Messages in this thread | | | Subject | [LKP] [locking/rwsem] 1a99367023f: no primary result change, +23.6% will-it-scale.time.system_time | From | Huang Ying <> | Date | Thu, 12 Mar 2015 14:35:40 +0800 |
| |
FYI, we noticed the below changes on
git://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git master commit 1a99367023f6ac664365a37fa508b059e31d0e88 ("locking/rwsem: Check for active lock before bailing on spinning")
There is some minor will-it-scale.per_thread_ops changes below (-1.8%), but not stable enough during bisect. So in general, there is no user visible change, just more system time.
testbox/testcase/testparams: ivb42/will-it-scale/performance-brk1
b3fd4f03ca0b9952 1a99367023f6ac664365a37fa5 ---------------- -------------------------- fail:runs %reproduction fail:runs | | | :4 25% 1:4 kmsg.Spurious_LAPIC_timer_interrupt_on_cpu %stddev %change %stddev \ | \ 308 ± 3% +23.6% 381 ± 1% will-it-scale.time.system_time 99 ± 3% +20.2% 119 ± 0% will-it-scale.time.percent_of_cpu_this_job_got 34098838 ± 1% +6.0% 36159517 ± 2% will-it-scale.time.voluntary_context_switches 314 ± 0% +2.5% 322 ± 0% will-it-scale.time.elapsed_time 314 ± 0% +2.5% 322 ± 0% will-it-scale.time.elapsed_time.max 0.61 ± 20% +428.8% 3.21 ± 5% perf-profile.cpu-cycles.osq_lock.rwsem_down_write_failed.call_rwsem_down_write_failed.sys_brk.system_call_fastpath 0.39 ± 23% +127.3% 0.88 ± 14% perf-profile.cpu-cycles.osq_unlock.rwsem_down_write_failed.call_rwsem_down_write_failed.sys_brk.system_call_fastpath 991202 ± 25% -47.8% 517752 ± 41% sched_debug.cpu#5.sched_count 481295 ± 25% -48.0% 250449 ± 42% sched_debug.cpu#5.sched_goidle 963157 ± 25% -47.9% 501898 ± 42% sched_debug.cpu#5.nr_switches 5.03 ± 16% +133.3% 11.73 ± 7% perf-profile.cpu-cycles.rwsem_spin_on_owner.rwsem_down_write_failed.call_rwsem_down_write_failed.sys_brk.system_call_fastpath 185603 ± 45% +99.3% 369978 ± 34% sched_debug.cpu#9.ttwu_count 17 ± 20% +75.0% 29 ± 35% sched_debug.cfs_rq[33]:/.load 1.07 ± 13% +88.8% 2.02 ± 17% perf-profile.cpu-cycles.__kernel_text_address.print_context_stack.dump_trace.save_stack_trace_tsk.__account_scheduler_latency 2.41 ± 9% +92.7% 4.64 ± 12% perf-profile.cpu-cycles._raw_spin_lock_irqsave.rwsem_wake.call_rwsem_wake.sys_brk.system_call_fastpath 1201 ± 30% -45.8% 651 ± 21% cpuidle.C3-IVT.usage 1.92 ± 3% -39.8% 1.16 ± 19% perf-profile.cpu-cycles._raw_spin_lock.try_to_wake_up.wake_up_process.__rwsem_do_wake.rwsem_wake 1.10 ± 10% +93.6% 2.12 ± 5% perf-profile.cpu-cycles.up_write.vma_adjust.vma_merge.do_brk.sys_brk 6 ± 17% +92.3% 12 ± 22% sched_debug.cfs_rq[6]:/.runnable_load_avg 2.02 ± 13% +95.2% 3.94 ± 20% perf-profile.cpu-cycles.print_context_stack.dump_trace.save_stack_trace_tsk.__account_scheduler_latency.enqueue_entity 6 ± 36% +52.0% 9 ± 17% sched_debug.cpu#6.cpu_load[2] 2.63 ± 14% +95.0% 5.13 ± 18% perf-profile.cpu-cycles.__account_scheduler_latency.enqueue_entity.enqueue_task_fair.enqueue_task.activate_task 5 ± 20% +66.7% 8 ± 9% sched_debug.cpu#6.cpu_load[3] 2.41 ± 14% +93.1% 4.66 ± 19% perf-profile.cpu-cycles.save_stack_trace_tsk.__account_scheduler_latency.enqueue_entity.enqueue_task_fair.enqueue_task 2.34 ± 14% +94.5% 4.55 ± 18% perf-profile.cpu-cycles.dump_trace.save_stack_trace_tsk.__account_scheduler_latency.enqueue_entity.enqueue_task_fair 0.96 ± 13% +71.0% 1.65 ± 13% perf-profile.cpu-cycles.find_vma.sys_brk.system_call_fastpath 17462 ± 4% +15.1% 20096 ± 6% sched_debug.cfs_rq[4]:/.exec_clock 82 ± 24% +116.2% 177 ± 46% sched_debug.cfs_rq[27]:/.tg_load_contrib 155743 ± 31% +81.1% 281980 ± 34% sched_debug.cpu#14.sched_count 13.98 ± 6% +63.6% 22.87 ± 3% perf-profile.cpu-cycles.call_rwsem_down_write_failed.sys_brk.system_call_fastpath 13.94 ± 6% +63.5% 22.78 ± 3% perf-profile.cpu-cycles.rwsem_down_write_failed.call_rwsem_down_write_failed.sys_brk.system_call_fastpath 702 ± 12% -39.2% 427 ± 16% cpuidle.C1E-IVT.usage 103116 ± 29% +82.1% 187794 ± 22% sched_debug.cpu#41.sched_goidle 206574 ± 29% +82.0% 375906 ± 22% sched_debug.cpu#41.nr_switches 214754 ± 29% +79.4% 385314 ± 22% sched_debug.cpu#41.sched_count 5 ± 8% +52.4% 8 ± 8% sched_debug.cpu#6.cpu_load[4] 67108 ± 40% +86.7% 125260 ± 35% sched_debug.cpu#14.sched_goidle 134740 ± 40% +86.4% 251133 ± 35% sched_debug.cpu#14.nr_switches 1.42 ± 8% -33.0% 0.95 ± 6% perf-profile.cpu-cycles.cpuidle_select.cpu_startup_entry.start_secondary 1.27 ± 6% -34.1% 0.83 ± 7% perf-profile.cpu-cycles.menu_select.cpuidle_select.cpu_startup_entry.start_secondary 1.28 ± 7% +44.9% 1.85 ± 8% perf-profile.cpu-cycles.find_vma.do_munmap.sys_brk.system_call_fastpath 2.69 ± 4% +36.6% 3.68 ± 4% perf-profile.cpu-cycles.vma_adjust.vma_merge.do_brk.sys_brk.system_call_fastpath 40423 ± 2% +11.6% 45108 ± 3% sched_debug.cpu#6.nr_load_updates 1.24 ± 7% -31.6% 0.85 ± 11% perf-profile.cpu-cycles.check_preempt_curr.ttwu_do_wakeup.ttwu_do_activate.try_to_wake_up.wake_up_process 5.67 ± 5% -29.9% 3.98 ± 7% perf-profile.cpu-cycles.perf_event_mmap.do_brk.sys_brk.system_call_fastpath 3.09 ± 1% -30.0% 2.16 ± 3% perf-profile.cpu-cycles.dequeue_task_fair.dequeue_task.deactivate_task.__sched_text_start.schedule 2.08 ± 7% -29.9% 1.46 ± 12% perf-profile.cpu-cycles.ttwu_do_wakeup.ttwu_do_activate.try_to_wake_up.wake_up_process.__rwsem_do_wake 2.52 ± 1% -30.1% 1.76 ± 3% perf-profile.cpu-cycles.dequeue_entity.dequeue_task_fair.dequeue_task.deactivate_task.__sched_text_start 3.20 ± 2% -28.8% 2.28 ± 4% perf-profile.cpu-cycles.pick_next_task_fair.__sched_text_start.schedule.schedule_preempt_disabled.cpu_startup_entry 10.55 ± 6% -24.9% 7.92 ± 4% perf-profile.cpu-cycles.cpuidle_enter_state.cpuidle_enter.cpu_startup_entry.start_secondary 1.16 ± 2% -30.1% 0.81 ± 9% perf-profile.cpu-cycles.free_pgtables.unmap_region.do_munmap.sys_brk.system_call_fastpath 1.10 ± 15% -28.9% 0.78 ± 5% perf-profile.cpu-cycles.unmap_page_range.unmap_single_vma.unmap_vmas.unmap_region.do_munmap 18033 ± 2% +10.5% 19932 ± 6% sched_debug.cfs_rq[6]:/.exec_clock 5.61 ± 1% -27.5% 4.07 ± 4% perf-profile.cpu-cycles.__sched_text_start.schedule.schedule_preempt_disabled.cpu_startup_entry.start_secondary 5.74 ± 2% -27.5% 4.16 ± 5% perf-profile.cpu-cycles.schedule_preempt_disabled.cpu_startup_entry.start_secondary 3.30 ± 1% -28.5% 2.36 ± 4% perf-profile.cpu-cycles.dequeue_task.deactivate_task.__sched_text_start.schedule.rwsem_down_write_failed 5.67 ± 1% -27.3% 4.12 ± 4% perf-profile.cpu-cycles.schedule.schedule_preempt_disabled.cpu_startup_entry.start_secondary 3.31 ± 1% -28.5% 2.37 ± 3% perf-profile.cpu-cycles.deactivate_task.__sched_text_start.schedule.rwsem_down_write_failed.call_rwsem_down_write_failed 4.74 ± 5% -29.1% 3.36 ± 6% perf-profile.cpu-cycles.perf_event_aux.perf_event_mmap.do_brk.sys_brk.system_call_fastpath 10.92 ± 5% -24.4% 8.25 ± 4% perf-profile.cpu-cycles.cpuidle_enter.cpu_startup_entry.start_secondary 6.64 ± 2% -28.1% 4.77 ± 3% perf-profile.cpu-cycles.schedule.rwsem_down_write_failed.call_rwsem_down_write_failed.sys_brk.system_call_fastpath 6.51 ± 2% -28.2% 4.67 ± 4% perf-profile.cpu-cycles.__sched_text_start.schedule.rwsem_down_write_failed.call_rwsem_down_write_failed.sys_brk 1.77 ± 3% -29.4% 1.25 ± 9% perf-profile.cpu-cycles.set_next_entity.pick_next_task_fair.__sched_text_start.schedule.schedule_preempt_disabled 1.32 ± 5% +33.2% 1.77 ± 9% perf-profile.cpu-cycles.up_write.sys_brk.system_call_fastpath 205 ± 11% +20.3% 247 ± 13% sched_debug.cpu#33.ttwu_local 5926 ± 3% +38.9% 8234 ± 23% sched_debug.cfs_rq[20]:/.exec_clock 244 ± 9% -26.4% 179 ± 8% sched_debug.cpu#26.ttwu_local 354306 ± 9% -20.2% 282834 ± 2% cpuidle.C6-IVT.usage 0.98 ± 9% -25.8% 0.73 ± 7% perf-profile.cpu-cycles.update_cfs_shares.dequeue_entity.dequeue_task_fair.dequeue_task.deactivate_task 17515 ± 3% +10.5% 19349 ± 3% sched_debug.cfs_rq[10]:/.exec_clock 0.79 ± 15% -27.5% 0.57 ± 13% perf-profile.cpu-cycles.resched_curr.ttwu_do_wakeup.ttwu_do_activate.try_to_wake_up.wake_up_process 3.13 ± 2% +27.9% 4.00 ± 4% perf-profile.cpu-cycles.vma_merge.do_brk.sys_brk.system_call_fastpath 29.84 ± 2% -25.2% 22.32 ± 3% perf-profile.cpu-cycles.start_secondary 29.69 ± 2% -25.2% 22.21 ± 3% perf-profile.cpu-cycles.cpu_startup_entry.start_secondary 1.93 ± 5% -24.9% 1.45 ± 8% perf-profile.cpu-cycles.perf_event_aux_ctx.perf_event_aux.perf_event_mmap.do_brk.sys_brk 3.67 ± 4% -23.4% 2.81 ± 3% perf-profile.cpu-cycles.tick_nohz_stop_sched_tick.__tick_nohz_idle_enter.tick_nohz_idle_enter.cpu_startup_entry.start_secondary 308 ± 3% +23.6% 381 ± 1% time.system_time 1.35 ± 11% -22.7% 1.05 ± 3% perf-profile.cpu-cycles.unmap_single_vma.unmap_vmas.unmap_region.do_munmap.sys_brk 4.09 ± 3% -22.8% 3.16 ± 3% perf-profile.cpu-cycles.__tick_nohz_idle_enter.tick_nohz_idle_enter.cpu_startup_entry.start_secondary 1.05 ± 8% -19.7% 0.84 ± 2% perf-profile.cpu-cycles.lapic_next_deadline.clockevents_program_event.tick_program_event.__hrtimer_start_range_ns.hrtimer_start_range_ns 4.38 ± 3% -22.9% 3.38 ± 4% perf-profile.cpu-cycles.tick_nohz_idle_enter.cpu_startup_entry.start_secondary 99 ± 3% +20.2% 119 ± 0% time.percent_of_cpu_this_job_got 4.31 ± 4% -22.4% 3.35 ± 3% perf-profile.cpu-cycles.tick_nohz_idle_exit.cpu_startup_entry.start_secondary 5808 ± 4% +26.5% 7349 ± 14% sched_debug.cfs_rq[13]:/.exec_clock 2.05 ± 7% -21.1% 1.61 ± 2% perf-profile.cpu-cycles.hrtimer_start_range_ns.tick_nohz_restart.tick_nohz_idle_exit.cpu_startup_entry.start_secondary 1.58 ± 6% +21.9% 1.92 ± 8% perf-profile.cpu-cycles.anon_vma_clone.__split_vma.do_munmap.sys_brk.system_call_fastpath 1.43 ± 10% -20.5% 1.14 ± 2% perf-profile.cpu-cycles.tick_program_event.__hrtimer_start_range_ns.hrtimer_start_range_ns.tick_nohz_restart.tick_nohz_idle_exit 9.35 ± 1% -19.6% 7.52 ± 5% perf-profile.cpu-cycles.intel_idle.cpuidle_enter_state.cpuidle_enter.cpu_startup_entry.start_secondary 6081 ± 7% +14.6% 6970 ± 5% sched_debug.cfs_rq[15]:/.exec_clock 6.83 ± 3% +17.1% 8.00 ± 10% perf-profile.cpu-cycles.enqueue_entity.enqueue_task_fair.enqueue_task.activate_task.ttwu_do_activate 5956 ± 10% +17.5% 7000 ± 8% sched_debug.cfs_rq[17]:/.exec_clock 2.02 ± 6% -20.7% 1.60 ± 2% perf-profile.cpu-cycles.__hrtimer_start_range_ns.hrtimer_start_range_ns.tick_nohz_restart.tick_nohz_idle_exit.cpu_startup_entry 2.57 ± 4% -19.3% 2.08 ± 4% perf-profile.cpu-cycles.hrtimer_start.tick_nohz_stop_sched_tick.__tick_nohz_idle_enter.tick_nohz_idle_enter.cpu_startup_entry 246343 ± 1% +9.9% 270748 ± 3% sched_debug.cfs_rq[14]:/.min_vruntime 1.40 ± 9% -20.3% 1.11 ± 2% perf-profile.cpu-cycles.clockevents_program_event.tick_program_event.__hrtimer_start_range_ns.hrtimer_start_range_ns.tick_nohz_restart 17631 ± 1% +15.3% 20334 ± 6% sched_debug.cfs_rq[8]:/.exec_clock 2.53 ± 4% -19.1% 2.05 ± 4% perf-profile.cpu-cycles.__hrtimer_start_range_ns.hrtimer_start.tick_nohz_stop_sched_tick.__tick_nohz_idle_enter.tick_nohz_idle_enter 2222 ± 7% +8.8% 2419 ± 7% sched_debug.cpu#35.curr->pid 2.95 ± 7% -18.9% 2.40 ± 2% perf-profile.cpu-cycles.tick_nohz_restart.tick_nohz_idle_exit.cpu_startup_entry.start_secondary 6200 ± 7% +17.9% 7311 ± 4% sched_debug.cfs_rq[14]:/.exec_clock 7.35 ± 3% +13.7% 8.36 ± 9% perf-profile.cpu-cycles.enqueue_task_fair.enqueue_task.activate_task.ttwu_do_activate.try_to_wake_up 30922 ± 2% +14.9% 35531 ± 1% sched_debug.cpu#15.nr_load_updates 5940 ± 2% +28.2% 7617 ± 22% sched_debug.cfs_rq[18]:/.exec_clock 1.13 ± 18% +39.4% 1.57 ± 12% perf-profile.cpu-cycles.down_write.sys_brk.system_call_fastpath 10.54 ± 3% -12.9% 9.19 ± 3% perf-profile.cpu-cycles.do_brk.sys_brk.system_call_fastpath 40246 ± 1% +13.2% 45575 ± 3% sched_debug.cpu#8.nr_load_updates 30869 ± 2% +19.1% 36767 ± 4% sched_debug.cpu#20.nr_load_updates 2178 ± 4% +11.7% 2433 ± 6% sched_debug.cpu#33.curr->pid 17639 ± 2% +13.5% 20020 ± 6% sched_debug.cfs_rq[11]:/.exec_clock 39954 ± 2% +13.5% 45359 ± 2% sched_debug.cpu#4.nr_load_updates 2483 ± 2% -9.9% 2238 ± 4% time.involuntary_context_switches 31335 ± 1% +13.6% 35597 ± 3% sched_debug.cpu#13.nr_load_updates 30990 ± 3% +13.9% 35313 ± 2% sched_debug.cpu#17.nr_load_updates 246393 ± 3% +13.8% 280499 ± 6% sched_debug.cfs_rq[18]:/.min_vruntime 31272 ± 2% +14.6% 35823 ± 1% sched_debug.cpu#14.nr_load_updates 242514 ± 1% +13.0% 274042 ± 4% sched_debug.cfs_rq[13]:/.min_vruntime 17452 ± 4% +14.8% 20038 ± 4% sched_debug.cfs_rq[9]:/.exec_clock 39962 ± 3% +13.9% 45502 ± 3% sched_debug.cpu#9.nr_load_updates 31046 ± 1% +16.6% 36199 ± 5% sched_debug.cpu#18.nr_load_updates 62.38 ± 1% +14.6% 71.51 ± 1% perf-profile.cpu-cycles.sys_brk.system_call_fastpath 62.72 ± 1% +14.4% 71.76 ± 1% perf-profile.cpu-cycles.system_call_fastpath 39742 ± 2% +11.1% 44168 ± 1% sched_debug.cpu#10.nr_load_updates 30858596 ± 1% +11.6% 34423247 ± 3% cpuidle.C1-IVT.usage 3.52 ± 4% -8.0% 3.24 ± 3% perf-profile.cpu-cycles.unmap_region.do_munmap.sys_brk.system_call_fastpath 243796 ± 1% +10.9% 270426 ± 2% sched_debug.cfs_rq[16]:/.min_vruntime 16.93 ± 2% -13.5% 14.65 ± 6% perf-profile.cpu-cycles.try_to_wake_up.wake_up_process.__rwsem_do_wake.rwsem_wake.call_rwsem_wake 29303 ± 4% +11.6% 32702 ± 4% sched_debug.cpu#12.nr_load_updates 245510 ± 0% +10.3% 270675 ± 2% sched_debug.cfs_rq[19]:/.min_vruntime 244024 ± 1% +10.4% 269379 ± 1% sched_debug.cfs_rq[15]:/.min_vruntime 17.91 ± 2% -12.8% 15.62 ± 5% perf-profile.cpu-cycles.wake_up_process.__rwsem_do_wake.rwsem_wake.call_rwsem_wake.sys_brk 15043 ± 3% -8.3% 13799 ± 4% slabinfo.kmalloc-512.num_objs 246096 ± 0% +11.1% 273409 ± 3% sched_debug.cfs_rq[12]:/.min_vruntime 18.14 ± 2% -12.5% 15.87 ± 5% perf-profile.cpu-cycles.__rwsem_do_wake.rwsem_wake.call_rwsem_wake.sys_brk.system_call_fastpath 17738 ± 2% +11.4% 19752 ± 5% sched_debug.cfs_rq[1]:/.exec_clock 31513 ± 1% +13.4% 35747 ± 3% sched_debug.cpu#16.nr_load_updates 14995 ± 3% -8.2% 13765 ± 4% slabinfo.kmalloc-512.active_objs 39689 ± 2% +12.7% 44717 ± 2% sched_debug.cpu#11.nr_load_updates 31173 ± 2% +10.8% 34530 ± 0% sched_debug.cpu#19.nr_load_updates 2900 ± 2% +8.2% 3137 ± 6% slabinfo.kmalloc-2048.active_objs 50519 ± 2% +9.3% 55204 ± 0% sched_debug.cpu#43.nr_load_updates 754899 ± 3% -7.2% 700567 ± 5% sched_debug.cpu#35.avg_idle 2189 ± 6% -6.5% 2046 ± 4% sched_debug.cpu#47.curr->pid 245137 ± 1% +10.9% 271884 ± 1% sched_debug.cfs_rq[20]:/.min_vruntime 252683 ± 2% +6.8% 269903 ± 2% sched_debug.cfs_rq[22]:/.min_vruntime 250553 ± 4% +7.3% 268896 ± 2% sched_debug.cfs_rq[21]:/.min_vruntime 40942 ± 4% +10.5% 45255 ± 2% sched_debug.cpu#1.nr_load_updates 19657 ± 4% -9.8% 17725 ± 5% vmstat.system.in 27.10 ± 0% -0.7% 26.90 ± 0% turbostat.%Busy 4.10 ± 0% -2.4% 4.00 ± 0% turbostat.RAMWatt
testbox/testcase/testparams: lituya/will-it-scale/performance-brk1
b3fd4f03ca0b9952 1a99367023f6ac664365a37fa5 ---------------- -------------------------- 239 ± 1% +32.0% 316 ± 3% will-it-scale.time.system_time 80 ± 1% +30.4% 105 ± 3% will-it-scale.time.percent_of_cpu_this_job_got 52295908 ± 1% -5.4% 49462338 ± 0% will-it-scale.time.voluntary_context_switches 728289 ± 0% -1.8% 715194 ± 0% will-it-scale.per_thread_ops 63 ± 48% -36.9% 40 ± 7% sched_debug.cpu#12.load 223957 ± 16% -62.8% 83209 ± 16% cpuidle.C6-HSW.usage 31 ± 16% +116.1% 67 ± 34% sched_debug.cpu#14.load 80 ± 34% -60.7% 31 ± 20% sched_debug.cpu#2.load 73 ± 25% -53.4% 34 ± 12% sched_debug.cfs_rq[2]:/.load 300986 ± 24% -40.3% 179777 ± 42% sched_debug.cfs_rq[4]:/.min_vruntime 346 ± 33% +91.1% 662 ± 25% cpuidle.POLL.usage 1212812 ± 35% -44.7% 670407 ± 25% sched_debug.cpu#2.ttwu_count 144641 ± 35% -62.3% 54518 ± 15% sched_debug.cpu#6.ttwu_local 33 ± 25% +90.2% 63 ± 34% sched_debug.cfs_rq[14]:/.load 1377774 ± 40% +210.8% 4282777 ± 48% sched_debug.cpu#9.sched_count 34 ± 10% +109.4% 72 ± 43% sched_debug.cpu#14.cpu_load[0] 681074 ± 40% +210.0% 2111486 ± 49% sched_debug.cpu#9.sched_goidle 1362573 ± 40% +210.0% 4223660 ± 49% sched_debug.cpu#9.nr_switches 327 ± 7% +81.2% 593 ± 14% sched_debug.cfs_rq[14]:/.tg_load_contrib 588875 ± 12% +78.6% 1051474 ± 12% sched_debug.cpu#6.sched_count 292062 ± 13% +77.6% 518637 ± 12% sched_debug.cpu#6.sched_goidle 585096 ± 13% +77.5% 1038414 ± 12% sched_debug.cpu#6.nr_switches 262640 ± 6% -41.6% 153289 ± 11% sched_debug.cfs_rq[6]:/.min_vruntime 148498 ± 46% +113.4% 316963 ± 13% sched_debug.cpu#1.ttwu_local 1385681 ± 22% +86.1% 2578972 ± 18% sched_debug.cpu#8.ttwu_count 296 ± 9% +80.1% 533 ± 17% sched_debug.cfs_rq[14]:/.blocked_load_avg 24472 ± 25% -40.1% 14663 ± 48% sched_debug.cfs_rq[4]:/.exec_clock 32 ± 7% +79.7% 57 ± 34% sched_debug.cpu#14.cpu_load[1] 1650425 ± 13% -37.7% 1027432 ± 29% sched_debug.cpu#14.ttwu_count 57 ± 14% +36.2% 78 ± 10% sched_debug.cpu#0.load 43412 ± 13% -26.2% 32048 ± 22% sched_debug.cfs_rq[2]:/.exec_clock 33 ± 6% +67.7% 55 ± 21% sched_debug.cpu#13.cpu_load[0] 64 ± 17% -22.5% 50 ± 19% sched_debug.cpu#9.cpu_load[0] 53 ± 14% +34.4% 72 ± 5% sched_debug.cfs_rq[0]:/.load 31 ± 7% +53.5% 48 ± 22% sched_debug.cpu#14.cpu_load[2] 29 ± 10% +47.9% 43 ± 15% sched_debug.cpu#13.cpu_load[1] 32 ± 16% -36.9% 20 ± 24% sched_debug.cpu#4.cpu_load[1] 30 ± 5% +36.9% 41 ± 10% sched_debug.cpu#14.cpu_load[4] 31 ± 5% +40.5% 44 ± 14% sched_debug.cpu#14.cpu_load[3] 520038 ± 13% -19.4% 419303 ± 18% sched_debug.cfs_rq[2]:/.min_vruntime 469998 ± 13% +31.3% 616962 ± 12% sched_debug.cfs_rq[10]:/.min_vruntime 36098 ± 10% -40.6% 21432 ± 4% sched_debug.cpu#6.nr_load_updates 21 ± 26% -33.7% 14 ± 13% sched_debug.cpu#4.cpu_load[4] 1178 ± 12% +43.7% 1694 ± 13% sched_debug.cpu#14.curr->pid 21211 ± 10% -38.2% 13103 ± 12% sched_debug.cfs_rq[6]:/.exec_clock 39866 ± 16% +32.5% 52814 ± 16% sched_debug.cfs_rq[10]:/.exec_clock 24 ± 23% -35.1% 15 ± 15% sched_debug.cpu#4.cpu_load[3] 1.38 ± 12% -15.0% 1.17 ± 4% perf-profile.cpu-cycles.avc_has_perm_noaudit.cred_has_capability.selinux_capable.selinux_vm_enough_memory.security_vm_enough_memory_mm 239 ± 1% +32.0% 316 ± 3% time.system_time 394 ± 7% +26.1% 497 ± 5% sched_debug.cfs_rq[14]:/.tg_runnable_contrib 18088 ± 7% +26.4% 22861 ± 5% sched_debug.cfs_rq[14]:/.avg->runnable_avg_sum 39 ± 17% -33.8% 26 ± 29% sched_debug.cpu#4.cpu_load[0] 1228 ± 4% +32.1% 1622 ± 6% sched_debug.cpu#0.curr->pid 80 ± 1% +30.4% 105 ± 3% time.percent_of_cpu_this_job_got 27 ± 18% -34.9% 17 ± 19% sched_debug.cpu#4.cpu_load[2] 401 ± 9% -14.3% 344 ± 11% sched_debug.cfs_rq[2]:/.tg_runnable_contrib 6.50 ± 1% -20.3% 5.19 ± 7% time.user_time 502752 ± 2% +29.0% 648466 ± 3% sched_debug.cfs_rq[14]:/.min_vruntime 53 ± 12% +23.6% 65 ± 4% sched_debug.cfs_rq[0]:/.runnable_load_avg 353 ± 12% +28.8% 455 ± 7% sched_debug.cfs_rq[12]:/.tg_runnable_contrib 16234 ± 12% +28.8% 20903 ± 7% sched_debug.cfs_rq[12]:/.avg->runnable_avg_sum 43966 ± 4% +26.9% 55773 ± 4% sched_debug.cfs_rq[14]:/.exec_clock 1344 ± 8% +15.4% 1552 ± 4% sched_debug.cpu#11.curr->pid 1080 ± 1% -15.6% 912 ± 4% time.involuntary_context_switches 433 ± 3% +16.3% 504 ± 2% sched_debug.cfs_rq[9]:/.tg_runnable_contrib 19913 ± 3% +16.5% 23191 ± 2% sched_debug.cfs_rq[9]:/.avg->runnable_avg_sum 54233 ± 6% +14.0% 61828 ± 6% sched_debug.cpu#14.nr_load_updates 48889 ± 7% +17.7% 57564 ± 2% sched_debug.cfs_rq[9]:/.exec_clock 59096 ± 3% +11.9% 66139 ± 4% sched_debug.cpu#9.nr_load_updates 53 ± 12% +18.8% 63 ± 7% sched_debug.cpu#0.cpu_load[0] 13853 ± 10% +35.6% 18786 ± 10% vmstat.system.in 346546 ± 1% -5.3% 328077 ± 0% vmstat.system.cs 1146 ± 0% +4.2% 1195 ± 0% turbostat.Avg_MHz 34.76 ± 0% +4.2% 36.23 ± 0% turbostat.%Busy
ivb42: Ivytown Ivy Bridge-EP Memory: 64G
lituya: Grantley Haswell Memory: 16G
will-it-scale.time.percent_of_cpu_this_job_got
125 ++--------------------------------------------------------------------+ O | 120 ++ O O O O | 115 ++ O O O O O O | | O O O O O | 110 ++ O O O | | | 105 ++ * | | + + | 100 ++ .*..*.. + + .* 95 ++ *...*.. ..*.. .. . ..*..*.. + *...*. | | .. *. * *.. .*. . .* | 90 ++..* *...*. *. | *. | 85 ++--------------------------------------------------------------------+
[*] 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
--- testcase: will-it-scale default-monitors: wait: pre-test uptime: iostat: vmstat: numa-numastat: numa-vmstat: numa-meminfo: proc-vmstat: proc-stat: meminfo: slabinfo: interrupts: lock_stat: latency_stats: softirqs: bdi_dev_mapping: diskstats: nfsstat: cpuidle: cpufreq-stats: turbostat: pmeter: sched_debug: interval: 10 default_watchdogs: watch-oom: watchdog: cpufreq_governor: performance commit: 5aaa6b9f28b1c0c7623dcfd8d87e1d0e8bd4dde6 model: Ivytown Ivy Bridge-EP nr_cpu: 48 memory: 64G rootfs: debian-x86_64-2015-02-07.cgz perf-profile: freq: 800 will-it-scale: test: brk1 testbox: ivb42 tbox_group: ivb42 kconfig: x86_64-rhel enqueue_time: 2015-03-08 22:23:51.602528045 +08:00 head_commit: 5aaa6b9f28b1c0c7623dcfd8d87e1d0e8bd4dde6 base_commit: 9eccca0843205f87c00404b663188b88eb248051 branch: linux-devel/devel-hourly-2015030907 kernel: "/kernel/x86_64-rhel/5aaa6b9f28b1c0c7623dcfd8d87e1d0e8bd4dde6/vmlinuz-4.0.0-rc3-01051-g5aaa6b9" user: lkp queue: cyclic result_root: "/result/ivb42/will-it-scale/performance-brk1/debian-x86_64-2015-02-07.cgz/x86_64-rhel/5aaa6b9f28b1c0c7623dcfd8d87e1d0e8bd4dde6/0" job_file: "/lkp/scheduled/ivb42/cyclic_will-it-scale-performance-brk1-debian-x86_64.cgz-x86_64-rhel-HEAD-5aaa6b9f28b1c0c7623dcfd8d87e1d0e8bd4dde6-0-20150308-40925-1ugo6xp.yaml" dequeue_time: 2015-03-09 08:52:18.512200890 +08:00 job_state: finished loadavg: 29.17 17.88 7.43 1/420 10748 start_time: '1425862384' end_time: '1425862718' version: "/lkp/lkp/.src-20150308-175746" echo performance > /sys/devices/system/cpu/cpu0/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu1/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu10/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu11/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu12/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu13/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu14/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu15/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu16/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu17/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu18/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu19/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu2/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu20/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu21/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu22/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu23/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu24/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu25/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu26/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu27/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu28/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu29/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu3/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu30/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu31/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu32/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu33/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu34/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu35/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu36/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu37/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu38/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu39/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu4/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu40/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu41/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu42/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu43/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu44/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu45/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu46/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu47/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu5/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu6/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu7/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu8/cpufreq/scaling_governor echo performance > /sys/devices/system/cpu/cpu9/cpufreq/scaling_governor ./runtest.py brk1 25 both 1 12 24 36 48 _______________________________________________ LKP mailing list LKP@linux.intel.com
| |