lkml.org 
[lkml]   [2010]   [Jan]   [21]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
Date
From
Subject[tip:perf/urgent] perf: Fix perf_event_do_pending() fallback callsite
Commit-ID:  fe432200abb0d64f409895168d9ad8fbb9d8e6c6
Gitweb: http://git.kernel.org/tip/fe432200abb0d64f409895168d9ad8fbb9d8e6c6
Author: Peter Zijlstra <a.p.zijlstra@chello.nl>
AuthorDate: Mon, 18 Jan 2010 09:08:26 +0100
Committer: Ingo Molnar <mingo@elte.hu>
CommitDate: Thu, 21 Jan 2010 13:40:39 +0100

perf: Fix perf_event_do_pending() fallback callsite

Paul questioned the context in which we should call
perf_event_do_pending(). After looking at that I found that it should be
called from IRQ context these days, however the fallback call-site is
placed in softirq context. Ammend this by placing the callback in the IRQ
timer path.

Reported-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1263374859.4244.192.camel@laptop>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
kernel/timer.c | 3 +--
1 files changed, 1 insertions(+), 2 deletions(-)

diff --git a/kernel/timer.c b/kernel/timer.c
index 15533b7..c61a794 100644
--- a/kernel/timer.c
+++ b/kernel/timer.c
@@ -1198,6 +1198,7 @@ void update_process_times(int user_tick)
run_local_timers();
rcu_check_callbacks(cpu, user_tick);
printk_tick();
+ perf_event_do_pending();
scheduler_tick();
run_posix_cpu_timers(p);
}
@@ -1209,8 +1210,6 @@ static void run_timer_softirq(struct softirq_action *h)
{
struct tvec_base *base = __get_cpu_var(tvec_bases);

- perf_event_do_pending();
-
hrtimer_run_pending();

if (time_after_eq(jiffies, base->timer_jiffies))

\
 
 \ /
  Last update: 2010-01-21 14:59    [W:1.147 / U:0.156 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site