sched: Tweak wake_idx

When merging select_task_rq_fair() and sched_balance_self() we lost
the use of wake_idx, restore that and set them to 0 to make wake
balancing more aggressive.

Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
This commit is contained in:
Peter Zijlstra 2009-09-03 13:16:51 +02:00 committed by Ingo Molnar
parent d7c33c4930
commit 78e7ed53c9
7 changed files with 28 additions and 11 deletions

View file

@ -120,7 +120,7 @@ int arch_update_cpu_topology(void);
.imbalance_pct = 125, \
.cache_nice_tries = 1, \
.busy_idx = 2, \
.wake_idx = 1, \
.wake_idx = 0, \
.forkexec_idx = 1, \
\
.flags = 1*SD_LOAD_BALANCE \
@ -152,7 +152,7 @@ int arch_update_cpu_topology(void);
.busy_idx = 2, \
.idle_idx = 1, \
.newidle_idx = 2, \
.wake_idx = 1, \
.wake_idx = 0, \
.forkexec_idx = 1, \
\
.flags = 1*SD_LOAD_BALANCE \