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  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date:   Sat, 31 Jul 2021 14:38:18 +0800
From:   kernel test robot <oliver.sang@...el.com>
To:     Gabriel Krisman Bertazi <krisman@...labora.com>
Cc:     0day robot <lkp@...el.com>,
        Gabriel Krisman Bertazi <krisman@...labora.com>,
        LKML <linux-kernel@...r.kernel.org>, lkp@...ts.01.org,
        ying.huang@...el.com, feng.tang@...el.com,
        zhengjun.xing@...ux.intel.com, jack@...e.com, amir73il@...il.com,
        djwong@...nel.org, tytso@....edu, david@...morbit.com,
        dhowells@...hat.com, khazhy@...gle.com,
        linux-fsdevel@...r.kernel.org, linux-ext4@...r.kernel.org,
        kernel@...labora.com
Subject: [fsnotify]  4c40d6efc8:  unixbench.score -3.3% regression



Greeting,

FYI, we noticed a -3.3% regression of unixbench.score due to commit:


commit: 4c40d6efc8b22b88a45c335ffd6d25b55d769f5b ("[PATCH v4 08/16] fsnotify: pass arguments of fsnotify() in struct fsnotify_event_info")
url: https://github.com/0day-ci/linux/commits/Gabriel-Krisman-Bertazi/File-system-wide-monitoring/20210721-001444
base: https://git.kernel.org/cgit/linux/kernel/git/jack/linux-fs.git fsnotify

in testcase: unixbench
on test machine: 96 threads 2 sockets Intel(R) Xeon(R) CPU @ 2.30GHz with 128G memory
with following parameters:

	runtime: 300s
	nr_task: 1
	test: pipe
	cpufreq_governor: performance
	ucode: 0x4003006

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 significant impact on the following tests:

+------------------+-------------------------------------------------------------------------------------+
| testcase: change | will-it-scale: will-it-scale.per_thread_ops -1.3% regression                        |
| test machine     | 192 threads 4 sockets Intel(R) Xeon(R) Platinum 9242 CPU @ 2.30GHz with 192G memory |
| test parameters  | cpufreq_governor=performance                                                        |
|                  | mode=thread                                                                         |
|                  | nr_task=100%                                                                        |
|                  | test=eventfd1                                                                       |
|                  | ucode=0x5003006                                                                     |
+------------------+-------------------------------------------------------------------------------------+


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


Details are as below:
-------------------------------------------------------------------------------------------------->


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 split-job --compatible job.yaml  # generate the yaml file for lkp run
        bin/lkp run                    generated-yaml-file

=========================================================================================
compiler/cpufreq_governor/kconfig/nr_task/rootfs/runtime/tbox_group/test/testcase/ucode:
  gcc-9/performance/x86_64-rhel-8.3/1/debian-10.4-x86_64-20200603.cgz/300s/lkp-csl-2sp4/pipe/unixbench/0x4003006

commit: 
  263b74f276 ("fsnotify: Add helper to detect overflow_event")
  4c40d6efc8 ("fsnotify: pass arguments of fsnotify() in struct fsnotify_event_info")

263b74f2761d777d 4c40d6efc8b22b88a45c335ffd6 
---------------- --------------------------- 
         %stddev     %change         %stddev
             \          |                \  
      1554            -3.3%       1503        unixbench.score
 7.551e+08            -3.4%  7.295e+08        unixbench.workload
      0.00 ±158%   +4075.0%       0.02 ± 37%  perf-sched.wait_time.avg.ms.preempt_schedule_common.__cond_resched.copy_page_from_iter.pipe_write.new_sync_write
      0.00 ±158%   +4075.0%       0.02 ± 37%  perf-sched.wait_time.max.ms.preempt_schedule_common.__cond_resched.copy_page_from_iter.pipe_write.new_sync_write
    318916 ±169%     -95.4%      14689 ±  5%  perf-stat.i.dTLB-load-misses
   5208687 ±  7%     +26.3%    6579325 ±  2%  perf-stat.i.iTLB-load-misses
    991.91 ±  5%     -17.0%     822.95 ±  2%  perf-stat.i.instructions-per-iTLB-miss
      0.02 ±170%      -0.0        0.00 ±  5%  perf-stat.overall.dTLB-load-miss-rate%
    931.13 ±  7%     -22.5%     721.29 ±  2%  perf-stat.overall.instructions-per-iTLB-miss
      2491            +1.8%       2535        perf-stat.overall.path-length
    318091 ±169%     -95.4%      14652 ±  5%  perf-stat.ps.dTLB-load-misses
   5195369 ±  7%     +26.3%    6562257 ±  2%  perf-stat.ps.iTLB-load-misses
      0.82 ± 13%      +0.2        1.02 ±  8%  perf-profile.calltrace.cycles-pp.mutex_lock.pipe_write.new_sync_write.vfs_write.ksys_write
      0.69 ±  8%      +0.2        0.91 ± 11%  perf-profile.calltrace.cycles-pp.security_file_permission.vfs_write.ksys_write.do_syscall_64.entry_SYSCALL_64_after_hwframe
      1.05 ± 10%      +0.3        1.33 ±  6%  perf-profile.calltrace.cycles-pp.security_file_permission.vfs_read.ksys_read.do_syscall_64.entry_SYSCALL_64_after_hwframe
      0.39 ± 63%      +0.3        0.69 ± 11%  perf-profile.calltrace.cycles-pp.common_file_perm.security_file_permission.vfs_write.ksys_write.do_syscall_64
      0.16 ±158%      +0.5        0.68 ±  6%  perf-profile.calltrace.cycles-pp.common_file_perm.security_file_permission.vfs_read.ksys_read.do_syscall_64
      0.64 ± 16%      -0.2        0.46 ± 17%  perf-profile.children.cycles-pp.anon_pipe_buf_release
      0.34 ± 47%      -0.2        0.19 ± 10%  perf-profile.children.cycles-pp.wait_for_xmitr
      0.35 ± 46%      -0.2        0.20 ± 10%  perf-profile.children.cycles-pp.serial8250_console_write
      0.33 ± 47%      -0.1        0.19 ±  9%  perf-profile.children.cycles-pp.serial8250_console_putchar
      0.21 ± 29%      -0.1        0.13 ± 22%  perf-profile.children.cycles-pp.enqueue_hrtimer
      0.18 ± 31%      -0.1        0.11 ± 31%  perf-profile.children.cycles-pp.timerqueue_add
      0.30 ± 14%      -0.1        0.23 ± 15%  perf-profile.children.cycles-pp.update_blocked_averages
      0.10 ± 21%      -0.1        0.04 ± 91%  perf-profile.children.cycles-pp.menu_reflect
      1.37 ±  8%      +0.4        1.76 ±  6%  perf-profile.children.cycles-pp.common_file_perm
      2.27 ±  8%      +0.5        2.82 ±  5%  perf-profile.children.cycles-pp.security_file_permission
      0.00            +1.5        1.47 ± 14%  perf-profile.children.cycles-pp.__fsnotify
      0.63 ± 16%      -0.2        0.45 ± 17%  perf-profile.self.cycles-pp.anon_pipe_buf_release
      0.11 ± 32%      -0.0        0.07 ± 23%  perf-profile.self.cycles-pp.sysvec_apic_timer_interrupt
      0.47 ±  9%      +0.1        0.58 ± 12%  perf-profile.self.cycles-pp.new_sync_write
      0.43 ± 14%      +0.1        0.58 ± 12%  perf-profile.self.cycles-pp.security_file_permission
      0.86 ± 17%      +0.4        1.28 ±  8%  perf-profile.self.cycles-pp.common_file_perm
      0.00            +1.5        1.45 ± 14%  perf-profile.self.cycles-pp.__fsnotify


                                                                                
                                  unixbench.score                               
                                                                                
  1580 +--------------------------------------------------------------------+   
       |                                                                    |   
  1560 |-+                     .+...+...+..  ..+                            |   
       |                   ..+.            +.                               |   
  1540 |-+              .+.                                                 |   
       |...+..+...+...+.                                                    |   
  1520 |-+                                                                  |   
       |                        O   O   O  O                 O   O          |   
  1500 |-+                                     O   O  O   O          O  O   |   
       |                                                                    |   
  1480 |-+            O                                                     |   
       |   O  O   O          O                                              |   
  1460 |-+               O                                                  |   
       |                                                                    |   
  1440 +--------------------------------------------------------------------+   
                                                                                
                                                                                
[*] bisect-good sample
[O] bisect-bad  sample

***************************************************************************************************
lkp-csl-2ap2: 192 threads 4 sockets Intel(R) Xeon(R) Platinum 9242 CPU @ 2.30GHz with 192G memory
=========================================================================================
compiler/cpufreq_governor/kconfig/mode/nr_task/rootfs/tbox_group/test/testcase/ucode:
  gcc-9/performance/x86_64-rhel-8.3/thread/100%/debian-10.4-x86_64-20200603.cgz/lkp-csl-2ap2/eventfd1/will-it-scale/0x5003006

commit: 
  263b74f276 ("fsnotify: Add helper to detect overflow_event")
  4c40d6efc8 ("fsnotify: pass arguments of fsnotify() in struct fsnotify_event_info")

263b74f2761d777d 4c40d6efc8b22b88a45c335ffd6 
---------------- --------------------------- 
         %stddev     %change         %stddev
             \          |                \  
 3.014e+08            -1.3%  2.974e+08        will-it-scale.192.threads
   1569651            -1.3%    1549107        will-it-scale.per_thread_ops
 3.014e+08            -1.3%  2.974e+08        will-it-scale.workload
    747.17 ± 37%   +1300.3%      10462 ±145%  softirqs.CPU13.NET_RX
      1212 ± 41%   +1341.6%      17474 ±133%  interrupts.34:PCI-MSI.524292-edge.eth0-TxRx-3
      1212 ± 41%   +1341.6%      17474 ±133%  interrupts.CPU13.34:PCI-MSI.524292-edge.eth0-TxRx-3
      0.01 ±  8%     +32.5%       0.02 ± 15%  perf-sched.sch_delay.avg.ms.pipe_read.new_sync_read.vfs_read.ksys_read
      2.86 ±162%    +256.5%      10.20 ± 38%  perf-sched.wait_time.avg.ms.preempt_schedule_common.__cond_resched.__do_fault.do_fault.__handle_mm_fault
     38371            -1.5%      37785        proc-vmstat.nr_slab_reclaimable
     71698            -1.8%      70437        proc-vmstat.nr_slab_unreclaimable
      5621 ±  6%     -12.4%       4926 ±  3%  slabinfo.Acpi-State.active_objs
      5621 ±  6%     -12.4%       4926 ±  3%  slabinfo.Acpi-State.num_objs
      5395 ±  7%     -13.4%       4670 ±  7%  slabinfo.files_cache.active_objs
      5395 ±  7%     -13.4%       4670 ±  7%  slabinfo.files_cache.num_objs
   1070171 ±  2%      +8.4%    1160265 ±  2%  perf-stat.i.cache-misses
   7414136 ±  2%     +11.7%    8284549 ±  2%  perf-stat.i.cache-references
    741668 ±  3%     -12.8%     647061 ±  4%  perf-stat.i.cycles-between-cache-misses
 1.598e+11            +1.5%  1.621e+11        perf-stat.i.dTLB-loads
    165125            -2.3%     161301        perf-stat.i.dTLB-store-misses
 5.501e+08 ±  3%      +4.5%  5.746e+08        perf-stat.i.iTLB-load-misses
      1005 ±  4%      -4.3%     962.30        perf-stat.i.instructions-per-iTLB-miss
    237876 ±  2%     +10.3%     262365 ±  3%  perf-stat.i.node-load-misses
     91893 ±  2%      +4.9%      96434        perf-stat.i.node-store-misses
      0.01 ±  2%     +11.4%       0.02 ±  2%  perf-stat.overall.MPKI
    497110 ±  3%      -6.9%     462766 ±  2%  perf-stat.overall.cycles-between-cache-misses
      0.00            -0.0        0.00        perf-stat.overall.dTLB-store-miss-rate%
      1004 ±  4%      -4.5%     960.10        perf-stat.overall.instructions-per-iTLB-miss
    551166            +1.3%     558404        perf-stat.overall.path-length
   1107917 ±  3%      +7.4%    1190240 ±  2%  perf-stat.ps.cache-misses
   7527216 ±  2%     +11.4%    8382817 ±  2%  perf-stat.ps.cache-references
 1.592e+11            +1.5%  1.616e+11        perf-stat.ps.dTLB-loads
    164932            -2.3%     161085        perf-stat.ps.dTLB-store-misses
 5.483e+08 ±  3%      +4.5%  5.729e+08        perf-stat.ps.iTLB-load-misses
    237049 ±  2%     +10.4%     261604 ±  4%  perf-stat.ps.node-load-misses
     91568 ±  2%      +4.9%      96075        perf-stat.ps.node-store-misses
      9.39            -0.3        9.05        perf-profile.calltrace.cycles-pp.eventfd_write.vfs_write.ksys_write.do_syscall_64.entry_SYSCALL_64_after_hwframe
     11.74            -0.2       11.57        perf-profile.calltrace.cycles-pp.eventfd_read.new_sync_read.vfs_read.ksys_read.do_syscall_64
      7.75            -0.1        7.65        perf-profile.calltrace.cycles-pp.__entry_text_start.__libc_read
      7.76            -0.1        7.66        perf-profile.calltrace.cycles-pp.__entry_text_start.__libc_write
      7.46            -0.1        7.37        perf-profile.calltrace.cycles-pp._copy_to_iter.eventfd_read.new_sync_read.vfs_read.ksys_read
      1.54            -0.1        1.45        perf-profile.calltrace.cycles-pp.syscall_return_via_sysret.__libc_write
      1.54            -0.1        1.46        perf-profile.calltrace.cycles-pp.syscall_return_via_sysret.__libc_read
      2.59            -0.1        2.53        perf-profile.calltrace.cycles-pp.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.__libc_read
      1.82            -0.0        1.78        perf-profile.calltrace.cycles-pp.exit_to_user_mode_prepare.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.__libc_read
      2.58            -0.0        2.54        perf-profile.calltrace.cycles-pp.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.__libc_write
      1.81            -0.0        1.78        perf-profile.calltrace.cycles-pp.exit_to_user_mode_prepare.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.__libc_write
      1.06            -0.0        1.03        perf-profile.calltrace.cycles-pp.copy_user_enhanced_fast_string._copy_from_user.eventfd_write.vfs_write.ksys_write
      1.74            -0.0        1.70        perf-profile.calltrace.cycles-pp._raw_spin_lock_irq.eventfd_read.new_sync_read.vfs_read.ksys_read
      1.34            -0.0        1.32        perf-profile.calltrace.cycles-pp.fput_many.ksys_write.do_syscall_64.entry_SYSCALL_64_after_hwframe.__libc_write
      0.55            -0.0        0.53        perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_safe_stack.__libc_read
      0.64            +0.0        0.67        perf-profile.calltrace.cycles-pp.__might_sleep.__might_fault._copy_from_user.eventfd_write.vfs_write
      1.91            +0.0        1.95        perf-profile.calltrace.cycles-pp.__might_fault._copy_to_iter.eventfd_read.new_sync_read.vfs_read
      1.99            +0.1        2.04        perf-profile.calltrace.cycles-pp.__might_fault._copy_from_user.eventfd_write.vfs_write.ksys_write
     42.27            +0.1       42.38        perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.__libc_read
     46.96            +0.1       47.07        perf-profile.calltrace.cycles-pp.__libc_write
      0.68 ±  6%      +0.1        0.82 ±  4%  perf-profile.calltrace.cycles-pp.__x64_sys_write.do_syscall_64.entry_SYSCALL_64_after_hwframe.__libc_write
     40.77            +0.2       40.94        perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.__libc_read
     36.01            +0.3       36.27        perf-profile.calltrace.cycles-pp.ksys_read.do_syscall_64.entry_SYSCALL_64_after_hwframe.__libc_read
     26.10            +0.3       26.36        perf-profile.calltrace.cycles-pp.ksys_write.do_syscall_64.entry_SYSCALL_64_after_hwframe.__libc_write
     32.26            +0.3       32.55        perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.__libc_write
     30.79            +0.3       31.13        perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.__libc_write
     28.76            +0.6       29.33        perf-profile.calltrace.cycles-pp.vfs_read.ksys_read.do_syscall_64.entry_SYSCALL_64_after_hwframe.__libc_read
      4.33 ±  4%      +0.8        5.08 ±  2%  perf-profile.calltrace.cycles-pp.security_file_permission.vfs_write.ksys_write.do_syscall_64.entry_SYSCALL_64_after_hwframe
     18.63 ±  2%      +0.8       19.40        perf-profile.calltrace.cycles-pp.vfs_write.ksys_write.do_syscall_64.entry_SYSCALL_64_after_hwframe.__libc_write
      7.72 ±  3%      +0.8        8.49        perf-profile.calltrace.cycles-pp.security_file_permission.vfs_read.ksys_read.do_syscall_64.entry_SYSCALL_64_after_hwframe
      0.00            +1.7        1.66 ±  3%  perf-profile.calltrace.cycles-pp.__fsnotify.vfs_write.ksys_write.do_syscall_64.entry_SYSCALL_64_after_hwframe
      0.00            +2.0        1.96        perf-profile.calltrace.cycles-pp.__fsnotify.vfs_read.ksys_read.do_syscall_64.entry_SYSCALL_64_after_hwframe
      0.00            +2.0        2.03 ±  2%  perf-profile.calltrace.cycles-pp.__fsnotify.security_file_permission.vfs_read.ksys_read.do_syscall_64
      7.05 ±  2%      -7.0        0.00        perf-profile.children.cycles-pp.fsnotify
      9.54            -0.3        9.22        perf-profile.children.cycles-pp.eventfd_write
      8.66            -0.2        8.42        perf-profile.children.cycles-pp.syscall_return_via_sysret
     11.96            -0.2       11.78        perf-profile.children.cycles-pp.eventfd_read
     57.15            -0.2       56.98        perf-profile.children.cycles-pp.__libc_read
     10.01            -0.1        9.88        perf-profile.children.cycles-pp.__entry_text_start
      3.74            -0.1        3.66        perf-profile.children.cycles-pp.exit_to_user_mode_prepare
      5.51            -0.1        5.43        perf-profile.children.cycles-pp.syscall_exit_to_user_mode
      3.62            -0.1        3.56        perf-profile.children.cycles-pp._raw_spin_lock_irq
      2.30            -0.0        2.26        perf-profile.children.cycles-pp.copy_user_enhanced_fast_string
      3.27            -0.0        3.22        perf-profile.children.cycles-pp.copy_user_generic_unrolled
      2.64            -0.0        2.62        perf-profile.children.cycles-pp.fput_many
      2.06            +0.0        2.09        perf-profile.children.cycles-pp.___might_sleep
      1.27            +0.0        1.32        perf-profile.children.cycles-pp.__might_sleep
      4.17            +0.1        4.26        perf-profile.children.cycles-pp.__might_fault
     47.14            +0.1       47.25        perf-profile.children.cycles-pp.__libc_write
      0.72 ±  5%      +0.1        0.84 ±  3%  perf-profile.children.cycles-pp.__x64_sys_write
     26.35            +0.2       26.59        perf-profile.children.cycles-pp.ksys_write
     36.20            +0.3       36.48        perf-profile.children.cycles-pp.ksys_read
     74.75            +0.4       75.16        perf-profile.children.cycles-pp.entry_SYSCALL_64_after_hwframe
      0.48 ±  3%      +0.5        1.00 ±  3%  perf-profile.children.cycles-pp.apparmor_file_permission
     71.83            +0.6       72.38        perf-profile.children.cycles-pp.do_syscall_64
     18.86 ±  2%      +0.7       19.59        perf-profile.children.cycles-pp.vfs_write
     12.35 ±  3%      +1.4       13.79        perf-profile.children.cycles-pp.security_file_permission
      0.00            +5.8        5.80        perf-profile.children.cycles-pp.__fsnotify
      6.73 ±  2%      -6.7        0.00        perf-profile.self.cycles-pp.fsnotify
      2.20 ±  2%      -0.3        1.93 ±  2%  perf-profile.self.cycles-pp.eventfd_write
      8.57            -0.2        8.34        perf-profile.self.cycles-pp.syscall_return_via_sysret
      2.96            -0.1        2.81        perf-profile.self.cycles-pp.entry_SYSCALL_64_after_hwframe
      1.51            -0.1        1.41 ±  2%  perf-profile.self.cycles-pp.ksys_write
      3.32            -0.1        3.25        perf-profile.self.cycles-pp.exit_to_user_mode_prepare
      4.50            -0.1        4.43        perf-profile.self.cycles-pp.__entry_text_start
      2.57            -0.1        2.50        perf-profile.self.cycles-pp.eventfd_read
      2.53            -0.1        2.46        perf-profile.self.cycles-pp._copy_to_iter
      3.45            -0.1        3.39        perf-profile.self.cycles-pp._raw_spin_lock_irq
      2.52            -0.0        2.47        perf-profile.self.cycles-pp.fput_many
      0.83            -0.0        0.79        perf-profile.self.cycles-pp._copy_from_user
      2.10            -0.0        2.06        perf-profile.self.cycles-pp.copy_user_enhanced_fast_string
      0.92            -0.0        0.89        perf-profile.self.cycles-pp.syscall_exit_to_user_mode_prepare
      2.01            +0.0        2.04        perf-profile.self.cycles-pp.___might_sleep
      1.11            +0.0        1.16        perf-profile.self.cycles-pp.__might_sleep
      0.70 ±  4%      +0.1        0.80 ±  2%  perf-profile.self.cycles-pp.__x64_sys_write
      2.99 ±  2%      +0.4        3.36 ±  2%  perf-profile.self.cycles-pp.vfs_read
      2.53 ±  3%      +0.5        3.03 ±  4%  perf-profile.self.cycles-pp.vfs_write
      0.38 ±  4%      +0.5        0.89 ±  3%  perf-profile.self.cycles-pp.apparmor_file_permission
      2.64 ± 13%      +0.9        3.54 ±  9%  perf-profile.self.cycles-pp.security_file_permission
      0.00            +5.6        5.59        perf-profile.self.cycles-pp.__fsnotify





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.


---
0DAY/LKP+ Test Infrastructure                   Open Source Technology Center
https://lists.01.org/hyperkitty/list/lkp@lists.01.org       Intel Corporation

Thanks,
Oliver Sang


View attachment "config-5.13.0-rc5-00009-g4c40d6efc8b2" of type "text/plain" (174036 bytes)

View attachment "job-script" of type "text/plain" (8097 bytes)

View attachment "job.yaml" of type "text/plain" (5454 bytes)

View attachment "reproduce" of type "text/plain" (276 bytes)

Powered by blists - more mailing lists