Re: [PATCH 1/5] perf/arm-cci: Fix CPU hotplug race avoidance

2019-02-10 Thread Thomas Gleixner
On Mon, 4 Feb 2019, Robin Murphy wrote: > + > + cpus_read_lock(); > + cci_pmu->cpu = smp_processor_id(); That wants to be raw_smp_processor_id() because this is preemptible context and debug_smp_processor_id() will complain. Thanks, tglx

Re: [PATCH 1/5] perf/arm-cci: Fix CPU hotplug race avoidance

2019-02-05 Thread Suzuki K Poulose
Robin, On 04/02/2019 17:09, Robin Murphy wrote: The arm-cci probe logic faces a cyclic dependency wherein it has to pick a valid CPU to associate with before registering the PMU device, has to have the PMU state initialised before handling hotplug events in case it must be migrated, but has to

Re: [PATCH 1/5] perf/arm-cci: Fix CPU hotplug race avoidance

2019-02-05 Thread Corentin Labbe
On Mon, Feb 04, 2019 at 05:09:04PM +, Robin Murphy wrote: > The arm-cci probe logic faces a cyclic dependency wherein it has to pick > a valid CPU to associate with before registering the PMU device, has to > have the PMU state initialised before handling hotplug events in case it > must be

[PATCH 1/5] perf/arm-cci: Fix CPU hotplug race avoidance

2019-02-04 Thread Robin Murphy
The arm-cci probe logic faces a cyclic dependency wherein it has to pick a valid CPU to associate with before registering the PMU device, has to have the PMU state initialised before handling hotplug events in case it must be migrated, but has to have the hotplug notifier registered before the