| 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: Mon, 8 Aug 2016 16:29:25 +0800
From: kernel test robot <xiaolong.ye@...el.com>
To: Mel Gorman <mgorman@...hsingularity.net>
Cc: Linus Torvalds <torvalds@...ux-foundation.org>,
Vlastimil Babka <vbabka@...e.cz>,
Hillf Danton <hillf.zj@...baba-inc.com>,
Johannes Weiner <hannes@...xchg.org>,
Joonsoo Kim <iamjoonsoo.kim@....com>,
Michal Hocko <mhocko@...nel.org>,
Minchan Kim <minchan@...nel.org>,
Rik van Riel <riel@...riel.com>,
Andrew Morton <akpm@...ux-foundation.org>,
LKML <linux-kernel@...r.kernel.org>, lkp@...org
Subject: [lkp] [mm, page_alloc] e6cbd7f2ef: pixz.throughput -5.1% regression
FYI, we noticed a -5.1% regression of pixz.throughput due to commit:
commit e6cbd7f2efb433d717af72aa8510a9db6f7a7e05 ("mm, page_alloc: remove fair zone allocation policy")
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git master
in testcase: pixz
on test machine: 48 threads Ivytown Ivy Bridge-EP with 64G memory
with following parameters:
nr_threads: 100%
cpufreq_governor: performance
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.
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
=========================================================================================
compiler/cpufreq_governor/kconfig/nr_threads/rootfs/tbox_group/testcase:
gcc-6/performance/x86_64-rhel/100%/debian-x86_64-2015-02-07.cgz/ivb43/pixz
commit:
e5146b12e2 ("mm, vmscan: add classzone information to tracepoints")
e6cbd7f2ef ("mm, page_alloc: remove fair zone allocation policy")
e5146b12e2d02af0 e6cbd7f2efb433d717af72aa85
---------------- --------------------------
%stddev %change %stddev
\ | \
70271863 ± 0% -5.1% 66679781 ± 0% pixz.throughput
5375256 ± 2% -3.5% 5185887 ± 0% pixz.time.involuntary_context_switches
7473 ± 42% -72.2% 2075 ± 16% latency_stats.max.call_rwsem_down_read_failed.__do_page_fault.do_page_fault.page_fault.copy_page_to_iter.pipe_read.__vfs_read.vfs_read.SyS_read.entry_SYSCALL_64_fastpath
1.63 ± 9% -24.2% 1.23 ± 12% perf-profile.cycles-pp.native_irq_return_iret
1571 ± 14% +492.3% 9310 ± 93% softirqs.NET_RX
20.30 ± 0% -7.8% 18.72 ± 0% turbostat.RAMWatt
4025 ± 0% -100.0% 0.00 ± -1% numa-vmstat.node0.nr_alloc_batch
4112 ± 1% -100.0% 0.00 ± -1% numa-vmstat.node1.nr_alloc_batch
480396 ± 4% +24.2% 596533 ± 11% cpuidle.C1-IVT.usage
367345 ± 6% +18.3% 434581 ± 9% cpuidle.C1E-IVT.time
2597 ± 3% +12.3% 2916 ± 7% cpuidle.C1E-IVT.usage
5020 ± 4% +20.6% 6052 ± 8% cpuidle.POLL.usage
536237 ± 0% +977.7% 5779033 ± 0% meminfo.Active
32820 ± 1% +17508.1% 5779015 ± 0% meminfo.Active(anon)
503416 ± 0% -100.0% 18.00 ± 0% meminfo.Active(file)
5819833 ± 0% -91.3% 503416 ± 0% meminfo.Inactive(file)
8190 ± 1% -100.0% 0.00 ± -1% proc-vmstat.nr_alloc_batch
3541203 ± 2% -11.3% 3140364 ± 3% proc-vmstat.numa_hit
3541183 ± 2% -11.3% 3140347 ± 3% proc-vmstat.numa_local
887680 ± 5% -100.0% 0.00 ± 0% proc-vmstat.pgalloc_dma32
11.88 ± 14% +23.5% 14.67 ± 8% sched_debug.cfs_rq:/.runnable_load_avg.min
209709 ± 44% -74.2% 54100 ± 58% sched_debug.cpu.avg_idle.min
174488 ± 11% +24.7% 217532 ± 2% sched_debug.cpu.avg_idle.stddev
6.14 ± 6% +25.2% 7.69 ± 9% sched_debug.cpu.clock.stddev
6.14 ± 6% +25.2% 7.69 ± 9% sched_debug.cpu.clock_task.stddev
6020 ± 4% +27.4% 7669 ± 10% sched_debug.cpu.sched_goidle.avg
25808 ± 12% +85.0% 47758 ± 14% sched_debug.cpu.sched_goidle.max
5604 ± 9% +67.2% 9368 ± 8% sched_debug.cpu.sched_goidle.stddev
4.679e+12 ± 0% -5.4% 4.426e+12 ± 0% perf-stat.branch-instructions
6.646e+10 ± 0% -5.4% 6.286e+10 ± 0% perf-stat.branch-misses
8.562e+10 ± 0% -5.4% 8.096e+10 ± 0% perf-stat.cache-misses
1.278e+11 ± 0% -5.2% 1.212e+11 ± 0% perf-stat.cache-references
5.929e+12 ± 0% -5.6% 5.595e+12 ± 0% perf-stat.dTLB-loads
1.053e+12 ± 0% -5.0% 1.001e+12 ± 0% perf-stat.dTLB-stores
84.74 ± 0% -2.2% 82.84 ± 1% perf-stat.iTLB-load-miss-rate
25042894 ± 10% +13.1% 28320339 ± 3% perf-stat.iTLB-loads
2.238e+13 ± 0% -5.4% 2.117e+13 ± 0% perf-stat.instructions
0.49 ± 0% -4.8% 0.47 ± 0% perf-stat.ipc
7.467e+10 ± 0% -5.7% 7.043e+10 ± 0% perf-stat.node-loads
1.176e+10 ± 0% -5.2% 1.114e+10 ± 0% perf-stat.node-stores
pixz.throughput
7.1e+07 ++-----------*---------------------------------------------------+
7.05e+07 ++*.*. *.*. + + .* .*.*. .*.* .*.*. .*.*.* .*.*.* *.** |
* * * *.* *.* * * * * + + + .*
7e+07 ++ * *.* |
6.95e+07 ++ |
| |
6.9e+07 ++ |
6.85e+07 ++ |
6.8e+07 ++ |
| |
6.75e+07 ++ |
6.7e+07 ++ O O O O |
| O O O O O O OO O O O O O O O O |
6.65e+07 O+ O O O O O O |
6.6e+07 ++---------------------------------------------------------------+
[*] bisect-good sample
[O] bisect-bad sample
Thanks,
Xiaolong
View attachment "config-4.7.0-05952-ge6cbd7f" of type "text/plain" (152075 bytes)
View attachment "job.yaml" of type "text/plain" (3531 bytes)
View attachment "reproduce" of type "text/plain" (103 bytes)
Powered by blists - more mailing lists