August 2019 - LKP - Ml01.01.Org

[locking/rwsem] 7813430057: vm-scalability.median -42.5% regression

by kernel test robot

Greeting, FYI, we noticed a -42.5% regression of vm-scalability.median due to commit: commit: 78134300579a45f527ca173ec8fdb4701b69f16e ("locking/rwsem: Don't call owner_on_cpu() on read-owner") https://kernel.googlesource.com/pub/scm/linux/kernel/git/torvalds/linux.git master in testcase: vm-scalability on test machine: 88 threads Intel(R) Xeon(R) CPU E5-2699 v4 @ 2.20GHz with 128G memory with following parameters: runtime: 300s test: small-allocs-mt cpufreq_governor: performance ucode: 0xb000036 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: --------------------------------------------------------------------------------------------------> To reproduce: git clone https://github.com/intel/lkp-tests.git cd lkp-tests bin/lkp install job.yaml # job file is attached in this email bin/lkp run job.yaml ========================================================================================= compiler/cpufreq_governor/kconfig/rootfs/runtime/tbox_group/test/testcase/ucode: gcc-7/performance/x86_64-rhel-7.6/debian-x86_64-2019-05-14.cgz/300s/lkp-bdw-ep4/small-allocs-mt/vm-scalability/0xb000036 commit: f9adc23ee9 ("futex: Cleanup generic SMP variant of arch_futex_atomic_op_inuser()") 7813430057 ("locking/rwsem: Don't call owner_on_cpu() on read-owner") f9adc23ee91e6f56 78134300579a45f527ca173ec8f ---------------- --------------------------- fail:runs %reproduction fail:runs | | | 1:4 -16% 0:4 perf-profile.children.cycles-pp.error_entry 0:4 -12% 0:4 perf-profile.self.cycles-pp.error_entry %stddev %change %stddev \ | \ 10295 ± 10% -42.5% 5918 ± 3% vm-scalability.median 904912 ± 10% -42.5% 520174 ± 3% vm-scalability.throughput 60474168 ± 10% -42.5% 34747417 ± 3% vm-scalability.time.minor_page_faults 551.25 ± 7% +22.3% 674.25 ± 3% vm-scalability.time.percent_of_cpu_this_job_got 1523 ± 6% +29.0% 1965 ± 3% vm-scalability.time.system_time 161.90 ± 13% -51.5% 78.55 ± 4% vm-scalability.time.user_time 32285656 ± 13% -74.5% 8235971 ± 3% vm-scalability.time.voluntary_context_switches 2.721e+08 ± 10% -42.5% 1.563e+08 ± 3% vm-scalability.workload 37054490 ± 39% -62.6% 13844254 ± 62% cpuidle.C3.usage 5.34 ± 6% +2.0 7.38 ± 2% mpstat.cpu.all.sys% 0.59 ± 12% -0.3 0.32 ± 4% mpstat.cpu.all.usr% 93.50 -1.6% 92.00 vmstat.cpu.id 209869 ± 13% -73.9% 54735 ± 3% vmstat.system.cs 29124 ± 9% -42.5% 16745 ± 4% numa-vmstat.node0.nr_page_table_pages 91040 ± 8% -35.0% 59192 ± 2% numa-vmstat.node0.nr_slab_unreclaimable 29038 ± 11% -42.4% 16734 ± 4% numa-vmstat.node1.nr_page_table_pages 88159 ± 10% -35.8% 56580 ± 4% numa-vmstat.node1.nr_slab_unreclaimable 2763837 ± 3% -12.8% 2410986 meminfo.Memused 232955 ± 10% -42.6% 133795 ± 4% meminfo.PageTables 716735 ± 9% -35.3% 463602 ± 3% meminfo.SUnreclaim 788566 ± 8% -32.1% 535591 ± 3% meminfo.Slab 11681 ± 5% -19.5% 9400 meminfo.max_used_kB 240.75 ± 5% +19.5% 287.75 ± 2% turbostat.Avg_MHz 8.68 ± 4% +1.4 10.12 ± 3% turbostat.Busy% 37053983 ± 39% -62.6% 13843816 ± 62% turbostat.C3 54.52 ± 6% -25.0% 40.91 ± 14% turbostat.CPU%c1 12.86 ± 3% -6.7% 12.00 ± 2% turbostat.RAMWatt 116649 ± 9% -42.5% 67086 ± 4% numa-meminfo.node0.PageTables 364217 ± 8% -34.9% 237199 ± 2% numa-meminfo.node0.SUnreclaim 401594 ± 5% -30.7% 278271 ± 2% numa-meminfo.node0.Slab 116289 ± 11% -42.3% 67047 ± 3% numa-meminfo.node1.PageTables 352683 ± 10% -35.7% 226762 ± 4% numa-meminfo.node1.SUnreclaim 387136 ± 11% -33.4% 257677 ± 4% numa-meminfo.node1.Slab 740.00 ± 3% +10.8% 820.00 ± 3% slabinfo.kmem_cache_node.active_objs 783.25 ± 3% +10.2% 863.00 ± 3% slabinfo.kmem_cache_node.num_objs 2947330 ± 11% -43.3% 1670662 ± 4% slabinfo.vm_area_struct.active_objs 73683 ± 11% -43.3% 41766 ± 4% slabinfo.vm_area_struct.active_slabs 2947343 ± 11% -43.3% 1670680 ± 4% slabinfo.vm_area_struct.num_objs 73683 ± 11% -43.3% 41766 ± 4% slabinfo.vm_area_struct.num_slabs 58082 ± 10% -42.3% 33491 ± 4% proc-vmstat.nr_page_table_pages 179181 ± 9% -35.3% 115969 ± 3% proc-vmstat.nr_slab_unreclaimable 8.75 ±133% +8705.7% 770.50 ±149% proc-vmstat.numa_hint_faults_local 1030609 ± 2% -10.7% 920710 proc-vmstat.numa_hit 1014892 ± 2% -11.0% 903417 proc-vmstat.numa_local 8262 ±160% +287.5% 32019 ± 65% proc-vmstat.numa_pte_updates 4248 ± 4% +14.4% 4860 ± 5% proc-vmstat.pgactivate 1237129 ± 3% -13.4% 1071267 ± 2% proc-vmstat.pgalloc_normal 61279755 ± 10% -42.0% 35555364 ± 3% proc-vmstat.pgfault 1153261 ± 4% -15.6% 972958 ± 12% proc-vmstat.pgfree 12276 ± 8% +14.2% 14023 ± 4% sched_debug.cfs_rq:/.exec_clock.max 816.73 ± 7% +29.3% 1056 ± 2% sched_debug.cfs_rq:/.exec_clock.stddev 90308 ± 8% +33.2% 120279 ± 13% sched_debug.cfs_rq:/.min_vruntime.avg 122001 ± 6% +30.7% 159494 ± 4% sched_debug.cfs_rq:/.min_vruntime.max 8108 ± 6% +35.2% 10962 ± 2% sched_debug.cfs_rq:/.min_vruntime.stddev 3.80 ± 28% +49.4% 5.67 ± 2% sched_debug.cfs_rq:/.runnable_load_avg.avg 19.20 ± 24% +33.7% 25.68 ± 11% sched_debug.cfs_rq:/.runnable_load_avg.stddev 50559 ± 12% +32.5% 67006 ± 2% sched_debug.cfs_rq:/.spread0.max 8108 ± 6% +35.2% 10962 ± 2% sched_debug.cfs_rq:/.spread0.stddev 715.29 ± 11% +31.0% 937.06 ± 3% sched_debug.cfs_rq:/.util_avg.max 0.62 ± 96% -100.0% 0.00 sched_debug.cfs_rq:/.util_avg.min 136.54 ± 16% +49.4% 204.01 ± 11% sched_debug.cfs_rq:/.util_avg.stddev 248.08 ± 6% +44.4% 358.25 ± 19% sched_debug.cfs_rq:/.util_est_enqueued.max 472136 ± 20% +74.7% 824658 ± 3% sched_debug.cpu.avg_idle.avg 48568 ± 17% -49.5% 24506 ± 29% sched_debug.cpu.avg_idle.min 161426 ± 19% +27.8% 206264 ± 4% sched_debug.cpu.avg_idle.stddev 2.64 ± 3% +7.0% 2.83 ± 3% sched_debug.cpu.clock.stddev 2.64 ± 3% +7.0% 2.83 ± 3% sched_debug.cpu.clock_task.stddev 362539 ± 13% -75.3% 89466 ± 12% sched_debug.cpu.nr_switches.avg 385763 ± 12% -71.0% 111713 ± 13% sched_debug.cpu.nr_switches.max 339521 ± 14% -78.2% 74098 ± 14% sched_debug.cpu.nr_switches.min 8.45 ± 6% -17.0% 7.01 ± 9% sched_debug.cpu.nr_uninterruptible.stddev 363377 ± 13% -75.2% 90290 ± 13% sched_debug.cpu.sched_count.avg 408058 ± 13% -65.0% 142960 ± 15% sched_debug.cpu.sched_count.max 339524 ± 14% -78.1% 74224 ± 14% sched_debug.cpu.sched_count.min 180453 ± 13% -75.7% 43933 ± 13% sched_debug.cpu.sched_goidle.avg 191582 ± 12% -71.6% 54363 ± 14% sched_debug.cpu.sched_goidle.max 169052 ± 14% -78.5% 36425 ± 14% sched_debug.cpu.sched_goidle.min 183653 ± 13% -75.1% 45783 ± 13% sched_debug.cpu.ttwu_count.avg 202354 ± 13% -72.8% 55054 ± 15% sched_debug.cpu.ttwu_count.max 168679 ± 13% -75.2% 41903 ± 13% sched_debug.cpu.ttwu_count.min 8022 ± 24% -75.1% 2000 ± 12% sched_debug.cpu.ttwu_count.stddev 1218 ± 6% +20.2% 1464 ± 6% sched_debug.cpu.ttwu_local.max 164.12 ± 17% +41.9% 232.82 ± 3% sched_debug.cpu.ttwu_local.stddev 8.44 ± 11% -36.3% 5.38 ± 11% perf-stat.i.MPKI 3.178e+09 ± 4% +15.6% 3.674e+09 perf-stat.i.branch-instructions 9232346 ± 8% -38.0% 5722174 ± 9% perf-stat.i.cache-misses 1.006e+08 ± 15% -26.2% 74257221 ± 6% perf-stat.i.cache-references 212153 ± 13% -73.9% 55331 ± 3% perf-stat.i.context-switches 2.115e+10 ± 5% +18.7% 2.511e+10 ± 2% perf-stat.i.cpu-cycles 482.97 ± 22% -73.1% 129.87 ± 5% perf-stat.i.cpu-migrations 2713 ± 9% +98.8% 5392 ± 16% perf-stat.i.cycles-between-cache-misses 0.20 ± 4% -0.1 0.09 ± 2% perf-stat.i.dTLB-load-miss-rate% 6517636 ± 10% -48.8% 3334950 ± 2% perf-stat.i.dTLB-load-misses 3.377e+09 ± 6% +26.6% 4.274e+09 ± 2% perf-stat.i.dTLB-loads 1125113 ± 8% -25.5% 838294 ± 4% perf-stat.i.dTLB-store-misses 7.672e+08 ± 5% -21.6% 6.017e+08 perf-stat.i.dTLB-stores 40.86 ± 11% +10.2 51.09 ± 6% perf-stat.i.iTLB-load-miss-rate% 1489763 ± 10% -34.9% 970146 ± 6% perf-stat.i.iTLB-load-misses 2248250 ± 14% -58.7% 928739 ± 6% perf-stat.i.iTLB-loads 1.262e+10 ± 4% +23.4% 1.558e+10 perf-stat.i.instructions 8679 ± 6% +87.6% 16280 ± 5% perf-stat.i.instructions-per-iTLB-miss 0.59 +4.0% 0.62 perf-stat.i.ipc 200189 ± 10% -41.7% 116680 ± 2% perf-stat.i.minor-faults 3344481 ± 10% -52.0% 1606079 ± 3% perf-stat.i.node-load-misses 84375 ± 20% -56.2% 36926 ± 37% perf-stat.i.node-loads 1565682 ± 10% -47.0% 830068 ± 3% perf-stat.i.node-store-misses 869275 ± 12% -46.0% 469789 ± 3% perf-stat.i.node-stores 200189 ± 10% -41.7% 116680 ± 2% perf-stat.i.page-faults 7.94 ± 11% -39.9% 4.77 ± 6% perf-stat.overall.MPKI 1.68 -3.8% 1.61 perf-stat.overall.cpi 2299 ± 4% +92.8% 4434 ± 11% perf-stat.overall.cycles-between-cache-misses 0.19 ± 4% -0.1 0.08 ± 2% perf-stat.overall.dTLB-load-miss-rate% 40.08 ± 10% +11.0 51.08 ± 5% perf-stat.overall.iTLB-load-miss-rate% 8525 ± 6% +89.0% 16114 ± 5% perf-stat.overall.instructions-per-iTLB-miss 0.60 +4.0% 0.62 perf-stat.overall.ipc 14280 ± 5% +112.6% 30354 ± 2% perf-stat.overall.path-length 3.167e+09 ± 4% +15.6% 3.662e+09 perf-stat.ps.branch-instructions 9201666 ± 8% -38.0% 5703837 ± 9% perf-stat.ps.cache-misses 1.002e+08 ± 15% -26.2% 74012239 ± 6% perf-stat.ps.cache-references 211434 ± 13% -73.9% 55142 ± 3% perf-stat.ps.context-switches 2.108e+10 ± 5% +18.7% 2.502e+10 ± 2% perf-stat.ps.cpu-cycles 481.54 ± 22% -73.1% 129.46 ± 5% perf-stat.ps.cpu-migrations 6495785 ± 10% -48.8% 3323804 ± 2% perf-stat.ps.dTLB-load-misses 3.365e+09 ± 6% +26.6% 4.26e+09 ± 2% perf-stat.ps.dTLB-loads 1121400 ± 8% -25.5% 835514 ± 4% perf-stat.ps.dTLB-store-misses 7.647e+08 ± 5% -21.6% 5.997e+08 perf-stat.ps.dTLB-stores 1484812 ± 10% -34.9% 966928 ± 6% perf-stat.ps.iTLB-load-misses 2240685 ± 14% -58.7% 925639 ± 6% perf-stat.ps.iTLB-loads 1.258e+10 ± 4% +23.4% 1.552e+10 perf-stat.ps.instructions 199509 ± 10% -41.7% 116282 ± 2% perf-stat.ps.minor-faults 3333192 ± 10% -52.0% 1600811 ± 3% perf-stat.ps.node-load-misses 84222 ± 20% -56.0% 37029 ± 37% perf-stat.ps.node-loads 1560349 ± 10% -47.0% 827246 ± 3% perf-stat.ps.node-store-misses 866336 ± 12% -46.0% 468214 ± 3% perf-stat.ps.node-stores 199509 ± 10% -41.7% 116282 ± 2% perf-stat.ps.page-faults 3.862e+12 ± 4% +22.8% 4.743e+12 ± 2% perf-stat.total.instructions 15090 ± 9% -30.9% 10424 ± 6% softirqs.CPU0.RCU 14722 ± 5% -28.2% 10570 ± 4% softirqs.CPU1.RCU 14675 ± 6% -25.8% 10884 ± 4% softirqs.CPU10.RCU 14751 ± 4% -27.5% 10698 ± 8% softirqs.CPU11.RCU 14902 ± 8% -27.8% 10765 ± 7% softirqs.CPU12.RCU 14421 ± 3% -26.2% 10637 ± 7% softirqs.CPU13.RCU 14747 ± 5% -27.1% 10757 ± 6% softirqs.CPU14.RCU 14293 -27.0% 10439 ± 7% softirqs.CPU15.RCU 14692 ± 8% -28.9% 10443 ± 6% softirqs.CPU16.RCU 13768 ± 5% -25.4% 10275 ± 6% softirqs.CPU17.RCU 13952 ± 6% -28.1% 10030 ± 6% softirqs.CPU18.RCU 13840 ± 3% -27.2% 10079 ± 6% softirqs.CPU19.RCU 19253 ± 4% -19.3% 15537 ± 4% softirqs.CPU2.RCU 14429 ± 5% -27.2% 10501 ± 7% softirqs.CPU20.RCU 13775 ± 4% -20.6% 10943 ± 12% softirqs.CPU21.RCU 15718 ± 7% -28.4% 11254 ± 6% softirqs.CPU22.RCU 14218 ± 4% -24.3% 10769 ± 10% softirqs.CPU23.RCU 13918 ± 4% -27.7% 10065 ± 7% softirqs.CPU24.RCU 13805 ± 5% -24.3% 10447 ± 16% softirqs.CPU25.RCU 14106 ± 5% -27.2% 10269 ± 8% softirqs.CPU26.RCU 14291 ± 5% -28.4% 10237 ± 7% softirqs.CPU27.RCU 14123 ± 4% -23.4% 10821 ± 7% softirqs.CPU28.RCU 13957 ± 5% -26.7% 10235 ± 7% softirqs.CPU29.RCU 14378 ± 3% -22.0% 11219 ± 15% softirqs.CPU3.RCU 14266 ± 3% -22.5% 11060 ± 12% softirqs.CPU30.RCU 13865 ± 4% -30.4% 9645 ± 7% softirqs.CPU31.RCU 14048 ± 4% -29.3% 9932 ± 5% softirqs.CPU32.RCU 14296 ± 4% -29.6% 10069 ± 12% softirqs.CPU33.RCU 14998 ± 8% -26.4% 11037 ± 11% softirqs.CPU34.RCU 14342 ± 5% -27.1% 10453 ± 5% softirqs.CPU35.RCU 14865 ± 4% -26.4% 10942 ± 3% softirqs.CPU36.RCU 14698 ± 4% -26.3% 10834 ± 4% softirqs.CPU37.RCU 14639 ± 5% -24.6% 11036 ± 6% softirqs.CPU38.RCU 13850 ± 8% -23.8% 10559 ± 5% softirqs.CPU39.RCU 14451 ± 3% -24.8% 10868 ± 6% softirqs.CPU4.RCU 14373 ± 5% -23.8% 10958 ± 2% softirqs.CPU40.RCU 14794 ± 5% -25.7% 10990 ± 2% softirqs.CPU41.RCU 14561 ± 5% -25.4% 10866 ± 3% softirqs.CPU42.RCU 14823 ± 4% -24.0% 11261 ± 10% softirqs.CPU43.RCU 11986 ± 3% -24.8% 9019 ± 2% softirqs.CPU44.RCU 14143 ± 5% -26.1% 10446 ± 6% softirqs.CPU45.RCU 14225 ± 4% -26.3% 10487 ± 8% softirqs.CPU46.RCU 15111 ± 10% -27.9% 10899 ± 7% softirqs.CPU47.RCU 14841 ± 10% -28.5% 10611 ± 6% softirqs.CPU48.RCU 14323 ± 5% -22.3% 11123 ± 3% softirqs.CPU49.RCU 14523 ± 3% -26.5% 10680 ± 5% softirqs.CPU5.RCU 14354 ± 4% -27.4% 10414 ± 6% softirqs.CPU50.RCU 14319 ± 4% -22.5% 11099 ± 12% softirqs.CPU51.RCU 13868 ± 5% -28.7% 9882 ± 9% softirqs.CPU52.RCU 14449 ± 4% -26.9% 10559 ± 11% softirqs.CPU53.RCU 14865 ± 8% -30.3% 10366 ± 7% softirqs.CPU54.RCU 14631 ± 2% -23.3% 11227 ± 15% softirqs.CPU55.RCU 15058 ± 4% -28.2% 10810 ± 7% softirqs.CPU56.RCU 14406 ± 4% -25.8% 10685 ± 6% softirqs.CPU57.RCU 14768 ± 3% -26.1% 10908 ± 6% softirqs.CPU58.RCU 14789 ± 6% -26.6% 10857 ± 5% softirqs.CPU59.RCU 14731 ± 5% -26.6% 10812 ± 6% softirqs.CPU6.RCU 13350 ± 2% -24.3% 10108 ± 7% softirqs.CPU60.RCU 13291 ± 5% -26.8% 9732 ± 8% softirqs.CPU61.RCU 13438 ± 6% -27.1% 9803 ± 8% softirqs.CPU62.RCU 13352 ± 4% -26.0% 9882 ± 6% softirqs.CPU63.RCU 13710 ± 5% -25.2% 10260 ± 6% softirqs.CPU64.RCU 13618 ± 4% -29.0% 9667 ± 8% softirqs.CPU65.RCU 14217 ± 6% -25.3% 10615 ± 7% softirqs.CPU66.RCU 14102 ± 5% -26.7% 10333 ± 8% softirqs.CPU67.RCU 14124 ± 4% -27.1% 10295 ± 8% softirqs.CPU68.RCU 13718 ± 5% -32.5% 9256 ± 9% softirqs.CPU69.RCU 14427 ± 4% -25.3% 10776 ± 4% softirqs.CPU7.RCU 13928 ± 4% -26.2% 10273 ± 9% softirqs.CPU70.RCU 13906 ± 6% -26.0% 10295 ± 6% softirqs.CPU71.RCU 14128 ± 5% -26.1% 10438 ± 8% softirqs.CPU72.RCU 14056 ± 4% -27.3% 10212 ± 7% softirqs.CPU73.RCU 14121 ± 5% -27.0% 10306 ± 10% softirqs.CPU74.RCU 13156 ± 4% -24.5% 9933 ± 15% softirqs.CPU75.RCU 13735 ± 5% -28.0% 9894 ± 15% softirqs.CPU76.RCU 13347 ± 4% -22.9% 10286 ± 3% softirqs.CPU77.RCU 13722 ± 9% -27.4% 9968 ± 7% softirqs.CPU78.RCU 13310 ± 5% -27.4% 9660 ± 6% softirqs.CPU79.RCU 14992 ± 11% -31.8% 10220 ± 9% softirqs.CPU8.RCU 13988 ± 5% -27.7% 10118 ± 5% softirqs.CPU80.RCU 13918 ± 4% -27.0% 10156 ± 5% softirqs.CPU81.RCU 13849 ± 5% -27.0% 10110 ± 5% softirqs.CPU82.RCU 14076 ± 6% -27.2% 10248 softirqs.CPU83.RCU 13473 ± 5% -24.3% 10205 softirqs.CPU84.RCU 13814 ± 5% -25.2% 10334 ± 5% softirqs.CPU85.RCU 13938 ± 4% -27.7% 10077 ± 4% softirqs.CPU86.RCU 14634 ± 9% -30.4% 10189 ± 5% softirqs.CPU87.RCU 13692 ± 11% -25.4% 10219 ± 9% softirqs.CPU9.RCU 1255273 ± 4% -26.4% 923618 ± 6% softirqs.RCU 308.00 ± 38% -41.3% 180.75 ± 7% interrupts.36:PCI-MSI.1572867-edge.eth0-TxRx-3 296.75 ± 31% -31.6% 203.00 ± 12% interrupts.42:PCI-MSI.1572873-edge.eth0-TxRx-9 196.50 ± 10% -17.4% 162.25 ± 3% interrupts.48:PCI-MSI.1572879-edge.eth0-TxRx-15 5727 ± 7% -38.4% 3530 ± 19% interrupts.CPU0.RES:Rescheduling_interrupts 5869 ± 14% -43.1% 3342 ± 10% interrupts.CPU1.RES:Rescheduling_interrupts 5447 ± 9% -40.8% 3226 ± 7% interrupts.CPU10.RES:Rescheduling_interrupts 40.50 ±101% +126.5% 91.75 ± 38% interrupts.CPU10.TLB:TLB_shootdowns 5312 ± 11% -40.9% 3137 ± 15% interrupts.CPU11.RES:Rescheduling_interrupts 49.00 ± 63% +78.1% 87.25 ± 27% interrupts.CPU11.TLB:TLB_shootdowns 5528 ± 7% -40.0% 3316 ± 15% interrupts.CPU12.RES:Rescheduling_interrupts 5491 ± 8% -44.6% 3042 ± 6% interrupts.CPU13.RES:Rescheduling_interrupts 5147 ± 8% -44.2% 2873 ± 14% interrupts.CPU14.RES:Rescheduling_interrupts 196.50 ± 10% -17.4% 162.25 ± 3% interrupts.CPU15.48:PCI-MSI.1572879-edge.eth0-TxRx-15 5139 ± 7% -50.1% 2566 ± 15% interrupts.CPU15.RES:Rescheduling_interrupts 5509 ± 11% -36.7% 3489 ± 10% interrupts.CPU16.RES:Rescheduling_interrupts 5478 ± 5% -44.1% 3065 ± 16% interrupts.CPU17.RES:Rescheduling_interrupts 5258 ± 12% -36.9% 3319 ± 9% interrupts.CPU18.RES:Rescheduling_interrupts 5652 ± 11% -35.5% 3643 ± 17% interrupts.CPU19.RES:Rescheduling_interrupts 5383 ± 14% -45.1% 2954 ± 13% interrupts.CPU2.RES:Rescheduling_interrupts 5356 ± 9% -46.0% 2894 ± 14% interrupts.CPU20.RES:Rescheduling_interrupts 5763 ± 19% -50.8% 2836 ± 11% interrupts.CPU21.RES:Rescheduling_interrupts 5639 ± 7% -39.4% 3418 ± 12% interrupts.CPU22.RES:Rescheduling_interrupts 5579 ± 6% -33.4% 3713 ± 25% interrupts.CPU23.RES:Rescheduling_interrupts 912.75 ± 33% +45.7% 1329 ± 6% interrupts.CPU24.NMI:Non-maskable_interrupts 912.75 ± 33% +45.7% 1329 ± 6% interrupts.CPU24.PMI:Performance_monitoring_interrupts 5440 ± 6% -39.2% 3305 ± 13% interrupts.CPU24.RES:Rescheduling_interrupts 5716 ± 7% -36.5% 3628 ± 9% interrupts.CPU25.RES:Rescheduling_interrupts 5544 ± 8% -37.3% 3474 ± 4% interrupts.CPU26.RES:Rescheduling_interrupts 948.00 ± 35% +67.1% 1583 ± 17% interrupts.CPU27.NMI:Non-maskable_interrupts 948.00 ± 35% +67.1% 1583 ± 17% interrupts.CPU27.PMI:Performance_monitoring_interrupts 5776 ± 7% -31.0% 3985 ± 14% interrupts.CPU27.RES:Rescheduling_interrupts 972.75 ± 24% +57.4% 1531 ± 14% interrupts.CPU28.NMI:Non-maskable_interrupts 972.75 ± 24% +57.4% 1531 ± 14% interrupts.CPU28.PMI:Performance_monitoring_interrupts 5638 ± 6% -40.9% 3334 ± 15% interrupts.CPU28.RES:Rescheduling_interrupts 5537 ± 5% -30.1% 3869 ± 22% interrupts.CPU29.RES:Rescheduling_interrupts 308.00 ± 38% -41.3% 180.75 ± 7% interrupts.CPU3.36:PCI-MSI.1572867-edge.eth0-TxRx-3 5446 ± 14% -39.7% 3282 ± 8% interrupts.CPU3.RES:Rescheduling_interrupts 1284 ± 3% +21.6% 1561 ± 5% interrupts.CPU30.NMI:Non-maskable_interrupts 1284 ± 3% +21.6% 1561 ± 5% interrupts.CPU30.PMI:Performance_monitoring_interrupts 5483 ± 12% -43.1% 3119 ± 15% interrupts.CPU30.RES:Rescheduling_interrupts 5823 ± 7% -42.4% 3356 ± 14% interrupts.CPU31.RES:Rescheduling_interrupts 28.25 ± 89% +187.6% 81.25 ± 40% interrupts.CPU31.TLB:TLB_shootdowns 5776 ± 7% -33.6% 3836 ± 13% interrupts.CPU32.RES:Rescheduling_interrupts 5675 ± 10% -45.1% 3117 ± 12% interrupts.CPU33.RES:Rescheduling_interrupts 1170 ± 3% +23.4% 1443 ± 11% interrupts.CPU34.NMI:Non-maskable_interrupts 1170 ± 3% +23.4% 1443 ± 11% interrupts.CPU34.PMI:Performance_monitoring_interrupts 5754 ± 9% -43.7% 3240 ± 14% interrupts.CPU34.RES:Rescheduling_interrupts 1252 ± 6% +15.6% 1447 ± 8% interrupts.CPU35.NMI:Non-maskable_interrupts 1252 ± 6% +15.6% 1447 ± 8% interrupts.CPU35.PMI:Performance_monitoring_interrupts 5266 ± 8% -41.2% 3099 ± 11% interrupts.CPU35.RES:Rescheduling_interrupts 5239 ± 10% -41.2% 3082 ± 23% interrupts.CPU36.RES:Rescheduling_interrupts 5576 ± 12% -43.0% 3175 ± 11% interrupts.CPU37.RES:Rescheduling_interrupts 5176 ± 6% -36.5% 3287 ± 16% interrupts.CPU38.RES:Rescheduling_interrupts 5559 ± 6% -35.5% 3583 ± 19% interrupts.CPU39.RES:Rescheduling_interrupts 5568 ± 9% -41.9% 3233 ± 15% interrupts.CPU4.RES:Rescheduling_interrupts 1096 ± 21% +37.8% 1511 ± 11% interrupts.CPU40.NMI:Non-maskable_interrupts 1096 ± 21% +37.8% 1511 ± 11% interrupts.CPU40.PMI:Performance_monitoring_interrupts 5151 ± 8% -34.8% 3360 ± 10% interrupts.CPU40.RES:Rescheduling_interrupts 5181 ± 9% -42.5% 2981 ± 16% interrupts.CPU41.RES:Rescheduling_interrupts 1139 ± 28% +36.7% 1557 ± 6% interrupts.CPU42.NMI:Non-maskable_interrupts 1139 ± 28% +36.7% 1557 ± 6% interrupts.CPU42.PMI:Performance_monitoring_interrupts 4909 ± 10% -36.5% 3117 ± 18% interrupts.CPU42.RES:Rescheduling_interrupts 5249 ± 5% -37.2% 3296 ± 9% interrupts.CPU43.RES:Rescheduling_interrupts 1146 ± 29% +31.0% 1501 ± 8% interrupts.CPU44.NMI:Non-maskable_interrupts 1146 ± 29% +31.0% 1501 ± 8% interrupts.CPU44.PMI:Performance_monitoring_interrupts 5653 ± 20% -45.4% 3084 ± 7% interrupts.CPU44.RES:Rescheduling_interrupts 5853 ± 5% -48.0% 3042 ± 3% interrupts.CPU45.RES:Rescheduling_interrupts 32.75 ± 47% +221.4% 105.25 ± 45% interrupts.CPU45.TLB:TLB_shootdowns 5898 ± 10% -43.1% 3358 ± 10% interrupts.CPU46.RES:Rescheduling_interrupts 5392 ± 9% -39.0% 3290 ± 18% interrupts.CPU47.RES:Rescheduling_interrupts 5924 ± 6% -46.5% 3170 ± 11% interrupts.CPU48.RES:Rescheduling_interrupts 5938 ± 5% -41.4% 3480 ± 8% interrupts.CPU49.RES:Rescheduling_interrupts 31.00 ± 63% +177.4% 86.00 ± 39% interrupts.CPU49.TLB:TLB_shootdowns 5543 ± 9% -42.4% 3193 ± 7% interrupts.CPU5.RES:Rescheduling_interrupts 5277 ± 5% -41.4% 3092 ± 3% interrupts.CPU50.RES:Rescheduling_interrupts 6101 ± 9% -44.3% 3399 ± 10% interrupts.CPU51.RES:Rescheduling_interrupts 5834 ± 12% -40.3% 3481 ± 6% interrupts.CPU52.RES:Rescheduling_interrupts 6214 ± 9% -40.2% 3717 ± 12% interrupts.CPU53.RES:Rescheduling_interrupts 25.75 ± 95% +198.1% 76.75 ± 46% interrupts.CPU53.TLB:TLB_shootdowns 6278 ± 11% -52.7% 2967 ± 13% interrupts.CPU54.RES:Rescheduling_interrupts 5668 ± 11% -43.4% 3205 ± 4% interrupts.CPU55.RES:Rescheduling_interrupts 5820 ± 8% -39.9% 3500 ± 3% interrupts.CPU56.RES:Rescheduling_interrupts 5928 ± 6% -43.5% 3350 ± 8% interrupts.CPU57.RES:Rescheduling_interrupts 35.00 ± 34% +127.1% 79.50 ± 29% interrupts.CPU57.TLB:TLB_shootdowns 5632 ± 11% -44.7% 3115 ± 16% interrupts.CPU58.RES:Rescheduling_interrupts 5440 ± 8% -47.4% 2863 ± 8% interrupts.CPU59.RES:Rescheduling_interrupts 5485 ± 12% -44.5% 3044 ± 8% interrupts.CPU6.RES:Rescheduling_interrupts 1153 ± 25% +32.2% 1525 ± 7% interrupts.CPU60.NMI:Non-maskable_interrupts 1153 ± 25% +32.2% 1525 ± 7% interrupts.CPU60.PMI:Performance_monitoring_interrupts 5514 ± 6% -45.5% 3006 ± 11% interrupts.CPU60.RES:Rescheduling_interrupts 5600 ± 7% -36.7% 3547 ± 14% interrupts.CPU61.RES:Rescheduling_interrupts 5612 ± 7% -30.8% 3885 ± 27% interrupts.CPU62.RES:Rescheduling_interrupts 1214 ± 5% +22.3% 1485 ± 5% interrupts.CPU63.NMI:Non-maskable_interrupts 1214 ± 5% +22.3% 1485 ± 5% interrupts.CPU63.PMI:Performance_monitoring_interrupts 6114 ± 11% -45.5% 3335 ± 13% interrupts.CPU63.RES:Rescheduling_interrupts 5297 ± 14% -45.4% 2891 ± 17% interrupts.CPU64.RES:Rescheduling_interrupts 1309 ± 10% +23.4% 1616 ± 9% interrupts.CPU65.NMI:Non-maskable_interrupts 1309 ± 10% +23.4% 1616 ± 9% interrupts.CPU65.PMI:Performance_monitoring_interrupts 5366 ± 8% -49.0% 2738 ± 17% interrupts.CPU65.RES:Rescheduling_interrupts 1267 ± 6% +23.6% 1566 ± 13% interrupts.CPU66.NMI:Non-maskable_interrupts 1267 ± 6% +23.6% 1566 ± 13% interrupts.CPU66.PMI:Performance_monitoring_interrupts 5244 ± 12% -37.8% 3261 ± 15% interrupts.CPU66.RES:Rescheduling_interrupts 5758 ± 11% -41.1% 3390 ± 23% interrupts.CPU67.RES:Rescheduling_interrupts 5453 ± 7% -35.8% 3501 ± 12% interrupts.CPU68.RES:Rescheduling_interrupts 5786 ± 11% -41.0% 3413 ± 20% interrupts.CPU69.RES:Rescheduling_interrupts 5628 ± 10% -45.4% 3072 ± 6% interrupts.CPU7.RES:Rescheduling_interrupts 1199 ± 3% +23.0% 1474 ± 7% interrupts.CPU70.NMI:Non-maskable_interrupts 1199 ± 3% +23.0% 1474 ± 7% interrupts.CPU70.PMI:Performance_monitoring_interrupts 5714 ± 7% -32.3% 3870 ± 13% interrupts.CPU70.RES:Rescheduling_interrupts 1242 ± 2% +22.5% 1521 ± 8% interrupts.CPU71.NMI:Non-maskable_interrupts 1242 ± 2% +22.5% 1521 ± 8% interrupts.CPU71.PMI:Performance_monitoring_interrupts 5650 ± 10% -38.1% 3496 ± 13% interrupts.CPU71.RES:Rescheduling_interrupts 73.75 ± 33% -55.3% 33.00 ± 40% interrupts.CPU71.TLB:TLB_shootdowns 1298 ± 14% +16.8% 1515 ± 12% interrupts.CPU72.NMI:Non-maskable_interrupts 1298 ± 14% +16.8% 1515 ± 12% interrupts.CPU72.PMI:Performance_monitoring_interrupts 5255 ± 6% -45.6% 2857 ± 15% interrupts.CPU72.RES:Rescheduling_interrupts 5522 ± 7% -37.1% 3473 ± 19% interrupts.CPU73.RES:Rescheduling_interrupts 1255 ± 5% +19.3% 1497 ± 10% interrupts.CPU74.NMI:Non-maskable_interrupts 1255 ± 5% +19.3% 1497 ± 10% interrupts.CPU74.PMI:Performance_monitoring_interrupts 5540 ± 8% -39.1% 3371 ± 18% interrupts.CPU74.RES:Rescheduling_interrupts 5489 ± 3% -45.6% 2986 ± 17% interrupts.CPU75.RES:Rescheduling_interrupts 5618 ± 7% -37.9% 3489 ± 22% interrupts.CPU76.RES:Rescheduling_interrupts 1284 ± 13% +30.2% 1673 ± 12% interrupts.CPU77.NMI:Non-maskable_interrupts 1284 ± 13% +30.2% 1673 ± 12% interrupts.CPU77.PMI:Performance_monitoring_interrupts 5621 ± 6% -29.8% 3943 ± 19% interrupts.CPU77.RES:Rescheduling_interrupts 1166 ± 4% +24.9% 1456 ± 9% interrupts.CPU78.NMI:Non-maskable_interrupts 1166 ± 4% +24.9% 1456 ± 9% interrupts.CPU78.PMI:Performance_monitoring_interrupts 5624 ± 9% -36.3% 3585 ± 19% interrupts.CPU78.RES:Rescheduling_interrupts 30.75 ± 81% +166.7% 82.00 ± 47% interrupts.CPU78.TLB:TLB_shootdowns 5642 ± 10% -33.2% 3772 ± 15% interrupts.CPU79.RES:Rescheduling_interrupts 5223 ± 8% -35.4% 3374 ± 8% interrupts.CPU8.RES:Rescheduling_interrupts 1309 ± 10% +23.4% 1615 ± 11% interrupts.CPU80.NMI:Non-maskable_interrupts 1309 ± 10% +23.4% 1615 ± 11% interrupts.CPU80.PMI:Performance_monitoring_interrupts 5464 ± 12% -35.6% 3521 ± 20% interrupts.CPU80.RES:Rescheduling_interrupts 1254 ± 5% +28.0% 1606 ± 5% interrupts.CPU81.NMI:Non-maskable_interrupts 1254 ± 5% +28.0% 1606 ± 5% interrupts.CPU81.PMI:Performance_monitoring_interrupts 5620 ± 7% -40.9% 3323 ± 9% interrupts.CPU81.RES:Rescheduling_interrupts 1210 ± 2% +20.2% 1454 ± 7% interrupts.CPU82.NMI:Non-maskable_interrupts 1210 ± 2% +20.2% 1454 ± 7% interrupts.CPU82.PMI:Performance_monitoring_interrupts 5422 ± 8% -37.6% 3382 ± 20% interrupts.CPU82.RES:Rescheduling_interrupts 5565 ± 4% -36.0% 3559 ± 14% interrupts.CPU83.RES:Rescheduling_interrupts 5603 ± 7% -33.3% 3739 ± 16% interrupts.CPU84.RES:Rescheduling_interrupts 5346 ± 5% -34.6% 3498 ± 15% interrupts.CPU85.RES:Rescheduling_interrupts 5122 ± 8% -41.4% 3000 ± 11% interrupts.CPU86.RES:Rescheduling_interrupts 5468 ± 7% -35.8% 3509 ± 12% interrupts.CPU87.RES:Rescheduling_interrupts 296.75 ± 31% -31.6% 203.00 ± 12% interrupts.CPU9.42:PCI-MSI.1572873-edge.eth0-TxRx-9 5324 ± 10% -42.6% 3053 ± 7% interrupts.CPU9.RES:Rescheduling_interrupts 102430 ± 6% +16.8% 119675 ± 6% interrupts.NMI:Non-maskable_interrupts 102430 ± 6% +16.8% 119675 ± 6% interrupts.PMI:Performance_monitoring_interrupts 488877 ± 8% -40.4% 291343 ± 3% interrupts.RES:Rescheduling_interrupts 67.21 ± 2% -13.1 54.11 ± 3% perf-profile.calltrace.cycles-pp.secondary_startup_64 66.51 ± 2% -12.9 53.62 ± 3% perf-profile.calltrace.cycles-pp.start_secondary.secondary_startup_64 66.51 ± 2% -12.9 53.62 ± 3% perf-profile.calltrace.cycles-pp.cpu_startup_entry.start_secondary.secondary_startup_64 66.47 ± 2% -12.9 53.59 ± 3% perf-profile.calltrace.cycles-pp.do_idle.cpu_startup_entry.start_secondary.secondary_startup_64 51.36 -8.5 42.83 ± 3% perf-profile.calltrace.cycles-pp.intel_idle.cpuidle_enter_state.cpuidle_enter.do_idle.cpu_startup_entry 56.55 -7.7 48.81 ± 4% perf-profile.calltrace.cycles-pp.cpuidle_enter.do_idle.cpu_startup_entry.start_secondary.secondary_startup_64 56.16 -7.7 48.47 ± 4% perf-profile.calltrace.cycles-pp.cpuidle_enter_state.cpuidle_enter.do_idle.cpu_startup_entry.start_secondary 3.27 ± 17% -2.3 0.98 ± 17% perf-profile.calltrace.cycles-pp.sched_ttwu_pending.do_idle.cpu_startup_entry.start_secondary.secondary_startup_64 2.45 ± 16% -2.1 0.31 ±100% perf-profile.calltrace.cycles-pp.__account_scheduler_latency.enqueue_entity.enqueue_task_fair.activate_task.ttwu_do_activate 2.72 ± 10% -1.9 0.77 ± 12% perf-profile.calltrace.cycles-pp.rwsem_wake.vm_mmap_pgoff.ksys_mmap_pgoff.do_syscall_64.entry_SYSCALL_64_after_hwframe 2.73 ± 17% -1.9 0.83 ± 17% perf-profile.calltrace.cycles-pp.ttwu_do_activate.sched_ttwu_pending.do_idle.cpu_startup_entry.start_secondary 2.72 ± 17% -1.9 0.83 ± 17% perf-profile.calltrace.cycles-pp.activate_task.ttwu_do_activate.sched_ttwu_pending.do_idle.cpu_startup_entry 2.71 ± 17% -1.9 0.83 ± 17% perf-profile.calltrace.cycles-pp.enqueue_task_fair.activate_task.ttwu_do_activate.sched_ttwu_pending.do_idle 2.52 ± 17% -1.7 0.78 ± 16% perf-profile.calltrace.cycles-pp.enqueue_entity.enqueue_task_fair.activate_task.ttwu_do_activate.sched_ttwu_pending 2.30 ± 10% -1.6 0.67 ± 12% perf-profile.calltrace.cycles-pp.wake_up_q.rwsem_wake.vm_mmap_pgoff.ksys_mmap_pgoff.do_syscall_64 2.08 ± 17% -1.6 0.47 ± 59% perf-profile.calltrace.cycles-pp.schedule.rwsem_down_read_slowpath.__do_page_fault.do_page_fault.page_fault 2.26 ± 10% -1.6 0.66 ± 13% perf-profile.calltrace.cycles-pp.try_to_wake_up.wake_up_q.rwsem_wake.vm_mmap_pgoff.ksys_mmap_pgoff 2.31 ± 18% -1.6 0.71 ± 12% perf-profile.calltrace.cycles-pp.schedule_idle.do_idle.cpu_startup_entry.start_secondary.secondary_startup_64 2.29 ± 18% -1.6 0.71 ± 11% perf-profile.calltrace.cycles-pp.__schedule.schedule_idle.do_idle.cpu_startup_entry.start_secondary 2.04 ± 17% -1.6 0.47 ± 59% perf-profile.calltrace.cycles-pp.__schedule.schedule.rwsem_down_read_slowpath.__do_page_fault.do_page_fault 10.72 ± 4% -1.4 9.31 ± 10% perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe 10.71 ± 4% -1.4 9.31 ± 10% perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe 10.47 ± 4% -1.3 9.16 ± 11% perf-profile.calltrace.cycles-pp.ksys_mmap_pgoff.do_syscall_64.entry_SYSCALL_64_after_hwframe 10.39 ± 4% -1.3 9.12 ± 11% perf-profile.calltrace.cycles-pp.vm_mmap_pgoff.ksys_mmap_pgoff.do_syscall_64.entry_SYSCALL_64_after_hwframe 1.63 ± 13% -1.0 0.67 ± 9% perf-profile.calltrace.cycles-pp.do_mmap.vm_mmap_pgoff.ksys_mmap_pgoff.do_syscall_64.entry_SYSCALL_64_after_hwframe 1.09 ± 14% -0.8 0.29 ±100% perf-profile.calltrace.cycles-pp.__handle_mm_fault.handle_mm_fault.__do_page_fault.do_page_fault.page_fault 1.28 ± 14% -0.7 0.62 ± 11% perf-profile.calltrace.cycles-pp.handle_mm_fault.__do_page_fault.do_page_fault.page_fault 2.83 ± 8% -0.6 2.25 ± 8% perf-profile.calltrace.cycles-pp.menu_select.do_idle.cpu_startup_entry.start_secondary.secondary_startup_64 0.80 ± 7% -0.3 0.46 ± 59% perf-profile.calltrace.cycles-pp.get_next_timer_interrupt.tick_nohz_next_event.tick_nohz_get_sleep_length.menu_select.do_idle 1.42 ± 10% -0.3 1.09 ± 9% perf-profile.calltrace.cycles-pp.tick_nohz_get_sleep_length.menu_select.do_idle.cpu_startup_entry.start_secondary 1.10 ± 8% -0.3 0.85 ± 10% perf-profile.calltrace.cycles-pp.tick_nohz_next_event.tick_nohz_get_sleep_length.menu_select.do_idle.cpu_startup_entry 0.63 ± 13% +0.2 0.82 ± 14% perf-profile.calltrace.cycles-pp.update_process_times.tick_sched_handle.tick_sched_timer.__hrtimer_run_queues.hrtimer_interrupt 0.70 ± 14% +0.2 0.90 ± 14% perf-profile.calltrace.cycles-pp.tick_sched_handle.tick_sched_timer.__hrtimer_run_queues.hrtimer_interrupt.smp_apic_timer_interrupt 0.88 ± 13% +0.2 1.10 ± 11% perf-profile.calltrace.cycles-pp.tick_sched_timer.__hrtimer_run_queues.hrtimer_interrupt.smp_apic_timer_interrupt.apic_timer_interrupt 1.46 ± 9% +0.3 1.79 ± 7% perf-profile.calltrace.cycles-pp.__hrtimer_run_queues.hrtimer_interrupt.smp_apic_timer_interrupt.apic_timer_interrupt.cpuidle_enter_state 0.74 ± 20% +1.6 2.31 ± 10% perf-profile.calltrace.cycles-pp.rwsem_spin_on_owner.rwsem_optimistic_spin.rwsem_down_read_slowpath.__do_page_fault.do_page_fault 5.93 ± 2% +1.7 7.62 ± 12% perf-profile.calltrace.cycles-pp.down_write_killable.vm_mmap_pgoff.ksys_mmap_pgoff.do_syscall_64.entry_SYSCALL_64_after_hwframe 5.86 ± 2% +1.7 7.58 ± 12% perf-profile.calltrace.cycles-pp.rwsem_down_write_slowpath.down_write_killable.vm_mmap_pgoff.ksys_mmap_pgoff.do_syscall_64 4.87 ± 4% +2.2 7.11 ± 13% perf-profile.calltrace.cycles-pp.rwsem_optimistic_spin.rwsem_down_write_slowpath.down_write_killable.vm_mmap_pgoff.ksys_mmap_pgoff 3.71 ± 5% +2.3 6.04 ± 14% perf-profile.calltrace.cycles-pp.osq_lock.rwsem_optimistic_spin.rwsem_down_write_slowpath.down_write_killable.vm_mmap_pgoff 19.29 ± 10% +16.1 35.34 ± 4% perf-profile.calltrace.cycles-pp.page_fault 19.24 ± 10% +16.1 35.32 ± 4% perf-profile.calltrace.cycles-pp.do_page_fault.page_fault 19.09 ± 10% +16.2 35.26 ± 4% perf-profile.calltrace.cycles-pp.__do_page_fault.do_page_fault.page_fault 15.78 ± 15% +17.7 33.52 ± 4% perf-profile.calltrace.cycles-pp.rwsem_down_read_slowpath.__do_page_fault.do_page_fault.page_fault 8.85 ± 29% +18.5 27.35 ± 5% perf-profile.calltrace.cycles-pp.osq_lock.rwsem_optimistic_spin.rwsem_down_read_slowpath.__do_page_fault.do_page_fault 10.65 ± 28% +20.5 31.17 ± 4% perf-profile.calltrace.cycles-pp.rwsem_optimistic_spin.rwsem_down_read_slowpath.__do_page_fault.do_page_fault.page_fault 67.21 ± 2% -13.1 54.11 ± 3% perf-profile.children.cycles-pp.secondary_startup_64 67.21 ± 2% -13.1 54.11 ± 3% perf-profile.children.cycles-pp.cpu_startup_entry 67.23 ± 2% -13.1 54.13 ± 3% perf-profile.children.cycles-pp.do_idle 66.51 ± 2% -12.9 53.62 ± 3% perf-profile.children.cycles-pp.start_secondary 51.47 -8.5 42.96 ± 3% perf-profile.children.cycles-pp.intel_idle 57.16 -7.9 49.27 ± 4% perf-profile.children.cycles-pp.cpuidle_enter 57.14 -7.9 49.26 ± 4% perf-profile.children.cycles-pp.cpuidle_enter_state 4.84 ± 16% -3.3 1.55 ± 13% perf-profile.children.cycles-pp.__schedule 4.04 ± 14% -2.8 1.24 ± 14% perf-profile.children.cycles-pp.ttwu_do_activate 4.04 ± 14% -2.8 1.24 ± 14% perf-profile.children.cycles-pp.activate_task 4.03 ± 14% -2.8 1.23 ± 14% perf-profile.children.cycles-pp.enqueue_task_fair 3.78 ± 14% -2.6 1.16 ± 13% perf-profile.children.cycles-pp.enqueue_entity 3.36 ± 16% -2.4 1.00 ± 17% perf-profile.children.cycles-pp.sched_ttwu_pending 3.20 ± 8% -2.1 1.08 ± 11% perf-profile.children.cycles-pp.rwsem_wake 2.59 ± 15% -1.8 0.79 ± 14% perf-profile.children.cycles-pp.__account_scheduler_latency 2.72 ± 9% -1.8 0.93 ± 11% perf-profile.children.cycles-pp.wake_up_q 2.67 ± 9% -1.8 0.92 ± 11% perf-profile.children.cycles-pp.try_to_wake_up 2.56 ± 14% -1.7 0.85 ± 13% perf-profile.children.cycles-pp.schedule 2.34 ± 18% -1.6 0.72 ± 12% perf-profile.children.cycles-pp.schedule_idle 2.00 ± 14% -1.5 0.48 ± 6% perf-profile.children.cycles-pp.switch_mm_irqs_off 10.72 ± 4% -1.4 9.31 ± 10% perf-profile.children.cycles-pp.entry_SYSCALL_64_after_hwframe 1.90 ± 30% -1.4 0.50 ± 14% perf-profile.children.cycles-pp.native_queued_spin_lock_slowpath 10.71 ± 4% -1.4 9.31 ± 10% perf-profile.children.cycles-pp.do_syscall_64 1.93 ± 28% -1.4 0.54 ± 9% perf-profile.children.cycles-pp._raw_spin_lock_irq 10.47 ± 4% -1.3 9.16 ± 11% perf-profile.children.cycles-pp.ksys_mmap_pgoff 10.40 ± 4% -1.3 9.12 ± 11% perf-profile.children.cycles-pp.vm_mmap_pgoff 1.60 ± 15% -1.1 0.51 ± 15% perf-profile.children.cycles-pp.stack_trace_save_tsk 1.64 ± 13% -1.0 0.67 ± 9% perf-profile.children.cycles-pp.do_mmap 1.37 ± 17% -0.9 0.45 ± 17% perf-profile.children.cycles-pp.arch_stack_walk 1.02 ± 13% -0.8 0.19 ± 6% perf-profile.children.cycles-pp.switch_mm 1.14 ± 14% -0.8 0.34 ± 16% perf-profile.children.cycles-pp.dequeue_task_fair 1.06 ± 13% -0.7 0.31 ± 14% perf-profile.children.cycles-pp.dequeue_entity 1.17 ± 9% -0.7 0.50 ± 10% perf-profile.children.cycles-pp.swapgs_restore_regs_and_return_to_usermode 1.29 ± 14% -0.7 0.62 ± 12% perf-profile.children.cycles-pp.handle_mm_fault 0.95 ± 16% -0.7 0.29 ± 15% perf-profile.children.cycles-pp.unwind_next_frame 1.02 ± 12% -0.6 0.40 ± 6% perf-profile.children.cycles-pp.update_load_avg 2.88 ± 8% -0.6 2.29 ± 8% perf-profile.children.cycles-pp.menu_select 1.10 ± 14% -0.6 0.53 ± 10% perf-profile.children.cycles-pp.__handle_mm_fault 0.86 ± 15% -0.6 0.30 ± 8% perf-profile.children.cycles-pp.get_unmapped_area 0.82 ± 15% -0.5 0.29 ± 8% perf-profile.children.cycles-pp.arch_get_unmapped_area_topdown 0.97 ± 13% -0.5 0.45 ± 8% perf-profile.children.cycles-pp._raw_spin_lock_irqsave 0.90 ± 10% -0.5 0.39 ± 8% perf-profile.children.cycles-pp.prepare_exit_to_usermode 0.79 ± 18% -0.5 0.28 ± 22% perf-profile.children.cycles-pp.pick_next_task_fair 0.78 ± 15% -0.5 0.28 ± 9% perf-profile.children.cycles-pp.unmapped_area_topdown 0.57 ± 9% -0.4 0.18 ± 20% perf-profile.children.cycles-pp.select_task_rq_fair 0.72 ± 12% -0.4 0.34 ± 10% perf-profile.children.cycles-pp.mmap_region 0.49 ± 10% -0.3 0.14 ± 10% perf-profile.children.cycles-pp.rwsem_mark_wake 0.71 ± 11% -0.3 0.38 ± 8% perf-profile.children.cycles-pp._raw_spin_lock 1.45 ± 10% -0.3 1.11 ± 9% perf-profile.children.cycles-pp.tick_nohz_get_sleep_length 0.44 ± 15% -0.3 0.14 ± 15% perf-profile.children.cycles-pp.update_curr 0.38 ± 15% -0.3 0.08 ± 17% perf-profile.children.cycles-pp.load_new_mm_cr3 0.43 ± 17% -0.3 0.14 ± 11% perf-profile.children.cycles-pp.finish_task_switch 0.43 ± 19% -0.3 0.15 ± 10% perf-profile.children.cycles-pp.___perf_sw_event 0.41 ± 18% -0.3 0.14 ± 23% perf-profile.children.cycles-pp.set_next_entity 0.51 ± 13% -0.2 0.26 ± 8% perf-profile.children.cycles-pp.do_anonymous_page 1.12 ± 8% -0.2 0.87 ± 10% perf-profile.children.cycles-pp.tick_nohz_next_event 0.31 ± 14% -0.2 0.07 ± 30% perf-profile.children.cycles-pp.tick_nohz_idle_exit 0.82 ± 7% -0.2 0.59 ± 14% perf-profile.children.cycles-pp.get_next_timer_interrupt 0.40 ± 10% -0.2 0.18 ± 9% perf-profile.children.cycles-pp.vma_link 0.38 ± 11% -0.2 0.16 ± 5% perf-profile.children.cycles-pp.update_rq_clock 0.34 ± 13% -0.2 0.12 ± 6% perf-profile.children.cycles-pp.__update_load_avg_se 0.29 ± 13% -0.2 0.09 ± 9% perf-profile.children.cycles-pp.__switch_to 0.21 ± 5% -0.2 0.03 ±100% perf-profile.children.cycles-pp.__list_del_entry_valid 0.27 ± 13% -0.2 0.09 ± 14% perf-profile.children.cycles-pp.update_cfs_group 0.31 ± 9% -0.2 0.13 ± 5% perf-profile.children.cycles-pp.__update_load_avg_cfs_rq 0.25 ± 10% -0.2 0.07 ± 26% perf-profile.children.cycles-pp.update_cfs_rq_h_load 0.26 ± 19% -0.2 0.09 ± 17% perf-profile.children.cycles-pp.__unwind_start 0.54 ± 3% -0.2 0.37 ± 2% perf-profile.children.cycles-pp.native_irq_return_iret 0.22 ± 17% -0.2 0.06 ± 16% perf-profile.children.cycles-pp.__switch_to_asm 0.26 ± 15% -0.2 0.10 ± 10% perf-profile.children.cycles-pp.__perf_sw_event 0.34 ± 13% -0.2 0.18 ± 10% perf-profile.children.cycles-pp.down_read_trylock 0.31 ± 11% -0.1 0.17 ± 4% perf-profile.children.cycles-pp.up_read 0.25 ± 14% -0.1 0.11 ± 22% perf-profile.children.cycles-pp.nr_iowait_cpu 0.31 ± 9% -0.1 0.17 ± 14% perf-profile.children.cycles-pp.find_vma 0.26 ± 14% -0.1 0.12 ± 14% perf-profile.children.cycles-pp.update_ts_time_stats 0.46 ± 2% -0.1 0.32 ± 5% perf-profile.children.cycles-pp.native_write_msr 0.18 ± 24% -0.1 0.04 ± 58% perf-profile.children.cycles-pp.orc_find 0.20 ± 21% -0.1 0.07 ± 23% perf-profile.children.cycles-pp.__orc_find 0.34 ± 7% -0.1 0.21 ± 5% perf-profile.children.cycles-pp.sched_clock 0.33 ± 6% -0.1 0.20 ± 5% perf-profile.children.cycles-pp.native_sched_clock 0.30 ± 9% -0.1 0.17 ± 16% perf-profile.children.cycles-pp.down_read 0.19 ± 18% -0.1 0.06 ± 11% perf-profile.children.cycles-pp.ttwu_do_wakeup 0.34 ± 10% -0.1 0.22 ± 7% perf-profile.children.cycles-pp.sched_clock_cpu 0.51 ± 6% -0.1 0.40 ± 14% perf-profile.children.cycles-pp.__next_timer_interrupt 0.20 ± 11% -0.1 0.09 ± 9% perf-profile.children.cycles-pp.vma_interval_tree_insert 0.13 ± 13% -0.1 0.03 ±100% perf-profile.children.cycles-pp.pick_next_task_idle 0.27 ± 9% -0.1 0.17 ± 8% perf-profile.children.cycles-pp.__lock_text_start 0.15 ± 9% -0.1 0.06 ± 20% perf-profile.children.cycles-pp.stack_trace_consume_entry_nosched 0.18 ± 14% -0.1 0.10 ± 15% perf-profile.children.cycles-pp.vmacache_find 0.13 ± 28% -0.1 0.05 ± 63% perf-profile.children.cycles-pp.unwind_get_return_address 0.28 ± 15% -0.1 0.20 ± 10% perf-profile.children.cycles-pp.hrtimer_next_event_without 0.11 ± 33% -0.1 0.04 ±102% perf-profile.children.cycles-pp.__kernel_text_address 0.35 ± 6% -0.1 0.28 ± 9% perf-profile.children.cycles-pp.read_tsc 0.17 ± 13% -0.1 0.10 ± 18% perf-profile.children.cycles-pp.hrtimer_get_next_event 0.12 ± 18% -0.1 0.06 ± 16% perf-profile.children.cycles-pp.perf_event_mmap 0.09 ± 34% -0.1 0.03 ±102% perf-profile.children.cycles-pp.kernel_text_address 0.17 ± 16% -0.1 0.11 ± 9% perf-profile.children.cycles-pp.rcu_idle_exit 0.12 ± 10% -0.0 0.08 ± 13% perf-profile.children.cycles-pp.rcu_eqs_enter 0.10 ± 18% -0.0 0.06 ± 13% perf-profile.children.cycles-pp.rcu_eqs_exit 0.10 ± 18% -0.0 0.07 ± 15% perf-profile.children.cycles-pp.rcu_dynticks_eqs_exit 0.22 ± 5% +0.1 0.30 ± 15% perf-profile.children.cycles-pp.tick_irq_enter 0.13 ± 27% +0.1 0.22 ± 11% perf-profile.children.cycles-pp.rcu_sched_clock_irq 0.28 +0.1 0.38 ± 15% perf-profile.children.cycles-pp.irq_enter 0.71 ± 13% +0.2 0.95 ± 11% perf-profile.children.cycles-pp.update_process_times 0.78 ± 13% +0.2 1.03 ± 11% perf-profile.children.cycles-pp.tick_sched_handle 0.96 ± 12% +0.3 1.24 ± 10% perf-profile.children.cycles-pp.tick_sched_timer 1.59 ± 9% +0.4 1.97 ± 7% perf-profile.children.cycles-pp.__hrtimer_run_queues 1.57 ± 10% +0.7 2.31 ± 3% perf-profile.children.cycles-pp.rwsem_spin_on_owner 5.93 ± 2% +1.7 7.62 ± 12% perf-profile.children.cycles-pp.down_write_killable 5.86 ± 2% +1.7 7.59 ± 12% perf-profile.children.cycles-pp.rwsem_down_write_slowpath 19.30 ± 10% +16.0 35.35 ± 4% perf-profile.children.cycles-pp.page_fault 19.24 ± 10% +16.1 35.33 ± 4% perf-profile.children.cycles-pp.do_page_fault 19.09 ± 10% +16.2 35.27 ± 4% perf-profile.children.cycles-pp.__do_page_fault 15.78 ± 15% +17.7 33.52 ± 4% perf-profile.children.cycles-pp.rwsem_down_read_slowpath 12.57 ± 22% +20.8 33.40 ± 4% perf-profile.children.cycles-pp.osq_lock 15.96 ± 20% +23.3 39.22 ± 3% perf-profile.children.cycles-pp.rwsem_optimistic_spin 50.41 -7.7 42.72 ± 3% perf-profile.self.cycles-pp.intel_idle 1.90 ± 30% -1.4 0.50 ± 14% perf-profile.self.cycles-pp.native_queued_spin_lock_slowpath 1.62 ± 14% -1.2 0.40 ± 8% perf-profile.self.cycles-pp.switch_mm_irqs_off 0.79 ± 15% -0.5 0.26 ± 12% perf-profile.self.cycles-pp.__schedule 0.78 ± 15% -0.5 0.27 ± 8% perf-profile.self.cycles-pp.unmapped_area_topdown 0.81 ± 11% -0.4 0.36 ± 7% perf-profile.self.cycles-pp.prepare_exit_to_usermode 0.85 ± 11% -0.4 0.42 ± 8% perf-profile.self.cycles-pp.rwsem_down_read_slowpath 0.52 ± 11% -0.4 0.14 ± 9% perf-profile.self.cycles-pp.__account_scheduler_latency 0.53 ± 13% -0.4 0.15 ± 14% perf-profile.self.cycles-pp.unwind_next_frame 0.61 ± 8% -0.3 0.27 ± 7% perf-profile.self.cycles-pp._raw_spin_lock_irqsave 0.47 ± 10% -0.3 0.15 ± 10% perf-profile.self.cycles-pp.try_to_wake_up 0.66 ± 10% -0.3 0.34 ± 6% perf-profile.self.cycles-pp._raw_spin_lock 0.54 ± 15% -0.3 0.24 ± 14% perf-profile.self.cycles-pp.__handle_mm_fault 0.38 ± 15% -0.3 0.08 ± 17% perf-profile.self.cycles-pp.load_new_mm_cr3 0.40 ± 16% -0.3 0.13 ± 12% perf-profile.self.cycles-pp.finish_task_switch 0.37 ± 21% -0.2 0.13 ± 10% perf-profile.self.cycles-pp.___perf_sw_event 0.40 ± 13% -0.2 0.17 ± 5% perf-profile.self.cycles-pp.update_load_avg 1.16 ± 8% -0.2 0.94 ± 7% perf-profile.self.cycles-pp.menu_select 0.43 ± 11% -0.2 0.22 ± 8% perf-profile.self.cycles-pp.do_idle 0.28 ± 14% -0.2 0.08 ± 11% perf-profile.self.cycles-pp.enqueue_entity 0.32 ± 13% -0.2 0.12 ± 5% perf-profile.self.cycles-pp.__update_load_avg_se 0.28 ± 12% -0.2 0.09 ± 9% perf-profile.self.cycles-pp.__switch_to 0.28 ± 9% -0.2 0.09 ± 20% perf-profile.self.cycles-pp.select_task_rq_fair 0.24 ± 14% -0.2 0.06 ± 60% perf-profile.self.cycles-pp.enqueue_task_fair 0.21 ± 5% -0.2 0.03 ±100% perf-profile.self.cycles-pp.__list_del_entry_valid 0.41 ± 7% -0.2 0.23 ± 4% perf-profile.self.cycles-pp._raw_spin_lock_irq 0.27 ± 12% -0.2 0.09 ± 14% perf-profile.self.cycles-pp.update_cfs_group 0.25 ± 10% -0.2 0.07 ± 26% perf-profile.self.cycles-pp.update_cfs_rq_h_load 0.30 ± 9% -0.2 0.12 ± 6% perf-profile.self.cycles-pp.__update_load_avg_cfs_rq 0.54 ± 3% -0.2 0.37 ± 2% perf-profile.self.cycles-pp.native_irq_return_iret 0.25 ± 10% -0.2 0.08 ± 16% perf-profile.self.cycles-pp.update_curr 0.29 ± 11% -0.2 0.13 ± 6% perf-profile.self.cycles-pp.update_rq_clock 0.24 ± 18% -0.2 0.08 ± 22% perf-profile.self.cycles-pp.set_next_entity 0.28 ± 10% -0.2 0.12 ± 16% perf-profile.self.cycles-pp.swapgs_restore_regs_and_return_to_usermode 0.21 ± 5% -0.2 0.05 perf-profile.self.cycles-pp.stack_trace_save_tsk 0.22 ± 17% -0.2 0.06 ± 16% perf-profile.self.cycles-pp.__switch_to_asm 0.23 ± 19% -0.2 0.07 ± 19% perf-profile.self.cycles-pp.pick_next_task_fair 0.33 ± 13% -0.2 0.18 ± 10% perf-profile.self.cycles-pp.down_read_trylock 0.31 ± 11% -0.1 0.17 ± 3% perf-profile.self.cycles-pp.up_read 0.20 ± 11% -0.1 0.06 ± 14% perf-profile.self.cycles-pp.dequeue_entity 0.24 ± 15% -0.1 0.11 ± 19% perf-profile.self.cycles-pp.nr_iowait_cpu 0.46 ± 2% -0.1 0.32 ± 5% perf-profile.self.cycles-pp.native_write_msr 0.20 ± 20% -0.1 0.07 ± 23% perf-profile.self.cycles-pp.__orc_find 0.18 ± 23% -0.1 0.04 ± 58% perf-profile.self.cycles-pp.orc_find 0.31 ± 6% -0.1 0.18 ± 6% perf-profile.self.cycles-pp.native_sched_clock 0.27 ± 10% -0.1 0.15 ± 18% perf-profile.self.cycles-pp.down_read 0.20 ± 11% -0.1 0.08 ± 5% perf-profile.self.cycles-pp.vma_interval_tree_insert 0.14 ± 5% -0.1 0.04 ± 58% perf-profile.self.cycles-pp.rwsem_mark_wake 0.17 ± 8% -0.1 0.08 ± 8% perf-profile.self.cycles-pp.rwsem_down_write_slowpath 0.18 ± 14% -0.1 0.10 ± 15% perf-profile.self.cycles-pp.vmacache_find 0.15 ± 27% -0.1 0.08 ± 20% perf-profile.self.cycles-pp.__do_page_fault 0.34 ± 5% -0.1 0.27 ± 10% perf-profile.self.cycles-pp.read_tsc 0.22 ± 8% -0.1 0.15 ± 11% perf-profile.self.cycles-pp.__lock_text_start 0.28 ± 7% -0.1 0.21 ± 17% perf-profile.self.cycles-pp.find_next_bit 0.12 ± 6% -0.1 0.07 ± 10% perf-profile.self.cycles-pp.find_vma 0.09 ± 17% -0.1 0.04 ± 60% perf-profile.self.cycles-pp.handle_mm_fault 0.10 ± 15% -0.0 0.05 ± 8% perf-profile.self.cycles-pp.do_syscall_64 0.23 ± 7% -0.0 0.19 ± 11% perf-profile.self.cycles-pp.__next_timer_interrupt 0.07 ± 11% -0.0 0.03 ±102% perf-profile.self.cycles-pp.get_next_timer_interrupt 0.12 ± 12% -0.0 0.08 ± 8% perf-profile.self.cycles-pp.rcu_eqs_enter 0.10 ± 18% -0.0 0.07 ± 10% perf-profile.self.cycles-pp.rcu_dynticks_eqs_exit 0.12 ± 37% +0.0 0.15 ± 35% perf-profile.self.cycles-pp.update_blocked_averages 0.11 ± 26% +0.1 0.18 ± 16% perf-profile.self.cycles-pp.rcu_sched_clock_irq 1.13 ± 4% +0.2 1.36 ± 3% perf-profile.self.cycles-pp.rwsem_spin_on_owner 2.04 ± 23% +2.2 4.23 ± 8% perf-profile.self.cycles-pp.rwsem_optimistic_spin 12.51 ± 22% +20.7 33.24 ± 4% perf-profile.self.cycles-pp.osq_lock vm-scalability.median 12000 +-+-----------------------------------------------------------------+ |.+ :: +..+ :: +.+ :: + | 11000 +-++ : : +.. .+ : : : : : : : : : : + | 10000 +-+ + : : : + : .+ : : : : : : : : : : .. :| | + : : : .+.+.. .+. + : + : : + : : + :| 9000 +-+ :: +. + + :: :: | | + + + | 8000 +-+ | | | 7000 +-+ O O | 6000 +-+ O O | O O O O O O O O O O O | 5000 +-+ O | | O | 4000 +-+-----------------------------------------------------------------+ vm-scalability.time.user_time 220 +-+-------------------------------------------------------------------+ | | 200 +-+ + + + +.. + | 180 +-+ :: : + :: : :: + | | + : : .+ : +.. : : : + : : : : + | 160 +-+ + : : .+. : : : : : + : : : : + :| | + : + : .+.. .+. : + : : + : : :+ :| 140 +-+ :.. : .+ +.+. + : :: + | | + +. + + | 120 +-+ | 100 +-+ | | O O | 80 +-+ O O O O | O O O O O O O O | 60 +-+O-O----------------------O-----------------------------------------+ vm-scalability.time.voluntary_context_switches 4e+07 +-+---------------------------------------------------------------+ |.+ +..+ +..+ + + | 3.5e+07 +-+: + + : : + : : + :: : :| | : : : .. : + : : : + : : : + : : : :| 3e+07 +-+ : : : .+.+ : .. + .+..+. : :: + : :: + : :: | | + +. +.+ + + + + + + + | 2.5e+07 +-+ | | | 2e+07 +-+ | | | 1.5e+07 +-+ | | | 1e+07 +-+ O O O O | O O O O O O O O O O O O O | 5e+06 +-+---------------------------------------------------------------+ [*] 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, Rong Chen

10 months

1
0
0 / 0

[sched] b7d48fb89a: WARNING:at_kernel/sched/sched.h:#pick_next_task_fair

by kernel test robot

FYI, we noticed the following commit (built with gcc-7): commit: b7d48fb89aa672ee3acaac2864a3bfda81606ef1 ("sched: Add task_struct pointer to sched_class::set_curr_task") https://kernel.googlesource.com/pub/scm/linux/kernel/git/peterz/queue.git sched/core in testcase: rcutorture with following parameters: runtime: 300s test: cpuhotplug torture_type: srcud test-description: rcutorture is rcutorture kernel module load/unload test. test-url: https://www.kernel.org/doc/Documentation/RCU/torture.txt on test machine: qemu-system-x86_64 -enable-kvm -cpu SandyBridge -smp 2 -m 4G caused below changes (please refer to attached dmesg/kmsg for entire log/backtrace): +------------------------------------------------------+------------+------------+ | | 0f922d0d29 | b7d48fb89a | +------------------------------------------------------+------------+------------+ | boot_successes | 0 | 0 | | boot_failures | 8 | 8 | | BUG:kernel_reboot-without-warning_in_test_stage | 8 | | | WARNING:at_kernel/sched/sched.h:#pick_next_task_fair | 0 | 8 | | RIP:pick_next_task_fair | 0 | 8 | | WARNING:at_kernel/sched/sched.h:#sched_cpu_dying | 0 | 8 | | RIP:sched_cpu_dying | 0 | 8 | +------------------------------------------------------+------------+------------+ If you fix the issue, kindly add following tag Reported-by: kernel test robot <rong.a.chen(a)intel.com> [ 38.967800] WARNING: CPU: 1 PID: 15 at kernel/sched/sched.h:1754 pick_next_task_fair+0x758/0x780 [ 38.971162] Modules linked in: rcutorture torture intel_rapl_msr intel_rapl_common crct10dif_pclmul crc32_pclmul crc32c_intel bochs_drm ghash_clmulni_intel drm_vram_helper ppdev ttm snd_pcm snd_timer snd drm_kms_helper soundcore pcspkr drm joydev serio_raw parport_pc floppy parport qemu_fw_cfg ata_generic i2c_piix4 pata_acpi [ 38.978871] CPU: 1 PID: 15 Comm: migration/1 Not tainted 5.3.0-rc1-00087-gb7d48fb89aa672 #1 [ 38.980246] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.10.2-1 04/01/2014 [ 38.981631] RIP: 0010:pick_next_task_fair+0x758/0x780 [ 38.982468] Code: 48 0f af 85 18 01 00 00 48 83 c1 01 48 f7 f1 48 89 83 10 0a 00 00 e9 66 fa ff ff bf 02 00 00 00 e8 8d 3a ff ff e9 04 fa ff ff <0f> 0b e9 19 fb ff ff 80 3d 65 5e 61 01 00 0f 85 30 fc ff ff e8 bf [ 38.985513] RSP: 0000:ffffb98c8008bcf8 EFLAGS: 00010002 [ 38.986376] RAX: ffffffffa2011e00 RBX: ffff8b7ffbb2b3c0 RCX: ffffffffa264d540 [ 38.987544] RDX: ffffb98c8008bd80 RSI: 0000000000000001 RDI: ffff8b7ffbb2b3c0 [ 38.988728] RBP: ffffb98c8008bdc0 R08: 0000000912a90774 R09: 0000000000000004 [ 38.989913] R10: ffffb98c8008bd38 R11: 0000000000000003 R12: ffff8b7ffbb2b3c0 [ 38.991079] R13: ffffffffa264d540 R14: ffffb98c8008bd80 R15: 0000000000000000 [ 38.992251] FS: 0000000000000000(0000) GS:ffff8b7ffbb00000(0000) knlGS:0000000000000000 [ 38.993579] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 38.994522] CR2: 000000005788d734 CR3: 000000007e104000 CR4: 00000000000406e0 [ 38.995689] Call Trace: [ 38.996124] ? sched_cpu_starting+0xf0/0xf0 [ 38.996823] sched_cpu_dying+0x2a9/0x430 [ 38.997484] ? sched_cpu_starting+0xf0/0xf0 [ 38.998231] cpuhp_invoke_callback+0x86/0x5d0 [ 38.998957] ? cpu_disable_common+0x217/0x230 [ 38.999684] take_cpu_down+0x60/0xb0 [ 39.000283] multi_cpu_stop+0x6b/0x100 [ 39.000909] ? stop_machine_yield+0x10/0x10 [ 39.001612] cpu_stopper_thread+0x94/0x100 [ 39.002293] ? smpboot_thread_fn+0x2f/0x1e0 [ 39.002988] ? smpboot_thread_fn+0x74/0x1e0 [ 39.003683] ? smpboot_thread_fn+0x14e/0x1e0 [ 39.004393] smpboot_thread_fn+0x149/0x1e0 [ 39.005073] ? sort_range+0x20/0x20 [ 39.005662] kthread+0x11e/0x140 [ 39.006203] ? kthread_park+0xa0/0xa0 [ 39.006814] ret_from_fork+0x35/0x40 [ 39.007411] ---[ end trace baf171d5e73cc4a2 ]--- To reproduce: # build kernel cd linux cp config-5.3.0-rc1-00087-gb7d48fb89aa672 .config make HOSTCC=gcc-7 CC=gcc-7 ARCH=x86_64 olddefconfig prepare modules_prepare bzImage git clone https://github.com/intel/lkp-tests.git cd lkp-tests bin/lkp qemu -k <bzImage> job-script # job-script is attached in this email Thanks, Rong Chen

10 months

1
0
0 / 0

[sched/fair] ac1f97be8d: WARNING:at_kernel/sched/sched.h:#dl_server_init

by kernel test robot

FYI, we noticed the following commit (built with gcc-7): commit: ac1f97be8df98718a3d545cb6bb6ee4159e4300c ("sched/fair: Add trivial fair server") https://kernel.googlesource.com/pub/scm/linux/kernel/git/peterz/queue.git sched/wip-deadline in testcase: boot on test machine: qemu-system-x86_64 -enable-kvm -cpu SandyBridge -smp 2 -m 4G caused below changes (please refer to attached dmesg/kmsg for entire log/backtrace): +---------------------------------------------------------+------------+------------+ | | f4fa60e1b1 | ac1f97be8d | +---------------------------------------------------------+------------+------------+ | boot_successes | 5 | 0 | | boot_failures | 6 | 14 | | BUG:kernel_reboot-without-warning_in_test_stage | 6 | | | WARNING:at_kernel/sched/sched.h:#dl_server_init | 0 | 14 | | RIP:dl_server_init | 0 | 14 | | WARNING:at_kernel/sched/deadline.c:#task_non_contending | 0 | 3 | | RIP:task_non_contending | 0 | 3 | | kernel_BUG_at_kernel/sched/deadline.c | 0 | 3 | | invalid_opcode:#[##] | 0 | 3 | | RIP:enqueue_dl_entity | 0 | 3 | | RIP:default_idle | 0 | 2 | | Kernel_panic-not_syncing:Fatal_exception_in_interrupt | 0 | 2 | | Kernel_panic-not_syncing:Fatal_exception | 0 | 1 | +---------------------------------------------------------+------------+------------+ If you fix the issue, kindly add following tag Reported-by: kernel test robot <lkp(a)intel.com> [ 3.215886] WARNING: CPU: 0 PID: 0 at kernel/sched/sched.h:1108 dl_server_init+0x68/0xca [ 3.237973] Modules linked in: [ 3.242573] CPU: 0 PID: 0 Comm: swapper Not tainted 5.3.0-rc1-00388-gac1f97be8df987 #4 [ 3.254600] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.10.2-1 04/01/2014 [ 3.267232] RIP: 0010:dl_server_init+0x68/0xca [ 3.274057] Code: 0b 83 bd 20 0a 00 00 01 4c 8b 63 48 77 1e 80 3d 3b 6e 6b 01 00 75 15 48 c7 c7 93 24 34 82 c6 05 2b 6e 6b 01 01 e8 d1 38 fc ff <0f> 0b 4c 39 a5 28 0a 00 00 79 0e 48 c7 c7 45 d6 32 82 e8 32 5f 01 [ 3.302613] RSP: 0000:ffffffff82603eb8 EFLAGS: 00010082 [ 3.310703] RAX: 0000000000000000 RBX: ffff88813fc2a2e8 RCX: 0000000000000000 [ 3.321660] RDX: 0000000000000026 RSI: ffffffff82c4afa6 RDI: ffffffff82c4b3a6 [ 3.332536] RBP: ffff88813fc29a00 R08: ffffffff82c4af80 R09: 0000000000000026 [ 3.343234] R10: 0000000000000000 R11: ffffffff82603ccf R12: 0000000000000000 [ 3.354133] R13: 0000000000029a00 R14: ffff88813fc29bc0 R15: ffff88813fc29a40 [ 3.365277] FS: 0000000000000000(0000) GS:ffff88813fc00000(0000) knlGS:0000000000000000 [ 3.377714] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 3.386474] CR2: ffff88813ffff000 CR3: 0000000002612000 CR4: 00000000000006b0 [ 3.397423] Call Trace: [ 3.401223] sched_init+0x37e/0x3fe [ 3.406629] start_kernel+0x267/0x50a [ 3.412198] ? x86_family+0x5/0x1d [ 3.417442] secondary_startup_64+0xb6/0xc0 [ 3.423800] random: get_random_bytes called from init_oops_id+0x22/0x31 with crng_init=0 [ 3.436167] ---[ end trace b2352b5ced6eae1c ]--- To reproduce: # build kernel cd linux cp config-5.3.0-rc1-00388-gac1f97be8df987 .config make HOSTCC=gcc-7 CC=gcc-7 ARCH=x86_64 olddefconfig prepare modules_prepare bzImage git clone https://github.com/intel/lkp-tests.git cd lkp-tests bin/lkp qemu -k <bzImage> job-script # job-script is attached in this email Thanks, lkp

10 months

1
0
0 / 0

[compat_ioctl] 97ca70e16d: mdadm-selftests.enchmarks/mdadm-selftests/tests/06name.fail

by kernel test robot

FYI, we noticed the following commit (built with gcc-7): commit: 97ca70e16dc76bad69ad774cf0bd0286c97d63b4 ("compat_ioctl: scsi: move ioctl handling into drivers") https://kernel.googlesource.com/pub/scm/linux/kernel/git/arnd/playground.git to-build in testcase: mdadm-selftests with following parameters: disk: 1HDD test_prefix: 06 on test machine: qemu-system-x86_64 -enable-kvm -cpu SandyBridge -smp 2 -m 4G caused below changes (please refer to attached dmesg/kmsg for entire log/backtrace): If you fix the issue, kindly add following tag Reported-by: kernel test robot <rong.a.chen(a)intel.com> Welcome to fdisk (util-linux 2.29.2). Changes will remain in memory only, until you decide to write them. Be careful before using the write command. Device does not contain a recognized partition table. Created a new DOS disklabel with disk identifier 0x8b600b10. Command (m for help): Partition type p primary (0 primary, 0 extended, 4 free) e extended (container for logical partitions) Select (default p): Partition number (1-4, default 1): First sector (2048-536870911, default 2048): Last sector, +sectors or +size{K,M,G,T,P} (2048-536870911, default 536870911): Created a new partition 1 of type 'Linux' and of size 5 GiB. Command (m for help): The partition table has been altered. Calling ioctl() to re-read partition table. Syncing disks. 2019-07-28 07:22:06 mkdir -p /var/tmp 2019-07-28 07:22:06 mke2fs -t ext3 -b 1024 -J size=1 -q /dev/vda1 2019-07-28 07:22:08 mount -t ext3 /dev/vda1 /var/tmp sed -e 's/{DEFAULT_METADATA}/1.2/g' \ -e 's,{MAP_PATH},/run/mdadm/map,g' mdadm.8.in > mdadm.8 /usr/bin/install -D -m 644 mdadm.8 /usr/share/man/man8/mdadm.8 /usr/bin/install -D -m 644 mdmon.8 /usr/share/man/man8/mdmon.8 /usr/bin/install -D -m 644 md.4 /usr/share/man/man4/md.4 /usr/bin/install -D -m 644 mdadm.conf.5 /usr/share/man/man5/mdadm.conf.5 /usr/bin/install -D -m 644 udev-md-raid-creating.rules /lib/udev/rules.d/01-md-raid-creating.rules /usr/bin/install -D -m 644 udev-md-raid-arrays.rules /lib/udev/rules.d/63-md-raid-arrays.rules /usr/bin/install -D -m 644 udev-md-raid-assembly.rules /lib/udev/rules.d/64-md-raid-assembly.rules /usr/bin/install -D -m 644 udev-md-clustered-confirm-device.rules /lib/udev/rules.d/69-md-clustered-confirm-device.rules /usr/bin/install -D -m 755 mdadm /sbin/mdadm /usr/bin/install -D -m 755 mdmon /sbin/mdmon Testing on linux-5.3.0-rc1-g97ca70e16dc76b kernel /lkp/benchmarks/mdadm-selftests/tests/06name... FAILED - see /var/tmp/06name.log and /var/tmp/fail06name.log for details 06name TIMEOUT Testing on linux-5.3.0-rc1-g97ca70e16dc76b kernel /lkp/benchmarks/mdadm-selftests/tests/06sysfs... succeeded 06sysfs TIMEOUT Testing on linux-5.3.0-rc1-g97ca70e16dc76b kernel /lkp/benchmarks/mdadm-selftests/tests/06wrmostly... FAILED - see /var/tmp/06wrmostly.log and /var/tmp/fail06wrmostly.log for details 06wrmostly TIMEOUT To reproduce: # build kernel cd linux cp config-5.3.0-rc1-g97ca70e16dc76b .config make HOSTCC=gcc-7 CC=gcc-7 ARCH=x86_64 olddefconfig prepare modules_prepare bzImage modules make HOSTCC=gcc-7 CC=gcc-7 ARCH=x86_64 INSTALL_MOD_PATH=<mod-install-dir> modules_install cd <mod-install-dir> find lib/ | cpio -o -H newc --quiet | gzip > modules.cgz git clone https://github.com/intel/lkp-tests.git cd lkp-tests bin/lkp qemu -k <bzImage> -m modules.cgz job-script # job-script is attached in this email Thanks, Rong Chen

10 months

4
3
0 / 0

[mm] 6471384af2: kernel_BUG_at_mm/slub.c

by kernel test robot

FYI, we noticed the following commit (built with gcc-7): commit: 6471384af2a6530696fc0203bafe4de41a23c9ef ("mm: security: introduce init_on_alloc=1 and init_on_free=1 boot options") https://kernel.googlesource.com/pub/scm/linux/kernel/git/torvalds/linux.git master in testcase: trinity with following parameters: runtime: 300s test-description: Trinity is a linux system call fuzz tester. test-url: http://codemonkey.org.uk/projects/trinity/ on test machine: qemu-system-x86_64 -enable-kvm -cpu SandyBridge -smp 2 -m 8G caused below changes (please refer to attached dmesg/kmsg for entire log/backtrace): +------------------------------------------+------------+------------+ | | ba5c5e4a5d | 6471384af2 | +------------------------------------------+------------+------------+ | boot_successes | 8 | 0 | | boot_failures | 2 | 15 | | invoked_oom-killer:gfp_mask=0x | 1 | | | Mem-Info | 1 | | | kernel_BUG_at_security/keys/keyring.c | 1 | | | invalid_opcode:#[##] | 1 | 15 | | RIP:__key_link_begin | 1 | | | Kernel_panic-not_syncing:Fatal_exception | 1 | 15 | | kernel_BUG_at_mm/slub.c | 0 | 15 | | RIP:kfree | 0 | 15 | +------------------------------------------+------------+------------+ If you fix the issue, kindly add following tag Reported-by: kernel test robot <rong.a.chen(a)intel.com> [ 4.478342] kernel BUG at mm/slub.c:306! [ 4.482437] invalid opcode: 0000 [#1] PREEMPT PTI [ 4.485750] CPU: 0 PID: 0 Comm: swapper Not tainted 5.2.0-05754-g6471384a #4 [ 4.490635] RIP: 0010:kfree+0x58a/0x5c0 [ 4.493679] Code: 48 83 05 78 37 51 02 01 0f 0b 48 83 05 7e 37 51 02 01 48 83 05 7e 37 51 02 01 48 83 05 7e 37 51 02 01 48 83 05 d6 37 51 02 01 <0f> 0b 48 83 05 d4 37 51 02 01 48 83 05 d4 37 51 02 01 48 83 05 d4 [ 4.506827] RSP: 0000:ffffffff82603d90 EFLAGS: 00010002 [ 4.510475] RAX: ffff8c3976c04320 RBX: ffff8c3976c04300 RCX: 0000000000000000 [ 4.515420] RDX: ffff8c3976c04300 RSI: 0000000000000000 RDI: ffff8c3976c04320 [ 4.520331] RBP: ffffffff82603db8 R08: 0000000000000000 R09: 0000000000000000 [ 4.525288] R10: ffff8c3976c04320 R11: ffffffff8289e1e0 R12: ffffd52cc8db0100 [ 4.530180] R13: ffff8c3976c01a00 R14: ffffffff810f10d4 R15: ffff8c3976c04300 [ 4.535079] FS: 0000000000000000(0000) GS:ffffffff8266b000(0000) knlGS:0000000000000000 [ 4.540628] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 4.544593] CR2: ffff8c397ffff000 CR3: 0000000125020000 CR4: 00000000000406b0 [ 4.549558] Call Trace: [ 4.551266] apply_wqattrs_prepare+0x154/0x280 [ 4.554357] apply_workqueue_attrs_locked+0x4e/0xe0 [ 4.557728] apply_workqueue_attrs+0x36/0x60 [ 4.560654] alloc_workqueue+0x25a/0x6d0 [ 4.563381] ? kmem_cache_alloc_trace+0x1e3/0x500 [ 4.566628] ? __mutex_unlock_slowpath+0x44/0x3f0 [ 4.569875] workqueue_init_early+0x246/0x348 [ 4.573025] start_kernel+0x3c7/0x7ec [ 4.575558] x86_64_start_reservations+0x40/0x49 [ 4.578738] x86_64_start_kernel+0xda/0xe4 [ 4.581600] secondary_startup_64+0xb6/0xc0 [ 4.584473] Modules linked in: [ 4.586620] ---[ end trace f67eb9af4d8d492b ]--- To reproduce: # build kernel cd linux cp config-5.2.0-05754-g6471384a .config make HOSTCC=gcc-7 CC=gcc-7 ARCH=x86_64 olddefconfig prepare modules_prepare bzImage git clone https://github.com/intel/lkp-tests.git cd lkp-tests bin/lkp qemu -k <bzImage> job-script # job-script is attached in this email Thanks, Rong Chen

10 months

6
7
0 / 0

Re: [LKP] [drm/mgag200] 90f479ae51: vm-scalability.median -18.8% regression

by Rong Chen

Hi, On 7/31/19 6:21 PM, Michel Dänzer wrote: > On 2019-07-31 11:25 a.m., Huang, Ying wrote: >> Hi, Daniel, >> >> Daniel Vetter <daniel(a)ffwll.ch> writes: >> >>> On Tue, Jul 30, 2019 at 10:27 PM Dave Airlie <airlied(a)gmail.com> wrote: >>>> On Wed, 31 Jul 2019 at 05:00, Daniel Vetter <daniel(a)ffwll.ch> wrote: >>>>> On Tue, Jul 30, 2019 at 8:50 PM Thomas Zimmermann <tzimmermann(a)suse.de> wrote: >>>>>> Hi >>>>>> >>>>>> Am 30.07.19 um 20:12 schrieb Daniel Vetter: >>>>>>> On Tue, Jul 30, 2019 at 7:50 PM Thomas Zimmermann <tzimmermann(a)suse.de> wrote: >>>>>>>> Am 29.07.19 um 11:51 schrieb kernel test robot: >>>>>>>>> Greeting, >>>>>>>>> >>>>>>>>> FYI, we noticed a -18.8% regression of vm-scalability.median due to commit:> >>>>>>>>> >>>>>>>>> commit: 90f479ae51afa45efab97afdde9b94b9660dd3e4 ("drm/mgag200: Replace struct mga_fbdev with generic framebuffer emulation") >>>>>>>>> https://kernel.googlesource.com/pub/scm/linux/kernel/git/next/linux-next.git master >>>>>>>> Daniel, Noralf, we may have to revert this patch. >>>>>>>> >>>>>>>> I expected some change in display performance, but not in VM. Since it's >>>>>>>> a server chipset, probably no one cares much about display performance. >>>>>>>> So that seemed like a good trade-off for re-using shared code. >>>>>>>> >>>>>>>> Part of the patch set is that the generic fb emulation now maps and >>>>>>>> unmaps the fbdev BO when updating the screen. I guess that's the cause >>>>>>>> of the performance regression. And it should be visible with other >>>>>>>> drivers as well if they use a shadow FB for fbdev emulation. >>>>>>> For fbcon we should need to do any maps/unamps at all, this is for the >>>>>>> fbdev mmap support only. If the testcase mentioned here tests fbdev >>>>>>> mmap handling it's pretty badly misnamed :-) And as long as you don't >>>>>>> have an fbdev mmap there shouldn't be any impact at all. >>>>>> The ast and mgag200 have only a few MiB of VRAM, so we have to get the >>>>>> fbdev BO out if it's not being displayed. If not being mapped, it can be >>>>>> evicted and make room for X, etc. >>>>>> >>>>>> To make this work, the BO's memory is mapped and unmapped in >>>>>> drm_fb_helper_dirty_work() before being updated from the shadow FB. [1] >>>>>> That fbdev mapping is established on each screen update, more or less. >>>>>> From my (yet unverified) understanding, this causes the performance >>>>>> regression in the VM code. >>>>>> >>>>>> The original code in mgag200 used to kmap the fbdev BO while it's being >>>>>> displayed; [2] and the drawing code only mapped it when necessary (i.e., >>>>>> not being display). [3] >>>>> Hm yeah, this vmap/vunmap is going to be pretty bad. We indeed should >>>>> cache this. >>>>> >>>>>> I think this could be added for VRAM helpers as well, but it's still a >>>>>> workaround and non-VRAM drivers might also run into such a performance >>>>>> regression if they use the fbdev's shadow fb. >>>>> Yeah agreed, fbdev emulation should try to cache the vmap. >>>>> >>>>>> Noralf mentioned that there are plans for other DRM clients besides the >>>>>> console. They would as well run into similar problems. >>>>>> >>>>>>>> The thing is that we'd need another generic fbdev emulation for ast and >>>>>>>> mgag200 that handles this issue properly. >>>>>>> Yeah I dont think we want to jump the gun here. If you can try to >>>>>>> repro locally and profile where we're wasting cpu time I hope that >>>>>>> should sched a light what's going wrong here. >>>>>> I don't have much time ATM and I'm not even officially at work until >>>>>> late Aug. I'd send you the revert and investigate later. I agree that >>>>>> using generic fbdev emulation would be preferable. >>>>> Still not sure that's the right thing to do really. Yes it's a >>>>> regression, but vm testcases shouldn run a single line of fbcon or drm >>>>> code. So why this is impacted so heavily by a silly drm change is very >>>>> confusing to me. We might be papering over a deeper and much more >>>>> serious issue ... >>>> It's a regression, the right thing is to revert first and then work >>>> out the right thing to do. >>> Sure, but I have no idea whether the testcase is doing something >>> reasonable. If it's accidentally testing vm scalability of fbdev and >>> there's no one else doing something this pointless, then it's not a >>> real bug. Plus I think we're shooting the messenger here. >>> >>>> It's likely the test runs on the console and printfs stuff out while running. >>> But why did we not regress the world if a few prints on the console >>> have such a huge impact? We didn't get an entire stream of mails about >>> breaking stuff ... >> The regression seems not related to the commit. But we have retested >> and confirmed the regression. Hard to understand what happens. > Does the regressed test cause any output on console while it's > measuring? If so, it's probably accidentally measuring fbcon/DRM code in > addition to the workload it's trying to measure. > Sorry, I'm not familiar with DRM, we enabled the console to output logs, and attached please find the log file. "Command line: ... console=tty0 earlyprintk=ttyS0,115200 console=ttyS0,115200 vga=normal rw" Best Regards, Rong Chen

10 months

3
8
0 / 0

Re: [LKP] [drm/mgag200] 90f479ae51: vm-scalability.median -18.8% regression

by Thomas Zimmermann

Hi Am 01.08.19 um 15:30 schrieb Michel Dänzer: > On 2019-08-01 8:19 a.m., Rong Chen wrote: >> Hi, >> >> On 7/31/19 6:21 PM, Michel Dänzer wrote: >>> On 2019-07-31 11:25 a.m., Huang, Ying wrote: >>>> Hi, Daniel, >>>> >>>> Daniel Vetter <daniel(a)ffwll.ch> writes: >>>> >>>>> On Tue, Jul 30, 2019 at 10:27 PM Dave Airlie <airlied(a)gmail.com> wrote: >>>>>> On Wed, 31 Jul 2019 at 05:00, Daniel Vetter <daniel(a)ffwll.ch> wrote: >>>>>>> On Tue, Jul 30, 2019 at 8:50 PM Thomas Zimmermann >>>>>>> <tzimmermann(a)suse.de> wrote: >>>>>>>> Hi >>>>>>>> >>>>>>>> Am 30.07.19 um 20:12 schrieb Daniel Vetter: >>>>>>>>> On Tue, Jul 30, 2019 at 7:50 PM Thomas Zimmermann >>>>>>>>> <tzimmermann(a)suse.de> wrote: >>>>>>>>>> Am 29.07.19 um 11:51 schrieb kernel test robot: >>>>>>>>>>> Greeting, >>>>>>>>>>> >>>>>>>>>>> FYI, we noticed a -18.8% regression of vm-scalability.median >>>>>>>>>>> due to commit:> >>>>>>>>>>> >>>>>>>>>>> commit: 90f479ae51afa45efab97afdde9b94b9660dd3e4 >>>>>>>>>>> ("drm/mgag200: Replace struct mga_fbdev with generic >>>>>>>>>>> framebuffer emulation") >>>>>>>>>>> https://kernel.googlesource.com/pub/scm/linux/kernel/git/next/linux-next.git >>>>>>>>>>> master >>>>>>>>>> Daniel, Noralf, we may have to revert this patch. >>>>>>>>>> >>>>>>>>>> I expected some change in display performance, but not in VM. >>>>>>>>>> Since it's >>>>>>>>>> a server chipset, probably no one cares much about display >>>>>>>>>> performance. >>>>>>>>>> So that seemed like a good trade-off for re-using shared code. >>>>>>>>>> >>>>>>>>>> Part of the patch set is that the generic fb emulation now maps >>>>>>>>>> and >>>>>>>>>> unmaps the fbdev BO when updating the screen. I guess that's >>>>>>>>>> the cause >>>>>>>>>> of the performance regression. And it should be visible with other >>>>>>>>>> drivers as well if they use a shadow FB for fbdev emulation. >>>>>>>>> For fbcon we should need to do any maps/unamps at all, this is >>>>>>>>> for the >>>>>>>>> fbdev mmap support only. If the testcase mentioned here tests fbdev >>>>>>>>> mmap handling it's pretty badly misnamed :-) And as long as you >>>>>>>>> don't >>>>>>>>> have an fbdev mmap there shouldn't be any impact at all. >>>>>>>> The ast and mgag200 have only a few MiB of VRAM, so we have to >>>>>>>> get the >>>>>>>> fbdev BO out if it's not being displayed. If not being mapped, it >>>>>>>> can be >>>>>>>> evicted and make room for X, etc. >>>>>>>> >>>>>>>> To make this work, the BO's memory is mapped and unmapped in >>>>>>>> drm_fb_helper_dirty_work() before being updated from the shadow >>>>>>>> FB. [1] >>>>>>>> That fbdev mapping is established on each screen update, more or >>>>>>>> less. >>>>>>>> From my (yet unverified) understanding, this causes the performance >>>>>>>> regression in the VM code. >>>>>>>> >>>>>>>> The original code in mgag200 used to kmap the fbdev BO while it's >>>>>>>> being >>>>>>>> displayed; [2] and the drawing code only mapped it when necessary >>>>>>>> (i.e., >>>>>>>> not being display). [3] >>>>>>> Hm yeah, this vmap/vunmap is going to be pretty bad. We indeed should >>>>>>> cache this. >>>>>>> >>>>>>>> I think this could be added for VRAM helpers as well, but it's >>>>>>>> still a >>>>>>>> workaround and non-VRAM drivers might also run into such a >>>>>>>> performance >>>>>>>> regression if they use the fbdev's shadow fb. >>>>>>> Yeah agreed, fbdev emulation should try to cache the vmap. >>>>>>> >>>>>>>> Noralf mentioned that there are plans for other DRM clients >>>>>>>> besides the >>>>>>>> console. They would as well run into similar problems. >>>>>>>> >>>>>>>>>> The thing is that we'd need another generic fbdev emulation for >>>>>>>>>> ast and >>>>>>>>>> mgag200 that handles this issue properly. >>>>>>>>> Yeah I dont think we want to jump the gun here. If you can try to >>>>>>>>> repro locally and profile where we're wasting cpu time I hope that >>>>>>>>> should sched a light what's going wrong here. >>>>>>>> I don't have much time ATM and I'm not even officially at work until >>>>>>>> late Aug. I'd send you the revert and investigate later. I agree >>>>>>>> that >>>>>>>> using generic fbdev emulation would be preferable. >>>>>>> Still not sure that's the right thing to do really. Yes it's a >>>>>>> regression, but vm testcases shouldn run a single line of fbcon or >>>>>>> drm >>>>>>> code. So why this is impacted so heavily by a silly drm change is >>>>>>> very >>>>>>> confusing to me. We might be papering over a deeper and much more >>>>>>> serious issue ... >>>>>> It's a regression, the right thing is to revert first and then work >>>>>> out the right thing to do. >>>>> Sure, but I have no idea whether the testcase is doing something >>>>> reasonable. If it's accidentally testing vm scalability of fbdev and >>>>> there's no one else doing something this pointless, then it's not a >>>>> real bug. Plus I think we're shooting the messenger here. >>>>> >>>>>> It's likely the test runs on the console and printfs stuff out >>>>>> while running. >>>>> But why did we not regress the world if a few prints on the console >>>>> have such a huge impact? We didn't get an entire stream of mails about >>>>> breaking stuff ... >>>> The regression seems not related to the commit. But we have retested >>>> and confirmed the regression. Hard to understand what happens. >>> Does the regressed test cause any output on console while it's >>> measuring? If so, it's probably accidentally measuring fbcon/DRM code in >>> addition to the workload it's trying to measure. >>> >> >> Sorry, I'm not familiar with DRM, we enabled the console to output logs, >> and attached please find the log file. >> >> "Command line: ... console=tty0 earlyprintk=ttyS0,115200 >> console=ttyS0,115200 vga=normal rw" > > I assume the > > user :notice: [ xxx.xxxx] xxxxxxxxx bytes / xxxxxxx usecs = xxxxx KB/s > > lines are generated by the test? > > If so, unless the test is intended to measure console performance, it > should be fixed not to generate output to console (while it's measuring). Yes, the test prints quite a lot of text to the console. It shouldn't do that. Best regards Thomas > > -- Thomas Zimmermann Graphics Driver Developer SUSE Linux GmbH, Maxfeldstrasse 5, 90409 Nuernberg, Germany GF: Felix Imendörffer, Mary Higgins, Sri Rasiah HRB 21284 (AG Nürnberg)

10 months

1
0
0 / 0

2020

2019

2018

2017

2016

2015

2014

2013

LKP August 2019