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]
Date:   Mon, 22 Aug 2022 10:10:25 +0800
From:   kernel test robot <yujie.liu@...el.com>
To:     Al Viro <viro@...iv.linux.org.uk>
CC:     <lkp@...ts.01.org>, kernel test robot <lkp@...el.com>,
        "Christian Brauner (Microsoft)" <brauner@...nel.org>,
        LKML <linux-kernel@...r.kernel.org>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
        <linux-nvme@...ts.infradead.org>, <linux-aio@...ck.org>,
        <linux-fsdevel@...r.kernel.org>, Huang Ying <ying.huang@...el.com>,
        <feng.tang@...el.com>, <zhengjun.xing@...ux.intel.com>,
        <fengwei.yin@...el.com>
Subject: 164f4064ca: unixbench.score 1.4% improvement

Greeting,

FYI, we noticed a 1.4% improvement of unixbench.score due to commit:


commit: 164f4064ca81eefcea29f7f5dcf394f92be1d0c0 ("keep iocb_flags() result cached in struct file")
https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git master

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

	runtime: 300s
	nr_task: 1
	test: fstime
	cpufreq_governor: performance
	ucode: 0xd000363

test-description: UnixBench is the original BYTE UNIX benchmark suite aims to test performance of Unix-like system.
test-url: https://github.com/kdlucas/byte-unixbench

In addition to that, the commit also has impact on the following tests:

+------------------+---------------------------------------------------------------------------------+
| testcase: change | unixbench: unixbench.score 2.1% improvement                                     |
| test machine     | 128 threads 2 sockets Intel(R) Xeon(R) Gold 6338 CPU @ 2.00GHz with 256G memory |
| test parameters  | cpufreq_governor=performance                                                    |
|                  | nr_task=1                                                                       |
|                  | runtime=300s                                                                    |
|                  | test=fsbuffer                                                                   |
|                  | ucode=0xd000363                                                                 |
+------------------+---------------------------------------------------------------------------------+

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/kconfig/nr_task/rootfs/runtime/tbox_group/test/testcase/ucode:
   gcc-11/performance/x86_64-rhel-8.3/1/debian-11.1-x86_64-20220510.cgz/300s/lkp-icl-2sp2/fstime/unixbench/0xd000363

commit:
   91b94c5d6a ("iocb: delay evaluation of IS_SYNC(...) until we want to check IOCB_DSYNC")
   164f4064ca ("keep iocb_flags() result cached in struct file")

91b94c5d6ae55d11 164f4064ca81eefcea29f7f5dcf
---------------- ---------------------------
          %stddev     %change         %stddev
              \          |                \
       3500            +1.4%       3548        unixbench.score
  6.657e+08            +1.4%  6.747e+08        unixbench.workload
       0.02 ±118%      +0.0        0.07 ± 17%  perf-profile.self.cycles-pp.rcu_eqs_enter
   39089557            +1.8%   39808281        perf-stat.i.cache-references
   39018465            +1.8%   39737152        perf-stat.ps.cache-references


***************************************************************************************************
lkp-icl-2sp2: 128 threads 2 sockets Intel(R) Xeon(R) Gold 6338 CPU @ 2.00GHz with 256G memory
=========================================================================================
compiler/cpufreq_governor/kconfig/nr_task/rootfs/runtime/tbox_group/test/testcase/ucode:
   gcc-11/performance/x86_64-rhel-8.3/1/debian-11.1-x86_64-20220510.cgz/300s/lkp-icl-2sp2/fsbuffer/unixbench/0xd000363

commit:
   91b94c5d6a ("iocb: delay evaluation of IS_SYNC(...) until we want to check IOCB_DSYNC")
   164f4064ca ("keep iocb_flags() result cached in struct file")

91b94c5d6ae55d11 164f4064ca81eefcea29f7f5dcf
---------------- ---------------------------
          %stddev     %change         %stddev
              \          |                \
       2322            +2.1%       2370        unixbench.score
  1.845e+08            +2.1%  1.883e+08        unixbench.workload
       0.03 ±  3%      +0.0        0.04 ± 12%  mpstat.cpu.all.soft%
      45.67 ± 56%   +7248.9%       3356 ± 98%  proc-vmstat.numa_hint_faults_local
  1.022e+09            +1.3%  1.036e+09        perf-stat.i.branch-instructions
  1.455e+09            +1.6%  1.478e+09        perf-stat.i.dTLB-loads
  7.985e+08            +1.8%  8.125e+08        perf-stat.i.dTLB-stores
      25.59            +1.6%      25.99        perf-stat.i.metric.M/sec
      12988            -1.4%      12803        perf-stat.overall.path-length
   1.02e+09            +1.3%  1.034e+09        perf-stat.ps.branch-instructions
  1.452e+09            +1.6%  1.475e+09        perf-stat.ps.dTLB-loads
  7.971e+08            +1.7%  8.109e+08        perf-stat.ps.dTLB-stores
       0.25 ± 12%      -0.1        0.15 ± 15%  perf-profile.children.cycles-pp.error_entry
       0.52 ±  6%      -0.1        0.43 ±  8%  perf-profile.children.cycles-pp.update_sd_lb_stats
       0.54 ±  8%      -0.1        0.46 ±  9%  perf-profile.children.cycles-pp.find_busiest_group
       0.09 ± 18%      -0.0        0.05 ± 47%  perf-profile.children.cycles-pp.trigger_load_balance
       0.09 ± 21%      -0.0        0.06 ± 46%  perf-profile.children.cycles-pp.update_group_capacity
       0.11 ±  9%      +0.0        0.15 ± 17%  perf-profile.children.cycles-pp.cpuidle_governor_latency_req
       0.02 ±142%      +0.0        0.07 ± 26%  perf-profile.children.cycles-pp.cpuidle_reflect
       0.05 ± 50%      +0.1        0.10 ± 36%  perf-profile.children.cycles-pp.tick_program_event
       0.02 ±141%      +0.1        0.09 ± 11%  perf-profile.children.cycles-pp.rcu_eqs_enter
       0.72 ±  7%      +0.1        0.80 ±  5%  perf-profile.children.cycles-pp.native_sched_clock
       0.89 ±  6%      +0.1        1.00 ±  4%  perf-profile.children.cycles-pp.sched_clock_cpu
       0.59 ±  6%      +0.1        0.70 ± 14%  perf-profile.children.cycles-pp._raw_spin_lock_irqsave
       0.23 ± 10%      -0.1        0.13 ± 14%  perf-profile.self.cycles-pp.error_entry
       0.09 ± 21%      -0.0        0.05 ± 46%  perf-profile.self.cycles-pp.update_group_capacity
       0.04 ± 45%      +0.0        0.06 ±  7%  perf-profile.self.cycles-pp.tick_nohz_tick_stopped
       0.06 ± 14%      +0.0        0.09 ± 11%  perf-profile.self.cycles-pp.cpuidle_enter
       0.06 ± 14%      +0.0        0.10 ± 18%  perf-profile.self.cycles-pp.rcu_idle_exit
       0.52 ±  6%      +0.0        0.56 ±  3%  perf-profile.self.cycles-pp._raw_spin_lock_irqsave
       0.28 ±  5%      +0.0        0.32 ±  7%  perf-profile.self.cycles-pp.irq_enter_rcu
       0.09 ± 27%      +0.1        0.17 ± 15%  perf-profile.self.cycles-pp.acpi_idle_enter
       0.01 ±223%      +0.1        0.09 ± 36%  perf-profile.self.cycles-pp.tick_program_event
       0.00            +0.1        0.08 ± 14%  perf-profile.self.cycles-pp.rcu_eqs_enter
       0.68 ±  7%      +0.1        0.78 ±  6%  perf-profile.self.cycles-pp.native_sched_clock
       0.38 ±  9%      +0.1        0.53 ± 13%  perf-profile.self.cycles-pp.sysvec_apic_timer_interrupt


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-rc1-00006-g164f4064ca81" of type "text/plain" (170371 bytes)

View attachment "job-script" of type "text/plain" (8302 bytes)

View attachment "job.yaml" of type "text/plain" (5660 bytes)

View attachment "reproduce" of type "text/plain" (290 bytes)

Powered by blists - more mailing lists