Messages in this thread | | | Date | Thu, 12 Jan 2023 15:25:42 -0500 | Subject | Re: rseq CPU ID not correct on 6.0 kernels for pinned threads | From | Mathieu Desnoyers <> |
| |
On 2023-01-12 11:33, Florian Weimer wrote: > * 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.
If it's not cpu hotunplug, then perhaps something like systemd modifies the AllowedCPUs of your cpuset concurrently ?
Thanks,
Mathieu
-- Mathieu Desnoyers EfficiOS Inc. https://www.efficios.com
| |