[linux-next:master] [syscore] a97fbc3ee3: will-it-scale.per_process_ops 3.2% regression
kernel test robot <[email protected]>
| Newsgroups | gmane.linux.ports.tegra,gmane.linux.ports.arm.kernel,gmane.linux.kernel.samsung-soc,gmane.linux.ports.mips,gmane.linux.ports.ppc64.devel,gmane.linux.ports.sh.devel,gmane.linux.kernel.perf.user,gmane.linux.kernel,gmane.comp.emulators.kvm.devel,gmane.linux.kernel.pci,gmane.linux.acpi.devel,gmane.linux.ports.arm.rockchip,gmane.linux.power-management.general,gmane.linux.kernel.gpio,gmane.linux.ports.arm.mediatek,gmane.linux.ports.riscv,gmane.linux.leds,gmane.comp.emulators.xen.devel |
|---|---|
| Message-ID | <[email protected]> |
Hello,
kernel test robot noticed a 3.2% regression of will-it-scale.per_process_ops on:
commit: a97fbc3ee3e2a536fafaff04f21f45472db71769 ("syscore: Pass context data to callbacks")
https://git.kernel.org/cgit/linux/kernel/git/next/linux-next.git master
[still regression on linux-next/master ef68bf704646690aba5e81c2f7be8d6ef13d7ad8]
testcase: will-it-scale
config: x86_64-rhel-9.4
compiler: gcc-14
test machine: 64 threads 2 sockets Intel(R) Xeon(R) Gold 6346 CPU @ 3.10GHz (Ice Lake) with 256G memory
parameters:
nr_task: 100%
mode: process
test: signal1
cpufreq_governor: performance
If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <[email protected]>
| Closes: https://lore.kernel.org/oe-lkp/202511281432.47ac7e38-lkp-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org
Details are as below:
-------------------------------------------------------------------------------------------------->
The kernel config and materials to reproduce are available at:
https://download.01.org/0day-ci/archive/20251128/202511281432.47ac7e38-lkp-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org
=========================================================================================
compiler/cpufreq_governor/kconfig/mode/nr_task/rootfs/tbox_group/test/testcase:
gcc-14/performance/x86_64-rhel-9.4/process/100%/debian-13-x86_64-20250902.cgz/lkp-icl-2sp7/signal1/will-it-scale
commit:
v6.18-rc1
a97fbc3ee3 ("syscore: Pass context data to callbacks")
v6.18-rc1 a97fbc3ee3e2a536fafaff04f21
---------------- ---------------------------
%stddev %change %stddev
\ | \
71066 ±105% +119.1% 155709 ± 64% numa-meminfo.node1.AnonHugePages
34.80 ±105% +118.6% 76.08 ± 64% numa-vmstat.node1.nr_anon_transparent_hugepages
0.34 +0.2 0.52 ± 25% mpstat.cpu.all.idle%
0.01 +0.0 0.01 ± 3% mpstat.cpu.all.soft%
142.02 ± 5% +20.2% 170.65 ± 9% sched_debug.cfs_rq:/.runnable_avg.stddev
7.16 ± 5% +52.9% 10.94 ± 8% sched_debug.cpu.clock.stddev
1233 ± 4% -22.2% 960.17 ± 8% perf-c2c.DRAM.remote
2429 ± 3% -18.0% 1993 perf-c2c.HITM.local
1010 ± 5% -29.8% 709.00 ± 2% perf-c2c.HITM.remote
8382642 -3.2% 8118230 will-it-scale.64.processes
130978 -3.2% 126846 will-it-scale.per_process_ops
8382642 -3.2% 8118230 will-it-scale.workload
285565 -1.6% 280908 proc-vmstat.nr_active_anon
28063 +4.4% 29306 proc-vmstat.nr_mapped
117169 -3.9% 112656 proc-vmstat.nr_shmem
285565 -1.6% 280908 proc-vmstat.nr_zone_active_anon
38231 ± 33% -48.6% 19647 ± 35% proc-vmstat.numa_hint_faults
19706 ± 75% -65.5% 6790 ± 15% proc-vmstat.numa_hint_faults_local
9.391e+09 -3.1% 9.102e+09 perf-stat.i.branch-instructions
0.48 -0.0 0.45 perf-stat.i.branch-miss-rate%
45545162 -10.7% 40674481 perf-stat.i.branch-misses
11.08 ± 8% -2.3 8.81 ± 12% perf-stat.i.cache-miss-rate%
1811 +6.2% 1922 perf-stat.i.context-switches
5.09 +3.2% 5.26 perf-stat.i.cpi
117.53 -2.1% 115.06 perf-stat.i.cpu-migrations
216404 ± 2% +86.7% 404107 ± 4% perf-stat.i.cycles-between-cache-misses
3.87e+10 -3.1% 3.752e+10 perf-stat.i.instructions
0.20 -3.1% 0.19 perf-stat.i.ipc
0.49 -0.0 0.45 perf-stat.overall.branch-miss-rate%
5.10 +3.2% 5.26 perf-stat.overall.cpi
0.20 -3.1% 0.19 perf-stat.overall.ipc
9.361e+09 -3.1% 9.068e+09 perf-stat.ps.branch-instructions
45428240 -10.7% 40562730 perf-stat.ps.branch-misses
1806 +6.2% 1917 perf-stat.ps.context-switches
117.13 -2.1% 114.62 perf-stat.ps.cpu-migrations
3.858e+10 -3.1% 3.738e+10 perf-stat.ps.instructions
1.166e+13 -3.0% 1.131e+13 perf-stat.total.instructions
32.30 -6.5 25.85 perf-profile.calltrace.cycles-pp.get_signal.arch_do_signal_or_restart.exit_to_user_mode_loop.do_syscall_64.entry_SYSCALL_64_after_hwframe
31.42 -6.4 25.04 perf-profile.calltrace.cycles-pp.dequeue_signal.get_signal.arch_do_signal_or_restart.exit_to_user_mode_loop.do_syscall_64
30.28 -6.3 23.98 perf-profile.calltrace.cycles-pp.dec_rlimit_put_ucounts.__sigqueue_free.dequeue_signal.get_signal.arch_do_signal_or_restart
30.32 -6.3 24.02 perf-profile.calltrace.cycles-pp.__sigqueue_free.dequeue_signal.get_signal.arch_do_signal_or_restart.exit_to_user_mode_loop
37.60 -6.0 31.57 perf-profile.calltrace.cycles-pp.arch_do_signal_or_restart.exit_to_user_mode_loop.do_syscall_64.entry_SYSCALL_64_after_hwframe
37.68 -6.0 31.66 perf-profile.calltrace.cycles-pp.exit_to_user_mode_loop.do_syscall_64.entry_SYSCALL_64_after_hwframe
90.92 -0.5 90.43 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe
90.63 -0.5 90.14 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe
0.62 +0.0 0.65 perf-profile.calltrace.cycles-pp.check_xstate_in_sigframe.__fpu_restore_sig.fpu__restore_sig.restore_sigcontext.__x64_sys_rt_sigreturn
0.72 ± 2% +0.0 0.75 perf-profile.calltrace.cycles-pp.kmem_cache_alloc_noprof.__send_signal_locked.do_send_sig_info.do_send_specific.__x64_sys_tgkill
0.67 +0.0 0.70 perf-profile.calltrace.cycles-pp.__fpu_restore_sig.fpu__restore_sig.restore_sigcontext.__x64_sys_rt_sigreturn.do_syscall_64
2.54 +0.0 2.58 perf-profile.calltrace.cycles-pp.fpu__restore_sig.restore_sigcontext.__x64_sys_rt_sigreturn.do_syscall_64.entry_SYSCALL_64_after_hwframe
1.15 +0.0 1.19 perf-profile.calltrace.cycles-pp.fpu__clear_user_states.arch_do_signal_or_restart.exit_to_user_mode_loop.do_syscall_64.entry_SYSCALL_64_after_hwframe
1.78 +0.0 1.83 perf-profile.calltrace.cycles-pp.clear_bhb_loop
0.56 +0.1 0.61 perf-profile.calltrace.cycles-pp.rseq_ip_fixup.__rseq_handle_notify_resume.arch_do_signal_or_restart.exit_to_user_mode_loop.do_syscall_64
1.46 +0.1 1.51 perf-profile.calltrace.cycles-pp.__restore_fpregs_from_user.restore_fpregs_from_user.fpu__restore_sig.restore_sigcontext.__x64_sys_rt_sigreturn
1.62 +0.1 1.68 perf-profile.calltrace.cycles-pp.arch_do_signal_or_restart.exit_to_user_mode_loop.do_syscall_64.entry_SYSCALL_64_after_hwframe.handler
3.06 +0.1 3.11 perf-profile.calltrace.cycles-pp.restore_sigcontext.__x64_sys_rt_sigreturn.do_syscall_64.entry_SYSCALL_64_after_hwframe
1.69 +0.1 1.75 perf-profile.calltrace.cycles-pp.exit_to_user_mode_loop.do_syscall_64.entry_SYSCALL_64_after_hwframe.handler
1.91 +0.1 1.98 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.handler
1.88 +0.1 1.96 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.handler
0.96 +0.1 1.04 perf-profile.calltrace.cycles-pp.save_xstate_epilog.get_sigframe.x64_setup_rt_frame.arch_do_signal_or_restart.exit_to_user_mode_loop
0.93 +0.1 1.02 perf-profile.calltrace.cycles-pp.__rseq_handle_notify_resume.arch_do_signal_or_restart.exit_to_user_mode_loop.do_syscall_64.entry_SYSCALL_64_after_hwframe
4.46 +0.1 4.57 perf-profile.calltrace.cycles-pp.__x64_sys_rt_sigreturn.do_syscall_64.entry_SYSCALL_64_after_hwframe
2.18 +0.2 2.35 perf-profile.calltrace.cycles-pp.copy_fpstate_to_sigframe.get_sigframe.x64_setup_rt_frame.arch_do_signal_or_restart.exit_to_user_mode_loop
4.88 +0.3 5.14 perf-profile.calltrace.cycles-pp.handler
3.66 +0.3 3.95 perf-profile.calltrace.cycles-pp.get_sigframe.x64_setup_rt_frame.arch_do_signal_or_restart.exit_to_user_mode_loop.do_syscall_64
4.16 +0.3 4.50 perf-profile.calltrace.cycles-pp.x64_setup_rt_frame.arch_do_signal_or_restart.exit_to_user_mode_loop.do_syscall_64.entry_SYSCALL_64_after_hwframe
0.00 +0.5 0.51 perf-profile.calltrace.cycles-pp.restore_altstack.__x64_sys_rt_sigreturn.do_syscall_64.entry_SYSCALL_64_after_hwframe
0.00 +0.5 0.52 perf-profile.calltrace.cycles-pp.__get_user_nocheck_8.__x64_sys_rt_sigreturn.do_syscall_64.entry_SYSCALL_64_after_hwframe
45.91 +5.3 51.19 perf-profile.calltrace.cycles-pp.do_send_sig_info.do_send_specific.__x64_sys_tgkill.do_syscall_64.entry_SYSCALL_64_after_hwframe
46.73 +5.3 52.02 perf-profile.calltrace.cycles-pp.do_send_specific.__x64_sys_tgkill.do_syscall_64.entry_SYSCALL_64_after_hwframe
47.27 +5.3 52.57 perf-profile.calltrace.cycles-pp.__x64_sys_tgkill.do_syscall_64.entry_SYSCALL_64_after_hwframe
45.10 +5.3 50.43 perf-profile.calltrace.cycles-pp.__send_signal_locked.do_send_sig_info.do_send_specific.__x64_sys_tgkill.do_syscall_64
43.31 +5.3 48.65 perf-profile.calltrace.cycles-pp.inc_rlimit_get_ucounts.sig_get_ucounts.__send_signal_locked.do_send_sig_info.do_send_specific
43.39 +5.3 48.73 perf-profile.calltrace.cycles-pp.sig_get_ucounts.__send_signal_locked.do_send_sig_info.do_send_specific.__x64_sys_tgkill
32.36 -6.5 25.90 perf-profile.children.cycles-pp.get_signal
31.44 -6.4 25.06 perf-profile.children.cycles-pp.dequeue_signal
30.29 -6.3 23.99 perf-profile.children.cycles-pp.dec_rlimit_put_ucounts
30.33 -6.3 24.03 perf-profile.children.cycles-pp.__sigqueue_free
39.26 -6.0 33.30 perf-profile.children.cycles-pp.arch_do_signal_or_restart
39.38 -5.9 33.43 perf-profile.children.cycles-pp.exit_to_user_mode_loop
93.64 -0.3 93.30 perf-profile.children.cycles-pp.entry_SYSCALL_64_after_hwframe
93.28 -0.3 92.95 perf-profile.children.cycles-pp.do_syscall_64
0.36 ± 11% -0.1 0.25 perf-profile.children.cycles-pp.recalc_sigpending
0.61 ± 11% -0.1 0.50 ± 6% perf-profile.children.cycles-pp._raw_spin_lock_irq
0.43 -0.1 0.38 perf-profile.children.cycles-pp.fpregs_mark_activate
0.44 -0.0 0.40 perf-profile.children.cycles-pp.complete_signal
0.27 ± 5% -0.0 0.23 ± 3% perf-profile.children.cycles-pp.sysvec_apic_timer_interrupt
0.25 ± 6% -0.0 0.22 ± 5% perf-profile.children.cycles-pp.__sysvec_apic_timer_interrupt
0.25 ± 6% -0.0 0.22 ± 5% perf-profile.children.cycles-pp.hrtimer_interrupt
0.25 -0.0 0.22 ± 5% perf-profile.children.cycles-pp._raw_spin_lock_irqsave
0.29 ± 5% -0.0 0.26 ± 3% perf-profile.children.cycles-pp.asm_sysvec_apic_timer_interrupt
0.31 ± 2% -0.0 0.28 perf-profile.children.cycles-pp.set_current_blocked
0.05 +0.0 0.06 perf-profile.children.cycles-pp.fpu__alloc_mathframe
0.06 +0.0 0.07 perf-profile.children.cycles-pp.__cond_resched
0.20 +0.0 0.21 perf-profile.children.cycles-pp.syscall_return_via_sysret
0.24 +0.0 0.25 perf-profile.children.cycles-pp.__get_user_nocheck_4
0.39 +0.0 0.40 perf-profile.children.cycles-pp.entry_SYSCALL_64_safe_stack
0.72 +0.0 0.74 perf-profile.children.cycles-pp.entry_SYSRETQ_unsafe_stack
0.20 +0.0 0.22 perf-profile.children.cycles-pp.__get_user_8
0.22 ± 2% +0.0 0.24 perf-profile.children.cycles-pp.__put_user_8
0.46 ± 3% +0.0 0.48 perf-profile.children.cycles-pp.__memcg_slab_post_alloc_hook
0.69 +0.0 0.71 perf-profile.children.cycles-pp.__fpu_restore_sig
0.27 +0.0 0.30 perf-profile.children.cycles-pp.rseq_get_rseq_cs
0.63 +0.0 0.66 perf-profile.children.cycles-pp.check_xstate_in_sigframe
0.74 +0.0 0.77 perf-profile.children.cycles-pp.kmem_cache_alloc_noprof
0.50 +0.0 0.53 perf-profile.children.cycles-pp.restore_altstack
0.32 +0.0 0.35 perf-profile.children.cycles-pp.rseq_update_cpu_node_id
0.36 +0.0 0.39 perf-profile.children.cycles-pp.__put_user_nocheck_4
0.69 +0.0 0.73 perf-profile.children.cycles-pp.arch_exit_to_user_mode_prepare
0.37 +0.0 0.40 perf-profile.children.cycles-pp.__put_user_nocheck_8
2.06 +0.0 2.10 perf-profile.children.cycles-pp.__getpid
0.66 +0.0 0.70 perf-profile.children.cycles-pp._copy_from_user
2.55 +0.0 2.59 perf-profile.children.cycles-pp.fpu__restore_sig
0.87 +0.0 0.91 perf-profile.children.cycles-pp.its_return_thunk
1.21 +0.0 1.25 perf-profile.children.cycles-pp.fpu__clear_user_states
1.48 +0.0 1.52 perf-profile.children.cycles-pp.__restore_fpregs_from_user
2.39 +0.1 2.44 perf-profile.children.cycles-pp.clear_bhb_loop
0.09 ± 4% +0.1 0.14 ± 4% perf-profile.children.cycles-pp.shmem_file_write_iter
0.08 +0.1 0.14 ± 5% perf-profile.children.cycles-pp.generic_perform_write
0.00 +0.1 0.06 ± 8% perf-profile.children.cycles-pp.shmem_get_folio_gfp
0.10 ± 3% +0.1 0.16 ± 3% perf-profile.children.cycles-pp.perf_mmap__push
0.00 +0.1 0.06 ± 6% perf-profile.children.cycles-pp.shmem_write_begin
0.58 +0.1 0.64 perf-profile.children.cycles-pp.rseq_ip_fixup
3.08 +0.1 3.13 perf-profile.children.cycles-pp.restore_sigcontext
0.10 ± 3% +0.1 0.16 ± 4% perf-profile.children.cycles-pp.record__mmap_read_evlist
0.09 ± 5% +0.1 0.15 ± 4% perf-profile.children.cycles-pp.record__pushfn
0.12 ± 4% +0.1 0.18 ± 4% perf-profile.children.cycles-pp.handle_internal_command
0.12 ± 4% +0.1 0.18 ± 4% perf-profile.children.cycles-pp.main
0.12 ± 4% +0.1 0.18 ± 4% perf-profile.children.cycles-pp.run_builtin
0.11 ± 6% +0.1 0.17 ± 5% perf-profile.children.cycles-pp.cmd_record
0.98 +0.1 1.06 perf-profile.children.cycles-pp.__get_user_nocheck_8
1.36 +0.1 1.45 perf-profile.children.cycles-pp.native_irq_return_iret
1.00 +0.1 1.09 perf-profile.children.cycles-pp.save_xstate_epilog
0.95 +0.1 1.04 perf-profile.children.cycles-pp.__rseq_handle_notify_resume
4.52 +0.1 4.64 perf-profile.children.cycles-pp.__x64_sys_rt_sigreturn
2.27 +0.2 2.45 perf-profile.children.cycles-pp.copy_fpstate_to_sigframe
3.68 +0.2 3.86 perf-profile.children.cycles-pp.handler
3.68 +0.3 3.98 perf-profile.children.cycles-pp.get_sigframe
4.21 +0.3 4.55 perf-profile.children.cycles-pp.x64_setup_rt_frame
45.96 +5.3 51.25 perf-profile.children.cycles-pp.do_send_sig_info
46.77 +5.3 52.07 perf-profile.children.cycles-pp.do_send_specific
47.30 +5.3 52.60 perf-profile.children.cycles-pp.__x64_sys_tgkill
45.18 +5.3 50.51 perf-profile.children.cycles-pp.__send_signal_locked
43.40 +5.3 48.74 perf-profile.children.cycles-pp.sig_get_ucounts
43.32 +5.3 48.66 perf-profile.children.cycles-pp.inc_rlimit_get_ucounts
30.28 -6.3 23.98 perf-profile.self.cycles-pp.dec_rlimit_put_ucounts
0.34 ± 12% -0.1 0.23 ± 2% perf-profile.self.cycles-pp.recalc_sigpending
0.58 ± 12% -0.1 0.48 ± 6% perf-profile.self.cycles-pp._raw_spin_lock_irq
0.31 -0.1 0.25 ± 2% perf-profile.self.cycles-pp.complete_signal
0.36 -0.1 0.31 perf-profile.self.cycles-pp.fpregs_mark_activate
0.23 ± 2% -0.0 0.20 ± 3% perf-profile.self.cycles-pp._raw_spin_lock_irqsave
0.20 +0.0 0.21 perf-profile.self.cycles-pp.arch_do_signal_or_restart
0.07 +0.0 0.08 perf-profile.self.cycles-pp.collect_signal
0.40 +0.0 0.41 perf-profile.self.cycles-pp.__getpid
0.20 +0.0 0.21 perf-profile.self.cycles-pp.syscall_return_via_sysret
0.23 +0.0 0.24 ± 2% perf-profile.self.cycles-pp.__get_user_nocheck_4
0.15 ± 2% +0.0 0.16 ± 2% perf-profile.self.cycles-pp.__x64_sys_rt_sigreturn
0.26 +0.0 0.28 perf-profile.self.cycles-pp.kmem_cache_free
0.10 +0.0 0.11 ± 4% perf-profile.self.cycles-pp.signal_setup_done
0.70 +0.0 0.72 perf-profile.self.cycles-pp.entry_SYSRETQ_unsafe_stack
0.18 ± 2% +0.0 0.20 perf-profile.self.cycles-pp.__get_user_8
0.21 ± 3% +0.0 0.23 perf-profile.self.cycles-pp.__put_user_8
0.25 ± 3% +0.0 0.27 ± 2% perf-profile.self.cycles-pp.__memcg_slab_post_alloc_hook
0.47 +0.0 0.49 perf-profile.self.cycles-pp.get_signal
0.52 +0.0 0.55 perf-profile.self.cycles-pp.its_return_thunk
0.37 +0.0 0.40 perf-profile.self.cycles-pp.get_sigframe
0.34 +0.0 0.37 perf-profile.self.cycles-pp.__put_user_nocheck_4
0.31 +0.0 0.34 perf-profile.self.cycles-pp.rseq_update_cpu_node_id
0.34 +0.0 0.38 perf-profile.self.cycles-pp.__put_user_nocheck_8
0.37 ± 2% +0.0 0.41 perf-profile.self.cycles-pp.save_xstate_epilog
0.64 +0.0 0.68 perf-profile.self.cycles-pp.arch_exit_to_user_mode_prepare
0.64 +0.0 0.68 perf-profile.self.cycles-pp._copy_from_user
0.53 +0.0 0.57 perf-profile.self.cycles-pp.x64_setup_rt_frame
1.46 +0.0 1.51 perf-profile.self.cycles-pp.__restore_fpregs_from_user
2.36 +0.1 2.41 perf-profile.self.cycles-pp.clear_bhb_loop
1.00 +0.1 1.06 perf-profile.self.cycles-pp.fpu__clear_user_states
0.94 +0.1 1.01 perf-profile.self.cycles-pp.__get_user_nocheck_8
1.36 +0.1 1.45 perf-profile.self.cycles-pp.native_irq_return_iret
1.45 +0.1 1.55 perf-profile.self.cycles-pp.copy_fpstate_to_sigframe
43.32 +5.3 48.66 perf-profile.self.cycles-pp.inc_rlimit_get_ucounts
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/wiki