From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id F0CC4C433F5 for ; Fri, 22 Oct 2021 12:01:19 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id D4F82610EA for ; Fri, 22 Oct 2021 12:01:19 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S230519AbhJVMDf (ORCPT ); Fri, 22 Oct 2021 08:03:35 -0400 Received: from mout.gmx.net ([212.227.17.21]:45367 "EHLO mout.gmx.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S230155AbhJVMD3 (ORCPT ); Fri, 22 Oct 2021 08:03:29 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=gmx.net; s=badeba3b8450; t=1634904042; bh=CV0bzqAO8xhY3j1WtbeSZ+D5Dn8OhUt0AEt1/gMj7+U=; h=X-UI-Sender-Class:Subject:From:To:Cc:Date:In-Reply-To:References; b=G6XJ6weTRoS/MvonxLAV4cyBsej5LR0tCDtcT8u9+E+G/0PDBCBzJprrr4b0TUdCT NaOg6kYR97HEmtAd1RVBRXbCqj5mL4uiVl31/IOLDoM9Yl8M1fzjgZqzZScTUTWDD1 t00TQcbSfB5s+eyfeh5ntcIVUhP0pMyPnQdju4CU= X-UI-Sender-Class: 01bb95c1-4bf8-414a-932a-4f6e2808ef9c Received: from homer.fritz.box ([185.221.150.113]) by mail.gmx.net (mrgmx104 [212.227.17.168]) with ESMTPSA (Nemesis) id 1MxUrx-1mtWEa3Z9t-00xtRH; Fri, 22 Oct 2021 14:00:41 +0200 Message-ID: <2fd273f7284c78b65927b2b2572a66985cee0199.camel@gmx.de> Subject: Re: [PATCH 1/2] sched/fair: Couple wakee flips with heavy wakers From: Mike Galbraith To: Mel Gorman Cc: Peter Zijlstra , Ingo Molnar , Vincent Guittot , Valentin Schneider , Aubrey Li , Barry Song , Srikar Dronamraju , LKML Date: Fri, 22 Oct 2021 14:00:38 +0200 In-Reply-To: <20211022110534.GJ3959@techsingularity.net> References: <20211021145603.5313-1-mgorman@techsingularity.net> <20211021145603.5313-2-mgorman@techsingularity.net> <37d8c167df66a1ead16b699115548ca376494c0c.camel@gmx.de> <20211022110534.GJ3959@techsingularity.net> Content-Type: text/plain; charset="UTF-8" User-Agent: Evolution 3.42.0 MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-Provags-ID: V03:K1:Lz2y05m0t9AO/lc7DM+cX8VtslQiwOxSUt32F3LvNV1JGJRZI/t Ib/HxfG11By5l8GBM3FNgywxdZlngcszMfhuPdGQxWFiEa7M84RhciWp0Q5l8dOkTKVgpb9 40rZymR42Qek2ClqO+LcBCd76llDIg5ETDksBGjscpvx2soAapSrLPzzgK2KZeSpx+LwzWy mZynS62cmseDCKo3j9izA== X-UI-Out-Filterresults: notjunk:1;V03:K0:XgErVOZclIw=:q3TMD6jxCCH+zC+R0+Y9hj +Uhmv/LwxP9sJE6OfCFEQXd8uhdprmXSNkFwidCM/uB76ezodckodZJO1b5So907IEiJe/uUk /knVw57kW3hUqyOYCiC7nBcO5FNQjEh8NCLQaqk7ZVY7CBoIKiUdjNflNN84o2sQK6+5/cWo2 4yHnpQs2YWl/j5+Kpv82zrCajNvLmya4Bhj+Ds49ZHzgFzzoJqhRHmJFI2Z4SZ/hyBZH3swMW ILvAt/rCe/uxQsoJnnUO/4GMnCFwJet7vfUsdnyLQrz5qXKDm2BsJNpIVBwJKFbnc8XvFSoUT yJHtGg1S3/O91rlMFkOwmJOdlaRwt7Gcv/MdKCzhaTRiZZ56BiOvc4/pn9sHX2iVNDwMqY9Ve Wy7Tl2YBkx7B8q58mzOVwM4W+fzqY0DMK+GntgplyOVjFLjwwb60cGNjT6nw8QyGlYrK4/Ikk ngvPA20dluYuTTLPGzkYlrVv4ftKCPPiI/mD66z2QkrwXrvCwPmpDbb9BsvHYxaw1PBkcPKP0 rBmy7gUX58CRlIZoNJQh8clv5LOITPescc3fkRdRTBJ/7VDFJ/KNwsGyMBws/3bDqhGX5p/pW /iuMtUCiYDdbfYddiKMj13TY0kLwP6tiDbgY8iWGxyJqJpihSD/Yv4mx+jMWEr2aNhAnTVFAR /qJeuB8YSGdwyo6cK5YYhYY9Fa18/ukXprqmJrEPEOX4yqUEaglgFRjQMqn94dQhJ2HdYF0gp DuWgmNIN4vLfDZFnB6r1rOjukX8vcOeTX+j82xch+7vA48bzOwPC6v4ELCxDhem8ecK62QPd6 se/SAbpuPXnVw2Ur3EJXiv6lXyhiDDIHsRyXmSta9qbLpR/NRaXoCGuMl6AJ/HhzGthhJHyRT dKyAIthXruDEiBy4zfCc+VRl+amOmR2IYeMhhftQCBrbnlS4xPpQog/nAWgDg29wSjMIe0zHy XdSRwnMMH0YTXfNN31YB9z5nG65GEn0WgizJzo1ipdKJTonCne0RB+2ibZM0zxAHknynW8xiC +1tW+sgot7SXqjs26gVV4OeDG9IdMhWNFou0Rbh6iTD9GeNiXstHt8kdnABEp4O/h4WsBaFvF 8Rd4oIijh0qSik= Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Fri, 2021-10-22 at 12:05 +0100, Mel Gorman wrote: > On Fri, Oct 22, 2021 at 12:26:08PM +0200, Mike Galbraith wrote: > > > > > Hm.=C2=A0 If patchlet repeatably impacts buddy pairs one way or the ot= her, > > it should probably be tossed out the nearest window. > > > > I don't see how buddy pairing would be impacted although there is likely > differences in the degree tasks get preempted due to pulling tasks. Hohum, numbers get to say whatever (weirdness) they want to I suppose. btw, below are some desktop impact numbers I had collected. box =3D i7-4790 quad+smt desktop vs massive_intr 8 9999 (8 x 8ms run/1ms sleep, for 9999 secs.. eff= ectively forever) perf sched record -a -- su mikeg -c 'firefox https://www.youtube.com/watch= ?v=3Daqz-KE-bpKQ'& sleep 300 && killall perf firefox runtime runtime sum delay = sum delay sum delay switches desktop patch/features total util massive_intr util total = massive_intr desktop total/massive util virgin/stock 2267347.921 ms 94.4% 1932675.152 ms 80.5% 158611.016= ms 133309.938 ms 25301.078 ms 594780/441157 13.9% virgin/-wa_weight 2236871.408 ms 93.2% 1881464.401 ms 78.3% 255785.391= ms 243958.616 ms 11826.775 ms 1525470/1424083 14.8% -1.34% -1.2% -2.2% = -13.474 s +0.9% wake_wide/stock 2254335.961 ms 93.9% 1917834.157 ms 79.9% 164766.194= ms 141974.540 ms 22791.654 ms 720711/599064 14.0% While patchlet mitigated the stacking somewhat, it wasn't meaningful to my test load (BigBuckBunny for the 10001th time). PELT clearly stacks up the desktop pretty damn badly, but gets away with it.. in this case. OTOH, killing stacking graveyard dead via NO_WA_WEIGHT would have certainly dinged up cache quite a bit for the compute load had it been something other than a synthetic CPU burner, so there's a brownie point for PELT in the mix to go along with its stacking demerit. Given there was zero perceptible difference, the only thing patchlet really did was to give me a warm fuzzy knowing it was in there fighting the good fight against obscene *looking* stacking (with potential). -Mike