lkml.org 
[lkml]   [2009]   [Oct]   [26]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
SubjectRe: [RFC V2 PATCH 1/5] cfq-iosched: adapt slice to number of processes doing I/O
From
On Mon, Oct 26, 2009 at 4:10 PM, Jeff Moyer <jmoyer@redhat.com> wrote:
> Corrado Zoccolo <czoccolo@gmail.com> writes:
>>
>> What about:
>> /* interested queues (we consider only the ones with the same priority class) */
>> unsigned iq = cfq_get_avg_queues(cfqd, cfq_class_rt(cfqq));
>> unsigned sync_slice = cfqd->cfq_slice[1];
>> unsigned expected_latency = sync_slice * iq;
>> if (expected_latency > target_latency) {
>>       unsigned base_low_slice = 2 * cfqd->cfq_slice_idle;
>>       /* scale low_slice according to IO priority and sync vs async */
>>       unsigned low_slice = min(slice,  base_low_slice * slice / sync_slice);
>>       /* the adapted slice value is scaled to fit all iqs into the target latency */
>>       slice = max(slice * cfq_target_latency / expected_latency, low_slice);
>> }
>
> Yeah, that looks a ton better.  Thanks for doing that!

Jens, if you like the new version of the code (integrating this last
change), I can rebase my patch set on top of Jeff's, since it is
already queued for 2.6.33 .

Corrado

>
> Cheers,
> Jeff
>
--
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/

\
 
 \ /
  Last update: 2009-10-26 16:19    [W:0.067 / U:0.744 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site