Skip to contentSkip to conversation

perf_events: improve task_sched_in()

PATCH In review [email protected] 2 participants 3 messages Mar 11, 2010

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 perf_events: improve task_sched_in()

This patch is an optimization in perf_event_task_sched_in() to avoid scheduling the events twice in a row. Without it, the perf_disable()/perf_enable() pair is invoked twice, thereby pinned events counts while scheduling flexible events and we go throuh hw_perf_enable() twice. By encapsulating, the whole sequence into perf_disable()/perf_enable() we ensure, hw_perf_enable() is going to be invoked only once because of the refcount protection.

--

 perf_event.c |    4 ++++
 1 file changed, 4 insertions(+)
kernel/perf_event.c +4 -0
@@ -1382,6 +1382,8 @@ void perf_event_task_sched_in(struct task_struct *task)
 	if (cpuctx->task_ctx == ctx)
 		return;
 
+	perf_disable();
+
 	/*
 	 * We want to keep the following priority order:
 	 * cpu pinned (that don't need to move), task pinned,
@@ -1394,6 +1396,8 @@ void perf_event_task_sched_in(struct task_struct *task)
 	ctx_sched_in(ctx, cpuctx, EVENT_FLEXIBLE);
 
 	cpuctx->task_ctx = ctx;
+
+	perf_enable();
 }
 
 #define MAX_INTERRUPTS (~0ULL)
Acked-by on v1 1/1
#1 lore raw
Quoted 6 lines · On Wed, 2010-03-10 at 22:26 -0800, [email protected] wrote:

This patch is an optimization in perf_event_task_sched_in() to avoid scheduling the events twice in a row. Without it, the perf_disable()/perf_enable() pair is invoked twice, thereby pinned events counts while scheduling flexible events and we go throuh hw_perf_enable() twice. By encapsulating, the whole sequence into perf_disable()/perf_enable() we ensure, hw_perf_enable() is going to be invoked only once because of the refcount protection.

Agreed, this makes perfect sense.

Quoted 25 lines

--

 perf_event.c |    4 ++++
 1 file changed, 4 insertions(+)
kernel/perf_event.c +4 -0
@@ -1382,6 +1382,8 @@ void perf_event_task_sched_in(struct task_struct *task)
 	if (cpuctx->task_ctx == ctx)
 		return;
 
+	perf_disable();
+
 	/*
 	 * We want to keep the following priority order:
 	 * cpu pinned (that don't need to move), task pinned,
@@ -1394,6 +1396,8 @@ void perf_event_task_sched_in(struct task_struct *task)
 	ctx_sched_in(ctx, cpuctx, EVENT_FLEXIBLE);
 
 	cpuctx->task_ctx = ctx;
+
+	perf_enable();
 }
 
 #define MAX_INTERRUPTS (~0ULL)
#2 lore raw

[tip:perf/core] perf_events: Improve task_sched_in()

In reply to [email protected]
Commit-ID:  9b33fa6ba0e2f90fdf407501db801c2511121564
Gitweb:     http://git.kernel.org/tip/9b33fa6ba0e2f90fdf407501db801c2511121564
Author:     [email protected] <[email protected]>
AuthorDate: Wed, 10 Mar 2010 22:26:05 -0800
Committer:  Ingo Molnar <[email protected]>
CommitDate: Thu, 11 Mar 2010 15:23:28 +0100

perf_events: Improve task_sched_in()

This patch is an optimization in perf_event_task_sched_in() to avoid scheduling the events twice in a row.

Without it, the perf_disable()/perf_enable() pair is invoked twice, thereby pinned events counts while scheduling flexible events and we go throuh hw_perf_enable() twice.

By encapsulating, the whole sequence into perf_disable()/perf_enable() we ensure, hw_perf_enable() is going to be invoked only once because of the refcount protection.

LKML-Reference: <[email protected]>

---
 kernel/perf_event.c |    4 ++++
 1 files changed, 4 insertions(+), 0 deletions(-)
kernel/perf_event.c +4 -0
index 52c69a3..3853d49 100644
@@ -1368,6 +1368,8 @@ void perf_event_task_sched_in(struct task_struct *task)
 	if (cpuctx->task_ctx == ctx)
 		return;
 
+	perf_disable();
+
 	/*
 	 * We want to keep the following priority order:
 	 * cpu pinned (that don't need to move), task pinned,
@@ -1380,6 +1382,8 @@ void perf_event_task_sched_in(struct task_struct *task)
 	ctx_sched_in(ctx, cpuctx, EVENT_FLEXIBLE);
 
 	cpuctx->task_ctx = ctx;
+
+	perf_enable();
 }
 
 #define MAX_INTERRUPTS (~0ULL)