Hello, kernel test robot noticed a 7.0% regression of phoronix-test-suite.ncnn.VulkanGPU.ms on: commit: 5e2bb14a5498d7b3197773937cc24fa4ed576836 ("random: Comment out write throttling in add_hwgenerator_randomness()") git://git.kernel.org/cgit/linux/kernel/git/jarkko/linux-tpmdd.git tpm1-s3-bug testcase: phoronix-test-suite test machine: 96 threads 2 sockets Intel(R) Xeon(R) Gold 6252 CPU @ 2.10GHz (Cascade Lake) with 512G memory parameters: test: ncnn-1.1.0 option_a: Vulkan GPU cpufreq_governor: performance test-description: The Phoronix Test Suite is the most comprehensive testing and benchmarking platform available that provides an extensible framework for which new tests can be easily added. test-url: http://www.phoronix-test-suite.com/ If you fix the issue, kindly add following tag | Reported-by: kernel test robot | Link: https://lore.kernel.org/oe-lkp/202304301154.c4ad1c70-oliver.sang@intel.com Details are as below: --------------------------------------------------------------------------------------------------> To reproduce: git clone https://github.com/intel/lkp-tests.git cd lkp-tests sudo bin/lkp install job.yaml # job file is attached in this email bin/lkp split-job --compatible job.yaml # generate the yaml file for lkp run sudo bin/lkp run generated-yaml-file # if come across any failure that blocks the test, # please remove ~/.lkp and /lkp dir to run from a clean state. ========================================================================================= compiler/cpufreq_governor/kconfig/option_a/rootfs/tbox_group/test/testcase: gcc-11/performance/x86_64-rhel-8.3/Vulkan GPU/debian-x86_64-phoronix/lkp-csl-2sp7/ncnn-1.1.0/phoronix-test-suite commit: bd8621ca15 ("tpm: Add !tpm_amd_is_rng_defective() to the hwrng_unregister() call site") 5e2bb14a54 ("random: Comment out write throttling in add_hwgenerator_randomness()") bd8621ca1510e6e8 5e2bb14a5498d7b3197773937cc ---------------- --------------------------- %stddev %change %stddev \ | \ 33.51 +7.0% 35.85 phoronix-test-suite.ncnn.VulkanGPU.ms 425.70 +28.0% 544.97 ± 38% phoronix-test-suite.time.elapsed_time 425.70 +28.0% 544.97 ± 38% phoronix-test-suite.time.elapsed_time.max 195937 +390.1% 960259 ± 40% phoronix-test-suite.time.involuntary_context_switches 36633 +29.9% 47584 ± 40% phoronix-test-suite.time.user_time 131816 +43.5% 189109 ± 34% phoronix-test-suite.time.voluntary_context_switches 2239 ± 2% +131.9% 5194 vmstat.system.cs 586.83 ± 2% +42.9% 838.83 ± 41% meminfo.Mlocked 654.83 ± 2% +38.5% 906.83 ± 38% meminfo.Unevictable 778215 ± 18% +44.0% 1120918 ± 28% numa-numastat.node0.local_node 825029 ± 17% +40.9% 1162569 ± 28% numa-numastat.node0.numa_hit 825451 ± 17% +40.9% 1162666 ± 28% numa-vmstat.node0.numa_hit 778637 ± 18% +44.0% 1121080 ± 28% numa-vmstat.node0.numa_local 2010685 +20.7% 2427619 ± 27% proc-vmstat.numa_hit 1942066 +21.5% 2359002 ± 28% proc-vmstat.numa_local 2816890 +19.2% 3358964 ± 29% proc-vmstat.pgfault 340470 ± 2% +22.7% 417901 ± 34% proc-vmstat.pgreuse 3533056 +25.2% 4423296 ± 35% proc-vmstat.unevictable_pgs_scanned 169760 ± 3% +77.0% 300470 ± 27% turbostat.C1E 0.15 ± 2% +0.1 0.22 ± 7% turbostat.C1E% 2.49 ± 3% -28.6% 1.78 ± 13% turbostat.CPU%c6 1.43 ± 3% -30.7% 0.99 ± 16% turbostat.Pkg%pc2 0.64 ± 5% -74.1% 0.16 ± 10% turbostat.Pkg%pc6 16010254 +44.6% 23143673 ± 41% sched_debug.cfs_rq:/.min_vruntime.avg 16099048 +44.6% 23287103 ± 41% sched_debug.cfs_rq:/.min_vruntime.max 15404106 +44.4% 22249932 ± 40% sched_debug.cfs_rq:/.min_vruntime.min 138799 ± 2% +46.2% 202956 ± 41% sched_debug.cfs_rq:/.min_vruntime.stddev 529368 ± 14% +54.8% 819201 ± 44% sched_debug.cfs_rq:/.spread0.avg 617788 ± 12% +55.8% 962287 ± 43% sched_debug.cfs_rq:/.spread0.max 138600 ± 2% +46.3% 202772 ± 41% sched_debug.cfs_rq:/.spread0.stddev 230591 +32.4% 305356 ± 32% sched_debug.cpu.clock.avg 230602 +32.4% 305367 ± 32% sched_debug.cpu.clock.max 230576 +32.4% 305345 ± 32% sched_debug.cpu.clock.min 228524 +32.5% 302741 ± 32% sched_debug.cpu.clock_task.avg 229014 +32.4% 303296 ± 32% sched_debug.cpu.clock_task.max 212880 +35.0% 287493 ± 34% sched_debug.cpu.clock_task.min 5245 +197.4% 15596 ± 34% sched_debug.cpu.nr_switches.avg 25964 ± 15% +327.8% 111075 ± 18% sched_debug.cpu.nr_switches.max 4319 ± 8% +323.3% 18285 ± 16% sched_debug.cpu.nr_switches.stddev 230577 +32.4% 305344 ± 32% sched_debug.cpu_clk 226197 +33.1% 300964 ± 33% sched_debug.ktime 231388 +32.3% 306167 ± 32% sched_debug.sched_clk 14.00 ± 2% -12.1% 12.30 ± 2% perf-stat.i.MPKI 5.913e+09 +4.4% 6.17e+09 perf-stat.i.branch-instructions 0.68 ± 5% -0.1 0.55 ± 7% perf-stat.i.branch-miss-rate% 11783773 -6.4% 11033798 ± 3% perf-stat.i.branch-misses 1.113e+08 ± 2% -5.6% 1.051e+08 ± 3% perf-stat.i.cache-misses 3.569e+08 -5.1% 3.386e+08 perf-stat.i.cache-references 2175 +136.9% 5153 perf-stat.i.context-switches 7.93 -1.8% 7.79 perf-stat.i.cpi 2.309e+11 +1.2% 2.337e+11 perf-stat.i.cpu-cycles 82.87 ± 3% +26.5% 104.81 ± 5% perf-stat.i.cpu-migrations 0.06 ± 9% -0.0 0.04 ± 7% perf-stat.i.dTLB-load-miss-rate% 1.711e+09 -5.4% 1.618e+09 ± 2% perf-stat.i.dTLB-stores 2404437 +1.2% 2434030 perf-stat.i.metric.GHz 6059 -4.1% 5809 ± 4% perf-stat.i.minor-faults 6062 -4.1% 5811 ± 4% perf-stat.i.page-faults 10.78 -36.8% 6.81 ± 70% perf-stat.overall.MPKI 0.20 -0.1 0.12 ± 70% perf-stat.overall.branch-miss-rate% 11758585 -37.6% 7335914 ± 70% perf-stat.ps.branch-misses 1.111e+08 ± 2% -36.3% 70703074 ± 70% perf-stat.ps.cache-misses 3.561e+08 -36.2% 2.273e+08 ± 70% perf-stat.ps.cache-references 9.611e+09 ± 2% -34.6% 6.283e+09 ± 70% perf-stat.ps.dTLB-loads 1.707e+09 -35.9% 1.093e+09 ± 70% perf-stat.ps.dTLB-stores 2.92 ± 7% -43.9% 1.63 ± 71% perf-stat.ps.major-faults 6047 -36.4% 3845 ± 71% perf-stat.ps.minor-faults 6050 -36.4% 3846 ± 71% perf-stat.ps.page-faults 25.42 ± 11% -5.1 20.30 ± 4% perf-profile.calltrace.cycles-pp.perf_callchain.perf_prepare_sample.perf_event_output_forward.__perf_event_overflow.perf_tp_event 8.88 ± 12% -1.8 7.10 ± 3% perf-profile.calltrace.cycles-pp.perf_callchain_user.get_perf_callchain.perf_callchain.perf_prepare_sample.perf_event_output_forward 4.32 ± 11% -1.4 2.88 ± 9% perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe 6.43 ± 11% -1.3 5.11 ± 5% perf-profile.calltrace.cycles-pp.__unwind_start.perf_callchain_kernel.get_perf_callchain.perf_callchain.perf_prepare_sample 5.64 ± 11% -1.2 4.42 ± 4% perf-profile.calltrace.cycles-pp.unwind_next_frame.__unwind_start.perf_callchain_kernel.get_perf_callchain.perf_callchain 3.88 ± 10% -1.2 2.70 ± 8% perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe 1.35 ± 22% -0.8 0.58 ± 75% perf-profile.calltrace.cycles-pp.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe 2.06 ± 10% -0.4 1.66 ± 8% perf-profile.calltrace.cycles-pp.__kernel_text_address.unwind_get_return_address.perf_callchain_kernel.get_perf_callchain.perf_callchain 0.79 ± 14% -0.4 0.39 ± 71% perf-profile.calltrace.cycles-pp.stack_access_ok.unwind_next_frame.__unwind_start.perf_callchain_kernel.get_perf_callchain 1.58 ± 12% -0.3 1.26 ± 9% perf-profile.calltrace.cycles-pp.kernel_text_address.__kernel_text_address.unwind_get_return_address.perf_callchain_kernel.get_perf_callchain 0.89 ± 11% -0.3 0.58 ± 45% perf-profile.calltrace.cycles-pp.search_exception_tables.fixup_exception.kernelmode_fixup_or_oops.exc_page_fault.asm_exc_page_fault 0.86 ± 13% -0.3 0.56 ± 45% perf-profile.calltrace.cycles-pp.search_extable.search_exception_tables.fixup_exception.kernelmode_fixup_or_oops.exc_page_fault 0.80 ± 13% -0.3 0.53 ± 45% perf-profile.calltrace.cycles-pp.bsearch.search_extable.search_exception_tables.fixup_exception.kernelmode_fixup_or_oops 1.09 ± 11% -0.3 0.83 ± 13% perf-profile.calltrace.cycles-pp.kernelmode_fixup_or_oops.exc_page_fault.asm_exc_page_fault.__get_user_nocheck_8.perf_callchain_user 1.00 ± 12% -0.2 0.76 ± 14% perf-profile.calltrace.cycles-pp.fixup_exception.kernelmode_fixup_or_oops.exc_page_fault.asm_exc_page_fault.__get_user_nocheck_8 1.02 ± 8% -0.1 0.88 ± 6% perf-profile.calltrace.cycles-pp.orc_find.unwind_next_frame.__unwind_start.perf_callchain_kernel.get_perf_callchain 0.00 +0.7 0.71 ± 11% perf-profile.calltrace.cycles-pp.try_to_wake_up.hrtimer_wakeup.__hrtimer_run_queues.hrtimer_interrupt.__sysvec_apic_timer_interrupt 0.00 +0.7 0.71 ± 11% perf-profile.calltrace.cycles-pp.hrtimer_wakeup.__hrtimer_run_queues.hrtimer_interrupt.__sysvec_apic_timer_interrupt.sysvec_apic_timer_interrupt 0.37 ± 70% +0.8 1.18 ± 4% perf-profile.calltrace.cycles-pp.exit_to_user_mode_prepare.irqentry_exit_to_user_mode.asm_sysvec_apic_timer_interrupt 0.00 +1.0 1.04 ± 4% perf-profile.calltrace.cycles-pp.__schedule.schedule.exit_to_user_mode_loop.exit_to_user_mode_prepare.irqentry_exit_to_user_mode 0.00 +1.1 1.05 ± 4% perf-profile.calltrace.cycles-pp.schedule.exit_to_user_mode_loop.exit_to_user_mode_prepare.irqentry_exit_to_user_mode.asm_sysvec_apic_timer_interrupt 0.00 +1.1 1.07 ± 4% perf-profile.calltrace.cycles-pp.exit_to_user_mode_loop.exit_to_user_mode_prepare.irqentry_exit_to_user_mode.asm_sysvec_apic_timer_interrupt 0.00 +1.4 1.36 ± 8% perf-profile.calltrace.cycles-pp.ioread8.tpm_tcg_read_bytes.tpm_tis_status.wait_for_tpm_stat.tpm_tis_send_data 0.00 +1.4 1.36 ± 8% perf-profile.calltrace.cycles-pp.tpm_tcg_read_bytes.tpm_tis_status.wait_for_tpm_stat.tpm_tis_send_data.tpm_tis_send_main 0.00 +1.4 1.37 ± 8% perf-profile.calltrace.cycles-pp.tpm_tis_status.wait_for_tpm_stat.tpm_tis_send_data.tpm_tis_send_main.tpm_try_transmit 0.00 +1.9 1.87 ± 9% perf-profile.calltrace.cycles-pp.wait_for_tpm_stat.tpm_tis_send_data.tpm_tis_send_main.tpm_try_transmit.tpm_transmit 0.00 +2.2 2.19 ± 10% perf-profile.calltrace.cycles-pp.tpm_tis_send_data.tpm_tis_send_main.tpm_try_transmit.tpm_transmit.tpm_transmit_cmd 0.00 +2.2 2.20 ± 10% perf-profile.calltrace.cycles-pp.tpm_tis_send_main.tpm_try_transmit.tpm_transmit.tpm_transmit_cmd.tpm2_get_random 0.00 +3.2 3.15 ± 12% perf-profile.calltrace.cycles-pp.tpm_tcg_read_bytes.recv_data.tpm_tis_recv.tpm_try_transmit.tpm_transmit 0.00 +3.2 3.15 ± 12% perf-profile.calltrace.cycles-pp.ioread8.tpm_tcg_read_bytes.recv_data.tpm_tis_recv.tpm_try_transmit 0.00 +3.5 3.49 ± 13% perf-profile.calltrace.cycles-pp.recv_data.tpm_tis_recv.tpm_try_transmit.tpm_transmit.tpm_transmit_cmd 0.00 +3.6 3.60 ± 13% perf-profile.calltrace.cycles-pp.tpm_tis_recv.tpm_try_transmit.tpm_transmit.tpm_transmit_cmd.tpm2_get_random 0.00 +6.4 6.39 ± 11% perf-profile.calltrace.cycles-pp.tpm_try_transmit.tpm_transmit.tpm_transmit_cmd.tpm2_get_random.tpm_get_random 0.00 +6.4 6.40 ± 11% perf-profile.calltrace.cycles-pp.tpm_transmit.tpm_transmit_cmd.tpm2_get_random.tpm_get_random.hwrng_fillfn 0.00 +6.4 6.40 ± 11% perf-profile.calltrace.cycles-pp.tpm_transmit_cmd.tpm2_get_random.tpm_get_random.hwrng_fillfn.kthread 0.00 +6.4 6.42 ± 11% perf-profile.calltrace.cycles-pp.tpm2_get_random.tpm_get_random.hwrng_fillfn.kthread.ret_from_fork 0.00 +6.6 6.57 ± 11% perf-profile.calltrace.cycles-pp.tpm_get_random.hwrng_fillfn.kthread.ret_from_fork 0.00 +6.6 6.60 ± 11% perf-profile.calltrace.cycles-pp.hwrng_fillfn.kthread.ret_from_fork 0.92 ± 18% +6.7 7.59 ± 8% perf-profile.calltrace.cycles-pp.ret_from_fork 0.92 ± 18% +6.7 7.59 ± 8% perf-profile.calltrace.cycles-pp.kthread.ret_from_fork 9.06 ± 12% -1.5 7.51 ± 4% perf-profile.children.cycles-pp.perf_callchain_user 8.70 ± 12% -1.5 7.22 ± 5% perf-profile.children.cycles-pp.__get_user_nocheck_8 1.74 ± 19% -0.7 1.08 ± 27% perf-profile.children.cycles-pp.syscall_exit_to_user_mode 1.91 ± 10% -0.4 1.55 ± 8% perf-profile.children.cycles-pp.kernel_text_address 1.58 ± 14% -0.2 1.35 ± 10% perf-profile.children.cycles-pp.stack_access_ok 1.13 ± 11% -0.2 0.90 ± 15% perf-profile.children.cycles-pp.kernelmode_fixup_or_oops 0.90 ± 11% -0.2 0.67 ± 17% perf-profile.children.cycles-pp.search_extable 0.86 ± 13% -0.2 0.66 ± 15% perf-profile.children.cycles-pp.bsearch 0.46 ± 19% -0.1 0.33 ± 6% perf-profile.children.cycles-pp.enqueue_hrtimer 0.22 ± 52% -0.1 0.08 ± 63% perf-profile.children.cycles-pp.syscall_enter_from_user_mode 0.34 ± 19% -0.1 0.22 ± 19% perf-profile.children.cycles-pp.futex_wake 0.57 ± 14% -0.1 0.46 ± 9% perf-profile.children.cycles-pp.update_irq_load_avg 0.44 ± 10% -0.1 0.34 ± 12% perf-profile.children.cycles-pp.hrtimer_update_next_event 0.13 ± 30% -0.1 0.04 ± 72% perf-profile.children.cycles-pp.trigger_load_balance 0.01 ±223% +0.1 0.11 ± 28% perf-profile.children.cycles-pp.select_idle_sibling 0.06 ± 79% +0.1 0.17 ± 16% perf-profile.children.cycles-pp.select_task_rq 0.11 ± 15% +0.1 0.23 ± 26% perf-profile.children.cycles-pp.select_task_rq_fair 0.00 +0.1 0.12 ± 23% perf-profile.children.cycles-pp.__tpm_tis_request_locality 0.00 +0.1 0.12 ± 23% perf-profile.children.cycles-pp.check_locality 0.00 +0.1 0.13 ± 24% perf-profile.children.cycles-pp.tpm_tis_request_locality 0.00 +0.1 0.13 ± 25% perf-profile.children.cycles-pp.tpm_chip_start 0.00 +0.1 0.14 ± 20% perf-profile.children.cycles-pp.tpm_try_get_ops 0.00 +0.1 0.14 ± 21% perf-profile.children.cycles-pp.tpm_find_get_ops 0.43 ± 10% +0.3 0.69 ± 14% perf-profile.children.cycles-pp.dequeue_task_fair 0.38 ± 12% +0.3 0.67 ± 14% perf-profile.children.cycles-pp.dequeue_entity 0.00 +0.3 0.29 ± 15% perf-profile.children.cycles-pp.ioread32 0.00 +0.3 0.29 ± 15% perf-profile.children.cycles-pp.get_burstcount 0.10 ± 29% +0.3 0.44 ± 13% perf-profile.children.cycles-pp.put_prev_entity 0.28 ± 16% +0.4 0.67 ± 13% perf-profile.children.cycles-pp.perf_trace_sched_wakeup_template 0.63 ± 24% +0.5 1.09 ± 15% perf-profile.children.cycles-pp.pick_next_task_fair 0.37 ± 33% +0.5 0.86 ± 21% perf-profile.children.cycles-pp.perf_trace_sched_switch 0.49 ± 8% +0.5 0.99 ± 13% perf-profile.children.cycles-pp.ttwu_do_activate 0.13 ± 34% +0.6 0.73 ± 10% perf-profile.children.cycles-pp.hrtimer_wakeup 0.50 ± 9% +0.6 1.15 ± 8% perf-profile.children.cycles-pp.try_to_wake_up 0.00 +0.8 0.75 ± 16% perf-profile.children.cycles-pp.usleep_range_state 0.01 ±223% +0.8 0.78 ± 15% perf-profile.children.cycles-pp.schedule_hrtimeout_range_clock 0.75 ± 10% +0.9 1.64 ± 14% perf-profile.children.cycles-pp.exit_to_user_mode_prepare 0.54 ± 12% +1.0 1.50 ± 15% perf-profile.children.cycles-pp.exit_to_user_mode_loop 1.64 ± 14% +1.5 3.14 ± 4% perf-profile.children.cycles-pp.schedule 1.82 ± 17% +1.6 3.37 ± 6% perf-profile.children.cycles-pp.__schedule 0.00 +2.1 2.06 ± 10% perf-profile.children.cycles-pp.wait_for_tpm_stat 0.00 +2.1 2.14 ± 11% perf-profile.children.cycles-pp.tpm_tis_status 0.00 +2.2 2.19 ± 10% perf-profile.children.cycles-pp.tpm_tis_send_data 0.00 +2.2 2.20 ± 10% perf-profile.children.cycles-pp.tpm_tis_send_main 0.00 +3.5 3.49 ± 13% perf-profile.children.cycles-pp.recv_data 0.00 +3.6 3.60 ± 13% perf-profile.children.cycles-pp.tpm_tis_recv 0.00 +5.4 5.39 ± 12% perf-profile.children.cycles-pp.ioread8 0.00 +5.7 5.69 ± 12% perf-profile.children.cycles-pp.tpm_tcg_read_bytes 0.00 +6.4 6.39 ± 11% perf-profile.children.cycles-pp.tpm_try_transmit 0.00 +6.4 6.40 ± 11% perf-profile.children.cycles-pp.tpm_transmit 0.00 +6.4 6.40 ± 11% perf-profile.children.cycles-pp.tpm_transmit_cmd 0.00 +6.4 6.42 ± 11% perf-profile.children.cycles-pp.tpm2_get_random 0.00 +6.6 6.57 ± 11% perf-profile.children.cycles-pp.tpm_get_random 0.00 +6.6 6.60 ± 11% perf-profile.children.cycles-pp.hwrng_fillfn 0.92 ± 18% +6.7 7.59 ± 8% perf-profile.children.cycles-pp.kthread 0.96 ± 17% +6.7 7.64 ± 8% perf-profile.children.cycles-pp.ret_from_fork 1.60 ± 20% -0.7 0.90 ± 32% perf-profile.self.cycles-pp.syscall_exit_to_user_mode 1.39 ± 15% -0.2 1.18 ± 10% perf-profile.self.cycles-pp.stack_access_ok 0.97 ± 12% -0.2 0.81 ± 10% perf-profile.self.cycles-pp.perf_callchain_kernel 0.64 ± 7% -0.2 0.48 ± 13% perf-profile.self.cycles-pp.hrtimer_interrupt 0.41 ± 19% -0.1 0.28 ± 23% perf-profile.self.cycles-pp._raw_spin_lock_irq 0.21 ± 52% -0.1 0.08 ± 61% perf-profile.self.cycles-pp.syscall_enter_from_user_mode 0.56 ± 13% -0.1 0.46 ± 9% perf-profile.self.cycles-pp.update_irq_load_avg 0.13 ± 28% -0.1 0.04 ± 72% perf-profile.self.cycles-pp.trigger_load_balance 0.12 ± 12% -0.1 0.04 ±112% perf-profile.self.cycles-pp.entry_SYSCALL_64_after_hwframe 0.15 ± 14% -0.1 0.09 ± 25% perf-profile.self.cycles-pp.futex_wake 0.32 ± 13% -0.1 0.26 ± 8% perf-profile.self.cycles-pp.__perf_event_header__init_id 0.01 ±223% +0.1 0.06 ± 28% perf-profile.self.cycles-pp.get_mem_cgroup_from_mm 0.02 ±143% +0.1 0.15 ± 20% perf-profile.self.cycles-pp.__schedule 0.00 +0.3 0.29 ± 15% perf-profile.self.cycles-pp.ioread32 0.00 +5.4 5.38 ± 12% perf-profile.self.cycles-pp.ioread8 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. -- 0-DAY CI Kernel Test Service https://github.com/intel/lkp-tests