Re: [PATCH] perf/x86/intel: Add proper condition to run sched_task callbacks

[Jiri Olsa]

On Tue, Jul 18, 2017 at 02:37:34PM +0200, Peter Zijlstra wrote:
> On Tue, Jul 18, 2017 at 11:29:32AM +0200, Jiri Olsa wrote:
> 
> > because we have 2 places using the same callback
> >   - PEBS drain for free running counters
> >   - LBR save/store
> > 
> > both of them called from intel_pmu_sched_task
> > 
> > so let's say PEBS drain setup the callback for the event,
> > but in the callback itself (intel_pmu_sched_task) we will
> > also run the code for LBR save/restore, which we did not
> > ask for, but the code in intel_pmu_sched_task does not
> > check for that
> 
> Ah fair enough; Changelog confused me.

just when I thought that my english is getting bedder ;-)

v2 attached with new changelog 

jirka


---
We have 2 functions using the same sched_task callback:
  - PEBS drain for free running counters
  - LBR save/store

Both of them are called from intel_pmu_sched_task and
either of them can be unwillingly triggered when the
other one is configured to run.

Let's say there's PEBS drain configured in sched_task
callback for the event, but in the callback itself
(intel_pmu_sched_task) we will also run the code for
LBR save/restore, which we did not ask for, but the
code in intel_pmu_sched_task does not check for that.

This can lead to extra cycles in some perf monitoring,
like when we monitor PEBS event without LBR data.

  # perf record --no-timestamp -c 10000 -e cycles:p ./perf bench sched pipe -l 1000000

  (We need PEBS, non freq/non timestamp event to enable
   the sched_task callback)

The perf stat of cycles and msr:write_msr for above
command before the change:
  ...
  Performance counter stats for './perf record --no-timestamp -c 10000 -e cycles:p \
                                 ./perf bench sched pipe -l 1000000' (5 runs):

    18,519,557,441      cycles:k
        91,195,527      msr:write_msr

      29.334476406 seconds time elapsed

And after the change:
  ...
  Performance counter stats for './perf record --no-timestamp -c 10000 -e cycles:p \
                                 ./perf bench sched pipe -l 1000000' (5 runs):

    18,704,973,540      cycles:k
        27,184,720      msr:write_msr

      16.977875900 seconds time elapsed

There's no affect on cycles:k because the sched_task happens
with events switched off, however the msr:write_msr tracepoint
counter together with almost 50% of time speedup show the
improvement.

Monitoring LBR event and having extra PEBS drain processing
in sched_task callback showed just a little speedup, because
the drain function does not do much extra work in case there
is no PEBS data.

Adding conditions to recognize the configured work that needs
to be done in the x86_pmu's sched_task callback.

Suggested-by: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
Signed-off-by: Jiri Olsa <jolsa@xxxxxxxxxx>
---
 arch/x86/events/intel/core.c |  6 ++----
 arch/x86/events/intel/ds.c   | 14 ++++++++------
 arch/x86/events/intel/lbr.c  |  4 ++++
 3 files changed, 14 insertions(+), 10 deletions(-)

diff --git a/arch/x86/events/intel/core.c b/arch/x86/events/intel/core.c
index ede97710c2f4..98b0f0729527 100644
--- a/arch/x86/events/intel/core.c
+++ b/arch/x86/events/intel/core.c
@@ -3397,10 +3397,8 @@ static void intel_pmu_cpu_dying(int cpu)
 static void intel_pmu_sched_task(struct perf_event_context *ctx,
 				 bool sched_in)
 {
-	if (x86_pmu.pebs_active)
-		intel_pmu_pebs_sched_task(ctx, sched_in);
-	if (x86_pmu.lbr_nr)
-		intel_pmu_lbr_sched_task(ctx, sched_in);
+	intel_pmu_pebs_sched_task(ctx, sched_in);
+	intel_pmu_lbr_sched_task(ctx, sched_in);
 }
 
 PMU_FORMAT_ATTR(offcore_rsp, "config1:0-63");
diff --git a/arch/x86/events/intel/ds.c b/arch/x86/events/intel/ds.c
index 6dc8a59e1bfb..a322fed5f8ed 100644
--- a/arch/x86/events/intel/ds.c
+++ b/arch/x86/events/intel/ds.c
@@ -606,12 +606,6 @@ static inline void intel_pmu_drain_pebs_buffer(void)
 	x86_pmu.drain_pebs(&regs);
 }
 
-void intel_pmu_pebs_sched_task(struct perf_event_context *ctx, bool sched_in)
-{
-	if (!sched_in)
-		intel_pmu_drain_pebs_buffer();
-}
-
 /*
  * PEBS
  */
@@ -822,6 +816,14 @@ static inline bool pebs_needs_sched_cb(struct cpu_hw_events *cpuc)
 	return cpuc->n_pebs && (cpuc->n_pebs == cpuc->n_large_pebs);
 }
 
+void intel_pmu_pebs_sched_task(struct perf_event_context *ctx, bool sched_in)
+{
+	struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events);
+
+	if (!sched_in && pebs_needs_sched_cb(cpuc))
+		intel_pmu_drain_pebs_buffer();
+}
+
 static inline void pebs_update_threshold(struct cpu_hw_events *cpuc)
 {
 	struct debug_store *ds = cpuc->ds;
diff --git a/arch/x86/events/intel/lbr.c b/arch/x86/events/intel/lbr.c
index eb261656a320..955457a30197 100644
--- a/arch/x86/events/intel/lbr.c
+++ b/arch/x86/events/intel/lbr.c
@@ -380,8 +380,12 @@ static void __intel_pmu_lbr_save(struct x86_perf_task_context *task_ctx)
 
 void intel_pmu_lbr_sched_task(struct perf_event_context *ctx, bool sched_in)
 {
+	struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events);
 	struct x86_perf_task_context *task_ctx;
 
+	if (!cpuc->lbr_users)
+		return;
+
 	/*
 	 * If LBR callstack feature is enabled and the stack was saved when
 	 * the task was scheduled out, restore the stack. Otherwise flush
-- 
2.9.4