* [linus:master] [net_tstamp] 4aecca4c76: redis.get_total_throughput_rps 1.5% improvement
@ 2025-02-05 6:13 kernel test robot
0 siblings, 0 replies; only message in thread
From: kernel test robot @ 2025-02-05 6:13 UTC (permalink / raw)
To: Vadim Fedorenko
Cc: oe-lkp, lkp, linux-kernel, Jakub Kicinski, Willem de Bruijn,
Jason Xing, netdev, linux-alpha, linux-mips, linux-parisc,
sparclinux, linux-arch, oliver.sang
Hello,
kernel test robot noticed a 1.5% improvement of redis.get_total_throughput_rps on:
commit: 4aecca4c76808f3736056d18ff510df80424bc9f ("net_tstamp: add SCM_TS_OPT_ID to provide OPT_ID in control message")
https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git master
testcase: redis
config: x86_64-rhel-9.4
compiler: gcc-12
test machine: 64 threads 2 sockets Intel(R) Xeon(R) Gold 6346 CPU @ 3.10GHz (Ice Lake) with 256G memory
parameters:
all: 1
sc_overcommit_memory: 1
sc_somaxconn: 65535
thp_enabled: never
thp_defrag: never
cluster: cs-localhost
cpu_node_bind: even
nr_processes: 4
test: set,get
data_size: 1024
n_client: 5
requests: 68000000
n_pipeline: 3
key_len: 68000000
cpufreq_governor: performance
Details are as below:
-------------------------------------------------------------------------------------------------->
The kernel config and materials to reproduce are available at:
https://download.01.org/0day-ci/archive/20250205/202502051330.4d2f403b-lkp@intel.com
=========================================================================================
all/cluster/compiler/cpu_node_bind/cpufreq_governor/data_size/kconfig/key_len/n_client/n_pipeline/nr_processes/requests/rootfs/sc_overcommit_memory/sc_somaxconn/tbox_group/test/testcase/thp_defrag/thp_enabled:
1/cs-localhost/gcc-12/even/performance/1024/x86_64-rhel-9.4/68000000/5/3/4/68000000/debian-12-x86_64-20240206.cgz/1/65535/lkp-icl-2sp7/set,get/redis/never/never
commit:
34ea1df802 ("Merge branch 'net-mlx5-hw-counters-refactor'")
4aecca4c76 ("net_tstamp: add SCM_TS_OPT_ID to provide OPT_ID in control message")
34ea1df802f79d44 4aecca4c76808f3736056d18ff5
---------------- ---------------------------
%stddev %change %stddev
\ | \
18491785 +2.1% 18880098 proc-vmstat.numa_hint_faults
18483590 +2.0% 18850441 proc-vmstat.numa_hint_faults_local
8589 ± 97% +255.7% 30553 ± 17% proc-vmstat.numa_pages_migrated
21039386 +2.2% 21505792 proc-vmstat.numa_pte_updates
8589 ± 97% +255.7% 30553 ± 17% proc-vmstat.pgmigrate_success
25696 ± 12% +14.4% 29397 proc-vmstat.pgreuse
252371 +1.5% 256108 redis.get_avg_throughput_rps
67.36 -1.5% 66.38 redis.get_avg_time_sec
1009486 +1.5% 1024432 redis.get_total_throughput_rps
269.45 -1.5% 265.52 redis.get_total_time_sec
257.67 -1.1% 254.83 redis.time.percent_of_cpu_this_job_got
337.27 -2.4% 329.05 redis.time.system_time
3.957e+09 +1.3% 4.008e+09 perf-stat.i.branch-instructions
38469227 +1.6% 39070923 perf-stat.i.branch-misses
32.20 +0.8 33.01 perf-stat.i.cache-miss-rate%
136208 +1.2% 137857 perf-stat.i.context-switches
1.34 -1.0% 1.32 perf-stat.i.cpi
1.948e+10 +1.3% 1.974e+10 perf-stat.i.instructions
9.12 +2.2% 9.32 perf-stat.i.metric.K/sec
224090 +2.5% 229667 perf-stat.i.minor-faults
224090 +2.5% 229667 perf-stat.i.page-faults
1.33 -34.1% 0.88 ± 70% perf-stat.overall.cpi
714.76 -33.9% 472.47 ± 70% perf-stat.overall.cycles-between-cache-misses
1.095e+08 -34.2% 72001076 ± 70% perf-stat.ps.cache-references
15.93 -0.8 15.15 perf-profile.calltrace.cycles-pp.tcp_write_xmit.__tcp_push_pending_frames.tcp_sendmsg_locked.tcp_sendmsg.sock_write_iter
15.95 -0.7 15.22 perf-profile.calltrace.cycles-pp.__tcp_push_pending_frames.tcp_sendmsg_locked.tcp_sendmsg.sock_write_iter.vfs_write
14.40 -0.7 13.72 perf-profile.calltrace.cycles-pp.tcp_write_xmit.__tcp_push_pending_frames.tcp_sendmsg_locked.tcp_sendmsg.__sys_sendto
14.43 -0.6 13.79 perf-profile.calltrace.cycles-pp.__tcp_push_pending_frames.tcp_sendmsg_locked.tcp_sendmsg.__sys_sendto.__x64_sys_sendto
21.35 -0.6 20.74 perf-profile.calltrace.cycles-pp.tcp_sendmsg_locked.tcp_sendmsg.sock_write_iter.vfs_write.ksys_write
21.50 -0.5 20.96 perf-profile.calltrace.cycles-pp.tcp_sendmsg.sock_write_iter.vfs_write.ksys_write.do_syscall_64
16.98 -0.5 16.44 perf-profile.calltrace.cycles-pp.tcp_sendmsg_locked.tcp_sendmsg.__sys_sendto.__x64_sys_sendto.do_syscall_64
17.15 -0.5 16.66 perf-profile.calltrace.cycles-pp.tcp_sendmsg.__sys_sendto.__x64_sys_sendto.do_syscall_64.entry_SYSCALL_64_after_hwframe
21.61 -0.5 21.14 perf-profile.calltrace.cycles-pp.sock_write_iter.vfs_write.ksys_write.do_syscall_64.entry_SYSCALL_64_after_hwframe
22.26 -0.4 21.84 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.write
21.76 -0.4 21.34 perf-profile.calltrace.cycles-pp.vfs_write.ksys_write.do_syscall_64.entry_SYSCALL_64_after_hwframe.write
22.24 -0.4 21.82 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.write
21.95 -0.4 21.53 perf-profile.calltrace.cycles-pp.ksys_write.do_syscall_64.entry_SYSCALL_64_after_hwframe.write
22.65 -0.4 22.24 perf-profile.calltrace.cycles-pp.write
17.28 -0.4 16.87 perf-profile.calltrace.cycles-pp.__sys_sendto.__x64_sys_sendto.do_syscall_64.entry_SYSCALL_64_after_hwframe.__send
17.30 -0.4 16.92 perf-profile.calltrace.cycles-pp.__x64_sys_sendto.do_syscall_64.entry_SYSCALL_64_after_hwframe.__send
17.49 -0.4 17.12 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.__send
17.51 -0.4 17.14 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.__send
17.92 -0.4 17.57 perf-profile.calltrace.cycles-pp.__send
0.57 +0.0 0.62 ± 3% perf-profile.calltrace.cycles-pp.skb_do_copy_data_nocache.tcp_sendmsg_locked.tcp_sendmsg.sock_write_iter.vfs_write
0.74 ± 2% +0.1 0.79 ± 3% perf-profile.calltrace.cycles-pp.tcp_stream_alloc_skb.tcp_sendmsg_locked.tcp_sendmsg.__sys_sendto.__x64_sys_sendto
1.34 +0.1 1.40 perf-profile.calltrace.cycles-pp.__inet_lookup_skb.tcp_v4_rcv.ip_protocol_deliver_rcu.ip_local_deliver_finish.__netif_receive_skb_one_core
3.85 +0.1 3.94 perf-profile.calltrace.cycles-pp.cpuidle_idle_call.do_idle.cpu_startup_entry.start_secondary.common_startup_64
1.87 ± 3% +0.1 1.97 perf-profile.calltrace.cycles-pp.intel_idle.cpuidle_enter_state.cpuidle_enter.cpuidle_idle_call.do_idle
5.14 +0.1 5.24 perf-profile.calltrace.cycles-pp.cpu_startup_entry.start_secondary.common_startup_64
5.14 +0.1 5.25 perf-profile.calltrace.cycles-pp.start_secondary.common_startup_64
5.54 +0.1 5.64 perf-profile.calltrace.cycles-pp.common_startup_64
5.13 +0.1 5.24 perf-profile.calltrace.cycles-pp.do_idle.cpu_startup_entry.start_secondary.common_startup_64
10.08 +0.1 10.20 perf-profile.calltrace.cycles-pp.do_epoll_ctl.__x64_sys_epoll_ctl.do_syscall_64.entry_SYSCALL_64_after_hwframe.epoll_ctl
10.48 +0.1 10.61 perf-profile.calltrace.cycles-pp.__x64_sys_epoll_ctl.do_syscall_64.entry_SYSCALL_64_after_hwframe.epoll_ctl
11.15 +0.1 11.28 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.epoll_ctl
11.03 +0.1 11.18 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.epoll_ctl
6.79 +0.2 6.95 perf-profile.calltrace.cycles-pp.dictFind
12.44 +0.2 12.62 perf-profile.calltrace.cycles-pp.epoll_ctl
15.91 +0.2 16.10 perf-profile.calltrace.cycles-pp.tcp_v4_rcv.ip_protocol_deliver_rcu.ip_local_deliver_finish.__netif_receive_skb_one_core.process_backlog
16.00 +0.2 16.21 perf-profile.calltrace.cycles-pp.ip_protocol_deliver_rcu.ip_local_deliver_finish.__netif_receive_skb_one_core.process_backlog.__napi_poll
16.03 +0.2 16.24 perf-profile.calltrace.cycles-pp.ip_local_deliver_finish.__netif_receive_skb_one_core.process_backlog.__napi_poll.net_rx_action
16.78 +0.2 17.01 perf-profile.calltrace.cycles-pp.process_backlog.__napi_poll.net_rx_action.handle_softirqs.do_softirq
16.54 +0.2 16.78 perf-profile.calltrace.cycles-pp.__netif_receive_skb_one_core.process_backlog.__napi_poll.net_rx_action.handle_softirqs
16.80 +0.2 17.04 perf-profile.calltrace.cycles-pp.__napi_poll.net_rx_action.handle_softirqs.do_softirq.__local_bh_enable_ip
20.70 +0.3 20.96 perf-profile.calltrace.cycles-pp.net_rx_action.handle_softirqs.do_softirq.__local_bh_enable_ip.__dev_queue_xmit
21.08 +0.3 21.34 perf-profile.calltrace.cycles-pp.handle_softirqs.do_softirq.__local_bh_enable_ip.__dev_queue_xmit.ip_finish_output2
21.15 +0.3 21.42 perf-profile.calltrace.cycles-pp.do_softirq.__local_bh_enable_ip.__dev_queue_xmit.ip_finish_output2.__ip_queue_xmit
21.23 +0.3 21.50 perf-profile.calltrace.cycles-pp.__local_bh_enable_ip.__dev_queue_xmit.ip_finish_output2.__ip_queue_xmit.__tcp_transmit_skb
22.73 +0.3 23.03 perf-profile.calltrace.cycles-pp.__dev_queue_xmit.ip_finish_output2.__ip_queue_xmit.__tcp_transmit_skb.tcp_write_xmit
23.44 +0.3 23.77 perf-profile.calltrace.cycles-pp.__ip_queue_xmit.__tcp_transmit_skb.tcp_write_xmit.__tcp_push_pending_frames.tcp_sendmsg_locked
22.94 +0.3 23.28 perf-profile.calltrace.cycles-pp.ip_finish_output2.__ip_queue_xmit.__tcp_transmit_skb.tcp_write_xmit.__tcp_push_pending_frames
26.32 +0.4 26.68 perf-profile.calltrace.cycles-pp.__tcp_transmit_skb.tcp_write_xmit.__tcp_push_pending_frames.tcp_sendmsg_locked.tcp_sendmsg
30.37 -1.5 28.92 perf-profile.children.cycles-pp.tcp_write_xmit
30.39 -1.4 29.03 perf-profile.children.cycles-pp.__tcp_push_pending_frames
38.37 -1.1 37.23 perf-profile.children.cycles-pp.tcp_sendmsg_locked
38.66 -1.0 37.67 perf-profile.children.cycles-pp.tcp_sendmsg
1.32 -0.9 0.38 ± 2% perf-profile.children.cycles-pp.tcp_event_new_data_sent
1.80 ± 2% -0.9 0.88 ± 2% perf-profile.children.cycles-pp.tcp_check_space
1.19 ± 2% -0.9 0.27 ± 4% perf-profile.children.cycles-pp.__mod_timer
1.22 ± 2% -0.9 0.30 ± 3% perf-profile.children.cycles-pp.sk_reset_timer
66.87 -0.5 66.34 perf-profile.children.cycles-pp.do_syscall_64
67.19 -0.5 66.67 perf-profile.children.cycles-pp.entry_SYSCALL_64_after_hwframe
21.61 -0.5 21.15 perf-profile.children.cycles-pp.sock_write_iter
21.82 -0.4 21.39 perf-profile.children.cycles-pp.vfs_write
22.02 -0.4 21.60 perf-profile.children.cycles-pp.ksys_write
17.29 -0.4 16.88 perf-profile.children.cycles-pp.__sys_sendto
22.78 -0.4 22.37 perf-profile.children.cycles-pp.write
17.31 -0.4 16.94 perf-profile.children.cycles-pp.__x64_sys_sendto
18.00 -0.4 17.65 perf-profile.children.cycles-pp.__send
0.23 ± 5% -0.1 0.16 ± 6% perf-profile.children.cycles-pp.tcp_event_data_recv
0.12 ± 4% +0.0 0.13 ± 3% perf-profile.children.cycles-pp.validate_xmit_skb
0.38 ± 2% +0.0 0.40 ± 3% perf-profile.children.cycles-pp.syscall_return_via_sysret
0.30 ± 4% +0.0 0.32 ± 3% perf-profile.children.cycles-pp.pick_next_task_fair
0.51 ± 2% +0.0 0.54 ± 2% perf-profile.children.cycles-pp._copy_from_iter
0.18 ± 5% +0.0 0.21 ± 2% perf-profile.children.cycles-pp.tcp_schedule_loss_probe
1.35 +0.1 1.40 perf-profile.children.cycles-pp.__inet_lookup_skb
1.49 +0.1 1.56 perf-profile.children.cycles-pp.tcp_stream_alloc_skb
0.76 +0.1 0.83 perf-profile.children.cycles-pp.skb_do_copy_data_nocache
4.02 +0.1 4.09 perf-profile.children.cycles-pp.cpuidle_enter
4.00 +0.1 4.07 perf-profile.children.cycles-pp.cpuidle_enter_state
0.24 ± 5% +0.1 0.32 ± 5% perf-profile.children.cycles-pp.release_sock
4.22 +0.1 4.32 perf-profile.children.cycles-pp.cpuidle_idle_call
1.93 ± 2% +0.1 2.03 perf-profile.children.cycles-pp.intel_idle
5.53 +0.1 5.63 perf-profile.children.cycles-pp.do_idle
5.14 +0.1 5.25 perf-profile.children.cycles-pp.start_secondary
5.54 +0.1 5.64 perf-profile.children.cycles-pp.common_startup_64
5.54 +0.1 5.64 perf-profile.children.cycles-pp.cpu_startup_entry
10.12 +0.1 10.24 perf-profile.children.cycles-pp.do_epoll_ctl
10.50 +0.1 10.63 perf-profile.children.cycles-pp.__x64_sys_epoll_ctl
12.76 +0.2 12.93 perf-profile.children.cycles-pp.epoll_ctl
6.88 +0.2 7.05 perf-profile.children.cycles-pp.dictFind
15.94 +0.2 16.13 perf-profile.children.cycles-pp.tcp_v4_rcv
16.04 +0.2 16.25 perf-profile.children.cycles-pp.ip_local_deliver_finish
16.02 +0.2 16.23 perf-profile.children.cycles-pp.ip_protocol_deliver_rcu
16.55 +0.2 16.78 perf-profile.children.cycles-pp.__netif_receive_skb_one_core
16.81 +0.2 17.05 perf-profile.children.cycles-pp.__napi_poll
16.78 +0.2 17.02 perf-profile.children.cycles-pp.process_backlog
21.54 +0.3 21.79 perf-profile.children.cycles-pp.handle_softirqs
20.72 +0.3 20.98 perf-profile.children.cycles-pp.net_rx_action
21.16 +0.3 21.43 perf-profile.children.cycles-pp.do_softirq
21.31 +0.3 21.60 perf-profile.children.cycles-pp.__local_bh_enable_ip
22.76 +0.3 23.06 perf-profile.children.cycles-pp.__dev_queue_xmit
22.96 +0.3 23.29 perf-profile.children.cycles-pp.ip_finish_output2
23.46 +0.3 23.80 perf-profile.children.cycles-pp.__ip_queue_xmit
26.38 +0.3 26.72 perf-profile.children.cycles-pp.__tcp_transmit_skb
1.13 ± 2% -1.0 0.18 ± 3% perf-profile.self.cycles-pp.__mod_timer
1.79 ± 2% -0.9 0.87 ± 2% perf-profile.self.cycles-pp.tcp_check_space
0.22 ± 6% -0.1 0.15 ± 7% perf-profile.self.cycles-pp.tcp_event_data_recv
0.48 +0.0 0.50 ± 2% perf-profile.self.cycles-pp.mod_objcg_state
0.32 ± 2% +0.0 0.34 perf-profile.self.cycles-pp.call
0.50 ± 2% +0.0 0.52 ± 2% perf-profile.self.cycles-pp._copy_from_iter
0.17 ± 4% +0.0 0.19 ± 3% perf-profile.self.cycles-pp.ip_finish_output2
0.11 ± 9% +0.0 0.14 ± 3% perf-profile.self.cycles-pp.tcp_event_new_data_sent
0.27 ± 4% +0.0 0.30 ± 3% perf-profile.self.cycles-pp.__alloc_skb
0.21 +0.0 0.24 ± 3% perf-profile.self.cycles-pp.kfree_skbmem
0.36 ± 3% +0.0 0.40 perf-profile.self.cycles-pp._raw_spin_lock_bh
0.56 ± 2% +0.0 0.60 ± 2% perf-profile.self.cycles-pp.kmem_cache_free
0.13 ± 5% +0.0 0.17 ± 5% perf-profile.self.cycles-pp.vfs_write
0.00 +0.1 0.06 ± 9% perf-profile.self.cycles-pp.__x64_sys_sendto
0.08 ± 8% +0.1 0.14 ± 4% perf-profile.self.cycles-pp.sock_write_iter
0.02 ± 99% +0.1 0.09 ± 11% perf-profile.self.cycles-pp.__sys_sendto
3.40 +0.1 3.50 perf-profile.self.cycles-pp.tcp_sendmsg_locked
1.93 ± 2% +0.1 2.03 perf-profile.self.cycles-pp.intel_idle
0.00 +0.1 0.11 ± 5% perf-profile.self.cycles-pp.__tcp_push_pending_frames
6.66 +0.1 6.78 perf-profile.self.cycles-pp.dictFind
0.00 +0.1 0.13 ± 2% perf-profile.self.cycles-pp.tcp_sendmsg
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
^ permalink raw reply [flat|nested] only message in thread
only message in thread, other threads:[~2025-02-05 6:14 UTC | newest]
Thread overview: (only message) (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2025-02-05 6:13 [linus:master] [net_tstamp] 4aecca4c76: redis.get_total_throughput_rps 1.5% improvement kernel test robot
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.