FYI, we noticed the below changes on
https://git.kernel.org/pub/scm/linux/kernel/git/maz/arm-platforms.git irq/wire-msi-bridge
commit 460ec9d78057bf462f5b5cd0a140e32252d81405 ("genirq/msi: Make the .prepare callback reusable")
=========================================================================================
compiler/kconfig/rootfs/tbox_group/test/testcase:
gcc-4.9/x86_64-rhel/debian-x86_64-2015-02-07.cgz/nhm-white2/context1/unixbench
commit:
c175e94e50249623f7e0c9dc087798afd2e541ae
460ec9d78057bf462f5b5cd0a140e32252d81405
c175e94e50249623 460ec9d78057bf462f5b5cd0a1
---------------- --------------------------
%stddev %change %stddev
\ | \
1421 ? 0% +8.5% 1543 ? 0% unixbench.score
64569793 ? 0% +7.0% 69093806 ? 0% unixbench.time.voluntary_context_switches
18591611 ?114% -62.0% 7060772 ?104% latency_stats.avg.nfs_wait_on_request.nfs_updatepage.nfs_write_end.generic_perform_write.__generic_file_write_iter.generic_file_write_iter.nfs_file_write.__vfs_write.vfs_write.SyS_write.entry_SYSCALL_64_fastpath
25939 ? 4% -7.2% 24059 ? 5% softirqs.SCHED
10.80 ? 1% +11.5% 12.04 ? 1% time.user_time
3971 ? 14% -18.9% 3220 ? 5% vmstat.system.in
60.25 ? 20% +28.6% 77.50 ? 15% sched_debug.cfs_rq[0]:/.load_avg
34.75 ? 6% +50.4% 52.25 ? 23% sched_debug.cfs_rq[0]:/.runnable_load_avg
60.25 ? 20% +28.6% 77.50 ? 15% sched_debug.cfs_rq[0]:/.tg_load_avg_contrib
440.00 ? 15% +32.9% 584.75 ? 11% sched_debug.cfs_rq[0]:/.util_avg
32425 ? 4% -15.2% 27496 ? 20% sched_debug.cfs_rq[5]:/.exec_clock
53.00 ? 16% +72.6% 91.50 ? 21% sched_debug.cfs_rq[5]:/.load_avg
53.00 ? 16% +73.1% 91.75 ? 21% sched_debug.cfs_rq[5]:/.tg_load_avg_contrib
80.25 ? 24% +28.0% 102.75 ? 14% sched_debug.cfs_rq[6]:/.load_avg
37.00 ? 8% +39.9% 51.75 ? 29% sched_debug.cfs_rq[6]:/.runnable_load_avg
80.25 ? 24% +28.7% 103.25 ? 15% sched_debug.cfs_rq[6]:/.tg_load_avg_contrib
36.00 ? 3% +27.1% 45.75 ? 14% sched_debug.cpu#0.cpu_load[1]
36.25 ? 4% +22.8% 44.50 ? 8% sched_debug.cpu#0.cpu_load[2]
35.75 ? 7% +24.5% 44.50 ? 9% sched_debug.cpu#0.cpu_load[3]
35.50 ? 9% +28.9% 45.75 ? 12% sched_debug.cpu#0.cpu_load[4]
2030062 ? 4% -16.0% 1705635 ? 5% sched_debug.cpu#1.sched_goidle
35.00 ? 11% +14.3% 40.00 ? 5% sched_debug.cpu#3.cpu_load[0]
32.75 ? 23% +25.2% 41.00 ? 4% sched_debug.cpu#3.cpu_load[2]
32.50 ? 24% +29.2% 42.00 ? 6% sched_debug.cpu#3.cpu_load[3]
32.00 ? 26% +34.4% 43.00 ? 10% sched_debug.cpu#3.cpu_load[4]
35.50 ? 13% +18.3% 42.00 ? 5% sched_debug.cpu#4.cpu_load[4]
38729 ? 6% -17.3% 32035 ? 20% sched_debug.cpu#5.nr_load_updates
1357857 ? 24% -46.7% 724368 ? 35% sched_debug.cpu#5.sched_goidle
nhm-white2: Nehalem
Memory: 4G
unixbench.score
1700 O+O--OO--------------------------------------------------------------+
| O |
1650 ++ |
| |
1600 ++ OO |
| OO O O OO O OO O |
1550 ++ O O O O O OO O |
| OO |
1500 ++ |
| |
1450 ++ |
| *.**.*.**.**. *.**.**.*
1400 *+**.**.**.**.**.* **.**.**.**.**.*.**.**.**.**.* |
| |
1350 ++-------------------------------------------------------------------+
unixbench.time.voluntary_context_switches
7.6e+07 O+--O-------------------------------------------------------------+
|O O O |
7.4e+07 ++ |
| |
7.2e+07 ++ O |
| O O O |
7e+07 ++ O O O OO OOO O OO O O |
| O O O OO |
6.8e+07 ++ |
| |
6.6e+07 ++ .* |
| **.*** *.* * .**.*|
6.4e+07 **. *.* .**.* .* + *.* .**. *.**. **. *.* .**.* + * *
| * * * * * * * * * * |
6.2e+07 ++----------------------------------------------------------------+
[*] 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