Messages in this thread | | | Date | Wed, 8 May 2024 11:03:22 -0700 | Subject | Re: [PATCH v2 1/2] x86/fpu: Extend kernel_fpu_begin_mask() to initialize AMX state | From | "Chang S. Bae" <> |
| |
On 5/8/2024 7:40 AM, Dave Hansen wrote: > On 5/7/24 16:53, Chang S. Bae wrote: > >> However, due to resource constraints in storage, AMX state is excluded >> from the scope of state recovery. Consequently, AMX state must be in its >> initialized state for the IFS test to run. > > This doesn't mention how this issue got introduced. Are we all bad at > reading the SDM? :)
Ah, I'd rather zap out this SDM sentence.
>> When AMX workloads are running, an active user AMX state remains even >> after a context switch, optimizing to reduce the state reload cost. In >> such cases, the test cannot proceed if it is scheduled. > > This is a bit out of the blue. What does scheduling have do do with IFS?
$ echo <cpu#> > /sys/devices/virtual/misc/intel_ifs_0/run_test
Then, run_test_store() -> do_core_test() -> ifs_test_core() -> stop_core_cpuslocked() -> stop_cpus() -> queue_stop_cpus_work() -> cpu_stop_queue_work() -> wake_q_add() -> wake_up_q()
So, the CPU stopper threads for <cpu#> and its sibling to execute doscan() are queued up with the highest priority.
queue_stop_cpus_work() has
for_each_cpu(cpu, cpumask) { work = &per_cpu(cpu_stopper.stop_work, cpu); work->fn = fn; work->arg = arg; work->done = done; work->caller = _RET_IP_; if (cpu_stop_queue_work(cpu, work)) queued = true; }
Those threads are created during early boot via smpboot_register_percpu_thread().
> I'm not sure those last two paragraphs add much value. I'd try to > banish most of that content to *after* you talk about the solution. Or > maybe put it in the cover letter.
It looks like lots of distractions coming from bunch of alternatives in different levels.
Thanks, Chang
PS: Let me respond the solution discussion separately. I do want to experiment the init-track behavior a bit.
| |