linux-kernel - Re: Abysmal scheduler performance in Linus' tree?

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: <20170906161422.kvqe5y6b3ng6ouyu@hirez.programming.kicks-ass.net>
Date:   Wed, 6 Sep 2017 18:14:22 +0200
From:   Peter Zijlstra <peterz@...radead.org>
To:     Andy Lutomirski <luto@...capital.net>
Cc:     Andy Lutomirski <luto@...nel.org>, Ingo Molnar <mingo@...nel.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: Abysmal scheduler performance in Linus' tree?

On Wed, Sep 06, 2017 at 11:03:33AM +0200, Peter Zijlstra wrote:
> On Wed, Sep 06, 2017 at 01:59:14AM -0700, Andy Lutomirski wrote:
> > > Any specifics on your setup?
> > 
> > On further fiddling, I only see this after a suspend and resume cycle.
> 
> Ah, ok. That's not something I otherwise test. Lets see if I can force
> this brick of mine through a suspend-resume cycle :-)

I'd be suspicious of these here commits:

  77d1dfda0e79 ("sched/topology, cpuset: Avoid spurious/wrong domain rebuilds")
  09e0dd8e0f2e ("sched/topology: Avoid pointless rebuild")
  bbdacdfed2f5 ("sched/debug: Optimize sched_domain sysctl generation")

I tested them with regular hotplug, but suspend resume always is a tad
funny.