Re: [PATCH v11] perf: Sharing PMU counters across compatible events

[Song Liu]

> On Mar 24, 2020, at 2:43 PM, Song Liu <songliubraving@xxxxxx> wrote:
> 
> 
> 
>> On Mar 24, 2020, at 2:41 PM, Song Liu <songliubraving@xxxxxx> wrote:
>> 
>> This patch tries to enable PMU sharing. When multiple perf_events are
>> counting the same metric, they can share the hardware PMU counter. We
>> call these events as "compatible events".
>> 
>> The PMU sharing are limited to events within the same perf_event_context
>> (ctx). When a event is installed or enabled, search the ctx for compatible
>> events. This is implemented in perf_event_setup_dup(). One of these
>> compatible events are picked as the master (stored in event->dup_master).
>> Similarly, when the event is removed or disabled, perf_event_remove_dup()
>> is used to clean up sharing.
>> 
>> If the master event is a cgroup event or a event in a group, it is
>> possible that some slave events are ACTIVE, but the master event is not.
>> To handle this scenario, we introduced PERF_EVENT_STATE_ENABLED. Also,
>> since PMU drivers write into event->count, master event needs another
>> variable (master_count) for the reading of this event.
>> 
>> On the critical paths (add, del read), sharing PMU counters doesn't
>> increase the complexity. Helper functions event_pmu_[add|del|read]() are
>> introduced to cover these cases. All these functions have O(1) time
>> complexity.
>> 
>> Cc: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
>> Cc: Arnaldo Carvalho de Melo <acme@xxxxxxxxxx>
>> Cc: Jiri Olsa <jolsa@xxxxxxxxxx>
>> Cc: Alexey Budankov <alexey.budankov@xxxxxxxxxxxxxxx>
>> Cc: Namhyung Kim <namhyung@xxxxxxxxxx>
>> Cc: Tejun Heo <tj@xxxxxxxxxx>
>> Reported-by: kernel test robot <rong.a.chen@xxxxxxxxx>
>> Signed-off-by: Song Liu <songliubraving@xxxxxx>
>> 
>> ---
>> Changes in v12:
>> Fix new failures perf_event_tests. (kernel test robot)
> 
> Typo in subject. This is actually v12. 

I was stupid... This version didn't work... Please discard it. 

Song