| 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
| ||
|
Message-ID: <20170215013408.GO18272@yexl-desktop>
Date: Wed, 15 Feb 2017 09:34:08 +0800
From: kernel test robot <xiaolong.ye@...el.com>
To: Johannes Weiner <hannes@...xchg.org>
Cc: Stephen Rothwell <sfr@...b.auug.org.au>,
Minchan Kim <minchan@...nel.org>,
Michal Hocko <mhocko@...e.com>, Mel Gorman <mgorman@...e.de>,
Hillf Danton <hillf.zj@...baba-inc.com>,
Rik van Riel <riel@...hat.com>,
Andrew Morton <akpm@...ux-foundation.org>,
LKML <linux-kernel@...r.kernel.org>, lkp@...org
Subject: [lkp-robot] [mm] e932fd3f17: 24% improvement of fio.write_bw_MBps
Greeting,
FYI, we noticed a 24% improvement of fio.write_bw_MBps due to commit:
commit: e932fd3f1772e1fbc0b90dcfb4fbc689729a48f8 ("mm: vmscan: kick flushers when we encounter dirty pages on the LRU")
https://git.kernel.org/cgit/linux/kernel/git/next/linux-next.git master
in testcase: fio-basic
on test machine: 16 threads Intel(R) Xeon(R) CPU D-1541 @ 2.10GHz with 8G memory
with following parameters:
runtime: 300s
disk: 1SSD
fs: btrfs
nr_task: 8
rw: randwrite
bs: 4k
ioengine: sync
test_size: 400g
cpufreq_governor: performance
test-description: Fio is a tool that will spawn a number of threads or processes doing a particular type of I/O action as specified by the user.
test-url: https://github.com/axboe/fio
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: fio-basic/300s-1SSD-btrfs-8-randwrite-4k-sync-400g-performance/lkp-bdw-de1
2b837027cdc60ac3 e932fd3f1772e1fbc0b90dcfb4
---------------- --------------------------
%stddev change %stddev
\ | \
4.07 ± 17% 70% 6.92 ± 23% fio.latency_10us%
0.29 ± 12% 28% 0.37 ± 10% fio.latency_100us%
66.58 ± 5% 24% 82.64 fio.write_bw_MBps
17043 ± 5% 24% 21155 fio.write_iops
43.93 ± 3% 11% 48.92 ± 3% fio.latency_250us%
10465 ± 5% -14% 8975 ± 10% fio.write_clat_stddev
469 ± 5% -20% 377 fio.write_clat_mean_us
570 ± 15% -26% 423 ± 3% fio.write_clat_90%_us
883 ± 23% -41% 519 ± 9% fio.write_clat_95%_us
6.24 ± 20% -44% 3.48 ± 11% fio.latency_750us%
2414 ± 14% -50% 1199 ± 27% fio.write_clat_99%_us
0.06 ± 24% -65% 0.02 ± 34% fio.latency_10ms%
2.64 ± 29% -67% 0.86 ± 38% fio.latency_2ms%
2.03 ± 33% -69% 0.62 ± 44% fio.latency_1000us%
40927209 ± 5% 11% 45440361 ± 35% fio.time.file_system_outputs
4913388 ± 4% 6% 5228420 ± 35% fio.time.voluntary_context_switches
56.25 ± 3% 5% 58.87 ± 35% fio.time.system_time
22 ± 4% 4% 23 ± 35% fio.time.percent_of_cpu_this_job_got
fail:runs %reproduction fail:runs
| | |
:10 11% 1:9 last_state.OOM
:10 11% 1:9 last_state.is_incomplete_run
:10 11% 1:9 dmesg.invoked_oom-killer:gfp_mask=0x
:10 11% 1:9 dmesg.Mem-Info
:10 11% 1:9 dmesg.Out_of_memory:Kill_process
562 ± 7% 24% 696 ± 3% turbostat.Avg_MHz
22.79 ± 7% 23% 28.09 ± 3% turbostat.%Busy
26.58 5% 28.03 turbostat.PkgWatt
10.78 4% 11.16 turbostat.RAMWatt
1659 ±300% 2e+05 187515 ± 9% latency_stats.avg.balance_dirty_pages.balance_dirty_pages_ratelimited.__btrfs_btree_balance_dirty.[btrfs].btrfs_btree_balance_dirty.[btrfs].__btrfs_buffered_write.[btrfs].btrfs_file_write_iter.[btrfs].__vfs_write.vfs_write.SyS_write.entry_SYSCALL_64_fastpath
1659 ±300% 2e+05 203157 latency_stats.max.balance_dirty_pages.balance_dirty_pages_ratelimited.__btrfs_btree_balance_dirty.[btrfs].btrfs_btree_balance_dirty.[btrfs].__btrfs_buffered_write.[btrfs].btrfs_file_write_iter.[btrfs].__vfs_write.vfs_write.SyS_write.entry_SYSCALL_64_fastpath
6217883 ±157% 1e+08 1.231e+08 ± 51% latency_stats.sum.balance_dirty_pages.balance_dirty_pages_ratelimited.__btrfs_buffered_write.[btrfs].btrfs_file_write_iter.[btrfs].__vfs_write.vfs_write.SyS_write.entry_SYSCALL_64_fastpath
1659 ±300% 3e+06 3100846 ± 57% latency_stats.sum.balance_dirty_pages.balance_dirty_pages_ratelimited.__btrfs_btree_balance_dirty.[btrfs].btrfs_btree_balance_dirty.[btrfs].__btrfs_buffered_write.[btrfs].btrfs_file_write_iter.[btrfs].__vfs_write.vfs_write.SyS_write.entry_SYSCALL_64_fastpath
182633 ± 8% 12% 204509 ± 5% vmstat.io.bo
314624 ± 6% 23% 387499 ± 3% vmstat.system.cs
36487 12% 40702 vmstat.system.in
16960 ± 5% 25% 21138 iostat.sda.w/s
46.04 ± 11% 16% 53.46 ± 7% iostat.sda.avgqu-sz
204068 ± 7% 11% 225809 ± 4% iostat.sda.wkB/s
8928 ± 8% 27% 11302 ± 3% perf-stat.instructions-per-iTLB-miss
2.7e+12 ± 7% 24% 3.344e+12 ± 4% perf-stat.cpu-cycles
95320324 ± 6% 23% 1.171e+08 ± 3% perf-stat.context-switches
3.384e+11 ± 6% 22% 4.135e+11 ± 3% perf-stat.branch-instructions
1.724e+12 ± 6% 21% 2.093e+12 ± 3% perf-stat.instructions
4.565e+11 ± 7% 21% 5.54e+11 perf-stat.dTLB-loads
1.656e+09 ± 5% 20% 1.993e+09 ± 4% perf-stat.iTLB-loads
2.698e+11 ± 5% 20% 3.225e+11 perf-stat.dTLB-stores
3.769e+10 ± 5% 16% 4.39e+10 ± 3% perf-stat.cache-misses
3.769e+10 ± 5% 16% 4.39e+10 ± 3% perf-stat.cache-references
3.547e+09 ± 4% 15% 4.086e+09 perf-stat.branch-misses
502899 ± 3% 13% 565994 ± 3% perf-stat.cpu-migrations
4.906e+08 ± 3% 6% 5.212e+08 ± 3% perf-stat.dTLB-load-misses
1.05 -6% 0.99 perf-stat.branch-miss-rate%
0.02 ± 3% -10% 0.02 perf-stat.dTLB-store-miss-rate%
0.11 ± 8% -13% 0.09 ± 3% perf-stat.dTLB-load-miss-rate%
10.49 ± 6% -19% 8.51 ± 3% perf-stat.iTLB-load-miss-rate%
[*] 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
View attachment "config-4.10.0-rc7-00263-ge932fd3" of type "text/plain" (155610 bytes)
View attachment "job-script" of type "text/plain" (7031 bytes)
View attachment "job.yaml" of type "text/plain" (4690 bytes)
View attachment "reproduce" of type "text/plain" (431 bytes)
Powered by blists - more mailing lists