linux-kernel - Re: rseq CPU ID not correct on 6.0 kernels for pinned threads

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: <871qnzpv9l.fsf@oldenburg.str.redhat.com>
Date:   Thu, 12 Jan 2023 17:33:58 +0100
From:   Florian Weimer <fweimer@...hat.com>
To:     Mathieu Desnoyers <mathieu.desnoyers@...icios.com>
Cc:     linux-kernel@...r.kernel.org, Ingo Molnar <mingo@...hat.com>,
        Peter Zijlstra <peterz@...radead.org>,
        libc-alpha@...rceware.org
Subject: Re: rseq CPU ID not correct on 6.0 kernels for pinned threads

* Mathieu Desnoyers:

> As you also point out, it can also be caused by some other task
> modifying the affinity of your task concurrently. You could print
> the result of sched_getaffinity on error to get a better idea of
> the expected vs actual mask.
>
> Lastly, it could be caused by CPU hotplug which would set all bits
> in the affinity mask as a fallback. As you mention it should not be
> the cause there.
>
> Can you share your kernel configuration ?

Attached.

cpupower frequency-info says:

analyzing CPU 0:
  driver: intel_cpufreq
  CPUs which run at the same hardware frequency: 0
  CPUs which need to have their frequency coordinated by software: 0
  maximum transition latency: 20.0 us
  hardware limits: 800 MHz - 4.60 GHz
  available cpufreq governors: conservative ondemand userspace powersave performance schedutil
  current policy: frequency should be within 800 MHz and 4.60 GHz.
                  The governor "schedutil" may decide which speed to use
                  within this range.
  current CPU frequency: Unable to call hardware
  current CPU frequency: 3.20 GHz (asserted by call to kernel)
  boost state support:
    Supported: yes
    Active: yes

And I have: kernel.sched_energy_aware = 1

> Is this on a physical machine or in a virtual machine ?

I think it happened on both.

I added additional error reporting to the test (running on kernel
6.0.18-300.fc37.x86_64), and it seems that there is something that is
mucking with affinity masks:

info: Detected CPU set size (in bits): 64
info: Maximum test CPU: 19
error: Pinned thread 17 ran on impossible cpu 7
info: getcpu reported CPU 7, node 0
info: CPU affinity mask: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
error: Pinned thread 3 ran on impossible cpu 13
info: getcpu reported CPU 13, node 0
info: CPU affinity mask: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
info: Main thread ran on 2 CPU(s) of 20 available CPU(s)
info: Other threads ran on 20 CPU(s)

For each of these threads, the affinity mask should be a singleton set.
Now I need to find out if there is a process that changes affinity
settings.

Thanks,
Florian

View attachment "config" of type "text/plain" (166344 bytes)