On Thu 31-05-12 16:18:10, Dmitry Monakhov wrote: > > If remount failed due to some reason we have to fallback quota to it's initial state > otherwise it will state in suspended state which is incorrect. > > #TEST_CASE > dd if=/dev/zero of=/tmp/imgfile bs=1M count=100 > mkfs.ext4 -F /tmp/imgfile > mount /tmp/imgfile /mnt -oloop,quota > quotacheck -cug /mnt > quotaon /mnt > # remount to RO state , and explicitly provide bad option > mount /mnt -oro,remount,some-bad-option > # At this point fs is in broken state because it is RW, but quota is in > # suspended state. Later activity will trigger NULL pointer dereference > mount /mnt -oro,remount,some-bad-option > > Bug was silently fixed in: e0ccfd959cd890 v2.6.34-7101-ge0ccfd9 while migrating > to newer quota approach. All previous kernel versions are affected. > > https://jira.sw.ru/browse/PSBM-13231 > > Signed-off-by: Dmitry Monakhov <dmonakhov@xxxxxxxxxx> The fix looks good so you can add: Acked-by: Jan Kara <jack@xxxxxxx> Honza > --- > fs/super.c | 6 +++++- > 1 files changed, 5 insertions(+), 1 deletions(-) > > diff --git a/fs/super.c b/fs/super.c > index 1fb63e3..e20367b 100644 > --- a/fs/super.c > +++ b/fs/super.c > @@ -607,8 +607,12 @@ int do_remount_sb(struct super_block *sb, int flags, void *data, int force) > > if (sb->s_op->remount_fs) { > retval = sb->s_op->remount_fs(sb, &flags, data); > - if (retval) > + if (retval) { > + /* Remount failed, fallback quota to original state */ > + if (remount_ro) > + vfs_dq_quota_on_remount(sb); > return retval; > + } > } > sb->s_flags = (sb->s_flags & ~MS_RMT_MASK) | (flags & MS_RMT_MASK); > if (remount_rw) -- Jan Kara <jack@xxxxxxx> SUSE Labs, CR -- To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
