Re: [PATCH 7/7] md/raid5: fix handling of degraded stripes in batches.

[NeilBrown <neilb@suse.de>]

[-- Attachment #1: Type: text/plain, Size: 2327 bytes --]

On Fri, 8 May 2015 12:12:23 -0700 Shaohua Li <shli@kernel.org> wrote:

> On Fri, May 08, 2015 at 06:56:12PM +1000, NeilBrown wrote:
> > There is no need for special handling of stripe-batches when the array
> > is degraded.
> > 
> > There may be if there is a failure in the batch, but STRIPE_DEGRADED
> > does not imply an error.
> > 
> > So don't set STRIPE_BATCH_ERR in ops_run_io just because the array is
> > degraded.
> > This actually causes a bug: the STRIPE_DEGRADED flag gets cleared in
> > check_break_stripe_batch_list() and so the bitmap bit gets cleared
> > when it shouldn't.
> > 
> > So in check_break_stripe_batch_list(), split the batch up completely -
> > again STRIPE_DEGRADED isn't meaningful.
> > 
> > Also don't set STRIPE_BATCH_ERR when there is a write error to a
> > replacement device.  This simply removes the replacement device and
> > requires no extra handling.
> > 
> > Signed-off-by: NeilBrown <neilb@suse.de>
> > ---
> >  drivers/md/raid5.c |   17 +++--------------
> >  1 file changed, 3 insertions(+), 14 deletions(-)
> > 
> > diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c
> > index 3873eaa6fa2e..1ba97fdc6df1 100644
> > --- a/drivers/md/raid5.c
> > +++ b/drivers/md/raid5.c
> > @@ -1078,9 +1078,6 @@ again:
> >  			pr_debug("skip op %ld on disc %d for sector %llu\n",
> >  				bi->bi_rw, i, (unsigned long long)sh->sector);
> >  			clear_bit(R5_LOCKED, &sh->dev[i].flags);
> > -			if (sh->batch_head)
> > -				set_bit(STRIPE_BATCH_ERR,
> > -					&sh->batch_head->state);
> >  			set_bit(STRIPE_HANDLE, &sh->state);
> >  		}
> 
> Patches look good to me. I had a question here. Is it possible some stripes in
> a batch become degraded here but some not? Seems possible, then the batch
> should be splitted too.

Why?

I don't really understand the purpose of splitting up the batch.
The only possible error handling on a full-stripe write is:
 - fail a device, or
 - record a bad-block.

The first case affects all stripes in a batch equally so there is no need to
split it up.
The second case it is probably best to record the bad blocks while iterating
through the batch in handle_stripe_clean_event().

What exactly do you expect to happen after the stripes in a batch after they
have been split up?

Thanks,
NeilBrown

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 811 bytes --]