FYI, we noticed the below changes on
https://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git master
commit 8fb47ff100af48034eb11449a81f42af24213987 ("cpufreq: governor: Replace timers
with utilization update callbacks")
=========================================================================================
compiler/kconfig/nr_task/rootfs/tbox_group/test/testcase:
gcc-4.9/x86_64-rhel/1/debian-x86_64-2015-02-07.cgz/nhm-white/shell1/unixbench
commit:
402c43ed2d7433d7a0a4cf2767bedc239f680155
8fb47ff100af48034eb11449a81f42af24213987
402c43ed2d7433d7 8fb47ff100af48034eb11449a8
---------------- --------------------------
%stddev %change %stddev
\ | \
1555 ± 0% -3.2% 1506 ± 0% unixbench.score
286697 ± 0% -7.0% 266678 ± 0% unixbench.time.involuntary_context_switches
17730295 ± 0% -3.1% 17172749 ± 0% unixbench.time.minor_page_faults
616646 ± 0% -3.1% 597250 ± 0% unixbench.time.voluntary_context_switches
288.75 ± 0% -3.0% 280.00 ± 0% turbostat.Avg_MHz
1956505 ± 12% -39.9% 1176054 ± 8% cpuidle.C1E-NHM.time
7033 ± 4% -18.0% 5765 ± 1% cpuidle.C1E-NHM.usage
14204 ± 0% -2.8% 13801 ± 0% vmstat.system.cs
1628 ± 0% +43.3% 2332 ± 0% vmstat.system.in
2855 ± 1% -11.1% 2538 ± 1% sched_debug.cfs_rq:/.exec_clock.stddev
43.25 ± 19% -39.3% 26.25 ± 18% sched_debug.cfs_rq:/.load_avg.2
20.00 ± 5% +135.0% 47.00 ± 68% sched_debug.cfs_rq:/.load_avg.7
15987 ± 3% -14.5% 13672 ± 1% sched_debug.cfs_rq:/.min_vruntime.stddev
0.33 ± 14% +38.1% 0.45 ± 5% sched_debug.cfs_rq:/.nr_running.avg
2.00 ± 79% +450.0% 11.00 ± 19% sched_debug.cfs_rq:/.runnable_load_avg.2
5.55 ± 19% +34.1% 7.44 ± 10% sched_debug.cfs_rq:/.runnable_load_avg.avg
-33720 ± -1% -17.3% -27894 ± -4% sched_debug.cfs_rq:/.spread0.4
-33413 ± -3% -13.7% -28847 ± -3% sched_debug.cfs_rq:/.spread0.5
-33146 ± -3% -18.0% -27185 ± -1% sched_debug.cfs_rq:/.spread0.6
-33527 ± -1% -12.8% -29230 ± -2% sched_debug.cfs_rq:/.spread0.7
-17639 ± -2% -16.2% -14779 ± -4% sched_debug.cfs_rq:/.spread0.avg
-34541 ± -1% -14.2% -29641 ± -2% sched_debug.cfs_rq:/.spread0.min
15990 ± 3% -14.5% 13674 ± 1% sched_debug.cfs_rq:/.spread0.stddev
708107 ± 10% +22.0% 863894 ± 5% sched_debug.cpu.avg_idle.0
13049 ± 44% -37.5% 8162 ± 38% sched_debug.cpu.curr->pid.6
22214 ± 0% -13.1% 19305 ± 0% sched_debug.cpu.curr->pid.max
9922 ± 4% -7.6% 9164 ± 1% sched_debug.cpu.curr->pid.stddev
29.00 ± 12% +162.1% 76.00 ± 36% sched_debug.cpu.load.2
124.75 ± 61% -63.1% 46.00 ± 42% sched_debug.cpu.load.3
15.50 ± 92% +343.5% 68.75 ± 59% sched_debug.cpu.load.4
54.67 ± 33% +43.6% 78.50 ± 3% sched_debug.cpu.load.7
0.35 ± 11% +28.9% 0.45 ± 11% sched_debug.cpu.nr_running.avg
8.52 ± 15% +29.7% 11.06 ± 10% sched_debug.cpu.nr_uninterruptible.stddev
25833 ± 0% +9.4% 28264 ± 2% sched_debug.cpu.ttwu_local.3
18030 ± 2% +10.7% 19951 ± 1% sched_debug.cpu.ttwu_local.5
17450 ± 1% +9.8% 19158 ± 1% sched_debug.cpu.ttwu_local.6
nhm-white: Nehalem
Memory: 6G
unixbench.score
1560 ++-------------------------------------------------------------------+
| .*. .*. .*. .*.*.*.. .*. .*. .*.*.*. .*.*.*.*.* |
1550 *+*.*.*.* * *.*.*.* * * *.* * * |
1540 ++ |
| |
1530 ++ |
| |
1520 ++ |
| |
1510 ++ O O O O O O O |
1500 ++ O O O
O O O O O O O O O O O O O O O |
1490 ++ O O O O O O O O O |
| O |
1480 ++-------------------------------------------------------------------+
unixbench.time.minor_page_faults
1.78e+07 ++---------------------------------------------------------------+
| .**.*.* |
1.77e+07 *+*.*.*.*.**.*.*.*.*.*.*.*.*.*.*.**.*.*.*.*.*.*.*.*.* |
1.76e+07 ++ |
| |
1.75e+07 ++ |
1.74e+07 ++ |
| |
1.73e+07 ++ |
1.72e+07 ++ |
| O O O OO O O O O O
1.71e+07 O+O O O O O O |
1.7e+07 ++ O O OO O O O O O O |
| O O O O O OO O |
1.69e+07 ++---------------------------------------------------------------+
unixbench.time.voluntary_context_switches
620000 ++-----------------------------------------------------------------+
| .*. |
615000 *+*.*.*.*.*.*.*. .*.*.*.*.*.*.**. .*.*.*.*.*.*.*.*.* *.* |
| *.*.* * |
| |
610000 ++ |
| |
605000 ++ |
| |
600000 ++ |
| O O O |
| O O O O O O O
595000 ++ O O O O O O |
O O O O O O O O O O O |
590000 ++----------------O-------O-O-O-O-OO-O-----------------------------+
unixbench.time.involuntary_context_switches
290000 ++-----------------------------------------------------------------+
*.*.*.*. .*.*.*.*.*. .*. .*. .*.*.*.* |
285000 ++ * *.*.*.*.*.* **.*.*.*.*.*.* *.* |
| |
280000 ++ |
| |
275000 ++ |
| |
270000 ++ |
| O O O O O O O O O
265000 ++ O |
O O O O |
260000 ++ O O O |
| O O O O O O O O O O OO O O O O O O |
255000 ++-----------------------------------------------------------------+
vmstat.system.in
2400 ++-------------------------------------------------------------------+
| O O O O O O O O O O O O O O O
2200 O+O O O O O O O O O O O O O O O O O O O |
| |
2000 ++ |
| |
1800 ++ |
| |
1600 *+*.*.*.*.*.*.*.*.*.*.*.* *.* *.*.* *.* *.*.*.* |
| : : + : : : + : |
1400 ++ : : *..* : : : *.* : |
| * : : * : : |
1200 ++ :: :: |
| * * |
1000 ++-------------------------------------------------------------------+
[*] bisect-good sample
[O] bisect-bad sample
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
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.
Thanks,
Ying Huang