On 2024-03-06 13:27:06 [-0300], Arnaldo Carvalho de Melo wrote: > That its what introduces perf_pending_task(), I'm now unsure we can just > disable migration, as event_sched_out() seems to require being called > under a raw_spin_lock and that disables preemption... Not sure what the best course of action is here but based on what I learned last time you reported this I think we need delayed signals… Let me look into this. We had it and then removed it because we had no users of it at some point but probably nobody took perf into account. > - Arnaldo Sebastian
