All of lore.kernel.org
 help / color / mirror / Atom feed
From: Peter Zijlstra <a.p.zijlstra@chello.nl>
To: Oleg Nesterov <oleg@redhat.com>, Jiri Olsa <jolsa@redhat.com>,
	Ingo Molnar <mingo@elte.hu>
Cc: linux-kernel@vger.kernel.org,
	Stephane Eranian <eranian@google.com>,
	Peter Zijlstra <a.p.zijlstra@chello.nl>
Subject: [RFC][PATCH 6/9] perf: Change ctx::is_active semantics
Date: Sat, 09 Apr 2011 21:17:45 +0200	[thread overview]
Message-ID: <20110409192141.930282378@chello.nl> (raw)
In-Reply-To: 20110409191739.813727025@chello.nl

[-- Attachment #1: perf-is_active.patch --]
[-- Type: text/plain, Size: 2210 bytes --]

Instead of tracking if a context is active or not, track which events
of the context are active. By making it a bitmask of
EVENT_PINNED|EVENT_FLEXIBLE we can simplify some of the scheduling
routines since it can avoid adding events that are already active.

Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
---
 kernel/perf_event.c |   99 +++++++++++++++++++++++++---------------------------
 1 file changed, 48 insertions(+), 51 deletions(-)

Index: linux-2.6/kernel/perf_event.c
===================================================================
--- linux-2.6.orig/kernel/perf_event.c
+++ linux-2.6/kernel/perf_event.c
@@ -1780,8 +1775,9 @@ static void ctx_sched_out(struct perf_ev
 			  enum event_type_t event_type)
 {
 	struct perf_event *event;
+	int is_active = ctx->is_active;
 
-	ctx->is_active = 0;
+	ctx->is_active &= ~event_type;
 	if (likely(!ctx->nr_events))
 		return;
 
@@ -1791,12 +1787,12 @@ static void ctx_sched_out(struct perf_ev
 		return;
 
 	perf_pmu_disable(ctx->pmu);
-	if (event_type & EVENT_PINNED) {
+	if ((is_active & EVENT_PINNED) && (event_type & EVENT_PINNED)) {
 		list_for_each_entry(event, &ctx->pinned_groups, group_entry)
 			group_sched_out(event, cpuctx, ctx);
 	}
 
-	if (event_type & EVENT_FLEXIBLE) {
+	if ((is_active & EVENT_FLEXIBLE) && (event_type & EVENT_FLEXIBLE)) {
 		list_for_each_entry(event, &ctx->flexible_groups, group_entry)
 			group_sched_out(event, cpuctx, ctx);
 	}
@@ -2075,8 +2071,9 @@ ctx_sched_in(struct perf_event_context *
 	     struct task_struct *task)
 {
 	u64 now;
+	int is_active = ctx->is_active;
 
-	ctx->is_active = 1;
+	ctx->is_active |= event_type;
 	if (likely(!ctx->nr_events))
 		return;
 
@@ -2087,11 +2084,11 @@ ctx_sched_in(struct perf_event_context *
 	 * First go through the list and put on any pinned groups
 	 * in order to give them the best chance of going on.
 	 */
-	if (event_type & EVENT_PINNED)
+	if (!(is_active & EVENT_PINNED) && (event_type & EVENT_PINNED))
 		ctx_pinned_sched_in(ctx, cpuctx);
 
 	/* Then walk through the lower prio flexible groups */
-	if (event_type & EVENT_FLEXIBLE)
+	if (!(is_active & EVENT_FLEXIBLE) && (event_type & EVENT_FLEXIBLE))
 		ctx_flexible_sched_in(ctx, cpuctx);
 }
 



  parent reply	other threads:[~2011-04-09 19:23 UTC|newest]

Thread overview: 25+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-04-09 19:17 [RFC][PATCH 0/9] perf: Rework event scheduling Peter Zijlstra
2011-04-09 19:17 ` [RFC][PATCH 1/9] perf: Optimize ctx_sched_out Peter Zijlstra
2011-05-28 16:38   ` [tip:perf/core] perf: Optimize ctx_sched_out() tip-bot for Peter Zijlstra
2011-04-09 19:17 ` [RFC][PATCH 2/9] perf: Clean up ctx reference counting Peter Zijlstra
2011-04-11  6:05   ` Lin Ming
2011-04-11  8:35     ` Peter Zijlstra
2011-05-28 16:39   ` [tip:perf/core] perf: Clean up 'ctx' " tip-bot for Peter Zijlstra
2011-04-09 19:17 ` [RFC][PATCH 3/9] perf: Change event scheduling locking Peter Zijlstra
2011-05-28 16:39   ` [tip:perf/core] perf: Optimize " tip-bot for Peter Zijlstra
2011-04-09 19:17 ` [RFC][PATCH 4/9] perf: Remove task_ctx_sched_in Peter Zijlstra
2011-05-28 16:39   ` [tip:perf/core] perf: Remove task_ctx_sched_in() tip-bot for Peter Zijlstra
2011-04-09 19:17 ` [RFC][PATCH 5/9] perf: Simplify and fix __perf_install_in_context Peter Zijlstra
2011-04-10  8:13   ` Peter Zijlstra
2011-04-11  8:44     ` Lin Ming
2011-04-11  8:50       ` Peter Zijlstra
2011-04-11  8:12   ` Lin Ming
2011-05-28 16:40   ` [tip:perf/core] perf: Simplify and fix __perf_install_in_context() tip-bot for Peter Zijlstra
2011-04-09 19:17 ` Peter Zijlstra [this message]
2011-05-28 16:40   ` [tip:perf/core] perf: Change and simplify ctx::is_active semantics tip-bot for Peter Zijlstra
2011-04-09 19:17 ` [RFC][PATCH 7/9] perf: Collect the schedule in rules in one function Peter Zijlstra
2011-05-28 16:41   ` [tip:perf/core] perf: Collect the schedule-in " tip-bot for Peter Zijlstra
2011-04-09 19:17 ` [RFC][PATCH 8/9] perf: Change close() semantics for group events Peter Zijlstra
2011-05-28 16:41   ` [tip:perf/core] " tip-bot for Peter Zijlstra
2011-04-09 19:17 ` [RFC][PATCH 9/9] perf: De-schedule a task context when removing the last event Peter Zijlstra
2011-05-28 16:42   ` [tip:perf/core] " tip-bot for Peter Zijlstra

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20110409192141.930282378@chello.nl \
    --to=a.p.zijlstra@chello.nl \
    --cc=eranian@google.com \
    --cc=jolsa@redhat.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@elte.hu \
    --cc=oleg@redhat.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.