Re: [PATCH 1/3] vfs: Do not ignore return code from s_op->sync_fs

[Vivek Goyal]

On Tue, Dec 22, 2020 at 12:23:11PM +1100, NeilBrown wrote:

[...]
> > diff --git a/fs/sync.c b/fs/sync.c
> > index 1373a610dc78..b5fb83a734cd 100644
> > --- a/fs/sync.c
> > +++ b/fs/sync.c
> > @@ -30,14 +30,18 @@
> >   */
> >  static int __sync_filesystem(struct super_block *sb, int wait)
> >  {
> > +	int ret, ret2;
> > +
> >  	if (wait)
> >  		sync_inodes_sb(sb);
> >  	else
> >  		writeback_inodes_sb(sb, WB_REASON_SYNC);
> >  
> >  	if (sb->s_op->sync_fs)
> > -		sb->s_op->sync_fs(sb, wait);
> > -	return __sync_blockdev(sb->s_bdev, wait);
> > +		ret = sb->s_op->sync_fs(sb, wait);
> > +	ret2 = __sync_blockdev(sb->s_bdev, wait);
> > +
> > +	return ret ? ret : ret2;
> 
> I'm surprised that the compiler didn't complain that 'ret' might be used
> uninitialized.

Indeed. That "ret" can be used uninitialized. Here is the fixed patch.


Subject: vfs: Do not ignore return code from s_op->sync_fs

Current implementation of __sync_filesystem() ignores the
return code from ->sync_fs(). I am not sure why that's the case.

Ignoring ->sync_fs() return code is problematic for overlayfs where
it can return error if sync_filesystem() on upper super block failed.
That error will simply be lost and sycnfs(overlay_fd), will get
success (despite the fact it failed).

Al Viro noticed that there are other filesystems which can sometimes
return error in ->sync_fs() and these errors will be ignored too.

fs/btrfs/super.c:2412:  .sync_fs        = btrfs_sync_fs,
fs/exfat/super.c:204:   .sync_fs        = exfat_sync_fs,
fs/ext4/super.c:1674:   .sync_fs        = ext4_sync_fs,
fs/f2fs/super.c:2480:   .sync_fs        = f2fs_sync_fs,
fs/gfs2/super.c:1600:   .sync_fs                = gfs2_sync_fs,
fs/hfsplus/super.c:368: .sync_fs        = hfsplus_sync_fs,
fs/nilfs2/super.c:689:  .sync_fs        = nilfs_sync_fs,
fs/ocfs2/super.c:139:   .sync_fs        = ocfs2_sync_fs,
fs/overlayfs/super.c:399:	.sync_fs        = ovl_sync_fs,
fs/ubifs/super.c:2052:  .sync_fs       = ubifs_sync_fs,

Hence, this patch tries to fix it and capture error returned
by ->sync_fs() and return to caller. I am specifically interested
in syncfs() path and return error to user.

I am assuming that we want to continue to call __sync_blockdev()
despite the fact that there have been errors reported from
->sync_fs(). So this patch continues to call __sync_blockdev()
even if ->sync_fs() returns an error.

Al noticed that there are few other callsites where ->sync_fs() error
code is being ignored. 

sync_fs_one_sb(): For this it seems desirable to ignore the return code.

dquot_disable(): Jan Kara mentioned that ignoring return code here is fine
		 because we don't want to fail dquot_disable() just beacuse
		 caches might be incoherent.

dquot_quota_sync(): Jan thinks that it might make some sense to capture
		    return code here. But I am leaving it untouched for
		   now. When somebody needs it, they can easily fix it.

Signed-off-by: Vivek Goyal <vgoyal@xxxxxxxxxx>
---
 fs/sync.c |    8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

Index: redhat-linux/fs/sync.c
===================================================================
--- redhat-linux.orig/fs/sync.c	2020-12-22 09:56:04.543483440 -0500
+++ redhat-linux/fs/sync.c	2020-12-22 10:01:28.560483440 -0500
@@ -30,14 +30,18 @@
  */
 static int __sync_filesystem(struct super_block *sb, int wait)
 {
+	int ret = 0, ret2;
+
 	if (wait)
 		sync_inodes_sb(sb);
 	else
 		writeback_inodes_sb(sb, WB_REASON_SYNC);
 
 	if (sb->s_op->sync_fs)
-		sb->s_op->sync_fs(sb, wait);
-	return __sync_blockdev(sb->s_bdev, wait);
+		ret = sb->s_op->sync_fs(sb, wait);
+	ret2 = __sync_blockdev(sb->s_bdev, wait);
+
+	return ret ? ret : ret2;
 }
 
 /*