linux-kernel - Re: [PATCH] writeback, cgroup: remove wb from offline list before releasing refcnt

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: <YPsU9/8ZijTjHTWh@carbon.dhcp.thefacebook.com>
Date:   Fri, 23 Jul 2021 12:13:59 -0700
From:   Roman Gushchin <guro@...com>
To:     Adam Williamson <awilliam@...hat.com>
CC:     <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] writeback, cgroup: remove wb from offline list before
 releasing refcnt

On Fri, Jul 23, 2021 at 11:33:43AM -0700, Adam Williamson wrote:
> Hi folks! I'm not subscribed to the list and am trying to send this
> using git-send-email, apologies for any weirdness related to that.
> 
> Just wanted to confirm the issue that this patch attempts to address,
> and ask if it can be reviewed/moved along. I look after Fedora's
> openQA automated test instance, and in tests of recent Fedora Rawhide
> composes, we're seeing at least one failure per compose that's caused
> by this crash (it usually prevents the system shutting down or
> rebooting correctly in a test which requires that to happen).
> 
> I can't actually confirm the fix as I can't really easily change our
> tests to run on a non-official kernel build, and our kernel maintainer
> (reasonably) says he won't backport the patch until it's at least got
> some review. But I'm definitely seeing the problem!
> 
> Thanks folks.
> -- 
> Adam Williamson
> Fedora QA
> IRC: adamw | Twitter: adamw_ha
> https://www.happyassassin.net 
> 

Hello, Adam!

The patch was picked up by Andrew Morton and will be soon merged into the Linus's tree.
Currently you can find it in the "next" tree (can't provide a hash because it's not stable
there).

Please, note that there it another fix called "writeback, cgroup: do not reparent dax inodes",
which you likely want to backport too.

Both fixes got some testing and reviews.

Thanks!