linux-kernel - Re: [PATCH V2 16/25] perf/x86: Register hybrid PMUs

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 for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20210311195408.GJ4746@worktop.programming.kicks-ass.net>
Date:   Thu, 11 Mar 2021 20:54:08 +0100
From:   Peter Zijlstra <peterz@...radead.org>
To:     Andi Kleen <ak@...ux.intel.com>
Cc:     kan.liang@...ux.intel.com, mingo@...nel.org,
        linux-kernel@...r.kernel.org, acme@...nel.org, tglx@...utronix.de,
        bp@...en8.de, namhyung@...nel.org, jolsa@...hat.com,
        yao.jin@...ux.intel.com, alexander.shishkin@...ux.intel.com,
        adrian.hunter@...el.com
Subject: Re: [PATCH V2 16/25] perf/x86: Register hybrid PMUs

On Thu, Mar 11, 2021 at 09:53:58AM -0800, Andi Kleen wrote:
> > AFAICT we could register them all here. That instantly fixes that
> > CPU_STARTING / CPU_DEAD fail elsewhere in this patch.
> 
> This would mean a system that only has Atoms or only has big cores
> would still show the other CPU's PMU. We expect those to exist.

Well, barring enumeration in ACPI/SRAT or thereabout, there's simply no
way to know.

Also, what's the point of making an alderlake with only big cores? Isn't
that what we're supposed to call saphire rappids? Same with only Atoms,
we're supposed to call that a temont.

Anyway, if people want to do weird things and not enumerate it sanely,
they'll get whatever we can make of it.