commit
507e1231 (timer stats: Optimize by adding quick check to avoid
function calls) introduced a regression in /proc/timer_list.
/proc/timer_list shows now
#0: <
c27d46b0>, tick_sched_timer, S:01, <(null)>, /-1
instead of
#0: <
c27d46b0>, tick_sched_timer, S:01, hrtimer_start, swapper/0
Revert the hrtimer quick check for now. The optimization needs more
thought, but this is neither 2.6.32-rc7 nor stable material.
[ tglx: - Removed unrelated changes from the original patch
- Prevent unneccesary call to timer_stats_update_stats
- massaged the changelog ]
Signed-off-by: Feng Tang <feng.tang@intel.com>
LKML-Reference: <alpine.LFD.2.00.
0911181933540.24119@localhost.localdomain>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: stable@kernel.org
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
static inline void timer_stats_account_hrtimer(struct hrtimer *timer)
{
- if (likely(!timer->start_site))
+ if (likely(!timer_stats_active))
return;
timer_stats_update_stats(timer, timer->start_pid, timer->start_site,
timer->function, timer->start_comm, 0);
static inline void timer_stats_hrtimer_set_start_info(struct hrtimer *timer)
{
- if (likely(!timer_stats_active))
- return;
__timer_stats_hrtimer_set_start_info(timer, __builtin_return_address(0));
}