From: Kefeng Wang <wangkefeng.wang@huawei.com>
To: kernel test robot <oliver.sang@intel.com>
Cc: <oe-lkp@lists.linux.dev>, <lkp@intel.com>,
<linux-kernel@vger.kernel.org>,
Andrew Morton <akpm@linux-foundation.org>,
Matthew Wilcox <willy@infradead.org>,
"Huang, Ying" <ying.huang@intel.com>,
David Hildenbrand <david@redhat.com>,
Hugh Dickins <hughd@google.com>,
Mike Kravetz <mike.kravetz@oracle.com>, Zi Yan <ziy@nvidia.com>,
<linux-mm@kvack.org>, <feng.tang@intel.com>,
<fengwei.yin@intel.com>
Subject: Re: [linus:master] [mm] 728be28fae: stress-ng.rawsock.ops_per_sec 92.8% improvement
Date: Wed, 31 Jan 2024 20:51:02 +0800 [thread overview]
Message-ID: <ebdc0ef1-e17f-4382-af22-f46fafc93a67@huawei.com> (raw)
In-Reply-To: <202401311429.be2be5fc-oliver.sang@intel.com>
On 2024/1/31 15:20, kernel test robot wrote:
>
>
> Hello,
>
> kernel test robot noticed a 92.8% improvement of stress-ng.rawsock.ops_per_sec on:
>
>
> commit: 728be28fae8c838d52c91dce4867133798146357 ("mm: migrate: remove THP mapcount check in numamigrate_isolate_page()")
> https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git master
>
> testcase: stress-ng
> test machine: 64 threads 2 sockets Intel(R) Xeon(R) Gold 6346 CPU @ 3.10GHz (Ice Lake) with 256G memory
> parameters:
>
> nr_threads: 100%
> testtime: 60s
> class: network
> test: rawsock
> cpufreq_governor: performance
>
>
> we doubt the performance improvement is caused by:
>
> 81.65 -6.0 75.61 mpstat.cpu.all.soft%
>
> but we still cannot connect this with code change, so this report is just FYI
> what we observed in our stress-ng tests.
The change is only related to the PMD huge page numa fault，with it，it
seems that there are more numa folio migration，see numa_pages_migrated
and pgmigrate_success，2498->81759, but I don't know why it reduce 6%
for testcase, also skb should not use PMD huge page...
>
>
>
> Details are as below:
> -------------------------------------------------------------------------------------------------->
>
>
> The kernel config and materials to reproduce are available at:
> https://download.01.org/0day-ci/archive/20240131/202401311429.be2be5fc-oliver.sang@intel.com
>
> =========================================================================================
> class/compiler/cpufreq_governor/kconfig/nr_threads/rootfs/tbox_group/test/testcase/testtime:
> network/gcc-12/performance/x86_64-rhel-8.3/100%/debian-11.1-x86_64-20220510.cgz/lkp-icl-2sp8/rawsock/stress-ng/60s
>
> commit:
> a8ac4a767d ("mm: migrate: remove PageTransHuge check in numamigrate_isolate_page()")
> 728be28fae ("mm: migrate: remove THP mapcount check in numamigrate_isolate_page()")
>
> a8ac4a767dcd9d87 728be28fae8c838d52c91dce486
> ---------------- ---------------------------
> %stddev %change %stddev
> \ | \
> 2.893e+08 ± 3% +10.9% 3.208e+08 ± 6% cpuidle..time
> 327582 ± 3% +17.2% 383952 ± 4% cpuidle..usage
> 0.02 ± 33% +21362.2% 4.86 ± 24% vmstat.procs.b
> 6087 ± 4% +37.6% 8376 ± 4% vmstat.system.cs
> 6.18 ± 11% +0.5 6.71 ± 8% mpstat.cpu.all.idle%
> 0.02 ± 41% +0.1 0.14 ± 12% mpstat.cpu.all.iowait%
> 0.72 +0.0 0.72 mpstat.cpu.all.irq%
> 81.65 -6.0 75.61 mpstat.cpu.all.soft%
> 10.81 ± 2% +5.1 15.92 mpstat.cpu.all.sys%
> 0.62 ± 3% +0.3 0.89 mpstat.cpu.all.usr%
> 0.62 ± 3% +0.3 0.89 mpstat.cpu.all.usr%
> 294865 ± 5% +17.5% 346453 ± 6% turbostat.C1
> 6.38 ± 6% +0.7 7.08 ± 7% turbostat.C1%
> 114.67 ±172% +1663.8% 2022 ± 28% turbostat.POLL
> 581.00 ± 8% -13.3% 503.83 ± 5% perf-c2c.DRAM.local
> 6865 ± 5% +23.4% 8471 ± 4% perf-c2c.DRAM.remote
> 11672 ± 4% +9.5% 12781 ± 5% perf-c2c.HITM.local
> 2365 ± 14% +77.4% 4195 ± 7% perf-c2c.HITM.remote
> 14037 ± 5% +20.9% 16977 ± 5% perf-c2c.HITM.total
> 1.80 ± 4% +96.8% 3.54 ± 2% stress-ng.rawsock.MB_recv'd_per_sec
> 2.686e+08 ± 4% +92.8% 5.177e+08 ± 2% stress-ng.rawsock.ops
> 4365441 ± 4% +92.8% 8416438 ± 2% stress-ng.rawsock.ops_per_sec
> 157379 ± 3% +20.4% 189468 ± 4% stress-ng.time.involuntary_context_switches
> 35115 +53.9% 54057 ± 2% stress-ng.time.minor_page_faults
> 696.83 ± 2% +50.5% 1048 stress-ng.time.percent_of_cpu_this_job_got
> 430.92 ± 2% +49.4% 643.73 stress-ng.time.system_time
> 12.80 ± 4% +88.0% 24.08 ± 3% stress-ng.time.user_time
> 42314 ± 2% +168.1% 113467 ± 7% stress-ng.time.voluntary_context_switches
> 120896 ± 2% -2.4% 118016 proc-vmstat.nr_inactive_anon
> 55524 -1.1% 54939 proc-vmstat.nr_slab_unreclaimable
> 120896 ± 2% -2.4% 118016 proc-vmstat.nr_zone_inactive_anon
> 14930 ± 3% -24.7% 11241 ± 42% proc-vmstat.numa_hint_faults_local
> 2138444 -5.4% 2021976 proc-vmstat.numa_hit
> 2072186 -6.5% 1937601 proc-vmstat.numa_local
> 2498 ± 24% +3172.5% 81759 ± 7% proc-vmstat.numa_pages_migrated
> 274877 ± 2% +39.0% 382002 ± 3% proc-vmstat.numa_pte_updates
> 3790765 -4.8% 3607274 proc-vmstat.pgalloc_normal
> 398562 +5.7% 421173 proc-vmstat.pgfault
> 3640715 -4.9% 3462273 proc-vmstat.pgfree
> 2498 ± 24% +3172.5% 81759 ± 7% proc-vmstat.pgmigrate_success
> 316024 ± 3% +57.4% 497273 sched_debug.cfs_rq:/.avg_vruntime.avg
> 527031 ± 6% +31.9% 695410 ± 8% sched_debug.cfs_rq:/.avg_vruntime.max
> 225478 ± 4% +71.8% 387338 ± 4% sched_debug.cfs_rq:/.avg_vruntime.min
> 377256 ± 18% +50.1% 566136 ± 13% sched_debug.cfs_rq:/.left_vruntime.max
> 1122 ± 67% -73.0% 302.59 ± 84% sched_debug.cfs_rq:/.load_avg.avg
> 316031 ± 3% +57.3% 497273 sched_debug.cfs_rq:/.min_vruntime.avg
> 527031 ± 6% +31.9% 695409 ± 8% sched_debug.cfs_rq:/.min_vruntime.max
> 225478 ± 4% +71.8% 387338 ± 4% sched_debug.cfs_rq:/.min_vruntime.min
> 377256 ± 18% +50.1% 566136 ± 13% sched_debug.cfs_rq:/.right_vruntime.max
> 347.20 ± 9% +55.5% 540.03 ± 28% sched_debug.cfs_rq:/.util_avg.stddev
> 369.30 ± 5% +44.1% 532.30 ± 7% sched_debug.cfs_rq:/.util_est_enqueued.avg
> 712094 -18.8% 578101 ± 4% sched_debug.cpu.avg_idle.avg
> 5954 ± 2% +21.0% 7207 ± 4% sched_debug.cpu.nr_switches.avg
> 1891 ± 6% +70.6% 3227 ± 5% sched_debug.cpu.nr_switches.min
> 21.17 ± 28% +195.7% 62.58 ± 10% sched_debug.cpu.nr_uninterruptible.max
> -16.92 +340.4% -74.50 sched_debug.cpu.nr_uninterruptible.min
> 6.02 ± 8% +364.8% 28.01 ± 14% sched_debug.cpu.nr_uninterruptible.stddev
> 2.94 ± 37% +113.3% 6.27 ± 24% perf-sched.sch_delay.avg.ms.__cond_resched.aa_sk_perm.security_socket_recvmsg.sock_recvmsg.__sys_recvfrom
> 2.51 ± 48% +176.8% 6.96 ± 26% perf-sched.sch_delay.avg.ms.exit_to_user_mode_loop.exit_to_user_mode_prepare.irqentry_exit_to_user_mode.asm_sysvec_apic_timer_interrupt
> 0.83 ± 17% +247.9% 2.89 ± 6% perf-sched.sch_delay.avg.ms.exit_to_user_mode_loop.exit_to_user_mode_prepare.syscall_exit_to_user_mode.do_syscall_64
> 0.64 ± 17% +185.9% 1.82 ± 15% perf-sched.sch_delay.avg.ms.schedule_timeout.__skb_wait_for_more_packets.__skb_recv_datagram.skb_recv_datagram
> 0.07 ±133% +368.9% 0.34 ± 37% perf-sched.sch_delay.avg.ms.schedule_timeout.rcu_gp_fqs_loop.rcu_gp_kthread.kthread
> 1.60 ± 17% +87.9% 3.00 ± 6% perf-sched.sch_delay.avg.ms.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
> 126.33 ± 14% +134.5% 296.21 ± 18% perf-sched.sch_delay.max.ms.__cond_resched.aa_sk_perm.security_socket_recvmsg.sock_recvmsg.__sys_recvfrom
> 88.34 ± 20% +150.1% 220.96 ± 23% perf-sched.sch_delay.max.ms.exit_to_user_mode_loop.exit_to_user_mode_prepare.irqentry_exit_to_user_mode.asm_sysvec_apic_timer_interrupt
> 1.05 ±223% +2503.1% 27.44 ±114% perf-sched.sch_delay.max.ms.exit_to_user_mode_loop.exit_to_user_mode_prepare.irqentry_exit_to_user_mode.asm_sysvec_reschedule_ipi
> 198.20 ± 16% +140.4% 476.55 ± 23% perf-sched.sch_delay.max.ms.exit_to_user_mode_loop.exit_to_user_mode_prepare.syscall_exit_to_user_mode.do_syscall_64
> 169.34 ± 39% +99.6% 337.96 ± 28% perf-sched.sch_delay.max.ms.schedule_timeout.__skb_wait_for_more_packets.__skb_recv_datagram.skb_recv_datagram
> 16.98 ±128% +539.8% 108.67 ± 52% perf-sched.sch_delay.max.ms.schedule_timeout.rcu_gp_fqs_loop.rcu_gp_kthread.kthread
> 124.78 ± 22% +123.0% 278.26 ± 67% perf-sched.sch_delay.max.ms.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
> 1.08 ± 8% +146.1% 2.67 ± 9% perf-sched.total_sch_delay.average.ms
> 474.33 ± 75% +65.7% 786.13 ± 39% perf-sched.total_sch_delay.max.ms
> 64.64 ± 6% -36.3% 41.20 ± 11% perf-sched.total_wait_and_delay.average.ms
> 23788 ± 10% +49.8% 35644 ± 12% perf-sched.total_wait_and_delay.count.ms
> 63.55 ± 6% -39.4% 38.54 ± 12% perf-sched.total_wait_time.average.ms
> 13.44 ± 6% +17.6% 15.81 ± 8% perf-sched.wait_and_delay.avg.ms.__cond_resched.__wait_for_common.affine_move_task.__set_cpus_allowed_ptr.__sched_setaffinity
> 77.35 ± 7% -51.5% 37.54 ± 15% perf-sched.wait_and_delay.avg.ms.__cond_resched.aa_sk_perm.security_socket_recvmsg.sock_recvmsg.__sys_recvfrom
> 4.60 ± 22% -100.0% 0.00 perf-sched.wait_and_delay.avg.ms.__cond_resched.kmem_cache_alloc_node.__alloc_skb.alloc_skb_with_frags.sock_alloc_send_pskb
> 24.90 ± 4% -27.8% 17.98 ± 10% perf-sched.wait_and_delay.avg.ms.exit_to_user_mode_loop.exit_to_user_mode_prepare.syscall_exit_to_user_mode.do_syscall_64
> 19.65 ± 16% -64.7% 6.95 ± 17% perf-sched.wait_and_delay.avg.ms.schedule_timeout.__skb_wait_for_more_packets.__skb_recv_datagram.skb_recv_datagram
> 138.88 ± 3% -15.8% 116.89 ± 6% perf-sched.wait_and_delay.avg.ms.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
> 367.83 ± 10% -20.6% 292.17 ± 7% perf-sched.wait_and_delay.count.__cond_resched.__wait_for_common.affine_move_task.__set_cpus_allowed_ptr.__sched_setaffinity
> 289.00 ± 15% +70.2% 491.83 ± 18% perf-sched.wait_and_delay.count.__cond_resched.aa_sk_perm.security_socket_recvmsg.sock_recvmsg.__sys_recvfrom
> 145.50 ± 11% -100.0% 0.00 perf-sched.wait_and_delay.count.__cond_resched.kmem_cache_alloc_node.__alloc_skb.alloc_skb_with_frags.sock_alloc_send_pskb
> 2074 ± 17% +325.3% 8822 ± 25% perf-sched.wait_and_delay.count.schedule_timeout.__skb_wait_for_more_packets.__skb_recv_datagram.skb_recv_datagram
> 300.28 ± 12% +97.5% 592.92 ± 18% perf-sched.wait_and_delay.max.ms.__cond_resched.aa_sk_perm.security_socket_recvmsg.sock_recvmsg.__sys_recvfrom
> 129.21 ± 21% -100.0% 0.00 perf-sched.wait_and_delay.max.ms.__cond_resched.kmem_cache_alloc_node.__alloc_skb.alloc_skb_with_frags.sock_alloc_send_pskb
> 412.03 ± 23% +64.2% 676.47 ± 27% perf-sched.wait_and_delay.max.ms.schedule_timeout.__skb_wait_for_more_packets.__skb_recv_datagram.skb_recv_datagram
> 46.21 ±107% +293.5% 181.84 ± 20% perf-sched.wait_and_delay.max.ms.schedule_timeout.rcu_gp_fqs_loop.rcu_gp_kthread.kthread
> 13.35 ± 6% +16.4% 15.55 ± 9% perf-sched.wait_time.avg.ms.__cond_resched.__wait_for_common.affine_move_task.__set_cpus_allowed_ptr.__sched_setaffinity
> 74.41 ± 9% -58.0% 31.27 ± 17% perf-sched.wait_time.avg.ms.__cond_resched.aa_sk_perm.security_socket_recvmsg.sock_recvmsg.__sys_recvfrom
> 4.60 ± 22% +81.2% 8.33 ± 23% perf-sched.wait_time.avg.ms.__cond_resched.kmem_cache_alloc_node.__alloc_skb.alloc_skb_with_frags.sock_alloc_send_pskb
> 24.51 ± 14% -28.9% 17.44 ± 24% perf-sched.wait_time.avg.ms.do_wait.kernel_wait4.__do_sys_wait4.do_syscall_64
> 61.41 ± 27% -53.0% 28.86 ± 14% perf-sched.wait_time.avg.ms.exit_to_user_mode_loop.exit_to_user_mode_prepare.irqentry_exit_to_user_mode.asm_sysvec_apic_timer_interrupt
> 0.10 ±200% +38886.9% 39.12 ±110% perf-sched.wait_time.avg.ms.exit_to_user_mode_loop.exit_to_user_mode_prepare.irqentry_exit_to_user_mode.asm_sysvec_reschedule_ipi
> 24.07 ± 5% -37.3% 15.09 ± 12% perf-sched.wait_time.avg.ms.exit_to_user_mode_loop.exit_to_user_mode_prepare.syscall_exit_to_user_mode.do_syscall_64
> 19.01 ± 16% -73.0% 5.13 ± 19% perf-sched.wait_time.avg.ms.schedule_timeout.__skb_wait_for_more_packets.__skb_recv_datagram.skb_recv_datagram
> 137.28 ± 2% -17.0% 113.90 ± 6% perf-sched.wait_time.avg.ms.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
> 43.37 ±144% +253.4% 153.28 ± 42% perf-sched.wait_time.max.ms.__cond_resched.__wait_for_common.stop_two_cpus.migrate_swap.task_numa_migrate
> 0.00 ±159% +3.3e+05% 3.89 ±194% perf-sched.wait_time.max.ms.__cond_resched.stop_one_cpu.sched_exec.bprm_execve.part
> 1.63 ±221% +16137.0% 264.50 ±131% perf-sched.wait_time.max.ms.exit_to_user_mode_loop.exit_to_user_mode_prepare.irqentry_exit_to_user_mode.asm_sysvec_reschedule_ipi
> 2.405e+10 -4.6% 2.293e+10 perf-stat.i.branch-instructions
> 0.28 ± 3% +0.1 0.37 ± 2% perf-stat.i.branch-miss-rate%
> 51251291 ± 2% +36.9% 70177276 perf-stat.i.branch-misses
> 22.15 +3.6 25.72 perf-stat.i.cache-miss-rate%
> 1.025e+09 ± 7% -16.7% 8.538e+08 ± 8% perf-stat.i.cache-references
> 6143 ± 3% +39.5% 8569 ± 5% perf-stat.i.context-switches
> 1.62 +4.7% 1.70 perf-stat.i.cpi
> 566.63 +137.0% 1343 ± 4% perf-stat.i.cpu-migrations
> 3.81e+10 -5.6% 3.598e+10 perf-stat.i.dTLB-loads
> 0.00 ± 8% +0.0 0.00 ± 22% perf-stat.i.dTLB-store-miss-rate%
> 76418 ± 11% +48.1% 113175 ± 25% perf-stat.i.dTLB-store-misses
> 2.412e+10 -5.0% 2.292e+10 perf-stat.i.dTLB-stores
> 1.326e+11 -4.8% 1.263e+11 perf-stat.i.instructions
> 0.62 -4.6% 0.59 perf-stat.i.ipc
> 1436 ± 2% -13.5% 1242 ± 12% perf-stat.i.metric.K/sec
> 1359 -5.4% 1287 perf-stat.i.metric.M/sec
> 4398 ± 2% +7.9% 4746 ± 2% perf-stat.i.minor-faults
> 91.42 +1.9 93.36 perf-stat.i.node-load-miss-rate%
> 42725236 ± 3% +14.4% 48889377 ± 5% perf-stat.i.node-load-misses
> 3460356 ± 8% -25.1% 2591983 ± 13% perf-stat.i.node-loads
> 84.11 +1.3 85.44 perf-stat.i.node-store-miss-rate%
> 38774401 ± 6% +58.9% 61593486 ± 4% perf-stat.i.node-store-misses
> 6145595 ± 2% +38.9% 8536000 perf-stat.i.node-stores
> 4398 ± 2% +7.9% 4747 ± 2% perf-stat.i.page-faults
> 0.21 ± 2% +0.1 0.30 perf-stat.overall.branch-miss-rate%
> 22.71 +3.9 26.62 ± 2% perf-stat.overall.cache-miss-rate%
> 1.64 +4.3% 1.71 perf-stat.overall.cpi
> 0.00 ± 11% +0.0 0.00 ± 26% perf-stat.overall.dTLB-store-miss-rate%
> 0.61 -4.2% 0.58 perf-stat.overall.ipc
> 92.52 +2.5 94.99 perf-stat.overall.node-load-miss-rate%
> 86.30 +1.5 87.82 perf-stat.overall.node-store-miss-rate%
> 2.375e+10 -4.6% 2.266e+10 perf-stat.ps.branch-instructions
> 50207120 ± 2% +37.4% 68982796 perf-stat.ps.branch-misses
> 1.012e+09 ± 7% -16.6% 8.446e+08 ± 9% perf-stat.ps.cache-references
> 5893 ± 3% +39.8% 8240 ± 5% perf-stat.ps.context-switches
> 545.78 +140.7% 1313 ± 4% perf-stat.ps.cpu-migrations
> 3.764e+10 -5.5% 3.556e+10 perf-stat.ps.dTLB-loads
> 74052 ± 11% +48.6% 110013 ± 25% perf-stat.ps.dTLB-store-misses
> 2.383e+10 -4.9% 2.265e+10 perf-stat.ps.dTLB-stores
> 1.31e+11 -4.7% 1.248e+11 perf-stat.ps.instructions
> 4213 ± 2% +8.1% 4553 perf-stat.ps.minor-faults
> 42211937 ± 3% +14.6% 48392298 ± 5% perf-stat.ps.node-load-misses
> 3418585 ± 8% -24.9% 2565985 ± 13% perf-stat.ps.node-loads
> 38311557 ± 6% +59.1% 60965372 ± 4% perf-stat.ps.node-store-misses
> 6068466 ± 2% +39.1% 8443820 perf-stat.ps.node-stores
> 4214 ± 2% +8.1% 4553 perf-stat.ps.page-faults
> 8.441e+12 -4.5% 8.062e+12 perf-stat.total.instructions
> 88.83 -6.8 82.04 perf-profile.calltrace.cycles-pp.sendto.stress_rawsock_child
> 88.75 -6.8 81.97 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.sendto.stress_rawsock_child
> 88.92 -6.8 82.14 perf-profile.calltrace.cycles-pp.stress_rawsock_child
> 88.71 -6.8 81.93 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.sendto.stress_rawsock_child
> 88.63 -6.8 81.86 perf-profile.calltrace.cycles-pp.__x64_sys_sendto.do_syscall_64.entry_SYSCALL_64_after_hwframe.sendto.stress_rawsock_child
> 88.59 -6.8 81.84 perf-profile.calltrace.cycles-pp.__sys_sendto.__x64_sys_sendto.do_syscall_64.entry_SYSCALL_64_after_hwframe.sendto
> 88.36 -6.7 81.64 perf-profile.calltrace.cycles-pp.sock_sendmsg.__sys_sendto.__x64_sys_sendto.do_syscall_64.entry_SYSCALL_64_after_hwframe
> 88.24 -6.7 81.55 perf-profile.calltrace.cycles-pp.raw_sendmsg.sock_sendmsg.__sys_sendto.__x64_sys_sendto.do_syscall_64
> 87.30 -6.6 80.74 perf-profile.calltrace.cycles-pp.raw_send_hdrinc.raw_sendmsg.sock_sendmsg.__sys_sendto.__x64_sys_sendto
> 85.83 -6.4 79.46 perf-profile.calltrace.cycles-pp.ip_finish_output2.raw_send_hdrinc.raw_sendmsg.sock_sendmsg.__sys_sendto
> 85.79 -6.4 79.42 perf-profile.calltrace.cycles-pp.__dev_queue_xmit.ip_finish_output2.raw_send_hdrinc.raw_sendmsg.sock_sendmsg
> 85.43 -6.3 79.13 perf-profile.calltrace.cycles-pp.__local_bh_enable_ip.__dev_queue_xmit.ip_finish_output2.raw_send_hdrinc.raw_sendmsg
> 85.42 -6.3 79.12 perf-profile.calltrace.cycles-pp.do_softirq.__local_bh_enable_ip.__dev_queue_xmit.ip_finish_output2.raw_send_hdrinc
> 85.40 -6.3 79.10 perf-profile.calltrace.cycles-pp.__do_softirq.do_softirq.__local_bh_enable_ip.__dev_queue_xmit.ip_finish_output2
> 85.09 -6.3 78.81 perf-profile.calltrace.cycles-pp.net_rx_action.__do_softirq.do_softirq.__local_bh_enable_ip.__dev_queue_xmit
> 85.05 -6.3 78.78 perf-profile.calltrace.cycles-pp.process_backlog.__napi_poll.net_rx_action.__do_softirq.do_softirq
> 85.06 -6.3 78.78 perf-profile.calltrace.cycles-pp.__napi_poll.net_rx_action.__do_softirq.do_softirq.__local_bh_enable_ip
> 84.98 -6.3 78.70 perf-profile.calltrace.cycles-pp.__netif_receive_skb_one_core.process_backlog.__napi_poll.net_rx_action.__do_softirq
> 84.84 -6.3 78.59 perf-profile.calltrace.cycles-pp.ip_local_deliver_finish.__netif_receive_skb_one_core.process_backlog.__napi_poll.net_rx_action
> 84.54 -6.2 78.32 perf-profile.calltrace.cycles-pp.ip_protocol_deliver_rcu.ip_local_deliver_finish.__netif_receive_skb_one_core.process_backlog.__napi_poll
> 84.02 -6.2 77.83 perf-profile.calltrace.cycles-pp.raw_v4_input.ip_protocol_deliver_rcu.ip_local_deliver_finish.__netif_receive_skb_one_core.process_backlog
> 63.80 -4.6 59.18 perf-profile.calltrace.cycles-pp.raw_rcv.raw_v4_input.ip_protocol_deliver_rcu.ip_local_deliver_finish.__netif_receive_skb_one_core
> 5.12 ± 4% -1.7 3.43 ± 8% perf-profile.calltrace.cycles-pp.__xfrm_policy_check2.raw_rcv.raw_v4_input.ip_protocol_deliver_rcu.ip_local_deliver_finish
> 5.24 ± 3% -1.1 4.13 ± 2% perf-profile.calltrace.cycles-pp.ipv4_pktinfo_prepare.raw_rcv.raw_v4_input.ip_protocol_deliver_rcu.ip_local_deliver_finish
> 3.47 ± 4% -0.9 2.55 ± 5% perf-profile.calltrace.cycles-pp.sk_filter_trim_cap.sock_queue_rcv_skb_reason.raw_rcv.raw_v4_input.ip_protocol_deliver_rcu
> 6.09 -0.6 5.50 perf-profile.calltrace.cycles-pp.skb_clone.raw_v4_input.ip_protocol_deliver_rcu.ip_local_deliver_finish.__netif_receive_skb_one_core
> 4.82 -0.6 4.26 ± 2% perf-profile.calltrace.cycles-pp.kfree_skb_reason.raw_rcv.raw_v4_input.ip_protocol_deliver_rcu.ip_local_deliver_finish
> 43.83 -0.6 43.27 perf-profile.calltrace.cycles-pp.sock_queue_rcv_skb_reason.raw_rcv.raw_v4_input.ip_protocol_deliver_rcu.ip_local_deliver_finish
> 2.28 ± 2% -0.5 1.75 ± 4% perf-profile.calltrace.cycles-pp.rcu_core.__do_softirq.run_ksoftirqd.smpboot_thread_fn.kthread
> 2.28 ± 2% -0.5 1.75 ± 4% perf-profile.calltrace.cycles-pp.__do_softirq.run_ksoftirqd.smpboot_thread_fn.kthread.ret_from_fork
> 2.28 ± 2% -0.5 1.75 ± 4% perf-profile.calltrace.cycles-pp.run_ksoftirqd.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
> 2.27 ± 2% -0.5 1.74 ± 4% perf-profile.calltrace.cycles-pp.rcu_do_batch.rcu_core.__do_softirq.run_ksoftirqd.smpboot_thread_fn
> 2.25 ± 2% -0.5 1.73 ± 4% perf-profile.calltrace.cycles-pp.dst_destroy.rcu_do_batch.rcu_core.__do_softirq.run_ksoftirqd
> 2.28 ± 2% -0.5 1.76 ± 4% perf-profile.calltrace.cycles-pp.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
> 2.29 ± 2% -0.5 1.77 ± 4% perf-profile.calltrace.cycles-pp.kthread.ret_from_fork.ret_from_fork_asm
> 2.29 ± 2% -0.5 1.77 ± 4% perf-profile.calltrace.cycles-pp.ret_from_fork.ret_from_fork_asm
> 2.29 ± 2% -0.5 1.77 ± 4% perf-profile.calltrace.cycles-pp.ret_from_fork_asm
> 2.92 -0.5 2.41 perf-profile.calltrace.cycles-pp.kmem_cache_free.raw_rcv.raw_v4_input.ip_protocol_deliver_rcu.ip_local_deliver_finish
> 2.03 ± 2% -0.5 1.54 ± 4% perf-profile.calltrace.cycles-pp.percpu_counter_add_batch.dst_destroy.rcu_do_batch.rcu_core.__do_softirq
> 3.92 -0.4 3.49 ± 2% perf-profile.calltrace.cycles-pp.skb_release_data.kfree_skb_reason.raw_rcv.raw_v4_input.ip_protocol_deliver_rcu
> 1.53 ± 2% -0.4 1.13 ± 5% perf-profile.calltrace.cycles-pp._raw_spin_lock.percpu_counter_add_batch.dst_destroy.rcu_do_batch.rcu_core
> 4.85 ± 2% -0.4 4.46 perf-profile.calltrace.cycles-pp.kmem_cache_alloc.skb_clone.raw_v4_input.ip_protocol_deliver_rcu.ip_local_deliver_finish
> 1.51 ± 2% -0.4 1.11 ± 5% perf-profile.calltrace.cycles-pp.native_queued_spin_lock_slowpath._raw_spin_lock.percpu_counter_add_batch.dst_destroy.rcu_do_batch
> 2.76 -0.2 2.59 ± 2% perf-profile.calltrace.cycles-pp.dst_release.ipv4_pktinfo_prepare.raw_rcv.raw_v4_input.ip_protocol_deliver_rcu
> 0.92 -0.1 0.78 perf-profile.calltrace.cycles-pp.security_sock_rcv_skb.sk_filter_trim_cap.sock_queue_rcv_skb_reason.raw_rcv.raw_v4_input
> 0.72 -0.1 0.62 ± 4% perf-profile.calltrace.cycles-pp.sock_alloc_send_pskb.raw_send_hdrinc.raw_sendmsg.sock_sendmsg.__sys_sendto
> 0.74 ± 4% +0.4 1.11 perf-profile.calltrace.cycles-pp.copyout._copy_to_iter.__skb_datagram_iter.skb_copy_datagram_iter.raw_recvmsg
> 0.79 ± 4% +0.4 1.22 perf-profile.calltrace.cycles-pp._copy_to_iter.__skb_datagram_iter.skb_copy_datagram_iter.raw_recvmsg.inet_recvmsg
> 0.39 ± 71% +0.5 0.90 ± 4% perf-profile.calltrace.cycles-pp.___slab_alloc.kmem_cache_alloc.skb_clone.raw_v4_input.ip_protocol_deliver_rcu
> 0.95 ± 3% +0.6 1.50 ± 2% perf-profile.calltrace.cycles-pp.skb_release_data.consume_skb.raw_recvmsg.inet_recvmsg.sock_recvmsg
> 9.05 +0.6 9.63 perf-profile.calltrace.cycles-pp.__skb_clone.raw_v4_input.ip_protocol_deliver_rcu.ip_local_deliver_finish.__netif_receive_skb_one_core
> 0.00 +0.6 0.58 ± 8% perf-profile.calltrace.cycles-pp.__unfreeze_partials.raw_recvmsg.inet_recvmsg.sock_recvmsg.__sys_recvfrom
> 0.83 ± 5% +0.6 1.43 ± 2% perf-profile.calltrace.cycles-pp.check_heap_object.__check_object_size.simple_copy_to_iter.__skb_datagram_iter.skb_copy_datagram_iter
> 5.27 +0.7 6.00 perf-profile.calltrace.cycles-pp.__copy_skb_header.__skb_clone.raw_v4_input.ip_protocol_deliver_rcu.ip_local_deliver_finish
> 0.00 +0.8 0.76 perf-profile.calltrace.cycles-pp.__slab_free.raw_recvmsg.inet_recvmsg.sock_recvmsg.__sys_recvfrom
> 0.00 +0.8 0.77 ± 2% perf-profile.calltrace.cycles-pp.sock_rfree.skb_release_head_state.consume_skb.raw_recvmsg.inet_recvmsg
> 0.00 +0.8 0.79 ± 2% perf-profile.calltrace.cycles-pp.move_addr_to_user.__sys_recvfrom.__x64_sys_recvfrom.do_syscall_64.entry_SYSCALL_64_after_hwframe
> 0.00 +0.8 0.80 ± 2% perf-profile.calltrace.cycles-pp.skb_release_head_state.consume_skb.raw_recvmsg.inet_recvmsg.sock_recvmsg
> 1.00 ± 6% +0.8 1.83 ± 2% perf-profile.calltrace.cycles-pp.__check_object_size.simple_copy_to_iter.__skb_datagram_iter.skb_copy_datagram_iter.raw_recvmsg
> 1.30 ± 5% +0.9 2.17 ± 2% perf-profile.calltrace.cycles-pp.native_queued_spin_lock_slowpath._raw_spin_lock_irqsave.__skb_try_recv_datagram.__skb_recv_datagram.skb_recv_datagram
> 1.07 ± 5% +0.9 1.96 ± 2% perf-profile.calltrace.cycles-pp.simple_copy_to_iter.__skb_datagram_iter.skb_copy_datagram_iter.raw_recvmsg.inet_recvmsg
> 1.37 ± 4% +1.0 2.33 ± 2% perf-profile.calltrace.cycles-pp.consume_skb.raw_recvmsg.inet_recvmsg.sock_recvmsg.__sys_recvfrom
> 1.80 ± 5% +1.4 3.19 ± 2% perf-profile.calltrace.cycles-pp._raw_spin_lock_irqsave.__skb_try_recv_datagram.__skb_recv_datagram.skb_recv_datagram.raw_recvmsg
> 1.44 ± 7% +1.4 2.84 ± 2% perf-profile.calltrace.cycles-pp.sock_def_readable.__sock_queue_rcv_skb.sock_queue_rcv_skb_reason.raw_rcv.raw_v4_input
> 2.01 ± 5% +1.5 3.47 ± 2% perf-profile.calltrace.cycles-pp.__skb_datagram_iter.skb_copy_datagram_iter.raw_recvmsg.inet_recvmsg.sock_recvmsg
> 2.04 ± 5% +1.5 3.51 ± 2% perf-profile.calltrace.cycles-pp.skb_copy_datagram_iter.raw_recvmsg.inet_recvmsg.sock_recvmsg.__sys_recvfrom
> 1.97 ± 5% +1.6 3.53 ± 2% perf-profile.calltrace.cycles-pp.__skb_try_recv_datagram.__skb_recv_datagram.skb_recv_datagram.raw_recvmsg.inet_recvmsg
> 2.03 ± 6% +1.6 3.67 ± 2% perf-profile.calltrace.cycles-pp.__skb_recv_datagram.skb_recv_datagram.raw_recvmsg.inet_recvmsg.sock_recvmsg
> 2.05 ± 5% +1.7 3.70 ± 2% perf-profile.calltrace.cycles-pp.skb_recv_datagram.raw_recvmsg.inet_recvmsg.sock_recvmsg.__sys_recvfrom
> 8.31 ± 3% +1.7 9.99 perf-profile.calltrace.cycles-pp.native_queued_spin_lock_slowpath._raw_spin_lock_irqsave.__sock_queue_rcv_skb.sock_queue_rcv_skb_reason.raw_rcv
> 9.31 ± 3% +2.2 11.47 perf-profile.calltrace.cycles-pp._raw_spin_lock_irqsave.__sock_queue_rcv_skb.sock_queue_rcv_skb_reason.raw_rcv.raw_v4_input
> 7.09 ± 5% +5.7 12.79 ± 2% perf-profile.calltrace.cycles-pp.raw_recvmsg.inet_recvmsg.sock_recvmsg.__sys_recvfrom.__x64_sys_recvfrom
> 7.13 ± 5% +5.7 12.88 ± 2% perf-profile.calltrace.cycles-pp.inet_recvmsg.sock_recvmsg.__sys_recvfrom.__x64_sys_recvfrom.do_syscall_64
> 7.29 ± 5% +5.9 13.20 ± 2% perf-profile.calltrace.cycles-pp.sock_recvmsg.__sys_recvfrom.__x64_sys_recvfrom.do_syscall_64.entry_SYSCALL_64_after_hwframe
> 8.08 ± 5% +6.7 14.74 ± 2% perf-profile.calltrace.cycles-pp.__sys_recvfrom.__x64_sys_recvfrom.do_syscall_64.entry_SYSCALL_64_after_hwframe.recvfrom
> 8.11 ± 5% +6.7 14.79 ± 2% perf-profile.calltrace.cycles-pp.__x64_sys_recvfrom.do_syscall_64.entry_SYSCALL_64_after_hwframe.recvfrom.stress_oomable_child
> 8.20 ± 5% +6.8 14.98 ± 2% perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.recvfrom.stress_oomable_child
> 8.26 ± 5% +6.8 15.08 ± 2% perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.recvfrom.stress_oomable_child
> 8.47 ± 5% +7.0 15.50 ± 2% perf-profile.calltrace.cycles-pp.recvfrom.stress_oomable_child
> 8.72 ± 5% +7.3 16.00 ± 2% perf-profile.calltrace.cycles-pp.stress_oomable_child
> 88.87 -6.8 82.07 perf-profile.children.cycles-pp.sendto
> 87.77 -6.8 80.98 perf-profile.children.cycles-pp.__do_softirq
> 88.92 -6.8 82.14 perf-profile.children.cycles-pp.stress_rawsock_child
> 88.64 -6.8 81.87 perf-profile.children.cycles-pp.__x64_sys_sendto
> 88.60 -6.7 81.85 perf-profile.children.cycles-pp.__sys_sendto
> 88.36 -6.7 81.65 perf-profile.children.cycles-pp.sock_sendmsg
> 88.24 -6.7 81.56 perf-profile.children.cycles-pp.raw_sendmsg
> 87.31 -6.6 80.74 perf-profile.children.cycles-pp.raw_send_hdrinc
> 85.83 -6.4 79.46 perf-profile.children.cycles-pp.ip_finish_output2
> 85.79 -6.4 79.42 perf-profile.children.cycles-pp.__dev_queue_xmit
> 85.10 -6.3 78.81 perf-profile.children.cycles-pp.net_rx_action
> 85.06 -6.3 78.78 perf-profile.children.cycles-pp.process_backlog
> 85.06 -6.3 78.78 perf-profile.children.cycles-pp.__napi_poll
> 84.98 -6.3 78.71 perf-profile.children.cycles-pp.__netif_receive_skb_one_core
> 85.51 -6.3 79.24 perf-profile.children.cycles-pp.__local_bh_enable_ip
> 85.49 -6.3 79.22 perf-profile.children.cycles-pp.do_softirq
> 84.84 -6.3 78.59 perf-profile.children.cycles-pp.ip_local_deliver_finish
> 84.55 -6.2 78.32 perf-profile.children.cycles-pp.ip_protocol_deliver_rcu
> 84.52 -6.2 78.30 perf-profile.children.cycles-pp.raw_v4_input
> 64.71 -4.8 59.94 perf-profile.children.cycles-pp.raw_rcv
> 5.21 ± 4% -1.7 3.52 ± 8% perf-profile.children.cycles-pp.__xfrm_policy_check2
> 5.44 ± 3% -1.1 4.32 ± 2% perf-profile.children.cycles-pp.ipv4_pktinfo_prepare
> 3.66 ± 4% -0.9 2.72 ± 4% perf-profile.children.cycles-pp.sk_filter_trim_cap
> 6.31 -0.6 5.68 perf-profile.children.cycles-pp.skb_clone
> 5.12 -0.6 4.50 ± 2% perf-profile.children.cycles-pp.kfree_skb_reason
> 44.12 -0.6 43.52 perf-profile.children.cycles-pp.sock_queue_rcv_skb_reason
> 2.28 ± 2% -0.5 1.75 ± 4% perf-profile.children.cycles-pp.run_ksoftirqd
> 2.28 ± 2% -0.5 1.76 ± 4% perf-profile.children.cycles-pp.smpboot_thread_fn
> 2.29 ± 2% -0.5 1.77 ± 4% perf-profile.children.cycles-pp.kthread
> 2.29 ± 2% -0.5 1.77 ± 4% perf-profile.children.cycles-pp.ret_from_fork
> 2.29 ± 2% -0.5 1.77 ± 4% perf-profile.children.cycles-pp.ret_from_fork_asm
> 2.53 ± 2% -0.5 2.04 ± 4% perf-profile.children.cycles-pp.dst_destroy
> 2.55 ± 3% -0.5 2.06 ± 4% perf-profile.children.cycles-pp.rcu_do_batch
> 2.55 ± 2% -0.5 2.07 ± 4% perf-profile.children.cycles-pp.rcu_core
> 5.38 -0.5 4.90 perf-profile.children.cycles-pp.kmem_cache_alloc
> 2.40 ± 2% -0.5 1.93 ± 4% perf-profile.children.cycles-pp.percpu_counter_add_batch
> 1.73 ± 2% -0.4 1.35 ± 5% perf-profile.children.cycles-pp._raw_spin_lock
> 3.32 -0.4 2.95 perf-profile.children.cycles-pp.kmem_cache_free
> 2.92 -0.2 2.73 ± 2% perf-profile.children.cycles-pp.dst_release
> 1.22 -0.2 1.05 perf-profile.children.cycles-pp.security_sock_rcv_skb
> 0.70 -0.1 0.58 ± 2% perf-profile.children.cycles-pp.memcg_slab_post_alloc_hook
> 0.73 -0.1 0.62 ± 4% perf-profile.children.cycles-pp.sock_alloc_send_pskb
> 0.36 ± 3% -0.1 0.28 ± 9% perf-profile.children.cycles-pp.skb_set_owner_w
> 0.48 ± 2% -0.1 0.40 ± 3% perf-profile.children.cycles-pp.kfree_skbmem
> 0.41 -0.1 0.35 ± 2% perf-profile.children.cycles-pp.skb_push
> 0.35 ± 3% -0.1 0.29 ± 3% perf-profile.children.cycles-pp.should_failslab
> 0.41 -0.1 0.35 ± 2% perf-profile.children.cycles-pp.apparmor_socket_sock_rcv_skb
> 0.52 ± 2% -0.0 0.47 ± 3% perf-profile.children.cycles-pp.ip_mc_sf_allow
> 0.25 -0.0 0.20 ± 3% perf-profile.children.cycles-pp.dev_hard_start_xmit
> 0.23 -0.0 0.19 ± 4% perf-profile.children.cycles-pp.loopback_xmit
> 0.38 ± 2% -0.0 0.35 perf-profile.children.cycles-pp.__ip_select_ident
> 0.10 ± 7% -0.0 0.06 ± 11% perf-profile.children.cycles-pp.sock_wfree
> 0.23 ± 2% -0.0 0.20 ± 3% perf-profile.children.cycles-pp.dst_alloc
> 0.24 -0.0 0.21 ± 3% perf-profile.children.cycles-pp.rt_dst_alloc
> 0.07 -0.0 0.05 ± 8% perf-profile.children.cycles-pp.validate_xmit_skb
> 0.06 +0.0 0.07 ± 6% perf-profile.children.cycles-pp.stress_rawsock_client
> 0.27 ± 5% +0.0 0.30 perf-profile.children.cycles-pp.sysvec_apic_timer_interrupt
> 0.14 ± 6% +0.0 0.17 ± 3% perf-profile.children.cycles-pp.rt_set_nexthop
> 0.03 ± 70% +0.0 0.06 ± 7% perf-profile.children.cycles-pp.__cond_resched
> 0.06 ± 7% +0.0 0.10 ± 4% perf-profile.children.cycles-pp.syscall_return_via_sysret
> 0.06 ± 8% +0.0 0.10 ± 3% perf-profile.children.cycles-pp._raw_spin_unlock_irqrestore
> 0.06 ± 9% +0.1 0.11 ± 3% perf-profile.children.cycles-pp.__skb_try_recv_from_queue
> 0.11 ± 5% +0.1 0.16 ± 2% perf-profile.children.cycles-pp.syscall_exit_to_user_mode
> 0.00 +0.1 0.05 ± 8% perf-profile.children.cycles-pp.import_single_range
> 0.07 ± 7% +0.1 0.12 ± 3% perf-profile.children.cycles-pp.__virt_addr_valid
> 0.00 +0.1 0.06 ± 9% perf-profile.children.cycles-pp.stress_hash_mulxror32
> 0.01 ±223% +0.1 0.06 ± 7% perf-profile.children.cycles-pp.exit_to_user_mode_prepare
> 0.00 +0.1 0.06 ± 6% perf-profile.children.cycles-pp.syscall_enter_from_user_mode
> 0.00 +0.1 0.06 ± 6% perf-profile.children.cycles-pp.check_stack_object
> 0.08 ± 10% +0.1 0.14 ± 5% perf-profile.children.cycles-pp.put_cpu_partial
> 0.07 ± 5% +0.1 0.15 ± 12% perf-profile.children.cycles-pp.__check_heap_object
> 0.15 ± 2% +0.1 0.23 ± 6% perf-profile.children.cycles-pp.aa_sk_perm
> 0.09 ± 10% +0.1 0.17 ± 2% perf-profile.children.cycles-pp.__put_user_nocheck_4
> 0.12 ± 5% +0.1 0.20 ± 2% perf-profile.children.cycles-pp.__fget_light
> 0.10 ± 5% +0.1 0.20 ± 2% perf-profile.children.cycles-pp.__get_user_4
> 0.14 ± 6% +0.1 0.24 ± 4% perf-profile.children.cycles-pp.entry_SYSRETQ_unsafe_stack
> 0.15 ± 7% +0.1 0.26 ± 2% perf-profile.children.cycles-pp.sockfd_lookup_light
> 0.19 ± 5% +0.1 0.31 ± 2% perf-profile.children.cycles-pp.entry_SYSCALL_64
> 0.13 ± 7% +0.1 0.26 ± 2% perf-profile.children.cycles-pp.stress_rawsock_server
> 0.16 ± 19% +0.1 0.29 ± 4% perf-profile.children.cycles-pp._copy_to_user
> 0.13 ± 9% +0.1 0.26 ± 6% perf-profile.children.cycles-pp.security_socket_recvmsg
> 0.15 ± 8% +0.2 0.36 ± 3% perf-profile.children.cycles-pp.is_vmalloc_addr
> 0.20 ± 14% +0.2 0.40 ± 7% perf-profile.children.cycles-pp.get_partial_node
> 0.99 ± 2% +0.3 1.31 perf-profile.children.cycles-pp.skb_release_head_state
> 0.70 ± 8% +0.3 1.04 ± 4% perf-profile.children.cycles-pp.___slab_alloc
> 0.25 ± 13% +0.4 0.60 ± 8% perf-profile.children.cycles-pp.__unfreeze_partials
> 0.84 ± 3% +0.4 1.21 perf-profile.children.cycles-pp.__slab_free
> 0.40 ± 6% +0.4 0.78 ± 2% perf-profile.children.cycles-pp.sock_rfree
> 0.75 ± 4% +0.4 1.15 perf-profile.children.cycles-pp.copyout
> 0.43 ± 8% +0.4 0.83 ± 2% perf-profile.children.cycles-pp.move_addr_to_user
> 0.79 ± 4% +0.4 1.23 perf-profile.children.cycles-pp._copy_to_iter
> 9.26 +0.5 9.81 perf-profile.children.cycles-pp.__skb_clone
> 1.02 ± 4% +0.7 1.73 ± 2% perf-profile.children.cycles-pp.check_heap_object
> 5.36 +0.7 6.08 perf-profile.children.cycles-pp.__copy_skb_header
> 1.23 ± 5% +0.9 2.09 ± 2% perf-profile.children.cycles-pp.__check_object_size
> 1.08 ± 5% +0.9 1.97 ± 2% perf-profile.children.cycles-pp.simple_copy_to_iter
> 1.57 ± 3% +0.9 2.51 ± 2% perf-profile.children.cycles-pp.consume_skb
> 1.44 ± 6% +1.4 2.84 ± 2% perf-profile.children.cycles-pp.sock_def_readable
> 2.02 ± 5% +1.5 3.48 ± 2% perf-profile.children.cycles-pp.__skb_datagram_iter
> 2.04 ± 5% +1.5 3.52 ± 2% perf-profile.children.cycles-pp.skb_copy_datagram_iter
> 1.99 ± 5% +1.6 3.57 ± 2% perf-profile.children.cycles-pp.__skb_try_recv_datagram
> 2.04 ± 5% +1.6 3.68 ± 2% perf-profile.children.cycles-pp.__skb_recv_datagram
> 2.06 ± 5% +1.7 3.72 ± 2% perf-profile.children.cycles-pp.skb_recv_datagram
> 11.40 ± 3% +2.5 13.86 perf-profile.children.cycles-pp.native_queued_spin_lock_slowpath
> 11.34 ± 3% +3.9 15.27 perf-profile.children.cycles-pp._raw_spin_lock_irqsave
> 7.10 ± 5% +5.7 12.82 ± 2% perf-profile.children.cycles-pp.raw_recvmsg
> 7.14 ± 5% +5.8 12.89 ± 2% perf-profile.children.cycles-pp.inet_recvmsg
> 7.30 ± 5% +5.9 13.22 ± 2% perf-profile.children.cycles-pp.sock_recvmsg
> 8.10 ± 5% +6.7 14.77 ± 2% perf-profile.children.cycles-pp.__sys_recvfrom
> 8.11 ± 5% +6.7 14.80 ± 2% perf-profile.children.cycles-pp.__x64_sys_recvfrom
> 8.58 ± 5% +7.1 15.71 ± 2% perf-profile.children.cycles-pp.recvfrom
> 8.72 ± 5% +7.3 16.00 ± 2% perf-profile.children.cycles-pp.stress_oomable_child
> 28.65 -3.1 25.55 perf-profile.self.cycles-pp.__sock_queue_rcv_skb
> 5.01 ± 4% -1.7 3.34 ± 8% perf-profile.self.cycles-pp.__xfrm_policy_check2
> 4.32 ± 5% -1.4 2.93 ± 6% perf-profile.self.cycles-pp.raw_v4_input
> 2.50 ± 7% -0.9 1.58 ± 8% perf-profile.self.cycles-pp.ipv4_pktinfo_prepare
> 2.44 ± 5% -0.8 1.66 ± 8% perf-profile.self.cycles-pp.sk_filter_trim_cap
> 3.96 -0.7 3.28 ± 2% perf-profile.self.cycles-pp.kmem_cache_alloc
> 3.21 -0.4 2.85 perf-profile.self.cycles-pp.kmem_cache_free
> 1.72 -0.2 1.52 perf-profile.self.cycles-pp.raw_rcv
> 3.90 -0.2 3.72 perf-profile.self.cycles-pp.__skb_clone
> 2.77 -0.2 2.60 ± 2% perf-profile.self.cycles-pp.dst_release
> 1.01 -0.2 0.85 perf-profile.self.cycles-pp.skb_clone
> 0.92 -0.1 0.79 perf-profile.self.cycles-pp.security_sock_rcv_skb
> 0.86 -0.1 0.74 ± 2% perf-profile.self.cycles-pp.sock_queue_rcv_skb_reason
> 0.67 ± 2% -0.1 0.56 ± 2% perf-profile.self.cycles-pp.kfree_skb_reason
> 0.58 -0.1 0.48 ± 2% perf-profile.self.cycles-pp.memcg_slab_post_alloc_hook
> 0.66 ± 3% -0.1 0.58 ± 2% perf-profile.self.cycles-pp.percpu_counter_add_batch
> 0.35 ± 2% -0.1 0.27 ± 9% perf-profile.self.cycles-pp.skb_set_owner_w
> 0.26 ± 3% -0.1 0.18 ± 6% perf-profile.self.cycles-pp.raw_sendmsg
> 0.36 ± 2% -0.1 0.30 ± 4% perf-profile.self.cycles-pp.kfree_skbmem
> 0.34 ± 3% -0.1 0.29 ± 2% perf-profile.self.cycles-pp.ip_mc_sf_allow
> 0.31 -0.0 0.27 perf-profile.self.cycles-pp.skb_push
> 0.38 -0.0 0.34 perf-profile.self.cycles-pp.skb_release_head_state
> 0.10 ± 8% -0.0 0.06 ± 11% perf-profile.self.cycles-pp.sock_wfree
> 0.35 ± 2% -0.0 0.32 ± 2% perf-profile.self.cycles-pp.__ip_select_ident
> 0.20 ± 2% -0.0 0.18 ± 2% perf-profile.self.cycles-pp.apparmor_socket_sock_rcv_skb
> 0.12 ± 3% -0.0 0.10 ± 3% perf-profile.self.cycles-pp.should_failslab
> 0.11 -0.0 0.09 ± 4% perf-profile.self.cycles-pp.__sys_sendto
> 0.06 ± 6% +0.0 0.08 ± 4% perf-profile.self.cycles-pp.do_syscall_64
> 0.06 ± 9% +0.0 0.09 ± 4% perf-profile.self.cycles-pp.entry_SYSCALL_64
> 0.08 ± 7% +0.0 0.12 ± 3% perf-profile.self.cycles-pp.entry_SYSCALL_64_after_hwframe
> 0.06 ± 6% +0.0 0.10 ± 4% perf-profile.self.cycles-pp.syscall_return_via_sysret
> 0.06 ± 9% +0.0 0.10 ± 4% perf-profile.self.cycles-pp.__skb_try_recv_from_queue
> 0.06 ± 6% +0.1 0.11 ± 3% perf-profile.self.cycles-pp.__virt_addr_valid
> 0.00 +0.1 0.05 perf-profile.self.cycles-pp.check_stack_object
> 0.00 +0.1 0.05 perf-profile.self.cycles-pp.sockfd_lookup_light
> 0.03 ± 70% +0.1 0.09 ± 5% perf-profile.self.cycles-pp.simple_copy_to_iter
> 0.07 ± 10% +0.1 0.12 ± 3% perf-profile.self.cycles-pp.move_addr_to_user
> 0.00 +0.1 0.06 ± 9% perf-profile.self.cycles-pp.syscall_exit_to_user_mode
> 0.00 +0.1 0.06 ± 8% perf-profile.self.cycles-pp.security_socket_recvmsg
> 0.04 ± 45% +0.1 0.10 ± 6% perf-profile.self.cycles-pp.__skb_recv_datagram
> 0.12 ± 3% +0.1 0.18 ± 7% perf-profile.self.cycles-pp.aa_sk_perm
> 0.08 ± 10% +0.1 0.14 ± 4% perf-profile.self.cycles-pp.put_cpu_partial
> 0.02 ±141% +0.1 0.08 ± 4% perf-profile.self.cycles-pp._raw_spin_unlock_irqrestore
> 0.10 ± 10% +0.1 0.16 ± 3% perf-profile.self.cycles-pp.__check_object_size
> 0.07 ± 8% +0.1 0.14 ± 11% perf-profile.self.cycles-pp.__check_heap_object
> 0.00 +0.1 0.07 ± 5% perf-profile.self.cycles-pp.sock_recvmsg
> 0.00 +0.1 0.07 ± 6% perf-profile.self.cycles-pp.inet_recvmsg
> 0.01 ±223% +0.1 0.08 ± 5% perf-profile.self.cycles-pp._copy_to_iter
> 0.12 ± 6% +0.1 0.20 ± 2% perf-profile.self.cycles-pp.__fget_light
> 0.08 ± 8% +0.1 0.16 ± 2% perf-profile.self.cycles-pp.__put_user_nocheck_4
> 0.07 ± 8% +0.1 0.15 ± 4% perf-profile.self.cycles-pp.is_vmalloc_addr
> 0.09 ± 7% +0.1 0.18 ± 2% perf-profile.self.cycles-pp.__get_user_4
> 0.14 ± 12% +0.1 0.23 ± 5% perf-profile.self.cycles-pp.get_partial_node
> 0.13 ± 5% +0.1 0.23 ± 4% perf-profile.self.cycles-pp.entry_SYSRETQ_unsafe_stack
> 0.12 ± 8% +0.1 0.22 ± 2% perf-profile.self.cycles-pp.recvfrom
> 0.10 ± 10% +0.1 0.21 ± 12% perf-profile.self.cycles-pp.__skb_try_recv_datagram
> 0.12 ± 11% +0.1 0.23 ± 5% perf-profile.self.cycles-pp.__unfreeze_partials
> 0.12 ± 4% +0.1 0.24 ± 2% perf-profile.self.cycles-pp.stress_rawsock_server
> 0.15 ± 19% +0.1 0.28 ± 4% perf-profile.self.cycles-pp._copy_to_user
> 0.46 ± 6% +0.1 0.60 ± 2% perf-profile.self.cycles-pp.___slab_alloc
> 0.15 ± 6% +0.1 0.30 ± 2% perf-profile.self.cycles-pp.__skb_datagram_iter
> 0.25 ± 6% +0.2 0.48 ± 2% perf-profile.self.cycles-pp.__sys_recvfrom
> 0.84 ± 3% +0.4 1.20 perf-profile.self.cycles-pp.__slab_free
> 0.39 ± 7% +0.4 0.77 ± 3% perf-profile.self.cycles-pp.sock_rfree
> 0.74 ± 4% +0.4 1.14 ± 2% perf-profile.self.cycles-pp.copyout
> 0.87 ± 5% +0.6 1.43 ± 2% perf-profile.self.cycles-pp.check_heap_object
> 0.70 ± 6% +0.7 1.38 ± 3% perf-profile.self.cycles-pp.raw_recvmsg
> 5.24 +0.7 5.98 perf-profile.self.cycles-pp.__copy_skb_header
> 1.63 ± 4% +1.1 2.73 ± 2% perf-profile.self.cycles-pp._raw_spin_lock_irqsave
> 1.43 ± 7% +1.4 2.81 ± 2% perf-profile.self.cycles-pp.sock_def_readable
> 11.39 ± 3% +2.4 13.84 perf-profile.self.cycles-pp.native_queued_spin_lock_slowpath
>
>
>
>
> 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.
>
>
prev parent reply other threads:[~2024-01-31 12:51 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-01-31 7:20 [linus:master] [mm] 728be28fae: stress-ng.rawsock.ops_per_sec 92.8% improvement kernel test robot
2024-01-31 12:51 ` Kefeng Wang [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=ebdc0ef1-e17f-4382-af22-f46fafc93a67@huawei.com \
--to=wangkefeng.wang@huawei.com \
--cc=akpm@linux-foundation.org \
--cc=david@redhat.com \
--cc=feng.tang@intel.com \
--cc=fengwei.yin@intel.com \
--cc=hughd@google.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=lkp@intel.com \
--cc=mike.kravetz@oracle.com \
--cc=oe-lkp@lists.linux.dev \
--cc=oliver.sang@intel.com \
--cc=willy@infradead.org \
--cc=ying.huang@intel.com \
--cc=ziy@nvidia.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).