linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [PATCH] perf: remove deprecated warn_on_once()
@ 2012-02-02 11:04 Stephane Eranian
  2012-02-04 13:55 ` [tip:perf/urgent] perf: Remove deprecated WARN_ON_ONCE() tip-bot for Stephane Eranian
  0 siblings, 1 reply; 2+ messages in thread
From: Stephane Eranian @ 2012-02-02 11:04 UTC (permalink / raw)
  To: linux-kernel; +Cc: peterz, mingo


With the new throttling/unthrottling code introduced with
commit e050e3f, we hit two warn_on_once() in:
  - intel_pmu_pebs_enable()
  - intel_pmu_lbr_enable()

The assertions are no longer problematic. There is a valid
path where they can trigger but it is harmless.

The assertion can be triggered with:
  $ perf record -e instructions:pp ....

Leading to paths:

  intel_pmu_pebs_enable
  intel_pmu_enable_event
  x86_perf_event_set_period
  x86_pmu_start
  perf_adjust_freq_unthr_context
  perf_event_task_tick
  scheduler_tick

And:

  intel_pmu_lbr_enable
  intel_pmu_enable_event
  x86_perf_event_set_period
  x86_pmu_start
  perf_adjust_freq_unthr_context.
  perf_event_task_tick
  scheduler_tick

cpuc->enabled is always on because when we get to
perf_adjust_freq_unthr_context() the PMU is not totally
disabled. Furthermore when we need to adjust a period,
we only stop the event we need to change and not the
entire PMU. Thus, when we re-enable, cpuc->enabled is
already set. Note that when we stop the event, both
pebs and lbr are stopped if necessary (and possible).

Signed-off-by: Stephane Eranian <eranian@google.com>
---

diff --git a/arch/x86/kernel/cpu/perf_event_intel_ds.c b/arch/x86/kernel/cpu/perf_event_intel_ds.c
index 73da6b6..d6bd49f 100644
--- a/arch/x86/kernel/cpu/perf_event_intel_ds.c
+++ b/arch/x86/kernel/cpu/perf_event_intel_ds.c
@@ -439,7 +439,6 @@ void intel_pmu_pebs_enable(struct perf_event *event)
 	hwc->config &= ~ARCH_PERFMON_EVENTSEL_INT;
 
 	cpuc->pebs_enabled |= 1ULL << hwc->idx;
-	WARN_ON_ONCE(cpuc->enabled);
 
 	if (x86_pmu.intel_cap.pebs_trap && event->attr.precise_ip > 1)
 		intel_pmu_lbr_enable(event);
diff --git a/arch/x86/kernel/cpu/perf_event_intel_lbr.c b/arch/x86/kernel/cpu/perf_event_intel_lbr.c
index 3fab3de..47a7e63 100644
--- a/arch/x86/kernel/cpu/perf_event_intel_lbr.c
+++ b/arch/x86/kernel/cpu/perf_event_intel_lbr.c
@@ -72,8 +72,6 @@ void intel_pmu_lbr_enable(struct perf_event *event)
 	if (!x86_pmu.lbr_nr)
 		return;
 
-	WARN_ON_ONCE(cpuc->enabled);
-
 	/*
 	 * Reset the LBR stack if we changed task context to
 	 * avoid data leaks.

^ permalink raw reply related	[flat|nested] 2+ messages in thread

* [tip:perf/urgent] perf: Remove deprecated WARN_ON_ONCE()
  2012-02-02 11:04 [PATCH] perf: remove deprecated warn_on_once() Stephane Eranian
@ 2012-02-04 13:55 ` tip-bot for Stephane Eranian
  0 siblings, 0 replies; 2+ messages in thread
From: tip-bot for Stephane Eranian @ 2012-02-04 13:55 UTC (permalink / raw)
  To: linux-tip-commits; +Cc: linux-kernel, eranian, hpa, mingo, tglx, mingo

Commit-ID:  84f2b9b2edc09595569c7397cc3c888764ffd78b
Gitweb:     http://git.kernel.org/tip/84f2b9b2edc09595569c7397cc3c888764ffd78b
Author:     Stephane Eranian <eranian@google.com>
AuthorDate: Thu, 2 Feb 2012 12:04:01 +0100
Committer:  Ingo Molnar <mingo@elte.hu>
CommitDate: Fri, 3 Feb 2012 08:24:40 +0100

perf: Remove deprecated WARN_ON_ONCE()

With the new throttling/unthrottling code introduced with
commit:

  e050e3f0a71b ("perf: Fix broken interrupt rate throttling")

we occasionally hit two WARN_ON_ONCE() checks in:

  - intel_pmu_pebs_enable()
  - intel_pmu_lbr_enable()
  - x86_pmu_start()

The assertions are no longer problematic. There is a valid
path where they can trigger but it is harmless.

The assertion can be triggered with:

  $ perf record -e instructions:pp ....

Leading to paths:

  intel_pmu_pebs_enable
  intel_pmu_enable_event
  x86_perf_event_set_period
  x86_pmu_start
  perf_adjust_freq_unthr_context
  perf_event_task_tick
  scheduler_tick

And:

  intel_pmu_lbr_enable
  intel_pmu_enable_event
  x86_perf_event_set_period
  x86_pmu_start
  perf_adjust_freq_unthr_context.
  perf_event_task_tick
  scheduler_tick

cpuc->enabled is always on because when we get to
perf_adjust_freq_unthr_context() the PMU is not totally
disabled. Furthermore when we need to adjust a period,
we only stop the event we need to change and not the
entire PMU. Thus, when we re-enable, cpuc->enabled is
already set. Note that when we stop the event, both
pebs and lbr are stopped if necessary (and possible).

Signed-off-by: Stephane Eranian <eranian@google.com>
Cc: peterz@infradead.org
Link: http://lkml.kernel.org/r/20120202110401.GA30911@quad
Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 arch/x86/kernel/cpu/perf_event.c           |    3 ---
 arch/x86/kernel/cpu/perf_event_intel_ds.c  |    1 -
 arch/x86/kernel/cpu/perf_event_intel_lbr.c |    2 --
 3 files changed, 0 insertions(+), 6 deletions(-)

diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c
index 5adce10..2a30e5a 100644
--- a/arch/x86/kernel/cpu/perf_event.c
+++ b/arch/x86/kernel/cpu/perf_event.c
@@ -986,9 +986,6 @@ static void x86_pmu_start(struct perf_event *event, int flags)
 	struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events);
 	int idx = event->hw.idx;
 
-	if (WARN_ON_ONCE(!(event->hw.state & PERF_HES_STOPPED)))
-		return;
-
 	if (WARN_ON_ONCE(idx == -1))
 		return;
 
diff --git a/arch/x86/kernel/cpu/perf_event_intel_ds.c b/arch/x86/kernel/cpu/perf_event_intel_ds.c
index 73da6b6..d6bd49f 100644
--- a/arch/x86/kernel/cpu/perf_event_intel_ds.c
+++ b/arch/x86/kernel/cpu/perf_event_intel_ds.c
@@ -439,7 +439,6 @@ void intel_pmu_pebs_enable(struct perf_event *event)
 	hwc->config &= ~ARCH_PERFMON_EVENTSEL_INT;
 
 	cpuc->pebs_enabled |= 1ULL << hwc->idx;
-	WARN_ON_ONCE(cpuc->enabled);
 
 	if (x86_pmu.intel_cap.pebs_trap && event->attr.precise_ip > 1)
 		intel_pmu_lbr_enable(event);
diff --git a/arch/x86/kernel/cpu/perf_event_intel_lbr.c b/arch/x86/kernel/cpu/perf_event_intel_lbr.c
index 3fab3de..47a7e63 100644
--- a/arch/x86/kernel/cpu/perf_event_intel_lbr.c
+++ b/arch/x86/kernel/cpu/perf_event_intel_lbr.c
@@ -72,8 +72,6 @@ void intel_pmu_lbr_enable(struct perf_event *event)
 	if (!x86_pmu.lbr_nr)
 		return;
 
-	WARN_ON_ONCE(cpuc->enabled);
-
 	/*
 	 * Reset the LBR stack if we changed task context to
 	 * avoid data leaks.

^ permalink raw reply related	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2012-02-04 13:55 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2012-02-02 11:04 [PATCH] perf: remove deprecated warn_on_once() Stephane Eranian
2012-02-04 13:55 ` [tip:perf/urgent] perf: Remove deprecated WARN_ON_ONCE() tip-bot for Stephane Eranian

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).