Re: [PATCH] perf cgroups: Fix perf_cgroup_switch schedule in warning

[Peter Zijlstra]

On Tue, 2012-10-02 at 15:34 +0200, Stephane Eranian wrote:
> And adding a comment in cgroup_switch()
> 
> +               /* ensure we process each cpuctx only once */
> +               if (cpuctx->active_pmu != pmu)
> +                       continue;
> 
---
Subject: perf: Fix perf_cgroup_switch for sw-events
From: Peter Zijlstra <a.p.zijlstra@xxxxxxxxx>
Date: Tue Oct 02 15:41:23 CEST 2012

Jiri reported that he could trigger the WARN_ON_ONCE() in
perf_cgroup_switch() using sw-events. This is because sw-events share
a cpuctx with multiple PMUs.

Use the ->unique_pmu pointer to limit the pmu iteration to unique
cpuctx instances.

Reported-and-Tested-by: Jiri Olsa <jolsa@xxxxxxxxxx>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@xxxxxxxxx>
---
 kernel/events/core.c |    9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)

--- a/kernel/events/core.c
+++ b/kernel/events/core.c
@@ -372,6 +372,8 @@ void perf_cgroup_switch(struct task_stru
 
 	list_for_each_entry_rcu(pmu, &pmus, entry) {
 		cpuctx = this_cpu_ptr(pmu->pmu_cpu_context);
+		if (cpuctx->unique_pmu != pmu)
+			continue; /* ensure we process each cpuctx once */
 
 		/*
 		 * perf_cgroup_events says at least one
@@ -395,9 +397,10 @@ void perf_cgroup_switch(struct task_stru
 
 			if (mode & PERF_CGROUP_SWIN) {
 				WARN_ON_ONCE(cpuctx->cgrp);
-				/* set cgrp before ctxsw in to
-				 * allow event_filter_match() to not
-				 * have to pass task around
+				/*
+				 * set cgrp before ctxsw in to allow
+				 * event_filter_match() to not have to pass
+				 * task around
 				 */
 				cpuctx->cgrp = perf_cgroup_from_task(task);
 				cpu_ctx_sched_in(cpuctx, EVENT_ALL, task);

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/