linux-kernel - [perf/x86/intel] fc17db8aa4: sched: RT throttling activated

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

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [day] [month] [year] [list]

Message-ID: <20210107080713.GC22733@xsang-OptiPlex-9020>
Date:   Thu, 7 Jan 2021 16:07:13 +0800
From:   kernel test robot <oliver.sang@...el.com>
To:     Stephane Eranian <eranian@...gle.com>
Cc:     Peter Zijlstra <peterz@...radead.org>,
        Namhyung Kim <namhyung@...nel.org>,
        LKML <linux-kernel@...r.kernel.org>, lkp@...ts.01.org,
        lkp@...el.com
Subject: [perf/x86/intel]  fc17db8aa4: sched: RT throttling activated


Greeting,

FYI, we noticed the following commit (built with gcc-9):

commit: fc17db8aa4c53cbd2d5469bb0521ea0f0a6dbb27 ("perf/x86/intel: Check PEBS status correctly")
https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git master


in testcase: fileio
version: 
with following parameters:

	period: 600s
	nr_threads: 100%
	disk: 1HDD
	fs: xfs
	size: 64G
	filenum: 1024f
	rwmode: seqrd
	iomode: sync
	cpufreq_governor: performance
	ucode: 0x11

test-description: fileio is a subtest of SysBench benchmark suite to measure file IO performance.
test-url: https://github.com/akopytov/sysbench


on test machine: 288 threads Intel(R) Xeon Phi(TM) CPU 7295 @ 1.50GHz with 80G memory

caused below changes (please refer to attached dmesg/kmsg for entire log/backtrace):




If you fix the issue, kindly add following tag
Reported-by: kernel test robot <oliver.sang@...el.com>


[  758.164955] Threads started!
[  758.164969]
[  758.177013]
[  758.177024]
[  804.470082] sched: RT throttling activated
[  806.473158] clocksource: timekeeping watchdog on CPU287: Marking clocksource 'tsc' as unstable because the skew is too large:
[  806.488569] clocksource:                       'hpet' wd_now: ada63507 wd_last: f626c00c mask: ffffffff
[  806.502032] clocksource:                       'tsc' cs_now: 10041e998914 cs_last: f50ae553aad mask: ffffffffffffffff
[  806.516915] tsc: Marking TSC unstable due to clocksource watchdog
[  814.452051] TSC found unstable after boot, most likely due to broken BIOS. Use 'tsc=unstable'.
[  814.466606] sched_clock: Marking unstable (813719139048, 732811744)<-(814741144871, -289175402)
[  815.475981] clocksource: Switched to clocksource hpet


pleased be noted above "sched: RT throttling activated" is persistent in all our 10 tests.
but not observed in runs for parent, such like below:
[  747.861410] Threads started!    <----- no problem after this, and just see results below
[  747.861422]
[  747.872247]
[  747.872358]
[ 1347.670691]
[ 1347.670702]
[ 1347.679946] Throughput:
[ 1347.679956]
[ 1347.698157]          read:  IOPS=374466.02 5851.03 MiB/s (6135.25 MB/s)
[ 1347.698171]
[ 1347.719242]          write: IOPS=0.00 0.00 MiB/s (0.00 MB/s)
[ 1347.719256]
[ 1347.736059]          fsync: IOPS=0.00



To reproduce:

        git clone https://github.com/intel/lkp-tests.git
        cd lkp-tests
        bin/lkp install job.yaml  # job file is attached in this email
        bin/lkp run     job.yaml



Thanks,
Oliver Sang


View attachment "config-5.10.0-rc2-00381-gfc17db8aa4c5" of type "text/plain" (171049 bytes)

View attachment "job-script" of type "text/plain" (8253 bytes)

Download attachment "dmesg.xz" of type "application/x-xz" (36908 bytes)

View attachment "job.yaml" of type "text/plain" (5575 bytes)

View attachment "reproduce" of type "text/plain" (693 bytes)