Greeting, FYI, we noticed a -9.3% regression of will-it-scale.per_process_ops due to commit: commit: 715fcf3c48c283e5e55993743bd9f814334cb4da ("vfs: Replace array of file pointers with an IDR") git://git.infradead.org/users/willy/linux-dax.git fd-alloc in testcase: will-it-scale on test machine: 48 threads 2 sockets Intel(R) Xeon(R) CPU E5-2697 v2 @ 2.70GHz with 64G memory with following parameters: test: open1 cpufreq_governor: performance test-description: Will It Scale takes a testcase and runs it from 1 through to n parallel copies to see if the testcase will scale. It builds both a process and threads based test in order to see any differences between the two. test-url: https://github.com/antonblanchard/will-it-scale In addition to that, the commit also has significant impact on the following tests: +------------------+----------------------------------------------------------------------+ | testcase: change | unixbench: unixbench.score -29.3% regression | | test machine | 88 threads Intel(R) Xeon(R) CPU E5-2699 v4 @ 2.20GHz with 48G memory | | test parameters | cpufreq_governor=performance | | | nr_task=1 | | | runtime=300s | | | test=syscall | +------------------+----------------------------------------------------------------------+ 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 testcase/path_params/tbox_group/run: will-it-scale/open1-performance/ivb42 4ddb7a6f75c4ebed 715fcf3c48c283e5e55993743b ---------------- -------------------------- %stddev %change %stddev \ | \ 679845 ± 2% -9.3% 616414 ± 1% will-it-scale.per_process_ops 242333 ± 0% -15.3% 205190 ± 1% will-it-scale.per_thread_ops 32038 ± 1% -5.8% 30188 ± 1% will-it-scale.time.involuntary_context_switches 3665 ± 0% +1.1% 3706 ± 0% will-it-scale.time.system_time 30.18 ± 3% -50.1% 15.07 ± 1% will-it-scale.time.user_time 30.18 ± 3% -50.1% 15.07 ± 1% time.user_time 362665 ± 1% -10.3% 325264 ± 0% meminfo.SUnreclaim 415508 ± 1% -10.0% 374099 ± 1% meminfo.Slab 210651 ± 0% -10.5% 188603 ± 0% numa-meminfo.node0.SUnreclaim 237325 ± 1% -10.0% 213542 ± 0% numa-meminfo.node0.Slab 55.93 ± 0% -1.6% 55.03 ± 0% turbostat.%Busy 1664 ± 0% -1.7% 1637 ± 0% turbostat.Avg_MHz 8.20 ± 0% -4.3% 7.84 ± 1% turbostat.RAMWatt 17438940 ± 2% -9.1% 15853343 ± 2% numa-numastat.node0.local_node 17439736 ± 2% -9.1% 15859555 ± 2% numa-numastat.node0.numa_hit 11435732 ± 2% -7.5% 10581271 ± 2% numa-numastat.node1.local_node 11444275 ± 2% -7.5% 10584394 ± 2% numa-numastat.node1.numa_hit 52465 ± 0% -10.0% 47224 ± 1% numa-vmstat.node0.nr_slab_unreclaimable 8160356 ± 2% -7.8% 7521524 ± 2% numa-vmstat.node0.numa_hit 8056954 ± 2% -8.0% 7414199 ± 2% numa-vmstat.node0.numa_local 4061449 ± 4% -5.3% 3845192 ± 4% numa-vmstat.node1.numa_hit 4035672 ± 4% -5.2% 3823927 ± 3% numa-vmstat.node1.numa_local 91034 ± 0% -9.5% 82410 ± 1% proc-vmstat.nr_slab_unreclaimable 28881534 ± 2% -8.4% 26441574 ± 2% proc-vmstat.numa_hit 28872195 ± 2% -8.5% 26432239 ± 2% proc-vmstat.numa_local 1.138e+08 ± 2% -8.6% 1.04e+08 ± 2% proc-vmstat.pgalloc_normal will-it-scale.time.user_time 35 ++---------------------------------------------------------------------+ | .* .*. .*. | 30 *+*.*.* : * *.*.*.*.*.*.*. .* **.*.* *.* | | : : * : : | 25 ++ : : : : | | : : : : | 20 ++ : : : : | | : : : : O | 15 O+O O :O:O O O O O O O O O: :O O O O O O O O O O O | : : : : | 10 ++ : : : : | | :: :: | 5 ++ : : | | : : | 0 ++--O---O-*---O-----------O-------O--O---O---O---O---O-------O-------O-+ will-it-scale.per_thread_ops 300000 ++-----------------------------------------------------------------+ | | 250000 *+ .*.** .*. .*.* * .*. | | * : * *.*.**.*.*.* : :* *.*.*.* | | : O O O O O O : O O O O OO O 200000 O+O O : : O O O O: : O O O O O | | : : : : | 150000 ++ : : : : | | : : : : | 100000 ++ : : : : | | : : : : | | :: :: | 50000 ++ : : | | : : | 0 ++--O--O-*---O----------O-------O--O---O---O---O--O-------O------O-+ [*] 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. Thanks, Xiaolong