Re: [PATCH v2] Btrfs: remove BUG() in raid56

[Liu Bo <bo.li.liu@oracle.com>]

On Tue, Jul 26, 2016 at 06:58:42PM +0200, David Sterba wrote:
> On Wed, Jun 29, 2016 at 05:57:56PM -0700, Liu Bo wrote:
> > This BUG() has been triggered by a fuzz testing image, but in fact
> > btrfs can handle this gracefully by returning -EIO.
> > 
> > Thus, use btrfs_warn to give us more debugging information than a 
> > single BUG() and return error properly.
> > 
> > Signed-off-by: Liu Bo <bo.li.liu@oracle.com>
> > ---
> > v2: - use btrfs_warn with more debugging information instead of WARN_ONCE.
> >     - change the patch title.
> > 
> >  fs/btrfs/raid56.c | 5 ++++-
> >  1 file changed, 4 insertions(+), 1 deletion(-)
> > 
> > diff --git a/fs/btrfs/raid56.c b/fs/btrfs/raid56.c
> > index f8b6d41..5f4712c 100644
> > --- a/fs/btrfs/raid56.c
> > +++ b/fs/btrfs/raid56.c
> > @@ -2139,7 +2139,10 @@ int raid56_parity_recover(struct btrfs_root *root, struct bio *bio,
> >  
> >  	rbio->faila = find_logical_bio_stripe(rbio, bio);
> >  	if (rbio->faila == -1) {
> > -		BUG();
> > +		btrfs_warn(root->fs_info,
> > +	"rbio->faila is -1: (bio has logical %llu len %llu, bbio has map_type %llu)",
> 
> That's rather cryptic message for a casual user, can it be prepended by
> a short summary what actually happened? Like "bad stripe for parity" or
> whatever seems more appropriate to you. Also the changelog could
> describe the error condition.

Good point, I'll update it.

Thanks,

-liubo