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 for Android: free password hash cracker in your pocket
[<prev] [next>] [day] [month] [year] [list]
Date:   Fri, 14 Oct 2022 13:39:59 +0800
From:   kernel test robot <oliver.sang@...el.com>
To:     "Darrick J. Wong" <djwong@...nel.org>
CC:     <lkp@...ts.01.org>, <lkp@...el.com>,
        Dave Chinner <dchinner@...hat.com>,
        <linux-kernel@...r.kernel.org>, <linux-xfs@...r.kernel.org>,
        <ying.huang@...el.com>, <feng.tang@...el.com>,
        <zhengjun.xing@...ux.intel.com>, <fengwei.yin@...el.com>
Subject: [xfs]  2ed5b09b3e:  aim7.jobs-per-min 13.9% improvement


Greeting,

FYI, we noticed a 13.9% improvement of aim7.jobs-per-min due to commit:


commit: 2ed5b09b3e8fc274ae8fecd6ab7c5106a364bed1 ("xfs: make inode attribute forks a permanent part of struct xfs_inode")
https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git master

in testcase: aim7
on test machine: 128 threads 2 sockets Intel(R) Xeon(R) Gold 6338 CPU @ 2.00GHz (Ice Lake) with 256G memory
with following parameters:

	disk: 1BRD_48G
	fs: xfs
	test: disk_rd
	load: 9000
	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 https://github.com/intel/lkp-tests.git
        cd lkp-tests
        sudo bin/lkp install job.yaml           # job file is attached in this email
        bin/lkp split-job --compatible job.yaml # generate the yaml file for lkp run
        sudo bin/lkp run generated-yaml-file

        # if come across any failure that blocks the test,
        # please remove ~/.lkp and /lkp dir to run from a clean state.

=========================================================================================
compiler/cpufreq_governor/disk/fs/kconfig/load/rootfs/tbox_group/test/testcase:
  gcc-11/performance/1BRD_48G/xfs/x86_64-rhel-8.3/9000/debian-11.1-x86_64-20220510.cgz/lkp-icl-2sp2/disk_rd/aim7

commit: 
  732436ef91 ("xfs: convert XFS_IFORK_PTR to a static inline helper")
  2ed5b09b3e ("xfs: make inode attribute forks a permanent part of struct xfs_inode")

732436ef916b4f33 2ed5b09b3e8fc274ae8fecd6ab7 
---------------- --------------------------- 
         %stddev     %change         %stddev
             \          |                \  
    518200           +13.9%     590052 ±  4%  aim7.jobs-per-min
    104.74           -12.0%      92.22 ±  4%  aim7.time.elapsed_time
    104.74           -12.0%      92.22 ±  4%  aim7.time.elapsed_time.max
   1268209 ±  3%     -21.5%     995207 ±  7%  aim7.time.involuntary_context_switches
     11845           -13.5%      10249 ±  5%  aim7.time.system_time
     96456 ±  2%      -6.4%      90310 ±  5%  aim7.time.voluntary_context_switches
    128756 ±  4%     -12.4%     112833 ±  5%  meminfo.AnonHugePages
     12.03 ±  3%      +2.4       14.42 ±  2%  mpstat.cpu.all.idle%
     15141 ±  2%      -8.0%      13934 ±  2%  vmstat.system.cs
     13.59 ±  2%     +18.4%      16.08 ±  2%  iostat.cpu.idle
     85.83            -3.0%      83.26        iostat.cpu.system
      0.94 ± 12%     -20.7%       0.74 ± 12%  sched_debug.cfs_rq:/.h_nr_running.stddev
      1714 ± 11%     -21.0%       1354 ± 12%  sched_debug.cfs_rq:/.runnable_avg.avg
      4309 ± 11%     -22.9%       3321 ± 13%  sched_debug.cfs_rq:/.runnable_avg.max
      0.93 ± 12%     -20.7%       0.74 ± 12%  sched_debug.cpu.nr_running.stddev
   1052162 ±  3%      -8.0%     967689 ±  2%  proc-vmstat.numa_hit
    935991 ±  3%      -9.0%     852146 ±  2%  proc-vmstat.numa_local
     83514 ± 22%     -38.3%      51505 ± 34%  proc-vmstat.numa_pte_updates
   1063781 ±  3%      -9.1%     967461 ±  2%  proc-vmstat.pgalloc_normal
    979883 ±  3%      -9.4%     887981 ±  2%  proc-vmstat.pgfault
    814030            -6.5%     760890 ±  2%  proc-vmstat.pgfree
     15209 ±  2%      -9.2%      13811 ±  2%  perf-stat.i.context-switches
     16.08           -14.6%      13.72 ±  5%  perf-stat.i.cpi
 2.882e+11            -4.3%  2.758e+11        perf-stat.i.cpu-cycles
      0.11 ±  5%     +16.1%       0.12 ±  9%  perf-stat.i.ipc
      2.24            -4.4%       2.15 ±  2%  perf-stat.i.metric.GHz
     17.40           -12.3%      15.26 ±  5%  perf-stat.overall.cpi
      0.00 ±  2%      -0.0        0.00 ±  4%  perf-stat.overall.dTLB-store-miss-rate%
      0.06           +14.3%       0.07 ±  4%  perf-stat.overall.ipc
     15213 ±  2%      -8.4%      13934 ±  2%  perf-stat.ps.context-switches
  2.92e+11            -2.5%  2.846e+11        perf-stat.ps.cpu-cycles
     26.57 ±  4%     -26.5        0.09 ±223%  perf-profile.calltrace.cycles-pp.touch_atime.filemap_read.xfs_file_buffered_read.xfs_file_read_iter.new_sync_read
     26.31 ±  4%     -26.2        0.09 ±223%  perf-profile.calltrace.cycles-pp.atime_needs_update.touch_atime.filemap_read.xfs_file_buffered_read.xfs_file_read_iter
     44.22           -24.2       20.01 ±  4%  perf-profile.calltrace.cycles-pp.filemap_read.xfs_file_buffered_read.xfs_file_read_iter.new_sync_read.vfs_read
      6.70 ±  5%      -2.8        3.85 ± 11%  perf-profile.calltrace.cycles-pp.filemap_get_read_batch.filemap_get_pages.filemap_read.xfs_file_buffered_read.xfs_file_read_iter
     98.12            -0.5       97.64        perf-profile.calltrace.cycles-pp.xfs_file_buffered_read.xfs_file_read_iter.new_sync_read.vfs_read.ksys_read
     98.20            -0.5       97.74        perf-profile.calltrace.cycles-pp.xfs_file_read_iter.new_sync_read.vfs_read.ksys_read.do_syscall_64
     98.34            -0.4       97.92        perf-profile.calltrace.cycles-pp.new_sync_read.vfs_read.ksys_read.do_syscall_64.entry_SYSCALL_64_after_hwframe
     98.68            -0.4       98.32        perf-profile.calltrace.cycles-pp.vfs_read.ksys_read.do_syscall_64.entry_SYSCALL_64_after_hwframe.read
     98.83            -0.3       98.49        perf-profile.calltrace.cycles-pp.ksys_read.do_syscall_64.entry_SYSCALL_64_after_hwframe.read
     98.99            -0.3       98.68        perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.read
     99.09            -0.3       98.81        perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.read
     99.40            -0.2       99.19        perf-profile.calltrace.cycles-pp.read
     34.96            +9.8       44.78        perf-profile.calltrace.cycles-pp.down_read.xfs_ilock.xfs_file_buffered_read.xfs_file_read_iter.new_sync_read
     35.32           +10.0       45.32        perf-profile.calltrace.cycles-pp.xfs_ilock.xfs_file_buffered_read.xfs_file_read_iter.new_sync_read.vfs_read
     18.23           +13.3       31.54 ±  4%  perf-profile.calltrace.cycles-pp.up_read.xfs_iunlock.xfs_file_buffered_read.xfs_file_read_iter.new_sync_read
     18.45           +13.6       32.07 ±  4%  perf-profile.calltrace.cycles-pp.xfs_iunlock.xfs_file_buffered_read.xfs_file_read_iter.new_sync_read.vfs_read
     26.54 ±  4%     -26.2        0.32 ± 30%  perf-profile.children.cycles-pp.atime_needs_update
     26.58 ±  4%     -26.2        0.36 ± 26%  perf-profile.children.cycles-pp.touch_atime
     44.28           -24.2       20.09 ±  4%  perf-profile.children.cycles-pp.filemap_read
      6.74 ±  5%      -2.9        3.88 ± 10%  perf-profile.children.cycles-pp.filemap_get_read_batch
     98.15            -0.5       97.66        perf-profile.children.cycles-pp.xfs_file_buffered_read
     98.21            -0.5       97.76        perf-profile.children.cycles-pp.xfs_file_read_iter
     98.36            -0.4       97.94        perf-profile.children.cycles-pp.new_sync_read
     98.70            -0.4       98.34        perf-profile.children.cycles-pp.vfs_read
     98.84            -0.3       98.51        perf-profile.children.cycles-pp.ksys_read
     99.56            -0.2       99.37        perf-profile.children.cycles-pp.read
     99.37            -0.1       99.24        perf-profile.children.cycles-pp.do_syscall_64
     99.43            -0.1       99.31        perf-profile.children.cycles-pp.entry_SYSCALL_64_after_hwframe
      0.05            +0.0        0.06 ±  7%  perf-profile.children.cycles-pp.do_filp_open
      0.05            +0.0        0.06 ±  7%  perf-profile.children.cycles-pp.path_openat
      0.06 ±  6%      +0.0        0.07 ±  5%  perf-profile.children.cycles-pp.exit_to_user_mode_loop
      0.05            +0.0        0.06 ±  7%  perf-profile.children.cycles-pp.times
      0.08 ±  5%      +0.0        0.10 ±  9%  perf-profile.children.cycles-pp.__might_sleep
      0.06 ±  9%      +0.0        0.07 ±  5%  perf-profile.children.cycles-pp.__close
      0.06            +0.0        0.08 ±  6%  perf-profile.children.cycles-pp.open64
      0.10 ±  5%      +0.0        0.11 ±  6%  perf-profile.children.cycles-pp.exit_to_user_mode_prepare
      0.06 ±  6%      +0.0        0.08        perf-profile.children.cycles-pp.__x64_sys_openat
      0.06 ±  6%      +0.0        0.08        perf-profile.children.cycles-pp.do_sys_openat2
      0.16 ±  2%      +0.0        0.19 ±  6%  perf-profile.children.cycles-pp.entry_SYSRETQ_unsafe_stack
      0.05 ±  8%      +0.0        0.08 ± 20%  perf-profile.children.cycles-pp.generic_file_read_iter
      0.14 ±  4%      +0.0        0.17 ±  5%  perf-profile.children.cycles-pp.syscall_exit_to_user_mode
      0.08 ±  8%      +0.0        0.11 ±  6%  perf-profile.children.cycles-pp.write
      0.11 ±  5%      +0.0        0.15 ± 16%  perf-profile.children.cycles-pp.__fdget_pos
      0.00            +0.1        0.05        perf-profile.children.cycles-pp._raw_spin_lock
      0.11 ±  7%      +0.1        0.16 ± 10%  perf-profile.children.cycles-pp.xas_load
      0.00            +0.1        0.06 ±  9%  perf-profile.children.cycles-pp.thread_group_cputime_adjusted
      0.00            +0.1        0.06 ±  8%  perf-profile.children.cycles-pp.do_sys_times
      0.00            +0.1        0.06 ±  8%  perf-profile.children.cycles-pp.task_work_run
      0.00            +0.1        0.06 ±  9%  perf-profile.children.cycles-pp.__x64_sys_times
      0.00            +0.1        0.07 ± 16%  perf-profile.children.cycles-pp.xas_start
      0.14 ± 21%      +0.1        0.23 ± 42%  perf-profile.children.cycles-pp.ksys_write
     35.29           +10.0       45.28        perf-profile.children.cycles-pp.down_read
     35.33           +10.0       45.33        perf-profile.children.cycles-pp.xfs_ilock
     18.40 ±  2%     +13.5       31.89 ±  4%  perf-profile.children.cycles-pp.up_read
     18.46           +13.6       32.10 ±  4%  perf-profile.children.cycles-pp.xfs_iunlock
     26.37 ±  4%     -26.1        0.22 ± 42%  perf-profile.self.cycles-pp.atime_needs_update
      6.60 ±  5%      -2.9        3.70 ± 11%  perf-profile.self.cycles-pp.filemap_get_read_batch
      0.06 ±  7%      +0.0        0.08 ±  7%  perf-profile.self.cycles-pp.__cond_resched
      0.07 ±  5%      +0.0        0.09 ±  6%  perf-profile.self.cycles-pp.xas_load
      0.07            +0.0        0.10 ± 21%  perf-profile.self.cycles-pp.xfs_file_read_iter
      0.07 ±  7%      +0.0        0.10 ± 10%  perf-profile.self.cycles-pp.do_syscall_64
      0.04 ± 44%      +0.0        0.08 ±  9%  perf-profile.self.cycles-pp.xfs_file_buffered_read
      0.12 ±  7%      +0.0        0.16 ±  4%  perf-profile.self.cycles-pp.new_sync_read
      0.02 ±141%      +0.1        0.08 ± 20%  perf-profile.self.cycles-pp.generic_file_read_iter
      0.00            +0.1        0.07 ± 15%  perf-profile.self.cycles-pp.xas_start
      0.06 ±  8%      +0.2        0.21 ± 12%  perf-profile.self.cycles-pp.xfs_iunlock
      3.85 ±  4%      +3.2        7.09 ±  6%  perf-profile.self.cycles-pp.filemap_read
     35.06           +10.0       45.02        perf-profile.self.cycles-pp.down_read
     18.33 ±  2%     +13.4       31.76 ±  4%  perf-profile.self.cycles-pp.up_read




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.


-- 
0-DAY CI Kernel Test Service
https://01.org/lkp



View attachment "config-5.19.0-rc5-00034-g2ed5b09b3e8f" of type "text/plain" (163182 bytes)

View attachment "job-script" of type "text/plain" (8047 bytes)

View attachment "job.yaml" of type "text/plain" (5455 bytes)

View attachment "reproduce" of type "text/plain" (731 bytes)

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ