linux-xfs.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* Re: yet another approach to fix loop autoclear for xfstets xfs/049
       [not found] <20220126155040.1190842-1-hch@lst.de>
@ 2022-01-26 19:38 ` Darrick J. Wong
  2022-01-27 16:50   ` Darrick J. Wong
  0 siblings, 1 reply; 2+ messages in thread
From: Darrick J. Wong @ 2022-01-26 19:38 UTC (permalink / raw)
  To: Christoph Hellwig
  Cc: Jens Axboe, Tetsuo Handa, Jan Kara, linux-block, Ming Lei, xfs

On Wed, Jan 26, 2022 at 04:50:32PM +0100, Christoph Hellwig wrote:
> Hi Jens, hi Tetsuo,
> 
> this series uses the approach from Tetsuo to delay the destroy_workueue
> cll, extended by a delayed teardown of the workers to fix a potential
> racewindow then the workqueue can be still round after finishing the
> commands.  It then also removed the queue freeing in release that is
> not needed to fix the dependency chain for that (which can't be
> reported by lockdep) as well.

[add xfs to cc list]

This fixes all the regressions I've been seeing in xfs/049 and xfs/073,
thank you.  I'll give this a spin with the rest of fstests overnight.

--D

^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: yet another approach to fix loop autoclear for xfstets xfs/049
  2022-01-26 19:38 ` yet another approach to fix loop autoclear for xfstets xfs/049 Darrick J. Wong
@ 2022-01-27 16:50   ` Darrick J. Wong
  0 siblings, 0 replies; 2+ messages in thread
From: Darrick J. Wong @ 2022-01-27 16:50 UTC (permalink / raw)
  To: Christoph Hellwig
  Cc: Jens Axboe, Tetsuo Handa, Jan Kara, linux-block, Ming Lei, xfs

On Wed, Jan 26, 2022 at 11:38:40AM -0800, Darrick J. Wong wrote:
> On Wed, Jan 26, 2022 at 04:50:32PM +0100, Christoph Hellwig wrote:
> > Hi Jens, hi Tetsuo,
> > 
> > this series uses the approach from Tetsuo to delay the destroy_workueue
> > cll, extended by a delayed teardown of the workers to fix a potential
> > racewindow then the workqueue can be still round after finishing the
> > commands.  It then also removed the queue freeing in release that is
> > not needed to fix the dependency chain for that (which can't be
> > reported by lockdep) as well.
> 
> [add xfs to cc list]
> 
> This fixes all the regressions I've been seeing in xfs/049 and xfs/073,
> thank you.  I'll give this a spin with the rest of fstests overnight.

After an overnight run with 5.17-rc1 + {xfs,iomap}-for-next + this patchset,
fstests is back to normal.

Tested-by: Darrick J. Wong <djwong@kernel.org>

--D

> 
> --D

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2022-01-27 16:50 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
     [not found] <20220126155040.1190842-1-hch@lst.de>
2022-01-26 19:38 ` yet another approach to fix loop autoclear for xfstets xfs/049 Darrick J. Wong
2022-01-27 16:50   ` Darrick J. Wong

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).