FYI, we noticed the below changes on git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git master commit 04b2fa9f8f36ec6fb6fd1c9dc9df6fff0cd27323 ("fs: split generic and aio kiocb") testbox/testcase/testparams: ivb42/unixbench/performance-pipe 599bd19bdc4c6b20 04b2fa9f8f36ec6fb6fd1c9dc9 ---------------- -------------------------- %stddev %change %stddev \ | \ 1733 ± 0% +2.2% 1770 ± 0% unixbench.score 21746 ± 5% -9.4% 19710 ± 5% numa-meminfo.node0.SReclaimable 20531 ± 6% +9.5% 22492 ± 5% numa-meminfo.node1.SReclaimable 5436 ± 5% -9.4% 4927 ± 5% numa-vmstat.node0.nr_slab_reclaimable 5132 ± 6% +9.6% 5622 ± 5% numa-vmstat.node1.nr_slab_reclaimable 22344 ± 41% -49.2% 11343 ± 21% sched_debug.cfs_rq[0]:/.min_vruntime 8908 ± 10% -23.9% 6780 ± 20% sched_debug.cfs_rq[19]:/.min_vruntime 2 ± 30% -63.6% 1 ± 0% sched_debug.cfs_rq[25]:/.nr_spread_over 245 ± 8% +37.6% 337 ± 18% sched_debug.cfs_rq[26]:/.avg->runnable_avg_sum 4354 ± 19% -23.6% 3327 ± 4% sched_debug.cfs_rq[3]:/.exec_clock 11391 ± 13% -36.9% 7185 ± 20% sched_debug.cfs_rq[4]:/.min_vruntime 558 ± 17% -30.7% 387 ± 23% sched_debug.cfs_rq[4]:/.exec_clock 2679 ± 6% +43.9% 3856 ± 13% sched_debug.cpu#13.sched_count 1038 ± 4% +54.8% 1607 ± 19% sched_debug.cpu#13.ttwu_count 720 ± 21% +22.2% 880 ± 9% sched_debug.cpu#33.nr_switches 726 ± 21% +22.3% 888 ± 9% sched_debug.cpu#33.sched_count 4524 ± 1% +79.6% 8127 ± 45% sched_debug.cpu#33.nr_load_updates 326 ± 21% +21.3% 396 ± 11% sched_debug.cpu#33.sched_goidle 598 ± 10% +27.8% 764 ± 15% sched_debug.cpu#35.nr_switches 606 ± 10% +27.0% 770 ± 15% sched_debug.cpu#35.sched_count 7807 ± 41% -42.6% 4477 ± 2% sched_debug.cpu#35.nr_load_updates 269 ± 11% +27.5% 344 ± 13% sched_debug.cpu#35.sched_goidle 4923 ± 36% -43.3% 2790 ± 11% sched_debug.cpu#4.sched_count 501 ± 17% -25.3% 374 ± 28% sched_debug.cpu#44.sched_goidle 396 ± 6% +17.7% 466 ± 8% sched_debug.cpu#46.sched_goidle 833 ± 10% +39.4% 1162 ± 4% sched_debug.cpu#8.sched_goidle 249 ± 39% +113.6% 533 ± 10% sched_debug.cpu#8.ttwu_local 1738 ± 10% +40.7% 2446 ± 5% sched_debug.cpu#8.nr_switches 670 ± 11% +36.0% 911 ± 8% sched_debug.cpu#8.ttwu_count 0.00 ± 20% +55.2% 0.00 ± 39% sched_debug.rt_rq[44]:/.rt_time ivb42: Ivytown Ivy Bridge-EP Memory: 64G unixbench.score 1790 ++-------------------------------------------------------------------+ | O O O O O O | 1780 O+ O O O O O O O O O | 1770 ++ O O O O O O O O O O O O O O | O O O O | 1760 ++ O | 1750 ++ | | | 1740 ++ | 1730 ++ *.* *.*.*.* | *. + : : | 1720 ++*.*.*.* : .*.*.*.* : | 1710 ++ :.* + .*. .*.*.*.*.*.*.*.* | | * * *.*. | 1700 ++-------------------------------------------------------------------+ [*] bisect-good sample [O] bisect-bad sample To reproduce: apt-get install ruby git clone git://git.kernel.org/pub/scm/linux/kernel/git/wfg/lkp-tests.git cd lkp-tests bin/setup-local job.yaml # the job file attached in this email bin/run-local 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