| From | David Carrillo-Cisneros <> | Subject | [PATCH v2 19/32] x86/intel/cqm: use PERF_CGROUP_NO_RECURSION in CQM | Date | Wed, 11 May 2016 16:02:19 -0700 |
| |
The CQM hardware is not compatible with the way generic code handles cgroup hierarchies (simultaneously adding the events of for all ancestors of the current cgroup). This version of Intel's CQM driver handles cgroup hierarchy internally.
Set PERF_CGROUP_NO_RECURSION for llc_occupancy events to signal perf's generic code to not add events for ancestors of current cgroup.
Reviewed-by: Stephane Eranian <eranian@google.com> Signed-off-by: David Carrillo-Cisneros <davidcc@google.com> --- arch/x86/events/intel/cqm.c | 8 ++++++++ 1 file changed, 8 insertions(+)
diff --git a/arch/x86/events/intel/cqm.c b/arch/x86/events/intel/cqm.c index a5bec07..b241268 100644 --- a/arch/x86/events/intel/cqm.c +++ b/arch/x86/events/intel/cqm.c @@ -2476,6 +2476,14 @@ static int intel_cqm_event_init(struct perf_event *event) INIT_LIST_HEAD(&event->hw.cqm_event_groups_entry); INIT_LIST_HEAD(&event->hw.cqm_event_group_entry); + /* + * CQM driver handles cgroup recursion and since only noe + * RMID can be programmed at the time in each core, then + * it is incompatible with the way generic code handles + * cgroup hierarchies. + */ + event->pmu_event_flags |= PERF_CGROUP_NO_RECURSION; + mutex_lock(&cqm_mutex); -- 2.8.0.rc3.226.g39d4020
|