lkml.org 
[lkml]   [2011]   [Jul]   [29]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
Date
From
Subject[RFC PATCH] sched: Kill WAKEUP_PREEMPT
On Fri, Jul 29, 2011 at 09:56:58AM +0200, Peter Zijlstra wrote:
> I've queued Lin's patch as I don't see the point of this thing either,
> normally WAKEUP_PREEMPT is enabled so it says || 0 which is kinda
> useless :-)
>
> And I'm starting to think we should just kill all of WAKEUP_PREEMPT I
> don't think we ever want to disable it anyway..

Someting like this?

---
From: Yong Zhang <yong.zhang0@gmail.com>
Subject: [RFC PATCH] sched: Kill WAKEUP_PREEMPT
Per Peter Zijlstra:
> And I'm starting to think we should just kill all of WAKEUP_PREEMPT I
> don't think we ever want to disable it anyway..

Signed-off-by: Yong Zhang <yong.zhang0@gmail.com>
---
kernel/sched_fair.c | 8 +-------
kernel/sched_features.h | 5 -----
2 files changed, 1 insertions(+), 12 deletions(-)
diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c
index 46b7855..3c58042 100644
--- a/kernel/sched_fair.c
+++ b/kernel/sched_fair.c
@@ -1114,9 +1114,6 @@ check_preempt_tick(struct cfs_rq *cfs_rq, struct sched_entity *curr)
* narrow margin doesn't have to wait for a full slice.
* This also mitigates buddy induced latencies under load.
*/
- if (!sched_feat(WAKEUP_PREEMPT))
- return;
-
if (delta_exec < sysctl_sched_min_granularity)
return;

@@ -1252,7 +1249,7 @@ entity_tick(struct cfs_rq *cfs_rq, struct sched_entity *curr, int queued)
return;
#endif

- if (cfs_rq->nr_running > 1 || !sched_feat(WAKEUP_PREEMPT))
+ if (cfs_rq->nr_running > 1)
check_preempt_tick(cfs_rq, curr);
}

@@ -1918,9 +1915,6 @@ static void check_preempt_wakeup(struct rq *rq, struct task_struct *p, int wake_
return;


- if (!sched_feat(WAKEUP_PREEMPT))
- return;
-
update_curr(cfs_rq);
find_matching_se(&se, &pse);
BUG_ON(!pse);
diff --git a/kernel/sched_features.h b/kernel/sched_features.h
index 1e7066d..878a5f9 100644
--- a/kernel/sched_features.h
+++ b/kernel/sched_features.h
@@ -12,11 +12,6 @@ SCHED_FEAT(GENTLE_FAIR_SLEEPERS, 1)
SCHED_FEAT(START_DEBIT, 1)

/*
- * Should wakeups try to preempt running tasks.
- */
-SCHED_FEAT(WAKEUP_PREEMPT, 1)
-
-/*
* Based on load and program behaviour, see if it makes sense to place
* a newly woken task on the same cpu as the task that woke it --
* improve cache locality. Typically used with SYNC wakeups as
--
1.7.4.1

--
Only stand for myself


\
 
 \ /
  Last update: 2011-07-29 10:23    [from the cache]
©2003-2011 Jasper Spaans