linux-kernel - Re: [PATCH] genirq: Add default affinity mask command line option

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: <alpine.DEB.2.11.1602032013060.25254@nanos>
Date:	Wed, 3 Feb 2016 20:15:15 +0100 (CET)
From:	Thomas Gleixner <tglx@...utronix.de>
To:	Rik van Riel <riel@...hat.com>
cc:	LKML <linux-kernel@...r.kernel.org>,
	Ingo Molnar <mingo@...nel.org>,
	Peter Zijlstra <peterz@...radead.org>,
	Frederic Weisbecker <fweisbec@...il.com>,
	Chris Metcalf <cmetcalf@...hip.com>,
	Christoph Lameter <cl@...ux.com>
Subject: Re: [PATCH] genirq: Add default affinity mask command line option

On Wed, 3 Feb 2016, Rik van Riel wrote:
> On 02/03/2016 01:52 PM, Thomas Gleixner wrote:
> > From: Thomas Gleixner <tglx@...utronix.de> Date: Fri May 25
> > 16:59:47 2012 +0200 Subject: genirq: Add default affinity mask
> > command line option
> > 
> > If we isolate CPUs, then we don't want random device interrupts on
> > them. Even w/o the user space irq balancer enabled we can end up
> > with irqs on non boot cpus and chasing newly requested interrupts
> > is a tedious task.
> 
> The userspace irqbalance daemon has code nowadays to prevent
> it from putting irqs on isolated CPUs.

I know, but a lot of systems do not enable it at all.

> > Allow to restrict the default irq affinity mask.
> 
> However, not placing them on there at boot time is a great
> enhancement, too.

That's the main plan :)
 
> Would it make sense to key off the isolated_cpus mask
> if isolated_cpus= was specified, but no irqaffinity
> mask was specified?

That might be an add on.
 
> Is there any use case where we would want irqs on
> isolated cpus by default, and could not bear the
> cost of reassigning them from userspace after boot?

I don't think so. Those interrupts you want on an isolated cpu are obviously
hand selected.

Thanks,

	tglx