From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753390Ab3LVI50 (ORCPT ); Sun, 22 Dec 2013 03:57:26 -0500 Received: from moutng.kundenserver.de ([212.227.126.187]:62397 "EHLO moutng.kundenserver.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751135Ab3LVI5Z (ORCPT ); Sun, 22 Dec 2013 03:57:25 -0500 Message-ID: <1387702631.5412.101.camel@marge.simpson.net> Subject: Re: [PATCH] rcu: Eliminate softirq processing from rcutree From: Mike Galbraith To: Sebastian Andrzej Siewior Cc: linux-rt-users@vger.kernel.org, Steven Rostedt , linux-kernel@vger.kernel.org, Thomas Gleixner , "Paul E. McKenney" Date: Sun, 22 Dec 2013 09:57:11 +0100 In-Reply-To: <1387681647.5412.25.camel@marge.simpson.net> References: <20131221193900.GA8427@linutronix.de> <1387681647.5412.25.camel@marge.simpson.net> Content-Type: text/plain; charset="UTF-8" X-Mailer: Evolution 3.2.3 Content-Transfer-Encoding: 7bit Mime-Version: 1.0 X-Provags-ID: V02:K0:EatatWPMffuAJ1Fj4jwQnhix30SkObingYGMWTDXKlt GrulPyzCpeTvUF3QRQ+d1RhTuRwD1GBDjirdgTrYNpEk5Fxx7V t5HEhsPWc9B6ipGxe0/LURgvM6bd8lzs1x3UaNs0upOxjf/Wa9 mLI9ckjFzT/pnW1DeeEuZruYfCZwpDlYIWZtcdSYUQRSdl/2yn dOV9gE9G651aTvV7gYWkVclgkAb+tjbGUCvwWIOEcoQ81IRNjA LpkpX3ydxlvRzEUclOk423z3Nv6YhI7v3IplvWSJvlZrJCK26I zgXiVkZM4RGLXmajmfsS0ZKZonF1yBOZ7WPB9+Oc//pbyRXvPR +og/c5tH0UlYjBOI0vfARIyGHlyRZ7iDlIehrYCbPJZEixJJEa tBt8xmf1W9iMQ== Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Sun, 2013-12-22 at 04:07 +0100, Mike Galbraith wrote: > On Sat, 2013-12-21 at 20:39 +0100, Sebastian Andrzej Siewior wrote: > > From: "Paul E. McKenney" > > > > Running RCU out of softirq is a problem for some workloads that would > > like to manage RCU core processing independently of other softirq work, > > for example, setting kthread priority. This commit therefore moves the > > RCU core work from softirq to a per-CPU/per-flavor SCHED_OTHER kthread > > named rcuc. The SCHED_OTHER approach avoids the scalability problems > > that appeared with the earlier attempt to move RCU core processing to > > from softirq to kthreads. That said, kernels built with RCU_BOOST=y > > will run the rcuc kthreads at the RCU-boosting priority. > > I'll take this for a spin on my 64 core test box. > > I'm pretty sure I'll still end up having to split softirq threads again > though, as big box has been unable to meet jitter requirements without, > and last upstream rt kernel tested still couldn't. Still can't fwiw, but whatever, back to $subject. I'll let the box give RCU something to do for a couple days. No news is good news. -Mike 30 minute isolated core jitter test says tinkering will definitely be required. 3.0-rt does single digit worst case on same old box. Darn. (test is imperfect, but good enough) FREQ=960 FRAMES=1728000 LOOP=50000 using CPUs 4 - 23 FREQ=1000 FRAMES=1800000 LOOP=48000 using CPUs 24 - 43 FREQ=300 FRAMES=540000 LOOP=160000 using CPUs 44 - 63 on your marks... get set... POW! Cpu Frames Min Max(Frame) Avg Sigma LastTrans Fliers(Frames) 4 1727979 0.0159 181.66 (1043545)0.4492 0.5876 0 (0) 16 (828505,828506,859225,859226,889945,..1043546) 5 1727980 0.0159 181.90 (1013305)0.4560 0.6118 0 (0) 16 (798265,798266,828985,828986,859705,..1013306) 6 1727981 0.0159 189.05 (1013785)0.3691 0.6225 0 (0) 16 (798745,798746,829465,829466,860185,..1013786) 7 1727982 0.0159 177.88 (983546) 0.2885 0.5269 0 (0) 16 (768505,768506,799225,799226,829945,..983546) 8 1727984 0.0159 192.63 (984025) 0.3131 0.6307 0 (0) 18 (738265,738266,768985,768986,799705,..984026) 9 1727985 0.0159 16.43 (801406) 0.6562 0.5794 0 (0) 10 1727986 0.0159 186.94 (954266) 0.3514 0.6252 0 (0) 16 (739225,739226,769945,769946,800665,..954266) 11 1727987 0.0159 194.06 (954745) 0.4341 0.6547 0 (0) 18 (708985,708986,739705,739706,770425,..954746) 12 1727989 0.0159 13.61 (67116) 0.3364 0.4294 0 (0) 13 1727990 0.0159 186.19 (894265) 0.3955 0.6113 0 (0) 16 (679225,679226,709945,709946,740665,..894266) 14 1727991 0.0159 192.18 (894746) 0.4410 0.6449 0 (0) 18 (648985,648986,679705,679706,710425,..894746) 15 1727993 0.0159 183.36 (833786) 0.5582 0.6655 0 (0) 16 (618745,618746,649465,649466,680185,..833786) 16 1727994 0.0159 193.61 (895706) 0.6073 0.7382 0 (0) 17 (649945,680665,680666,711385,711386,..895706) 17 1727995 0.0159 36.94 (739943) 0.7135 0.7543 0 (0) 6 (173558,173559,739943,739944,1224751,1224752) 18 1727996 0.0159 167.39 (835226) 0.8385 0.8287 0 (0) 16 (620185,620186,650905,650906,681625,..835226) 19 1727997 0.0159 172.84 (804985) 0.5110 0.6959 0 (0) 17 (589946,620665,620666,651385,651386,..835706) 20 1727999 0.0159 180.47 (774745) 0.7566 0.7562 0 (0) 16 (559705,559706,590425,590426,621145,..774746) 21 1728000 0.0159 169.74 (744505) 0.7719 0.8154 0 (0) 16 (560185,560186,590905,590906,621625,..775226) 22 1728000 0.0159 194.80 (836667) 0.6799 0.7063 0 (0) 16 (590906,590907,622105,622106,652346,..836667) 23 1728000 0.0159 183.12 (745466) 0.6733 0.7091 0 (0) 16 (530425,530426,561145,561146,591865,..745466) 24 1800000 0.0725 7.46 (132730) 0.5375 0.4462 0 (0) 25 1800000 0.0725 7.23 (132730) 0.5725 0.4816 0 (0) 26 1800000 0.0725 7.23 (132730) 0.5119 0.4194 0 (0) 27 1800000 0.0725 4.93 (132730) 0.4102 0.3379 0 (0) 28 1800000 0.0725 5.08 (444312) 0.4275 0.3510 0 (0) 29 1800000 0.0725 6.75 (132717) 0.5501 0.5232 0 (0) 30 1800000 0.0725 11.61 (12026) 0.3811 0.3934 0 (0) 31 1800000 0.0725 11.61 (12526) 0.4054 0.4551 0 (0) 32 1800000 0.0725 50.95 (13026) 0.6015 0.5617 0 (0) 31 (13026,13027,45026,45027,77026,..909027) 33 1800000 0.0725 62.63 (13526) 0.5643 0.5922 0 (0) 112 (13526,13527,45526,45527,77526,..1773527) 34 1800000 0.0725 70.26 (14026) 0.3698 0.6132 0 (0) 112 (14026,14027,46026,46027,78026,..1774027) 35 1800000 0.0725 84.57 (14526) 0.6490 0.7981 0 (0) 112 (14526,14527,46526,46527,78526,..1774527) 36 1800000 0.0725 81.94 (943026) 0.3917 0.6387 0 (0) 112 (15026,15027,47026,47027,79026,..1775027) 37 1800000 0.0725 93.86 (15526) 0.6346 0.8580 0 (0) 112 (15526,15527,47526,47527,79526,..1775527) 38 1800000 0.0725 82.66 (144026) 0.4776 0.7459 0 (0) 112 (16026,16027,48026,48027,80026,..1776027) 39 1800000 0.0725 96.63 (16527) 0.4559 0.6881 0 (0) 112 (16526,16527,48526,48527,80526,..1776527) 40 1800000 0.0725 169.44 (17026) 0.4103 1.2801 0 (0) 112 (17026,17027,49026,49027,81026,..1777027) 41 1800000 0.0725 172.07 (145526) 0.6840 1.4300 0 (0) 112 (17526,17527,49526,49527,81526,..1777527) 42 1800000 0.0725 180.41 (18026) 0.5174 1.4290 0 (0) 112 (18026,18027,50026,50027,82026,..1778027) 43 1800000 0.0725 193.52 (466526) 1.4156 2.1665 0 (0) 112 (18526,18527,50526,50527,82526,..1778527) 44 540000 0.0032 11.92 (465332) 0.6862 0.7986 0 (0) 45 540000 0.0032 14.30 (452401) 1.2460 0.9368 0 (0) 46 540000 0.0032 13.35 (452402) 1.1379 0.9079 0 (0) 47 540000 0.0032 12.64 (457991) 1.0116 0.8752 0 (0) 48 540000 0.0032 10.49 (412210) 0.6312 0.7135 0 (0) 49 540000 0.0032 10.01 (412210) 0.5195 0.6649 0 (0) 50 540000 0.0032 9.53 (382891) 0.4586 0.6074 0 (0) 51 540000 0.0032 10.02 (374093) 0.5366 0.6469 0 (0) 52 540000 0.0032 13.35 (360263) 0.6725 0.7738 0 (0) 53 540000 0.0032 10.73 (344411) 1.0863 0.9166 0 (0) 54 540000 0.0032 12.16 (345014) 0.8779 0.7472 0 (0) 55 540000 0.0032 11.45 (339953) 0.8549 0.7650 0 (0) 56 540000 0.0032 11.45 (220829) 0.6262 0.7004 0 (0) 57 540000 0.0032 9.77 (209623) 0.5978 0.6652 0 (0) 58 540000 0.0032 9.78 (188624) 0.5481 0.6939 0 (0) 59 540000 0.0032 5.96 (124615) 1.1515 0.8079 0 (0) 60 540000 0.0032 13.11 (162649) 0.7490 0.7415 0 (0) 61 540000 0.0032 11.92 (161653) 0.7996 0.8211 0 (0) 62 540000 0.0032 13.12 (163325) 0.6313 0.8024 0 (0) 63 540000 0.0032 9.30 (182608) 0.5861 0.6881 0 (0)