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:   Wed, 28 Feb 2018 09:35:34 +0800
From:   kernel test robot <shun.hao@...el.com>
To:     Mel Gorman <mgorman@...hsingularity.net>
Cc:     Ingo Molnar <mingo@...nel.org>,
        Peter Zijlstra <peterz@...radead.org>,
        Linus Torvalds <torvalds@...ux-foundation.org>,
        Matt Fleming <matt@...eblueprint.co.uk>,
        Mike Galbraith <efault@....de>,
        Thomas Gleixner <tglx@...utronix.de>,
        LKML <linux-kernel@...r.kernel.org>,
        linux-kernel@...r.kernel.org, lkp@...org
Subject: [lkp-robot] [sched/fair]  806486c377:  fsmark.app_overhead 39.4%
 improvement

Greeting,

FYI, we noticed a 39.4% improvement of fsmark.app_overhead due to commit:


commit: 806486c377e33ab662de6d47902e9e2a32b79368 ("sched/fair: Do not migrate if the prev_cpu is idle")
https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git master

in testcase: fsmark
on test machine: 4 threads Intel(R) Core(TM) i5-2300 CPU @ 2.80GHz with 4G memory
with following parameters:

	iterations: 1x
	nr_threads: 32t
	disk: 1HDD
	fs: ext4
	filesize: 16MB
	test_size: 60G
	sync_method: fsyncBeforeClose
	nr_directories: 16d
	nr_files_per_directory: 256fpd
	cpufreq_governor: performance

test-description: The fsmark is a file system benchmark to test synchronous write workloads, for example, mail servers workload.
test-url: https://sourceforge.net/projects/fsmark/



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 run     job.yaml

=========================================================================================
compiler/cpufreq_governor/disk/filesize/fs/iterations/kconfig/nr_directories/nr_files_per_directory/nr_threads/rootfs/sync_method/tbox_group/test_size/testcase:
  gcc-7/performance/1HDD/16MB/ext4/1x/x86_64-rhel-7.2/16d/256fpd/32t/debian-x86_64-2016-08-31.cgz/fsyncBeforeClose/lkp-sb02/60G/fsmark

commit: 
  3b76c4a339 ("sched/fair: Restructure wake_affine*() to return a CPU id")
  806486c377 ("sched/fair: Do not migrate if the prev_cpu is idle")

3b76c4a33959ca98 806486c377e33ab662de6d4790 
---------------- -------------------------- 
       fail:runs  %reproduction    fail:runs
           |             |             |    
          0:4           -1%           0:4     perf-profile.children.cycles-pp.schedule_timeout
         %stddev     %change         %stddev
             \          |                \  
    538776 ± 28%     +39.4%     751196 ± 11%  fsmark.app_overhead
     57.14            +1.7%      58.08        fsmark.time.system_time
    292997 ± 13%    +205.4%     894676 ±  3%  fsmark.time.voluntary_context_switches
     75688 ±  9%     +16.9%      88488 ±  5%  meminfo.DirectMap4k
     18.47 ± 48%     -11.1        7.40 ± 14%  mpstat.cpu.idle%
     18446 ± 12%     +24.3%      22936 ±  2%  vmstat.system.cs
      1064 ±  8%     +23.9%       1318 ± 11%  proc-vmstat.kswapd_low_wmark_hit_quickly
      1471 ±  5%     +15.5%       1699 ±  6%  proc-vmstat.pageoutrun
 1.082e+09 ±  3%      -3.3%  1.047e+09        perf-stat.cache-misses
   9715021 ±  2%     +15.0%   11175392 ±  2%  perf-stat.context-switches
     27238 ±  4%     -39.0%      16602 ±  2%  perf-stat.cpu-migrations
 2.654e+08 ± 31%     -44.6%  1.469e+08 ± 11%  cpuidle.C3.time
    287354 ± 37%     -46.0%     155067 ± 10%  cpuidle.C3.usage
 3.639e+08 ± 61%     -67.4%  1.188e+08 ± 20%  cpuidle.C6.time
    403340 ± 67%     -67.9%     129309 ± 18%  cpuidle.C6.usage
    225.50 ± 58%     +66.0%     374.25 ± 12%  turbostat.Avg_MHz
      8.05 ± 58%      +5.1       13.20 ± 11%  turbostat.Busy%
    872073 ± 58%     +50.8%    1315008 ±  2%  turbostat.C1E
      0.35 ± 59%    +144.6%       0.85 ± 31%  turbostat.Pkg%pc3
    130.50 ±  7%     -17.8%     107.30 ± 10%  sched_debug.cpu.cpu_load[2].stddev
     80.59 ±  3%     -12.9%      70.22 ± 10%  sched_debug.cpu.cpu_load[3].stddev
      7.46 ± 21%    +142.8%      18.11 ± 76%  sched_debug.cpu.cpu_load[4].min
    174257 ± 16%    +136.5%     412171 ±  8%  sched_debug.cpu.nr_switches.min
    511.80 ±  4%     -66.6%     170.89 ± 32%  sched_debug.cpu.nr_uninterruptible.max
   -752.16           -80.6%    -145.58        sched_debug.cpu.nr_uninterruptible.min
    524.29 ±  7%     -76.0%     125.75 ± 25%  sched_debug.cpu.nr_uninterruptible.stddev
    158052 ± 19%    +153.6%     400742 ±  8%  sched_debug.cpu.sched_count.min
     74333 ± 22%    +164.3%     196451 ±  9%  sched_debug.cpu.sched_goidle.min
      3.54 ± 12%      -2.6        0.96 ±102%  perf-profile.calltrace.cycles-pp.ret_from_intr.cpuidle_enter_state.do_idle.cpu_startup_entry.start_secondary
      6.38 ± 16%      -2.4        3.96 ± 33%  perf-profile.calltrace.cycles-pp.kswapd.kthread.ret_from_fork
      6.36 ± 16%      -2.4        3.94 ± 33%  perf-profile.calltrace.cycles-pp.shrink_node.kswapd.kthread.ret_from_fork
      6.23 ± 16%      -2.4        3.87 ± 33%  perf-profile.calltrace.cycles-pp.shrink_node_memcg.shrink_node.kswapd.kthread.ret_from_fork
      6.23 ± 16%      -2.4        3.87 ± 33%  perf-profile.calltrace.cycles-pp.shrink_inactive_list.shrink_node_memcg.shrink_node.kswapd.kthread
      6.85 ± 12%      -2.3        4.57 ± 29%  perf-profile.calltrace.cycles-pp.ret_from_fork
      6.85 ± 12%      -2.3        4.57 ± 29%  perf-profile.calltrace.cycles-pp.kthread.ret_from_fork
      5.43 ± 18%      -2.2        3.26 ± 35%  perf-profile.calltrace.cycles-pp.shrink_page_list.shrink_inactive_list.shrink_node_memcg.shrink_node.kswapd
      5.72 ± 24%      -2.0        3.73 ± 34%  perf-profile.calltrace.cycles-pp.pagecache_get_page.grab_cache_page_write_begin.ext4_da_write_begin.generic_perform_write.__generic_file_write_iter
      2.83 ± 45%      -2.0        0.84 ± 49%  perf-profile.calltrace.cycles-pp.wait_for_xmitr.serial8250_console_putchar.uart_console_write.serial8250_console_write.console_unlock
      2.46 ± 37%      -1.8        0.66 ± 18%  perf-profile.calltrace.cycles-pp.uart_console_write.serial8250_console_write.console_unlock.vprintk_emit.printk
      2.46 ± 37%      -1.8        0.67 ± 19%  perf-profile.calltrace.cycles-pp.serial8250_console_write.console_unlock.vprintk_emit.printk.irq_work_run_list
      2.44 ± 38%      -1.8        0.65 ± 16%  perf-profile.calltrace.cycles-pp.serial8250_console_putchar.uart_console_write.serial8250_console_write.console_unlock.vprintk_emit
      2.71 ± 37%      -1.6        1.11 ± 67%  perf-profile.calltrace.cycles-pp.printk.irq_work_run_list.irq_work_run.smp_irq_work_interrupt.irq_work_interrupt
      2.71 ± 37%      -1.6        1.11 ± 67%  perf-profile.calltrace.cycles-pp.vprintk_emit.printk.irq_work_run_list.irq_work_run.smp_irq_work_interrupt
      2.71 ± 37%      -1.6        1.11 ± 67%  perf-profile.calltrace.cycles-pp.console_unlock.vprintk_emit.printk.irq_work_run_list.irq_work_run
      4.65 ± 11%      -1.5        3.12 ± 32%  perf-profile.calltrace.cycles-pp.ext4_block_write_begin.ext4_da_write_begin.generic_perform_write.__generic_file_write_iter.ext4_file_write_iter
      2.75 ± 48%      -1.2        1.56 ± 27%  perf-profile.calltrace.cycles-pp.ext4_do_update_inode.ext4_mark_iloc_dirty.ext4_mark_inode_dirty.ext4_dirty_inode.__mark_inode_dirty
      2.84 ± 19%      -1.1        1.76 ± 31%  perf-profile.calltrace.cycles-pp.block_write_end.generic_write_end.ext4_da_write_end.generic_perform_write.__generic_file_write_iter
      3.54 ± 12%      -1.0        2.50 ± 26%  perf-profile.calltrace.cycles-pp.do_IRQ.ret_from_intr.cpuidle_enter_state.do_idle.cpu_startup_entry
      1.68 ±  4%      -0.9        0.81 ± 58%  perf-profile.calltrace.cycles-pp.create_empty_buffers.ext4_block_write_begin.ext4_da_write_begin.generic_perform_write.__generic_file_write_iter
      1.54 ± 29%      -0.8        0.73 ± 63%  perf-profile.calltrace.cycles-pp.__remove_mapping.shrink_page_list.shrink_inactive_list.shrink_node_memcg.shrink_node
      1.29 ± 12%      -0.7        0.59 ± 61%  perf-profile.calltrace.cycles-pp.alloc_page_buffers.create_empty_buffers.ext4_block_write_begin.ext4_da_write_begin.generic_perform_write
      1.23 ± 11%      -0.7        0.56 ± 60%  perf-profile.calltrace.cycles-pp.alloc_buffer_head.alloc_page_buffers.create_empty_buffers.ext4_block_write_begin.ext4_da_write_begin
      1.19 ± 12%      -0.7        0.53 ± 60%  perf-profile.calltrace.cycles-pp.kmem_cache_alloc.alloc_buffer_head.alloc_page_buffers.create_empty_buffers.ext4_block_write_begin
      1.73 ± 23%      -0.6        1.10 ± 32%  perf-profile.calltrace.cycles-pp.intel_idle.cpuidle_enter_state.do_idle.cpu_startup_entry.start_secondary
      1.27 ± 19%      -0.6        0.65 ± 64%  perf-profile.calltrace.cycles-pp.try_to_free_buffers.shrink_page_list.shrink_inactive_list.shrink_node_memcg.shrink_node
      0.84 ± 24%      -0.5        0.33 ±100%  perf-profile.calltrace.cycles-pp.free_unref_page_commit.free_unref_page_list.shrink_page_list.shrink_inactive_list.shrink_node_memcg
      0.96 ± 22%      -0.5        0.50 ± 61%  perf-profile.calltrace.cycles-pp.free_unref_page_list.shrink_page_list.shrink_inactive_list.shrink_node_memcg.shrink_node
      0.77 ± 14%      -0.3        0.46 ± 58%  perf-profile.calltrace.cycles-pp.isolate_lru_pages.shrink_inactive_list.shrink_node_memcg.shrink_node.kswapd
      6.42 ± 16%      -2.5        3.94 ± 33%  perf-profile.children.cycles-pp.shrink_node
      6.30 ± 16%      -2.4        3.87 ± 33%  perf-profile.children.cycles-pp.shrink_inactive_list
      6.38 ± 16%      -2.4        3.96 ± 33%  perf-profile.children.cycles-pp.kswapd
      6.29 ± 16%      -2.4        3.87 ± 33%  perf-profile.children.cycles-pp.shrink_node_memcg
      6.87 ± 12%      -2.3        4.57 ± 29%  perf-profile.children.cycles-pp.ret_from_fork
      6.85 ± 12%      -2.3        4.57 ± 29%  perf-profile.children.cycles-pp.kthread
      5.48 ± 18%      -2.2        3.29 ± 36%  perf-profile.children.cycles-pp.shrink_page_list
      4.72 ± 12%      -1.6        3.15 ± 32%  perf-profile.children.cycles-pp.ext4_block_write_begin
      3.61 ± 21%      -1.3        2.30 ± 30%  perf-profile.children.cycles-pp.serial8250_console_putchar
      3.31 ± 17%      -1.3        2.01 ± 25%  perf-profile.children.cycles-pp.irq_work_interrupt
      3.31 ± 17%      -1.3        2.01 ± 25%  perf-profile.children.cycles-pp.smp_irq_work_interrupt
      3.31 ± 17%      -1.3        2.01 ± 25%  perf-profile.children.cycles-pp.irq_work_run
      3.31 ± 17%      -1.3        2.01 ± 25%  perf-profile.children.cycles-pp.printk
      3.31 ± 17%      -1.3        2.01 ± 25%  perf-profile.children.cycles-pp.vprintk_emit
      3.62 ± 21%      -1.3        2.33 ± 29%  perf-profile.children.cycles-pp.wait_for_xmitr
      2.36 ± 27%      -0.9        1.43 ± 25%  perf-profile.children.cycles-pp.intel_idle
      1.75 ±  4%      -0.8        0.93 ± 29%  perf-profile.children.cycles-pp.kmem_cache_alloc
      1.69 ±  3%      -0.8        0.94 ± 30%  perf-profile.children.cycles-pp.create_empty_buffers
      1.33 ± 11%      -0.6        0.69 ± 32%  perf-profile.children.cycles-pp.alloc_page_buffers
      1.29 ±  8%      -0.6        0.68 ± 32%  perf-profile.children.cycles-pp.alloc_buffer_head
      1.29 ± 19%      -0.5        0.76 ± 36%  perf-profile.children.cycles-pp.try_to_free_buffers
      1.22 ± 17%      -0.5        0.76 ± 28%  perf-profile.children.cycles-pp.delay_tsc
      1.02 ± 20%      -0.4        0.66 ± 32%  perf-profile.children.cycles-pp.free_unref_page_list
      0.87 ± 22%      -0.3        0.52 ± 37%  perf-profile.children.cycles-pp.free_unref_page_commit
      0.56 ± 44%      -0.3        0.22 ± 31%  perf-profile.children.cycles-pp.scsi_run_queue
      0.75 ± 25%      -0.3        0.42 ± 29%  perf-profile.children.cycles-pp.clear_page_dirty_for_io
      1.06 ± 18%      -0.3        0.72 ± 25%  perf-profile.children.cycles-pp._raw_spin_lock_irqsave
      0.61 ± 20%      -0.3        0.29 ± 49%  perf-profile.children.cycles-pp.page_fault
      0.61 ± 21%      -0.3        0.29 ± 47%  perf-profile.children.cycles-pp.do_page_fault
      0.61 ± 21%      -0.3        0.29 ± 47%  perf-profile.children.cycles-pp.__do_page_fault
      0.74 ± 27%      -0.3        0.43 ± 38%  perf-profile.children.cycles-pp.free_pcppages_bulk
      0.55 ± 46%      -0.3        0.25 ± 37%  perf-profile.children.cycles-pp.__slab_alloc
      0.65 ± 35%      -0.3        0.36 ± 43%  perf-profile.children.cycles-pp.drop_buffers
      0.70 ± 22%      -0.2        0.45 ± 42%  perf-profile.children.cycles-pp._raw_read_lock
      0.45 ± 31%      -0.2        0.21 ± 13%  perf-profile.children.cycles-pp.radix_tree_tag_set
      0.55 ± 29%      -0.2        0.31 ± 16%  perf-profile.children.cycles-pp.__tick_nohz_idle_enter
      0.78 ± 11%      -0.2        0.55 ± 22%  perf-profile.children.cycles-pp.isolate_lru_pages
      0.53 ± 20%      -0.2        0.31 ± 35%  perf-profile.children.cycles-pp.account_page_dirtied
      0.43 ±  9%      -0.2        0.22 ± 31%  perf-profile.children.cycles-pp.blk_run_queue
      0.44 ±  9%      -0.2        0.24 ± 28%  perf-profile.children.cycles-pp.__blk_run_queue
      0.44 ±  9%      -0.2        0.24 ± 28%  perf-profile.children.cycles-pp.scsi_request_fn
      0.44 ± 36%      -0.2        0.24 ± 42%  perf-profile.children.cycles-pp.find_get_entry
      0.30 ± 19%      -0.2        0.13 ± 41%  perf-profile.children.cycles-pp.mem_cgroup_commit_charge
      0.32 ± 31%      -0.2        0.15 ± 16%  perf-profile.children.cycles-pp.dec_zone_page_state
      0.32 ± 17%      -0.2        0.16 ± 25%  perf-profile.children.cycles-pp.blk_peek_request
      0.29 ± 40%      -0.2        0.13 ± 44%  perf-profile.children.cycles-pp.blk_queue_split
      0.39 ± 32%      -0.2        0.23 ± 24%  perf-profile.children.cycles-pp.execve
      0.40 ± 11%      -0.2        0.24 ± 37%  perf-profile.children.cycles-pp.__getblk_gfp
      0.30 ± 15%      -0.1        0.16 ± 27%  perf-profile.children.cycles-pp.scsi_prep_fn
      0.27 ± 15%      -0.1        0.13 ± 37%  perf-profile.children.cycles-pp.__radix_tree_create
      0.20 ± 26%      -0.1        0.07 ± 60%  perf-profile.children.cycles-pp.__mod_node_page_state
      0.23 ± 29%      -0.1        0.11 ± 59%  perf-profile.children.cycles-pp.do_filp_open
      0.17 ± 40%      -0.1        0.05 ± 62%  perf-profile.children.cycles-pp.__brelse
      0.22 ± 32%      -0.1        0.11 ± 59%  perf-profile.children.cycles-pp.path_openat
      0.21 ± 19%      -0.1        0.10 ± 46%  perf-profile.children.cycles-pp.do_sys_open
      0.27 ±  9%      -0.1        0.16 ± 32%  perf-profile.children.cycles-pp.ktime_get
      0.24 ± 24%      -0.1        0.13 ± 31%  perf-profile.children.cycles-pp.blk_rq_map_sg
      0.24 ± 29%      -0.1        0.14 ± 39%  perf-profile.children.cycles-pp.ahci_single_level_irq_intr
      0.24 ±  9%      -0.1        0.14 ± 19%  perf-profile.children.cycles-pp.clockevents_program_event
      0.15 ± 26%      -0.1        0.05 ± 74%  perf-profile.children.cycles-pp.ahci_handle_port_interrupt
      0.24 ±  5%      -0.1        0.14 ± 43%  perf-profile.children.cycles-pp.drm_fb_helper_cfb_imageblit
      0.24 ±  5%      -0.1        0.14 ± 43%  perf-profile.children.cycles-pp.cfb_imageblit
      0.14 ± 30%      -0.1        0.06 ± 58%  perf-profile.children.cycles-pp.__wake_up_common_lock
      0.24 ± 20%      -0.1        0.15 ± 23%  perf-profile.children.cycles-pp.search_binary_handler
      0.23 ± 21%      -0.1        0.15 ± 23%  perf-profile.children.cycles-pp.load_elf_binary
      0.27 ± 12%      -0.1        0.19 ± 32%  perf-profile.children.cycles-pp.bit_putcs
      0.18 ±  8%      -0.1        0.12 ± 30%  perf-profile.children.cycles-pp.__list_add_valid
      0.15 ± 19%      -0.1        0.10 ± 31%  perf-profile.children.cycles-pp.lapic_next_deadline
      0.08 ± 26%      +0.1        0.14 ± 27%  perf-profile.children.cycles-pp.kjournald2
      0.08 ± 26%      +0.1        0.14 ± 27%  perf-profile.children.cycles-pp.jbd2_journal_commit_transaction
      0.01 ±173%      +0.1        0.08 ± 24%  perf-profile.children.cycles-pp.rcu_process_callbacks
      0.00            +0.1        0.10 ± 26%  perf-profile.children.cycles-pp.filemap_fdatawait_keep_errors
      0.32 ± 49%      +0.5        0.79 ± 32%  perf-profile.children.cycles-pp.wake_up_page_bit
      2.34 ± 27%      -0.9        1.42 ± 25%  perf-profile.self.cycles-pp.intel_idle
      2.06 ± 16%      -0.7        1.36 ± 33%  perf-profile.self.cycles-pp.block_write_end
      1.06 ± 29%      -0.5        0.59 ± 32%  perf-profile.self.cycles-pp.get_page_from_freelist
      1.22 ± 17%      -0.5        0.76 ± 28%  perf-profile.self.cycles-pp.delay_tsc
      0.78 ± 12%      -0.3        0.51 ± 22%  perf-profile.self.cycles-pp.ext4_finish_bio
      0.69 ± 22%      -0.3        0.43 ± 42%  perf-profile.self.cycles-pp._raw_read_lock
      0.57 ± 24%      -0.2        0.32 ± 41%  perf-profile.self.cycles-pp.free_pcppages_bulk
      0.59 ± 20%      -0.2        0.35 ± 25%  perf-profile.self.cycles-pp.ext4_bio_write_page
      0.45 ± 31%      -0.2        0.21 ± 13%  perf-profile.self.cycles-pp.radix_tree_tag_set
      0.47 ± 29%      -0.2        0.26 ± 31%  perf-profile.self.cycles-pp.ext4_journal_check_start
      0.39 ± 20%      -0.2        0.21 ± 23%  perf-profile.self.cycles-pp.__test_set_page_writeback
      0.32 ± 31%      -0.2        0.15 ± 16%  perf-profile.self.cycles-pp.dec_zone_page_state
      0.36 ± 26%      -0.1        0.22 ± 26%  perf-profile.self.cycles-pp.mpage_map_and_submit_buffers
      0.20 ± 26%      -0.1        0.07 ± 60%  perf-profile.self.cycles-pp.__mod_node_page_state
      0.32 ± 17%      -0.1        0.20 ± 35%  perf-profile.self.cycles-pp.cpuidle_enter_state
      0.17 ± 40%      -0.1        0.05 ± 62%  perf-profile.self.cycles-pp.__brelse
      0.16 ± 31%      -0.1        0.06 ± 60%  perf-profile.self.cycles-pp.balance_dirty_pages_ratelimited
      0.21 ± 10%      -0.1        0.11 ± 34%  perf-profile.self.cycles-pp.__radix_tree_create
      0.23 ± 26%      -0.1        0.12 ± 33%  perf-profile.self.cycles-pp.blk_rq_map_sg
      0.24 ± 33%      -0.1        0.13 ± 17%  perf-profile.self.cycles-pp.__remove_mapping
      0.24 ± 19%      -0.1        0.14 ± 28%  perf-profile.self.cycles-pp.jbd2__journal_start
      0.24 ±  5%      -0.1        0.14 ± 43%  perf-profile.self.cycles-pp.cfb_imageblit
      0.22 ±  9%      -0.1        0.14 ± 38%  perf-profile.self.cycles-pp.__find_get_block
      0.12 ± 35%      -0.1        0.04 ±107%  perf-profile.self.cycles-pp.__lru_cache_add
      0.11 ± 27%      -0.1        0.04 ±107%  perf-profile.self.cycles-pp.ktime_get
      0.15 ± 22%      -0.1        0.07 ± 65%  perf-profile.self.cycles-pp.pagecache_get_page
      0.11 ± 17%      -0.1        0.04 ±100%  perf-profile.self.cycles-pp.entry_SYSCALL_64_fastpath
      0.18 ±  8%      -0.1        0.12 ± 30%  perf-profile.self.cycles-pp.__list_add_valid
      0.12 ± 24%      -0.1        0.07 ± 19%  perf-profile.self.cycles-pp.mem_cgroup_commit_charge
      0.08 ± 21%      -0.1        0.03 ±102%  perf-profile.self.cycles-pp.ext4_mark_inode_dirty
      0.05 ± 62%      +0.1        0.13 ± 27%  perf-profile.self.cycles-pp.try_to_wake_up
     34.50 ± 37%     +19.8       54.29 ± 23%  perf-profile.self.cycles-pp.poll_idle


                                                                                
                              perf-stat.cpu-migrations                          
                                                                                
  30000 +-+-----------------------------------------------------------------+   
        |     +     +                   +         +                   +.    |   
  28000 +-+  + +   + :            +    + + .+    + +                 +  +.+ |   
  26000 +-+.+   +.+  :           + + .+   +  + .+   +. .+.+.+.     .+      +|   
        | +           +   +.+.+.+   +         +       +       +.+.+         |   
  24000 +-+            + +                                                  |   
        |               +                                                   |   
  22000 +-+                                                                 |   
        |                                                                   |   
  20000 +-+                                                                 |   
  18000 +-+                       O   O                                     |   
        O O     O O O O   O   O O   O   O     O                             |   
  16000 +-+ O O             O             O O   O                           |   
        |               O                                                   |   
  14000 +-+-----------------------------------------------------------------+   
                                                                                
                                                                                                                                                                
                       fsmark.time.voluntary_context_switches                   
                                                                                
   1e+06 +-+----------------------------------------------------------------+   
         |                         O   O    O                               |   
  900000 +-+         O     O         O   O O    O                           |   
         O   O O O O   O O   O O O            O                             |   
  800000 +-O                                                                |   
  700000 +-+                                                                |   
         |                                                                  |   
  600000 +-+                                                                |   
         |                                                                  |   
  500000 +-+                                                                |   
  400000 +-+                                                                |   
         |.+.     .+.+.       .+    .+.      .+.     .+. .+.       .+.+.    |   
  300000 +-+ +.+.+     +.+.+.+  + .+   +.+.++   +.+.+   +   +.+.+.+     +.  |   
         |                       +                                        +.|   
  200000 +-+----------------------------------------------------------------+   
                                                                                
                                                                                                                                                                
                           time.voluntary_context_switches                      
                                                                                
   1e+06 +-+----------------------------------------------------------------+   
         |                         O   O    O                               |   
  900000 +-+         O     O         O   O O    O                           |   
         O   O O O O   O O   O O O            O                             |   
  800000 +-O                                                                |   
  700000 +-+                                                                |   
         |                                                                  |   
  600000 +-+                                                                |   
         |                                                                  |   
  500000 +-+                                                                |   
  400000 +-+                                                                |   
         |.+.     .+.+.       .+    .+.      .+.     .+. .+.       .+.+.    |   
  300000 +-+ +.+.+     +.+.+.+  + .+   +.+.++   +.+.+   +   +.+.+.+     +.  |   
         |                       +                                        +.|   
  200000 +-+----------------------------------------------------------------+   
                                                                                
                                                                                
[*] 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.



View attachment "config-4.15.0-01969-g806486c" of type "text/plain" (164733 bytes)

View attachment "job.yaml" of type "text/plain" (5356 bytes)

View attachment "reproduce" of type "text/plain" (1057 bytes)

Powered by blists - more mailing lists