Re: 2.6.16-rc1: 28ms latency when process with lots of swapped memory exits

[Lee Revell <rlrevell@joe-job.com>]

On Tue, 2006-03-14 at 18:40 +0000, Hugh Dickins wrote:
> On Tue, 14 Mar 2006, Lee Revell wrote:
> 
> > I've been testign for weeks with 2.6.16-rc1 + the latency trace patch
> > and the longest latencies measured were 10-15ms due to the well known
> > rt_run_flush issue.  Today I got one twice as long, when a Firefox
> > process with a bunch of acroreads in tabs, from a new code path.
> > 
> > It seems to trigger when a process with a large amount of memory swapped
> > out exits.
> > 
> > Can this be solved with a cond_resched?
> 
> Not that easily, I think.
> 
> Are you testing with CONFIG_PREEMPT=y, as I'd expect?  I thought
> cond_resched() adds nothing to that case (and we keep on intending
> but forgetting to make it compile away to nothing in that case).
> Or am I confused?
> 

Thanks for the explanation - I am the one confused (PREEMPT is on so
cond_resched would be pointless).

Lee

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/