *bh; + struct list_head list; +}; + +void reiserfs_free_jh(struct buffer_head *bh); +int reiserfs_add_tail_list(struct inode *inode, struct buffer_head *bh); +int reiserfs_add_ordered_list(struct inode *inode, struct buffer_head *bh); +int journal_mark_dirty(struct reiserfs_transaction_handle *, + struct super_block *, struct buffer_head *bh); + +static inline int reiserfs_file_data_log(struct inode *inode) +{ + if (reiserfs_data_log(inode->i_sb) || + (REISERFS_I(inode)->i_flags & i_data_log)) + return 1; + return 0; +} + +static inline int reiserfs_transaction_running(struct super_block *s) +{ + struct reiserfs_transaction_handle *th = current->journal_info; + if (th && th->t_super == s) + return 1; + if (th && th->t_super == NULL) + BUG(); + return 0; +} + +static inline int reiserfs_transaction_free_space(struct reiserfs_transaction_handle *th) +{ + return th->t_blocks_allocated - th->t_blocks_logged; +} + +struct reiserfs_transaction_handle *reiserfs_persistent_transaction(struct + super_block + *, + int count); +int reiserfs_end_persistent_transaction(struct reiserfs_transaction_handle *); +int reiserfs_commit_page(struct inode *inode, struct page *page, + unsigned from, unsigned to); +int reiserfs_flush_old_commits(struct super_block *); +int reiserfs_commit_for_inode(struct inode *); +int reiserfs_inode_needs_commit(struct inode *); +void reiserfs_update_inode_transaction(struct inode *); +void reiserfs_wait_on_write_block(struct super_block *s); +void reiserfs_block_writes(struct reiserfs_transaction_handle *th); +void reiserfs_allow_writes(struct super_block *s); +void reiserfs_check_lock_depth(struct super_block *s, char *caller); +int reiserfs_prepare_for_journal(struct super_block *, struct buffer_head *bh, + int wait); +void reiserfs_restore_prepared_buffer(struct super_block *, + struct buffer_head *bh); +int journal_init(struct super_block *, const char *j_dev_name, int old_format, + unsigned int); +int journal_release(struct reiserfs_transaction_handle *, struct super_block *); +int journal_release_error(struct reiserfs_transaction_handle *, + struct super_block *); +int journal_end(struct reiserfs_transaction_handle *, struct super_block *, + unsigned long); +int journal_end_sync(struct reiserfs_transaction_handle *, struct super_block *, + unsigned long); +int journal_mark_freed(struct reiserfs_transaction_handle *, + struct super_block *, b_blocknr_t blocknr); +int journal_transaction_should_end(struct reiserfs_transaction_handle *, int); +int reiserfs_in_journal(struct super_block *sb, unsigned int bmap_nr, + int bit_nr, int searchall, b_blocknr_t *next); +int journal_begin(struct reiserfs_transaction_handle *, + struct super_block *sb, unsigned long); +int journal_join_abort(struct reiserfs_transaction_handle *, + struct super_block *sb, unsigned long); +void reiserfs_abort_journal(struct super_block *sb, int errno); +void reiserfs_abort(struct super_block *sb, int errno, const char *fmt, ...); +int reiserfs_allocate_list_bitmaps(struct super_block *s, + struct reiserfs_list_bitmap *, unsigned int); + +void add_save_link(struct reiserfs_transaction_handle *th, + struct inode *inode, int truncate); +int remove_save_link(struct inode *inode, int truncate); + +/* objectid.c */ +__u32 reiserfs_get_unused_objectid(struct reiserfs_transaction_handle *th); +void reiserfs_release_objectid(struct reiserfs_transaction_handle *th, + __u32 objectid_to_release); +int reiserfs_convert_objectid_map_v1(struct super_block *); + +/* stree.c */ +int B_IS_IN_TREE(const struct buffer_head *); +extern void copy_item_head(struct item_head *to, + const struct item_head *from); + +// first key is in cpu form, second - le +extern int comp_short_keys(const struct reiserfs_key *le_key, + const struct cpu_key *cpu_key); +extern void le_key2cpu_key(struct cpu_key *to, const struct reiserfs_key *from); + +// both are in le form +extern int comp_le_keys(const struct reiserfs_key *, + const struct reiserfs_key *); +extern int comp_short_le_keys(const struct reiserfs_key *, + const struct reiserfs_key *); + +// +// get key version from on disk key - kludge +// +static inline int le_key_version(const struct reiserfs_key *key) +{ + int type; + + type = offset_v2_k_type(&(key->u.k_offset_v2)); + if (type != TYPE_DIRECT && type != TYPE_INDIRECT + && type != TYPE_DIRENTRY) + return KEY_FORMAT_3_5; + + return KEY_FORMAT_3_6; + +} + +static inline void copy_key(struct reiserfs_key *to, + const struct reiserfs_key *from) +{ + memcpy(to, from, KEY_SIZE); +} + +int comp_items(const struct item_head *stored_ih, const struct treepath *path); +const struct reiserfs_key *get_rkey(const struct treepath *chk_path, + const struct super_block *sb); +int search_by_key(struct super_block *, const struct cpu_key *, + struct treepath *, int); +#define search_item(s,key,path) search_by_key (s, key, path, DISK_LEAF_NODE_LEVEL) +int search_for_position_by_key(struct super_block *sb, + const struct cpu_key *cpu_key, + struct treepath *search_path); +extern void decrement_bcount(struct buffer_head *bh); +void decrement_counters_in_path(struct treepath *search_path); +void pathrelse(struct treepath *search_path); +int reiserfs_check_path(struct treepath *p); +void pathrelse_and_restore(struct super_block *s, struct treepath *search_path); + +int reiserfs_insert_item(struct reiserfs_transaction_handle *th, + struct treepath *path, + const struct cpu_key *key, + struct item_head *ih, + struct inode *inode, const char *body); + +int reiserfs_paste_into_item(struct reiserfs_transaction_handle *th, + struct treepath *path, + const struct cpu_key *key, + struct inode *inode, + const char *body, int paste_size); + +int reiserfs_cut_from_item(struct reiserfs_transaction_handle *th, + struct treepath *path, + struct cpu_key *key, + struct inode *inode, + struct page *page, loff_t new_file_size); + +int reiserfs_delete_item(struct reiserfs_transaction_handle *th, + struct treepath *path, + const struct cpu_key *key, + struct inode *inode, struct buffer_head *un_bh); + +void reiserfs_delete_solid_item(struct reiserfs_transaction_handle *th, + struct inode *inode, struct reiserfs_key *key); +int reiserfs_delete_object(struct reiserfs_transaction_handle *th, + struct inode *inode); +int reiserfs_do_truncate(struct reiserfs_transaction_handle *th, + struct inode *inode, struct page *, + int update_timestamps); + +#define i_block_size(inode) ((inode)->i_sb->s_blocksize) +#define file_size(inode) ((inode)->i_size) +#define tail_size(inode) (file_size (inode) & (i_block_size (inode) - 1)) + +#define tail_has_to_be_packed(inode) (have_large_tails ((inode)->i_sb)?\ +!STORE_TAIL_IN_UNFM_S1(file_size (inode), tail_size(inode), inode->i_sb->s_blocksize):have_small_tails ((inode)->i_sb)?!STORE_TAIL_IN_UNFM_S2(file_size (inode), tail_size(inode), inode->i_sb->s_blocksize):0 ) + +void padd_item(char *item, int total_length, int length); + +/* inode.c */ +/* args for the create parameter of reiserfs_get_block */ +#define GET_BLOCK_NO_CREATE 0 /* don't create new blocks or convert tails */ +#define GET_BLOCK_CREATE 1 /* add anything you need to find block */ +#define GET_BLOCK_NO_HOLE 2 /* return -ENOENT for file holes */ +#define GET_BLOCK_READ_DIRECT 4 /* read the tail if indirect item not found */ +#define GET_BLOCK_NO_IMUX 8 /* i_mutex is not held, don't preallocate */ +#define GET_BLOCK_NO_DANGLE 16 /* don't leave any transactions running */ + +void reiserfs_read_locked_inode(struct inode *inode, + struct reiserfs_iget_args *args); +int reiserfs_find_actor(struct inode *inode, void *p); +int reiserfs_init_locked_inode(struct inode *inode, void *p); +void reiserfs_evict_inode(struct inode *inode); +int reiserfs_write_inode(struct inode *inode, struct writeback_control *wbc); +int reiserfs_get_block(struct inode *inode, sector_t block, + struct buffer_head *bh_result, int create); +struct dentry *reiserfs_fh_to_dentry(struct super_block *sb, struct fid *fid, + int fh_len, int fh_type); +struct dentry *reiserfs_fh_to_parent(struct super_block *sb, struct fid *fid, + int fh_len, int fh_type); +int reiserfs_encode_fh(struct dentry *dentry, __u32 * data, int *lenp, + int connectable); + +int reiserfs_truncate_file(struct inode *, int update_timestamps); +void make_cpu_key(struct cpu_key *cpu_key, struct inode *inode, loff_t offset, + int type, int key_length); +void make_le_item_head(struct item_head *ih, const struct cpu_key *key, + int version, + loff_t offset, int type, int length, int entry_count); +struct inode *reiserfs_iget(struct super_block *s, const struct cpu_key *key); + +struct reiserfs_security_handle; +int reiserfs_new_inode(struct reiserfs_transaction_handle *th, + struct inode *dir, umode_t mode, + const char *symname, loff_t i_size, + struct dentry *dentry, struct inode *inode, + struct reiserfs_security_handle *security); + +void reiserfs_update_sd_size(struct reiserfs_transaction_handle *th, + struct inode *inode, loff_t size); + +static inline void reiserfs_update_sd(struct reiserfs_transaction_handle *th, + struct inode *inode) +{ + reiserfs_update_sd_size(th, inode, inode->i_size); +} + +void sd_attrs_to_i_attrs(__u16 sd_attrs, struct inode *inode); +void i_attrs_to_sd_attrs(struct inode *inode, __u16 * sd_attrs); +int reiserfs_setattr(struct dentry *dentry, struct iattr *attr); + +int __reiserfs_write_begin(struct page *page, unsigned from, unsigned len); + +/* namei.c */ +void set_de_name_and_namelen(struct reiserfs_dir_entry *de); +int search_by_entry_key(struct super_block *sb, const struct cpu_key *key, + struct treepath *path, struct reiserfs_dir_entry *de); +struct dentry *reiserfs_get_parent(struct dentry *); + +#ifdef CONFIG_REISERFS_PROC_INFO +int reiserfs_proc_info_init(struct super_block *sb); +int reiserfs_proc_info_done(struct super_block *sb); +int reiserfs_proc_info_global_init(void); +int reiserfs_proc_info_global_done(void); + +#define PROC_EXP( e ) e + +#define __PINFO( sb ) REISERFS_SB(sb) -> s_proc_info_data +#define PROC_INFO_MAX( sb, field, value ) \ + __PINFO( sb ).field = \ + max( REISERFS_SB( sb ) -> s_proc_info_data.field, value ) +#define PROC_INFO_INC( sb, field ) ( ++ ( __PINFO( sb ).field ) ) +#define PROC_INFO_ADD( sb, field, val ) ( __PINFO( sb ).field += ( val ) ) +#define PROC_INFO_BH_STAT( sb, bh, level ) \ + PROC_INFO_INC( sb, sbk_read_at[ ( level ) ] ); \ + PROC_INFO_ADD( sb, free_at[ ( level ) ], B_FREE_SPACE( bh ) ); \ + PROC_INFO_ADD( sb, items_at[ ( level ) ], B_NR_ITEMS( bh ) ) +#else +static inline int reiserfs_proc_info_init(struct super_block *sb) +{ + return 0; +} + +static inline int reiserfs_proc_info_done(struct super_block *sb) +{ + return 0; +} + +static inline int reiserfs_proc_info_global_init(void) +{ + return 0; +} + +static inline int reiserfs_proc_info_global_done(void) +{ + return 0; +} + +#define PROC_EXP( e ) +#define VOID_V ( ( void ) 0 ) +#define PROC_INFO_MAX( sb, field, value ) VOID_V +#define PROC_INFO_INC( sb, field ) VOID_V +#define PROC_INFO_ADD( sb, field, val ) VOID_V +#define PROC_INFO_BH_STAT(sb, bh, n_node_level) VOID_V +#endif + +/* dir.c */ +extern const struct inode_operations reiserfs_dir_inode_operations; +extern const struct inode_operations reiserfs_symlink_inode_operations; +extern const struct inode_operations reiserfs_special_inode_operations; +extern const struct file_operations reiserfs_dir_operations; +int reiserfs_readdir_dentry(struct dentry *, void *, filldir_t, loff_t *); + +/* tail_conversion.c */ +int direct2indirect(struct reiserfs_transaction_handle *, struct inode *, + struct treepath *, struct buffer_head *, loff_t); +int indirect2direct(struct reiserfs_transaction_handle *, struct inode *, + struct page *, struct treepath *, const struct cpu_key *, + loff_t, char *); +void reiserfs_unmap_buffer(struct buffer_head *); + +/* file.c */ +extern const struct inode_operations reiserfs_file_inode_operations; +extern const struct file_operations reiserfs_file_operations; +extern const struct address_space_operations reiserfs_address_space_operations; + +/* fix_nodes.c */ + +int fix_nodes(int n_op_mode, struct tree_balance *tb, + struct item_head *ins_ih, const void *); +void unfix_nodes(struct tree_balance *); + +/* prints.c */ +void __reiserfs_panic(struct super_block *s, const char *id, + const char *function, const char *fmt, ...) + __attribute__ ((noreturn)); +#define reiserfs_panic(s, id, fmt, args...) \ + __reiserfs_panic(s, id, __func__, fmt, ##args) +void __reiserfs_error(struct super_block *s, const char *id, + const char *function, const char *fmt, ...); +#define reiserfs_error(s, id, fmt, args...) \ + __reiserfs_error(s, id, __func__, fmt, ##args) +void reiserfs_info(struct super_block *s, const char *fmt, ...); +void reiserfs_debug(struct super_block *s, int level, const char *fmt, ...); +void print_indirect_item(struct buffer_head *bh, int item_num); +void store_print_tb(struct tree_balance *tb); +void print_cur_tb(char *mes); +void print_de(struct reiserfs_dir_entry *de); +void print_bi(struct buffer_info *bi, char *mes); +#define PRINT_LEAF_ITEMS 1 /* print all items */ +#define PRINT_DIRECTORY_ITEMS 2 /* print directory items */ +#define PRINT_DIRECT_ITEMS 4 /* print contents of direct items */ +void print_block(struct buffer_head *bh, ...); +void print_bmap(struct super_block *s, int silent); +void print_bmap_block(int i, char *data, int size, int silent); +/*void print_super_block (struct super_block * s, char * mes);*/ +void print_objectid_map(struct super_block *s); +void print_block_head(struct buffer_head *bh, char *mes); +void check_leaf(struct buffer_head *bh); +void check_internal(struct buffer_head *bh); +void print_statistics(struct super_block *s); +char *reiserfs_hashname(int code); + +/* lbalance.c */ +int leaf_move_items(int shift_mode, struct tree_balance *tb, int mov_num, + int mov_bytes, struct buffer_head *Snew); +int leaf_shift_left(struct tree_balance *tb, int shift_num, int shift_bytes); +int leaf_shift_right(struct tree_balance *tb, int shift_num, int shift_bytes); +void leaf_delete_items(struct buffer_info *cur_bi, int last_first, int first, + int del_num, int del_bytes); +void leaf_insert_into_buf(struct buffer_info *bi, int before, + struct item_head *inserted_item_ih, + const char *inserted_item_body, int zeros_number); +void leaf_paste_in_buffer(struct buffer_info *bi, int pasted_item_num, + int pos_in_item, int paste_size, const char *body, + int zeros_number); +void leaf_cut_from_buffer(struct buffer_info *bi, int cut_item_num, + int pos_in_item, int cut_size); +void leaf_paste_entries(struct buffer_info *bi, int item_num, int before, + int new_entry_count, struct reiserfs_de_head *new_dehs, + const char *records, int paste_size); +/* ibalance.c */ +int balance_internal(struct tree_balance *, int, int, struct item_head *, + struct buffer_head **); + +/* do_balance.c */ +void do_balance_mark_leaf_dirty(struct tree_balance *tb, + struct buffer_head *bh, int flag); +#define do_balance_mark_internal_dirty do_balance_mark_leaf_dirty +#define do_balance_mark_sb_dirty do_balance_mark_leaf_dirty + +void do_balance(struct tree_balance *tb, struct item_head *ih, + const char *body, int flag); +void reiserfs_invalidate_buffer(struct tree_balance *tb, + struct buffer_head *bh); + +int get_left_neighbor_position(struct tree_balance *tb, int h); +int get_right_neighbor_position(struct tree_balance *tb, int h); +void replace_key(struct tree_balance *tb, struct buffer_head *, int, + struct buffer_head *, int); +void make_empty_node(struct buffer_info *); +struct buffer_head *get_FEB(struct tree_balance *); + +/* bitmap.c */ + +/* structure contains hints for block allocator, and it is a container for + * arguments, such as node, search path, transaction_handle, etc. */ +struct __reiserfs_blocknr_hint { + struct inode *inode; /* inode passed to allocator, if we allocate unf. nodes */ + sector_t block; /* file offset, in blocks */ + struct in_core_key key; + struct treepath *path; /* search path, used by allocator to deternine search_start by + * various ways */ + struct reiserfs_transaction_handle *th; /* transaction handle is needed to log super blocks and + * bitmap blocks changes */ + b_blocknr_t beg, end; + b_blocknr_t search_start; /* a field used to transfer search start value (block number) + * between different block allocator procedures + * (determine_search_start() and others) */ + int prealloc_size; /* is set in determine_prealloc_size() function, used by underlayed + * function that do actual allocation */ + + unsigned formatted_node:1; /* the allocator uses different polices for getting disk space for + * formatted/unformatted blocks with/without preallocation */ + unsigned preallocate:1; +}; + +typedef struct __reiserfs_blocknr_hint reiserfs_blocknr_hint_t; + +int reiserfs_parse_alloc_options(struct super_block *, char *); +void reiserfs_init_alloc_options(struct super_block *s); + +/* + * given a directory, this will tell you what packing locality + * to use for a new object underneat it. The locality is returned + * in disk byte order (le). + */ +__le32 reiserfs_choose_packing(struct inode *dir); + +int reiserfs_init_bitmap_cache(struct super_block *sb); +void reiserfs_free_bitmap_cache(struct super_block *sb); +void reiserfs_cache_bitmap_metadata(struct super_block *sb, struct buffer_head *bh, struct reiserfs_bitmap_info *info); +struct buffer_head *reiserfs_read_bitmap_block(struct super_block *sb, unsigned int bitmap); +int is_reusable(struct super_block *s, b_blocknr_t block, int bit_value); +void reiserfs_free_block(struct reiserfs_transaction_handle *th, struct inode *, + b_blocknr_t, int for_unformatted); +int reiserfs_allocate_blocknrs(reiserfs_blocknr_hint_t *, b_blocknr_t *, int, + int); +static inline int reiserfs_new_form_blocknrs(struct tree_balance *tb, + b_blocknr_t * new_blocknrs, + int amount_needed) +{ + reiserfs_blocknr_hint_t hint = { + .th = tb->transaction_handle, + .path = tb->tb_path, + .inode = NULL, + .key = tb->key, + .block = 0, + .formatted_node = 1 + }; + return reiserfs_allocate_blocknrs(&hint, new_blocknrs, amount_needed, + 0); +} + +static inline int reiserfs_new_unf_blocknrs(struct reiserfs_transaction_handle + *th, struct inode *inode, + b_blocknr_t * new_blocknrs, + struct treepath *path, + sector_t block) +{ + reiserfs_blocknr_hint_t hint = { + .th = th, + .path = path, + .inode = inode, + .block = block, + .formatted_node = 0, + .preallocate = 0 + }; + return reiserfs_allocate_blocknrs(&hint, new_blocknrs, 1, 0); +} + +#ifdef REISERFS_PREALLOCATE +static inline int reiserfs_new_unf_blocknrs2(struct reiserfs_transaction_handle + *th, struct inode *inode, + b_blocknr_t * new_blocknrs, + struct treepath *path, + sector_t block) +{ + reiserfs_blocknr_hint_t hint = { + .th = th, + .path = path, + .inode = inode, + .block = block, + .formatted_node = 0, + .preallocate = 1 + }; + return reiserfs_allocate_blocknrs(&hint, new_blocknrs, 1, 0); +} + +void reiserfs_discard_prealloc(struct reiserfs_transaction_handle *th, + struct inode *inode); +void reiserfs_discard_all_prealloc(struct reiserfs_transaction_handle *th); +#endif + +/* hashes.c */ +__u32 keyed_hash(const signed char *msg, int len); +__u32 yura_hash(const signed char *msg, int len); +__u32 r5_hash(const signed char *msg, int len); + +#define reiserfs_set_le_bit __set_bit_le +#define reiserfs_test_and_set_le_bit __test_and_set_bit_le +#define reiserfs_clear_le_bit __clear_bit_le +#define reiserfs_test_and_clear_le_bit __test_and_clear_bit_le +#define reiserfs_test_le_bit test_bit_le +#define reiserfs_find_next_zero_le_bit find_next_zero_bit_le + +/* sometimes reiserfs_truncate may require to allocate few new blocks + to perform indirect2direct conversion. People probably used to + think, that truncate should work without problems on a filesystem + without free disk space. They may complain that they can not + truncate due to lack of free disk space. This spare space allows us + to not worry about it. 500 is probably too much, but it should be + absolutely safe */ +#define SPARE_SPACE 500 + +/* prototypes from ioctl.c */ +long reiserfs_ioctl(struct file *filp, unsigned int cmd, unsigned long arg); +long reiserfs_compat_ioctl(struct file *filp, + unsigned int cmd, unsigned long arg); +int reiserfs_unpack(struct inode *inode, struct file *filp); diff --git a/fs/reiserfs/resize.c b/fs/reiserfs/resize.c index e1415ad93251..9a17f63c3fd7 100644 --- a/fs/reiserfs/resize.c +++ b/fs/reiserfs/resize.c @@ -13,7 +13,7 @@ #include #include #include -#include +#include "reiserfs.h" #include int reiserfs_resize(struct super_block *s, unsigned long block_count_new) diff --git a/fs/reiserfs/stree.c b/fs/reiserfs/stree.c index 313d39d639eb..f1b68afc268f 100644 --- a/fs/reiserfs/stree.c +++ b/fs/reiserfs/stree.c @@ -51,7 +51,7 @@ #include #include #include -#include +#include "reiserfs.h" #include #include diff --git a/fs/reiserfs/super.c b/fs/reiserfs/super.c index cf68a6ba0ec6..8b7616ef06d8 100644 --- a/fs/reiserfs/super.c +++ b/fs/reiserfs/super.c @@ -16,7 +16,7 @@ #include #include #include -#include +#include "reiserfs.h" #include "acl.h" #include "xattr.h" #include diff --git a/fs/reiserfs/tail_conversion.c b/fs/reiserfs/tail_conversion.c index d7f6e51bef2a..32f9a80d292f 100644 --- a/fs/reiserfs/tail_conversion.c +++ b/fs/reiserfs/tail_conversion.c @@ -5,7 +5,7 @@ #include #include #include -#include +#include "reiserfs.h" /* access to tail : when one is going to read tail it must make sure, that is not running. direct2indirect and indirect2direct can not run concurrently */ diff --git a/fs/reiserfs/xattr.c b/fs/reiserfs/xattr.c index 61c9b5633e27..46fc1c20a6b1 100644 --- a/fs/reiserfs/xattr.c +++ b/fs/reiserfs/xattr.c @@ -33,7 +33,7 @@ * The xattrs themselves are protected by the xattr_sem. */ -#include +#include "reiserfs.h" #include #include #include diff --git a/fs/reiserfs/xattr.h b/fs/reiserfs/xattr.h index ccd146bb0665..f59626c5d33b 100644 --- a/fs/reiserfs/xattr.h +++ b/fs/reiserfs/xattr.h @@ -2,7 +2,6 @@ #include #include #include -#include struct inode; struct dentry; diff --git a/fs/reiserfs/xattr_acl.c b/fs/reiserfs/xattr_acl.c index f09094057eaa..44474f9b990d 100644 --- a/fs/reiserfs/xattr_acl.c +++ b/fs/reiserfs/xattr_acl.c @@ -1,7 +1,7 @@ #include #include #include -#include +#include "reiserfs.h" #include #include #include diff --git a/fs/reiserfs/xattr_security.c b/fs/reiserfs/xattr_security.c index 6104066e44c4..800a3cef6f62 100644 --- a/fs/reiserfs/xattr_security.c +++ b/fs/reiserfs/xattr_security.c @@ -1,4 +1,4 @@ -#include +#include "reiserfs.h" #include #include #include diff --git a/fs/reiserfs/xattr_trusted.c b/fs/reiserfs/xattr_trusted.c index f995b258c141..a0035719f66b 100644 --- a/fs/reiserfs/xattr_trusted.c +++ b/fs/reiserfs/xattr_trusted.c @@ -1,4 +1,4 @@ -#include +#include "reiserfs.h" #include #include #include diff --git a/fs/reiserfs/xattr_user.c b/fs/reiserfs/xattr_user.c index 748978db6f03..8667491ae7c3 100644 --- a/fs/reiserfs/xattr_user.c +++ b/fs/reiserfs/xattr_user.c @@ -1,4 +1,4 @@ -#include +#include "reiserfs.h" #include #include #include diff --git a/include/linux/reiserfs_fs.h b/include/linux/reiserfs_fs.h index 2213ddcce20c..ea3700cd7367 100644 --- a/include/linux/reiserfs_fs.h +++ b/include/linux/reiserfs_fs.h @@ -1,32 +1,12 @@ /* * Copyright 1996, 1997, 1998 Hans Reiser, see reiserfs/README for licensing and copyright details */ - - /* this file has an amazingly stupid - name, yura please fix it to be - reiserfs.h, and merge all the rest - of our .h files that are in this - directory into it. */ - #ifndef _LINUX_REISER_FS_H #define _LINUX_REISER_FS_H #include #include -#ifdef __KERNEL__ -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#endif - /* * include/linux/reiser_fs.h * @@ -43,2318 +23,4 @@ #define REISERFS_IOC_GETVERSION FS_IOC_GETVERSION #define REISERFS_IOC_SETVERSION FS_IOC_SETVERSION -#ifdef __KERNEL__ -/* the 32 bit compat definitions with int argument */ -#define REISERFS_IOC32_UNPACK _IOW(0xCD, 1, int) -#define REISERFS_IOC32_GETFLAGS FS_IOC32_GETFLAGS -#define REISERFS_IOC32_SETFLAGS FS_IOC32_SETFLAGS -#define REISERFS_IOC32_GETVERSION FS_IOC32_GETVERSION -#define REISERFS_IOC32_SETVERSION FS_IOC32_SETVERSION - -/* - * Locking primitives. The write lock is a per superblock - * special mutex that has properties close to the Big Kernel Lock - * which was used in the previous locking scheme. - */ -void reiserfs_write_lock(struct super_block *s); -void reiserfs_write_unlock(struct super_block *s); -int reiserfs_write_lock_once(struct super_block *s); -void reiserfs_write_unlock_once(struct super_block *s, int lock_depth); - -#ifdef CONFIG_REISERFS_CHECK -void reiserfs_lock_check_recursive(struct super_block *s); -#else -static inline void reiserfs_lock_check_recursive(struct super_block *s) { } -#endif - -/* - * Several mutexes depend on the write lock. - * However sometimes we want to relax the write lock while we hold - * these mutexes, according to the release/reacquire on schedule() - * properties of the Bkl that were used. - * Reiserfs performances and locking were based on this scheme. - * Now that the write lock is a mutex and not the bkl anymore, doing so - * may result in a deadlock: - * - * A acquire write_lock - * A acquire j_commit_mutex - * A release write_lock and wait for something - * B acquire write_lock - * B can't acquire j_commit_mutex and sleep - * A can't acquire write lock anymore - * deadlock - * - * What we do here is avoiding such deadlock by playing the same game - * than the Bkl: if we can't acquire a mutex that depends on the write lock, - * we release the write lock, wait a bit and then retry. - * - * The mutexes concerned by this hack are: - * - The commit mutex of a journal list - * - The flush mutex - * - The journal lock - * - The inode mutex - */ -static inline void reiserfs_mutex_lock_safe(struct mutex *m, - struct super_block *s) -{ - reiserfs_lock_check_recursive(s); - reiserfs_write_unlock(s); - mutex_lock(m); - reiserfs_write_lock(s); -} - -static inline void -reiserfs_mutex_lock_nested_safe(struct mutex *m, unsigned int subclass, - struct super_block *s) -{ - reiserfs_lock_check_recursive(s); - reiserfs_write_unlock(s); - mutex_lock_nested(m, subclass); - reiserfs_write_lock(s); -} - -static inline void -reiserfs_down_read_safe(struct rw_semaphore *sem, struct super_block *s) -{ - reiserfs_lock_check_recursive(s); - reiserfs_write_unlock(s); - down_read(sem); - reiserfs_write_lock(s); -} - -/* - * When we schedule, we usually want to also release the write lock, - * according to the previous bkl based locking scheme of reiserfs. - */ -static inline void reiserfs_cond_resched(struct super_block *s) -{ - if (need_resched()) { - reiserfs_write_unlock(s); - schedule(); - reiserfs_write_lock(s); - } -} - -struct fid; - -/* in reading the #defines, it may help to understand that they employ - the following abbreviations: - - B = Buffer - I = Item header - H = Height within the tree (should be changed to LEV) - N = Number of the item in the node - STAT = stat data - DEH = Directory Entry Header - EC = Entry Count - E = Entry number - UL = Unsigned Long - BLKH = BLocK Header - UNFM = UNForMatted node - DC = Disk Child - P = Path - - These #defines are named by concatenating these abbreviations, - where first comes the arguments, and last comes the return value, - of the macro. - -*/ - -#define USE_INODE_GENERATION_COUNTER - -#define REISERFS_PREALLOCATE -#define DISPLACE_NEW_PACKING_LOCALITIES -#define PREALLOCATION_SIZE 9 - -/* n must be power of 2 */ -#define _ROUND_UP(x,n) (((x)+(n)-1u) & ~((n)-1u)) - -// to be ok for alpha and others we have to align structures to 8 byte -// boundary. -// FIXME: do not change 4 by anything else: there is code which relies on that -#define ROUND_UP(x) _ROUND_UP(x,8LL) - -/* debug levels. Right now, CONFIG_REISERFS_CHECK means print all debug -** messages. -*/ -#define REISERFS_DEBUG_CODE 5 /* extra messages to help find/debug errors */ - -void __reiserfs_warning(struct super_block *s, const char *id, - const char *func, const char *fmt, ...); -#define reiserfs_warning(s, id, fmt, args...) \ - __reiserfs_warning(s, id, __func__, fmt, ##args) -/* assertions handling */ - -/** always check a condition and panic if it's false. */ -#define __RASSERT(cond, scond, format, args...) \ -do { \ - if (!(cond)) \ - reiserfs_panic(NULL, "assertion failure", "(" #cond ") at " \ - __FILE__ ":%i:%s: " format "\n", \ - in_interrupt() ? -1 : task_pid_nr(current), \ - __LINE__, __func__ , ##args); \ -} while (0) - -#define RASSERT(cond, format, args...) __RASSERT(cond, #cond, format, ##args) - -#if defined( CONFIG_REISERFS_CHECK ) -#define RFALSE(cond, format, args...) __RASSERT(!(cond), "!(" #cond ")", format, ##args) -#else -#define RFALSE( cond, format, args... ) do {;} while( 0 ) -#endif - -#define CONSTF __attribute_const__ -/* - * Disk Data Structures - */ - -/***************************************************************************/ -/* SUPER BLOCK */ -/***************************************************************************/ - -/* - * Structure of super block on disk, a version of which in RAM is often accessed as REISERFS_SB(s)->s_rs - * the version in RAM is part of a larger structure containing fields never written to disk. - */ -#define UNSET_HASH 0 // read_super will guess about, what hash names - // in directories were sorted with -#define TEA_HASH 1 -#define YURA_HASH 2 -#define R5_HASH 3 -#define DEFAULT_HASH R5_HASH - -struct journal_params { - __le32 jp_journal_1st_block; /* where does journal start from on its - * device */ - __le32 jp_journal_dev; /* journal device st_rdev */ - __le32 jp_journal_size; /* size of the journal */ - __le32 jp_journal_trans_max; /* max number of blocks in a transaction. */ - __le32 jp_journal_magic; /* random value made on fs creation (this - * was sb_journal_block_count) */ - __le32 jp_journal_max_batch; /* max number of blocks to batch into a - * trans */ - __le32 jp_journal_max_commit_age; /* in seconds, how old can an async - * commit be */ - __le32 jp_journal_max_trans_age; /* in seconds, how old can a transaction - * be */ -}; - -/* this is the super from 3.5.X, where X >= 10 */ -struct reiserfs_super_block_v1 { - __le32 s_block_count; /* blocks count */ - __le32 s_free_blocks; /* free blocks count */ - __le32 s_root_block; /* root block number */ - struct journal_params s_journal; - __le16 s_blocksize; /* block size */ - __le16 s_oid_maxsize; /* max size of object id array, see - * get_objectid() commentary */ - __le16 s_oid_cursize; /* current size of object id array */ - __le16 s_umount_state; /* this is set to 1 when filesystem was - * umounted, to 2 - when not */ - char s_magic[10]; /* reiserfs magic string indicates that - * file system is reiserfs: - * "ReIsErFs" or "ReIsEr2Fs" or "ReIsEr3Fs" */ - __le16 s_fs_state; /* it is set to used by fsck to mark which - * phase of rebuilding is done */ - __le32 s_hash_function_code; /* indicate, what hash function is being use - * to sort names in a directory*/ - __le16 s_tree_height; /* height of disk tree */ - __le16 s_bmap_nr; /* amount of bitmap blocks needed to address - * each block of file system */ - __le16 s_version; /* this field is only reliable on filesystem - * with non-standard journal */ - __le16 s_reserved_for_journal; /* size in blocks of journal area on main - * device, we need to keep after - * making fs with non-standard journal */ -} __attribute__ ((__packed__)); - -#define SB_SIZE_V1 (sizeof(struct reiserfs_super_block_v1)) - -/* this is the on disk super block */ -struct reiserfs_super_block { - struct reiserfs_super_block_v1 s_v1; - __le32 s_inode_generation; - __le32 s_flags; /* Right now used only by inode-attributes, if enabled */ - unsigned char s_uuid[16]; /* filesystem unique identifier */ - unsigned char s_label[16]; /* filesystem volume label */ - __le16 s_mnt_count; /* Count of mounts since last fsck */ - __le16 s_max_mnt_count; /* Maximum mounts before check */ - __le32 s_lastcheck; /* Timestamp of last fsck */ - __le32 s_check_interval; /* Interval between checks */ - char s_unused[76]; /* zero filled by mkreiserfs and - * reiserfs_convert_objectid_map_v1() - * so any additions must be updated - * there as well. */ -} __attribute__ ((__packed__)); - -#define SB_SIZE (sizeof(struct reiserfs_super_block)) - -#define REISERFS_VERSION_1 0 -#define REISERFS_VERSION_2 2 - -// on-disk super block fields converted to cpu form -#define SB_DISK_SUPER_BLOCK(s) (REISERFS_SB(s)->s_rs) -#define SB_V1_DISK_SUPER_BLOCK(s) (&(SB_DISK_SUPER_BLOCK(s)->s_v1)) -#define SB_BLOCKSIZE(s) \ - le32_to_cpu ((SB_V1_DISK_SUPER_BLOCK(s)->s_blocksize)) -#define SB_BLOCK_COUNT(s) \ - le32_to_cpu ((SB_V1_DISK_SUPER_BLOCK(s)->s_block_count)) -#define SB_FREE_BLOCKS(s) \ - le32_to_cpu ((SB_V1_DISK_SUPER_BLOCK(s)->s_free_blocks)) -#define SB_REISERFS_MAGIC(s) \ - (SB_V1_DISK_SUPER_BLOCK(s)->s_magic) -#define SB_ROOT_BLOCK(s) \ - le32_to_cpu ((SB_V1_DISK_SUPER_BLOCK(s)->s_root_block)) -#define SB_TREE_HEIGHT(s) \ - le16_to_cpu ((SB_V1_DISK_SUPER_BLOCK(s)->s_tree_height)) -#define SB_REISERFS_STATE(s) \ - le16_to_cpu ((SB_V1_DISK_SUPER_BLOCK(s)->s_umount_state)) -#define SB_VERSION(s) le16_to_cpu ((SB_V1_DISK_SUPER_BLOCK(s)->s_version)) -#define SB_BMAP_NR(s) le16_to_cpu ((SB_V1_DISK_SUPER_BLOCK(s)->s_bmap_nr)) - -#define PUT_SB_BLOCK_COUNT(s, val) \ - do { SB_V1_DISK_SUPER_BLOCK(s)->s_block_count = cpu_to_le32(val); } while (0) -#define PUT_SB_FREE_BLOCKS(s, val) \ - do { SB_V1_DISK_SUPER_BLOCK(s)->s_free_blocks = cpu_to_le32(val); } while (0) -#define PUT_SB_ROOT_BLOCK(s, val) \ - do { SB_V1_DISK_SUPER_BLOCK(s)->s_root_block = cpu_to_le32(val); } while (0) -#define PUT_SB_TREE_HEIGHT(s, val) \ - do { SB_V1_DISK_SUPER_BLOCK(s)->s_tree_height = cpu_to_le16(val); } while (0) -#define PUT_SB_REISERFS_STATE(s, val) \ - do { SB_V1_DISK_SUPER_BLOCK(s)->s_umount_state = cpu_to_le16(val); } while (0) -#define PUT_SB_VERSION(s, val) \ - do { SB_V1_DISK_SUPER_BLOCK(s)->s_version = cpu_to_le16(val); } while (0) -#define PUT_SB_BMAP_NR(s, val) \ - do { SB_V1_DISK_SUPER_BLOCK(s)->s_bmap_nr = cpu_to_le16 (val); } while (0) - -#define SB_ONDISK_JP(s) (&SB_V1_DISK_SUPER_BLOCK(s)->s_journal) -#define SB_ONDISK_JOURNAL_SIZE(s) \ - le32_to_cpu ((SB_ONDISK_JP(s)->jp_journal_size)) -#define SB_ONDISK_JOURNAL_1st_BLOCK(s) \ - le32_to_cpu ((SB_ONDISK_JP(s)->jp_journal_1st_block)) -#define SB_ONDISK_JOURNAL_DEVICE(s) \ - le32_to_cpu ((SB_ONDISK_JP(s)->jp_journal_dev)) -#define SB_ONDISK_RESERVED_FOR_JOURNAL(s) \ - le16_to_cpu ((SB_V1_DISK_SUPER_BLOCK(s)->s_reserved_for_journal)) - -#define is_block_in_log_or_reserved_area(s, block) \ - block >= SB_JOURNAL_1st_RESERVED_BLOCK(s) \ - && block < SB_JOURNAL_1st_RESERVED_BLOCK(s) + \ - ((!is_reiserfs_jr(SB_DISK_SUPER_BLOCK(s)) ? \ - SB_ONDISK_JOURNAL_SIZE(s) + 1 : SB_ONDISK_RESERVED_FOR_JOURNAL(s))) - -int is_reiserfs_3_5(struct reiserfs_super_block *rs); -int is_reiserfs_3_6(struct reiserfs_super_block *rs); -int is_reiserfs_jr(struct reiserfs_super_block *rs); - -/* ReiserFS leaves the first 64k unused, so that partition labels have - enough space. If someone wants to write a fancy bootloader that - needs more than 64k, let us know, and this will be increased in size. - This number must be larger than than the largest block size on any - platform, or code will break. -Hans */ -#define REISERFS_DISK_OFFSET_IN_BYTES (64 * 1024) -#define REISERFS_FIRST_BLOCK unused_define -#define REISERFS_JOURNAL_OFFSET_IN_BYTES REISERFS_DISK_OFFSET_IN_BYTES - -/* the spot for the super in versions 3.5 - 3.5.10 (inclusive) */ -#define REISERFS_OLD_DISK_OFFSET_IN_BYTES (8 * 1024) - -/* reiserfs internal error code (used by search_by_key and fix_nodes)) */ -#define CARRY_ON 0 -#define REPEAT_SEARCH -1 -#define IO_ERROR -2 -#define NO_DISK_SPACE -3 -#define NO_BALANCING_NEEDED (-4) -#define NO_MORE_UNUSED_CONTIGUOUS_BLOCKS (-5) -#define QUOTA_EXCEEDED -6 - -typedef __u32 b_blocknr_t; -typedef __le32 unp_t; - -struct unfm_nodeinfo { - unp_t unfm_nodenum; - unsigned short unfm_freespace; -}; - -/* there are two formats of keys: 3.5 and 3.6 - */ -#define KEY_FORMAT_3_5 0 -#define KEY_FORMAT_3_6 1 - -/* there are two stat datas */ -#define STAT_DATA_V1 0 -#define STAT_DATA_V2 1 - -static inline struct reiserfs_inode_info *REISERFS_I(const struct inode *inode) -{ - return container_of(inode, struct reiserfs_inode_info, vfs_inode); -} - -static inline struct reiserfs_sb_info *REISERFS_SB(const struct super_block *sb) -{ - return sb->s_fs_info; -} - -/* Don't trust REISERFS_SB(sb)->s_bmap_nr, it's a u16 - * which overflows on large file systems. */ -static inline __u32 reiserfs_bmap_count(struct super_block *sb) -{ - return (SB_BLOCK_COUNT(sb) - 1) / (sb->s_blocksize * 8) + 1; -} - -static inline int bmap_would_wrap(unsigned bmap_nr) -{ - return bmap_nr > ((1LL << 16) - 1); -} - -/** this says about version of key of all items (but stat data) the - object consists of */ -#define get_inode_item_key_version( inode ) \ - ((REISERFS_I(inode)->i_flags & i_item_key_version_mask) ? KEY_FORMAT_3_6 : KEY_FORMAT_3_5) - -#define set_inode_item_key_version( inode, version ) \ - ({ if((version)==KEY_FORMAT_3_6) \ - REISERFS_I(inode)->i_flags |= i_item_key_version_mask; \ - else \ - REISERFS_I(inode)->i_flags &= ~i_item_key_version_mask; }) - -#define get_inode_sd_version(inode) \ - ((REISERFS_I(inode)->i_flags & i_stat_data_version_mask) ? STAT_DATA_V2 : STAT_DATA_V1) - -#define set_inode_sd_version(inode, version) \ - ({ if((version)==STAT_DATA_V2) \ - REISERFS_I(inode)->i_flags |= i_stat_data_version_mask; \ - else \ - REISERFS_I(inode)->i_flags &= ~i_stat_data_version_mask; }) - -/* This is an aggressive tail suppression policy, I am hoping it - improves our benchmarks. The principle behind it is that percentage - space saving is what matters, not absolute space saving. This is - non-intuitive, but it helps to understand it if you consider that the - cost to access 4 blocks is not much more than the cost to access 1 - block, if you have to do a seek and rotate. A tail risks a - non-linear disk access that is significant as a percentage of total - time cost for a 4 block file and saves an amount of space that is - less significant as a percentage of space, or so goes the hypothesis. - -Hans */ -#define STORE_TAIL_IN_UNFM_S1(n_file_size,n_tail_size,n_block_size) \ -(\ - (!(n_tail_size)) || \ - (((n_tail_size) > MAX_DIRECT_ITEM_LEN(n_block_size)) || \ - ( (n_file_size) >= (n_block_size) * 4 ) || \ - ( ( (n_file_size) >= (n_block_size) * 3 ) && \ - ( (n_tail_size) >= (MAX_DIRECT_ITEM_LEN(n_block_size))/4) ) || \ - ( ( (n_file_size) >= (n_block_size) * 2 ) && \ - ( (n_tail_size) >= (MAX_DIRECT_ITEM_LEN(n_block_size))/2) ) || \ - ( ( (n_file_size) >= (n_block_size) ) && \ - ( (n_tail_size) >= (MAX_DIRECT_ITEM_LEN(n_block_size) * 3)/4) ) ) \ -) - -/* Another strategy for tails, this one means only create a tail if all the - file would fit into one DIRECT item. - Primary intention for this one is to increase performance by decreasing - seeking. -*/ -#define STORE_TAIL_IN_UNFM_S2(n_file_size,n_tail_size,n_block_size) \ -(\ - (!(n_tail_size)) || \ - (((n_file_size) > MAX_DIRECT_ITEM_LEN(n_block_size)) ) \ -) - -/* - * values for s_umount_state field - */ -#define REISERFS_VALID_FS 1 -#define REISERFS_ERROR_FS 2 - -// -// there are 5 item types currently -// -#define TYPE_STAT_DATA 0 -#define TYPE_INDIRECT 1 -#define TYPE_DIRECT 2 -#define TYPE_DIRENTRY 3 -#define TYPE_MAXTYPE 3 -#define TYPE_ANY 15 // FIXME: comment is required - -/***************************************************************************/ -/* KEY & ITEM HEAD */ -/***************************************************************************/ - -// -// directories use this key as well as old files -// -struct offset_v1 { - __le32 k_offset; - __le32 k_uniqueness; -} __attribute__ ((__packed__)); - -struct offset_v2 { - __le64 v; -} __attribute__ ((__packed__)); - -static inline __u16 offset_v2_k_type(const struct offset_v2 *v2) -{ - __u8 type = le64_to_cpu(v2->v) >> 60; - return (type <= TYPE_MAXTYPE) ? type : TYPE_ANY; -} - -static inline void set_offset_v2_k_type(struct offset_v2 *v2, int type) -{ - v2->v = - (v2->v & cpu_to_le64(~0ULL >> 4)) | cpu_to_le64((__u64) type << 60); -} - -static inline loff_t offset_v2_k_offset(const struct offset_v2 *v2) -{ - return le64_to_cpu(v2->v) & (~0ULL >> 4); -} - -static inline void set_offset_v2_k_offset(struct offset_v2 *v2, loff_t offset) -{ - offset &= (~0ULL >> 4); - v2->v = (v2->v & cpu_to_le64(15ULL << 60)) | cpu_to_le64(offset); -} - -/* Key of an item determines its location in the S+tree, and - is composed of 4 components */ -struct reiserfs_key { - __le32 k_dir_id; /* packing locality: by default parent - directory object id */ - __le32 k_objectid; /* object identifier */ - union { - struct offset_v1 k_offset_v1; - struct offset_v2 k_offset_v2; - } __attribute__ ((__packed__)) u; -} __attribute__ ((__packed__)); - -struct in_core_key { - __u32 k_dir_id; /* packing locality: by default parent - directory object id */ - __u32 k_objectid; /* object identifier */ - __u64 k_offset; - __u8 k_type; -}; - -struct cpu_key { - struct in_core_key on_disk_key; - int version; - int key_length; /* 3 in all cases but direct2indirect and - indirect2direct conversion */ -}; - -/* Our function for comparing keys can compare keys of different - lengths. It takes as a parameter the length of the keys it is to - compare. These defines are used in determining what is to be passed - to it as that parameter. */ -#define REISERFS_FULL_KEY_LEN 4 -#define REISERFS_SHORT_KEY_LEN 2 - -/* The result of the key compare */ -#define FIRST_GREATER 1 -#define SECOND_GREATER -1 -#define KEYS_IDENTICAL 0 -#define KEY_FOUND 1 -#define KEY_NOT_FOUND 0 - -#define KEY_SIZE (sizeof(struct reiserfs_key)) -#define SHORT_KEY_SIZE (sizeof (__u32) + sizeof (__u32)) - -/* return values for search_by_key and clones */ -#define ITEM_FOUND 1 -#define ITEM_NOT_FOUND 0 -#define ENTRY_FOUND 1 -#define ENTRY_NOT_FOUND 0 -#define DIRECTORY_NOT_FOUND -1 -#define REGULAR_FILE_FOUND -2 -#define DIRECTORY_FOUND -3 -#define BYTE_FOUND 1 -#define BYTE_NOT_FOUND 0 -#define FILE_NOT_FOUND -1 - -#define POSITION_FOUND 1 -#define POSITION_NOT_FOUND 0 - -// return values for reiserfs_find_entry and search_by_entry_key -#define NAME_FOUND 1 -#define NAME_NOT_FOUND 0 -#define GOTO_PREVIOUS_ITEM 2 -#define NAME_FOUND_INVISIBLE 3 - -/* Everything in the filesystem is stored as a set of items. The - item head contains the key of the item, its free space (for - indirect items) and specifies the location of the item itself - within the block. */ - -struct item_head { - /* Everything in the tree is found by searching for it based on - * its key.*/ - struct reiserfs_key ih_key; - union { - /* The free space in the last unformatted node of an - indirect item if this is an indirect item. This - equals 0xFFFF iff this is a direct item or stat data - item. Note that the key, not this field, is used to - determine the item type, and thus which field this - union contains. */ - __le16 ih_free_space_reserved; - /* Iff this is a directory item, this field equals the - number of directory entries in the directory item. */ - __le16 ih_entry_count; - } __attribute__ ((__packed__)) u; - __le16 ih_item_len; /* total size of the item body */ - __le16 ih_item_location; /* an offset to the item body - * within the block */ - __le16 ih_version; /* 0 for all old items, 2 for new - ones. Highest bit is set by fsck - temporary, cleaned after all - done */ -} __attribute__ ((__packed__)); -/* size of item header */ -#define IH_SIZE (sizeof(struct item_head)) - -#define ih_free_space(ih) le16_to_cpu((ih)->u.ih_free_space_reserved) -#define ih_version(ih) le16_to_cpu((ih)->ih_version) -#define ih_entry_count(ih) le16_to_cpu((ih)->u.ih_entry_count) -#define ih_location(ih) le16_to_cpu((ih)->ih_item_location) -#define ih_item_len(ih) le16_to_cpu((ih)->ih_item_len) - -#define put_ih_free_space(ih, val) do { (ih)->u.ih_free_space_reserved = cpu_to_le16(val); } while(0) -#define put_ih_version(ih, val) do { (ih)->ih_version = cpu_to_le16(val); } while (0) -#define put_ih_entry_count(ih, val) do { (ih)->u.ih_entry_count = cpu_to_le16(val); } while (0) -#define put_ih_location(ih, val) do { (ih)->ih_item_location = cpu_to_le16(val); } while (0) -#define put_ih_item_len(ih, val) do { (ih)->ih_item_len = cpu_to_le16(val); } while (0) - -#define unreachable_item(ih) (ih_version(ih) & (1 << 15)) - -#define get_ih_free_space(ih) (ih_version (ih) == KEY_FORMAT_3_6 ? 0 : ih_free_space (ih)) -#define set_ih_free_space(ih,val) put_ih_free_space((ih), ((ih_version(ih) == KEY_FORMAT_3_6) ? 0 : (val))) - -/* these operate on indirect items, where you've got an array of ints -** at a possibly unaligned location. These are a noop on ia32 -** -** p is the array of __u32, i is the index into the array, v is the value -** to store there. -*/ -#define get_block_num(p, i) get_unaligned_le32((p) + (i)) -#define put_block_num(p, i, v) put_unaligned_le32((v), (p) + (i)) - -// -// in old version uniqueness field shows key type -// -#define V1_SD_UNIQUENESS 0 -#define V1_INDIRECT_UNIQUENESS 0xfffffffe -#define V1_DIRECT_UNIQUENESS 0xffffffff -#define V1_DIRENTRY_UNIQUENESS 500 -#define V1_ANY_UNIQUENESS 555 // FIXME: comment is required - -// -// here are conversion routines -// -static inline int uniqueness2type(__u32 uniqueness) CONSTF; -static inline int uniqueness2type(__u32 uniqueness) -{ - switch ((int)uniqueness) { - case V1_SD_UNIQUENESS: - return TYPE_STAT_DATA; - case V1_INDIRECT_UNIQUENESS: - return TYPE_INDIRECT; - case V1_DIRECT_UNIQUENESS: - return TYPE_DIRECT; - case V1_DIRENTRY_UNIQUENESS: - return TYPE_DIRENTRY; - case V1_ANY_UNIQUENESS: - default: - return TYPE_ANY; - } -} - -static inline __u32 type2uniqueness(int type) CONSTF; -static inline __u32 type2uniqueness(int type) -{ - switch (type) { - case TYPE_STAT_DATA: - return V1_SD_UNIQUENESS; - case TYPE_INDIRECT: - return V1_INDIRECT_UNIQUENESS; - case TYPE_DIRECT: - return V1_DIRECT_UNIQUENESS; - case TYPE_DIRENTRY: - return V1_DIRENTRY_UNIQUENESS; - case TYPE_ANY: - default: - return V1_ANY_UNIQUENESS; - } -} - -// -// key is pointer to on disk key which is stored in le, result is cpu, -// there is no way to get version of object from key, so, provide -// version to these defines -// -static inline loff_t le_key_k_offset(int version, - const struct reiserfs_key *key) -{ - return (version == KEY_FORMAT_3_5) ? - le32_to_cpu(key->u.k_offset_v1.k_offset) : - offset_v2_k_offset(&(key->u.k_offset_v2)); -} - -static inline loff_t le_ih_k_offset(const struct item_head *ih) -{ - return le_key_k_offset(ih_version(ih), &(ih->ih_key)); -} - -static inline loff_t le_key_k_type(int version, const struct reiserfs_key *key) -{ - return (version == KEY_FORMAT_3_5) ? - uniqueness2type(le32_to_cpu(key->u.k_offset_v1.k_uniqueness)) : - offset_v2_k_type(&(key->u.k_offset_v2)); -} - -static inline loff_t le_ih_k_type(const struct item_head *ih) -{ - return le_key_k_type(ih_version(ih), &(ih->ih_key)); -} - -static inline void set_le_key_k_offset(int version, struct reiserfs_key *key, - loff_t offset) -{ - (version == KEY_FORMAT_3_5) ? (void)(key->u.k_offset_v1.k_offset = cpu_to_le32(offset)) : /* jdm check */ - (void)(set_offset_v2_k_offset(&(key->u.k_offset_v2), offset)); -} - -static inline void set_le_ih_k_offset(struct item_head *ih, loff_t offset) -{ - set_le_key_k_offset(ih_version(ih), &(ih->ih_key), offset); -} - -static inline void set_le_key_k_type(int version, struct reiserfs_key *key, - int type) -{ - (version == KEY_FORMAT_3_5) ? - (void)(key->u.k_offset_v1.k_uniqueness = - cpu_to_le32(type2uniqueness(type))) - : (void)(set_offset_v2_k_type(&(key->u.k_offset_v2), type)); -} - -static inline void set_le_ih_k_type(struct item_head *ih, int type) -{ - set_le_key_k_type(ih_version(ih), &(ih->ih_key), type); -} - -static inline int is_direntry_le_key(int version, struct reiserfs_key *key) -{ - return le_key_k_type(version, key) == TYPE_DIRENTRY; -} - -static inline int is_direct_le_key(int version, struct reiserfs_key *key) -{ - return le_key_k_type(version, key) == TYPE_DIRECT; -} - -static inline int is_indirect_le_key(int version, struct reiserfs_key *key) -{ - return le_key_k_type(version, key) == TYPE_INDIRECT; -} - -static inline int is_statdata_le_key(int version, struct reiserfs_key *key) -{ - return le_key_k_type(version, key) == TYPE_STAT_DATA; -} - -// -// item header has version. -// -static inline int is_direntry_le_ih(struct item_head *ih) -{ - return is_direntry_le_key(ih_version(ih), &ih->ih_key); -} - -static inline int is_direct_le_ih(struct item_head *ih) -{ - return is_direct_le_key(ih_version(ih), &ih->ih_key); -} - -static inline int is_indirect_le_ih(struct item_head *ih) -{ - return is_indirect_le_key(ih_version(ih), &ih->ih_key); -} - -static inline int is_statdata_le_ih(struct item_head *ih) -{ - return is_statdata_le_key(ih_version(ih), &ih->ih_key); -} - -// -// key is pointer to cpu key, result is cpu -// -static inline loff_t cpu_key_k_offset(const struct cpu_key *key) -{ - return key->on_disk_key.k_offset; -} - -static inline loff_t cpu_key_k_type(const struct cpu_key *key) -{ - return key->on_disk_key.k_type; -} - -static inline void set_cpu_key_k_offset(struct cpu_key *key, loff_t offset) -{ - key->on_disk_key.k_offset = offset; -} - -static inline void set_cpu_key_k_type(struct cpu_key *key, int type) -{ - key->on_disk_key.k_type = type; -} - -static inline void cpu_key_k_offset_dec(struct cpu_key *key) -{ - key->on_disk_key.k_offset--; -} - -#define is_direntry_cpu_key(key) (cpu_key_k_type (key) == TYPE_DIRENTRY) -#define is_direct_cpu_key(key) (cpu_key_k_type (key) == TYPE_DIRECT) -#define is_indirect_cpu_key(key) (cpu_key_k_type (key) == TYPE_INDIRECT) -#define is_statdata_cpu_key(key) (cpu_key_k_type (key) == TYPE_STAT_DATA) - -/* are these used ? */ -#define is_direntry_cpu_ih(ih) (is_direntry_cpu_key (&((ih)->ih_key))) -#define is_direct_cpu_ih(ih) (is_direct_cpu_key (&((ih)->ih_key))) -#define is_indirect_cpu_ih(ih) (is_indirect_cpu_key (&((ih)->ih_key))) -#define is_statdata_cpu_ih(ih) (is_statdata_cpu_key (&((ih)->ih_key))) - -#define I_K_KEY_IN_ITEM(ih, key, n_blocksize) \ - (!COMP_SHORT_KEYS(ih, key) && \ - I_OFF_BYTE_IN_ITEM(ih, k_offset(key), n_blocksize)) - -/* maximal length of item */ -#define MAX_ITEM_LEN(block_size) (block_size - BLKH_SIZE - IH_SIZE) -#define MIN_ITEM_LEN 1 - -/* object identifier for root dir */ -#define REISERFS_ROOT_OBJECTID 2 -#define REISERFS_ROOT_PARENT_OBJECTID 1 - -extern struct reiserfs_key root_key; - -/* - * Picture represents a leaf of the S+tree - * ______________________________________________________ - * | | Array of | | | - * |Block | Object-Item | F r e e | Objects- | - * | head | Headers | S p a c e | Items | - * |______|_______________|___________________|___________| - */ - -/* Header of a disk block. More precisely, header of a formatted leaf - or internal node, and not the header of an unformatted node. */ -struct block_head { - __le16 blk_level; /* Level of a block in the tree. */ - __le16 blk_nr_item; /* Number of keys/items in a block. */ - __le16 blk_free_space; /* Block free space in bytes. */ - __le16 blk_reserved; - /* dump this in v4/planA */ - struct reiserfs_key blk_right_delim_key; /* kept only for compatibility */ -}; - -#define BLKH_SIZE (sizeof(struct block_head)) -#define blkh_level(p_blkh) (le16_to_cpu((p_blkh)->blk_level)) -#define blkh_nr_item(p_blkh) (le16_to_cpu((p_blkh)->blk_nr_item)) -#define blkh_free_space(p_blkh) (le16_to_cpu((p_blkh)->blk_free_space)) -#define blkh_reserved(p_blkh) (le16_to_cpu((p_blkh)->blk_reserved)) -#define set_blkh_level(p_blkh,val) ((p_blkh)->blk_level = cpu_to_le16(val)) -#define set_blkh_nr_item(p_blkh,val) ((p_blkh)->blk_nr_item = cpu_to_le16(val)) -#define set_blkh_free_space(p_blkh,val) ((p_blkh)->blk_free_space = cpu_to_le16(val)) -#define set_blkh_reserved(p_blkh,val) ((p_blkh)->blk_reserved = cpu_to_le16(val)) -#define blkh_right_delim_key(p_blkh) ((p_blkh)->blk_right_delim_key) -#define set_blkh_right_delim_key(p_blkh,val) ((p_blkh)->blk_right_delim_key = val) - -/* - * values for blk_level field of the struct block_head - */ - -#define FREE_LEVEL 0 /* when node gets removed from the tree its - blk_level is set to FREE_LEVEL. It is then - used to see whether the node is still in the - tree */ - -#define DISK_LEAF_NODE_LEVEL 1 /* Leaf node level. */ - -/* Given the buffer head of a formatted node, resolve to the block head of that node. */ -#define B_BLK_HEAD(bh) ((struct block_head *)((bh)->b_data)) -/* Number of items that are in buffer. */ -#define B_NR_ITEMS(bh) (blkh_nr_item(B_BLK_HEAD(bh))) -#define B_LEVEL(bh) (blkh_level(B_BLK_HEAD(bh))) -#define B_FREE_SPACE(bh) (blkh_free_space(B_BLK_HEAD(bh))) - -#define PUT_B_NR_ITEMS(bh, val) do { set_blkh_nr_item(B_BLK_HEAD(bh), val); } while (0) -#define PUT_B_LEVEL(bh, val) do { set_blkh_level(B_BLK_HEAD(bh), val); } while (0) -#define PUT_B_FREE_SPACE(bh, val) do { set_blkh_free_space(B_BLK_HEAD(bh), val); } while (0) - -/* Get right delimiting key. -- little endian */ -#define B_PRIGHT_DELIM_KEY(bh) (&(blk_right_delim_key(B_BLK_HEAD(bh)))) - -/* Does the buffer contain a disk leaf. */ -#define B_IS_ITEMS_LEVEL(bh) (B_LEVEL(bh) == DISK_LEAF_NODE_LEVEL) - -/* Does the buffer contain a disk internal node */ -#define B_IS_KEYS_LEVEL(bh) (B_LEVEL(bh) > DISK_LEAF_NODE_LEVEL \ - && B_LEVEL(bh) <= MAX_HEIGHT) - -/***************************************************************************/ -/* STAT DATA */ -/***************************************************************************/ - -// -// old stat data is 32 bytes long. We are going to distinguish new one by -// different size -// -struct stat_data_v1 { - __le16 sd_mode; /* file type, permissions */ - __le16 sd_nlink; /* number of hard links */ - __le16 sd_uid; /* owner */ - __le16 sd_gid; /* group */ - __le32 sd_size; /* file size */ - __le32 sd_atime; /* time of last access */ - __le32 sd_mtime; /* time file was last modified */ - __le32 sd_ctime; /* time inode (stat data) was last changed (except changes to sd_atime and sd_mtime) */ - union { - __le32 sd_rdev; - __le32 sd_blocks; /* number of blocks file uses */ - } __attribute__ ((__packed__)) u; - __le32 sd_first_direct_byte; /* first byte of file which is stored - in a direct item: except that if it - equals 1 it is a symlink and if it - equals ~(__u32)0 there is no - direct item. The existence of this - field really grates on me. Let's - replace it with a macro based on - sd_size and our tail suppression - policy. Someday. -Hans */ -} __attribute__ ((__packed__)); - -#define SD_V1_SIZE (sizeof(struct stat_data_v1)) -#define stat_data_v1(ih) (ih_version (ih) == KEY_FORMAT_3_5) -#define sd_v1_mode(sdp) (le16_to_cpu((sdp)->sd_mode)) -#define set_sd_v1_mode(sdp,v) ((sdp)->sd_mode = cpu_to_le16(v)) -#define sd_v1_nlink(sdp) (le16_to_cpu((sdp)->sd_nlink)) -#define set_sd_v1_nlink(sdp,v) ((sdp)->sd_nlink = cpu_to_le16(v)) -#define sd_v1_uid(sdp) (le16_to_cpu((sdp)->sd_uid)) -#define set_sd_v1_uid(sdp,v) ((sdp)->sd_uid = cpu_to_le16(v)) -#define sd_v1_gid(sdp) (le16_to_cpu((sdp)->sd_gid)) -#define set_sd_v1_gid(sdp,v) ((sdp)->sd_gid = cpu_to_le16(v)) -#define sd_v1_size(sdp) (le32_to_cpu((sdp)->sd_size)) -#define set_sd_v1_size(sdp,v) ((sdp)->sd_size = cpu_to_le32(v)) -#define sd_v1_atime(sdp) (le32_to_cpu((sdp)->sd_atime)) -#define set_sd_v1_atime(sdp,v) ((sdp)->sd_atime = cpu_to_le32(v)) -#define sd_v1_mtime(sdp) (le32_to_cpu((sdp)->sd_mtime)) -#define set_sd_v1_mtime(sdp,v) ((sdp)->sd_mtime = cpu_to_le32(v)) -#define sd_v1_ctime(sdp) (le32_to_cpu((sdp)->sd_ctime)) -#define set_sd_v1_ctime(sdp,v) ((sdp)->sd_ctime = cpu_to_le32(v)) -#define sd_v1_rdev(sdp) (le32_to_cpu((sdp)->u.sd_rdev)) -#define set_sd_v1_rdev(sdp,v) ((sdp)->u.sd_rdev = cpu_to_le32(v)) -#define sd_v1_blocks(sdp) (le32_to_cpu((sdp)->u.sd_blocks)) -#define set_sd_v1_blocks(sdp,v) ((sdp)->u.sd_blocks = cpu_to_le32(v)) -#define sd_v1_first_direct_byte(sdp) \ - (le32_to_cpu((sdp)->sd_first_direct_byte)) -#define set_sd_v1_first_direct_byte(sdp,v) \ - ((sdp)->sd_first_direct_byte = cpu_to_le32(v)) - -/* inode flags stored in sd_attrs (nee sd_reserved) */ - -/* we want common flags to have the same values as in ext2, - so chattr(1) will work without problems */ -#define REISERFS_IMMUTABLE_FL FS_IMMUTABLE_FL -#define REISERFS_APPEND_FL FS_APPEND_FL -#define REISERFS_SYNC_FL FS_SYNC_FL -#define REISERFS_NOATIME_FL FS_NOATIME_FL -#define REISERFS_NODUMP_FL FS_NODUMP_FL -#define REISERFS_SECRM_FL FS_SECRM_FL -#define REISERFS_UNRM_FL FS_UNRM_FL -#define REISERFS_COMPR_FL FS_COMPR_FL -#define REISERFS_NOTAIL_FL FS_NOTAIL_FL - -/* persistent flags that file inherits from the parent directory */ -#define REISERFS_INHERIT_MASK ( REISERFS_IMMUTABLE_FL | \ - REISERFS_SYNC_FL | \ - REISERFS_NOATIME_FL | \ - REISERFS_NODUMP_FL | \ - REISERFS_SECRM_FL | \ - REISERFS_COMPR_FL | \ - REISERFS_NOTAIL_FL ) - -/* Stat Data on disk (reiserfs version of UFS disk inode minus the - address blocks) */ -struct stat_data { - __le16 sd_mode; /* file type, permissions */ - __le16 sd_attrs; /* persistent inode flags */ - __le32 sd_nlink; /* number of hard links */ - __le64 sd_size; /* file size */ - __le32 sd_uid; /* owner */ - __le32 sd_gid; /* group */ - __le32 sd_atime; /* time of last access */ - __le32 sd_mtime; /* time file was last modified */ - __le32 sd_ctime; /* time inode (stat data) was last changed (except changes to sd_atime and sd_mtime) */ - __le32 sd_blocks; - union { - __le32 sd_rdev; - __le32 sd_generation; - //__le32 sd_first_direct_byte; - /* first byte of file which is stored in a - direct item: except that if it equals 1 - it is a symlink and if it equals - ~(__u32)0 there is no direct item. The - existence of this field really grates - on me. Let's replace it with a macro - based on sd_size and our tail - suppression policy? */ - } __attribute__ ((__packed__)) u; -} __attribute__ ((__packed__)); -// -// this is 44 bytes long -// -#define SD_SIZE (sizeof(struct stat_data)) -#define SD_V2_SIZE SD_SIZE -#define stat_data_v2(ih) (ih_version (ih) == KEY_FORMAT_3_6) -#define sd_v2_mode(sdp) (le16_to_cpu((sdp)->sd_mode)) -#define set_sd_v2_mode(sdp,v) ((sdp)->sd_mode = cpu_to_le16(v)) -/* sd_reserved */ -/* set_sd_reserved */ -#define sd_v2_nlink(sdp) (le32_to_cpu((sdp)->sd_nlink)) -#define set_sd_v2_nlink(sdp,v) ((sdp)->sd_nlink = cpu_to_le32(v)) -#define sd_v2_size(sdp) (le64_to_cpu((sdp)->sd_size)) -#define set_sd_v2_size(sdp,v) ((sdp)->sd_size = cpu_to_le64(v)) -#define sd_v2_uid(sdp) (le32_to_cpu((sdp)->sd_uid)) -#define set_sd_v2_uid(sdp,v) ((sdp)->sd_uid = cpu_to_le32(v)) -#define sd_v2_gid(sdp) (le32_to_cpu((sdp)->sd_gid)) -#define set_sd_v2_gid(sdp,v) ((sdp)->sd_gid = cpu_to_le32(v)) -#define sd_v2_atime(sdp) (le32_to_cpu((sdp)->sd_atime)) -#define set_sd_v2_atime(sdp,v) ((sdp)->sd_atime = cpu_to_le32(v)) -#define sd_v2_mtime(sdp) (le32_to_cpu((sdp)->sd_mtime)) -#define set_sd_v2_mtime(sdp,v) ((sdp)->sd_mtime = cpu_to_le32(v)) -#define sd_v2_ctime(sdp) (le32_to_cpu((sdp)->sd_ctime)) -#define set_sd_v2_ctime(sdp,v) ((sdp)->sd_ctime = cpu_to_le32(v)) -#define sd_v2_blocks(sdp) (le32_to_cpu((sdp)->sd_blocks)) -#define set_sd_v2_blocks(sdp,v) ((sdp)->sd_blocks = cpu_to_le32(v)) -#define sd_v2_rdev(sdp) (le32_to_cpu((sdp)->u.sd_rdev)) -#define set_sd_v2_rdev(sdp,v) ((sdp)->u.sd_rdev = cpu_to_le32(v)) -#define sd_v2_generation(sdp) (le32_to_cpu((sdp)->u.sd_generation)) -#define set_sd_v2_generation(sdp,v) ((sdp)->u.sd_generation = cpu_to_le32(v)) -#define sd_v2_attrs(sdp) (le16_to_cpu((sdp)->sd_attrs)) -#define set_sd_v2_attrs(sdp,v) ((sdp)->sd_attrs = cpu_to_le16(v)) - -/***************************************************************************/ -/* DIRECTORY STRUCTURE */ -/***************************************************************************/ -/* - Picture represents the structure of directory items - ________________________________________________ - | Array of | | | | | | - | directory |N-1| N-2 | .... | 1st |0th| - | entry headers | | | | | | - |_______________|___|_____|________|_______|___| - <---- directory entries ------> - - First directory item has k_offset component 1. We store "." and ".." - in one item, always, we never split "." and ".." into differing - items. This makes, among other things, the code for removing - directories simpler. */ -#define SD_OFFSET 0 -#define SD_UNIQUENESS 0 -#define DOT_OFFSET 1 -#define DOT_DOT_OFFSET 2 -#define DIRENTRY_UNIQUENESS 500 - -/* */ -#define FIRST_ITEM_OFFSET 1 - -/* - Q: How to get key of object pointed to by entry from entry? - - A: Each directory entry has its header. This header has deh_dir_id and deh_objectid fields, those are key - of object, entry points to */ - -/* NOT IMPLEMENTED: - Directory will someday contain stat data of object */ - -struct reiserfs_de_head { - __le32 deh_offset; /* third component of the directory entry key */ - __le32 deh_dir_id; /* objectid of the parent directory of the object, that is referenced - by directory entry */ - __le32 deh_objectid; /* objectid of the object, that is referenced by directory entry */ - __le16 deh_location; /* offset of name in the whole item */ - __le16 deh_state; /* whether 1) entry contains stat data (for future), and 2) whether - entry is hidden (unlinked) */ -} __attribute__ ((__packed__)); -#define DEH_SIZE sizeof(struct reiserfs_de_head) -#define deh_offset(p_deh) (le32_to_cpu((p_deh)->deh_offset)) -#define deh_dir_id(p_deh) (le32_to_cpu((p_deh)->deh_dir_id)) -#define deh_objectid(p_deh) (le32_to_cpu((p_deh)->deh_objectid)) -#define deh_location(p_deh) (le16_to_cpu((p_deh)->deh_location)) -#define deh_state(p_deh) (le16_to_cpu((p_deh)->deh_state)) - -#define put_deh_offset(p_deh,v) ((p_deh)->deh_offset = cpu_to_le32((v))) -#define put_deh_dir_id(p_deh,v) ((p_deh)->deh_dir_id = cpu_to_le32((v))) -#define put_deh_objectid(p_deh,v) ((p_deh)->deh_objectid = cpu_to_le32((v))) -#define put_deh_location(p_deh,v) ((p_deh)->deh_location = cpu_to_le16((v))) -#define put_deh_state(p_deh,v) ((p_deh)->deh_state = cpu_to_le16((v))) - -/* empty directory contains two entries "." and ".." and their headers */ -#define EMPTY_DIR_SIZE \ -(DEH_SIZE * 2 + ROUND_UP (strlen (".")) + ROUND_UP (strlen (".."))) - -/* old format directories have this size when empty */ -#define EMPTY_DIR_SIZE_V1 (DEH_SIZE * 2 + 3) - -#define DEH_Statdata 0 /* not used now */ -#define DEH_Visible 2 - -/* 64 bit systems (and the S/390) need to be aligned explicitly -jdm */ -#if BITS_PER_LONG == 64 || defined(__s390__) || defined(__hppa__) -# define ADDR_UNALIGNED_BITS (3) -#endif - -/* These are only used to manipulate deh_state. - * Because of this, we'll use the ext2_ bit routines, - * since they are little endian */ -#ifdef ADDR_UNALIGNED_BITS - -# define aligned_address(addr) ((void *)((long)(addr) & ~((1UL << ADDR_UNALIGNED_BITS) - 1))) -# define unaligned_offset(addr) (((int)((long)(addr) & ((1 << ADDR_UNALIGNED_BITS) - 1))) << 3) - -# define set_bit_unaligned(nr, addr) \ - __test_and_set_bit_le((nr) + unaligned_offset(addr), aligned_address(addr)) -# define clear_bit_unaligned(nr, addr) \ - __test_and_clear_bit_le((nr) + unaligned_offset(addr), aligned_address(addr)) -# define test_bit_unaligned(nr, addr) \ - test_bit_le((nr) + unaligned_offset(addr), aligned_address(addr)) - -#else - -# define set_bit_unaligned(nr, addr) __test_and_set_bit_le(nr, addr) -# define clear_bit_unaligned(nr, addr) __test_and_clear_bit_le(nr, addr) -# define test_bit_unaligned(nr, addr) test_bit_le(nr, addr) - -#endif - -#define mark_de_with_sd(deh) set_bit_unaligned (DEH_Statdata, &((deh)->deh_state)) -#define mark_de_without_sd(deh) clear_bit_unaligned (DEH_Statdata, &((deh)->deh_state)) -#define mark_de_visible(deh) set_bit_unaligned (DEH_Visible, &((deh)->deh_state)) -#define mark_de_hidden(deh) clear_bit_unaligned (DEH_Visible, &((deh)->deh_state)) - -#define de_with_sd(deh) test_bit_unaligned (DEH_Statdata, &((deh)->deh_state)) -#define de_visible(deh) test_bit_unaligned (DEH_Visible, &((deh)->deh_state)) -#define de_hidden(deh) !test_bit_unaligned (DEH_Visible, &((deh)->deh_state)) - -extern void make_empty_dir_item_v1(char *body, __le32 dirid, __le32 objid, - __le32 par_dirid, __le32 par_objid); -extern void make_empty_dir_item(char *body, __le32 dirid, __le32 objid, - __le32 par_dirid, __le32 par_objid); - -/* array of the entry headers */ - /* get item body */ -#define B_I_PITEM(bh,ih) ( (bh)->b_data + ih_location(ih) ) -#define B_I_DEH(bh,ih) ((struct reiserfs_de_head *)(B_I_PITEM(bh,ih))) - -/* length of the directory entry in directory item. This define - calculates length of i-th directory entry using directory entry - locations from dir entry head. When it calculates length of 0-th - directory entry, it uses length of whole item in place of entry - location of the non-existent following entry in the calculation. - See picture above.*/ -/* -#define I_DEH_N_ENTRY_LENGTH(ih,deh,i) \ -((i) ? (deh_location((deh)-1) - deh_location((deh))) : (ih_item_len((ih)) - deh_location((deh)))) -*/ -static inline int entry_length(const struct buffer_head *bh, - const struct item_head *ih, int pos_in_item) -{ - struct reiserfs_de_head *deh; - - deh = B_I_DEH(bh, ih) + pos_in_item; - if (pos_in_item) - return deh_location(deh - 1) - deh_location(deh); - - return ih_item_len(ih) - deh_location(deh); -} - -/* number of entries in the directory item, depends on ENTRY_COUNT being at the start of directory dynamic data. */ -#define I_ENTRY_COUNT(ih) (ih_entry_count((ih))) - -/* name by bh, ih and entry_num */ -#define B_I_E_NAME(bh,ih,entry_num) ((char *)(bh->b_data + ih_location(ih) + deh_location(B_I_DEH(bh,ih)+(entry_num)))) - -// two entries per block (at least) -#define REISERFS_MAX_NAME(block_size) 255 - -/* this structure is used for operations on directory entries. It is - not a disk structure. */ -/* When reiserfs_find_entry or search_by_entry_key find directory - entry, they return filled reiserfs_dir_entry structure */ -struct reiserfs_dir_entry { - struct buffer_head *de_bh; - int de_item_num; - struct item_head *de_ih; - int de_entry_num; - struct reiserfs_de_head *de_deh; - int de_entrylen; - int de_namelen; - char *de_name; - unsigned long *de_gen_number_bit_string; - - __u32 de_dir_id; - __u32 de_objectid; - - struct cpu_key de_entry_key; -}; - -/* these defines are useful when a particular member of a reiserfs_dir_entry is needed */ - -/* pointer to file name, stored in entry */ -#define B_I_DEH_ENTRY_FILE_NAME(bh,ih,deh) (B_I_PITEM (bh, ih) + deh_location(deh)) - -/* length of name */ -#define I_DEH_N_ENTRY_FILE_NAME_LENGTH(ih,deh,entry_num) \ -(I_DEH_N_ENTRY_LENGTH (ih, deh, entry_num) - (de_with_sd (deh) ? SD_SIZE : 0)) - -/* hash value occupies bits from 7 up to 30 */ -#define GET_HASH_VALUE(offset) ((offset) & 0x7fffff80LL) -/* generation number occupies 7 bits starting from 0 up to 6 */ -#define GET_GENERATION_NUMBER(offset) ((offset) & 0x7fLL) -#define MAX_GENERATION_NUMBER 127 - -#define SET_GENERATION_NUMBER(offset,gen_number) (GET_HASH_VALUE(offset)|(gen_number)) - -/* - * Picture represents an internal node of the reiserfs tree - * ______________________________________________________ - * | | Array of | Array of | Free | - * |block | keys | pointers | space | - * | head | N | N+1 | | - * |______|_______________|___________________|___________| - */ - -/***************************************************************************/ -/* DISK CHILD */ -/***************************************************************************/ -/* Disk child pointer: The pointer from an internal node of the tree - to a node that is on disk. */ -struct disk_child { - __le32 dc_block_number; /* Disk child's block number. */ - __le16 dc_size; /* Disk child's used space. */ - __le16 dc_reserved; -}; - -#define DC_SIZE (sizeof(struct disk_child)) -#define dc_block_number(dc_p) (le32_to_cpu((dc_p)->dc_block_number)) -#define dc_size(dc_p) (le16_to_cpu((dc_p)->dc_size)) -#define put_dc_block_number(dc_p, val) do { (dc_p)->dc_block_number = cpu_to_le32(val); } while(0) -#define put_dc_size(dc_p, val) do { (dc_p)->dc_size = cpu_to_le16(val); } while(0) - -/* Get disk child by buffer header and position in the tree node. */ -#define B_N_CHILD(bh, n_pos) ((struct disk_child *)\ -((bh)->b_data + BLKH_SIZE + B_NR_ITEMS(bh) * KEY_SIZE + DC_SIZE * (n_pos))) - -/* Get disk child number by buffer header and position in the tree node. */ -#define B_N_CHILD_NUM(bh, n_pos) (dc_block_number(B_N_CHILD(bh, n_pos))) -#define PUT_B_N_CHILD_NUM(bh, n_pos, val) \ - (put_dc_block_number(B_N_CHILD(bh, n_pos), val)) - - /* maximal value of field child_size in structure disk_child */ - /* child size is the combined size of all items and their headers */ -#define MAX_CHILD_SIZE(bh) ((int)( (bh)->b_size - BLKH_SIZE )) - -/* amount of used space in buffer (not including block head) */ -#define B_CHILD_SIZE(cur) (MAX_CHILD_SIZE(cur)-(B_FREE_SPACE(cur))) - -/* max and min number of keys in internal node */ -#define MAX_NR_KEY(bh) ( (MAX_CHILD_SIZE(bh)-DC_SIZE)/(KEY_SIZE+DC_SIZE) ) -#define MIN_NR_KEY(bh) (MAX_NR_KEY(bh)/2) - -/***************************************************************************/ -/* PATH STRUCTURES AND DEFINES */ -/***************************************************************************/ - -/* Search_by_key fills up the path from the root to the leaf as it descends the tree looking for the - key. It uses reiserfs_bread to try to find buffers in the cache given their block number. If it - does not find them in the cache it reads them from disk. For each node search_by_key finds using - reiserfs_bread it then uses bin_search to look through that node. bin_search will find the - position of the block_number of the next node if it is looking through an internal node. If it - is looking through a leaf node bin_search will find the position of the item which has key either - equal to given key, or which is the maximal key less than the given key. */ - -struct path_element { - struct buffer_head *pe_buffer; /* Pointer to the buffer at the path in the tree. */ - int pe_position; /* Position in the tree node which is placed in the */ - /* buffer above. */ -}; - -#define MAX_HEIGHT 5 /* maximal height of a tree. don't change this without changing JOURNAL_PER_BALANCE_CNT */ -#define EXTENDED_MAX_HEIGHT 7 /* Must be equals MAX_HEIGHT + FIRST_PATH_ELEMENT_OFFSET */ -#define FIRST_PATH_ELEMENT_OFFSET 2 /* Must be equal to at least 2. */ - -#define ILLEGAL_PATH_ELEMENT_OFFSET 1 /* Must be equal to FIRST_PATH_ELEMENT_OFFSET - 1 */ -#define MAX_FEB_SIZE 6 /* this MUST be MAX_HEIGHT + 1. See about FEB below */ - -/* We need to keep track of who the ancestors of nodes are. When we - perform a search we record which nodes were visited while - descending the tree looking for the node we searched for. This list - of nodes is called the path. This information is used while - performing balancing. Note that this path information may become - invalid, and this means we must check it when using it to see if it - is still valid. You'll need to read search_by_key and the comments - in it, especially about decrement_counters_in_path(), to understand - this structure. - -Paths make the code so much harder to work with and debug.... An -enormous number of bugs are due to them, and trying to write or modify -code that uses them just makes my head hurt. They are based on an -excessive effort to avoid disturbing the precious VFS code.:-( The -gods only know how we are going to SMP the code that uses them. -znodes are the way! */ - -#define PATH_READA 0x1 /* do read ahead */ -#define PATH_READA_BACK 0x2 /* read backwards */ - -struct treepath { - int path_length; /* Length of the array above. */ - int reada; - struct path_element path_elements[EXTENDED_MAX_HEIGHT]; /* Array of the path elements. */ - int pos_in_item; -}; - -#define pos_in_item(path) ((path)->pos_in_item) - -#define INITIALIZE_PATH(var) \ -struct treepath var = {.path_length = ILLEGAL_PATH_ELEMENT_OFFSET, .reada = 0,} - -/* Get path element by path and path position. */ -#define PATH_OFFSET_PELEMENT(path, n_offset) ((path)->path_elements + (n_offset)) - -/* Get buffer header at the path by path and path position. */ -#define PATH_OFFSET_PBUFFER(path, n_offset) (PATH_OFFSET_PELEMENT(path, n_offset)->pe_buffer) - -/* Get position in the element at the path by path and path position. */ -#define PATH_OFFSET_POSITION(path, n_offset) (PATH_OFFSET_PELEMENT(path, n_offset)->pe_position) - -#define PATH_PLAST_BUFFER(path) (PATH_OFFSET_PBUFFER((path), (path)->path_length)) - /* you know, to the person who didn't - write this the macro name does not - at first suggest what it does. - Maybe POSITION_FROM_PATH_END? Or - maybe we should just focus on - dumping paths... -Hans */ -#define PATH_LAST_POSITION(path) (PATH_OFFSET_POSITION((path), (path)->path_length)) - -#define PATH_PITEM_HEAD(path) B_N_PITEM_HEAD(PATH_PLAST_BUFFER(path), PATH_LAST_POSITION(path)) - -/* in do_balance leaf has h == 0 in contrast with path structure, - where root has level == 0. That is why we need these defines */ -#define PATH_H_PBUFFER(path, h) PATH_OFFSET_PBUFFER (path, path->path_length - (h)) /* tb->S[h] */ -#define PATH_H_PPARENT(path, h) PATH_H_PBUFFER (path, (h) + 1) /* tb->F[h] or tb->S[0]->b_parent */ -#define PATH_H_POSITION(path, h) PATH_OFFSET_POSITION (path, path->path_length - (h)) -#define PATH_H_B_ITEM_ORDER(path, h) PATH_H_POSITION(path, h + 1) /* tb->S[h]->b_item_order */ - -#define PATH_H_PATH_OFFSET(path, n_h) ((path)->path_length - (n_h)) - -#define get_last_bh(path) PATH_PLAST_BUFFER(path) -#define get_ih(path) PATH_PITEM_HEAD(path) -#define get_item_pos(path) PATH_LAST_POSITION(path) -#define get_item(path) ((void *)B_N_PITEM(PATH_PLAST_BUFFER(path), PATH_LAST_POSITION (path))) -#define item_moved(ih,path) comp_items(ih, path) -#define path_changed(ih,path) comp_items (ih, path) - -/***************************************************************************/ -/* MISC */ -/***************************************************************************/ - -/* Size of pointer to the unformatted node. */ -#define UNFM_P_SIZE (sizeof(unp_t)) -#define UNFM_P_SHIFT 2 - -// in in-core inode key is stored on le form -#define INODE_PKEY(inode) ((struct reiserfs_key *)(REISERFS_I(inode)->i_key)) - -#define MAX_UL_INT 0xffffffff -#define MAX_INT 0x7ffffff -#define MAX_US_INT 0xffff - -// reiserfs version 2 has max offset 60 bits. Version 1 - 32 bit offset -#define U32_MAX (~(__u32)0) - -static inline loff_t max_reiserfs_offset(struct inode *inode) -{ - if (get_inode_item_key_version(inode) == KEY_FORMAT_3_5) - return (loff_t) U32_MAX; - - return (loff_t) ((~(__u64) 0) >> 4); -} - -/*#define MAX_KEY_UNIQUENESS MAX_UL_INT*/ -#define MAX_KEY_OBJECTID MAX_UL_INT - -#define MAX_B_NUM MAX_UL_INT -#define MAX_FC_NUM MAX_US_INT - -/* the purpose is to detect overflow of an unsigned short */ -#define REISERFS_LINK_MAX (MAX_US_INT - 1000) - -/* The following defines are used in reiserfs_insert_item and reiserfs_append_item */ -#define REISERFS_KERNEL_MEM 0 /* reiserfs kernel memory mode */ -#define REISERFS_USER_MEM 1 /* reiserfs user memory mode */ - -#define fs_generation(s) (REISERFS_SB(s)->s_generation_counter) -#define get_generation(s) atomic_read (&fs_generation(s)) -#define FILESYSTEM_CHANGED_TB(tb) (get_generation((tb)->tb_sb) != (tb)->fs_gen) -#define __fs_changed(gen,s) (gen != get_generation (s)) -#define fs_changed(gen,s) \ -({ \ - reiserfs_cond_resched(s); \ - __fs_changed(gen, s); \ -}) - -/***************************************************************************/ -/* FIXATE NODES */ -/***************************************************************************/ - -#define VI_TYPE_LEFT_MERGEABLE 1 -#define VI_TYPE_RIGHT_MERGEABLE 2 - -/* To make any changes in the tree we always first find node, that - contains item to be changed/deleted or place to insert a new - item. We call this node S. To do balancing we need to decide what - we will shift to left/right neighbor, or to a new node, where new - item will be etc. To make this analysis simpler we build virtual - node. Virtual node is an array of items, that will replace items of - node S. (For instance if we are going to delete an item, virtual - node does not contain it). Virtual node keeps information about - item sizes and types, mergeability of first and last items, sizes - of all entries in directory item. We use this array of items when - calculating what we can shift to neighbors and how many nodes we - have to have if we do not any shiftings, if we shift to left/right - neighbor or to both. */ -struct virtual_item { - int vi_index; // index in the array of item operations - unsigned short vi_type; // left/right mergeability - unsigned short vi_item_len; /* length of item that it will have after balancing */ - struct item_head *vi_ih; - const char *vi_item; // body of item (old or new) - const void *vi_new_data; // 0 always but paste mode - void *vi_uarea; // item specific area -}; - -struct virtual_node { - char *vn_free_ptr; /* this is a pointer to the free space in the buffer */ - unsigned short vn_nr_item; /* number of items in virtual node */ - short vn_size; /* size of node , that node would have if it has unlimited size and no balancing is performed */ - short vn_mode; /* mode of balancing (paste, insert, delete, cut) */ - short vn_affected_item_num; - short vn_pos_in_item; - struct item_head *vn_ins_ih; /* item header of inserted item, 0 for other modes */ - const void *vn_data; - struct virtual_item *vn_vi; /* array of items (including a new one, excluding item to be deleted) */ -}; - -/* used by directory items when creating virtual nodes */ -struct direntry_uarea { - int flags; - __u16 entry_count; - __u16 entry_sizes[1]; -} __attribute__ ((__packed__)); - -/***************************************************************************/ -/* TREE BALANCE */ -/***************************************************************************/ - -/* This temporary structure is used in tree balance algorithms, and - constructed as we go to the extent that its various parts are - needed. It contains arrays of nodes that can potentially be - involved in the balancing of node S, and parameters that define how - each of the nodes must be balanced. Note that in these algorithms - for balancing the worst case is to need to balance the current node - S and the left and right neighbors and all of their parents plus - create a new node. We implement S1 balancing for the leaf nodes - and S0 balancing for the internal nodes (S1 and S0 are defined in - our papers.)*/ - -#define MAX_FREE_BLOCK 7 /* size of the array of buffers to free at end of do_balance */ - -/* maximum number of FEB blocknrs on a single level */ -#define MAX_AMOUNT_NEEDED 2 - -/* someday somebody will prefix every field in this struct with tb_ */ -struct tree_balance { - int tb_mode; - int need_balance_dirty; - struct super_block *tb_sb; - struct reiserfs_transaction_handle *transaction_handle; - struct treepath *tb_path; - struct buffer_head *L[MAX_HEIGHT]; /* array of left neighbors of nodes in the path */ - struct buffer_head *R[MAX_HEIGHT]; /* array of right neighbors of nodes in the path */ - struct buffer_head *FL[MAX_HEIGHT]; /* array of fathers of the left neighbors */ - struct buffer_head *FR[MAX_HEIGHT]; /* array of fathers of the right neighbors */ - struct buffer_head *CFL[MAX_HEIGHT]; /* array of common parents of center node and its left neighbor */ - struct buffer_head *CFR[MAX_HEIGHT]; /* array of common parents of center node and its right neighbor */ - - struct buffer_head *FEB[MAX_FEB_SIZE]; /* array of empty buffers. Number of buffers in array equals - cur_blknum. */ - struct buffer_head *used[MAX_FEB_SIZE]; - struct buffer_head *thrown[MAX_FEB_SIZE]; - int lnum[MAX_HEIGHT]; /* array of number of items which must be - shifted to the left in order to balance the - current node; for leaves includes item that - will be partially shifted; for internal - nodes, it is the number of child pointers - rather than items. It includes the new item - being created. The code sometimes subtracts - one to get the number of wholly shifted - items for other purposes. */ - int rnum[MAX_HEIGHT]; /* substitute right for left in comment above */ - int lkey[MAX_HEIGHT]; /* array indexed by height h mapping the key delimiting L[h] and - S[h] to its item number within the node CFL[h] */ - int rkey[MAX_HEIGHT]; /* substitute r for l in comment above */ - int insert_size[MAX_HEIGHT]; /* the number of bytes by we are trying to add or remove from - S[h]. A negative value means removing. */ - int blknum[MAX_HEIGHT]; /* number of nodes that will replace node S[h] after - balancing on the level h of the tree. If 0 then S is - being deleted, if 1 then S is remaining and no new nodes - are being created, if 2 or 3 then 1 or 2 new nodes is - being created */ - - /* fields that are used only for balancing leaves of the tree */ - int cur_blknum; /* number of empty blocks having been already allocated */ - int s0num; /* number of items that fall into left most node when S[0] splits */ - int s1num; /* number of items that fall into first new node when S[0] splits */ - int s2num; /* number of items that fall into second new node when S[0] splits */ - int lbytes; /* number of bytes which can flow to the left neighbor from the left */ - /* most liquid item that cannot be shifted from S[0] entirely */ - /* if -1 then nothing will be partially shifted */ - int rbytes; /* number of bytes which will flow to the right neighbor from the right */ - /* most liquid item that cannot be shifted from S[0] entirely */ - /* if -1 then nothing will be partially shifted */ - int s1bytes; /* number of bytes which flow to the first new node when S[0] splits */ - /* note: if S[0] splits into 3 nodes, then items do not need to be cut */ - int s2bytes; - struct buffer_head *buf_to_free[MAX_FREE_BLOCK]; /* buffers which are to be freed after do_balance finishes by unfix_nodes */ - char *vn_buf; /* kmalloced memory. Used to create - virtual node and keep map of - dirtied bitmap blocks */ - int vn_buf_size; /* size of the vn_buf */ - struct virtual_node *tb_vn; /* VN starts after bitmap of bitmap blocks */ - - int fs_gen; /* saved value of `reiserfs_generation' counter - see FILESYSTEM_CHANGED() macro in reiserfs_fs.h */ -#ifdef DISPLACE_NEW_PACKING_LOCALITIES - struct in_core_key key; /* key pointer, to pass to block allocator or - another low-level subsystem */ -#endif -}; - -/* These are modes of balancing */ - -/* When inserting an item. */ -#define M_INSERT 'i' -/* When inserting into (directories only) or appending onto an already - existent item. */ -#define M_PASTE 'p' -/* When deleting an item. */ -#define M_DELETE 'd' -/* When truncating an item or removing an entry from a (directory) item. */ -#define M_CUT 'c' - -/* used when balancing on leaf level skipped (in reiserfsck) */ -#define M_INTERNAL 'n' - -/* When further balancing is not needed, then do_balance does not need - to be called. */ -#define M_SKIP_BALANCING 's' -#define M_CONVERT 'v' - -/* modes of leaf_move_items */ -#define LEAF_FROM_S_TO_L 0 -#define LEAF_FROM_S_TO_R 1 -#define LEAF_FROM_R_TO_L 2 -#define LEAF_FROM_L_TO_R 3 -#define LEAF_FROM_S_TO_SNEW 4 - -#define FIRST_TO_LAST 0 -#define LAST_TO_FIRST 1 - -/* used in do_balance for passing parent of node information that has - been gotten from tb struct */ -struct buffer_info { - struct tree_balance *tb; - struct buffer_head *bi_bh; - struct buffer_head *bi_parent; - int bi_position; -}; - -static inline struct super_block *sb_from_tb(struct tree_balance *tb) -{ - return tb ? tb->tb_sb : NULL; -} - -static inline struct super_block *sb_from_bi(struct buffer_info *bi) -{ - return bi ? sb_from_tb(bi->tb) : NULL; -} - -/* there are 4 types of items: stat data, directory item, indirect, direct. -+-------------------+------------+--------------+------------+ -| | k_offset | k_uniqueness | mergeable? | -+-------------------+------------+--------------+------------+ -| stat data | 0 | 0 | no | -+-------------------+------------+--------------+------------+ -| 1st directory item| DOT_OFFSET |DIRENTRY_UNIQUENESS| no | -| non 1st directory | hash value | | yes | -| item | | | | -+-------------------+------------+--------------+------------+ -| indirect item | offset + 1 |TYPE_INDIRECT | if this is not the first indirect item of the object -+-------------------+------------+--------------+------------+ -| direct item | offset + 1 |TYPE_DIRECT | if not this is not the first direct item of the object -+-------------------+------------+--------------+------------+ -*/ - -struct item_operations { - int (*bytes_number) (struct item_head * ih, int block_size); - void (*decrement_key) (struct cpu_key *); - int (*is_left_mergeable) (struct reiserfs_key * ih, - unsigned long bsize); - void (*print_item) (struct item_head *, char *item); - void (*check_item) (struct item_head *, char *item); - - int (*create_vi) (struct virtual_node * vn, struct virtual_item * vi, - int is_affected, int insert_size); - int (*check_left) (struct virtual_item * vi, int free, - int start_skip, int end_skip); - int (*check_right) (struct virtual_item * vi, int free); - int (*part_size) (struct virtual_item * vi, int from, int to); - int (*unit_num) (struct virtual_item * vi); - void (*print_vi) (struct virtual_item * vi); -}; - -extern struct item_operations *item_ops[TYPE_ANY + 1]; - -#define op_bytes_number(ih,bsize) item_ops[le_ih_k_type (ih)]->bytes_number (ih, bsize) -#define op_is_left_mergeable(key,bsize) item_ops[le_key_k_type (le_key_version (key), key)]->is_left_mergeable (key, bsize) -#define op_print_item(ih,item) item_ops[le_ih_k_type (ih)]->print_item (ih, item) -#define op_check_item(ih,item) item_ops[le_ih_k_type (ih)]->check_item (ih, item) -#define op_create_vi(vn,vi,is_affected,insert_size) item_ops[le_ih_k_type ((vi)->vi_ih)]->create_vi (vn,vi,is_affected,insert_size) -#define op_check_left(vi,free,start_skip,end_skip) item_ops[(vi)->vi_index]->check_left (vi, free, start_skip, end_skip) -#define op_check_right(vi,free) item_ops[(vi)->vi_index]->check_right (vi, free) -#define op_part_size(vi,from,to) item_ops[(vi)->vi_index]->part_size (vi, from, to) -#define op_unit_num(vi) item_ops[(vi)->vi_index]->unit_num (vi) -#define op_print_vi(vi) item_ops[(vi)->vi_index]->print_vi (vi) - -#define COMP_SHORT_KEYS comp_short_keys - -/* number of blocks pointed to by the indirect item */ -#define I_UNFM_NUM(ih) (ih_item_len(ih) / UNFM_P_SIZE) - -/* the used space within the unformatted node corresponding to pos within the item pointed to by ih */ -#define I_POS_UNFM_SIZE(ih,pos,size) (((pos) == I_UNFM_NUM(ih) - 1 ) ? (size) - ih_free_space(ih) : (size)) - -/* number of bytes contained by the direct item or the unformatted nodes the indirect item points to */ - -/* get the item header */ -#define B_N_PITEM_HEAD(bh,item_num) ( (struct item_head * )((bh)->b_data + BLKH_SIZE) + (item_num) ) - -/* get key */ -#define B_N_PDELIM_KEY(bh,item_num) ( (struct reiserfs_key * )((bh)->b_data + BLKH_SIZE) + (item_num) ) - -/* get the key */ -#define B_N_PKEY(bh,item_num) ( &(B_N_PITEM_HEAD(bh,item_num)->ih_key) ) - -/* get item body */ -#define B_N_PITEM(bh,item_num) ( (bh)->b_data + ih_location(B_N_PITEM_HEAD((bh),(item_num)))) - -/* get the stat data by the buffer header and the item order */ -#define B_N_STAT_DATA(bh,nr) \ -( (struct stat_data *)((bh)->b_data + ih_location(B_N_PITEM_HEAD((bh),(nr))) ) ) - - /* following defines use reiserfs buffer header and item header */ - -/* get stat-data */ -#define B_I_STAT_DATA(bh, ih) ( (struct stat_data * )((bh)->b_data + ih_location(ih)) ) - -// this is 3976 for size==4096 -#define MAX_DIRECT_ITEM_LEN(size) ((size) - BLKH_SIZE - 2*IH_SIZE - SD_SIZE - UNFM_P_SIZE) - -/* indirect items consist of entries which contain blocknrs, pos - indicates which entry, and B_I_POS_UNFM_POINTER resolves to the - blocknr contained by the entry pos points to */ -#define B_I_POS_UNFM_POINTER(bh,ih,pos) le32_to_cpu(*(((unp_t *)B_I_PITEM(bh,ih)) + (pos))) -#define PUT_B_I_POS_UNFM_POINTER(bh,ih,pos, val) do {*(((unp_t *)B_I_PITEM(bh,ih)) + (pos)) = cpu_to_le32(val); } while (0) - -struct reiserfs_iget_args { - __u32 objectid; - __u32 dirid; -}; - -/***************************************************************************/ -/* FUNCTION DECLARATIONS */ -/***************************************************************************/ - -#define get_journal_desc_magic(bh) (bh->b_data + bh->b_size - 12) - -#define journal_trans_half(blocksize) \ - ((blocksize - sizeof (struct reiserfs_journal_desc) + sizeof (__u32) - 12) / sizeof (__u32)) - -/* journal.c see journal.c for all the comments here */ - -/* first block written in a commit. */ -struct reiserfs_journal_desc { - __le32 j_trans_id; /* id of commit */ - __le32 j_len; /* length of commit. len +1 is the commit block */ - __le32 j_mount_id; /* mount id of this trans */ - __le32 j_realblock[1]; /* real locations for each block */ -}; - -#define get_desc_trans_id(d) le32_to_cpu((d)->j_trans_id) -#define get_desc_trans_len(d) le32_to_cpu((d)->j_len) -#define get_desc_mount_id(d) le32_to_cpu((d)->j_mount_id) - -#define set_desc_trans_id(d,val) do { (d)->j_trans_id = cpu_to_le32 (val); } while (0) -#define set_desc_trans_len(d,val) do { (d)->j_len = cpu_to_le32 (val); } while (0) -#define set_desc_mount_id(d,val) do { (d)->j_mount_id = cpu_to_le32 (val); } while (0) - -/* last block written in a commit */ -struct reiserfs_journal_commit { - __le32 j_trans_id; /* must match j_trans_id from the desc block */ - __le32 j_len; /* ditto */ - __le32 j_realblock[1]; /* real locations for each block */ -}; - -#define get_commit_trans_id(c) le32_to_cpu((c)->j_trans_id) -#define get_commit_trans_len(c) le32_to_cpu((c)->j_len) -#define get_commit_mount_id(c) le32_to_cpu((c)->j_mount_id) - -#define set_commit_trans_id(c,val) do { (c)->j_trans_id = cpu_to_le32 (val); } while (0) -#define set_commit_trans_len(c,val) do { (c)->j_len = cpu_to_le32 (val); } while (0) - -/* this header block gets written whenever a transaction is considered fully flushed, and is more recent than the -** last fully flushed transaction. fully flushed means all the log blocks and all the real blocks are on disk, -** and this transaction does not need to be replayed. -*/ -struct reiserfs_journal_header { - __le32 j_last_flush_trans_id; /* id of last fully flushed transaction */ - __le32 j_first_unflushed_offset; /* offset in the log of where to start replay after a crash */ - __le32 j_mount_id; - /* 12 */ struct journal_params jh_journal; -}; - -/* biggest tunable defines are right here */ -#define JOURNAL_BLOCK_COUNT 8192 /* number of blocks in the journal */ -#define JOURNAL_TRANS_MAX_DEFAULT 1024 /* biggest possible single transaction, don't change for now (8/3/99) */ -#define JOURNAL_TRANS_MIN_DEFAULT 256 -#define JOURNAL_MAX_BATCH_DEFAULT 900 /* max blocks to batch into one transaction, don't make this any bigger than 900 */ -#define JOURNAL_MIN_RATIO 2 -#define JOURNAL_MAX_COMMIT_AGE 30 -#define JOURNAL_MAX_TRANS_AGE 30 -#define JOURNAL_PER_BALANCE_CNT (3 * (MAX_HEIGHT-2) + 9) -#define JOURNAL_BLOCKS_PER_OBJECT(sb) (JOURNAL_PER_BALANCE_CNT * 3 + \ - 2 * (REISERFS_QUOTA_INIT_BLOCKS(sb) + \ - REISERFS_QUOTA_TRANS_BLOCKS(sb))) - -#ifdef CONFIG_QUOTA -#define REISERFS_QUOTA_OPTS ((1 << REISERFS_USRQUOTA) | (1 << REISERFS_GRPQUOTA)) -/* We need to update data and inode (atime) */ -#define REISERFS_QUOTA_TRANS_BLOCKS(s) (REISERFS_SB(s)->s_mount_opt & REISERFS_QUOTA_OPTS ? 2 : 0) -/* 1 balancing, 1 bitmap, 1 data per write + stat data update */ -#define REISERFS_QUOTA_INIT_BLOCKS(s) (REISERFS_SB(s)->s_mount_opt & REISERFS_QUOTA_OPTS ? \ -(DQUOT_INIT_ALLOC*(JOURNAL_PER_BALANCE_CNT+2)+DQUOT_INIT_REWRITE+1) : 0) -/* same as with INIT */ -#define REISERFS_QUOTA_DEL_BLOCKS(s) (REISERFS_SB(s)->s_mount_opt & REISERFS_QUOTA_OPTS ? \ -(DQUOT_DEL_ALLOC*(JOURNAL_PER_BALANCE_CNT+2)+DQUOT_DEL_REWRITE+1) : 0) -#else -#define REISERFS_QUOTA_TRANS_BLOCKS(s) 0 -#define REISERFS_QUOTA_INIT_BLOCKS(s) 0 -#define REISERFS_QUOTA_DEL_BLOCKS(s) 0 -#endif - -/* both of these can be as low as 1, or as high as you want. The min is the -** number of 4k bitmap nodes preallocated on mount. New nodes are allocated -** as needed, and released when transactions are committed. On release, if -** the current number of nodes is > max, the node is freed, otherwise, -** it is put on a free list for faster use later. -*/ -#define REISERFS_MIN_BITMAP_NODES 10 -#define REISERFS_MAX_BITMAP_NODES 100 - -#define JBH_HASH_SHIFT 13 /* these are based on journal hash size of 8192 */ -#define JBH_HASH_MASK 8191 - -#define _jhashfn(sb,block) \ - (((unsigned long)sb>>L1_CACHE_SHIFT) ^ \ - (((block)<<(JBH_HASH_SHIFT - 6)) ^ ((block) >> 13) ^ ((block) << (JBH_HASH_SHIFT - 12)))) -#define journal_hash(t,sb,block) ((t)[_jhashfn((sb),(block)) & JBH_HASH_MASK]) - -// We need these to make journal.c code more readable -#define journal_find_get_block(s, block) __find_get_block(SB_JOURNAL(s)->j_dev_bd, block, s->s_blocksize) -#define journal_getblk(s, block) __getblk(SB_JOURNAL(s)->j_dev_bd, block, s->s_blocksize) -#define journal_bread(s, block) __bread(SB_JOURNAL(s)->j_dev_bd, block, s->s_blocksize) - -enum reiserfs_bh_state_bits { - BH_JDirty = BH_PrivateStart, /* buffer is in current transaction */ - BH_JDirty_wait, - BH_JNew, /* disk block was taken off free list before - * being in a finished transaction, or - * written to disk. Can be reused immed. */ - BH_JPrepared, - BH_JRestore_dirty, - BH_JTest, // debugging only will go away -}; - -BUFFER_FNS(JDirty, journaled); -TAS_BUFFER_FNS(JDirty, journaled); -BUFFER_FNS(JDirty_wait, journal_dirty); -TAS_BUFFER_FNS(JDirty_wait, journal_dirty); -BUFFER_FNS(JNew, journal_new); -TAS_BUFFER_FNS(JNew, journal_new); -BUFFER_FNS(JPrepared, journal_prepared); -TAS_BUFFER_FNS(JPrepared, journal_prepared); -BUFFER_FNS(JRestore_dirty, journal_restore_dirty); -TAS_BUFFER_FNS(JRestore_dirty, journal_restore_dirty); -BUFFER_FNS(JTest, journal_test); -TAS_BUFFER_FNS(JTest, journal_test); - -/* -** transaction handle which is passed around for all journal calls -*/ -struct reiserfs_transaction_handle { - struct super_block *t_super; /* super for this FS when journal_begin was - called. saves calls to reiserfs_get_super - also used by nested transactions to make - sure they are nesting on the right FS - _must_ be first in the handle - */ - int t_refcount; - int t_blocks_logged; /* number of blocks this writer has logged */ - int t_blocks_allocated; /* number of blocks this writer allocated */ - unsigned int t_trans_id; /* sanity check, equals the current trans id */ - void *t_handle_save; /* save existing current->journal_info */ - unsigned displace_new_blocks:1; /* if new block allocation occurres, that block - should be displaced from others */ - struct list_head t_list; -}; - -/* used to keep track of ordered and tail writes, attached to the buffer - * head through b_journal_head. - */ -struct reiserfs_jh { - struct reiserfs_journal_list *jl; - struct buffer_head *bh; - struct list_head list; -}; - -void reiserfs_free_jh(struct buffer_head *bh); -int reiserfs_add_tail_list(struct inode *inode, struct buffer_head *bh); -int reiserfs_add_ordered_list(struct inode *inode, struct buffer_head *bh); -int journal_mark_dirty(struct reiserfs_transaction_handle *, - struct super_block *, struct buffer_head *bh); - -static inline int reiserfs_file_data_log(struct inode *inode) -{ - if (reiserfs_data_log(inode->i_sb) || - (REISERFS_I(inode)->i_flags & i_data_log)) - return 1; - return 0; -} - -static inline int reiserfs_transaction_running(struct super_block *s) -{ - struct reiserfs_transaction_handle *th = current->journal_info; - if (th && th->t_super == s) - return 1; - if (th && th->t_super == NULL) - BUG(); - return 0; -} - -static inline int reiserfs_transaction_free_space(struct reiserfs_transaction_handle *th) -{ - return th->t_blocks_allocated - th->t_blocks_logged; -} - -struct reiserfs_transaction_handle *reiserfs_persistent_transaction(struct - super_block - *, - int count); -int reiserfs_end_persistent_transaction(struct reiserfs_transaction_handle *); -int reiserfs_commit_page(struct inode *inode, struct page *page, - unsigned from, unsigned to); -int reiserfs_flush_old_commits(struct super_block *); -int reiserfs_commit_for_inode(struct inode *); -int reiserfs_inode_needs_commit(struct inode *); -void reiserfs_update_inode_transaction(struct inode *); -void reiserfs_wait_on_write_block(struct super_block *s); -void reiserfs_block_writes(struct reiserfs_transaction_handle *th); -void reiserfs_allow_writes(struct super_block *s); -void reiserfs_check_lock_depth(struct super_block *s, char *caller); -int reiserfs_prepare_for_journal(struct super_block *, struct buffer_head *bh, - int wait); -void reiserfs_restore_prepared_buffer(struct super_block *, - struct buffer_head *bh); -int journal_init(struct super_block *, const char *j_dev_name, int old_format, - unsigned int); -int journal_release(struct reiserfs_transaction_handle *, struct super_block *); -int journal_release_error(struct reiserfs_transaction_handle *, - struct super_block *); -int journal_end(struct reiserfs_transaction_handle *, struct super_block *, - unsigned long); -int journal_end_sync(struct reiserfs_transaction_handle *, struct super_block *, - unsigned long); -int journal_mark_freed(struct reiserfs_transaction_handle *, - struct super_block *, b_blocknr_t blocknr); -int journal_transaction_should_end(struct reiserfs_transaction_handle *, int); -int reiserfs_in_journal(struct super_block *sb, unsigned int bmap_nr, - int bit_nr, int searchall, b_blocknr_t *next); -int journal_begin(struct reiserfs_transaction_handle *, - struct super_block *sb, unsigned long); -int journal_join_abort(struct reiserfs_transaction_handle *, - struct super_block *sb, unsigned long); -void reiserfs_abort_journal(struct super_block *sb, int errno); -void reiserfs_abort(struct super_block *sb, int errno, const char *fmt, ...); -int reiserfs_allocate_list_bitmaps(struct super_block *s, - struct reiserfs_list_bitmap *, unsigned int); - -void add_save_link(struct reiserfs_transaction_handle *th, - struct inode *inode, int truncate); -int remove_save_link(struct inode *inode, int truncate); - -/* objectid.c */ -__u32 reiserfs_get_unused_objectid(struct reiserfs_transaction_handle *th); -void reiserfs_release_objectid(struct reiserfs_transaction_handle *th, - __u32 objectid_to_release); -int reiserfs_convert_objectid_map_v1(struct super_block *); - -/* stree.c */ -int B_IS_IN_TREE(const struct buffer_head *); -extern void copy_item_head(struct item_head *to, - const struct item_head *from); - -// first key is in cpu form, second - le -extern int comp_short_keys(const struct reiserfs_key *le_key, - const struct cpu_key *cpu_key); -extern void le_key2cpu_key(struct cpu_key *to, const struct reiserfs_key *from); - -// both are in le form -extern int comp_le_keys(const struct reiserfs_key *, - const struct reiserfs_key *); -extern int comp_short_le_keys(const struct reiserfs_key *, - const struct reiserfs_key *); - -// -// get key version from on disk key - kludge -// -static inline int le_key_version(const struct reiserfs_key *key) -{ - int type; - - type = offset_v2_k_type(&(key->u.k_offset_v2)); - if (type != TYPE_DIRECT && type != TYPE_INDIRECT - && type != TYPE_DIRENTRY) - return KEY_FORMAT_3_5; - - return KEY_FORMAT_3_6; - -} - -static inline void copy_key(struct reiserfs_key *to, - const struct reiserfs_key *from) -{ - memcpy(to, from, KEY_SIZE); -} - -int comp_items(const struct item_head *stored_ih, const struct treepath *path); -const struct reiserfs_key *get_rkey(const struct treepath *chk_path, - const struct super_block *sb); -int search_by_key(struct super_block *, const struct cpu_key *, - struct treepath *, int); -#define search_item(s,key,path) search_by_key (s, key, path, DISK_LEAF_NODE_LEVEL) -int search_for_position_by_key(struct super_block *sb, - const struct cpu_key *cpu_key, - struct treepath *search_path); -extern void decrement_bcount(struct buffer_head *bh); -void decrement_counters_in_path(struct treepath *search_path); -void pathrelse(struct treepath *search_path); -int reiserfs_check_path(struct treepath *p); -void pathrelse_and_restore(struct super_block *s, struct treepath *search_path); - -int reiserfs_insert_item(struct reiserfs_transaction_handle *th, - struct treepath *path, - const struct cpu_key *key, - struct item_head *ih, - struct inode *inode, const char *body); - -int reiserfs_paste_into_item(struct reiserfs_transaction_handle *th, - struct treepath *path, - const struct cpu_key *key, - struct inode *inode, - const char *body, int paste_size); - -int reiserfs_cut_from_item(struct reiserfs_transaction_handle *th, - struct treepath *path, - struct cpu_key *key, - struct inode *inode, - struct page *page, loff_t new_file_size); - -int reiserfs_delete_item(struct reiserfs_transaction_handle *th, - struct treepath *path, - const struct cpu_key *key, - struct inode *inode, struct buffer_head *un_bh); - -void reiserfs_delete_solid_item(struct reiserfs_transaction_handle *th, - struct inode *inode, struct reiserfs_key *key); -int reiserfs_delete_object(struct reiserfs_transaction_handle *th, - struct inode *inode); -int reiserfs_do_truncate(struct reiserfs_transaction_handle *th, - struct inode *inode, struct page *, - int update_timestamps); - -#define i_block_size(inode) ((inode)->i_sb->s_blocksize) -#define file_size(inode) ((inode)->i_size) -#define tail_size(inode) (file_size (inode) & (i_block_size (inode) - 1)) - -#define tail_has_to_be_packed(inode) (have_large_tails ((inode)->i_sb)?\ -!STORE_TAIL_IN_UNFM_S1(file_size (inode), tail_size(inode), inode->i_sb->s_blocksize):have_small_tails ((inode)->i_sb)?!STORE_TAIL_IN_UNFM_S2(file_size (inode), tail_size(inode), inode->i_sb->s_blocksize):0 ) - -void padd_item(char *item, int total_length, int length); - -/* inode.c */ -/* args for the create parameter of reiserfs_get_block */ -#define GET_BLOCK_NO_CREATE 0 /* don't create new blocks or convert tails */ -#define GET_BLOCK_CREATE 1 /* add anything you need to find block */ -#define GET_BLOCK_NO_HOLE 2 /* return -ENOENT for file holes */ -#define GET_BLOCK_READ_DIRECT 4 /* read the tail if indirect item not found */ -#define GET_BLOCK_NO_IMUX 8 /* i_mutex is not held, don't preallocate */ -#define GET_BLOCK_NO_DANGLE 16 /* don't leave any transactions running */ - -void reiserfs_read_locked_inode(struct inode *inode, - struct reiserfs_iget_args *args); -int reiserfs_find_actor(struct inode *inode, void *p); -int reiserfs_init_locked_inode(struct inode *inode, void *p); -void reiserfs_evict_inode(struct inode *inode); -int reiserfs_write_inode(struct inode *inode, struct writeback_control *wbc); -int reiserfs_get_block(struct inode *inode, sector_t block, - struct buffer_head *bh_result, int create); -struct dentry *reiserfs_fh_to_dentry(struct super_block *sb, struct fid *fid, - int fh_len, int fh_type); -struct dentry *reiserfs_fh_to_parent(struct super_block *sb, struct fid *fid, - int fh_len, int fh_type); -int reiserfs_encode_fh(struct dentry *dentry, __u32 * data, int *lenp, - int connectable); - -int reiserfs_truncate_file(struct inode *, int update_timestamps); -void make_cpu_key(struct cpu_key *cpu_key, struct inode *inode, loff_t offset, - int type, int key_length); -void make_le_item_head(struct item_head *ih, const struct cpu_key *key, - int version, - loff_t offset, int type, int length, int entry_count); -struct inode *reiserfs_iget(struct super_block *s, const struct cpu_key *key); - -struct reiserfs_security_handle; -int reiserfs_new_inode(struct reiserfs_transaction_handle *th, - struct inode *dir, umode_t mode, - const char *symname, loff_t i_size, - struct dentry *dentry, struct inode *inode, - struct reiserfs_security_handle *security); - -void reiserfs_update_sd_size(struct reiserfs_transaction_handle *th, - struct inode *inode, loff_t size); - -static inline void reiserfs_update_sd(struct reiserfs_transaction_handle *th, - struct inode *inode) -{ - reiserfs_update_sd_size(th, inode, inode->i_size); -} - -void sd_attrs_to_i_attrs(__u16 sd_attrs, struct inode *inode); -void i_attrs_to_sd_attrs(struct inode *inode, __u16 * sd_attrs); -int reiserfs_setattr(struct dentry *dentry, struct iattr *attr); - -int __reiserfs_write_begin(struct page *page, unsigned from, unsigned len); - -/* namei.c */ -void set_de_name_and_namelen(struct reiserfs_dir_entry *de); -int search_by_entry_key(struct super_block *sb, const struct cpu_key *key, - struct treepath *path, struct reiserfs_dir_entry *de); -struct dentry *reiserfs_get_parent(struct dentry *); - -#ifdef CONFIG_REISERFS_PROC_INFO -int reiserfs_proc_info_init(struct super_block *sb); -int reiserfs_proc_info_done(struct super_block *sb); -int reiserfs_proc_info_global_init(void); -int reiserfs_proc_info_global_done(void); - -#define PROC_EXP( e ) e - -#define __PINFO( sb ) REISERFS_SB(sb) -> s_proc_info_data -#define PROC_INFO_MAX( sb, field, value ) \ - __PINFO( sb ).field = \ - max( REISERFS_SB( sb ) -> s_proc_info_data.field, value ) -#define PROC_INFO_INC( sb, field ) ( ++ ( __PINFO( sb ).field ) ) -#define PROC_INFO_ADD( sb, field, val ) ( __PINFO( sb ).field += ( val ) ) -#define PROC_INFO_BH_STAT( sb, bh, level ) \ - PROC_INFO_INC( sb, sbk_read_at[ ( level ) ] ); \ - PROC_INFO_ADD( sb, free_at[ ( level ) ], B_FREE_SPACE( bh ) ); \ - PROC_INFO_ADD( sb, items_at[ ( level ) ], B_NR_ITEMS( bh ) ) -#else -static inline int reiserfs_proc_info_init(struct super_block *sb) -{ - return 0; -} - -static inline int reiserfs_proc_info_done(struct super_block *sb) -{ - return 0; -} - -static inline int reiserfs_proc_info_global_init(void) -{ - return 0; -} - -static inline int reiserfs_proc_info_global_done(void) -{ - return 0; -} - -#define PROC_EXP( e ) -#define VOID_V ( ( void ) 0 ) -#define PROC_INFO_MAX( sb, field, value ) VOID_V -#define PROC_INFO_INC( sb, field ) VOID_V -#define PROC_INFO_ADD( sb, field, val ) VOID_V -#define PROC_INFO_BH_STAT(sb, bh, n_node_level) VOID_V -#endif - -/* dir.c */ -extern const struct inode_operations reiserfs_dir_inode_operations; -extern const struct inode_operations reiserfs_symlink_inode_operations; -extern const struct inode_operations reiserfs_special_inode_operations; -extern const struct file_operations reiserfs_dir_operations; -int reiserfs_readdir_dentry(struct dentry *, void *, filldir_t, loff_t *); - -/* tail_conversion.c */ -int direct2indirect(struct reiserfs_transaction_handle *, struct inode *, - struct treepath *, struct buffer_head *, loff_t); -int indirect2direct(struct reiserfs_transaction_handle *, struct inode *, - struct page *, struct treepath *, const struct cpu_key *, - loff_t, char *); -void reiserfs_unmap_buffer(struct buffer_head *); - -/* file.c */ -extern const struct inode_operations reiserfs_file_inode_operations; -extern const struct file_operations reiserfs_file_operations; -extern const struct address_space_operations reiserfs_address_space_operations; - -/* fix_nodes.c */ - -int fix_nodes(int n_op_mode, struct tree_balance *tb, - struct item_head *ins_ih, const void *); -void unfix_nodes(struct tree_balance *); - -/* prints.c */ -void __reiserfs_panic(struct super_block *s, const char *id, - const char *function, const char *fmt, ...) - __attribute__ ((noreturn)); -#define reiserfs_panic(s, id, fmt, args...) \ - __reiserfs_panic(s, id, __func__, fmt, ##args) -void __reiserfs_error(struct super_block *s, const char *id, - const char *function, const char *fmt, ...); -#define reiserfs_error(s, id, fmt, args...) \ - __reiserfs_error(s, id, __func__, fmt, ##args) -void reiserfs_info(struct super_block *s, const char *fmt, ...); -void reiserfs_debug(struct super_block *s, int level, const char *fmt, ...); -void print_indirect_item(struct buffer_head *bh, int item_num); -void store_print_tb(struct tree_balance *tb); -void print_cur_tb(char *mes); -void print_de(struct reiserfs_dir_entry *de); -void print_bi(struct buffer_info *bi, char *mes); -#define PRINT_LEAF_ITEMS 1 /* print all items */ -#define PRINT_DIRECTORY_ITEMS 2 /* print directory items */ -#define PRINT_DIRECT_ITEMS 4 /* print contents of direct items */ -void print_block(struct buffer_head *bh, ...); -void print_bmap(struct super_block *s, int silent); -void print_bmap_block(int i, char *data, int size, int silent); -/*void print_super_block (struct super_block * s, char * mes);*/ -void print_objectid_map(struct super_block *s); -void print_block_head(struct buffer_head *bh, char *mes); -void check_leaf(struct buffer_head *bh); -void check_internal(struct buffer_head *bh); -void print_statistics(struct super_block *s); -char *reiserfs_hashname(int code); - -/* lbalance.c */ -int leaf_move_items(int shift_mode, struct tree_balance *tb, int mov_num, - int mov_bytes, struct buffer_head *Snew); -int leaf_shift_left(struct tree_balance *tb, int shift_num, int shift_bytes); -int leaf_shift_right(struct tree_balance *tb, int shift_num, int shift_bytes); -void leaf_delete_items(struct buffer_info *cur_bi, int last_first, int first, - int del_num, int del_bytes); -void leaf_insert_into_buf(struct buffer_info *bi, int before, - struct item_head *inserted_item_ih, - const char *inserted_item_body, int zeros_number); -void leaf_paste_in_buffer(struct buffer_info *bi, int pasted_item_num, - int pos_in_item, int paste_size, const char *body, - int zeros_number); -void leaf_cut_from_buffer(struct buffer_info *bi, int cut_item_num, - int pos_in_item, int cut_size); -void leaf_paste_entries(struct buffer_info *bi, int item_num, int before, - int new_entry_count, struct reiserfs_de_head *new_dehs, - const char *records, int paste_size); -/* ibalance.c */ -int balance_internal(struct tree_balance *, int, int, struct item_head *, - struct buffer_head **); - -/* do_balance.c */ -void do_balance_mark_leaf_dirty(struct tree_balance *tb, - struct buffer_head *bh, int flag); -#define do_balance_mark_internal_dirty do_balance_mark_leaf_dirty -#define do_balance_mark_sb_dirty do_balance_mark_leaf_dirty - -void do_balance(struct tree_balance *tb, struct item_head *ih, - const char *body, int flag); -void reiserfs_invalidate_buffer(struct tree_balance *tb, - struct buffer_head *bh); - -int get_left_neighbor_position(struct tree_balance *tb, int h); -int get_right_neighbor_position(struct tree_balance *tb, int h); -void replace_key(struct tree_balance *tb, struct buffer_head *, int, - struct buffer_head *, int); -void make_empty_node(struct buffer_info *); -struct buffer_head *get_FEB(struct tree_balance *); - -/* bitmap.c */ - -/* structure contains hints for block allocator, and it is a container for - * arguments, such as node, search path, transaction_handle, etc. */ -struct __reiserfs_blocknr_hint { - struct inode *inode; /* inode passed to allocator, if we allocate unf. nodes */ - sector_t block; /* file offset, in blocks */ - struct in_core_key key; - struct treepath *path; /* search path, used by allocator to deternine search_start by - * various ways */ - struct reiserfs_transaction_handle *th; /* transaction handle is needed to log super blocks and - * bitmap blocks changes */ - b_blocknr_t beg, end; - b_blocknr_t search_start; /* a field used to transfer search start value (block number) - * between different block allocator procedures - * (determine_search_start() and others) */ - int prealloc_size; /* is set in determine_prealloc_size() function, used by underlayed - * function that do actual allocation */ - - unsigned formatted_node:1; /* the allocator uses different polices for getting disk space for - * formatted/unformatted blocks with/without preallocation */ - unsigned preallocate:1; -}; - -typedef struct __reiserfs_blocknr_hint reiserfs_blocknr_hint_t; - -int reiserfs_parse_alloc_options(struct super_block *, char *); -void reiserfs_init_alloc_options(struct super_block *s); - -/* - * given a directory, this will tell you what packing locality - * to use for a new object underneat it. The locality is returned - * in disk byte order (le). - */ -__le32 reiserfs_choose_packing(struct inode *dir); - -int reiserfs_init_bitmap_cache(struct super_block *sb); -void reiserfs_free_bitmap_cache(struct super_block *sb); -void reiserfs_cache_bitmap_metadata(struct super_block *sb, struct buffer_head *bh, struct reiserfs_bitmap_info *info); -struct buffer_head *reiserfs_read_bitmap_block(struct super_block *sb, unsigned int bitmap); -int is_reusable(struct super_block *s, b_blocknr_t block, int bit_value); -void reiserfs_free_block(struct reiserfs_transaction_handle *th, struct inode *, - b_blocknr_t, int for_unformatted); -int reiserfs_allocate_blocknrs(reiserfs_blocknr_hint_t *, b_blocknr_t *, int, - int); -static inline int reiserfs_new_form_blocknrs(struct tree_balance *tb, - b_blocknr_t * new_blocknrs, - int amount_needed) -{ - reiserfs_blocknr_hint_t hint = { - .th = tb->transaction_handle, - .path = tb->tb_path, - .inode = NULL, - .key = tb->key, - .block = 0, - .formatted_node = 1 - }; - return reiserfs_allocate_blocknrs(&hint, new_blocknrs, amount_needed, - 0); -} - -static inline int reiserfs_new_unf_blocknrs(struct reiserfs_transaction_handle - *th, struct inode *inode, - b_blocknr_t * new_blocknrs, - struct treepath *path, - sector_t block) -{ - reiserfs_blocknr_hint_t hint = { - .th = th, - .path = path, - .inode = inode, - .block = block, - .formatted_node = 0, - .preallocate = 0 - }; - return reiserfs_allocate_blocknrs(&hint, new_blocknrs, 1, 0); -} - -#ifdef REISERFS_PREALLOCATE -static inline int reiserfs_new_unf_blocknrs2(struct reiserfs_transaction_handle - *th, struct inode *inode, - b_blocknr_t * new_blocknrs, - struct treepath *path, - sector_t block) -{ - reiserfs_blocknr_hint_t hint = { - .th = th, - .path = path, - .inode = inode, - .block = block, - .formatted_node = 0, - .preallocate = 1 - }; - return reiserfs_allocate_blocknrs(&hint, new_blocknrs, 1, 0); -} - -void reiserfs_discard_prealloc(struct reiserfs_transaction_handle *th, - struct inode *inode); -void reiserfs_discard_all_prealloc(struct reiserfs_transaction_handle *th); -#endif - -/* hashes.c */ -__u32 keyed_hash(const signed char *msg, int len); -__u32 yura_hash(const signed char *msg, int len); -__u32 r5_hash(const signed char *msg, int len); - -#define reiserfs_set_le_bit __set_bit_le -#define reiserfs_test_and_set_le_bit __test_and_set_bit_le -#define reiserfs_clear_le_bit __clear_bit_le -#define reiserfs_test_and_clear_le_bit __test_and_clear_bit_le -#define reiserfs_test_le_bit test_bit_le -#define reiserfs_find_next_zero_le_bit find_next_zero_bit_le - -/* sometimes reiserfs_truncate may require to allocate few new blocks - to perform indirect2direct conversion. People probably used to - think, that truncate should work without problems on a filesystem - without free disk space. They may complain that they can not - truncate due to lack of free disk space. This spare space allows us - to not worry about it. 500 is probably too much, but it should be - absolutely safe */ -#define SPARE_SPACE 500 - -/* prototypes from ioctl.c */ -long reiserfs_ioctl(struct file *filp, unsigned int cmd, unsigned long arg); -long reiserfs_compat_ioctl(struct file *filp, - unsigned int cmd, unsigned long arg); -int reiserfs_unpack(struct inode *inode, struct file *filp); - -#endif /* __KERNEL__ */ - #endif /* _LINUX_REISER_FS_H */ From 765fd6b23d8b1130bd9e2c995040b137faa0992d Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 17 Mar 2012 01:19:24 -0400 Subject: [PATCH 34/64] kill reiserfs_fs_{i,sb}.h Signed-off-by: Al Viro --- fs/reiserfs/reiserfs.h | 599 ++++++++++++++++++++++++++++++++- include/linux/reiserfs_fs_i.h | 63 ---- include/linux/reiserfs_fs_sb.h | 554 ------------------------------ 3 files changed, 597 insertions(+), 619 deletions(-) delete mode 100644 include/linux/reiserfs_fs_i.h delete mode 100644 include/linux/reiserfs_fs_sb.h diff --git a/fs/reiserfs/reiserfs.h b/fs/reiserfs/reiserfs.h index b3865c84f54c..445d768eea44 100644 --- a/fs/reiserfs/reiserfs.h +++ b/fs/reiserfs/reiserfs.h @@ -12,8 +12,6 @@ #include #include #include -#include -#include /* the 32 bit compat definitions with int argument */ #define REISERFS_IOC32_UNPACK _IOW(0xCD, 1, int) @@ -22,6 +20,603 @@ #define REISERFS_IOC32_GETVERSION FS_IOC32_GETVERSION #define REISERFS_IOC32_SETVERSION FS_IOC32_SETVERSION +struct reiserfs_journal_list; + +/** bitmasks for i_flags field in reiserfs-specific part of inode */ +typedef enum { + /** this says what format of key do all items (but stat data) of + an object have. If this is set, that format is 3.6 otherwise + - 3.5 */ + i_item_key_version_mask = 0x0001, + /** If this is unset, object has 3.5 stat data, otherwise, it has + 3.6 stat data with 64bit size, 32bit nlink etc. */ + i_stat_data_version_mask = 0x0002, + /** file might need tail packing on close */ + i_pack_on_close_mask = 0x0004, + /** don't pack tail of file */ + i_nopack_mask = 0x0008, + /** If those is set, "safe link" was created for this file during + truncate or unlink. Safe link is used to avoid leakage of disk + space on crash with some files open, but unlinked. */ + i_link_saved_unlink_mask = 0x0010, + i_link_saved_truncate_mask = 0x0020, + i_has_xattr_dir = 0x0040, + i_data_log = 0x0080, +} reiserfs_inode_flags; + +struct reiserfs_inode_info { + __u32 i_key[4]; /* key is still 4 32 bit integers */ + /** transient inode flags that are never stored on disk. Bitmasks + for this field are defined above. */ + __u32 i_flags; + + __u32 i_first_direct_byte; // offset of first byte stored in direct item. + + /* copy of persistent inode flags read from sd_attrs. */ + __u32 i_attrs; + + int i_prealloc_block; /* first unused block of a sequence of unused blocks */ + int i_prealloc_count; /* length of that sequence */ + struct list_head i_prealloc_list; /* per-transaction list of inodes which + * have preallocated blocks */ + + unsigned new_packing_locality:1; /* new_packig_locality is created; new blocks + * for the contents of this directory should be + * displaced */ + + /* we use these for fsync or O_SYNC to decide which transaction + ** needs to be committed in order for this inode to be properly + ** flushed */ + unsigned int i_trans_id; + struct reiserfs_journal_list *i_jl; + atomic_t openers; + struct mutex tailpack; +#ifdef CONFIG_REISERFS_FS_XATTR + struct rw_semaphore i_xattr_sem; +#endif + struct inode vfs_inode; +}; + +typedef enum { + reiserfs_attrs_cleared = 0x00000001, +} reiserfs_super_block_flags; + +/* struct reiserfs_super_block accessors/mutators + * since this is a disk structure, it will always be in + * little endian format. */ +#define sb_block_count(sbp) (le32_to_cpu((sbp)->s_v1.s_block_count)) +#define set_sb_block_count(sbp,v) ((sbp)->s_v1.s_block_count = cpu_to_le32(v)) +#define sb_free_blocks(sbp) (le32_to_cpu((sbp)->s_v1.s_free_blocks)) +#define set_sb_free_blocks(sbp,v) ((sbp)->s_v1.s_free_blocks = cpu_to_le32(v)) +#define sb_root_block(sbp) (le32_to_cpu((sbp)->s_v1.s_root_block)) +#define set_sb_root_block(sbp,v) ((sbp)->s_v1.s_root_block = cpu_to_le32(v)) + +#define sb_jp_journal_1st_block(sbp) \ + (le32_to_cpu((sbp)->s_v1.s_journal.jp_journal_1st_block)) +#define set_sb_jp_journal_1st_block(sbp,v) \ + ((sbp)->s_v1.s_journal.jp_journal_1st_block = cpu_to_le32(v)) +#define sb_jp_journal_dev(sbp) \ + (le32_to_cpu((sbp)->s_v1.s_journal.jp_journal_dev)) +#define set_sb_jp_journal_dev(sbp,v) \ + ((sbp)->s_v1.s_journal.jp_journal_dev = cpu_to_le32(v)) +#define sb_jp_journal_size(sbp) \ + (le32_to_cpu((sbp)->s_v1.s_journal.jp_journal_size)) +#define set_sb_jp_journal_size(sbp,v) \ + ((sbp)->s_v1.s_journal.jp_journal_size = cpu_to_le32(v)) +#define sb_jp_journal_trans_max(sbp) \ + (le32_to_cpu((sbp)->s_v1.s_journal.jp_journal_trans_max)) +#define set_sb_jp_journal_trans_max(sbp,v) \ + ((sbp)->s_v1.s_journal.jp_journal_trans_max = cpu_to_le32(v)) +#define sb_jp_journal_magic(sbp) \ + (le32_to_cpu((sbp)->s_v1.s_journal.jp_journal_magic)) +#define set_sb_jp_journal_magic(sbp,v) \ + ((sbp)->s_v1.s_journal.jp_journal_magic = cpu_to_le32(v)) +#define sb_jp_journal_max_batch(sbp) \ + (le32_to_cpu((sbp)->s_v1.s_journal.jp_journal_max_batch)) +#define set_sb_jp_journal_max_batch(sbp,v) \ + ((sbp)->s_v1.s_journal.jp_journal_max_batch = cpu_to_le32(v)) +#define sb_jp_jourmal_max_commit_age(sbp) \ + (le32_to_cpu((sbp)->s_v1.s_journal.jp_journal_max_commit_age)) +#define set_sb_jp_journal_max_commit_age(sbp,v) \ + ((sbp)->s_v1.s_journal.jp_journal_max_commit_age = cpu_to_le32(v)) + +#define sb_blocksize(sbp) (le16_to_cpu((sbp)->s_v1.s_blocksize)) +#define set_sb_blocksize(sbp,v) ((sbp)->s_v1.s_blocksize = cpu_to_le16(v)) +#define sb_oid_maxsize(sbp) (le16_to_cpu((sbp)->s_v1.s_oid_maxsize)) +#define set_sb_oid_maxsize(sbp,v) ((sbp)->s_v1.s_oid_maxsize = cpu_to_le16(v)) +#define sb_oid_cursize(sbp) (le16_to_cpu((sbp)->s_v1.s_oid_cursize)) +#define set_sb_oid_cursize(sbp,v) ((sbp)->s_v1.s_oid_cursize = cpu_to_le16(v)) +#define sb_umount_state(sbp) (le16_to_cpu((sbp)->s_v1.s_umount_state)) +#define set_sb_umount_state(sbp,v) ((sbp)->s_v1.s_umount_state = cpu_to_le16(v)) +#define sb_fs_state(sbp) (le16_to_cpu((sbp)->s_v1.s_fs_state)) +#define set_sb_fs_state(sbp,v) ((sbp)->s_v1.s_fs_state = cpu_to_le16(v)) +#define sb_hash_function_code(sbp) \ + (le32_to_cpu((sbp)->s_v1.s_hash_function_code)) +#define set_sb_hash_function_code(sbp,v) \ + ((sbp)->s_v1.s_hash_function_code = cpu_to_le32(v)) +#define sb_tree_height(sbp) (le16_to_cpu((sbp)->s_v1.s_tree_height)) +#define set_sb_tree_height(sbp,v) ((sbp)->s_v1.s_tree_height = cpu_to_le16(v)) +#define sb_bmap_nr(sbp) (le16_to_cpu((sbp)->s_v1.s_bmap_nr)) +#define set_sb_bmap_nr(sbp,v) ((sbp)->s_v1.s_bmap_nr = cpu_to_le16(v)) +#define sb_version(sbp) (le16_to_cpu((sbp)->s_v1.s_version)) +#define set_sb_version(sbp,v) ((sbp)->s_v1.s_version = cpu_to_le16(v)) + +#define sb_mnt_count(sbp) (le16_to_cpu((sbp)->s_mnt_count)) +#define set_sb_mnt_count(sbp, v) ((sbp)->s_mnt_count = cpu_to_le16(v)) + +#define sb_reserved_for_journal(sbp) \ + (le16_to_cpu((sbp)->s_v1.s_reserved_for_journal)) +#define set_sb_reserved_for_journal(sbp,v) \ + ((sbp)->s_v1.s_reserved_for_journal = cpu_to_le16(v)) + +/* LOGGING -- */ + +/* These all interelate for performance. +** +** If the journal block count is smaller than n transactions, you lose speed. +** I don't know what n is yet, I'm guessing 8-16. +** +** typical transaction size depends on the application, how often fsync is +** called, and how many metadata blocks you dirty in a 30 second period. +** The more small files (<16k) you use, the larger your transactions will +** be. +** +** If your journal fills faster than dirty buffers get flushed to disk, it must flush them before allowing the journal +** to wrap, which slows things down. If you need high speed meta data updates, the journal should be big enough +** to prevent wrapping before dirty meta blocks get to disk. +** +** If the batch max is smaller than the transaction max, you'll waste space at the end of the journal +** because journal_end sets the next transaction to start at 0 if the next transaction has any chance of wrapping. +** +** The large the batch max age, the better the speed, and the more meta data changes you'll lose after a crash. +** +*/ + +/* don't mess with these for a while */ + /* we have a node size define somewhere in reiserfs_fs.h. -Hans */ +#define JOURNAL_BLOCK_SIZE 4096 /* BUG gotta get rid of this */ +#define JOURNAL_MAX_CNODE 1500 /* max cnodes to allocate. */ +#define JOURNAL_HASH_SIZE 8192 +#define JOURNAL_NUM_BITMAPS 5 /* number of copies of the bitmaps to have floating. Must be >= 2 */ + +/* One of these for every block in every transaction +** Each one is in two hash tables. First, a hash of the current transaction, and after journal_end, a +** hash of all the in memory transactions. +** next and prev are used by the current transaction (journal_hash). +** hnext and hprev are used by journal_list_hash. If a block is in more than one transaction, the journal_list_hash +** links it in multiple times. This allows flush_journal_list to remove just the cnode belonging +** to a given transaction. +*/ +struct reiserfs_journal_cnode { + struct buffer_head *bh; /* real buffer head */ + struct super_block *sb; /* dev of real buffer head */ + __u32 blocknr; /* block number of real buffer head, == 0 when buffer on disk */ + unsigned long state; + struct reiserfs_journal_list *jlist; /* journal list this cnode lives in */ + struct reiserfs_journal_cnode *next; /* next in transaction list */ + struct reiserfs_journal_cnode *prev; /* prev in transaction list */ + struct reiserfs_journal_cnode *hprev; /* prev in hash list */ + struct reiserfs_journal_cnode *hnext; /* next in hash list */ +}; + +struct reiserfs_bitmap_node { + int id; + char *data; + struct list_head list; +}; + +struct reiserfs_list_bitmap { + struct reiserfs_journal_list *journal_list; + struct reiserfs_bitmap_node **bitmaps; +}; + +/* +** one of these for each transaction. The most important part here is the j_realblock. +** this list of cnodes is used to hash all the blocks in all the commits, to mark all the +** real buffer heads dirty once all the commits hit the disk, +** and to make sure every real block in a transaction is on disk before allowing the log area +** to be overwritten */ +struct reiserfs_journal_list { + unsigned long j_start; + unsigned long j_state; + unsigned long j_len; + atomic_t j_nonzerolen; + atomic_t j_commit_left; + atomic_t j_older_commits_done; /* all commits older than this on disk */ + struct mutex j_commit_mutex; + unsigned int j_trans_id; + time_t j_timestamp; + struct reiserfs_list_bitmap *j_list_bitmap; + struct buffer_head *j_commit_bh; /* commit buffer head */ + struct reiserfs_journal_cnode *j_realblock; + struct reiserfs_journal_cnode *j_freedlist; /* list of buffers that were freed during this trans. free each of these on flush */ + /* time ordered list of all active transactions */ + struct list_head j_list; + + /* time ordered list of all transactions we haven't tried to flush yet */ + struct list_head j_working_list; + + /* list of tail conversion targets in need of flush before commit */ + struct list_head j_tail_bh_list; + /* list of data=ordered buffers in need of flush before commit */ + struct list_head j_bh_list; + int j_refcount; +}; + +struct reiserfs_journal { + struct buffer_head **j_ap_blocks; /* journal blocks on disk */ + struct reiserfs_journal_cnode *j_last; /* newest journal block */ + struct reiserfs_journal_cnode *j_first; /* oldest journal block. start here for traverse */ + + struct block_device *j_dev_bd; + fmode_t j_dev_mode; + int j_1st_reserved_block; /* first block on s_dev of reserved area journal */ + + unsigned long j_state; + unsigned int j_trans_id; + unsigned long j_mount_id; + unsigned long j_start; /* start of current waiting commit (index into j_ap_blocks) */ + unsigned long j_len; /* length of current waiting commit */ + unsigned long j_len_alloc; /* number of buffers requested by journal_begin() */ + atomic_t j_wcount; /* count of writers for current commit */ + unsigned long j_bcount; /* batch count. allows turning X transactions into 1 */ + unsigned long j_first_unflushed_offset; /* first unflushed transactions offset */ + unsigned j_last_flush_trans_id; /* last fully flushed journal timestamp */ + struct buffer_head *j_header_bh; + + time_t j_trans_start_time; /* time this transaction started */ + struct mutex j_mutex; + struct mutex j_flush_mutex; + wait_queue_head_t j_join_wait; /* wait for current transaction to finish before starting new one */ + atomic_t j_jlock; /* lock for j_join_wait */ + int j_list_bitmap_index; /* number of next list bitmap to use */ + int j_must_wait; /* no more journal begins allowed. MUST sleep on j_join_wait */ + int j_next_full_flush; /* next journal_end will flush all journal list */ + int j_next_async_flush; /* next journal_end will flush all async commits */ + + int j_cnode_used; /* number of cnodes on the used list */ + int j_cnode_free; /* number of cnodes on the free list */ + + unsigned int j_trans_max; /* max number of blocks in a transaction. */ + unsigned int j_max_batch; /* max number of blocks to batch into a trans */ + unsigned int j_max_commit_age; /* in seconds, how old can an async commit be */ + unsigned int j_max_trans_age; /* in seconds, how old can a transaction be */ + unsigned int j_default_max_commit_age; /* the default for the max commit age */ + + struct reiserfs_journal_cnode *j_cnode_free_list; + struct reiserfs_journal_cnode *j_cnode_free_orig; /* orig pointer returned from vmalloc */ + + struct reiserfs_journal_list *j_current_jl; + int j_free_bitmap_nodes; + int j_used_bitmap_nodes; + + int j_num_lists; /* total number of active transactions */ + int j_num_work_lists; /* number that need attention from kreiserfsd */ + + /* debugging to make sure things are flushed in order */ + unsigned int j_last_flush_id; + + /* debugging to make sure things are committed in order */ + unsigned int j_last_commit_id; + + struct list_head j_bitmap_nodes; + struct list_head j_dirty_buffers; + spinlock_t j_dirty_buffers_lock; /* protects j_dirty_buffers */ + + /* list of all active transactions */ + struct list_head j_journal_list; + /* lists that haven't been touched by writeback attempts */ + struct list_head j_working_list; + + struct reiserfs_list_bitmap j_list_bitmap[JOURNAL_NUM_BITMAPS]; /* array of bitmaps to record the deleted blocks */ + struct reiserfs_journal_cnode *j_hash_table[JOURNAL_HASH_SIZE]; /* hash table for real buffer heads in current trans */ + struct reiserfs_journal_cnode *j_list_hash_table[JOURNAL_HASH_SIZE]; /* hash table for all the real buffer heads in all + the transactions */ + struct list_head j_prealloc_list; /* list of inodes which have preallocated blocks */ + int j_persistent_trans; + unsigned long j_max_trans_size; + unsigned long j_max_batch_size; + + int j_errno; + + /* when flushing ordered buffers, throttle new ordered writers */ + struct delayed_work j_work; + struct super_block *j_work_sb; + atomic_t j_async_throttle; +}; + +enum journal_state_bits { + J_WRITERS_BLOCKED = 1, /* set when new writers not allowed */ + J_WRITERS_QUEUED, /* set when log is full due to too many writers */ + J_ABORTED, /* set when log is aborted */ +}; + +#define JOURNAL_DESC_MAGIC "ReIsErLB" /* ick. magic string to find desc blocks in the journal */ + +typedef __u32(*hashf_t) (const signed char *, int); + +struct reiserfs_bitmap_info { + __u32 free_count; +}; + +struct proc_dir_entry; + +#if defined( CONFIG_PROC_FS ) && defined( CONFIG_REISERFS_PROC_INFO ) +typedef unsigned long int stat_cnt_t; +typedef struct reiserfs_proc_info_data { + spinlock_t lock; + int exiting; + int max_hash_collisions; + + stat_cnt_t breads; + stat_cnt_t bread_miss; + stat_cnt_t search_by_key; + stat_cnt_t search_by_key_fs_changed; + stat_cnt_t search_by_key_restarted; + + stat_cnt_t insert_item_restarted; + stat_cnt_t paste_into_item_restarted; + stat_cnt_t cut_from_item_restarted; + stat_cnt_t delete_solid_item_restarted; + stat_cnt_t delete_item_restarted; + + stat_cnt_t leaked_oid; + stat_cnt_t leaves_removable; + + /* balances per level. Use explicit 5 as MAX_HEIGHT is not visible yet. */ + stat_cnt_t balance_at[5]; /* XXX */ + /* sbk == search_by_key */ + stat_cnt_t sbk_read_at[5]; /* XXX */ + stat_cnt_t sbk_fs_changed[5]; + stat_cnt_t sbk_restarted[5]; + stat_cnt_t items_at[5]; /* XXX */ + stat_cnt_t free_at[5]; /* XXX */ + stat_cnt_t can_node_be_removed[5]; /* XXX */ + long int lnum[5]; /* XXX */ + long int rnum[5]; /* XXX */ + long int lbytes[5]; /* XXX */ + long int rbytes[5]; /* XXX */ + stat_cnt_t get_neighbors[5]; + stat_cnt_t get_neighbors_restart[5]; + stat_cnt_t need_l_neighbor[5]; + stat_cnt_t need_r_neighbor[5]; + + stat_cnt_t free_block; + struct __scan_bitmap_stats { + stat_cnt_t call; + stat_cnt_t wait; + stat_cnt_t bmap; + stat_cnt_t retry; + stat_cnt_t in_journal_hint; + stat_cnt_t in_journal_nohint; + stat_cnt_t stolen; + } scan_bitmap; + struct __journal_stats { + stat_cnt_t in_journal; + stat_cnt_t in_journal_bitmap; + stat_cnt_t in_journal_reusable; + stat_cnt_t lock_journal; + stat_cnt_t lock_journal_wait; + stat_cnt_t journal_being; + stat_cnt_t journal_relock_writers; + stat_cnt_t journal_relock_wcount; + stat_cnt_t mark_dirty; + stat_cnt_t mark_dirty_already; + stat_cnt_t mark_dirty_notjournal; + stat_cnt_t restore_prepared; + stat_cnt_t prepare; + stat_cnt_t prepare_retry; + } journal; +} reiserfs_proc_info_data_t; +#else +typedef struct reiserfs_proc_info_data { +} reiserfs_proc_info_data_t; +#endif + +/* reiserfs union of in-core super block data */ +struct reiserfs_sb_info { + struct buffer_head *s_sbh; /* Buffer containing the super block */ + /* both the comment and the choice of + name are unclear for s_rs -Hans */ + struct reiserfs_super_block *s_rs; /* Pointer to the super block in the buffer */ + struct reiserfs_bitmap_info *s_ap_bitmap; + struct reiserfs_journal *s_journal; /* pointer to journal information */ + unsigned short s_mount_state; /* reiserfs state (valid, invalid) */ + + /* Serialize writers access, replace the old bkl */ + struct mutex lock; + /* Owner of the lock (can be recursive) */ + struct task_struct *lock_owner; + /* Depth of the lock, start from -1 like the bkl */ + int lock_depth; + + /* Comment? -Hans */ + void (*end_io_handler) (struct buffer_head *, int); + hashf_t s_hash_function; /* pointer to function which is used + to sort names in directory. Set on + mount */ + unsigned long s_mount_opt; /* reiserfs's mount options are set + here (currently - NOTAIL, NOLOG, + REPLAYONLY) */ + + struct { /* This is a structure that describes block allocator options */ + unsigned long bits; /* Bitfield for enable/disable kind of options */ + unsigned long large_file_size; /* size started from which we consider file to be a large one(in blocks) */ + int border; /* percentage of disk, border takes */ + int preallocmin; /* Minimal file size (in blocks) starting from which we do preallocations */ + int preallocsize; /* Number of blocks we try to prealloc when file + reaches preallocmin size (in blocks) or + prealloc_list is empty. */ + } s_alloc_options; + + /* Comment? -Hans */ + wait_queue_head_t s_wait; + /* To be obsoleted soon by per buffer seals.. -Hans */ + atomic_t s_generation_counter; // increased by one every time the + // tree gets re-balanced + unsigned long s_properties; /* File system properties. Currently holds + on-disk FS format */ + + /* session statistics */ + int s_disk_reads; + int s_disk_writes; + int s_fix_nodes; + int s_do_balance; + int s_unneeded_left_neighbor; + int s_good_search_by_key_reada; + int s_bmaps; + int s_bmaps_without_search; + int s_direct2indirect; + int s_indirect2direct; + /* set up when it's ok for reiserfs_read_inode2() to read from + disk inode with nlink==0. Currently this is only used during + finish_unfinished() processing at mount time */ + int s_is_unlinked_ok; + reiserfs_proc_info_data_t s_proc_info_data; + struct proc_dir_entry *procdir; + int reserved_blocks; /* amount of blocks reserved for further allocations */ + spinlock_t bitmap_lock; /* this lock on now only used to protect reserved_blocks variable */ + struct dentry *priv_root; /* root of /.reiserfs_priv */ + struct dentry *xattr_root; /* root of /.reiserfs_priv/xattrs */ + int j_errno; +#ifdef CONFIG_QUOTA + char *s_qf_names[MAXQUOTAS]; + int s_jquota_fmt; +#endif + char *s_jdev; /* Stored jdev for mount option showing */ +#ifdef CONFIG_REISERFS_CHECK + + struct tree_balance *cur_tb; /* + * Detects whether more than one + * copy of tb exists per superblock + * as a means of checking whether + * do_balance is executing concurrently + * against another tree reader/writer + * on a same mount point. + */ +#endif +}; + +/* Definitions of reiserfs on-disk properties: */ +#define REISERFS_3_5 0 +#define REISERFS_3_6 1 +#define REISERFS_OLD_FORMAT 2 + +enum reiserfs_mount_options { +/* Mount options */ + REISERFS_LARGETAIL, /* large tails will be created in a session */ + REISERFS_SMALLTAIL, /* small (for files less than block size) tails will be created in a session */ + REPLAYONLY, /* replay journal and return 0. Use by fsck */ + REISERFS_CONVERT, /* -o conv: causes conversion of old + format super block to the new + format. If not specified - old + partition will be dealt with in a + manner of 3.5.x */ + +/* -o hash={tea, rupasov, r5, detect} is meant for properly mounting +** reiserfs disks from 3.5.19 or earlier. 99% of the time, this option +** is not required. If the normal autodection code can't determine which +** hash to use (because both hashes had the same value for a file) +** use this option to force a specific hash. It won't allow you to override +** the existing hash on the FS, so if you have a tea hash disk, and mount +** with -o hash=rupasov, the mount will fail. +*/ + FORCE_TEA_HASH, /* try to force tea hash on mount */ + FORCE_RUPASOV_HASH, /* try to force rupasov hash on mount */ + FORCE_R5_HASH, /* try to force rupasov hash on mount */ + FORCE_HASH_DETECT, /* try to detect hash function on mount */ + + REISERFS_DATA_LOG, + REISERFS_DATA_ORDERED, + REISERFS_DATA_WRITEBACK, + +/* used for testing experimental features, makes benchmarking new + features with and without more convenient, should never be used by + users in any code shipped to users (ideally) */ + + REISERFS_NO_BORDER, + REISERFS_NO_UNHASHED_RELOCATION, + REISERFS_HASHED_RELOCATION, + REISERFS_ATTRS, + REISERFS_XATTRS_USER, + REISERFS_POSIXACL, + REISERFS_EXPOSE_PRIVROOT, + REISERFS_BARRIER_NONE, + REISERFS_BARRIER_FLUSH, + + /* Actions on error */ + REISERFS_ERROR_PANIC, + REISERFS_ERROR_RO, + REISERFS_ERROR_CONTINUE, + + REISERFS_USRQUOTA, /* User quota option specified */ + REISERFS_GRPQUOTA, /* Group quota option specified */ + + REISERFS_TEST1, + REISERFS_TEST2, + REISERFS_TEST3, + REISERFS_TEST4, + REISERFS_UNSUPPORTED_OPT, +}; + +#define reiserfs_r5_hash(s) (REISERFS_SB(s)->s_mount_opt & (1 << FORCE_R5_HASH)) +#define reiserfs_rupasov_hash(s) (REISERFS_SB(s)->s_mount_opt & (1 << FORCE_RUPASOV_HASH)) +#define reiserfs_tea_hash(s) (REISERFS_SB(s)->s_mount_opt & (1 << FORCE_TEA_HASH)) +#define reiserfs_hash_detect(s) (REISERFS_SB(s)->s_mount_opt & (1 << FORCE_HASH_DETECT)) +#define reiserfs_no_border(s) (REISERFS_SB(s)->s_mount_opt & (1 << REISERFS_NO_BORDER)) +#define reiserfs_no_unhashed_relocation(s) (REISERFS_SB(s)->s_mount_opt & (1 << REISERFS_NO_UNHASHED_RELOCATION)) +#define reiserfs_hashed_relocation(s) (REISERFS_SB(s)->s_mount_opt & (1 << REISERFS_HASHED_RELOCATION)) +#define reiserfs_test4(s) (REISERFS_SB(s)->s_mount_opt & (1 << REISERFS_TEST4)) + +#define have_large_tails(s) (REISERFS_SB(s)->s_mount_opt & (1 << REISERFS_LARGETAIL)) +#define have_small_tails(s) (REISERFS_SB(s)->s_mount_opt & (1 << REISERFS_SMALLTAIL)) +#define replay_only(s) (REISERFS_SB(s)->s_mount_opt & (1 << REPLAYONLY)) +#define reiserfs_attrs(s) (REISERFS_SB(s)->s_mount_opt & (1 << REISERFS_ATTRS)) +#define old_format_only(s) (REISERFS_SB(s)->s_properties & (1 << REISERFS_3_5)) +#define convert_reiserfs(s) (REISERFS_SB(s)->s_mount_opt & (1 << REISERFS_CONVERT)) +#define reiserfs_data_log(s) (REISERFS_SB(s)->s_mount_opt & (1 << REISERFS_DATA_LOG)) +#define reiserfs_data_ordered(s) (REISERFS_SB(s)->s_mount_opt & (1 << REISERFS_DATA_ORDERED)) +#define reiserfs_data_writeback(s) (REISERFS_SB(s)->s_mount_opt & (1 << REISERFS_DATA_WRITEBACK)) +#define reiserfs_xattrs_user(s) (REISERFS_SB(s)->s_mount_opt & (1 << REISERFS_XATTRS_USER)) +#define reiserfs_posixacl(s) (REISERFS_SB(s)->s_mount_opt & (1 << REISERFS_POSIXACL)) +#define reiserfs_expose_privroot(s) (REISERFS_SB(s)->s_mount_opt & (1 << REISERFS_EXPOSE_PRIVROOT)) +#define reiserfs_xattrs_optional(s) (reiserfs_xattrs_user(s) || reiserfs_posixacl(s)) +#define reiserfs_barrier_none(s) (REISERFS_SB(s)->s_mount_opt & (1 << REISERFS_BARRIER_NONE)) +#define reiserfs_barrier_flush(s) (REISERFS_SB(s)->s_mount_opt & (1 << REISERFS_BARRIER_FLUSH)) + +#define reiserfs_error_panic(s) (REISERFS_SB(s)->s_mount_opt & (1 << REISERFS_ERROR_PANIC)) +#define reiserfs_error_ro(s) (REISERFS_SB(s)->s_mount_opt & (1 << REISERFS_ERROR_RO)) + +void reiserfs_file_buffer(struct buffer_head *bh, int list); +extern struct file_system_type reiserfs_fs_type; +int reiserfs_resize(struct super_block *, unsigned long); + +#define CARRY_ON 0 +#define SCHEDULE_OCCURRED 1 + +#define SB_BUFFER_WITH_SB(s) (REISERFS_SB(s)->s_sbh) +#define SB_JOURNAL(s) (REISERFS_SB(s)->s_journal) +#define SB_JOURNAL_1st_RESERVED_BLOCK(s) (SB_JOURNAL(s)->j_1st_reserved_block) +#define SB_JOURNAL_LEN_FREE(s) (SB_JOURNAL(s)->j_journal_len_free) +#define SB_AP_BITMAP(s) (REISERFS_SB(s)->s_ap_bitmap) + +#define SB_DISK_JOURNAL_HEAD(s) (SB_JOURNAL(s)->j_header_bh->) + +/* A safe version of the "bdevname", which returns the "s_id" field of + * a superblock or else "Null superblock" if the super block is NULL. + */ +static inline char *reiserfs_bdevname(struct super_block *s) +{ + return (s == NULL) ? "Null superblock" : s->s_id; +} + +#define reiserfs_is_journal_aborted(journal) (unlikely (__reiserfs_is_journal_aborted (journal))) +static inline int __reiserfs_is_journal_aborted(struct reiserfs_journal + *journal) +{ + return test_bit(J_ABORTED, &journal->j_state); +} + /* * Locking primitives. The write lock is a per superblock * special mutex that has properties close to the Big Kernel Lock diff --git a/include/linux/reiserfs_fs_i.h b/include/linux/reiserfs_fs_i.h deleted file mode 100644 index 97959bdfe214..000000000000 --- a/include/linux/reiserfs_fs_i.h +++ /dev/null @@ -1,63 +0,0 @@ -#ifndef _REISER_FS_I -#define _REISER_FS_I - -#include - -struct reiserfs_journal_list; - -/** bitmasks for i_flags field in reiserfs-specific part of inode */ -typedef enum { - /** this says what format of key do all items (but stat data) of - an object have. If this is set, that format is 3.6 otherwise - - 3.5 */ - i_item_key_version_mask = 0x0001, - /** If this is unset, object has 3.5 stat data, otherwise, it has - 3.6 stat data with 64bit size, 32bit nlink etc. */ - i_stat_data_version_mask = 0x0002, - /** file might need tail packing on close */ - i_pack_on_close_mask = 0x0004, - /** don't pack tail of file */ - i_nopack_mask = 0x0008, - /** If those is set, "safe link" was created for this file during - truncate or unlink. Safe link is used to avoid leakage of disk - space on crash with some files open, but unlinked. */ - i_link_saved_unlink_mask = 0x0010, - i_link_saved_truncate_mask = 0x0020, - i_has_xattr_dir = 0x0040, - i_data_log = 0x0080, -} reiserfs_inode_flags; - -struct reiserfs_inode_info { - __u32 i_key[4]; /* key is still 4 32 bit integers */ - /** transient inode flags that are never stored on disk. Bitmasks - for this field are defined above. */ - __u32 i_flags; - - __u32 i_first_direct_byte; // offset of first byte stored in direct item. - - /* copy of persistent inode flags read from sd_attrs. */ - __u32 i_attrs; - - int i_prealloc_block; /* first unused block of a sequence of unused blocks */ - int i_prealloc_count; /* length of that sequence */ - struct list_head i_prealloc_list; /* per-transaction list of inodes which - * have preallocated blocks */ - - unsigned new_packing_locality:1; /* new_packig_locality is created; new blocks - * for the contents of this directory should be - * displaced */ - - /* we use these for fsync or O_SYNC to decide which transaction - ** needs to be committed in order for this inode to be properly - ** flushed */ - unsigned int i_trans_id; - struct reiserfs_journal_list *i_jl; - atomic_t openers; - struct mutex tailpack; -#ifdef CONFIG_REISERFS_FS_XATTR - struct rw_semaphore i_xattr_sem; -#endif - struct inode vfs_inode; -}; - -#endif diff --git a/include/linux/reiserfs_fs_sb.h b/include/linux/reiserfs_fs_sb.h deleted file mode 100644 index 8c9e85c64b46..000000000000 --- a/include/linux/reiserfs_fs_sb.h +++ /dev/null @@ -1,554 +0,0 @@ -/* Copyright 1996-2000 Hans Reiser, see reiserfs/README for licensing - * and copyright details */ - -#ifndef _LINUX_REISER_FS_SB -#define _LINUX_REISER_FS_SB - -#ifdef __KERNEL__ -#include -#include -#include -#include -#endif - -typedef enum { - reiserfs_attrs_cleared = 0x00000001, -} reiserfs_super_block_flags; - -/* struct reiserfs_super_block accessors/mutators - * since this is a disk structure, it will always be in - * little endian format. */ -#define sb_block_count(sbp) (le32_to_cpu((sbp)->s_v1.s_block_count)) -#define set_sb_block_count(sbp,v) ((sbp)->s_v1.s_block_count = cpu_to_le32(v)) -#define sb_free_blocks(sbp) (le32_to_cpu((sbp)->s_v1.s_free_blocks)) -#define set_sb_free_blocks(sbp,v) ((sbp)->s_v1.s_free_blocks = cpu_to_le32(v)) -#define sb_root_block(sbp) (le32_to_cpu((sbp)->s_v1.s_root_block)) -#define set_sb_root_block(sbp,v) ((sbp)->s_v1.s_root_block = cpu_to_le32(v)) - -#define sb_jp_journal_1st_block(sbp) \ - (le32_to_cpu((sbp)->s_v1.s_journal.jp_journal_1st_block)) -#define set_sb_jp_journal_1st_block(sbp,v) \ - ((sbp)->s_v1.s_journal.jp_journal_1st_block = cpu_to_le32(v)) -#define sb_jp_journal_dev(sbp) \ - (le32_to_cpu((sbp)->s_v1.s_journal.jp_journal_dev)) -#define set_sb_jp_journal_dev(sbp,v) \ - ((sbp)->s_v1.s_journal.jp_journal_dev = cpu_to_le32(v)) -#define sb_jp_journal_size(sbp) \ - (le32_to_cpu((sbp)->s_v1.s_journal.jp_journal_size)) -#define set_sb_jp_journal_size(sbp,v) \ - ((sbp)->s_v1.s_journal.jp_journal_size = cpu_to_le32(v)) -#define sb_jp_journal_trans_max(sbp) \ - (le32_to_cpu((sbp)->s_v1.s_journal.jp_journal_trans_max)) -#define set_sb_jp_journal_trans_max(sbp,v) \ - ((sbp)->s_v1.s_journal.jp_journal_trans_max = cpu_to_le32(v)) -#define sb_jp_journal_magic(sbp) \ - (le32_to_cpu((sbp)->s_v1.s_journal.jp_journal_magic)) -#define set_sb_jp_journal_magic(sbp,v) \ - ((sbp)->s_v1.s_journal.jp_journal_magic = cpu_to_le32(v)) -#define sb_jp_journal_max_batch(sbp) \ - (le32_to_cpu((sbp)->s_v1.s_journal.jp_journal_max_batch)) -#define set_sb_jp_journal_max_batch(sbp,v) \ - ((sbp)->s_v1.s_journal.jp_journal_max_batch = cpu_to_le32(v)) -#define sb_jp_jourmal_max_commit_age(sbp) \ - (le32_to_cpu((sbp)->s_v1.s_journal.jp_journal_max_commit_age)) -#define set_sb_jp_journal_max_commit_age(sbp,v) \ - ((sbp)->s_v1.s_journal.jp_journal_max_commit_age = cpu_to_le32(v)) - -#define sb_blocksize(sbp) (le16_to_cpu((sbp)->s_v1.s_blocksize)) -#define set_sb_blocksize(sbp,v) ((sbp)->s_v1.s_blocksize = cpu_to_le16(v)) -#define sb_oid_maxsize(sbp) (le16_to_cpu((sbp)->s_v1.s_oid_maxsize)) -#define set_sb_oid_maxsize(sbp,v) ((sbp)->s_v1.s_oid_maxsize = cpu_to_le16(v)) -#define sb_oid_cursize(sbp) (le16_to_cpu((sbp)->s_v1.s_oid_cursize)) -#define set_sb_oid_cursize(sbp,v) ((sbp)->s_v1.s_oid_cursize = cpu_to_le16(v)) -#define sb_umount_state(sbp) (le16_to_cpu((sbp)->s_v1.s_umount_state)) -#define set_sb_umount_state(sbp,v) ((sbp)->s_v1.s_umount_state = cpu_to_le16(v)) -#define sb_fs_state(sbp) (le16_to_cpu((sbp)->s_v1.s_fs_state)) -#define set_sb_fs_state(sbp,v) ((sbp)->s_v1.s_fs_state = cpu_to_le16(v)) -#define sb_hash_function_code(sbp) \ - (le32_to_cpu((sbp)->s_v1.s_hash_function_code)) -#define set_sb_hash_function_code(sbp,v) \ - ((sbp)->s_v1.s_hash_function_code = cpu_to_le32(v)) -#define sb_tree_height(sbp) (le16_to_cpu((sbp)->s_v1.s_tree_height)) -#define set_sb_tree_height(sbp,v) ((sbp)->s_v1.s_tree_height = cpu_to_le16(v)) -#define sb_bmap_nr(sbp) (le16_to_cpu((sbp)->s_v1.s_bmap_nr)) -#define set_sb_bmap_nr(sbp,v) ((sbp)->s_v1.s_bmap_nr = cpu_to_le16(v)) -#define sb_version(sbp) (le16_to_cpu((sbp)->s_v1.s_version)) -#define set_sb_version(sbp,v) ((sbp)->s_v1.s_version = cpu_to_le16(v)) - -#define sb_mnt_count(sbp) (le16_to_cpu((sbp)->s_mnt_count)) -#define set_sb_mnt_count(sbp, v) ((sbp)->s_mnt_count = cpu_to_le16(v)) - -#define sb_reserved_for_journal(sbp) \ - (le16_to_cpu((sbp)->s_v1.s_reserved_for_journal)) -#define set_sb_reserved_for_journal(sbp,v) \ - ((sbp)->s_v1.s_reserved_for_journal = cpu_to_le16(v)) - -/* LOGGING -- */ - -/* These all interelate for performance. -** -** If the journal block count is smaller than n transactions, you lose speed. -** I don't know what n is yet, I'm guessing 8-16. -** -** typical transaction size depends on the application, how often fsync is -** called, and how many metadata blocks you dirty in a 30 second period. -** The more small files (<16k) you use, the larger your transactions will -** be. -** -** If your journal fills faster than dirty buffers get flushed to disk, it must flush them before allowing the journal -** to wrap, which slows things down. If you need high speed meta data updates, the journal should be big enough -** to prevent wrapping before dirty meta blocks get to disk. -** -** If the batch max is smaller than the transaction max, you'll waste space at the end of the journal -** because journal_end sets the next transaction to start at 0 if the next transaction has any chance of wrapping. -** -** The large the batch max age, the better the speed, and the more meta data changes you'll lose after a crash. -** -*/ - -/* don't mess with these for a while */ - /* we have a node size define somewhere in reiserfs_fs.h. -Hans */ -#define JOURNAL_BLOCK_SIZE 4096 /* BUG gotta get rid of this */ -#define JOURNAL_MAX_CNODE 1500 /* max cnodes to allocate. */ -#define JOURNAL_HASH_SIZE 8192 -#define JOURNAL_NUM_BITMAPS 5 /* number of copies of the bitmaps to have floating. Must be >= 2 */ - -/* One of these for every block in every transaction -** Each one is in two hash tables. First, a hash of the current transaction, and after journal_end, a -** hash of all the in memory transactions. -** next and prev are used by the current transaction (journal_hash). -** hnext and hprev are used by journal_list_hash. If a block is in more than one transaction, the journal_list_hash -** links it in multiple times. This allows flush_journal_list to remove just the cnode belonging -** to a given transaction. -*/ -struct reiserfs_journal_cnode { - struct buffer_head *bh; /* real buffer head */ - struct super_block *sb; /* dev of real buffer head */ - __u32 blocknr; /* block number of real buffer head, == 0 when buffer on disk */ - unsigned long state; - struct reiserfs_journal_list *jlist; /* journal list this cnode lives in */ - struct reiserfs_journal_cnode *next; /* next in transaction list */ - struct reiserfs_journal_cnode *prev; /* prev in transaction list */ - struct reiserfs_journal_cnode *hprev; /* prev in hash list */ - struct reiserfs_journal_cnode *hnext; /* next in hash list */ -}; - -struct reiserfs_bitmap_node { - int id; - char *data; - struct list_head list; -}; - -struct reiserfs_list_bitmap { - struct reiserfs_journal_list *journal_list; - struct reiserfs_bitmap_node **bitmaps; -}; - -/* -** one of these for each transaction. The most important part here is the j_realblock. -** this list of cnodes is used to hash all the blocks in all the commits, to mark all the -** real buffer heads dirty once all the commits hit the disk, -** and to make sure every real block in a transaction is on disk before allowing the log area -** to be overwritten */ -struct reiserfs_journal_list { - unsigned long j_start; - unsigned long j_state; - unsigned long j_len; - atomic_t j_nonzerolen; - atomic_t j_commit_left; - atomic_t j_older_commits_done; /* all commits older than this on disk */ - struct mutex j_commit_mutex; - unsigned int j_trans_id; - time_t j_timestamp; - struct reiserfs_list_bitmap *j_list_bitmap; - struct buffer_head *j_commit_bh; /* commit buffer head */ - struct reiserfs_journal_cnode *j_realblock; - struct reiserfs_journal_cnode *j_freedlist; /* list of buffers that were freed during this trans. free each of these on flush */ - /* time ordered list of all active transactions */ - struct list_head j_list; - - /* time ordered list of all transactions we haven't tried to flush yet */ - struct list_head j_working_list; - - /* list of tail conversion targets in need of flush before commit */ - struct list_head j_tail_bh_list; - /* list of data=ordered buffers in need of flush before commit */ - struct list_head j_bh_list; - int j_refcount; -}; - -struct reiserfs_journal { - struct buffer_head **j_ap_blocks; /* journal blocks on disk */ - struct reiserfs_journal_cnode *j_last; /* newest journal block */ - struct reiserfs_journal_cnode *j_first; /* oldest journal block. start here for traverse */ - - struct block_device *j_dev_bd; - fmode_t j_dev_mode; - int j_1st_reserved_block; /* first block on s_dev of reserved area journal */ - - unsigned long j_state; - unsigned int j_trans_id; - unsigned long j_mount_id; - unsigned long j_start; /* start of current waiting commit (index into j_ap_blocks) */ - unsigned long j_len; /* length of current waiting commit */ - unsigned long j_len_alloc; /* number of buffers requested by journal_begin() */ - atomic_t j_wcount; /* count of writers for current commit */ - unsigned long j_bcount; /* batch count. allows turning X transactions into 1 */ - unsigned long j_first_unflushed_offset; /* first unflushed transactions offset */ - unsigned j_last_flush_trans_id; /* last fully flushed journal timestamp */ - struct buffer_head *j_header_bh; - - time_t j_trans_start_time; /* time this transaction started */ - struct mutex j_mutex; - struct mutex j_flush_mutex; - wait_queue_head_t j_join_wait; /* wait for current transaction to finish before starting new one */ - atomic_t j_jlock; /* lock for j_join_wait */ - int j_list_bitmap_index; /* number of next list bitmap to use */ - int j_must_wait; /* no more journal begins allowed. MUST sleep on j_join_wait */ - int j_next_full_flush; /* next journal_end will flush all journal list */ - int j_next_async_flush; /* next journal_end will flush all async commits */ - - int j_cnode_used; /* number of cnodes on the used list */ - int j_cnode_free; /* number of cnodes on the free list */ - - unsigned int j_trans_max; /* max number of blocks in a transaction. */ - unsigned int j_max_batch; /* max number of blocks to batch into a trans */ - unsigned int j_max_commit_age; /* in seconds, how old can an async commit be */ - unsigned int j_max_trans_age; /* in seconds, how old can a transaction be */ - unsigned int j_default_max_commit_age; /* the default for the max commit age */ - - struct reiserfs_journal_cnode *j_cnode_free_list; - struct reiserfs_journal_cnode *j_cnode_free_orig; /* orig pointer returned from vmalloc */ - - struct reiserfs_journal_list *j_current_jl; - int j_free_bitmap_nodes; - int j_used_bitmap_nodes; - - int j_num_lists; /* total number of active transactions */ - int j_num_work_lists; /* number that need attention from kreiserfsd */ - - /* debugging to make sure things are flushed in order */ - unsigned int j_last_flush_id; - - /* debugging to make sure things are committed in order */ - unsigned int j_last_commit_id; - - struct list_head j_bitmap_nodes; - struct list_head j_dirty_buffers; - spinlock_t j_dirty_buffers_lock; /* protects j_dirty_buffers */ - - /* list of all active transactions */ - struct list_head j_journal_list; - /* lists that haven't been touched by writeback attempts */ - struct list_head j_working_list; - - struct reiserfs_list_bitmap j_list_bitmap[JOURNAL_NUM_BITMAPS]; /* array of bitmaps to record the deleted blocks */ - struct reiserfs_journal_cnode *j_hash_table[JOURNAL_HASH_SIZE]; /* hash table for real buffer heads in current trans */ - struct reiserfs_journal_cnode *j_list_hash_table[JOURNAL_HASH_SIZE]; /* hash table for all the real buffer heads in all - the transactions */ - struct list_head j_prealloc_list; /* list of inodes which have preallocated blocks */ - int j_persistent_trans; - unsigned long j_max_trans_size; - unsigned long j_max_batch_size; - - int j_errno; - - /* when flushing ordered buffers, throttle new ordered writers */ - struct delayed_work j_work; - struct super_block *j_work_sb; - atomic_t j_async_throttle; -}; - -enum journal_state_bits { - J_WRITERS_BLOCKED = 1, /* set when new writers not allowed */ - J_WRITERS_QUEUED, /* set when log is full due to too many writers */ - J_ABORTED, /* set when log is aborted */ -}; - -#define JOURNAL_DESC_MAGIC "ReIsErLB" /* ick. magic string to find desc blocks in the journal */ - -typedef __u32(*hashf_t) (const signed char *, int); - -struct reiserfs_bitmap_info { - __u32 free_count; -}; - -struct proc_dir_entry; - -#if defined( CONFIG_PROC_FS ) && defined( CONFIG_REISERFS_PROC_INFO ) -typedef unsigned long int stat_cnt_t; -typedef struct reiserfs_proc_info_data { - spinlock_t lock; - int exiting; - int max_hash_collisions; - - stat_cnt_t breads; - stat_cnt_t bread_miss; - stat_cnt_t search_by_key; - stat_cnt_t search_by_key_fs_changed; - stat_cnt_t search_by_key_restarted; - - stat_cnt_t insert_item_restarted; - stat_cnt_t paste_into_item_restarted; - stat_cnt_t cut_from_item_restarted; - stat_cnt_t delete_solid_item_restarted; - stat_cnt_t delete_item_restarted; - - stat_cnt_t leaked_oid; - stat_cnt_t leaves_removable; - - /* balances per level. Use explicit 5 as MAX_HEIGHT is not visible yet. */ - stat_cnt_t balance_at[5]; /* XXX */ - /* sbk == search_by_key */ - stat_cnt_t sbk_read_at[5]; /* XXX */ - stat_cnt_t sbk_fs_changed[5]; - stat_cnt_t sbk_restarted[5]; - stat_cnt_t items_at[5]; /* XXX */ - stat_cnt_t free_at[5]; /* XXX */ - stat_cnt_t can_node_be_removed[5]; /* XXX */ - long int lnum[5]; /* XXX */ - long int rnum[5]; /* XXX */ - long int lbytes[5]; /* XXX */ - long int rbytes[5]; /* XXX */ - stat_cnt_t get_neighbors[5]; - stat_cnt_t get_neighbors_restart[5]; - stat_cnt_t need_l_neighbor[5]; - stat_cnt_t need_r_neighbor[5]; - - stat_cnt_t free_block; - struct __scan_bitmap_stats { - stat_cnt_t call; - stat_cnt_t wait; - stat_cnt_t bmap; - stat_cnt_t retry; - stat_cnt_t in_journal_hint; - stat_cnt_t in_journal_nohint; - stat_cnt_t stolen; - } scan_bitmap; - struct __journal_stats { - stat_cnt_t in_journal; - stat_cnt_t in_journal_bitmap; - stat_cnt_t in_journal_reusable; - stat_cnt_t lock_journal; - stat_cnt_t lock_journal_wait; - stat_cnt_t journal_being; - stat_cnt_t journal_relock_writers; - stat_cnt_t journal_relock_wcount; - stat_cnt_t mark_dirty; - stat_cnt_t mark_dirty_already; - stat_cnt_t mark_dirty_notjournal; - stat_cnt_t restore_prepared; - stat_cnt_t prepare; - stat_cnt_t prepare_retry; - } journal; -} reiserfs_proc_info_data_t; -#else -typedef struct reiserfs_proc_info_data { -} reiserfs_proc_info_data_t; -#endif - -/* reiserfs union of in-core super block data */ -struct reiserfs_sb_info { - struct buffer_head *s_sbh; /* Buffer containing the super block */ - /* both the comment and the choice of - name are unclear for s_rs -Hans */ - struct reiserfs_super_block *s_rs; /* Pointer to the super block in the buffer */ - struct reiserfs_bitmap_info *s_ap_bitmap; - struct reiserfs_journal *s_journal; /* pointer to journal information */ - unsigned short s_mount_state; /* reiserfs state (valid, invalid) */ - - /* Serialize writers access, replace the old bkl */ - struct mutex lock; - /* Owner of the lock (can be recursive) */ - struct task_struct *lock_owner; - /* Depth of the lock, start from -1 like the bkl */ - int lock_depth; - - /* Comment? -Hans */ - void (*end_io_handler) (struct buffer_head *, int); - hashf_t s_hash_function; /* pointer to function which is used - to sort names in directory. Set on - mount */ - unsigned long s_mount_opt; /* reiserfs's mount options are set - here (currently - NOTAIL, NOLOG, - REPLAYONLY) */ - - struct { /* This is a structure that describes block allocator options */ - unsigned long bits; /* Bitfield for enable/disable kind of options */ - unsigned long large_file_size; /* size started from which we consider file to be a large one(in blocks) */ - int border; /* percentage of disk, border takes */ - int preallocmin; /* Minimal file size (in blocks) starting from which we do preallocations */ - int preallocsize; /* Number of blocks we try to prealloc when file - reaches preallocmin size (in blocks) or - prealloc_list is empty. */ - } s_alloc_options; - - /* Comment? -Hans */ - wait_queue_head_t s_wait; - /* To be obsoleted soon by per buffer seals.. -Hans */ - atomic_t s_generation_counter; // increased by one every time the - // tree gets re-balanced - unsigned long s_properties; /* File system properties. Currently holds - on-disk FS format */ - - /* session statistics */ - int s_disk_reads; - int s_disk_writes; - int s_fix_nodes; - int s_do_balance; - int s_unneeded_left_neighbor; - int s_good_search_by_key_reada; - int s_bmaps; - int s_bmaps_without_search; - int s_direct2indirect; - int s_indirect2direct; - /* set up when it's ok for reiserfs_read_inode2() to read from - disk inode with nlink==0. Currently this is only used during - finish_unfinished() processing at mount time */ - int s_is_unlinked_ok; - reiserfs_proc_info_data_t s_proc_info_data; - struct proc_dir_entry *procdir; - int reserved_blocks; /* amount of blocks reserved for further allocations */ - spinlock_t bitmap_lock; /* this lock on now only used to protect reserved_blocks variable */ - struct dentry *priv_root; /* root of /.reiserfs_priv */ - struct dentry *xattr_root; /* root of /.reiserfs_priv/xattrs */ - int j_errno; -#ifdef CONFIG_QUOTA - char *s_qf_names[MAXQUOTAS]; - int s_jquota_fmt; -#endif - char *s_jdev; /* Stored jdev for mount option showing */ -#ifdef CONFIG_REISERFS_CHECK - - struct tree_balance *cur_tb; /* - * Detects whether more than one - * copy of tb exists per superblock - * as a means of checking whether - * do_balance is executing concurrently - * against another tree reader/writer - * on a same mount point. - */ -#endif -}; - -/* Definitions of reiserfs on-disk properties: */ -#define REISERFS_3_5 0 -#define REISERFS_3_6 1 -#define REISERFS_OLD_FORMAT 2 - -enum reiserfs_mount_options { -/* Mount options */ - REISERFS_LARGETAIL, /* large tails will be created in a session */ - REISERFS_SMALLTAIL, /* small (for files less than block size) tails will be created in a session */ - REPLAYONLY, /* replay journal and return 0. Use by fsck */ - REISERFS_CONVERT, /* -o conv: causes conversion of old - format super block to the new - format. If not specified - old - partition will be dealt with in a - manner of 3.5.x */ - -/* -o hash={tea, rupasov, r5, detect} is meant for properly mounting -** reiserfs disks from 3.5.19 or earlier. 99% of the time, this option -** is not required. If the normal autodection code can't determine which -** hash to use (because both hashes had the same value for a file) -** use this option to force a specific hash. It won't allow you to override -** the existing hash on the FS, so if you have a tea hash disk, and mount -** with -o hash=rupasov, the mount will fail. -*/ - FORCE_TEA_HASH, /* try to force tea hash on mount */ - FORCE_RUPASOV_HASH, /* try to force rupasov hash on mount */ - FORCE_R5_HASH, /* try to force rupasov hash on mount */ - FORCE_HASH_DETECT, /* try to detect hash function on mount */ - - REISERFS_DATA_LOG, - REISERFS_DATA_ORDERED, - REISERFS_DATA_WRITEBACK, - -/* used for testing experimental features, makes benchmarking new - features with and without more convenient, should never be used by - users in any code shipped to users (ideally) */ - - REISERFS_NO_BORDER, - REISERFS_NO_UNHASHED_RELOCATION, - REISERFS_HASHED_RELOCATION, - REISERFS_ATTRS, - REISERFS_XATTRS_USER, - REISERFS_POSIXACL, - REISERFS_EXPOSE_PRIVROOT, - REISERFS_BARRIER_NONE, - REISERFS_BARRIER_FLUSH, - - /* Actions on error */ - REISERFS_ERROR_PANIC, - REISERFS_ERROR_RO, - REISERFS_ERROR_CONTINUE, - - REISERFS_USRQUOTA, /* User quota option specified */ - REISERFS_GRPQUOTA, /* Group quota option specified */ - - REISERFS_TEST1, - REISERFS_TEST2, - REISERFS_TEST3, - REISERFS_TEST4, - REISERFS_UNSUPPORTED_OPT, -}; - -#define reiserfs_r5_hash(s) (REISERFS_SB(s)->s_mount_opt & (1 << FORCE_R5_HASH)) -#define reiserfs_rupasov_hash(s) (REISERFS_SB(s)->s_mount_opt & (1 << FORCE_RUPASOV_HASH)) -#define reiserfs_tea_hash(s) (REISERFS_SB(s)->s_mount_opt & (1 << FORCE_TEA_HASH)) -#define reiserfs_hash_detect(s) (REISERFS_SB(s)->s_mount_opt & (1 << FORCE_HASH_DETECT)) -#define reiserfs_no_border(s) (REISERFS_SB(s)->s_mount_opt & (1 << REISERFS_NO_BORDER)) -#define reiserfs_no_unhashed_relocation(s) (REISERFS_SB(s)->s_mount_opt & (1 << REISERFS_NO_UNHASHED_RELOCATION)) -#define reiserfs_hashed_relocation(s) (REISERFS_SB(s)->s_mount_opt & (1 << REISERFS_HASHED_RELOCATION)) -#define reiserfs_test4(s) (REISERFS_SB(s)->s_mount_opt & (1 << REISERFS_TEST4)) - -#define have_large_tails(s) (REISERFS_SB(s)->s_mount_opt & (1 << REISERFS_LARGETAIL)) -#define have_small_tails(s) (REISERFS_SB(s)->s_mount_opt & (1 << REISERFS_SMALLTAIL)) -#define replay_only(s) (REISERFS_SB(s)->s_mount_opt & (1 << REPLAYONLY)) -#define reiserfs_attrs(s) (REISERFS_SB(s)->s_mount_opt & (1 << REISERFS_ATTRS)) -#define old_format_only(s) (REISERFS_SB(s)->s_properties & (1 << REISERFS_3_5)) -#define convert_reiserfs(s) (REISERFS_SB(s)->s_mount_opt & (1 << REISERFS_CONVERT)) -#define reiserfs_data_log(s) (REISERFS_SB(s)->s_mount_opt & (1 << REISERFS_DATA_LOG)) -#define reiserfs_data_ordered(s) (REISERFS_SB(s)->s_mount_opt & (1 << REISERFS_DATA_ORDERED)) -#define reiserfs_data_writeback(s) (REISERFS_SB(s)->s_mount_opt & (1 << REISERFS_DATA_WRITEBACK)) -#define reiserfs_xattrs_user(s) (REISERFS_SB(s)->s_mount_opt & (1 << REISERFS_XATTRS_USER)) -#define reiserfs_posixacl(s) (REISERFS_SB(s)->s_mount_opt & (1 << REISERFS_POSIXACL)) -#define reiserfs_expose_privroot(s) (REISERFS_SB(s)->s_mount_opt & (1 << REISERFS_EXPOSE_PRIVROOT)) -#define reiserfs_xattrs_optional(s) (reiserfs_xattrs_user(s) || reiserfs_posixacl(s)) -#define reiserfs_barrier_none(s) (REISERFS_SB(s)->s_mount_opt & (1 << REISERFS_BARRIER_NONE)) -#define reiserfs_barrier_flush(s) (REISERFS_SB(s)->s_mount_opt & (1 << REISERFS_BARRIER_FLUSH)) - -#define reiserfs_error_panic(s) (REISERFS_SB(s)->s_mount_opt & (1 << REISERFS_ERROR_PANIC)) -#define reiserfs_error_ro(s) (REISERFS_SB(s)->s_mount_opt & (1 << REISERFS_ERROR_RO)) - -void reiserfs_file_buffer(struct buffer_head *bh, int list); -extern struct file_system_type reiserfs_fs_type; -int reiserfs_resize(struct super_block *, unsigned long); - -#define CARRY_ON 0 -#define SCHEDULE_OCCURRED 1 - -#define SB_BUFFER_WITH_SB(s) (REISERFS_SB(s)->s_sbh) -#define SB_JOURNAL(s) (REISERFS_SB(s)->s_journal) -#define SB_JOURNAL_1st_RESERVED_BLOCK(s) (SB_JOURNAL(s)->j_1st_reserved_block) -#define SB_JOURNAL_LEN_FREE(s) (SB_JOURNAL(s)->j_journal_len_free) -#define SB_AP_BITMAP(s) (REISERFS_SB(s)->s_ap_bitmap) - -#define SB_DISK_JOURNAL_HEAD(s) (SB_JOURNAL(s)->j_header_bh->) - -/* A safe version of the "bdevname", which returns the "s_id" field of - * a superblock or else "Null superblock" if the super block is NULL. - */ -static inline char *reiserfs_bdevname(struct super_block *s) -{ - return (s == NULL) ? "Null superblock" : s->s_id; -} - -#define reiserfs_is_journal_aborted(journal) (unlikely (__reiserfs_is_journal_aborted (journal))) -static inline int __reiserfs_is_journal_aborted(struct reiserfs_journal - *journal) -{ - return test_bit(J_ABORTED, &journal->j_state); -} - -#endif /* _LINUX_REISER_FS_SB */ From 640045a1cd21d002caafa9eb9ed66cc31b88076a Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 17 Mar 2012 01:37:51 -0400 Subject: [PATCH 35/64] spufs: make register_filesystem the last potential failure exit Signed-off-by: Al Viro --- arch/powerpc/platforms/cell/spufs/inode.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/arch/powerpc/platforms/cell/spufs/inode.c b/arch/powerpc/platforms/cell/spufs/inode.c index 17b3211e3641..4a3d90e0d426 100644 --- a/arch/powerpc/platforms/cell/spufs/inode.c +++ b/arch/powerpc/platforms/cell/spufs/inode.c @@ -828,19 +828,19 @@ static int __init spufs_init(void) ret = spu_sched_init(); if (ret) goto out_cache; - ret = register_filesystem(&spufs_type); - if (ret) - goto out_sched; ret = register_spu_syscalls(&spufs_calls); if (ret) - goto out_fs; + goto out_sched; + ret = register_filesystem(&spufs_type); + if (ret) + goto out_syscalls; spufs_init_isolated_loader(); return 0; -out_fs: - unregister_filesystem(&spufs_type); +out_syscalls: + unregister_spu_syscalls(&spufs_calls); out_sched: spu_sched_exit(); out_cache: From ea23b453103e7495ef0a90c86b92f27e17092bfa Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 17 Mar 2012 01:50:32 -0400 Subject: [PATCH 36/64] ibmasmfs: make register_filesystem the last failure exit Signed-off-by: Al Viro --- drivers/misc/ibmasm/module.c | 11 +++++------ 1 file changed, 5 insertions(+), 6 deletions(-) diff --git a/drivers/misc/ibmasm/module.c b/drivers/misc/ibmasm/module.c index 1ccedb71e728..168d8008f460 100644 --- a/drivers/misc/ibmasm/module.c +++ b/drivers/misc/ibmasm/module.c @@ -211,18 +211,17 @@ static void __exit ibmasm_exit (void) static int __init ibmasm_init(void) { - int result; + int result = pci_register_driver(&ibmasm_driver); + if (result) + return result; result = ibmasmfs_register(); if (result) { + pci_unregister_driver(&ibmasm_driver); err("Failed to register ibmasmfs file system"); return result; } - result = pci_register_driver(&ibmasm_driver); - if (result) { - ibmasmfs_unregister(); - return result; - } + ibmasm_register_panic_notifier(); info(DRIVER_DESC " version " DRIVER_VERSION " loaded"); return 0; From 2452992aa78315bf0f510620d22fae8ecb40a2eb Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 17 Mar 2012 02:13:52 -0400 Subject: [PATCH 37/64] make simple_pin_fs() pass MS_KERNMOUNT - it's a kernel-internal one Signed-off-by: Al Viro --- fs/libfs.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/libfs.c b/fs/libfs.c index 7c895a763a1e..722e0d5ba182 100644 --- a/fs/libfs.c +++ b/fs/libfs.c @@ -534,7 +534,7 @@ int simple_pin_fs(struct file_system_type *type, struct vfsmount **mount, int *c spin_lock(&pin_fs_lock); if (unlikely(!*mount)) { spin_unlock(&pin_fs_lock); - mnt = vfs_kern_mount(type, 0, type->name, NULL); + mnt = vfs_kern_mount(type, MS_KERNMOUNT, type->name, NULL); if (IS_ERR(mnt)) return PTR_ERR(mnt); spin_lock(&pin_fs_lock); From 01e0fe0b10d8673554525984947cc5de21492462 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 17 Mar 2012 02:17:32 -0400 Subject: [PATCH 38/64] usbfs: kill racy detection of simple_pin_fs() can check MS_KERNMOUNT in flags now Signed-off-by: Al Viro --- drivers/usb/core/inode.c | 10 +--------- 1 file changed, 1 insertion(+), 9 deletions(-) diff --git a/drivers/usb/core/inode.c b/drivers/usb/core/inode.c index 06c175aee6a3..cefa0c8b5b6a 100644 --- a/drivers/usb/core/inode.c +++ b/drivers/usb/core/inode.c @@ -50,7 +50,6 @@ static const struct file_operations default_file_operations; static struct vfsmount *usbfs_mount; static int usbfs_mount_count; /* = 0 */ -static int ignore_mount = 0; static struct dentry *devices_usbfs_dentry; static int num_buses; /* = 0 */ @@ -256,7 +255,7 @@ static int remount(struct super_block *sb, int *flags, char *data) * i.e. it's a simple_pin_fs from create_special_files, * then ignore it. */ - if (ignore_mount) + if (*flags & MS_KERNMOUNT) return 0; if (parse_options(sb, data)) { @@ -582,11 +581,6 @@ static int create_special_files (void) struct dentry *parent; int retval; - /* the simple_pin_fs calls will call remount with no options - * without this flag that would overwrite the real mount options (if any) - */ - ignore_mount = 1; - /* create the devices special file */ retval = simple_pin_fs(&usb_fs_type, &usbfs_mount, &usbfs_mount_count); if (retval) { @@ -594,8 +588,6 @@ static int create_special_files (void) goto exit; } - ignore_mount = 0; - parent = usbfs_mount->mnt_root; devices_usbfs_dentry = fs_create_file ("devices", listmode | S_IFREG, parent, From 2226a288fac462ebc98e40da007842f92a7e4799 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 17 Mar 2012 02:37:05 -0400 Subject: [PATCH 39/64] 9p: make register_filesystem() the last failure exit Signed-off-by: Al Viro --- fs/9p/v9fs.c | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-) diff --git a/fs/9p/v9fs.c b/fs/9p/v9fs.c index 1964f98e74be..b85efa773949 100644 --- a/fs/9p/v9fs.c +++ b/fs/9p/v9fs.c @@ -594,21 +594,21 @@ static int __init init_v9fs(void) int err; pr_info("Installing v9fs 9p2000 file system support\n"); /* TODO: Setup list of registered trasnport modules */ - err = register_filesystem(&v9fs_fs_type); - if (err < 0) { - pr_err("Failed to register filesystem\n"); - return err; - } err = v9fs_cache_register(); if (err < 0) { pr_err("Failed to register v9fs for caching\n"); - goto out_fs_unreg; + return err; } err = v9fs_sysfs_init(); if (err < 0) { pr_err("Failed to register with sysfs\n"); + goto out_cache; + } + err = register_filesystem(&v9fs_fs_type); + if (err < 0) { + pr_err("Failed to register filesystem\n"); goto out_sysfs_cleanup; } @@ -617,8 +617,8 @@ static int __init init_v9fs(void) out_sysfs_cleanup: v9fs_sysfs_cleanup(); -out_fs_unreg: - unregister_filesystem(&v9fs_fs_type); +out_cache: + v9fs_cache_unregister(); return err; } From ca7068c41a8d48ef35d1a3bba422ebcedace8513 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 17 Mar 2012 02:52:29 -0400 Subject: [PATCH 40/64] anon_inodes: move allocation of anon_inode into ->mount() Signed-off-by: Al Viro --- fs/anon_inodes.c | 109 ++++++++++++++++++++++++----------------------- 1 file changed, 56 insertions(+), 53 deletions(-) diff --git a/fs/anon_inodes.c b/fs/anon_inodes.c index f11e43ed907d..28d39fb84ae3 100644 --- a/fs/anon_inodes.c +++ b/fs/anon_inodes.c @@ -39,19 +39,6 @@ static const struct dentry_operations anon_inodefs_dentry_operations = { .d_dname = anon_inodefs_dname, }; -static struct dentry *anon_inodefs_mount(struct file_system_type *fs_type, - int flags, const char *dev_name, void *data) -{ - return mount_pseudo(fs_type, "anon_inode:", NULL, - &anon_inodefs_dentry_operations, ANON_INODE_FS_MAGIC); -} - -static struct file_system_type anon_inode_fs_type = { - .name = "anon_inodefs", - .mount = anon_inodefs_mount, - .kill_sb = kill_anon_super, -}; - /* * nop .set_page_dirty method so that people can use .page_mkwrite on * anon inodes. @@ -65,6 +52,62 @@ static const struct address_space_operations anon_aops = { .set_page_dirty = anon_set_page_dirty, }; +/* + * A single inode exists for all anon_inode files. Contrary to pipes, + * anon_inode inodes have no associated per-instance data, so we need + * only allocate one of them. + */ +static struct inode *anon_inode_mkinode(struct super_block *s) +{ + struct inode *inode = new_inode_pseudo(s); + + if (!inode) + return ERR_PTR(-ENOMEM); + + inode->i_ino = get_next_ino(); + inode->i_fop = &anon_inode_fops; + + inode->i_mapping->a_ops = &anon_aops; + + /* + * Mark the inode dirty from the very beginning, + * that way it will never be moved to the dirty + * list because mark_inode_dirty() will think + * that it already _is_ on the dirty list. + */ + inode->i_state = I_DIRTY; + inode->i_mode = S_IRUSR | S_IWUSR; + inode->i_uid = current_fsuid(); + inode->i_gid = current_fsgid(); + inode->i_flags |= S_PRIVATE; + inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME; + return inode; +} + +static struct dentry *anon_inodefs_mount(struct file_system_type *fs_type, + int flags, const char *dev_name, void *data) +{ + struct dentry *root; + root = mount_pseudo(fs_type, "anon_inode:", NULL, + &anon_inodefs_dentry_operations, ANON_INODE_FS_MAGIC); + if (!IS_ERR(root)) { + struct super_block *s = root->d_sb; + anon_inode_inode = anon_inode_mkinode(s); + if (IS_ERR(anon_inode_inode)) { + dput(root); + deactivate_locked_super(s); + root = ERR_CAST(anon_inode_inode); + } + } + return root; +} + +static struct file_system_type anon_inode_fs_type = { + .name = "anon_inodefs", + .mount = anon_inodefs_mount, + .kill_sb = kill_anon_super, +}; + /** * anon_inode_getfile - creates a new file instance by hooking it up to an * anonymous inode, and a dentry that describe the "class" @@ -180,38 +223,6 @@ err_put_unused_fd: } EXPORT_SYMBOL_GPL(anon_inode_getfd); -/* - * A single inode exists for all anon_inode files. Contrary to pipes, - * anon_inode inodes have no associated per-instance data, so we need - * only allocate one of them. - */ -static struct inode *anon_inode_mkinode(void) -{ - struct inode *inode = new_inode_pseudo(anon_inode_mnt->mnt_sb); - - if (!inode) - return ERR_PTR(-ENOMEM); - - inode->i_ino = get_next_ino(); - inode->i_fop = &anon_inode_fops; - - inode->i_mapping->a_ops = &anon_aops; - - /* - * Mark the inode dirty from the very beginning, - * that way it will never be moved to the dirty - * list because mark_inode_dirty() will think - * that it already _is_ on the dirty list. - */ - inode->i_state = I_DIRTY; - inode->i_mode = S_IRUSR | S_IWUSR; - inode->i_uid = current_fsuid(); - inode->i_gid = current_fsgid(); - inode->i_flags |= S_PRIVATE; - inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME; - return inode; -} - static int __init anon_inode_init(void) { int error; @@ -224,16 +235,8 @@ static int __init anon_inode_init(void) error = PTR_ERR(anon_inode_mnt); goto err_unregister_filesystem; } - anon_inode_inode = anon_inode_mkinode(); - if (IS_ERR(anon_inode_inode)) { - error = PTR_ERR(anon_inode_inode); - goto err_mntput; - } - return 0; -err_mntput: - kern_unmount(anon_inode_mnt); err_unregister_filesystem: unregister_filesystem(&anon_inode_fs_type); err_exit: From 54bf586e1f51018ba7624c851e7aa14cee0548d2 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 17 Mar 2012 02:55:49 -0400 Subject: [PATCH 41/64] autofs: set things up *before* registering fs type it's not a serious race, but we really want misc device before anybody gets to mount this sucker. Signed-off-by: Al Viro --- fs/autofs4/init.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/fs/autofs4/init.c b/fs/autofs4/init.c index c038727b4050..cddc74b9cdb2 100644 --- a/fs/autofs4/init.c +++ b/fs/autofs4/init.c @@ -31,11 +31,11 @@ static int __init init_autofs4_fs(void) { int err; + autofs_dev_ioctl_init(); + err = register_filesystem(&autofs_fs_type); if (err) - return err; - - autofs_dev_ioctl_init(); + autofs_dev_ioctl_exit(); return err; } From 8fc3dc5a3a17aa2b353886422bd89420619af211 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 17 Mar 2012 03:05:16 -0400 Subject: [PATCH 42/64] __register_binfmt() made void Just don't pass NULL to it - nobody does, anyway. Signed-off-by: Al Viro --- arch/alpha/kernel/binfmt_loader.c | 3 ++- arch/x86/ia32/ia32_aout.c | 3 ++- fs/binfmt_aout.c | 3 ++- fs/binfmt_elf.c | 3 ++- fs/binfmt_elf_fdpic.c | 3 ++- fs/binfmt_em86.c | 3 ++- fs/binfmt_flat.c | 3 ++- fs/binfmt_misc.c | 7 ++----- fs/binfmt_script.c | 3 ++- fs/binfmt_som.c | 3 ++- fs/exec.c | 6 ++---- include/linux/binfmts.h | 10 +++++----- 12 files changed, 27 insertions(+), 23 deletions(-) diff --git a/arch/alpha/kernel/binfmt_loader.c b/arch/alpha/kernel/binfmt_loader.c index 3fcfad410130..d1f474d1d44d 100644 --- a/arch/alpha/kernel/binfmt_loader.c +++ b/arch/alpha/kernel/binfmt_loader.c @@ -46,6 +46,7 @@ static struct linux_binfmt loader_format = { static int __init init_loader_binfmt(void) { - return insert_binfmt(&loader_format); + insert_binfmt(&loader_format); + return 0; } arch_initcall(init_loader_binfmt); diff --git a/arch/x86/ia32/ia32_aout.c b/arch/x86/ia32/ia32_aout.c index 39e49091f648..cdfc8dc43670 100644 --- a/arch/x86/ia32/ia32_aout.c +++ b/arch/x86/ia32/ia32_aout.c @@ -519,7 +519,8 @@ out: static int __init init_aout_binfmt(void) { - return register_binfmt(&aout_format); + register_binfmt(&aout_format); + return 0; } static void __exit exit_aout_binfmt(void) diff --git a/fs/binfmt_aout.c b/fs/binfmt_aout.c index 1ff94054d35a..a543364ba29b 100644 --- a/fs/binfmt_aout.c +++ b/fs/binfmt_aout.c @@ -454,7 +454,8 @@ out: static int __init init_aout_binfmt(void) { - return register_binfmt(&aout_format); + register_binfmt(&aout_format); + return 0; } static void __exit exit_aout_binfmt(void) diff --git a/fs/binfmt_elf.c b/fs/binfmt_elf.c index 07d096c49920..f8ac4251877e 100644 --- a/fs/binfmt_elf.c +++ b/fs/binfmt_elf.c @@ -2077,7 +2077,8 @@ out: static int __init init_elf_binfmt(void) { - return register_binfmt(&elf_format); + register_binfmt(&elf_format); + return 0; } static void __exit exit_elf_binfmt(void) diff --git a/fs/binfmt_elf_fdpic.c b/fs/binfmt_elf_fdpic.c index 30745f459faf..e7afcb67a2d3 100644 --- a/fs/binfmt_elf_fdpic.c +++ b/fs/binfmt_elf_fdpic.c @@ -91,7 +91,8 @@ static struct linux_binfmt elf_fdpic_format = { static int __init init_elf_fdpic_binfmt(void) { - return register_binfmt(&elf_fdpic_format); + register_binfmt(&elf_fdpic_format); + return 0; } static void __exit exit_elf_fdpic_binfmt(void) diff --git a/fs/binfmt_em86.c b/fs/binfmt_em86.c index b8e8b0acf9bd..2790c7e1912e 100644 --- a/fs/binfmt_em86.c +++ b/fs/binfmt_em86.c @@ -100,7 +100,8 @@ static struct linux_binfmt em86_format = { static int __init init_em86_binfmt(void) { - return register_binfmt(&em86_format); + register_binfmt(&em86_format); + return 0; } static void __exit exit_em86_binfmt(void) diff --git a/fs/binfmt_flat.c b/fs/binfmt_flat.c index 1bffbe0ed778..68affab88146 100644 --- a/fs/binfmt_flat.c +++ b/fs/binfmt_flat.c @@ -950,7 +950,8 @@ static int load_flat_binary(struct linux_binprm * bprm, struct pt_regs * regs) static int __init init_flat_binfmt(void) { - return register_binfmt(&flat_format); + register_binfmt(&flat_format); + return 0; } /****************************************************************************/ diff --git a/fs/binfmt_misc.c b/fs/binfmt_misc.c index a9198dfd5f85..1ffb60355cae 100644 --- a/fs/binfmt_misc.c +++ b/fs/binfmt_misc.c @@ -726,11 +726,8 @@ static struct file_system_type bm_fs_type = { static int __init init_misc_binfmt(void) { int err = register_filesystem(&bm_fs_type); - if (!err) { - err = insert_binfmt(&misc_format); - if (err) - unregister_filesystem(&bm_fs_type); - } + if (!err) + insert_binfmt(&misc_format); return err; } diff --git a/fs/binfmt_script.c b/fs/binfmt_script.c index 396a9884591f..d3b8c1f63155 100644 --- a/fs/binfmt_script.c +++ b/fs/binfmt_script.c @@ -105,7 +105,8 @@ static struct linux_binfmt script_format = { static int __init init_script_binfmt(void) { - return register_binfmt(&script_format); + register_binfmt(&script_format); + return 0; } static void __exit exit_script_binfmt(void) diff --git a/fs/binfmt_som.c b/fs/binfmt_som.c index cc8560f6c9b0..ec15972dd98a 100644 --- a/fs/binfmt_som.c +++ b/fs/binfmt_som.c @@ -289,7 +289,8 @@ static int load_som_library(struct file *f) static int __init init_som_binfmt(void) { - return register_binfmt(&som_format); + register_binfmt(&som_format); + return 0; } static void __exit exit_som_binfmt(void) diff --git a/fs/exec.c b/fs/exec.c index 153dee14fe55..2c5ae338773c 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -79,15 +79,13 @@ static atomic_t call_count = ATOMIC_INIT(1); static LIST_HEAD(formats); static DEFINE_RWLOCK(binfmt_lock); -int __register_binfmt(struct linux_binfmt * fmt, int insert) +void __register_binfmt(struct linux_binfmt * fmt, int insert) { - if (!fmt) - return -EINVAL; + BUG_ON(!fmt); write_lock(&binfmt_lock); insert ? list_add(&fmt->lh, &formats) : list_add_tail(&fmt->lh, &formats); write_unlock(&binfmt_lock); - return 0; } EXPORT_SYMBOL(__register_binfmt); diff --git a/include/linux/binfmts.h b/include/linux/binfmts.h index 0092102db2de..366422bc1633 100644 --- a/include/linux/binfmts.h +++ b/include/linux/binfmts.h @@ -92,17 +92,17 @@ struct linux_binfmt { unsigned long min_coredump; /* minimal dump size */ }; -extern int __register_binfmt(struct linux_binfmt *fmt, int insert); +extern void __register_binfmt(struct linux_binfmt *fmt, int insert); /* Registration of default binfmt handlers */ -static inline int register_binfmt(struct linux_binfmt *fmt) +static inline void register_binfmt(struct linux_binfmt *fmt) { - return __register_binfmt(fmt, 0); + __register_binfmt(fmt, 0); } /* Same as above, but adds a new binfmt at the top of the list */ -static inline int insert_binfmt(struct linux_binfmt *fmt) +static inline void insert_binfmt(struct linux_binfmt *fmt) { - return __register_binfmt(fmt, 1); + __register_binfmt(fmt, 1); } extern void unregister_binfmt(struct linux_binfmt *); From 81d44ed159e3e81f7e62cee2d0fe68aae0c95e78 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 17 Mar 2012 16:13:25 -0400 Subject: [PATCH 43/64] configfs: don't bother with checks for mkdir/rmdir/unlink/symlink in root just give root directory separate inode_operations without all those methods... Signed-off-by: Al Viro --- fs/configfs/configfs_internal.h | 1 + fs/configfs/dir.c | 13 +++++-------- fs/configfs/mount.c | 2 +- fs/configfs/symlink.c | 6 ------ 4 files changed, 7 insertions(+), 15 deletions(-) diff --git a/fs/configfs/configfs_internal.h b/fs/configfs/configfs_internal.h index ede857d20a04..1b7fdc0a6a09 100644 --- a/fs/configfs/configfs_internal.h +++ b/fs/configfs/configfs_internal.h @@ -89,6 +89,7 @@ extern const struct file_operations configfs_dir_operations; extern const struct file_operations configfs_file_operations; extern const struct file_operations bin_fops; extern const struct inode_operations configfs_dir_inode_operations; +extern const struct inode_operations configfs_root_inode_operations; extern const struct inode_operations configfs_symlink_inode_operations; extern const struct dentry_operations configfs_dentry_ops; diff --git a/fs/configfs/dir.c b/fs/configfs/dir.c index 5ddd7ebd9dcd..b0fbcbeb03ee 100644 --- a/fs/configfs/dir.c +++ b/fs/configfs/dir.c @@ -1183,11 +1183,6 @@ static int configfs_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode struct module *subsys_owner = NULL, *new_item_owner = NULL; char *name; - if (dentry->d_parent == configfs_sb->s_root) { - ret = -EPERM; - goto out; - } - sd = dentry->d_parent->d_fsdata; /* @@ -1359,9 +1354,6 @@ static int configfs_rmdir(struct inode *dir, struct dentry *dentry) struct module *subsys_owner = NULL, *dead_item_owner = NULL; int ret; - if (dentry->d_parent == configfs_sb->s_root) - return -EPERM; - sd = dentry->d_fsdata; if (sd->s_type & CONFIGFS_USET_DEFAULT) return -EPERM; @@ -1459,6 +1451,11 @@ const struct inode_operations configfs_dir_inode_operations = { .setattr = configfs_setattr, }; +const struct inode_operations configfs_root_inode_operations = { + .lookup = configfs_lookup, + .setattr = configfs_setattr, +}; + #if 0 int configfs_rename_dir(struct config_item * item, const char *new_name) { diff --git a/fs/configfs/mount.c b/fs/configfs/mount.c index 07f60455f1c1..eb41adc28cfe 100644 --- a/fs/configfs/mount.c +++ b/fs/configfs/mount.c @@ -82,7 +82,7 @@ static int configfs_fill_super(struct super_block *sb, void *data, int silent) inode = configfs_new_inode(S_IFDIR | S_IRWXU | S_IRUGO | S_IXUGO, &configfs_root); if (inode) { - inode->i_op = &configfs_dir_inode_operations; + inode->i_op = &configfs_root_inode_operations; inode->i_fop = &configfs_dir_operations; /* directory inodes start off with i_nlink == 2 (for "." entry) */ inc_nlink(inode); diff --git a/fs/configfs/symlink.c b/fs/configfs/symlink.c index 0f3eb41d9201..2817153d33c2 100644 --- a/fs/configfs/symlink.c +++ b/fs/configfs/symlink.c @@ -141,10 +141,6 @@ int configfs_symlink(struct inode *dir, struct dentry *dentry, const char *symna struct config_item *target_item = NULL; struct config_item_type *type; - ret = -EPERM; /* What lack-of-symlink returns */ - if (dentry->d_parent == configfs_sb->s_root) - goto out; - sd = dentry->d_parent->d_fsdata; /* * Fake invisibility if dir belongs to a group/default groups hierarchy @@ -198,8 +194,6 @@ int configfs_unlink(struct inode *dir, struct dentry *dentry) if (!(sd->s_type & CONFIGFS_ITEM_LINK)) goto out; - BUG_ON(dentry->d_parent == configfs_sb->s_root); - sl = sd->s_element; parent_item = configfs_get_config_item(dentry->d_parent); From b7c177fcd2022ca8572284deb8f9b6ab5730eafb Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 17 Mar 2012 16:24:54 -0400 Subject: [PATCH 44/64] configfs: kill configfs_sb Signed-off-by: Al Viro --- fs/configfs/configfs_internal.h | 3 +-- fs/configfs/dir.c | 29 +++++++++++++++++------------ fs/configfs/inode.c | 9 ++++----- fs/configfs/mount.c | 4 +--- fs/configfs/symlink.c | 6 +++--- 5 files changed, 26 insertions(+), 25 deletions(-) diff --git a/fs/configfs/configfs_internal.h b/fs/configfs/configfs_internal.h index 1b7fdc0a6a09..37121c2be110 100644 --- a/fs/configfs/configfs_internal.h +++ b/fs/configfs/configfs_internal.h @@ -63,7 +63,7 @@ extern struct kmem_cache *configfs_dir_cachep; extern int configfs_is_root(struct config_item *item); -extern struct inode * configfs_new_inode(umode_t mode, struct configfs_dirent *); +extern struct inode * configfs_new_inode(umode_t mode, struct configfs_dirent *, struct super_block *); extern int configfs_create(struct dentry *, umode_t mode, int (*init)(struct inode *)); extern int configfs_inode_init(void); extern void configfs_inode_exit(void); @@ -84,7 +84,6 @@ extern int configfs_pin_fs(void); extern void configfs_release_fs(void); extern struct rw_semaphore configfs_rename_sem; -extern struct super_block * configfs_sb; extern const struct file_operations configfs_dir_operations; extern const struct file_operations configfs_file_operations; extern const struct file_operations bin_fops; diff --git a/fs/configfs/dir.c b/fs/configfs/dir.c index b0fbcbeb03ee..54c59a7e37ce 100644 --- a/fs/configfs/dir.c +++ b/fs/configfs/dir.c @@ -1079,6 +1079,7 @@ int configfs_depend_item(struct configfs_subsystem *subsys, int ret; struct configfs_dirent *p, *root_sd, *subsys_sd = NULL; struct config_item *s_item = &subsys->su_group.cg_item; + struct dentry *root; /* * Pin the configfs filesystem. This means we can safely access @@ -1093,9 +1094,10 @@ int configfs_depend_item(struct configfs_subsystem *subsys, * subsystem is really registered, and so we need to lock out * configfs_[un]register_subsystem(). */ - mutex_lock(&configfs_sb->s_root->d_inode->i_mutex); + root = configfs_mount->mnt_root; + mutex_lock(&root->d_inode->i_mutex); - root_sd = configfs_sb->s_root->d_fsdata; + root_sd = root->d_fsdata; list_for_each_entry(p, &root_sd->s_children, s_sibling) { if (p->s_type & CONFIGFS_DIR) { @@ -1129,7 +1131,7 @@ int configfs_depend_item(struct configfs_subsystem *subsys, out_unlock_dirent_lock: spin_unlock(&configfs_dirent_lock); out_unlock_fs: - mutex_unlock(&configfs_sb->s_root->d_inode->i_mutex); + mutex_unlock(&root->d_inode->i_mutex); /* * If we succeeded, the fs is pinned via other methods. If not, @@ -1543,6 +1545,7 @@ static inline unsigned char dt_type(struct configfs_dirent *sd) static int configfs_readdir(struct file * filp, void * dirent, filldir_t filldir) { struct dentry *dentry = filp->f_path.dentry; + struct super_block *sb = dentry->d_sb; struct configfs_dirent * parent_sd = dentry->d_fsdata; struct configfs_dirent *cursor = filp->private_data; struct list_head *p, *q = &cursor->s_sibling; @@ -1605,7 +1608,7 @@ static int configfs_readdir(struct file * filp, void * dirent, filldir_t filldir ino = inode->i_ino; spin_unlock(&configfs_dirent_lock); if (!inode) - ino = iunique(configfs_sb, 2); + ino = iunique(sb, 2); if (filldir(dirent, name, len, filp->f_pos, ino, dt_type(next)) < 0) @@ -1677,6 +1680,7 @@ int configfs_register_subsystem(struct configfs_subsystem *subsys) struct config_group *group = &subsys->su_group; struct qstr name; struct dentry *dentry; + struct dentry *root; struct configfs_dirent *sd; err = configfs_pin_fs(); @@ -1686,18 +1690,18 @@ int configfs_register_subsystem(struct configfs_subsystem *subsys) if (!group->cg_item.ci_name) group->cg_item.ci_name = group->cg_item.ci_namebuf; - sd = configfs_sb->s_root->d_fsdata; + root = configfs_mount->mnt_root; + sd = root->d_fsdata; link_group(to_config_group(sd->s_element), group); - mutex_lock_nested(&configfs_sb->s_root->d_inode->i_mutex, - I_MUTEX_PARENT); + mutex_lock_nested(&root->d_inode->i_mutex, I_MUTEX_PARENT); name.name = group->cg_item.ci_name; name.len = strlen(name.name); name.hash = full_name_hash(name.name, name.len); err = -ENOMEM; - dentry = d_alloc(configfs_sb->s_root, &name); + dentry = d_alloc(root, &name); if (dentry) { d_add(dentry, NULL); @@ -1714,7 +1718,7 @@ int configfs_register_subsystem(struct configfs_subsystem *subsys) } } - mutex_unlock(&configfs_sb->s_root->d_inode->i_mutex); + mutex_unlock(&root->d_inode->i_mutex); if (err) { unlink_group(group); @@ -1728,13 +1732,14 @@ void configfs_unregister_subsystem(struct configfs_subsystem *subsys) { struct config_group *group = &subsys->su_group; struct dentry *dentry = group->cg_item.ci_dentry; + struct dentry *root = dentry->d_sb->s_root; - if (dentry->d_parent != configfs_sb->s_root) { + if (dentry->d_parent != root) { printk(KERN_ERR "configfs: Tried to unregister non-subsystem!\n"); return; } - mutex_lock_nested(&configfs_sb->s_root->d_inode->i_mutex, + mutex_lock_nested(&root->d_inode->i_mutex, I_MUTEX_PARENT); mutex_lock_nested(&dentry->d_inode->i_mutex, I_MUTEX_CHILD); mutex_lock(&configfs_symlink_mutex); @@ -1751,7 +1756,7 @@ void configfs_unregister_subsystem(struct configfs_subsystem *subsys) d_delete(dentry); - mutex_unlock(&configfs_sb->s_root->d_inode->i_mutex); + mutex_unlock(&root->d_inode->i_mutex); dput(dentry); diff --git a/fs/configfs/inode.c b/fs/configfs/inode.c index 3ee36d418863..8cf21ef902fc 100644 --- a/fs/configfs/inode.c +++ b/fs/configfs/inode.c @@ -44,8 +44,6 @@ static struct lock_class_key default_group_class[MAX_LOCK_DEPTH]; #endif -extern struct super_block * configfs_sb; - static const struct address_space_operations configfs_aops = { .readpage = simple_readpage, .write_begin = simple_write_begin, @@ -132,9 +130,10 @@ static inline void set_inode_attr(struct inode * inode, struct iattr * iattr) inode->i_ctime = iattr->ia_ctime; } -struct inode *configfs_new_inode(umode_t mode, struct configfs_dirent * sd) +struct inode *configfs_new_inode(umode_t mode, struct configfs_dirent *sd, + struct super_block *s) { - struct inode * inode = new_inode(configfs_sb); + struct inode * inode = new_inode(s); if (inode) { inode->i_ino = get_next_ino(); inode->i_mapping->a_ops = &configfs_aops; @@ -192,7 +191,7 @@ int configfs_create(struct dentry * dentry, umode_t mode, int (*init)(struct ino if (dentry) { if (!dentry->d_inode) { struct configfs_dirent *sd = dentry->d_fsdata; - if ((inode = configfs_new_inode(mode, sd))) { + if ((inode = configfs_new_inode(mode, sd, dentry->d_sb))) { if (dentry->d_parent && dentry->d_parent->d_inode) { struct inode *p_inode = dentry->d_parent->d_inode; p_inode->i_mtime = p_inode->i_ctime = CURRENT_TIME; diff --git a/fs/configfs/mount.c b/fs/configfs/mount.c index eb41adc28cfe..cc829fc85d77 100644 --- a/fs/configfs/mount.c +++ b/fs/configfs/mount.c @@ -38,7 +38,6 @@ #define CONFIGFS_MAGIC 0x62656570 struct vfsmount * configfs_mount = NULL; -struct super_block * configfs_sb = NULL; struct kmem_cache *configfs_dir_cachep; static int configfs_mnt_count = 0; @@ -77,10 +76,9 @@ static int configfs_fill_super(struct super_block *sb, void *data, int silent) sb->s_magic = CONFIGFS_MAGIC; sb->s_op = &configfs_ops; sb->s_time_gran = 1; - configfs_sb = sb; inode = configfs_new_inode(S_IFDIR | S_IRWXU | S_IRUGO | S_IXUGO, - &configfs_root); + &configfs_root, sb); if (inode) { inode->i_op = &configfs_root_inode_operations; inode->i_fop = &configfs_dir_operations; diff --git a/fs/configfs/symlink.c b/fs/configfs/symlink.c index 2817153d33c2..cc9f2546ea4a 100644 --- a/fs/configfs/symlink.c +++ b/fs/configfs/symlink.c @@ -110,13 +110,13 @@ out: static int get_target(const char *symname, struct path *path, - struct config_item **target) + struct config_item **target, struct super_block *sb) { int ret; ret = kern_path(symname, LOOKUP_FOLLOW|LOOKUP_DIRECTORY, path); if (!ret) { - if (path->dentry->d_sb == configfs_sb) { + if (path->dentry->d_sb == sb) { *target = configfs_get_config_item(path->dentry); if (!*target) { ret = -ENOENT; @@ -158,7 +158,7 @@ int configfs_symlink(struct inode *dir, struct dentry *dentry, const char *symna !type->ct_item_ops->allow_link) goto out_put; - ret = get_target(symname, &path, &target_item); + ret = get_target(symname, &path, &target_item, dentry->d_sb); if (ret) goto out_put; From 16d13b59b5b85ebc91de6c889716fa6e7766237f Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 17 Mar 2012 16:41:55 -0400 Subject: [PATCH 45/64] configfs: sanitize configfs_create() Signed-off-by: Al Viro --- fs/configfs/inode.c | 55 ++++++++++++++++++++++----------------------- 1 file changed, 27 insertions(+), 28 deletions(-) diff --git a/fs/configfs/inode.c b/fs/configfs/inode.c index 8cf21ef902fc..0074362d9f7f 100644 --- a/fs/configfs/inode.c +++ b/fs/configfs/inode.c @@ -187,36 +187,35 @@ static void configfs_set_inode_lock_class(struct configfs_dirent *sd, int configfs_create(struct dentry * dentry, umode_t mode, int (*init)(struct inode *)) { int error = 0; - struct inode * inode = NULL; - if (dentry) { - if (!dentry->d_inode) { - struct configfs_dirent *sd = dentry->d_fsdata; - if ((inode = configfs_new_inode(mode, sd, dentry->d_sb))) { - if (dentry->d_parent && dentry->d_parent->d_inode) { - struct inode *p_inode = dentry->d_parent->d_inode; - p_inode->i_mtime = p_inode->i_ctime = CURRENT_TIME; - } - configfs_set_inode_lock_class(sd, inode); - goto Proceed; - } - else - error = -ENOMEM; - } else - error = -EEXIST; - } else - error = -ENOENT; - goto Done; + struct inode *inode = NULL; + struct configfs_dirent *sd; + struct inode *p_inode; - Proceed: - if (init) + if (!dentry) + return -ENOENT; + + if (dentry->d_inode) + return -EEXIST; + + sd = dentry->d_fsdata; + inode = configfs_new_inode(mode, sd, dentry->d_sb); + if (!inode) + return -ENOMEM; + + p_inode = dentry->d_parent->d_inode; + p_inode->i_mtime = p_inode->i_ctime = CURRENT_TIME; + configfs_set_inode_lock_class(sd, inode); + + if (init) { error = init(inode); - if (!error) { - d_instantiate(dentry, inode); - if (S_ISDIR(mode) || S_ISLNK(mode)) - dget(dentry); /* pin link and directory dentries in core */ - } else - iput(inode); - Done: + if (error) { + iput(inode); + return error; + } + } + d_instantiate(dentry, inode); + if (S_ISDIR(mode) || S_ISLNK(mode)) + dget(dentry); /* pin link and directory dentries in core */ return error; } From 0dd6c08a0042ed83037cf5c772d9066e33046427 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 17 Mar 2012 16:49:20 -0400 Subject: [PATCH 46/64] configfs: configfs_create_dir() has parent dentry in dentry->d_parent no need to play sick games with parent item, internal mount, etc. Signed-off-by: Al Viro --- fs/configfs/dir.c | 20 +++++--------------- 1 file changed, 5 insertions(+), 15 deletions(-) diff --git a/fs/configfs/dir.c b/fs/configfs/dir.c index 54c59a7e37ce..6c560e77965c 100644 --- a/fs/configfs/dir.c +++ b/fs/configfs/dir.c @@ -264,11 +264,13 @@ static int init_symlink(struct inode * inode) return 0; } -static int create_dir(struct config_item * k, struct dentry * p, - struct dentry * d) +static int create_dir(struct config_item *k, struct dentry *d) { int error; umode_t mode = S_IFDIR| S_IRWXU | S_IRUGO | S_IXUGO; + struct dentry *p = d->d_parent; + + BUG_ON(!k); error = configfs_dirent_exists(p->d_fsdata, d->d_name.name); if (!error) @@ -304,19 +306,7 @@ static int create_dir(struct config_item * k, struct dentry * p, static int configfs_create_dir(struct config_item * item, struct dentry *dentry) { - struct dentry * parent; - int error = 0; - - BUG_ON(!item); - - if (item->ci_parent) - parent = item->ci_parent->ci_dentry; - else if (configfs_mount) - parent = configfs_mount->mnt_root; - else - return -EFAULT; - - error = create_dir(item,parent,dentry); + int error = create_dir(item, dentry); if (!error) item->ci_dentry = dentry; return error; From 2a152ad3a58508b06b9e0482e68117a79bbb27ce Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 17 Mar 2012 16:53:29 -0400 Subject: [PATCH 47/64] make configfs_pin_fs() return root dentry on success ... and make configfs_mnt static Signed-off-by: Al Viro --- fs/configfs/configfs_internal.h | 3 +-- fs/configfs/dir.c | 14 ++++++-------- fs/configfs/mount.c | 7 ++++--- 3 files changed, 11 insertions(+), 13 deletions(-) diff --git a/fs/configfs/configfs_internal.h b/fs/configfs/configfs_internal.h index 37121c2be110..b5f0a3b91f18 100644 --- a/fs/configfs/configfs_internal.h +++ b/fs/configfs/configfs_internal.h @@ -58,7 +58,6 @@ struct configfs_dirent { extern struct mutex configfs_symlink_mutex; extern spinlock_t configfs_dirent_lock; -extern struct vfsmount * configfs_mount; extern struct kmem_cache *configfs_dir_cachep; extern int configfs_is_root(struct config_item *item); @@ -80,7 +79,7 @@ extern const unsigned char * configfs_get_name(struct configfs_dirent *sd); extern void configfs_drop_dentry(struct configfs_dirent *sd, struct dentry *parent); extern int configfs_setattr(struct dentry *dentry, struct iattr *iattr); -extern int configfs_pin_fs(void); +extern struct dentry *configfs_pin_fs(void); extern void configfs_release_fs(void); extern struct rw_semaphore configfs_rename_sem; diff --git a/fs/configfs/dir.c b/fs/configfs/dir.c index 6c560e77965c..7e6c52d8a207 100644 --- a/fs/configfs/dir.c +++ b/fs/configfs/dir.c @@ -1075,16 +1075,15 @@ int configfs_depend_item(struct configfs_subsystem *subsys, * Pin the configfs filesystem. This means we can safely access * the root of the configfs filesystem. */ - ret = configfs_pin_fs(); - if (ret) - return ret; + root = configfs_pin_fs(); + if (IS_ERR(root)) + return PTR_ERR(root); /* * Next, lock the root directory. We're going to check that the * subsystem is really registered, and so we need to lock out * configfs_[un]register_subsystem(). */ - root = configfs_mount->mnt_root; mutex_lock(&root->d_inode->i_mutex); root_sd = root->d_fsdata; @@ -1673,14 +1672,13 @@ int configfs_register_subsystem(struct configfs_subsystem *subsys) struct dentry *root; struct configfs_dirent *sd; - err = configfs_pin_fs(); - if (err) - return err; + root = configfs_pin_fs(); + if (IS_ERR(root)) + return PTR_ERR(root); if (!group->cg_item.ci_name) group->cg_item.ci_name = group->cg_item.ci_namebuf; - root = configfs_mount->mnt_root; sd = root->d_fsdata; link_group(to_config_group(sd->s_element), group); diff --git a/fs/configfs/mount.c b/fs/configfs/mount.c index cc829fc85d77..aee0a7ebbd8e 100644 --- a/fs/configfs/mount.c +++ b/fs/configfs/mount.c @@ -37,7 +37,7 @@ /* Random magic number */ #define CONFIGFS_MAGIC 0x62656570 -struct vfsmount * configfs_mount = NULL; +static struct vfsmount *configfs_mount = NULL; struct kmem_cache *configfs_dir_cachep; static int configfs_mnt_count = 0; @@ -115,10 +115,11 @@ static struct file_system_type configfs_fs_type = { .kill_sb = kill_litter_super, }; -int configfs_pin_fs(void) +struct dentry *configfs_pin_fs(void) { - return simple_pin_fs(&configfs_fs_type, &configfs_mount, + int err = simple_pin_fs(&configfs_fs_type, &configfs_mount, &configfs_mnt_count); + return err ? ERR_PTR(err) : configfs_mount->mnt_root; } void configfs_release_fs(void) From 76bf09fcf76da1a2d777d9827d5a0a6a4cec0dcb Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 17 Mar 2012 18:14:34 -0400 Subject: [PATCH 48/64] jfs: mising cleanup on register_filesystem() failure Signed-off-by: Al Viro --- fs/jfs/super.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/fs/jfs/super.c b/fs/jfs/super.c index b3bb95504479..4a82950f412f 100644 --- a/fs/jfs/super.c +++ b/fs/jfs/super.c @@ -860,8 +860,14 @@ static int __init init_jfs_fs(void) jfs_proc_init(); #endif - return register_filesystem(&jfs_fs_type); + rc = register_filesystem(&jfs_fs_type); + if (!rc) + return 0; +#ifdef PROC_FS_JFS + jfs_proc_clean(); +#endif + kthread_stop(jfsSyncThread); kill_committask: for (i = 0; i < commit_threads; i++) kthread_stop(jfsCommitThread[i]); From 03e897a1edb9604c299e1a7646ade0ca90fe1e02 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 17 Mar 2012 18:16:24 -0400 Subject: [PATCH 49/64] logfs: missing cleanup on register_filesystem() failure Signed-off-by: Al Viro --- fs/logfs/super.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/fs/logfs/super.c b/fs/logfs/super.c index 7de18c3021fe..97bca623d893 100644 --- a/fs/logfs/super.c +++ b/fs/logfs/super.c @@ -626,7 +626,10 @@ static int __init logfs_init(void) if (ret) goto out2; - return register_filesystem(&logfs_fs_type); + ret = register_filesystem(&logfs_fs_type); + if (!ret) + return 0; + logfs_destroy_inode_cache(); out2: logfs_compr_exit(); out1: From f4c5499d2cbe54f0470764acfd0faf4e965aa7e9 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 17 Mar 2012 18:19:57 -0400 Subject: [PATCH 50/64] ntfs: forgets to unregister sysctls on register_filesystem() failure Signed-off-by: Al Viro --- fs/ntfs/super.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/fs/ntfs/super.c b/fs/ntfs/super.c index 22020d8b1ed2..3502a9e8f0a9 100644 --- a/fs/ntfs/super.c +++ b/fs/ntfs/super.c @@ -3159,6 +3159,8 @@ static int __init init_ntfs_fs(void) } printk(KERN_CRIT "NTFS: Failed to register NTFS filesystem driver!\n"); + /* Unregister the ntfs sysctls. */ + ntfs_sysctl(0); sysctl_err_out: kmem_cache_destroy(ntfs_big_inode_cache); big_inode_err_out: From 0794f569ec307dc25bbb12456ef75aa71f72f744 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 17 Mar 2012 21:29:13 -0400 Subject: [PATCH 51/64] ecryptfs: make register_filesystem() the last potential failure exit Signed-off-by: Al Viro --- fs/ecryptfs/main.c | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-) diff --git a/fs/ecryptfs/main.c b/fs/ecryptfs/main.c index 6e0e017e6932..68954937a071 100644 --- a/fs/ecryptfs/main.c +++ b/fs/ecryptfs/main.c @@ -794,15 +794,10 @@ static int __init ecryptfs_init(void) "Failed to allocate one or more kmem_cache objects\n"); goto out; } - rc = register_filesystem(&ecryptfs_fs_type); - if (rc) { - printk(KERN_ERR "Failed to register filesystem\n"); - goto out_free_kmem_caches; - } rc = do_sysfs_registration(); if (rc) { printk(KERN_ERR "sysfs registration failed\n"); - goto out_unregister_filesystem; + goto out_free_kmem_caches; } rc = ecryptfs_init_kthread(); if (rc) { @@ -823,19 +818,24 @@ static int __init ecryptfs_init(void) "rc = [%d]\n", rc); goto out_release_messaging; } + rc = register_filesystem(&ecryptfs_fs_type); + if (rc) { + printk(KERN_ERR "Failed to register filesystem\n"); + goto out_destroy_crypto; + } if (ecryptfs_verbosity > 0) printk(KERN_CRIT "eCryptfs verbosity set to %d. Secret values " "will be written to the syslog!\n", ecryptfs_verbosity); goto out; +out_destroy_crypto: + ecryptfs_destroy_crypto(); out_release_messaging: ecryptfs_release_messaging(); out_destroy_kthread: ecryptfs_destroy_kthread(); out_do_sysfs_unregistration: do_sysfs_unregistration(); -out_unregister_filesystem: - unregister_filesystem(&ecryptfs_fs_type); out_free_kmem_caches: ecryptfs_free_kmem_caches(); out: From 342827d7d19cb52b562bb3efeb4d4b672d008c35 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 17 Mar 2012 22:03:58 -0400 Subject: [PATCH 52/64] ocfs2: fix leaks on failure exits in module_init Signed-off-by: Al Viro --- fs/ocfs2/super.c | 40 +++++++++++++++++++++------------------- 1 file changed, 21 insertions(+), 19 deletions(-) diff --git a/fs/ocfs2/super.c b/fs/ocfs2/super.c index 337687c3e233..68f4541c2db9 100644 --- a/fs/ocfs2/super.c +++ b/fs/ocfs2/super.c @@ -1624,21 +1624,17 @@ static int __init ocfs2_init(void) init_waitqueue_head(&ocfs2__ioend_wq[i]); status = init_ocfs2_uptodate_cache(); - if (status < 0) { - mlog_errno(status); - goto leave; - } + if (status < 0) + goto out1; status = ocfs2_initialize_mem_caches(); - if (status < 0) { - mlog_errno(status); - goto leave; - } + if (status < 0) + goto out2; ocfs2_wq = create_singlethread_workqueue("ocfs2_wq"); if (!ocfs2_wq) { status = -ENOMEM; - goto leave; + goto out3; } ocfs2_debugfs_root = debugfs_create_dir("ocfs2", NULL); @@ -1650,17 +1646,23 @@ static int __init ocfs2_init(void) ocfs2_set_locking_protocol(); status = register_quota_format(&ocfs2_quota_format); -leave: - if (status < 0) { - ocfs2_free_mem_caches(); - exit_ocfs2_uptodate_cache(); - mlog_errno(status); - } + if (status < 0) + goto out4; + status = register_filesystem(&ocfs2_fs_type); + if (!status) + return 0; - if (status >= 0) { - return register_filesystem(&ocfs2_fs_type); - } else - return -1; + unregister_quota_format(&ocfs2_quota_format); +out4: + destroy_workqueue(ocfs2_wq); + debugfs_remove(ocfs2_debugfs_root); +out3: + ocfs2_free_mem_caches(); +out2: + exit_ocfs2_uptodate_cache(); +out1: + mlog_errno(status); + return status; } static void __exit ocfs2_exit(void) From 104bb37d3e0eb576b6aa61b9ed909e118e026bd2 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sun, 8 Jan 2012 16:13:28 -0500 Subject: [PATCH 53/64] gadgetfs: list_for_each_safe() misuse really weirdly spelled "while the list is non-empty, pick its first element, remove it from the list and free it" kind of loop... Signed-off-by: Al Viro --- drivers/usb/gadget/inode.c | 9 +++------ 1 file changed, 3 insertions(+), 6 deletions(-) diff --git a/drivers/usb/gadget/inode.c b/drivers/usb/gadget/inode.c index c95eea43b637..f701afb349ee 100644 --- a/drivers/usb/gadget/inode.c +++ b/drivers/usb/gadget/inode.c @@ -1569,20 +1569,18 @@ delegate: static void destroy_ep_files (struct dev_data *dev) { - struct list_head *entry, *tmp; - DBG (dev, "%s %d\n", __func__, dev->state); /* dev->state must prevent interference */ restart: spin_lock_irq (&dev->lock); - list_for_each_safe (entry, tmp, &dev->epfiles) { + while (!list_empty(&dev->epfiles)) { struct ep_data *ep; struct inode *parent; struct dentry *dentry; /* break link to FS */ - ep = list_entry (entry, struct ep_data, epfiles); + ep = list_first_entry (&dev->epfiles, struct ep_data, epfiles); list_del_init (&ep->epfiles); dentry = ep->dentry; ep->dentry = NULL; @@ -1605,8 +1603,7 @@ restart: dput (dentry); mutex_unlock (&parent->i_mutex); - /* fds may still be open */ - goto restart; + spin_lock_irq (&dev->lock); } spin_unlock_irq (&dev->lock); } From ea6fb41752dda2e450bd8a318c7400f8cfa9111e Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 23 Feb 2012 01:33:19 -0500 Subject: [PATCH 54/64] um: embed ->stub_pages[] into mmu_context seriously, kmalloc() for two-element array of pointers? Signed-off-by: Al Viro --- arch/um/include/asm/mmu.h | 2 +- arch/um/kernel/skas/mmu.c | 23 +++++------------------ 2 files changed, 6 insertions(+), 19 deletions(-) diff --git a/arch/um/include/asm/mmu.h b/arch/um/include/asm/mmu.h index 30509b9f37fd..53e8b498ebba 100644 --- a/arch/um/include/asm/mmu.h +++ b/arch/um/include/asm/mmu.h @@ -12,7 +12,7 @@ typedef struct mm_context { struct mm_id id; struct uml_arch_mm_context arch; - struct page **stub_pages; + struct page *stub_pages[2]; } mm_context_t; extern void __switch_mm(struct mm_id * mm_idp); diff --git a/arch/um/kernel/skas/mmu.c b/arch/um/kernel/skas/mmu.c index 1aee587e9c5d..db2ef422c1b8 100644 --- a/arch/um/kernel/skas/mmu.c +++ b/arch/um/kernel/skas/mmu.c @@ -92,8 +92,6 @@ int init_new_context(struct task_struct *task, struct mm_struct *mm) goto out_free; } - to_mm->stub_pages = NULL; - return 0; out_free: @@ -120,29 +118,20 @@ void arch_dup_mmap(struct mm_struct *oldmm, struct mm_struct *mm) if (ret) goto out; - pages = kmalloc(2 * sizeof(struct page *), GFP_KERNEL); - if (pages == NULL) { - printk(KERN_ERR "arch_dup_mmap failed to allocate 2 page " - "pointers\n"); - goto out; - } - - pages[0] = virt_to_page(&__syscall_stub_start); - pages[1] = virt_to_page(mm->context.id.stack); - mm->context.stub_pages = pages; + mm->context.stub_pages[0] = virt_to_page(&__syscall_stub_start); + mm->context.stub_pages[1] = virt_to_page(mm->context.id.stack); /* dup_mmap already holds mmap_sem */ err = install_special_mapping(mm, STUB_START, STUB_END - STUB_START, VM_READ | VM_MAYREAD | VM_EXEC | - VM_MAYEXEC | VM_DONTCOPY, pages); + VM_MAYEXEC | VM_DONTCOPY, + mm->context.stub_pages); if (err) { printk(KERN_ERR "install_special_mapping returned %d\n", err); - goto out_free; + goto out; } return; -out_free: - kfree(pages); out: force_sigsegv(SIGSEGV, current); } @@ -151,8 +140,6 @@ void arch_exit_mmap(struct mm_struct *mm) { pte_t *pte; - if (mm->context.stub_pages != NULL) - kfree(mm->context.stub_pages); pte = virt_to_pte(mm, STUB_CODE); if (pte != NULL) pte_clear(mm, STUB_CODE, pte); From ac2aca2817ca10ad30f86d0d3caa41ea8cc4f0f0 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 23 Feb 2012 01:37:19 -0500 Subject: [PATCH 55/64] um: uml_dup_mmap() relies on ->mmap_sem being held, but activate_mm() doesn't hold it ... while calling uml_dup_mmap() Signed-off-by: Al Viro --- arch/um/include/asm/mmu_context.h | 11 +++++++++-- arch/um/kernel/skas/mmu.c | 2 +- 2 files changed, 10 insertions(+), 3 deletions(-) diff --git a/arch/um/include/asm/mmu_context.h b/arch/um/include/asm/mmu_context.h index 591b3d8d7614..aa4a743dc4ab 100644 --- a/arch/um/include/asm/mmu_context.h +++ b/arch/um/include/asm/mmu_context.h @@ -9,7 +9,7 @@ #include #include -extern void arch_dup_mmap(struct mm_struct *oldmm, struct mm_struct *mm); +extern void uml_setup_stubs(struct mm_struct *mm); extern void arch_exit_mmap(struct mm_struct *mm); #define deactivate_mm(tsk,mm) do { } while (0) @@ -23,7 +23,9 @@ static inline void activate_mm(struct mm_struct *old, struct mm_struct *new) * when the new ->mm is used for the first time. */ __switch_mm(&new->context.id); - arch_dup_mmap(old, new); + down_write(&new->mmap_sem); + uml_setup_stubs(new); + up_write(&new->mmap_sem); } static inline void switch_mm(struct mm_struct *prev, struct mm_struct *next, @@ -39,6 +41,11 @@ static inline void switch_mm(struct mm_struct *prev, struct mm_struct *next, } } +static inline void arch_dup_mmap(struct mm_struct *oldmm, struct mm_struct *mm) +{ + uml_setup_stubs(mm); +} + static inline void enter_lazy_tlb(struct mm_struct *mm, struct task_struct *tsk) { diff --git a/arch/um/kernel/skas/mmu.c b/arch/um/kernel/skas/mmu.c index db2ef422c1b8..4947b319f53a 100644 --- a/arch/um/kernel/skas/mmu.c +++ b/arch/um/kernel/skas/mmu.c @@ -101,7 +101,7 @@ int init_new_context(struct task_struct *task, struct mm_struct *mm) return ret; } -void arch_dup_mmap(struct mm_struct *oldmm, struct mm_struct *mm) +void uml_setup_stubs(struct mm_struct *mm) { struct page **pages; int err, ret; From e59cc473cc603d562f2c80c12c943ef2a8cde6b2 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Wed, 7 Dec 2011 13:17:19 -0500 Subject: [PATCH 56/64] trim includes in inode.c Signed-off-by: Al Viro --- fs/inode.c | 12 +----------- 1 file changed, 1 insertion(+), 11 deletions(-) diff --git a/fs/inode.c b/fs/inode.c index 8b612813a6a7..9f4f5fecc096 100644 --- a/fs/inode.c +++ b/fs/inode.c @@ -2,29 +2,19 @@ * (C) 1997 Linus Torvalds * (C) 1999 Andrea Arcangeli (dynamic inode allocation) */ +#include #include #include -#include -#include -#include -#include -#include #include -#include -#include #include #include #include -#include #include #include #include #include -#include #include #include -#include -#include #include /* for inode_has_buffers */ #include #include "internal.h" From 19e5109fef2c368ab3f8a5157270f87f4a7c0326 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 23 Feb 2012 22:29:17 -0500 Subject: [PATCH 57/64] take removal of PF_FORKNOEXEC to flush_old_exec() Signed-off-by: Al Viro --- arch/x86/ia32/ia32_aout.c | 1 - fs/binfmt_aout.c | 1 - fs/binfmt_elf.c | 2 -- fs/binfmt_elf_fdpic.c | 3 --- fs/binfmt_flat.c | 1 - fs/binfmt_som.c | 1 - fs/exec.c | 2 +- 7 files changed, 1 insertion(+), 10 deletions(-) diff --git a/arch/x86/ia32/ia32_aout.c b/arch/x86/ia32/ia32_aout.c index cdfc8dc43670..4c2e59a420b9 100644 --- a/arch/x86/ia32/ia32_aout.c +++ b/arch/x86/ia32/ia32_aout.c @@ -323,7 +323,6 @@ static int load_aout_binary(struct linux_binprm *bprm, struct pt_regs *regs) } install_exec_creds(bprm); - current->flags &= ~PF_FORKNOEXEC; if (N_MAGIC(ex) == OMAGIC) { unsigned long text_addr, map_size; diff --git a/fs/binfmt_aout.c b/fs/binfmt_aout.c index a543364ba29b..4d5e6d26578c 100644 --- a/fs/binfmt_aout.c +++ b/fs/binfmt_aout.c @@ -267,7 +267,6 @@ static int load_aout_binary(struct linux_binprm * bprm, struct pt_regs * regs) } install_exec_creds(bprm); - current->flags &= ~PF_FORKNOEXEC; if (N_MAGIC(ex) == OMAGIC) { unsigned long text_addr, map_size; diff --git a/fs/binfmt_elf.c b/fs/binfmt_elf.c index f8ac4251877e..81878b78c9d4 100644 --- a/fs/binfmt_elf.c +++ b/fs/binfmt_elf.c @@ -712,7 +712,6 @@ static int load_elf_binary(struct linux_binprm *bprm, struct pt_regs *regs) goto out_free_dentry; /* OK, This is the point of no return */ - current->flags &= ~PF_FORKNOEXEC; current->mm->def_flags = def_flags; /* Do this immediately, since STACK_TOP as used in setup_arg_pages @@ -934,7 +933,6 @@ static int load_elf_binary(struct linux_binprm *bprm, struct pt_regs *regs) #endif /* ARCH_HAS_SETUP_ADDITIONAL_PAGES */ install_exec_creds(bprm); - current->flags &= ~PF_FORKNOEXEC; retval = create_elf_tables(bprm, &loc->elf_ex, load_addr, interp_load_addr); if (retval < 0) { diff --git a/fs/binfmt_elf_fdpic.c b/fs/binfmt_elf_fdpic.c index e7afcb67a2d3..c64bf5ee2df4 100644 --- a/fs/binfmt_elf_fdpic.c +++ b/fs/binfmt_elf_fdpic.c @@ -335,8 +335,6 @@ static int load_elf_fdpic_binary(struct linux_binprm *bprm, current->mm->context.exec_fdpic_loadmap = 0; current->mm->context.interp_fdpic_loadmap = 0; - current->flags &= ~PF_FORKNOEXEC; - #ifdef CONFIG_MMU elf_fdpic_arch_lay_out_mm(&exec_params, &interp_params, @@ -414,7 +412,6 @@ static int load_elf_fdpic_binary(struct linux_binprm *bprm, #endif install_exec_creds(bprm); - current->flags &= ~PF_FORKNOEXEC; if (create_elf_fdpic_tables(bprm, current->mm, &exec_params, &interp_params) < 0) goto error_kill; diff --git a/fs/binfmt_flat.c b/fs/binfmt_flat.c index 68affab88146..04f61f0bdfde 100644 --- a/fs/binfmt_flat.c +++ b/fs/binfmt_flat.c @@ -902,7 +902,6 @@ static int load_flat_binary(struct linux_binprm * bprm, struct pt_regs * regs) libinfo.lib_list[j].start_data:UNLOADED_LIB; install_exec_creds(bprm); - current->flags &= ~PF_FORKNOEXEC; set_binfmt(&flat_format); diff --git a/fs/binfmt_som.c b/fs/binfmt_som.c index ec15972dd98a..e4fc746629a7 100644 --- a/fs/binfmt_som.c +++ b/fs/binfmt_som.c @@ -225,7 +225,6 @@ load_som_binary(struct linux_binprm * bprm, struct pt_regs * regs) goto out_free; /* OK, This is the point of no return */ - current->flags &= ~PF_FORKNOEXEC; current->personality = PER_HPUX; setup_new_exec(bprm); diff --git a/fs/exec.c b/fs/exec.c index 2c5ae338773c..60478a0e7a37 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -1110,7 +1110,7 @@ int flush_old_exec(struct linux_binprm * bprm) bprm->mm = NULL; /* We're using it now */ set_fs(USER_DS); - current->flags &= ~(PF_RANDOMIZE | PF_KTHREAD); + current->flags &= ~(PF_RANDOMIZE | PF_FORKNOEXEC | PF_KTHREAD); flush_thread(); current->personality &= ~bprm->per_clear; From 8f82ecae53347ec65721e1c0683c3f26647e3c97 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Mon, 13 Feb 2012 20:57:12 -0500 Subject: [PATCH 58/64] qnx4: get rid of qnx4_bread/qnx4_getblk pointless, since the only caller will want the physical block number anyway; might as well call qnx4_block_map() and use sb_bread() Signed-off-by: Al Viro --- fs/qnx4/inode.c | 32 -------------------------------- fs/qnx4/namei.c | 5 +++-- fs/qnx4/qnx4.h | 2 -- 3 files changed, 3 insertions(+), 36 deletions(-) diff --git a/fs/qnx4/inode.c b/fs/qnx4/inode.c index 14a0ba0f0c1c..7900a813d90f 100644 --- a/fs/qnx4/inode.c +++ b/fs/qnx4/inode.c @@ -52,38 +52,6 @@ static int qnx4_remount(struct super_block *sb, int *flags, char *data) return 0; } -static struct buffer_head *qnx4_getblk(struct inode *inode, int nr, - int create) -{ - struct buffer_head *result = NULL; - - if ( nr >= 0 ) - nr = qnx4_block_map( inode, nr ); - if (nr) { - result = sb_getblk(inode->i_sb, nr); - return result; - } - return NULL; -} - -struct buffer_head *qnx4_bread(struct inode *inode, int block, int create) -{ - struct buffer_head *bh; - - bh = qnx4_getblk(inode, block, create); - if (!bh || buffer_uptodate(bh)) { - return bh; - } - ll_rw_block(READ, 1, &bh); - wait_on_buffer(bh); - if (buffer_uptodate(bh)) { - return bh; - } - brelse(bh); - - return NULL; -} - static int qnx4_get_block( struct inode *inode, sector_t iblock, struct buffer_head *bh, int create ) { unsigned long phys; diff --git a/fs/qnx4/namei.c b/fs/qnx4/namei.c index e8eb8784ed30..a512c0b30e8e 100644 --- a/fs/qnx4/namei.c +++ b/fs/qnx4/namei.c @@ -68,7 +68,9 @@ static struct buffer_head *qnx4_find_entry(int len, struct inode *dir, block = offset = blkofs = 0; while (blkofs * QNX4_BLOCK_SIZE + offset < dir->i_size) { if (!bh) { - bh = qnx4_bread(dir, blkofs, 0); + block = qnx4_block_map(dir, blkofs); + if (block) + bh = sb_bread(dir->i_sb, block); if (!bh) { blkofs++; continue; @@ -76,7 +78,6 @@ static struct buffer_head *qnx4_find_entry(int len, struct inode *dir, } *res_dir = (struct qnx4_inode_entry *) (bh->b_data + offset); if (qnx4_match(len, name, bh, &offset)) { - block = qnx4_block_map( dir, blkofs ); *ino = block * QNX4_INODES_PER_BLOCK + (offset / QNX4_DIR_ENTRY_SIZE) - 1; return bh; diff --git a/fs/qnx4/qnx4.h b/fs/qnx4/qnx4.h index 33a60858203b..244d4620189b 100644 --- a/fs/qnx4/qnx4.h +++ b/fs/qnx4/qnx4.h @@ -27,8 +27,6 @@ extern struct dentry *qnx4_lookup(struct inode *dir, struct dentry *dentry, stru extern unsigned long qnx4_count_free_blocks(struct super_block *sb); extern unsigned long qnx4_block_map(struct inode *inode, long iblock); -extern struct buffer_head *qnx4_bread(struct inode *, int, int); - extern const struct inode_operations qnx4_dir_inode_operations; extern const struct file_operations qnx4_dir_operations; extern int qnx4_is_free(struct super_block *sb, long block); From 7cd916f6ea86e8538b1e8136847dfb941813fe94 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Tue, 14 Feb 2012 17:48:20 -0500 Subject: [PATCH 59/64] qnx4: new helper - try_extent() checking if an extent is the one we are looking for is done twice in qnx4_block_map(); gather that code into a helper function. Signed-off-by: Al Viro --- fs/qnx4/inode.c | 23 +++++++++++++++-------- 1 file changed, 15 insertions(+), 8 deletions(-) diff --git a/fs/qnx4/inode.c b/fs/qnx4/inode.c index 7900a813d90f..552e994e3aa1 100644 --- a/fs/qnx4/inode.c +++ b/fs/qnx4/inode.c @@ -66,23 +66,31 @@ static int qnx4_get_block( struct inode *inode, sector_t iblock, struct buffer_h return 0; } +static inline u32 try_extent(qnx4_xtnt_t *extent, u32 *offset) +{ + u32 size = le32_to_cpu(extent->xtnt_size); + if (*offset < size) + return le32_to_cpu(extent->xtnt_blk) + *offset - 1; + *offset -= size; + return 0; +} + unsigned long qnx4_block_map( struct inode *inode, long iblock ) { int ix; - long offset, i_xblk; - unsigned long block = 0; + long i_xblk; struct buffer_head *bh = NULL; struct qnx4_xblk *xblk = NULL; struct qnx4_inode_entry *qnx4_inode = qnx4_raw_inode(inode); u16 nxtnt = le16_to_cpu(qnx4_inode->di_num_xtnts); + u32 offset = iblock; + u32 block = try_extent(&qnx4_inode->di_first_xtnt, &offset); - if ( iblock < le32_to_cpu(qnx4_inode->di_first_xtnt.xtnt_size) ) { + if (block) { // iblock is in the first extent. This is easy. - block = le32_to_cpu(qnx4_inode->di_first_xtnt.xtnt_blk) + iblock - 1; } else { // iblock is beyond first extent. We have to follow the extent chain. i_xblk = le32_to_cpu(qnx4_inode->di_xblk); - offset = iblock - le32_to_cpu(qnx4_inode->di_first_xtnt.xtnt_size); ix = 0; while ( --nxtnt > 0 ) { if ( ix == 0 ) { @@ -98,12 +106,11 @@ unsigned long qnx4_block_map( struct inode *inode, long iblock ) return -EIO; } } - if ( offset < le32_to_cpu(xblk->xblk_xtnts[ix].xtnt_size) ) { + block = try_extent(&xblk->xblk_xtnts[ix], &offset); + if (block) { // got it! - block = le32_to_cpu(xblk->xblk_xtnts[ix].xtnt_blk) + offset - 1; break; } - offset -= le32_to_cpu(xblk->xblk_xtnts[ix].xtnt_size); if ( ++ix >= xblk->xblk_num_xtnts ) { i_xblk = le32_to_cpu(xblk->xblk_next_xblk); ix = 0; From f3922382ce930e76773fb06416a7a6081a8702ad Mon Sep 17 00:00:00 2001 From: Matthew Garrett Date: Tue, 13 Mar 2012 12:10:34 -0400 Subject: [PATCH 60/64] hfsplus: initialise userflags The userflags field was being written to the filesystem without being initialised. Make sure it's clear, since otherwise files end up with garbage attributes. Signed-off-by: Matthew Garrett Signed-off-by: Christoph Hellwig Signed-off-by: Al Viro --- fs/hfsplus/inode.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/fs/hfsplus/inode.c b/fs/hfsplus/inode.c index 6643b242bdd7..82b69ee4dacc 100644 --- a/fs/hfsplus/inode.c +++ b/fs/hfsplus/inode.c @@ -193,6 +193,7 @@ static struct dentry *hfsplus_file_lookup(struct inode *dir, mutex_init(&hip->extents_lock); hip->extent_state = 0; hip->flags = 0; + hip->userflags = 0; set_bit(HFSPLUS_I_RSRC, &hip->flags); err = hfs_find_init(HFSPLUS_SB(sb)->cat_tree, &fd); @@ -400,6 +401,7 @@ struct inode *hfsplus_new_inode(struct super_block *sb, umode_t mode) atomic_set(&hip->opencnt, 0); hip->extent_state = 0; hip->flags = 0; + hip->userflags = 0; memset(hip->first_extents, 0, sizeof(hfsplus_extent_rec)); memset(hip->cached_extents, 0, sizeof(hfsplus_extent_rec)); hip->alloc_blocks = 0; From 0347b6e95ceeb648631f89a0ced001e90e4daaa9 Mon Sep 17 00:00:00 2001 From: Matthew Garrett Date: Thu, 2 Feb 2012 15:39:50 -0500 Subject: [PATCH 61/64] hfsplus: change finder_info to u32 The finder_info block in the hfsplus volume header is currently defined as an array of 8 bit values, but TN1150 defines it as being an array of 32 bit values. Fix for convenience. Signed-off-by: Matthew Garrett Signed-off-by: Christoph Hellwig Signed-off-by: Al Viro --- fs/hfsplus/hfsplus_raw.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/hfsplus/hfsplus_raw.h b/fs/hfsplus/hfsplus_raw.h index 927cdd6d5bf5..921967e5abb1 100644 --- a/fs/hfsplus/hfsplus_raw.h +++ b/fs/hfsplus/hfsplus_raw.h @@ -117,7 +117,7 @@ struct hfsplus_vh { __be32 write_count; __be64 encodings_bmp; - u8 finder_info[32]; + u32 finder_info[8]; struct hfsplus_fork_raw alloc_file; struct hfsplus_fork_raw ext_file; From a051f71ce97c53cde3ac64de64eb02d658d9308e Mon Sep 17 00:00:00 2001 From: Matthew Garrett Date: Mon, 6 Feb 2012 15:14:40 -0500 Subject: [PATCH 62/64] hfsplus: add an ioctl to bless files Making an hfsplus partition bootable requires the ability to "bless" a file by putting its inode number in the volume header. Doing this from userspace on a mounted filesystem is impractical since the kernel will write back the original values on unmount. Add an ioctl to allow userspace to update the volume header information based on the target file. Signed-off-by: Matthew Garrett Signed-off-by: Christoph Hellwig Signed-off-by: Al Viro --- Documentation/ioctl/ioctl-number.txt | 1 + fs/hfsplus/hfsplus_fs.h | 5 ++++ fs/hfsplus/ioctl.c | 34 ++++++++++++++++++++++++++++ 3 files changed, 40 insertions(+) diff --git a/Documentation/ioctl/ioctl-number.txt b/Documentation/ioctl/ioctl-number.txt index 4840334ea97b..37a424868eea 100644 --- a/Documentation/ioctl/ioctl-number.txt +++ b/Documentation/ioctl/ioctl-number.txt @@ -218,6 +218,7 @@ Code Seq#(hex) Include File Comments 'h' 00-7F conflict! Charon filesystem 'h' 00-1F linux/hpet.h conflict! +'h' 80-8F fs/hfsplus/ioctl.c 'i' 00-3F linux/i2o-dev.h conflict! 'i' 0B-1F linux/ipmi.h conflict! 'i' 80-8F linux/i8k.h diff --git a/fs/hfsplus/hfsplus_fs.h b/fs/hfsplus/hfsplus_fs.h index 21a5b7fc6db4..4e75ac646fea 100644 --- a/fs/hfsplus/hfsplus_fs.h +++ b/fs/hfsplus/hfsplus_fs.h @@ -316,6 +316,11 @@ static inline unsigned short hfsplus_min_io_size(struct super_block *sb) #define HFSPLUS_IOC_EXT2_SETFLAGS FS_IOC_SETFLAGS +/* + * hfs+-specific ioctl for making the filesystem bootable + */ +#define HFSPLUS_IOC_BLESS _IO('h', 0x80) + /* * Functions in any *.c used in other files */ diff --git a/fs/hfsplus/ioctl.c b/fs/hfsplus/ioctl.c index f66c7655b3f7..c640ba57074b 100644 --- a/fs/hfsplus/ioctl.c +++ b/fs/hfsplus/ioctl.c @@ -20,6 +20,38 @@ #include #include "hfsplus_fs.h" +/* + * "Blessing" an HFS+ filesystem writes metadata to the superblock informing + * the platform firmware which file to boot from + */ +static int hfsplus_ioctl_bless(struct file *file, int __user *user_flags) +{ + struct dentry *dentry = file->f_path.dentry; + struct inode *inode = dentry->d_inode; + struct hfsplus_sb_info *sbi = HFSPLUS_SB(inode->i_sb); + struct hfsplus_vh *vh = sbi->s_vhdr; + struct hfsplus_vh *bvh = sbi->s_backup_vhdr; + + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; + + mutex_lock(&sbi->vh_mutex); + + /* Directory containing the bootable system */ + vh->finder_info[0] = bvh->finder_info[0] = + cpu_to_be32(parent_ino(dentry)); + + /* Bootloader */ + vh->finder_info[1] = bvh->finder_info[1] = cpu_to_be32(inode->i_ino); + + /* Per spec, the OS X system folder - same as finder_info[0] here */ + vh->finder_info[5] = bvh->finder_info[5] = + cpu_to_be32(parent_ino(dentry)); + + mutex_unlock(&sbi->vh_mutex); + return 0; +} + static int hfsplus_ioctl_getflags(struct file *file, int __user *user_flags) { struct inode *inode = file->f_path.dentry->d_inode; @@ -108,6 +140,8 @@ long hfsplus_ioctl(struct file *file, unsigned int cmd, unsigned long arg) return hfsplus_ioctl_getflags(file, argp); case HFSPLUS_IOC_EXT2_SETFLAGS: return hfsplus_ioctl_setflags(file, argp); + case HFSPLUS_IOC_BLESS: + return hfsplus_ioctl_bless(file, argp); default: return -ENOTTY; } From 88187398cc5fa6650f38b9dcd5464667f468888f Mon Sep 17 00:00:00 2001 From: Al Viro Date: Tue, 20 Mar 2012 06:00:24 -0400 Subject: [PATCH 63/64] debugfs-related mode_t whack-a-mole all of those should be umode_t... Signed-off-by: Al Viro --- Documentation/filesystems/debugfs.txt | 2 +- drivers/mmc/card/block.c | 2 +- drivers/net/ethernet/brocade/bna/bnad_debugfs.c | 2 +- fs/debugfs/file.c | 2 +- fs/nfsd/fault_inject.c | 2 +- include/linux/debugfs.h | 4 ++-- 6 files changed, 7 insertions(+), 7 deletions(-) diff --git a/Documentation/filesystems/debugfs.txt b/Documentation/filesystems/debugfs.txt index 6872c91bce35..41865d62e5b9 100644 --- a/Documentation/filesystems/debugfs.txt +++ b/Documentation/filesystems/debugfs.txt @@ -133,7 +133,7 @@ file. void __iomem *base; }; - struct dentry *debugfs_create_regset32(const char *name, mode_t mode, + struct dentry *debugfs_create_regset32(const char *name, umode_t mode, struct dentry *parent, struct debugfs_regset32 *regset); diff --git a/drivers/mmc/card/block.c b/drivers/mmc/card/block.c index c6a383d0244d..e5a3c7b6dedb 100644 --- a/drivers/mmc/card/block.c +++ b/drivers/mmc/card/block.c @@ -1685,7 +1685,7 @@ static int mmc_add_disk(struct mmc_blk_data *md) if ((md->area_type & MMC_BLK_DATA_AREA_BOOT) && card->ext_csd.boot_ro_lockable) { - mode_t mode; + umode_t mode; if (card->ext_csd.boot_ro_lock & EXT_CSD_BOOT_WP_B_PWR_WP_DIS) mode = S_IRUGO; diff --git a/drivers/net/ethernet/brocade/bna/bnad_debugfs.c b/drivers/net/ethernet/brocade/bna/bnad_debugfs.c index 592ad3929f53..df18a8337624 100644 --- a/drivers/net/ethernet/brocade/bna/bnad_debugfs.c +++ b/drivers/net/ethernet/brocade/bna/bnad_debugfs.c @@ -531,7 +531,7 @@ static const struct file_operations bnad_debugfs_op_drvinfo = { struct bnad_debugfs_entry { const char *name; - mode_t mode; + umode_t mode; const struct file_operations *fops; }; diff --git a/fs/debugfs/file.c b/fs/debugfs/file.c index ef023eef0464..21e93605161c 100644 --- a/fs/debugfs/file.c +++ b/fs/debugfs/file.c @@ -611,7 +611,7 @@ static const struct file_operations fops_regset32 = { * %NULL or !%NULL instead as to eliminate the need for #ifdef in the calling * code. */ -struct dentry *debugfs_create_regset32(const char *name, mode_t mode, +struct dentry *debugfs_create_regset32(const char *name, umode_t mode, struct dentry *parent, struct debugfs_regset32 *regset) { diff --git a/fs/nfsd/fault_inject.c b/fs/nfsd/fault_inject.c index ce7f0758d84c..9559ce468732 100644 --- a/fs/nfsd/fault_inject.c +++ b/fs/nfsd/fault_inject.c @@ -72,7 +72,7 @@ int nfsd_fault_inject_init(void) { unsigned int i; struct nfsd_fault_inject_op *op; - mode_t mode = S_IFREG | S_IRUSR | S_IWUSR; + umode_t mode = S_IFREG | S_IRUSR | S_IWUSR; debug_dir = debugfs_create_dir("nfsd", NULL); if (!debug_dir) diff --git a/include/linux/debugfs.h b/include/linux/debugfs.h index 6169c26fd8c8..ae36b72c22f3 100644 --- a/include/linux/debugfs.h +++ b/include/linux/debugfs.h @@ -86,7 +86,7 @@ struct dentry *debugfs_create_blob(const char *name, umode_t mode, struct dentry *parent, struct debugfs_blob_wrapper *blob); -struct dentry *debugfs_create_regset32(const char *name, mode_t mode, +struct dentry *debugfs_create_regset32(const char *name, umode_t mode, struct dentry *parent, struct debugfs_regset32 *regset); @@ -208,7 +208,7 @@ static inline struct dentry *debugfs_create_blob(const char *name, umode_t mode, } static inline struct dentry *debugfs_create_regset32(const char *name, - mode_t mode, struct dentry *parent, + umode_t mode, struct dentry *parent, struct debugfs_regset32 *regset) { return ERR_PTR(-ENODEV); From 07c0c5d8b8c122b2f2df9ee574ac3083daefc981 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Tue, 20 Mar 2012 22:05:02 -0400 Subject: [PATCH 64/64] ext4: initialization of ext4_li_mtx needs to be done earlier Signed-off-by: Al Viro --- fs/ext4/super.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/fs/ext4/super.c b/fs/ext4/super.c index d2baea7bcf30..933900909ed0 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -5055,6 +5055,9 @@ static int __init ext4_init_fs(void) { int i, err; + ext4_li_info = NULL; + mutex_init(&ext4_li_mtx); + ext4_check_flag_values(); for (i = 0; i < EXT4_WQ_HASH_SZ; i++) { @@ -5093,8 +5096,6 @@ static int __init ext4_init_fs(void) if (err) goto out; - ext4_li_info = NULL; - mutex_init(&ext4_li_mtx); return 0; out: unregister_as_ext2();