linux-kernel - Re: sched/autogroup: race if !sysctl_sched_autogroup

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] [day] [month] [year] [list]

Message-ID: <20161114151447.GA23318@redhat.com>
Date:   Mon, 14 Nov 2016 16:14:48 +0100
From:   Oleg Nesterov <oleg@...hat.com>
To:     Mike Galbraith <efault@....de>
Cc:     Peter Zijlstra <peterz@...radead.org>,
        Ingo Molnar <mingo@...nel.org>,
        Linus Torvalds <torvalds@...ux-foundation.org>,
        hartsjc@...hat.com, vbendel@...hat.com, vlovejoy@...hat.com,
        linux-kernel@...r.kernel.org
Subject: Re: sched/autogroup: race if !sysctl_sched_autogroup_enabled ?

On 11/13, Mike Galbraith wrote:
>
> Autogroup never should have had knobs in the first place, and now, Oleg has
> discovered that the dynamic enable/disable capability either has become, or
> perhaps always was racy.

It was always racy afaics, even if sysctl_sched_autogroup_enabled doesn't
change. So this patch is not enough.

But the main reason I dislike it is that I can't backport it ;) rhel7 runs
with sysctl_sched_autogroup_enabled=0 by default, and users actually use
this knob to enable it dynamically.

Mike, et al, sorry for delay. I'll try to finally force myself to write the
comments and the changelog and send 2 (trivial) fixes today.

Oleg.