* [PATCH] O12.3 for interactivity
@ 2003-08-04 11:50 Con Kolivas
0 siblings, 0 replies; only message in thread
From: Con Kolivas @ 2003-08-04 11:50 UTC (permalink / raw)
To: linux kernel mailing list; +Cc: Andrew Morton
Here is a small patch against O12.2. This should not change peformance in any
way but removes redundant code (which we all love to do). Ingo pointed out
that new forks wont have any sleep time to earn as sleep_avg. It is also
extrememly unlikely that there won't be any sleep time with nanosecond
timing. Patch applies to 2.6.0-test2-mm4 or test2 patched with O12.2.
Con
diff -Naurp linux-2.6.0-test2-mm4/kernel/sched.c linux-2.6.0-test2-mm4-O12.3/kernel/sched.c
--- linux-2.6.0-test2-mm4/kernel/sched.c 2003-08-04 20:16:02.000000000 +1000
+++ linux-2.6.0-test2-mm4-O12.3/kernel/sched.c 2003-08-04 20:19:10.000000000 +1000
@@ -365,9 +365,6 @@ static void recalc_task_prio(task_t *p,
unsigned long long __sleep_time = now - p->timestamp;
unsigned long sleep_time;
- if (unlikely(!p->timestamp))
- __sleep_time = 0;
-
if (__sleep_time > NS_MAX_SLEEP_AVG)
sleep_time = NS_MAX_SLEEP_AVG;
else
@@ -420,8 +417,7 @@ static void recalc_task_prio(task_t *p,
p->interactive_credit++;
}
}
- } else if (!p->sleep_avg)
- p->interactive_credit--;
+ }
p->prio = effective_prio(p);
}
@@ -454,9 +450,6 @@ static inline void activate_task(task_t
*/
p->activated = 1;
- if (unlikely(!p->timestamp))
- p->activated = 0;
-
p->timestamp = now;
__activate_task(p, rq);
@@ -644,7 +637,6 @@ void wake_up_forked_process(task_t * p)
p->sleep_avg = JIFFIES_TO_NS(sleep_avg);
p->interactive_credit = 0;
- p->timestamp = 0;
p->prio = effective_prio(p);
set_task_cpu(p, smp_processor_id());
^ permalink raw reply [flat|nested] only message in thread
only message in thread, other threads:[~2003-08-04 11:44 UTC | newest]
Thread overview: (only message) (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2003-08-04 11:50 [PATCH] O12.3 for interactivity Con Kolivas
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).