atom feed67 messages in org.freebsd.freebsd-hackersRe: [RFT][patch] Scheduling for HTT a...
FromSent OnAttachments
Alexander MotinFeb 5, 2012 11:04 pm 
David XuFeb 5, 2012 11:59 pm 
Gary JennejohnFeb 6, 2012 2:08 am 
Alexander BestFeb 6, 2012 8:01 am 
Alexander MotinFeb 6, 2012 8:28 am 
Tijl CoosemansFeb 6, 2012 9:37 am 
Alexander MotinFeb 6, 2012 9:54 am 
Florian SmeetsFeb 6, 2012 11:07 am 
Alexander BestFeb 6, 2012 11:10 am 
Alexander MotinFeb 6, 2012 11:18 am 
Julian ElischerFeb 6, 2012 10:10 pm 
Ivan VorasFeb 8, 2012 3:06 am 
Andriy GaponFeb 11, 2012 5:34 am 
Alexander MotinFeb 11, 2012 6:21 am 
Konstantin BelousovFeb 11, 2012 7:35 am 
Andriy GaponFeb 11, 2012 9:04 am 
Alexander MotinFeb 13, 2012 11:56 am 
Jeff RobersonFeb 13, 2012 12:23 pm 
Alexander MotinFeb 13, 2012 12:54 pm 
Jeff RobersonFeb 13, 2012 1:39 pm 
Alexander MotinFeb 13, 2012 2:38 pm 
Alexander MotinFeb 15, 2012 11:46 am 
Jeff RobersonFeb 15, 2012 11:54 am 
Alexander MotinFeb 15, 2012 12:06 pm 
Alexander MotinFeb 15, 2012 8:41 pm 
Alexander MotinFeb 16, 2012 12:48 am 
Alexander MotinFeb 16, 2012 2:58 am 
Florian SmeetsFeb 16, 2012 1:28 pm 
Alexander MotinFeb 17, 2012 8:29 am 
Arnaud LacombeFeb 17, 2012 8:52 am 
Alexander MotinFeb 17, 2012 9:02 am 
George MitchellFeb 26, 2012 4:32 pm 
George MitchellFeb 26, 2012 4:37 pm 
Olivier SmedtsFeb 27, 2012 2:34 am 
George MitchellFeb 27, 2012 3:23 am 
Olivier SmedtsFeb 27, 2012 3:27 am 
Andriy GaponFeb 27, 2012 4:41 am 
George MitchellFeb 27, 2012 3:54 pm 
Adrian ChaddMar 2, 2012 3:05 pm 
George MitchellMar 2, 2012 4:14 pm 
Adrian ChaddMar 2, 2012 7:24 pm 
Alexander MotinMar 2, 2012 11:40 pm 
Ivan KlymenkoMar 3, 2012 12:18 am 
Adrian ChaddMar 3, 2012 12:59 am 
Alexander MotinMar 3, 2012 1:12 am 
Alexander MotinMar 3, 2012 4:53 am 
Ivan KlymenkoMar 3, 2012 7:25 am 
Alexander MotinMar 3, 2012 8:30 am 
Mario LoboMar 3, 2012 8:56 am 
Alexander MotinMar 3, 2012 9:56 am 
Ivan KlymenkoMar 3, 2012 11:15 am 
Arnaud LacombeApr 5, 2012 11:11 am 
Alexander MotinApr 5, 2012 11:45 am 
Attilio RaoApr 6, 2012 7:12 am 
Alexander MotinApr 6, 2012 7:26 am 
Attilio RaoApr 6, 2012 7:30 am 
Alexander MotinApr 6, 2012 7:40 am 
Alexander MotinApr 9, 2012 12:57 pm 
Arnaud LacombeApr 10, 2012 9:57 am 
Alexander MotinApr 10, 2012 10:18 am 
Alexander MotinApr 10, 2012 10:53 am 
Arnaud LacombeApr 10, 2012 11:45 am 
Alexander MotinApr 10, 2012 12:13 pm 
Mike MeyerApr 10, 2012 1:04 pm 
Arnaud LacombeApr 10, 2012 1:50 pm 
Mike MeyerApr 10, 2012 2:19 pm 
Adrian ChaddApr 11, 2012 3:19 pm 
Subject:Re: [RFT][patch] Scheduling for HTT and not only
From:Alexander Motin (ma@FreeBSD.org)
Date:Feb 15, 2012 8:41:08 pm
List:org.freebsd.freebsd-hackers

On 02/15/12 21:54, Jeff Roberson wrote:

On Wed, 15 Feb 2012, Alexander Motin wrote:

On 02/14/12 00:38, Alexander Motin wrote:

I see no much point in committing them sequentially, as they are quite orthogonal. I need to make one decision. I am going on small vacation next week. It will give time for thoughts to settle. May be I indeed just clean previous patch a bit and commit it when I get back. I've spent too much time trying to make these things formal and so far results are not bad, but also not so brilliant as I would like. May be it is indeed time to step back and try some more simple solution.

I've decided to stop those cache black magic practices and focus on things that really exist in this world -- SMT and CPU load. I've dropped most of cache related things from the patch and made the rest of things more strict and predictable: http://people.freebsd.org/~mav/sched.htt34.patch

This looks great. I think there is value in considering the other approach further but I would like to do this part first. It would be nice to also add priority as a greater influence in the load balancing as well.

I have feeling that for timeshare/idle threads balance should take into account not a priority, but a nice level. Priority is very unstable thing that is recalculated each time, while nice it is what really describes how much time thread should get in perspective and those values should be shuffled equally between CPUs. But I haven't thought about specific math yet.