linux-kernel - Re: [PATCH 2/2] workqueue:Fix affinity of an unbound worker of a node with 1 online CPU

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: <20160615160112.GC24102@mtj.duckdns.org>
Date:	Wed, 15 Jun 2016 12:01:12 -0400
From:	Tejun Heo <htejun@...il.com>
To:	Peter Zijlstra <peterz@...radead.org>
Cc:	Gautham R Shenoy <ego@...ux.vnet.ibm.com>,
	Thomas Gleixner <tglx@...utronix.de>,
	Michael Ellerman <mpe@...erman.id.au>,
	Abdul Haleem <abdhalee@...ux.vnet.ibm.com>,
	Aneesh Kumar <aneesh.kumar@...ux.vnet.ibm.com>,
	linuxppc-dev@...ts.ozlabs.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH 2/2] workqueue:Fix affinity of an unbound worker of a
 node with 1 online CPU

Hello,

On Wed, Jun 15, 2016 at 03:14:15PM +0200, Peter Zijlstra wrote:
> On Wed, Jun 15, 2016 at 06:20:33PM +0530, Gautham R Shenoy wrote:
> > We will no longer have the optimization in
> > restore_unbound_workers_cpumask() but I suppose we don't lose much by
> > resetting the affinity every time a CPU in the pool->attr->cpumask
> > comes online.
> 
> Right; optimizing hotplug really isn't worth it. The code needs to be
> simple and robust (ha! funny).

The only case it might matter is CPU hotplug being used aggressively
for power saving.  No idea how popular that is now tho.
set_cpus_allowed isn't that expensive and phones don't tend to have
massive number of kworkers, so hopefully it won't show up.

> In any case, Tejun, does this work for you?

I'm not sure about the reordering part but for setting affinity on
each onlining, no objection.  If it ever shows up as performance /
power regression, we can revisit it later.

Thanks.

-- 
tejun