Greeting, We noticed a -92.3% regression of vm-scalability.throughput due to commit: commit: 2f18d46683cb3047c41229d57cf7c6e2ee48676f ("rbd: refactor rbd_wait_state_locked()") https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git master in testcase: vm-scalability on test machine: 80 threads Skylake with 64G memory with following parameters: runtime: 300s test: lru-file-readonce cpufreq_governor: performance test-description: The motivation behind this suite is to exercise functions and regions of the mm/ of the Linux kernel which are of interest to us. test-url: https://git.kernel.org/cgit/linux/kernel/git/wfg/vm-scalability.git/ Details are as below: --------------------------------------------------------------------------------------------------> ========================================================================================= compiler/kconfig/rootfs/sleep/tbox_group/testcase: gcc-7/x86_64-rhel-7.2/debian-x86_64-2018-04-03.cgz/1/vm-lkp-wsx03-8G/boot commit: v4.17-rc1 2f18d46683 ("rbd: refactor rbd_wait_state_locked()") v4.17-rc1 2f18d46683cb3047c41229d57c ---------------- -------------------------- %stddev %change %stddev \ | \ 259103 +520.5% 1607621 vm-scalability.median 5.71 ± 4% -100.0% 0.00 vm-scalability.stddev 21004763 -92.3% 1607621 vm-scalability.throughput 240.41 +28.9% 309.83 vm-scalability.time.elapsed_time 240.41 +28.9% 309.83 vm-scalability.time.elapsed_time.max 439976 -85.3% 64590 ± 19% vm-scalability.time.involuntary_context_switches 62896 -89.1% 6836 vm-scalability.time.minor_page_faults 6698 -98.7% 89.00 vm-scalability.time.percent_of_cpu_this_job_got 15789 -98.4% 252.97 vm-scalability.time.system_time 316.29 -92.0% 25.44 vm-scalability.time.user_time 1292 -84.7% 197.50 ± 7% vm-scalability.time.voluntary_context_switches 4.295e+09 -88.8% 4.823e+08 vm-scalability.workload 2567 ± 65% -99.7% 8.98 boot-time.idle 278517 ± 3% -100.0% 0.00 interrupts.CAL:Function_call_interrupts 1076372 -86.0% 150299 ± 4% softirqs.RCU 493821 -100.0% 0.00 softirqs.SCHED 8295195 -97.9% 175669 ± 2% softirqs.TIMER 14.80 -14.2 0.58 ± 3% mpstat.cpu.idle% 0.00 ± 31% -0.0 0.00 mpstat.cpu.iowait% 0.01 ± 23% +0.1 0.09 ± 26% mpstat.cpu.soft% 1.67 +6.9 8.58 mpstat.cpu.usr% 72.60 ± 4% +142.1% 175.75 vmstat.memory.buff 3313366 ± 7% +50.6% 4988389 vmstat.memory.free 69.20 -98.6% 1.00 vmstat.procs.r 12208 ± 3% -66.1% 4143 ± 2% vmstat.system.cs 86363 -98.7% 1133 ± 2% vmstat.system.in 733300 ± 12% -100.0% 0.00 cpuidle.C1.time 28147 ± 6% -94.1% 1672 ± 2% cpuidle.C1.usage 15489018 ± 3% -100.0% 0.00 cpuidle.C1E.time 49325 ± 2% -100.0% 0.00 cpuidle.C1E.usage 2.758e+09 -100.0% 0.00 cpuidle.C6.time 2879847 -100.0% 0.00 cpuidle.C6.usage 8915 ± 10% -100.0% 0.00 cpuidle.POLL.time 540.00 ± 5% -100.0% 0.00 cpuidle.POLL.usage 240.41 +28.9% 309.83 time.elapsed_time 240.41 +28.9% 309.83 time.elapsed_time.max 439976 -85.3% 64590 ± 19% time.involuntary_context_switches 62896 -89.1% 6836 time.minor_page_faults 6698 -98.7% 89.00 time.percent_of_cpu_this_job_got 15789 -98.4% 252.97 time.system_time 316.29 -92.0% 25.44 time.user_time 1292 -84.7% 197.50 ± 7% time.voluntary_context_switches 5.317e+08 -100.0% 0.00 numa-numastat.node0.local_node 1.119e+08 -100.0% 0.00 numa-numastat.node0.numa_foreign 5.317e+08 -100.0% 0.00 numa-numastat.node0.numa_hit 29285955 ± 3% -100.0% 0.00 numa-numastat.node0.numa_miss 29289175 ± 3% -100.0% 0.00 numa-numastat.node0.other_node 4.015e+08 -100.0% 0.00 numa-numastat.node1.local_node 29285955 ± 3% -100.0% 0.00 numa-numastat.node1.numa_foreign 4.015e+08 -100.0% 0.00 numa-numastat.node1.numa_hit 1.119e+08 -100.0% 0.00 numa-numastat.node1.numa_miss 1.119e+08 -100.0% 0.00 numa-numastat.node1.other_node 2648 +16.3% 3079 turbostat.Avg_MHz 85.69 +13.8 99.50 turbostat.Busy% 25313 ± 6% -93.4% 1672 ± 2% turbostat.C1 45534 ± 3% -100.0% 0.00 turbostat.C1E 0.07 ± 6% -0.1 0.00 turbostat.C1E% 2877662 -100.0% 0.00 turbostat.C6 14.25 -14.2 0.00 turbostat.C6% 14.21 -96.5% 0.49 turbostat.CPU%c1 vm-scalability.throughput 2.2e+07 +-+---------------------------------------------------------------+ 2e+07 +-++..+..+..+..+..+..+..+..+..+..+..+..+..+..+..+..+..+ | | | 1.8e+07 +-+ | 1.6e+07 +-+ | 1.4e+07 +-+ | 1.2e+07 +-+ | | | 1e+07 +-+ | 8e+06 +-+ | 6e+06 +-+ | 4e+06 +-+ | | | 2e+06 O-+O O O O O O O O O O O O O O O O O O O O O O 0 +-+---------------------------------------------------------------+ vm-scalability.median 1.8e+06 +-+---------------------------------------------------------------+ | | 1.6e+06 O-+O O O O O O O O O O O O O O O O O O O O O O | | 1.4e+06 +-+ | 1.2e+06 +-+ | | | 1e+06 +-+ | | | 800000 +-+ | 600000 +-+ | | | 400000 +-+ | |..+.. .+..+..+..+..+.. .+..+.. .+..+..+..+.. .+..+ | 200000 +-+---------------------------------------------------------------+ vm-scalability.workload 4.5e+09 +-+---------------------------------------------------------------+ |..+..+..+..+..+..+..+..+..+..+..+..+..+..+..+..+..+..+ | 4e+09 +-+ | 3.5e+09 +-+ | | | 3e+09 +-+ | 2.5e+09 +-+ | | | 2e+09 +-+ | 1.5e+09 +-+ | | | 1e+09 +-+ | 5e+08 +-+O O O O O O O O O O O O O O O | O O O O O O O O 0 +-+---------------------------------------------------------------+ vm-scalability.time.user_time 350 +-+-------------------------------------------------------------------+ |..+..+...+..+..+..+..+..+...+..+..+..+..+...+..+..+..+..+ | 300 +-+ | | | 250 +-+ | | | 200 +-+ | | | 150 +-+ | | | 100 +-+ | | | 50 +-+ | O O O O O O O O O O O O O O O O O O O O O O O 0 +-+-------------------------------------------------------------------+ vm-scalability.time.system_time 16000 +-+-----------------------------------------------------------------+ | | 14000 +-+ | 12000 +-+ | | | 10000 +-+ | | | 8000 +-+ | | | 6000 +-+ | 4000 +-+ | | | 2000 +-+ | | | 0 O-+O--O--O--O--O---O--O--O--O--O--O--O--O--O--O--O---O--O--O--O--O--O vm-scalability.time.percent_of_cpu_this_job_got 7000 +-+------------------------------------------------------------------+ |..+..+..+...+..+..+..+..+..+..+...+..+..+..+..+..+..+..+ | 6000 +-+ | | | 5000 +-+ | | | 4000 +-+ | | | 3000 +-+ | | | 2000 +-+ | | | 1000 +-+ | | | 0 O-+O--O--O---O--O--O--O--O--O--O---O--O--O--O--O--O--O--O---O--O--O--O vm-scalability.time.elapsed_time 310 O-+O--O---O--O--O--O--O--O---O--O--O--O--O---O--O--O--O--O--O---O--O--O | | 300 +-+ | 290 +-+ | | | 280 +-+ | | | 270 +-+ | | | 260 +-+ | 250 +-+ | | | 240 +-++..+...+..+..+..+..+..+...+..+..+..+..+...+..+..+..+..+ | | | 230 +-+-------------------------------------------------------------------+ vm-scalability.time.elapsed_time.max 310 O-+O--O---O--O--O--O--O--O---O--O--O--O--O---O--O--O--O--O--O---O--O--O | | 300 +-+ | 290 +-+ | | | 280 +-+ | | | 270 +-+ | | | 260 +-+ | 250 +-+ | | | 240 +-++..+...+..+..+..+..+..+...+..+..+..+..+...+..+..+..+..+ | | | 230 +-+-------------------------------------------------------------------+ vm-scalability.time.minor_page_faults 70000 +-+-----------------------------------------------------------------+ |..+..+..+..+..+...+..+..+..+..+..+..+..+..+..+..+...+..+ | 60000 +-+ | | | 50000 +-+ | | | 40000 +-+ | | | 30000 +-+ | | | 20000 +-+ | | | 10000 +-+ | O O O O O O O O O O O O O O O O O O O O O O O 0 +-+-----------------------------------------------------------------+ vm-scalability.time.voluntary_context_switches 1400 +-+------------------------------------------------------------------+ |..+..+..+...+..+..+..+..+..+..+...+..+..+..+..+..+..+..+ | 1200 +-+ | | | 1000 +-+ | | | 800 +-+ | | | 600 +-+ | | | 400 +-+ | | | 200 O-+O O O O O O O O O O O O O O O O O O O O O O | | 0 +-+------------------------------------------------------------------+ vm-scalability.time.involuntary_context_switches 450000 +-+----------------------------------------------------------------+ | +. +. + | 400000 +-+ | 350000 +-+ | | | 300000 +-+ | 250000 +-+ | | | 200000 +-+ | 150000 +-+ | | | 100000 +-+ O O O | 50000 O-+ O O O O O O O O O O O O O O | O O O O O | 0 +-+----------------------------------------------------------------+ [*] 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