linux-kernel - Re: [sched] 10e7071b2f: BUG:kernel_NULL_pointer

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]

Date:   Thu, 7 Nov 2019 12:43:55 +0000
From:   Quentin Perret <qperret@...gle.com>
To:     Valentin Schneider <valentin.schneider@....com>
Cc:     kernel test robot <lkp@...el.com>,
        Peter Zijlstra <peterz@...radead.org>,
        Aaron Lu <aaron.lwe@...il.com>, Phil Auld <pauld@...hat.com>,
        Julien Desfossez <jdesfossez@...italocean.com>,
        Nishanth Aravamudan <naravamudan@...italocean.com>,
        LKML <linux-kernel@...r.kernel.org>,
        Linus Torvalds <torvalds@...ux-foundation.org>, lkp@...org
Subject: Re: [sched] 10e7071b2f: BUG:kernel_NULL_pointer_dereference,address

On Thursday 07 Nov 2019 at 12:37:09 (+0000), Valentin Schneider wrote:
> On 07/11/2019 12:15, Quentin Perret wrote:
> > On Thursday 07 Nov 2019 at 12:09:22 (+0000), Quentin Perret wrote:
> >> sched_move_task() follows what Peter called the 'change' pattern, so I'm
> >> thinking this is most likely the same issue. Dropping the lock causes an
> >> unmitigated race between sched_move_task() and pick_next_task_dl(), so
> >> hilarity ensues (set_next_task() being called twice for instance).
> > 
> > Bah, scratch that. 10e7071b2 is clearly before the pick_next_task()
> > rework, so that's not it :(
> > 
> 
> And besides we don't drop the lock until reaching pick_next_task_fair(),
> and the splat says it died on pick_next_task_dl() which happens earlier.

Right, with the rework of pick_next_task(), we would in fact drop the
lock before calling pick_next_task_dl(), which would explain the issue,
hence my confusion.

But that doesn't apply here, so this is another problem :(

Sorry for the noise,
Quentin