linux-kernel - Re: [PATCH v4.9] sched/fair: Overload-On-Wakeup fix.

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: <20170315122647.GQ3093@worktop>
Date:   Wed, 15 Mar 2017 13:26:47 +0100
From:   Peter Zijlstra <peterz@...radead.org>
To:     Jarmo Tiitto <jarmo.tiitto@...il.com>
Cc:     Ingo Molnar <mingo@...hat.com>, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v4.9] sched/fair: Overload-On-Wakeup fix.

On Tue, Mar 14, 2017 at 11:34:06PM +0200, Jarmo Tiitto wrote:
> Try improve multi-core scaling.
> The changes are experimental - the fix works
> but possible regressions are unknown.
> 
> Please read the orginal author's paper to understand
> the overall problem this patch tries to solve.
> 
> See the orignal patches at:
> https://github.com/jplozi/wastedcores/tree/master/patches
> 
> And the related paper at:
> http://www.i3s.unice.fr/~jplozi/wastedcores/
> 
> v1: the patch was re-written for v4.9 to remove goto statement.
> 
> v2: improved version from the wastedcores initial code. try take into
> 	account the cpu  wake-up latency and choose most recent idle cpu
> 	instead.
> 
> I have tested the code and it really does improve
> multi-core performance on my machine.

v4.9 is not the right kernel to send patches against (hint: v4.10 is not
either).

Also your Changelog is inadequate; it should be self contained, explain
the problem and describe the solution.