From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753961AbdCNCKU (ORCPT ); Mon, 13 Mar 2017 22:10:20 -0400 Received: from szxga01-in.huawei.com ([45.249.212.187]:4318 "EHLO dggrg01-dlp.huawei.com" rhost-flags-OK-FAIL-OK-FAIL) by vger.kernel.org with ESMTP id S1752044AbdCNCKT (ORCPT ); Mon, 13 Mar 2017 22:10:19 -0400 Subject: Re: [lkp-robot] [f2fs] 4ac912427c: -33.7% aim7.jobs-per-min regression To: Jaegeuk Kim , Thorsten Leemhuis , kernel test robot References: <20170308012154.GB17420@yexl-desktop> <09a8a523-d374-fed0-219e-5736a90ec4ea@leemhuis.info> <20170313192223.GC41055@jaegeuk.local> CC: , From: Chao Yu Message-ID: <19cd8c99-ed38-7a5a-24c3-ea81acefb395@huawei.com> Date: Tue, 14 Mar 2017 10:09:47 +0800 User-Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:45.0) Gecko/20100101 Thunderbird/45.0 MIME-Version: 1.0 In-Reply-To: <20170313192223.GC41055@jaegeuk.local> Content-Type: text/plain; charset="windows-1252" Content-Transfer-Encoding: 8bit X-Originating-IP: [10.134.22.195] X-CFilter-Loop: Reflected X-Mirapoint-Virus-RAPID-Raw: score=unknown(0), refid=str=0001.0A020202.58C750F4.0199,ss=1,re=0.000,recu=0.000,reip=0.000,cl=1,cld=1,fgs=0, ip=0.0.0.0, so=2014-11-16 11:51:01, dmn=2013-03-21 17:37:32 X-Mirapoint-Loop-Id: eba9d6884394e4c89ff36fe16283f54f Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 2017/3/14 3:22, Jaegeuk Kim wrote: > On 03/13, Thorsten Leemhuis wrote: >> @Chao Yu/@Jaegeuk Kim: I'm considering to add this to the regressions >> report for 4.11; or is there a reason why it shouldn't be considered a >> regression? Ciao, Thorsten > > Hi, > > I'm planning to submit f2fs updates for 4.11-rcX including a patch which > resolves this issue as well, as I expect. > > https://lkml.org/lkml/2017/3/7/813 Sorry for late repay. I expect below patches in Jaegeuk's tree could help to recover the performance as well f2fs: skip scanning free nid bitmap of full NAT blocks f2fs: combine nat_bits and free_nid_bitmap cache Xiaolong, Thorsten Could you help to test these patches? Thanks, > > Thanks, > >> >> On 08.03.2017 02:21, kernel test robot wrote: >>> >>> Greeting, >>> >>> We noticed a -33.7 regression of aim7.jobs-per-min due to commit: >>> >>> commit: 4ac912427c4214d8031d9ad6fbc3bc75e71512df ("f2fs: introduce free nid bitmap") >>> https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git master >>> >>> in testcase: aim7 >>> on test machine: 40 threads Intel(R) Xeon(R) CPU E5-2690 v2 @ 3.00GHz with 384G memory >>> with following parameters: >>> >>> disk: 1BRD_48G >>> fs: f2fs >>> test: disk_wrt >>> load: 3000 >>> cpufreq_governor: performance >>> >>> test-description: AIM7 is a traditional UNIX system level benchmark suite which is used to test and measure the performance of multiuser system. >>> test-url: https://sourceforge.net/projects/aimbench/files/aim-suite7/ >>> >>> >>> >>> Details are as below: >>> --------------------------------------------------------------------------------------------------> >>> >>> >>> To reproduce: >>> >>> git clone git://git.kernel.org/pub/scm/linux/kernel/git/wfg/lkp-tests.git >>> cd lkp-tests >>> bin/lkp install job.yaml # job file is attached in this email >>> bin/lkp run job.yaml >>> >>> testcase/path_params/tbox_group/run: aim7/1BRD_48G-f2fs-disk_wrt-3000-performance/lkp-ivb-ep01 >>> >>> ced2c7ea8e99b467 4ac912427c4214d8031d9ad6fb >>> ---------------- -------------------------- >>> %stddev change %stddev >>> \ | \ >>> 117419 ± 1% -33.7% 77863 ± 0% aim7.jobs-per-min >>> 153.78 ± 1% +50.6% 231.63 ± 0% aim7.time.elapsed_time >>> 153.78 ± 1% +50.6% 231.63 ± 0% aim7.time.elapsed_time.max >>> 805644 ± 3% +11.3% 896604 ± 0% aim7.time.involuntary_context_switches >>> 5408 ± 1% +15.4% 6240 ± 0% aim7.time.system_time >>> 5066069 ± 0% +10.5% 5600256 ± 9% meminfo.DirectMap2M >>> 135538 ± 8% -41.9% 78738 ± 8% meminfo.Dirty >>> 980.67 ± 16% -67.8% 315.50 ± 12% meminfo.Writeback >>> 71322 ± 10% -44.0% 39953 ± 1% numa-meminfo.node0.Dirty >>> 11158 ± 18% -27.1% 8132 ± 0% numa-meminfo.node0.Mapped >>> 56776 ± 6% -32.5% 38309 ± 0% numa-meminfo.node1.Dirty >>> 9684 ± 22% +30.9% 12676 ± 0% numa-meminfo.node1.Mapped >>> 6069 ± 57% -78.1% 1328 ± 18% softirqs.NET_RX >>> 619333 ± 4% +8.0% 669152 ± 3% softirqs.RCU >>> 128030 ± 2% +33.3% 170724 ± 0% softirqs.SCHED >>> 2331994 ± 1% +15.3% 2688290 ± 0% softirqs.TIMER >>> 7701 ± 1% -35.7% 4948 ± 3% vmstat.io.bo >>> 64.67 ± 2% -39.7% 39.00 ± 2% vmstat.procs.b >>> 333.33 ± 7% -48.5% 171.50 ± 2% vmstat.procs.r >>> 17530 ± 1% -23.4% 13425 ± 1% vmstat.system.cs >>> 47642 ± 1% -5.3% 45100 ± 1% vmstat.system.in >>> 33522 ± 4% -43.1% 19068 ± 0% proc-vmstat.nr_dirty >>> 236.00 ± 14% -66.1% 80.00 ± 3% proc-vmstat.nr_writeback >>> 33907 ± 4% -43.3% 19222 ± 0% proc-vmstat.nr_zone_write_pending >>> 28194 ± 10% +10.4% 31131 ± 6% proc-vmstat.pgactivate >>> 746402 ± 2% +24.6% 929960 ± 3% proc-vmstat.pgfault >>> 153.78 ± 1% +50.6% 231.63 ± 0% time.elapsed_time >>> 153.78 ± 1% +50.6% 231.63 ± 0% time.elapsed_time.max >>> 805644 ± 3% +11.3% 896604 ± 0% time.involuntary_context_switches >>> 3524 ± 0% -23.4% 2701 ± 0% time.percent_of_cpu_this_job_got >>> 5408 ± 1% +15.4% 6240 ± 0% time.system_time >>> 12.19 ± 1% +36.7% 16.66 ± 0% time.user_time >>> 48260939 ± 3% +12.1% 54110616 ± 2% cpuidle.C1-IVT.time >>> 33149237 ± 5% +52.6% 50597349 ± 1% cpuidle.C1E-IVT.time >>> 89642 ± 4% +52.8% 136976 ± 0% cpuidle.C1E-IVT.usage >>> 13534795 ± 6% +276.3% 50934566 ± 55% cpuidle.C3-IVT.time >>> 42893 ± 6% +138.8% 102439 ± 30% cpuidle.C3-IVT.usage >>> 6.431e+08 ± 2% +390.1% 3.152e+09 ± 10% cpuidle.C6-IVT.time >>> 802009 ± 2% +375.3% 3811880 ± 10% cpuidle.C6-IVT.usage >>> 1535987 ± 4% +156.3% 3936830 ± 4% cpuidle.POLL.time >>> 88.14 ± 0% -24.9% 66.17 ± 3% turbostat.%Busy >>> 2659 ± 0% -44.7% 1471 ± 3% turbostat.Avg_MHz >>> 3016 ± 0% -26.3% 2224 ± 0% turbostat.Bzy_MHz >>> 5.20 ± 5% +127.0% 11.80 ± 2% turbostat.CPU%c1 >>> >>> >>> >>> perf-stat.page-faults >>> >>> 1e+06 ++-----------------------------------------------------------------+ >>> 900000 O+O O O O O O O O | >>> | O O O O O O O O O O | >>> 800000 ++ .*.*. .*. .*. .*.. | >>> 700000 ++*.*.*.*.*..*.* *.*.*.*.*.*..*.* * *.*.*.*.* *.*.*.* * >>> | : : | >>> 600000 ++: : | >>> 500000 ++ : :| >>> 400000 ++ : :| >>> |: : :| >>> 300000 ++ : :| >>> 200000 ++ : :| >>> | : | >>> 100000 ++ : | >>> 0 *+------------------------------------O-O------------------------*-+ >>> >>> >>> perf-stat.minor-faults >>> >>> 1e+06 ++-----------------------------------------------------------------+ >>> 900000 O+O O O O O O O O | >>> | O O O O O O O O O O | >>> 800000 ++ .*.*. .*. .*. .*.. | >>> 700000 ++*.*.*.*.*..*.* *.*.*.*.*.*..*.* * *.*.*.*.* *.*.*.* * >>> | : : | >>> 600000 ++: : | >>> 500000 ++ : :| >>> 400000 ++ : :| >>> |: : :| >>> 300000 ++ : :| >>> 200000 ++ : :| >>> | : | >>> 100000 ++ : | >>> 0 *+------------------------------------O-O------------------------*-+ >>> >>> >>> aim7.jobs-per-min >>> >>> 140000 ++-----------------------------------------------------------------+ >>> | *. .*.. .*.*.*.*.*.*.*..*.*.*.*.*.*. | >>> 120000 ++: *.*.* *.*.* *.*.*.*.*..*.*.*. * >>> | : * | >>> 100000 ++: : | >>> |: : | >>> 80000 O+O O O O O O O O O O O O O O O O O O : :| >>> |: : :| >>> 60000 ++ : :| >>> |: : :| >>> 40000 ++ : :| >>> | : :| >>> 20000 ++ : | >>> | : | >>> 0 *+------------------------------------O-O------------------------*-+ >>> >>> >>> >>> [*] bisect-good sample >>> [O] bisect-bad sample >>> >>> >>> Disclaimer: >>> Results have been estimated based on internal Intel analysis and are provided >>> for informational purposes only. Any difference in system hardware or software >>> design or configuration may affect actual performance. >>> >>> >>> Thanks, >>> Xiaolong >>> > > . >