lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
Open Source and information security mailing list archives
| ||
|
Date: Tue, 04 Jun 2013 09:45:46 +0800 From: Alex Shi <alex.shi@...el.com> To: Peter Zijlstra <a.p.zijlstra@...llo.nl> CC: Paul Turner <pjt@...gle.com>, Ingo Molnar <mingo@...hat.com>, Thomas Gleixner <tglx@...utronix.de>, Andrew Morton <akpm@...ux-foundation.org>, Borislav Petkov <bp@...en8.de>, Namhyung Kim <namhyung@...nel.org>, Mike Galbraith <efault@....de>, Morten Rasmussen <morten.rasmussen@....com>, Vincent Guittot <vincent.guittot@...aro.org>, Preeti U Murthy <preeti@...ux.vnet.ibm.com>, Viresh Kumar <viresh.kumar@...aro.org>, LKML <linux-kernel@...r.kernel.org>, Mel Gorman <mgorman@...e.de>, Rik van Riel <riel@...hat.com>, Michael Wang <wangyun@...ux.vnet.ibm.com> Subject: Re: [PATCH v5 5/7] sched: compute runnable load avg in cpu_load and cpu_avg_load_per_task On 05/07/2013 02:17 PM, Alex Shi wrote: > On 05/06/2013 07:10 PM, Peter Zijlstra wrote: >>>> The runnable_avgs themselves actually have a fair bit of history in >>>> them already (50% is last 32ms); but given that they don't need to be >>>> cut-off to respond to load being migrated I'm guessing we could >>>> actually potentially get by with just "instaneous" and "use averages" >>>> where appropriate? >> Sure,. worth a try. If things fall over we can always look at it again. >> >>>> We always end up having to re-pick/tune them based on a variety of >>>> workloads; if we can eliminate them I think it would be a win. >> Agreed, esp. the plethora of weird idx things we currently have. If we need to >> re-introduce something it would likely only be the busy case and for that we >> can immediately link to the balance interval or so. >> >> >> > > I like to have try bases on this patchset. :) > > First, we can remove the idx, to check if the removing is fine for our > benchmarks, kbuild, dbench, tbench, hackbench, aim7, specjbb etc. > > If there are some regression. we can think more. > Peter, I just tried to remove the variety rq.cpu_load, by the following patch. Because forkexec_idx and busy_idx are all zero, after the patch system just keep cpu_load[0] and remove other values. I tried the patch base 3.10-rc3 and latest tip/sched/core with benchmark dbench,tbench, aim7,hackbench. and oltp of sysbench. Seems performance doesn't change clear. So, for my tested machines, core2, NHM, SNB, with 2 or 4 CPU sockets, and above tested benchmark. We are fine to remove the variety cpu_load. Don't know if there some other concerns on other scenarios. --- diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index 590d535..f0ca983 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -4626,7 +4626,7 @@ static inline void update_sd_lb_stats(struct lb_env *env, if (child && child->flags & SD_PREFER_SIBLING) prefer_sibling = 1; - load_idx = get_sd_load_idx(env->sd, env->idle); + load_idx = 0; //get_sd_load_idx(env->sd, env->idle); do { int local_group; -- Thanks Alex -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@...r.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists