linux-kernel - Re: [PATCH V2 1/7] ext4: remove the 'group' parameter of ext4_trim

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <YN7qL+AQ85bbJ1yo@localhost>
Date:   Fri, 2 Jul 2021 03:27:59 -0700
From:   Josh Triplett <josh@...htriplett.org>
To:     Theodore Ts'o <tytso@....edu>
Cc:     Wang Jianchao <jianchao.wan9@...il.com>,
        Andreas Dilger <adilger.kernel@...ger.ca>,
        linux-ext4@...r.kernel.org, linux-kernel@...r.kernel.org,
        lishujin@...ishou.com
Subject: Re: [PATCH V2 1/7] ext4: remove the 'group' parameter of
 ext4_trim_extent

On Wed, Jun 23, 2021 at 03:49:52PM -0400, Theodore Ts'o wrote:
> CC'ed Josh since he was testing this patchset.
> 
> On Wed, Jun 23, 2021 at 09:13:42AM -0400, Theodore Ts'o wrote:
> > 
> > The conflicts were relatively minor, but the obvious fix-ups resulted
> > in a large number of crashes caused by various stress tests, such as
> > generic/068 and generic/204.  I'm currently investigating to see what
> > I might have messed up when I tried applying these patches, as well as
> > running your patch set applied against 5.12 to make sure the problems
> > weren't introduced by the patch set itself.
> 
> I applied this patch set on top of 5.12, where it applied w/o any
> patch conflicts, and I'm still seeing large number of crashes[1].  The
> crashes are in a wide variety of tests, and many of the stack traces
> involve the ext4_discard_work workqueue.

I applied this on top of a96bfed64c8986d6404e553f18203cae1f5ac7e6, which
required some fixups. I didn't try to run any kind of intensive
stress-tests, but I've run the deletion performance test I mentioned, as
well as various software compiles and other miscellaneous work. I
haven't run into any apparent issues.