linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Oleg Nesterov <oleg@redhat.com>
To: Jan Kara <jack@suse.cz>
Cc: Al Viro <viro@zeniv.linux.org.uk>,
	Dave Chinner <david@fromorbit.com>,
	Nikolay Borisov <kernel@kyup.com>,
	"Paul E. McKenney" <paulmck@linux.vnet.ibm.com>,
	linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH V2 2/2] fs/super.c: don't fool lockdep in freeze_super() and thaw_super() paths
Date: Tue, 27 Sep 2016 19:29:01 +0200	[thread overview]
Message-ID: <20160927172901.GA11879@redhat.com> (raw)
In-Reply-To: <20160927065135.GA1139@quack2.suse.cz>

On 09/27, Jan Kara wrote:
>
> On Mon 26-09-16 18:55:25, Oleg Nesterov wrote:
> >
> > Heh ;) if only I knew how to test this... I ran the following script
> > under qemu
> >
> > 	mkfs.xfs -f /dev/vda
> > 	mkfs.xfs -f /dev/vdb
> >
> > 	mkdir -p TEST SCRATCH
> >
> > 	TEST_DEV=/dev/vda TEST_DIR=TEST SCRATCH_DEV=/dev/vdb SCRATCH_MNT=SCRATCH \
> > 	./check `grep -il freeze tests/*/???`
>
> You can run either:
>
> 	./check -g freeze

passed all 6 tests.

> to check just the freezing tests or
>
> 	./check
>
> to run all sensible tests which is what I'd do (but it will take couple of
> hours to pass). If that passes, chances are good there are no easy false
> positives.

It seems that generic/001 just hangs on my laptop. With or without this change.
Or perhaps I didn't wait enough... Or perhaps something is wrong with my very
limited testing environment. I'll reserve a testing machine tomorrow.

> > And yes, I'm afraid this change can uncover some false positives later.
> > But at the same time potentially it can find the real problems.
>
> Well, sure it's not an end of world if there is some false positive - we
> can just revert the change - but lockdep false positives are always
> annoying because they take time to analyze and until they are fixed, you
> are unable to see other probles found by lockdep...

Yes, yes, agreed.

> > It would be nice to remove another hack in __sb_start_write under
> > ifdef(CONFIG_LOCKDEP), but iirc XFS actually takes the same rw_sem twice
> > for reading, so we can't do this.
>
> Yes, and I don't really consider this a hack.

Ah, sorry, I didn't try to blame XFS/fs. I meant, this "force_trylock" hack
doesn't look nice. Perhaps we can use rwsem_acquire_nest() instead.

> Reviewed-by: Jan Kara <jack@suse.cz>

Thanks!

Oleg.

  parent reply	other threads:[~2016-09-27 17:29 UTC|newest]

Thread overview: 30+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-09-26 16:07 [PATCH 0/2] (Was: BUG_ON in rcu_sync_func triggered) Oleg Nesterov
2016-09-26 16:07 ` [PATCH 1/2] fs/super.c: fix race between freeze_super() and thaw_super() Oleg Nesterov
2016-09-26 16:11   ` Jan Kara
2016-09-26 16:08 ` [PATCH 2/2] fs/super.c: don't fool lockdep in freeze_super() and thaw_super() paths Oleg Nesterov
2016-09-26 16:18   ` Jan Kara
2016-09-26 16:55     ` [PATCH V2 " Oleg Nesterov
2016-09-27  6:51       ` Jan Kara
2016-09-27  7:14         ` Dave Chinner
2016-09-27 17:29         ` Oleg Nesterov [this message]
2016-09-30 17:14           ` Oleg Nesterov
2016-10-02 21:42             ` Dave Chinner
2016-10-03 16:44               ` Oleg Nesterov
2016-10-04 11:43                 ` Oleg Nesterov
2016-10-04 11:48                   ` Michal Hocko
2016-10-06 13:44                     ` Johannes Weiner
2016-10-07 16:52                       ` Oleg Nesterov
2016-10-04 16:58                   ` Oleg Nesterov
2016-10-04 20:03                     ` Dave Chinner
2016-10-05 16:33                       ` Oleg Nesterov
2016-10-04 19:44                   ` Dave Chinner
2016-10-05 16:44                     ` Oleg Nesterov
2016-10-06  7:27                       ` Jan Kara
2016-10-06 17:17                       ` Oleg Nesterov
2016-10-06 21:59                         ` Dave Chinner
2016-10-07 17:15                           ` Oleg Nesterov
2016-10-07 22:52                             ` Dave Chinner
2016-10-09 16:14                               ` Oleg Nesterov
2016-10-10  1:02                                 ` Dave Chinner
2016-10-13 16:58                                   ` Oleg Nesterov
2016-10-13 17:10 ` [PATCH 0/2] (Was: BUG_ON in rcu_sync_func triggered) Oleg Nesterov

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20160927172901.GA11879@redhat.com \
    --to=oleg@redhat.com \
    --cc=david@fromorbit.com \
    --cc=jack@suse.cz \
    --cc=kernel@kyup.com \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=paulmck@linux.vnet.ibm.com \
    --cc=viro@zeniv.linux.org.uk \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
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).