2008-07-25 09:13:29

by Andrew Morton

[permalink] [raw]
Subject: [patch 061/277] ext3: fix synchronization of quota files in journal=data mode

From: Jan Kara <[email protected]>

In journal=data mode, it is not enough to do write_inode_now as done in
vfs_quota_on() to write all data to their final location (which is needed for
quota_read to work correctly). Calling journal_flush() does its job.

Reported-by: Nick <[email protected]>
Cc: <[email protected]>
Signed-off-by: Jan Kara <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
---

fs/ext3/super.c | 35 +++++++++++++++++++++++++++--------
1 file changed, 27 insertions(+), 8 deletions(-)

diff -puN fs/ext3/super.c~ext3-fix-synchronization-of-quota-files-in-journal=data-mode fs/ext3/super.c
--- a/fs/ext3/super.c~ext3-fix-synchronization-of-quota-files-in-journal=data-mode
+++ a/fs/ext3/super.c
@@ -2759,23 +2759,42 @@ static int ext3_quota_on(struct super_bl

if (!test_opt(sb, QUOTA))
return -EINVAL;
- /* Not journalling quota or remount? */
- if ((!EXT3_SB(sb)->s_qf_names[USRQUOTA] &&
- !EXT3_SB(sb)->s_qf_names[GRPQUOTA]) || remount)
+ /* When remounting, no checks are needed and in fact, path is NULL */
+ if (remount)
return vfs_quota_on(sb, type, format_id, path, remount);
+
err = path_lookup(path, LOOKUP_FOLLOW, &nd);
if (err)
return err;
+
/* Quotafile not on the same filesystem? */
if (nd.path.mnt->mnt_sb != sb) {
path_put(&nd.path);
return -EXDEV;
}
- /* Quotafile not in fs root? */
- if (nd.path.dentry->d_parent->d_inode != sb->s_root->d_inode)
- printk(KERN_WARNING
- "EXT3-fs: Quota file not on filesystem root. "
- "Journalled quota will not work.\n");
+ /* Journaling quota? */
+ if (EXT3_SB(sb)->s_qf_names[type]) {
+ /* Quotafile not of fs root? */
+ if (nd.path.dentry->d_parent->d_inode != sb->s_root->d_inode)
+ printk(KERN_WARNING
+ "EXT3-fs: Quota file not on filesystem root. "
+ "Journaled quota will not work.\n");
+ }
+
+ /*
+ * When we journal data on quota file, we have to flush journal to see
+ * all updates to the file when we bypass pagecache...
+ */
+ if (ext3_should_journal_data(nd.path.dentry->d_inode)) {
+ /*
+ * We don't need to lock updates but journal_flush() could
+ * otherwise be livelocked...
+ */
+ journal_lock_updates(EXT3_SB(sb)->s_journal);
+ journal_flush(EXT3_SB(sb)->s_journal);
+ journal_unlock_updates(EXT3_SB(sb)->s_journal);
+ }
+
path_put(&nd.path);
return vfs_quota_on(sb, type, format_id, path, remount);
}
_