Re: CONFIG_NO_HZ breaks blktrace timestamps

[Ingo Molnar]

* Ingo Molnar <mingo@xxxxxxx> wrote:

> > With those patches, CONFIG_NO_HZ works just fine.
> 
> could you please try the two patches below, do they fix the problem as 
> well? They got a ton of testing in x86.git in the past ~2 months and 
> we could perhaps still push them into v2.6.24.

plus, if needed, perhaps this patch below from Guillaume ontop of it.

	Ingo

---------------->
Subject: sched: fix rq->clock warps on frequency changes
From: Guillaume Chazarain <guichaz@xxxxxxxx>

sched: fix rq->clock warps on frequency changes

Fix 2bacec8c318ca0418c0ee9ac662ee44207765dd4
(sched: touch softlockup watchdog after idling) that reintroduced warps
on frequency changes. touch_softlockup_watchdog() calls __update_rq_clock
that checks rq->clock for warps, so call it after adjusting rq->clock.

Signed-off-by: Guillaume Chazarain <guichaz@xxxxxxxx>
Signed-off-by: Ingo Molnar <mingo@xxxxxxx>
---
 kernel/sched.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

Index: linux-x86.q/kernel/sched.c
===================================================================
--- linux-x86.q.orig/kernel/sched.c
+++ linux-x86.q/kernel/sched.c
@@ -668,7 +668,6 @@ void sched_clock_idle_wakeup_event(u64 d
 	struct rq *rq = cpu_rq(smp_processor_id());
 	u64 now = sched_clock();
 
-	touch_softlockup_watchdog();
 	rq->idle_clock += delta_ns;
 	/*
 	 * Override the previous timestamp and ignore all
@@ -680,6 +679,7 @@ void sched_clock_idle_wakeup_event(u64 d
 	rq->prev_clock_raw = now;
 	rq->clock += delta_ns;
 	spin_unlock(&rq->lock);
+	touch_softlockup_watchdog();
 }
 EXPORT_SYMBOL_GPL(sched_clock_idle_wakeup_event);
 
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/