| 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
| ||
|
Date: Mon, 1 Aug 2016 09:59:10 +0800
From: kernel test robot <xiaolong.ye@...el.com>
To: Mikulas Patocka <mpatocka@...hat.com>
Cc: Stephen Rothwell <sfr@...b.auug.org.au>,
Hugh Dickins <hughd@...gle.com>,
Michal Hocko <mhocko@...nel.org>,
Alexander Viro <viro@...iv.linux.org.uk>,
Mel Gorman <mgorman@...e.de>,
Johannes Weiner <hannes@...xchg.org>,
Andrew Morton <akpm@...ux-foundation.org>,
LKML <linux-kernel@...r.kernel.org>, lkp@...org
Subject: [lkp] [mm] b6a927bd24: hackbench.throughput 7.9% improvement
FYI, we noticed a 7.9% improvement of hackbench.throughput due to commit:
commit b6a927bd2442259fc049b7ea6a451bb7436a4d22 ("mm: optimize copy_page_to/from_iter_iovec")
https://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git master
in testcase: hackbench
on test machine: 12 threads Westmere with 6G memory
with following parameters:
nr_threads: 50%
mode: process
ipc: pipe
cpufreq_governor: performance
In addition to that, the commit also has significant impact on the following tests:
+------------------+---------------------------------------------------+
| testcase: change | hackbench: hackbench.throughput 14.7% improvement |
| test machine | 12 threads Westmere with 6G memory |
| test parameters | cpufreq_governor=performance |
| | ipc=pipe |
| | mode=threads |
| | nr_threads=50% |
+------------------+---------------------------------------------------+
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.
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
=========================================================================================
compiler/cpufreq_governor/ipc/kconfig/mode/nr_threads/rootfs/tbox_group/testcase:
gcc-6/performance/pipe/x86_64-rhel/process/50%/debian-x86_64-2015-02-07.cgz/wsm/hackbench
commit:
9f80bee3fc ("mm: add cond_resched() to generic_swapfile_activate()")
b6a927bd24 ("mm: optimize copy_page_to/from_iter_iovec")
9f80bee3fce2deb4 b6a927bd2442259fc049b7ea6a
---------------- --------------------------
%stddev %change %stddev
\ | \
55711 ± 0% +7.9% 60104 ± 0% hackbench.throughput
2.42e+08 ± 1% -48.8% 1.239e+08 ± 1% hackbench.time.involuntary_context_switches
220043 ± 0% +7.9% 237518 ± 0% hackbench.time.minor_page_faults
1180 ± 0% -3.3% 1142 ± 0% hackbench.time.percent_of_cpu_this_job_got
6712 ± 0% -2.5% 6547 ± 0% hackbench.time.system_time
446.95 ± 0% -8.8% 407.41 ± 0% hackbench.time.user_time
9.334e+08 ± 1% -16.2% 7.823e+08 ± 0% hackbench.time.voluntary_context_switches
306575 ± 1% +181.5% 863126 ± 0% softirqs.SCHED
99.06 ± 0% -1.6% 97.46 ± 0% turbostat.%Busy
3480 ± 0% -1.6% 3424 ± 0% turbostat.Avg_MHz
0.80 ± 3% +196.9% 2.38 ± 0% turbostat.CPU%c1
141.50 ± 0% -20.8% 112.00 ± 1% vmstat.procs.r
1978628 ± 0% -16.7% 1648774 ± 0% vmstat.system.cs
387925 ± 0% -42.7% 222247 ± 1% vmstat.system.in
37048052 ± 0% +25.9% 46659497 ± 0% proc-vmstat.numa_hit
37048049 ± 0% +25.9% 46659496 ± 0% proc-vmstat.numa_local
37105421 ± 0% +25.9% 46721220 ± 0% proc-vmstat.pgalloc_normal
37091851 ± 0% +25.9% 46706252 ± 0% proc-vmstat.pgfree
55074252 ± 2% +219.7% 1.761e+08 ± 0% cpuidle.C1-NHM.time
15562151 ± 2% +278.9% 58960722 ± 0% cpuidle.C1-NHM.usage
600073 ± 6% +107.8% 1246685 ± 2% cpuidle.C1E-NHM.time
14646 ± 6% +159.7% 38043 ± 1% cpuidle.C1E-NHM.usage
6868049 ± 6% +25.0% 8581924 ± 0% cpuidle.C3-NHM.time
20157 ± 2% +48.7% 29970 ± 1% cpuidle.C3-NHM.usage
2874818 ± 18% +138.3% 6850072 ± 1% cpuidle.POLL.time
356651 ± 2% +285.6% 1375389 ± 0% cpuidle.POLL.usage
2.612e+12 ± 0% +6.3% 2.776e+12 ± 0% perf-stat.branch-instructions
1.51 ± 0% -3.5% 1.45 ± 0% perf-stat.branch-miss-rate
3.937e+10 ± 0% +2.5% 4.036e+10 ± 0% perf-stat.branch-misses
0.20 ± 3% +44.3% 0.29 ± 2% perf-stat.cache-miss-rate
3.071e+08 ± 2% +25.1% 3.842e+08 ± 2% perf-stat.cache-misses
1.544e+11 ± 0% -13.3% 1.339e+11 ± 0% perf-stat.cache-references
1.204e+09 ± 1% -16.3% 1.007e+09 ± 0% perf-stat.context-switches
7642375 ± 2% +299.0% 30490295 ± 1% perf-stat.cpu-migrations
2.519e+13 ± 0% -1.2% 2.488e+13 ± 0% perf-stat.cycles
4.084e+12 ± 0% +5.7% 4.318e+12 ± 0% perf-stat.dTLB-loads
5.787e+09 ± 7% +23.2% 7.129e+09 ± 14% perf-stat.dTLB-store-misses
2.509e+12 ± 0% +5.9% 2.658e+12 ± 0% perf-stat.dTLB-stores
4.113e+09 ± 0% -16.3% 3.441e+09 ± 0% perf-stat.iTLB-load-misses
2.757e+13 ± 0% +5.5% 2.908e+13 ± 0% perf-stat.instructions
6702 ± 0% +26.1% 8453 ± 0% perf-stat.instructions-per-iTLB-miss
1.09 ± 0% +6.8% 1.17 ± 0% perf-stat.ipc
856262 ± 0% +9.4% 936583 ± 0% perf-stat.minor-faults
1.073e+08 ± 2% +28.7% 1.381e+08 ± 2% perf-stat.node-loads
1.881e+08 ± 2% +22.9% 2.312e+08 ± 2% perf-stat.node-stores
856298 ± 0% +9.4% 936571 ± 0% perf-stat.page-faults
21870 ± 35% -67.1% 7188 ± 26% sched_debug.cfs_rq:/.load.min
16.51 ± 9% -26.4% 12.16 ± 20% sched_debug.cfs_rq:/.load_avg.stddev
5811549 ± 0% +22.9% 7144409 ± 0% sched_debug.cfs_rq:/.min_vruntime.avg
5882237 ± 0% +23.6% 7268954 ± 0% sched_debug.cfs_rq:/.min_vruntime.max
5746624 ± 0% +22.3% 7029885 ± 0% sched_debug.cfs_rq:/.min_vruntime.min
44003 ± 10% +76.3% 77590 ± 4% sched_debug.cfs_rq:/.min_vruntime.stddev
0.10 ± 64% +72.6% 0.17 ± 31% sched_debug.cfs_rq:/.nr_running.stddev
75.75 ± 3% -10.9% 67.47 ± 3% sched_debug.cfs_rq:/.runnable_load_avg.avg
49.02 ± 25% -38.1% 30.36 ± 14% sched_debug.cfs_rq:/.runnable_load_avg.min
120784 ± 25% +41.8% 171320 ± 12% sched_debug.cfs_rq:/.spread0.max
-14612 ±-112% +363.0% -67660 ±-32% sched_debug.cfs_rq:/.spread0.min
43936 ± 10% +76.4% 77519 ± 4% sched_debug.cfs_rq:/.spread0.stddev
22.80 ± 69% +115.8% 49.21 ± 6% sched_debug.cfs_rq:/.util_avg.stddev
8.17 ± 28% -41.0% 4.82 ± 25% sched_debug.cpu.clock.stddev
8.17 ± 28% -41.0% 4.82 ± 25% sched_debug.cpu.clock_task.stddev
103.23 ± 3% -10.9% 91.95 ± 6% sched_debug.cpu.cpu_load[1].max
102.30 ± 3% -12.1% 89.93 ± 4% sched_debug.cpu.cpu_load[2].max
100.89 ± 3% -13.5% 87.25 ± 3% sched_debug.cpu.cpu_load[3].max
12.14 ± 14% -26.7% 8.90 ± 14% sched_debug.cpu.cpu_load[3].stddev
99.59 ± 3% -14.1% 85.55 ± 2% sched_debug.cpu.cpu_load[4].max
11.81 ± 14% -29.9% 8.28 ± 12% sched_debug.cpu.cpu_load[4].stddev
18672 ± 28% -69.3% 5737 ± 14% sched_debug.cpu.load.min
0.00 ± 6% +15.7% 0.00 ± 5% sched_debug.cpu.next_balance.stddev
9.71 ± 11% -22.7% 7.51 ± 2% sched_debug.cpu.nr_running.avg
21.07 ± 1% -19.1% 17.05 ± 6% sched_debug.cpu.nr_running.max
2.80 ± 34% -78.0% 0.61 ± 32% sched_debug.cpu.nr_running.min
49639095 ± 0% -16.9% 41232200 ± 0% sched_debug.cpu.nr_switches.avg
52436741 ± 0% -15.6% 44250358 ± 0% sched_debug.cpu.nr_switches.max
47028331 ± 0% -18.1% 38524421 ± 1% sched_debug.cpu.nr_switches.min
1636426 ± 1% +17.4% 1920956 ± 4% sched_debug.cpu.nr_switches.stddev
0.19 ± 55% +1024.5% 2.09 ± 7% sched_debug.cpu.nr_uninterruptible.avg
299.00 ± 6% +476.4% 1723 ± 4% sched_debug.cpu.nr_uninterruptible.max
-279.82 ±-17% +516.9% -1726 ±-13% sched_debug.cpu.nr_uninterruptible.min
172.38 ± 11% +588.3% 1186 ± 11% sched_debug.cpu.nr_uninterruptible.stddev
0.00 ± -1% +Inf% 1.43 ± 5% perf-profile.cycles-pp.__fget_light.sys_write.entry_SYSCALL_64_fastpath
1.17 ± 7% -100.0% 0.00 ± -1% perf-profile.cycles-pp.__fget_light.sys_write.entry_SYSCALL_64_fastpath.__write_nocancel
1.88 ± 4% +76.7% 3.32 ± 2% perf-profile.cycles-pp.__mutex_lock_slowpath.mutex_lock.pipe_write.__vfs_write.vfs_write
32.48 ± 13% -100.0% 0.00 ± -1% perf-profile.cycles-pp.__read_nocancel
2.02 ± 11% -57.1% 0.87 ± 4% perf-profile.cycles-pp.__schedule.schedule.exit_to_usermode_loop.syscall_return_slowpath.entry_SYSCALL_64_fastpath
12.33 ± 3% -18.1% 10.10 ± 1% perf-profile.cycles-pp.__schedule.schedule.pipe_wait.pipe_read.__vfs_read
0.23 ±173% +563.8% 1.56 ± 2% perf-profile.cycles-pp.__switch_to
1.19 ± 16% -100.0% 0.00 ± -1% perf-profile.cycles-pp.__switch_to.__read_nocancel
4.81 ± 58% +418.8% 24.94 ± 0% perf-profile.cycles-pp.__vfs_read.vfs_read.sys_read.entry_SYSCALL_64_fastpath
20.99 ± 14% -100.0% 0.00 ± -1% perf-profile.cycles-pp.__vfs_read.vfs_read.sys_read.entry_SYSCALL_64_fastpath.__read_nocancel
7.23 ± 61% +521.3% 44.90 ± 0% perf-profile.cycles-pp.__vfs_write.vfs_write.sys_write.entry_SYSCALL_64_fastpath
36.95 ± 11% -100.0% 0.00 ± -1% perf-profile.cycles-pp.__vfs_write.vfs_write.sys_write.entry_SYSCALL_64_fastpath.__write_nocancel
20.65 ± 4% -22.9% 15.92 ± 2% perf-profile.cycles-pp.__wake_up_common.__wake_up_sync_key.pipe_write.__vfs_write.vfs_write
29.51 ± 3% -10.6% 26.39 ± 1% perf-profile.cycles-pp.__wake_up_sync_key.pipe_write.__vfs_write.vfs_write.sys_write
48.99 ± 11% -100.0% 0.00 ± -1% perf-profile.cycles-pp.__write_nocancel
4.44 ± 8% -21.6% 3.48 ± 4% perf-profile.cycles-pp._raw_spin_lock.try_to_wake_up.default_wake_function.autoremove_wake_function.__wake_up_common
7.41 ± 2% +27.9% 9.48 ± 2% perf-profile.cycles-pp._raw_spin_lock_irqsave.__wake_up_sync_key.pipe_write.__vfs_write.vfs_write
7.34 ± 4% -24.4% 5.55 ± 1% perf-profile.cycles-pp.activate_task.ttwu_do_activate.try_to_wake_up.default_wake_function.autoremove_wake_function
19.99 ± 4% -23.5% 15.29 ± 2% perf-profile.cycles-pp.autoremove_wake_function.__wake_up_common.__wake_up_sync_key.pipe_write.__vfs_write
1.02 ± 20% +179.2% 2.85 ± 4% perf-profile.cycles-pp.call_cpuidle.cpu_startup_entry.start_secondary
1.20 ± 11% -20.7% 0.96 ± 4% perf-profile.cycles-pp.check_preempt_curr.ttwu_do_wakeup.ttwu_do_activate.try_to_wake_up.default_wake_function
2.67 ± 5% +63.5% 4.36 ± 2% perf-profile.cycles-pp.copy_page_from_iter.pipe_write.__vfs_write.vfs_write.sys_write
0.96 ± 11% +113.0% 2.05 ± 3% perf-profile.cycles-pp.copy_page_from_iter_iovec.copy_page_from_iter.pipe_write.__vfs_write.vfs_write
3.69 ± 3% +38.7% 5.12 ± 1% perf-profile.cycles-pp.copy_page_to_iter.pipe_read.__vfs_read.vfs_read.sys_read
1.51 ± 6% +34.8% 2.04 ± 2% perf-profile.cycles-pp.copy_page_to_iter_iovec.copy_page_to_iter.pipe_read.__vfs_read.vfs_read
1.41 ± 10% +22.5% 1.73 ± 0% perf-profile.cycles-pp.copy_user_generic_string.copy_page_from_iter.pipe_write.__vfs_write.vfs_write
2.05 ± 2% +26.3% 2.59 ± 1% perf-profile.cycles-pp.copy_user_generic_string.copy_page_to_iter.pipe_read.__vfs_read.vfs_read
1.75 ± 21% +214.2% 5.49 ± 4% perf-profile.cycles-pp.cpu_startup_entry.start_secondary
1.02 ± 20% +180.0% 2.84 ± 4% perf-profile.cycles-pp.cpuidle_enter.call_cpuidle.cpu_startup_entry.start_secondary
1.01 ± 20% +180.6% 2.82 ± 4% perf-profile.cycles-pp.cpuidle_enter_state.cpuidle_enter.call_cpuidle.cpu_startup_entry.start_secondary
5.27 ± 2% -15.3% 4.46 ± 1% perf-profile.cycles-pp.deactivate_task.__schedule.schedule.pipe_wait.pipe_read
19.79 ± 4% -23.7% 15.11 ± 2% perf-profile.cycles-pp.default_wake_function.autoremove_wake_function.__wake_up_common.__wake_up_sync_key.pipe_write
3.25 ± 4% -10.5% 2.91 ± 2% perf-profile.cycles-pp.dequeue_entity.dequeue_task_fair.deactivate_task.__schedule.schedule
4.31 ± 4% -12.9% 3.75 ± 1% perf-profile.cycles-pp.dequeue_task_fair.deactivate_task.__schedule.schedule.pipe_wait
5.24 ± 4% -25.0% 3.93 ± 1% perf-profile.cycles-pp.enqueue_entity.enqueue_task_fair.activate_task.ttwu_do_activate.try_to_wake_up
6.45 ± 4% -25.1% 4.83 ± 1% perf-profile.cycles-pp.enqueue_task_fair.activate_task.ttwu_do_activate.try_to_wake_up.default_wake_function
15.07 ± 62% +492.9% 89.38 ± 0% perf-profile.cycles-pp.entry_SYSCALL_64_fastpath
28.73 ± 13% -100.0% 0.00 ± -1% perf-profile.cycles-pp.entry_SYSCALL_64_fastpath.__read_nocancel
46.81 ± 11% -100.0% 0.00 ± -1% perf-profile.cycles-pp.entry_SYSCALL_64_fastpath.__write_nocancel
1.76 ± 20% -100.0% 0.00 ± -1% perf-profile.cycles-pp.exit_to_usermode_loop.syscall_return_slowpath.entry_SYSCALL_64_fastpath.__write_nocancel
0.97 ± 20% +176.9% 2.69 ± 4% perf-profile.cycles-pp.intel_idle.cpuidle_enter_state.cpuidle_enter.call_cpuidle.cpu_startup_entry
0.90 ± 9% +34.1% 1.21 ± 2% perf-profile.cycles-pp.mutex_lock.pipe_read.__vfs_read.vfs_read.sys_read
3.84 ± 6% +40.2% 5.39 ± 1% perf-profile.cycles-pp.mutex_lock.pipe_write.__vfs_write.vfs_write.sys_write
1.81 ± 4% +73.1% 3.14 ± 2% perf-profile.cycles-pp.mutex_optimistic_spin.__mutex_lock_slowpath.mutex_lock.pipe_write.__vfs_write
0.12 ±173% +738.0% 1.05 ± 2% perf-profile.cycles-pp.mutex_spin_on_owner.isra.1.mutex_optimistic_spin.__mutex_lock_slowpath.mutex_lock.pipe_write
1.15 ± 9% +24.7% 1.43 ± 1% perf-profile.cycles-pp.mutex_unlock.__vfs_write.vfs_write.sys_write.entry_SYSCALL_64_fastpath
4.42 ± 8% -21.6% 3.46 ± 4% perf-profile.cycles-pp.native_queued_spin_lock_slowpath._raw_spin_lock.try_to_wake_up.default_wake_function.autoremove_wake_function
6.51 ± 2% +29.4% 8.42 ± 2% perf-profile.cycles-pp.native_queued_spin_lock_slowpath._raw_spin_lock_irqsave.__wake_up_sync_key.pipe_write.__vfs_write
1.03 ± 16% -87.1% 0.13 ±173% perf-profile.cycles-pp.pick_next_task_fair.__schedule.schedule.exit_to_usermode_loop.syscall_return_slowpath
1.91 ± 9% -13.6% 1.65 ± 1% perf-profile.cycles-pp.pick_next_task_fair.__schedule.schedule.pipe_wait.pipe_read
14.56 ± 3% -16.8% 12.12 ± 1% perf-profile.cycles-pp.pipe_wait.pipe_read.__vfs_read.vfs_read.sys_read
0.32 ±173% +1411.9% 4.76 ± 1% perf-profile.cycles-pp.rw_verify_area.vfs_read.sys_read.entry_SYSCALL_64_fastpath
4.01 ± 11% -100.0% 0.00 ± -1% perf-profile.cycles-pp.rw_verify_area.vfs_read.sys_read.entry_SYSCALL_64_fastpath.__read_nocancel
0.28 ±173% +1344.6% 4.04 ± 1% perf-profile.cycles-pp.rw_verify_area.vfs_write.sys_write.entry_SYSCALL_64_fastpath
3.43 ± 10% -100.0% 0.00 ± -1% perf-profile.cycles-pp.rw_verify_area.vfs_write.sys_write.entry_SYSCALL_64_fastpath.__write_nocancel
1.70 ± 20% -100.0% 0.00 ± -1% perf-profile.cycles-pp.schedule.exit_to_usermode_loop.syscall_return_slowpath.entry_SYSCALL_64_fastpath.__write_nocancel
12.78 ± 3% -18.0% 10.48 ± 1% perf-profile.cycles-pp.schedule.pipe_wait.pipe_read.__vfs_read.vfs_read
3.31 ± 8% -20.2% 2.64 ± 2% perf-profile.cycles-pp.select_idle_sibling.select_task_rq_fair.try_to_wake_up.default_wake_function.autoremove_wake_function
4.38 ± 3% -24.4% 3.31 ± 2% perf-profile.cycles-pp.select_task_rq_fair.try_to_wake_up.default_wake_function.autoremove_wake_function.__wake_up_common
2.69 ± 6% +11.3% 3.00 ± 2% perf-profile.cycles-pp.selinux_file_permission.security_file_permission.rw_verify_area.vfs_read.sys_read
1.76 ± 21% +214.6% 5.54 ± 4% perf-profile.cycles-pp.start_secondary
5.78 ± 61% +468.2% 32.84 ± 0% perf-profile.cycles-pp.sys_read.entry_SYSCALL_64_fastpath
27.55 ± 13% -100.0% 0.00 ± -1% perf-profile.cycles-pp.sys_read.entry_SYSCALL_64_fastpath.__read_nocancel
8.24 ± 63% +546.8% 53.26 ± 0% perf-profile.cycles-pp.sys_write.entry_SYSCALL_64_fastpath
43.99 ± 11% -100.0% 0.00 ± -1% perf-profile.cycles-pp.sys_write.entry_SYSCALL_64_fastpath.__write_nocancel
1.82 ± 20% -100.0% 0.00 ± -1% perf-profile.cycles-pp.syscall_return_slowpath.entry_SYSCALL_64_fastpath.__write_nocancel
0.82 ± 8% +18.2% 0.97 ± 4% perf-profile.cycles-pp.touch_atime.pipe_read.__vfs_read.vfs_read.sys_read
19.24 ± 4% -23.7% 14.68 ± 2% perf-profile.cycles-pp.try_to_wake_up.default_wake_function.autoremove_wake_function.__wake_up_common.__wake_up_sync_key
9.00 ± 3% -24.7% 6.78 ± 1% perf-profile.cycles-pp.ttwu_do_activate.try_to_wake_up.default_wake_function.autoremove_wake_function.__wake_up_common
5.55 ± 60% +464.9% 31.37 ± 0% perf-profile.cycles-pp.vfs_read.sys_read.entry_SYSCALL_64_fastpath
26.37 ± 13% -100.0% 0.00 ± -1% perf-profile.cycles-pp.vfs_read.sys_read.entry_SYSCALL_64_fastpath.__read_nocancel
7.94 ± 63% +542.4% 51.02 ± 0% perf-profile.cycles-pp.vfs_write.sys_write.entry_SYSCALL_64_fastpath
42.14 ± 11% -100.0% 0.00 ± -1% perf-profile.cycles-pp.vfs_write.sys_write.entry_SYSCALL_64_fastpath.__write_nocancel
0.87 ± 4% +114.7% 1.86 ± 3% perf-profile.func.cycles-pp.___might_sleep
0.97 ± 5% -43.8% 0.55 ± 3% perf-profile.func.cycles-pp.__enqueue_entity
1.13 ± 2% +100.2% 2.27 ± 5% perf-profile.func.cycles-pp.__might_sleep
0.99 ± 14% -100.0% 0.00 ± -1% perf-profile.func.cycles-pp.__read_nocancel
2.19 ± 2% -17.6% 1.80 ± 2% perf-profile.func.cycles-pp.__schedule
2.46 ± 2% -21.0% 1.94 ± 3% perf-profile.func.cycles-pp.__switch_to
1.05 ± 2% -10.7% 0.94 ± 3% perf-profile.func.cycles-pp._raw_spin_lock
1.00 ± 1% -18.3% 0.81 ± 1% perf-profile.func.cycles-pp._raw_spin_unlock_irqrestore
1.18 ± 5% -11.8% 1.04 ± 2% perf-profile.func.cycles-pp.copy_page_from_iter_iovec
1.68 ± 3% -35.0% 1.09 ± 1% perf-profile.func.cycles-pp.copy_page_to_iter_iovec
3.76 ± 1% +15.3% 4.34 ± 1% perf-profile.func.cycles-pp.copy_user_generic_string
1.00 ± 2% -23.8% 0.76 ± 3% perf-profile.func.cycles-pp.enqueue_task_fair
1.85 ± 5% -22.7% 1.43 ± 3% perf-profile.func.cycles-pp.idle_cpu
1.05 ± 19% +173.9% 2.89 ± 5% perf-profile.func.cycles-pp.intel_idle
0.73 ± 6% +35.7% 1.00 ± 4% perf-profile.func.cycles-pp.mutex_optimistic_spin
0.53 ± 4% +115.0% 1.14 ± 2% perf-profile.func.cycles-pp.mutex_spin_on_owner.isra.1
2.24 ± 3% +13.2% 2.54 ± 3% perf-profile.func.cycles-pp.pipe_write
1.00 ± 3% -19.1% 0.80 ± 1% perf-profile.func.cycles-pp.select_idle_sibling
1.44 ± 3% -23.3% 1.10 ± 2% perf-profile.func.cycles-pp.switch_mm_irqs_off
1.53 ± 2% -17.0% 1.27 ± 3% perf-profile.func.cycles-pp.update_cfs_shares
2.28 ± 2% -22.4% 1.77 ± 2% perf-profile.func.cycles-pp.update_curr
1.14 ± 4% +13.8% 1.29 ± 2% perf-profile.func.cycles-pp.vfs_write
***************************************************************************************************
wsm: 12 threads Westmere with 6G memory
=========================================================================================
compiler/cpufreq_governor/ipc/kconfig/mode/nr_threads/rootfs/tbox_group/testcase:
gcc-6/performance/pipe/x86_64-rhel/threads/50%/debian-x86_64-2015-02-07.cgz/wsm/hackbench
commit:
9f80bee3fc ("mm: add cond_resched() to generic_swapfile_activate()")
b6a927bd24 ("mm: optimize copy_page_to/from_iter_iovec")
9f80bee3fce2deb4 b6a927bd2442259fc049b7ea6a
---------------- --------------------------
%stddev %change %stddev
\ | \
49540 ± 0% +14.7% 56838 ± 0% hackbench.throughput
3.113e+08 ± 1% -54.4% 1.42e+08 ± 0% hackbench.time.involuntary_context_switches
18163 ± 0% +13.7% 20659 ± 0% hackbench.time.minor_page_faults
1184 ± 0% -3.7% 1141 ± 0% hackbench.time.percent_of_cpu_this_job_got
7096 ± 0% -5.2% 6727 ± 0% hackbench.time.system_time
313.25 ± 0% +9.8% 343.94 ± 0% hackbench.time.user_time
1.089e+09 ± 0% -24.9% 8.184e+08 ± 0% hackbench.time.voluntary_context_switches
210952 ± 1% +289.8% 822334 ± 0% softirqs.SCHED
59.25 ± 62% +158.6% 153.25 ± 12% slabinfo.PING.active_objs
59.25 ± 62% +158.6% 153.25 ± 12% slabinfo.PING.num_objs
99.21 ± 0% -1.8% 97.42 ± 0% turbostat.%Busy
3485 ± 0% -1.8% 3422 ± 0% turbostat.Avg_MHz
0.55 ± 2% +321.2% 2.34 ± 1% turbostat.CPU%c1
149.00 ± 0% -24.3% 112.75 ± 0% vmstat.procs.r
2265412 ± 0% -24.5% 1710775 ± 0% vmstat.system.cs
461497 ± 0% -45.0% 253630 ± 0% vmstat.system.in
29058831 ± 0% +46.0% 42439790 ± 0% proc-vmstat.numa_hit
29058829 ± 0% +46.0% 42439789 ± 0% proc-vmstat.numa_local
29106634 ± 0% +46.0% 42492773 ± 0% proc-vmstat.pgalloc_normal
665961 ± 0% +10.6% 736659 ± 0% proc-vmstat.pgfault
29098943 ± 0% +46.0% 42486010 ± 0% proc-vmstat.pgfree
38695009 ± 2% +352.1% 1.749e+08 ± 1% cpuidle.C1-NHM.time
10610968 ± 2% +448.1% 58159044 ± 1% cpuidle.C1-NHM.usage
469713 ± 3% +189.8% 1361024 ± 1% cpuidle.C1E-NHM.time
12352 ± 1% +253.4% 43655 ± 2% cpuidle.C1E-NHM.usage
10750094 ± 2% +25.6% 13505762 ± 5% cpuidle.C3-NHM.time
22293 ± 2% +63.4% 36432 ± 3% cpuidle.C3-NHM.usage
1585052 ± 7% +341.8% 7002525 ± 2% cpuidle.POLL.time
243417 ± 2% +457.1% 1356130 ± 1% cpuidle.POLL.usage
2.747e+12 ± 0% +4.6% 2.874e+12 ± 0% perf-stat.branch-instructions
1.58 ± 0% -8.5% 1.45 ± 0% perf-stat.branch-miss-rate
4.341e+10 ± 0% -4.2% 4.157e+10 ± 0% perf-stat.branch-misses
0.11 ± 7% +60.2% 0.17 ± 4% perf-stat.cache-miss-rate
1.757e+08 ± 8% +30.5% 2.293e+08 ± 4% perf-stat.cache-misses
1.658e+11 ± 0% -18.5% 1.351e+11 ± 0% perf-stat.cache-references
1.422e+09 ± 0% -25.2% 1.063e+09 ± 0% perf-stat.context-switches
10300111 ± 4% +205.2% 31439454 ± 1% perf-stat.cpu-migrations
2.602e+13 ± 0% -2.7% 2.531e+13 ± 0% perf-stat.cycles
4.208e+12 ± 0% +4.4% 4.394e+12 ± 0% perf-stat.dTLB-loads
2.586e+12 ± 0% +5.0% 2.715e+12 ± 0% perf-stat.dTLB-stores
4.483e+08 ± 10% +46.5% 6.569e+08 ± 4% perf-stat.iTLB-load-misses
2.849e+13 ± 0% +4.1% 2.967e+13 ± 0% perf-stat.instructions
64217 ± 10% -29.5% 45254 ± 4% perf-stat.instructions-per-iTLB-miss
1.09 ± 0% +7.1% 1.17 ± 0% perf-stat.ipc
657253 ± 0% +10.9% 728898 ± 0% perf-stat.minor-faults
47496119 ± 6% +24.3% 59022005 ± 3% perf-stat.node-loads
1.193e+08 ± 8% +32.2% 1.577e+08 ± 4% perf-stat.node-stores
657291 ± 0% +10.9% 728919 ± 0% perf-stat.page-faults
26174 ±122% +556.0% 171694 ± 26% sched_debug.cfs_rq:/.MIN_vruntime.avg
314072 ±122% +525.9% 1965737 ± 23% sched_debug.cfs_rq:/.MIN_vruntime.max
86805 ±122% +534.9% 551154 ± 24% sched_debug.cfs_rq:/.MIN_vruntime.stddev
28888 ± 21% -65.7% 9904 ± 24% sched_debug.cfs_rq:/.load.min
38526 ± 34% +89.4% 72969 ± 25% sched_debug.cfs_rq:/.load.stddev
26174 ±122% +556.0% 171694 ± 26% sched_debug.cfs_rq:/.max_vruntime.avg
314072 ±122% +525.9% 1965738 ± 23% sched_debug.cfs_rq:/.max_vruntime.max
86805 ±122% +534.9% 551154 ± 24% sched_debug.cfs_rq:/.max_vruntime.stddev
5291734 ± 0% +34.7% 7129082 ± 0% sched_debug.cfs_rq:/.min_vruntime.avg
5358421 ± 0% +35.3% 7249961 ± 0% sched_debug.cfs_rq:/.min_vruntime.max
5230492 ± 0% +34.1% 7016038 ± 0% sched_debug.cfs_rq:/.min_vruntime.min
38418 ± 9% +91.0% 73394 ± 5% sched_debug.cfs_rq:/.min_vruntime.stddev
0.80 ± 4% -34.3% 0.52 ± 28% sched_debug.cfs_rq:/.nr_running.min
0.07 ± 15% +136.0% 0.17 ± 17% sched_debug.cfs_rq:/.nr_running.stddev
74.38 ± 1% -9.4% 67.39 ± 1% sched_debug.cfs_rq:/.runnable_load_avg.avg
51.09 ± 6% -35.6% 32.89 ± 13% sched_debug.cfs_rq:/.runnable_load_avg.min
13.27 ± 6% +23.6% 16.40 ± 5% sched_debug.cfs_rq:/.runnable_load_avg.stddev
-46479 ±-42% +145.9% -114298 ±-16% sched_debug.cfs_rq:/.spread0.min
38342 ± 9% +91.3% 73355 ± 5% sched_debug.cfs_rq:/.spread0.stddev
5.40 ± 26% -34.1% 3.56 ± 10% sched_debug.cpu.clock.stddev
5.40 ± 26% -34.1% 3.56 ± 10% sched_debug.cpu.clock_task.stddev
71.80 ± 1% -17.6% 59.15 ± 4% sched_debug.cpu.cpu_load[0].avg
39.70 ± 25% -73.9% 10.36 ± 70% sched_debug.cpu.cpu_load[0].min
16.52 ± 11% +51.9% 25.09 ± 11% sched_debug.cpu.cpu_load[0].stddev
97.84 ± 2% -10.0% 88.07 ± 1% sched_debug.cpu.cpu_load[3].max
97.09 ± 1% -10.5% 86.89 ± 1% sched_debug.cpu.cpu_load[4].max
11.30 ± 8% -18.9% 9.16 ± 5% sched_debug.cpu.cpu_load[4].stddev
6920 ± 0% +11.0% 7683 ± 0% sched_debug.cpu.curr->pid.max
5165 ± 16% -37.4% 3233 ± 7% sched_debug.cpu.curr->pid.min
563.00 ± 47% +131.1% 1301 ± 3% sched_debug.cpu.curr->pid.stddev
26205 ± 35% -74.9% 6574 ± 16% sched_debug.cpu.load.min
10.76 ± 4% -28.9% 7.65 ± 3% sched_debug.cpu.nr_running.avg
19.95 ± 3% -11.6% 17.64 ± 7% sched_debug.cpu.nr_running.max
4.09 ± 28% -83.9% 0.66 ± 11% sched_debug.cpu.nr_running.min
4.72 ± 10% +14.5% 5.40 ± 5% sched_debug.cpu.nr_running.stddev
56812993 ± 0% -24.8% 42738623 ± 0% sched_debug.cpu.nr_switches.avg
59731856 ± 1% -24.1% 45359351 ± 0% sched_debug.cpu.nr_switches.max
53466489 ± 1% -25.1% 40041204 ± 1% sched_debug.cpu.nr_switches.min
0.12 ± 14% +1428.1% 1.85 ± 27% sched_debug.cpu.nr_uninterruptible.avg
206.09 ± 24% +758.7% 1769 ± 18% sched_debug.cpu.nr_uninterruptible.max
-180.75 ±-16% +879.8% -1770 ±-14% sched_debug.cpu.nr_uninterruptible.min
118.78 ± 23% +842.4% 1119 ± 11% sched_debug.cpu.nr_uninterruptible.stddev
0.00 ± 36% +87.6% 0.00 ± 6% sched_debug.rt_rq:/.rt_time.min
1.05 ± 10% -75.8% 0.26 ±100% perf-profile.cycles-pp.__enqueue_entity.enqueue_entity.enqueue_task_fair.activate_task.ttwu_do_activate
1.49 ± 24% +93.6% 2.88 ± 2% perf-profile.cycles-pp.__mutex_lock_slowpath.mutex_lock.pipe_write.__vfs_write.vfs_write
2.47 ± 12% -59.2% 1.01 ± 6% perf-profile.cycles-pp.__schedule.schedule.exit_to_usermode_loop.syscall_return_slowpath.entry_SYSCALL_64_fastpath
13.64 ± 4% -27.0% 9.96 ± 2% perf-profile.cycles-pp.__schedule.schedule.pipe_wait.pipe_read.__vfs_read
2.38 ± 4% -25.0% 1.79 ± 5% perf-profile.cycles-pp.__switch_to
26.07 ± 1% -8.9% 23.73 ± 1% perf-profile.cycles-pp.__vfs_read.vfs_read.sys_read.entry_SYSCALL_64_fastpath
24.20 ± 6% -31.2% 16.65 ± 2% perf-profile.cycles-pp.__wake_up_common.__wake_up_sync_key.pipe_write.__vfs_write.vfs_write
32.62 ± 2% -18.3% 26.63 ± 1% perf-profile.cycles-pp.__wake_up_sync_key.pipe_write.__vfs_write.vfs_write.sys_write
2.29 ± 7% -41.3% 1.35 ± 5% perf-profile.cycles-pp._raw_spin_lock.__schedule.schedule.pipe_wait.pipe_read
5.91 ± 8% -38.2% 3.65 ± 2% perf-profile.cycles-pp._raw_spin_lock.try_to_wake_up.default_wake_function.autoremove_wake_function.__wake_up_common
7.08 ± 15% +27.1% 8.99 ± 2% perf-profile.cycles-pp._raw_spin_lock_irqsave.__wake_up_sync_key.pipe_write.__vfs_write.vfs_write
8.45 ± 5% -31.9% 5.75 ± 1% perf-profile.cycles-pp.activate_task.ttwu_do_activate.try_to_wake_up.default_wake_function.autoremove_wake_function
23.54 ± 6% -31.8% 16.05 ± 2% perf-profile.cycles-pp.autoremove_wake_function.__wake_up_common.__wake_up_sync_key.pipe_write.__vfs_write
0.29 ±100% +835.6% 2.76 ± 3% perf-profile.cycles-pp.call_cpuidle.cpu_startup_entry.start_secondary
1.61 ± 8% -34.9% 1.05 ± 4% perf-profile.cycles-pp.check_preempt_curr.ttwu_do_wakeup.ttwu_do_activate.try_to_wake_up.default_wake_function
1.11 ± 8% -41.3% 0.65 ± 2% perf-profile.cycles-pp.check_preempt_wakeup.check_preempt_curr.ttwu_do_wakeup.ttwu_do_activate.try_to_wake_up
2.47 ± 2% +68.1% 4.15 ± 2% perf-profile.cycles-pp.copy_page_from_iter.pipe_write.__vfs_write.vfs_write.sys_write
0.97 ± 3% +104.4% 1.98 ± 3% perf-profile.cycles-pp.copy_page_from_iter_iovec.copy_page_from_iter.pipe_write.__vfs_write.vfs_write
3.43 ± 1% +38.9% 4.76 ± 0% perf-profile.cycles-pp.copy_page_to_iter.pipe_read.__vfs_read.vfs_read.sys_read
1.17 ± 3% +68.7% 1.96 ± 1% perf-profile.cycles-pp.copy_page_to_iter_iovec.copy_page_to_iter.pipe_read.__vfs_read.vfs_read
1.40 ± 1% +14.8% 1.61 ± 3% perf-profile.cycles-pp.copy_user_generic_string.copy_page_from_iter.pipe_write.__vfs_write.vfs_write
0.69 ± 64% +674.6% 5.34 ± 3% perf-profile.cycles-pp.cpu_startup_entry.start_secondary
0.29 ±100% +829.7% 2.74 ± 3% perf-profile.cycles-pp.cpuidle_enter.call_cpuidle.cpu_startup_entry.start_secondary
0.29 ±100% +839.7% 2.73 ± 3% perf-profile.cycles-pp.cpuidle_enter_state.cpuidle_enter.call_cpuidle.cpu_startup_entry.start_secondary
5.88 ± 5% -21.9% 4.59 ± 3% perf-profile.cycles-pp.deactivate_task.__schedule.schedule.pipe_wait.pipe_read
23.32 ± 6% -31.9% 15.88 ± 2% perf-profile.cycles-pp.default_wake_function.autoremove_wake_function.__wake_up_common.__wake_up_sync_key.pipe_write
3.73 ± 4% -20.3% 2.97 ± 2% perf-profile.cycles-pp.dequeue_entity.dequeue_task_fair.deactivate_task.__schedule.schedule
4.91 ± 4% -21.0% 3.88 ± 2% perf-profile.cycles-pp.dequeue_task_fair.deactivate_task.__schedule.schedule.pipe_wait
5.97 ± 5% -32.8% 4.01 ± 1% perf-profile.cycles-pp.enqueue_entity.enqueue_task_fair.activate_task.ttwu_do_activate.try_to_wake_up
7.43 ± 5% -32.7% 5.00 ± 1% perf-profile.cycles-pp.enqueue_task_fair.activate_task.ttwu_do_activate.try_to_wake_up.default_wake_function
2.83 ± 12% -59.0% 1.16 ± 6% perf-profile.cycles-pp.exit_to_usermode_loop.syscall_return_slowpath.entry_SYSCALL_64_fastpath
1.02 ± 7% +16.4% 1.19 ± 4% perf-profile.cycles-pp.file_has_perm.selinux_file_permission.security_file_permission.rw_verify_area.vfs_read
1.01 ± 9% +18.8% 1.20 ± 2% perf-profile.cycles-pp.fput.entry_SYSCALL_64_fastpath
1.87 ± 9% -33.6% 1.24 ± 4% perf-profile.cycles-pp.idle_cpu.select_idle_sibling.select_task_rq_fair.try_to_wake_up.default_wake_function
0.28 ±100% +832.1% 2.61 ± 3% perf-profile.cycles-pp.intel_idle.cpuidle_enter_state.cpuidle_enter.call_cpuidle.cpu_startup_entry
0.86 ± 9% +27.0% 1.09 ± 2% perf-profile.cycles-pp.mutex_lock.pipe_read.__vfs_read.vfs_read.sys_read
3.39 ± 13% +43.2% 4.85 ± 2% perf-profile.cycles-pp.mutex_lock.pipe_write.__vfs_write.vfs_write.sys_write
1.44 ± 24% +88.9% 2.72 ± 2% perf-profile.cycles-pp.mutex_optimistic_spin.__mutex_lock_slowpath.mutex_lock.pipe_write.__vfs_write
2.29 ± 7% -41.3% 1.34 ± 5% perf-profile.cycles-pp.native_queued_spin_lock_slowpath._raw_spin_lock.__schedule.schedule.pipe_wait
5.89 ± 8% -38.2% 3.64 ± 2% perf-profile.cycles-pp.native_queued_spin_lock_slowpath._raw_spin_lock.try_to_wake_up.default_wake_function.autoremove_wake_function
6.30 ± 16% +27.3% 8.02 ± 2% perf-profile.cycles-pp.native_queued_spin_lock_slowpath._raw_spin_lock_irqsave.__wake_up_sync_key.pipe_write.__vfs_write
1.44 ± 12% -57.4% 0.61 ± 4% perf-profile.cycles-pp.pick_next_task_fair.__schedule.schedule.exit_to_usermode_loop.syscall_return_slowpath
2.53 ± 4% -33.8% 1.67 ± 2% perf-profile.cycles-pp.pick_next_task_fair.__schedule.schedule.pipe_wait.pipe_read
24.25 ± 2% -10.7% 21.64 ± 0% perf-profile.cycles-pp.pipe_read.__vfs_read.vfs_read.sys_read.entry_SYSCALL_64_fastpath
16.17 ± 4% -26.0% 11.96 ± 2% perf-profile.cycles-pp.pipe_wait.pipe_read.__vfs_read.vfs_read.sys_read
4.06 ± 2% +13.3% 4.60 ± 2% perf-profile.cycles-pp.rw_verify_area.vfs_read.sys_read.entry_SYSCALL_64_fastpath
2.71 ± 12% -59.4% 1.10 ± 5% perf-profile.cycles-pp.schedule.exit_to_usermode_loop.syscall_return_slowpath.entry_SYSCALL_64_fastpath
14.16 ± 4% -27.0% 10.33 ± 2% perf-profile.cycles-pp.schedule.pipe_wait.pipe_read.__vfs_read.vfs_read
3.63 ± 2% +14.4% 4.15 ± 2% perf-profile.cycles-pp.security_file_permission.rw_verify_area.vfs_read.sys_read.entry_SYSCALL_64_fastpath
3.92 ± 7% -30.1% 2.74 ± 4% perf-profile.cycles-pp.select_idle_sibling.select_task_rq_fair.try_to_wake_up.default_wake_function.autoremove_wake_function
4.90 ± 8% -28.8% 3.49 ± 4% perf-profile.cycles-pp.select_task_rq_fair.try_to_wake_up.default_wake_function.autoremove_wake_function.__wake_up_common
2.55 ± 4% +15.9% 2.96 ± 2% perf-profile.cycles-pp.selinux_file_permission.security_file_permission.rw_verify_area.vfs_read.sys_read
1.20 ± 3% -34.1% 0.79 ± 2% perf-profile.cycles-pp.set_next_entity.pick_next_task_fair.__schedule.schedule.pipe_wait
0.70 ± 64% +678.1% 5.41 ± 3% perf-profile.cycles-pp.start_secondary
2.93 ± 12% -58.6% 1.21 ± 5% perf-profile.cycles-pp.syscall_return_slowpath.entry_SYSCALL_64_fastpath
22.67 ± 6% -32.1% 15.41 ± 2% perf-profile.cycles-pp.try_to_wake_up.default_wake_function.autoremove_wake_function.__wake_up_common.__wake_up_sync_key
10.34 ± 5% -31.2% 7.11 ± 2% perf-profile.cycles-pp.ttwu_do_activate.try_to_wake_up.default_wake_function.autoremove_wake_function.__wake_up_common
1.78 ± 7% -28.6% 1.27 ± 5% perf-profile.cycles-pp.ttwu_do_wakeup.ttwu_do_activate.try_to_wake_up.default_wake_function.autoremove_wake_function
1.44 ± 5% -17.0% 1.20 ± 2% perf-profile.cycles-pp.update_cfs_shares.dequeue_entity.dequeue_task_fair.deactivate_task.__schedule
1.92 ± 5% -29.2% 1.36 ± 4% perf-profile.cycles-pp.update_cfs_shares.enqueue_entity.enqueue_task_fair.activate_task.ttwu_do_activate
0.80 ± 5% +126.1% 1.80 ± 3% perf-profile.func.cycles-pp.___might_sleep
1.19 ± 11% -50.7% 0.59 ± 2% perf-profile.func.cycles-pp.__enqueue_entity
1.02 ± 2% +108.6% 2.12 ± 4% perf-profile.func.cycles-pp.__might_sleep
2.46 ± 3% -22.3% 1.92 ± 3% perf-profile.func.cycles-pp.__schedule
2.88 ± 4% -24.8% 2.16 ± 4% perf-profile.func.cycles-pp.__switch_to
1.15 ± 1% -14.6% 0.98 ± 2% perf-profile.func.cycles-pp._raw_spin_lock
1.04 ± 6% -17.1% 0.86 ± 4% perf-profile.func.cycles-pp._raw_spin_unlock_irqrestore
1.21 ± 3% -11.8% 1.06 ± 1% perf-profile.func.cycles-pp.copy_page_to_iter_iovec
3.54 ± 1% +11.1% 3.94 ± 0% perf-profile.func.cycles-pp.copy_user_generic_string
0.91 ± 7% -20.8% 0.72 ± 5% perf-profile.func.cycles-pp.dequeue_task_fair
1.21 ± 5% -32.2% 0.82 ± 3% perf-profile.func.cycles-pp.enqueue_task_fair
1.67 ± 1% +7.5% 1.79 ± 4% perf-profile.func.cycles-pp.entry_SYSCALL_64_fastpath
1.02 ± 9% +18.2% 1.20 ± 2% perf-profile.func.cycles-pp.fput
2.03 ± 9% -31.4% 1.39 ± 4% perf-profile.func.cycles-pp.idle_cpu
0.41 ± 52% +589.0% 2.81 ± 4% perf-profile.func.cycles-pp.intel_idle
0.42 ± 20% +147.0% 1.04 ± 4% perf-profile.func.cycles-pp.mutex_spin_on_owner.isra.1
15.32 ± 3% -10.8% 13.66 ± 2% perf-profile.func.cycles-pp.native_queued_spin_lock_slowpath
0.95 ± 5% -29.5% 0.67 ± 2% perf-profile.func.cycles-pp.pick_next_task_fair
2.24 ± 8% +18.0% 2.64 ± 5% perf-profile.func.cycles-pp.pipe_write
1.21 ± 3% -24.6% 0.91 ± 4% perf-profile.func.cycles-pp.select_idle_sibling
1.97 ± 0% +16.0% 2.29 ± 5% perf-profile.func.cycles-pp.selinux_file_permission
1.04 ± 5% -20.9% 0.82 ± 0% perf-profile.func.cycles-pp.update_cfs_rq_load_avg
1.62 ± 2% -18.8% 1.32 ± 4% perf-profile.func.cycles-pp.update_cfs_shares
2.51 ± 3% -31.4% 1.72 ± 3% perf-profile.func.cycles-pp.update_curr
1.10 ± 6% +11.6% 1.23 ± 4% perf-profile.func.cycles-pp.vfs_write
Thanks,
Xiaolong
View attachment "config-4.7.0-01848-gb6a927b" of type "text/plain" (151070 bytes)
View attachment "job.yaml" of type "text/plain" (3591 bytes)
View attachment "reproduce" of type "text/plain" (1923 bytes)
Powered by blists - more mailing lists