linux-kernel - Re: [PATCH v3] sched/topology: Enable topology_span

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: <20250205101600.GC7145@noisy.programming.kicks-ass.net>
Date: Wed, 5 Feb 2025 11:16:00 +0100
From: Peter Zijlstra <peterz@...radead.org>
To: K Prateek Nayak <kprateek.nayak@....com>
Cc: Naman Jain <namjain@...ux.microsoft.com>,
	Ingo Molnar <mingo@...hat.com>, Juri Lelli <juri.lelli@...hat.com>,
	Vincent Guittot <vincent.guittot@...aro.org>,
	Dietmar Eggemann <dietmar.eggemann@....com>,
	Steven Rostedt <rostedt@...dmis.org>,
	Ben Segall <bsegall@...gle.com>, Mel Gorman <mgorman@...e.de>,
	Valentin Schneider <vschneid@...hat.com>, stable@...r.kernel.org,
	linux-kernel@...r.kernel.org, Steve Wahl <steve.wahl@....com>,
	Saurabh Singh Sengar <ssengar@...ux.microsoft.com>,
	srivatsa@...il.mit.edu, Michael Kelley <mhklinux@...look.com>
Subject: Re: [PATCH v3] sched/topology: Enable topology_span_sane check only
 for debug builds

On Wed, Feb 05, 2025 at 03:43:54PM +0530, K Prateek Nayak wrote:
> Hello Peter,
> 
> Thank you for the background!
> 
> On 2/5/2025 3:25 PM, Peter Zijlstra wrote:
> > On Wed, Feb 05, 2025 at 03:18:24PM +0530, K Prateek Nayak wrote:
> > 
> > > Have there been any reports on an x86 system / VM where
> > > topology_span_sane() was tripped?
> > 
> > At the very least Intel SNC 'feature' tripped it at some point. They
> > figured it made sense to have the LLC span two nodes.
> > 
> > But I think there were some really dodgy VMs too.
> > 
> > But yeah, its not been often. But basically dodgy BIOS/VM data can mess
> > up things badly enough for it to trip.
> 
> Has it ever happened without tripping the topology_sane() check first
> on the x86 side?

That I can't remember, sorry :/