| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <20211210041514.GA11309@linux.intel.com>
Date: Fri, 10 Dec 2021 12:15:15 +0800
From: Carel Si <beibei.si@...el.com>
To: Dave Hansen <dave.hansen@...el.com>
Cc: kernel test robot <oliver.sang@...el.com>,
Thomas Gleixner <tglx@...utronix.de>,
Borislav Petkov <bp@...e.de>,
"Chang S. Bae" <chang.seok.bae@...el.com>,
LKML <linux-kernel@...r.kernel.org>, lkp@...ts.01.org,
lkp@...el.com, fengwei.yin@...el.com
Subject: Re: [LKP] Re: [x86/signal] 3aac3ebea0: will-it-scale.per_thread_ops
-11.9% regression
Hi Dave,
On Tue, Dec 07, 2021 at 03:14:38PM -0800, Dave Hansen wrote:
> On 12/6/21 5:21 PM, kernel test robot wrote:
> >
> > 1bdda24c4af64cd2 3aac3ebea08f2d342364f827c89
> > ---------------- ---------------------------
> > %stddev %change %stddev
> > \ | \
> > 980404 ± 3% -10.2% 880436 ± 2% will-it-scale.16.threads
> > 61274 ± 3% -10.2% 55027 ± 2% will-it-scale.per_thread_ops
> > 980404 ± 3% -10.2% 880436 ± 2% will-it-scale.workload
> > 9745749 ± 18% +26.8% 12356608 ± 4% meminfo.DirectMap2M
>
> Something else funky is going on here. Why would there all of a sudden
> be so many more 2M pages in the direct map? I also see gunk like
> interrupts on the network card going up. I can certainly see that
> happening if something else on the network was messing around.
>
> Granted, this was seen across several systems, but it's really odd. I
> guess I'll go try to dig up one of the actual ones where this was seen.
>
> I tried on a smaller Skylake system and I don't see any regression at
> all or any interesting delta in a perf profile.
>
> Oliver or Chang, could you try to reproduce this by hand on one of the
> suspect systems? Build:
>
> 1bdda24c4a ("signal: Add an optional check for altstack size")
>
> then run will-it-scale by hand. Then build:
>
> 3aac3ebea0 ("x86/signal: Implement sigaltstack size validation")
>
> and run it again. Also, do we see any higher core-count regressions?
> These all seem to happen with:
>
> mode=thread
> nr_task=16
>
> That's really odd to see that for these systems with probably ~50 cores
> each. I'd expect to see it get worse at higher core counts.
We tested 144 threads, it has -10.6% regression. Thanks.
=========================================================================================
compiler/cpufreq_governor/kconfig/mode/nr_task/rootfs/tbox_group/test/testcase/ucode:
gcc-9/performance/x86_64-rhel-8.3/thread/100%/debian-10.4-x86_64-20200603.cgz/lkp-hsw-4ex1/signal1/will-it-scale/0x16
commit:
1bdda24c4a ("signal: Add an optional check for altstack size")
3aac3ebea0 ("x86/signal: Implement sigaltstack size validation")
1bdda24c4af64cd2 3aac3ebea08f2d342364f827c89
---------------- ---------------------------
%stddev %change %stddev
\ | \
279993 -10.6% 250241 will-it-scale.144.threads
1943 -10.6% 1737 will-it-scale.per_thread_ops
279993 -10.6% 250241 will-it-scale.workload
3376415 ± 3% +1.7e+05 3546025 ± 4% syscalls.sys_getpid.noise.100%
32.92 -4.2 28.77 perf-profile.calltrace.cycles-pp.sigprocmask.__x64_sys_rt_sigprocmask.do_syscall_64.entry_SYSCALL_64_after_hwframe.raise
32.90 -4.2 28.75 perf-profile.calltrace.cycles-pp.__set_current_blocked.sigprocmask.__x64_sys_rt_sigprocmask.do_syscall_64.entry_SYSCALL_64_after_hwframe
32.96 -4.2 28.80 perf-profile.calltrace.cycles-pp.__x64_sys_rt_sigprocmask.do_syscall_64.entry_SYSCALL_64_after_hwframe.raise
32.60 -4.1 28.50 perf-profile.calltrace.cycles-pp._raw_spin_lock_irq.__set_current_blocked.sigprocmask.__x64_sys_rt_sigprocmask.do_syscall_64
32.52 -4.1 28.43 perf-profile.calltrace.cycles-pp.native_queued_spin_lock_slowpath._raw_spin_lock_irq.__set_current_blocked.sigprocmask.__x64_sys_rt_sigprocmask
32.80 -3.9 28.91 perf-profile.calltrace.cycles-pp.arch_do_signal_or_restart.exit_to_user_mode_prepare.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe
14.80 -2.5 12.30 perf-profile.calltrace.cycles-pp.__restore_rt
14.79 -2.5 12.29 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.__restore_rt
14.79 -2.5 12.29 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.__restore_rt
14.75 -2.5 12.25 perf-profile.calltrace.cycles-pp.__x64_sys_rt_sigreturn.do_syscall_64.entry_SYSCALL_64_after_hwframe.__restore_rt
14.71 -2.5 12.22 perf-profile.calltrace.cycles-pp.__set_current_blocked.__x64_sys_rt_sigreturn.do_syscall_64.entry_SYSCALL_64_after_hwframe.__restore_rt
14.57 -2.5 12.11 perf-profile.calltrace.cycles-pp._raw_spin_lock_irq.__set_current_blocked.__x64_sys_rt_sigreturn.do_syscall_64.entry_SYSCALL_64_after_hwframe
14.53 -2.5 12.07 perf-profile.calltrace.cycles-pp.native_queued_spin_lock_slowpath._raw_spin_lock_irq.__set_current_blocked.__x64_sys_rt_sigreturn.do_syscall_64
16.35 -2.0 14.33 perf-profile.calltrace.cycles-pp.handler
16.28 -2.0 14.27 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.handler
16.28 -2.0 14.27 perf-profile.calltrace.cycles-pp.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.handler
16.28 -2.0 14.27 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.handler
16.22 -2.0 14.21 perf-profile.calltrace.cycles-pp.exit_to_user_mode_prepare.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.handler
16.20 -2.0 14.19 perf-profile.calltrace.cycles-pp.__set_current_blocked.signal_setup_done.arch_do_signal_or_restart.exit_to_user_mode_prepare.syscall_exit_to_user_mode
16.20 -2.0 14.20 perf-profile.calltrace.cycles-pp.signal_setup_done.arch_do_signal_or_restart.exit_to_user_mode_prepare.syscall_exit_to_user_mode.do_syscall_64
16.06 -2.0 14.06 perf-profile.calltrace.cycles-pp._raw_spin_lock_irq.__set_current_blocked.signal_setup_done.arch_do_signal_or_restart.exit_to_user_mode_prepare
16.03 -2.0 14.04 perf-profile.calltrace.cycles-pp.native_queued_spin_lock_slowpath._raw_spin_lock_irq.__set_current_blocked.signal_setup_done.arch_do_signal_or_restart
17.06 -1.9 15.14 perf-profile.calltrace.cycles-pp.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.raise
16.50 -1.9 14.63 perf-profile.calltrace.cycles-pp.get_signal.arch_do_signal_or_restart.exit_to_user_mode_prepare.syscall_exit_to_user_mode.do_syscall_64
16.62 -1.9 14.76 perf-profile.calltrace.cycles-pp.exit_to_user_mode_prepare.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.raise
16.30 -1.8 14.45 perf-profile.calltrace.cycles-pp._raw_spin_lock_irq.get_signal.arch_do_signal_or_restart.exit_to_user_mode_prepare.syscall_exit_to_user_mode
16.27 -1.8 14.42 perf-profile.calltrace.cycles-pp.native_queued_spin_lock_slowpath._raw_spin_lock_irq.get_signal.arch_do_signal_or_restart.exit_to_user_mode_prepare
18.18 -1.5 16.70 perf-profile.calltrace.cycles-pp.__x64_sys_tgkill.do_syscall_64.entry_SYSCALL_64_after_hwframe.raise
18.16 -1.5 16.68 perf-profile.calltrace.cycles-pp.do_send_specific.do_tkill.__x64_sys_tgkill.do_syscall_64.entry_SYSCALL_64_after_hwframe
18.18 -1.5 16.69 perf-profile.calltrace.cycles-pp.do_tkill.__x64_sys_tgkill.do_syscall_64.entry_SYSCALL_64_after_hwframe.raise
18.10 -1.5 16.63 perf-profile.calltrace.cycles-pp.do_send_sig_info.do_send_specific.do_tkill.__x64_sys_tgkill.do_syscall_64
17.82 -1.4 16.38 perf-profile.calltrace.cycles-pp.__lock_task_sighand.do_send_sig_info.do_send_specific.do_tkill.__x64_sys_tgkill
17.82 -1.4 16.38 perf-profile.calltrace.cycles-pp._raw_spin_lock_irqsave.__lock_task_sighand.do_send_sig_info.do_send_specific.do_tkill
17.77 -1.4 16.33 perf-profile.calltrace.cycles-pp.native_queued_spin_lock_slowpath._raw_spin_lock_irqsave.__lock_task_sighand.do_send_sig_info.do_send_specific
68.68 +4.5 73.21 perf-profile.calltrace.cycles-pp.raise
68.45 +4.6 73.01 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.raise
68.47 +4.6 73.02 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.raise
0.00 +12.0 12.01 perf-profile.calltrace.cycles-pp.native_queued_spin_lock_slowpath._raw_spin_lock_irq.do_sigaltstack.restore_altstack.__x64_sys_rt_sigreturn
0.00 +12.0 12.04 perf-profile.calltrace.cycles-pp._raw_spin_lock_irq.do_sigaltstack.restore_altstack.__x64_sys_rt_sigreturn.do_syscall_64
0.00 +12.1 12.13 perf-profile.calltrace.cycles-pp.do_sigaltstack.restore_altstack.__x64_sys_rt_sigreturn.do_syscall_64.entry_SYSCALL_64_after_hwframe
0.00 +12.1 12.15 perf-profile.calltrace.cycles-pp.restore_altstack.__x64_sys_rt_sigreturn.do_syscall_64.entry_SYSCALL_64_after_hwframe.raise
0.00 +12.2 12.17 perf-profile.calltrace.cycles-pp.__x64_sys_rt_sigreturn.do_syscall_64.entry_SYSCALL_64_after_hwframe.raise
63.84 -8.7 55.19 perf-profile.children.cycles-pp.__set_current_blocked
32.92 -4.2 28.77 perf-profile.children.cycles-pp.sigprocmask
32.96 -4.2 28.80 perf-profile.children.cycles-pp.__x64_sys_rt_sigprocmask
33.34 -3.9 29.42 perf-profile.children.cycles-pp.syscall_exit_to_user_mode
32.80 -3.9 28.91 perf-profile.children.cycles-pp.arch_do_signal_or_restart
32.84 -3.9 28.98 perf-profile.children.cycles-pp.exit_to_user_mode_prepare
14.80 -2.5 12.30 perf-profile.children.cycles-pp.__restore_rt
16.35 -2.0 14.33 perf-profile.children.cycles-pp.handler
16.20 -2.0 14.20 perf-profile.children.cycles-pp.signal_setup_done
16.51 -1.9 14.64 perf-profile.children.cycles-pp.get_signal
18.18 -1.5 16.69 perf-profile.children.cycles-pp.do_tkill
18.18 -1.5 16.70 perf-profile.children.cycles-pp.__x64_sys_tgkill
18.16 -1.5 16.68 perf-profile.children.cycles-pp.do_send_specific
18.10 -1.5 16.63 perf-profile.children.cycles-pp.do_send_sig_info
17.82 -1.4 16.38 perf-profile.children.cycles-pp._raw_spin_lock_irqsave
17.82 -1.4 16.38 perf-profile.children.cycles-pp.__lock_task_sighand
0.05 -0.1 0.00 perf-profile.children.cycles-pp.trace_clock_x86_tsc
0.05 ± 7% -0.0 0.01 ±200% perf-profile.children.cycles-pp.copy_fpstate_to_sigframe
0.04 ± 44% -0.0 0.00 perf-profile.children.cycles-pp.ring_buffer_unlock_commit
0.22 ± 4% -0.0 0.19 ± 4% perf-profile.children.cycles-pp.syscall_trace_enter
0.20 ± 4% -0.0 0.17 ± 2% perf-profile.children.cycles-pp.ftrace_syscall_enter
0.19 ± 3% -0.0 0.17 ± 2% perf-profile.children.cycles-pp.trace_buffer_lock_reserve
0.15 ± 2% -0.0 0.13 ± 3% perf-profile.children.cycles-pp.recalc_sigpending
0.15 ± 3% -0.0 0.13 ± 3% perf-profile.children.cycles-pp.__set_task_blocked
0.15 ± 4% -0.0 0.13 perf-profile.children.cycles-pp.ring_buffer_lock_reserve
0.19 ± 3% -0.0 0.17 ± 2% perf-profile.children.cycles-pp.syscall_exit_to_user_mode_prepare
0.02 ± 99% -0.0 0.01 ±200% perf-profile.children.cycles-pp.__task_pid_nr_ns
0.10 ± 4% -0.0 0.09 perf-profile.children.cycles-pp.__rb_reserve_next
0.17 ± 2% -0.0 0.15 ± 3% perf-profile.children.cycles-pp.ftrace_syscall_exit
0.14 ± 5% -0.0 0.13 ± 3% perf-profile.children.cycles-pp._raw_spin_unlock_irqrestore
0.08 ± 5% -0.0 0.07 perf-profile.children.cycles-pp.dequeue_signal
0.10 ± 4% -0.0 0.09 ± 4% perf-profile.children.cycles-pp.syscall_return_via_sysret
0.08 ± 4% -0.0 0.07 ± 7% perf-profile.children.cycles-pp.__dequeue_signal
0.12 ± 6% -0.0 0.11 ± 3% perf-profile.children.cycles-pp.__entry_text_start
0.10 ± 3% -0.0 0.09 perf-profile.children.cycles-pp.__send_signal
0.03 ±100% -0.0 0.02 ±123% perf-profile.children.cycles-pp.__libc_start_main
0.03 ±100% -0.0 0.02 ±123% perf-profile.children.cycles-pp.main
0.03 ±100% -0.0 0.02 ±123% perf-profile.children.cycles-pp.run_builtin
0.01 ±223% -0.0 0.00 perf-profile.children.cycles-pp.perf_output_copy
0.01 ±223% -0.0 0.00 perf-profile.children.cycles-pp.security_task_kill
0.01 ±223% -0.0 0.00 perf-profile.children.cycles-pp.__libc_write
0.06 ± 6% -0.0 0.05 perf-profile.children.cycles-pp.trace_buffer_unlock_commit_regs
0.03 ±100% -0.0 0.02 ±123% perf-profile.children.cycles-pp.cmd_sched
0.03 ±100% -0.0 0.02 ±125% perf-profile.children.cycles-pp.cmd_record
0.68 -0.0 0.67 ± 2% perf-profile.children.cycles-pp.asm_sysvec_apic_timer_interrupt
0.03 ±100% -0.0 0.02 ±125% perf-profile.children.cycles-pp.record__finish_output
0.03 ±100% -0.0 0.02 ±125% perf-profile.children.cycles-pp.perf_session__process_events
0.06 ± 7% -0.0 0.06 perf-profile.children.cycles-pp.__setup_rt_frame
0.03 ±100% -0.0 0.02 ±125% perf-profile.children.cycles-pp.process_simple
0.62 -0.0 0.62 perf-profile.children.cycles-pp.sysvec_apic_timer_interrupt
0.06 ± 7% -0.0 0.06 perf-profile.children.cycles-pp.memcpy_erms
0.05 ± 7% -0.0 0.05 perf-profile.children.cycles-pp.native_irq_return_iret
0.05 ± 7% -0.0 0.05 perf-profile.children.cycles-pp.__sigqueue_alloc
0.14 ± 3% -0.0 0.14 ± 3% perf-profile.children.cycles-pp.unwind_next_frame
0.06 ± 6% -0.0 0.06 ± 6% perf-profile.children.cycles-pp.__get_user_nocheck_8
0.05 -0.0 0.05 perf-profile.children.cycles-pp.restore_sigcontext
0.06 ± 6% +0.0 0.06 ± 6% perf-profile.children.cycles-pp.__unwind_start
0.59 +0.0 0.59 perf-profile.children.cycles-pp.__sysvec_apic_timer_interrupt
0.06 ± 9% +0.0 0.06 ± 8% perf-profile.children.cycles-pp.asm_exc_page_fault
0.59 ± 2% +0.0 0.59 perf-profile.children.cycles-pp.hrtimer_interrupt
0.06 ± 6% +0.0 0.06 perf-profile.children.cycles-pp.perf_callchain_user
0.26 ± 4% +0.0 0.26 ± 3% perf-profile.children.cycles-pp.perf_prepare_sample
0.24 ± 3% +0.0 0.25 ± 3% perf-profile.children.cycles-pp.get_perf_callchain
0.06 ± 8% +0.0 0.06 perf-profile.children.cycles-pp.perf_output_sample
0.18 ± 3% +0.0 0.19 ± 2% perf-profile.children.cycles-pp.perf_callchain_kernel
0.35 ± 3% +0.0 0.36 ± 2% perf-profile.children.cycles-pp.update_curr
0.24 ± 3% +0.0 0.25 ± 2% perf-profile.children.cycles-pp.perf_callchain
0.43 ± 2% +0.0 0.43 perf-profile.children.cycles-pp.scheduler_tick
0.33 ± 3% +0.0 0.34 perf-profile.children.cycles-pp.perf_swevent_overflow
0.34 ± 3% +0.0 0.34 ± 2% perf-profile.children.cycles-pp.perf_trace_sched_stat_runtime
0.33 ± 3% +0.0 0.33 ± 2% perf-profile.children.cycles-pp.__perf_event_overflow
0.33 ± 3% +0.0 0.33 ± 2% perf-profile.children.cycles-pp.perf_event_output_forward
0.40 ± 2% +0.0 0.41 perf-profile.children.cycles-pp.task_tick_fair
0.47 ± 2% +0.0 0.48 perf-profile.children.cycles-pp.tick_sched_timer
0.46 ± 2% +0.0 0.47 perf-profile.children.cycles-pp.tick_sched_handle
0.45 ± 2% +0.0 0.46 perf-profile.children.cycles-pp.update_process_times
0.34 ± 2% +0.0 0.34 ± 2% perf-profile.children.cycles-pp.perf_tp_event
0.52 ± 2% +0.0 0.52 perf-profile.children.cycles-pp.__hrtimer_run_queues
0.00 +0.0 0.01 ±200% perf-profile.children.cycles-pp.ordered_events__queue
0.00 +0.0 0.01 ±200% perf-profile.children.cycles-pp.queue_event
99.63 +0.0 99.66 perf-profile.children.cycles-pp.entry_SYSCALL_64_after_hwframe
99.61 +0.0 99.65 perf-profile.children.cycles-pp.do_syscall_64
97.13 +0.2 97.31 perf-profile.children.cycles-pp.native_queued_spin_lock_slowpath
79.54 +1.6 81.16 perf-profile.children.cycles-pp._raw_spin_lock_irq
68.70 +4.5 73.22 perf-profile.children.cycles-pp.raise
14.79 +9.6 24.43 perf-profile.children.cycles-pp.__x64_sys_rt_sigreturn
0.00 +12.1 12.14 perf-profile.children.cycles-pp.do_sigaltstack
> _______________________________________________
> LKP mailing list -- lkp@...ts.01.org
> To unsubscribe send an email to lkp-leave@...ts.01.org
Powered by blists - more mailing lists